Sector_Tech_201104_JPM_194pg

8/7/2019 Sector_Tech_201104_JPM_194pg

1/194

Technology outlook

Uncertainty in supply chains and end-demand persists

New themes in 2011

See the end pages of this presentation for analyst certification and important disclosures, including non-US analyst disclosures.J.P. Morgan does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may

have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making theirinvestment decision.

Asia Pacific Equity Research

April 2011

JJ ParkAC

+82-2-758-5717

[email protected]

J.P. Morgan Securities (Far East) Ltd, Seoul Branch

Hyunjoon RohAC

+82-2-758-5712

[email protected]


Justin J. ParkAC

+82-2-758-5725

[email protected]


Rick Hsu AC

+886 2 2725 9874

[email protected]

J.P. Morgan Securities (Taiwan) Limited.

Narci Chang AC

+886 2 2725 9899

[email protected]


Cynthia Chou AC

+886 2 2725 9898

[email protected]



2/194

Memory market outlook

2


3/194

Annual DRAM+NAND revenue trend

US$ in billions, %

Source: WSTS, J.P. Morgan estimates

Global memory market: Summary

25 20 14 2129

11 15 1727 26

34 3124 22

39 33 360

00

00

12 5

7 1111 15

12 12

17 19 22

0% 0% 0% 0% 1%7%

13%

21% 20%

29% 25% 32%

33% 36%

31%

37% 37%

-5%

5%

15%

25%

35%

45%

0

10

20

30

4050

60

70

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011E 2012E

DRAM NAND NAND as % of Memory (RHS)

We forecast the global memory market (DRAM+NAND) revenue will increase to US$53 billion for 2011, followed by an

increase of 10% Y/Y in 2012.

- Gloomy demand outlook for PCs in 2011

- Sunny days ahead for NAND thanks to smartphones and tablets

- Increasing portion of Mobile DRAM to help reduce volatility of DRAM market

3


4/194

-60%

-40%

-20%

0%

20%40%

60%

80%

100%

0

4

8

12

16

20

1Q00 1Q01 1Q02 1Q03 1Q04 1Q05 1Q06 1Q07 1Q08 1Q09 1Q10 1Q11E 1Q12E

DRAM NAND Q/Q growth (RHS)

- 2011 DRAM outlook ASP stabilization to be achieved by content growthWe believe growing concerns over a potential shortfall in PC DRAM supply due to a shift in commodity PC DRAM capacity to

mDRAM product and supply-side disruptions post Japans earthquake are overblown and expect this recent recovery inDRAM prices to be short-lived.

We continue to put forward our thesis that DRAM ASP stabilization will be achieved fundamentally by the return of memorycontent growth at PC OEMs/ODMs.

However, we believe the recent hike in DRAM prices will further deter PC makers from adding back memory content.

- 2011 NAND flash outlook Remain positive due to secular growth in smartphones/tablets

According to our estimates, NAND should continue to witness a strong shipment growth of almost 70%+ Y/Y for 2011 and2012, respectively.

We continue to forecast a sequential decline in NAND ASPs, which would help to stimulate its demand for heavy memory-using devices.

Quarterly DRAM+NAND revenue trend

US$ in billions, %

Source: WSTS, J.P. Morgan estimates

Global memory market overview

4


5/194

Quarterly DRAM revenue trend

US$ in billions, %

Source: Companies, J.P. Morgan estimates Source: Companies, J.P. Morgan estimates

Q/Q, %

Annual DRAM revenue trend

DRAM: Revenue trends

- Meaningful revenue erosion for DRAM in 1Q11E

1Q11 DRAM market revenue is expected to decline by a meaningful 17% Q/Q, largely driven by significant erosion in DRAM

prices (down 25% Q/Q).

- DRAM prices to bottom by mid-2Q11; though we expect limited price recovery

Looking ahead, we expect DRAM prices to bottom out by mid-2Q11, largely driven by the return of memory content growth at

PC OEMs/ODMs.

However, contrary to the bulls' camp view; we only expect limited recovery in DRAM prices, largely due to the gloomy PC

demand outlook being echoed throughout the PC supply chain.

In short, we expect the global DRAM market to turn into green in 2Q11, rising by 6% Q/Q, largely due to strong bit-shipment

growth due to seasonal up-ticks in demand, increasing portion of mobile DRAM product and potential stabilization of PC

DRAM prices.

-30%

-15%

0%

15%

30%

45%

0

2

4

68

10

12

1Q09 3Q09 1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

DRAM Revenues Q/Q growth (RHS)

40%

-61%

36%

9%

61%

-5%

32%

-7%-23%

-7%

75%

-15%

9%

-80%

-50%

-20%

10%40%

70%

100%

0

10

20

30

40

50

2000 2002 2004 2006 2008 2010E 2012E

DRAM revenue s Y /Y growth (RHS)

5


6/194

Fully-loaded cost(US$)

Cash Cost(US$)

Node

Samsung 1.6 1.2 46nmHynix 1.7 1.3 44nmInotera 2.5 1.6 50nmNTC 2.5 1.6 50nmPowerchip** 2.5 1.6 63nmRexchip 1.8 0.9 40nm

Elpida (Hiroshima fab) 1.7 1.3 45nm

DRAM ASP trend by quarter

1Gb equivalent, %

Source: DRAMeXchange, J.P. Morgan estimates. Source: Company data, J.P. Morgan. **65nm extra-shrink version.

US$

Cost status for major DRAM makers (2Gb DDR3)

DRAM: ASP trends

- We expect limited price recovery in overall DRAM prices in 2H11

The recent weakness in DRAM spot trends reinforces our concerns about weak PC DRAM demand, due to a combination of

stagnating demand from the developed world, Chinese demand loosing steam, and concerns over limited memory content

growth due to expanding market for tablets.

Contrary to the bulls camps expectation of a V-shaped recovery in PC DRAM prices, we continue to forecast limited price

recovery in overall DRAM prices in 2H11.

- Major DRAM makers cost status

According to our checks with major DRAM makers, fully-loaded costs are hovering around mid-US$1.6-1.8 for major memory

makers, while cash costs are below US$1.6. Hence, major DRAM makers will continue to generate a decent cash profit

unless DRAM prices come below US$1.6, according to our estimates.

4%

8%

-9%

-21%-25%

-5%-2% -2%

-11%-9% -8% -9%

-30%

-20%

-10%

0%

10%

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

DRAM ASP (US$) ASP change, Q/Q (RHS)

6


7/194

Global PC shipments (incl. tablet PCs)

Units in millions, Y/Y

Source: Company data, J.P. Morgan estimates. Source: Company data, J.P. Morgan estimates.

Units in millions, Y/Y

Global PC shipments (excl. tablet PCs))

PC demand outlook

- PC (ex-tablets) unit growth expected to be 7% Y/Y in 2011

Our US PC team recently lowered its PC forecasts for 2011, largely driven by the weakening demand in China, softening

consumer demand, ongoing tablet invasion, decelerating enterprise PC growth and elongating PCs useful life.

Our industry checks also indicate growing concerns over stagnating PC demand from the developed world, while ongoing

tablet invasion seems to defer PC purchase cycle given increased PC usage life due to the reduced time spent on traditional

notebooks and limited per annum consumer budget for technology spending.

We see this as incrementally detrimental for DRAM volume growth and memory content increase, given the gloomy PC

demand outlook, and competitive pricing policies among PC vendors eyeing market share gain under the weak end-demand

scenario.

10%5%

20%

16%14%

0%

5%10%

15%

20%

25%

0

100200

300

400

500

600

2008 2009 2010E 2011E 2012E

Servers Desktops NBPC

Netbooks Tablet PCs Y/Y growth

10%

5%

14%

7% 7%

0%

5%

10%

15%

20%

0

100200

300

400

500

2008 2009 2010E 2011E 2012E

Servers Desktops NBPC

Netbooks Y/Y growth

7


8/194

Global DRAM shipment versus capacity trend

%

Source: Companies, J.P. Morgan estimates

DRAM: Demand by application

- mDRAM: A long-term growth trendAccording to our industry estimates, mobile DRAM would continue to expand its penetration in coming years, accounting for

around 13% and 16% of global DRAM demand in 2011 and 2012, respectively.

We believe mDRAM will have a long-term growth trend in the DRAM industry on the back of robust volume expectations for

modern tech gadgets such as smartphones and tablets.

- Reduced volatility in DRAM industry due to declining share of PC DRAM

PCs global DRAM demand share could continue to decline from 77% in 2010 to 74% by 2012E, according to our estimates.

We believe this would help in reducing the volatility in the DRAM industry given steady volume growth expectations for

smartphones/tablets/IETVs, and gloomy demand outlook for PCs.

78% 78% 76% 76% 75% 75% 76% 76% 75% 74% 74%73%

9% 10% 11% 12% 13% 13% 13% 14% 15% 15% 16% 18%

0%

20%

40%

60%

80%

100%

1Q10 2Q10 3Q10 4Q10 1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E

PC Mobile DRAM Consumer Electronics

Automotive Electronics Industrial Electronics Military/Civil Aerospace Electronics

8


9/194

Global DRAM shipment versus capacity trend

%


(on production-only basis), %

Global DRAM makers M/S trend

DRAM: Supply outlook

- DRAM supply outlookWe estimate that the global DRAM bit-shipment to grow by 45%+ Y/Y in 2011 and 2012, while its capacity is expected to grow

at a marginal 8% and 6% Y/Y in 2011 and 2012, respectively.

Hence, we continue to believe that most of the capex in the DRAM space in the next two years will be dedicated to

technology migration rather than capacity expansion.

- DRAM market share

As per our bottom-up analysis, SEC would continue to maintain its leadership position in the DRAM market, commanding a

market share of ca. 40% and 42% in 2011 and 2012 respectively. Among other major players, Hynix, Elpida (Japan), Micron

and Rexchip would continue to enjoy around ca. 20+%, 7%, 7% and 7% of DRAM market share in 2011E/2012E, respectively

(on a production-only basis).

91

63

22

42 47 4734

17

-10

28 6

-20

0

20

4060

80

100

2007 2008 2009 2010 2011E 2012E

DRAM shipment change, % DRAM capacity change, %

32% 38% 40%42%

23%22% 21% 20%

8% 9% 7% 7%8%

7% 7% 7%11% 9% 7% 7%

0%

20%

40%

60%

80%

100%

2009 2010 2011E 2012E

SEC Hynix Elpida Rexchip Micron Inotera NTC Others

9


10/194

DRAM supply/demand trend

256Mb equivalent million units


DRAM: Supply and demand dynamics

- Global DRAM supply and demand dynamicsAccording to our industry estimates, DRAM supply growth will continue to outpace demand growth until 2Q11, given the

seasonal pull-back in demand and double-digit supply growth.

Thus, we find it hard to believe the concerns over potential supply shortage in 1H11, given the DRAM supply growth is well

above that of demand growth in 1H11.

We estimate that the DRAM industry will likely see a meaningful increase in supply.

Going by conventional wisdom, DRAM prices should have continued trending down given the supply/end-demand mismatch

in the channel.

However, recent surge in DRAM prices implies inventory re-stocking in the channel, amongst fears of a possible shortage of

silicon wafers (due to Japan earthquake) limiting supply growth from 2Q11 onwards.

1%2%

3%

6%

0%

5%

-4% -4%-2%

-5%-4%

-1%

5% 5%

1%

-4%

5%

8%

4%1%

-10%

-5%

0%

5%

10%

0

10,000

20,000

30,000

40,000

50,000

1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

Demand (LHS) Supply (LHS) % Over./(Under) supply (inc. inv.) (RHS)

10


11/194

Memory content growth versus DRAM ASP change

Y/Y %

Source: Company data, J.P. Morgan estimates Source: Company data, J.P. Morgan estimates

Megabyte (MB)

Average memory content comparison

DRAM: Memory content trends

- Below-average industry content growth due to growing penetration of tabletsDespite the meaningful pullback in DRAM prices throughout 4Q10, PC makers showed resistance in adding back memory

and instead focused on offering lucrative price concessions to stimulate end-demand and gain market share in weakening

end-markets.

Though we expect healthy memory content growth momentum to return only in 2H11, desktops and NBPCs should continue

to experience above 30% Y/Y content growth in 2011/2012, respectively; however, annual memory content growth is

expected to remain below historical average levels (30%+) going forward due to the expanding market for tablet PCs.

- Memory content comparison

Tablets and smartphones would continue to enjoy handsome content growth, though their memory content would be much

lower compared to conventional PC devices.

According to our estimates, the average memory content for desktops and NBPC is estimated to be around 4.1-4.3GB and

5.4-5.7GB in 2011 and 2012, respectively, while that of smartphones and tablets would be around 0.5GB/0.7GB and

0.6GB/1GB in 2011 and 2012, respectively.

58% 45% 32% 30% 34% 70% 47% 16% 25% 25% 28%-3%

-22%

0%

-37%-11%

-51% -53%-23%

23%

-42%-26%

-60%

-40%

-20%

0%

20%

40%

60%

80%

2002 2004 2006 2008 2010 2012E

Memory content growth ASP change

10

76

0

5

10

15

0

1,500

3,000

4,500

6,000

2009 2010 2011E 2012E

Desktops NBPC Smartphones Tablets NBPC/Tablet DRAM content multiple

11


12/194

DRAM cost as a percentage of PC cost

%

Source: Company data, J.P. Morgan estimates Source: Company data, J.P. Morgan estimates

%

DRAM + panel as a percentage of PC cost

DRAM: PC BOM analysis

- How many dollars for DRAM?Due to the recent pullback in DRAM prices, our PC cost analysis suggests that DRAM cost as a percentage of PC cost has

trended below the previous 10-year historical average levels. Despite a meaningful price erosion for DRAM , PC makers are

still refraining themselves from adding memory content, amid the weakening end-demand environment.

- Recent price spike to further deter memory content addition

Moreover, a sudden spike in DRAM prices and concerns over PC makers margins under threat due to potential componentshortages from Japan could further deter PC makers from adding back memory content.

- Reiterate our expectation of healthy memory content growth in 2H11, driven by price elasticity

We continue to believe the recent recovery in DRAM prices to be short-lived and reiterate our expectation of healthy memory

content growth in 2H11, driven by price elasticity.

0%2%4%6%8%

10%12%14%16%

1Q001Q01 1Q021Q03 1Q041Q051Q061Q07 1Q081Q09 1Q101Q11E

Desktop NBPC (excluding Netbook)

Average in past 10 years

16%

17%

15%11%

9%11%

14%

16%18%

20%17%

13%12%

0%

5%

10%

15%

20%

25%

1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11E

NBPC (excl. Netbook)

12


13/194

-20%

-10%

0%

10%20%

30%

40%

0

1

2

3

4

5

6

7

1Q09 3Q09 1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

NAND Revenues Q/Q growth (RHS)

Quarterly NAND revenue trend

US$ in billions, %


US$ in billions, %

Annual NAND revenue trend

NAND: Revenue trends

- Marginal decline in 1Q11E revenues; supported by strong shipment growth1Q11 global NAND revenue is expected to decline marginally by 2% Q/Q, largely driven by robust shipment growth due to

smartphones and tablets, which would help offsetting the meaningful ASP erosion (down 15% Q/Q), in our view.

- NAND to continue to enjoy secular growth in coming years due to smartphones/tablets

Although we continue to expect a sequential price decline for NAND throughout 2011 and 2012, we expect price elasticity to

come into action which would help expand NANDs penetration/adoption into the ever-increasing smartphones, tablets and

SSD market.

We expect NAND market revenues to continue experiencing handsome growth trajectory, inclining by 11% Y/Y and 12% Y/Y

in 2011E and 2012E respectively.

Hence, NAND as a % of total memory market should increase to 37% in the next two years from 31% in 2010.

45%

63%

8%

26%

-17%

4%

39%11% 12%

-40%

-20%

0%

20%40%

60%

80%

0

5

10

15

20

25

2003 2005 2007 2009 2011E

NA ND Revenu es Y/Y gro wth (RHS)

13


14/194

NAND demand by application

Source: J.P. Morgan estimates.

Units in millions (1Gb equivalent)

NAND: Demand by application

- Robust growth assumptions for smartphones and tabletsBased on our estimates, smartphones and SSD will be the main drivers for NAND demand in coming years, commanding a

combined share of 53% and 66% of the global NAND demand in 2011 and 2012, respectively.

- MP3 and DSC demand is fading away

On the flip side, MP3Ps/PMPs proportion of NAND demand should continue to decline from 20% in 2008 to less than 7% in

2011. This is mainly due to the meaningful expansion in multimedia-enabled handsets, including smartphones, which usually

support MP3/PMP functions.

In addition, DSC will continue to lose its proportion of NAND demand given its maturing growth curve, in our view.

1Q10 2Q10 3Q10 4Q10 1Q11E 2Q11E 3Q11E 4Q11E 2009 2010 2011E 2012EDigital Camera (DSC) 2,476 2,410 3,141 3,363 2,713 3,102 4,107 4,557 9,554 11,390 14,479 18,192Seq chg % -24% -3% 30% 7% -19% 14% 32% 11% 59% 19% 27% 26%% of total 16% 16% 17% 15% 11% 11% 12% 11% 21% 16% 11% 8%Mobile Phones 4,655 4,867 6,666 7,783 9,319 11,075 12,613 14,698 14,198 23,971 47,706 89,851Seq chg % 3% 5% 37% 17% 20% 19% 14% 17% 72% 69% 99% 88%% of total 31% 32% 36% 35% 39% 38% 37% 35% 31% 34% 37% 38%

USB Flash Drive 1,641 1,644 1,838 1,839 1,946 2,035 2,292 2,449 5,201 6,962 8,721 12,690Seq chg % 12% 0% 12% 0% 6% 5% 13% 7% 63% 34% 25% 46%% of total 11% 11% 10% 8% 8% 7% 7% 6% 11% 10% 7% 5%

MP3P/PMP 1,784 1,670 1,799 2,685 1,898 1,920 2,073 3,137 6,950 7,939 9,028 10,680Seq chg % -29% -6% 8% 49% -29% 1% 8% 51% 19% 14% 14% 18%% of total 12% 11% 10% 12% 8% 7% 6% 8% 15% 11% 7% 5%SSD including tablets 1,232 1,639 2,128 3,121 4,480 6,757 9,188 12,156 2,076 8,120 32,580 79,059Seq chg % 53% 33% 30% 47% 44% 51% 36% 32% 182% 291% 301% 143%% of total 8% 11% 11% 14% 19% 23% 27% 29% 5% 11% 25% 34%Ohers 3,234 2,761 3,163 3,469 3,681 3,980 4,265 4,611 7,526 12,626 16,537 23,851Seq chg % 28% -15% 15% 10% 6% 8% 7% 8% 53% 68% 31% 44%% of total 22% 18% 17% 16% 15% 14% 12% 11% 17% 18% 13% 10%

Total 15,021 14,991 18,735 22,261 24,036 28,868 34,537 41,609 45,506 71,008 129,051 234,324

Seq chg % 0 0 25 19 8 20 20 20 57 56 82 82 14


15/194

Global NAND flash bit supply and capacity growth

Y/Y % growth


%

Global NAND makers market share trend

NAND: Supply outlook

- NAND supply outlookWe estimate that the global NAND supply will grow 76% Y/Y in 2011, followed by another strong growth of 80% Y/Y in 2012,

while its capacity is expected to grow at 19% and 23% Y/Y in 2011 and 2012, respectively.

- NAND market share

Our bottom-up analysis indicates that SEC will emerge as the leader in the NAND market in 2011.Given the aggressive plans by SEC and Hynix for NAND, we expect both companies to witness market share gain from

Toshiba.

According to our estimates, Micron (including IM Flash) will maintain its market share of around 17% in 2011 and 2012.

40 36 35 38 38

37 39 40 36 36

12 19 17 17 17

0

20

40

60

80

100

2008 2009 2010 2011E 2012E

SEC Tos hiba + SanD is k H ynix (inc l STM) M ic ron (inc l. IM F lash)

66%

76% 80%

20% 19% 23%

0%

20%

40%

60%

80%

100%

2010 2011E 2012E

NAND bit-supply growth (Y/Y) NAND wafer-in capacity growth (Y/Y)

15


16/194

J.P. Morgan researchInitial roll-up of estimated global tablet build plans

Source: Company reports and J.P. Morgan estimates.

Units in 000s

Tablets: Apple, Android and oversupply?

- 2H11 tablet oversupply bubble?We continue to expect tablets to witness a very strong growth in the next two years, given the ongoing attempts by major

handset and PC brands to replicate the success enjoyed by Apples iPad and further commoditization of Android OS due to

Googles rotating alpha partner policy.

However, our latest industry checks point to a significant discrepancy between our 2011 tablet demand assumptions and tech

supply chains build-up plans, indicating what could be a 2H11 tablet oversupply bubble in the making.

Given the disappointing initial sell-through numbers for non-Apple tablets, the rise in performance benchmarks post iPad 2

launch, and higher-than-expected launch price points, we expect a potential downside to the extremely optimistic consensus

estimates of 65~80 million tablet shipments for 2011, which could dilute the Streets NAND and mDRAM demand

expectations, in our view.

2011E Acer Apple Inc. 38,000

Asustek Dell Inc. 2,500Hewlett-Packard 4,250HTC 1,000Lenovo 1,500Motorola 2,750RIM 3,500Samsung 7,500Toshiba 2,500Others 6,000

Total 81,000

16


17/194

NAND demand from Apples products as a % of total

%


Million units (1Gb equivalent), %

NAND demand from Apple

NAND and Apple

- Apple continue to influence NAND demand dynamicsApple has previously dominated much of the NAND demand dynamics, largely owing to its significantly successful launch of

flagship products such as iPod, iPhone and most recently iPad.

- What it means for NAND?

We continue to believe that Apple will continue to play an important role in the robust growth of the NAND market, accounting

for over 23% and 18% of global NAND demand in 2011 and 2012, respectively.

However, we turn cautious about the seasonality the NAND market might suffer, given that the sales pattern of Apples

products is heavily biased towards 2H.

Moreover, any sign of demand weakness for Apples products could have an incremental negative impact on the NAND

industry as a whole.

11% 9% 8%5% 3% 2%

3%3% 7% 12% 13%

11%

4% 6%

6%

0%

5%

10%

15%

20%

25%

2007 2008 2009 2010E 2011E 2012E

iPad iPhone iPod

178%

-23%

116%

6%

96%

27%39%

10%32%

-0.5

0

0.5

1

1.5

2

0

5,000

10,000

15,000

20,000

25,000

30,000

1H08 2H08 1H09 2H09 1H10 2H10 1H11E 2H11E 1H12E 2H12E

NA ND d ema nd from A pple H/H cha ng e (RHS)

17


18/194

Global NAND supply and demand trends

Supply/demand (%), ASP change (Q/Q)


NAND: Supply and Demand dynamics

- Global NAND supply and demand dynamicsSimilar to DRAM, NAND supply growth would also continue to outpace the demand growth throughout 1H11, according to our

estimates.

With the scheduled launch of new products in tablet PCs and smartphone segments in 2H11, along with the seasonal pick-up

in demand, we expect NAND supply to become tight in 2H11.

- NAND market to become tight if 2011 tablet shipments reach ~ 66-70 million units

However, we continue to forecast NAND demand assumptions as a moving target, given increasing discrepancy over

possible tablet shipments in 2011.

As mentioned earlier, if 2011 tablet PC shipments reach ~ 66-70 million units, compared to our current assumption of 51

million units, assuming an average density of 36GB, supply in the NAND market could become tight and vice-versa.

-34%

13%

1%

-31% -29%

-19%-27%

-35%

29%

7%

-13%

-4%

7%

-8% -8%-14% -15%

-10%-3% -2%

-20%-14% -12%

-4%

-40%

-30%

-20%

-10%0%

10%

20%

30%

40%

-10%

-5%

0%

5%

10%

15%

20%

1Q07 3Q07 1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

Supply/demand (exclud ing inv,) ASP change, Q/Q (RHS)

18


19/194

mDRAM shipments as % of total DRAM shipments, by company (2009-12E)

%

Source: Company data, J.P. Morgan estimates.

Mobile DRAM: Market overview

- What is Mobile DRAM?Mobile DRAM, often referred to as low-power DRAM (LPDRAM), has emerged as a preferred DRAM alternative for handheld

devices, versus conventional PC DRAM, largely due to its low energy footprint compared to traditional PC DRAM.

At its core, mDRAM enjoys a relatively low-voltage rating (1.2V-1.8V for mDRAM versus 1.5V-2.5V for PC DRAM1), coupledwith advanced power-saving techniques, which enable embedded systems to enjoy longer battery life without compromisingon the processing power.

- How does mDRAM consume less power?

Most of the arguments in favor of application of mDRAM revolve around its attractive power usage profile, which is believedto stem from three advanced powersaving techniques:

Temperature compensated self refresh (TCSR): adjustment of internal self refresh rates using a temperature sensor inaccordance with the previously specified ambient temperatures.

Partial array self refresh (PASR): This technique aids in bypassing the refresh cycle of empty portions of DRAM chip.

Deep power down (DPD): helps the mDRAM to enter sleep mode when the chip is not in active use.

19


20/194

- Revenue outlook

We expect the global mDRAM market to experience robust revenue growth in the coming years, largely owing to the strongdemand from smartphones and tablets, along with the increasing number of suppliers.According to our estimates, mDRAM is expected to contribute 24% of the total DRAM market revenue in 2011, up from 15%in 2010, largely owing to strong volume growth, expectations of strong ASPs in 1H11, and shrinking commodity DRAMmarket size

However, mDRAM ASPs are expected to witness meaningful erosion in 2H11, largely due to the aggressive capacityconversion to mDRAM across the board (appreciable mDRAM capacity expected to come online in 2H11.

In terms of quarterly trend, we expect the mDRAM revenue growth momentum to turn weak and disappoint in 3Q11 and4Q11, before returning to traditional seasonality in 2012.

Mobile DRAM: Market outlook

mDRAM revenue trend by quarter


US$ in millions, % (Y/Y)

mDRAM demand as a % of total DRAM shipments


US$ in millions, %

20


21/194

mDRAM demand by application

%


- Smartphones and tablets should be the main drivers for mDRAM demand

We expect smartphones and tablets to be the main drivers for mDRAM demand in the coming years, commanding anestimated combined share of 67% and 77% of the global mDRAM demand in 2011 and 2012, respectively.

The new smart devices should continue to witness sound memory content growth in the coming years.

Mobile DRAM: Demand outlook

Mobile DRAM content in currently available smartphones and tablets

Source: gsmarena.com.

21


22/194

- mDRAM demand from smartphones and tablets

According to our forecasts, average mDRAM content in smartphones is expected to grow 42% Y/Y in 2011 (544MB),followed by another 37% Y/Y growth in 2012 (742 MB).

Tablets are also expected to show a handsome increase in DRAM content in 2011 (+98% Y/Y, 632MB) largely due to theimpressive line-up of tablet launches scheduled in early 2Q11.

These tablets aim to leverage the ever-increasing base of data-intensive applications and deliver a faster processing

experience to consumers

Mobile DRAM: Smartphones and tablets

mDRAM demand from smartphones


Units in millions (1 Gb equivalents), %

mDRAM demand from tablets


Units in millions (1Gb equivalents), %

22


23/194

- Downward pressure on mDRAM ASP

We expect mDRAM prices to face downward pressure going forward, declining 33% Y/Y in 2011 and 36% Y/Y in 2012,mainly driven by the aggressive capacity expansion across the board, coupled with active steps by set makers to standardizethe product to decrease dependence on selective players and initiate price competitiveness among memory makers.

On a quarter-over-quarter basis, we expect mDRAM ASPs to continue to decline throughout 2011, with the magnitude of thedecline to be much pronounced in 2H11.

- ASP premium over PC DRAMThe recent appreciable incline in mDRAM price premium over traditional PC DRAM has attracted memory makers toward

mDRAM product, given the potential strong demand in the coming quarters and higher margins over commodity DRAM.

We forecast that mDRAM will enjoy attractive price premiums in 1H11; however, we believe it will narrow throughout 2011,given the accelerated mDRAM price erosion throughout 2011. Of note, the ASP premium for mDRAM is expected to bearound 111% in 1Q11 and narrow to 26% in 4Q11.

Mobile DRAM: mDRAM ASP trend

mDRAM ASP trend, Q/Q %


US$ per 1Gb equiv., Q/Q

mDRAM ASP premium over blended DRAM ASP


US$ per 1Gb equivalent, %

23


24/194

- BOM analysis

Based on our BOM analysis, DRAM dollar cost for smartphones and tablet PCs for 2011 is expected to be around US$10.1(6% of smartphones BOM) and US$11.7 (5% of tablets BOM), respectively.

- High touch panel : Major deterrent in achieving attractive price points

We estimate touch panel cost will represent over 30% of tablet BOM and 10% of smartphones BOM, respectively, of whichITO film/glass sensors have the largest cost burden, according to our estimates.

Our recent checks indicate touch panel makers efforts towards reduction in BOM costs through either developing newtechnologies such as touch-on-lens (TPK and Wintek) or direct pattern window (Melfas) (which may eliminate the use of ITOfilm, or they may go for in-house production of ITO sensor).

We see these efforts as fairly positive, as appealing price points for these devices would drive their volume growth, thusbenefiting DRAM and NAND.

Mobile DRAM: Budget analysis

mDRAM dollar cost for smartphones and tablet PCs


US$

mDRAM cost as a percentage of total BOM


%

0

3

6

9

12

15

1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E

Smartphones Tablet PCs

0%

2%

4%

6%

8%

10 %

1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q 12E

Sm artphones T ablets

24


25/194

Mobile DRAM: Supply outlook

mDRAM supply trend by quarter


Units in millions (1 Gb equivalents), Q/QmDRAM supply as a % of total DRAM supply


US$ in millions (1 Gb equivalents), %

- Mobile DRAM supply outlook

We estimate that the global mDRAM supply will grow 92% Y/Y in 2011, followed by 75% Y/Y in 2012 and would compriseover 17% of total DRAM supply in 2012 from the mere 8% levels back in 2009.

- Standardization via PoP - beneficial for late entrants

PoP (package on package) has been attracting much traction of late due to the ever-increasing demand for high-densitymDRAM, which is best mounted on the top of the application processor, given space constraints in smartphones and tablets.

We see this increasing trend as the first visible sign of standardization for the mDRAM industry. We expect the reducedcustomization of product to benefit late entrants and set makers alike, given the lower barriers of entry for memory makers,and more bargaining power at the hands of set makers.

However, increasing standardization would negatively affect mDRAM margins in the future, in our view.

25


26/194

Key trends in display

26


27/194

TFT-LCD: Supply/demand outlook

Global TFT-LCD supply/demand dynamics


% (Supply/demand 1)

- Supply-demand dynamics under utilization rate adjustments We recently changed our major assumptions for 2010/2011 and introduced our 2012 assumptions after analyzing the supply

and demand dynamics.

Based on our revised supply and demand dynamics, we estimate the magnitude of oversupply to ease from 3Q10 onwardsgiven UT cuts by major panel makers.

Given flexible production schedules, however, the magnitude of supply should not be as severe as in past downturn cycles.

With lower retail prices, TV brands and OEMs should be able to clear inventories.

- Slight upward revision in demand assumptions

Largely driven by upward revision to our LCD TV shipments forecasts and size-mix change (moving to bigger-size screens),we revised upward our demand assumptions by 3% and 1% for 2010 and 2011, respectively.

On the other hand, we cut our monitor and NBPC shipment forecasts.

6%

-2% -3%-5% -4%

1%

8%

-4%-1% 0%

-4% -3% -3%

3%-1% -1%

4% 3%0% -1% -1%

3%-1%-1%

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

1Q07 3Q07 1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11E 3Q11E 1Q12E 3Q12E

Supply/demand analysis - Inv. Adj. Supply/demand (w/ revised util. rate)

27


28/194

TFT-LCD: demand trend by application (1)

- LCD TV will keep increasing its proportion in the LCD industryWe continue to expect that TVs will keep driving up demand, resulting in a higher proportion in the industry in terms of both

area and revenue.

We project that LCD TV shipments will increase to 246 million units in 2011 from 163 million units in 2009.

In 2011, LCD TV proportion will reach 67% and 68% in area and revenue, respectively, as per our estimates.

TFT-LCD demand by application

Source: J.P. Morgan estimates, DisplaySearch.

1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q

Total Demand (M2) 25,349 26,660 27,601 26,413 28,109 30,655 35,413 35,259 35,962 37,475 40,679 41,366 83,523 106,023 129,436 155,482

NBPC unit 49.4 47.5 51.9 54.6 52.2 52.4 61.8 65.4 61.4 61.6 70.5 72.8 169 203 232 266

Square meter 3,280 3,160 3,437 3,605 3,443 3,454 4,077 4,307 4,060 4,076 4,658 4,808 11,232 13,483 15,281 17,602

Tablet PC 38 105 153 203 254 286 413 603 524 564 648 730 499 1,556 2,466

DT Monitor 35.7 35.8 37.1 38.9 36.8 37.0 39.7 40.6 37.7 36.9 38.7 38.3 138.3 147.4 154.1 151.7

Diffusion rates, % 94% 94% 95% 95% 95% 95% 96% 96% 97% 97% 98% 98% 93% 94% 95% 97%

Monitor Standalone 12.7 11.1 9.5 10.1 11.4 13.6 13.1 12.1 15.8 18.4 16.4 14.4 39.7 43.4 50.2 65.0

Total LCD monitor 46.3 44.7 44.5 47.0 46.4 48.7 51.0 51.1 52.3 54.2 54.1 52.0 168 183 197 203

Square meter 5,128 5,002 5,021 5,323 5,263 5,530 5,847 5,900 5,994 6,193 6,230 6,039 18,544 20,474 22,541 24,456

LCD TV 50.9 54.1 56.5 49.9 53.5 58.6 68.3 65.9 66.1 67.3 72.3 73.0 163 211 246 279

Square meter 16,079 17,660 18,146 16,377 18,295 20,462 24,133 23,414 24,403 25,569 28,040 28,608 50,813 68,262 86,303 106,620

Others (Square meter) 824 733 843 904 854 923 943 1,034 982 1,072 1,103 1,180 2,934 3,305 3,755 4,337

Based on area

1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4QNBPC 13 12 12 14 12 11 12 12 11 11 11 12 13 13 12 11

Tablet PC 0 0 1 1 1 1 1 2 1 2 2 2 0 1 2

LCD Monitor 20 19 18 20 19 18 17 17 17 17 15 15 22 19 17 16

LCD TV 63 66 66 62 65 67 68 66 68 68 69 69 61 64 67 69

Based on revenue

NBPC 15 14 13 15 13 12 13 13 12 12 12 12 16 14 13 12

Tablet PC 0 0 1 1 1 1 2 3 2 2 2 3 0 1 2 2

LCD Monitor 20 19 17 19 17 17 16 16 15 15 14 13 21 19 16 14

LCD TV 64 67 69 64 68 69 70 68 70 71 71 71 63 66 68 71

2012E

2012E

2012E

2012E2009

20092010E

2010E

2011E

2011E

2010

2010

2011E

2011E

28


29/194

TFT-LCD: demand trend by application (2)

12% 20%33%

46% 49% 53% 63% 66% 68% 71%

0%

20%

40%

60%

80%

100%

2003

2004

2005

2006

2007

2008

2009

2010

2011E

2012E

NBPC LCD monitor LCD TV

- Strong LCD TV demand

We believe demand for LCD TVs will continue to rise but demand for NBPC will show a downward trend.

We expect the LCD TV proportion to reach 71% by FY12 from 53% in FY08.

- LCD TV blended ASP trend

Thanks to product mix changes, along with new features and functions, we estimate blended TV panel prices will be lessvolatile than in the past (when all product was plain-vanilla LCD TVs) and we expect relatively flat ASPs going forward.

Also, falling TV panel prices should stimulate end-demand since TV set makers will be able to lower retail prices.

TFT-LCD demand by application

Source: Company data, DisplaySearch, J.P. Morgan

%

LCD TV blended ASP


%

-20

-15

-10

-5

0

5

1015

1Q05 4Q05 3Q06 2Q07 1Q08 4Q08 3Q09 2Q10 1Q11E 4Q11E 3Q12E

-50

-40

-30

-20

-10

0

1020

Q/ Q change (LHS) Y/ Y change (RHS)

29


30/194

TFT-LCD: supply side assumptions

Global LCD supply growth (by area and units)


%, Y/Y


TFT-LCD supply side assumptions

Square meter ('000s), %

- We expect global supply growth to ease in FY10 and FY11We believe that the aggressive capex spending by major panel makers in 1H10 will translate into new capacity coming online in

future quarters. Taking this into consideration, we revised up supply assumptions for 2010 and 2011 by 8% and 6%, respectively.

Given UT adjustments by major panel makers, we are seeing some discrepancy between actual supply growth and input capacitygrowth.

1311 11

15

3026

18

12

7 7 5

0

20

40

60

80

100

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010E 2011E

0

5

10

15

20

25

30

35

Unit inc rease % (LHS) Area increase % (LHS) Area - Unit (RHS)

1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q

Total supply 24,845 26,868 29,114 30,299 31,088 32,287 34,439 35,990 36,794 38,135 39,531 40,495 89,409 111,126 133,804 154,954

Q/Q growth, % 0 8 8 4 3 4 7 5 2 4 4 2

Y/Y growth, % 25 25 25 22 25 20 18 19 18 18 15 13 24 24 20 16

Korea 50 49 50 50 50 51 51 51 - 33 29 31 48 50 51 50

Taiwan 37 37 36 35 35 34 33 33 - 3 16 30 39 36 34 33

China 3 3 3 3 3 4 4 5 - 3 6 6 4 3 4 7

Japan 10 10 11 12 12 12 11 11 - 62 49 34 9 11 11 11

2012E20092012E2011E2010E

2010E 2011E

30


31/194

-28%

-30% -29%

-32%-33%

-32%

-31%

-30%

-29%

-28%

-27%

-26%

-25%

15.6" NBPC 20" Monitor TV 32" TV 42"

TFT-LCD: strong blended TV panel prices

- Panel prices have shown a meaningful correction from recent peakPrice weakness has been seen in TFT-LCD panel price trends on concerns of above-average inventory build-up in the

channel and at set makers due to weaker-than-expected World Cup retail sales and back-to-school demand.

We view the recent panel price decline as a positive development for set makers/OEMs, and expect consumer demand toreturn on price elasticity, as reducing component prices would help set makers/OEMs lower their BOM and cut prices to

stimulate end-demand.

- Blended TV prices remain resilientIn light of the price premium of LED panels over LCD panels, selling a higher proportion of LED TV would improve the major

panel makers' blended ASP.

The LED TV penetration rate accelerated in 2H10 and this trend is likely to continue into 2011.

Hence, the magnitude of decline in blended ASP for major panel makers would be much lower than that for LCD TV in anapples-to apples comparison.

Source: DisplaySearch, J.P. Morgan. Prices as of March 2011.

%

Panel price changes

Source: DisplaySearch, J.P. Morgan.

US$, %

LED vs. LCD TV panel price trends

45%

38%40%

43%40%

43%

46%49% 48%

41%

35% 36%38% 38% 38%

30%

35%

40%

45%

50%

55%

0

100

200

300

400

500

600

Jan-10 Mar-10 May-10 Jul-10 Sep-10 Nov-10 Jan-11 Mar-11

40 - 42" Full HD 4 0 - 4 2" 1 20 Hz E dge-LED LED TV pan el' s price premium, % (RHS)

31


32/194

-80%

-60%

-40%

-20%

0%

20%

40%

Jan-0

9

Mar-09

May-0

9

Jul-09

Sep-0

9

Nov-0

9

Jan-1

0

Mar-10

May-1

0

Jul-10

Sep-1

0

Nov-1

0

Jan-1

1

Mar-11

NBPC 15.4"W MNT 19"W TV 32"

147

5343

145

52 46

0

20

40

60

80100

120

140

160

32" TV 19" Monitor 15.6" LED NBPC

Panel price (US$) Cash cost (US$)

Panel price vs. cash cost


Panel price Y/Y change

Moderate price decline from September

The latest trends show that the LCD TV panel prices are expected to stabilize in 1Q11, fueled by a pickup in LCD TV demand.

While there is some slowdown in demand in North America due to inventory in retailers, the leading brands have started to carryout their promotion campaigns to boost the demand for large-screen TVs.

The TV panel prices are close to cash-cost level and the increasing portion of LED TVs will help blended ASP hold up.

If the panel price approaches near cash costs, we expect the panel makers to reduce UT further, which will eventually result inpanel price stabilization.

In 4Q10, the production cuts initiated by panel makers helped alleviate downward pressure on panel price.

TFT-LCD: signals of panel price stabilization

Source: DisplaySearch, J.P. Morgan.

US$ %

32


33/194

Global 3D TV shipments trend

Positive outlook for global 3D TV market We estimate 3D TV shipments to reach more than 80 million units, rising at a CAGR of 80%+ from 4.2 million units

in 2010.

The majority of 3D TV sales in 2010 will happen in the mature TV regions including the US, Japan, and WesternEurope, where sizable markets exist for upgrading or replacing order, non-3D sets.

The critical issues for mass consumer acceptance include standardization, content availability and interoperability

of the 3D glasses or eyewear.

Given that all supply chains are participating in 3D market, we expect 3D TV to account for almost one-third oftotal LCD TV in the next couple of years.

3D TV: 3D TV market outlook (1)

Source: iSupply, J.P. Morgan estimates.

Unit in millions, %

33


34/194

Major panel makers 3D TV portion as % of total production in FY11E

Aggressive 3D TV target by major panel makers Recently, LCD TV panel manufacturers have introduced advanced features including 3D technology along with

ultra-slim form factors and direct-lit LED backlights.

We believe these enhancements in TV panels will stimulate replacement demand and create a market segment bysatisfying consumer preferences.

The global LCD TV panel makers started to offer their 3D panels with aggressive shipment target for 2011.

According to our analysis, the major panel suppliers' 3D TV panel portion as a percentage of total production isexpected to reach as high as 35%. For CMI, the company is planning to have all the new LED TV modelsequipped with 3D feature in order to distinguish itself from the competitors. AUO, similarly, targets to increase its3D TV panel portion up to 10% in 2011.

We are estimating Korean panel makers 3D TV portion to remain in a range of 15%-20% in 2011. Sharp is aimingto gain shares in 3D TV panel industry with 3D TV panel portion target of 10-20%.

3D TV: 3D TV market outlook (2)


%

Panel brands 3D TV panel portion as % of total production

SEC 15%

LGD 20%

AUO 10%

CMI 35%

Sharp 10-20%

34


35/194

AM-OLED Sector

35


36/194

- Opportunities and challenges We expect the equipment and OLED material market to witness strong growth thanks to:

Much higher CAPEX requirement than for LCD investment & on-going capacity expansion by AM-OLED manufacturers

Korea-centric nature of AM-OLED will benefit Korean equipment/material makers.

- Self-created competition

AM-OLED is cannibalizing existing LCD businesses; thus AM-OLED TV in near-term is an expensive approach

Contrary to market expectation, Samsung and LG will avoid unnecessary cannibalization of LED TV.

- Winners and losers

Samsung SDI as a key beneficiary of SMDs AM-OLED business despite concerns over the dilution of SDIs stake in SMD.

Cautious on Cheil Industries as we think the current valuation fully reflects positive expectations.

We like AM-OLED equipment and material makers as AM-OLED growth offers new growth drivers

OLED stock performance (relative to KOSPI)

Source: Bloomberg, J.P. Morgan.

OLED stock relative performance

AM-OLED: Investment summary

Source: Bloomberg, J.P. Morgan. *Note: OLED stocks aggregate market capitalization.

1 March, 2010 = 100%, Won in billion 1 March, 2010 = 100%

36


37/194

- Hyper growth outlook We estimate a significant 122% CAGR from 2010 to 2013, driven largely by growing penetration in smartphones & tablet PCs

AM-OLED market to reach US$4 billion in 2011 (up 200% Y/Y) and US$14 billion (up 48% Y/Y) by 2013

Samsung Mobile Display (SMD), LG Display and AUO are expected to be key players.

Smartphone and tablet PC application to account for 90% of the total AM-OLED demand (area base).

Global AM-OLED market forecast

AM-OLED: AM-OLED market outlook


1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E 2011E 2012E

Total supply (US$ mn) 556 777 1,123 1,541 1,967 2,364 2,827 3,169 3,997 10,328

Q/Q change (%) 26% 40% 45% 37% 28% 20% 20% 12%

Y/Y change (%) 142% 165% 212% 249% 254% 204% 152% 106% 202% 158%

Total supply (mn units, 4" equivalent) 23 33 50 71 95 120 149 172 178 536

Q/Q change (%) 29% 44% 51% 41% 34% 26% 24% 16%

Y/Y change (%) 166% 194% 249% 295% 311% 259% 195% 141% 239% 201%

Total demand (US$ mn) 536 785 1,136 1,598 1,854 2,214 2,597 3,131 4,055 9,795

Q/Q change (%) 16% 46% 45% 41% 16% 19% 17% 21%

Y/Y change (%) 140% 178% 215% 247% 246% 182% 129% 96% 205% 142%

Total demand (mn units, 4" equivalent) 22 34 51 74 90 112 137 170 181 509

Q/Q change (%) 19% 51% 51% 45% 22% 25% 22% 24%

Y/Y change (%) 163% 209% 253% 292% 302% 233% 168% 130% 243% 181%

Market Share (%) 1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E 2011E 2012E

SMD 93% 93% 83% 80% 80% 78% 79% 78% 85% 79%

LGD 7% 7% 7% 8% 8% 10% 9% 9% 8% 9%

AUO 0% 0% 0% 3% 4% 5% 5% 6% 1% 5%

China/Japan/Taiwan 0% 0% 10% 9% 8% 7% 6% 7% 6% 7%

Total 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

37


38/194

- Supply-constrained growth Supply-driven market dynamics for now. Supply and demand to remain balanced.

Due to a limited number of suppliers and high-entry barriers, we foresee demand growth being constrained by supply growth.

Quarterly OLED supply forecast (revenue) Quarterly OLED supply forecast (area)


US$ in millions, % 000 square meters, %

Quarterly OLED demand forecast (revenue)


Quarterly OLED demand forecast (unit)


US$ in millions, % Units in millions, 4" equivalent, %

38


39/194

- Supply: SMD to drive supply growth and LGD to follow We forecast sharp capacity growth, up 268% Y/Y in 2011, coming mainly from the ramp up of 5.5G capacity by SMD.

LGD and AUO to add meaningful capacity, mainly 4G, in 2012. We estimate 173% Y/Y growth in 2012.

AM-OLED capacity analysis

AM-OLED: Supply side analysis


000' square meters / month

Korea Generation Glass size 1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E 1Q13E 2Q13ESMD

A1 4G 730X920 36 39 40 40 40 40 40 40 40 40

Q/Q change (%) 20% 7% 3% 0% 0% 0% 0% 0% 0% 0%

Y/Y change (%) 135% 93% 67% 33% 11% 3% 0% 0% 0% 0%

A2 5.5G 1300X1500 0 20 39 59 88 117 156 185 195 195

Q/Q change (%) na na 100% 50% 50% 33% 33% 19% 5% 0%

Y/Y change (%) na na na na na 500% 300% 217% 122% 67%

A3 8G 2200X2500 0 0 0 0 0 0 0 0 0 28

Q/Q change (%) na na na na na na na na na na

Y/Y change (%) na na na na na na na na na naLGD

AP2-E2 4G 730X920 3 4 6 10 13 20 24 27 34 37

Q/Q change (%) na 50% 50% 67% 33% 50% 17% 14% 25% 10%

Y/Y change (%) na na na na 400% 400% 289% 167% 150% 83%

P9-OLED 8G 2200X2500 0 0 0 0 0 0 0 0 0 0

Q/Q change (%) na na na na na na na na na na

Y/Y change (%) na na na na na na na na na na

Taiwan Generation Glass size 1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E 1Q13E 2Q13E

AUO

Hsinchu 4G 730X920 0 0 0 3 7 10 13 17 20 24

Q/Q change (%) na na na na 100% 50% 33% 25% 20% 17%

Y/Y change (%) na na na na na na na 400% 200% 133%

Others Generation Glass size 1Q11E 2Q11E 3Q11E 4Q11E 1Q12E 2Q12E 3Q12E 4Q12E 1Q13E 2Q13E

China/Japan/Taiwan 4G 730X920 5 7 9 11 13 15 17 20 23 28

Q/Q change (%) na 40% 29% 22% 18% 15% 13% 18% 15% 22%

Y/Y change (%) na na na na 160% 114% 89% 82% 77% 87%

0 0 0 0 0 0 0 0 0 0

Total capacity 44 69 94 123 161 203 250 289 312 351

Q/Q change (%) 45% 58% 36% 31% 31% 26% 24% 16% 8% 13%Y/Y change (%) 185% 245% 290% 308% 267% 191% 165% 135% 94% 73%

39


40/194

- OLED capacity share trend

We expect SMD to continue to represent the lions share of the AM-OLED market for the next 12-18 months, with near 80%share in 2012, followed by LG Display and AUO, which combined are expected to gain 15% of the AM-OLED market by the endof 2012.

- 4G/4.5G to remain mainstream capacity for now

On area base capacity, by glass size generation, we forecast 4G/4.5G capacity to remain mainstream capacity, accounting for

65% of total capacity in 2011 and the rest with 5.5G.

Looking ahead into 2012 and 2013, our AM-OLED model suggests further share gain by SMDs 5.5G, despite the addition of 4Qcapacity and the small presence of 8G capacity expected towards the end of 2013.

OLED capacity share in FY11E


OLED capacity share trend by generation



%, based on total shipment (area base) %

40


41/194

- Sizeable investment to continue in 2012

Expect another round of meaningful OLED investment cycle in 2012 as SMD and other lead players are likely to compete togain and/or maintain market share in the new technology space.

The size of investment could vary depending on technology advancement in key areas such as crystallization and evaporation.

In the near-term, we expect AM-OLED CAPEX to remain higher than that of LCD until the technology matures further andadvances to find a cheaper solution than current AM-OLED manufacturing technologies

- 8G investment an expensive pre-requisite for AM-OLED TV growth

AM-OLED TV is unlikely to have a meaningful impact on AM-OLED demand until the end of 2013

5.5G glass size is not optimal for large size (42" and above) TVs

Commencement of 8G mass production is scheduled for 3Q13

8G capex for AM-OLED fab (60K/ month glass input capacity) could be as high as US$7.5 billion if we assume similar capex

increase as SMDs 5.5G AM-OLED compared with that of existing 6G TFT-LCD fab, based on our capex analysis.

OLED capex trend


SMDs capex for 5.5G for 2011E

AM-OLED: OLED capex analysis


US$ in millions Won in trillion

41


42/194


43/194

- OLED panel price premium over LCD

With the increasing influence of consumer electronics in the display space, price elasticity of demand for AM-OLED is expected toincrease as the technology further replaces small-medium LCD displays.

We forecast price premium of AM-OLED panel over LCD panel to narrow further, given our expectation of:

increasing production scalability of AM-OLED makers; and

set makers efforts to differentiate their products

As of 4Q10, our analysis suggests 64% price premium for 4" AM-OLED panel (US$24) over the same size TFT-LCD panel (US$15).

We estimate the price premium to decline to 40% level and 4" AM-OLED panel to reach US$22 by the end of 2011.

OLED panel price premium over LCD

AM-OLED: AM-OLED price outlook


US$, %

OLED panel price premium over LCD


US$, %

43


44/194

- Self-created competition

With their already-dominant M/S in TFT-LCD panel and TVs, AM-OLED puts both Samsung and LG in a unique competitivelandscape, in which one-sided focus on new technology could pose a threat to the existing technology and vice versa..

The cannibalization of existing LCD business and a negative impact on LCD food chain seems inevitable.

Although technology advancement in the crystallization process could help utilize existing TFT-LCD fabs, we expect agradual cannibalization in balancing the two technologies.

Hence, we believe the rate of AM-OLED penetration will be somewhat controlled, especially for TV application, to preventunnecessary cannibalization of the LED TV market.

While the demonstration of technology leadership in AM-OLED could provide high value add for Samsung and LGs brandequity as consumer electronics companies, aggressive investment in AM-OLED TV is an expensive approach for now.

TFT LCD panel makers market share trend


AMOLED makers market share trend

AM-OLED: Unique competitive landscape


% %

44


45/194

- LED TV is another challenge

We agree that AM-OLED technology will eventually replace current TFT-LCD technology for TV application and will become amainstream TV technology in the long-term.

We believe (1) improving technology; and (2) better affordability of LED TV will be key challenges for AM-OLED TV market growth.

For example, LED TV performance has already reached near an AM-OLED equivalent for picture quality in terms of color gamut,response time and resolution.

We expect the price gap between the two technologies will narrow further and become more affordable for average consumers.

42" LED TV price premium over LCD TV


42" LED TV panel price premium over LCD TV panel

AM-OLED: Unique competitive landscape (1)


US$, % %

45


46/194

- LED technology share and efficiency

In addition, edge-lit type LED TVs are becoming more competitive vs. direct-lit full LED TVs.

Our discussions with leading LED packagers suggest that the latest LED BLU technology is moving towards enablingpartial-local dimming for edge-lit TV which was believed to be possible in more expensive direct-lit full LED TVs.

We expect such a technology advancement will help LED TV to remain attractive for consumers compared to AM-OLEDtechnology.

Separately, improving LED chip efficiency (i.e. lumen/watt) has been helping LED TV makers to use less LED chips perTV (i.e. moving from 4 bar to 2 bar technology in edge-lit LED TV backlights), which we believe will make LED TV moreaffordable for average consumers.

LED TV penetration has reached above 20% level as of end-10 and is expected to reach above 50% by the end-11.

Edge-lit BLU technology share


Number of LED chips required per BLU



% Units, %

46

O ( )


47/194

- High CAPEX is an entry barrier

While we agree with the general concerns about low ROIC and longer return period for AMOLED investment than TFT-LCD, we believe high capex will be a major entry barrier, along with AM-OLED technology, for new entrants.

Our capex analysis suggests 5.5G (with 100K/month glass input capacity) AMOLED capex requirement is approximately80% higher than that of 6G TFT-LCD fab (1500x1850) with similar design capacity and the capex for 8G (2200x1250 with100K/month glass input capacity) could be as high as W8 trillion.

AM-OLED vs. LCD capex analysis


Source: Company data, J.P. Morgan estimates. *Note: 100,000 substrates per month fab.

US$ in millions, %

LCD capex trend

Source: Company data, J.P. Morgan estimates. *Note: 60,000 substrates per month fab, 5-mask process. Equipment includes array, cell, color filter and module tools.

US$ in millions

47

AM OLED U i titi l d (4)


48/194

- SMD capex breakdown

Breaking down the capex for 6G TFT-LCD fab, approximately 65% of the capex was spent on equipment and the rest wasspent mainly on facilities (including land).

Based on similar assumptions, we estimate approximately W2.9 trillion to be spent on equipment and the rest on facilities(including land) for SMDs 5.5G fab building.


SMD capex for 2011


SMDs capex for 5.5G (2011E)


Won in trillion

48

AM OLED C t it (1)


49/194

- Equipment opportunity much greater than OLED material

Our analysis of LCD historical capex suggests approximately 60% of capex allocation for equipment purchase.

As we view LCD and AM-OLED technology as similar in manufacturing nature, we apply similar capex allocation estimatefor AMOLED investment to gauge potential market opportunities for OLED equipment makers.

Based on our estimates, we expect the capex for AM-OLED to reach US$5.7 billion (up 390% Y/Y) in 2011.

Even with a more conservative percentage allocation assumption of 50% of total capex, the potential market opportunity is

likely to exceed US$2.8 billion or approximately 10x higher than that of the OLED material market in 2011.

We believe: (1) much higher capex requirement than LCD investment; and (2) ongoing capacity expansion by Koreanmanufacturers will benefit Korean equipment makers.

Our recent discussions with companies also indicate noticeable efforts by AM-OLED panel makers to localize keyequipment manufacturing.

AM-OLED: Capex opportunity (1)

OLED capex trend


SMDs capex for 5.5G (2011E)


Won in trillionUS$ in millions

49

AM OLED: Capex opportunity (2)


50/194

- Lessons from the 2009-2010 LED TV cycle

From LED TV cycle during 09-10, we identify interesting earnings trends in different parts of the supply chain:

Margin expansion for the companies including panel, LED chip/packaging and BLU makers, was short-lived;

Only sapphire ingot makers and equipment makers witnessed meaningful margin expansion; and

Branded TV makers with leadership in LED technology benefited the most in terms of brand recognition

AM-OLED: Capex opportunity (2)

Major TFT-LCD makers OPM trend SEMCOs S-LED OPM trend

%%

Major LED BLU makers OPM trend

Source: Company data, J.P. Morgan.

Major LED materials makers OPM trend


%%

Source: Company data, J.P. Morgan.Source: Company data, J.P. Morgan.

50

AM OLED: OLED materials market outlook (1)


51/194

- High growth opportunity but higher expectation

We expect the OLED material market to witness strong growth in coming years, reaching US$269 million in 2011 (up240% Y/Y) and W630 billion in 2012 (up 134% Y/Y).

Due to increasing production localization in Korea, we expect local OLED material suppliers to benefit the most. However,we are concerned about the overly optimistic expectation about the OLED material market.

Along with higher penetration, we expect AM-OLED components and material market to witness further meaningfulgrowth over the next two-three years.

We forecast the marketkey components and materials market combinedto reach US$568 million this year, up 119%Y/Y and US$1.3 billion by 2012, up 124% Y/Y.

Separately, we estimate the OLED material market to reach close to US$269 million in 2011, up 240% Y/Y and US$630million in 2012E, up 134% Y/Y.

AM-OLED: OLED materials market outlook (1)

Key AM-OLED component/material market


OLED material revenue as % of component/material

Source: KDIA, DisplayBank, J.P. Morgan estimates.

US$ in millions, %US$ in millions

51


52/194

AM-OLED: What is an AM-OLED?


53/194

OLEDs are a special class of LEDs, manufactured by placing thin layers of organic compounds between two conductors.

They are primarily based on the principle of electro-phosphorescence, wherein certain organic molecules are observed toemit light on application of electric voltage across them.

- AM-OLED versus TFT-LCD

In general, OLED incorporates a simpler bi-layer structure, consisting of conductive and emissive layers, which allowsthinner form factor as well as lower manufacturing costs compared to TFT-LCD.

The basic difference between the OLED and TFT-LCD display technology lies in the usage of backlight illumination.

In contrast to conventional TFT-LCD display, an OLED display does not require any other source of backlight illumination,given the self-emissive nature of organic molecules.

In OLED, electric current travels from cathode to anode through layers, while the flow of lights from BLU is controlled byliquid crystals and polarizers in LCD.

AM OLED: What is an AM OLED?

A structural comparison of OLED and LCD

Source: Company data, J.P. Morgan. 53

AM-OLED: Cost analysis on OLED vs LCD


54/194

- Optimizing OLED technology with lower manufacturing cost

The different structure of OLEDs lends itself to the cost advantages over flat panel display made with LCD technology.

The OLED stack structure is simpler than the structure of TFT-LCD and that provides cost-saving potential in the long run.

The OLED manufacturing cost is approximately 30% lower than that of TFT-LCD, largely due to the removal of BLU andcolor filter.

The BLU and color filter represent one of the highest costs within TFT-LCD components.

OLEDs do not require a separate light source such as backlighting units (BLU) due to their self-luminous structure;therefore they provide room for further cost reductions.

AM OLED: Cost analysis on OLED vs. LCD

Cost analysis on OLED vs. LCD


%

54

AM-OLED: AM-OLED structure


55/194

- AM-OLED materials

Hole injection layer (HIL): The hole injection layer is often placed in between the anode and the hole transport layer,where it accepts positively charged holes and effectively passes them over to the hole transport layer.

Hole transport layer (HTL): The hole-transport layer enhances the ability of the anode to deliver positively charged holesto the emission layer, where the recombination process takes place with the electrons.

Emission material layer (EML): The emissive layer is where the electron-hole recombination takes place to generate thelight source.

Electron transport layer (ETL): In order to achieve a balanced injection and transport of electrons into the emissive layer,additional organic layer is introduced into the device structure, called electron transport layer. The electron transport layeroptimizes the transport of electrons from cathode to the emissive layer.

Electron injection layer (EIL): The electron injection layer enhances the electron injection efficiency, resulting in improveddevice performance.

Passivation: The passivation layer can be adopted to protect OLED from moisture and oxygen, which offers theenhancement of lifetime and stability.

AM OLED: AM OLED structure

OLED structure


Recombination process in OLED device


%

55

AM-OLED: Introduction to manufacturing process


56/194

Key differences are: (1) elimination of color filter and backlight, and (2) the addition of evaporation and encapsulation

processes for AM-OLED due to its nature of self-emission of lights.

- Evaporation

The OLED evaporation condenses thin film layers on a glass substrate through vacuum evaporation of the organic matter.

The evaporation process largely calls for layer-by-layer deposition of high quality luminous RGB layers onto OLED substrate,which lends its better transmittance quality and attractive form factor.

- EncapsulationThe encapsulation is indispensable in OLED manufacturing process as it protects organic compounds from humidity.

In order to achieve the realization of OLEDs with a sufficient lifetime, the sealing of OLED display is required with extremelylow permeation for oxygen and humidity.

The glass lid is attached on top of the display substrate with power inside to absorb moisture that diffuses through the surface.

g p

Manufacturing process of OLED and LCD

Source: Company data, J.P. Morgan. 56


57/194


58/194


59/194

Rechargeable Batteries

59


60/194

RB: Electric vehicle battery supply matrix


61/194

- Korean makers have the best customer mix

Japanese battery makers are leading in terms of number of auto makers they are supplying to, yet (with the exception ofSanyo) this is rather constrained to Japanese auto makers.

Korean battery makers SDI and LG Chem, on the other hand, have an alliance with US / European auto makers besidesHyundai, which are more aggressive in EV development.

Nissan and Mitsubishi are aggressive in EV while Toyota and Honda are more conservative sticking to HEV.

Given the much larger battery size of EV comparing to HEV, AESC and GS Yuasa, suppliers of Nissan and Mitsubishishould be able to ramp up volume faster.

Sanyo is also well positioned with customer exposure to US and European auto makers.

Battery supply matrix

Source: J.P. Morgan.61

RB: LiB TAM to quadruple xEV & ESS to drive secular growth


62/194

- LiB to drive steady total addressable market (TAM) growth

In the near future, consumer LiB should continue to drive steady total addressable market (TAM) growth, led by:

Robust unit growth from end-demand

Meaningful battery density growth along with rapidly rising multimedia content

New device types such as e-books, tablets.

We forecast LiB TAM to more than quadruple in the upcoming decade, from US$10 billion (09) to US$43 billion (20).

- Higher value proposition for xEV to lift profitability outlook

We believe xEV and ESS should be the key growth drivers, going from virtually zero to US$13B each in 2020,representing a 10-year CAGR of 34% and 48%, respectively.

LiB becomes a critical and high-ticket building block within xEV its role being significantly elevated as part of thepower chain system instead of serving just in the backend for consumer electronic products.

The tight integration and heavy customization also minimize competition, forming high entire barriers for new entrants,and should result in sustainable healthy profitability for a prolonged period.

US$ in billions


0

10

20

30

40

50

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

HEV/ EV Consumer Storage Indus trial & Others

LiB total addressable market

%


LiB CAGR for 2010-2020

34%

5%

48%

10%

0%

10%

20%

30%

40%

50%

60%

HEV/EV Consumer Storage Industrial &

Others

62

RB: xEV (EV / PHEV / HEV)


63/194

- Battery technology maturing fast

No fundamental technology road blocks for xEV-used battery; evident with recent commercial launches of several PHEV/EVmodels such as Mitsubishi Miev, BYD F3DM/E6, Nissan Leaf.

Remaining issues such as cell uniformity and low production yield can be resolved through production process improvement.

- System integration complexity often drags out launch date

xEVs involves extensive change in power chain and body design, which require high level of customization for system

integration which calls for battery makers and auto makers to form JVs for effective collaboration.

Lack of standardization in xEV battery is a headache for auto makers: Different xEV batteries from different vendors havedifferent product characteristics (capacity, charging performance, life cycle) thus requiring effort from auto makers to evaluatebatteries.

Potential xEV LiB TAM (conservative forecast)

US$ in billions

Source: Company data, J.P. Morgan estimates. Source: J.P. Morgan based on interview with companies.

Most expect volume ramp-up by 2013-15

Auto makers lined-up Expected launch date Meaningful volume

Sanyo VW, Honda, Ford, PSA 2011 2015

GS Yuasa Mitsubishi, Honda 2009 (EV) / 2010(HEV)

AESC Nissan, Renault 2010

Samsung SDI BMW 2012 2013

LG Chemical GM, Hyundai 2010-11

BYD Own brand, Daimler 2010 2013

BAK Chery, FAW 2010

63

RB: xEV (EV / PHEV / HEV)


64/194

- High battery price to come down sharply by 2013

LiB makers agree that high price of xEV battery are mainly due to high start-up cost with small volume and expect batterycost to come down sharply over time as volume increases.

Sanyo believes that it is possible to meet the Japanese governments mandate of reaching US$300/kWh price target by

2015, from ~$1000/kWh now.

- Chinese has early price advantage, yet that could diminish if China market fails to take off

We expect the price gap between Chinese and Japanese/Korean to narrow over the next few years, yet the lower costadvantage by BYD through vertical integration should persist.

Such cost advantage may allow BYD to eventually penetrate into global auto makers, if it's quality can be validated throughits own xEV in the next 2-3 years.

LiB price (US$/KWh)

%

Source: J.P. Morgan based on interview with companies. Source: J.P. MorganNote: Year 0 is 1993 for NiMH, 1995 for LiB

Historical price trend for Ni-MH and consumer LiB

Current 2012/13 2015

Sanyo 1,000 300

GS Yuasa 1,300 600-650

AESC 1,300 650 300LG Chemical 900-1,000

BYD 500 200-300

BAK 400 2000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10

Ni-MH

LiB

Year

price decline accelerated

at year 2-4

64

Rechargeable Battery


65/194

Sanyo, 20%

Sony, 12%

SamsungSDI, 20%

LG Chem,15%

BYD, 7%

Others, 27%

Lithium battery cost structure

%

Lithium battery raw material cost breakdown

Lithium polymer battery capacity share (FY10E)


Lithium battery shipment share (FY10E)


% %

%


65

Rechargeable Battery


66/194

Separator market share (FY10E)

%

Electrolyte market share (FY10E)

Anode market share (FY10E)


Cathode market share (FY10E)


% %

%


66


67/194

LED market outlook

Source: Bloomberg for NR companies, J.P. Morgan estimates. *Note: Share prices as of April 7, 2011

Price Market cap

(LC) (US$ B) FY11E FY12E FY11E FY12E FY11E FY12E

SSC 046890 KQ NR 43,500 2.3 17.6 12.4 3.6 2.8 22.8 24.8

LG Innotek 011070 KS NR 115,000 2.1 15.4 8.9 1.5 1.3 10.6 16.4

SEMCO 009150 KS N 118,000 8.3 19.4 17.4 2.3 2.0 12.2 11.6

Cree CREE US OW 46.44 4.9 30.8 21.6 2.5 2.1 na na

Epistar 2448 TT OW 109.0 3.2 13.1 12.2 1.9 1.8 14.6 14.7

Everlight 2393 TT N 84.5 1.2 13.4 12.6 2.1 2.0 16.0 16.1

ROE(%)Company Ticker Rating

P/E(x) P/B(x)

67

- Hyper earnings growth witnessed in 2010 but

LED: Market review


68/194

Hyper earnings growth witnessed in 2010 but

LED chip makers and packagers with LED TV exposure a saw sharp growth in both top / bottom lines during 2010 thanksto Samsungs push for LED TV

However, margin expansion for LED chip makers was short-lived due to over-capacity. LED chip makers addedexcessive capacity during 1H10 on TV makers optimistic guidance. Looking back, the LED TV cycle benefitted MOCVDand ingot makers.

Source: Companies, DisplaySearch, J.P. Morgan estimates.Source: KDIA, DisplayBank, J.P. Morgan estimates.

Units in millionsUS$ in millions

0%

50%

100%

150%

200%

250%

300%

2009 2010

Epis tar Everlight C ree S-LED LG Innotek SSC


SEMCOs S-LED OPM trend

LED makers revenue growth in FY10 LED TV shipment targets by TV makers

MOCVD and ingot makers margin trend

68

Cost reduction cost reduction and cost reduction

LED: Market outlook


69/194

- Cost reduction, cost reduction and cost reduction

Panel makers are attempting to lower LED BLU cost by reducing the number of LED chips per panel along with:

Reducing number of LED bars, and enhancing aperture ratios and increasing light transmission rate.

The primary source of LED BLU price reductions will mainly come from Edge-type development from 4 light bars to 2 lightbars or 1 light bar. Reducing 4 light bars to 2 light bars will reduce the number of LED chips used by nearly 30%.

Reducing 2 light bars to 1 light bar will reduce the number of LED chips used by another 20% and lower the LED BLU costsby 15-20%.

Such efforts leave LED makers with homework to reduce cost and improve efficiency of LED chips. As a result, LED makersare adding more MOCVD and migrating to larger size wafer; i.e., better scalability.

Progress in low cost LED BLU

Source: Companies, DisplaySearch, J.P. Morgan estimates.

69

L tili ti j id ff t f 1H10 it i

LED: Low utilization remains a key concern


70/194

30%

10%13%

10%

34%

0%

10%

20%

30%

40%

Epistar Everlight SEMCO LG Innotek SeoulSemi

- Low utilization; a major side effect from 1H10 capacity expansion spree..

We are concerned about LED makers with heavy BLU exposure as we expect their utilization to remain lower thannormal level during 2011E.

Our concerns on production utilization stem mainly from:

Ongoing MOCVD equipment addition in 2011E

Increasing production conversion to 6" wafers from 2" and 4" wafers

- Limited revenue contribution from general lighting

Korean LED makers are targeting higher general lighting market exposure this year at ~10% of total revenue.

Unless we are assuming sharp pick up in LED TV demand, 2011E could be another tough year for Korean LED makers

Source: Company data, J.P. Morgan estimate

MOCVD

Source: J.P. Morgan estimate

General lighting revenue contribution in 11E%

0

500

1000

1500

2000

2008 2009 2010E 2011E

Nichia TG SEMCO LG Innotek Seoul Opto

Epistar ForEpi Huga Others

Units

70

Margin pressure clouds 1H11 outlook

LED: Margin outlook


71/194

- Margin pressure clouds 1H11 outlook

Margin pressure on LED component makers may intensify in 1H11, mainly due to margin contraction from set/panel makers

From 4Q10 to end of 2Q11, we forecast LED BLU selling prices to decline by 15-20% for 32-inch and 42-inch LED TVs

LED component costs account for 40% of the LED BLU BOM costs, which represent 40% of the LED TV panel cash costs andare the biggest portion in TV panel costs

42 FHD LED TVs BLU bill of materials

Source: Display Search, J.P. Morgan

71


72/194

- LED lighting is in early adoption stage

LED: LED lighting industry


73/194

- LED lighting is in early adoption stage

LED lighting market is still in the early stage and with 5% penetration rate in overall lighting market in 2010.

We look for a 6% CAGR for the Lighting industry over the next 5 years, as a positive mix and adoption of LED drive the lightsource business

- LED lighting market to play major roles in the eventual revenue level in 2012

We believe LED lights will represent 5-10% of all bulbs replaced in 12 with an associated revenue opportunity of $5-$19bn.

We expect lighting users with high energy or maintenance costs to be first adopters.

Energy efficiency subsidies, the availability of regionally certified (such as Energy Star) LED lighting products, as well as thediversity of products available, will likely play major roles in the eventual revenue level in 2012.

Annual volume of light bulbs/LED modules

Source: J.P. Morgan estimate

Annual sales of light bulbs/LED models


US$ in millionsUnits in millions

73

- High entry barrier for LED lighting industry

LED: Early mover advantage


74/194

High entry barrier for LED lighting industry

We expect LED chip and module manufacturing to become scale driven, cyclical and in the longer term commoditizing business notsuitable to the late entrants in the market.

The Lighting industry overall is likely to become more competitive with increased pressures on returns, particularly post 2015

Near-term growth looks best at Siemens.

- Philips is best positioned in LED lighting space

Philips is the only vertically integrated traditional and LED Lighting company; it is best positioned to develop LED lighting solutions.

- Cost and performance main drivers for fast adoption

Current high end LED lamps can produce 120/130 lumen per watt, a better ratio than fluorescent lamps, the previously mostefficiency traditional lighting technology.

Cost per lumen has been falling at around 20% per annum, i.e. an LED costs 10x less per decade per one unit of light it emits.

Philips Lighting growth model to 2020 ( mn)

Source: Roland Haitz & Lumileds

Light output per LED accelerated from its long term trend


Logarithmic scalemn

74


75/194

Disruption analysis after the earthquake

75

Japan Earthquake: Semiconductor supply chain

- BT Resin: High tightness expected


76/194

According to our checks, Mitsubishi Gas Chemical (MGC) is the sole supplier for BT resin, and has a patent on the compound.We believe the impact of BT resin shortage would not only be limited to the substrate and OSAT side, but spread across the

entire semiconductor food chain with an indirect impact on foundries and all related handset/smartphone vendors.

Korean players such as SEMCO, LG Innotek and Daeduck have also expressed concern about potential shortages of BT resin.

- Silicon wafers: We expect medium tightness in supply

Shin-Etsu and SUMCO accounted for 32% and 29%, respectively, of global supply of 300mm wafer in 2010.Shin-Etsus Shirakawa has received mechanical damage from the earthquake while SUMCOs Yonezawa plant and MEMCs

Utsunomiya plant are also currently suspended.

In our view, the wafer shortage may not be an immediate issue since semiconductor makers typically have one to two monthsinventory as backup, plus they can source from other wafer makers outside Japa

Sector_Tech_201104_JPM_194pg

Documents