European Metals & Mining: A Strange Love — How I Learned to Stop Worrying and Love the Ore JULY 2013 SEE DISCLOSURE APPENDIX OF THIS REPORT FOR IMPORTANT DISCLOSURES AND ANALYST CERTIFICATIONS Is $80/t iron ore a certainty? Perhaps not...the future may not be as dark as consensus believes Recent weakness in mining equities is consistent with iron ore at $80/t Aus FOB, or ~40% below both average five-year and 1H:13 prices; we show how the magnitude and speed of the implied price decline express exaggerated fears around GDP, demand growth and supply response China, the world's largest commodities consumer (56% of 2012 iron ore), must complete its transformation from a U.S. 1930s' equivalent level of industrialization before it faces an unprecedented demographic crunch; this supports iron ore demand for 7-12 years Maintaining "super-cycle" pricing, thus, depends on the supply side exercising sufficient capex restraint not to displace the bulk of high-cost producers; however, given our view on demand, we see room for the currently approved pipeline and prices supported above $120/t for seven years What if we are wrong? Even then, if iron ore fell to $80/t, three "automatic stabilizers" — the natural hedge between revenue and costs, value-in-use and iron ore demand elasticity — would offset some of the fall with a positive value impact of 5-20% for our coverage stocks For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
European Metals & Mining: A Strange Love — How I Learned to Stop Worrying and Love the Ore
JULY 2013
SEE DISCLOSURE APPENDIX OF THIS REPORT FOR IMPORTANT DISCLOSURES AND ANALYST CERTIFICATIONS
Is $80/t iron ore a certainty? Perhaps not...the future may not be as dark as consensus believes
Recent weakness in mining equities is consistent with iron ore at $80/t Aus FOB, or ~40% below both average five-year and 1H:13 prices; we show how the magnitude and speed of the implied price decline express exaggerated fears around GDP, demand growth and supply response
China, the world's largest commodities consumer (56% of 2012 iron ore), must complete its transformation from a U.S. 1930s' equivalent level of industrialization before it faces an unprecedented demographic crunch; this supports iron ore demand for 7-12 years
Maintaining "super-cycle" pricing, thus, depends on the supply side exercising sufficient capex restraint not to displace the bulk of high-cost producers; however, given our view on demand, we see room for the currently approved pipeline and prices supported above $120/t for seven years
What if we are wrong? Even then, if iron ore fell to $80/t, three "automatic stabilizers" — the natural hedge between revenue and costs, value-in-use and iron ore demand elasticity — would offset some of the fall with a positive value impact of 5-20% for our coverage stocks
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 1
Portfolio Manager's Summary
Iron ore has been the outperforming commodity of the last decade (rising 1,340% from its 2001 average to its 2011 peak). Since 2009, the iron ore price has averaged close to $130/t FOB Australia; however, recent weakness in mining equities suggests that the market believes that an imminent, severe and permanent downward revision to the value of this commodity is due. An analysis of equity values shows that the market is discounting a 40% correction in the price of iron ore down to levels close to $80/t. We do not believe that such a decline is necessarily justified and that it is, at the very least, inconsistent with the growth forecasts for the Chinese economy held by the IMF and the OECD. We believe that the price of iron ore is supported by the fundamental geological poverty of China and that any reacceleration of the Chinese economy must imply a reacceleration in metal demand and with it support for the price of raw materials.
That China is the world's largest consumer of commodities (56% of iron ore in 2012) is undeniable. Equally undeniable is the unprecedented rapidity with which China has attained this position. However, we also believe that the demographic crunch looming in 2020-25, as the birth-rate explosion of the 1950s and 1960s and subsequent "one child" policy successively age their way through the population, is a phenomenon that cannot be ignored. If China is to industrialize sufficiently to support its aging population before exhausting its supply of "cheap" labor, then it has a bare decade in which to complete a transformation from an agrarian society to one dominated by the service sector. China's embedded capital stock (~4.5 tons of steel per capita) and the composition of its labor force (with parity between primary and tertiary forms of employment) resemble that of the U.S. of the 1930s. Furthermore, while China's rate of capital stock formation is roughly 2x that of the U.S., this merely speaks of the increases in mining productivity that we have seen over the last 50 years. Productivity gains mean that China's capital stock is embedded at one-fourth the labor used to build up that of the U.S.
Maintaining so-called "super-cycle" pricing, thus, depends upon the supply side. Specifically, will the major Western miners exhibit sufficient capital discipline to limit the quantity of low-cost iron ore brought to market? This commodity can be mined by the likes of the "Big Three," for as low as $25-30/t cash costs. Will they race to compete on low-cost volume, thereby displacing the ~9% of 2012 global supply provided by high-cost ($120-$180/t) Chinese miners? Increasingly, the miners are becoming aware that prioritizing volume over value is a fool's paradigm and that just because you can build something does not mean that you should. In this Blackbook, we show how iron ore prices could well be supported above $120/t through to 2020.
Furthermore, even if we are wrong and iron ore prices do decline as the market is anticipating, three factors will offset their impact on company valuations: First, the natural hedge between costs and revenue (including but not limited to exchange rates); second, because not all iron ores are created equal, value-in-use adjustment reduces the ability for low-quality supply to compete and provides price support for the major miners with high-quality deposits; and third, the elasticity of Chinese iron ore demand. We believe that a major part of the slowdown in Chinese steel growth is the budgetary constraint of high iron ore prices; relax the constraint through lower iron ore price and iron ore demand should increase.
Upside/(Downside) to Target Price 56% 25% 102% 47% 53% ‐
MCAP (US$m) 25,217 145,992 27,483 78,780 77,034 ‐
Net Debt/(Cash) (US$m) 5,855 32,169 22,551 21,990 26,088 ‐
Minorities 17,642 1,349 5,032 490 (4,702) ‐
EV (US$m) 51,687 170,890 93,427 98,682 98,420 ‐
EBITDA ‐ US$m
2013E 10,052 25,614 15,902 24,183 22,523
2014E 11,833 38,119 20,203 31,348 26,230
2015E 15,590 45,903 26,252 38,698 32,415
EV/EBITDA ‐ US$m
2013E 5.1 6.7 5.8 4.1 4.4
2014E 4.4 4.5 4.2 3.1 3.8
2015E 3.3 3.7 3.3 2.6 3.0
EPS ‐ US$/share
2013E 1.90 2.70 0.43 5.33 2.28
2014E 2.57 3.73 0.63 7.34 2.55
2015E 3.75 4.73 0.97 9.52 3.58
Dividend Yield ‐ %
2013E 3.5% 4.2% 3.5% 3.9% 2.2%
2014E 2.3% 3.5% 4.9% 4.1% 3.4%
2015E 3.0% 3.6% 5.8% 5.1% 4.8%
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 5
Significant Research Conclusions
Iron ore has been the outperforming mining commodity for the last decade and is the critical source of earnings generation (~60% of EBITDA) for our coverage group. Recent weakness in mining company equities implies a very severe discount to spot iron ore price ($80/t versus an 1H:13 average US$129/t Australian FOB) in terms of future price level and the rapidity of decline. We do not believe this decline is a foregone conclusion. Furthermore, if it were to happen, there are a number of factors that would mitigate the severity of the valuation impact.
The industrial behemoth1 that is the Chinese economy has been powered by exploitation of China's most abundant resource: vast pools of surplus rural labor. However, this labor supply is not infinite. Over the next generation, China will experience a substantial increase in its old-age dependency ratio as the country ages sharply in response to earlier policy initiatives (namely, the "one child" policy) — see Exhibit 2. On top of this are the behavioral changes in a workforce increasingly composed of only children. The IMF sees cheap labor in China exhausted between 2020 and 2025 — concurrent with our forecast for peak raw materials demand in China (see Exhibit 3). We believe that the proximity of this labor exhaustion has contributed to the urgency of the Chinese governmental push to install a productive capital base and transition the economy away from subsistence agriculture to tertiary forms of value-added activity. The unprecedented rapidity of Chinese industrialization has rendered China the largest consumer of global commodities (56% of total iron ore demanded globally in 2012). In our view, this reflects the country's requirement to "grow rich before it grows old" and before the supply of cheap labor that has powered this industrialization is exhausted.
Exhibit 2 In 2000, China Was a Clear Outlier — the World's Largest Country Set to Age Rapidly and Yet Still at a Pre-Industrial Output Level
Exhibit 3 The Demographic Challenges Facing China Stand Behind the Requirement to Accelerate the Capital Stock Formation
Source: UN and Bernstein estimates and analysis. Note: 2015-40 numbers are UN estimates.
Source: UN and Bernstein analysis.
1 The original biblical Behemoth is described as having a tail like cedar, bones like iron, limbs as strong as copper, and sinews like stone, hence it seems an apropos metaphor for the economy of the world's largest consumer of commodities.
-10
0
10
20
30
40
50
60
-50 -40 -30 -20 -10 0 10
GD
P/C
apit
a —
US
$'00
0
Change in Old-Age Dependency Ratio 2000-2040 — Negative Indicates Aging Population
Aging of Population vs. Development of Economy Including ChinaSize of Bubble = Population
USA
Japan
ChinaSouth Korea
-10%
-5%
0%
5%
10%
15%
20%
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
Ch
ang
e in
Wo
rkin
g-A
ge
Po
pul
atio
n
China — Growth in Working-Age Population (15-64 Years)
Critical demographic inflection point ~2020‐2025
We Do Not Believe the Sustained Iron Ore Price Fall Implied by Mining Equity Valuations Is a Foregone Conclusion; Even If So, Several Factors Would Mitigate the Valuation Impact
If the World's Largest Consumer of Commodities Is to Grow Rich Before It Grows Old, China Must Largely Complete Its Unprecedentedly Rapid Industrialization by 2020-25
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
6 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
It is the urgency of China's industrialization, in our view — coupled with the fundamental geological scarcity of certain key raw materials domestically — that has been the key driver behind the emergence of the commodity "super-cycle." On the demand side, our secular commodity price view is supported by a fundamental analysis of industrialization patterns in 120 countries since 1900. On the supply side, we identify the emergence of a new class of marginal producers in 2003 (Chinese high-cost producers), their impact on the global cost curve, and the duration of supply side support for "super-cycle" pricing based on new greenfield and brownfield projects. We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25) and on the supply side until sufficient new low-cost Western production is brought on line to displace the marginal high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline) — see Exhibit 4.
The high rate of recent Chinese steel consumption, in both absolute magnitude and relative to the current consumption by other countries, is often interpreted to suggest overinvestment and that the Chinese economy may be unduly steel intensive — in which case, an immediate and sustained iron ore price decline to US$80/t might indeed be reasonable. However, many datasets look at only the last ~20 years and point out China's peak steel intensity of 60kg/000$GDP as being substantially higher than the peak rate of consumption during the industrialization of the U.S. They fail to consider the pronounced productivity increases in the mining sector over the last half-century. China has access to iron ore production that is 2.5-5x more efficient in man-hour terms than was available to the U.S. when it industrialized (see Exhibit 5). Three factors stand behind China’s access to this higher productivity iron ore mining (largely located in Australia): 1) globalization of the commodities industry (between 1965 and 2005, real freight costs fell 40%, while global trade in iron ore rose by 330% and coal by 1070%), 2) geological endowment (a miner in the U.S. needs to move ~2x ROM material as his Australian counterpart to generate the same amount of useful material), and 3) population densities (Western Australia ex-Perth has, on average, 0.23 people per square kilometer versus 67 in Minnesota and 175 in Michigan). Due to this increased productivity and China's access to it, the Middle Kingdom2 is embedding steel into its capital stock at a rate ~2x that of peak U.S. steel intensity — but at one-fourth of the labor that the U.S. required. This more rapid embedding of steel in China's capital stock reflects not inefficiency per se, but rather that the productivity of the world's best iron ore miners (Australians today) can be utilized to hasten the process of capital stock formation. This rapidity has been further enabled by the purchase, disassembly and reassembly of steel-making production from the West into China (e.g., the Famous Industrials Group industry relocation projects). Consequently, from a demand perspective, we consider the fear that the iron ore price will decline to US$80/t and remain there to be overdone.
2 The literal translation of Zhong-guo — the Chinese name for China, which can also be rendered as Middle State. American is Meiguo or Beautiful Kingdom/State.
We Attribute "Super-Cycle" Pricing in Iron Ore to the Emergence of High-Cost Chinese Marginal Producers; We See the "Super-Cycle" Continuing Until Those Producers Are Displaced by New Low-Cost Production
We Do Not Believe That the Rate of Chinese Steel Consumption Is Indicative of Overinvestment; China Is Embedding Steel Into Its Capital Stock at a Rate Approximately Double That of Peak U.S. Steel Intensity But at Around One-Fourth of the Labor That the U.S. Required
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 7
Exhibit 4 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines
Exhibit 5 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003
Note: There are a few other high-cost producers in the "missing Chinese mines" category but they are de minimis.
Source: AME and Bernstein estimates and analysis.
Source: WSA, NBS, BLS and Bernstein estimates and analysis.
On the supply side, we see the inability of the Chinese mining sector to transition from labor-intensive to capital-intensive modes of production during this rapid industrialization phase in China, supporting the current "super cycle" pricing. Costs on the supply side provide the long-term pricing structure, meaning pricing support will persist, in our view, so long as a significant proportion of high-cost marginal producers in China (~9% of 2011 ROE equivalent global supply, with marginal costs of production ~$120-180/t) are not displaced (see Exhibit 4). Accordingly, capital discipline, the single greatest source of value creation in the mining space, is particularly critical for iron ore (the fourth most crustally abundant element).
Here, we provide a case study on the possible 70Mtpa Pilbara iron ore expansion proposed by Rio Tinto. The price elasticity backed out from Rio Tinto's own cost curve implies a reduction of US$27/t from this project, resulting in an $18 billion value loss (not quite Alcan levels, but getting there). However, if management's 3% steel demand growth target came to pass, this would limit price declines from the project to US$14/t — meaning that expenditure of US$5 billion would generate a positive value of ~US$8 billion. Shareholders in Rio Tinto may be comfortable risking an $18 billion loss for the chance of an $8 billion gain, should China achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of Emperor Augustus and consider the whole expansion akin to fishing with a golden fish hook: the gain, if all goes well, is too small to offset the risk of loss if it does not. We further expand our analysis to demonstrate how globally superior value generation can result from the pursuit of locally optimal growth targets — but only in consolidated markets where industry incumbents are explicitly aware of the price destructive effects of excessive volume growth (see Exhibit 6 and Exhibit 7). It is this dynamic, which underlies our continued calls for capital discipline from the Western majors to avoid engaging in a massive value transfer from their own shareholders to the Chinese Communist Party by subsidizing China's continued industrialization.
170
160
150
140
130
120
110
100
80
70
0
60
180
40
30
20
10
0
50
1,000,000750,000500,000250,000
90
1,250,000
Cumulative global production– kt Fe
Missing Chinese
Mines
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
To
ns
pe
r Man
Ye
ar
Effective Chinese Productivity, Including Imports
USA Effective Chinese Productivity
Pricing Support in the Next Decade Thus Depends Upon Capital Discipline from the Western Miners as Our "Rio 360" Case Study Shows
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
8 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 6 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered
Exhibit 7 It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve; the Higher the Elasticity, the Higher the Margin and the Lower the Degree of Consolidation Required to Generate Superior Returns
Source: Bernstein estimates and analysis. Source: Bernstein estimates and analysis.
The iron ore price (the most important source of EBITDA generation for our coverage) has averaged US$135/t over the last five years and US$129/t 1H:13 Australian FOB — and yet mining company stocks have been among the worst performers of the year. We ran four fixed commodity price scenarios ( "Grizzly," "Bear," "Bull," and "De Niro" in order of increasing aggressiveness) and examined how sensitive the value generated by our DCF (out to 2030) was to these different scenarios. We then selected the scenario that most closely resembled current market conditions and flexed the prices of the two most important commodities (from an EBITDA generation perspective) — copper and iron ore — in order to determine what the market was pricing in for those commodities. We found that mining equity valuations were baking in $80/t iron ore, a ~40% discount to year-to-date and five-year average prices. We then asked ourselves what the world needs to look like for these prices to come about. Specifically, "what growth outlook for the world's largest commodity consumer, China, would be consistent with $80/t iron ore?" We calculated the Chinese GDP growth implied by consensus demand expectations in China, then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore. We found that this was consistent with ~7% Chinese GDP growth — some 1.3% below the IMF forecast and 0.5% below the official government guidance — as well as significant new supply: ~550Mtpa of new iron ore capacity out to 2020 (a ~50% increase on the 1,123mt produced in 2012 globally), of which ~300Mtpa would be surplus at this GDP growth rate and hence price destructive.
Using the model (discussed earlier) that we built to back out Chinese GDP and supply side response consistent with consensus demand and price expectations, we can also determine a consistent forecast for iron ore prices and steel demand growth across any GDP scenario for China. Our "bull" scenario takes the IMF's forecast from 2Q:12, which sees the Chinese economy accelerate back up to a trend rate of 8.5%. This would absorb the ~300Mtpa of price destructive oversupply implied in the consensus scenario (mentioned earlier), resulting in a tight market in which the miners do not compete on volume and in which a trend price of US$140/t could
0
20
40
60
80
100
120
140
0 25 50 75 100125150175200225250275300
Val
ue
(US
$bn
)
Iron Ore Volume Growth for New Entrant (Mt)
Value of Volume Growth for Different Price Elasticity
0.0 -0.5 -1.0 -1.5-2.0 -2.5 -3.0
"Normal" straight‐line relationship between volume and value...butonly if elasticity is
ignored
0
2
4
6
8
10
12
1 2 3 4 5 6 7 8 9 10EV/E
BIT
DA
Mu
ltip
le o
f In
du
stry
No. of Players in Industry (i.e., Degree of Consolidation)
Multiple Uplift Available in Consolidated Commodity Markets by Elasticity
e=-1 e=-2 e=-3
Mining Equities Have Recently Been Baking in $80/t Iron Ore (~40% Below Year-to-Date and Five-Year Average Prices) — This Is Consistent With Both ~7% Chinese GDP Growth (Below Both IMF Estimates and Chinese Targets) and ~550Mtpa New Iron Ore Capacity Out to 2020
We Construct Bull, Middle and Bear Scenarios for Chinese GDP and Supply Response to Bound the Possibilities for Iron Ore Prices Out to 2020 ($140/t down to $65/t)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 9
eventuate out to 2020 (see Exhibit 8). A "middle" scenario takes the Chinese target of 7.5% GDP growth as a floor (despite the fact that Chinese GDP has a tendency to miss on the upside). This yields a price of ~US$120/t out to 2020. Our China "bear" scenario sees GDP decline to 6% and no supply coming out of the market (a true worst-case scenario). This results in an iron ore price testing the US$65/t mark over a multi-year period.
Exhibit 8 Our Three Scenarios Set the Widest Plausible Band, in Our View, Around Implied Consensus Expectations; They Range from More Bullish Than the IMF's Current Forecast, Through Official Chinese GDP Targets, to a Bear Scenario in Which No Supply Is Removed from the Market Despite Prices Collapsing
Source: IMF and Bernstein estimates and analysis.
The previous sections looked at analysis conducted during 2Q:13. Here, we provide an overview of our updated iron ore price forecast. At the time of publication of this Blackbook, we predict $130/t (2013E) rising to a peak of $144/t (2016E). This is consistent with the IMF's current GDP forecast of 7.8% in 2013 rising to 8.5% in 2016 and average increase in supply of 88Mtpa 2013-16E. By 2020, we expect iron ore prices of $115/t consistent with of IMF forecasts for an average 8.5% Chinese GDP growth 2017-20E and a further 28Mtpa of incremental supply in the market.
Exhibit 9 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-2016
Source: Bloomberg L.P. and Bernstein estimates and analysis.
Our Current Forecast Is for Iron Ore That Peaks at $144/t in 2016 and Declines to $115/t by 2020
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
10 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
As JK Galbraith observed, "faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof." As an exercise in listening to Galbraith, we asked ourselves "what are the valuation impacts if we are wrong, and if $80/t iron ore becomes a near-term reality?" While examining this hypothetical situation, we came across three "automatic stabilizers" that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners and offer a realistic picture of how a sustained fall in iron ore prices would impact valuations: (1) the natural FX hedge embedded in the miners' cost structures; (2) value-in-use adjustments commanded by ores of differing qualities and (3) elasticity of iron ore demand and spare capacity in China's steel-making industry whose exploitation would become economical, if the iron price were to fall sufficiently. Across all of these stabilizers, those companies that are the most exposed to iron ore experience the greatest offsetting effect.
A natural hedge exists in the operating cost structure of the miners that softens the earnings impact of any sustained fall in commodity prices (see Exhibit 10 and Exhibit 11). In a world of falling commodity prices, we expect to see substantial weakening in producer currencies (AUD and BRL). Given that the miners' costs are denominated in local currency and their revenues in U.S. dollars, any producer-currency weakening will lead to an improvement in operating margin. With nearly 50% of the cost increases in iron ore over the last 10 years coming from the impact of currency movements, any currency weakening is a significant source of upside. We estimate that Vale and Rio would experience the greatest benefits (equivalent to 12% and 8% of June 30, 2013 share prices, respectively), while more diversified BHP and Anglo would experience 4% and neutral impacts, respectively.
Exhibit 10 Mining Costs and Revenue Exhibit a High Degree of Correlation
Exhibit 11 Currency Appreciation Drives a Significant Part of the Cost to Price Relationship, Particularly for the Australian Producers
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.
3 Note that we omit Glencore Xstrata from this automatic stabilizer valuation analysis as the company lacks direct iron ore production bar its 1Q:13 acquisition of a stake in Ferrous Resources Ltd. Note that all valuation offsets are expressed in percentage of the June 30, 2013 price in USD terms.
R² = 0.9048
0
5,000
10,000
15,000
20,000
25,000
0 20,000 40,000 60,000
To
tal C
ove
rage
Co
sts
($m
)
Total Coverage Revenue
Total Iron Ore Costs and Revenue
41%
17%
42%
Rio Tinto Iron Ore Cost Drivers
Producer Currency Appreciation Local Currency Inflation
Real Cost Increases
What If We Are Wrong and There Is a Sustained Iron Ore Price Fall (to $80/t)? If So, Three Factors Would Mitigate the Negative Value Impact for the Big Western Miners
We Estimate That Cost Savings, Combined With the Natural FX Hedge Embedded in Operating Costs, Could Offset Valuation Declines by an Equivalent of, on Average, 5% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale3
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 11
Not all iron ores were created equal. Substantial quality and value-in-use (VIU) heterogeneity exists in iron ore products. Removing homogeneity from a commodity industry, of course, has a "de-commoditizing" effect — iron ore grade accounts for 93% of the iron ore price variation around the 62% Fe benchmark. However, even after this effect is adjusted for, poorer-quality, low-grade iron ore still incurs a pricing penalty (see Exhibit 12). Inclusion of a VIU adjustment in the cost curve thus increases the marginal cost of production, reduces the ability for low-quality supply to compete and provides price support for the major miners with high-quality deposits. We estimate that Vale and Rio would experience the greatest benefits (equivalent to 11% and 9% of June 30, 2013 share prices), while more diversified BHP and Anglo would experience 4% and 6% offsets, respectively.
Exhibit 12 There Is a Significant Range of Variation Around the Benchmark 62% CIF Price Point, and Even After Adjusting for Fe Grade, Poor-Quality Ore Suffers a Significant Discount
Exhibit 13 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from the Budgetary Constraint Imposed by High Commodity Prices
Source: BAIINFO and Bernstein analysis. Source: Bloomberg L.P., AME, CRU, IMF and Bernstein estimates and analysis.
Chinese steel making has seen its EBITDA margins cut by ~70% over the last decade on the back of rising raw material prices (iron ore CAGR of 26%, PCI and coking coal CAGR of 14%) and essentially stagnant steel prices (local currency CAGR of 1%, USD CAGR of 4%). We believe that this decline stands behind the fall in Chinese steel production growth (down from ~20% p.a. 10 years ago to the ~6% p.a. seen today). Steel capital stock in China is still only at around one-third of the level seen in the West, and the urbanization and industrialization of China remains far from complete (see Exhibit 13). In a world where iron ore starts to fall toward a non-Chinese cost floor of US$80/t, there would be a significant reduction in the cost pressures facing the Chinese steel industry, which, we expect, would yield an increase in the Chinese steel trend growth and push iron ore price back up to US$95/t. Vale and Rio would experience the greatest benefits (equivalent to 31% and 24% of June 30, 2013 share prices), while more diversified BHP and Anglo would experience 10% and 16% offsets, respectively.
R² = 0.7734
-40
-30
-20
-10
0
10
20
30
50% 55% 60% 65% 70%
Iro
n O
re P
rem
ium
vs
. Be
nch
mar
k (U
S$/
t)
Fe Grade of Iron Ore
Price Impact of Quality in Iron Ore
-10%
-5%
0%
5%
10%
15%
20%
25%
30%
35%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Jan-
05
Aug
-05
Mar
-06
Oct
-06
Ma y
-07
Dec
-07
Jul-0
8
Feb-
09
Se p
-09
Apr
-10
Nov
-10
Jun-
11
Jan-
12
Au g
-12
Ste
el G
row
th
Ste
el E
BIT
DA
Mar
gin
Chinese Steel Growth and Profitability Since 2005
Margin, %Annualized Grow thSteel Grow th Ex Monetary Grow th
VIU or Heterogeneity in a Commodity Offers a Degree of Automatic Stabilization to the Iron Ore Price, Offsetting Valuation Declines in an Oversupply World by an Equivalent of, on Average, ~7% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale
Chinese Steel Growth Rate Declines Are a Consequence of the Budgetary Constraint of High Iron Ore Prices; Consequently, Low Iron Ore Prices Would See Demand Rise, Offsetting Valuation Declines by an Equivalent of, on Average, 20% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
12 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Mining companies are operationally and financially geared to their underlying commodity exposure, so we provide two valuation metrics (see the "Valuation and Risks" chapter of this Blackbook for more details): ~80% of weekly mining equity price moves can be explained by underlying
commodity price moves, so we use a regression-based trading model and our commodity price forecasts to help determine our 12-month price targets. If the regression remains stable or deviations appear temporary, the model determines the target price. If we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made.
We additionally provide a supplementary DCF-based valuation constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is applied for the remaining years of potentially exploitable material. We forecast our models in reporting currency (USD), convert to listing currency (GBP or Real), and round final DCF values in 25p/cent increments.
The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources), and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation). For more details on both sector risks and company-specific risks, see the risk section in the "Valuation and Risks" chapter.
We believe that there are sound fundamental reasons why commodity prices over the last five years have been at elevated levels relative to historical precedent. Moreover, as we have discussed, for iron ore4 these reasons are based on supply and demand dynamics that center around China, the largest consumer of commodities. On the demand side, the Chinese government is facing an imperative to industrialize at an unprecedented rate ahead of an equally unprecedented demographic shift. Meanwhile, the Chinese domestic mining industry, with its high-cost and labor-intensive production, has become more important to global supply. We believe that changing the cost structure and importance of this industry may take longer than the market currently anticipates. As a result, we do not believe that the sustained and near-term fall in commodity price implied by mining valuations is a foregone conclusion. Furthermore, in the event that we are wrong, the natural FX hedge embedded in operating costs, VIU adjustment and reduced cost pressures on the steel-making industry that would eventuate would all mitigate the valuation impact of such a fall. A key conclusion of our analysis is the sensitivity of the iron ore price forecast to two variables: China's macro policy (will the outcome be similar to that which the IMF predicts?) and the strategic choices of the incumbent iron ore miners (will they discipline capital in pursuit of value or
4 European Metals & Mining: Iron, Cold Iron, Is Master of Them All...or at Least 60% of EBITDA.
Valuation Methodology
Risks
Investment Conclusion
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 13
will they simply pursue volume at the cost of suppressing prices?). Both bear watching — and while the former is beyond the control of our coverage companies, the latter component of their destiny is firmly within their own control.
Rio Tinto (TP £41.25; unmodified from trading model prediction) — 73% of 2012 EBIT from iron ore, plus world-class copper exposure: We consider Rio Tinto the most attractive stock in our coverage. Rio owns some of the highest quality iron ore assets and (vitally) infrastructure globally. Rio also has exposure to some of the world's best operational copper assets — not to mention three of the world's best undeveloped copper deposits (Resolution and La Granja). Tier 1 assets in iron ore include Dampier and Cape Lambert (Australian Pilbara), while copper includes Escondida, Grasberg, Bingham Canyon and Oyu Tolgoi. The prospects of genuine capital discipline and cost cutting under new CEO Sam Walsh (who made his bones in low-cost brownfield Pilbara production), coupled with a more reasoned approach to volume growth, mean that we continue to see Rio Tinto as our top pick. Our trading model, used to set 12-month target prices, has shown no signs of the regression shifting (R-squared = 88%) nor is there anything in our fundamental research to date that causes us to anticipate it will, hence we have taken our target price from the unmodified regression model price.
Vale (TP BRL 46.50; unmodified from trading model prediction) — 95% of 2012 EBIT from iron ore, a commodity in which we see near-term upside: Vale is the world's largest iron ore producer and, in Carajas, has one of the most globally attractive iron ore assets. Vale is the most operationally geared miner to the iron ore price and we see an asymmetric risk to the upside in iron ore prices in the near term, hence from a pure value consideration, we are more favorably inclined to Vale now than previously. We do note the significant influence of the Brazilian government (5.5% directly through Golden Shares and ~34% indirectly through the strategic consortium of Valepar). The company is, in our view, most at risk should a reduced iron ore price eventuate, given its geographic longinquity from the world’s largest consumer of iron ore, China. Vale's regression continues to show instability in our trading model (despite R-squared of 86% in our most recent update). However, for now we have taken the unmodified price target generated by the trading model, but we will continue to monitor the regression.
BHP (TP £22.50; unmodified from trading model prediction) — highest-quality stock, Tier 1 assets and limited geographic risk: BHP Billiton not only has the longest corporate history of our coverage group (its component parts trace their origins back to 1851 and 1883, respectively), we view it as our highest quality stock. In 2012, the company generated the highest revenue of the "Big Three" miners in our coverage (US$72 billion versus US$56 billion for Rio and US$49 billion for Vale) and had the highest EBITDA margin (42% versus 39% for Vale and 36% for Rio). BHP Billiton is the second most diversified (from a commodity exposure perspective) company in our coverage, after Anglo American. However, where platinum has been a drag on Anglo's portfolio, BHP's has received a boost from "black gold" in its petroleum division (69% EBITDA margins in calendar 2012 versus 11% for Anglo's platinum division perspective). The strong commodity risk diversification of BHP Billiton is complemented with geographic diversification that, while skewed to Australia, is nonetheless low risk. We have not seen signs of the regression shifting (R-squared =78%) in our trading model, nor is there anything in our fundamental research to date that causes us to anticipate it will. Accordingly, we set our target price from the unmodified regression model price.
Anglo American (TP £20.25; unmodified from trading model prediction) — a soluble turnaround story: Anglo American has the most diversified portfolio of our coverage group from a commodity perspective and is significantly smaller than the "Big Three" of BHP, Rio and Vale due to its smaller iron ore exposure. That Anglo is ahead of Glencore Xstrata on this metric is, in large measure, due to the existence of Kumba in Anglo's portfolio. Kumba, however, like AmPlats and thermal coal, increases the company's country risk exposure due to its South
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
14 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
African location. Anglo American has the highest country risk premium in our coverage — 2.6% versus an average of 1.2% for the "Big Three." Given recent operational difficulties (including failure to deliver on Minas Rio and Los Bronces, not to mention the issues plaguing platinum), Anglo remains in our minds a turnaround story — one that will challenge new CEO Mark Cutifani, but one which we believe is solvable for a leader with his 36 years of operational expertise. Given the shift in the regression and the fundamental factors that have been weighing on the stock, as well as the upcoming 1H reporting (at which time Mr. Cutifani will present the results of his portfolio review), and that we consider Anglo a restructuring story, we consider there is a greater likelihood of regression instability going forward than for Rio or BHP and are monitoring the situation. However, for now, we have taken the unmodified price target generated by the trading model (R-squared = 85%) at present.
Glencore Xstrata (TP £5.25; unmodified from trading model prediction) — traders, copper and coal: Glencore Xstrata offers exposure for copper and coal bulls without a taste for iron.5 It also offers exposure to the sales and trading house that Mr. Ivan Glasenberg built. As such it has a profile that is distinct relative to the pure miners in our coverage. The impact of the high-turnover, low-margin trading business (2% EBITDA margin in 2012 versus 25% for Glencore Xstrata combined) is clearer even when contextualized against the average 33% for our coverage group. As an owner operator (Mr. Glasenberg holds 8% of the combined entity's paper), Mr. Glasenberg's incentives are squarely aligned with investors. It remains to be seen how he will staff the combined entity, following the departure of much of the senior management talent from Xstrata. We do consider that Glencore has some ways to go yet before it will be "institutional quality" on the metrics of reporting and governance. Furthermore, the company's operations in frontier jurisdictions like the DRC not only result in the second-highest country risk premium in our coverage but also carry headline risk — as does Glencore's trading division.6
For Glencore Xstrata, we have constructed a synthetic stock using Xstrata’s actual history and its recent relationship with Glencore. This synthetic shows, in our view, an acceptably high R-squared of 64%. For now, we have taken the unmodified price target of £5.25 but will continue to monitor the trading model.
5 Given that the company's only direct iron ore exposure is its 1Q:13 acquisition of a stake in Ferrous Resources. 6 To wit the April 21, 2013 story Glencore traded with Iranian supplier to nuclear program http://www.guardian.co.uk/business/2013/apr/21/glencore-trade-iran-supplier-nuclear
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 15
If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade
The industrial behemoth,7 that is the Chinese economy, has so far been powered by exploitation of China's most abundant resource: vast pools of surplus rural labor. However, this labor supply is not infinite. Over the next generation, China will experience a substantial increase in its old-age dependency ratio as the country ages sharply in response to earlier policy initiatives (namely, the "one child" policy). A January 2013 IMF working paper8 sees cheap labor in China becoming exhausted sometime between 2020 and 2025 — the same point at which we forecast that China's absolute demand for raw materials will peak. We believe that it is the proximity of this labor exhaustion that has contributed to the urgency of the Chinese government's push to install a productive capital base and transition the economy away from subsistence agriculture to tertiary forms of value-added activity. The unprecedented rapidity of Chinese industrialization has rendered China the largest consumer of global commodities (56% of total iron ore demand globally in 2012); in our view, this reflects the county's requirement to "grow rich before it grows old" and before the supply of cheap labor that has powered this industrialization is exhausted.
It is the rapidity of this industrialization, coupled with the fundamental geological scarcity of certain key raw materials domestically, that has been the key driver behind the emergence of the commodity "super-cycle." In this chapter, we show how our counter-consensus view on "stronger for longer" commodity prices is consistent with the imperative faced by the Chinese to build an industrial society in a little more than 20 years.
Global Demographics and the Impact of China A clear relationship exists between the wealth of societies and their demographic structure (see Exhibit 14). Rich industrialized economies tend to see falling birth rates and rapidly increasing dependency ratios.9 In these societies, old-age provision is not critically dependent on having a large number of children. Rather, the increase in productivity associated with industrialization enables the society to easily generate the output required to support what would otherwise be a significant burden. On the other hand, in societies relying on subsistence agriculture, the birth rate tends to stay high and the society does not age significantly over time.
The transition from an agrarian economy to an industrial society, dominated by tertiary forms of economic activity, requires significant consumption of steel and other raw materials. In effect, it represents the build-up of capital stock that increases the productivity of the workforce, thus liberating labor from the requirement to be engaged in primary enterprises in the process. We see this pattern clearly in the U.S., where the employment in agriculture peaked in 1910 at 11.6 million people before falling monotonically to just over 2 million people today (see Exhibit 15). This is despite the fact that over the same period U.S. production of wheat rose from 17mt in 1910 to over 60mt today (+253%). Likewise, the U.S.
7 The original biblical Behemoth is described as having a tail like cedar, bones like iron, limbs as strong as copper, and sinews like stone, hence it seems an apropos metaphor for the economy of the world's largest consumer of commodities. 8 "Chronicle of a Decline Foretold: Has China Reached the Lewis Turning Point?" IMF WP/13/26. 9 Defined as the number of non-productive workers (according to the UN, those aged 0-14 and 65 plus) relative to the working-age population. So, a dependency ratio of 50 implies 50 non-productive citizens supported by 100 people of working age.
Over the Next Generation, China's Old-Age Dependency Ratio Will Increase Sharply in Response to Earlier Policy Initiatives
In 2000, China's Demographic Composition Resembled That of a Rich Industrial Society, With a Falling Birth Rate and a Rapidly Aging Society, But Its Economy Was Still Agrarian
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
16 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
experienced similar increases in the production of other agricultural commodities (maize +315%; rice +1,970%; cattle +62%). As a result, the service sector employment rose substantially from 5.9 million people in 1910 to roughly 48 million today (see Exhibit 16). In 1850, U.S. agriculture employed 5x as many people as the service sector did (5 million versus 0.9 million), but it now employs only 0.05x as many (2.2 million versus 48 million) — a 100-fold decline over the space of the U.S.'s industrialization.
As the U.S. transitioned from an agrarian to an industrialized society, its capital stock rose. Rather than relying on the energy of human labor to drive output, it is far more productive to harness the forces of mechanization and industrialization (see Exhibit 17). Moreover, this transition is not isolated to the U.S., but has been replicated in one form or another across all countries that have industrialized.
Exhibit 14 There Is a Clear Relationship Between the Aging of a Population and Its Wealth…Usually, Birth Rates Fall as Societies Get Rich, With Aging Population Supported by the Increased Productivity Associated With Industrialization
Source: UN and Bernstein estimates and analysis.
R² = 0.4677
-10
0
10
20
30
40
50
60
-50 -40 -30 -20 -10 0 10
GD
P/C
apit
a —
US
$'00
0
Change in Old-Age Dependency Ratio 2000-2040 — Negative Indicates Aging Population
Aging of Population vs. Development of Economy Excluding ChinaSize of Bubble = Population
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 17
Exhibit 15 In 1910, the U.S. Began Transitioning from an Agricultural Economy…
Exhibit 16 …Toward One Driven by Services
Source: Mitchell and Bernstein analysis. Source: Mitchell and Bernstein analysis. Exhibit 17 We Believe That the Increases in Labor Productivity, Occasioned by the Increased
Intensity of Capital Stock (Machinery, Railroads, etc.), Enabled This Transition
Source: World Steel Association (WSA), UN, Mitchell and Bernstein estimates and analysis.
The stark outlier from the rule of an aging population equating to high output and increased prosperity, however, was China in 2000. In particular, its output in real terms stood at a little over US$2,000 per capita, and yet the country was set to experience a precipitous decline in the working-age population (see Exhibit 18). Little wonder that as few as six years ago, a Chinese demographer reportedly stated at Davos that "China might have to resort to mass suicide in the end, shoving pensioners onto the ice."10 Though stark hyperbole, this comment nevertheless illustrates the realization by the Chinese policy makers that a graying population in a society that has not yet industrialized is a recipe for social and economic disaster. Since 2000, China has embarked upon a program of urbanization unprecedented in human history. Agricultural employment in China peaked in 2000 and has been
10 As reported in the Daily Telegraph on February 6, 2013.
Ratio Primary to Tertiary Employment - USA Steel Capital Stock - USA
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
18 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
declining ever since, with first manufacturing and now services increasing in importance to the overall economy (see Exhibit 19 through Exhibit 22). Indeed, there are now more people employed in the service sector in China than in manufacturing and construction.
Meanwhile, the installation of the capital stock required to support an urban service-based society has continued apace (see Exhibit 23) — and has followed the same trajectory as that of the U.S. and other industrialized countries. There are now as many jobs in services in China as in farming, and the country's capital stock comprises ~4 tons of steel per capita. The U.S. reached the point of equivalence between employment in the service sector to agricultural labor in 1930 accompanied by a capital stock of ~5.5 tons of steel per capita. On this basis, Chinese industrialization has been more productive than that seen in the U.S. Both from the perspective of the labor pool composition and capital stock installation, China today resembles the U.S. of the 1930s (see Exhibit 24). The important difference between the two, however, is the speed of the transition toward an urban society. It took the U.S. 80 years to reduce the importance of agriculture from 5x that of services to parity. The Chinese accomplished this in 30 years.
Exhibit 18 In 2000, China Was a Clear Outlier from the Normal Pattern — the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-industrial Level of Output
Source: UN and Bernstein estimates and analysis.
-10
0
10
20
30
40
50
60
-50 -40 -30 -20 -10 0 10
GD
P/C
apit
a —
US
$'00
0
Change in Old-Age Dependency Ratio 2000-2040 — Negative Indicates Aging Population
Aging of Population vs. Development of Economy Including ChinaSize of Bubble = Population
USA
Japan
China
South Korea
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 19
Exhibit 19 It Was Only in the Year 2000 That China Started Reducing Its Framing Dependence
Exhibit 20 1970s Saw the Rise in Construction and Manufacturing Employment Start…
Source: NBS and Bernstein analysis. Source: NBS and Bernstein analysis. Exhibit 21 …Followed in Short Order by the Service
Sector
Exhibit 22 Today, More People Are Employed in Services Than Manufacturing
Source: NBS and Bernstein analysis. Source: NBS and Bernstein analysis.
0
50
100
150
200
250
300
350
400
Mill
ion
Wo
rker
s
Primary Employment — China
0
50
100
150
200
250
Mill
ion
Wo
rker
s
Secondary Employment — China
0
50
100
150
200
250
300
Mill
ion
Wo
rker
s
Tertiary Employment — China
0
100
200
300
400
500
600
700
800
Mill
ion
Wo
rker
s
Chinese Employment Breakdown
Primary Secondary Tertiary
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
20 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 23 The Pattern in China Is Nearly Identical to That Seen in the U.S.; Capitalization Drives Productivity Increases, Which Then Lead to the Emergence of Services
Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 24 Comparing China to the U.S. Suggests That the Structure of the Chinese Economy
Today Is Similar to the U.S. in the 1930s, Both in Labor and Capital Stock; However, the Transition Has Been Far More Rapid in China
Source: NBS, WSA, Mitchell and Bernstein estimates and analysis.
Ratio Primary to Tertiary Employment - USA Ratio Primary to Tertiary Employment - ChinaSteel Capital Stock - USA Steel Capital Stock - China
The Chinese economy iresembles that of the
US in the 1930s
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 21
Lewis Point and the End of Cheap Labor The Lewis point is defined as the point at which surplus agricultural labor becomes exhausted and industrial wages start to accelerate. This, in turn, lowers the industrial sector productivity, placing a drag on the level of investment in an economy. According to the argument, prior to this point, the pool of rural labor acts as a brake on wages in the industrial sector. This, in turn, increases its profitability, hence supporting a higher level of investment.
So far it would seem that (with the exception of raw materials) China has maintained an exceptionally rapid growth rate with (relatively) low inflation precisely because of this mechanism (or so the IMF would argue). However, the pool of cheap labor in the country is soon to be exhausted. In a January 2013 whitepaper, the IMF marks the point of exhaustion somewhere between 2020 and 2025 (see Exhibit 25). This, therefore, marks the point at which capital stock accumulation in China will come under the greatest pressure. Moreover, this point also coincides with our forecast for the peak demand for raw materials in China, and this coincidence is not accidental (see Exhibit 26). By 2020, the Chinese will have installed a capital stock comparable to that seen in most other industrialized nations. Consequently, China should be able to bear the acceleration in industrial wages with some degree of equanimity — principally because once the country's capital stock is already built, the decline in investment profitability will not matter as much as it would now.
To see this more clearly, let's consider Japan. It started to exhaust its working-age population some 20 years prior to China, with Japanese population in the 15-64 age group growth turning negative in 2000 (see Exhibit 27). However, at this point, Japan was already an industrialized nation with 14 tons of steel per capita in its capital stock.
Exhibit 25 The Demographic Challenges Facing China and the Emergence of Lewis Point Stand
Behind the Requirement to Accelerate the Capital Stock Formation
As China's Supply of Cheap Rural Labor Is Exhausted, Maintaining the Level of Capital Investment Will Become More Difficult; China's Transition Will Have to Be as Rapid as It Has Been So Far and Continue Until at Least 2020
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
22 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 26 The Urgency to Industrialize the Chinese Economy Is Evident in the Acceleration in Steel Demand in Comparison to the Historical Precedent in the West
Note: 2015-30 numbers are Bernstein estimates.
Source: WSA and Bernstein estimates and analysis. Exhibit 27 A Very Similar Demographic to China Was Seen in Japan Toward the End of the Last
Century; However, by the Time the Japanese Transition Occurred, Industrialization Was Complete and Output Stayed High
By 2025, Chinese steel per capita will stand at ~10 tons vs. ~4 today
When Japanhit the demographic ceiling, there was an installed stock of 14 tons per
capita
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 23
If our demand projections prove accurate, by the time China reaches this demographic tipping point, the country too will have industrialized (critically, it will not have done so if raw material demand goes ex-growth from this point). Consequently, China has a very limited window in which to achieve the necessary changes to the structure of its economy and society, and it is clearly seen in the evolution of China's dependency ratio over time. In the 1950s, the infant mortality rate in China stood at ~200 per 1000, but then fell dramatically over the next 20 years to a level well below 50. However, female fertility levels stayed at roughly the same level (around six births per woman) for a considerable period, thus resulting in an explosive population growth, an increase in life expectancy and a rapid fall in the child dependency ratio following the "one child" policy (see Exhibit 28).
The response of the "one child" policy to the problem subsequently sowed the seeds for an inevitable rise in the old-age dependency ratio (see Exhibit 29). In toto, these two effects generated a demographic "wave" of cheap labor moving through the Chinese economy between 1970 and 2020, as the children of the previous pro-growth policy initiatives matured and entered the workforce (see Exhibit 30).
Given that the first part of the boon of cheap labor was effectively squandered during the Cultural Revolution and subsequent political restructuring, only the period from 2000 to 2020 was left available to take advantage of reorientation of mass cheap rural labor in the Chinese economy. So while a similar demographic profile to China was seen 20 years earlier in Japan, the political situation in the Middle Kingdom and the sheer size of the country led to a radically different outcome (see Exhibit 31).
Further compounding this demographic trend is the accompanying behavioral trend. On a 1Q:13 trip to China, our team interviewed management across industries, including sewing machine and baby goods manufacturing, hospitality, fine dining, education, and mining to gather anecdotes about managing a workforce increasingly composed of only children. These "xiao huangdi" (little emperors) are the sole focus of often six doting relatives (two parents and four grandparents). We collected anecdotes that were similar across different industries: only children were often resistant to shift work or overtime, many expected to be allowed to go home for a lunch break, they would refuse to work if they came down with minor colds, if a family member grew ill they would leave without offering any notice, and if they changed jobs they would also often leave without giving notice. There was also discussion of a seasonal pattern of, if not outright productivity disruption, at least disruption from management's perspective: the month following Chinese New Year (and the payment of bonuses) many workers would shop around for offers of new employment at higher wages (particularly in Shanghai and Beijing). They would take these offers back to their current employer in order to negotiate increased compensation. The employers felt that they had no choice but to raise wages for those workers that had proven themselves to be reliable, as many workers were not reliable. Across all the conversations we had, one common thread emerged: behaviorally, management considered their respective workforces less reliable than they had been 5-10 years in the past. To the extent that our interviews can be extrapolated to the nation at large, this suggests that behavioral shifts in the workforce may further increase the time pressure upon the Chinese government to industrialize beyond that suggested by demographics alone.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
24 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 28 A Rapid Fall in the Child Dependency Ratio Following the "One Child" Policy…
Exhibit 29 …Will Eventually Be Offset by the Extraordinary Increase in the Old-Age Dependency Ratio
Note: 2020-60 are UN estimates.
Source: UN and Bernstein analysis.
Note: 2020-60 are UN estimates.
Source: UN and Bernstein analysis. Exhibit 30 The End of the Last Century Marked an
Opportunity to Deploy Abundant Labor in China to Develop Its Capital Stock…
Exhibit 31 …Mirroring Japan But Scaling Up More Than 10 Fold (127 Million vs. 1,345 Million People)
Note: 2020-60 are UN estimates.
Source: UN and Bernstein analysis.
Note: 2020-60 are UN estimates.
Source: UN and Bernstein analysis.
0
10
20
30
40
50
60
70
80
Po
pu
lati
on
65+
to 1
5-64
Child Dependency Ratio — China
0
10
20
30
40
50
60
Po
pu
lati
on
0-1
4 to
15-
64
Old-Age Dependency Ratio — China
30
40
50
60
70
80
90
Po
pu
lati
on
0-1
4 an
d 6
5+ to
15-
64
Total Dependency Ratio — China
30
40
50
60
70
80
90
100
Po
pu
lati
on
0-1
4 an
d 6
5+ to
15-
64
China and Japan Aging Compared
China Japan
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 25
Capital Accumulation and Real Commodity Prices Over 1930-1970, there was a clear link between the trend growth in real commodity prices and the growth in Western capital stock (see Exhibit 32). When the West went ex-growth for raw materials following the oil shocks of the 1970s, real commodity prices started their downward trend. This illustrates that falling real commodity prices do not constitute the norm. In fact, quite the opposite should be true, given that the miners start by accessing the most promising and easily exploited geology and move onto the harder, higher-cost deposits only subsequently. However, after the mid-1970s, productivity improvements flattened the mining cost curve. Given the commoditized nature of the industry and absent meaningful marketing power, the benefits of these cost improvements were transferred to raw material consumers through lower prices. Mining companies became their own worst enemies as they competed on volume and pursued simple "beggar thy neighbour" strategies.
This trend halted abruptly in 2003, as the world — driven by Chinese industrialization — re-entered the mode of capital accumulation. Part of our conviction in continued commodity price strength lies in the real and inevitable requirement we see to transform the Chinese economy. We believe that capital accumulation in China must continue to run for some time to come, and therefore demand for raw materials going ex-growth simply does not match the demographic reality faced by the Chinese policy makers.
As an example of this, we show our forecast for real copper prices and the continuation of capital stock accumulation in Exhibit 33. The market currently prices in significant declines in the real copper price going forward. In our view, this is at odds with the historical precedents of significant size economies trying to industrialize and the effect of such industrialization on commodity prices.
A further complication relates to the fact that it has been Chinese mining wage escalation that has driven the commodity "super-cycle" through its impact on the structure of the global cost curve (see Exhibit 34). As China approaches the Lewis point, Chinese wages should start to accelerate. If China has not been able to "unlock" Africa by the time this happens, and absent massive value destructive capital profligacy from the major miners, the chance of seeing further significant commodity price rises cannot be discounted.
The Period of Capital Accumulation in the West Saw a Trend of Increasing Real Commodity Prices; We See Chinese Capital Accumulation Continuing to at Least 2020
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
26 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 32 1930-1973 Saw Real Commodity Prices Grow as Capital Was Accumulated in Western Economies; Once Demand Went Flat, Productivity Improvements in Mining Drove Real Commodity Prices Down
Source: WSA, Bloomberg L.P., Mitchell and Bernstein estimates and analysis. Exhibit 33 The Emergence of a New Phase of Capital Accumulation Driven by China Has Seen
Significant Real Commodity Price Appreciation; Our Counter-Consensus View on Commodities Is Driven, in Part, by a Belief That This Will Continue Until the Chinese Capital Stock Accumulation Comes to an End
Note: 2015-20 are Bernstein estimates
Source: WSA, Bloomberg L.P., Mitchell and Bernstein estimates and analysis.
The Rise of China and the Impact of Real Commodity Prices
Real Cu Price Real Cu Price Forecast US Capital Stock China Capital Stock
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 27
Exhibit 34 Chinese Labor Costs Have Been Instrumental in Changing Global Commodity Cost Curve Structures and, With Them, the Price of Mined Raw Materials; What Happens Once China Moves Past the Lewis Point?
Source: IMF, NBS and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
28 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 29
Iron Ore "Super-Cycle" Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization
In our 2012 Blackbook, European Metals & Mining: Iron, Cold Iron, Is Master of Them All...or at Least 60% of EBITDA, we reviewed why iron ore is important to our coverage group, and discussed in length our supply and demand framework.11 Iron ore is the key commodity for our coverage group, generating, on average, 41% of revenue and 65% of EBITDA in 2012. This has risen from the average of 23% of revenue and 28% of EBITDA back in 2003, supported not only by rising iron ore prices12 but also by iron ore production increases of 11.6% CAGR across our coverage group.13 In our view, it is the urgency of China's industrialization (discussed in the previous chapter), coupled with the fundamental geological scarcity of certain key raw materials domestically, that has been the key driver behind the emergence of the commodity "super-cycle." On the demand side for iron ore, our secular commodity price view is supported by a fundamental analysis of industrialization patterns in 120 countries since 1900. On the supply side, we identify the emergence of a new class of marginal iron ore producers in 2003 (Chinese high-cost producers), their impact on the global cost curve, and the duration of supply side support for "super-cycle" pricing based on new greenfield and brownfield projects. We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25) and on the supply side until sufficient new low-cost Western production is brought on line to displace the marginal high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline).
Iron Ore — Key Commodity for Our Coverage Iron ore is the key commodity for our coverage group, generating, on average, 41% of revenue and 65% of EBITDA in 2012 (see Exhibit 35 and Exhibit 36). This has risen from the average of 23% of revenue and 28% of EBITDA back in 2003, supported not only by rising iron ore prices but also by iron ore production increases of 11.6% CAGR across our coverage group. Of the "Big Three,"14 BHP showed the highest CAGR for the period (8.1%), while Vale — which started from a production base 2.5x that of BHP's in 2003 — had the lowest CAGR at 6.9% (meaning that by 2012, Vale's production base was only 1.9x that of BHP's). Anglo American had the highest CAGR out of our coverage stocks exposed to direct iron ore production of 23.7% (see Exhibit 37 through Exhibit 40).
11 Please see the "China Growth Scenarios" and "Our Iron Ore Price Forecast" chapters for more information on our price forecast in the context of different scenarios discussed in this Blackbook. 12 2003-12 saw the IMF 62% FE monthly spot (CFR Tianjin port) rise at a CAGR of 128%, from an average $14/t to an average $129/t. 13 Excluding Glencore, which has no direct production exposure to iron ore, bar the company's 1Q:13 stake acquired in Ferrous Resources. 14 Rio Tinto, Vale and BHP Billiton.
We See Iron Ore "Super-Cycle" Pricing Continuing Until High-Cost Chinese Marginal Producers Are Displaced by New Low-Cost Production
Iron Ore’s Contribution to Our Sector’s EBITDA Has Increased from 28% to 65% (2003-12), Supported by a Price CAGR of 128% and Production Volume CAGR of 12% Over the Period
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
30 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 35 Across Our Coverage Group, Iron Ore Generated, on Average, 41% of Revenue…
Exhibit 36 ...And 65% of EBITDA in 2012
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis. Exhibit 37 Iron Ore Currently Accounts for 44% of Rio Tinto's Revenue and Has Been Driven by
a 7.6% Production CAGR
Source: Corporate reports and Bernstein estimates and analysis.
70%
44%
31%
18%
41%
2012 Revenue from Iron Ore
97%
76%
49%
37%
65%
2012 EBITDA from Iron Ore
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
Re
ven
ue
—$m
Pro
du
ctio
n —
kt
Rio Tinto Iron Ore
Rio Tinto Volume Rio Tinto Revenue
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 31
Exhibit 38 Iron Ore Currently Accounts for 31% of BHP Billiton's Revenue and Has Been Driven by an 8.1% Production CAGR
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 39 Iron Ore Currently Accounts for 70% of Vale's Revenue and Has Been Driven by a
6.9% Production CAGR
Source: Corporate reports and Bernstein estimates and analysis.
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
Re
ven
ue
—$m
Pro
du
ctio
n —
kt
BHP Iron Ore
BHP Volume BHP Revenue
0
5,000
10,000
15,000
20,000
25,000
30,000
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000
Rev
enu
e -$
m
Pro
du
ctio
n -
kt
Vale Iron Ore
Vale Volume Vale Revenue
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
32 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 40 Iron Ore Currently Accounts for 18% of Anglo American's Revenue and Has Been Driven by a 23.7% Production CAGR
Source: Corporate reports and Bernstein estimates and analysis.
Iron Ore Demand Iron ore demand can be understood only in the context of steel demand, as steel is the means by which iron ore is actually consumed in an economy. Through months of data analysis, we have assembled a database covering the consumption of metals and economic output for 120 countries since 1900. Our analysis identified distinct and repeatable patterns in every country that has successfully industrialized (as well as the anomalous patterns in those that have not).
First and foremost, it is important to distinguish between the rate of steel consumption (how much is being consumed/embedded today) and the level of steel consumption (how much has been consumed/embedded historically). We liken the situation to filling a bathtub: the speed of water coming out of the tap cannot tell you when the bathtub will overflow — you also need to know how full the bathtub is already and how much water it can hold.
Industrializing and industrialized economies move through three distinct phases in their rate of steel consumption. The same three phases are repeated across industrialized and industrializing countries in our dataset.
i. Development. The build-up of industrialization where steel use grows faster than the underlying GDP. This corresponds to the phase of capital-base installation, which will subsequently generate output.
ii. Peak. A peak in intensity where GDP growth and steel growth are matched. During this phase, we see the effect of diminishing returns — additional capital spending starts delivering incrementally less output.
iii. Decline. The development of an economy based on tertiary forms of value-add, i.e., forms of manufacturing that require little incremental consumption of natural resources (e.g., airbags in cars — no new steel is used but significant value is created). Once this occurs, overall economic growth will be faster than the growth in metal consumption.
As a country industrializes, it evolves through these three distinct phases, which result in a unique and consistently repeated structural form that appears across the development of Western Europe, North America and the developed Asian economies. Following the same three-phase trajectory for the evolution of
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
0
5,000
10,000
15,000
20,000
25,000
30,000
Re
ven
ue
—$m
Pro
du
ctio
n —
kt
Anglo Iron Ore
Anglo Volume Anglo Revenue
Our Dataset of 120 Countries Since 1900 Shows All Industrialized and Industrializing Countries Pass Through Distinct and Repeated Patterns of Steel Demand
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 33
the steel consumption rate within all industrialized and industrializing countries, China's steel intensity has passed from development through peak and is now in the decline phase — a period in which the rate of demand growth is decelerating. The requirement to continue to build the capital stock and complete the industrialization constrains how rapid the deceleration will be. The steady-state rate of consumption, once the process is completed, will be determined by whether the economy, once mature, adopts a more manufacturing or services-oriented focus (see Exhibit 41).15 Countries that have not industrialized (and historical data for industrialized countries prior to industrialization) show clustered and random (albeit typically low) rates of steel intensity.
In addition, three paradigms emerge for the evolution of the level of steel consumption across countries that have industrialized (or attempted to industrialize). These patterns show how the level of steel stock has varied by the stage of economic development in each country, which allows us to view how a country, becoming ever more productive (i.e., increasing total factor productivity with increasing economic development), looks through the lens of cumulative steel use. We dub these paradigms "Services and Technology" (exemplified by the U.S. and the U.K.), "Manufacturing" (exemplified by Germany and Japan) and "Inefficient" (exemplified by the former USSR)16 — see Exhibit 42.
Our framework shows that Chinese steel demand (46% of 2012 finished steel demand globally) is on a sustainable path (see Exhibit 43). We believe it is erroneous to conflate decelerating metal intensity with negative or flat metal demand growth. It is likewise erroneous to conclude that rapidity of development is an appropriate proxy for efficiency (or lack thereof) in capital allocation. However, these seem to be two of the more widely disseminated flaws in analyzing the Chinese steel industry. China’s current steel intensity (rate) is consistent with its state of industrialization and shows that the country is transitioning from Phase 2 (Peak) to Phase 3 (Decline) at a rate commensurate with that seen elsewhere. This is not a "structural shift" so much as a natural development to a state that is less steel intensive. The current trajectory sees China ending up less steel intensive than "Manufacturing" countries like South Korea but more so than "Services and Technology" countries like the U.S. To highlight the sustainability of China's demand, we would call attention to the fact that the economic activity of every person in the U.S. (or any other developed country) is supported by an installed capital base comprising 80kg of copper and 12,000kg of steel — more than triple the current capital base in China (26kg of copper and 4,500kg of steel per capita) — see Exhibit 44.
15 In order to render these comparable across differently populated countries, we adjust output and embedded stock on a per-capita basis.
16 As with the rate calculation, in order to render these comparable across differently populated countries, we adjust output and steel consumption on a per-capita basis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
34 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 41 Like Other Countries That Have Industrialized, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady-State Rate of Consumption Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing or Services-Oriented Focus
Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.
Exhibit 42 The "Services and Technology" Pattern (Embodied by the U.S.) Shows the Highest Output/Capital Stock (on a Per-Capita Basis), the "Manufacturing" Pattern (Embodied by Germany) Also Shows an Upward-Sloping Relationship Between Output and Capital Stock; Only the Failed "Inefficient" Pattern (Embodied by the former USSR) Shows a Flat and Negative Relationship Between Output and Embedded Capital Stock
Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 35
Exhibit 43 Our Analysis Derives a Consistent Relationship Between Steel Demand Growth and Overall Level of Economic Activity for China; It Does Not See Signs That the Output Generated by the Embedded Capital Stock Is "Inefficient"; Rather, It Is Increasing in a Manner Consistent With Development Into an Economy Midway Between the "Services and Technology" (Embodied by the U.S.) and the "Manufacturing" (Embodied by Germany) Patterns
Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.
0
5
10
15
20
25
30
35
40
45
50
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000
GD
P p
er
Cap
ita
—U
S$'
000
Installed Steel Stock (including depreciation) — kg per Capita
Steel Stock vs. Output Evolution
China Historical China Forecast - SCB USA Historical Linear (China Historical - Trend)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
36 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 44 A Central Plank of Our Thesis Rests on the Fact That Chinese Urbanization and Industrialization Are Still Far from Complete; in Terms of Capital Stock Accumulation, Labor Composition and Levels of Output, China Today Resembles the U.S. in the 1930s
Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.
R² = 0.6587
0
10
20
30
40
50
60
70
80
0 5,000 10,000 15,000 20,000
Ou
tpu
t —
GD
P/C
apit
a $'
000
(Rea
l 20
05 P
PP
$)
Capital Stock — kg Steel/Capita (including depreciation)
Capital Stock to Output by Country
China
UK
Kuwait
Germany
S Korea
Taiwan
USA
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 37
Iron Ore Supply We do not believe that the microeconomics of supply and demand have suddenly ceased to determine commodity prices. In other words, we reject "financialization" or other non-economic explanations of the "super-cycle." Rather, a shift in the source of the marginal unit of supply and lack of transparency into that marginal unit have obfuscated the underlying supply and demand dynamic. In 2003, Chinese producers emerged as the marginal high-cost supplier for metals, including iron ore, met coal and copper. However, a lack of visibility into the nature of these producers led to a breakdown in the representativeness of the dataset historically used to construct the global cost curve (see Exhibit 45 and Exhibit 46). We see a paradox in China: pre-industrial modes of natural resource production coexisting with industrial modes of production in other sectors. This has led to Chinese mining costs outstripping productivity gains, which has driven up both Chinese commodity imports and global commodity prices. Chinese labor productivity and costs and, specifically, the inability to displace increasingly expensive labor with capital represent the true cause of commodity "super-cycle" pricing. Our analysis has completed the global cost curve, showing that ~10% of global iron ore production in 2011 came from the "missing" Chinese mines (see Exhibit 47). From the supply side, the current "super-cycle" is supported so long as these high-cost marginal producers in China are not displaced. China lacks the geological endowment, experience in capital allocation processes and political will to drive a transition to capital-intensive modes of mining. That leaves, in our view, two routes for displacement — Chinese mining abroad (e.g., in Africa) and new low-cost supply brought on by Western majors. This means that the persistence of the supply side support for the "super-cycle" will be determined by the degree of capital discipline exercised by Western majors (see Exhibit 48). Currently, our medium-term iron ore price forecast is based on a "zero-capital discipline" assumption. That is, we assume that the miners will continue failing to prioritize their total portfolio value ahead of the marginal value of incremental growth. To the extent that Western majors bring on new, low-cost supply in copious amounts (referring here to the mega projects that have the ability to displace significant numbers of high-cost marginal suppliers in China), they effectively subsidize Chinese industrialization at the expense of future sector profitability. Thus, the fate of sector profitability rests squarely on the shoulders of the incumbents, who control the best portion of the world's production and new mega projects.
In 2003, Chinese Producers Emerged as the Marginal, High-Cost Suppliers of Iron Ore; But Lack of Visibility Into the Nature of These Producers Led to a Breakdown in the Representativeness of the Dataset Historically Used to Construct the Global Cost Curve — Our Reconstructed Global Cost Curve Includes the "Missing" Chinese Mines
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
38 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 45 At First Glance, the Old Heuristic of Marginal Cash Costs Appears to Have Broken Down…
Exhibit 46 …But the Fact That Our Visibility Into Costs in China Is Incomplete Provides a Better Answer
Source: AME, Bloomberg L.P. and Bernstein analysis. Source: AME and Bernstein analysis.
Exhibit 47 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis, Which Completes the Global Cost Curve, Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines
Exhibit 48 Our Analysis of the Cash Costs of 149 Green and Brownfield Projects Shows That the Majority Will Not Be Value Accretive at Iron Ore Prices Below $150/t Against a Fully Loaded Discount Rate
Note: There are a few other high-cost producers in the "missing Chinese mines" category but they are de minimis.
Source: AME and Bernstein estimates and analysis.
Source: AME, CRU and Bernstein analysis.
0
20
40
60
80
100
120
140
160
180
200
2005 2006 2007 2008 2009 2010 2011 2012
US
$/t
Iron Ore Costs and Price
C(90) Price
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Australia China
Percentage of "Costed" Iron Ore Supply
Known Unknown
170
160
150
140
130
120
110
100
80
70
0
60
180
40
30
20
10
0
50
1,000,000750,000500,000250,000
90
1,250,000
Cumulative global production– kt Fe
Missing Chinese
Mines
800,000600,000400,000200,0000
5,000
25,000
35,000
45,000
55,000
60,000
20,000
1,000,000
50,000
40,000
15,000
0
30,000
10,000
There are 16 projects globally that need to be understood/monitored to track the future of the iron
ore price. Capital discipline here is key to ensure that the market
remains well supported.
Cumulative production from new Greenfield projects – kt Fe
Pro
duc
tion-
kt F
e
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 39
Long-Term Price Forecast We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25), and on the supply side, until sufficient new low-cost Western production is brought on line to displace the marginal, high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline).
We believe that prices rise to the extent that marginal units of supply are called by demand and that prices fall to the extent that marginal units of supply are displaced. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of that industry. With this heuristic in place, we take inspiration from Fischer Black's work on business cycles and monetary policy17 and construct a modified framework, replicating his approach to business cycles but extending the lead time between demand signals and the supply side response. One consequence of the long lead time is inherent cyclicality in the mining industry. Furthermore, this cyclicality ensures that the incentive price (the price at which a miner is incentivized to bring on new capacity) is seldom achieved, but is rather over- or under-shot. This means that the value generated by investment (either by the mining companies themselves or by investors into mining companies) is almost entirely a call on the ability to act counter-cyclically.
Our supply and demand analysis gives rise to our counter-consensus positive iron ore price forecast (2013: +3%, 2014: +18% and 2015: +29%) — see Exhibit 49 and Exhibit 50. We expect iron ore to peak at $145/t in 2016 as supported by the marginal costs of Chinese iron ore production. We believe that the tendency to embed backwardation into commodity price forecasts and instantaneous reversion to some notional long-term mean has been the most obvious error in commodity price forecasting over the last 10 years. In our view, it is the supply side cost structure that determines the structural shape of commodity prices over the longer term; supply is sticky as projects have multi-year lead times, while demand gains and losses can be relatively instantaneous. All of which is another way of bringing us back to a principal tenet of this Blackbook: capital discipline is the key to price preservation.
Exhibit 49 We Embed a Contango Rather Than Backwardation Into Our Short- to Medium-Term Price Forecasts
Source: Bloomberg L.P., IHS Global Insight and Bernstein estimates and analysis.
Exhibit 50 Our Iron Ore Forecast Sees Considerable Upside in 2014-16E Relative to Consensus; We Would Be Even More Bullish If We Were Convinced the Western Majors Were Committed to a Higher Degree of Capital Discipline
* As of June 30, 2013
Source: Bloomberg L.P. and Bernstein estimates and analysis.
17 "Business Cycles and Equilibrium," published in 1987.
Iron Ore Fines FOB Aus 2013 H2 2014 H1 2014 H2 2015 H1 2015 H2 2016 H1 2016 H2 2017 H1 2017 H2 2018 H1 2018 H2
We Apply a Framework Taken from Fischer Black's Work on Business Cycles to Our Supply and Demand Analysis, Adapted to Long Lead Times in the Mining Sector; This Generates Our Counter-Consensus Positive View on Iron Ore Pricing
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
40 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 41
Two Times the Rate at One-Fourth the Labor: Chinese Steel Intensity Does Not Indicate Overinvestment
This chapter grew out of an analysis of productivity increases in mining, which we looked at to address frequent client concerns about the high rate of Chinese steel consumption in recent years. This rate, in both absolute magnitude and relative to the current consumption by other countries, is often interpreted to suggest overinvestment and that the Chinese economy may be unduly steel intensive — in which case, an immediate and sustained iron ore price decline to US$80/t might indeed be reasonable.18 Is China's steel intensity (annual rate of consumption) higher than that of the U.S.? Yes, of course, but...once one adjusts for relative labor productivity in iron ore mining, the answer becomes emphatically “no.” When the U.S. industrialized, steel and bulk commodities traded in local markets. In contrast, China's industrialization reflects a world of global bulk commodity markets, where the productivity of the world's best iron ore miners (Australians) — rather than just the geographically proximate producers — can be utilized to advance capital stock formation. In effect, China has been able to outsource its mining industry to Australia, and in doing so it has increased the effective productivity of its raw material consumption radically. Every million tons of steel embedded into the U.S. economy has consumed 475 man-years of labor in the form of iron ore. On the other hand, every million tons of steel embedded into China (so far) has consumed, on average, only 183 man-years of labor. So while China is embedding steel into its capital stock at approximately double the average speed of the U.S., the steel consumed reflects a use of resources (labor) only one-fourth as intense as that used during the U.S. industrialization. The wholesale acquisition of Western steel plants and an ability to replicate Western levels of steep productivity through the purchase, disassembly and reassembly of steel-making production in China enabled rapid Chinese steel growth (e.g., the Famous Industrials Group industry relocation projects). In essence, China combined the efficiency of large blast furnaces with cheap labor and raw materials to affect an incredibly high return to capital, which has enabled the rapidity of capital stock accumulation in China so far. Consequently, from a demand perspective, we consider the fear that the iron ore price will decline to US$80/t and remain there is overdone.
Steel Intensity and Productivity It is estimated that it took 30,000 people 30 years to build the Great Pyramid of Cheops — a total of 900,000 man-years of labor embedded into 6 million tons of Egyptian limestone (of course, today many would point out that the limestone intensity of Egypt was far higher than that seen in any previous period of temple building. Consequently, they would criticize the endeavor as being illustrative of massive capital misallocation and thereby predict the imminent collapse of Egyptian civilization some 3,000 years too early). The vast increases in labor productivity can be powerfully illustrated by pointing out that the same edifice today could be pulled together by 150 Australian iron ore miners in a year. This
18 In the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter, we demonstrate that the market is currently factoring in a long-term iron ore price of US$80/t or a 40% discount to the June 12, 2013 spot price.
Every Million Tons of Steel Embedded Into the U.S. Economy Has Consumed 475 Man-Years of Labor in the Form of Iron Ore Versus 183 in China; Absent Western Capital Ill-Discipline, We See Price Support for Iron Ore Until 2023
Neither Farmers Nor Hairdressers Consume a Huge Amount of Metal; Rather, Metal Is Required to Embed the Capital Stock That Allows the Transition from Primary to Tertiary Forms of Value-Add to Take Place
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
42 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
represents a 6,000-fold increase in the efficiency of human labor. We would understand this increase as being driven by the ability of capital stock (machinery) to convert high-intensity energy sources (oil) into useful work far more effectively than the human body can convert food into output. It is against this backdrop that we illustrate the change in the operating model of the Australian mining industry over the last 50 years in Exhibit 51 and Exhibit 52. These exhibits not only illustrate the transition in the mining industry, but also offer a qualitative impression of the main theme of this chapter — the impact that mining productivity improvements have had on the rapidity of capitalization and urbanization in China over the last decade. It is easy to forget how phenomenally effective Rio Tinto, BHP Billiton and the like are at the task of mining (of course, whether their economic analysis is of the same caliber is a different question). The engineering competence on display at well-run mining operations and the increase in this over time have significant implications on how we should understand the development of industrial societies.
Exhibit 53 illustrates the evolution of U.S. industrialization from 1900 to date, exemplified by the rate at which metal is embedded into the economy. The picture is of an agrarian society moving to an industrial society as its economy transitions from one dominated by farming to one that is services focused. At either end of this transition, metal intensity (rate of consumption/embedding) is low: neither farmers nor hairdressers consume a huge amount of metal. Rather, metal is required to embed the capital stock that enables the transition from primary to tertiary forms of value-add to take place.
In this transition, labor productivity plays two vital roles. First, improved agricultural productivity is required to generate a surplus level of output that frees labor from the task of feeding itself and enables it to branch out into other economic activities. The greater the agricultural productivity, the more labor can be freed from farming and moved into the production of raw materials. Next, the greater the level of productivity in mining (for example), the greater the output of raw materials that can be generated by this now-surplus, non-agricultural labor. The greater the output of these raw materials, the greater the rapidity with which a capital stock can be installed. And, of course, the faster the capital stock can be installed in a country (i.e., bridge, buildings, rolling stock, machinery, etc.), the faster the transition to a service-sector economy can be achieved. This is a relatively simple narrative, but one which we have seen repeated over and over in the database we built, which looks at the rates of metal consumption/embedding across 120 countries since 1900.19 Furthermore, we have the same narrative in the back of our minds when considering the current and future role for metals in the global economy.
19 Iron - Cold Iron - Is Master of Them All...(Or At Least 60% of EBITDA) - The Forms of Steel Use & Iron Ore Demand (2/5).
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 43
Exhibit 51 Mining in Western Australia in 1961...
Source: 1961 Report of the Department of Mines, Western Australia, Presented to Both Houses of Parliament in 1962.
Exhibit 52 ...And Today
Source: BHP Billiton.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
44 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 53 At Either End of the Industrialization Process, Metal Intensity (Rate of Consumption/Embedding) Is Low; Rather, Metal Is Required to Embed the Capital Stock That Enables the Transition from Primary to Tertiary Forms of Value-Add (or Farmers to Hairdressers)
Source: WSA, Mitchell, IMF and Bernstein estimates and analysis.
Exhibit 54 shows China's path of steel intensity and summarizes why there is significant anxiety about the steel-intensive nature of Chinese economic development. The peak steel intensity in China was ~65kg/$'000 of GDP — some 60% higher than that in the U.S. It should be recalled, however, that in the U.S. steel intensity peaked ~100% higher than the peak steel in France. Consequently, the dangers of drawing uncritical parallels between China and the U.S. are obvious, and yet the fact that Chinese metal intensity is high has led many to fear that it is "unsustainably" high and that a rapid retrenchment from current levels is imminent.
China's overall economic growth is decelerating. Moreover, the country's metal intensity (rate of consumption/embedding) will track below rather than above Chinese GDP growth. However, part of the reason for the decline in metal growth (and thereby indirectly GDP growth) in China is the impact of high commodity prices. To the extent that the world was willing to supply the Chinese economy with nearly "free" raw materials (i.e., iron ore at US$12/t), it showed itself quite able to absorb these raw materials at an unprecedentedly rapid rate. Clearly, there could be hysteresis in this relationship, but there is at least a prima facia case that lowering commodity prices will see Chinese growth re-accelerate.
That China's apparent steel intensity exceeds that seen in the U.S. belies the fact that far more real resources were embedded in the U.S. capital stock than are being embedded in China today. This comes down to the creation of a globalized commodity market and access to 2.5-5x more efficient (in man-hour terms) iron ore mining. This is made possible not only by geological and population differences, but also due to the capitalization of the mining industry. The West underwrote its own industrialization (and that of post-World War II Japan) through the removal of labor from the mining sector in favor of capital. We question how much more Western capital will be embedded into the sector. In the case of the iron-ore mega projects, capital spend represents a value transfer from Western shareholders to
0
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 50 55
Ste
el In
ten
sity
(kg
/$'0
00
GD
P)
GDP/Capita (US$'000)
U.S. Finished Steel Demand Intensity Since 1900
SI Actual SI Trend
2012
1900
1917
1975
1948
The Peak Steel Intensity in China's Ongoing Industrialization Was Some 60% Higher Than That Seen in the U.S.; But the Peak Only Indicates the Rapidity of the Industrialization Process —the U.S Peaked 100% Higher Than France When the Latter Industrialized
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 45
China through suppressing iron ore prices and thereby underwriting the continued industrialization of the Middle Kingdom.20
Exhibit 54 The Chinese Economy Has Seen Steel Intensity Peak at ~65kg/$'000 of Output (vs. 40kg/$000 U.S. Peak); Because China's Peak Was More Than 60% Higher Than the U.S., and Because Many Datasets Look at Only the Last ~20 Years of Data, This Has Spawned Fears That the High Rate of Steel Consumption in Recent Years Is Indicative of Overinvestment and That the Chinese Economy Is Unduly (and Inefficiently) Steel Intensive
Source: WSA, Mitchell, IMF and Bernstein estimates and analysis.
20 The literal translation of Zhong-guo — the Chinese name for China, which can also be rendered as Middle State. America is Meiguo, or Beautiful Kingdom/State.
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45 50
Ste
el In
ten
sity
(kg
/$'0
00
GD
P)
Output — GDP/Capita — $'000 (Real 2005 PPP)
Trend Steel Intensity Evolution
France USA China Japan South Korea China Actual
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
46 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Globalization, Freight and China's Industrialization During its industrialization, China has had access to Australian iron ore production (20,000 tons/man-year at peak versus the U.S. peak of ~9,000 tons/man-year). It is not capitalization that made the difference between the U.S. and Australian industries, as both countries have well-capitalized mining industries. Rather, three factors stand behind this higher Australian productivity and China’s access to it: globalization of the commodities industry, different geological endowment and different population densities.
Exhibit 55 and Exhibit 56 display the history of global freight rates, going back to 1750. This is a vitally important place to start when considering the history of bulk commodities economics as freight rates are the driving force behind the decline in regional commodity markets and the rise of a truly global trade in iron ore and coal. Between 1965 and 2005, real freight costs fell 40%, while global trade in iron ore rose 330% and coal 1,070%. Global markets arise when the value of freight is lower than the value of the commodity being traded. These markets fragment again, of course, when freight rates rise relative to the cost of the commodity.21 Prior to the advent of global trade in the bulk industrial commodities (and it is these commodities that form the bedrock of industrial development), the level of national raw material productivity was the critical determinant of the pace of industrial development. In Exhibit 57 through Exhibit 59, we summarize the economic history of the U.S. iron ore industry since 1900. Productivity in U.S. iron ore production peaked at nearly 9,000 tons/man-year, and the path to that level was by no means steady. For large parts of its history, the U.S. was producing iron ore at productivities of less than 4,000 tons/man-year.
Exhibit 55 Global Freight Rates Have Declined in Real Terms by an Average of 0.9% p.a. Over the Last 260 Years...
Source: Stopford and Bernstein estimates and analysis.
21 Consequently, it is ironic that, of the "Big Three" global miners, Vale seems determined to undermine the pricing support of its key commodity and increase the risk of fragmenting global iron ore markets. As the most geographically distant from the bulk of iron ore demand (Asia), we estimate that China alone represented 56% of 2012 iron ore demand and produced just 13%, importing the balance of 42%. Consequently, Vale's profits and shareholders are the most vulnerable if the company's own cheap supply removes the need for high-cost inefficient artisanal mining in China. Vale's determination to bring huge low-cost supply onstream seems to indicate that its strategic direction is determined by incentives different than those of other publicly listed companies, perhaps due to the significant ownership by the Brazilian government (5.5% direct through Golden Shares and 34% indirect through the strategic consortium of Valepar).
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
1750
1760
1770
1780
1790
1800
1810
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
Re
al 2
000$
Global Freight Index 1750-2012
Access to Iron Ore Production That Is 2.5-5x More Efficient Than During the U.S. Industrialization Has Helped Enable the Rapidity of China’s Ongoing Development
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 47
Exhibit 56 …Leading to the Emergence of a Truly Global Market for Bulk Commodities, Including Iron Ore and Coal; Consequently, Relative Global Competitiveness in the Production of These Commodities Began to Matter
Source: Stopford and Bernstein estimates and analysis. Exhibit 57 Before the Advent of Scrap-Based EAF Production, U.S. Iron Ore Production Peaked
in the 30 Years Between 1950 and 1980
Source: USGS and Bernstein estimates and analysis.
0
200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,00019
00
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
Re
al 2
000$
Global Freight Index 1900-2012
0
20
40
60
80
100
120
140
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
Iro
n O
re —
Mt
US Iron Ore Production
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
48 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 58 The Transition to EAF-Based Production Saw Declines in Both Iron Ore Production and Employment
Source: BLS and Bernstein estimates and analysis. Exhibit 59 Consequently, the U.S. Iron Ore Productivity Peaked at Nearly 9,000 Tons/Man-Year
(Although for Much of the U.S. Industrialization, Its Intensity Was Far Lower)
Source: Govt. of Western Australia and Bernstein estimates and analysis.
In contrast, we show the development of iron ore mining in Western Australia
(see Exhibit 60 through Exhibit 62). Unlike the iron ore industry in the U.S., in Australia, it has never been intended for domestic consumption and has always had a strong international market orientation (Western Australian Government lifted its embargo on iron ore exports in 1960), focusing originally on post-WWII reconstruction and the subsequent industrialization of Japan and now on China. This development operated hand-in-hand with the declines in the international
0
5
10
15
20
25
30
35
40
45
5019
00
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
Th
ou
san
d P
eop
le E
mp
loye
d
US Iron Ore Mining Employment
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
To
ns
pe
r Man
-Yea
r
US Iron Ore Mining Productivity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 49
freight costs, which enabled the economic movement of low-value material over significant distances. Consequently, the industrial development of the U.S. was predicated largely upon the productivity of its domestic mining industry. The U.S. did not and could not take advantage of the highest level of global raw materials production when it was industrializing and had to be content with the level of productivity that prevailed at the local or national level. So, while the U.S. iron ore intensity peaked at ~9,000 tons/man-year, Australia peaked at 20,000 tons/man-year. It is this level of productivity that the Chinese are also able to access (albeit indirectly) today, and the Japan and the Asian Tigers were able to enjoy during the later part of the 20th century.
Having said this, it is important to understand the difference between the relative productivities of Australia and the U.S. First of all, both American and Australian iron ore industries are very well capitalized, and it is capital rather than labor that stands behind raw material production (in contrast to China where mining is still labor driven). However, the two key differences are geology and population density. Different geological endowments: A miner in the U.S. needs to move about
twice as much run of mine (ROM) material as his Australian counterpart to generate the same amount of useful material. This is due to the difference between accessing direct shipping ore hematite (62% Fe grade) available throughout the Pilbara and magnetite (~35% Fe ROM) found in the U.S., which requires beneficiation and processing before it can be converted to pellets used by the U.S. steel makers.
Population density: Western Australia (ex-Perth) has, on average, 0.23 people per square kilometer versus 67 in Minnesota and 175 in Michigan. Virtually no one lives in Western Australia outside of Perth (see Exhibit 63). As a resource-intensive industry, mining requires significant land, water and power (not to mention there are those who have questioned the aesthetics of open-cut mining, as odd a suggestion as that may sound). A high population density always means that there will be competing calls on the resources that a miner needs as well as the potential for objections from local residents on aesthetic and environmental grounds. Consequently, it is much easier to mine in a sparsely populated area. On this note, China is 1,300 times more densely populated than the Pilbara!22
Before moving on to the labor productivity required to generate each ton of steel hitherto embedded in the Chinese capital stock, it is worth addressing why the declines in Pilbara mining productivity post 2006 are linked to the use of labor rather than declining asset productivity (which might initially appear to be the primary cause) — see Exhibit 62. It is easier to see what is going on if we look at the breakdown of production, employment and productivity by producer in the Pilbara (see Exhibit 64 through Exhibit 67). The large producers, Rio and BHP, have not suffered productivity declines anywhere near as severe as the average would suggest, though even these producers have suffered deterioration. However, for FMG, using the alternative surface mining technology, the productivity is about half that of the traditional mining methods. This indicates that the decline in productivity is largely attributable to the employment of labor on project and exploration sites for assets that are yet to produce rather than necessarily indicating a reduction in productivity from the mines already producing material. Consequently, Exhibit 68 shows the range of productivities that we use in the subsequent analysis.
22 Chinese statistic, excluding the provinces of Inner Mongolia, Xinjiang, Tibet, Qinghai and Gansu, as there is limited population and negligible mining.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
50 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 60 In Contrast to the U.S., Western Australian Iron Ore Industry Was Always Driven by Exports — First to Japan and Today to China; the Declines in Shipping Costs Facilitated This Greatly
Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 61 Superior Geology and Continuing Improvement in Mining and Logistics Technology
Have Led to Substantial Increases in Mining Productivity
Source: Govt. of Western Australia and Bernstein estimates and analysis.
0
50
100
150
200
250
300
350
400
45019
60
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
Iro
n O
re —
Mt
Western Australia (Pilbara) Iron Ore Production
0
5
10
15
20
25
30
35
40
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
Th
ou
san
d P
eop
le E
mp
loye
d
Western Australia (Pilbara) Iron Ore Mining Employment
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 51
Exhibit 62 In Addition, Low Population Density Renders Iron Ore Mining in Pilbara Hugely Productive...
Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 63 …And Provides a Significant Advantage Over the U.S. or China as Mining
Jurisdictions
Note: China excludes the "five provinces" of Inner Mongolia, Xinjiang, Tibet, Qinghai and Gansu.
Source: Wikipedia and Bernstein analysis.
0
5,000
10,000
15,000
20,000
25,000
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
To
ns
pe
r Man
-Yea
rWestern Australia (Pilbara) Iron Ore Mining Productivity
295
175
67
0.89 0.23
China Michigan Minnesota Western Aus WA (ex Perth)
De
ns
ity
—P
ers
ons
pe
r Sq
uar
e km
Population Density by Iron Ore Mining Region
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
52 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 64 The Consolidation of the Industry Can Be Seen in Both the Concentration of Employment...
Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 65 ...And Production in Western Australia
Source: Govt. of Western Australia and Bernstein estimates and analysis.
14,278
9,712
5,974
1,240 1,177611 374 325 254 171 135 115 68 8
Peo
ple
Em
plo
yed
Western Australia Iron Ore Mining Employment (2011)
0
50
100
150
200
250
Iro
n O
re —
Mt
Western Australia Iron Ore Mining Production (2011)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 53
Exhibit 66 Among the Majors, Rio and BHP Have Almost Identical Productivities and Are Almost Twice as Productive as Either FMG or Cliffs
Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 67 It Is Also Interesting to Note the Significant Declines in Productivity Seen at the
Company Level as Well as in the Industry Overall in Recent Years
Source: Govt. of Western Australia and Bernstein estimates and analysis.
26,901
16,397 16,011
8,7047,579 6,870 6,709 6,497
3,653 3,0772,035
To
ns
pe
r Man
-Yea
rWestern Australia Iron Ore Mining Productivity (2011)
-
20
40
60
80
100
120
140
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
100
= 20
02
Indexed Iron Ore Mining Labor Productivity
BHP Cliffs Resources Rio Tinto
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
54 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 68 However, the Decline in Productivity in Western Australia Is Due to the Inclusion of Labor That Is Allocated to Building Projects (With a Significant Production Lead Time) vs. Labor Allocated to Operations That Generate Current Volume
Source: Govt. of Western Australia and Bernstein estimates and analysis.
When turning to China and its recent rates of steel production and iron ore demand (see Exhibit 69 and Exhibit 70), it is important to stress how radically different its domestic iron ore production is from either the U.S. or Australia (see Exhibit 71 through Exhibit 75). Vast amounts of (hitherto) cheap labor historically stood behind the raw material production in China rather than the use of capital and machinery. However, post the late 1970s' reforms and the beginning of Chinese integration into the global economy (GATT and finally WTO membership), Chinese raw material consumption underwent a radical change. Imported raw materials rather than domestic production became the mainstay of industrial activity. The exports into China allowed the PRC to access the ever-increasing productivity of Australian iron ore miners (see Exhibit 76).
Since 1990, self-sufficiency in Chinese iron ore production has only been declining (see Exhibit 71). Given that the marginal producer of iron ore globally is in China (as discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter), it is the high costs of domestic Chinese production that are currently setting raw material prices. Removing these producers will catalyze the end of the current secular period of pricing support (short-term supply demand fluctuations notwithstanding). In this regard, the trend decline is remarkably stable; if history is projected forward (admittedly, always a dangerous move), it would see China falling to 10% self-sufficiency (and thus becoming irrelevant to price formation) only by 2023.23 Clearly, extrapolating trend ignores the fact that conditions change and that supply and demand may accelerate, but due to the stabilizing effect of demand elasticity it may well be that the path to oversupply takes longer to come into play than many realize (see Exhibit 72).
However that may be, while the U.S. was reliant on its own geology and mining productivity to stand behind its industrialization, the decline in global
23 This assumes straight-line trend of declining Chinese self-sufficiency due to continued importation of iron ore into China by the majors at the same historical absolute amount of increased tonnage (rather than historical average growth rate of importation). It does not take into consideration the actual new project pipeline.
0
5,000
10,000
15,000
20,000
25,00019
60
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
To
ns
pe
r Man
-Yea
r
Western Australia (Pilbara) Iron Ore Mining Productivity
Productivity From Operations Productivity Including Projects
China Is Embedding Approximately One-Fourth the Labor Into Its Capital Stock Than the U.S. Did, But at a Steel Intensity That Is Approximately Double That of Peak U.S. Steel Consumption
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 55
freight costs enabled China to outsource this to Australia and take advantage of the superior geology and mining environment in the Pilbara. If we calculate the average productivity of the iron ore that China has actually consumed in its industrialization, we can see that it actually outstrips that of the U.S. (see Exhibit 77). The weighted average productivity24 of iron ore produced over the industrial history of the U.S. stands at 3,350 tons per man-year. The average productivity that stands behind the Chinese economic development is 8,750 tons per man-year. To return to Cheops, today we would be able to build 6,000 Great Pyramids, where the ancient Egyptians built one, employing exactly the same amount of manpower. This testifies to the increase in labor productivity over the intervening 5,000 years. Pyramid building in toto may be unproductive insofar as it satisfies no real human need, but that cannot be deduced from the relative intensity with which one society constructs pyramids versus another at a very different point in time.
Turning to Exhibit 78 and Exhibit 79, we see an incredibly strong relationship between relative productivity in raw material production and relative intensity of steel use between the U.S. and China. Chinese steel intensity is not much greater than that seen in the U.S. (or indeed anywhere else). Peak U.S. iron ore production occurred in 1951, at which time a U.S. miner could produce at a rate of 3,150 tons per year and steel intensity was 33kg/$'000. Today, steel intensity in China is 63kg/$'000 but the effective productivity of iron ore production is 13,600 tons/man-year. China is therefore embedding approximately one-fourth the labor into its capital stock that the U.S. did, despite having higher steel intensity in absolute terms. The more rapid embedding of steel in China's capital stock reflects not inefficiency per se, but rather that the productivity of the world's best iron ore miners (Australians) as opposed to just the geographically proximate producers can be utilized to advance the process of capital stock formation.
Exhibit 69 China Has Seen Both Its Steel… Exhibit 70 …And Iron Ore Production Increase
Source: WSA, NBS and Bernstein estimates and analysis. Source: WSA, NBS and Bernstein estimates and analysis.
24 Weighted by tons of iron ore.
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
800,000
Cru
de
Ste
el (k
t)
Chinese Crude Steel Production
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
Iro
n O
re (k
t)
Chinese Iron Ore Demand
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
56 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 71 The Iron Ore Majors (Vale, Rio and BHP) Have Aggressively Grown Their Volume Share of the Chinese Iron Ore Market Over the Last Decade, While Chinese Self-Sufficiency Has Declined Significantly
Source: WSA, NBS and Bernstein estimates and analysis. Exhibit 72 Extrapolating the Trend Decline in Self-Sufficiency Forward Suggests the End of the
Chinese Iron Ore Industry (and With It the High-Cost Price Support for Iron Ore) in the Region from 2020 to 2025; Clearly, Acceleration in Supply from the Seaborne Market and Declining Steel Growth Rates Would Advance This
Note: 2014-20 are Bernstein estimates.
Source: WSA, NBS and Bernstein estimates and analysis.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
Ch
ine
se S
elf
-Suf
ficie
ncy
(%)
Iro
n O
re (k
t)
Chinese Iron Ore Supply
Imports Domestic Production Self-Suff iciency
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Se
lf-S
uff
icie
ncy
(%
)
Chinese Iron Ore Self-Sufficiency — Trend Decline
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 57
Exhibit 73 Chinese Growth in Iron Ore (on a Rich-Ore-Equivalent Basis) Peaked in 2007; Since Then, Declining Grades and Rising Costs Have Seen Useable Ore Production Decline
Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 74 An Ever-Increasing Amount of Labor Has to Be Allocated to China's Iron Ore
Production…
Source: NBS, WSA and Bernstein estimates and analysis.
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000
450,000
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Iro
n O
re (R
OE
Bas
is) —
kt
Chinese Domestic Iron Ore Production
0
100
200
300
400
500
600
700
800
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Th
ou
san
d P
eop
le E
mp
loye
d
Chinese Domestic Iron Ore Employment
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
58 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 75 …Leading to a Prolonged Decline in the Productivity of the Chinese Domestic Industry Since 2007
Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 76 An Australian Miner Is 35x as Productive as His Chinese Counterpart; However,
Given That Globalization of the Iron Ore Industry Has Enabled China to Outsource Its Iron Ore Mining to Australia, Does That Differential Matter?
Source: WSA, NBS, BLS and Bernstein estimates and analysis.
0
100
200
300
400
500
600
700
800
900
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
To
ns
pe
r Man
-Yea
r
Chinese Domestic Iron Ore Productivity
0
5,000
10,000
15,000
20,000
25,000
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
To
ns
per
Man
-Yea
r
Western Australia (Pilbara) Iron Ore Mining Productivity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 59
Exhibit 77 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003
Source: WSA, NBS, BLS and Bernstein estimates and analysis. Exhibit 78 We Compare the Rate at Which Steel Is Being Embedded in China Relative to Peak
U.S. Steel Intensity (Y Axis) and the Peak Iron Ore Production (in Man-Hour Terms) to Which China Has Access Relative to That Accessible to the U.S. During Its Industrialization (X Axis); However, Due to Productivity Improvements, the Real Resources Consumed to Embed Steel in the Chinese Economy Are Markedly Lower Than Was the Case During the Industrialization of the U.S.
Source: WSA, NBS, BLS and Bernstein estimates and analysis.
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
To
ns
pe
r Man
-Yea
rEffective Chinese Productivity, Including Imports
USA Effective Chinese Productivity
R² = 0.8458
80%
90%
100%
110%
120%
130%
140%
150%
160%
20% 40% 60% 80% 100% 120% 140% 160% 180%Ch
ine
se S
teel
Inte
nsit
y to
Pe
ak U
S S
tee
l In
tens
ity
Ratio of Chinese Iron Ore Productivity to Peak US Iron Ore Productivity
Relative Iron Ore Productivity to Relative Steel Intensity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
60 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 79 U.S. Iron Ore Production Peaked in 1951, With an American Miner Producing 3,150 Tons/Year and Steel Intensity of 33kg/$1,000; Today, China's Steel Intensity Is 63kg/$'000, While the Effective Productivity of Iron Ore Production Is 13,600 Tons/Man-Year
Source: WSA, NBS, BLS and Bernstein estimates and analysis.
Imitation is often touted as the sincerest form of flattery, but we would argue that the outright purchase and adoption of another's technological developments go at least one step further. We mentioned earlier in this chapter that post the Open Door Policy and the beginnings of Chinese integration into the global economy (GATT and finally WTO membership), Chinese raw material consumption underwent a radical change, with imported raw materials rather than domestic production becoming the mainstay of industrial activity. This was not the only radical change that enabled greater steel embedding into the Chinese economy. The Chinese also purchased steel-making technology from the West; one of the more "famous" examples of this is the Famous Industrials Group, which began importing precision measuring tools and lock cylinders into China in the 1990s. Famous expanded into industry relocation projects; over the last 10 years, the company has relocated departments and production, including six complete industrial plants and production lines from the West into China.25 Some examples include: 1995 saw the purchase of the Bergkamen Coal Washing plant from the Monopol
mine (Northern Rhur Valley), its subsequent disassembly and transport to China. Today, it is run by SDIC Xinji Energy Co., Ltd., a Chinese coal producer with mines in the Anhui province (see Exhibit 80).
Three years later, in 1998, Famous purchased the Sophia Jacoba coal washing plant from the Hückelhoven coal mine in the Rhine Valley. Around 100 Chinese experts spent eight months disassembling the plant before relocating it to Eastern China, where it is also operated today by SDIX Xinji Energy Co (see Exhibit 81).
2002 saw the purchase of the Kaiserstuhl Coking Plant from the Dortmunder Westfalenhuette. The following year saw its disassembly by 300 Chinese specialists followed by its 2004 reassembly in Jinin, Shandong province. The
Chinese to Peak US Steel Intensity Chinese To Peak US Iron Ore Productivity
The Rate at Which China Has Been Embedding Steel Has Furthermore Been Accelerated by the Purchase of Western Steel-Making Technology
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 61
plant, which had only been operational for eight years prior to its purchase, recommenced operations in 2006 (see Exhibit 82).
It is worth noting that the purchase of entire plants and production lines is not limited to the mining and steel-making sector; Famous has also relocated spring factories (for the manufacture of trucks and railway vehicles), underground refrigeration plants, foot processing and packaging factories, and automation and communications systems, to name a few of its non-mining projects (see Exhibit 83). This process of disassembly and reassembly means that in a bare manner of months, not only did China benefit from the specifically relocated department or production itself, but the "specialists" involved in the dis- and re-assembly are now able to replicate and reproduce that technology on a wider scale across China. This further enables the rapidity of the Chinese industrialization process.
Source: Famous Industrial Group GmbH. Source: Famous Industrial Group GmbH.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
62 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 63
Iron Ore Pricing Support Over the Next Decade Depends Upon Capital Discipline (Rio 360 Case Study)
On the supply side, as noted in the "Iron Ore 'Super-Cycle' Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization" chapter, we see the inability of the Chinese mining sector to transition from labor-intensive to capital-intensive modes of production as a consequence of poor geological endowment supporting the current "super-cycle" pricing (this is, of course, bearing in mind that this failure is occurring against the backdrop of the rapid industrialization of China). As we see the costs on the supply side providing the long-term pricing structure, pricing support will persist, in our view, so long as a significant proportion of high-cost marginal producers in China (~35% of 2011 rich ore equivalent [ROE] Chinese supply, with marginal costs of production ~$120-180/t) are not displaced. Should the Western miners actively cease to underwrite continued Chinese industrialization (as the West underwrote its own and that of Japan), we see pricing support for iron ore continuing for the duration of China's industrialization (discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter). If not, then it is 2015 that we see as the earliest time at which there will be sufficient new production coming onstream to displace high-cost producers.
We are aware that we are beginning to sound like a later day and mining-focused version of Cato the Elder (if such a thing can be imagined) and finish every piece with a cry that "capital is to be disciplined." Yet capital discipline is the single greatest source of value creation in the mining space — and this is particularly critical for iron ore, the fourth most crustally abundant element. For any large mining company, an "unrelenting focus" on value is synonymous with ensuring that capital is allocated to new volume only if it does not lead to price destruction. An “unrelenting focus” on shareholder value is diametrically opposed to any strategy designed (explicitly or not) to lower commodity prices.
This chapter continues with the "value maximizing growth strategies for large mining companies" theme. In particular, we focus on the possible 70Mtpa iron ore expansion by Rio Tinto as a case study and show how the decision to proceed (or not) with the project depends on how comfortable Rio Tinto owners are with the 3% steel demand growth target that the company's management clearly believes in. Shareholders in Rio Tinto may be comfortable risking an $18 billion loss on the chance of an $8 billion gain should China achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of the Emperor Augustus in considering the whole enterprise akin to fishing with a golden fish hook and that the gain, if all goes well, is too small to offset the risk of loss if it does not.
We further expand our analysis to demonstrate how globally superior value generation can result from the pursuit of locally optimal growth targets in consolidated markets, where industry incumbents are explicitly aware of the price destructive effects that excessive volume growth has. In itself, market consolidation and a "good" industry structure are not sufficient to drive superior returns. Instead, the miners require an approach to value that explicitly captures the price destructive effect of new volume. For us, the critical feature of the debate on Rio Tinto 360 is the emerging language from the company recognizing (for the first time) that the value of the investment proposition turns on the market impact that the extra tons
Capital Discipline Is the Single Greatest Source of Value Creation in the Mining Space; an "Unrelenting Focus" on Value Is Synonymous With Ensuring That Capital Is Allocated to New Volume Only If It Does Not Lead to Price Destruction
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
64 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
will have. In our view, it is a great leap forward and could precipitate a re-rating of the iron ore industry.
Capital Discipline Is Not the Same as Returns of Capital We have written previously and extensively on the need for capital discipline. In our call, Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value, we showed how natural limits to growth emerge in consolidated commodity industries. Moreover, we believe that a natural evolution from "growth" to "value" exists in any mining company as it grows bigger. This transition corresponds to the development of a mining company from the entrepreneurial activity of new asset creation for a junior to the custodial function of value and asset preservation for a major. That is, the most important skill set for a mining junior is necessarily focused on the engineering and project-management aspects of mining. In contrast, once established as a major producer, it is the commercial, strategic and economic abilities that the miner's management will have tested.
Much has been made of capital discipline in the last few months, with CEO after CEO lining up to claim this virtue and to extol his "unrelenting focus" on shareholder value creation. However, what does capital discipline mean in practice, and why should it actually matter? After all, doesn't capital provide the lifeblood for continuing activity?
In theory, and ignoring taxation implications, shareholders should be ambivalent between receiving a dividend, which can be reinvested at their cost of capital, and having that dividend retained in the company to finance growth. If anything, it is always possible to generate a synthetic dividend stream,26 which — transaction costs aside — should render dividend policy a value-neutral proposition. Given this, capital discipline becomes a virtue only if capital allocation has gone wrong. Consequently, we view capital discipline in mining as a tool for preventing excessive organic growth. High commodity price preservation through maintaining a grip on supply is the single greatest value driver that mining companies have at their disposal. The greater the capital distribution returned to shareholders, the less likely a catastrophic oversupply is to result. In our view, mining will always be a cyclical business. However, it is up to the incumbent players to determine whether to feed or dampen that cyclicality. An "unrelenting focus" on value — as far as any large-scale miner is concerned — is synonymous with ensuring that capital is allocated to new volume only if it does not lead to price destruction.
For us, capital discipline is not about returns of capital per se. Instead, we see it as a mechanism to ensure that new volume growth is more accurately aligned with the forward-looking demand environment. We prefer our companies to run the risk of undersupplying the market and seeing higher prices, with the threat of possible demand destruction, to them oversupplying the market and guaranteeing prices will collapse. In other words, given the inherently asymmetric risk profile facing the miners, we would like them to play the odds in a manner that favors value creation for their own bottom lines and their shareholders. Consequently, we are not indifferent to the ways the capital is to be returned. Specifically, for us, capital discipline is not the same as asset disposal; we certainly would not like to see the miners execute wholesale asset disposals at (possibly) "fire sale" prices. Offloading assets to competitors is a risky strategic enterprise, which could result in de-consolidation of the market, while giving a potential competitor access to geology that it would not otherwise have (as we discuss later in this chapter, industry consolidation is one of the few means by which superior returns can be created in commodity markets). Moreover, in absence of synergies, shifting assets from one hand to another would create value only if the markets were inefficient
26 By either selling stock outright or writing out of the money call options on it to achieve the required income. From a risk-return perspective all of these alternatives are equivalent.
Capital Discipline Becomes a Virtue Only When Capital Allocation Has Gone Wrong; the Greater the Capital Distribution Out of the Mining Space, the Less Likely a Catastrophic Oversupply Is to Result
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 65
over a sustained period. In our view, such belief cannot become the strategy for the industry as a whole. Finally, asset disposals masquerading as capital discipline may enable the miners to sidestep the real objective, i.e., rationing of supply. Consequently, it is not returns of capital per se that we look for, but returns of capital that would otherwise be expended to accelerate supply into a declining demand environment. Clearly, we are not suggesting that an asset should be kept even if a truly attractive price is offered — rather, lucrative asset disposal opportunities should be explored all the time. Nevertheless, in our view, such exploration is more likely to yield attractive results at the top of the cycle rather than the bottom of it.
The mining industry is facing a dilemma at the moment. Although equity prices are reflecting a very significant and irreversible decline in commodity prices (most notably, for iron ore it sees the long-term price of US$80/t), the price environment, while volatile, is not yet utterly unsupportive of earnings. The last time equity valuations were as low as currently, the iron ore price had hit just US$85/t and it looked as if it were game over for that particular commodity. So why, when the 1H:13 iron ore price averaged US$129/t Aus FOB, is this not being reflected in equity valuations?
We believe that this comes back, again, to capital discipline. For retained earnings to be valued, they must have the prospect of being returned to shareholders at some future — even if far removed — date (hopefully, magnified many times over). Currently, however, the market fears that greater mining operating cash flows will translate to proportionally higher organic growth plans. The greater the capital spend today, the greater the volume tomorrow and — in a slowing demand environment — the greater the likelihood of a sustained and severe supply demand imbalance and price destruction. Consequently, higher earnings today imply lower earnings tomorrow. Furthermore, however counter-intuitive it may seem, the longer commodity price environment stays strong, and the more mining companies will de-rate unless they alter their behavior and convince the market that they are truly earnest about capital discipline. The problem is not the price environment, rather it is the strategy of the miners and their reaction to that price environment. Simply converting ever-expanding operating cash flows into new sources of production must come to an end. In our view, it will happen either in a value-accretive manner, with the miners willingly forgoing organic growth opportunities, or violently, when commodity prices are driven down to such a level that new organic growth is impossible, while the promise of a return on historical investment is proven illusory.
We strongly believe that increasing the dividend by 15% or 50% is simply insufficient. For the miners to see a structural re-rating, there must be a clear change in strategy that sees returns to capital rather than expenditure of capex as the priority. In 2012, Rio Tinto spent ~US$18 billion on capex, ~US$6 billion on corporate tax and ~US$3 billion on dividends. Why not simply reverse the priority of capex and dividends (see Exhibit 84 and Exhibit 85)?
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
66 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 84 Assuming Disciplined Capex, We Expect Rio Tinto to Have Room to De-Gear Its Balance Sheet Significantly Over the Next Few Years
Exhibit 85 Curtailment of Capex and De-Gearing of the Balance Sheet Would Enable Rio to Raise Its Dividend from US$3.0 Billion in 2012; We See No Reason Why It Should Not Double Over Time
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
Rio Tinto Pilbara 360 Project — Does It Make Sense? In a meeting this past May that was closed to the media (and yet whose contents were known to the media almost immediately), Rio Tinto's management put forward a 70Mtpa Pilbara expansion to 360Mtpa to the company's board for approval in 4Q:13. We interpret this "leak" as an attempt by the company to gauge shareholder sentiment toward continued capital spend ahead of actually having to make the decision. While some pushback has resulted, rhetoric around value aside, we would expect management to execute this expansion if investor opposition grows muted. For us, the question is whether this expansion creates or destroys value for Rio Tinto and the mining sector.
In a recent interview, Rio's CEO Sam Walsh claimed that "the additional tonnage of 70 million tons was unlikely to result in a US$20/ton price decline."27 Critically, this statement represents the first time that the company explicitly recognized that the value of the investment may turn on the impact that it will have on the market. In our view, this recognition is a significant step forward for the industry, which moves the debate much closer to where it needs to be if returns from mining have any hope of being sustained into the medium term. While industry consolidation is a necessary requirement to see superior returns and genuine pricing power in a commodity industry, it is far from sufficient on its own. The necessary and sufficient condition is that there is industry consolidation as well as explicit awareness of the impact that new volume growth has on price. Pursuing industry consolidation — absent an understanding of how volume growth can act to undermine price and value — is merely a verisimilitudinous distraction. However, once this factor is understood, the prospect that superior returns can accrue naturally follows.
To make this point concrete, we look at the case study that Rio Tinto has provided in the form of its 360Mtpa Pilbara expansion. First of all, Rio Tinto, like
27 "Rio Said to Pursue $5 Billion Iron-Ore Project as Glut Looms," Bloomberg L.P., May 7, 2013.
2,775
45,182
38,577
18,861
4,071
8,807
19,41217,829
6,220
(3,729)
2006 A
2007 A
2008 A
2009 A
2010 A
2011 A
2012 A
2013 E
2014 E
2015 E
US
$m
Rio Tinto Net Debt
-
2%
4%
6%
8%
10%
12%
-
5,000
10,000
15,000
20,000
25,000
30,000
2012 A 2015 E
Div
iden
d Y
ield
(%)
US
$m
Rio Tinto Dividend Potential
Capex Tax Dividend Dividend Yield
Rio Tinto Owners Need to Get Comfortable With the Prospects for 3% Steel Demand Growth from China; It Is on This That the Value of the Expansion Turns; Will It Suppress Price Sufficiently to Render It Unattractive?
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 67
us, recognizes a significant inflection in the global cost curve (see Exhibit 86 and Exhibit 87). It should be noted that it is this inflection in a commodity cost curve that is responsible for both the margin that an industry generates as well as the volatility in that commodity's price. A steep cost curve implies significant industry profitability as well as significant risk; a flat cost curve has low profitability but a correspondingly low risk. As a result, the more valuable a commodity industry, the greater the requirement to ensure that volume growth is well managed and non-disruptive. Uninteresting and low-value businesses, on the other hand, can more or less look after themselves. Taking Rio Tinto's cost curve, we convert it from a discrete to a continuous mining output function for the industry as a whole, and from this curve, we calculate the elasticity of supply that is implicit in Rio Tinto's analysis. From this, we can see that Rio Tinto estimates the elasticity of supply to currently be somewhere between 3 and 4 (see Exhibit 88 and Exhibit 89).
Most importantly, an explicit awareness of price elasticity naturally generates the limits to organic growth. If one imagines an infinite reservoir of value-accretive projects, conventional wisdom would suggest that a company should execute as many of them as it can. Moreover, the same wisdom would suggest that there is a straight-line relationship between the number of projects that are executed (i.e., volume of output) and the value of that company (see Exhibit 90). However, this ignores that demand for a commodity is always finite and there is a negative relationship between commodity volumes and prices. The greater the volume in the market, the lower the price will be. Factoring that into the analysis of value maximization increasingly bends the straight line down into a parabola that shows a discrete point of maximum volume growth.
Exhibit 86 Rio Tinto's Cost Curve... Exhibit 87 ...Not Dissimilar to Our Own
Source: Rio Tinto. Source: AME, CRU, NBS and Bernstein estimates and analysis.
170
160
150
140
130
120
110
100
80
70
0
60
180
40
30
20
10
0
50
1,000,000750,000500,000250,000
90
1,250,000
Cumulative global production– kt Fe
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
68 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 88 Schematically, We Can Represent the Rio Tinto Cost Curve...
Source: Rio Tinto and Bernstein estimates and analysis. Exhibit 89 ...And So We Can Calculate the Elasticity of Supply That Rio Tinto Must Believe In
Source: Rio Tinto and Bernstein estimates and analysis.
0
20
40
60
80
100
120
140
160
180
200
0 200 400 600 800 1,000 1,200 1,400 1,600
Op
ex
(US
$/t)
Cumulative Iron Ore Production (mt)
Rio Tinto Iron Ore Supply
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Ela
stic
ity
Cumulative Iron Ore Production (mt)
Rio Tinto Iron Ore Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 69
Exhibit 90 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered
Source: Bernstein analysis and estimates.
Exhibit 91 summarizes this relationship. Given the elasticity that Rio Tinto
believes exists in the iron ore market, it ought to come to the conclusion that its expansion by 70Mtpa will lower the price of iron ore by ~US$27/t. Under such a scenario, it is simply expending US$5 billion of capital to reduce its total EBITDA by ~US$2.3 billion a year for a total value destruction of almost US$18 billion — not quite at Alcan levels, but getting there. This simple analysis can also be made more complex by building the full DCF model for the investment (or indeed by generating a closed-form algebraic solution to the problem as we did in Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value) but the essence stays the same. One could object that the analysis uses a price of US$140/t and that this is far too high given that we "know" that price will fall to US$80/t. However, this begs rather than answers the question of whether or not to proceed with the investment and is actually irrelevant to the analysis in hand, which turns on price differences, not price levels.
0
20
40
60
80
100
120
140
0 25 50 75 100 125 150 175 200 225 250 275 300
Val
ue
(US
$bn
)
Iron Ore Volume Growth for New Entrant (Mt)
Value of Volume Growth for Different Price Elasticity
0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0
"Normal" straight‐line relationship between volume and value...but
only if elasticity is ignored
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
70 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 91 Rio Tinto's Claim That It Does Not See a Price Fall of US$20/t on the Back of Its 70Mtpa Increase in Production Becomes the Key Variable in Deciding Whether or Not the Project Makes Sense; If the Project Induces a Greater Decline, It Will Be Net Value Destructive for the Company
Source: Rio Tinto and Bernstein estimates and analysis.
In any event, we also know that Rio Tinto believes in 3% iron ore demand growth from China. If we factor this in, the conclusion changes as shown in Exhibit 92. The extra demand growth is sufficient to limit the price declines to only ~US$14/t and so there is an incrementally positive contribution to EBITDA. That is, an expenditure of US$5 billion generates a positive value of ~US$8 billion. Clearly, everything here turns on whether the 3% demand growth in China is something that shareholders are happy with. Shareholders in Rio Tinto may be comfortable facing the loss of US$18 billion versus the gain of US$8 billion on the chance that China will achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of Emperor Augustus in considering the whole enterprise akin to fishing with a golden fish hook and that the gain, if all goes well, is too small to offset the risk of loss if it does not. Just as Augustus saw that there were limits to his own empire building and that, after all, Germany was just too small a prize to warrant the loss of Roman lives, so too we would caution that tons, once put in the market, are virtually impossible to remove. Much better to build only once there is clarity on the demand side from China. Festina lente, as Augustus would have it.
Original Volume Mt 290
Expanded Volume Mt 360
Change in Volume Mt 70
Current Chinese Iron Ore Demand Mt 1100
% Additional Supply from Rio % 6.4%
Price Elasticity - -3
Price Ex Expansion US$/t 140
Price Inc Expansion US$/t 113
Rio Tinto Unit Costs US$/t 35
Total Costs Ex Expansion US$bn 10.2
Total Costs Inc Expansion US$bn 12.6
Incremental EBITDA US$bn (2.3)
Multiple on EBITDA - 5.5
Less Capex Expended US$bn 5.0
Total Incremental Value US$bn (17.5)
Rio Tinto 360Mt Expansion Ex-Demand Growth
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 71
Exhibit 92 Rio Tinto Also Believes in a 3% Demand Growth World; This Would Be Sufficient to Limit the Price Impact of the Expansion to ~US$14/t and So Make It Appear Value Accretive (Though Clearly Not by the Degree That It Would Be If the Price Impact of the New Supply Were Neglected in Entirety)
Source: Rio Tinto and Bernstein estimates and analysis.
If we were asked to provide a guide as to our preferred course of action with
this project, we would suggest taking the time from now until 4Q to really understand the true value of expansion to 360Mtpa under a range of demand scenarios and price responses. When it comes to the 4Q decision point, we would prefer Rio Tinto to push this back to 2Q:14, under the pretext of undertaking further market studies. This keeps the threat of new supply in the market, but increases the likelihood of continuing market tightness for a longer period without unduly upsetting its largest customer — China. Post this point, Rio Tinto can always come back with a revised path to growth, which shows a markedly slower ascent to the 360mt target, if demand is really there for it. Moreover, we believe that Sam Walsh has given himself an "out" on the project by declaring it contingent upon there being no "significant change in the demand-supply situation" (ironic in some sense, given that this project is in itself a significant change in the demand-supply situation). We would like to see this "out" utilized and, by doing so, see Rio Tinto's management demonstrate how superior industry structure can drive superior returns. Positive industry structure is not in itself a standalone positive outcome; it enables a positive outcome only in so far as it supports superior value creation. There is no point in trying to preserve market share if this leads to substantial value destruction.
Original Volume Mt 290
Expanded Volume Mt 360
Change in Volume Mt 70
Current Chinese Iron Ore Demand Mt 1100
% Additional Supply from Rio % 6.4%
% Chinese Demand Growth % 3.0%
Price Elasticity - -3
Price Ex Expansion US$/t 140
Price Inc Expansion US$/t 126
Rio Tinto Unit Costs US$/t 35
Total Costs Ex Expansion US$bn 10.2
Total Costs Inc Expansion US$bn 12.6
Incremental EBITDA US$bn 2.3
Multiple on EBITDA - 5.5
Less Capex Expended US$bn 5.0
Total Incremental Value US$bn 7.5
Rio Tinto 360Mt Expansion Including Demand Growth
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
72 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
The Virtue of Consolidation — Why Does It Matter? This simple example provided by Rio Tinto can be usefully extended to consider the case of commodity industries in general. In Exhibit 93 and Exhibit 94, we set out two industry models (deliberately chosen to be caricatures of the iron ore industry) that both show the potential for significant oversupply with possible new project growth in excess of demand growth. Both industries are currently generating significant margins, as is shown by the price elasticity. The only difference between these two industries is the degree of consolidation — one is highly consolidated (four industry players or the Herfindahl-Herschman Index [HHI] of 0.25, the threshold of "high concentration") and the other fragmented. Under each market situation, we simply ask what the locally optimal value maximizing strategy is for each incumbent player and then calculate the global response as the sum of the local responses. We then consider four possible scenarios: industry fragmentation versus consolidation and an explicit understanding or not of the value impact of the elasticity of supply.
In the first scenario (see Exhibit 95), the industry is fragmented where each participant acts on the belief that the only change that building a project has is to increase volume for the participant, rather than to lower price as well. Under this scenario, each player thinks that its volume will be value enhancing and so executes the maximum possible growth. Each believes that it will be able to add an incremental US$18 billion of value locally (or 27% versus the starting value of US$66 billion) — but the global effect of this local value maximization is massive oversupply and the destruction of US$32 billion of value for each player. It is the realization of this outcome that should see a de-rating of the industry from say 6x EBITDA to closer to 3x EBITDA as the market prices in today what it believes will come to pass tomorrow.
Exhibit 93 In Understanding the Value Impact of
Consolidation, We Construct Two Simple Industries — One Fragmented...
Exhibit 94 ...The Other Consolidated, But in All Other Respects Identical
Source: Bernstein analysis and estimates. Source: Bernstein analysis and estimates.
Industry Structure
Number of Players # 10
Size of Market Mt 1000
Market Share Mt 100
New Projects Available Mt 400
Max Growth per Player Mt 40
Demand Growth Mt 200
Price US$/t 150
Elasticity of Supply # -2
Caital Intensity for Growth US$/t 200
Opex US$/t 40
Industry Structure
Number of Players # 4
Size of Market Mt 1000
Market Share Mt 250
New Projects Available Mt 400
Max Growth per Player Mt 100
Demand Growth Mt 200
Price US$/t 150
Elasticity of Supply # -2
Caital Intensity for Growth US$/t 200
Opex US$/t 40
Industry Consolidation Is a Necessary But Not a Sufficient Condition for a Commodity Industry to Generate Superior Returns; the Virtue of Commercial Awareness Must Be Added
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 73
Exhibit 95 For Each Industry, We Simply Look at the Consequences of Each Player Looking to Maximize Apparent Local Value; in This Case, "Apparent" Means Simply Looking at the Economics of Their Own Decision Making; It Is Easy to Show That in a Situation of Potential Oversupply, Local Value Maximization Leads to Global Value Destruction for a Fragmented Industry With No Market Impact Awareness...
Source: Bernstein analysis and estimates.
Under the second scenario (see Exhibit 96), the industry is still fragmented, but each participant has a more sophisticated understanding of value and recognizes that his individual project will impact price as well as volume. However, each project is simply too small to have much of an impact on price (in this case US$6/t) and so each player chooses to maximize volume growth just the same as before. While the prize that each player believes that it is aiming at is slightly lower than before (US$13 billion versus US$18 billion), it is still large enough to justify unconstrained volume growth. However, the net result is exactly the same as in the first example — massive oversupply, huge price cuts and value destruction, with the anticipation of such an outcome being presaged by an industry-wide de-rating.
Incumbent Player Total Industry
Player Ex Growth Total Industry Ex Growth
Volume Mt 100 Volume Mt 1000
Price US$/t 150 Price US$/t 150
Revenue US$bn 15 Revenue US$bn 150
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 4 Total Costs US$bn 40
EBITDA US$bn 11 EBITDA US$bn 110
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 66 EV US$bn 660
Player Inc Growth Total Industry Inc Growth
Original Volume Mt 100 Original Volume Mt 1000
Growth Volume Mt 40 Growth Volume Mt 400
Price US$/t 150 Price US$/t 90
Revenue US$bn 21 Revenue US$bn 126
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 6 Total Costs US$bn 56
EBITDA US$bn 15 EBITDA US$bn 70
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 92 EV US$bn 420
Less Capex US$bn 8 Less Capex US$bn 80
Total EV US$bn 84 Total EV US$bn 340
EV per Participant 34
Apparent Value Creation per Incumbent US$bn 18 Actual Value Creation per Incumbent US$bn -32
Fragmented Industry Excluding Awareness of Price Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
74 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 96 ...And This Global Value Destruction Will Occur Whether or Not Each Individual Player Is Aware of the Impact That Its Decisions Have on the Market Price; the Apparent Gain from Organic Growth in Each Scenario Is an Overwhelmingly Compelling Prize
Source: Bernstein analysis and estimates.
In the next scenario (see Exhibit 97), there is a consolidated market but each player is blithely unaware of the impact of its decisions on the market. Under this scenario, local value maximization is again realized through each player building in an unconstrained manner. It is clear that there is no difference in outcome between a fragmented and a consolidated industry, if the effect of price elasticity is neglected from the investment decision-making process (as seems to be the standard practice in mining). Industry consolidation is not an automatic good for shareholders — it is a positive only in so far as it yields a value outcome superior to that which would be achieved without consolidation. There is no point in trying to preserve an oligopoly if its preservation destroys rather than enhances value. In short, there is no such thing as a "good" industry structure if that structure does not return price power and superior value creation. And in mining, such an outcome cannot exist for as long as local value maximization is thought to be synonymous with ignoring the market impact of new volume. As so often in corporate life, the unanalyzed pursuit of "strategic" imperatives can easily lead to massive value destruction.
Incumbent Player Total Industry
Player Ex Growth Total Industry Ex Growth
Volume Mt 100 Volume Mt 1000
Price US$/t 150 Price US$/t 150
Revenue US$bn 15 Revenue US$bn 150
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 4 Total Costs US$bn 40
EBITDA US$bn 11 EBITDA US$bn 110
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 66 EV US$bn 660
Player Inc Growth Total Industry Inc Growth
Original Volume Mt 100 Original Volume Mt 1000
Growth Volume Mt 40 Growth Volume Mt 400
Price US$/t 144 Price US$/t 90
Revenue US$bn 20 Revenue US$bn 126
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 6 Total Costs US$bn 56
EBITDA US$bn 15 EBITDA US$bn 70
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 87 EV US$bn 420
Less Capex US$bn 8 Less Capex US$bn 80
Total EV US$bn 79 Total EV US$bn 340
EV per Participant 34
Apparent Value Creation per Incumbent US$bn 13 Actual Value Creation per Incumbent US$bn -32
Fragmented Industry Including Awareness of Price Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 75
Exhibit 97 In a Consolidated Industry, Where Each Player Acts as If It Was Unaware of the Impact That Its Decisions Have on the Market, There Is Absolutely No Difference in Outcome Versus a Fragmented Industry; Exactly the Same Volume Is Brought to the Market and Exactly the Same Value Destruction Results; Consolidation in Commodity Markets Is Necessary But Not Sufficient to Generate Superior Returns
Source: Bernstein analysis and estimates.
It is only when we come to the fourth and last scenario (see Exhibit 98) that a new feature emerges. Here again, local value maximization is sought; each agent individually builds as fast as it can commensurate with maximizing the value of its individual portfolio. However, because each player is independently aware of the fact that its decisions impact the market and because each has a large installed capital base, each comes to the conclusion that excessive growth will have a deleterious impact on value. Consequently, we see — for the first time — delivery of volume in a restrained manner. The effect of this at an industry-wide level is to generate value superior to that which each individual agent was targeting individually. Instead of massive oversupply, the "perfect" balance between new volume and price appreciation is met and industry-wide value is enhanced. Under this circumstance, the market should begin to re-rate rather than de-rate the sector, as awareness is formed about the emergent rationality of the industry and gets priced in.
We would like to stress no call is made here for concert or collusive action from the miners. Rather, a simple analysis of each agent's own local value maximization would be sufficient. In none of the foregoing analysis has anything other than local value maximization been the driving force behind decision making. There is no call to "look after" the market or to act to support price for other
Incumbent Player Total Industry
Player Ex Growth Total Industry Ex Growth
Volume Mt 250 Volume Mt 1000
Price US$/t 150 Price US$/t 150
Revenue US$bn 37.5 Revenue US$bn 150
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 10 Total Costs US$bn 40
EBITDA US$bn 27.5 EBITDA US$bn 110
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 165 EV US$bn 660
Player Inc Growth Total Industry Inc Growth
Original Volume Mt 250 Original Volume Mt 1000
Growth Volume Mt 100 Growth Volume Mt 400
Price US$/t 150 Price US$/t 90
Revenue US$bn 53 Revenue US$bn 126
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 14 Total Costs US$bn 56
EBITDA US$bn 39 EBITDA US$bn 70
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 231 EV US$bn 420
Less Capex US$bn 20 Less Capex US$bn 80
Total EV US$bn 211 Total EV US$bn 340
EV per Participant 85
Apparent Value Creation per Incumbent US$bn 46 Actual Value Creation per Incumbent US$bn -80
Consolidated Industry Excluding Awareness of Price Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
76 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
participants. However, one has to look only at a number of media articles on the topic of Rio Tinto's 360Mt expansion target to realize how pervasive the misunderstanding of this topic is. The common error is the mistaken belief that looking to preserve price is somehow putting global rather than local value objectives at the heart of a company's strategy. This is not the case. Rather, putting price preservation at the heart of strategy is something that every company ought to do, but it is something that becomes of critical importance when the company approaches the scale of the large diversified miners.
Exhibit 98 The Sufficient Condition for Superior Returns in Consolidated Markets Is an Awareness of the Impact That Volume Decisions Have on Price and the Fact That This Changes the Value of the Entire Portfolio, Not Just the Marginal Ton; Under Such Conditions, the Pursuit of Local Optimality Returns a Global Value Solution in Excess of What Is Optimal
Source: Bernstein analysis and estimates.
To summarize, industry consolidation is a necessary but not sufficient condition to enjoy superior returns in a commodity market (see Exhibit 99). Industry consolidation and an awareness of what local value maximization actually means are the necessary and sufficient conditions to enjoy pricing power and a superior return. The re-rating or de-rating that a sector should enjoy on the back of this is shown in Exhibit 100 through Exhibit 102. This shows how the rating varies by degree of market awareness and by degree of consolidation as well as certain
Incumbent Player Total Industry
Player Ex Growth Total Industry Ex Growth
Volume Mt 250 Volume Mt 1000
Price US$/t 150 Price US$/t 150
Revenue US$bn 37.5 Revenue US$bn 150
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 10 Total Costs US$bn 40
EBITDA US$bn 27.5 EBITDA US$bn 110
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 165 EV US$bn 660
Player Inc Growth Total Industry Inc Growth
Original Volume Mt 250 Original Volume Mt 1000
Growth Volume Mt 28 Growth Volume Mt 112
Price US$/t 157 Price US$/t 176
Revenue US$bn 44 Revenue US$bn 196
Unit Costs US$/t 40 Unit Costs US$/t 40
Total Costs US$bn 11 Total Costs US$bn 44
EBITDA US$bn 32 EBITDA US$bn 152
EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6
EV US$bn 194 EV US$bn 910
Less Capex US$bn 6 Less Capex US$bn 22
Total EV US$bn 189 Total EV US$bn 888
EV per Participant 222
Apparent Value Creation per Incumbent US$bn 24 Actual Value Creation per Incumbent US$bn 57
Consolidated Industry Including Awareness of Price Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 77
other variables.28 It is clear that iron ore is an industry where there ought to be pricing power and superior return potential.29 However, it is clear (see Exhibit 103) that currently the miners are not getting the benefit of an expectation of superior performance — quite the reverse. Given that the industry is sufficiently consolidated to enjoy a premium rating (and will remain so under most scenarios), the reason for its absence must relate to expectations of how the market incumbents understand their role as agents in charge of value maximization. As we look at the sector, the critical flaw we see is that of incorrect or too simplistic valuation methodologies as we lay out in Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value. The valuation policy error in mining companies results in poor capital discipline and an inability for globally superior outcomes to emerge from local value optimization. However, this is clearly beginning to change. Irrespective of the decision that is finally reached on Pilbara 360, the announcement from Rio Tinto's CEO Sam Walsh that is of key importance is that "the additional tonnage of 70 million tons was unlikely to result in a $20/t price decline."30 This is clearly an indication that the error in approach to value in at least one company is in the process of being rectified. It is for this reason that we continue to be optimistic about the future of the mining industry and continue to believe that emphasizing capital discipline is the key. It took, after all, 118 years to finally reduce Carthage; so perhaps a few years to see the emergence of real capital discipline is not too bad a result. We believe that a re-rating in this space is possibly precipitated by an appreciation that the defence of the iron ore price is a genuinely possible outcome. Exhibit 104 shows the impact of a re-rating around iron ore.
Exhibit 99 A Necessary Condition for Superior Price Performance Is That an Industry Be Consolidated, But It Is Not Sufficient; a Necessary and Sufficient Condition Is That There Is a Consolidated Industry Comprising Agents Who Understand the Price Impact of Their Own Decisions
Source: Bernstein analysis and estimates.
28 The consolidation could have been illustrated using the Herfindahl-Herschman Index, but the effect is much the same as if we just use the number of market participants to illustrate the point. 29 While iron ore may be a scarce commodity, cape size export ports and world-class bulk transport infrastructure are not, and these are the key elements in the iron ore business. 30 "Rio Said to Pursue $5 Billion Iron-Ore Project as Glut Looms," Bloomberg L.P., May 7, 2013.
Conditions for Superior Returns in
Commodity Industries
Consolidated Fragmented
Aware of Price Impact of Own
Decisions Unaware of Price
Impact of Own Decisions
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
78 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 100 There Are Really Only Three Players That Matter in Iron Ore (Four at a Push)...This Is a Sufficient Enough Degree of Consolidation That the Pursuit of Local Value Maximization Ought to Embed Capital Discipline in the Industry and Generate Superior Returns
Source: Bernstein analysis and estimates. Exhibit 101 While the Degree of Possible Oversupply Is Clearly an Important Variable...
Source: Bernstein analysis and estimates.
0
1
2
3
4
5
6
7
8
9
10
1 2 3 4 5 6 7 8 9 10
EV/E
BIT
DA
Mu
ltip
le o
f In
du
stry
Number of Players in Industry (i.e., Degree of Consolidation)
Multiple Uplift Available in Consolidated Commodity Markets Facing Oversupply
Including Market Aw areness Excluding Market Aw areness
0
1
2
3
4
5
6
7
8
9
10
1 2 3 4 5 6 7 8 9 10
EV/E
BIT
DA
Mu
ltip
le o
f In
du
stry
Number of Players in Industry (i.e., Degree of Consolidation)
Multiple Uplift Available in Consolidated Commodity Markets Facing Oversupply
100% Possible Oversupply 75% Possible Oversupply 50% Possible Oversupply
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 79
Exhibit 102 ...Even More Important Is the Elasticity of Supply, and Here Rio Tinto Itself Estimates This to Be About Three; Moreover, It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve and So Generates the Margins That an Industry Enjoys; the Higher the Elasticity, the Higher the Margin and Also the Lower the Degree of Consolidation Required to Generate Superior Returns
Source: Bernstein analysis and estimates. Exhibit 103 However, This Is Clearly Not What Is Being Discounted in the Stocks
Source: FactSet and Bernstein analysis.
0
2
4
6
8
10
12
1 2 3 4 5 6 7 8 9 10
EV/E
BIT
DA
Mu
ltip
le o
f In
du
stry
Number of Players in Industry (i.e., Degree of Consolidation)
Multiple Uplift Available in Consolidated Commodity Markets by Elasticity
e=-1 e=-2 e=-3
4.65.0 5.0
5.8
Vale Rio Anglo BHP
EV
/EB
ITD
A
Current Trading Multiples for Iron Ore Producers
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
80 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 104 We Are Beginning to See Greater Clarity from the Miners in Their Thinking on Capital Discipline and Price Preservation...All of Which Ought to Create a Greater Chance of Seeing Growth Restrained and the Industry Re-Rate, Generating Significant Upside for the Miners, Especially Rio Tinto and Vale
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Rio BHP Vale Anglo
Current EV (US$m) 106,719 186,765 106,936 57,856
Proportion of Price Target Attributable to Iron Ore (%) 42.8% 25.4% 40.3% 24.6%
Current EV/EBITDA (x) 4.97 5.79 4.58 5.04
Bernstein Iron Ore Multiple (x) 6.18 6.81 7.41 4.87
Multiple Possible on Iron Ore (x) 7.68 8.31 8.91 6.37
Value Impact (US$m) 11,075 10,447 8,725 4,378
Share Price Impact ($) 6.0 2.0 1.7 3.4
Current* Share Price ($) 42.28 27.19 13.37 19.56
Per Share Impact (%) 14% 7% 13% 18%
* July 12, 2013
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 81
What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?
The iron ore price (the most important source of mining EBITDA generation) has averaged US$135/t over the last five years and US$129/t average 1H:13 Aus FOB price — and yet mining company stocks have been among the worst performers of the year. This spring, we ran four fixed commodity price scenarios ("Grizzly," "Bear," "Bull" and "De Niro" in order of increasing aggressiveness) and examined how sensitive the value generated by our DCF (out to 2030) was to these different scenarios. The scenarios were bounded by the range of prices seen over the last five years (varying from the 25th percentile to the 80th percentile of the distribution), given the changes in the cost structure of the mining industry. We then selected the scenario that most closely resembled current market conditions and flexed the prices of the two most important commodities (from an EBITDA generation perspective for our coverage group) — copper and iron ore — in order to determine what the market was pricing in for these commodities. While the market still appeared comfortable about the longer-term copper prospects (baking into equity prices a 19% premium to the June 30, 2013 spot), the market was and remains deeply concerned about iron ore (baking in a 26% discount to the June 30, 2013 spot). Our analysis found that the market was pricing in iron ore at US$80/t and copper at US$8,000/t (see Exhibit 117 through Exhibit 121). We found — and continue to find — it difficult to believe that the pessimistic view on iron ore was tenable then or at the time of this Blackbook's publication.
We then asked ourselves "what Chinese growth outlook would be consistent with $80/t iron ore?" We calculated the Chinese GDP growth implied by consensus demand expectations in China, and then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore. We found that this was consistent with ~7% GDP growth — some 1.3% below the IMF trend forecast and 0.5% below the official government guidance — as well as significant new supply: ~550Mtpa of new iron ore capacity out to 2020 (a ~50% increase on the 1,123mt produced in 2012 globally), of which ~300Mtpa is surplus and hence price destructive.
Stress Testing DCF Valuations in Different Commodity Price Worlds At the time of this analysis in May, with the exception of the precious metals (most notably palladium), commodities were trading toward the bottom end of their historical five-year trading ranges (see Exhibit 105 through Exhibit 107), with the base metals and coal looking particularly depressed. We held all other variables constant (e.g., FX, capex and operating cost escalation) and focused on how our target price would vary with a changing long-term commodity price. Our "base case" factored in the most reasonable (in our view) temporal structure of the price line as a result of changing industry structures. Moreover, it is considered that, from a value perspective, the path taken to reach a long-term price is just as important as the overall price level.
Mining Equities Have Recently Been Baking in $80/t Iron Ore (~40% Below 1H:13 and Five-Year Average Prices) — This Is Consistent With Both ~7% Chinese GDP Growth (Below Both IMF Estimates and Chinese Targets) and ~550Mtpa New Iron Ore Capacity Out to 2020
Commodities Were Trading at Multi-Year Lows, So We Ran Four Commodity Price Scenarios to Examine the Sensitivity of Our DCF Valuation
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
82 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 105 Summary of Historical Commodity Prices
Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 106 Except for Some Precious Metals, Current Spot Prices of the Key Commodities for
the Miners in Our Coverage Were Substantially Below Their Most Recent Five-Year Mean
Current Spot Price — Percentile of Five-Year Price Distribution
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 83
Exhibit 107 Higher Average Year-to-Date Prices Reflect Higher Commodity Prices in the Beginning of the First Quarter of the Year
Source: Bloomberg L.P. and Bernstein estimates and analysis.
In order to stress test this case and our stock recommendations, we took four price scenarios with fixed commodity prices (we label them "Grizzly," "Bear," "Bull" and "De Niro" in order of increasing aggressiveness) and examined how the price target generated by our DCF31 would vary with each of them.
A brief overview of our scenarios for a selection of the most important commodities is given in Exhibit 108 and Exhibit 109 and explained in detail in "Appendix 1: From Grizzly to DeNiro." In order to understand what these price scenarios represent, one should contrast them against the range of prices that we have seen over the last five years. Exhibit 110 and Exhibit 111 show the distributions of historical copper and iron ore prices (for the other major commodities, these are shown in Appendix 1). All the price scenarios we have chosen are bounded by the range of prices seen over the last five years (varying from the 25th percentile of the distribution, on average, for the "Grizzly" scenario to the 80th percentile for the "De Niro" scenario). Given the change in the cost structure of the mining industry over the last decade, if a price distribution further back than the last five years is taken, we believe it would lead to an overly negative assessment of what the future could hold. A copper price of US$6,500/t is markedly different in terms of impact now than it would have been in 2003.
31 Please note that we provide two valuation metrics: a one-year target price based upon our regression model and an intrinsic value generated by a DCF out to 2030. It is the latter that was used throughout this analysis. For more details on our valuation methodology, please see the "Valuation and Risks"' chapter.
82%76%
67%
58% 58% 58%
48%
34% 32% 30%
22%16%
Average YTD Price — Percentile of Five-Year Price Distribution
Our Four Scenarios Ranged from 25th to 80th Percentile of Historical Price Ranges
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
84 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 108 Summary of Commodity Price Scenarios vs. Historical Distribution — Real Terms
Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 109 Summary of Commodity Price Scenarios vs. Current Model — Real Terms
Source: Bloomberg L.P. and Bernstein estimates and analysis.
Real LT Commodity prices Percentile of 5Yr Distribution
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 85
Exhibit 110 The "Fat Tail" in the Copper Price Distribution Relates to the Period Immediately After the Financial Crisis; We Would Interpret the Shape of the Price Distribution as Telling Us Something About the True Costs of Marginal Production
Source: Bloomberg L.P. and Bernstein analysis. Exhibit 111 The Shift in Pricing Regime from the Benchmark to the Spot Market as Well as the
Deterioration in Chinese Domestic Mined Iron Ore Grades Result in a Markedly Different Price Distribution for Iron Ore Than for Copper
Source: Bloomberg L.P. and Bernstein analysis.
Please refer to Appendix 1 of this Blackbook for other commodity price distributions.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
86 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 112 gives the high-level summary of how each of our stocks would respond to the valuation scenarios used. Vale is tied with Rio Tinto for the lowest downside in our Grizzly scenario and shows the highest upside in Bull and De Niro scenarios; the company with the greatest downside across all scenarios is Glencore. This was largely a consequence of our then-negative view on the long-term iron ore price32 (albeit with an interesting trajectory down to this level) versus copper and the high level of operational gearing in Vale on the back of its significant capex program (see Exhibit 113 through Exhibit 116).
For each of our coverage companies, we detail the pro forma financials for each of the scenarios, giving a summary income statement, balance sheet and cash flow. The full details of these scenarios are given in Appendix 1, but for the sake of brevity only the headline details have been included here. We also show how the SOTP DCF value varies by scenario for our coverage, as we base our target prices on EV of the aggregate assets held by each company.
Exhibit 112 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our
Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our DCF Value (as a Result of Our Bullish View on Copper in the Medium Term)
Source: Corporate reports and Bernstein estimates and analysis.
Exhibit 113 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest
Exhibit 114 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
32 Please see the "Our Iron Ore Price Forecast" chapter for our updated iron ore price forecast.
TP / NPV per Share vs. Current Model
Current Model
Grizzly Bear Bull De Niro Grizzly Bear Bull De Niro
Vale Showed the Greatest Upside Across All Scenarios; Glencore the Lowest
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 87
Exhibit 115 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices
Exhibit 116 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price
Source: Corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
What Iron Ore Price Is Being Baked In? In addition to stress testing our equity valuations, we also aimed to find out what iron ore and copper prices the market was pricing into the miner's equity prices. From the four scenarios run earlier, we selected the scenario that most closely resembled current market conditions ("Bear") and flexed the prices of the two most important commodities (from an EBITDA generation perspective for our coverage group) — copper and iron ore — in order to determine what the market was pricing in for those commodities. We found that mining equity valuations were baking in $80/t iron ore, a ~40% discount to year-to-date and five-year average prices (see Exhibit 117 through Exhibit 121).
Exhibit 117 As This Scenario Analysis Run in Late May Shows, Rio Tinto Offers the Most
Attractive Balance Between Risk and Potential Upside
Note: Price current as of May 21, 2013, when this analysis was run.
Source: Corporate reports and Bernstein estimates and analysis.
102%
31%
17%8%
-5%
Vale Rio Tinto Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — Bull
186%
79%
57%50%
31%
Vale Rio Tinto Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — De Niro
Rio Tinto (GBP), Current Price = £28.78 Iron Ore (US$/t)
TP 70 90 110 130
6,500 20.76 35.94 50.73 65.28
7,500 22.57 37.73 52.50 67.03
8,500 24.38 39.52 54.28 68.79
9,500 26.21 41.33 56.06 70.55Co
pp
er
(US
$/t)
Flexing Our Price Assumptions Revealed That Iron Ore Was Priced in at a ~40% Discount Year-to-Date and Versus the Five-Year Average
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
88 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 118 The Presence of Oil in the BHP Portfolio Adds a Third Critical Variable That Renders It Harder to Unpack the Implied Iron Ore and Copper Price from the Market Price; Oil Remains a Key Diversifier for the Highest Quality Stock in Our Coverage
Note: Price current as of May 21, 2013, when this analysis was run.
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 119 While the Market Price of Vale Might Seem to Imply a Higher Iron Ore Price, We
Believe That This Is a Function of Us Modeling a Higher Tax Rate for Vale Than the Market
Note: Price current as of May 21, 2013, when this analysis was run.
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 120 For Anglo, the Issues of Both Platinum and South African Political Risks Add a Layer
of Complexity to the Investment Proposition
Note: Price current as of May 21, 2013, when this analysis was run.
Source: Corporate reports and Bernstein estimates and analysis.
BHP Billiton (GBP), Current Price = £19.12 Iron Ore (US$/t)
TP 70 90 110 130
6,500 11.22 15.36 19.50 23.66
7,500 12.57 16.71 20.87 25.03
8,500 13.93 18.07 22.23 26.40
9,500 15.29 19.44 23.60 27.77Co
pp
er
(US
$/t)
Vale (BRL), Current Price = BRL32.1 Iron Ore (US$/t)
TP 70 90 110 130
6,500 13.14 33.15 53.15 73.14
7,500 14.37 34.38 54.38 74.37
8,500 15.40 35.41 55.41 75.41
9,500 16.39 36.41 56.41 76.40Co
pp
er
(US
$/t
)
Anglo (GBP), Current Price = £15.49 Iron Ore (US$/t)
TP 70 90 110 130
6,500 12.32 15.05 17.75 20.43
7,500 14.33 17.03 19.69 22.34
8,500 16.33 18.99 21.62 24.24
9,500 18.31 20.93 23.54 26.13Co
pp
er
(US
$/t
)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 89
Exhibit 121 Note That for Glencore Xstrata, We Run the Analysis on Our Combined Pro Forma Glencore Xstrata; Copper Is the Single Greatest Sensitivity for Glencore Xstrata; to See Outperformance in This "Not Yet Institutional Quality" (from a Governance and Reporting Perspective) Stock, It Is Necessary to Believe in a Strong Copper Price
Note: Price current as of May 21, 2013, when this analysis was run.
Source: Corporate reports and Bernstein estimates and analysis.
What GDP Outlook Is Consistent With $80/t Iron Ore? Given that the equity market appeared to be discounting a US$80/t iron ore world, we then asked ourselves what the world needs to look like for these prices to come about. Specifically, "what growth outlook for the world's largest commodity consumer, China, would be consistent with $80/t iron ore?"
We calculated the Chinese GDP growth implied by consensus demand expectations in China, and then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore.
We found that US$80/t iron ore is consistent with a ~7% GDP growth — some 1.3% below the IMF trend forecast and 0.5% below the official government guidance. Our framework demonstrates that consensus expectations therefore imply ~550Mtpa of new iron ore capacity out to 2020, of which ~300Mtpa comes in excess of the requirement and is price destructive. Yet more evidence of the pressing need for a more restrained approach to organic growth by the majors!
Because China is the largest consumer of iron ore globally, and is only one-third of the way through the most rapid industrialization ever seen globally, the most important demand-side question for mining stocks is "How does one estimate future steel intensity in China?" Exhibit 122 shows the history of Chinese steel intensity as well as a plausible forward-looking trend line. The problem, however, is that a lot of such lines might look equally convincing and yet each would represent a substantially different future for the Chinese steel. There must be more to determining the future of this key variable than the aesthetics of different curves, and yet so often this still is the basis of demand projections.
Simply examining steel intensity curves does not provide enough information to decide if China is "on track" relative to other industrialized countries. Instead, we require a means to link steel intensity today with steel intensity tomorrow. We find such a link in the constant relationship between an economy's steel capital stock and its overall economic activity (R-squared = 65%) — see Exhibit 122 and Exhibit 123. Although steel intensity has varied from industrialization to industrialization (the U.S. was twice as intensive as France, and China is again some 50% more intensive than the U.S.), it has always been the embedding of steel in physical infrastructure that enabled a transition from an agricultural economy to one dominated by services. We therefore use this relationship to forecast how China's capital stock is likely to evolve.
In order to ground our analysis, we take the trajectory of U.S. steel capital stock formation relative to output as a benchmark development path for any
Glen Xta (GBP), Current Glencore Price = £3.37 Iron Ore (US$/t)
Implied Glen-Xta NPV 70 90 110 130
6,500 2.49 2.49 2.49 2.49
7,500 3.05 3.05 3.05 3.05
8,500 3.61 3.61 3.61 3.61
9,500 4.17 4.17 4.17 4.17Co
pp
er
(US
$/t)
What Is the Implicit Chinese GDP Assumption Consistent With US$80/t Iron Ore? ~7% — Some 1.3% Below the IMF Forecast and 0.5% Below the Official Government Guidance
U.S. Industrialization Provides a Canonical Pattern from Which to Derive the Multiplier Between Economic Growth and Metal Demand Growth for China
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
90 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
economy (see Exhibit 124 and Exhibit 125). These exhibits show the increasing levels of productivity in the U.S. — that is, incrementally higher levels of output being generated from each additional unit of input. In fact, the relationship between steel and output measures the evolution of productivity within an economy. However, increases in productivity have only become evident after the U.S. reached a steel stock of 10,000kg/capita versus ~4,500kg/capita in China today. Although China remains more steel intensive than the U.S., efforts by the Chinese government to diversify its economy away from reliance upon cheap outsourced manufacturing result in a trajectory that is directionally converging with that of the U.S. (see Exhibit 126).
In our annual forecasts, we extend the pattern of convergence seen historically with ever-increasing productivity in the Chinese economy. While a case could be made that China should be more metal intensive than the U.S. (due to its population density and intensity of housing and the associated infrastructure), we believe that the U.S. represents a sensible base case. Moreover, a more metal-intensive China would only defer rather than obviate the need for the transition that was experienced by the U.S. In any event, this adds only upside risk to our steel forecast.
Exhibit 122 A Central Question for an Iron Ore Price Forecast Is How to Project China's Steel Intensity Into the Future? Looking at Historical Steel Intensity Alone Is Insufficient
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45 50
Ste
el In
ten
sity
(kg
/$'0
00
GD
P)
Output — GDP/Capita — $'000 (Real 2005 PPP)
Trend Steel Intensity Evolution
France USA China Japan South Korea China Actual
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 91
Exhibit 123 The Relationship Between Economies' Capital Stock and Level of Output (R-squared = 65%) Provides the Missing Data Point for Forecasting How the Capital Stock to Output Ratio Will Evolve in China
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
R² = 0.6532
0
10
20
30
40
50
60
70
80
0 5,000 10,000 15,000 20,000
Ou
tpu
t —
GD
P/C
apit
a $'
000
(Rea
l 20
05 P
PP
$)
Capital Stock — kg Steel/Capita (including depreciation)
Capital Stock to Output by Country
China 2012
UK
Kuwait
Germany
Taiwan
USA
China 2020
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
92 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 124 We Use the U.S. Development Path as a Basis for China's Industrialization Forecast
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 125 The Benefit of Increasing Productivity Over Time Is Easier to See If the Axes in the
Earlier Graph Are Inverted; in the U.S., the Real Advances Took Place Only Once the Capital Stock Had Been Built and Urbanization With the Accompanying Labor Force Transformation Had Been Completed
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
0
5
10
15
20
25
30
35
40
45
50
- 2,000 4,000 6,000 8,000 10,000 12,000 14,000
US
GD
P/C
ap
ita
$'0
00
(Re
al 2
005
PP
P)
Kg of Steel in Economy per Capita
US Steel Capital Stock
R² = 0.9595
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0 5 10 15 20 25 30 35 40 45 50
Kg
of
Ste
el in
Eco
no
my
pe
r C
apit
a
US GDP/Capita $'000 (Real 2005 PPP)
US Steel Capital Stock (Inverted Axes)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 93
Exhibit 126 China Is More Steel Intensive Than the U.S., But Shows Convergence
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
We use this convergence with the U.S. levels of steel productivity to calculate
a consistent steel intensity for any level of overall economic activity. For illustration only, we show a steel-intensity path that would enable China to reach roughly current levels of the U.S. output by 2020. It would require China to post GDP growth rates of ~20% for the next seven years, and, unsurprisingly, such a growth pattern would show a very rapid convergence to the U.S. (see Exhibit 127). In order to support this level of economic activity and to transition its large rural population into urban centers, China needs to embed steel. The total amount of new steel that would be required is supplied by the relationship between output and capital stock. Assuming this amount of new steel would have to be generated over the next seven years, we can then calculate the run rate of steel production (and hence steel intensity) for the economy as a whole. This, thus, links current and future steel intensities. Clearly, under the implausible/illustrative scenario of 20% GDP growth, there must be a correspondingly large increase in steel intensity (see Exhibit 128).
A further sensitivity should be flagged. The database of industrial data we have assembled for all countries currently reporting to the WSA, dating back to 1900, enables us to choose any historical paradigm of steel productivity as the basis for Chinese development. Consequently, we could show China converging to Japanese or German levels of steel productivity.
However, given the importance of the U.S. economy to global demand and global trade, we believe that the development patterns of countries other than the U.S. are less representative for China — the soon-to-be largest economy in the world.
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0 5 10 15 20 25 30 35 40 45 50
Kg
of
Ste
el p
er C
apit
a
GDP/Capita (Real 2005 PPP)
Steel Capital Stock — China vs. USA
USA China
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
94 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 127 For Illustration Only, We Show a Development Path That Would Enable China to Converge to the Current U.S. GDP Levels by 2020; It Would Require a 20% Growth Rate Each Year
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 128 Using Convergence as the "Missing" Data Enables Us to Calculate the Transition
Path Between Steel Intensities at Two Different Points in Time and Derive the Relationship Between Overall Economic and Metal Growth Rates
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0 5 10 15 20 25 30 35 40 45 50
Kg
of S
teel
per
Cap
ita
GDP/Capita (Real 2005 PPP)
Steel Capital Stock — China Forecast vs. USA
USA China Actual China Forecast
0
10
20
30
40
50
60
70
80
- 5 10 15 20 25 30 35 40 45 50 55
Ste
el In
ten
sity
—K
g/$
'000
Output — GDP per Capita $'000 (Real 2005 PPP)
Steel Intensity Forecast
China Actual China Forecast Original SCB Trend
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 95
At our September 2012 launch, we used the April 2012 IMF global GDP forecasts as the basis for our demand calculations. Since our launch, the Chinese economy has unquestionably slowed down. In April 2012, the IMF was forecasting an average annual growth of 8.7% over 2014-2016. This estimate was revised down to an average 8.5% over 2014-2016 as of 2Q:13 (see Exhibit 129 and Exhibit 130). Furthermore, at the time of our launch, we were happy to calibrate China's future steel intensity forecast looking at the transition that would be required out to 2020 and beyond (as shown in Exhibit 123). However, in the face of frequent growth revisions, looking at only longer-term trends is insufficient. Instead, an explicit year-by-year analysis and greater rigor is required to explicitly link economic activity to steel use.
Taking the April 2013 IMF GDP deck as a starting point, the path to convergence for the U.S. is shown in Exhibit 131. The resultant steel intensity is then shown in Exhibit 132. Evidently, steel intensity shows a marked departure from the too simple trend line we had drawn previously (yet again, this highlights the dangers, if any further highlighting was needed, of too simple curve plotting in commodity markets analysis). However, if the acceleration of the Chinese economy back up to a trend rate of 8.5% is believed — and this is the IMF's current base case for 2016-2019 — steel production will accelerate accordingly (see Exhibit 132). Given the relative immaturity of the Chinese economy, it is still at a very metal-intensive phase of its development. Consequently, accelerating growth requires acceleration in raw material consumption. The rising productivity of the U.S. economy, which enabled it to grow output with no corresponding increase in metal intensity, occurred only after its capital stock had been built. China is still some way from this level (~4,500kg of steel per capital in China versus 12,000kg in the U.S.). Consequently, in order to accelerate growth without accelerating metal intensity, China's economy would have to experience an unprecedented increase in metal productivity. In any event, redrawing the trend steel intensity line that incorporates these more rigorously derived data points results in the structure shown in Exhibit 133. Having derived the steel intensity for the Chinese economy, it is possible to calculate the total steel production and volume growth (see Exhibit 134). More importantly, the ratio between metal growth and economic growth can be calculated from the first principles (see Exhibit 135). Instead of asserting the relationship between metal growth and economic growth, which simply masks rather than solves the problem, this approach enables it to be derived consistently.
The IMF's GDP Deck Implies That China's Steel Intensity Would Need to Accelerate Again
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
96 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 129 Since the Time of Our Original Analysis, China's Economy Has Slowed Down Markedly
Exhibit 130 In the Space of a Year, the IMF Revised Its Growth Projections Down by 0.5% on Average Over 2014-16)
Source: IMF and Bernstein analysis. Source: IMF and Bernstein analysis. Exhibit 131 Taking the IMF's April 2013 Growth Forecast, We Estimate the Following
Convergence to the U.S. Steel Stock Levels for China…
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 97
Exhibit 132 …Which Would Imply the Following Trajectory for Steel Intensity; the Previous Steel Intensity Trend Line Was Extrapolated in Order to Keep the Long-Term Evolution of China Consistent With the U.S.; Now, We Calculate It Year-by-Year
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 133 We Refine Our Previous Steel Intensity Trend Line…
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
0
10
20
30
40
50
60
70
80
- 5 10 15 20 25 30 35 40 45 50 55
Ste
el In
ten
sit
y —
Kg
/$'0
00
Output — GDP per Capita $'000 (Real 2005 PPP)
Steel Intensity Forecast
China Actual China Forecast Original SCB Trend
0
10
20
30
40
50
60
70
80
- 5 10 15 20 25 30 35 40 45 50 55
Ste
el I
nte
ns
ity
—K
g/$
'000
Output — GDP per Capita $'000 (Real 2005 PPP)
Steel Intensity Forecast
China Actual China Forecast (IMF GDP) Original SCB Trend Rev ised IMF/SCB Trend
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
98 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 134 …And Generate the Following Steel Production Forecast; It Shows Declines in Crude Steel Growth Seen Post 2020; in Our View, It Is Unavoidable Under All Chinese Growth Scenarios
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 135 Instead of Having to Guess the Relationship Between Metal Growth and Overall
Economic Growth, We Can Derive It from the First Principles
Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.
Chinese Steel Growth to GDP Growth Multiplier — IMF Scenario
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 99
Inverting the earlier calculation, we can unpack the assumptions consistent with a particular view on China's expected GDP and, by extension, steel demand growth. We start with consensus expectations for steel production and price. We take the steel forecast of CRU — a widely used high-quality data provider to the mining industry (see Exhibit 136). It is one of the most frequently used forecasts in the industry.
We derive the GDP deck consistent with the steel demand evolution assumed by CRU (see Exhibit 137). The Chinese GDP growth rate that is consistent with consensus expectations for steel demand is therefore ~7% — some 1.3% below the IMF trend figure and 0.5% less than the official government target. Historically, the target has represented a floor to the achieved growth rather than an average expectation and is a figure that is not subject to frequent revisions, which might cause loss of face within the government. Moreover, the target aims to balance the competing objectives of avoiding inflation, while ensuring employment (clearly, something easier in theory than practice). Consequently, we believe that consensus expectations embed a bear case, rather than an equally weighted upside to downside scenario.
Nevertheless, the consequences that such steel demand path would have are evident in Exhibit 138 and Exhibit 139. The consensus expectation sees the decline in steel intensity seen over the last four years continuing into the future. We would like to stress that Chinese steel intensity peaked in 2009 and has been declining every year since — what we are witnessing in China and mining more broadly today is hardly a recent phenomenon.33 In any event, the implied consensus view on Chinese growth sees a metal growth multiplier tracking below one (see Exhibit 139).
We compare the IMF and consensus views on Chinese economic development and their implications for steel growth in Exhibit 140. The exhibit summarizes the sensitivity of the mining sector's value, which represents geared exposure to the iron ore price. It, in turn, represents geared exposure to Chinese steel demand, which then depends upon the overall growth of China's economy! Consequently, looking through these permutations and taking a view on how this will play out over the longer term is not an easy task, and we would interpret this as yet another reason for the miners to take a restrained approach to organic growth. The only ton that cannot be sold tomorrow is the one you sold today. Any doubt about the "natural" home for a ton suggests that it is better to leave it in the ground and wait till certainty is forthcoming. In 2010, Sam Walsh borrowed from Pliny the Elder and noted that "the only certainty is that nothing is certain."34 If 2010 was uncertain, then China is even more so today. We would therefore like to borrow a phrase from Pliny the Younger and note that one should "never do a thing concerning the rectitude of which you are in doubt." There is considerable doubt about the virtue of further iron ore growth.
33 And it is important to stress: peak steel intensity is only the year during which steel is embedded in the capital stock at the fastest rate. It does not mean that in absolute terms, the amount of steel embedded will decline. 34 MMC, December 2, 2010.
The Chinese GDP Growth Rate Consistent With Consensus (e) Steel Demand Is ~7%, or 1.3% Below the IMF and 0.5% Less Than the Official Government Target
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
100 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 136 Consensus Expectations for Chinese Steel Production Are Significantly Different from Those Implied by the IMF's GDP Deck
Source: CRU and Bernstein estimates and analysis. Exhibit 137 We Back Out the Scenario for China's GDP Evolution That Is Implied in the Steel
Growth Consensus; It Sees the Annual GDP Trend Growth of 7.1 vs. 8.5% Implied in the IMF Deck
Source: IMF, CRU and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 101
Exhibit 138 The Steel Intensity That Such a Trajectory Would Imply Is Markedly Lower Than the Scenario Consistent With the IMF GDP Deck
Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 139 Instead of a Recovery in the Metal to GDP Multiplier, We Would Continue to See
Declines
Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis.
0
10
20
30
40
50
60
70
80
- 5 10 15 20 25 30 35 40 45 50 55
Ste
el I
nte
nsi
ty -
Kg
/'$00
0
Output - GDP per Capita '$000 (Real 2005 PPP)
Implied ConsensusSteel Intensity Forecast
China Actual China Forecast Original SCB Trend
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
200
5A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
Chinese Steel Growth to GDP Growth Multiplier
IMF Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
102 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 140 Unsurprisingly, Given the Relative Immaturity of the Chinese Economy, Its Steel Growth Is Highly Leveraged (~4 to 1) to the Overall Economic Growth Rate
Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis.
In addition to having visibility into the consensus view on steel production in China, we have some understanding of the Chinese cost structure (see Exhibit 141 and Exhibit 142). Moreover, we know the consensus price expectation (see Exhibit 143) — this price line shows the decline to ~US$80/t (real) that is currently discounted in equity valuations, albeit US$80/t appears to be factored into equity prices with no "fade" down. Rather, the market is indicating that any excess profits generated by the miners above this long-run equilibrium will simply be squandered on mistaken organic growth projects. Given this, we derive a consistent supply side response that is implicitly factored into the commodity price expectations (see Exhibit 144 through Exhibit 146). Over the next seven years, the market "expects" an increase in supply of ~550Mtpa delivered into the Chinese steel market, of which ~300Mtpa will be superfluous to the Chinese demand growth needs. The 300Mtpa of production will have to compete for space in the market through the displacement of high-cost incumbent tons. Consequently, we expect this to yield a decline in price along the trajectory assumed by consensus.
China GDP and Steel Growth IMF vs. Implied Consensus
GDP Steel
$80/t Iron Ore at ~7% Chinese GDP Is Only Maintained If ~550Mtpa of New Supply (330Mtpa Surplus) Also Come Onstream
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 103
Exhibit 141 Using Our Knowledge of the Structure of the Chinese Domestic Mining Industry, We Can Generate an Exemplar Cost Curve for the Industry as a Whole
Source: WSA, IMF, CRU, AME and Bernstein estimates and analysis. Exhibit 142 This Enables Us to Understand the Elasticity of Supply
Source: WSA, IMF, CRU, AME and Bernstein estimates and analysis.
0
20
40
60
80
100
120
140
160
180
200
0 200 400 600 800 1000 1200 1400 1600
Op
ex
(US
$/t)
Cumulative Iron Ore Production (mt)
Idealized Iron Ore Supply
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
Elas
tici
ty
Cumulative Iron Ore Production (mt)
Idealized Iron Ore Elasticity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
104 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 143 Combining It With Consensus Steel Demand Expectations Yields Consensus Price Expectations…
Exhibit 144 …And Enables Us to Derive Consensus Supply Expectations
Source: Bloomberg L.P. Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Exhibit 145 Assuming the Consensus GDP Deck, Two-
Thirds of the Supply Expected to Come Onstream Is Not Needed…
Exhibit 146 …With Price Destructive Oversupply of ~300Mt by 2020
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
-
20
40
60
80
100
120
140
160
180
No
min
al (U
S$/
t)
Consensus Iron Ore Price Forecast
0
20
40
60
80
100
120
140
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
Su
pp
ly G
row
th o
f Iro
n O
re (M
t)
Implied Consensus Supply Expectations
Supply Required by Chinese Demand
Price Destructiv e Suppy in Excess of Demand
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
% of Excess Price Destructive Supply
Required Surplus
235
299
Total Iron Ore Supply by 2020 — Mt
Required Surplus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 105
China Growth Scenarios
Using the model we built to back out Chinese GDP and supply side response that is consistent with consensus demand and price expectations, we can determine a consistent forecast for iron ore prices and steel demand growth across any GDP scenario for China. Our "bull" scenario takes the IMF's forecast from just over a year ago (April 2012), which sees the Chinese economy accelerate back up to a trend rate of 8.5%. This would absorb the ~300Mtpa of price destructive oversupply implied in the consensus scenario (mentioned earlier), resulting in a tight market in which the miners do not compete on volume and in which a trend price of US$140/t could eventuate out to 2020. A "middle" scenario takes the Chinese target of 7.5% GDP growth as a floor (despite the fact that Chinese GDP has a tendency to miss on the upside). This yields a price of US$120/t out to 2020. Our China "bear" scenario sees GDP decline to 6% and no supply coming out of the market (a true worst-case scenario). This results in an iron ore price testing the US$65/t mark over a multi-year period.35
The model we discussed in the previous chapter can be used to derive a self-consistent GDP, steel demand, supply reaction and iron ore price forecast for any particular GDP or price assumption. For example, using the consensus supply we derived in that, we assume that the market has a good grip on the supply side but sees demand (i.e., Chinese GDP) as uncertain, hence we can flex GDP growth assumptions and derive a new price forecast. Exhibit 147 summarizes forecast prices under the IMF GDP deck.36 The higher steel intensity of this GDP view against a fixed supply side response leads to higher prices. Consequently, under this scenario, prices are expected to average ~US$140/t out to 2020 — a far cry from the US$80/t that equities are currently discounting.
If it became clear that this price path were indeed to be realized, one could reasonably expect a supply side response, whereby more new growth projects would be approved. As we have stressed before, there is an inverse relationship between apparent and actual attractiveness of commodity markets. The more negative consensus expectations are, the less capital flows into the sector and thus the lower future volumes will be. Unsurprisingly, the lower future volumes in a commodity industry result in a higher future price. Conversely, the stronger consensus expectations are, the higher the capital inflows and the higher the volume growth, leading to a weaker counterfactual price environment. Once again, it is only by understanding the implications of counter-consensual and counter-cyclical behavior that sustainable value creation can be delivered in this industry. For the purposes of this analysis, we assume that the long lead time in new project development and approval will lead to the supply side changing more slowly than the demand side. Consequently, at least to a first order approximation, the resulting price line will be robust.
35 Our iron ore price forecast is discussed in greater detail in the "Our Iron Ore Price Forecast" chapter. 36 Ibid.
We Construct Bull, Middle, and Bear Scenarios for Chinese GDP and Supply Side Response to Bound the Possibilities for Iron Ore Prices Out to 2020 ($140/t Down to $65/t)
We Can Determine a Consistent Forecast for Iron Ore Prices and Steel Demand Growth Across Any GDP Scenario for China
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
106 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 147 We Can Use the Implied Consensus Supply Expectations to Derive Consensus Steel Demand Scenario and Iron Ore Price Line for Any Chinese GDP Deck Assumptions; Steel Intensity Is Geared to China's GDP and the Iron Ore Price Is Geared to Steel Intensity; Risk and Uncertainty Pervades This Sector
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
We construct three scenarios to stress test the iron ore price against different future development paths for China and against the supply view discussed earlier and in the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter (see Exhibit 148 and Exhibit 149).
Scenario 1 — Back to the future: The IMF April 2012 view of GDP results in prices supported at an average ~US$140/t out to 2020. If the acceleration of the Chinese economy back up to a trend rate of 8.5% is to be believed, and this is the IMF base case, then that would require a corresponding acceleration in steel production. Given the relative immaturity of the Chinese economy, it is still at a very metal-intensive phase of its development. Consequently, accelerating growth requires acceleration in raw material consumption. The increasing productivity of the U.S. economy, which enabled it to grow output with no corresponding increase in metal intensity, occurred only post the conclusion of the capital stock build. China is still some way from this level (~4,500kg/capita versus 12,000kg/capita). For the Chinese economy to accelerate its growth without accelerating metal intensity, at this stage in its economic life, there would have to be an unprecedented increase in metal productivity. If we regard the implied supply expectations as being representative of what is most likely to eventuate in reality, such acceleration in steel intensity would mean that all new tons would be able to find a home. Instead of there being ~300Mtpa of price destructive oversupply, all of the ~550Mtpa of new supply would be able to satisfy true demand. So instead of having the emergence of the miners competing on volume, the market would stay tight and a trend price of US$140/t could eventuate out to 2020.
Scenario 2 — Ending with a whimper: Taking the Chinese base case of 7.5% GDP growth as a floor yields an average price of US$120/t out to 2020. There are, we believe, three factors that the Chinese government has at the heart of its industrial policy, all of which are focused on ensuring the maintenance of political stability. They are inflation (particularly for food), unemployment and corruption. The official GDP target for China is 7.5%. Historically, the official GDP target has been a floor rather than an average expectation, and is not subject to
Consensus @ IMF GDP 33 42 46 76 81 147 168 128 138 142 152 160 157 150 129 114
0
20
40
60
80
100
120
140
160
180
Iro
n O
re —
FO
B A
us
Iron Ore Price Scenarios
Chinese GDP Growth Scenarios: 8.5%, 7.5% and 6% (and No Supply Reduction) Generate 2020 Iron Ore Prices of $140/t, $120/t and $65/t, Respectively
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 107
frequent revisions that might cause loss of face within the government (remember that during the period of an official 8% growth target, China averaged 10.6%). Moreover, this growth target is set to try and balance the competing objectives of avoiding inflation, while ensuring employment (clearly something easier in theory than in practice). The previous notwithstanding, this scenario shows the impact of a gradual fade to the official target levels of growth would actually be on China. The implication of this would be US$120/t iron ore out to 2020.
Scenario 3 — It's all over now baby blue: GDP revisions continue their descent down to 6%, and there is no curtailment of supply (a true bear case), resulting in prices testing US$65/t. This is for the true China bears. It shows the impact of continued downward revisions to Chinese growth and no curtailment of supply. It is worth noting that in a world of 6% growth, we believe that not only would many of the high-cost producers ($120-180/t) close, but a number of projects in the 550Mtpa of new supply would be canceled. However, a thorough bear scenario should leave all this supply intact and that is what we do. Under this world, the iron ore price could test the US$65/t level, just as it did in 2008; however, unlike 2008, we do not believe that the miners would be saved by the Chinese government stimulus. Consequently, a protracted period of low-price levels would persist. Troublingly, the iron ore majors would still be cash positive at these price levels, and so cutting existing tonnage as a form of self-help might still not occur. Again, the warning for us is clear...far better not to install capacity today than have to deal with trying to remove it if the world turns against you. (Not that it is a particularly good parallel, but Platinum used to be Anglo American's most profitable division; however, the good times have passed and trying to fix the industry through removing ounces and labor is proving virtually impossible.)
For each scenario, we provide the following data points: Overall Chinese GDP deck and how it differs from economic growth
expectations implied in consensus (see Exhibit 150, Exhibit 151, Exhibit 158, Exhibit 159, Exhibit 166 and Exhibit 167);
Steel intensity and production that would result in for a given GDP deck expectation and assuming that China will converge to U.S. productivity levels (see Exhibit 152, Exhibit 153, Exhibit 160, Exhibit 161, Exhibit 168 and Exhibit 169);
Steel growth as well as steel-growth-to-economic growth multiple (see Exhibit 154, Exhibit 155, Exhibit 162, Exhibit 163, Exhibit 170 and Exhibit 171);
Proportion of new supply that will be required to satisfy growing demand as well as the proportion that comes in excess of it, and would hence act to destroy the price (see Exhibit 156, Exhibit 164 and Exhibit 172); and
Resultant iron ore price forecast (see Exhibit 157, Exhibit 165 and Exhibit 173).
Please see "Appendix 2: Growth Scenarios" for more detail on the scenarios in this chapter and also the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
108 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 148 In Order to "Stress Test" the Demand and Price Environment, We Construct Three Scenarios: (1) Slightly More Bullish Than the IMF (i.e., Back to the IMF April 2012 Figures); (2) One That Respects China's Official 7.5% GDP Growth Target; and (3) Significantly Below the Target and the Implied Consensus Growth Rates
Source: IMF and Bernstein estimates and analysis. Exhibit 149 The Resulting Price Paths Illustrate That the Risks for Long-Term Iron Ore Prices Are
to the Upside, Barring a Significant "Miss" on the Chinese Growth
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 109
Scenario 1 — "Back to the Future" A return to the world anticipated by the IMF in April 2012 would see a re-acceleration in the Chinese steel demand, with only 18% (rather than ~60%) of new supply being price destructive. Prices could stay at US$160/t for a sustained period of time.
Exhibit 150 Scenario 1 Uses a Slightly Stronger GDP
Deck Than the Current IMF Figures…
Exhibit 151 …And Nearly 2% Higher Than Consensus
Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 152 To Support This Level of Growth, Steel
Intensity Would Need to Rise
Exhibit 153 It Would Enable an Accelerated Capital Stock Growth Rate in China
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%
11.0%
12.0%
13.0%
14.0%
15.0%
Rea
l GD
P G
row
th
China GDP Deck
Scenario 1 Consensus
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
1.4%
1.6%
1.8%
200
5A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
Scenario 1 GDP vs. Implied Consensus
40
45
50
55
60
65
70
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Kg
per
$'0
00
of G
DP
Chinese Steel Intensity
Scenario 1 Consensus
-
200
400
600
800
1,000
1,200
Cru
de
Ste
el (M
t)
Chinese Crude Steel Production
Scenario 1 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
110 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 154 This Would See Steel Growth Approach 10%...
Exhibit 155 ...And a Steel-to-GDP Ratio Well Above 1
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Exhibit 156 Under This Scenario, Only 18% of New
Volume Is Price Destructive…
Exhibit 157 …Leading to a Continuation of High Prices Into the Medium Term
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
-5.0%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Chinese Crude Steel Production Growth
Scenario 1 Consensus
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Steel to GDP Growth
Scenario 1 Consensus
441
94
New Supply — Scenario 1 (Mtpa)
Required Additional Supply Price Destructiv e Supply
0
20
40
60
80
100
120
140
160
1802
005
A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
Iro
n O
re P
ric
e —
No
min
al —
US
$/t
FO
B A
us
Iron Ore Price
Scenario 1 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 111
Scenario 2 — Ending With a Whimper If we respect a trend decline down to the government's target annual GDP growth of 7.5%, we would still see iron ore prices track some US$15/t above consensus.
Exhibit 158 Scenario 2 Assumes the Official Annual
GDP Growth Figure of 7.5%...
Exhibit 159 …Which Is Still Significantly Ahead of Consensus Expectations
Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 160 Although It Can Be Achieved Without an
Upward Correction to Steel Intensity…
Exhibit 161 …It Still Anticipates an Above-Consensus Demand Growth…
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%
11.0%
12.0%
13.0%
14.0%
15.0%
Rea
l GD
P G
row
th
China GDP Deck
Scenario 2 Consensus
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
0.7%
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Scenario 2 GDP vs. Implied Consensus
40
45
50
55
60
65
70
200
5A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
kg
pe
r $
'00
0 o
f G
DP
Chinese Steel Intensity
Scenario 2 Consensus
-
200
400
600
800
1,000
1,200
Cru
de
Ste
el (M
t)
Chinese Crude Steel Production
Scenario 2 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
112 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 162 …With Well-Supported Steel Growth... Exhibit 163 ...And a Declining Steel-to-GDP Ratio
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Exhibit 164 Nevertheless, 200Mtpa of Supply Is Still
Unwarranted from a Value Perspective...
Exhibit 165 …Leading to Suppressed Yet Above-Consensus Iron Ore Prices
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Chinese Crude Steel Production Growth
Scenario 2 Consensus
0.0
0.5
1.0
1.5
2.0
2.5
Steel to GDP Growth
Scenario 2 Consensus
334
200
New Supply — Scenario 2 (Mtpa)
Required Additional Supply Price Destructiv e Supply
0
20
40
60
80
100
120
140
160
180
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Iro
n O
re P
ric
e —
No
min
al
—U
S$
/t F
OB
Au
sIron Ore Price
Scenario 2 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 113
Scenario 3 — "It's All Over Now Baby Blue" The third scenario sees all of the new supply growth as price destructive. It is consistent with China growing at 6% p.a. and would see iron ore prices test the US$65/t level.
Exhibit 166 Our Bear Case Assumes 6% GDP Growth
for China in the Long Term…
Exhibit 167 …Which Is ~1% Below the Current Implied Consensus Expectation
Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 168 This Would Lead to a Significant Reduction
in Steel Intensity...
Exhibit 169 ...And Would See Stagnant Production Into the Medium Term
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%
11.0%
12.0%
13.0%
14.0%
15.0%
Rea
l GD
P G
row
th
China GDP Deck
Scenario 3 Consensus
-1.2%
-1.0%
-0.8%
-0.6%
-0.4%
-0.2%
0.0%
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Scenario 3 GDP vs. Implied Consensus
40
45
50
55
60
65
70
200
5A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E
kg
pe
r $
'00
0 o
f G
DP
Chinese Steel Intensity
Scenario 3 Consensus
-
200
400
600
800
1,000
1,200
Cru
de
Ste
el (M
t)
Chinese Crude Steel Production
Scenario 3 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
114 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 170 Consequently, the Industry Would Effectively Go Ex Growth...
Exhibit 171 ...And Steel-to-GDP Ratio Would Fall Well Below 1
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Exhibit 172 In This Scenario, Virtually All Volume
Growth Is Price and Value Destructive…
Exhibit 173 …Iron Ore Tests a US$65/t Floor
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
-5.0%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Chinese Crude Steel Production Growth
Scenario 3 Consensus
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Steel to GDP Growth
Scenario 3 Consensus
6
528
New Supply — Scenario 3 (Mtpa)
Required Additional Supply Price Destructiv e Supply
0
20
40
60
80
100
120
140
160
18020
05A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Iro
n O
re P
rice
-N
om
inal
-U
S$/
t F
OB
Au
sIron Ore Price
Scenario 3 Consensus
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 115
Our Iron Ore Price Forecast
We believe that prices rise to the extent that marginal units of supply are called by demand and that prices fall to the extent that marginal units of supply are displaced. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of that industry. With this heuristic in place, we take inspiration from Fischer Black's "Business Cycles and Equilibrium," a book on business cycles and monetary policy,37 and construct a modified framework, replicating his approach to business cycles but extending the lead time between demand signals and the supply side response. One consequence of the long lead time is inherent cyclicality in the mining industry. Furthermore, this cyclicality ensures that the incentive price (the price at which a miner is incentivized to bring on new capacity) is seldom achieved, but is rather over- or under-shot. This means that the value generated by investment (either by the mining companies themselves or by investors into mining companies) is almost entirely a call on the ability to act counter-cyclically.
In previous chapters, we looked at scenario analysis conducted in mid-2Q:13. This chapter discusses our updated iron ore price forecast, distilled from the analytical frameworks of the previous chapters and their conclusions.
At the time of publication of this Blackbook, we predict $130/t (2013E) rising to a peak of $144/t (2016E). This is consistent with the IMF's current GDP forecast of 7.8% in 2013 rising to 8.5% in 2016 and average increase in supply of 88Mtpa 2013-16. By 2020, we expect iron ore prices of $115/t consistent with IMF forecasts for Chinese GDP growth 2017-20 and a further 28Mtpa of incremental supply in the market.
Our basic premise is that price signals refer to something in the real economy — namely, the marginal units of supply called to clear demand. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of the industry. We also note that the simple connection between a notional "oversupply" and falling prices or "undersupply" and rising prices embeds two implicit assumptions. The first is that price elasticity at the margin is high. The second is that cost escalation is low. In a world where these conditions do not hold, it is perfectly possible to have an oversupplied market with rising prices. We have to understand the structure of supply and demand at each point, and use that understanding to deduce the market clearing price. Prices rise to the extent that marginal units of supply are called by demand.
That call can originate from a number of sources, the most obvious of which is the increase in underlying economic growth. Other reasons include the depletion of existing non-marginal units of supply, as existing mines become exhausted.
Prices fall to the extent that marginal units of supply are displaced. The displacement can include both temporary falls in demand (e.g., through destocking and inventory management programs) and secular declines in demand (as a result of substitution or falling overall growth rates). The displacement can also occur as a consequence of the growth in non-marginal units of supply (which is typically described as an "oversupplied" industry).
While our interpretation of what supply and demand imply is perhaps a little different, the building blocks are traditional enough. The first element in our model is demand. We show how raw material consumption fits into the overall structure of economic output for a country. We do not believe that raw material consumption
37 Published in 1987.
We Apply a Fischer Black Framework Taken from Fischer Black's Work to Our Supply and Demand Analysis, Adapted to Long Lead Times in the Mining Sector; This Generates Our Counter-Consensus Positive View on Iron Ore Pricing
Our Basic Premise Is That Price Signals Refer to Something in the Real Economy
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
116 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
is an effect of economic growth — rather it is the cause of it. We understand the economy as primarily a mechanism for the conversion or processing of raw materials. Economic growth is then a measure of its relative intensity. This activity requires a scale of capital stock (machinery and infrastructure) capable of displacing human labor with the far more concentrated forms of energy contained in (primarily) fossil fuels. It is this displacement of labor by capital and high-intensity energy that generates high levels of productivity in the primary and secondary industry that we associate with developed economies. The displaced labor then moves into tertiary forms of value-add, leading to a society and an economy characterized by its service sector. However, the essential point is that the transition to a service sector cannot be occasioned without the development of a capital stock of sufficient scale.
With the previous discussion in mind, it is clear that China is currently at a point in its economic life where it is still installing this capital stock. Investment now accounts for 45% of Chinese GDP compared to 35% just 10 years ago (see Exhibit 174 and Exhibit 175). This, in turn, corresponds to the fact that over the last 10 years, secondary forms of economic activity in China have been the fastest-growing sector, outstripping growth in the tertiary forms of value-add by 1.4% per annum (see Exhibit 176). This implies a very significant capital stock formation in China. Moreover, it is the rapidity of this stock formation that stands behind the rapid acceleration in China's output over the last 10 years. We use this fundamental relationship between stock and output as the basis for our demand-side analysis (see Exhibit 177 through Exhibit 179). At this stage in China's development, the economy remains very metal intensive. Consequently, any re-acceleration in GDP back to the trend levels forecast by the IMF (8.5%) or OECD (8.4%) must translate into a corresponding acceleration in metal demand. In developed economies, accelerating GDP means accelerating consumption. However, the composition of the Chinese economy is such that consumption alone cannot account for economic growth. Rather, it must be supplemented with investment, which, in turn, means metal.
Exhibit 174 There Has Been a Significant Change in the
Structure of the Chinese Economy Over the Last 10 Years...
Exhibit 175 ...With Consumption Losing Out to Investment; Now Investment Accounts for Around Half of All Activity
Source: Bloomberg L.P. and Bernstein estimates and analysis. Source: Bloomberg L.P. and Bernstein estimates and analysis.
45%
16%
35%
2%2%
Composition of Chinese GDP (2002)
Household Consumption Govt Consumption
Fixed Capital Formation Inventory Formation
Net Exports
35%
14%
46%
2% 3%
Composition of Chinese GDP (2012)
Household Consumption Govt Consumption
Fixed Capital Formation Inventory Formation
Net Exports
At This Stage in China's Development, Its Economy Remains Metal Intensive; Any Re-Acceleration in China's GDP Back to Trend Levels Forecast by the IMF (8.5%) or OECD (8.4%) Must Translate Into a Corresponding Acceleration in Metal Demand
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 117
Exhibit 176 Secondary Forms of Economic Activity Have Grown on Average 1.4% Faster Than the Tertiary Forms Over the Course of China's Industrialization (10.7% vs. 9.3%)
Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 177 However, the Net Result of All of This Activity Is That China Is on Trend to Catch Up
With the U.S. in Terms of the Overall Steel Productivity
Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
118 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 178 China Currently Has ~35% of the Steel Stock of the U.S.; We Expect It to Rise to ~69% by 2020
Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis. Exhibit 179 We Use This Relationship Between Steel Stock and Output to Link the Steel Intensity
(i.e., Rate of Steel Consumption) Between Different Periods in a Country's Economic Development
Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis.
However, there is more than just the steel demand figure that needs to be considered. Specifically, it has to be put in the context of an overarching view on the development of the Chinese economy. Would the figure that we generate for steel demand imply, for example, an increase or a decrease in the fixed-asset investment (FAI) ratio for China? Exhibit 180 shows that the increase in steel demand that we forecast is commensurate with a decline in the FAI ratio in China back down to ~38% of GDP. As such, it corresponds to the sought-after "rebalancing" of the Chinese economy. In order to explicate the commensurability of increasing steel use with falling FAI, it is worth bearing in mind that the high
0
100
200
300
400
500
600
700
800
900
1000
0%
10%
20%
30%
40%
50%
60%
70%
80%
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Fin
ish
ed
Ste
el D
ema
nd
(M
t)
Ch
ines
e C
ap
ital
Sto
ck
as
% o
f US
AChinese Finished Steel Demand and Capital Stock
Finished Steel Demand Capital Stock as % of USA
0
10
20
30
40
50
60
70
80
- 5 10 15 20 25 30 35 40 45 50 55
Ste
el I
nte
nsi
ty —
Kg
/$'0
00
Output — GDP per Capita $'000 (Real 2005 PPP)
Steel Intensity Forecast
China Actual China Forecast (IMF GDP) SCB Trend
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 119
FAI ratio in China has very little to do with raw material consumption per se. An analysis of the most important raw materials in China establishes that only about 12% of the FAI is tied to raw material consumption (see Exhibit 181), even given the most aggressive assumption that 100% of metal demand in China occurs as investment rather than consumption. Our forecast for steel use sees this 12% number declining over time (see Exhibit 182). Furthermore, it is clear that only three metals really matter to the Chinese economy considered as a whole: steel, copper and aluminum. Between them, they account for ~80% of the FAI attributable to metal consumption (see Exhibit 183 and Exhibit 184).
Exhibit 180 Such a Trajectory of Steel Use Implies FAI Falling from 48% to 38% of GDP, in Effect
Achieving the Policy Aim of Rebalancing the Chinese Economy
Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis. Exhibit 181 However, the Vast Majority of the Investment in China Relates Not to Raw Material
Consumption Per Se, But Rather the Capitalization of Labor Associated With That Investment
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis.
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
0
5,000
10,000
15,000
20,000
25,000
30,000
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Composition of Chinese GDP FAI and Consumption
FAI in GDP Non FAI in GDP FAI as Proportion of Total
12%
88%
Composition of FAI in 2012
Raw Materials in FAI Non-Raw Materials in FAI
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
120 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 182 Under Our Scenario for Raw Material Consumption (With Constant Prices), the Raw Material Component of FAI Falls from 12.4% (Historical Average) to 8.9% Over the Forecast Period
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis. Exhibit 183 Steel Is by Far the Most Important Raw Material Component of FAI; Copper and
Aluminum Are Also Significant
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis.
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
16.0%20
05A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Raw Material Consumption Contribution to FAI
61%
6%
10%
2%2%1%
18%
Raw Material Breakdown in Chinese FAI
% Steel of FAI % Aluminum of FAI % Copper of FAI % Zinc of FAI % Nickel of FAI % Lead of FAI % Cement of FAI
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 121
Exhibit 184 The Proportion of Raw Materials in FAI Has Stayed Very Stable Over Time
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis.
The real issue with the FAI ratio is not raw material consumption but the capitalization of labor. Exhibit 185 shows how the labor pool in China has moved over time as well as our forecast evolution of it going forward. Exhibit 186 shows a continuation of the trend rates of loss of labor from agriculture, as the de-agrarianization of China continues. We also extend the trend rate adding labor to the service sector. This leaves the secondary sector employment as the balancing item in the labor pool in the context of an overall contraction in the workforce in China from 2014. Consequently, this sees the secondary sector experiencing significant labor losses by ~2016 and is again in line with economic "rebalancing."
Exhibit 187 shows, for interest, the process of capital stock formation and de-agrarianization in China. The capital stock in China is on track to be some 15% larger than that in the U.S. before the transition to a U.S.-style service sector economy is complete. The main takeaway from this analysis, however, is shown in Exhibit 188. It illustrates how rising steel demand with falling FAI in China can be squared through increasing labor productivity in the secondary sector. The increases in steel demand are more than offset by the declines in the labor employed in the secondary sector. It is this productivity increase that also corresponds to the increase in the overall monetary value of the Chinese economy, with more goods circulating ever more efficiently through society.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Composition of Raw Materials in FAI Over Time
% Steel of FAI % Aluminum of FAI % Copper of FAI % Zinc of FAI % Nickel of FAI % Lead of FAI % Cement of FAI
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
122 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 185 Against This Picture of Raw Material Use, the Increase in the Importance of Investment in China Is Due to the Movement of Labor Into Tertiary Forms of Activity
Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis.
Exhibit 186 Against a Backdrop of Continued De-Agrarianization, a Declining Labor Pool and a Continued Move Into Tertiary Forms of Employment, the Component of the Labor Force in Secondary Forms Must Decline Over Time
Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis.
0
100
200
300
400
500
600
700
800
900
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Eco
no
mic
ally
Act
ive
Pe
ople
(mn)
Primary Labor Secondary Labor Tertiary Labor
-8.0%
-6.0%
-4.0%
-2.0%
0.0%
2.0%
4.0%
6.0%
8.0%
20
06A
20
07A
20
08A
20
09A
20
10A
20
11A
20
12A
20
13E
20
14E
20
15E
20
16E
20
17E
20
18E
20
19E
20
20E
Change in Chinese Labor Force Structure
Change in Primary Labor Change in Secondary Labor Change inTertiary Labor
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 123
Exhibit 187 The Trajectory of De-Agrarianization Has China Requiring 15% More Steel Capital Stock Than the U.S. to Effect the Transition to an Economy Dominated by the Service Sector
Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis. Exhibit 188 This Deceleration in Steel Demand and Change in Composition of the Workforce
Imply a Significant Increase in Steel Use Productivity; If Steel Productivity Were Held Flat, There Would Be Significant Upside to Steel Demand
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie, IMF, World Bank, NBS and Bernstein estimates and analysis.
R² = 0.9987
0
5
10
15
20
25
30
35
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%
Imp
ort
anc
e o
f Ag
ric
ult
ure
in C
hin
a R
ela
tive
to U
SA
Capital Stock of China Relative to USA
De-Agrarianization of China Implied in Capital Stock Accumulation
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
0
50
100
150
200
250
300
200
5A
200
6A
200
7A
200
8A
200
9A
201
0A
201
1A
201
2A
201
3E
201
4E
201
5E
201
6E
201
7E
201
8E
201
9E
202
0E P
rod
uct
ivit
y o
f Sec
on
dar
y L
ab
or
—kg
Ste
el/C
apit
a
Lab
or
in t
he
Sec
on
da
ry S
ect
or
Steel Use Productivity
Secondary Labor Steel Productiv ity
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
124 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
On the supply side, we use our analysis of the main mining companies (Vale, Rio Tinto, BHP Billiton and Anglo) as well as third-party data for Fortescue Metals Group (ticker FMG; not covered) to derive the supply side response from the incumbent producers. We supplement this with an estimate of iron ore supply from new entrants and other non-major sources of supply (see Exhibit 190 through Exhibit 192). Then, we estimate what we believe is a representative proxy for the cost structure of the iron ore industry landed China. Clearly, this involves the conversion of a discrete set of individual mines into a continuous output function for the iron ore industry as a whole. However, given the inherent uncertainty in any price forecast, we believe that this step is warranted. We then estimate how the cost structure would evolve if price equilibrium were to be maintained, i.e., if supply changes were to exactly equal demand changes. We supplement this with an estimate of how the structure would change as a consequence of the cost escalation that we continue to see in the Chinese domestic industry. In this regard, it is useful to quickly summarize the two types of movements within the cost structure of a commodity market. Rotations of the cost curve — change margin but may or may not change price.
This change implies differential in cost escalation within the industry and hence a change in competitive position, privileging some players and disadvantaging others. This creation of relative advantage generates margin.
Translations — change price but leave margin unchanged. Translations of the cost curve imply that all players have been impacted equally by any new industry development. Consequently, relative competitive position is unchanged and the change has no impact on the margin.
Margin in mining is created by geological and mining method differentiation. This differentiation then translates into cost-curve rotations, which lead to margin increases and (assuming that the change has not taken place through an unwarranted expenditure of capital) value. For example, in iron ore this has arisen as a consequence of the difference between truck and shovel operations in Australia (mining high-grade hematite) versus wheelbarrow and muscle operations in China (mining low-grade magnetite). We show the impact of the cost-curve changes in Exhibit 189.
Exhibit 189 Against This Demand Environment, We Look at Cost Curves of Iron Ore Landed China; the New Low-Cost Supply Additions Rotate the Cost Curve Downwards; Cost Inflation Translates the Cost Curve Upwards; It Is the Interplay Between Cost Curve Translation and Rotation That Generates Margin and Ultimately Price
Note: 2016 and 2020 numbers are Bernstein estimates.
Source: CRU, AME and Bernstein estimates and analysis.
We Look at the Supply Response from the "Big 4+1" to Understand How the Cost Curve (Landed China) Will Change Over Time; This Cost Curve Dynamic Drives the Expected Price Performance
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 125
Exhibit 190 Our Analysis of the Major Miners Informs Our Supply Side Projections
Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 191 We Allow for Significant Growth from the
Juniors
Exhibit 192 FMG Appears to Have the Most Aggressive Growth Plans
Source: CRU, corporate reports and Bernstein estimates and analysis. Source: CRU, corporate reports and Bernstein estimates and analysis.
-
200
400
600
800
1,000
1,200
1,400
1,600
2005
A
2006
A
2007
A
2008
A
2009
A
2010
A
2011
A
2012
A
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
2019
E
2020
E
Mt I
ron
Ore
"Big Five" Iron Ore Supply
Rio BHP FMG Vale Anglo
- 50 100 150 200 250
Other
BHP
Vale
Rio
FMG
Anglo
Iron Ore (Mt)
Nw Iron Ore Supply to 2020E
0.0% 2.0% 4.0% 6.0% 8.0% 10.0%
FMG
BHP
Rio
Anglo
Vale
Volume Growth CAGR to 2020E
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
126 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Having calculated a supply and a demand response, we look at how the balance in the iron ore market will change over time. It should be stressed that this balance is purely notional. Iron ore is not copper — we do not expect to see 200mt (say) of "spare" iron ore on board a queue of capesize ships outside Qingdao. Rather, there will be changes to system-wide utilization rates, and these changes will be reflected in the corresponding cost structures of the various steps along the steel value chain. These balances are shown in Exhibit 193 for us and Exhibit 194 for consensus. Exhibit 195 then compares our estimates to consensus.
We are more bullish than consensus on demand — a positive for price. Consequently, either consensus has not understood what it will take to re-accelerate the Chinese economy correctly, or the IMF and OECD have underestimated the Chinese dislike of paying for raw materials. We also remain ahead of consensus on the supply side — a negative for price. At the time of our launch in September 2012, we expected a low-cost supply response of over one billion tons by 2020 and a "catastrophic" real price collapse down to ~US$75/t. We are now more inclined to believe that a degree of capital discipline is entering the industry, and that radical overbuild of mining capacity will be avoided.
However, supply and demand are only one component of any price analysis — the other is cost. If the cost structure of an industry is changing faster than the rate at which marginal units of supply are either called or displaced, and if the elasticity of supply at the margin is low, then oversupply can be accompanied by rising prices and vice versa. Despite the simple continuous output function that we have used in this analysis, we believe that there is a good reason to believe that the elasticity of supply in China (once one is at a high-enough cost level) is actually relatively flat. Once you are prepared to mine below, say, 15% Fe ROM, there is an abundance of such material that can be exploited even in China. And if the method of exploitation is similar, then the cost structure of the industry at the margin ought to be similar as well, leading to a low elasticity of supply. However, this would represent an upside to our price forecast, as it is already ahead of consensus. Consequently, we thought it best not to model this effect in our current version of the analysis. However, the change in elasticity at the margin may be structured — cost escalation in China is real enough (see Exhibit 196).
Putting all of these elements together generates our counter-consensus iron ore price forecast (see Exhibit 197 and Exhibit 198). Moreover, we are able to disaggregate the difference between our view and the view implicit in the consensus price forecast, as shown in Exhibit 199. Approximately half of the variance to consensus is attributable to our understanding of cash cost escalation in the Chinese mines, and the other half is due to a difference in supply and demand balance. Exhibit 200 through Exhibit 203 provide the full details of this analysis.
Three Variables Need to Be Considered in Any Price Forecast — Supply, Demand and Cost; However, We Believe That Cost Is Commonly Omitted Through a Desire Not to Confuse Nominal and Real Drivers of Price Appreciation; Yet Differential Real Cost Escalation Is a Key Source of Value Creation in Mining
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 127
Exhibit 193 We Then Look at Changes to the Supply and Demand Situation to Imply a Change in the Overall "Balance" of the Iron Ore Industry...
Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 194 ...Which Is, Naturally Enough, Different from Consensus
Source: CRU, corporate reports and Bernstein estimates and analysis.
(150)
(100)
(50)
-
50
100
150
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Iro
n O
re (M
t)SCB S-D Iron Ore Balance
SCB Change in Demand SCB Change in Supply SCB Change in Balance
(150)
(100)
(50)
-
50
100
150
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Iro
n O
re (M
t)
Consensus S-D Iron Ore Balance
Consensus Change in Demand Consensus Change in Supply Consensus Change in Balance
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
128 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 195 We Expect a Stronger Demand Environment and a Stronger Supply Side Response Than Consensus
Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 196 A Critical Difference Is in the Impact of Expected Continued Cost Escalation on a
Non-Static Chinese Iron Ore Industry
Source: NBS and Bernstein estimates and analysis.
(150)
(100)
(50)
-
50
100
150
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Iro
n O
re (M
t)Consensus vs. SCB Comparison S-D Iron Ore Balance
Comparison Change in Demand Comparison Change in Supply Comparison Change in Balance
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Chinese Iron Ore Cost Escalator
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 129
Exhibit 197 It Is a Combination of the Differences in Supply and Demand Forecasts as Well as Expected Real Cost Escalation in China That Gives Rise to Our Above-Consensus Price Forecast
Source: NBS, Bloomberg L.P., CRU and Bernstein estimates and analysis. Exhibit 198 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of
7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-16E
Source: Bloomberg L.P. and Bernstein estimates and analysis.
-
20
40
60
80
100
120
140
160
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Iro
n O
re (
US
$/t
)
Resulting Price Path and Comparison to Consensus
Price Dif f erence SCB Price Forecast Consensus Price Forecast
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
130 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 199 Roughly 50% of the Upside of Our Forecast vs. Consensus Is Attributable to Differences in Our Cost Escalation Assumptions and 50% Due to S-D Differences
Source: NBS, Bloomberg L.P., CRU and Bernstein estimates and analysis.
(5)
-
5
10
15
20
25
30
2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E
Iro
n O
re (U
S$/
t)
Explanation of Price Differential vs. Consensus
Cost Supply & Demand
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
E
UR
OP
EA
N M
ET
AL
S &
MIN
ING
: A S
TR
AN
GE
LO
VE
— H
OW
I LE
AR
NE
D T
O S
TO
P W
OR
RY
ING
AN
D L
OV
E T
HE
OR
E 131
Exhibit 200 The First Element of an Iron Ore Price Forecast, in Our View, Is the Calibration of a Consistent View on Steel Intensity and Steel Productivity
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie, IMF, World Bank, NBS and Bernstein estimates and analysis.
Exhibit 202 We Look at How the Cost Curve in China Would Change If There Were to Be No Impact on Price; We Then Calculate the Impact of a Notional "Over Supply" or "Under Supply" on That Cost Curve, Particularly Relative to Consensus Expectations
Source: Bloomberg L.P., AME, CRU, Wood Mackenzie, IMF, World Bank, NBS and Bernstein estimates and analysis.
SCB vs. Consensus - Due to Chinese Cost Escalation US$/t 4 9 13 14 16 19 21 24
SCB vs. Consensus - Due to S&D Differences US$/t () 10 18 24 24 22 14 2
SCB vs. Consensus - Due to Chinese Cost Escalation % 111.0% 49.3% 40.9% 36.8% 39.3% 45.4% 59.8% 93.3%
SCB vs. Consensus - Due to S&D Differences % -11.0% 50.7% 59.1% 63.2% 60.7% 54.6% 40.2% 6.7%
For the exclusive use of JA
SO
N LA
PO
RT
E at P
ER
RY
CA
PIT
AL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 135
Local Currency Costs and USD Revenues: The Stabilizing Effect of the "Natural Hedge" in Costs
As JK Galbraith observed, "faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof." As an exercise in listening to Galbraith, we asked ourselves "what are the valuation impacts if we are wrong with regards to the iron ore price, and if $80/t iron ore becomes a near-term reality?" While examining this hypothetical situation, we came across three "automatic stabilizers" that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners and offer a realistic picture of how a sustained fall in iron ore prices would impact valuations: (1) the natural FX hedge embedded in the miners' cost structures; (2) value-in-use (VIU) adjustments commanded by ores of differing qualities; and (3) elasticity of iron ore demand and spare capacity in China's steel-making industry whose exploitation would become economical, were the iron price to fall sufficiently. Across all of these stabilizers, the companies that are the most exposed to iron ore experience the greatest offsetting effect.
After several years of double-digit cost growth, the miners (particularly Rio Tinto and BHP Billiton) are winning the battle to contain further operating cost increases. Furthermore, in a world of falling commodity prices, we expect to see substantial weakening in producer currencies (AUD and BRL). Given that the miners' costs are denominated in local currency and their revenues in USD, any currency weakening will lead to an improvement in operating margin. Consequently, with nearly 50% of the cost increases in iron ore over the last 10 years coming from the impact of currency movements, any currency weakening is a significant source of upside. Our analysis shows that there is a natural hedge in the operating cost structure of the miners that softens the earnings impact of any sustained fall in commodity prices. In a world where iron ore falls to US$80/t, we estimate that Vale and Rio would experience the greatest benefits from this natural hedge, given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.
FX as an Automatic Stabilizer The period from 2003 to 2012 saw the IMF 62% FE monthly spot (CFR Tianjin port) rise at a CAGR of 128%, from an average of US$14/t to an average of US$129/t. Over the same period, iron ore revenue and EBITDA for our coverage group rose at a CAGR of 133% and 141%, respectively. The combination of rising prices and volume increases has led to iron ore EBITDA margin expansion from, on average, 34% to 56%. While margin growth has been impressive, it could have been considerably more so, had the miners' costs not tracked prices upward.
Currencies in key iron ore producing countries show a strong correlation to iron ore prices. The R-squared for the Australian dollar and iron ore prices is 71% (see Exhibit 204). Similarly, for the Brazilian real, it is 57% (see Exhibit 205). The relationship in less iron-ore-intensive economies is weaker, e.g., the R-squared to the South African rand is only 15% (see Exhibit 206). It is therefore evident that iron ore price and cost rises are strongly linked through the medium of currency movement. Consequently, the R-squared between iron ore costs and revenue across our coverage group is 90% (see Exhibit 207). Fortunately, in presence of falling iron ore prices, the effect of this natural hedge is reversed.
In a World Where Iron Ore Falls to $80/t Iron Ore, We Estimate That Cost Savings — Combined With the Natural FX Hedge Embedded in Miners' Operating Cost Structure — Could Offset Valuation Declines by the Equivalent of, on Average, 6% of June 30, 2013 Share Prices for Rio, BHP, Vale and Anglo
The Strong Correlation Between Iron Ore Costs and Revenues Has Dampened Miners' Margin Expansion in a Rising Iron Ore Price Environment; If Iron Ore Prices Fall, the Effect of This Natural Hedge Is Reversed
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
136 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
As a consequence of mining being a commoditized business, there is a very strong natural hedge between its revenues and costs.38 Part of the relationship is driven by currency movements, which are influenced by balance of payment forces arising from strengthening commodity prices in producer countries. Specifically, mining companies based in a jurisdiction that uses a non-USD currency receive large amounts of USD as payments for the commodities they export. In order to compensate the labor and cover other mining costs in the mining jurisdiction, these companies have to convert the dollars they received into the local currency. In doing so, they increase both supply of dollars and demand for the local currency, thus inducing USD to depreciate vis-a-vis the local currency. Depreciation of the dollar then makes mining costs expressed in USD terms higher than they would be otherwise and reduces the miners' margins.
The second component of the natural hedge arises from the fact that significant elements in mining costs (like diesel or steel in grinding media or rubber for tires) are tied to the same macroeconomic drivers that influence commodity prices themselves. Looking at total iron ore costs, we find that local currency appreciation accounts for 42% of Rio Tinto’s iron ore costs and 51% of BHP's (see Exhibit 208 and Exhibit 209).
This also explains why hedging in mining is fraught with danger and, in our view, increases the level of operational risk rather than decreases it. If only the revenue is hedged, then the natural hedge between costs and revenue is broken. In addition, trying to hedge the costs of production as well as the revenue is extremely difficult. Moreover, there is always a price to be paid to set up any hedge, especially one as complicated as what we believe would be required to hedge all of the mining costs' components. There are numerous examples in mining where a naïve forward sale of material (often executed under the auspices of a risk management program) has destroyed significant value through breaking this natural hedge.
Exhibit 204 There Is a Strong Correlation Between the
Price of Iron Ore and the Strength of the Australian Dollar, With a High Iron Ore Price Equating to a Strong Local Currency
Exhibit 205 Likewise, the Relationship Exists (Albeit Less Pronounced) for the Brazilian Real
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.
38 In the "Not All Iron Ores Were Created Equal: The Stabilizing Effect of Value in Use" chapter, we discuss how the degree of commoditization across different ores can vary, resulting in price differentials for different ores.
R² = 0.7147
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
0 25 50 75 100 125 150 175
1 U
SD
= X
AU
D
Iron Ore (US$/t)
Australian Dollar vs. Iron Ore Price
R² = 0.5718
1.00
1.50
2.00
2.50
3.00
3.50
0 25 50 75 100 125 150 175
1 U
SD
= X
BR
L
Iron Ore (US$/t)
Brazilian Real vs. Iron Ore Price
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 137
Exhibit 206 Unfortunately for Anglo American, for the South African Rand, the Relationship Is Not in Evidence
Exhibit 207 Mining Costs and Revenue Exhibit a High Degree of Correlation
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis. Exhibit 208 Currency Appreciation Drives a Significant
Part of the Cost-to-Price Relationship…
Exhibit 209 ...Which Is Particularly True for the Australian Producers Rio Tinto and BHP Billiton
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
R² = 0.1516
5.00
5.50
6.00
6.50
7.00
7.50
8.00
8.50
9.00
9.50
0 25 50 75 100 125 150 175
1 U
SD
= X
ZA
R
Iron Ore (US$/t)
South African Rand vs. Iron Ore Price
R² = 0.9048
0
5,000
10,000
15,000
20,000
25,000
0 20,000 40,000 60,000
To
tal C
ove
rage
Co
sts
($m
)
Total Coverage Revenue
Total Iron Ore Costs and Revenue
41%
17%
42%
Rio Tinto Iron Ore Cost Drivers
Producer Currency Appreciation Local Currency Inflation
Real Cost Increases
51%
22%
27%
BHP Iron Ore Cost Drivers
Producer Currency Appreciation Local Currency Inflation
Real Cost Increases
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
138 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Costs as an Automatic Stabilizer Macro effects notwithstanding, the miners have made significant progress in controlling iron ore costs, with Rio Tinto leading the way (see Exhibit 210 through Exhibit 212). It has shown a deceleration in the relationship between rising iron ore prices and costs as well as a slowdown in total cost growth. On top of that, Rio is the lowest-cost iron ore producer in our coverage. A component of this cost control is Rio Tinto's "Mine of the Future" program, which looks to increase the level of automation and remote automation in mining, thus reducing the requirement for ever more expensive labor. Rio Tinto believes that it represents a once-in-a-century step change in the way mining is conducted. Whether it will truly rate alongside the introduction of steam shovels or froth floatation remains to be seen, but it clearly has significant potential. In iron ore, for example, the program has a vision of utilizing remote-controlled drills to enable drill-and-blast extraction of ore; satellite-guided haul trucks to transport ore at the mine site; driverless trains to carry production to port; and teleoperated ship loading. Rio Tinto believes that the majority of the mining operations could be controlled from a center in Perth — some 1,500km away from the actual mine sites.
We applaud Rio Tinto's vision for the mining industry. Moreover, we look forward to seeing how this may continue to support Rio's ability to contain operating costs, and hence generate value creation.
Rio's fellow antipodean BHP Billiton has also achieved a marked deceleration in both cost growth over time and cost growth relative to iron ore price growth (see Exhibit 213 and Exhibit 214). This slowdown in cost growth coincided with a pause in a period of impressive growth for the company (during which cost controls can often be overlooked in favor of hitting volume targets). We believe that BHP's total cost profile is positively impacted by the nature of the company's most important iron ore asset — Mt Whaleback. This mine was established in 1968 and is the biggest single-pit, open-cut iron ore mine in the world. It is more than 5km long and nearly 1.5km wide. We believe that having large assets with significant inherent optionality greatly facilitates value accretive growth and enables miners to capture genuine economies of scale — trying to get growth from aggregations of numerous smaller assets will always be far more challenging.
Nevertheless, the company still has ground to cover. Its December half-year reporting noted that $1.9 billion in “controllable” cost savings (i.e., not fuel or taxation) had been achieved, and incoming CEO Andrew MacKenzie commented that he would "extend our pressure on costs and drive us to the bottom of cost per mined ton."39
Vale has continued to struggle with cost controls. Unlike its companions in the "Big Three," which have seen cost increases slow down, Vale’s costs have been growing at an accelerating rate (see Exhibit 215 and Exhibit 216). Interestingly, this has left Vale facing opex that is just 7% cheaper than BHP and 21% more expensive than Rio’s (all in USD terms). In its 1Q:13 earnings report, Vale noted that "for the first time in many years," costs and expenses were "an important source of improvement," with cost cutting extending beyond the operational into SG&A. The company reported savings of US$880 million year-over-year and US$2,539 million quarter-over-quarter (with US$1,283 million of this quarterly decline coming from COGS). While Vale’s average cost per mined ton of iron ore could certainly benefit from the new low-cost production from Serra Sul (cash opex estimated at ~US$30/t), we believe this to be an extraordinarily misguided lens through which to view cost controls. In our view, miners must consider the price impact that any new incremental tonnage has upon the market, as it impacts the price for all tons already in the market. For a functionally "pure" iron ore play like Vale, this is even more critical than for one of the more diversified miners.
The Miners Have Taken Steps (of Varying Efficacy) to Curb Iron Ore Cost Increases and Achieve Leaner Cost Structures Than in the Past
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 139
Anglo American has seen its iron ore costs rise though 2012 but showed some signs of containment (see Exhibit 217 and Exhibit 218). The story seems to be midway between the Australian and Brazilian members of the "Big Three," which is perhaps unsurprising, given the series of operational disappointments and the turnaround nature of Anglo's investment case. Perhaps equally unsurprising in this light is that if one Googles “Anglo American cost controls,” the first hits are not stories about CEOs trying to rein in costs in light of new projects (as with the Big Three); instead, they are for jobs as cost controllers within the company (see Exhibit 220).
Exhibit 210 Rio Tinto Is the Lowest Cost Iron Ore Producer in Our Coverage; Anglo, as the
Highest, Experiences Cash Opex That Is 50% Higher Than Rio's
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 211 While Mining Cost Escalation Is a Common
Worry, It Appears That the Miners Have Started to Get It Back Under Control
Exhibit 212 Rio Tinto, in Particular, Is Leading the Way
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
550
15
30
45
50
0 100
35
200 300 400 500 600 700 800150 250 350 450
10
5
50
0
25
40
650 750
20
Rio Vale BHP
33
49
4340
Anglo
+50%
Cumulative Production (2012 – Mt)
Opex–US$/t
R² = 0.9302
0
5
10
15
20
25
30
35
40
0 50 100 150 200
Un
it C
os
t (U
S$/
t)
Iron Ore Price — US$/t
Rio Tinto Iron Ore Costs vs. Price
0
5
10
15
20
25
30
35
40
Un
it C
os
ts (U
S$/
t)
Rio Tinto Iron Ore Costs Over Time
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
140 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 213 BHP Has Gone Through a Similar Experience With Both the Cost Relationship to Iron Ore...
Exhibit 214 ...And Over Time, Showing a Marked Deceleration
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
Exhibit 215 For Vale, the Evidence of Cost Control Is
Less Convincing...
Exhibit 216 …As Can Be Seen in the Continued Cost Growth Over Time
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
R² = 0.8923
0
5
10
15
20
25
30
35
40
45
50
0 50 100 150 200
Un
it C
os
t (U
S$/
t)
Iron Ore Price — US$/t
BHP Billiton Iron Ore Costs vs. Price
0
5
10
15
20
25
30
35
40
45
50
Un
it C
os
ts (U
S$/
t)
BHP Iron Ore Costs Over Time
R² = 0.7575
0
10
20
30
40
50
60
70
0 50 100 150 200
Un
it C
os
t (U
S$/
t)
Iron Ore Price — US$/t
Vale Iron Ore Costs vs. Price
0
10
20
30
40
50
60
70
Un
it C
os
ts (U
S$/
t)
Vale Iron Ore Costs Over Time
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 141
Exhibit 217 For Anglo, the Situation Seems to Lie Somewhere Between Vale and the Australians…
Exhibit 218 …With Some Evidence of Cost Containment
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
Exhibit 219 An Internet Search for the Miners and "Cost Controls" Results in an Ad for a Cost Control Position as at Least One of the Top 10 Hits (as of April 30, 2013); Only Anglo Has All of Its First Five Hits as Job Openings
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
142 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Impact for Our Coverage Declining iron ore prices would see producer currencies weaken, which would lower USD costs and improve margins for the miners. Based on the historical relationship between the iron ore price and iron ore producer currencies, and bearing in mind the current cost structures of our coverage companies (see Exhibit 211, Exhibit 213, Exhibit 215 and Exhibit 217), we estimate that the “natural hedge effect” linked to an iron ore price decline to US$80/t would offset valuation declines by 12% for Vale, 8% for Rio, 4% for BHP and have a neutral effect upon Anglo (all percentages relative to June 30, 2013 prices — see Exhibit 220 and Exhibit 221).40
Exhibit 220 Given the Historical Relationship Between the Iron Ore Price and Producer
Currencies, a Return to US$80/t Iron Ore Would See Producer Currencies Weaken, Lowering US$ Costs and Improving Margin for the Miners
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 221 The Sensitivity to Such a Margin Impact Gives, on Average, ~6% Additional Upside
to the Australian and Brazilian Miners But Is Value Neutral for Anglo American
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
40 All valuation offsets are expressed as a percentage of the June 30, 2013 price, in USD terms.
12%
8%
4%
-1%Vale Rio BHP Anglo
% o
f Cur
rent
Sha
re P
rice
Share Price Impact of Curreny Effect of US$80/t Iron Ore
FX Impact Rio BHP Vale Anglo
Average $ Cost 32.7 42.5 40.0 49.1
Average LC Cost 32.8 42.6 70.7 374.7
US$ Cost at 80$ Fe 27.9 36.2 34.0 50.3
Margin Uplift per Ton ($) 4.9 6.3 6.0 -1.1
Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137
Current EV/EBITDA 5.0 5.8 4.3 5.1
EV Impact ($m) 6,024 5,880 7,787 -283
Per Share Impact ($) 3.26 1.11 1.52 -0.22
June 30, 2013 Price ($) 40.38 25.46 12.90 18.40
Per Share Impact (%) 8% 4% 12% ‐1%
In a World of US$80/t Iron Ore, We Estimate That This Natural Hedge Could Offset Valuation Declines by, on Average, 6% of June 30, 2013 Share Prices
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 143
Not All Iron Ores Were Created Equal: The Stabilizing Effect of Value in Use
Not all iron ores were created equal. This chapter examines the second "automatic stabilizer" we have identified that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners: Value-in-Use (VIU) and heterogeneity in iron ore products. The reference iron ore price is typically the 62% Fe landed China (CIF) price — the 62% grade is, of course, chosen to be broadly representative of much of the high-quality volume supplied into China from Australia and Brazil. However, a significant volume of Australian material (particularly for Fortescue Metals Group, not covered) suffers a substantial discount to the benchmark on the back of a lower Fe grade. Moreover, both Indian and Iranian material are important components of the overall supply mix into China (~10% of the total), and they also suffer a significant discount. This chapter examines the effective premium and discount on iron ores landed into China by region and by producer. We conclude that VIU offers a degree of automatic stabilization to the iron ore price, should a sustained period of "oversupply" begin to emerge. Furthermore, we expect the higher-quality producers to continue to produce in a declining price environment for longer. We believe that a not insubstantial portion of non-Chinese supply could become economically challenged and be pulled back from the market, should the reference price of iron ore experience sustained pressure. In a world where iron ore falls to US$80/t, we estimate that Vale and Rio would experience the greatest benefits from this natural hedge given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.
Not All Iron Ores Were Created Equal In terms of geological abundance, iron ore is not scarce; it is the fourth most abundant element in the Earth's crust, comprising ~5% of the total. However, to be economically useful, the iron contained in iron ore must generally be 4-14x more concentrated (so 20% and 70% Fe by mass) than the average geological abundance (see Exhibit 222). While by no means rendering iron ore a scarce commodity, this does make a considerable difference to the types of material that can be exploited in the production of steel.
From a geological perspective, the story of economically viable iron ore begins in the Mesoarchean era (3,200 million to 2,800 million years ago), with the development of Banded Iron Formations (BIFs). The originally reducing environment of the very early Earth's atmosphere was progressively converted into an oxidizing (oxygen rich) atmosphere through the photosynthesis of blue-green algae in the Earth's oceans.41 The increasing levels of oxygen in the atmosphere reacted with the iron dissolved in the Earth's oceans to form insoluble iron oxides. As the levels of oxygen in seawater rose, the solubility of the iron oxides decreased. Following this decrease in solubility, the oxides were precipitated out of the seawater to form solid bands of iron-rich material. These bands were laid down intermingled with layers of mud and clay resulting in the characteristic pattern as shown in Exhibit 223. The iron-rich material is interwoven with silica-rich layers of
41An atmospheric condition rich in reducing gases (e.g., hydrogen and carbon monoxide) in which oxygen and other oxidizing gases and vapors are removed, thereby preventing oxidation.
VIU Offers a Degree of Automatic Stabilization to the Iron Ore Price, Should a Sustained Period of "Oversupply" Begin to Emerge; in a World Where Iron Ore Falls to US$80/t, This Stabilizer Could Offset Valuation Declines by, on Average, 7% of June 30, 2013 Share Price for Rio, BHP, Vale and Anglo
Iron Ore Is the Fourth Most Abundant Element, Comprising 5% of the Earth's Crust by Mass; However, Not All Iron Ores Are Created Equal
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
144 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
flint and chert42 to create BIFs. The second step in the generation of exploitable iron ore was the enrichment of this basic BIF. This occurred through the leaching out of the silica-rich layers, which led to concentration and further oxidation of the iron ores that were left behind. The weathering of the BIFs over millions of years effectively dissolved away the impurities to leave behind an enriched and economically valuable form of iron ore. It is this iron ore that is mined today. However, the process of weathering and enrichment of BIFs is complicated; the multiplicity of geological processes at work means that there are large variations in the material that can pass under the generic description of "iron ore."
A brief outline of some of the more important types of iron ore is given in Exhibit 224. The maximum Fe grade occurs at ~70%. This corresponds to chemically pure hematite or magnetite; at this degree of concentration, the residual mass is the oxygen associated with the iron rather than any impurity of the ore. The 62% referred to in the Fe benchmark is equivalent to 88.6% pure hematite (i.e., 62/70). It is the residual 11.4%, rather than 38%, that is non-iron bearing material.
As a further example of the point, we can take a single product — such as Vale's high-quality Standard Sinter Fines (SSF) — and look at what other than iron itself is actually in the iron ore (see Exhibit 225).
Exhibit 222 Rusting Granite in Peterborough, New Hampshire; During Pre-Industrial and Early Industrial Periods, Local Blacksmiths Relied Upon "Bog Iron" and "Mountain Iron" (Like the Below); While the Highest-Quality Deposit Reportedly Contained Ore of 56-63%, the Grade Was Often Much Lower and Was Quickly Depleted in Concentrations of Useful Economic Value
Note: A Gazetteer of the State of New-Hampshire (1823) by John Farmer and Jacob B. Moore noted that ore considered to be "the richest" in the nation was "obtained from a mountain in the east part of Concord [changed to Lisbon in 1824]." By 1849, however, The American Railroad Journal (vol. XXII, no. 609) noted that "The ores of New Hampshire, like those of Maine, are generally situated so that the expenses of transportation have rendered them of little value. Only one furnace is in operation, at Franconia…The furnace does but a small business, making only two and a half tons of iron a day." This stone furnace (no longer in use) is the only surviving iron smelter in New Hampshire today.
Source: Bernstein.
42 A fine-grained silica-rich sedimentary rock that often contains small fossils.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 145
Exhibit 223 Where All the EBITDA Began — A Banded Iron Formation (BIF) Showing Characteristic Layering Patterns of Iron Ore and Gangue Material
Source: Wikimedia Commons. Exhibit 224 Not All Iron Ores Are Created Equal...
Source: Wikipedia and Bernstein analysis.
Ore Type Chemical Composition Max Fe
Magnetite Fe3O4 72.4%
Martite Fe3O4.Fe2O3 72.4% to 70.0%
Hematite Fe2O3 70.0%
Limonite 2Fe2O3.3H2O 59.8%
Goethite FeO.OH 62.9%
Siderite FeCO3 48.3%
Ilmenite FeO.TiO2 36.8%
Pyrite FeS2 46.7%
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
146 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 225 …And There Is a Significant "Tail" of Non-Iron Material in Iron Ores, All of Which Have a Value Implication
Source: IPCC and Bernstein analysis.
The purpose of this (very) brief geological digression was to make the point that not all iron ores were created equal. The 62% benchmark may generate a false impression of how homogenous — and hence how commoditized — iron ore is. The reality is that a variety of materials can be classified as "iron ore," all of which behave very differently in the blast furnace (BF). Consequently, each has a different VIU for a steel maker. The most important feature is that it is as costly to melt impurities as iron, and impurities yield no output benefit. Coke (the fuel source in BF steel making) is consumed to melt non-iron material lost to slag. Clearly, this is to the detriment of the steel maker.
Beyond wasting fuel, impurities also impact the resultant product. Many residual elements have a further deleterious effect on either the quality of the pig iron produced or the process of production. For example, phosphorous (P) renders steel brittle at low temperatures and so must be removed by a process of hot metal de-phosphorization, which incurs a cost. Alkali metals (K and Na) attack the refractory lining of the blast furnace, reducing campaign life as well as the strength of coke and pellets, and so lower overall productivity.
From an economic perspective, inclusion of a VIU adjustment into any cost curve tends to increase the marginal cost of production, thus putting upward pressure on the commodity's price. The removal of homogeneity in a commodity industry has, in some sense, a "de-commoditizing" effect — ceteris paribus, it increases the margin for advantaged players at the expense of those with disadvantaged ore assets (in the case of iron ore, a low Fe, high-contaminant profile). To the extent that these disadvantaged players are required to satisfy market demand, then the cost of the "benchmark" product must be high enough so that, once adjusted for VIU, it is still economic for their product to remain in the market.
Chemical Mass (% Wt)
Fe 64.50%
SiO2 4.50%
Al2O3 3.50%
Mn 0.150%
LOI (Loss on Ignition) 0.800%
P 0.027%
SiO2 0.007%
Na2O 0.015%
K2O 0.008%
TiO2 0.080%
CaO 0.020%
MgO 0.030%
From an Economic Perspective, Inclusion of a VIU Adjustment Into a Cost Curve Tends to Increase the Marginal Cost of Production, Thus Putting Upward Pressure on the Commodity's Price
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 147
VIU as an Automatic Stabilizer China is the world’s largest consumer of iron ore. In 2012, the Middle Kingdom accounted for 56% of global demand but produced just 13% of global supply. Australia and Brazil are the most important importers of iron ore to China (46% and 22% of 2012 imports, respectively, a proportion which has remained relatively constant over time) — see Exhibit 226 and Exhibit 227. The ore imported to satisfy this 42% gap between production and demand is of varying quality. Ore imported in March of this year, for example, ranges from 52% to 67% (see Exhibit 228). While this may not sound like a wide range, it is worth bearing in mind that the range is 3x the average concentration of iron ore in the Earth’s crust (5%). It is further worth considering that, as 70% is the maximum Fe grade, this range spans the equivalent of 74-96% pure hematite, with "impurities" ranging from 4% to 26%. These varying ores with their different grades and types of impurities, of course, command different premiums or discounts to the benchmark 62% Fe price (see Exhibit 229).
Unsurprisingly, iron ore grade accounts for 93% of the price variation (see Exhibit 230); the lowest grade imported (52%) commanded the greatest discount at 33%, while the highest grade imported (67%) commanded a 21% premium. However, even after adjusting for Fe grade, poorer-quality ores still command a significant discount (see Exhibit 231 through Exhibit 234). As a result, when the benchmark price of iron ore declines, exporters of poor-quality, low-grade iron ore into China can be expected to drop out faster than exporters of high-quality, high-grade iron ore. This functions like an automatic stabilizer, providing support not only to the iron ore price (thus slowing its rate of decline) but also to valuations of exporters from higher-quality deposits.
Exhibit 226 While Australia and Brazil Are the Most
Important Suppliers Into China, They Account for Only ~70% of the Total Non-Chinese Portion of Supply
Exhibit 227 The Overall Market Shares of Brazil and Australia Have Remained Roughly Constant Over Time; Perhaps This Indicates an Understanding from Incumbents That Competing on Volume in Commodity Markets Is Value Destructive
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.
46%
22%
5%
4%
2%2%2%
2%
15%
China Iron Ore Imports by Origin
Aus Brazil SA India Iran
Canada Ukraine Russia Other
0%
10%
20%
30%
40%
50%
60%
Jan-
04
Aug
-04
Mar
-05
Oct
-05
May
-06
Dec
-06
Jul-0
7
Feb-
08
Sep
-08
Apr
-09
Nov
-09
Jun-
10
Jan-
11
Aug
-11
Mar
-12
Oct
-12
Share of Chinese Iron Ore Imports
Australia Share of Imports Brazil Share of Imports
The Price Paid for Iron Ore Imported Into China Varies Considerably; This Acts Like an Automatic Stabilizer in a Falling Price Environment, Providing Support Not Only to the Iron Ore Price But Also to Producers of High-Quality Product
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
148 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 228 There Is a Considerable Variation in the Observed Grade of Iron Ore Landed Into China — It Is Still Far from Being a Homogeneous Product
Source: BAIINFO and Bernstein analysis. Exhibit 229 An Analysis of 208 Pricing Points Shows a Significant Range of Variation Around the
Benchmark 62% CIF Price Point
Source: BAIINFO and Bernstein analysis.
50%
52%
54%
56%
58%
60%
62%
64%
66%
68%
Australia Brazil Chile India Iran Malaysia Mexico Russia South Africa
Ukraine
Fe
Gra
de
Distribution of Observed Iron Ore Grades Landed in China During March
R² = 0.7734
-40
-30
-20
-10
0
10
20
30
50% 52% 54% 56% 58% 60% 62% 64% 66% 68%
Iro
n O
re P
rem
ium
vs
. Be
nch
mar
k —
US
$/t
Fe Grade of Iron Ore
Price Impact of Quality in Iron Ore
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 149
Exhibit 230 Unsurprisingly, Fe Grade Explains Most of the Variation in Price
Exhibit 231 Nevertheless, Even After Adjusting for Fe Grade, Poor Ores Suffer a Significant Discount
Source: BAIINFO and Bernstein analysis. Source: BAIINFO and Bernstein analysis. Exhibit 232 This Is True Across Australian Iron Ores... Exhibit 233 ...As Well as Those from India
Source: BAIINFO and Bernstein analysis. Source: BAIINFO and Bernstein analysis.
R² = 0.9294
-40
-30
-20
-10
0
10
20
30
50% 55% 60% 65% 70%
Pre
miu
m —
US
$/t
Fe Grade
Average Premium (US$/t)
R² = 0.6298
-25
-20
-15
-10
-5
0
5
10
15
20
50% 55% 60% 65% 70%P
rem
ium
—U
S$/
tFe Grade
Average Premium (US¢/Dmtu)
R² = 0.9084
-20
-15
-10
-5
0
5
10
15
20
50% 55% 60% 65%
Pre
miu
m —
US
$/t
Fe Grade
Premium for Australian Ore
R² = 0.9938
-40
-35
-30
-25
-20
-15
-10
-5
0
5
50% 55% 60% 65%
Pre
miu
m —
US
$/t
Fe Grade
Premium for Indian Ore
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
150 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 234 There Is a US$46/t Differential Across Ores from Different Supply Locations
Source: BAIINFO and Bernstein estimates and analysis.
Impact for Our Coverage There is US$25/t of iron ore differential across the majors importing into China, with Fortescue Metals Group (not covered) at the low end of the range and Anglo American at the high end (see Exhibit 235). Assuming that 50% of the quality-disadvantaged material moves out of the market as prices begin to fall, VIU stabilizer can be expected to provide ~US$5/t of offsetting upward price support (see Exhibit 236). We estimate that Vale and Rio would experience the greatest benefits (equivalent to 11% and 9% of June 30, 2013 share prices), while more diversified BHP and Anglo experience 4% and 6% offsets, respectively.
Exhibit 235 Across the "Major" Space, There Is a US$25/t Differential
Source: BAIINFO and Bernstein estimates and analysis.
‐35
‐10
15
‐30
10
5
0
‐5
0 5 10 15 20 25 30 35 40 45 50
‐20
‐15
‐25
55
IranMalaysia
‐46
SAChile
Ukraine Brazil MexicoRussia Aus India
Average Monthly Imports to China ‐Mt
Prem
ium Paid in China
–US$/t
‐8
‐6
‐4
650‐2
2
4
550
6
600
8
12
14
350 400 450 500 700 750 800 850
‐12
0
10
0 50 100 150 200 250 300
‐10
BHPBRioVale FMG
‐25
Anglo
Annual Production by Majors ‐Mt
Prem
ium Paid in China
–US$/t
In a World of US$80/t Iron Ore, VIU Automatic Stabilizer Could Offset Valuation Declines by, on Average, 7% of June 30, 2013 Share Prices
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 151
Exhibit 236 Assuming That 50% of the Quality-Disadvantaged Material Moves Out of the Market as Price Begins to Fall, It Can Be Expected to Provide ~US$5/t of Offsetting Upward Price Support, Adding, on Average, $1.75/Share Across Our Coverage (as of June 30, 2013 Prices)
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
VIU Impact Rio BHP Vale Anglo
Average $ Cost 32.7 42.5 40.0 49.1
Margin Uplift per Ton ($) 5.4 5.4 5.4 5.4
Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137
Current EV/EBITDA 5.0 5.8 4.6 5.0
EV Impact ($m) 6,583 5,010 7,457 1,330
Per Share Impact ($) 3.56 0.94 1.46 1.04
June 30th, 2013 Price ($) 40.38 25.46 12.90 18.40
Per Share Impact (%) 9% 4% 11% 6%
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
152 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 153
If Prices Fall While Capital Stock Is Accumulating, Demand Accelerates: The Demand Elasticity Stabilizer
This chapter examines the third and final "automatic stabilizer" we have identified that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners: the elasticity of demand for iron ore. Chinese steel making has seen its EBITDA margins cut by ~70% over the last decade on the back of rising raw material prices (iron ore CAGR 26%, PCI and coking coal CAGR 14%) and essentially stagnant steel prices (1% local currency CAGR, 4% USD CAGR). Commodity price escalation has driven steel making from being highly profitable to break even. We believe that this decline coincides with (or rather is instrumental in causing) the declines in Chinese steel production growth (down from ~20% p.a. 10 years ago to the ~6% p.a. seen today). However, steel capital stock in China is still only around one-third the level of that in the West, and the urbanization and industrialization of China remains far from complete. It is inconsistent to believe that China is undergoing a period of ongoing capital stock accumulation, while also believing that iron ore prices would fall as a consequence of low Chinese steel growth. Consequently, if iron ore prices were low for a sustained period of time on the back of an "oversupplied" market, demand would rise, which would itself unwind some of the apparent "oversupply." Compared to consensus expectations of a long-term iron ore price of 80US$/t, we believe that this feedback loop could generate long-term iron ore prices some 15US$/t higher than would otherwise be the case. This would see demand-supply equilibrium restored at a level closer to US$95/t. We estimate that Vale and Rio would experience the greatest benefits from this natural hedge, given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.
China's Industrialization and Demographics — A Recap
According to our supply and demand overview discussed in the "Iron Ore 'Super-Cycle' Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization" chapter, through months of data gathering, we assembled a database covering the consumption of metals and economic output for 120 countries since 1900. Our iron ore demand framework looks at two distinct components: the level of installed stock (here, cumulative steel intensity, i.e., total historical steel installed in an economy less depreciation) and the rate of steel consumption (here, current steel intensity). We look at both of these on a per-capita basis against GDP per capita, which enables us to account for not only the stage of industrialization (and output generation) of a given economy but also its importance to the overall global steel (and hence iron ore) consumption. Industrializing and industrialized economies move through three distinct phases of their rate of steel consumption (see Exhibit 237). The same three phases in the rate of steel consumption are repeated across industrialized and industrializing countries in our dataset. We categorize them as "Development," "Peak" and "Decline/Services." Furthermore, three paradigms emerge for the evolution of the level of steel consumption across countries that have industrialized (or attempted to). Two of these paradigms are successful ("Services and Technology," exemplified by the U.S. and the U.K., and "Manufacturing," exemplified by
If Iron Ore Prices Were Continuously Low on the Back of an "Oversupplied" Market, Demand Would Rise, Which Would Itself Unwind Some of the Apparent "Oversupply"; This Automatic Stabilizer Could Offset Valuation Declines by, on Average, 20% of the June 30, 2013 Share Prices for Vale, Rio, BHP and Anglo
Far from Being Overbuilt, China Resembles the U.S. in the 1930s, Has Just One-Third the per Capita Level of Embedded Steel as the Present Day U.S. and Is Facing Tremendous Pressure to Industrialize Before Its Old-Age Dependency Ratio Skyrockets
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
154 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Germany and Japan), while one is not sustainable ("Inefficient," exemplified by the former USSR).
The 120 countries in our dataset display a strongly correlated relationship (R-squared = 66%) between levels of installed capital and GDP. Furthermore, and perhaps more importantly, as the outliers can be explained by individual economic circumstances (e.g., manufacturing hubs or endowment with oil), this framework generates a powerful predictive tool for looking at the sustainability of the current and future demand in the context of historical development. Our framework shows that Chinese steel demand (46% of 2012 finished steel demand globally) is on a sustainable path. China’s current steel intensity ("rate") is consistent with its stage of industrialization. It shows China transitioning from Phase 2 (peak) to Phase 3 (services) at a rate commensurate with that seen elsewhere. This is not a "structural shift" so much as a natural development to an economic state that is less steel intensive. To highlight the sustainability of China's demand, we would call attention to the fact that the economic activity of every person in the U.S. (or any other developed country) is supported by an installed capital base comprising 80kg of copper and 12,000kg of steel — well more than double the current capital stock in China (26kg of copper and 4,500kg of steel per capita).
Furthermore, as discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter, a legacy of the "one child" policy is that China's aging population resembles an industrialized nation's demographics. However, it is supported by only a pre-industrial level of output. By 2020, China's working-age population will have peaked at close to 1 billion people. Over the next generation, 260 million people will leave the Chinese labor force, leading to a 90% increase in the old-age dependency ratio (see Exhibit 238). While such demographic challenges are not new, they have hitherto been associated with industrial rather than agrarian economies. By the time Japan's working population started to decline at the end of the 1990s, the country had an installed capital stock equivalent to nearly 14,000kg of steel per capita. With China's current capital stock standing at 4,500kg, our demand projections anticipate that this will approach 10,000kg by the time China's population starts to age dramatically (2020). In this regard, it is important to note that there is an important distinction between high young-age and old-age dependency ratios. Young-age dependency ratio = investment. I feed my children today on the
expectation that in the future they will feed not only themselves but will generate a surplus that will enable me to consume. The resources that are allocated toward a society's children are the quintessential investment in the future productive capacity of that society.
Old-age dependency ratio = consumption. Other than the social and moral imperative, it is not clear why I would want to feed granny. There is no hope that feeding granny will ever pay for itself. Allocating resources to the non-productive members of society can never generate a return in excess of the resources that are thereby consumed.
This is not an attempt to make any far-reaching social point. We emphasize that, from an economic perspective, one form of dependency ratio is "affordable" in the sense of offering a future return, while the other is not. If a society wishes to move to a gerontocracy, the imperative is to make sure that there is a level of productive capacity high enough to withstand the necessary intergenerational redistribution of the calls on that capacity without triggering social unrest.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 155
Exhibit 237 Following the Same Three-Phase Trajectory for the Evolution of the Steel Consumption Rate Within All Industrialized and Industrializing Countries, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady State Rate of Consumption, Once the Process Is Completed, Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing- or Services-Oriented Focus
Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45 50
Ste
el In
ten
sity
(kg
/$'0
00
GD
P)
Output — GDP/Capita — $'000 (Real 2005 PPP)
Trend Steel Intensity Evolution
France USA China Japan South Korea China Actual
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
156 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 238 In 2000, China Represented a Clear Outlier from the Normal Pattern: the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-Industrial Level of Output; by 2020, China's Working-Age Population Will Have Peaked; Over the Next Generation, 260 Million People Will Leave the Chinese Labor Force, Leading to a 90% Increase in the Old-Age Dependency Ratio
Source: UN and Bernstein estimates and analysis.
-10
0
10
20
30
40
50
60
-50 -40 -30 -20 -10 0 10
GD
P/C
apit
a —
US
$'00
0
Change in Old-Age Dependency Ratio 2000-2040 — Negative Indicates Aging Population
Aging of Population vs. Development of Economy Including ChinaSize of Bubble = Population
USA
Japan
China
South Korea
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 157
Lower Iron Ore Price as a Stimulus for Chinese Steel Making Over the last decade, the Chinese domestic price of steel has grown at a CAGR of 1% (4% in USD terms), while its input costs have risen at, on average, 8.5%. In USD terms, iron ore costs have had a CAGR of 26% over the period, while for both PCI and coking coal the CAGR has been 14% (see Exhibit 239 through Exhibit 243). As a result, raw material input costs per ton of steel have risen from less than US$100/t in 2003 to more than US$300/t today (see Exhibit 244) and comprise nearly three-fourths of the overall costs of steel making (see Exhibit 245 and Exhibit 246). The rising input costs, combined with more slowly rising steel prices over the last decade, have, unsurprisingly, seen steel-making EBITDA margins cut to one-third of their previous levels — from an average of 62% in 2003 to 21% by the end of 1Q:13 (see Exhibit 247) — and profit margins all but wiped out.
While at first glance Chinese crude steel production appears to have grown steadily (see Exhibit 248), this belies the fact that steel production growth has been declining. Concurrent with the rising input cost pressures, the annualized crude steel production growth has declined from ~20% to ~6% (see Exhibit 249 and Exhibit 250). The correlation between steel growth and steel profitability has an R-squared of 52% (see Exhibit 251).
Exhibit 239 Over the Last 10 Years, the Domestic Steel
Price in China Has Been Remarkably Constant...
Exhibit 240 ...Whereas the Price of Raw Materials Has Accelerated Sharply; It Is True for All — Iron Ore...
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P., IMF and Bernstein analysis.
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
Jan-
03
Sep
-03
May
-…
Jan-
05
Sep
-05
May
-…
Jan-
07
Sep
-07
May
-…
Jan-
09
Sep
-09
May
-…
Jan-
11
Sep
-11
May
-…
Jan-
13
HR
C —
RM
B/t
Chinese Domestic Steel Price
0
20
40
60
80
100
120
140
160
180
200Ja
n-0
3
Sep
-03
May
-04
Jan
-05
Sep
-05
Ma y
-06
Jan
-07
Se p
-07
May
-08
Jan
-09
Se p
-09
May
-10
Jan
-11
Sep
-11
Ma y
-12
Jan
-13
US
$/t
Iron Ore Price
In the West, a Negative Step Change in Steel Demand Growth Occurred Only Once Urbanization Was Complete; We Believe That If the Price of Iron Ore Were to Fall, the Demand for Steel Would Rise
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
158 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 241 ...Key Fuel and Reductants Used in Steel Making — HCC...
Exhibit 242 ...And Its Replacement PCI
Source: CRU and Bernstein analysis. Source: CRU and Bernstein analysis. Exhibit 243 The Rise in Iron Ore Price Has Done the Most to Increase the Costs of Chinese Steel
Production, Reflecting a Transfer of Value from the Chinese Industrial Sector to the Western-Owned Miners
Source: AME, CRU, IMF and Bernstein analysis.
0
50
100
150
200
250
300
350Ja
n-0
3A
ug-0
3M
ar-0
4O
ct-0
4M
ay-0
5D
ec-0
5Ju
l-06
Feb
-07
Sep
-07
Ap
r-08
No
v-08
Jun
-09
Jan
-10
Aug
-10
Mar
-11
Oct
-11
May
-12
Dec
-12
US
$/t
Hard Coking Coal (HCC) Price
0
50
100
150
200
250
300
Jan
-03
Sep
-03
May
-04
Jan
-05
Sep
-05
Ma y
-06
Jan
-07
Sep
-07
Ma y
-08
Jan
-09
Se p
-09
Ma y
-10
Jan
-11
Sep
-11
Ma y
-12
Jan
-13
US
$/t
PCI Price
26%
14% 14%
6%5% 5%
4% 4% 3% 3%
Iron Ore Cost
PCI Cost Coking Coal Cost
Steel Making
Iron Making Sintering Admin & Support
Rolling Casting & Refining
Coke Making
CA
GR
Chinese Steel-Making Cost Escalation (CAGR 2003-2013)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 159
Exhibit 244 Raw Materials Fed Into Steel Making Used to Cost Less Than 100US$/t; It Is Now Well Over 300US$/t
Source: AME, CRU, IMF and Bernstein analysis. Exhibit 245 Currently, the Raw Material Cost of Steel
Making Accounts for Nearly Three Quarters of the Overall Costs in China
Exhibit 246 There Has Also Been a Slower, But Nonetheless Pronounced Rise in the Non-Raw Materials Costs of Steel Making in China
Source: AME, CRU, IMF and Bernstein analysis. Source: AME and Bernstein analysis.
0
100
200
300
400
500
600Ja
n-03
May
-03
Sep
-03
Jan-
04
May
-04
Sep
-04
Jan-
05
May
-05
Sep
-05
Jan-
06
May
-06
Sep
-06
Jan-
07
May
-07
Sep
-07
Jan-
08
Ma y
-08
Sep
-08
Jan-
09
May
-09
Sep
-09
Jan-
10
Ma y
-10
Sep
-10
Jan-
11
Ma y
-11
Se p
-11
Jan-
12
May
-12
Sep
-12
Jan-
13
US
$/t
Chinese Steel-Making Raw Materials Cost
Iron Ore Cost Coking Coal Cost PCI Cost
45%
21%
3%
31%
Steel Making Costs by Element (2013)
Iron Ore Cost Coking Coal Cost PCI Cost Other Costs
5540
35
24
27
14
15
9
11
7
7
4
2
2
2013 2003
US
$/t S
tee
l
Non-Raw Material Cost of Steel Making
Casting & Refining Rolling Steel MakingIron Making Admin & Support SinteringCoke Making
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
160 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 247 Unsurprisingly, Margins in the Chinese Steel-Making Industry Have Declined on the Back of Higher Raw Material Prices
Source: AME, CRU, IMF and Bernstein analysis. Exhibit 248 While, at First Glance, the Growth in Chinese Steel Production Appears Monotonic...
Source: Bloomberg L.P., AME, CRU, IMF and Bernstein analysis.
0%
10%
20%
30%
40%
50%
60%
70%
80%Ja
n-03
May
-03
Se p
-03
Jan-
04
May
-04
Sep
-04
Jan-
05
Ma y
-05
Sep
-05
Jan-
06
Ma y
-06
Sep
-06
Jan-
07
Ma y
-07
Sep
-07
Jan-
08
Ma y
-08
Se p
-08
Jan-
09
May
-09
Se p
-09
Jan-
10
Ma y
-10
Se p
-10
Jan-
11
May
-11
Se p
-11
Jan-
12
May
-12
Se p
-12
Jan-
13
HR
C E
BIT
DA
Mar
gin
Chinese Steel-Making Margin
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
Jan-
03
May
-03
Se p
-03
Jan-
04
May
-04
Sep
-04
Jan-
05
May
-05
Se p
-05
Jan-
06
Ma y
-06
Sep
-06
Jan-
07
May
-07
Sep
-07
Jan-
08
May
-08
Sep
-08
Jan-
09
May
-09
Sep
-09
Jan-
10
May
-10
Se p
-10
Jan-
11
May
-11
Sep
-11
Jan-
12
Ma y
-12
Se p
-12
Jan-
13
Cru
de
Ste
el —
kt
Chinese Monthly Crude Steel Production
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 161
Exhibit 249 ...This Belies a Marked Slowdown in Chinese Steel Production Rates from ~20% at the Start of the Last Decade to Closer to 6% Today, Concurrent With the Declining Steel-Making Margins
Source: Bloomberg L.P. and Bernstein analysis. Exhibit 250 Growth and Profitability Declined Concurrently
Source: Bloomberg L.P., AME, CRU, IMF and Bernstein analysis.
-10%
-5%
0%
5%
10%
15%
20%
25%
30%
35%Ja
n-05
Apr
-05
Jul-0
5
Oct
-05
Jan-
06
Apr
-06
Jul-0
6
Oct
-06
Jan-
07
Apr
-07
Jul-0
7
Oct
-07
Jan-
08
Apr
-08
Jul-0
8
Oct
-08
Jan-
09
Apr
-09
Jul-0
9
Oct
-09
Jan-
10
Apr
-10
Jul-1
0
Oct
-10
Jan-
11
Apr
-11
Jul-1
1
Oct
-11
Jan-
12
Apr
-12
Jul-1
2
Oct
-12
Jan-
13
An
nu
aliz
ed T
wo-
Yea
r MoM
CA
GR
Crude Steel Production Growth Rate
-10%
-5%
0%
5%
10%
15%
20%
25%
30%
35%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Jan-
05
Apr
-05
Jul-0
5
Oct
-05
Jan-
06
Apr
-06
Jul-0
6
Oct
-06
Jan-
07
Apr
-07
Jul-0
7
Oct
-07
Jan-
08
Apr
-08
Jul-0
8
Oct
-08
Jan-
09
Apr
-09
Jul-0
9
Oct
-09
Jan-
10
Apr
-10
Jul-1
0
Oct
-10
Jan-
11
Apr
-11
Jul-1
1
Oct
-11
Jan-
12
Apr
-12
Jul-1
2
Oct
-12
Jan-
13
Ste
el G
row
th
Ste
el E
BIT
DA
Mar
gin
Chinese Steel Growth and Profitability Since 2005
Margin, % Annualized Grow th
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
162 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 251 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from High Commodity Prices; the West Experienced a Step Change in Steel Demand Growth Only After Urbanization Was Complete; However, This Is Far from Complete in China, Combined With the Imminent Demographic Shift Enables Us to Conclude That in the Presence of Falling Iron Price, Steel Demand Would Rise
Source: AME, CRU, IMF and Bernstein analysis.
While it is clear that correlation does not necessarily imply causation, we believe that a significant part of the slowdown in Chinese steel growth was a result of the budgetary constraint imposed by high commodity prices. In the West, a negative step change in steel demand growth occurred only once urbanization was complete. Urbanization is far from complete in China. Consequently, we believe that if the price of iron ore were to fall, the demand for steel would rise (particularly bearing in mind the hard stop the demographic shift in China is imposing upon the industrialization process). Consequently, we believe that it is incorrect to interpret the step change seen in OECD steel demand growth post the 1970s' oil shocks as representative of the situation in China today. By the time the OECD-trend steel demand declined, urbanization was complete, and this is not the case in China today. Rather, when China "unleashed" the forces of private enterprise at the turn of the century, it enabled the industry to actively chase the enormous profitability that was available as a consequence of combining the productivity of large blast furnaces with the vast pools of surplus Chinese labor and the availability of incredibly cheap iron ore on the international commodity markets. As commodity prices re-equilibrated themselves, this excess return was passed from Chinese steel makers on to Australian (and Brazilian) iron ore producers. The effect of this is, naturally enough, to reduce the incentive to grow steel output, while increasing the incentive to grow iron ore output. To the extent that the incumbent iron ore producers wish to unwind the structural advantage they currently possess through continuing the poorly thoughtout strategy of value-destructive capital profligacy, they have at least this one consolation: lowering the price of iron ore ought to encourage a higher level of Chinese demand.
A complicating factor in the analysis of steel elasticity is the effect of Chinese monetary stimulus of 2009. A strong relationship exists between lagged steel growth and the growth in the money supply (R-squared = 55%) — see Exhibit 252 through Exhibit 254. If one strips out the effect of Chinese monetary stimulus in 2009, the correlation between steel growth and EBITDA margin increases (R-
R² = 0.5187
-5%
0%
5%
10%
15%
20%
25%
30%
35%
0% 10% 20% 30% 40% 50% 60% 70% 80%
An
nu
aliz
ed S
teel
Gro
wth
Rat
e
Chinese Steel EBITDA Margin
Chinese Steel Growth vs. Chinese Steel Profitability
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 163
squared = 67%) — see Exhibit 255 and Exhibit 256. Stripping out the effect of M2 growth suppresses the steel growth rate that one would associate with any given industry profitability. Consequently, as well as increasing the overall correlation, stripping out government stimulus embeds a level of conservatism in the forward-looking counterfactual demand growth that we calculate for the falling iron ore price scenario. A further level of conservatism is factored in by allowing for a declining steel price in our analysis (down to an HRC price of US$600/t), which reflects the expectation of continuing steel price declines on the back of sustained "oversupply" in steel. On the other hand, were the steel price to stay higher than this, the effect would be to further improve the profitability (and growth) of the Chinese steel industry.
Exhibit 252 A Complicating Factor in the Analysis Steel
Elasticity Is the Effect of Chinese Monetary Stimulus in 2009
Exhibit 253 There Is a Strong Relationship Between Lagged Steel Growth and the Growth in the Supply of Money...
Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis. Exhibit 254 ...We Strip This Effect Out of the Analysis of Steel Demand Growth...
Source: Bloomberg L.P. and Bernstein analysis.
0%
5%
10%
15%
20%
25%
30%
35%
Jan
-07
Jun
-07
No
v-07
Ap
r-08
Sep
-08
Feb
-09
Jul-
09
Dec
-09
May
-10
Oct
-10
Mar
-11
Aug
-11
Jan
-12
Jun
-12
No
v-12
M2
Yo
Y S
up
ply
Gro
wth
Chinese Money Supply
0%
5%
10%
15%
20%
25%
0%
5%
10%
15%
20%
25%
30%
35%
Jan
-09
Ap
r-09
Jul-
09
Oct
-09
Jan
-10
Ap
r-10
Jul-
10
Oct
-10
Jan
-11
Ap
r-11
Jul-
11
Oct
-11
Ste
el
M2
M2 and Lagged Steel
M2 Growth Steel Growth
R² = 0.5537
0%
5%
10%
15%
20%
25%
0% 10% 20% 30% 40%
Lag
ged
Ste
el G
row
th
M2 Supply Growth
Chinese Money Supply vs. Steel Growth
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
164 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 255 ...The Effect of This Would Be to Lower the Level of Steel Production for Any Given Profitability...
Source: Bloomberg L.P., AME, CRU, IMF and Bernstein estimates and analysis. Exhibit 256 ...Which Increases the Strength of the Relationship
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 165
Impact on Our Coverage Valuation Accepting that the Chinese government is industrializing China as fast as it can, thanks to demographic pressures (and from the equivalent of a capital stock base akin to that of the U.S. in the 1930s), and seeing the margin compression that has occurred in the steel industry, as iron ore prices (and other inputs) have risen over the last decade, mean accepting that falling iron ore prices would lead to increased steel growth. In a world where iron ore falls to $80/t, we estimate this reduced cost pressure would increase steel trend growth rates by 3% — an increase in demand, which, we estimate, would in fact push iron ore price back up to $95/t (see Exhibit 257 and Exhibit 258).
Companies with the highest iron ore exposure would of course benefit the most; we estimate the offset for Vale would be the equivalent of 31% of its June 30, 2013 share price, Rio 24%, Anglo 16%, and BHP 10% (see Exhibit 259).
Exhibit 257 It Is Inconsistent to Simultaneously Believe That Iron Ore Prices Would Fall as a
Consequence of Low Chinese Steel Growth and That China Is Undergoing a Period of Ongoing Capital Stock Accumulation; Consequently, We Believe That Falling Iron Ore Prices Would Lead to Increased Steel Growth of an Additional 3% Impact on Trend Steel Growth Rates, Even Given a Declining Steel Price
Source: Bloomberg L.P. and Bernstein estimates and analysis.
Average 2011 to Present Forecast at 80US$/t Fe
Steel Price - $/t 673 600
Iron Ore Price - $/t 148 80
Iron Ore Cost - $/t 237 128
Fuel Costs - $/t 172 172
Other Costs - $/t 152 152
Total Costs - $/t 561 452
Margin - $/t 112 148
Margin - % 16.7% 24.7%
Best Fit Steel Growth Rate - % 8.1% 10.9%
Observed Steel Growth Rate - % 8.0% N/A
Impact of Stimulus - % -1.9% -1.9%
Trend Growth Rate - % 6.3% 9.0%
Impact of Lower Iron Ore Prices - % 2.7%
Compared to Consensus Expectations of a Long-Term Iron Ore Price of US$80/t (Which Is Already Being Discounted by Mining Equities), We Estimate That This Automatic Stabilizer Could Provide an Additional US$15/t of Support Leading to an Equilibrium Demand-Supply Situation Closer to US$95/t
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
166 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 258 The Impact of the Additional Steel Demand Helping Offset the Impact of "Oversupply" from New Projects Could Support a Higher Price Level Than Would Otherwise Be Achieved
Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 259 The Impact of Incorporating This Effect Could See Iron Ore Prices Stabilize Some
15US$/t Higher Than Would Be the Case If the Elasticity of Demand Is Ignored
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
0
20
40
60
80
100
120
140
1600 50 100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1,00
0
1,05
0
1,10
0
1,15
0
1,20
0
1,25
0
1,30
0
1,35
0
1,40
0
1,45
0
1,50
0
1,55
0
1,60
0
Iro
n O
re (U
S$/
t)
Current and Future Iron Ore Cost Curves
Current Cost Curve Future Cost Curve
Price equillibrium higher as a conseuence of incorporating
demand elasticity
Price equilibrium higher as a conseuence of incorporating
demand elasticity
Lowering priceincreases trend
demand
Elasticity Impact Rio BHP Vale Anglo
Average $ Cost 32.7 42.5 40.0 49.1
Margin Uplift per Ton ($) 14.9 14.9 14.9 14.9
Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137
Current EV/EBITDA 5.0 5.8 4.6 5.0
EV Impact ($m) 18,255 13,893 20,680 3,689
Per Share Impact ($) 9.89 2.61 4.04 2.89
June 30, 2013 Price ($) 40.38 25.46 12.90 18.40
Per Share Impact (%) 24% 10% 31% 16%
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 167
Valuation and Risks
We provide two valuation metrics: As mining companies are operationally and financially geared to their underlying
commodity baskets (~80% of weekly equity price moves can be explained by moves in the underlying commodity prices), we use a regression-based trading model and our forward-looking commodity price forecasts to help determine our 12-month target prices (and to generate six monthly price forecasts). To the extent that the regression holds, and the parameters of the regression have not significantly shifted, we take the target price from the trading model. To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Note that we round final target prices in 25p/cent increments. Exhibit 260 summarizes our ratings and target prices as of June 30, 2013.
In addition to the target price (and short-term price forecasts generated by our trading model), we provide a supplementary valuation based on DCF. Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). Our DCF model constructed in nominal local currency terms out to 2030, over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash-flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (GBP or Real) at an average exchange rate, and round final DCF values in 25p/cent increments.
Investing in the Miners Is Always a Call on the Future Prices of Their Commodity Exposure
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
168 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
The Trading Model Sets 12-Month Target Prices The value of the miners is essentially a call on the future commodity price deck; each of the miners shows a very strong relationship to their underlying commodity baskets. Our regression-based trading model looks at the relationship between a company's historical share price performance and its underlying commodity exposure. It fits the best estimate relationship between the mining company valuation and its commodity basket historical performance, and then rolls that relationship forward over time. We do this to capture the fact that the parameterization of the fit between the miners and the commodity is not fixed.
Regression-based trading models must be taken, of course, with a grain of salt. A strong regression or R-squared fit between two variables can of course be preserved, while changing the parameterization of that fit. In other words, that a relationship exists may always be true, but the nature of that relationship may change over time. It is the changing nature of the relationship between a mining equity and its underlying commodity basket that expresses the de-rating or re-rating of a sector.
Accordingly, we monitor how the results generated by this model change over time. Does the best-fit change (and if so, for one stock or multiple stocks)? And is the equity deviating from its historical relationship (or is the R-squared changing over time)? To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the precise adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Exhibit 261 through Exhibit 280 summarize our ratings and target prices as of June 30, 2013.43
43 In order to generate a regression analysis for the new Glencore Xstrata entity, we calculate a synthetic price for the entity based on the trading
history of Xstrata and the relationship that existed between Glencore and Xstrata post the Glencore listing but before the Xstrata offer. Clearly, this is a less reliable basis than for the other miners but we believe that it continues to have some validity.
Our Trading Model Starts With Regression Analysis — But We Caution That One Must Interpret the Regressions Before Setting Target Prices
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 169
Exhibit 260 Summary of Our Coverage
Source: Bloomberg L.P., FactSet and Bernstein estimates and analysis.
BHP’s regression has been tight, not only in our most recent model update conducted June 5, 2013, but in our previous trading model updates. The R-squared of the most recent update is 78% (see Exhibit 261 and Exhibit 262). At the time of our update, BHP was moderately overvalued relative to its best-fit regression (see Exhibit 263). We find this unsurprising given the faltering market sentiment and the perception (justified in our view) of BHP as the highest quality stock in our coverage. Our trading model generates a 12-month target price of £22.50 (see Exhibit 264). We have not seen signs of the regression shifting nor is there anything in our fundamental research to date that causes us to anticipate it will; hence, we set our target price from the unmodified regression model price.
7/12/2013 Anglo BHPB Rio Tinto Vale
£ £ £ BRL
Price 12.95 18.00 27.99 30.32
Price Target 20.25 22.50 41.25 46.50
Potential Up/Downside 56% 25% 47% 53%
EV/EBITDA ‐ Current 5.09 5.67 4.89 4.30
EV/EBITDA ‐ 5 Yr. Avg. 4.9 5.5 4.8 5.0
EV/EBITDA Target 8.4 8.6 9.8 8.2
PE ‐ Current 8.9 9.3 7.1 5.5
PE ‐ 5 Yr. Avg. 9.2 9.4 8.0 7.6
PE Target 16.1 14.0 16.8 21.7
2013 EPS ‐ Consensus 2.11 2.47 5.65 2.48
2013 EPS ‐ SCB 1.90 2.70 5.33 2.28
% SCB vs. Consensus (9.8%) 9.4% (5.6%) (8.2%)
2014 EPS ‐ Consensus 2.63 2.84 6.44 2.89
2014 EPS ‐ SCB 2.57 3.73 7.34 2.55
% SCB vs. Consensus (2.2%) 31.3% 14.0% (11.8%)
*Note all EPS figures are in reporting currency (USD)
BHP: TP £22.50 (Unmodified from Trading Model Prediction)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
170 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 261 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong...
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 262 ...And Explains the Majority of the Historical Performance
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 171
Exhibit 263 BHP Is Currently Slightly Overvalued Relative to Its Historical Best Fit, an Unsurprising Development in the Current Tremulous Market Environment
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 264 We See Considerable Upside in the Share Price
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Like BHP, Rio Tinto’s regression has also been tight in both our historical and most recent trading model update (June 5, 2013). The R-squared of the most recent update is 88% (see Exhibit 265 and Exhibit 266). At the time of our update, Rio was slightly overvalued relative to its best-fit regression (see Exhibit 267), but by less than 5%. We consider this unsurprising, given the recent slides in both iron ore prices and Rio Tinto’s share price (bearing in mind that iron ore generated ~90% of Rio Tinto’s 2012 EBIT). Our trading model generates a 12-month target price of £41.25 (see Exhibit 268). We have not seen signs of the regression shifting nor is
Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP
Rio Tinto: TP £41.25 (Unmodified from Trading Model Prediction)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
172 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
there anything in our fundamental research to date that causes us to anticipate it will; hence, we set our target price from the unmodified regression model price.
Exhibit 265 The Relationship Between the Miners and Their Underlying Commodity Exposure Is
Incredibly Strong…
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 266 …And Explains the Majority of the Historical Performance
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 173
Exhibit 267 Rio Is Currently Slightly Overvalued Relative to Its Historical Best Fit, But Less Than 5%, and Hence Less Than High-Quality BHP
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 268 We See Considerable Upside in the Share Price
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
In our previous trading model update, we noted that Anglo American’s equity price had declined from fairly valued at the start of the year relative to its commodity basket to 12% below the basket’s spot price (see European Metals & Mining: Timing Is Everything. Predictions From Our Short-Term Trading Model).
At the time we commented that whether the deviation was temporary or resulted in a permanent de-rating would depend, in our view, upon the resolution of the company's operating disappointments and South African risk. With our June 5 update, the regression shifted and the new fit has a tight fit and an R-squared of
Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP
Anglo American: TP £20.25 (Unmodified from Trading Model Prediction)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
174 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
85% (see Exhibit 269 and Exhibit 270); however, the stock is still slightly undervalued relative to its best-fit regression (see Exhibit 271). Our trading model generates a 12-month price target of £20.25, which we have taken unmodified (see Exhibit 272). Given the shift in the regression and the fundamental factors that have been weighing on the stock, the upcoming 1H reporting (at which time Mr. Cutifani will present the results of his portfolio review), and that we consider Anglo a restructuring story, we consider there is a greater likelihood of regression instability going forward than for Rio or BHP. We are monitoring the situation.
Exhibit 269 The Relationship Between the Miners and Their Underlying Commodity Exposure Is
Incredibly Strong…
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 270 …And Explains the Majority of the Historical Performance
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 175
Exhibit 271 Anglo Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But We Note the Regression Has Shown Instability and a Decision to Invest in the Stock Is Contingent Upon a Belief in the Fundamental Restructuring Story as Well as Commodity Prices
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 272 We See Considerable Upside in the Share Price
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
176 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Vale continues to show instability in its regression relationship. This was something we had noted in our previous trading model update on May 16, 2013 as well (see European Metals & Mining: Timing Is Everything. Predictions From Our Short-Term Trading Model). At that time, we commented that the equity had previously been more than 20% undervalued at the start of the year; however, our then-updated regression found the best fit at 3% undervalued in January, and yielded a 3% discount to the May 16, 2013 price of its spot basket. Our fundamental concerns about this company cause us to question whether this regression shift may be indicative of a de-rating.
Our most recent regression analysis finds best fit at 5% overvalued in January 2013. The R-squared of the most recent update is 86% (see Exhibit 273 and Exhibit 274). At the time of our update, Vale was 8% undervalued relative to its best-fit regression (see Exhibit 275). As with Anglo American, we believe there are fundamental factors at play here. Though not a restructuring story like Anglo, this essentially pure-play iron ore company is subject to significant influence by the Brazilian government (through direct ownership of Golden Shares as well as ownership by the strategic consortium of Valepar). Furthermore, the company is, in our view, most at risk should a reduced iron ore price eventuate, given its geographic longinquity from the world’s largest consumer of iron ore, China. As a result, we consider that there is a greater likelihood of regression instability going forward than for Rio or BHP and are monitoring the situation. However, for now, we have taken the unmodified target price generated by the trading model of BRL 46.50 (see Exhibit 276).
Exhibit 273 The Relationship Between the Miners and Their Underlying Commodity Exposure Is
Incredibly Strong…
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
Vale: TP BRL 46.50 (Unmodified from Trading Model Prediction)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 177
Exhibit 274 …And Explains the Majority of the Historical Performance
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 275 Vale Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But as With
Anglo, We Caution That the Regression Has Been Unstable and That There Are Risks Not Captured by It (Specifically the Significant Influence of the Brazilian Government)
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
178 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 276 We See Considerable Upside in the Share Price
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
For Glencore Xstrata, we have constructed a synthetic stock using Xstrata’s actual history and its recent relationship with Glencore. This synthetic stock shows, in our view, an acceptably high R-squared of 64% (see Exhibit 277 through Exhibit 279). We note that the difficulties of using this model and the synthetic stock for setting a target price include not only the short duration of Glencore’s trading history (having IPO'd only in May 2011) but also the distortive effects that the merger between the two companies had on their respective share prices. Furthermore, the marketing arm of Glencore (93% of 2012 revenues but 42% of operating profit that year) adds a layer of differentiation relative to the "pure" miners in our coverage. For now, we have taken the unmodified target price of £5.25 but will continue to monitor the trading model (see Exhibit 280).
0
10
20
30
40
50
60
70
80
Sh
are
Pri
ce
(£)
Vale Share Price Forecast
Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP
Glencore: TP £5.25 (Unmodified from Trading Model Prediction)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 179
Exhibit 277 For Glencore Xstrata, the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 278 And in Glencore Xstrata (Though for Glencore Xstrata the Historical Relationship Is
Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore)
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
nGlencore Xstrata Share Price vs. Best-Fit Regression
Undervalued Relative to Commodity Basket
Overvalued Relative to Commodity Basket
2
3
4
5
6
7
8
9
10
May -11 Nov -11 May -12 Nov -12 May -13 Nov -13 May -14 Nov -14 May -15
Sh
are
Pri
ce
(£)
Glencore Xstrata Share Price Forecast
Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 181
DCF Generates an Intrinsic Value Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). This is not only because we believe a DCF captures the future cash flows (and after all, what is the value of a mining company and what cash flows can be generated by its assets) but also because it forces us to consider the fundamental factors impacting the company's assets (on an asset-by-asset basis where possible, division-by-division where not) and exercise discipline — particularly important in a sector whose equity returns at least in the near term can be highly trading oriented.
Our DCF model constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual LOM may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (GBP or real) at an average exchange rate, and round final DCF values in 25p/cent increments.
In our call European Metals & Mining: Changing Price Targets and Upgrading Vale to Outperform, we discussed how the DCF overstates the sensitivity of mining stock prices to changes in commodity price assumptions (by a factor of 3-6x). This is in part due to the different time frames considered by market participants (ranging from a matter of days to perhaps a decade at the outside) and the DCF (which discretely models cash flows out to 2030). However, other confounding factors may be: A standard DCF model ignores the covariance between commodity prices
themselves and between commodity prices and the drivers of mining operating costs (such as FX and diesel). Such covariance between cost and price has the effect of desensitizing value to price changes.
The sensitivity analysis that is conducted using a DCF model ignores the degree of mean reversion in commodity prices. Again, this effect desensitizes value to price changes; in the extreme of instantaneous and mean reversion to a fixed long-term price, there would be no relationship between prices and mining valuations. It is worth noting that this effect applies that even if a backwardated price is used to mimic mean reversion, a backwardated curve that moves varies by a fixed quantum would show exactly the same behavior (as far as sensitivity is concerned) as a flat or contangoed price line.
Market misevaluation of risk. The final explanation is that the market places too high a discount rate on future earnings and too low a discount rate on current earnings. This has the same effect of increasing the apparent degree of commodity price backwardation in the miners and would reduce the sensitivity of the equity to changes in the commodity price basket.
DCF Generates an Intrinsic Value for the Stock Today, But Is Silent on When the Stock May Achieve That Value
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
182 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 281 BHP DCF Value
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 282 Rio Tinto DCF Value
Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.
US$m DCF Terminal Value Total Per Share (US$) Per Share (£) 2013 EBITDA EV/EBITDA
Petroleum 90,058 - 90,058 16.92 11.10 11,592 7.8
Aluminum 5,089 942 6,032 1.13 .74 165 36.5
Base Metals 39,562 8,147 47,709 8.97 5.88 4,689 10.2
Glencore Enterprise Value 118,270 17,006 135,276 10.15 6.33 14,058
Less Minorities (6,998) (.53) (.33)
Less Central Costs - - -
Less (Net Debt)/Net Cash (34,538) (2.59) (1.62)
Plus Corporate & Other
Equity Value 93,739 7.03 58,441.02
Shares Outstanding 13,326
Per Share NPV (USD) 7.03
Exchange Rate 1.60
Per Share NPV (GBP) 4.39
Without Synergies
Glencore Enterprise Value 114,029 16,246 130,275 9.78 6.41 13,308
Less Minorities (6,998) (.53) (.34)
Less Central Costs - - -
Less (Net Debt)/Net Cash (34,538) (2.59) (1.70)
Plus Corporate & Other
Equity Value 88,739 6.66 58,216.90
Shares Outstanding 13,326
Per Share NPV (USD) 6.66
Exchange Rate 1.52
Per Share NPV (GBP) 4.37
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 185
Risks to Our Sector and Stocks The four most significant risks facing the major mining houses in Europe are lack of capital discipline, operating cost inflation, a sustained downturn in the Chinese economy and resource nationalism. Capital discipline: We believe that mined commodity prices will stay high and
will continue to trend higher until such point that the massive amounts of labor currently employed in the Chinese mining industry are displaced by capital. This can happen either through a reform of the domestic mining industry in China or through the displacement of that industry by supply increases in other geographies. We do not believe that the natural resource endowment of China will allow for a rapid (if any) domestic reform, as this requires the existence of massive high-grade, long-life deposits (such as copper in Chile or iron ore in Brazil and Australia). Consequently, the duration of the current pricing environment comes down to the extent to which the Western capital deployed by the major mining houses to increase low-cost commodity production will displace the requirement for Chinese domestic production.
Operating cost inflation: U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years. Part of this can be explained by movements in exchange rates and part by the prevailing inflationary environment in producer geographies. However, there has still been a very significant increase in underlying real costs. Should this persist or accelerate, then it has the capacity to erode value.
Chinese economic risks: China is important in commodities as both the major source of demand growth and as the location of the marginal units of supply. We believe that the current slowdown in the Chinese economy has been largely self-induced in an effort to contain food inflation and, as a consequence, a political stasis ahead of the handover of power at the end of the year. However, a more long-lived slowdown has the capacity to move the commodity markets into oversupply on a sustained basis — particularly if new supply is not curtailed.
Resource nationalism: Finally, we note with concern the trend toward global fragmentation and the ever greater desire to extract value from the mining sector. We believe that this is ultimately a self-defeating strategy by host governments but it is one with an impressively long pedigree. Persistent macroeconomic headwinds will make this an ever more attractive option.
Rio Tinto PLC: Company-specific risks include any sustained downturn in iron ore prices, as the company is the second most exposed of our coverage group to iron ore (after Vale). Any relaxation of capital discipline particularly around the Simandou project in Guinea would also be, in our view, a negative catalyst. Further execution delays in the commissioning of the Oyu Tolgoi copper project in Mongolia or significant revenue grabs from the Mongolian government could also be a risk.
BHP Billiton PLC: Company-specific risks include continued weakness in the price of natural gas in the U.S. and iron ore. Repeats of the weather-induced volume losses in BHP’s metallurgical coal operations as well as continued labor-related disruptions in these assets could also prove an impediment to our target price.
Vale SA: Company-specific risks include any sustained downturn in iron ore price, as the company derives nearly the entirety of its value from exposure to iron ore. The continuation of disruptions to the output of nickel and attendant cost pressures are also a risk. The commissioning of Goro (VNC) is an issue that needs to be addressed, as is the performance at Onca Puma.
Anglo American PLC: For Anglo American in particular, inability to improve the efficiency of its platinum operations and continued margin pressure arising from South African labor inflation poses downside risk, as does the potential for increased union militancy in South Africa (and again in platinum, in particular). A continued deterioration in labor unrest along with the attendant physical hazards, delays and expenses could weigh on results. Further delays at the Minas Rio iron
Risks
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
186 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
ore project in Brazil would also be a significant negative catalyst. Failure to properly integrate De Beers into the Anglo American portfolio could again risk value loss.
Glencore Xstrata PLC: The greatest unknown, in our minds, for Glencore relates to its marketing activities. We believe there is insufficient transparency to assess both embedded risk in the trading book and persistence of edge. We also note that Glencore requires high levels of working capital and remains vulnerable to large swings in cash-flow generation as a result. We note as a result of its operations in frontier jurisdictions, as well as the unknown nature of embedded risk and persistence of edge in the marketing book, headline risk remains a significant concern for Glencore. Lastly, we do not consider Glencore yet "institutional quality" in terms of its transparency or governance.
Post merger, we see challenges facing the combined Glencore-Xstrata entity, specifically integration of Xstrata's significantly larger operating business into Glencore's management structures while avoiding disruption to a number of critical projects, particularly given the anticipated dismissal of many Xstrata personnel. Furthermore, the choice for a new chairman will be critical to ensuring that the corporate governance and minority shareholder interests are well protected as Glencore enters a new stage in its life as one of the largest publicly listed mining companies in the world.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 187
Company Impact
Iron ore is the single most important commodity in the mining space, representing 58% of EBITDA generation (see Exhibit 286). Furthermore, it has been one of the strongest performing commodities throughout the mining "super-cycle" (see Exhibit 287), having risen by 1,340% from its 2001 average price to its March 2011 peak.44 Consequently, having a view on this commodity is essential for understanding the bulk of the value contained within and generated by our coverage group (see Exhibit 288).45
Although it represents the lion's share of our coverage EBITDA generation, iron ore's importance to each individual miner varies substantially across companies. Vale and Rio Tinto have the largest exposure, with iron ore contributing 97% and 75% to their 2012 EBITDA, respectively. As the most diversified miners, Anglo American and BHP Billiton saw 37% and 55% of EBITDA, respectively, coming from the cold metal. Finally, mining operations of Glencore Xstrata have no direct production exposure to iron ore bar the 1Q:13 acquisition of a minority stake in Ferrous Resources.
This chapter is divided into three sections: highly concentrated iron ore plays (Rio and Vale), diversified miners with moderate exposure (Anglo American and BHP Billiton) and companies with no direct production exposure (Glencore Xstrata).
Exhibit 286 Iron Ore Represents, by Far, the Most Significant Contribution to the Cash Generation of the Miners…
Exhibit 287 …And Has Been, Along With Copper, the Strongest Performer of the Major Commodities
Source: Corporate reports and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.
44 The IMF 62% Fe monthly spot (CFR Tianjin port), 2001 average = US$13/t, March 2011 = US$187/t. 45 Glencore Xstrata is the only company in our coverage without any direct production exposure to iron ore.
58%
18%
8%
15%2012 EBITDA Breakdown
Iron Ore Base Metals Coal Other
0
100
200
300
400
500
600
700
800
900
1000
Jun
e 2
000
= 10
0
Indexed Commodity Price Performance
Iron Ore Copper HCC
Iron Ore Currently Accounts for ~60% of the EBITDA of the Five Major Mining Houses; It Has Been the Outperforming Commodity in the Last 10 Years
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
188 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 288 Vale Represents Essentially an Iron Ore Pure Play in Both the Near and Longer Term, While We See Rio Diversifying in the Medium Term; BHP and Anglo American Have Exposure to a Well-Diversified Basket of Commodities; Glencore Xstrata Is the Only Company in Our Coverage Without Any Direct Production Exposure to Iron Ore*
* Excluding the 1Q:13 acquisition of a minority stake in Ferrous Resources.
Source: Corporate reports and Bernstein analysis.
Vale and Rio Tinto — Highly Concentrated Iron Ore Plays In our coverage group, Vale and Rio Tinto are the most heavily exposed to iron ore (97% and 93% of 2012 EBIT, respectively).
We consider Rio Tinto the most attractive stock in our coverage. Not only does the company have significant iron ore exposure (93% of 2012 EBIT), it does not suffer from the same risks — quantifiable and unquantifiable — that we see impacting Vale (which generated 97% of its 2012 EBIT from iron ore). Exhibit 289 and Exhibit 290 illustrate the 2012 breakdown of Rio Tinto's and Vale's revenues and operating profits by commodity. Exhibit 291 through Exhibit 294 show a historical evolution of revenues and EBITDA by commodity since 2001.
0%
20%
40%
60%
80%
100%
120%
Anglo BHP Rio Vale Glencore Xstrata
Contribution of Iron Ore to EBITDA (2012)
Iron Ore Other
Both Rio and Vale Derive the Bulk of Their Earnings (93% and 97% of 2012 EBIT, Respectively) from Iron Ore; We Prefer Rio, Which Also Benefits from World-Class Copper Exposure
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 189
Exhibit 289 Vale Is the Most Heavily Exposed to Iron Ore of Our Coverage Group (a Trend We Do Not See Changing)…
Exhibit 290 …Followed by Rio Tinto, Which Is Augmenting Its Portfolio With Some of the World's Best Copper Projects (the Metal on Which We Are the Most Bullish)
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
Exhibit 291 Historical Evolution of Rio Tinto's Revenue by Business Unit
Source: Corporate reports and Bernstein analysis.
70%
-4%
1%3%
8%
4%
15%
97%
2%1%
100%
EBIT
3%
0% 2%
-2%
Group Revenue
Others
Manganese & Ferroalloys
Coal
Logistics
Fertilizers
Base Metals
Iron Ore
Vale
7%
-7%
10%
9%4%
12%
7%
100%
EBIT
93%
Group Revenue
3%
0%
18%
44%Other
Diamonds & Minerals
Energy
Copper
Aluminium
Iron Ore
Rio Tinto
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
US
$m
Revenue by Business Unit
Iron Ore Aluminum Copper Energy Diamonds & Minerals Other
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
190 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 292 Historical Evolution of Rio Tinto's EBITDA by Business Unit
Source: Corporate reports and Bernstein analysis.
Exhibit 293 Historical Evolution of Vale's Revenue by Business Unit
Source: Corporate reports and Bernstein analysis.
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
US
$mEBITDA by Business Unit
Iron Ore Aluminum Copper Energy Diamonds & Minerals Other
Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 191
Exhibit 294 Historical Evolution of Vale's EBITDA by Business Unit
Source: Corporate reports and Bernstein analysis.
Vale is the world's largest iron ore producer and, in Carajas, has one of the
world's most attractive iron ore assets. However, it is the most operationally geared miner to the iron ore price (partially as a consequence of geographical distance to China, partially as a consequence of the operating cost position of some of the Southern mines and the pellet operations, and partially as a consequence of its significant capital spend). Consequently, any change in the long-term price of iron ore has a huge impact on the DCF value of Vale. Moreover, we also believe that there is more to the story of iron ore prices than simple monotonic declines to some low flat price. Consequently, there is still an asymmetric risk to the upside for iron ore prices in the short term. So from a pure value consideration, we are more favorably inclined to Vale now than previously. Moreover, it has a diversification strategy, which is beginning to move the company away from being a single-commodity pure play and most importantly increasing the company's exposure to copper with Salobo and Lubambe.
That said, we still favor Rio Tinto for those wishing to gain iron ore exposure. Rio Tinto owns some of the highest-quality iron ore assets and (vitally) infrastructure globally. Rio has offered a slightly better "bang for your iron ore buck" on correlation and absolute return bases — at less than half the capex per incremental ton of production and without the risk of significant involvement by the Brazilian government. Rio also has exposure to some of the world's best operational copper assets — not to mention three of the world's best undeveloped copper deposits. Tier 1 assets in iron ore include Dampier and Cape Lambert (Australian Pilbara), while copper includes Escondida, Grasberg and Bingham Canyon. The company also has some of the most promising and largest copper prospects in Oyu Tolgoi, Resolution and La Granja.
Rio has commodity diversification into copper and mineral sand, rather than nickel, and has demonstrated an ability to deliver iron ore growth at considerably lower capital intensity than Vale. That, coupled with the prospects of genuine capital discipline and cost cutting under new CEO Sam Walsh (who made his bones in low-cost brownfield Pilbara production) and a more reasoned approach to volume growth, means that we continue to see Rio Tinto as our top pick. Exhibit 295 through Exhibit 298 show our revenue and EBITDA forecasts for Rio Tinto and Vale.
Iron Ore Aluminum Copper Energy Diamonds & Minerals Other
(10,000)
-
10,000
20,000
30,000
40,000
50,000
EBIT
DA
(U
S$m
)
Rio Tinto EBITDA
Iron Ore Aluminum Copper Energy Diamonds & Minerals Other
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
2011
2012
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
Re
ven
ue
(US
$m)
Vale Revenue
Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics Others
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,00020
11
2012
2013
E
2014
E
2015
E
2016
E
2017
E
2018
E
EBIT
DA
(U
S$m
)
Vale EBITDA
Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics Others
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 193
BHP Billiton and Anglo American — The Diversified Miners BHP Billiton not only has the longest corporate history of our coverage group (its component parts trace their origins back to 1851 and 1883, respectively), we view it as the highest quality stock in our coverage. In 2012, the company generated the highest revenue of the "Big Three" miners in our coverage (US$72 billion versus US$56 billion for Rio and US$49 billion for Vale) and had the highest EBITDA margin (42% versus 39% for Vale and 36% for Rio).
BHP Billiton has a well-diversified portfolio from a commodity exposure perspective. Lagging only Anglo American, it is the second most diversified company in our coverage (see Exhibit 299 through Exhibit 302). Both Anglo and BHP have exposure to a commodity that behaves differently from either the ferrous or the base metals. For Anglo, it is platinum, and for BHP Billion, it is petroleum. However, where platinum has been a drag on Anglo's portfolio, BHP has received a boost from "black gold" related in its petroleum division (69% EBITDA margins in calendar 2012 versus 11% for Anglo's platinum division) — Exhibit 303 and Exhibit 304. We attribute BHP's high margins relative to peers to the Tier 1 nature of the company’s assets and the operational excellence with which they are run. Consequently, it is BHP Billiton that is the most diversified miner from an earnings perspective.
The strong commodity risk diversification of BHP Billiton is complemented with geographic diversification that, while skewed to Australia, is nonetheless low risk. A significant portion of BHP's assets is located either in Australia (iron ore, coal and petroleum), North America (petroleum) or Chile (copper). There is limited emerging market asset exposure and, consequently, limited political risk in the portfolio (see Exhibit 300). That said, we have noted with concern the deterioration in the attractiveness of Australia as a location for new mining investment over the last few years. We believe that this provides a reminder that no country is ever entirely "safe." Nevertheless, we continue to see significant revenue and EBITDA growth for "The Big Australian" (see Exhibit 305 and Exhibit 306).
Exhibit 299 The Diversity in Exposure Makes BHP One of the Lowest-Risk Miners from a Commodity Perspective…
Exhibit 300 …Complemented by Low Political Risk and a Diverse Range of Operating Geographies
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
7%
9%
19%
26%
19%
20%
29%
52%
0%
3%
0%
4%
0%
8%
3%
100%
EBIT
-2%1%
Group Revenue
1%
Diamonds
Manganese
Stainless Steel Materials
Aluminum
Energy Coal
Metallurgical Coal
Petroleum
Base Metals
Iron Ore
BHP Billiton
55%
0%
15%
13%
8%1%
8%
Net Assets by Location
Australia UK North America South America
Southern Africa Rest of World Unallocated
BHP Billiton Is the Highest Quality Mining Stock in Our Coverage, With Genuine Tier 1 Assets (e.g., Escondida, Mt Newman, Olympic Dam and Peak Downs) Diversified Across a Range of Commodities and With Limited Geographic Risk
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
194 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 301 Historical Evolution of BHP's Revenue by Business Unit
Source: Corporate reports and Bernstein analysis.
Exhibit 302 Historical Evolution of BHP's EBITDA by Business Unit
Source: Corporate reports and Bernstein analysis.
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
Re
ven
ue
(US
$m)
Revenue by Business Unit
Petroleum Aluminum Base Metals Diamonds & Specialty ProductsStainless Steel Materials Iron Ore Manganese Metallurgical CoalEnergy Coal Other Group & Inter Segment
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
EB
ITD
A (U
S$m
)
EBITDA by Business Unit
Petroleum Aluminum Base Metals
Diamonds & Specialty Products Stainless Steel Materials Iron Ore
Manganese Metallurgical Coal Energy Coal
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 195
Exhibit 303 BHP's Petroleum Exposure Is Its Great Diversifier (Generating 69% EBITDA Margins in 2012)…
Exhibit 304 …While Anglo American's Great Diversifier, Platinum, Came in at Margins of Just 11%
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
Exhibit 305 We Also See Revenue and EBITDA Growth…
Exhibit 306 …For BHP Billiton
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
Anglo American has the most diversified portfolio of our coverage group from a commodity perspective and is significantly smaller than the "Big Three" of BHP, Rio and Vale. There is a distinct reduction in scale when stepping down from this triumvirate of the "Big Three" miners to Glencore Xstrata and Anglo (see Exhibit 307).46 While it is tempting to talk of the five largest global miners as peers, we believe that the gulf between them is now so large as to make that unconvincing.
The reason for the emergence of the "Big Three" and for the relative ranking in EBITDA margin performance is attributable to iron ore. Those miners that have
46 We believe that Glencore Xstrata would need to merge with Anglo American to create a competitor to the "Big Three."
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
-
20
40
60
80
100
120
140D
ec-0
1D
ec-0
2D
ec-0
3D
ec-0
4D
ec-0
5D
ec-0
6D
ec-0
7D
ec-0
8D
ec-0
9D
ec-1
0D
ec-1
1D
ec-1
2
US
$m
Pe
tro
leum
Pro
duct
ion
(mm
boe)
Petroleum
Petroleum Production (LHS) Petroleum Revenue (RHS)
Manganese Metallurgical CoalEnergy Coal Other Group & Inter Segment
(10,000)
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
2011 A
2012 A
2013 E
2014 E
2015 E
2016 E
2017 E
2018 E
EBIT
DA
(U
S$m
)
BHP Billiton EBITDA
Petroleum Aluminum
Base Metals Diamonds & Specialty Products
Stainless Steel Materials Iron Ore
Manganese Metallurgical Coal
Energy Coal
Anglo American Has the Most Diversified Portfolio in Our Coverage — It Remains a Turnaround Story in Our Minds
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
196 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
exposure to this commodity enjoy the highest margins and have seen the fastest revenue growth. To the extent that the iron ore margins are diluted by other less well-performing commodities, the lower the overall EBITDA margin will be (see Exhibit 308). In general, Anglo benefits from having a well-diversified portfolio of different commodity exposures (see Exhibit 309, Exhibit 311 and Exhibit 312). In this regard, it is probably the most diversified of the companies in our coverage and has the lowest pure commodity risk. This lower commodity risk is, however, partially offset by greater country risk. Anglo remains heavily exposed to South Africa through Kumba, Anglo Platinum and its thermal coal operations. We also see the company becoming increasingly reliant on South America for future earnings growth. Geographically, Anglo American lacks a really significant presence in Australia, with antipodean exposure confined to metallurgical coal (a key differentiator from BHP Billiton, the second most diversified of our coverage companies from a commodity perspective) — Exhibit 310.
The performance at Anglo Platinum is of particular concern to us, as the overall level of flat production belies a significant shift in the source of the material that stands behind this. Following the requirements of South African Black Economic Empowerment (BEE), we see significant material now coming via concentrate purchases from JV partners and associates (see Exhibit 314). Only ~75% of Anglo Platinum's refined output comes from its own mined material (see Exhibit 313). This change has a significant impact on margins, as we believe that Anglo Platinum purchases material for about 90% of the price of the contained metal. Given that refining costs must still be borne, we see purchases of material as being dilutive of the margins that can be achieved from mining.
Anglo remains in our minds a turnaround story — one that will challenge new CEO Mark Cutifani, but one which we believe is solvable for a leader with his 36 years of operational expertise. We are looking forward to the results of his strategic review.47 As an operationally experienced miner across a host of commodities and geographies, including deep precious metal mining in Africa, we believe he is well prepared to address the issues currently facing Anglo Platinum and the group as a whole. Going forward, we will be particularly interested to hear more from the company about the progress of its Platinum business restructuring and the development of Minas Rio, and we expect to hear more details in his portfolio review, to be presented at the upcoming half-yearly reporting. Exhibit 315 and Exhibit 316 show our revenue and EBITDA forecasts.
47 Anglo's upcoming reporting schedule is as follows: 2Q production reports for Anglo and listed subsidiaries Kumba and Anglo Platinum are scheduled for Thursday, July 18, 2013. Anglo Platinum's 1H earnings report is scheduled for Monday, July 22, 2013; Kumba's for Tuesday, July 23, 2013; and Anglo American's for Friday, July 26, 2013. All reporting times 7am BST.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 197
Exhibit 307 Anglo and Glencore Xstrata Represent the Tier 2 of Mining Companies Behind the "Big Three" of Vale, Rio and BHP
Exhibit 308 The Differentiating Feature of the "Big Three," Both in Terms of Size and Margin, Is the Presence of Significant Iron Ore Exposure
Note: Only the revenue and EBITDA for Glencore's Mining and Energy Industrial activities are included.
Source: Corporate reports and Bernstein analysis.
Note: Only the revenue and EBITDA for Glencore's Mining and Energy Industrial activities are included.
Source: Corporate reports and Bernstein analysis.
Exhibit 309 Anglo Is the Most Diversified and the Lowest-Risk Miner from a Commodity Exposure Perspective
Exhibit 310 The Relative Lack of "Safe" Australian Exposure Has Always Been a Risk Compared to the Other Miners
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
Re
ven
ue
(US
$m) D
ece
mbe
r Ye
ar E
nd
Sector Revenue and EBITDA
Revenue EBITDA
0%
5%
10%
15%
20%
25%
30%
35%
40%
Vale Rio Tinto BHP Billiton
Anglo Glencore Xstrata
Sector EBITDA Margin
12%
7%
12%
8%
11%
13%
0%
100%
EBIT
-3%-2%
27%
49%
Group Revenue
11%
17%
1%
16%
21%
Other
Platinum
Nickel
Metallurgical Coal
Diamonds
Thermal Coal
Copper
Iron Ore and Manganese
Anglo American
36%
1%46%
2%
10%
5%Net Assets by Location
South Africa Other Africa South America
North America Australasia Other
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
198 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 311 Historical Evolution of Anglo American's Revenue by Business Unit
Source: Corporate reports and Bernstein analysis.
Exhibit 312 Historical Evolution of Anglo American's EBITDA by Business Unit
Source: Corporate reports and Bernstein analysis.
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
US
$mRevenue by Business Unit
Platinum Diamonds Copper Nickel Iron Ore and Manganese Metallurgical Coal Thermal Coal Other Mining & Industrial Other
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A
US
$m
EBITDA by Business Unit
Platinum Diamonds Copper Nickel Iron Ore and Manganese Metallurgical Coal Thermal Coal Other Mining & Industrial Other
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 199
Exhibit 313 An Increasing Proportion of Anglo Platinum's Refined Production Now Comes from Non-Mined Sources…
Exhibit 314 …The Majority Is Purchased from Related Joint Venture Partners and Associates
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
Exhibit 315 Our Forecast for Anglo American Sees Revenue Growth…
Exhibit 316 …As Well as EBITDA Growth as Part of the Turnaround Story
Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.
0200400600800
1,0001,2001,4001,6001,800
Jun-
2001
Jun-
2002
Jun-
2003
Jun-
2004
Jun-
2005
Jun-
2006
Jun-
2007
Jun-
2008
Jun-
2009
Jun-
2010
Jun-
2011
Jun-
2012
Jun-
2013
ko
z P
t
Refined Platinum Production by Source
Series1 Series2
74%
11%
11%4%
Refined Pt by Source, 2012
Refined Pt from Mined Production Conc from JVs
Conc from Assocs Conc from Third Party
-
10,000
20,000
30,000
40,000
50,000
60,000
2010 A
2011 A
2012 A
2013 E
2014 E
2015 E
2016 E
2017 E
2018 E
Re
ven
ue
(US
$m)
Anglo American Revenue
Platinum Diamonds Copper
Nickel Iron Ore and Manganese Metallurgical Coal
Thermal Coal Other Mining & Industrial Other
-
5,000
10,000
15,000
20,000
25,000
2010 A
2011 A
2012 A
2013 E
2014 E
2015 E
2016 E
2017 E
2018 E
EBIT
DA
(U
S$m
)
Anglo American EBITDA
Platinum Diamonds Copper
Nickel Iron Ore and Manganese Metallurgical Coal
Thermal Coal Other Mining & Industrial Other
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
200 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Glencore Xstrata Has No Direct Iron Ore Production Exposure Glencore Xstrata offers exposure for copper and coal bulls without a taste for iron (given that the company's only direct iron ore exposure is its 3Q:13 acquisition of a stake in Ferrous Resources). Glencore Xstrata also offers exposure to the sales and trading house that CEO Ivan Glasenberg built. As such, it has a profile that is distinct relative to the pure miners in our coverage (see Exhibit 317 and Exhibit 318). The impact of the high-turnover, low-margin trading business (2% EBITDA margin in 2012 versus 25% for Glencore Xstrata combined) is even clearer when contextualized against the average 33% for our coverage group (see Exhibit 319 and Exhibit 320).
On the industrial-assets side, Glencore Xstrata is predominantly a copper and coal play, with zinc a tertiary significant exposure (see Exhibit 321 through Exhibit 323). The copper and coal assets from the erstwhile Xstrata portfolio contain some particularly compelling operations. In Latin American copper, Collahuasi, Antamina and Alumbrera are particularly attractive operating assets in our view, while in coal, we consider Okay Creek, Rolleston, Bulfa, Mangoola, Ulan (all Australia) and Cerrejon (Columbia) to be gems.
It is worth noting that, from a strategic perspective, we prefer pure mining sector plays to diversified natural resources companies (hence our positive view on the strategic repositioning that Anglo American has undertaken). To the extent that mining companies have energy — specifically, oil and gas — production in their portfolios and can run those at similar or better operating profit margins to the oil and gas majors, we are not unduly troubled by the inherent diversification from mining (despite the different skill sets and technical expertise required by the two disciplines). Agriculturals are, however, not an obvious fit to us, in terms of operational skill sets, time horizons for management, margins, etc. We understand Glencore’s historical reasons for developing this portion of its asset portfolio, growing as it did from a trading company. However, we remain concerned that the diversity of the group splits management's focus, diverting energy and attention away from the most attractive mining elements of the portfolio. We see this as a negative for the company.
Copper- and Coal-Heavy Xstrata Is the Miner for Sector Bulls Without a Taste for Iron — But One Has to Have An Appetite for Headline Risk
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 201
Exhibit 317 Glencore Xstrata's High-Revenue, Low-Margin Trading Business Renders Its Group Profile Very Different from the Pure Miners in Our Coverage…
Exhibit 318 …Though Excluding the Marketing Division Reveals a Profile That Is Similar to the Rest of the Miners
Source: Corporate reports and Bernstein estimates and analysis.
Note: Only the revenue and EBITDA for Glencore Xstrata's Industrial Activities are included.
Source: Corporate reports and Bernstein estimates and analysis.
Exhibit 319 The Impact of the Marketing Division Can Be Seen in Total Company Margins…
Exhibit 320 …While the Margins for Only Industrial Activities More Closely Resemble the Pure Miners in Our Coverage
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
0
50,000
100,000
150,000
200,000
250,000
Re
ven
ue
(US
$m) D
ece
mbe
r Ye
ar E
nd
Sector Revenue and EBITDA, 2012
Revenue EBITDA
010,00020,00030,00040,00050,00060,00070,00080,000
Rev
enu
e (U
S$m
) Dec
emb
er Y
ear E
nd
Sector Revenue and EBITDA, 2012
Revenue EBITDA
0%
5%
10%
15%
20%
25%
30%
35%
40%
Sector EBITDA Margin, 2012
0%
5%
10%
15%
20%
25%
30%
35%
40%
Vale Rio Tinto BHP Billiton
Anglo Glencore Xstrata
Industrial Activities
Sector EBITDA Margin, 2012
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
202 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 321 Historical Pro Forma Evolution of Glencore Xstrata's Industrial Activities EBITDA
Source: Corporate reports and Bernstein estimates and analysis.
Exhibit 322 Glencore Xstrata's Industrial Activities Are Predominantly a Copper, Coal and Zinc Play…
Exhibit 323 …As Can Also Be Seen When Looking at the Company in Toto (Inclusive of Marketing)
Note: Only Industrial Activities shown, marketing and associates have been omitted.
Source: Corporate reports and Bernstein estimates and analysis.
Source: Corporate reports and Bernstein estimates and analysis.
We consider the recently completed merger between Glencore and Xstrata to
be the logical conclusion of a journey nearly two decades in the making, one that was kicked off by the 1994 management buyout at Glencore and which catapulted Mr. Ivan Glasenberg into a senior role at the company he now runs. Mr. Glasenberg has proved to be a quick learner in his transition from a prominent-yet-private
(500)
1,500
3,500
5,500
7,500
9,500
11,500
13,500
15,500
2008 A 2009 A 2010 A 2011 A
Glencore-Xstrata Pro Forma EBITDA — Industrial Activities (US$mln)
Glencore Xstrata Pro Forma EBITDA Industrial Activities (2012)
Zinc, 15%
Copper, 36%
Nickel, 8%Alumina & Aluminum, 0%
Coal, 27%
Oil & Oil Products, 0%
Agricultural Products, 0%
Alloys, 2%
Marketing, 17%
Associates, 1%
Other, -7%
Glencore Xstrata Pro Forma EBITDA (2012)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 203
master trader to the CEO of a publicly listed company — as evidenced to investors who watched him navigate 2H:12 demands from shareholders for an improved share ratio and alterations to management compensation packages. His decision to prolong this merger process in order to "negotiate" with a regulator that does not oversee direct operations but represents the interests of his largest consumer showed an awareness of the importance of customer relationship management. The concessions that he granted addressed the concerns of the Chinese and left Mr. Glasenberg several options to unload greenfield projects that he was not particularly interested in pursuing — or the disposal of an admittedly costly crown jewel at a potentially lucrative price.
We hope that Mr. Glasenberg will show an equally steep learning curve when it comes to transparency of reporting and governance. We consider that Glencore has some way to go before it will be "institutional quality" on these metrics. Furthermore, potential investors should note that Glencore Xstrata's country risk premium is the second highest in our coverage group (2%, behind only Anglo American's 2.6% versus an average of 1.2% for the "Big Three"). This is a function of the company's operations in frontier jurisdictions like the DRC. However prudently one conducts business in these jurisdictions, they carry not only an elevated risk that host governments may make increased grabs for a share of the revenues, but also headline risk — as does Glencore's trading division.48
Nevertheless, as an owner operator (he holds 8% of the combined entity's paper), Mr. Glasenberg's incentives are squarely aligned with investors — indeed, media coverage of his "rich" 2012 compensation package seemed to fail to grasp what a good thing it is indeed that the bulk of his compensation comes from company dividends. We have long been calling for capex restraint to support commodity prices, and have highlighted how increased dividends can not only offer a boost to the sector in terms of supporting commodity prices, but also can lead to equity rerating. It was appropriate for the sector to pursue an aggressive growth strategy in the last half of the last century; this century brings a new environment to which mining companies must respond if they are to be responsible custodians of the value created by previous generations of management.49 Mr. Glasenberg seems to be one of the few industry leaders who truly understand this.
It remains to be seen how he will staff the combined entity, following the departure of much of the senior management talent from Xstrata. But whether through good luck or perspicacity (or a little of both), he finds himself facing this dilemma at a time when other major miners are letting project managers go in favor of cost cutting. So yet again, Mr. Glasenberg finds himself able to pick up attractive assets at even more attractive prices. We will be watching to see how the organization unfolds — will it be run to maximize LOM asset value, trade flow for the marketing division, or run in some hybrid fashion? Only time will tell. Time too will tell whether the senior traders at Glencore, on whom the strength of the reputation of its marketing division was built, will stick around as their post-IPO options vest — or whether this merger will help them monetize an exit strategy.
Looking farther ahead, the next logical move in Mr. Glasenberg's trajectory is, in our view, a bid for Anglo American. The synergies from the portfolios — particularly in coal — are attractive. Furthermore, "Glencore Xstrata Anglo" would form, in our view, a credible challenger to the "Big Three" of Vale, Rio and BHP, allowing the combined entity to more effectively compete for assets — and it would be in Glencore’s interests to be part of this transformation. But that is, in our view, a question for 2014-15.
48 To wit the April 21, 2013 story concerning Glencore's reported trading with Iranian supplier to nuclear program http://www.guardian.co.uk/business/2013/apr/21/glencore-trade-iran-supplier-nuclear. 49 Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
204 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 205
Appendix 1: From Grizzly to De Niro
This appendix contains the analysis from the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities" chapter of this Blackbook and is divided into three parts.
Section 1 contains a summary of the equity scenario analysis — namely, how each of the equities in our coverage react under the "Grizzly," "Bear," "Bull," and "De Niro" scenarios relative to our current base scenario at the time the analysis was conducted (May 2013).
Section 2 offers the complete commodity prices applied across the four scenarios.
Section 3 shows the five-year historical spot price distribution of the key commodities used in these scenarios.
Please note that all analysis in this appendix, as in the corresponding chapter, was conducted in May 2013. All references to current share prices, spot prices and models are as of May 21, 2013.
Equity Scenario Analysis — Summary Exhibit 324 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our
Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our Current Price Target (as a Result of Our Bullish View on Copper in the Medium Term)
Source: Corporate reports and Bernstein estimates and analysis.
TP / NPV per Share vs. Current Model
Current Model
Grizzly Bear Bull De Niro Grizzly Bear Bull De Niro
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
206 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 325 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest
Exhibit 326 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 327 Our Target Prices Are the Equivalent of the
"Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices
Exhibit 328 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
-67% -67%
-72%
-74%
-77%
Rio Tinto Vale Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — Grizzly
18%
-18%
-26%
-33%
-41%
Vale Rio Tinto Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — Bear
102%
31%
17%8%
-5%
Vale Rio Tinto Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — Bull
186%
79%
57%50%
31%
Vale Rio Tinto Anglo American
BHP Billiton
Glen Xta Pro Forma
TP vs. Current Model — De Niro
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 207
Exhibit 329 Rio Tinto Scenario Summary vs. Consensus
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 330 BHP Billiton Scenario Summary vs. Consensus
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 331 Anglo American Scenario Summary vs. Consensus
Source: Corporate reports and Bernstein estimates and analysis.
vs. Consensus
Rio TintoCurrent Model
Consensus Grizzly Bear Bull De Niro Grizzly Bear Bull De Niro
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 209
Exhibit 334 Both the "Grizzly"... Exhibit 335 ...And "Bear" Scenarios Show Revenue Below Consensus Expectations
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 336 Over the Next Two Years What We Call the
"Bull" Scenario Is Closest to Consensus Expectation
Exhibit 337 While Under the Strongest Price Scenario, There Is Upside to Consensus
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
-12%-13%
-17%
-22%-25%
-20%-19%
-18%
-23% -24%
BHP Billiton
Vale Glen Xta Pro Forma
Anglo American
Rio Tinto
Revenue vs. Consensus — Grizzly
2013E 2014E
-1%
-5%
-7%
-12% -12%
-1%
-6%
-8%
-10%
-13%
Vale BHP Billiton
Glen Xta Pro Forma
Rio Tinto Anglo American
Revenue vs. Consensus — Bear
2013E 2014E
12%
3%2%
1%
-2%
17%
2%
9%
3%
-3%
Vale Glen Xta Pro Forma
BHP Billiton
Rio Tinto Anglo American
Revenue vs. Consensus — Bull
2013E 2014E
24%
14% 13%
10%8%
34%
16%
12%
23%
8%
Vale Rio Tinto Glen Xta Pro Forma
BHP Billiton
Anglo American
Revenue vs. Consensus — De Niro
2013E 2014E
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
210 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 338 The Same Feature Is Seen at the EBITDA Line...
Exhibit 339 ...The Lack of Coverage of Glencore Makes Comparison More Difficult
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 340 Again, the "Bull" Scenario Appears Closest
to Consensus
Exhibit 341 Considerable Upside Under Stronger Commodity Scenarios Emphasizes the Degree of Operationally Geared Exposure the Miners Offer to Commodity Prices
Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.
-23%
-35%-38%
-56%-61%
-34%
-51% -50%
-67% -66%
Glen Xta Pro Forma
Vale BHP Billiton
Anglo American
Rio Tinto
EBITDA vs. Consensus — Grizzly
2013E 2014E
15%
-7%
-19% -21%
-26%
3%
-11%
-31%
-18%
-30%
Glen Xta Pro Forma
Vale Anglo American
BHP Billiton
Rio Tinto
EBITDA vs. Consensus — Bear
2013E 2014E
54%
20%17%
9%
-4%
40%
29%
6% 7%
14%
Glen Xta Pro Forma
Vale Anglo American
Rio Tinto BHP Billiton
EBITDA vs. Consensus — Bull
2013E 2014E
93%
54%48%
44%
13%
76%
42%
69%
43% 46%
Glen Xta Pro Forma
Anglo American
Vale Rio Tinto BHP Billiton
EBITDA vs. Consensus — De Niro
2013E 2014E
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 211
Rio Tinto Rio Tinto is primarily an iron ore business differentiated from its peers BHP Billiton and Vale by lack of petroleum on one side and by its high-quality (and growing) copper business on the other. Only under the most bullish price scenarios does aluminum offer diversification for the group on any measure other than simple revenue. Rio Tinto is our top pick, and, as this scenario analysis shows, the company offers the most attractive balance between upside potential and downside risk.
Exhibit 342 Rio Tinto Scenario Analysis Summary
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 343 Rio Tinto EV by Division
Source: Corporate reports and Bernstein estimates and analysis.
Rio Tinto Implied Share Price (GBP) vs. Current Model
Current Model 45.75 -
Grizzly 15.23 -67%
Bear 37.73 -18%
Bull 59.99 31%
De Niro 81.69 79%
174,957
77,886
150,387
221,759
291,325
(50,000)
-
50,000
100,000
150,000
200,000
250,000
300,000
350,000
Current Model Grizzly Bear Bull De Niro
US
$m
Rio Tinto EV
Iron Ore Aluminum Copper Energy Diamonds & Minerals Other Total
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
212 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 344 Rio Tinto — Current Model — Pro Forma Financials
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 345 Rio Tinto — Current Model — Valuation
Source: Corporate reports and Bernstein estimates and analysis.
Pro Forma Financials
Dec year-end 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A 2013 E 2014 E 2015 E
Rio Enterprise Value 228,447 62,878 291,325 157.77 99.58 30,267 9.6
Less Minorities (31,279) (16.94) (10.69)
Less Central Costs - - -
Less (Net Debt)/Net Cash (21,058) (11.40) (7.20)
Equity Value 238,988 129.43 81.69
Shares Outstanding 1,847
Share Price (USD) 129.43
Exchange Rate 1.58
Implied Share Price (GBP) 81.69
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 217
BHP Billiton BHP Billiton is the highest quality stock in our coverage, and has the second most diversified portfolio after Anglo American. But where Anglo has the drag of platinum, BHP has a petroleum kicker (run at 77% EBITDA margins in 2011 and 73% in 2012). BHP fares second worst in the Grizzly scenario (after Glencore) but only third best in the De Niro scenario (the inverse of Anglo’s third worst in Grizzly and second best in De Niro). The presence of petroleum means that, uniquely for BHP, there is a third key explanatory variable (besides copper and iron ore).
Energy Coal 10,899 3,446 14,346 2.70 1.70 1,873 7.7
BHPB Enterprise Value 302,599 56,646 359,245 67.51 42.61 46,335 7.8
Less Minorities (3,691) (.69) (.44)
Less Central Costs (9,755) (1.83) (1.16)
Less (Net Debt)/Net Cash (29,772) (5.60) (3.53)
Equity Value 316,027 59.39 37.49
Shares Outstanding 5,321
Share Price (USD) 59.39
Exchange Rate 1.58
Implied Share Price (£) 37.49
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
224 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Anglo American Anglo American has the most commodity-diversified portfolio in our coverage and the second-highest country risk (after Glencore), thanks to its South African exposure. Where BHP’s analysis confirms a third explanatory variable (oil), Anglo’s is impacted by the interlinked risks of platinum and South African country risk. The company fares third worst in the Grizzly scenario and second best in De Niro. We will look with interest to see how the Anglo turnaround story under new CEO Mark Cutifani plays out and how resultant changes in exposure may alter the potential upside/downside of this analysis.
Exhibit 366 Anglo American Scenario Analysis Summary
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 367 Anglo American EV by Division
Source: Corporate reports and Bernstein estimates and analysis.
Anglo American Implied Share Price (GBP) vs. Current Model
Current Model 23.00 -
Grizzly 6.51 -72%
Bear 17.03 -26%
Bull 26.84 17%
De Niro 36.22 57%
71,480
30,791
57,712
82,848
106,894
(20,000)
-
20,000
40,000
60,000
80,000
100,000
120,000
Current Model Grizzly Bear Bull De Niro
US
$m
Anglo American EV
Platinum Diamonds Copper Nickel
Iron Ore - South Africa Iron Ore - Brazil Manganese Metallurgical Coal
Thermal Coal Other - Non Mining Assets
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 225
Exhibit 368 Anglo American — Current Model — Pro Forma Financials
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 369 Anglo American — Current Model — Valuation
Source: Corporate reports and Bernstein estimates and analysis.
Pro Forma Financials
Dec year-end 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A 2013 E 2014 E 2015 E
Anglo Enterprise Value 91,938 14,957 106,894 83.64 52.79 8,815 12.1
Less Minorities (25,902) (20.27) (12.79)
Less Central Costs (2,277) (1.78) (1.12)
Less (Net Debt)/Net Cash (5,365) (4.20) (2.65)
Equity Value 73,350 57.39 36.22
Shares Outstanding 1,278
Share Price (USD) 57.39
Exchange Rate 1.58
Implied Share Price (GBP) 36.22
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
230 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Glencore Xstrata Plc Pro Forma The combined Glencore Xstrata entity (pro forma analysis) shows the greatest downside and least upside in the Grizzly and De Niro scenarios. This is because Glencore Xstrata is the play for investors who want sector exposure with minimal iron ore (Xstrata has none, Glencore has in 1Q:13 acquired a stake in Carl Ichan-backed Ferrous Resources, which produced just 3.2Mt in 2012 and plans to boost production to 5Mt in 2013, and reach 17Mt by 2016). Note that while we have outperform ratings on both Glencore and Xstrata, we have no rating on the combined entity as it remains pro forma until its listing.
Exhibit 378 Glencore Xstrata Pro Forma Scenario Analysis Summary
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 379 Glencore Xstrata Pro Forma EV by Division
Source: Corporate reports and Bernstein estimates and analysis.
Glencore Xstrata Pro Forma NPV per share vs. Current Model
Glencore Enterprise Value 159,169 27,955 187,125 14.04 8.75 14,135
Less Minorities (10,038) (.75) (.47)
Less Central Costs - - -
Less (Net Debt)/Net Cash (32,890) (2.47) (1.54)
Plus Corporate & Other
Equity Value 144,197 10.82 6.75
Shares Outstanding 13,326
Per Share NPV (USD) 10.82
Exchange Rate 1.60
Per Share NPV (GBP) 6.75
Without Synergies
Glencore Enterprise Value 154,142 27,122 181,264 13.60 8.59 13,385
Less Minorities (10,038) (.75) (.48)
Less Central Costs - - -
Less (Net Debt)/Net Cash (32,890) (2.47) (1.56)
Plus Corporate & Other
Equity Value 138,337 10.38 6.55
Shares Outstanding 13,326
Per Share NPV (USD) 10.38
Exchange Rate 1.58
Per Share NPV (GBP) 6.55
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 241
Vale Vale is the "pure" iron ore play in our coverage, with 70% of FY 2012 revenues coming from iron ore. As such, and given iron ore’s historical volatility, Vale has a downside tied with Rio Tinto in the Grizzly and Bear Scenarios, but the greatest upside in the Bull and De Niro scenarios. Vale remains our least preferred stock — the company incurs more capex/ton for its new iron ore exposure than Rio Tinto, and we question how the significant involvement of the Brazilian government (directly, through Golden Shares, and indirectly, through Valepar) influences the company’s strategic direction vis-a-vis shareholder interests.
Exhibit 390 Vale Scenario Analysis Summary
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 391 Vale Pro Forma EV by Division
Source: Corporate reports and Bernstein estimates and analysis.
Vale Implied Share Price (BRL) vs. Current Model
Current Model 29.25 -
Grizzly 9.73 -67%
Bear 34.38 18%
Bull 58.97 102%
De Niro 83.54 186%
102,468
51,601
115,951
180,127
244,263
(50,000)
-
50,000
100,000
150,000
200,000
250,000
300,000
Current Model Grizzly Bear Bull De Niro
US
$m
Vale EV
Iron Ore Manganese&Ferroalloys Coal Base Metals Fertilizers Other
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
242 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Exhibit 392 Vale — Current Model — Pro Forma Financials
Source: Corporate reports and Bernstein estimates and analysis. Exhibit 393 Vale — Current Model — Valuation
Source: Corporate reports and Bernstein estimates and analysis.
Pro-Forma Financials
Dec year-end 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A 2013 E 2014 E 2015 E
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 253
Appendix 2: Growth Scenarios
This appendix provides full details on the internally consistent scenarios in the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" and "China Growth Scenarios" chapters. It shows the relationship between Chinese GDP, supply, demand and resultant iron ore price in five scenarios. These scenarios are:
1) Consensus implied scenario ($80/t iron ore) 2) April 2013 IMF GDP forecast 3) "Back to the Future" based upon IMF 2012 forecast 4) "Out With a Whimper" taking the Chinese official GDP target 5) "It's All Over Now Baby Blue," with 6% GDP growth and no supply
curtailment
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
254 E
UR
OP
EA
N M
ET
AL
S &
MIN
ING
: A S
TR
AN
GE
LO
VE
— H
OW
I LE
AR
NE
D T
O S
TO
P W
OR
RY
ING
AN
D L
OV
E T
HE
OR
E
Exhibit 413 Consensus Scenario
Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.
(FCF = Operating cash flow less tax paid, less capex)
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 277
Index of Exhibits
1 Financial Overview 4
2 In 2000, China Was a Clear Outlier — the World's Largest Country Set to Age Rapidly and Yet Still at a Pre-Industrial Output Level 5
3 The Demographic Challenges Facing China Stand Behind the Requirement to Accelerate the Capital Stock Formation 5
4 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines 7
5 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003 7
6 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered 8
7 It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve; the Higher the Elasticity, the Higher the Margin and the Lower the Degree of Consolidation Required to Generate Superior Returns 8
8 Our Three Scenarios Set the Widest Plausible Band, in Our View, Around Implied Consensus Expectations; They Range from More Bullish Than the IMF's Current Forecast, Through Official Chinese GDP Targets, to a Bear Scenario in Which No Supply Is Removed from the Market Despite Prices Collapsing 9
9 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-2016 9
10 Mining Costs and Revenue Exhibit a High Degree of Correlation 10
11 Currency Appreciation Drives a Significant Part of the Cost to Price Relationship, Particularly for the Australian Producers 10
12 There Is a Significant Range of Variation Around the Benchmark 62% CIF Price Point, and Even After Adjusting for Fe Grade, Poor-Quality Ore Suffers a Significant Discount 11
13 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from the Budgetary Constraint Imposed by High Commodity Prices 11
14 There Is a Clear Relationship Between the Aging of a Population and Its Wealth…Usually, Birth Rates Fall as Societies Get Rich, With Aging Population Supported by the Increased Productivity Associated With Industrialization 16
15 In 1910, the U.S. Began Transitioning from an Agricultural Economy… 17
16 …Toward One Driven by Services 17
17 We Believe That the Increases in Labor Productivity, Occasioned by the Increased Intensity of Capital Stock (Machinery, Railroads, etc.), Enabled This Transition 17
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
278 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
18 In 2000, China Was a Clear Outlier from the Normal Pattern — the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-industrial Level of Output 18
19 It Was Only in the Year 2000 That China Started Reducing Its Framing Dependence 19
20 1970s Saw the Rise in Construction and Manufacturing Employment Start… 19
21 …Followed in Short Order by the Service Sector 19
22 Today, More People Are Employed in Services Than Manufacturing 19
23 The Pattern in China Is Nearly Identical to That Seen in the U.S.; Capitalization Drives Productivity Increases, Which Then Lead to the Emergence of Services 20
24 Comparing China to the U.S. Suggests That the Structure of the Chinese Economy Today Is Similar to the U.S. in the 1930s, Both in Labor and Capital Stock; However, the Transition Has Been Far More Rapid in China 20
25 The Demographic Challenges Facing China and the Emergence of Lewis Point Stand Behind the Requirement to Accelerate the Capital Stock Formation 21
26 The Urgency to Industrialize the Chinese Economy Is Evident in the Acceleration in Steel Demand in Comparison to the Historical Precedent in the West 22
27 A Very Similar Demographic to China Was Seen in Japan Toward the End of the Last Century; However, by the Time the Japanese Transition Occurred, Industrialization Was Complete and Output Stayed High 22
28 A Rapid Fall in the Child Dependency Ratio Following the "One Child" Policy… 24
29 …Will Eventually Be Offset by the Extraordinary Increase in the Old-Age Dependency Ratio 24
30 The End of the Last Century Marked an Opportunity to Deploy Abundant Labor in China to Develop Its Capital Stock… 24
31 …Mirroring Japan But Scaling Up More Than 10 Fold (127 Million vs. 1,345 Million People) 24
32 1930-1973 Saw Real Commodity Prices Grow as Capital Was Accumulated in Western Economies; Once Demand Went Flat, Productivity Improvements in Mining Drove Real Commodity Prices Down 26
33 The Emergence of a New Phase of Capital Accumulation Driven by China Has Seen Significant Real Commodity Price Appreciation; Our Counter-Consensus View on Commodities Is Driven, in Part, by a Belief That This Will Continue Until the Chinese Capital Stock Accumulation Comes to an End 26
34 Chinese Labor Costs Have Been Instrumental in Changing Global Commodity Cost Curve Structures and, With Them, the Price of Mined Raw Materials; What Happens Once China Moves Past the Lewis Point? 27
35 Across Our Coverage Group, Iron Ore Generated, on Average, 41% of Revenue… 30
36 ...And 65% of EBITDA in 2012 30
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 279
37 Iron Ore Currently Accounts for 44% of Rio Tinto's Revenue and Has Been Driven by a 7.6% Production CAGR 30
38 Iron Ore Currently Accounts for 31% of BHP Billiton's Revenue and Has Been Driven by an 8.1% Production CAGR 31
39 Iron Ore Currently Accounts for 70% of Vale's Revenue and Has Been Driven by a 6.9% Production CAGR 31
40 Iron Ore Currently Accounts for 18% of Anglo American's Revenue and Has Been Driven by a 23.7% Production CAGR 32
41 Like Other Countries That Have Industrialized, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady-State Rate of Consumption Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing or Services-Oriented Focus 34
42 The "Services and Technology" Pattern (Embodied by the U.S.) Shows the Highest Output/Capital Stock (on a Per-Capita Basis), the "Manufacturing" Pattern (Embodied by Germany) Also Shows an Upward-Sloping Relationship Between Output and Capital Stock; Only the Failed "Inefficient" Pattern (Embodied by the former USSR) Shows a Flat and Negative Relationship Between Output and Embedded Capital Stock 34
43 Our Analysis Derives a Consistent Relationship Between Steel Demand Growth and Overall Level of Economic Activity for China; It Does Not See Signs That the Output Generated by the Embedded Capital Stock Is "Inefficient"; Rather, It Is Increasing in a Manner Consistent With Development Into an Economy Midway Between the "Services and Technology" (Embodied by the U.S.) and the "Manufacturing" (Embodied by Germany) Patterns 35
44 A Central Plank of Our Thesis Rests on the Fact That Chinese Urbanization and Industrialization Are Still Far from Complete; in Terms of Capital Stock Accumulation, Labor Composition and Levels of Output, China Today Resembles the U.S. in the 1930s 36
45 At First Glance, the Old Heuristic of Marginal Cash Costs Appears to Have Broken Down… 38
46 …But the Fact That Our Visibility Into Costs in China Is Incomplete Provides a Better Answer 38
47 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis, Which Completes the Global Cost Curve, Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines 38
48 Our Analysis of the Cash Costs of 149 Green and Brownfield Projects Shows That the Majority Will Not Be Value Accretive at Iron Ore Prices Below $150/t Against a Fully Loaded Discount Rate 38
49 We Embed a Contango Rather Than Backwardation Into Our Short- to Medium-Term Price Forecasts 39
50 Our Iron Ore Forecast Sees Considerable Upside in 2014-16E Relative to Consensus; We Would Be Even More Bullish If We Were Convinced the Western Majors Were Committed to a Higher Degree of Capital Discipline 39
51 Mining in Western Australia in 1961... 43
52 ...And Today 43
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
280 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
53 At Either End of the Industrialization Process, Metal Intensity (Rate of Consumption/Embedding) Is Low; Rather, Metal Is Required to Embed the Capital Stock That Enables the Transition from Primary to Tertiary Forms of Value-Add (or Farmers to Hairdressers) 44
54 The Chinese Economy Has Seen Steel Intensity Peak at ~65kg/$'000 of Output (vs. 40kg/$000 U.S. Peak); Because China's Peak Was More Than 60% Higher Than the U.S., and Because Many Datasets Look at Only the Last ~20 Years of Data, This Has Spawned Fears That the High Rate of Steel Consumption in Recent Years Is Indicative of Overinvestment and That the Chinese Economy Is Unduly (and Inefficiently) Steel Intensive 45
55 Global Freight Rates Have Declined in Real Terms by an Average of 0.9% p.a. Over the Last 260 Years... 46
56 …Leading to the Emergence of a Truly Global Market for Bulk Commodities, Including Iron Ore and Coal; Consequently, Relative Global Competitiveness in the Production of These Commodities Began to Matter 47
57 Before the Advent of Scrap-Based EAF Production, U.S. Iron Ore Production Peaked in the 30 Years Between 1950 and 1980 47
58 The Transition to EAF-Based Production Saw Declines in Both Iron Ore Production and Employment 48
59 Consequently, the U.S. Iron Ore Productivity Peaked at Nearly 9,000 Tons/Man-Year (Although for Much of the U.S. Industrialization, Its Intensity Was Far Lower) 48
60 In Contrast to the U.S., Western Australian Iron Ore Industry Was Always Driven by Exports — First to Japan and Today to China; the Declines in Shipping Costs Facilitated This Greatly 50
61 Superior Geology and Continuing Improvement in Mining and Logistics Technology Have Led to Substantial Increases in Mining Productivity 50
62 In Addition, Low Population Density Renders Iron Ore Mining in Pilbara Hugely Productive... 51
63 …And Provides a Significant Advantage Over the U.S. or China as Mining Jurisdictions 51
64 The Consolidation of the Industry Can Be Seen in Both the Concentration of Employment... 52
65 ...And Production in Western Australia 52
66 Among the Majors, Rio and BHP Have Almost Identical Productivities and Are Almost Twice as Productive as Either FMG or Cliffs 53
67 It Is Also Interesting to Note the Significant Declines in Productivity Seen at the Company Level as Well as in the Industry Overall in Recent Years 53
68 However, the Decline in Productivity in Western Australia Is Due to the Inclusion of Labor That Is Allocated to Building Projects (With a Significant Production Lead Time) vs. Labor Allocated to Operations That Generate Current Volume 54
69 China Has Seen Both Its Steel… 55
70 …And Iron Ore Production Increase 55
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 281
71 The Iron Ore Majors (Vale, Rio and BHP) Have Aggressively Grown Their Volume Share of the Chinese Iron Ore Market Over the Last Decade, While Chinese Self-Sufficiency Has Declined Significantly 56
72 Extrapolating the Trend Decline in Self-Sufficiency Forward Suggests the End of the Chinese Iron Ore Industry (and With It the High-Cost Price Support for Iron Ore) in the Region from 2020 to 2025; Clearly, Acceleration in Supply from the Seaborne Market and Declining Steel Growth Rates Would Advance This 56
73 Chinese Growth in Iron Ore (on a Rich-Ore-Equivalent Basis) Peaked in 2007; Since Then, Declining Grades and Rising Costs Have Seen Useable Ore Production Decline 57
74 An Ever-Increasing Amount of Labor Has to Be Allocated to China's Iron Ore Production… 57
75 …Leading to a Prolonged Decline in the Productivity of the Chinese Domestic Industry Since 2007 58
76 An Australian Miner Is 35x as Productive as His Chinese Counterpart; However, Given That Globalization of the Iron Ore Industry Has Enabled China to Outsource Its Iron Ore Mining to Australia, Does That Differential Matter? 58
77 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003 59
78 We Compare the Rate at Which Steel Is Being Embedded in China Relative to Peak U.S. Steel Intensity (Y Axis) and the Peak Iron Ore Production (in Man-Hour Terms) to Which China Has Access Relative to That Accessible to the U.S. During Its Industrialization (X Axis); However, Due to Productivity Improvements, the Real Resources Consumed to Embed Steel in the Chinese Economy Are Markedly Lower Than Was the Case During the Industrialization of the U.S. 59
79 U.S. Iron Ore Production Peaked in 1951, With an American Miner Producing 3,150 Tons/Year and Steel Intensity of 33kg/$1,000; Today, China's Steel Intensity Is 63kg/$'000, While the Effective Productivity of Iron Ore Production Is 13,600 Tons/Man-Year 60
80 Bergkamen Coal Washing Plant 61
81 Hückelhoven Coal Washing Plant 61
82 Kaiserstuhl Coking Plant 61
83 Werdohl Leaf Spring Factory 61
84 Assuming Disciplined Capex, We Expect Rio Tinto to Have Room to De-Gear Its Balance Sheet Significantly Over the Next Few Years 66
85 Curtailment of Capex and De-Gearing of the Balance Sheet Would Enable Rio to Raise Its Dividend from US$3.0 Billion in 2012; We See No Reason Why It Should Not Double Over Time 66
86 Rio Tinto's Cost Curve... 67
87 ...Not Dissimilar to Our Own 67
88 Schematically, We Can Represent the Rio Tinto Cost Curve... 68
89 ...And So We Can Calculate the Elasticity of Supply That Rio Tinto Must Believe In 68
90 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered 69
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
282 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
91 Rio Tinto's Claim That It Does Not See a Price Fall of US$20/t on the Back of Its 70Mtpa Increase in Production Becomes the Key Variable in Deciding Whether or Not the Project Makes Sense; If the Project Induces a Greater Decline, It Will Be Net Value Destructive for the Company 70
92 Rio Tinto Also Believes in a 3% Demand Growth World; This Would Be Sufficient to Limit the Price Impact of the Expansion to ~US$14/t and So Make It Appear Value Accretive (Though Clearly Not by the Degree That It Would Be If the Price Impact of the New Supply Were Neglected in Entirety) 71
93 In Understanding the Value Impact of Consolidation, We Construct Two Simple Industries — One Fragmented... 72
94 ...The Other Consolidated, But in All Other Respects Identical 72
95 For Each Industry, We Simply Look at the Consequences of Each Player Looking to Maximize Apparent Local Value; in This Case, "Apparent" Means Simply Looking at the Economics of Their Own Decision Making; It Is Easy to Show That in a Situation of Potential Oversupply, Local Value Maximization Leads to Global Value Destruction for a Fragmented Industry With No Market Impact Awareness... 73
96 ...And This Global Value Destruction Will Occur Whether or Not Each Individual Player Is Aware of the Impact That Its Decisions Have on the Market Price; the Apparent Gain from Organic Growth in Each Scenario Is an Overwhelmingly Compelling Prize 74
97 In a Consolidated Industry, Where Each Player Acts as If It Was Unaware of the Impact That Its Decisions Have on the Market, There Is Absolutely No Difference in Outcome Versus a Fragmented Industry; Exactly the Same Volume Is Brought to the Market and Exactly the Same Value Destruction Results; Consolidation in Commodity Markets Is Necessary But Not Sufficient to Generate Superior Returns 75
98 The Sufficient Condition for Superior Returns in Consolidated Markets Is an Awareness of the Impact That Volume Decisions Have on Price and the Fact That This Changes the Value of the Entire Portfolio, Not Just the Marginal Ton; Under Such Conditions, the Pursuit of Local Optimality Returns a Global Value Solution in Excess of What Is Optimal 76
99 A Necessary Condition for Superior Price Performance Is That an Industry Be Consolidated, But It Is Not Sufficient; a Necessary and Sufficient Condition Is That There Is a Consolidated Industry Comprising Agents Who Understand the Price Impact of Their Own Decisions 77
100 There Are Really Only Three Players That Matter in Iron Ore (Four at a Push)...This Is a Sufficient Enough Degree of Consolidation That the Pursuit of Local Value Maximization Ought to Embed Capital Discipline in the Industry and Generate Superior Returns 78
101 While the Degree of Possible Oversupply Is Clearly an Important Variable... 78
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 283
102 ...Even More Important Is the Elasticity of Supply, and Here Rio Tinto Itself Estimates This to Be About Three; Moreover, It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve and So Generates the Margins That an Industry Enjoys; the Higher the Elasticity, the Higher the Margin and Also the Lower the Degree of Consolidation Required to Generate Superior Returns 79
103 However, This Is Clearly Not What Is Being Discounted in the Stocks 79
104 We Are Beginning to See Greater Clarity from the Miners in Their Thinking on Capital Discipline and Price Preservation...All of Which Ought to Create a Greater Chance of Seeing Growth Restrained and the Industry Re-Rate, Generating Significant Upside for the Miners, Especially Rio Tinto and Vale 80
105 Summary of Historical Commodity Prices 82
106 Except for Some Precious Metals, Current Spot Prices of the Key Commodities for the Miners in Our Coverage Were Substantially Below Their Most Recent Five-Year Mean 82
107 Higher Average Year-to-Date Prices Reflect Higher Commodity Prices in the Beginning of the First Quarter of the Year 83
108 Summary of Commodity Price Scenarios vs. Historical Distribution — Real Terms 84
109 Summary of Commodity Price Scenarios vs. Current Model — Real Terms 84
110 The "Fat Tail" in the Copper Price Distribution Relates to the Period Immediately After the Financial Crisis; We Would Interpret the Shape of the Price Distribution as Telling Us Something About the True Costs of Marginal Production 85
111 The Shift in Pricing Regime from the Benchmark to the Spot Market as Well as the Deterioration in Chinese Domestic Mined Iron Ore Grades Result in a Markedly Different Price Distribution for Iron Ore Than for Copper 85
112 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our DCF Value (as a Result of Our Bullish View on Copper in the Medium Term) 86
113 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest 86
114 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent 86
115 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices 87
116 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price 87
117 As This Scenario Analysis Run in Late May Shows, Rio Tinto Offers the Most Attractive Balance Between Risk and Potential Upside 87
118 The Presence of Oil in the BHP Portfolio Adds a Third Critical Variable That Renders It Harder to Unpack the Implied Iron Ore and Copper Price from the Market Price; Oil Remains a Key Diversifier for the Highest Quality Stock in Our Coverage 88
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
284 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
119 While the Market Price of Vale Might Seem to Imply a Higher Iron Ore Price, We Believe That This Is a Function of Us Modeling a Higher Tax Rate for Vale Than the Market 88
120 For Anglo, the Issues of Both Platinum and South African Political Risks Add a Layer of Complexity to the Investment Proposition 88
121 Note That for Glencore Xstrata, We Run the Analysis on Our Combined Pro Forma Glencore Xstrata; Copper Is the Single Greatest Sensitivity for Glencore Xstrata; to See Outperformance in This "Not Yet Institutional Quality" (from a Governance and Reporting Perspective) Stock, It Is Necessary to Believe in a Strong Copper Price 89
122 A Central Question for an Iron Ore Price Forecast Is How to Project China's Steel Intensity Into the Future? Looking at Historical Steel Intensity Alone Is Insufficient 90
123 The Relationship Between Economies' Capital Stock and Level of Output (R-squared = 65%) Provides the Missing Data Point for Forecasting How the Capital Stock to Output Ratio Will Evolve in China 91
124 We Use the U.S. Development Path as a Basis for China's Industrialization Forecast 92
125 The Benefit of Increasing Productivity Over Time Is Easier to See If the Axes in the Earlier Graph Are Inverted; in the U.S., the Real Advances Took Place Only Once the Capital Stock Had Been Built and Urbanization With the Accompanying Labor Force Transformation Had Been Completed 92
126 China Is More Steel Intensive Than the U.S., But Shows Convergence 93
127 For Illustration Only, We Show a Development Path That Would Enable China to Converge to the Current U.S. GDP Levels by 2020; It Would Require a 20% Growth Rate Each Year 94
128 Using Convergence as the "Missing" Data Enables Us to Calculate the Transition Path Between Steel Intensities at Two Different Points in Time and Derive the Relationship Between Overall Economic and Metal Growth Rates 94
129 Since the Time of Our Original Analysis, China's Economy Has Slowed Down Markedly 96
130 In the Space of a Year, the IMF Revised Its Growth Projections Down by 0.5% on Average Over 2014-16) 96
131 Taking the IMF's April 2013 Growth Forecast, We Estimate the Following Convergence to the U.S. Steel Stock Levels for China… 96
132 …Which Would Imply the Following Trajectory for Steel Intensity; the Previous Steel Intensity Trend Line Was Extrapolated in Order to Keep the Long-Term Evolution of China Consistent With the U.S.; Now, We Calculate It Year-by-Year 97
133 We Refine Our Previous Steel Intensity Trend Line… 97
134 …And Generate the Following Steel Production Forecast; It Shows Declines in Crude Steel Growth Seen Post 2020; in Our View, It Is Unavoidable Under All Chinese Growth Scenarios 98
135 Instead of Having to Guess the Relationship Between Metal Growth and Overall Economic Growth, We Can Derive It from the First Principles 98
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 285
136 Consensus Expectations for Chinese Steel Production Are Significantly Different from Those Implied by the IMF's GDP Deck 100
137 We Back Out the Scenario for China's GDP Evolution That Is Implied in the Steel Growth Consensus; It Sees the Annual GDP Trend Growth of 7.1 vs. 8.5% Implied in the IMF Deck 100
138 The Steel Intensity That Such a Trajectory Would Imply Is Markedly Lower Than the Scenario Consistent With the IMF GDP Deck 101
139 Instead of a Recovery in the Metal to GDP Multiplier, We Would Continue to See Declines 101
140 Unsurprisingly, Given the Relative Immaturity of the Chinese Economy, Its Steel Growth Is Highly Leveraged (~4 to 1) to the Overall Economic Growth Rate 102
141 Using Our Knowledge of the Structure of the Chinese Domestic Mining Industry, We Can Generate an Exemplar Cost Curve for the Industry as a Whole 103
142 This Enables Us to Understand the Elasticity of Supply 103
143 Combining It With Consensus Steel Demand Expectations Yields Consensus Price Expectations… 104
144 …And Enables Us to Derive Consensus Supply Expectations 104
145 Assuming the Consensus GDP Deck, Two-Thirds of the Supply Expected to Come Onstream Is Not Needed… 104
146 …With Price Destructive Oversupply of ~300Mt by 2020 104
147 We Can Use the Implied Consensus Supply Expectations to Derive Consensus Steel Demand Scenario and Iron Ore Price Line for Any Chinese GDP Deck Assumptions; Steel Intensity Is Geared to China's GDP and the Iron Ore Price Is Geared to Steel Intensity; Risk and Uncertainty Pervades This Sector 106
148 In Order to "Stress Test" the Demand and Price Environment, We Construct Three Scenarios: (1) Slightly More Bullish Than the IMF (i.e., Back to the IMF April 2012 Figures); (2) One That Respects China's Official 7.5% GDP Growth Target; and (3) Significantly Below the Target and the Implied Consensus Growth Rates 108
149 The Resulting Price Paths Illustrate That the Risks for Long-Term Iron Ore Prices Are to the Upside, Barring a Significant "Miss" on the Chinese Growth 108
150 Scenario 1 Uses a Slightly Stronger GDP Deck Than the Current IMF Figures… 109
151 …And Nearly 2% Higher Than Consensus 109
152 To Support This Level of Growth, Steel Intensity Would Need to Rise 109
153 It Would Enable an Accelerated Capital Stock Growth Rate in China 109
154 This Would See Steel Growth Approach 10%... 110
155 ...And a Steel-to-GDP Ratio Well Above 1 110
156 Under This Scenario, Only 18% of New Volume Is Price Destructive… 110
157 …Leading to a Continuation of High Prices Into the Medium Term 110
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
286 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
158 Scenario 2 Assumes the Official Annual GDP Growth Figure of 7.5%... 111
159 …Which Is Still Significantly Ahead of Consensus Expectations 111
160 Although It Can Be Achieved Without an Upward Correction to Steel Intensity… 111
161 …It Still Anticipates an Above-Consensus Demand Growth… 111
162 …With Well-Supported Steel Growth... 112
163 ...And a Declining Steel-to-GDP Ratio 112
164 Nevertheless, 200Mtpa of Supply Is Still Unwarranted from a Value Perspective... 112
165 …Leading to Suppressed Yet Above-Consensus Iron Ore Prices 112
166 Our Bear Case Assumes 6% GDP Growth for China in the Long Term… 113
167 …Which Is ~1% Below the Current Implied Consensus Expectation 113
168 This Would Lead to a Significant Reduction in Steel Intensity... 113
169 ...And Would See Stagnant Production Into the Medium Term 113
170 Consequently, the Industry Would Effectively Go Ex Growth... 114
171 ...And Steel-to-GDP Ratio Would Fall Well Below 1 114
172 In This Scenario, Virtually All Volume Growth Is Price and Value Destructive… 114
173 …Iron Ore Tests a US$65/t Floor 114
174 There Has Been a Significant Change in the Structure of the Chinese Economy Over the Last 10 Years... 116
175 ...With Consumption Losing Out to Investment; Now Investment Accounts for Around Half of All Activity 116
176 Secondary Forms of Economic Activity Have Grown on Average 1.4% Faster Than the Tertiary Forms Over the Course of China's Industrialization (10.7% vs. 9.3%) 117
177 However, the Net Result of All of This Activity Is That China Is on Trend to Catch Up With the U.S. in Terms of the Overall Steel Productivity 117
178 China Currently Has ~35% of the Steel Stock of the U.S.; We Expect It to Rise to ~69% by 2020 118
179 We Use This Relationship Between Steel Stock and Output to Link the Steel Intensity (i.e., Rate of Steel Consumption) Between Different Periods in a Country's Economic Development 118
180 Such a Trajectory of Steel Use Implies FAI Falling from 48% to 38% of GDP, in Effect Achieving the Policy Aim of Rebalancing the Chinese Economy 119
181 However, the Vast Majority of the Investment in China Relates Not to Raw Material Consumption Per Se, But Rather the Capitalization of Labor Associated With That Investment 119
182 Under Our Scenario for Raw Material Consumption (With Constant Prices), the Raw Material Component of FAI Falls from 12.4% (Historical Average) to 8.9% Over the Forecast Period 120
183 Steel Is by Far the Most Important Raw Material Component of FAI; Copper and Aluminum Are Also Significant 120
184 The Proportion of Raw Materials in FAI Has Stayed Very Stable Over Time 121
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 287
185 Against This Picture of Raw Material Use, the Increase in the Importance of Investment in China Is Due to the Movement of Labor Into Tertiary Forms of Activity 122
186 Against a Backdrop of Continued De-Agrarianization, a Declining Labor Pool and a Continued Move Into Tertiary Forms of Employment, the Component of the Labor Force in Secondary Forms Must Decline Over Time 122
187 The Trajectory of De-Agrarianization Has China Requiring 15% More Steel Capital Stock Than the U.S. to Effect the Transition to an Economy Dominated by the Service Sector 123
188 This Deceleration in Steel Demand and Change in Composition of the Workforce Imply a Significant Increase in Steel Use Productivity; If Steel Productivity Were Held Flat, There Would Be Significant Upside to Steel Demand 123
189 Against This Demand Environment, We Look at Cost Curves of Iron Ore Landed China; the New Low-Cost Supply Additions Rotate the Cost Curve Downwards; Cost Inflation Translates the Cost Curve Upwards; It Is the Interplay Between Cost Curve Translation and Rotation That Generates Margin and Ultimately Price 124
190 Our Analysis of the Major Miners Informs Our Supply Side Projections 125
191 We Allow for Significant Growth from the Juniors 125
192 FMG Appears to Have the Most Aggressive Growth Plans 125
193 We Then Look at Changes to the Supply and Demand Situation to Imply a Change in the Overall "Balance" of the Iron Ore Industry... 127
194 ...Which Is, Naturally Enough, Different from Consensus 127
195 We Expect a Stronger Demand Environment and a Stronger Supply Side Response Than Consensus 128
196 A Critical Difference Is in the Impact of Expected Continued Cost Escalation on a Non-Static Chinese Iron Ore Industry 128
197 It Is a Combination of the Differences in Supply and Demand Forecasts as Well as Expected Real Cost Escalation in China That Gives Rise to Our Above-Consensus Price Forecast 129
198 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-16E 129
199 Roughly 50% of the Upside of Our Forecast vs. Consensus Is Attributable to Differences in Our Cost Escalation Assumptions and 50% Due to S-D Differences 130
200 The First Element of an Iron Ore Price Forecast, in Our View, Is the Calibration of a Consistent View on Steel Intensity and Steel Productivity 131
201 We Calibrate This Against the Overall Composition of Economic Activity and the Labor Force in China 132
202 We Look at How the Cost Curve in China Would Change If There Were to Be No Impact on Price; We Then Calculate the Impact of a Notional "Over Supply" or "Under Supply" on That Cost Curve, Particularly Relative to Consensus Expectations 133
203 This Then Drives Our Price Line 134
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
288 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
204 There Is a Strong Correlation Between the Price of Iron Ore and the Strength of the Australian Dollar, With a High Iron Ore Price Equating to a Strong Local Currency 136
205 Likewise, the Relationship Exists (Albeit Less Pronounced) for the Brazilian Real 136
206 Unfortunately for Anglo American, for the South African Rand, the Relationship Is Not in Evidence 137
207 Mining Costs and Revenue Exhibit a High Degree of Correlation 137
208 Currency Appreciation Drives a Significant Part of the Cost-to-Price Relationship… 137
209 ...Which Is Particularly True for the Australian Producers Rio Tinto and BHP Billiton 137
210 Rio Tinto Is the Lowest Cost Iron Ore Producer in Our Coverage; Anglo, as the Highest, Experiences Cash Opex That Is 50% Higher Than Rio's 139
211 While Mining Cost Escalation Is a Common Worry, It Appears That the Miners Have Started to Get It Back Under Control 139
212 Rio Tinto, in Particular, Is Leading the Way 139
213 BHP Has Gone Through a Similar Experience With Both the Cost Relationship to Iron Ore... 140
214 ...And Over Time, Showing a Marked Deceleration 140
215 For Vale, the Evidence of Cost Control Is Less Convincing... 140
216 …As Can Be Seen in the Continued Cost Growth Over Time 140
217 For Anglo, the Situation Seems to Lie Somewhere Between Vale and the Australians… 141
218 …With Some Evidence of Cost Containment 141
219 An Internet Search for the Miners and "Cost Controls" Results in an Ad for a Cost Control Position as at Least One of the Top 10 Hits (as of April 30, 2013); Only Anglo Has All of Its First Five Hits as Job Openings 141
220 Given the Historical Relationship Between the Iron Ore Price and Producer Currencies, a Return to US$80/t Iron Ore Would See Producer Currencies Weaken, Lowering US$ Costs and Improving Margin for the Miners 142
221 The Sensitivity to Such a Margin Impact Gives, on Average, ~6% Additional Upside to the Australian and Brazilian Miners But Is Value Neutral for Anglo American 142
222 Rusting Granite in Peterborough, New Hampshire; During Pre-Industrial and Early Industrial Periods, Local Blacksmiths Relied Upon "Bog Iron" and "Mountain Iron" (Like the Below); While the Highest-Quality Deposit Reportedly Contained Ore of 56-63%, the Grade Was Often Much Lower and Was Quickly Depleted in Concentrations of Useful Economic Value 144
223 Where All the EBITDA Began — A Banded Iron Formation (BIF) Showing Characteristic Layering Patterns of Iron Ore and Gangue Material 145
224 Not All Iron Ores Are Created Equal... 145
225 …And There Is a Significant "Tail" of Non-Iron Material in Iron Ores, All of Which Have a Value Implication 146
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 289
226 While Australia and Brazil Are the Most Important Suppliers Into China, They Account for Only ~70% of the Total Non-Chinese Portion of Supply 147
227 The Overall Market Shares of Brazil and Australia Have Remained Roughly Constant Over Time; Perhaps This Indicates an Understanding from Incumbents That Competing on Volume in Commodity Markets Is Value Destructive 147
228 There Is a Considerable Variation in the Observed Grade of Iron Ore Landed Into China — It Is Still Far from Being a Homogeneous Product 148
229 An Analysis of 208 Pricing Points Shows a Significant Range of Variation Around the Benchmark 62% CIF Price Point 148
230 Unsurprisingly, Fe Grade Explains Most of the Variation in Price 149
231 Nevertheless, Even After Adjusting for Fe Grade, Poor Ores Suffer a Significant Discount 149
232 This Is True Across Australian Iron Ores... 149
233 ...As Well as Those from India 149
234 There Is a US$46/t Differential Across Ores from Different Supply Locations 150
235 Across the "Major" Space, There Is a US$25/t Differential 150
236 Assuming That 50% of the Quality-Disadvantaged Material Moves Out of the Market as Price Begins to Fall, It Can Be Expected to Provide ~US$5/t of Offsetting Upward Price Support, Adding, on Average, $1.75/Share Across Our Coverage (as of June 30, 2013 Prices) 151
237 Following the Same Three-Phase Trajectory for the Evolution of the Steel Consumption Rate Within All Industrialized and Industrializing Countries, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady State Rate of Consumption, Once the Process Is Completed, Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing- or Services-Oriented Focus 155
238 In 2000, China Represented a Clear Outlier from the Normal Pattern: the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-Industrial Level of Output; by 2020, China's Working-Age Population Will Have Peaked; Over the Next Generation, 260 Million People Will Leave the Chinese Labor Force, Leading to a 90% Increase in the Old-Age Dependency Ratio 156
239 Over the Last 10 Years, the Domestic Steel Price in China Has Been Remarkably Constant... 157
240 ...Whereas the Price of Raw Materials Has Accelerated Sharply; It Is True for All — Iron Ore... 157
241 ...Key Fuel and Reductants Used in Steel Making — HCC... 158
242 ...And Its Replacement PCI 158
243 The Rise in Iron Ore Price Has Done the Most to Increase the Costs of Chinese Steel Production, Reflecting a Transfer of Value from the Chinese Industrial Sector to the Western-Owned Miners 158
244 Raw Materials Fed Into Steel Making Used to Cost Less Than 100US$/t; It Is Now Well Over 300US$/t 159
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
290 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
245 Currently, the Raw Material Cost of Steel Making Accounts for Nearly Three Quarters of the Overall Costs in China 159
246 There Has Also Been a Slower, But Nonetheless Pronounced Rise in the Non-Raw Materials Costs of Steel Making in China 159
247 Unsurprisingly, Margins in the Chinese Steel-Making Industry Have Declined on the Back of Higher Raw Material Prices 160
248 While, at First Glance, the Growth in Chinese Steel Production Appears Monotonic... 160
249 ...This Belies a Marked Slowdown in Chinese Steel Production Rates from ~20% at the Start of the Last Decade to Closer to 6% Today, Concurrent With the Declining Steel-Making Margins 161
250 Growth and Profitability Declined Concurrently 161
251 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from High Commodity Prices; the West Experienced a Step Change in Steel Demand Growth Only After Urbanization Was Complete; However, This Is Far from Complete in China, Combined With the Imminent Demographic Shift Enables Us to Conclude That in the Presence of Falling Iron Price, Steel Demand Would Rise 162
252 A Complicating Factor in the Analysis Steel Elasticity Is the Effect of Chinese Monetary Stimulus in 2009 163
253 There Is a Strong Relationship Between Lagged Steel Growth and the Growth in the Supply of Money... 163
254 ...We Strip This Effect Out of the Analysis of Steel Demand Growth... 163
255 ...The Effect of This Would Be to Lower the Level of Steel Production for Any Given Profitability... 164
256 ...Which Increases the Strength of the Relationship 164
257 It Is Inconsistent to Simultaneously Believe That Iron Ore Prices Would Fall as a Consequence of Low Chinese Steel Growth and That China Is Undergoing a Period of Ongoing Capital Stock Accumulation; Consequently, We Believe That Falling Iron Ore Prices Would Lead to Increased Steel Growth of an Additional 3% Impact on Trend Steel Growth Rates, Even Given a Declining Steel Price 165
258 The Impact of the Additional Steel Demand Helping Offset the Impact of "Oversupply" from New Projects Could Support a Higher Price Level Than Would Otherwise Be Achieved 166
259 The Impact of Incorporating This Effect Could See Iron Ore Prices Stabilize Some 15US$/t Higher Than Would Be the Case If the Elasticity of Demand Is Ignored 166
260 Summary of Our Coverage 169
261 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong... 170
262 ...And Explains the Majority of the Historical Performance 170
263 BHP Is Currently Slightly Overvalued Relative to Its Historical Best Fit, an Unsurprising Development in the Current Tremulous Market Environment 171
264 We See Considerable Upside in the Share Price 171
265 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong… 172
266 …And Explains the Majority of the Historical Performance 172
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 291
267 Rio Is Currently Slightly Overvalued Relative to Its Historical Best Fit, But Less Than 5%, and Hence Less Than High-Quality BHP 173
268 We See Considerable Upside in the Share Price 173
269 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong… 174
270 …And Explains the Majority of the Historical Performance 174
271 Anglo Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But We Note the Regression Has Shown Instability and a Decision to Invest in the Stock Is Contingent Upon a Belief in the Fundamental Restructuring Story as Well as Commodity Prices 175
272 We See Considerable Upside in the Share Price 175
273 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong… 176
274 …And Explains the Majority of the Historical Performance 177
275 Vale Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But as With Anglo, We Caution That the Regression Has Been Unstable and That There Are Risks Not Captured by It (Specifically the Significant Influence of the Brazilian Government) 177
276 We See Considerable Upside in the Share Price 178
277 For Glencore Xstrata, the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore 179
278 And in Glencore Xstrata (Though for Glencore Xstrata the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore) 179
279 Glencore Xstrata Seems Fairly Valued Relative to the Historical Best Fit of This Synthetic Stock 180
280 We See Considerable Upside in the Share Price 180
281 BHP DCF Value 182
282 Rio Tinto DCF Value 182
283 Anglo American DCF Value 183
284 Vale DCF Value 183
285 Glencore Xstrata DCF Value (With and Without Synergies) 184
286 Iron Ore Represents, by Far, the Most Significant Contribution to the Cash Generation of the Miners… 187
287 …And Has Been, Along With Copper, the Strongest Performer of the Major Commodities 187
288 Vale Represents Essentially an Iron Ore Pure Play in Both the Near and Longer Term, While We See Rio Diversifying in the Medium Term; BHP and Anglo American Have Exposure to a Well-Diversified Basket of Commodities; Glencore Xstrata Is the Only Company in Our Coverage Without Any Direct Production Exposure to Iron Ore* 188
289 Vale Is the Most Heavily Exposed to Iron Ore of Our Coverage Group (a Trend We Do Not See Changing)… 189
290 …Followed by Rio Tinto, Which Is Augmenting Its Portfolio With Some of the World's Best Copper Projects (the Metal on Which We Are the Most Bullish) 189
291 Historical Evolution of Rio Tinto's Revenue by Business Unit 189
292 Historical Evolution of Rio Tinto's EBITDA by Business Unit 190
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
292 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
293 Historical Evolution of Vale's Revenue by Business Unit 190
294 Historical Evolution of Vale's EBITDA by Business Unit 191
295 Our Forecasts for Rio Tinto's Revenues… 192
296 …And EBITDA Through 2018 See Considerable Growth 192
297 Similarly, Our Forecasts for Vale's Revenues… 192
298 …And EBITDA Also Sees Considerable Growth 192
299 The Diversity in Exposure Makes BHP One of the Lowest-Risk Miners from a Commodity Perspective… 193
300 …Complemented by Low Political Risk and a Diverse Range of Operating Geographies 193
301 Historical Evolution of BHP's Revenue by Business Unit 194
302 Historical Evolution of BHP's EBITDA by Business Unit 194
303 BHP's Petroleum Exposure Is Its Great Diversifier (Generating 69% EBITDA Margins in 2012)… 195
304 …While Anglo American's Great Diversifier, Platinum, Came in at Margins of Just 11% 195
305 We Also See Revenue and EBITDA Growth… 195
306 …For BHP Billiton 195
307 Anglo and Glencore Xstrata Represent the Tier 2 of Mining Companies Behind the "Big Three" of Vale, Rio and BHP 197
308 The Differentiating Feature of the "Big Three," Both in Terms of Size and Margin, Is the Presence of Significant Iron Ore Exposure 197
309 Anglo Is the Most Diversified and the Lowest-Risk Miner from a Commodity Exposure Perspective 197
310 The Relative Lack of "Safe" Australian Exposure Has Always Been a Risk Compared to the Other Miners 197
311 Historical Evolution of Anglo American's Revenue by Business Unit 198
312 Historical Evolution of Anglo American's EBITDA by Business Unit 198
313 An Increasing Proportion of Anglo Platinum's Refined Production Now Comes from Non-Mined Sources… 199
314 …The Majority Is Purchased from Related Joint Venture Partners and Associates 199
315 Our Forecast for Anglo American Sees Revenue Growth… 199
316 …As Well as EBITDA Growth as Part of the Turnaround Story 199
317 Glencore Xstrata's High-Revenue, Low-Margin Trading Business Renders Its Group Profile Very Different from the Pure Miners in Our Coverage… 201
318 …Though Excluding the Marketing Division Reveals a Profile That Is Similar to the Rest of the Miners 201
319 The Impact of the Marketing Division Can Be Seen in Total Company Margins… 201
320 …While the Margins for Only Industrial Activities More Closely Resemble the Pure Miners in Our Coverage 201
321 Historical Pro Forma Evolution of Glencore Xstrata's Industrial Activities EBITDA 202
322 Glencore Xstrata's Industrial Activities Are Predominantly a Copper, Coal and Zinc Play… 202
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 293
323 …As Can Also Be Seen When Looking at the Company in Toto (Inclusive of Marketing) 202
324 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our Current Price Target (as a Result of Our Bullish View on Copper in the Medium Term) 205
325 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest 206
326 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent 206
327 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices 206
328 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price 206
329 Rio Tinto Scenario Summary vs. Consensus 207
330 BHP Billiton Scenario Summary vs. Consensus 207
331 Anglo American Scenario Summary vs. Consensus 207
332 Glencore Xstrata Pro Forma Scenario Summary vs. Consensus 208
333 Vale Scenario Summary vs. Consensus 208
334 Both the "Grizzly"... 209
335 ...And "Bear" Scenarios Show Revenue Below Consensus Expectations 209
336 Over the Next Two Years What We Call the "Bull" Scenario Is Closest to Consensus Expectation 209
337 While Under the Strongest Price Scenario, There Is Upside to Consensus 209
338 The Same Feature Is Seen at the EBITDA Line... 210
339 ...The Lack of Coverage of Glencore Makes Comparison More Difficult 210
340 Again, the "Bull" Scenario Appears Closest to Consensus 210
341 Considerable Upside Under Stronger Commodity Scenarios Emphasizes the Degree of Operationally Geared Exposure the Miners Offer to Commodity Prices 210
342 Rio Tinto Scenario Analysis Summary 211
343 Rio Tinto EV by Division 211
344 Rio Tinto — Current Model — Pro Forma Financials 212
345 Rio Tinto — Current Model — Valuation 212
346 Rio Tinto — Grizzly — Pro Forma Financials 213
347 Rio Tinto — Grizzly — Valuation 213
348 Rio Tinto — Bear — Pro Forma Financials 214
349 Rio Tinto — Bear — Valuation 214
350 Rio Tinto — Bull — Pro Forma Financials 215
351 Rio Tinto — Bull — Valuation 215
352 Rio Tinto — De Niro — Pro Forma Financials 216
353 Rio Tinto — De Niro — Valuation 216
354 BHP Billiton Scenario Analysis Summary 217
355 BHP Billiton EV by Division 217
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
294 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
356 BHP Billiton — Current Model — Pro Forma Financials 218
357 BHP Billiton — Current Model — Valuation 218
358 BHP Billiton — Grizzly — Pro Forma Financials 219
359 BHP Billiton — Grizzly — Valuation 219
360 BHP Billiton — Bear — Pro Forma Financials 220
361 BHP Billiton — Bear — Valuation 221
362 BHP Billiton — Bull — Pro Forma Financials 222
363 BHP Billiton — Bull — Valuation 222
364 BHP Billiton — De Niro — Pro Forma Financials 223
365 BHP Billiton — De Niro — Valuation 223
366 Anglo American Scenario Analysis Summary 224
367 Anglo American EV by Division 224
368 Anglo American — Current Model — Pro Forma Financials 225
369 Anglo American — Current Model — Valuation 225
370 Anglo American — Grizzly — Pro Forma Financials 226
371 Anglo American — Grizzly — Valuation 226
372 Anglo American — Bear — Pro Forma Financials 227
373 Anglo American — Bear — Valuation 227
374 Anglo American — Bull — Pro Forma Financials 228
375 Anglo American — Bull — Valuation 228
376 Anglo American — De Niro — Pro Forma Financials 229
377 Anglo American — De Niro — Valuation 229
378 Glencore Xstrata Pro Forma Scenario Analysis Summary 230
379 Glencore Xstrata Pro Forma EV by Division 230
380 Glencore Xstrata Pro Forma — Current Model — Pro Forma Financials 231
381 Glencore Xstrata Pro Forma — Current Model — Valuation 232
382 Glencore Xstrata Pro Forma — Grizzly — Pro Forma Financials 233
383 Glencore Xstrata — Grizzly — Valuation 234
384 Glencore Xstrata Pro Forma — Bear — Pro Forma Financials 235
385 Glencore Xstrata Pro Forma — Bear — Valuation 236
386 Glencore Xstrata — Bull — Pro Forma Financials 237
387 Glencore Xstrata Pro Forma — Bull — Valuation 238
388 Glencore Xstrata — De Niro — Pro Forma Financials 239
389 Glencore Xstrata — De Niro — Valuation 240
390 Vale Scenario Analysis Summary 241
391 Vale Pro Forma EV by Division 241
392 Vale — Current Model — Pro Forma Financials 242
393 Vale — Current Model — Valuation 242
394 Vale — Grizzly — Pro Forma Financials 243
395 Vale — Grizzly — Valuation 243
396 Vale — Bear — Pro Forma Financials 244
397 Vale — Bear — Valuation 244
398 Vale — Bull — Pro Forma Financials 245
399 Vale — Bull — Valuation 245
400 Vale — De Niro — Pro Forma Financials 246
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 295
401 Vale — De Niro — Valuation 246
402 Long-Term Real Commodity Price Scenarios 247
403 Nickel Spot Price Distribution 248
404 Zinc Spot Price Distribution 248
405 Aluminum Spot Price Distribution 249
406 Platinum Spot Price Distribution 249
407 Palladium Spot Price Distribution 250
408 Gold Spot Price Distribution 250
409 Silver Spot Price Distribution 251
410 Thermal Coal Spot Price Distribution 251
411 Molybdenum Spot Price Distribution 252
412 Uranium Spot Price Distribution 252
413 Consensus Scenario 254
414 IMF Scenario 255
415 Back to the Future 256
416 Ending With a Whimper 257
417 It's All Over Now Baby Blue 258
418 Rio Tinto Income Statement 259
419 Rio Tinto Balance Sheet 260
420 Rio Tinto Cash Flow Statement 261
421 BHP Income Statement 262
422 BHP Balance Sheet 263
423 BHP Cash Flow Statement 264
424 Vale Income Statement 265
425 Vale Balance Sheet 266
426 Vale Cash Flow Statement 267
427 Anglo American Income Statement 268
428 Anglo American Balance Sheet 269
429 Anglo American Cash Flow Statement 270
430 Glencore Xstrata Pro Forma Income Statement (Including Synergies) 271
431 Glencore Xstrata Pro Forma Balance Sheet (Including Synergies) 272
432 Glencore Xstrata Pro Forma Cash Flow Statement (Including Synergies) 273
433 Glencore Xstrata Pro Forma Income Statement (Excluding Synergies) 274
434 Glencore Xstrata Pro Forma Balance Sheet (Excluding Synergies) 275
435 Glencore Xstrata Pro Forma Cash Flow (Excluding Synergies) 276
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
296 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
Disclosure Appendix
VALUATION METHODOLOGY
Mining companies are operationally and financially geared to their underlying commodity exposure, so we provide two valuation metrics (see the "Valuation and Risks" chapter of this Blackbook for more details):
• ~80% of weekly mining equity price moves can be explained by underlying commodity price moves, so we use a regression-based trading model and our commodity price forecasts to help determine our 12-month price targets. If the regression remains stable or deviations appear temporary, the model determines the target price. If we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made.
• We additionally provide a supplementary DCF-based valuation constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is applied for the remaining years of potentially exploitable material. We forecast our models in reporting currency (USD), convert to listing currency (GBP or Real), and round final DCF values in 25p/cent increments.
RISKS
The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources), and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation). For more details on both sector risks and company-specific risks, see the risk section in the "Valuation and Risks" chapter.
SRO REQUIRED DISCLOSURES
References to "Bernstein" relate to Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and Sanford C. Bernstein (business registration number 53193989L), a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, collectively.
Bernstein analysts are compensated based on aggregate contributions to the research franchise as measured by account penetration, productivity and proactivity of investment ideas. No analysts are compensated based on performance in, or contributions to, generating investment banking revenues.
Bernstein rates stocks based on forecasts of relative performance for the next 6-12 months versus the S&P 500 for stocks listed on the U.S. and Canadian exchanges, versus the MSCI Pan Europe Index for stocks listed on the European exchanges (except for Russian companies), versus the MSCI Emerging Markets Index for Russian companies and stocks listed on emerging markets exchanges outside of the Asia Pacific region, and versus the MSCI Asia Pacific ex-Japan Index for stocks listed on the Asian (ex-Japan) exchanges - unless otherwise specified. We have three categories of ratings:
Outperform: Stock will outpace the market index by more than 15 pp in the year ahead.
Market-Perform: Stock will perform in line with the market index to within +/-15 pp in the year ahead.
Underperform: Stock will trail the performance of the market index by more than 15 pp in the year ahead.
Not Rated: The stock Rating, Target Price and estimates (if any) have been suspended temporarily.
As of 07/11/2013, Bernstein's ratings were distributed as follows: Outperform - 39.6% (0.9% banking clients) ; Market-Perform - 47.2% (0.0% banking clients); Underperform - 13.2% (0.0% banking clients); Not Rated - 0.0% (0.0% banking clients). The numbers in parentheses represent the percentage of companies in each category to whom Bernstein provided investment banking services within the last twelve (12) months.
Accounts over which Bernstein and/or their affiliates exercise investment discretion own more than 1% of the outstanding common stock of the following companies RIO.LN / Rio Tinto PLC, BLT.LN / BHP Billiton PLC.
This research publication covers six or more companies. For price chart disclosures, please visit www.bernsteinresearch.com, you can also write to either: Sanford C. Bernstein & Co. LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105 or Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom; or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE 297
12-Month Rating History as of 07/14/2013
Ticker Rating Changes
AAL.LN O (IC) 09/05/12 O (DC) 07/29/11
BBL O (IC) 09/26/12 O (DC) 07/29/11
BHP O (IC) 09/05/12
BHP.AU O (IC) 09/26/12
BLT.LN O (IC) 09/05/12 O (DC) 07/29/11
GLEN.LN O (RC) 02/13/13 M (IC) 09/05/12
RIO O (IC) 09/05/12 O (DC) 07/29/11
RIO.LN O (IC) 09/05/12 O (DC) 07/29/11
VALE O (RC) 06/07/13 U (IC) 09/05/12
VALE3.BZ O (RC) 06/07/13 U (IC) 09/05/12
Rating Guide: O - Outperform, M - Market-Perform, U - Underperform, N - Not Rated
Rating Actions: IC - Initiated Coverage, DC - Dropped Coverage, RC - Rating Change
OTHER DISCLOSURES
A price movement of a security which may be temporary will not necessarily trigger a recommendation change. Bernstein will advise as and when coverage of securities commences and ceases. Bernstein has no policy or standard as to the frequency of any updates or changes to its coverage policies. Although the definition and application of these methods are based on generally accepted industry practices and models, please note that there is a range of reasonable variations within these models. The application of models typically depends on forecasts of a range of economic variables, which may include, but not limited to, interest rates, exchange rates, earnings, cash flows and risk factors that are subject to uncertainty and also may change over time. Any valuation is dependent upon the subjective opinion of the analysts carrying out this valuation.
This document may not be passed on to any person in the United Kingdom (i) who is a retail client (ii) unless that person or entity qualifies as an authorised person or exempt person within the meaning of section 19 of the UK Financial Services and Markets Act 2000 (the "Act"), or qualifies as a person to whom the financial promotion restriction imposed by the Act does not apply by virtue of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005, or is a person classified as an "professional client" for the purposes of the Conduct of Business Rules of the Financial Conduct Authority.
To our readers in the United States: Sanford C. Bernstein & Co., LLC is distributing this publication in the United States and accepts responsibility for its contents. Any U.S. person receiving this publication and wishing to effect securities transactions in any security discussed herein should do so only through Sanford C. Bernstein & Co., LLC.
To our readers in the United Kingdom: This publication has been issued or approved for issue in the United Kingdom by Sanford C. Bernstein Limited, authorised and regulated by the Financial Conduct Authority and located at 50 Berkeley Street, London W1J 8SB, +44 (0)20-7170-5000.
To our readers in member states of the EEA: This publication is being distributed in the EEA by Sanford C. Bernstein Limited, which is authorised and regulated in the United Kingdom by the Financial Conduct Authority and holds a passport under the Markets in Financial Instruments Directive.
To our readers in Hong Kong: This publication is being distributed in Hong Kong by Sanford C. Bernstein (Hong Kong) Limited which is licensed and regulated by the Hong Kong Securities and Futures Commission (Central Entity No. AXC846). This publication is solely for professional investors only, as defined in the Securities and Futures Ordinance (Cap. 571).
To our readers in Singapore: This publication is being distributed in Singapore by Sanford C. Bernstein, a unit of AllianceBernstein (Singapore) Ltd., only to accredited investors or institutional investors, as defined in the Securities and Futures Act (Chapter 289). Recipients in Singapore should contact AllianceBernstein (Singapore) Ltd. in respect of matters arising from, or in connection with, this publication. AllianceBernstein (Singapore) Ltd. is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C. It is regulated by the Monetary Authority of Singapore and located at 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712, +65-62304600. The business name "Sanford C. Bernstein" is registered under business registration number 53193989L.
To our readers in Australia: Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited and Sanford C. Bernstein (Hong Kong) Limited are exempt from the requirement to hold an Australian financial services licence under the Corporations Act 2001 in respect of the provision of the following financial services to wholesale clients:
providing financial product advice;
dealing in a financial product;
making a market for a financial product; and
providing a custodial or depository service.
Sanford C. Bernstein & Co., LLC., Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited and AllianceBernstein (Singapore) Ltd. are regulated by, respectively, the Securities and Exchange Commission under U.S. laws, by the Financial Conduct Authority under U.K. laws, by the Hong Kong Securities and Futures Commission under Hong Kong laws, and by the Monetary Authority of Singapore under Singapore laws, all of which differ from Australian laws.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
298 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING
AND LOVE THE ORE
One or more of the officers, directors, or employees of Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, and/or their affiliates may at any time hold, increase or decrease positions in securities of any company mentioned herein.
Bernstein or its affiliates may provide investment management or other services to the pension or profit sharing plans, or employees of any company mentioned herein, and may give advice to others as to investments in such companies. These entities may effect transactions that are similar to or different from those recommended herein.
Bernstein Research Publications are disseminated to our customers through posting on the firm's password protected website, www.bernsteinresearch.com. Additionally, Bernstein Research Publications are available through email, postal mail and commercial research portals. If you wish to alter your current distribution method, please contact your salesperson for details.
Bernstein and/or its affiliates do and seek to do business with companies covered in its research publications. As a result, investors should be aware that Bernstein and/or its affiliates may have a conflict of interest that could affect the objectivity of this publication. Investors should consider this publication as only a single factor in making their investment decisions.
This publication has been published and distributed in accordance with Bernstein's policy for management of conflicts of interest in investment research, a copy of which is available from Sanford C. Bernstein & Co., LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105, Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom, or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712. Additional disclosures and information regarding Bernstein's business are available on our website www.bernsteinresearch.com.
CERTIFICATIONS
I/(we), Paul Gait, Senior Analyst(s)/Analyst(s), certify that all of the views expressed in this publication accurately reflect my/(our) personal views about any and all of the subject securities or issuers and that no part of my/(our) compensation was, is, or will be, directly or indirectly, related to the specific recommendations or views in this publication.
Approved By: NK
Copyright 2013, Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and AllianceBernstein (Singapore) Ltd., subsidiaries of AllianceBernstein L.P. ~1345 Avenue of the Americas ~ NY, NY 10105 ~212/756-4400. All rights reserved.
This publication is not directed to, or intended for distribution to or use by, any person or entity who is a citizen or resident of, or located in any locality, state, country or other jurisdiction where such distribution, publication, availability or use would be contrary to law or regulation or which would subject Bernstein or any of their subsidiaries or affiliates to any registration or licensing requirement within such jurisdiction. This publication is based upon public sources we believe to be reliable, but no representation is made by us that the publication is accurate or complete. We do not undertake to advise you of any change in the reported information or in the opinions herein. This publication was prepared and issued by Bernstein for distribution to eligible counterparties or professional clients. This publication is not an offer to buy or sell any security, and it does not constitute investment, legal or tax advice. The investments referred to herein may not be suitable for you. Investors must make their own investment decisions in consultation with their professional advisors in light of their specific circumstances. The value of investments may fluctuate, and investments that are denominated in foreign currencies may fluctuate in value as a result of exposure to exchange rate movements. Information about past performance of an investment is not necessarily a guide to, indicator of, or assurance of, future performance.
For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013
EUROPEAN METALS & MINING: A STRANGE LOVE – HOW I LEARNED TO STOP WORRYING AND LOVE THE ORE