7/30/2019 SEB report: Expanding the universe
1/12
X-asset themes#8, 2012: Expanding the universe
Traditional assets work in the long run,but they dont diversify macro risks well.Expanding the universe can increaserisk-adjusted return, but risks remain.
25 NOVEMBER 2012
7/30/2019 SEB report: Expanding the universe
2/12
2
X-asset themes
CONTENTS
Summary: Expanding the universe................ ................ ................ ................ ................ ................. ................ ................ ................ ......... 3Traditional assets long-term track record with flaws........................................................................................................................4Expanding the universe with non-traditional assets ............................................................................................................................7Constructing long-tem portfolios using the full tool-box....................................................................................................................9Cyclical risk in a balanced, expanded portfolio...................................................................................................................................10
Lead analyst on this study: Kristina Styf
THE SEB X-ASSET TEAMThomas Thygesen+45 33281008
Kristina Styf+46 8 50623048
Jakob Lage Hansen+45 33281469
Johan Lundgren+46 8 50623246
KEY CONCLUSIONS: EXPANDING THE UNIVERSE
TRADITIONAL ASSETS - LONG-TERM TRACK RECORD WITH FLAWS Traditional assets appeal to investors due totheir long track records, transparency and their intuitive roles in the economic system. However, both bonds and
equities have extended periods of real losses and are not good at diversifying macro risks: equities only perform wellin a favourable growth environment and bonds need low and stable inflation.
EXPANDING THE UNIVERSE WITH NON-TRADITIONAL ASSETS In this study we expand the investment universe toinclude commodity futures, hedge funds and alternative betas. They have less historical backing and risk-adjustedreturns are similar to traditional assets, but they offer powerful diversification effects. Commodities hedge inflation
risks, while diversifying hedge fund strategies and alternative betas have acyclical, uncorrelated returns
CONSTRUCTING LONG-TERM PORTFOLIOS USING THE FULL TOOL BOX Using the expanded universe as basis forportfolio construction, we get a high optimal allocation to non-traditional assets for all risk levels. In a 15% Value- at-Risk long-term portfolio, the optimal weight is above 50%. Alternative assets displace fixed income assets, but allowan increase in the equity allocation that leads to an expected extra annual return of around 0.6%.
CYCLICAL RISK IN AN EXPANDED PORTFOLIO Alternative assets improve long-term portfolio returns, but notenough to eliminate periods of sub-trend returns of a cyclical nature. Using the SEB Waves cyclical framework, wefind that while alternative returns are less correlated to macro risks, the expanded portfolio still suffers systematicepisodes of sub-par performance during cyclical setbacks on all horizons.
7/30/2019 SEB report: Expanding the universe
3/12
3
X-asset themes
Summary: Expanding the universe
THE CASE FOR A DEEPER ALLOCATION FRAMEWORK WITH MORE ASSET CLASSESAsset allocation models have traditionally focused on a limited set of traditional asset classes: bonds and equities.This is partly for historical economic reasons, as these capital markets were developed early in the capitalist
development and fulfil a clear and crucial role in the economic system. They are also very good long-term investmentswith a history of positive risk premiums extending over more than a century.
The long track records, transparency and relatively intuitive explanations behind traditional assets appeal to investors.Nonetheless, both bonds and equities have, with regular intervals, seen very long periods with flat or negative realreturns, and they struggle to diversify systematic macro risks: equities need a favourable growth environment, bondsneed low and stable inflation. These facts are largely forgotten during long secular bull markets, but equities havenow delivered poor results for more than 10 years, and with zero rates and bond yields at all-time lows, investors have
strong reasons to look for a new investment approach.
Chart 1. Real return and risk, 1927-2011 Chart 2. 12M correlations with macro factors, 1961-2011
In this note, we expand the investment universe to include commodity futures, hedge funds and alternative betas.This work builds on earlier in-depth studies of historical returns, but here we tie all the strains together, analysing all
the different asset classes on the same basis as we look at traditional assets. These assets lack the transparency andtrack record of traditional assets, but they turn out to be less susceptible to the risks that hurt traditional assets. Non-traditional assets are thus not attractive because of their high returns, but because they offer a range of uncorrelatedsources of return the holy grail of asset allocation. Commodities stand out as the only asset class providing inflationprotection. Hedge funds as a group do not offer true diversification, but our exposure is comprised of strategies withproven diversifying effects, and the seven alternative betas we have identified complement each other as well as thetraditional assets on all horizons.
Chart 3. Optimal portfolio traditional assets, 15% VaR Chart 4. Optimal portfolio extended universe, 15% VaR
10%
31%
33%
26%
T-bills
Gvt.bonds
Credit
Equities
Source:GFD, Ecowin and SEBX-asset
The stronger diversification achieved by including non-traditional assets increases the risk-adjusted return of abalanced portfolio, with significant exposure to alternative beta at all risk levels. Based on our forward-lookingestimates, we find an optimal allocation to non-traditional assets of around 54% for a 15% Value- at-Risk long-termportfolio. Alternatives only displace fixed income assets, but allow an increase in the equity allocation that leads to an
expected extra annual return of around 0.6%. However, while the long-term risk-adjusted return improves, includingnon-traditional assets in a balanced portfolio does not eliminate losses in episodes of market distress.
0%
1%
2%
3%
4%
5%
6%
7%
0% 5% 10% 15% 20%
Standard deviation
Realreturn
Source:GFD, Ecowin,Fama & French and SEBX-asset
T-bills
Gvt.bonds
Credit
Equities
Hedge funds combo
Market neutrals
Macro
CTA
Gvt.bonds curvature
Credit premiumDefensive sectors/market
High/Low dividendSmall/Large cap
FX carry
Value/Growth
Alternative betas
Commodities
Note: The combined "Alternative betas" hasbeen constructed asan equal weighted basket
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
-0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00
OECD LEI
US CPI
Source: GFD,Ecowin, Fama &French and SEB X-asset
T-bills
Gvt.bondsCredit
Equities
Commodities
Hedge funds combo
Alternative betasDefensive sectors/market
Gvt.bonds curvatureHigh/Low dividend
Value/Growth
FX carry
Small/Large cap
Credit premium
MarketMacro
CTA
Note: The combined "Alternative betas" has been constructed asan equal weighted basket
32%
44%
2%0%
5% 5%
12%
T-billsGvt.bonds
Credit
Equities
Commodties
Hedge funds
Alternative betas
Source:GFD,Ecowin,Fama & French and SEBX-asset
7/30/2019 SEB report: Expanding the universe
4/12
4
X-asset themes
Traditional assets long-term track record with flawsTraditional assets like government bonds, corporate bonds and equities were introduced early in the industrialdevelopment and still play a clear and crucial economic role. They thus offer an unparalelled long-term track recordwith two centuries of risk premiums ranging from 1-3% for Treasuries and credit to 4-6% for equities. They cover awide range of risks and thus allow investors to combine them for almost any risk level. But from an asset allocation
perspective, they also have important drawbacks: they all suffer from long periods of low or negative real returns andthey are only suited for a narrow range of macro climates on both long and short time horizons.
T-BILLS: RISK-FREE RATE WITH REAL RISKT-bills are a (nominally) risk-free benchmark and anchor for risky assets offering superior liquidity. Nominal losses areextremely rare as rates almost never fall below zero, and real returns are for a majority of developed countries positive
over the very long-term. However, returns are very vulnerable to both long- and short-term inflation.
Chart 5. Real T-bill risk and return, 1900-2011 Chart 6. Real T-bill return, 1835-2011
-4%
-3%
-2%
-1%
0%
1%
2%
3%
0% 5% 10% 15% 20%
Standard deviation
Realreturn
Australia
Belgium
Canada
Denmark
Finland
France
Germany
Ireland
Italy
Japan
Netherlands
New Zealand
Norway
Spain
Sweden
SwitzerlandUKUSA/World/World ex USA
South Africa
Source:Global InvestmentReturns Yearbook 2012
Real US t-bill returns have averaged 1.6% since 1835, but it has not been stable. In the 1800s it was above 3%, butafter a trend break in early 20th century, in connection with the collapse of the Gold Standard, real returns haveaveraged 0.5% with several periods of major losses. For a global comparison we use numbers from Global InvestmentReturns Yearbook (CS 2012) covering data for 19 countries 1900-2011. Their real T-bill returns range from -3.5 to 2%,
with volatility between 4% and 14%. T-bills can thus suffer irrecoverable losses from extreme inflation/currencyshocks, but global exposure can mitigate concentration risk. Our long-term estimate for the real T-bill return is 0.75%.
GOVERNMENT BONDS STABLE RETURNS WITH DEEP INFLATION RISKGovernment bonds offer a safe nominal return, but inflation can lead to real return drawdowns that outlast even themost patient investor. Short-term government bonds are a low risk investment: governments are assumed not todefault (although they sometimes do), so the main risk is inflation which normally moves at a slow speed. However,long bonds have a duration that expose them to slow, but persistent changes in inflation. Historically, both return andrisk have increased with duration with the risk adjusted return rising fastest at the front end of the curve.
Chart 7. Real 10Y US Treasury return, 1820-2012 Chart 8. 10Y rolling correlation to CPI & OECD LEI, 1961-2011
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
CPI
Source: GFD, Ecowin and SEB X-assetSince 1820 10-year US government bonds have had a real annual return of 3%, with a trend change around 1910 afterwhich the return dropped from 4% to 2%, in our view caused by the collapse in the gold standard that also triggered
a sharp decline in real T-bill returns. Bond returns have had an unstable relationship with growth, but they haveinvariably been negatively correlated to inflation. Higher and more variable inflation in the 20th century has thus
reduced the level of real bond returns, even though the risk premium tripled from 0.3% to 0.9% after 1910.
7/30/2019 SEB report: Expanding the universe
5/12
5
X-asset themes
Chart 9. Real bond market risk and return, 1900-2011 Chart 10. US 10Y Treasury risk premium, 1835-2012
-2%
-1%
0%
1%
2%
3%
4%
0% 5% 10% 15% 20% 25%
Standard deviation
Rea
lreturn Australia
Belgium
Canada
Denmark
Finland
France
Germany
Ireland
Italy
Japan
Netherlands
New ZealandNorwaySouth Africa
Spain
Sweden
Switzerland
UK
USA
World
World ex US
Source:Global Investment ReturnsYearbook 2011
Bonds have limited short-term risk the largest 12M real loss is around 20% but the past 100 years have seen twolong periods of sustained, inflation driven losses for US bond-holders, with negative returns in one case extendingover more than 40 years. Meanwhile, 6 of 19 countries in GIRY experienced outright losses over the full period since1900, the same six countries that showed a negative real t-bill return. However, even in these countries the losseswere smaller than those for T-bills, so all 19 countries in GIRY have had positive government bond risk premiums
between 0 and 2.5%. Our forward-looking estimate is for a long-term risk premium of 1.25% and a real return of 2%.
EQUITIES OUTPACING THE ECONOMY - WHEN IT GROWSEquities have an unparalleled long-term return history, but returns are highly variable over all operational horizons.Equity returns consist of a direct return from dividends and volatile changes in the stock price. Over long horizonsstock prices track earnings, so returns come mainly from the dividend yield and earnings growth. Since 1875 each hascontributed around half of the S&P 500s total return, but the contributions have not been stable: price increases
have picked up with nominal growth in the 20 th century, but this has been off-set by a decline in the dividend yield.
Chart 11. S&P 500 real total return, 1820-2012 Chart 12. 10Y rolling correlation to CPI & OECD LEI, 1961-2011
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
CPI
Source: GFD, Ecowin and SEB X-assetReal total S&P 500 return has thus been remarkably stable at around 6% over the past 200 years, but as Chart 11shows the return has been extremely volatile. 12-month losses have exceeded 40% more than once, and there areregular periods of negative returns lasting as long as 20 years. Real equity returns have had a variable correlation to
inflation, with the correlation turning positive when inflation is low, but they have always been negatively correlatedto growth. Indeed, our cyclical analysis shows equity losses are clustered in economic downturns on all time horizons.
Chart 13. Real equity market risk and return, 1900-2011 Chart 14. S&P 500 risk premium, 1835-2012
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10% 15% 20% 25% 30% 35% 40%
Standard deviation
Realreturn
Australia
Belgium
Canada
DenmarkFinland
Fran ce Germany
Ireland
Italy
Japan
Netherlands
New Zealand
Norway
South Africa
Spain
Sweden
Switzerland
UK
USA
World
World ex US
Source: Global InvestmentReturns Yearbook2012
7/30/2019 SEB report: Expanding the universe
6/12
6
X-asset themes
As was the case for bonds, the higher variability of real returns in the 20th century has been accompanied by anincrease in the risk premium. The excess S&P 500 return over T-bills rose from 3% between 1835 and 1910 to morethan 5% after 1910, even though the real return for equities didnt budge. The rise of paper money and inflation has
clearly reduced long-term returns for fixed income assets, but does not appear to have a similar effect on equities.
Over the very long term, equity returns are actually less risky than bond returns: the real returns since 1900 have been
positive for all the 19 countries in the GIRY sample even those that had negative long-term bond returns and riskpremiums are in a range from 2.5-6.5%. Our forward-looking long-term estimate for real equity return is 5.75%,
consistent with a 5% risk premium over T-bills.
CREDIT COMBINING BOND AND EQUITY CHARACTERISTICSPrivate sector bonds share characteristics from both the bond and equity market: they include the fixed nominalpayments of government bonds, but add a yield spread to cover default risk.Corporate bond returns can thus bedivided into a relatively stable return of a similar duration government bond and a more volatile excess return driven
by changes in preceived default risk as well as by actual defaults and rating changes.
Chart 15. Accumulated real credit return, 1915-2012 Chart 16. 5Y rolling correlation to CPI & OECD LEI, 1961-2011
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
CPI
Source: GFD, Ecowin and SEB X-assetEven after factoring in defaults, the long-term excess return is positive, but the total real return has nonethelesssuffered during both short and long periods of government bond losses. The biggest 12-month real losses of around
25% have thus coincided with extreme losses for government bonds. The excess return is correlated with equity beta,as equity capital is a buffer for credit losses, and credit risk can rise in a non-linear way relative to the risk of similargovernment bonds in periods with very low equity prices and high volatility, limiting the ability to hedge against
extreme equity losses. Credit has underperformed government bonds by more than 15% during equity crashes.
Due to the conflicting forces combined in the two parts of corporate bond returns, credits link to macro factors isunstable: the correlation to growth indicators ranges from almost perfect positive correlation to almost perfectnegative correlation, while the correlation to inflation is dominated by the negative effect from duration risk.
Chart 17. Credit vs Treasury return, 1915-2011 Chart 18. Excess return and S&P 500 return, 1915-2012
In practice, the total return from a credit investment will depend on a range of factors with the average credit ratingand the duration of the bonds in front. Our analysis, based on long-term US investment grade bonds, shows a realreturn from US credit of 3.3% since 1915, 1.3% more than government bonds, and a risk premium over T-bills of
around 3%. Our forward-looking long-term estimate is a risk premium of 2.5% and a real return of 3.25%.
7/30/2019 SEB report: Expanding the universe
7/12
7
X-asset themes
Expanding the universe with non-traditional assets
The common thread from the analysis of the traditional assets is clear: they all have long track records and aretransparent, liquid and credible markets. Nonetheless, they have all seen very long periods with flat or negative realreturns, and they struggle to diversify systematic macro risk in particular, balanced portfolios tend to do well onlywhen growth is strong and inflation is neither too high nor too low. These conditions have not been in place for morethan a decade, and this naturally increases the interest in finding more robust alternatives. In this section weintroduce alternative asset classes: commodity futures, hedge funds and alternative betas. They do not offer higher
long-term risk-adjusted returns, but have other advantages in a portfolio context.
COMMODITIES FUTURES MAINLY FOR INFLATION PROTECTIONUnlike traditional assets, investing in commodity futures does not provide capital for corporates they are essentiallybets on the future price of commodity spot prices. Long-term, commodity futures offer a positive risk premium iffutures prices are set lower than the expected future spot price, and this has historically been the case becausecommodity producers are willing or forced to pay for downside price protection. Short-term commodity returns aremainly driven by unexpected spot price changes and hence have a strong correlation with actual price developments.As a result, commodities stand out among all assets as the only one to show a clear, positive correlation to inflation.
Chart 19. Historical real risk and return, 1915-2012 Chart 20. 12M correlations with macro factors, 1970-2012
Our commodity index is calculated from three sources: 1915-1950 data is based on reconstructed futures contractsfrom GFD, 1951-1969 results from a study by Gorton & Rouwenhorst (2005), and starting 1970 it is an equal weightedtotal return index based on four GSCI commodity sub-groups precious metals, industrial metals, agriculture andenergy. Academic studies show a risk premium from commodities of around 5-7% over the last 50 years, but this wasa favorable period for commodity returns. Using long-term data, we estimate the historical risk premium to have been3%since 1915, at a risk level a bit below equities, and this is also our long-term forward-looking estimate.
Chart 21. Distribution of 12 months real returns, 1915-2011 Chart 22. 5Y rolling correlation to CPI & OECD LEI, 1970-2011
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
CPI
Source: GFD,Ecowin and SEB X-asset The distribution of commodity returns clearly deviates from the normal distribution and follow that of equity returnsrather well with a similar 12 month average. Both have rather fat tails with maximum 12-month losses well above30%, but equities have more extreme observations on both the upside and downside. Correlations with macro risksare very different from the traditional assets: real commodity futures returns are almost invariably positively
correlated with inflation, while the correlation to growth indicators is more variable. Although commodities have arather low return per unit of risk compared with equities, a small, diversified exposure is thus a valuable addition totraditional assets in a balanced portfolio thanks to the diversification benefits.
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
-0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80
OECD LEI
US CPI
Commodities
Credit
T-bills
Gvt.bonds
Equities
Industrial metals
Precious metals
Energy
Agriculture
Source: GFD,Ecowin and SEB X-asset
0%
10%
20%
30%
40%
50%
40%
Equities CommoditiesSource:GFD, Ecowinand SEB X-asset
0%
2%
4%
6%
8%
10%
12%
0% 5% 10% 15% 20% 25% 30% 35%
Standard deviation
Realreturn
T-bills
Gvt.bonds
CreditEquities
Commodities
Source:GFD, Ecowin,SEB X-asset
Industrial metalsPrecious metals
Agriculture
Energy
7/30/2019 SEB report: Expanding the universe
8/12
8
X-asset themes
HEDGE FUNDS ACTIVE DIVERSIFYING STRATEGIES WITH ALPHA POTENTIALHedge funds add value through skill and access to deep risk premiums. Many hedge funds also have the possibility touse leverage which is not always accessible for individual investors. The hedge fund universe return is dominated byequity beta and has not delivered the promised uncorrelated returns. This means selection is key hedge funds canadd diversification in a portfolio, but it is essential to identify hedge funds with targeted characteristics. Our hedgefund exposure is comprised of three strategies, based upon a X-asset study (Hedge funds avoiding a simplisticapproach, Hansen 2012): equity market neutral, macro and CTA funds. Each strategy has been a better macro hedge
historically and a combination of them has attractive diversification characteristics unlike hedge funds on average.
Chart 23. Real total return and risk, 1915-2012 Chart 24. 12M correlations with macro factors, 1961-2012
Due to limited history of around 20 years of realized returns, our analysis uses risk factor replication of the beta returnof hedge funds by several risk factors. The three strategies complement each other in balanced portfolios of most risklevels and would be included even with a conservative forward-looking estimate without alpha assumptions due totheir attractive diversification characteristics. We find that a basket of the strategies would have had a real return of2.5% and a risk premium of 2.1% since 1915 close to uncorrelated with macro risks. Our estimate for the forward-
looking risk premium of a balanced portfolio of diversifying hedge funds is 2% at a risk between Treasuries and credit.
Chart 25. Distribution of 12 months real returns, 1915-2011 Chart 26. 5Y rolling correlation to CPI & OECD LEI, 1961-2011
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.81.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
CPI
Source: GFD, Ecowin and SEB X-asset The distribution of returns from our hedge fund basket is more concentrated than that of equities, with much smaller
tails. Real 12-month losses peak around 20%, similar to government bonds. But unlike government bonds, the hedgefund basket is not negatively correlated to inflation. Correlations with macro risk factors are less volatile than thosefor traditional assets and average correlation is very close to the center of Chart 26, highlighting the attractivediversification characteristics.
ALTERNATIVE BETAS PASSIVE DIVERSIFYING STRATEGIESAlternative betas are risk premiums embedded within capital markets that are not fully captured by traditional assetbetas. Alternative betas have raised hopes of uncorrelated returns, but they must lift the same burden of proof asother assets. In order to put them on a level playing field with other assets, we have required that they must haveproven performance over many decades, an intuitive explanation and backing from academic studies.
In this analysis companion study, The role of Alternative Betas in long-term portfolios, we identify seven riskpremiums satisfying these requirements: Government bonds curvature, Credit premium, FX carry, Defensive sectors,
Value premium, Dividend premium and Small cap premium. They have all been analysed on a stand-alone,unleveraged basis using >85 years of data and are calculated as long/short baskets with collateral interest. Costs have
been included, between 40-75 bps depending on underlying assets, for all alternative betas except curvature.
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
-0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80
OECD LEI
US CPI
Source: GFD,Ecowin, Fama & French and SEBX-asset
T-bills
Gvt.bondsCredit
Equities
Hedge funds combo
Market neutralsMacro
CTAHedge funds universe
0%
10%
20%
30%
40%
50%
40%
Equit ies Hedge funds comboSource:GFD, Ecowinand SEB X-asset
0%
1%
2%
3%
4%
5%
6%
7%
0% 5% 10% 15% 20%
Standard deviation
Realreturn
Source:GFD, Ecowin,Fama & French and SEBX-asset
T-bills
Gvt.bonds
Credit
Equities
Hedge funds combo
Market neutrals
Macro
CTA
Hedge funds universe
7/30/2019 SEB report: Expanding the universe
9/12
9
X-asset themes
Chart 27. Real total return and risk, 1927-2012 Chart 28. 12M correlations with macro factors, 1961-2011
As with commodity futures, a majority of the individual alternative betas have rather low historical risk-adjustedreturn but attractive correlation benefits. All but one of the seven strategies have historically been below the risk-return trend line of traditional assets. Like equities the performance from individual alternative betas vary over time.Some strategies have delivered flat or even negative returns over periods as long as 10-20 years. On the upside thealternative betas are close to uncorrelated with inflation with a tilt towards negative correlation to leading indicators.
They are also internally independent, so an en equal weighted basket has historically had a real return of 2.2% with arisk around 5% like government bonds. Our forward-looking risk premium estimate is 1.6%.
Chart 29. Distribution of 12 months real returns, 1927-2011 Chart 30. 12M correlations with macro factors, 1961-2011
12 month alternative beta returns are clustered around zero with relatively small tails and mainly on the upside. Over40% of the observations are within the 0-5% return bucket and more than 90% of the observations are coveredwithin -5% to +10% return. 12-month real losses peak below 15%, somewhat lower than for government bonds.Correlations with macro factors for the equal weighted alternative beta basket has been variable, but 10-yearobservations are generally scattered evenly around the centre, with the average being very close to uncorrelated to
both inflation and leading indicators as well as to the traditional asset class returns.
Constructing long-tem portfolios using the full tool-boxBased on these forward looking risk and return estimates and historical correlations since 1927 we optimise a long-term passive portfolio including both traditional assets and our expanded universe. As a group, the new assets offer
similar risk-adjusted returns to traditional assets, but we can now populate a larger part of the correlation chart.
Chart 31. Forward-looking real return and risk Chart 32. 12M correlation with macro factors, 1961-2011
0%
1%
2%
3%
4%
5%
6%
0% 5% 10% 15% 20%
Standard deviation
Realreturn
Source:GFD, Ecowin,Fama & French and SEBX-asset
T-bills
Gvt.bonds
Credit
Equities
Hedge funds combo
Alternative betas
Commodities
*Trading costsof 75 bps hasbeen included forequity related premiums,
25bps for the creditpremium and 50 bpsfor FX carry
Note: The combined "Alternatives" has been constructed by equal weighting the alternative betas
0%
10%
20%
30%
40%
50%
40%
E qui ti es A lte rn ativ e be ta sSource:GFD, Ecowin,Fama & French and SEBX-assetNote: The combined "Alternative betas" hasbeen constructed asan equal weighted basket
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
OECD LEI
US CPI
Source: GFD, Ecowin, Fama &French and SEB X-asset
0%
1%
2%
3%
4%
5%
6%
7%
0% 5% 10% 15% 20%
Standard deviation
Real
return
Source:GFD, Ecowin,Fama & French and SEBX-asset
T-bills
Gvt.bonds
Credit
Equities
Gvt.bonds curvature*
Credit premium*
Defensive sectors/market*
High/Low dividend*Small/Large cap*
FX carry*
Value/Growth*
Alternative betas
*Trading costsof 75 bps hasbeen included forequity related premiums,
40bps for the creditpremium and 50 bpsfor FX carry
Note: the combined "Alternative betas" has been constructed as an equal weighted basket
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
-0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80
OECD LEI
US CPI
Source: GFD, Ecowin, Fama & French and SEB X-asset
T-bills
Gvt.bondsCredit
Equities
Alternative betas
Defensive sectors/market
Gvt.bonds curvatureHigh/Low dividend
Value/GrowthFX carry
Small/Large cap
Credit premium
Note: The combined "Alternative betas" has been constructed as an equal weighted basket
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
-0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80
OECD LEI
US CPI
Source:GFD,Ecowin,Fama & French and SEBX-asset
T-bills
Gvt.bonds Credit
Equities
Commodities
Hedge funds comboAlternative betas
Note: The combined "Alternatives" hasbeen constructed byeq ual weighting the alternative betas
7/30/2019 SEB report: Expanding the universe
10/12
10
X-asset themes
We use resampled mean variance optimisation which, as the name suggests, uses resampling in order to avoid someof the well know problems with traditional Markowitz optimisation. Portfolios are optimised with a 12 month timehorizon and a 98% confidence level. The three hedge fund strategies and the seven alternative betas enter theoptimisation as individual strategies. No cost assumptions are included for traditional assets and commodities. Forhedge funds and alternative betas costs are taken into consideration depending on the underlying asset class.
Chart 33. Allocation along the efficient frontier Chart 34. Optimal portfolio extended universe, 15% VaR
Taken as a group, alternative betas have a 25-70% allocation along the risk spectrum thanks to their attractivediversification characteristics. Hedge funds have a rather stable allocation along the efficient frontier at all but thehighest risk levels, while the commodity exposure increases with risk level. In a portfolio at 15% Value-at-Risk non-traditional assets make up 54% of an unconstrained portfolio. They push out fixed income assets - the exposure to t-bills and government bonds is reduced to close to zero and credits to less than half of that in a traditional assets onlyportfolio. Due to the uncorrelated nature of the independent alternative returns sources, we think its fair to call theexpanded portfolio less concentrated, with increased diversification allowing a 6%-point increase in equity allocationwhich creates an additional annual return of 60 bps compared to a traditional portfolio at the same level of risk.
Cyclical risk in a balanced, expanded portfolioIncluding non traditional assets in a balanced portfolio is not likely to eliminate losses in episodes of market distress.
While alternative returns are less correlated to macro risks and can even hedge away some inflation risk, they stillleave a limited capacity to hedge risks from low growth and high inflation.
Chart 35. Real portfolio return by decade, 1930-2009 Chart 36. 12-months rolling real return, USD, 1927-2011
-2%
0%
2%
4%
6%
8%
10%
0% 5% 10% 15% 20% 25% 30% 35%
Value-at-Risk
Realreturn
Source: GFD,Ecowin, Fama & Frenchand SEBX-asset
-45%
-30%
-15%
0%
15%
30%
45%
60%
75%
1928 1938 1948 1958 1968 1978 1988 1998 2008
Traditional portfolio Expanded portfolioSource: GFD,Ecowin, Fama & French and SEB X-asset
Chart 35 shows portfolio return and risk by decade, and even in this limited sample the portfolio has experienced onedecade with no return and one decade with twice the expected risk. Short-term risk has only been marginally reduced,particularly in the high-inflation 1970s, but tail risk remains and real 12-month losses still break the 15% limit during
periods of extreme market stress such as the early 1930s and 2008 more often than they are supposed to.
Adding alternative assets thus improves long-term risk-adjusted returns, but not enough to eliminate recurringperiods of sub-trend returns over several horizons. In order to get a better understanding of the nature of this risk, weuse the SEB Waves cyclical allocation framework, our systematic macro-based cyclical risk analysis based on threemacro cycles across three different time horizons. Each cycle is broken down into four phases using historical macro
indicators and we then analyse phase-dependent investment returns to identify macro-based clusters of risk. Theunderlying idea is that investors should be willing to pay more for risk that is not clustered in the same periods asother assets, and in particular for losses that are not clustered together with losses for the most risky assets.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
5.0% 7.5% 10.0% 12.5% 15.0% 17.5% 20.0% 22.5% 25.0%
Value-at-Risk
T-bills Gvt.bonds Credit Equities Commodties Hedge funds Alternative betas
Source: GFD, Ecowin,Fama & French and SEB X-asset
Note: The alternative betas and the hedge fund strategies have been i ncluded individually
Value-at-Risk based on 98%confidence level and a 12 month time horizon
32%
44%
2%0%
5% 5%
12%
T-bills
Gvt.bonds
Credit
Equities
Commodties
Hedge funds
Alternative betas
Source:GFD,Ecowin,Fama & French and SEBX-asset
7/30/2019 SEB report: Expanding the universe
11/12
11
X-asset themes
Our structural wave reflects a technology cycle of secular bull and bear markets, where each phase on average lasts10-15 years. Traditional assets have their main drawbacks during crises, while non-traditional assets have stable andpositive return in all phases but Golden age boom where returns are positive but lower. The expanded portfolioposts a positive real return in all phases, but in the high inflation capacity shortage phase the real return has been lessthan 1% while the low-inflation overcapacity crisis phase risk has significantly exceeded the long-term average.
Chart 37. Total real return in structural phases, 1927-2011 Chart 38. Portfolio performance, structural phases, 1927-2011
The strategic wave is based on the output gap cycle: growth above trend or below trend and easing or tighteningcredit conditions determine four phases of the cycle with duration of 1-3 years. Traditional portfolio losses areconcentrated in the first part of a recession. Alternative betas are basically a-cyclical with positive and stable return inall strategic phases, while both commodities and hedge funds have losses in early recession. Improved diversificationis not enough to prevent the expanded portfolio from systematically posting average real losses and elevated risk in
the early recession, mainly due to the significant losses from equities.
Chart 39. Total real return in strategic phases, 1970-2011 Chart 40. Portfolio performance, strategic phases, 1970-2011
The tactical wave is based on the manufacturing inventory cycle, with leading indicators defining 3-6 month phases.Tactical asset class behaviour and portfolio performance is similar to strategic cycle: the sequence of relative returnsis similar, it just happens faster. The most critical phase is early downturn where equities have sharp losses and allnon-traditional assets hold up well. Again, the expanded portfolio has systematic losses at elevated risk in this phase.
Chart 41. Total real return in tactical phases, 1970-2011 Chart 42. Portfolio performance, tactical phases, 1970-2011
Active asset allocation, including the expanded universe, is the subject of a coming study. The study will include adeeper analysis of systematic variations of especially non-traditional assets, the implications on portfolio performance
and how phase dependent overlays can dampen portfolio drawdowns and reduce short-term risks.
-5%
0%
5%
10%
15%
20%
Capacity shortage
crisis
New era boom Overcapacity crisis Golden age boom
T-bills Gvt.bonds Credit Equities Commodities Hedge funds Alternative betasSource: GFD,Ecowin, Fama & French and SEB X-asset
-20%
-15%
-10%
-5%
0%
5%
10%
15%
20%
25%
30%
Early recession Late recession Early expansion Late expansion
T-bills Gvt.bonds Credit Equities Commodities Hedge funds Alternative betas
: , ,
Source: GFD,Ecowin, Fama & French and SEB X-asset
-20%
-15%
-10%
-5%
0%
5%
10%
15%
20%
25%
30%
Early downturn Late downturn Early upswing Late upswing
T-bills Gvt.bonds Credit Equities Commodities Hedge funds Alternative betas
:
Source: GFD,Ecowin, Fama & French and SEB X-asset
0%
2%
4%
6%
8%
10%
12%
0% 5% 10% 15%
Standard deviation
Realreturn
Source:GFD, Ecowin,Fama & French and SEB X-asset
Capacity shortage crisis
Golden age boom
New era boom
Overca acit crisis
Full sample
-4%
-2%
0%
2%
4%
6%
8%
10%
12%
0% 5% 10% 15%
Standard deviation
Realretu
rn
Source: GFD, Ecowin, Fama &French and SEB X-asset
Early recession
Late expansionLate recession
Early expansion
Full sample
-5%
0%
5%
10%
15%
20%
0% 5% 10% 15%
Standard deviation
Realreturn
Source: GFD, Ecowin,Fama & French and SEB X-asset
Early downturn
Late upswing
Late downturn
Early upswing
Full sample
7/30/2019 SEB report: Expanding the universe
12/12
www.seb.se/mb
STOCKHOLM
Kungstrdgrdsgatan 8Merchant Banking DivisionSE-106 40 StockholmTelephone: +46 8 763 80 00
CAPITAL MARKET BOND SALES
Telephone: +46 8 506 231 70
ENSKILDA EQUITY SALES
Telephone +46 8 522 29 500
COMMODITIES SALES
Telephone +46 8 763 91 93
FX SALES
Telephone: +46 8 506 23 140
EMERGING MARKETS CROSS ASSETS
Telephone +46 8 506 231 10
GLOBAL FINANCIAL SOLUTIONS
Telephone +46 8 506 233 71
WEALTH MANAGEMENT DIVISION
Sveavgen 8SE 106 40 StockholmTelephone: +46 8 763 50 00
LONDON
2 Cannon StreetGB-London EC4M 6XXTelephone: +44 20 7246 4000Registered in England under
NFC014326BR000979
SEB FUTURES DESK
Telephone: +44 207 489 8964
OSLO
Filipstad Brygge 1NO-0123 OsloTelephone: + 47 22 82 72 03
LUXEMBOURG
6 A, Circuit de la Foire InternationaleL-2014 LuxembourgTelephone: +352 26 23 1
NEW YORK ENSKILDA
245 Park Avenue, 42nd Floor L- NewYork, NY 10167Telephone: +1 (212) 692 4760
COPENHAGEN
Bernstorffsgade 50DK-1577 Copenhagen VTelephone: +45 3328 1400
FRANKFURTMerchant Banking DivisionUlmenstrasse 30DE-60283 FrankfurtTelephone: +49 69 258-0
WEALTH MANAGEMENT DIVISION
Rotfeder-Ring 7DE-60283 Frankfurt am MainTelephone: +49 69 27299-1000
HELSINKI
Unioninkatu 30P.O Box 630FI-00100 HelsinkiTelephone: +358 9 616 280 00
DISCLAIMERImportant: This statement affects your rights
The information in this document has been compiled by Skandinaviska Enskilda Banken AB (publ) (SEB AB) and SEB A/S, a subsidiary to SEB AB. It is
produced for private information of recipients and SEB is not soliciting any action based upon it. All information has been compiled in good faith from
sources believed to be reliable. However, no representation or warranty, expressed or implied, is made with respect to the completeness or accuracy of its
contents and the information is not to be relied upon as authoritative. Recipients are urged to base any investment decisions upon such investigations as
they deem necessary. To the extent permitted by applicable law, no liability whatsoever is accepted for any direct or consequential loss arising from use of
this document or its contents. Any presented performance data is un-audited.
Information relating to taxes may become outdated and may not fit your individual circumstances. Investment products produce a return linked to risk.
Their value may fall as well as rise, and historic returns are no guarantee of future returns; in some cases, losses can exceed the initial amount invested.
Where either funds or you invest in securities denominated in a foreign currency, changes in exchange rates can impact the return. Your attention is drawn
to the fact that SEB, a member of, or any entity associated with, SEB or its affiliates, of ficers, directors, employees or shareholders of such members may
from time to time have holdings in the securities mentioned herein.
THIS INFORMATION IS NOT INTENDED TO BE PUBLISHED OR DISTRIBUTED IN THE UNITED STATES.SEB AB is incorporated in Stockholm, Sweden, with limited liability.
SEB AB is regulated by Finansinspektionen (the Swedish Financial Supervisory Authority).
SEB A/S is incorporated in Copenhagen, Denmark, with limited liability
SEB A/S is regulated by Finanstillsynet (the Danish Financial Supervisory Authority)
Confidentiality Notice
This information is confidential and may not be reproduced or redistributed to any person.