goldman sachs-转型中国：探索太阳能行业的理性化之路－展望，下一个十年

2012 年 9 月 11 日

中国太阳能行业展望：下一个十年

转型中国：探索太阳能行业的理性化之路 (摘要) 证券研究报告

全球 DRAM（内存）行业和中国 CRT(显像管)电视机行业对于太阳能行业的中期前景具有借鉴意

义我们认为，中国太阳能行业的投资周期需要被打破，

基本面整合才能够启动。我们在本文中分析了经历过

初期快速增长的科技行业（DRAM 和中国 CRT 电视机

行业），以了解行业赢家在理性化道路上的成功经

验，并以此为基础，描绘出太阳能上游行业理性化

（借鉴 DRAM 行业）和中游整合（参考电视机行业）

的潜在路径，并采用 GS SUSTAIN 框架来筛选潜在的

行业领军企业。

短期内，行业理性化是大势所趋，但需要由地方

政府和银行率先行动宏观环境可能会给中国太阳能行业带来的一项积极影

响是，长期持续的下行周期最终将促使行业趋于理性

化。中国太阳能投资周期能否被打破在很大程度上取

决于地方政府和银行率先行动的意愿。在政治前景更

为明朗之后（我们认为可预见性将在 10 月中旬的十八

大之后数月有所改善），我们才能够预测行业整合的

时间点。

在短期产能理性化之前，产品均价/盈利仍将承压

我们预计，在中小型企业开始退出供应链之前，多晶

硅产品均价将进一步下跌。我们预计 2012/13 年多晶

硅均价为每千克 22 美元/18 美元，并推出了每千克 15美元的 2014 年预测。我们还将 2012/13 年太阳能组件

均价预测进一步下调至每瓦 70 美分/57 美分，并推出

了每瓦 50 美分的 2014 年预测。

买入阳光电源；卖出英利绿色能源和京运通；相

对于产品制造商，仍相对看好国内逆变器生产商我们将研究范围内太阳能产品制造商和设备生产商的

12 个月目标价格下调了 24%-60%，以计入我们认为产

品均价将进一步下降和行业下行周期持续的预期。总

体而言，阳光电源仍是我们的首选买入股。我们恢复

覆盖晶澳太阳能，给予中性评级，并维持对保利协

鑫、天合光能和精功科技的中性评级。我们对英利绿

色能源和京运通的评级为卖出。我们对该行业的投资

观点面临的风险包括：全球需求高于预期、产品均价

降幅小于预期。

* 全文翻译随后提供

转型中国转型中国是我们在 2012 年新推出的系列主题报告。该

系列报告将关注那些我们认为在未来几年将因竞争格局

变化、新技术推出或全球环境改变等因素而出现重大变

革的行业。

研究范围概览

资料来源：Datastream、高华证券研究预测

宋赟波, CFA (研究助理) +86(21)2401-8925 [email protected] 北京高华证券有限责任公司胡玲玲 (分析师) +86(10)6627-3520 [email protected] 北京高华证券有限责任公司执业证书编号: S1420511100002

北京高华证券有限责任公司及其关联机构与其研究报告所分析的企业存在业务关系，并且继续寻求发展这些关系。因此，投资者应当考虑到本公司可能存在可能影响本报告客观性的利益冲突，不应视本报告为作出投资决策的唯一因素。有关分析师的申明和其他重要信息，见信息披露附录，或请与您的投资代表联系。

北京高华证券有限责任公司投资研究

Price Potential

7-Sep-12 up/downside

Solar product manufacturing sector

GCL-Poly 3800.HK HK$ Neutral 1.10 1.19 -8%

Trina TSL US$ Neutral 4.00 4.02 0%

JA Solar JASO US$ Neutral 1.00 0.80 25%

Yingli YGE US$ Sell 0.80 1.67 -52%

Solar equipment sector

Sungrow Power 300274.SZ Rmb Buy 11.00 9.12 21%

Jing Gong 002006.SZ Rmb Neutral 8.40 9.25 -9%

Jinyuntong 601908.SS Rmb Sell 2.40 6.77 -65%

RatingCompany Ticker Currency 12-m TP

2012 年 9 月 11 日中国：清洁能源: 太阳能

高华证券投资研究 2

Table of contents

Executive summary: Rationalization will come in two phases, but timing is hard to predict 3

Phase I: Rationalization is inevitable; reduced access to funding the likely catalyst 9

Phase II: Cost/scale leadership is essential for upstream; lessons from DRAM experience 13

GS SUSTAIN: Identifying potential winners in solar upstream 22

Phase II: R&D/financial strength essential for midstream; China’s CRT-TV transformation 28

GS SUSTAIN: Identifying potential winners in solar midstream 37

Equipment: earnings growth for inverters, extended down-cycle for furnaces 39

Changes to earnings metrics, estimates, target prices and valuations 40

JA Solar – Solid balance sheet to weather the turbulence; reinstate with Neutral 45

Appendix I: Goldman Sachs global solar demand model 48

Appendix II: Levelised cost of energy (LCOE) analysis by geographic coverage 49

Appendix III: Key assumptions for our LCOE model 50

Appendix IV: The ESG risk evaluation indicators of GS SUSTAIN 51

Appendix V: GCL-Poly key data 52

Disclosure Appendix 54

Prices in this report are as of close of September 07, 2012, unless other specified.

Gao Hua Securities acknowledges the role of Brian Lee, CFA and Kenneth Ho of Goldman Sachs in the preparation of this product.

The author would like to thank Shuai Liu for his contribution to this report.

Transforming China

Transforming China is a new series of thematic reports we are introducing in 2012. The series will focus on industries which we believe will undergo significant transformation over the next few years due to a change in the competitive landscape, the introduction of new technology or a change in the global environment etc.

Leveraging GS’ global industry expertise and resources, we hope to bring fresh perspectives or unique insights into the analysis of these trends or events. In addition, we will utilize our GS SUSTAIN industry framework to identify companies with superior and sustainable business models that we think can generate industry leading returns and become long term industry winners. Combined with our industry valuation analysis, we hope to provide actionable investment ideas that we believe will generate alpha in the timeframe presented.



Executive summary: Rationalization will come in two phases, but timing is hard to predict

Deterioration of ASP/earnings has been a theme for the solar industry since October 2010, over which time the solar names under GH coverage have corrected by 89% to date. We are increasingly concerned about the ability of Chinese solar names to survive amid

the challenges of the current environment:

Trade frictions between the US, China and EU;

Highly stressed earnings outlook and the ongoing extended down-cycle; and

Poor visibility on timing for rationalization and consolidation of overcapacity in China.

Ongoing solar trade friction between the US, China and EU could be a near-term swing factor for the fundamental competitiveness

of solar companies globally, with potential tariff barriers eventually driving the reallocation of global effective solar capacity. Tariffs

might also decrease the grid parity competitiveness of solar energy vs. other energy sources to a certain extent. Furthermore, since

fossil fuel prices have come down, grid parity has become a near-term moving target in many regions. So, while grid parity in major

electricity markets would be a milestone for the PV industry, it would not necessarily represent an inflection point for overall PV

growth. Therefore, we remain cautious on our demand forecasts over the next few years (Appendix I).

In this report we outline our vision for the medium-term rationalization of the solar industry. While calling the precise timing and path of the industry's necessary restructuring is difficult, we believe it will be catalyzed by reduced access to funding, triggering a period of dramatic consolidation and capacity reduction. We draw upon analysis of the restructuring of the DRAM and China CRT-TV markets to define an outlook for the solar industry and identify the characteristics of the likely long-term winners.

Phase I: Local governments & banks initiate the process to break the over-investment cycle If there is any upside in the currently poor macro environment, it will be from the ongoing extended solar industry down-cycle finally triggering rationalization. However, putting the brakes on what appears to be a perpetual solar investment cycle (Exhibit 2) hinges largely on the will of local governments and banks to address their accommodative policies for lending, taxes and land. The exact timing for this requires some political insight, but should have greater visibility post upcoming Communist Party congress in mid-October. As of now, debt profiles and credit metrics give us some indication of where risks might lie.

Phase II: Fundamental rationalization driven by divergence of technology, scale and cost leadership This is where we draw upon lessons from other technologies that saw early-stage secular growth and peaked out impressively before rationalizing. DRAM provides a useful comparison for solar upstream, while China’s CRT-TV sector (gone but not forgotten) offers insight for the solar midstream. Although we maintain our near-term bearish view on China solar, we believe Phase II will see the industry return to fundamentals, with clear industry leaders emerging from the uncertainty. Our DRAM/CRT-TV studies helped to identify the nature and likelihood of industry trends, which in turn revealed the characteristics of future industry leaders.



Exhibit 1: Rationalization is inevitable due to excessive overcapacity Global solar industry’s supply & demand vs. solar module cost

Exhibit 2: China solar’s cycle of overinvestment – monitoring the behavior of Chinese companies is key to understanding the near-term pricing fundamentals

Source: Goldman Sachs Research estimates, Gao Hua Securities Research estimates. Source: Gao Hua Securities Research.

Exhibit 3: Identifying potential leaders within the solar industry – lessons learned from DRAM and China CRT-TV

Source: Gao Hua Securities Research.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

4Q

12

E m

od

ule

co

st (

US

$ /

Wa

tt)

2012E module capacity (GW)

Rest of AsiaUS

EU

Other module

manufacturers in China

Tier-1 Chinese players



GS C-Si module

demand forecast

2013E

GS C-Si module

demand forecast

2014E

GS module ASP

forecast by 4Q2012

Companies invested in capacity in anticipation of much

larger demand for modules. As costs fell governments

lowered subsidies reducing demand, which was further

dampened by low availability of credit for commercial solar power generation. The shift in the industry's

supply curve to the right, concurrent with the demand

curve's move to the left has resulted in capacity almost

2x forecast demand.

Accommodative local policies for lending,

taxes and land support aggressive capacity

addition

Dumping modules in international

marketsAnti-dumping

investigations and duties

Dumping modules in domestic market

Market accessFinancial strength

Cost and scale leadership

•Technology developer

•Lowest cost per watt

•Scale production

•Integration

•Location

•Cost of capital

•CROCI returns

•Market share growth

•Balance sheet strength

•Sustainable investment

•Restructuring potential

•Valuation (P/E, P/FCF, DCF, EV/GCI)

•Global footprint

•Brand recognition

•Policy support

•New market entry

•Employee recruitment

•Distribution and installer

networks

Potential

winners in the

solar industry

Wafer

Polysilicon

Higher capex & fixed cost

Assertive investment, economic scale,

technology advantage & cost leadership

Forward integration helps to identify demand

visibility

Strong balance sheet is needed to weather

through the high cyclicality

The DRAM

industry

Low entry/exiting barriers

Brand /channel building is expensive

Migrating from vertical integration to

specializations

Strong balance sheet is needed to weather

through the cyclicality

Cell

Module

The CRT era

of the

Chinese TV

industry

Solar upstream

Solar midstream

Cross industry studying – lessons learned from other technology industries



The transition from secular to cyclical growth in any tech sector is usually driven by technology, economic scale and cost leadership. Near-term demand elasticity often leads to medium-term overcapacity and commoditization. We previously identified the following characteristics of a structural solar winner: 1) R&D, cost and scale leaderships; 2) broad market access; and 3) a strong balance sheet. Industry rationalization can take years and every sector is different, depending on factors like supply chain, competitive landscape and demand cyclicality. However, we found the solar upstream segment (poly to wafer) has similarities to the DRAM industry, while solar midstream (wafer to module) has similarities to China’s CRT-TV industry (Exhibit 3). We believe the consolidation paths of both industries offer valuable insight into how China’s solar industry will transform in the medium term.

Solar upstream: cost/scale leadership essential – a case study of DRAM’s four-decade history The four-decade evolution of the DRAM industry offers a useful analogy in exploring potential consolidation paths for the solar upstream segment. We believe these two industries share much common ground, including large exposure to commodity projects, higher capex & fixed costs, and a structure of downstream integration providing higher order visibility.

We consider solar upstream to be at a mid-to-late stage of consolidation, which appears to be progressing at an accelerated pace compared to the DRAM industry. Our conclusion is based on the following:

1) Demand elasticity is more effective throughout cycles in DRAM vs. solar, thus historically driving more consistent new capex in the DRAM industry. Solar demand is tied to energy demand and is driven by many factors (e.g., availability of grid/funding systems, fossil fuel prices, regulatory stability, etc.) rather than just falling ASPs. Therefore, the low visibility of PV demand might discourage new capex to flow into solar upstream (mainly due to its higher capex and fixed cost as % of total cost), as seen in the lack of new capacity plans announced recently by tier-1 poly makers.

2) Poly price continues to reach new lows, although we see limited downside risk remaining. Poly ASP is now close to marginal manufacturers’ variable cost (COGS minus D&A). However, we would need a clear sign of new capacity being cancelled before we finally call a floor for the poly price.

Near-term view:

Prior to Phase I consolidation and the exit of marginal players from the supply chain, we expect poly ASP to decline

steadily over 2012-2014E. Hence, we adjust our poly ASP forecast to US$22/US$18kg in 2012/13, and introduce our 2014

poly ASP forecast of US$15/kg.

Our GS SUSTAIN framework analysis shows GS/GH covered tier-1 poly makers – GCL-Poly, Wacker and OCI – have some

potential winning characteristics. GCL is positioned better at cost & scale leadership/market access/downstream

integration advantages, but appears to have a weaker financial position vs. global peers.

Solar midstream: R&D & balance sheet strength are critical – a case study on the transformation of the TV industry c-Si module has dominated end-market demand for years, mainly due to its cost leadership vs. other technologies. However, its cost

leadership was due mainly to its raw material-based cost structure, which is another driver of the currently excessive oversupply situation.

On the upside, the uniqueness of c-Si technology (i.e., lower capex therefore lower fixed cost per watt, or low elasticity between utilization rate and per unit cost) has made it a more favored technology in a highly cyclical industry (Exhibits 46/47).



The transformational years for China’s CRT-TV industry seem to have much in common with today’s excessive overcapacity in China’s solar midstream segment. The geographical migration of the CRT-TV supply chain demonstrated the importance of technology leadership, followed by scale leadership.

Will we see a similar outcome for Chinese solar manufacturers? Although we think more new entrants to the supply chain are likely as new commercial solar technology emerges, we have yet to see a new technology with the potential to trigger a migration. Of course, just as we couldn’t foresee LCD/LED TV in the 1980s, a solar supply chain migration trigger is difficult to anticipate.

Assuming c-Si PV module remains the leading technology, Chinese solar manufacturers should remain at the global technology forefront, in our view. However, amid the challenges of the current environment, we consider balance sheet strength to be the most critical characteristic for the survival of Chinese midstream solar companies, as the industry heads towards multiple-phase consolidation, and an ongoing need to invest in R&D.

Near-term view:

ASP/earnings remain under pressure across the value chain. We adjust down our module ASP forecasts to US70¢/57¢ per

watt in 2012/13, compared to our previous forecast of US79¢/US67¢ per watt. Meanwhile, we introduce our 2014 module

ASP forecast of US50¢/W.

Our GS SUSTAIN framework analysis shows that among GS/GH covered midstream global solar manufacturers, First

Solar appears to be best positioned for the long term. We are yet to see signs of structural rationalization, while lower

replacement costs are making the kick-starting of consolidation even more difficult. We thus consider financial strength

the most important metric, indicating an ability to facilitate corporate activity and to pursue R&D amid industry turbulence.

Solar equipment: The extended down-cycle has hurt order intakes for furnace makers. Prior to Phase I rationalization, therefore, we believe only downstream inverter makers will see earnings growth due to domestic market growth.

Earnings/TP adjustments: We lower our sector EPS by up to 1,327% over 2012-2014E to reflect expectations of further ASP

erosion and an extended down-cycle. Meanwhile, we have lowered cost structures per TSL/STP/YGE’s most recent guidance and

our poly price outlook. Overall, we are lowering our target prices across the sector by 24%-60%.

We resume coverage on JA Solar with Neutral. We believe it has the strongest balance sheet among its peers, which should give it greater flexibility amid a rapidly transforming industry landscape. Despite the implied 25% upside to our target price, we resume coverage at Neutral to reflect the stock's volatile trading history and small market capitalization. We maintain Buy on Sungrow, Neutral ratings on GCL-Poly, Trina and Jing Gong, and Sell ratings on Yingli and Jingyuntong.

We apply a 0.46X industry average ValRatio to derive the target EV/GCI underlying our target prices. Sungrow offers 21% potential

upside at its revised TP, and JA Solar offers 25% potential upside, while the remainder of our China solar coverage target prices

imply 0% to 65% downside.



Exhibit 4: We lower target prices for our solar coverage, and resume coverage on JA Solar with Neutral Rating and target price changes for our solar coverage

Source: Datastream, Gao Hua Securities Research estimates.

Price Potential

New Old Change New Old % Change 7-Sep-12 up/downside 2012E 2013E


GCL-Poly 3800.HK HK$ 2.4 Neutral Neutral 1.10 1.70 -35% 1.19 -8% 0.9x 0.9x

Trina TSL US$ 0.3 Neutral Neutral 4.00 5.50 -27% 4.02 0% 0.3x 0.3x

JA Solar JASO US$ 0.2 Neutral Not Rated N.A. 1.00 N.A. N.A. 0.80 25% 0.2x 0.2x

Yingli YGE US$ 0.3 Sell Sell 0.80 2.00 -60% 1.67 -52% 0.2x 0.2x


Sungrow Power 300274.SZ Rmb 0.5 Buy Buy 11.00 14.80 -26% 9.12 21% 1.8x 1.7x

Jing Gong 002006.SZ Rmb 0.7 Neutral Neutral 8.40 11.00 -24% 9.25 -9% 2.9x 2.6x

Jinyuntong 601908.SS Rmb 0.9 Sell Sell 2.40 5.30 -55% 6.77 -65% 0.6x 0.5x

Implied P/BCompany Ticker Currency

Rating 12-m target priceMarket cap

(US$ bn)



Exhibit 5: We are still fairly bearish on this sector across GS/GH global solar coverage

Note: OCI is covered by Seung Shin; European stocks are covered by Benjamin Moore; US stocks by Brian Lee. Source: Datastream, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Stock Target Target price PriceMarket

cap

Company Ticker rating price period 7-Sep-2012 (US$bn) 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Asia - offshore

GCL-Poly 3800.HK Neutral 1.10HK$ 12-m 1.19 -8% 2.4 (0.06) 0.00 0.04 N.A. N.A. 32 0.9 0.9 0.9 (5) 0 3 11 9 8

OCI 010060.KS Neutral 178,000Won 12-m 191,500 -7% 4.0 6,851 10,850 10,704 28 18 18 1.3 1.3 1.2 5 7 7 7 6 5

Trina TSL Neutral 4.00US$ 12-m 4.02 0% 0.3 (2.62) (0.71) (0.42) N.A. N.A. N.A. 0.3 0.3 0.3 (18) (6) (4) N.A. 10 8

JA Solar JASO Neutral 1.00US$ 12-m 0.80 25% 0.2 (1.00) (0.36) (0.32) N.A. N.A. N.A. 0.2 0.2 0.2 (20) (8) (7) N.A. 12 10

Yingli YGE Sell 0.80US$ 12-m 1.67 -52% 0.3 (1.92) (1.21) (0.95) N.A. N.A. N.A. 0.4 0.5 0.7 (39) (30) (32) N.A. 30 21

Average 28 18 25 0.6 0.6 0.7 (15) (7) (6.6) 9 13 10

Asia - onshore

Sungrow 300274.SZ Buy 11.00Rmb 12-m 9.12 21% 0.5 0.42 0.52 0.67 22 17 14 1.5 1.4 1.2 7 8 10 16 11 8

JG 002006.SZ Neutral 8.40Rmb 12-m 9.25 -9% 0.7 0.13 0.40 0.47 73 23 20 3.2 2.8 2.5 4 13 13 58 31 27

JYT 601908.SS Sell 2.40Rmb 12-m 6.77 -65% 0.9 0.07 0.12 0.14 95 55 49 1.6 1.5 1.5 2 3 3 N.A. N.A. N.A.

Average 63 32 27 2.1 1.9 1.7 4 8 9 37 21 17

Europe

Wacker WCHG.DE Sell 45.00EUR 6-m 53.83 -16% 3.4 2.93 3.49 3.28 26 15 16 1.0 1.0 1.0 4 7 6 6 5 5

Average 26 15 16 1.0 1.0 1.0 4 7 6 6 5 5

US

First Solar FSLR Neutral 15.00US$ 6-m 20.44 -27% 1.8 4.59 4.04 1.76 4 5 12 0.5 0.4 0.4 11 9 4 2 1 N.A.

MEMC WFR Neutral 2.50US$ 6-m 3.19 -22% 0.7 (0.12) 0.25 0.30 N.A. 13 11 1.3 1.2 1.1 (4) 10 11 4 3 2

STR STRI Neutral 3.50US$ 6-m 3.07 14% 0.1 0.05 0.20 0.35 58 15 9 0.5 0.5 0.4 (3) (1) 1 N.A. 3 2

SunPower SPWR Sell 4.00US$ 6-m 4.46 -10% 0.5 (0.15) 0.25 0.50 N.A. 18 9 0.5 0.5 0.5 (1) 3 6 8 6 5

Average 31 13 10 0.7 0.6 0.6 1 5 5 5 3 3

Potential up/downside

P/E (X)EPS (reporting currency) ROE (%) EV / EBITDA (X)P/B (X)



Phase I: Rationalization is inevitable; reduced access to funding the likely catalyst

Although Chinese companies were relative latecomers to solar technology commercialization, they have dominated c-Si module capacity since 2007/2008 and now have the most comprehensive c-Si module supply chain (including raw materials) in the world. Consolidation and

rationalization has historically been the domain of Western companies in this sector, mainly due to higher cost disadvantage, but

the depth and duration of the down-cycle suggests listed Chinese companies may now have to participate, too.

Exhibit 6: Competitive landscape of the global solar industry – many players have chosen to exit the supply chain

Source: Company data, Gao Hua Securities Research estimates.

Solar companies 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012EGCL GCL GCL GCL GCL GCL GCL

DAQO DAQO DAQO DAQO DAQO

ReneSola ReneSola ReneSola ReneSola ReneSola ReneSola ReneSola

LDK LDK LDK LDK LDK LDK

Solarfun Solarfun Solarfun Hanwha Hanwha Hanwha Hanwha Hanwha Hanwha

Suntech Suntech Suntech Suntech Suntech Suntech Suntech

Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar

Trina Trina Trina Trina Trina Trina

JA Solar JA Solar JA Solar JA Solar JA Solar JA Solar

Yingli Yingli Yingli Yingli Yingli

Jinko Jinko Jinko Jinko Jinko Jinko

China Sunergy China Sunergy China Sunergy China Sunergy China Sunergy China Sunergy

Erdos Erdos

Eging Eging Eging Eging

Sunflower Sunflower Sunflower Sunflower

Chao Ri Chao Ri Chao Ri

Hengdian Hengdian

Risen Risen Risen Risen

Hareon Hareon

TBEA TBEA TBEA

Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri

Estimated number of other

Chinese companies0 0 20-50 50-80 100+ 200+ 400+ 300+ 300+ 500+ 400+ 50-100

Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama Tokuyama

OCI OCI OCI OCI OCI

Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech

E-Ton E-Ton E-Ton E-Ton E-Ton E-Ton E-Ton

Gintech Gintech Gintech Gintech Gintech Gintech

Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier

Honda Soltec Honda Soltec Honda Soltec Honda Soltec Honda Soltec Honda Soltec Honda Soltec

Wacker Wacker Wacker Wacker Wacker Wacker Wacker Wacker Wacker Wacker Wacker Wacker

REC REC REC REC REC REC REC

Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld

Ersol Ersol Ersol Ersol Ersol Ersol Ersol Ersol Bosch Solar Bosch Solar Bosch Solar Bosch Solar

Solaria Energia Solaria Energia Solaria Energia Solaria Energia Solaria Energia Solaria Energia Solaria Energia

Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth

Avancis Avancis Avancis Avancis Avancis Avancis Avancis

Solon Solon Solon Solon Solon Solon Solon Solon Solon

Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells

Sunways Sunways Sunways Sunways Sunways Sunways Sunways

Centrosolar Centrosolar Centrosolar

PV Crystalox PV Crystalox PV Crystalox PV Crystalox PV Crystalox

Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture

Solibro Solibro Solibro Solibro

MEMC MEMC MEMC MEMC MEMC MEMC MEMC

Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock

Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar

Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar

MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé

First Solar First Solar First Solar First Solar First Solar First Solar First Solar

Stion Stion Stion Stion Stion Stion Stion

NuvoSun NuvoSun NuvoSun NuvoSun NuvoSun

Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen

SunPower SunPower SunPower SunPower SunPower

Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion Energy Conversion

Arise Arise Arise Arise Arise Arise Arise Arise Arise Arise

Solyndra Solyndra Solyndra Solyndra Solyndra Solyndra Solyndra

Global Solar Global Solar Global Solar Global Solar Global Solar Global Solar

Asia (except china)

China (offshore listed)

China (onshore listed)

Europe

North America



Local banks extending the depth and duration of the down-cycle by maintaining access to credit, in our view. This can be traced to policies dating back to 2005/06 that were originally intended to help China quickly establish a world-class solar product manufacturing supply chain.

Unfortunately, global solar demand is essentially subsidized in any new emerging markets, so demand is vulnerable and highly correlated to local macro economic conditions. Access to finance and grid connectivity are imperative for sustainable solar demand, along with falling costs relative to competing sources of power. Therefore, subsidized new capacity and low entry barriers led to an overcrowded supply chain quickly in mid to late 2010, and demand started to show signs of vulnerability in early 2011.

Near term, we believe a reassessment of current solar policy is needed. Until we see a clear pullback in access to funding, earnings deterioration is likely to persist. Reduced access to capital would impact the creditworthiness/solvency of a highly leveraged and

listed solar company, in our view. Although the unsustainability of the China solar investment cycle is well-recognized by the Chinese business/political community, in our view, socioeconomic concerns seem to carry greater weight at this juncture, even though most Chinese listed companies have reported negative earnings since 2Q2011. While we can’t predict the exact timing of rationalization kicking off in the next 6 to 12 months, we have looked to listed solar companies’ debt profiles, credit metrics outstanding bond yields for indications of risks that may lie ahead.

Exhibits 7-8 detail Chinese solar companies’ outstanding bonds due up to end-2013 and highlight the divergence in yields between offshore and onshore debt. Suntech’s offshore CB due in March 2013 stands out as the largest debt in both size and yield. The very large divergence between on/offshore bond yields suggest the offshore capital market is pricing in a high-risk outlook for offshore debt, with Suntech’s current CB yield of 285% ranking well above its peers.

Exhibit 7: Outstanding solar bonds with maturity between now and Dec. 13EChinese solar makers’ bond amounts and maturity dates

Exhibit 8: Significant divergence between on/offshore bond yields Chinese solar companies’ on/offshore bond yields comparison

Note: Bond amount is using the data of Sep 06, 2012.

Note: Yield is using the data of Sep 06, 2012. Source: Company data, Bloomberg, Gao Hua Securities Research. Source: Company data, Bloomberg, Gao Hua Securities Research.

Company Bond amount Maturity

Offshore bond

Yingli US$ 1 mn Dec, 2012

Suntech US$ 541 mn Mar, 2013

LDK US$ 3 mn Apr, 2013

LDK US$ 21 mn Apr, 2013

JA Solar US$ 220 mn May, 2013

China Sunergy US$ 28 mn Jun, 2013

Trina US$ 113 mn Jul, 2013

Onshore bond

LDK Rmb 400 mn Oct, 2012

Jinko Rmb 300 mn Apr, 2013

Bond name Yield

Offshore bond

Suntech USD CB 2013 284.8%

LDK USD CB 2013 93.8%

China Sunergy USD CB 2013 49.7%

Jinko USD CB 2014 29.7%

Trina USD CB 2013 20.1%

ReneSolar USD CB 2016 17.7%

JA Solar USD CB 2013 16.3%

Hanwah USD CB 2015 10.4%

Onshore bond

LDK Domestic Bond 2014 6.8%

Jinko Domestic Bond 2013 6.2%

Yingli Domestic Bond 2016 6.1%

GCL-Poly Domestic Bond 2015 5.8%



The credit metric assessment in Exhibit 9 provides an indication as to where risks lie. Offshore listed Chinese companies are financing

through new debt, leading to cash deficits in 2012, while onshore listed equipment makers seem to face less financial stress.

Offshore listed names have a heavy reliance on short-term debt, with eight companies having over 50% of their debt coming due in the next 12 months. Even if we assume all their short-term debt can be rolled over, lower EBIDTA/interest ratios (LDK is lowest among upstream names) indicate higher likelihood of cash deficit or need for new funding for ongoing capex and working capital.

Exhibit 9: Debt profile and key credit metric of listed Chinese solar companies should give us some indication of where risks lay

Note: (1) NC= Not Covered; (2) LDK, Suntech and Hanwha’s key balance sheet items are derived from 1Q12 financial data. Source: Company data, Bloomberg, Gao Hua Securities Research.

We estimate that the total outstanding bank loans of listed offshore Chinese companies to be US$19bn. Furthermore, including A-

share listed and non-listed solar companies, the aggregate total would be roughly estimated to be US$23bn. Under various default scenarios, the NPL ratio of China banks would likely increase slightly (Exhibits 10/11). From a policy perspective, Premier Wen Jiabao’s call last March for an end to the “blind expansion” of solar/wind investment could have been taken as a signal to China’s banks to rein in solar lending. However, we think this might not take place until after next month’s 18th National Congress Meeting of the Communist Party of China, where the Congress will elect the 18th Central Committee of the Communist Party of China.

Onshore Offshore Cash ST debt LT debtST as % of

total debt

Net debt /

equity (%)

Gross debt /

EBITDA

EBITDA /

Interest

Capex

(US$, mn)

Free cash

flow

Funding

source

Chinese solar upstream makers

GCL-Poly 3800.HK Neutral 80% 20% 1,133 2,097 2,954 42% 145% 3.7 7.8 1,925 Negative Debt

LDK* LDK NC 95% 5% 136 2,251 1,173 66% 487% (15.1) (1.2) 1,114 Negative Debt

Renesola SOL NC 88% 12% 314 691 242 74% 119% 9.7 2.5 153 Negative Debt

DAQO DAQO NC 100% 0% 66 120 224 35% 65% 2.7 12.2 252 Negative Debt

Average 91% 9% 412 1,290 1,148 54% 204% 0.2 5.3 861

Chinese solar midstream makers

TrinaSolar TSL Neutral 63% 37% 660 734 569 56% 63% 11.3 2.9 361 Negative Debt

JA Solar JASO Neutral 79% 21% 589 382 672 36% 26% 18.5 1.0 341 Negative Debt

Yingli YGE Sell 95% 5% 882 1,304 1,387 48% 180% 20 1.2 769 Negative Debt

Suntech* STP NC 61% 39% 474 2,086 178 92% 221% 29.0 0.6 367 Negative Debt

Hanwha* HSOL NC 45% 55% 303 483 232 68% 65% (61.7) (4.3) 387 Negative Debt

China Sunergy CSUN NC 95% 5% 205 503 129 80% 432% (12.1) (1.9) 60 Negative Debt

Jinko JKS NC 80% 20% 73 629 107 85% 160% 7.5 3.5 312 Negative Debt

Canadian Solar CSIQ NC 100% 0% 330 928 137 87% 152% 15.1 1.4 205 Negative Debt

Average 77% 23% 440 881 426 69% 162% 3.4 0.5 350

Chinese solar equipment makers

Sungrow 300274.SZ Buy 100% 0% 190 45 0 100% (50%) 1.1 97.8 13 Negative Equity

Jing Gong 002006.SZ Neutral 100% 0% 75 113 0 100% 20% 1.3 18.6 20 Negative Equity

Jingyuntong 601908.SS Sell 100% 0% 346 73 13 85% (46%) 0.6 31.9 31 Negative Equity

Average 100% 0% 204 77 4 95% (25%) 1.0 49.4 21

Debt profile (2Q12) Credit matrix (2011)Key balance sheet items (2Q12, US$ mn)

Company Ticker Stock rating



Exhibit 10: Our definition of at-risk capacity – net debt/equity ratio greater than 150%, or EBITDA/interest expense less than 1X Chinese solar companies’ net debt/equity and EBITDA/interest expense

Exhibit 11: If Chinese solar companies default on their debt, it might only have a slight impact on total Chinese banks’ NPL ratios Analysis of Chinese banks’ NPL ratios in different solar loan default scenarios

Note: Hanwha, Suntech and LDK’s net debt/equity are derived from 1Q12 financial data.

Source: Company data, Gao Hua Securities Research estimates. Source: Company data, CEIC, Gao Hua Securities Research estimates.

Cash deficit analysis is the most critical assessment for our coverage. The cash deficit status of a company in the next two years (Exhibit 12) is an indicator of financial stability and, hence, the possibility for improved competitiveness in a highly challenging environment.

Exhibit 12: YGE could remain in a cash deficit for the next three years


Net debt / Equity EBITDA / Interest expense

2Q12 2011

Sungrow -50% 97.8

Jingyuntong -46% 31.9

Jing Gong 20% 18.6

JA Solar 26% 1.0

Trina 63% 2.9

DAQO 65% 12.2

Renesola 119% 2.5

GCL-Poly 145% 7.8

Hanwha* 65% -4.3

Canadian Solar 152% 1.4

Jinko 160% 3.5

Yingli 180% 1.2

Suntech* 221% 0.6

China Sunergy 432% -1.9

LDK* 487% -1.2

Company

1.77%

1.79%

1.81%

1.75%

1.76%

1.77%

1.78%

1.79%

1.80%

1.81%

1.82%

Actual ratio at the end of 2011 If solar companies with credit

risk default their total debt

If all Chinese solar companies

default on their debt

China banking industry NPL ratio

(US$ mn)

Company 2011 2012E 2013E 2014E 2011 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Tier-1 polysilicon makers

GCL Neutral 886 1,017 976 958 1,491 1,983 1,725 1,468 1.8 2.2 2.5 133% 125% 113% (90) 411 433 (657) (193) (193) 747 (217) (239)

GS covered off-shore listed midstream solar companies

TSL Neutral 817 712 648 573 389 589 489 389 (2.8) 1.2 1.6 60% 57% 55% (155) 66 55 (150) (30) (30) 305 (36) (25)

JASO Neutral 618 689 364 371 84 84 (96) (96) (0.2) 0.9 1.1 16% 34% 34% 111 (105) 46 (40) (40) (40) (71) 145 (6)

YGE Sell 664 539 630 502 1,329 1,329 1,329 1,329 (0.4) 0.5 0.8 208% 262% 349% (123) 41 (77) (400) (50) (50) 523 9 127

Cash flow from operations Cash flow from investment Cash deficit

Rating

Year-end cash balanceEBITDA/

int. exp.

Net debt/

equityST debt



Phase II: Cost/scale leadership is essential for upstream; lessons from DRAM experience Four decades of DRAM industry evolution provide us with a useful roadmap for the consolidation of solar upstream. Similarities between the two sectors include large exposure to commodity projects, high capex and fixed costs, and downstream integration enabling high order visibility. The DRAM (Dynamic Random Access Memory) market size in USD peaked as early as 1995, although DRAM remains widely used in IT applications since then. Global DRAM industry consolidation has taken roughly two decades, and the number of pure DRAM makers declined from 18 to six names (Exhibit 13). In 1H12 we are finally seeing some signs of supply/demand balance. From 1997 to 2011, the DRAM industry spent capex of US$153bn but earned only an aggregate US$21.9bn in operating profit. Excluding Samsung and Hynix, other DRAM makers suffered a US$14.0bn operating loss over the same period. The profitability outlook for Taiwan’s DRAM industry is the weakest vs. its global peers, as it has seen large losses since 2007.

Exhibit 13: The market size (in USD) of the DRAM sector peaked as early as 1995, with the ensuing consolidation taking almost two decades Global DRAM market revenue vs. DRAM makers’ migration (1974-2012)

Source: Company data, Gao Hua Securities Research.

Progressing faster than DRAM, solar upstream appears to be at mid-to-late stage of consolidation In the relatively short history of the solar industry, the upstream segment has seen rationalization twice (Exhibit 14). During the 2008-09 downturn, companies were much smaller and could not sustain the poly price collapse (from US$400/kg in 3Q08 to US$83/kg in 2Q09)

18

6

0

5

10

15

20

25

30

0

5

10

15

20

25

30

35

40

45

1974

197

5

19

76

197

7

197

8

197

9

198

0

198

1

198

2

198

3

198

4

198

5

198

6

198

7

198

8

198

9

199

0

199

1

199

2

199

3

199

4

199

5

199

6

199

7

199

8

199

9

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

2010

2011

Number of global DRAM players (RHS) Global DRAM market value (LHS)

Beginning of the Japan era,

driven by government

support and strong domestic

demand - vertically

integrated electronics

producers

Commoditization era

and launch of Win95.

Three Korean companies

take 32% of market share

Japan hit by recession and its

domestic DRAM market

weakens. Korea era begins

Taiwanese companies

enter supply chain

(US$ bn)

US era of the DRAM

industry, driven by

technology innovation,

but dragged by high

CAPEX of DRAM

investment

Almost two decades'consolidation



combined with higher fixed costs, and many of them exited the supply chain. In the current wave of rationalization, the number of Chinese poly makers has declined from over 50 in mid-2011 to four or five currently, including listed names GCL-Poly, ReneSola, DAQO, LDK and TBEA.

Exhibit 14: Solar upstream is at mid-to-late stage of DRAM consolidation process, although we think its addressable market size peaked in 2010-11

Source: Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Two key lessons from comparison of DRAM and solar upstream

DRAM lesson #1: Demand elasticity Demand elasticity works more effectively throughout cycles in the DRAM industry than in the solar industry. Hence, higher demand visibility has encouraged more new capex in DRAM throughout cycles, leading to extended down-cycles.

DRAM industry development has been characterized by relentless technology advancements (in line with Moore's Law – Gordon Moore’s prediction in 1965 that the number of transistors per area would double every 18 months to two years), resulting in costs and prices declining at a CAGR of 34% from 1989 to May 2012 (Exhibit 15). With continuously cheaper prices, DRAM shipment volumes expanded at a CAGR of 49% from 1985 to 2011 (Exhibit 16), representing the transformational product of the consumer electronics industry – from desktops to notebooks, mobile handsets, smartphones, tablets, etc.

68

18

0

10

20

30

40

50

60

70

80

90

100

0

5

10

15

20

25

30

35

40

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012E 2013E 2014E 2015E

Number of global poly makers (RHS) Global solar market value (LHS)(US$ bn)

European

financial crisis

Global financial crisis

and Spanish pullback

cap solar installation

Strong demand

from Spain and

Germany

Peak of solar market value,

driven by strong German and

Italian demand Germany and Italy

cut FiT significantly

Addressable market value

will continue to decline,

based on our demand and

module ASP forecasts

First wave of poly

industry

consolidation



Exhibit 15: Persistent DRAM ASP decline in line with Moore’s Law (numbers of transistors per area double every 18 months to two years) DRAM ASP trend (1989-2012)

Exhibit 16: Demand elasticity (falling ASP/higher demand) works well in the DRAM industry Global DRAM shipment trend (1985-2011)

Source: DRAMeXchange, Gao Hua Securities Research. Source: World Semiconductor Trade Statistics, Gao Hua Securities Research.

Solar demand dynamic is different – falling ASPs might impact subsidy schemes, discouraging upstream new capex In the solar industry, however, the demand dynamic amid falling ASPs is different and – to a certain extent – dysfunctional. Falling ASPs typically cause subsidy schemes in a particular market to be revised, thus discouraging new capex in high capex segments, such as solar upstream. Prior to the revision of the subsidy schemes, there would typically be a strong, short-term demand elasticity effect. For example, growth in the solar space has outstripped all but the most bullish forecasts in the past few years, driven mainly by higher electricity price regions such as Spain, Italy and Germany; however, FiT subsidy schemes have been revised down over time, due to lower module and thus lower total installation cost. Hence, the degree of ASP erosion has not corrected to reflect higher demand volume in 2011/12E (Exhibits 19/20).

Grid parity is a moving target Aside from the availability of funding and a well-built grid system, the two key criteria for secular PV growth in the eyes of many solar supporters are grid parity and, in turn, the volume inflection point.

In our view, grid parity is not a threshold after which PV will quickly become a dominant energy source, because established infrastructure and utility business models will take time to change. Grid parity is a moving target and will happen in different geographies at different times.

Competing energy sources will adjust to the challenge from PV and other alternatives. Hydraulic fracking – especially within the US – has caused natural gas production to increase dramatically and resulted in a correction in gas prices to the extent they now rival coal. Declining fossil fuel prices have also increased the challenge to grid parity. In addition, the grid infrastructure is not optimized for distributing PV energy,

7,349

0.5

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

DRAM ASP(US$/Gb)

CAGR of ASPdelcine - 34%

(1989 - 2012)

1

22,476

0

5,000

10,000

15,000

20,000

25,000

30,000

DRAM global annual shipment(mn Gb)

The CAGR of DRAM globalshipment is 49%

(1985 - 2011)

Mobile EraMobile --> Smartphone

Notebook --> Tablet

PC EraDesktop --> Notebook



which is intermittent by nature. Most grids are designed to deliver electricity from power plants to industrial and population centers. This could be addressed in the future when energy storage technology is more mature. However, currently, these technologies are still too expensive today with limited scale feasibility to ensure grid stability.

Therefore, while grid parity in major electricity markets would represent a milestone for the PV industry, it would not necessarily represent an inflection point for overall PV growth. For example, LCOE (Levelised Cost of Energy) of China energy sources has changed recently, given that the coal price has come down significantly (Exhibit 17). China is poised to see the highest PV installation growth globally (Exhibit 18), but its legacy infrastructure may cause challenges for PV power in accessing the grid. Implementing infrastructure changes for distributed generation will take time and significant investment. Reality, therefore, might slow growth momentum.

Therefore, we remain quite cautious on our demand forecasts over the next few years (Exhibit 21/22). Meanwhile, year to

date, we have heard little in the way of new greenfield capacity development in 2013 from global tier-1/2 poly makers,

although construction work on a few projects is still under way.

Exhibit 17: LCOE cost analysis of most energy sources in China declined yoy LCOE cost analysis of different energy sources in China

Exhibit 18: We expect China demand growth to outpace global demand growth China solar demand growth vs. global (2010-2015E)

Source: Goldman Sachs Research estimates, Gao Hua Securities Research estimates. Source: Company data, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

0.86

0.59

0.52

0.36

0.28 0.24

0.77

0.53 0.52

0.31 0.28

0.24

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

Solar Wind Gas Coal Hydro Nuclear

2011 2012E(Rmb/kWh)

21 23 22 21

23 25

1

3 6 8

9

10

0

5

10

15

20

25

30

35

40

2010 2011 2012E 2013E 2014E 2015E

China PV demand Ex-China global PV demand(GW)



Exhibit 19: Poly ASP reached a new historical low recently Poly ASP trend (2H03-1H12)

Exhibit 20: Slower demand growth outlook, despite poly ASP dropping to historically low level... Global solar-graded poly shipment (2004-2015E)

Source: PVinsights, Gao Hua Securities Research.

Source: PVinsights, Gao Hua Securities Research estimates. Exhibit 21: …due to sustainable demand driven by stability of 1) policy schemes, 2) financial lending systems; 3) grid systems in a particular marketGlobal solar PV market demand forecast

Exhibit 22: By 2013, we expect Asia to be the largest solar market globally Global solar demand growth perspective by regions and countries



DRAM lesson #2: Pricing trends Poly ASP has dropped close to marginal companies’ variable costs (COGS minus D&A), but we would want to see some cancellations of ongoing new capacity projects (e.g., ReneSola’s phase II 6000MT, DQ’s Xingjiang 6000MT, TBEA’s Xingjiang 12K MT) before we call a bottom for poly prices. Similar to the DRAM experience, due to high capex and fixed cost (Exhibits 23), marginal companies tend to hang in there – either struggling through by cutting costs further or trying to complete projects already in construction – until ASP falls below variable cost for an extended period of time.

4050

7080

90

110

300

350

400

500

200

102

62 55

7860

3922

0

100

200

300

400

500

2H03 1H04 2H04 1H05 2H05 1H06 2H06 1H07 2H07 1H08 2H08 1H09 2H09 1H10 2H10 1H11 2H11 1H12

(US$/kg)

Poly price has dropped to

historical low

3

1620

24

41

51

147152

157 154162

167

0

20

40

60

80

100

120

140

160

180

200

2004 2005 2006 2007 2008 2009 2010 2011 2012E 2013E 2014E 2015E

(KMT) Slower period of PV

global installation

Fast growth period

of PV global

installation

0.3 0.3 0.4 0.6 1.1 1.5 2.0 2.9

5.7

8.1

21.2

25.6

28.2 28.8

31.9

35.0

41%

20%

31%35%

77%

39%35%

49%

95%

40%

163%

21%

10%

2%

11% 9%

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012E 2013E 2014E 2015E

Global module shipment Global module shipment yoy growth (RHS)(GW)

Driven by Germany

demand

Driven by Spain

demand

Driven by Germany &

Italy

demand

Germany & Italy have cut solar FiT

significantly

(GW) 2011 2012E 2013E 2014E 2015E

China 2.7 6.0 8.0 9.0 10.0

Japan 1.5 2.2 5.0 5.0 2.7

Asia 5.4 10.1 15.5 17.6 18.1

Growth 150% 88% 53% 14% 3%

Germany 7.5 6.5 4.5 4.5 5.0

Italy 5.7 4.0 0.7 0.7 1.0

Europe 17.1 13.8 7.6 7.4 8.6

Growth -3% -20% -45% -3% 17%

USA 2.0 3.0 4.0 5.0 6.0

North America 2.5 3.7 5.0 6.0 7.0

Growth 108% 48% 35% 20% 17%

Global total 25.6 28.2 28.8 31.9 35.0

Growth 21% 10% 2% 11% 9%



The historical DRAM pricing trend is similar. During the oversupply period of the DRAM industry cycle, prices fell sharply – often right through total cost, which is equal to fixed cost (SG&A plus R&D) plus variable cost (COGS minus D&A). Eventually, ASP hits a marginal company’s variable cost line (COGS minus D&A) (Exhibit 24). Hence, a marginal company’s variable cost provides a clear floor for

ASPs. At this point, the marginal player with the worst cost structure has to reduce production by either closing inefficient lines or

reducing utilization. We would expect to see a similar pattern in the solar upstream segment – poly ASP falling to marginal companies’ variable cost levels (Exhibits 25/26).

Exhibit 23: DRAM and solar upstream both have high entry/exit barriers due to persistently high capex and fixed cost DRAM industry and solar upstream segment’s cost structure and capex

Exhibit 24: Marginal DRAM companies’ cash cost is floor for ASP downsideGlobal DRAM companies’ cost breakdown

Note: 1) Fixed cost includes SG&A, R&D and D&A costs; 2) Variable cost equal to reported COGS, excluding D&A costs; 3) Deriving DRAM industry average from key DRAM producers’ company data, including Samsung, Hynix, Micron, Elpida, Nanya, ProMOS and Powerchip; 4) Deriving solar industry average from key poly makers’ company data, including GCL-Poly, OCI, Wacker and REC. Source: Company data, World Semiconductor Trade Statistics, Gao Hua Securities Research.

Source: Company data, World Semiconductor Trade Statistics, Goldman Sachs Research estimates, Gao Hua Securities Research.

Exhibit 25: Poly ASP to fall to marginal players‘ variable cost level in 2012E…2012E polysilicon cost curve

Exhibit 26: ... as well as in 2013E 2013E polysilicon cost curve

Source: Company data, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.


0%

10%

20%

30%

40%

50%

60%

70%

80%

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

DRAM industry capex/revenue

Solar upstream capex/revenue

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Industry Supply (Gbs)

Samsung

DRAM cash cost (US$/Gb)

Elpida, Micron

Taiwanese

players

Hynix

Width of wafer line at 3x nm level



ASP floor

Fixed

cost

Variablecost

GS

estimate

of 2012

global

demand

0

5

10

15

20

25

30

35

40

45

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44

Poly cash cost (US$/kg)

4Q12E capacity in wafer (GW)

Wacker

Hemlock

GCL OCI

MEMC

Tokuyama (supply

Japan market only)

ASP

GS forecast of global C-Si

module demand in 2012

Estimated

WIP inventory

(5-10 GW)

DAQO's

0.7 GW capacity

ReneSola's

0.7 GW capacity

TBEA's

0.3 GW capacity

LDK's

3.4 GW capacity

Yingli's

0.5 GW capacity

REC

Variable cost

Fixed cost

0

5

10

15

20

25

30

35

40

45

50

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 51

Poly cash cost (US$/kg)

4Q13E capacity in wafer (GW)

Wacker

GCL OCI

MEMC

REC

ASP

GS forecast of global C-Si

module demand in 2013

Estimated

WIP inventory

(5-10 GW)

ReneSola's

1.7 GW capacity

TBEA's

2.3 GW capacity

DAQO's

1.3 GW capacity

LDK's

3.4 GW capacity

Yingli's

0.5 GW capacity

Hemlock

Tokuyama (supply

Japan market only)

Fixed cost

Variable cost



Downside risk looks limited for poly prices, as further cost reduction in polysilicon production is likely to be derived mainly from location advantage (cheaper electricity).

Over 90% of global poly production is based on Modified Siemens Method (MSM), with the rest based on Fluidized Bed Reactor (FBR) method. Given that electricity cost accounts for 40%+ of total production cost via MSM (on a per unit basis), poly makers are trying to lower electricity cost by: 1) locating new capacity near cheaper electricity sources/abundant coal supply, such as in Xinjiang, Mongolia and Shanxi provinces; 2) upgrading MSM to Hydrochlorination from Hydrogenation, which could reduce electricity usage by more than 50%; and 3) combining MSM and FBR methods together in one plant, which can increase the recycling efficiency of the production process and lower the average working temperature to lower electricity usage further (Exhibit 27).

Therefore, prior to Phase I consolidation, and considering marginal players’ ongoing projects and tier-1 poly makers’ ongoing/debottlenecking projects scheduled to start from 2014, we expect poly ASP to decline steadily over 2013/14E. We adjust our poly ASP forecasts to US$22/US$18kg in 2012/13, and introduce our 2014 poly ASP forecast of US$15/kg (Exhibit 29).

Exhibit 27: While tier-1 poly makers have announced limited new capacity additions, we expect marginal companies to struggle prior to Phase I consolidation Global major poly makers’ profiles

Note: (1) Variable cost equals COGS minus depreciation and amortization; (2) Fixed cost equals SG&A and R&D cost at operating level. Source: Company data, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Tier-1/2 global poly makers

GCL- Poly 65,000 11.4 14.0 2.3 65,000 11.4 13.0 1.9Using in-house electricity; targeting to utilize

GCL Method to decrease cost further. N.A.

OCI 42,000 7.2 14.0 2.1 52,000 8.9 13.0 1.9 N.A.Announced additional debottlenecking of

P3 with 10,000 MT in June, 2014E.

Wacker 52,000 8.8 15.5 2.5 55,000 10.5 15.0 2.0 N.A.

Continue to press on with development of

new site (with new technology), with 1/3rd

lower cost per unit of energy due to different

location as well as significant cost saving

afterwards 2014E.

Hemlock 46,000 5.2 18.0 2.5 50,000 5.6 17.0 2.3 N.A. N.A.

REC 21,500 3.5 17.3 2.3 21,700 3.6 15.0 2.0FBR production target of 15,000MT in 2012E,

40% above design capacity. N.A.

MEMC 8,425 0.7 26.5 2.4 9,425 0.7 25.0 2.2 N.A. N.A.

Tokuyama 9,200 1.6 23.0 2.5 17,200 3.0 22.0 2.3 N.A.

Malaysia project in progress with total

production capacity of 20,000 tons when

phase 2 is completed.Total global tier-1 players 244,125 38.4 18.3 2.4 270,325 43.6 17.1 2.1

Tier-2 Chinese poly makers

DAQO 4,300 0.7 22.0 3.3 7,300 1.3 18.0 2.3Locating new capacity in Xinjiang to enjoy

lower electricity price.Adding 3K MT in Xinjiang.

ReneSola 4,000 0.7 22.0 3.4 10,000 1.7 19.0 2.5Developing new poly manufacturing

technology.

Adding 6K MT with modified Siemen

method (Hydrochlorination).

TBEA (located in Xinjiang) 1,500 0.3 23.0 3.0 13,500 2.3 18.0 2.2Locating new capacity in Xinjiang to enjoy

lower electricity price.Adding 12K MT in Xinjiang

Others (located in Xinjiang) 3,000 0.5 22.5 3.1 3,000 0.5 19.9 2.5 N.A. N.A.

Total tier-2 Chinese players 12,800 2.2 22.4 3.2 33,800 5.8 18.7 2.4 N.A. N.A.

Higher cost listed poly makers

LDK 20,000 3.4 35.0 N.A. 20,000 3.4 N.A. N.A. N.A. N.A.

Yingli 3,000 0.5 40.0 N.A. 3,000 0.5 N.A. N.A. N.A. N.A.

4Q13E

Estimated

variable cost

(US$/kg)

Estimated fixed

cost (US$/kg)

Estimated

variable cost

(US$/kg)

Estimated fixed

cost (US$/kg)

Company and comparison Existing capacity's further cost reduction

potential Further capacity expansion plan

Capacity (MT)

Capacity in

solar graded

wafer (GW)

Capacity in

solar graded

wafer (GW)

Capacity (MT)

4Q12E



Exhibit 28: Electricity is the largest cost element of Modified Siemens MethodPoly production cost breakdown by different technologies

Exhibit 29: We expect poly price to continue to fall, but at a slower rate Global poly spot price (1Q09-2014E)

Source: Company data, Gao Hua Securities Research estimates. Source: PVnews, Gao Hua Securities Research estimates.

Characteristics of winners in the DRAM industry After almost 40 years of fierce competition and low industry returns, the global DRAM industry has consolidated into a rational structure: Samsung, Hynix and Micron. Oligopolistic economics prevailed in this scenario, with healthy profits for the remaining companies, particularly Samsung (Exhibits 30/31). In the next two pages we identify two key elements of Samsung’s successful DRAM strategy.

Exhibit 30: Samsung has consistently gained market share Samsung’s DRAM market share (2001-2011)

Exhibit 31: Samsung has consistently maintained higher profitability Samsung’s operating margin vs. industry average

Note: Deriving industry average from key DRAM producers’ company data, including Samsung, Hynix, Micron, Elpida, Nanya, ProMOS and Powerchip.

Source: Company data, Goldman Sachs Research, Gao Hua Securities Research.


Electricity

consumption

(KWh/kg)

Electricity

cost (%)

Raw material

cost (%)

Maintenance

and other cost

(%)

Modified Siemens

method

(Hydrochlorination)

60 47% 51% 1% GCL-Poly 14.0

Modified Siemens

method

(Hydrogenation)

140 58% 30% 12% DAQO 22.0

Fluidized Bed Reactor

method20 21% 62% 17% REC 12.5

4Q12E target

cash cost

(US$/kg)

Technology

Cost breakdown (4Q12E)Sampling

poly

makers

121

83

67

56 55 55

70

85

7469

55

33

2218

15

0

20

40

60

80

100

120

1Q09 2Q09 3Q09 4Q09 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 2012E 2013E 2014E

(US$/kg)

2322

ASP declined by

25% quarterly(4Q10 - 4Q11)

We expect ASP to

decline by a

CAGR of 17% (2012E -2014E)

GS forecast

(previous) GS forecast

(current)

ASP declined by

23% quarterly(1Q09- 4Q09)

28%

34%32% 31%

33%30%

29%31%

34%

40%

44%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Samsung's DRAM market share

-8%-13%

16%

7%

17%

-8%

-52%

-70%

-3% -9%

34%

27%

48%

28%

37%

9%

-8%

1%

42%

24%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Industry average operating margin (excluding Samsung)

Samsung's operating margin



1) Aggressive investment in low-cost capacity and technology (Exhibits 32/33): Samsung's substantial gross margin advantage is a product of its cost leadership, which stems directly from technology leadership and scale, which in turn is due to aggressive R&D and capex throughout the cycle. Samsung is consistently ahead of the pack in introducing and ramping next-generation process technology.

Exhibit 32: Samsung’s DRAM strategy: invest in the down-cycle Samsung’s capex as % of total industry capex vs. industry operating margin (2001-2011)

Exhibit 33: Samsung’s DRAM strategy: drive new product cost structures Samsung’s DRAM technology roadmap vs. industry

Source: Company data, Goldman Sachs Research, Gao Hua Securities Research. Source: Company data, Goldman Sachs Research, Gao Hua Securities Research.

2) Product diversification & downstream integration (Exhibit 34): Samsung’s diversified business portfolio provides a profit buffer for its DRAM business unit during down-cycles, while its downstream-integrated business model (into different consumer electronic devices) offers better DRAM volume visibility. Three decades of PC-driven DRAM growth ended in the early 2000s. Mobile DRAM and NAND have surged in the past five years, but the business model of selling only standalone discrete memory chips has limitations, as tablet/smartphone makers are demanding complete memory solutions (memory + controller + packaging). Samsung differentiated itself by adopting this new product solution faster than its peers, due mainly to its higher visibility on demand for various electronic devices.

26%

22%

32%

27%

22%

30%

22%

43%

33%

50%

31%

-70%

-60%

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

0%

10%

20%

30%

40%

50%

60%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Samsung's capex as % total industry capex Industry operating margin

Samsung invests at

bottom of the cycles

Wafer line width (nm)

350~400

1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011

1200~1500

1500~2000

800~1000

4x~5x

1800~2000

1500~1800

1000~1200

400~500

600~800

300~350

250~300

200~250

150~200

120~150

9x~120

9x

6x

5x

3x

6x~7x

8x~9x

100~120

120~150

150~200

200~250

250~300

350~400

400~500

500~800

6 m 8 m 11 m 20 m 26 m 26 m 30 m 29 m 30 m 30 m

Product life cycle (months)

SamsungIndustry average1800~20001500~2000

Higher cost

Lower cost



Exhibit 34: Samsung’s DRAM strategy: downstream integration brings higher order visibility Samsung’s revenue breakdown vs. DRAM demand breakdown by appliances


GS SUSTAIN: Identifying potential winners in solar upstream

Is there a future “Samsung” in the solar industry? We apply the GS SUSTAIN framework to our China solar upstream coverage to identify those stocks with standout longer-term potential.

GS SUSTAIN aims to generate outperformance by identifying stocks that can sustain long-term industry-leading CROCI by integrating analysis of return on capital based on cash returns on cash invested (CROCI), industry positioning and management’s engagement in mitigating risks related to environmental, social and governance (ESG) factors. Leveraging this framework, we analyze China’s solar companies to appraise which businesses are best placed to benefit from rationalization and consolidation within the global industry.

Although meaningful returns will require capacity reduction and/or structural acceleration in demand growth, we believe the polysilicon segment has already entered the consolidation phase. We define four measurements of industry positioning (Exhibit 35), assess the winning characteristics of four polysilicon makers (GCL-Poly, Wacker, OCI and MEMC) under GH/GS coverage in the winning table (Exhibit 37), and present their relative competitiveness in the winners’ circle (Exhibit 36).

PC, 50%

Mobile, 15%

Graphic card/TV,

15%

Tablet, 2%

Others, 18%

Global DRAM demand breakdown by appliances (2011)

DRAM, 9%

PC, 7%

Mobile, 32%

TV, 21%

Tablet, 1%

LCD panel, 18%

Others, 12%

Samsung's revenue breakdown by appliances (2011)



Exhibit 35: Structural changes in the solar industry and analysis of drivers of cash returns and growth of the solar upstream segment


Drivers of corporate performance

Solar upstream

Return on capitalIndustry positioningManagement quality GS SUSTAIN

CROCI

Cash return on cash invested

Cost postion Downstream integration

ESG

Environmental, social and

governance issues

Themes

Investment in smart grids

Solar is an intermittent source of electricity supply. Full scale uptake will require a major upgrade to the power grid. Infrastructure changes

take time to implement, and require significant investments.

Stable regulation, pricing and funding

Successful project financing of industrial power generation capacity

requires stable and smaller scale projects, in turn requiring regulatory

stability, transparent pricing and access to funding.

Over-capacity

High FiTs, subsidies and favorable access to capital have

resulted in overinvestment in capacity since 2009/10.

Meaningful recovery in returns will require capacity reduction /

structural acceleration in demand growth

Investing at down-cycles for

further R&D development, and scale leadership

are two most important drivers tolead the cost

curve

Integrating into EPC & developers segment provides

higher order visibility and

earnings upside

Sustained investment in technology and leadingthe

cost curve

Polysilicon’s manufacturing costs have fallen 48% since 1Q2010.

With a capacity overhang of almost 50% and steep industry

cost curve, consistent R&D development, hence, driving down costs is fundamental to maintain

profitability.

Financial strength

Entering a potentially

protracted phase of restructuring players with strong B/S’s are

better placed



Exhibit 36: Winners’ circle – three top-tier global poly makers

Note: MEMC scored below median on all three metrics.


These three top-tier global poly makers potentially have winning characteristics, namely the ability to generate better-than-peers’ returns on capital and to position operating assets to sustain these returns – both long-standing traits of industry leaders post industry consolidation.

Exhibit 37: Winning table – three tier-1 polysilicon makers


Management of ESG issues – ESG (based on 2011 data) above sector median

Industry positioning: leaders on a combination of:– High-potential growth markets– Attractive market structure– Favourable business mix and regulatory outlook- Operating margins

Returns – CROCI (2012-14E) above sector median

Management quality Industry positioning

Return on capital

GCL-PolyOCI

Wacker

Overall

Industry

Positioning

ESG

2012-14E Percentile

R&D cost

/revenue

(1H12)

Percentile

Cash cost

(4Q12E,

US$/kg)

PercentileCapacity

(4Q12E)Percentile

Net

debt/equity

(2013E)

Percentile

EBITDA/

interest

expense

(2013E)

PercentileCash deficit

(2013E)Percentile

Capacity of

downstream

projects on

pipeline (MW)

Percentile Percentile

Score as % of

maximum

(2010/11)

Percentile

(2010/11)

GCL-Poly 8% 67% 0.7% 0% 14.0 100% 65,000 100% 125% 0% 2.2 0% 746.8 33% 3,000 100% 100% 40% 0%

OCI 9% 100% 1.2% 33% 15.0 67% 42,000 33% 20% 67% 11.8 67% -47.8 67% 400 33% 33% 56% 67%

MEMC 7% 33% 2.9% 67% 26.5 0% 8,425 0% 26% 33% 4.0 33% 774.9 0% 2,900 67% 0% 44% 33%

Wacker 4% 0% 3.6% 100% 15.5 33% 52,000 67% 18% 100% 12.7 100% -49.5 100% 0 0% 67% 64% 100%

Management quality

CROCI Financial strength Downstream integration

Industry positioning

Cost positionCompany

Return on capital



Overall, we are unable to identify a clear “future Samsung” candidate from the solar upstream at this stage (Exhibit 36/37),

due mainly to the fact that the solar midstream segment has not yet entered Phase I of consolidation. We expect the

continued depressed earnings/cash outlook of the midstream segment to negatively impact upstream poly makers, even

tier-1 players.

While the exact timing of consolidation for the industry appears to rely more on political events than industry fundamentals, we believe GCL-Poly is the most likely candidate among our GH solar coverage to participate in the early stages of the process, mainly due to GCL’s strong leadership in both cost structure and scale, which we consider the two most important factors to assess the long-term competitiveness of global poly makers at the beginning of consolidation, especially given our view that poly ASP will continue to trend down slightly in 2013/14. Currently, GCL appears to lead the matrix for both cost and scale leadership (Exhibit 38).

Exhibit 38: In the early days of consolidation, we consider cost and scale leadership to be most important Global poly makers’ poly cash cost vs. capacity


Furthermore, GCL’s better-than-peers’ downstream integration and market access advantages should help the company to gain relatively stronger volume visibility.

Downstream integration: GCL is the only fully integrated poly to wafer maker, as well as EPC/solar farm developer globally. Its cost leadership in the wafer segment and large presence at the downstream EPC/solar farm business segment should leverage up its poly sale volumes.

Market access advantage: With the majority of low-cost c-Si module makers, as well as the most competitive raw material supply chain, based in China, China looks set to become the largest PV installation market in the world, which implies a location advantage for GCL vs. its global peers in term of demand sources.

GCL

OCI Wacker

Hemlock

REC

MEMC

ReneSola

DAQO

Tokuyama

TBEA

LDK

YGE

0

5

10

15

20

25

30

35

40

45

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000

Poly capacity (2012E, MT)

Po

ly c

ash

co

st (4

Q12E

,U

S$/k

g)

Cost & scaleleadership



Exhibit 39: GCL-Poly’s downstream integration strategy is ahead of its peers, in our view, due partly to 1) China being the key base for solar wafer/cell/module capacity; and 2) China’s potential to become the largest source of end demand

Source: Goldman Sachs Research, Gao Hua Securities Research.

However, GCL’s weaker financial position is a critical disadvantage vs. its peers. GCL has the highest % of revenue generated from the solar business, so its profitability is more sensitive to poly price movements. In addition, GCL has the highest leverage among its peers (Exhibit 40), raising concerns about its financial flexibility during an extended down-cycle.

Exhibit 40: Due to its higher % of solar business, GCL’s earnings look more vulnerable vs. other global peers Poly makers’ ASP sensitivity analysis


Polysilicon Wafer Cell Module

GCL-Poly

EPC &

Project

developer

GCL-Poly

OCI OCI

Wacker

Hemlock

2012E China capacity as % of global capacity

Solar value chain

Low cost Asian

solar upstream

companies

Low cost global

solar upstream

companies

27% 75% 60% 64% 21%

2012E China demand

as % of global demand

Gross profit

(overall)

Gross profit

(solar)EPS ROE CROCI

GCL 80% 145% +/- 4.1% +/- 5.3% +/-16.2% +/- 67 bps +/- 33 bps

OCI 45% 19% +/- 2.1% +/- 4.4% +/- 11.0% +/- 54 bps +/- 24 bps

Wacker 29% 16% +/-1.7% +/- 2.9% +/- 7.6% +/- 32 bps +/- 15 bps

Company

Changes for up/down US$1/kg polysilicon ASPSolar revenue as

% of total (2011)

Net debt / equity

(2Q12)



GS SUSTAIN: An overview of our ESG scorecard While an ability to generate strong returns on capital and position operating assets to sustain these are long-standing traits of industry leaders, a shift in society’s expectations of corporations now requires true industry leaders to demonstrate engagement in mitigating environmental, social and governance risks. As China transitions from being the workshop of the world into a more mature domestic demand story, we are already seeing increased focus on environmental protection, product quality, labor relations and governance manifest itself in rising costs and in some instances value destruction when companies mismanage these risks.

Our ESG scorecard uses publically disclosed data and is scored objectively. For more details, see Disclosing China’s potential --- from

good to great, November 5, 2010. Exhibit 41 summarizes the scorecard for the global solar upstream and midstream segments. A high score signals management is disclosing the existence of protocols and performance consistent with mitigating ESG risks. A low score will demonstrate either weak engagement or weak disclosure of engagement. We believe that if a company is not articulating the existence of policies, targets for improvement and measurement of these targets that it may not be engaged in managing these risks. (Appendix IV).

Currently, the level of disclosure of how Chinese corporations are engaging in mitigating these risks remains considerably below that of Western companies, preventing conclusive analysis. That said, we believe appraising the current performance allows us to identify early adopters and provide visibility on areas companies could improve performance and those simply impacted due to low disclosure.

As discussed above, limited disclosure of management engagement with ESG factors results in overall low scores across our coverage. Our scoring is 1 (weak) to 5 (good) with conditional formatting showing full disclosure (white), no disclosure (grey) or partial disclosure (blue). The grey “1’s” indicate the number of tests negatively impacted by no disclosure with all companies disclosing no data on 30%-60% of data analyzed.

Exhibit 41: GS SUSTAIN analysis of global solar coverage


Company

Ove

rall

ES

G(a

s %

of m

axim

um)

Inde

pend

ent b

oard

le

ader

ship

Inde

pend

ent B

oard

di

rect

ors

& co

mm

ittee

s

Inde

pend

ent a

udito

rs

CEO

com

pens

atio

n

Sha

re-b

ased

com

pens

atio

n

Min

ority

sha

reho

lder

s' ri

ghts

Gov

erna

nce

Rep

ortin

g fo

r sus

tain

abilit

y

Lead

ersh

ip fo

r sus

tain

abilit

y in

itiat

ives

Soci

al -

Lead

ersh

ip

Em

ploy

ee c

ompe

nsat

ion

Em

ploy

ee p

rodu

ctiv

ity

Empl

oyee

trai

ning

Fata

litie

s

Gen

der d

iver

sity

Hea

lth a

nd s

afet

y m

anag

emen

t

LTI

Soci

al -

Em

ploy

ees

Busi

ness

eth

ics

and

corr

uptio

n

Com

mun

ity in

vest

men

t

R&

D in

vest

men

t

Sup

ply

chai

n m

anag

emen

t

Soc

ial -

Sta

keho

lder

s

Ene

rgy

cons

umpt

ion

/ GC

I

Envi

ronm

enta

l pol

icy

&

targ

ets

GH

G e

mis

sion

s / G

CI

Was

te g

ener

atio

n / G

CI

Wat

er c

onsu

mpt

ion

/ GC

I

Envi

ronm

ent

GCL-Poly Energy Holdings 40% 1 2 3 3 3 2 14 2 1 3 3 5 1 1 3 3 1 17 2 1 2 2 7 1 3 1 1 1 7OCI Company 56% 4 2 1 1 3 2 13 3 2 5 3 5 5 5 1 4 1 24 3 3 2 2 10 5 4 2 2 2 15MEMC Electronic Materials, Inc. 44% 4 5 4 5 3 3 24 2 2 4 1 4 3 1 2 4 5 20 5 1 3 3 12 2 4 2 2 2 12Wacker Chemie AG 64% 4 1 3 5 1 1 15 4 3 7 5 5 5 5 1 5 3 29 5 3 3 3 14 2 4 1 3 2 12Yingli Green Energy 32% 1 3 1 1 3 2 11 2 1 3 1 3 3 1 1 1 1 11 2 1 2 1 6 1 3 1 1 1 7Trina Solar 43% 1 5 1 1 5 3 16 2 1 3 2 3 3 1 2 4 1 16 2 1 2 2 7 1 5 1 1 1 9JA Solar Holdings 40% 4 1 1 1 3 2 12 2 1 3 1 3 3 1 2 1 1 12 2 1 3 3 9 1 1 1 1 1 5First Solar, Inc. 51% 4 5 4 2 3 2 20 1 1 2 1 5 1 5 2 4 5 23 3 1 3 2 9 1 3 1 1 1 7SunPower Corp. 38% 1 1 1 1 3 1 8 2 1 3 4 4 3 5 4 4 1 25 2 1 4 2 9 1 4 3 4 1 13



Phase II: R&D/financial strength essential for midstream; China’s CRT-TV transformation C-Si module has dominated end-market demand for years, mainly due to its cost leadership over other technologies. The collapse of poly prices in late 2008 helped the raw materials-based cost structure of c-Si module (Exhibit 42) and made the thin-film module cost structure less competitive. The cost component of thin-film technology accounts for less than 10% of total module cost. Therefore, thin-film makers only have two ways to reduce costs: economic scale and conversion efficiency. These two approaches work at a much slower pace than raw material price erosion, giving the cost structure advantage to c-Si module.

Exhibit 42: Solar midstream has excessive oversupply, while c-Si PV module is the dominant technology due to cost leadership

Source: Company data, Goldman Sachs Research, Gao Hua Securities Research estimates.

China

Solarfun Solarfun Solarfun Hanwha Hanwha Hanwha Hanwha Hanwha Hanwha

Suntech Suntech Suntech Suntech Suntech Suntech Suntech

Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar Canadian Solar

Trina Trina Trina Trina Trina Trina

JA Solar JA Solar JA Solar JA Solar JA Solar JA Solar

China Sunergy China Sunergy China Sunergy China Sunergy China Sunergy China Sunergy

Yingli Yingli Yingli Yingli Yingli Yingli

Jinko Jinko Jinko Jinko Jinko Jinko

Estimated number of other players in China 10 20 100 400 300 300 500 400 100

Europe

Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld Solarworld

Ersol Ersol Ersol Ersol Ersol Ersol Ersol Ersol Ersol Bosch Solar Bosch Solar Bosch Solar

Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells Q-Cells

Solaria Energia Solaria Energia Solaria Energia Solaria Energia Solaria Energia Solaria Energia

North America

Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen Evergreen

Arise Arise Arise Arise Arise Arise Arise Arise Arise Arise

Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar Daystar

Taiwan

Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech Motech

Gintech Gintech Gintech Gintech Gintech Gintech

Estimated number of other players globally 100 20

US

First Solar First Solar First Solar First Solar First Solar First Solar First Solar (CdTe)

Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar Nanosolar

MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé MiaSolé

Solyndra Solyndra Solyndra Solyndra Solyndra Solyndra Solyndra

Global Solar Global Solar Global Solar Global Solar Global Solar Global Solar

Europe

Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth Wurth

Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture Soltecture

Solibro Solibro Solibro Solibro

Avancis Avancis Avancis Avancis Avancis Avancis Avancis

Japan

Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier Solar Frontier

Estimated number of other players globally 200 80

C-Si based cell/module

Thin film (CIGS)

2.12

2.43

2.95 2.953.12

1.58

1.251.15

0.770.62 0.55

2.94

1.591.40

1.231.08

0.870.77 0.74 0.71 0.64 0.59

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

0

5

10

15

20

25

30

35

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012E 2013E 2014E 2015E

Global solar installation (thin-film) Global solar installation (c-Si) Module cost (C-Si) Module cost (thin-film)(GW)$/W



C-Si downside: cost leadership drives overcapacity The cost leadership of c-Si technology is another driver of the excessive oversupply situation and lack of consolidation activity due to: 1) relatively lower fixed cost creating lower entry barriers; 2) scale benefits appearing to be marginal as total capacity approaches 1GW (Exhibit 44); and 3) a large supply chain of commoditized manufacturing equipment for c-Si module has allowed equipment costs to come down dramatically, which lowers the replacement cost for existing capacity.

Exhibit 43: Solar supply chain is mainly on a raw materials-based cost structure Module cost breakdown (1Q12 vs. 4Q12E)

Exhibit 44: Impact of c-Si module economic scale appears to be marginal Yingli and Trina’s unit non-silicon module cost vs. module capacity change (2006-2011)

Source: Company data, Gao Hua Securities Research estimates. Source: Company data, Gao Hua Securities Research estimates.

Exhibit 45: Cost structure differences among tier-1/2 Chinese players are insignificant Solar module makers’ cost curve


50%46%

60%

54%

70% 67%

45% 36%

0.00

0.05

0.10

0.15

0.20

0.25

0.30

1Q12 4Q12E 1Q12 4Q12E 1Q12 4Q12E 1Q12 4Q12E

Electricity Raw material Depreciation OthersUS$/W


R² = 0.8772

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

0 100 200 300 400 500 600 700 800 900

Module capacity (MW)

Un

it n

on

-sil

ico

n m

od

ule

co

st (U

S$/W

)

Module capacity (MW)

Un

it n

on

-sil

ico

n m

od

ule

co

st (U

S$/W

)

Yingli Trina

US$ 0.07/W

US$ 0.19/W

US$ 0.10/W

Unit non-silicon module

cost doesn’t decline

proportionally according

to magnitude of capacity

expansion

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

4Q1

2E m

od

ule

co

st (

US

$ /

Wat

t)

2012E module capacity (GW)

Rest of Asia US EU




GS C-Si module

demand forecast

2013E

GS module ASP

forecast 4Q2012

Other module

manufacturers in China

GS C-Si module

demand forecast

2014E

GS C-Si module

demand forecast

2012E



C-Si upside: cost structure more favorable Cost leadership is a double-edged sword for c-Si technology. Its unique characteristics – raw materials-based with lower capex cost structure, therefore lower fixed cost per watt, or low elasticity between utilization rate and per unit cost – makes it a more favored technology in a highly cyclical industry (Exhibits 46/47).

Where depreciation of c-Si capital equipment amounts to 15% of module costs, it can account for 25% of a fully utilized thin-film manufacturer’s costs (Exhibit 46). A large supply chain of commoditized manufacturing equipment for c-Si has allowed equipment costs to come down dramatically. With current global annual production capacity of c-Si PV modules above 50GW, utilization rates have fallen dramatically. However, the actual effect of low utilization on costs is limited by the relatively low capex for c-Si manufacturing (unlike thin film manufacturing, where capex can easily approach and even exceed US$1 per watt). While demand markets are still sorting themselves out, expansion of scale to reduce costs may only increase cash burn while crystalline competitors continue to push product at ASP close to or even below cash cost per watt.

Exhibit 46: Low utilization has relatively less incremental impact on module cost for c-Si module compared to thin film due to its lower capex Incremental module cost vs. capex under different utilization rate

Exhibit 47: c-Si module is not a leading technology in terms of conversion efficiency Efficiency roadmap for c-Si cell/module vs. thin film

Note: CPV=Concentrated photovoltaic.

Source: Company data, Gao Hua Securities Research estimates. Source: Company data, Gao Hua Securities Research.

Transformational era of China CRT-TV holds similarities to solar midstream’s current excessive oversupply situation Prior to 1978, development in China’s electronics industry was sluggish to non-existent due to internal and external issues as well as problems on both the demand and supply sides. From the 1980s, the government played an important role in developing the TV industry via two approaches:

1. Accelerating the process of integration with global TV development to obtain up-to-date technology.

2. Developing its own technology using the country’s existing resources.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

100% Utilization rate 75% Utilization rate 50% Utilization rate

Capex (US$/W)

Incre

men

tal m

od

ule

co

st(U

S$/W

)

c-Si module

Thin film

CPV*

2005 2010 2012Highest lab

efficiency

Concentrated photovoltaic

CPV 30.0% 35.0% 38.0% 41.0%

c-Si based cell/module

Mono-crystalline 16.7% 17.8% 19.5% 25.0%

Multi-crystalline 14.9% 15.9% 17.5% 20.4%

Thin film

CdTe 8.7% 11.6% 12.6% 17.3%

CIGS 10.0% 12.0% 13.3% 20.3%

μc-Si/a-Si 6.7% 8.5% 10.2% 12.5%

Conversion efficiency

Technology



Various policy measures were implemented to these ends, including financial incentives & assistance packages, tax deductions, tariff deductions or waiver on imports of locally needed technologies, export tax rebates and tax concession on profits. These measures significantly improved China’s TV production capacity and the international competitiveness of its TV industry. After a few years, the majority of components and materials for TV sets were produced in China, thereby considerably reducing the cost of sets as well as reducing China’s imports of TV components. Hence, by the end of the 1980s, China had 200-300 CRT-TV makers, taking roughly 80% of domestic market share. In 1989, however, the industry entered a decade-long pricing war (Exhibit 48), and by 2001 industry profitability was in negative territory (Exhibit 51).

Globally, Chinese TV makers’ competitiveness was hurt by its lack of a technology edge. With the dawn of the flat-TV era in the early 2000s, consumer preferences changed and excess capacity was forced to exit the supply chain rather than be consolidated. Over the next decade, Chinese TV makers lost market share to technology/scale leaders in Korea, Taiwan and Japan.

Exhibit 48: After a decade-long pricing war, Chinese CRT-TV makers lost market share to technology & scale leaders in Japan/Korea/Taiwan Evolutionary roadmap during Chinese CRT-TV era of development & consolidation


Geographical migration of capacity in TV supply chain demonstrates the importance of technology leadership, followed by

scale leadership. Technology development provides the strongest competitive motivator. In the CRT-TV study, we found that domestic TV companies were driven to catch up with overseas competitors by strengthening their R&D abilities to eventually achieve technological superiority. Meanwhile, foreign TV giants had the advantage of being ahead in panel technology as well as economic scale leadership, giving

Listed companies

Changhong Changhong Changhong Changhong ChanghongChanghongChanghong Changhong Changhong ChanghongChanghongChanghong Changhong Changhong Changhong Changhong Changhong Changhong Changhong Changhong ChanghongChanghong Changhong

Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense Hisense

TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL TCL

Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka Konka

Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima Prima

Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth Skyworth

Non-listed companies

Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing Jin Xing

Fei Yue Fei Yue Fei Yue Fei Yue Fei Yue Fei Yue Fei Yue

Panda Panda Panda Panda Panda Panda Panda Panda Panda Panda Panda Panda

Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri Fu Ri

GLH GLH GLH GLH GLH GLH GLH GLH GLH GLH GLH GLHLe Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua Le Hua

Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan Mu Dan

Number of other non-listed companies

220 240 200 170 150 120 91 85 85 82 80 80 77 68 65 60 55 40 35 25 20 18 15

0

20

40

60

80

100

120

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

China TV set annual production (mn units)

The decade of Chinese CRT-TV pricing war / industry rationalization

1st pricing war

LCD-TVCRT to LCD TV

2nd pricing war

3rd pricing war

4th pricing war

5th pricing war

6th pricing war

Chinese TV makers lost market

share to technology/scale

industry leaders in

Japan/Korea/Taiwan

Adopting up-

to-date

technology

from abroad

Supported by

government

subsidy/

buyout

programs

Building up a

self-sufficient

domestic CRT-

TV

component

value chain

Chinese TV

makers

lagged on

new

technology

development

70% of

domestic

players

eventually

were forced

to exit supply

chain

Chinese TV

makers’ cost

structures

were higher

than global

peers due to

lack of

economic

scale



them the initiative in price competition. Combined with the overseas supply chain advantage for panel technology, previously the edge of China CRT-TV makers, China’s TV industry was not able to defend against market share loss in a highly commoditized industry (Exhibit 50). Exhibit 49: Geographical capacity migration of global TV industry driven by technology innovation

Source: Gao Hua Securities Research estimates.

Exhibit 50: Chinese TV makers lost market share to global peers China TV market share breakdown by foreign and domestic makers (1980-2011)

Exhibit 51: Chinese TV makers’ profitability deteriorated from early 90s China TV manufacturing industry operating margins (1992-2011)

Note: Operating margins of Hisense, TCL, Changhong, Konka and Skyworth used to calculate China TV industry average.

Source: Company data, Gao Hua Securities Research. Source: Company data, Gao Hua Securities Research.

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015E 2020E 2025E

CRT-TV (Black &

White)

CRT-TV (Color)

LCD-TV

(CCFL, LED)

US TV makers took the

largest market share

Technology edge the primary driver of global TV capacity migration

OLED-TV

Japanese TV makers entered

TV industry in the 60s and

quickly gained market share

from US/European TV makers

via technology & better

Chinese TV makers adopted up-

to-date technology by importing

CRT-TV production lines from

abroad. Under government

subsidy programs, China

established its own design,

component supply chain and final

goods assembling production

bases nationwide

Japanese/Korean/Taiwanese TV makers

regained market share from Chinese

companies via technology innovation &

better economic scale

5%

20%

73%80%

21%

95%

80%

28%20%

79%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1980 1989 1996 2000 ..... ..... 2011

Foreign TV makers' market share in China

Domestic TV makers' (supplied by domestic component maker) market share in China

Due to lack of technology

and production scale of

LCD/LED panels,

Chinese TV makers lost market share

11%

16%

14%14%

13% 13%

10%

6%

4%

-3%

1%

2%2%

0% 0%

1%2% 2% 2%

3%

-5%

0%

5%

10%

15%

20%

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

China TV manufacturing industry operating margin

Decade-longCRT-TV pricing

war in China

(1989 -2000)

Margins recovered due to industry consolidation and

better product mix, offset by

higher cost base due to LCD/LED panels imports



Lessons for Chinese solar midstream We can’t say with certainty that the solar midstream experience will mirror that of CRT-TV. We fully expect to see more newcomers to the solar supply chain as new commercial technology emerges. However, just as the future impact of LCD/LED TV wasn’t anticipated in the 1980s, we cannot identify a migration trigger for the solar supply chain at this stage.

Assuming c-Si PV module remains the leading technology, Chinese solar manufacturers are likely to remain technology leaders. Chinese companies’ mastery of multi-crystalline ingots/wafers production techniques – such as nucleation control and furnace optimization – has been evident for several years already. R&D investment has escalated in recent years, with most leading Chinese companies having large research teams (50+ persons), including scientists from abroad (e.g., National Renewable Energy Laboratory at New South Wales, Australia). Given Chinese c-Si PV makers’ dominant positioning, US/Japan/EU companies have adopted and developed mainly thin-film technologies in the past (Exhibit 52).

Exhibit 52: Roadmap of solar technologies development. Chinese companies are still leading in c-Si PV technology

Note: “D” = Development; “T” = Testing; “P” = Production. Source: Company data, Gao Hua Securities Research estimates.

Improving Cutting raw

Technology 1990s 2000 2005 2010 2012 2013E 2015E efficiency material cost

Polysilicon

Siemens process D T

Modified Siemens process

(Hydrochlorination)D T √

Silane method D T √

Fluidized bed method D T √

Siemens-FBR combined method D P √

Wafer

Multi-crystalline D T

Mono-crystalline D T √

Quasi-mono D T √

High efficiency multi D T √

N-type mono D T √

Thinner wafer D T √

Direct wafer D P √

Nucleation control/furnace

optimization D T √

New cutting edge technology D P √

Cell/module (C-si)

Copper paste D P √

Heterojunction D P √

Selective Emitter D T √

Metal Wrap Through D T √

Back-Surface Field D √

Cell/module (Thin film)

CIGS D T

CdTe D T

Thin film

c-Si

Development roadmap

1950s-1980s

P

P

P

P

P

P

P

T

T

T

T

P

P

T

P

P

P

P

T P

P



Solar midstream survival checklist: balance sheet strength, continuous R&D Given the current negative cash flow across the industry, cash burn rates differ according to companies’ financial strength and working capital management; stronger balance sheets should enable greater flexibility during a period of consolidation. Although the technology gap between leaders and laggards is still not significant among Chinese listed solar manufacturers (Exhibit 54), we believe persistence in pursuing R&D will be a key differentiator between market share gainers and losers in the future (Exhibit 53).

Exhibit 53: Balance sheet reflects potential to remain competitive: JASO & TSL appear to have stronger balance sheets Offshore listed Chinese solar makers’ market cap vs. net debt/equity ratio (1Q12)


Exhibit 54: Although the gap between technology leaders and laggards is not significant, R&D persistence is critical Solar players’ high efficiency products’ profile

Note: (1) MWT: Metallization Wrap-Through; (2) PERL: Passivity Emitter and Rear Locally diffuse. Source: Company data, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Further cost reductions are possible from continued R&D for c-Si technology. Chinese equipment companies and silicon production companies started to work together in 2009. It is well-known that many companies in China and Taiwan can produce high-efficiency crystalline silicon wafers using casting, modified VGF, or modified Bridgman furnaces. Since late 2010, rapid advancements in quasi-single

24 33

80 87

128 135

167 173

202

307

391

234%

50%

147%

57%

122% 102%

38%

221%

349%

150%

45%

0%

50%

100%

150%

200%

250%

300%

350%

400%

0

50

100

150

200

250

300

350

400

450

CSUN DQ JKS HSOL CSIQ SOL JASO STP LDK YGE TSL

Maket cap (LHS) Net debt/equity (RHS)(US$ mn)

2011 2012E

Wafer

ReneSola Quasi-mono (Virtus I, II) 18.0% 18.5%

Yingli N-type mono (PANDA I, II) 19.0% 20.5%

LDK High efficiency Multi (M2) 16.3% 17.3%

Cell/Module

17.6%(Multi) 18.5%(Multi)

18.8%(Mono) 19.5%(Mono)

China Sunergy Selective Emitter (QSAR) 18.4% (Mono) 19.1% (Mono)

Trina (Honey) N.A. 17.4% (Multi)

PERL* (Pluto)

Selective Emitter

Cast-mono

17.5%(Multi) 18.5%(Multi)

19.0%(Mono) 19.5%(Mono)

Heterojunction N.A. 20.0% (Lab)

Jinko Selective Emitter (WING) 18.6% (Mono) 19.0% (Mono)

Company TechnologyProducts' efficiency

JA Solar MWT* (Cypress)

CanadianSolarMWT

19.0% (Mono) 19.5% (Mono)Suntech



crystalline technology have been adopted by many offshore listed Chinese solar companies. In 2011/12, more specially designed furnaces are being developed by production companies in China, offering dramatic enhancements to areas like energy efficiency, throughput, and quality. More importantly, furnace manufacturing, crucible technology, silicon production, wafer and cell processing are being integrated together.

Based on recent news flow and industry surveys, we see possible areas for further cost reduction (Exhibit 55). In the future, crucible and wafer slicing using direct wafer technology may no longer be needed – the development of new technology to industrial scale production is three to five years away. In the next two to three years, we expect production of high-efficiency solar cells with different technologies and at lower raw material cost will be one of the key areas of competition for the industry. Different technologies are available, but financial stability is critical for a company’s ability to pursue an edge here, particularly during periods of turbulence.

Exhibit 55: Further cost reduction through the solar value chain by technology developments


Earnings across the value chain remain under pressure as we await the beginning of Phase I of rationalization and consolidation. We therefore do not expect poly ASP to trough until late 2012/early 2013, while module ASP should remain under pressure in 2013E (barring the start of Phase I). Hence, we lower our module ASP forecasts further to US60¢/US57¢ per watt in 4Q12 and 2013, from US70¢/US67¢ per watt (Exhibits 56/58). We introduce our 2014 module ASP forecast of US50¢/W. Given that ASP has dropped close to most Chinese midstream players’ cash cost (Exhibit 56), and considering the current industry-wide 5-10% S&A (including R&D cost as % of total revenue, we expect most midstream solar manufacturers to struggle to break even at the EBITDA level (Exhibit 57).


Improving efficiency:

(1) High efficiency multi

& Quasi-mono: increasing

cell conversion efficiency.

Cutting cost:

(2) Direct wafer: cutting

step of ingot.

(3) Reducing cutting cost,

20% potential.

Cutting cost:

Copper paste:

reducing silver usage,

30% cost reduction

potential.

Improving efficiency:

(1) Metal Wrap Through /

Heterojunction / Selective

Emitter: increasing cell

conversion efficiency.

Cutting cost:

(2) New back sheet

technology: 20% cost

reduction potential.

Cutting cost:

(1) Hydrochlorination:

50% electricity cost

reduction potential.

(2) Fluidized Bed

Reactor combined with

Siemens method: 30%

cost reduction potential.

Solar value chain

cost breakdown

Technology

development

Electricity,

45%

Raw

material,

23%

Others,

32%

Electricity,

25%

Raw

material,

50%

Others,

32%

Electricity,

10%

Raw

material,

60%

Others,

32%

Electricity,

10%

Raw

material,

70%

Others,

32%



Exhibit 56: Module ASP to remain under pressure in 2012-2014E Solar module ASP trend (1Q09-2014E)

Exhibit 57: .. and solar midstream companies’ EBIT remain under pressure Manufacturers’ all in-house module production cost

Source: PV news, Goldman Sachs estimates, Gao Hua Securities Research estimates. Source: PV news, Goldman Sachs estimates, Gao Hua Securities Research estimates.

Exhibit 58: We expect ASPs to remain under pressure across the supply chain as long as solar investment cycle in China continues Solar ASP estimates across supply chain

Source: PV news, Goldman Sachs estimates, Gao Hua Securities Research estimates.

1.65

1.73

1.82

1.71

1.54

1.26

0.89

0.81

0.73

0.640.60

0.570.500.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

2Q

10

3Q

10

4Q

10

1Q

11

2Q

11

3Q

11

4Q

11

1Q

12

2Q

12

3Q

12

E

4Q

12

E

20

13E

20

14E

US$/W

0.770.70

0.67

GS solar module ASP forecast (previous)

GS solar module ASP forecast (current)

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

2.20

US$/W

Operating cost/W

Cash cost/W

COGS/W

Module ASP

Manufacturers'

all in-house

module production cost

2011A 2012E

1Q11A 2Q11A 3Q11A 4Q11A 1Q12A 2Q12A 3Q12E 4Q12E

Polysilicon (US$/kg) 66 74 69 55 33 58 27 23 20 19 22 18 15

% change yoy -19% 35% 25% -21% -61% -13% -64% -67% -64% -43% -62% -19% -17%

Wafer (US$/W) 0.83 0.87 0.69 0.54 0.34 0.61 0.31 0.26 0.25 0.24 0.26 0.22 0.20

% change yoy -17% 13% -16% -36% -61% -26% -65% -63% -54% -29% -57% -16% -10%

Cell (US$/W) 1.32 1.21 0.95 0.70 0.53 0.85 0.50 0.44 0.40 0.38 0.43 0.36 0.32

% change yoy -17% -3% -27% -48% -62% -36% -59% -53% -43% -29% -49% -17% -10%

Module (US$/W) 1.74 1.71 1.54 1.26 0.89 1.35 0.81 0.73 0.64 0.60 0.70 0.57 0.50

% change yoy -17% -2% -7% -27% -51% -22% -52% -53% -49% -33% -48% -18% -12%

Thin-film module (US$/W) 1.58 1.53 1.29 1.10 0.85 1.26 0.75 0.59 0.57 0.55 0.59 0.52 0.47

% change yoy -14% -14% -19% -28% -42% -20% -51% -55% -49% -36% -53% -13% -9%

2014E 2010A 2011A 2012E 2013E



GS SUSTAIN: Identifying potential winners in solar midstream

In line with our upstream segment analysis, we apply the GS SUSTAIN framework to indentify potential winners in our midstream coverage.

We are yet to see signs of structural rationalization, while lower replacement costs are making the kick-starting of consolidation even more difficult. We thus consider financial strength the most important metric, indicating an ability to facilitate corporate activity and to

pursue R&D amid industry turbulence. We outline five measures of industry positioning (Exhibit 59), assess the winning characteristics of cell/module makers under GH/GS coverage (Trina, JA Solar, Yingli, First Solar, Sunpower) in the winning table (Exhibit 61), and present their relative competitiveness in the winners’ circle (Exhibit 60). For an overview of ESG scorecard, please refer to page 28 and Exhibit 41.

Exhibit 59: Structural changes in the solar industry and analysis of drivers of cash returns and growth of the solar midstream segment


Drivers of corporate performance

Solar midstream

Return on capitalIndustry positioningManagement

quality GS SUSTAIN

CROCI

Cash return on cash invested

Distribution channel

Access to growth

ESG

Environmental, social and

governance issues

ThemesC-Si module non-silicon costs

have fallen 55% since 1Q2010. with limited but possible further

cost reduction. Hence,persistent R&D development in the next generation of lower cost based technology is the primary

long-term winner solution.

Stable regulation, pricing and funding

Successful project financing of industrial power generation capacity

requires stable and smaller scale projects, in turn requiring regulatory

stability, transparent pricing and access to funding.

Over-capacity

High FiTs, subsidies and favorable access to capital have

resulted in overinvestment in capacity since 2009/10.

Meaningful recovery in returns will require capacity reduction /

structural acceleration in demand grow.

A global footprint and

extensive distribution

and installer network

Innovation

Technology edge

Financial strength Cost position

Investing at down-cycles

for further R&D

development; hence,

leading the cost curve

Entering a potentially protracted phase of

restructuring players with strong B/S’s are better

placed

Creating product

differentiation in a

commodity industry

Obtaining unique

access to local

downstream project

developers/IPPs

Investment in smart grids

Solar is an intermittent source of electricity supply. Full scale uptake will require a major upgrade to the power grid. Infrastructure changes take time to implement, and require

significant investments.



Exhibit 60: Winners’ circle – five global cell/module makers


Exhibit 61: Winning table – global cell/module makers

Source: Goldman Sachs Research, Gao Hua Securities Research estimates.

Management of ESG issues – ESG (based on 2011 data) above sector median

Industry positioning: leaders on a combination of:– High-potential growth markets– Attractive market structure– Favourable business mix and regulatory outlook- Operating margins

Returns – CROCI (2012-14E) above sector median

Management quality Industry positioning

Return on capital

First Solar

Yingli

SunPowerJA Solar

Trina

Overall Industry

PositioningESG

2012-14E Percentile

Extensive

distibution

network

PercentileTechnology

leadershipPercentile

Leading in

the cost curvePercentile

Net

debt/equity

(2013E)

Percentile

EBITDA/

interest exp.

(2013E)

Percentile

Cash deficit

(US$ mn,

2013E)

PercentileProduction

differentiationPercentile

Access to

downstream

project

Percentile Percentile

Score as % of

maximum

(2010/11)

Percentile

(2010/11)

Yingli 2% 0% 5 50% 4 25% 3 0% 280% 0% 0.5 0% 523 0% 5 100% 4 25% 50% 32% 0%

Trina 3% 25% 5 50% 3 0% 4 50% 69% 25% 1.2 50% 305 50% 3 0% 3 0% 0% 43% 75%

JA Solar 5% 75% 4 0% 4 25% 5 100% 34% 75% 0.9 25% -71 75% 3 0% 4 25% 25% 40% 50%

First Solar 25% 100% 4 0% 4 25% 4 50% -33% 100% 213.6 100% -244 100% 4 50% 5 75% 75% 51% 100%

SunPower 4% 50% 5 50% 5 100% 3 0% 67% 50% 5.0 75% 355 25% 4 50% 5 75% 100% 38% 25%

Company

Return on capital Management quality

CROCI Financial strength InnovationCost positionDistribution channel Access to growth

Industry positioning



Equipment: earnings growth for inverters, extended down-cycle for furnaces The extended down-cycle makes prospects for the recovery of wafer furnace makers (JG and JYT) more challenging in 2012E. We thus

lower our shipment forecasts and expect ASP to decline faster than unit cost for JG and JYT (Exhibit 64). We might see some signs

of recovery in 2013, due to upgrades and enhancement orders, led by better technology providers like JG with better customer

profiles (see 2012 outlook: Focusing on cash/structural winners; initiating China solar equipment makers, December 19, 2011).

Among our equipment coverage, we expect to see earnings growth only at inverter makers due to volume upside potential from

secular growth in the domestic market. This will likely be offset by softer gross margins due to fierce competition.

Sungrow’s strong balance sheet and leading inverter production scale could provide a competitive advantage. With 14 years of experience

in the inverter business, Sungrow listed on the A-share market at end-2011. While mainstream buyers of solar inverters in China are

the big five IPPs and SOEs, extended cash payment terms seem to be a common business practice. With Rmb1.3 bn in funds raised

from its IPO, Sungrow would appear to have relatively strong financial resources vs. newcomers in the domestic inverter market.

Furthermore, the most compelling domestic competitors (big power grid equipment SOEs, such as Xuji and Nari, which might have

better access to downstream project developers) have made no clear indication to allocate adequate resources to this niche market.

We thus maintain our view that Sungrow will hold its 30%+ share of the China market in 2012/14E.

Exhibit 62: Extended down-cycle is lowering new capacity of wafer China wafer capacity addition yoy (2007-2013E)

Exhibit 63: We expect Sungrow’s ASP to decline faster than global peer SMA Solar Technology’s Inverter ASP forecasts (1Q09-2013E)

Source: Goldman Sachs Research estimates, Gao Hua Securities Research estimates. Source: Company data, Goldman Sachs estimates, Gao Hua Securities Research estimates.

1,216

2,715

4,180

9,090

14,340

1,120 800 2,100

199%

149%

92%104%

81%

3% 2%6%

0%

50%

100%

150%

200%

250%

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

2007 2008 2009 2010 2011 2012E 2013E 2014E

China wafer annual new additional capacity (LHS)

Growth of China wafer annual new additional capacity (RHS)(MW) 4.70

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

SMA's consensus ASP forecast:

-23%/-6% in 12/13E

GH Sungrow ASP forecast (new):

yoy growth -26%/-12% in 12/13E

SMA's ASP

0.75

0.56

GH Sungrow ASP

forecast (old)

(Rmb/W)



Changes to earnings metrics, estimates, target prices and valuations Solar product manufacturers Module ASP: We have lowered our 2012-14 ASP forecasts by 7%-19% and lowered our 2012-14 unit cost forecasts up to 11% to reflect

our expectations of persistent overcapacity and an extended midstream down-cycle. Wafer ASP: We have lowered GCL’s 2012-14 wafer ASP forecasts to US$0.20-0.26/W. Shipments: We have slightly revised our shipment expectations by -4% to -13% per updated company guidance.

Solar equipment manufacturers Ingot furnace makers (JG/JYT): 1) We have lowered our 2012-14 ASP forecasts by 11%-43% and lowered our 2012-14 unit cost

forecasts by 3%-26% to reflect more intense competition; 2) we have lowered our 2012-14 shipment forecasts by 22%-53% to reflect our view of an extended down-cycle, and slower order intake recovery in 2012-14. Inverter makers (Sungrow): 1) We have lowered our 2012-14 ASP forecasts by 22%-38% and lowered our 2012-14 unit cost forecasts by 27%-37% to reflect fierce competition from domestic inverter companies; 2) we have revised our shipment expectations by +3% to -12% per updated company guidance and demand uptick from domestic PV market.

Exhibit 64: Solar makers: ASPs declining faster than costs being reduced; adjusting shipments per company guidance Estimate changes for Chinese solar product/equipment makers’ ASP/unit cost/shipments (2012E-2014E)


2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Wafer

GCL 0.26 0.22 0.20 0.26 0.25 0.25 0% -10% -20%

Module

YGE 0.74 0.57 0.50 0.80 0.67 0.61 -7% -15% -19%

TSL 0.74 0.58 0.51 0.80 0.67 0.62 -8% -13% -18%

JASO 0.68 0.57 0.50 N.A. N.A. N.A. N.A. N.A. N.A.

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Wafer

GCL 0.22 0.19 0.16 0.23 0.20 0.18 -2% -5% -11%

Module

YGE 0.73 0.54 0.46 0.73 0.58 0.52 0% -7% -11%

TSL 0.67 0.54 0.47 0.72 0.58 0.52 -7% -6% -10%

JASO 0.65 0.55 0.48 N.A. N.A. N.A. N.A. N.A. N.A.

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Wafer

GCL 5,842 6,400 6,400 6,364 7,000 7,000 -8% -9% -9%

Module

YGE 2,200 2,310 2,541 2,300 2,650 2,700 -4% -13% -6%

TSL 1,750 2,013 2,314 2,000 2,300 2,645 -13% -13% -13%

JASO 1,734 2,200 2,400 N.A. N.A. N.A. N.A. N.A. N.A.

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Sungrow 0.50 0.36 0.32 0.64 0.56 0.51 -22% -36% -38%

JG 1.81 1.30 1.11 2.03 1.87 1.72 -11% -30% -36%

JYT 1.64 1.09 0.93 2.00 1.80 1.62 -18% -40% -43%

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Sungrow 0.30 0.24 0.22 0.41 0.37 0.34 -27% -37% -37%

JG 1.14 0.93 0.84 1.25 1.19 1.13 -9% -21% -26%

JYT 1.17 0.97 0.88 1.20 1.14 1.08 -3% -15% -19%

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Sungrow 1,667 2,500 3,100 1,900 2,450 3,000 -12% 2% 3%

JG 130 150 200 220 320 370 -41% -53% -46%

JYT 101 100 120 130 150 180 -22% -33% -33%

Solar product maker ASP

(US$/W)

New estimates Old estimates % diff. between new and old estimates

Solar product maker unit

cost (US$/W)


Solar product maker

shipments (MW)


Solar equipment maker

shipments (MW)



unit cost (Rmb/W)



ASP (Rmb/W)


Adjusting down ASPs, faster than cost reduction;

adjusting down shipments to reflect sluggish solar

demand growth

Adjusting down ASPs, faster than cost reduction; adjusting

down furnace makers’ shipments to reflect weaken solar demand; but adjusting up inverter maker’s shipment

to reflect China demand upside



Exhibit 65: We lower target prices for our solar coverage, and resume coverage on JA Solar with Neutral Rating and target price changes for our solar coverage

Source: Datastream, Gao Hua Securities Research estimates.

Exhibit 66: Our earnings estimates are consistently below consensus estimates across our coverage group 2012-14E EPS, GH vs. consensus

Source: Bloomberg, Wind, Gao Hua Securities Research estimates.

Price Potential

New Old Change New Old % Change 7-Sep-12 up/downside 2012E 2013E


GCL-Poly 3800.HK HK$ 2.4 Neutral Neutral 1.10 1.70 -35% 1.19 -8% 0.9x 0.9x

Trina TSL US$ 0.3 Neutral Neutral 4.00 5.50 -27% 4.02 0% 0.3x 0.3x

JA Solar JASO US$ 0.2 Neutral Not Rated N.A. 1.00 N.A. N.A. 0.80 25% 0.2x 0.2x

Yingli YGE US$ 0.3 Sell Sell 0.80 2.00 -60% 1.67 -52% 0.2x 0.2x


Sungrow Power 300274.SZ Rmb 0.5 Buy Buy 11.00 14.80 -26% 9.12 21% 1.8x 1.7x

Jing Gong 002006.SZ Rmb 0.7 Neutral Neutral 8.40 11.00 -24% 9.25 -9% 2.9x 2.6x

Jinyuntong 601908.SS Rmb 0.9 Sell Sell 2.40 5.30 -55% 6.77 -65% 0.6x 0.5x

Implied P/BCompany Ticker Currency

Rating 12-m target priceMarket cap

(US$ bn)

2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E


GCL-Poly HK$ -0.06 0.00 0.04 0.05 0.15 0.20 -223% -98% -81% -0.01 0.06 0.14 -354% -94% -73%

Trina US$ -2.62 -0.71 -0.42 -1.48 0.20 0.48 -77% -447% -187% -2.41 -0.57 0.85 -9% -24% -150%

JA Solar US$ -1.00 -0.36 -0.32 N.A. N.A. N.A. N.A. N.A. N.A. -0.75 -0.31 0.14 -33% -18% -325%

Yingli US$ -1.92 -1.21 -0.95 -1.19 -0.18 0.08 -61% -587% -1327% -0.92 -0.44 0.25 -108% -177% -480%


Sungrow Power Rmb 0.42 0.52 0.67 0.59 0.77 0.85 -30% -33% -20% 0.55 0.66 0.73 -24% -21% -8%

Jing Gong Rmb 0.13 0.40 0.47 0.50 0.60 0.66 -75% -33% -29% 0.50 0.81 1.05 -75% -51% -55%

Jingyuntong Rmb 0.07 0.12 0.14 0.24 0.28 0.22 -70% -56% -36% 0.30 0.55 0.69 -77% -77% -80%

% of diff. with consensusCompany Currency

New EPS estimates Old EPS estimates % of diff. with old estimates Consensus estimates



We use a cash-return (CROCI) based valuation framework as our primary valuation methodology to derive our ratings and 12-month target prices. This method suggests Yingli is overvalued within our China solar coverage, while GCL-Poly, Trina and JA Solar are fairly valued.

Exhibit 67: Solar stocks are trading at historically lowest P/B Exhibit 68: Cash-based methodology suggests YGE is overvalued

Note: Spots with red color means companies under GH-coverage. Source: Company data, Bloomberg, Gao Hua Securities Research estimates. Source: Bloomberg, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Exhibit 69: We are still fairly bearish on this sector across GS/GH global solar coverage

Note: OCI is covered by Seung Shin; European stocks are covered by Benjamin Moore; US stocks by Brian Lee.. Source: Datastream, Goldman Sachs Research estimates, Gao Hua Securities Research estimates.

Consistent with revisions across Goldman Sachs global solar coverage, we have updated our 2012-14 CROCI estimates and implied 2013E EV/GCI multiples for our China coverage. We have adjusted our Director’s Cut-based target prices for Chinese offshore names accordingly.

0.0X

1.0X

2.0X

3.0X

4.0X

5.0X

6.0X

7.0X

Max P/B = 5.9X

Avg. P/B = 1.9X

Min P/B = 0.4X

Offshore listed Chinese solar companies 12-m forward P/B

Wacker

MEMC

SunPower

STR

GCL-Poly

Trina

Yingli

JA Solar

OCI

y = 0.5209x + 0.365

R² = 0.5224

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

2012

-14E

EV

/GC

I (X

)

2012-14E CROCI/WACC (X)

Stock Target Target price PriceMarket

cap

Company Ticker rating price period 7-Sep-2012 (US$bn) 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E 2012E 2013E 2014E

Asia - offshore

GCL-Poly 3800.HK Neutral 1.10HK$ 12-m 1.19 -8% 2.4 (0.06) 0.00 0.04 N.A. N.A. 32 0.9 0.9 0.9 (5) 0 3 11 9 8

OCI 010060.KS Neutral 178,000Won 12-m 191,500 -7% 4.0 6,851 10,850 10,704 28 18 18 1.3 1.3 1.2 5 7 7 7 6 5

Trina TSL Neutral 4.00US$ 12-m 4.02 0% 0.3 (2.62) (0.71) (0.42) N.A. N.A. N.A. 0.3 0.3 0.3 (18) (6) (4) N.A. 10 8

JA Solar JASO Neutral 1.00US$ 12-m 0.80 25% 0.2 (1.00) (0.36) (0.32) N.A. N.A. N.A. 0.2 0.2 0.2 (20) (8) (7) N.A. 12 10

Yingli YGE Sell 0.80US$ 12-m 1.67 -52% 0.3 (1.92) (1.21) (0.95) N.A. N.A. N.A. 0.4 0.5 0.7 (39) (30) (32) N.A. 30 21

Average 28 18 25 0.6 0.6 0.7 (15) (7) (6.6) 9 13 10

Asia - onshore

Sungrow 300274.SZ Buy 11.00Rmb 12-m 9.12 21% 0.5 0.42 0.52 0.67 22 17 14 1.5 1.4 1.2 7 8 10 16 11 8

JG 002006.SZ Neutral 8.40Rmb 12-m 9.25 -9% 0.7 0.13 0.40 0.47 73 23 20 3.2 2.8 2.5 4 13 13 58 31 27

JYT 601908.SS Sell 2.40Rmb 12-m 6.77 -65% 0.9 0.07 0.12 0.14 95 55 49 1.6 1.5 1.5 2 3 3 N.A. N.A. N.A.

Average 63 32 27 2.1 1.9 1.7 4 8 9 37 21 17

Europe

Wacker WCHG.DE Sell 45.00EUR 6-m 53.83 -16% 3.4 2.93 3.49 3.28 26 15 16 1.0 1.0 1.0 4 7 6 6 5 5

Average 26 15 16 1.0 1.0 1.0 4 7 6 6 5 5

US

First Solar FSLR Neutral 15.00US$ 6-m 20.44 -27% 1.8 4.59 4.04 1.76 4 5 12 0.5 0.4 0.4 11 9 4 2 1 N.A.

MEMC WFR Neutral 2.50US$ 6-m 3.19 -22% 0.7 (0.12) 0.25 0.30 N.A. 13 11 1.3 1.2 1.1 (4) 10 11 4 3 2

STR STRI Neutral 3.50US$ 6-m 3.07 14% 0.1 0.05 0.20 0.35 58 15 9 0.5 0.5 0.4 (3) (1) 1 N.A. 3 2

SunPower SPWR Sell 4.00US$ 6-m 4.46 -10% 0.5 (0.15) 0.25 0.50 N.A. 18 9 0.5 0.5 0.5 (1) 3 6 8 6 5

Average 31 13 10 0.7 0.6 0.6 1 5 5 5 3 3

Potential up/downside

P/E (X)EPS (reporting currency) ROE (%) EV / EBITDA (X)P/B (X)



Exhibit 70: We expect to see some recovery signs starting from 2013E CROCI of GS/GH global solar coverage (2009-2014E)


Exhibit 71: Target price calculation: China solar product manufacturers Rating and 12-month target prices for our offshore solar coverage


Solar product manufacturing Segment 2009 2010 2011 2012E 2013E 2014E

China

GCL-Poly Upstream + Downstream 27% 48% 18% 8% 8% 9%

Trina Midstream 35% 47% 12% -2% 6% 7%

JA Solar Midstream 19% 41% 9% 0% 7% 8%

Yingli Midstream 14% 26% 8% -1% 3% 4%

Korea

OCI Upstream + Downstream 9% 15% 26% 9% 10% 9%

Europe

Wacker Upstream 9% 8% 5% 4% 4% 4%

US

First Solar Midstream + Downstream 21% 26% -23% 32% 33% 11%

MEMC Upstream 3% 10% -9% 5% 8% 9%

SunPower Midstream 14% 13% -2% 1% 5% 7%

STR Holdings Midstream 11% 15% 6% 3% 7% 8%

Solar equipment Segment 2009 2010 2011 2012E 2013E 2014E

China

Sungrow Downstream 37% 244% 58% 18% 18% 21%

Jing Gong Midstream 57% 74% 49% 5% 12% 13%

Jingyuntong Midstream 36% 109% 25% -4% -7% -5%

Implied Implied

2012E P/B 2013E P/B

GCL-Poly HK$ Neutral 8.2% 43,880 15,857 1.1 1.2 -8% 0.9x 0.9x

Trina US$ Neutral 2.1% 821 279 4.0 4.0 0% 0.3x 0.3x

JA Solar US$ Neutral 4.3% 430 195 1.0 0.8 25% 0.2x 0.2x

Yingli US$ Sell 1.5% 2,285 124 0.8 1.7 -52% 0.2x 0.2x

Company Currency Rating

2012-14E

average

CROCI

Potential

up/downside

2013E Implied

Market cap

(mn)

2013E

Implied EV

(mn)

Target priceCurrent

price



Our back-testing for A-share solar companies shows that a P/E vs. EPS growth valuation methodology appears to effectively capture risk/reward in a fast-growing but rapidly changing industry (see 2012 outlook: Focusing on cash/structural winners; initiating China solar equipment makers, December 19, 2011).

Our A-share solar universe includes makers of solar equipment, modules and inverters. The average 2013E P/E is 32X, while the average 2013E-2014E EPS growth rate is 33%. JYT looks overvalued at 56X 2013E P/E, while Sungrow looks undervalued at 17X 2013E vs. its 2013E-2014E EPS growth. We derive our 12-month target prices from the regression of 2013E P/E and 2013E-2014E EPS growth.

Exhibit 72: A-share solar companies’ 2011E-2014E P/E vs. EPS growth Exhibit 73: We believe Sungrow is undervalued while JYT is overvalued

Note: NC = Not Covered. Estimates for NC stocks are from Wind.

Note: Spots with red color means companies under GH-coverage. Source: Wind, Gao Hua Securities Research estimates. Source: Wind, Datastream, Gao Hua Securities Research.

Exhibit 74: Our target prices imply 65% downside potential for JYT and 21% upside potential for Sungrow Rating and 12-month target prices for our onshore solar equipment makers coverage


Current

Company Ticker Rating Price 2011 2012E 2013E 2014E 2011 2012E 2013E 2014E

Solar equipment companies

Jing Gong 002006.SZ Neutral 9.25 7X 73X 23X 20X 109% -90% 217% 17%

Jingyuntong 601908.SS Sell 6.77 13X 95X 56X 48X -33% -87% 69% 17%

Jingsheng 300316.SZ NC 26.03 10X 22X 8X 8X 102% -53% 161% 11%

Solar module companies

Risen 300118.SZ NC 5.10 21X 85X 43X 26X -75% -75% 100% 67%

Chaori 002506.SZ NC 5.87 37X 84X 65X 31X -71% -56% 29% 111%

Solar inverter companies

Sungrow 300274.SZ Buy 9.12 9X 22X 17X 14X 17% -57% 26% 28%

Innovance 300124.SZ NC 20.80 24X 39X 31X 26X 26% -40% 26% 20%

INVT 002334.SZ NC 13.13 36X 33X 29X 24X -36% 11% 13% 22%

Rongxin 002123.SZ NC 10.43 17X 21X 14X 14X 13% -17% 44% 4%

Average 19X 53X 32X 23X 6% -52% 76% 33%

EPS growthP/E

Jing Gong

Jingyuntong

Jingsheng

Risen

Chaori

Sungrow

Innovance

INVT

Rongxin

y = 40.837x + 18.495

R² = 0.5259

0X

10X

20X

30X

40X

50X

60X

70X

0% 20% 40% 60% 80% 100% 120%

2013E P/E (X)

2013-2014E EPS CAGR

2013E-2014E 2013E 2013E Target Current Potential

EPS growth Target P/E EPS (Rmb) price (Rmb) price (Rmb) up/downside

Sungrow 300274.SZ Buy 28% 21x 0.52 11.00 9.12 21%

Jing Gong 002006.SZ Neutral 17% 21x 0.40 8.40 9.25 -9%

Jingyuntong 601908.SS Sell 17% 20x 0.12 2.40 6.77 -65%

Company Ticker Rating



JA Solar – Solid balance sheet to weather the turbulence; reinstate with Neutral

What's changed We resume coverage on JA Solar (JASO) with a Neutral rating and a 12-month Director’s Cut-based target price of US$1.00. Year to date, JASO’s share price has contracted by 30%, vs. our GS coverage universe performance of -41%. Historically, JASO has traded at a 30%-40% discount to industry average valuation, due mainly to its cell - focused business model, which makes its earnings more vulnerable than peers to a cyclical downturn. JASO is currently trading at 0.2X 2013E P/B vs. 0.4X – the average 2013E P/B of Yingli and Trina. Hence, we think JASO is fairly valued. Given the low visibility on the ongoing industry down-cycle, we believe a strong balance sheet is the most important indicator to assess a company’s ability to survive an extended period of turbulence. In fact, a strong balance sheet should allow a company to explore M&A activity during rationalization/consolidation, in our view. We consider JASO’s balance sheet to be the strongest among its peers, giving it a competitive advantage in a structural deteriorated industry.

Implications We see the following potential catalysts for JA Solar over the next 12 months: (1) Current trade frictions between the US, EU and China could result in global solar migration, in which Chinese companies could move some capacity outside of China to avoid potential tariff barriers. Companies with greater financial strength, such as JASO (Exhibit 9) should have the flexibility to explore a wider variety of solutions amid the challenging macro environment. (2) JASO has been steadily increasing its degree of vertical integration by adding more module/wafer capacity. Hence, as reported at 2Q, module shipment accounted for more than 55% of shipments, and the company guides for this to reach 70% of total shipments by end-2012. Meanwhile, JASO has been able to maintain its tier-1 non-silicon cost structure. Therefore, we expect its margin outlook to improve in 2013E.

Valuation Our 12-month Director’s Cut-based target price of US$1.00 implies 25% upside potential. The stock is currently trading at a 2013E P/B of 0.2X and a 2013E EV/EBITDA of 12X versus peer averages of 0.6X and 13X.

Key risks Better- or worse-than-expected global solar demand; higher- or lower-than-expected ASPs; higher- or lower-than-expected costs.

Growth

Returns *

Multiple

Volatility Volatility

Multiple

Returns *

Growth

Investment Profile

Low High

Percentile 20th 40th 60th 80th 100th

* Returns = Return on Capital For a complete description of the investment

profile measures please refer to the

disclosure section of this document.

JA Solar Holdings (JASO)

Asia Pacific Alternative Energy Peer Group Average

Key data Current

Price ($) 0.80

12 month price target ($) 1.00

Market cap ($ mn / US$ mn) 156.6 / 156.6

Foreign ownership (%) --

12/11 12/12E 12/13E 12/14E

EPS ($) (0.54) (1.00) (0.36) (0.32)

EPS growth (%) (132.9) (85.9) 63.4 13.5

EPS (diluted) ($) (0.54) (1.00) (0.36) (0.32)

EPS (basic pre-ex) ($) (0.54) (1.00) (0.36) (0.32)

P/E (X) NM NM NM NM

P/B (X) 0.1 0.2 0.2 0.2

EV/EBITDA (X) 97.0 NM 12.4 10.2

Dividend yield (%) 0.0 0.0 0.0 0.0

ROE (%) (8.7) (20.3) (8.3) (7.4)

CROCI (%) 9.3 0.5 7.2 7.6

2,300

2,500

2,700

2,900

3,100

3,300

3,500

3,700

3,900

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Sep-11 Dec-11 Mar-12 Jun-12

Price performance chart

JA Solar Holdings (L) NASDAQ Composite (R)

Share price performance (%) 3 month 6 month 12 month

Absolute (19.2) (54.0) (75.2)

Rel. to NASDAQ Composite (27.1) (57.0) (79.8)

Source: Company data, Goldman Sachs Research estimates, FactSet. Price as of 9/07/2012 close.

INVESTMENT LIST MEMBERSHIP

Neutral

Coverage View: Neutral



Exhibit 75: JA Solar financial summary and P/B vs. ROE chart


Profit model ($ mn) 12/11 12/12E 12/13E 12/14E Balance sheet ($ mn) 12/11 12/12E 12/13E 12/14E

Total revenue 1,705.3 976.4 950.8 927.2 Cash & equivalents 617.9 689.1 364.4 370.9

Cost of goods sold (1,632.0) (952.7) (911.4) (894.4) Accounts receivable 197.8 202.1 196.8 191.9

SG&A (81.0) (109.8) (72.9) (55.4) Inventory 116.1 193.6 185.2 181.8

R&D (59.1) (11.4) (9.5) (9.3) Other current assets 293.1 293.1 293.1 293.1

Other operating profit/(expense) 0.0 0.0 0.0 0.0 Total current assets 1,224.9 1,377.9 1,039.5 1,037.6

EBITDA 10.2 (14.4) 35.7 43.0 Net PP&E 810.2 767.2 728.4 693.6

Depreciation & amortization (77.0) (83.0) (78.7) (74.8) Net intangibles 0.0 0.0 0.0 0.0

EBIT (66.8) (97.5) (43.0) (31.9) Total investments 230.8 230.8 230.8 230.8

Interest income 0.0 0.0 13.8 7.3 Other long-term assets 75.4 75.4 75.4 75.4

Interest expense (59.4) (82.6) (51.9) (45.6) Total assets 2,341.3 2,451.3 2,074.1 2,037.4

Income/(loss) from uncons. subs. 0.0 0.0 0.0 0.0

Others 45.7 (2.5) 0.0 0.0 Accounts payable 115.2 385.8 225.2 215.4

Pretax profits (80.5) (182.6) (81.1) (70.1) Short-term debt 84.2 84.2 (95.8) (95.8)

Income tax (9.2) (12.7) 9.7 8.4 Other current liabilities 328.9 328.9 328.9 328.9

Minorities 0.0 0.0 0.0 0.0 Total current liabilities 528.3 798.9 458.3 448.6

Long-term debt 746.8 746.8 746.8 746.8

Net income pre-preferred dividends (89.7) (195.3) (71.4) (61.7) Other long-term liabilities 25.6 25.6 25.6 25.6

Preferred dividends 0.0 0.0 0.0 0.0 Total long-term liabilities 772.4 772.4 772.4 772.4

Net income (pre-exceptionals) (89.7) (195.3) (71.4) (61.7) Total liabilities 1,300.8 1,571.3 1,230.8 1,221.0

Post-tax exceptionals 0.0 0.0 0.0 0.0

Net income (89.7) (195.3) (71.4) (61.7) Preferred shares 0.0 0.0 0.0 0.0

Total common equity 1,040.5 880.0 843.3 816.4

EPS (basic, pre-except) ($) (0.54) (1.00) (0.36) (0.32) Minority interest 0.0 0.0 0.0 0.0

EPS (basic, post-except) ($) (0.54) (1.00) (0.36) (0.32)

EPS (diluted, post-except) ($) (0.54) (1.00) (0.36) (0.32) Total liabilities & equity 2,341.3 2,451.3 2,074.1 2,037.4

DPS ($) 0.00 0.00 0.00 0.00

Dividend payout ratio (%) 0.0 0.0 0.0 0.0 BVPS ($) 6.40 4.50 4.31 4.17

Free cash flow yield (%) (35.7) 45.5 (92.4) 4.1

Growth & margins (%) 12/11 12/12E 12/13E 12/14E Ratios 12/11 12/12E 12/13E 12/14E

Sales growth (4.3) (42.7) (2.6) (2.5) CROCI (%) 9.3 0.5 7.2 7.6

EBITDA growth (97.0) (241.3) 347.3 20.5 ROE (%) (8.7) (20.3) (8.3) (7.4)

EBIT growth (122.3) (45.9) 55.8 26.0 ROA (%) (4.4) (8.2) (3.2) (3.0)

Net income growth (133.7) (117.8) 63.4 13.5 ROACE (%) (2.6) (9.9) (3.1) (2.1)

EPS growth (132.9) (85.9) 63.4 13.5 Inventory days 35.8 59.3 75.9 74.9

Gross margin 4.3 2.4 4.1 3.5 Receivables days 36.5 74.7 76.6 76.5

EBITDA margin 0.6 (1.5) 3.8 4.6 Payable days 30.4 96.0 122.3 89.9

EBIT margin (3.9) (10.0) (4.5) (3.4) Net debt/equity (%) 20.5 16.1 34.0 34.3

Interest cover - EBIT (X) NM NM NM NM

Cash flow statement ($ mn) 12/11 12/12E 12/13E 12/14E Valuation 12/11 12/12E 12/13E 12/14E

Net income pre-preferred dividends (89.7) (195.3) (71.4) (61.7)

D&A add-back 77.0 83.0 78.7 74.8 P/E (analyst) (X) NM NM NM NM

Minorities interests add-back 0.0 0.0 0.0 0.0 P/B (X) 0.1 0.2 0.2 0.2

Net (inc)/dec working capital 27.9 188.7 (146.9) (1.4) EV/EBITDA (X) 97.0 NM 12.4 10.2

Other operating cash flow 57.4 34.8 34.8 34.8 EV/GCI (X) 0.8 0.3 0.4 0.4

Cash flow from operations 72.7 111.2 (104.7) 46.5 Dividend yield (%) 0.0 0.0 0.0 0.0

Capital expenditures (350.1) (40.0) (40.0) (40.0)

Acquisitions 0.0 0.0 0.0 0.0

Divestitures 0.0 0.0 0.0 0.0

Others 22.7 0.0 0.0 0.0

Cash flow from investments (327.5) (40.0) (40.0) (40.0)

Dividends paid (common & pref) 0.0 0.0 0.0 0.0

Inc/(dec) in debt 0.0 0.0 (180.0) 0.0

Common stock issuance (repurchase) (0.9) 0.0 0.0 0.0

Other financing cash flows 526.8 0.0 0.0 0.0

Cash flow from financing 526.0 0.0 (180.0) 0.0

Total cash flow 271.2 71.2 (324.7) 6.5 Note: Last actual year may include reported and estimated data.

Source: Company data, Goldman Sachs Research estimates.

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%

0.0X

0.2X

0.4X

0.6X

0.8X

1.0X

1.2X

1.4X

1.6X

1.8X 12-m forward P/B ROE

JA Solar (JASO) P/B vs. ROE



Exhibit 76: Solar investment views – Valuations and key risks


Stock 12-month investment stance 12-month price targets; key risks

GCL-Poly

3800.HK (Neutral)

Sep 07 close: HK$ 1.19

(1) Some at-risk marginal capacity to finally exit the supply chain, as poly price reaches

bottom; (2) GCL to deliver results on its solar farm EPC and system integration businesses,

enhancing earnings/reducing its high gearing; (3) China potentially responding to

antidumping (AD) duties issued by the US Department of Commerce by imposing its own

tariffs on imported solar raw materials, including polysilicon.

Director's Cut-based 12-month target price of HK$1.10. Adjusting down TP to

reflect our expectations of further ASP erosion and a continued industry down-

cycle. Key risks: Better- or worse-than-expected global solar demand; higher- or

lower-than-expected ASPs; higher- or lower-than-expected costs.

Trina

TSL (Neutral)

Sep 07 close: US$ 4.03

(1) Better than peers balance sheet suggests a stronger ability to survive an extended period

of turbulence, as well as provide flexibility to explore M&A activity during

rationalization/consolidation; (2) conservative management track record.

Director's Cut-based 12-month target price of US$4.00. Adjusting down TP to


cycle. Key risks: Better- or worse-than-expected global solar demand; higher- or

lower-than-expected ASPs; higher- or lower-than-expected costs.

JA Solar

JASO (Neutral)


(1) Better than peers balance sheet suggests stronger ability to survive an extended period of

turbulence, as well as provide flexibility to explore M&A activity during

rationalization/consolidation; (2) JASO has been persistently increasing degree of vertical

integration by adding more module/wafer capacity; hence, we expect it to structurally gain

the opportunity to improve margin outlook in 2013E.

Resuming coverage with a Director's Cut-based 12-month target price of US$1.00.

We believe its balance sheet is the strongest among its peers, giving it a

competitive advantage amid a structurally deteriorated industry. Key risks: Better-

or worse-than-expected global solar demand; higher- or lower-than-expected

ASPs; higher- or lower-than-expected costs.

Yingli

YGE (Sell)


(1) High gearing ratio indicates potential credit risk; (2) High SG&A expense impacts earnings

outlook.

Director's Cut-based 12-month target price of US$0.8. Adjusting down TP to reflect

our expectations of further ASP erosion and a continued industry down-cycle. Key

upside risks: Higher-than-expected ASP, shipments and global demand; better-

than-expected cost structure reduction.

Sungrow

300274.SZ (Buy)

Sep 07 close: Rmb 9.12

(1) Sungrow’s strong balance sheet and leading inverter production scale could provide a

competitive advantage; (2) Sungrow appears to be well positioned to benefit from the secular

growth outlook of domestic PV market.

P/E vs. EPS-based 12-month target price of Rmb11.00. Adjusting down TP to


cycle. Key downside risks: Greater than expected decline in ASP; limited room for

cost reduction and fiercer than expected competition from competitors as the

domestic PV market picks up.

Jing Gong

002006.SZ (Neutral)


(1) Continuous investment in R&D and long history of solar equipment business experience

might help Jing Gong to get some replacement/upgrade orders even in an extended down-

cycle; (2) Solid growth and stable margins for its non-solar equipment business, which could

help JG to offset some earnings vulnerability from its solar equipment business.

P/E vs. EPS-based 12-month target price of Rmb 8.40. Adjusting down TP to reflect


risks: Better- or worse-than-expected solar demand; higher- or lower-than-

expected ASPs; higher- or lower-than-expected costs.

Jingyuntong

601908.SS (Sell)


(1) Shipment/ASP to decline faster than peers, mainly due to its poor customer profile and

lack of R&D advantages; (2) Significant increase of working capital should bring higher

financial burden to the company.

P/E vs. EPS-based 12-month target price of Rmb 2.40. Adjusting down TP to reflect


upside risks: Higher than expected ASP/ shipments; lower than expected unit cost.



Appendix I: Goldman Sachs global solar demand model Exhibit 77: GS global solar module demand (in MW)


Module Demand (MW) 2005 2006 2007 2008 2009 2010 2011

AsiaJapan 321 296 230 230 450 1,000 1,500 2,200 5,000 5,000 2,700

Korea 5 21 60 270 415 240 130 300 210 330 350

China - - 6 9 88 500 2,700 6,000 8,000 9,000 10,000

India - - - - 90 18 400 750 1,500 2,500 4,000

Taiwan - - - - - 8 50 50 50 50 50

Australia - 10 20 50 90 383 600 800 700 700 1,000

Asia total 326 328 316 559 1,132 2,149 5,380 10,100 15,460 17,580 18,100Growth 1% -3% 77% 102% 90% 150% 88% 53% 14% 3%

EuropeGermany 912 1,003 1,200 1,500 3,800 7,200 7,500 6,500 4,500 4,500 5,000

Italy 36 60 90 260 720 6,000 5,700 4,000 700 700 1,000

Czech Republic - - - 51 400 1,820 - - - - -

France 27 43 45 100 200 719 1,000 1,000 500 500 500

Spain 24 47 545 2,500 270 500 250 200 50 200 500

Belgium - 2 18 48 413 350 500 100 100 100 100

Luxembourg 23 31 42 57 80 108 124 150 173 198 228

Portugal 2 4 10 50 50 100 100 100 150 200 200

Greece - 5 2 20 35 150 300 150 150 250 300

UK - 3 4 4 4 96 760 300 120 130 140

Bulgaria - - - - 20 28 90 126 77 108 151

Romania 5 200 150 150 150 150

Turkey - - - - - 3 100 350 300 100 100

Europe - other - - - 125 300 639 499 650 600 216 244

Europe total 1,024 1,199 1,955 4,715 6,292 17,719 17,123 13,776 7,570 7,352 8,613 Growth 17% 63% 141% 33% 113% -3% -20% -45% -3% 17%

North AmericaUSA 108 141 268 350 540 1,000 2,000 3,000 4,000 5,000 6,000

Canada 5 5 10 10 40 200 500 700 1,000 1,000 1,000

North America total 113 146 278 360 580 1,200 2,500 3,700 5,000 6,000 7,000 Growth 30% 90% 30% 61% 107% 108% 48% 35% 20% 17%

Others - 300 390 100 50 150 571 592 763 1,000 1,250

Total 1,463 1,973 2,939 5,735 8,054 21,218 25,574 28,168 28,793 31,932 34,963Growth 35% 49% 95% 40% 163% 21% 10% 2% 11% 9%

Global cell/module capacity (MW) 1,505 2,869 5,367 9,671 14,946 32,888 52,231 54,874 56,154 57,753 58,626Utilization rate 97% 69% 55% 59% 54% 65% 49% 51% 51% 55% 60%

2015E2012E 2013E 2014E



Appendix II: Levelised cost of energy (LCOE) analysis by geographic coverage Exhibit 78: Japan has among the highest electricity prices globally, a key driver for grid parity Global LCOE comparison (July 2012)

Note: The US averages for commercial/industrial and retail grid parity encompass individual regions with different levels of solar insolation. Some regions within the US – e.g. the Southwest – are likely significantly less than five years away from grid parity. Source: Goldman Sachs Research estimates, Gao Hua Securities Research.

Wholesale "parity" by region

Wholesale price ($/kWh) Implied Assumed % Years from

Country Base PeakLCOE

($/kWh) WACCsystem capex

($/Wp)Current system

capex ($/Wp)annual capital cost decline

wholesale "parity"

Japan $0.17 $0.21 $0.21 7.8% $2.18 $2.25 15% 0.2Italy $0.10 $0.12 $0.12 9.2% $1.48 $2.25 15% 2.6Germany $0.07 $0.08 $0.08 5.6% $0.69 $2.25 15% 7.3Spain $0.07 $0.07 $0.07 7.8% $0.84 $2.25 15% 6.1France $0.07 $0.09 $0.09 7.8% $0.76 $2.25 15% 6.7China $0.06 $0.06 $0.06 7.8% $0.62 $2.25 15% 7.9Saudi Arabia $0.03 $0.03 $0.03 7.8% $0.21 $2.25 15% 14.7Russia $0.03 $0.05 $0.05 7.8% $0.19 $2.25 15% 15.2

Retail "parity" by region -- commercial and industrial

Retail price ($/kWh) Implied Assumed % Years from

Country C&ILCOE




retail "parity"

Italy $0.14 $0.14 9.2% $1.77 $2.75 15% 2.7Germany $0.12 $0.12 5.6% $1.22 $2.75 15% 5.0Spain $0.14 $0.14 7.8% $1.99 $2.75 15% 2.0France $0.09 $0.09 7.8% $0.76 $2.75 15% 7.9USA average $0.10 $0.10 6.3% $1.24 $2.75 15% 4.9

Retail "parity" by region -- residential

Retail price ($/kWh) Implied Assumed % Years from

Country ResidentialLCOE




retail "parity"

Japan $0.26 $0.26 7.8% $2.77 $3.50 15% 1.4Italy $0.18 $0.18 9.2% $2.36 $3.50 15% 2.4Germany $0.18 $0.18 5.6% $2.02 $3.50 15% 3.4Spain $0.19 $0.19 7.8% $2.81 $3.50 15% 1.4France $0.12 $0.12 7.8% $1.11 $3.50 15% 7.1USA average $0.12 $0.12 6.3% $1.55 $3.50 15% 5.0



Appendix III: Key assumptions for our LCOE model Exhibit 79: Among the major solar markets, Italy, Spain and the Southwestern US have among the best solar resources, while Germany has relatively weak solar resources and thus a relatively low load factor Global solar resources comparison (July 2012)

Source: European Commission Joint Research Center Institute for Environment and Sustainability, Goldman Sachs Research.

RegionInsolation

(kWh/m2/day)

Peak sunshine

hoursDC to AC

AdjustmentAdjusted

peak hoursLoad

factor Wh/WpAustralia 5.5 5.5 83% 4.6 19.0% 1,664

Brazil 4.0 4.0 83% 3.3 14.0% 1,226

China - East 3.9 3.9 83% 3.2 13.0% 1,139

China - West 4.5 4.5 83% 3.7 16.0% 1,402

France - Paris 3.5 3.5 83% 2.9 12.0% 1,051

Germany - Berlin 3.2 3.2 83% 2.6 11.0% 964

Germany - South 3.1 3.1 83% 2.6 11.0% 964

Greece 4.8 4.8 83% 4.0 17.0% 1,489

India 5.0 5.0 83% 4.2 17.0% 1,489

Italy - Center 4.6 4.6 83% 3.8 16.0% 1,402

Italy - South 5.2 5.2 83% 4.3 18.0% 1,577

Japan 3.5 3.5 83% 2.9 12.0% 1,051

Mexico 4.5 4.5 83% 3.7 16.0% 1,402

North Africa 6.5 6.5 83% 5.4 22.0% 1,927

Peru 5.0 5.0 83% 4.2 17.0% 1,489

Russia 3.0 3.0 83% 2.5 10.0% 876

Saudi Arabia 5.5 5.5 83% 4.6 19.0% 1,664

Spain - Center 5.2 5.2 83% 4.3 18.0% 1,577

Spain - South 5.3 5.3 83% 4.4 18.0% 1,577

UK - London 3.1 3.1 83% 2.6 11.0% 964

US - Average 4.5 4.5 83% 3.7 16.0% 1,402

US - California 5.6 5.6 83% 4.6 19.0% 1,664

US - Southwest 6.5 6.5 83% 5.4 22.0% 1,927



Appendix IV: The ESG risk evaluation indicators of GS SUSTAIN

Exhibit 80: The ESG risk evaluation indicators of GS SUSTAIN

Source: Goldman Sachs Research.

Criteria Description Purpose Weighting

Independent Board leadership Separation of CEO and Chairman roles and appointment of independent Lead Director Maintain balance of power

Independent Board directors & committees Percentage of independent, non-executive directors and wholly independent compensation and nomination committees Shareholder representation

Independent auditors Audit committee independence and ratio of non-audit to audit fees paid to the assigned auditor Independence of audit process

CEO compensation CEO compensation (including salary, bonus, stock grants and options) as a percentage of net income Management incentives

Share-based compensation Fair value of share-based compensation expense as percentage of equity Transparency

Minority shareholders' rights Block ownership greater than 5%, staggered Board, poison pill, unequal voting rights and other provisions Strength of individual shareholders

Reporting and assurance of ESG performance Number of years of reporting on environmental and social ("ES") issues and external assurance of data Transparency

Leadership responsibility for ESG performance Compensation link and responsibility of Board, senior executives to environmental and social performance Integration of ES issues into strategy

Employee compensation Total payroll costs divided by average number of employees Employee incentives

Employee productivity Cash flow per employee Labour efficiency

Gender diversity Gender diversity of total workforce, senior executives, and Board directors Quality of workplace

Employee training Institutionalized training programme, amount of resources used for training, hours or spend Employee incentives

Health and safety management Behaviour based health & safety policy, health & safety risk assessment, and pandemics policy Quality of workplace

Fatalities Total employee and contractor fatalities and rate per 50,000 employees Quality of workplace

Health & safety performance - LTI Lost time injuries of employees and contractors per million hours worked Quality of workplace

Research and development Research and development expenditure relative to cash flow Product innovation

Community investment Community investment as a percentage of cash flow Brand, impact on communities

Supply chain management Supplier guidelines, suppliers: assess for environment, suppliers: assess for human rights, and % of suppliers assessed disclosed Quality of supply chain

Business ethics and corruption Procedures for stakeholder dialogue, Whistleblower mechanisms, UN declaration of human rights, bribery prohibition License to operate

Environmental policy & targets Climate change targets, policy to increase use of power from renewable sources, recycling program Environmental awareness

Energy consumption Energy consumption as a ratio of gross cash invested Energy efficiency

GHG emissions Greenhouse gas emissions as a ratio of gross cash invested Impact of operations

Water consumption Water consumption as a ratio of gross cash invested Water efficiency

Waste generation Waste generation as a ratio of gross cash invested Impact of operations

Envi

ronm

ent

21%

Cor

pora

te g

over

nanc

eLe

ader

ship

Stak

ehol

ders

25%

8%

29%

17%

Empl

oyee

s

Soci

al



Appendix V: GCL-Poly key data Exhibit 81: GCL-Poly financial summary and PB vs. ROE chart


Profit model (HK$ mn) 12/11 12/12E 12/13E 12/14E Balance sheet (HK$ mn) 12/11 12/12E 12/13E 12/14E

Total revenue 25,505.6 25,623.7 27,901.1 28,073.2 Cash & equivalents 6,882.7 7,887.9 7,573.0 7,428.9

Cost of goods sold (17,039.3) (22,417.3) (24,315.2) (23,853.0) Accounts receivable 7,040.1 12,462.7 12,230.6 12,306.0

SG&A (1,674.0) (1,844.1) (1,528.6) (1,684.4) Inventory 3,626.7 5,433.9 5,995.5 5,881.6

R&D -- -- -- -- Other current assets 4,465.8 4,465.8 4,465.8 4,465.8

Other operating profit/(expense) 0.0 0.0 0.0 0.0 Total current assets 22,015.3 30,250.3 30,264.9 30,082.3

EBITDA 8,880.4 4,260.3 5,108.0 5,514.7 Net PP&E 41,181.3 43,441.2 42,445.2 41,518.0

Depreciation & amortization (2,088.0) (2,897.9) (3,050.7) (2,979.0) Net intangibles 1,061.7 1,003.8 949.1 897.4

EBIT 6,792.4 1,362.3 2,057.3 2,535.7 Total investments 388.4 388.4 388.4 388.4

Interest income 95.9 101.5 116.3 111.6 Other long-term assets 2,841.6 2,841.6 2,341.6 1,841.6

Interest expense (1,166.3) (2,526.0) (2,456.0) (2,316.0) Total assets 67,488.2 77,925.3 76,389.3 74,727.6

Income/(loss) from uncons. subs. 0.0 0.0 0.0 0.0

Others 117.2 308.1 460.0 400.0 Accounts payable 7,628.1 12,249.6 12,657.2 12,416.6

Pretax profits 5,839.1 (754.2) 177.6 731.3 Short-term debt 11,582.4 15,382.4 13,382.4 11,382.4

Income tax (1,269.2) (127.1) (71.2) (102.4) Other current liabilities 2,241.5 2,241.5 2,241.5 2,241.5

Minorities (295.1) (103.1) (50.0) (50.0) Total current liabilities 21,452.1 29,873.5 28,281.2 26,040.6

Long-term debt 17,703.9 20,703.9 20,703.9 20,703.9

Net income pre-preferred dividends 4,274.9 (984.4) 56.4 578.9 Other long-term liabilities 6,198.2 6,198.2 6,198.2 6,198.2

Preferred dividends 0.0 0.0 0.0 0.0 Total long-term liabilities 23,902.0 26,902.0 26,902.0 26,902.0

Net income (pre-exceptionals) 4,274.9 (984.4) 56.4 578.9 Total liabilities 45,354.1 56,775.6 55,183.2 52,942.7

Post-tax exceptionals 0.0 0.0 0.0 0.0

Net income 4,274.9 (984.4) 56.4 578.9 Preferred shares 0.0 0.0 0.0 0.0

Total common equity 20,567.1 19,582.7 19,639.0 20,218.0

EPS (basic, pre-except) (HK$) 0.28 (0.06) 0.00 0.04 Minority interest 1,567.0 1,567.0 1,567.0 1,567.0

EPS (basic, post-except) (HK$) 0.28 (0.06) 0.00 0.04

EPS (diluted, post-except) (HK$) 0.28 (0.06) 0.00 0.04 Total liabilities & equity 67,488.2 77,925.3 76,389.3 74,727.6

DPS (HK$) 0.00 0.00 0.00 0.00

Dividend payout ratio (%) 0.0 0.0 0.0 0.0 BVPS (HK$) 1.33 1.27 1.27 1.31

Free cash flow yield (%) (21.8) (29.0) 5.9 6.8

Growth & margins (%) 12/11 12/12E 12/13E 12/14E Ratios 12/11 12/12E 12/13E 12/14E

Sales growth 38.1 0.5 8.9 0.6 CROCI (%) 18.0 8.1 7.7 9.0

EBITDA growth 23.6 (52.0) 19.9 8.0 ROE (%) 23.3 (4.9) 0.3 2.9

EBIT growth 17.8 (79.9) 51.0 23.3 ROA (%) 7.9 (1.4) 0.1 0.8

Net income growth 6.2 (123.0) 105.7 927.2 ROACE (%) 15.6 3.3 3.1 5.4

EPS growth 6.2 (123.0) 105.7 927.2 Inventory days 56.5 73.8 85.8 90.9

Gross margin 33.2 12.5 12.9 15.0 Receivables days 67.3 138.9 161.5 159.5

EBITDA margin 34.8 16.6 18.3 19.6 Payable days 126.6 161.8 186.9 191.8

EBIT margin 26.6 5.3 7.4 9.0 Net debt/equity (%) 101.2 133.3 125.0 113.2

Interest cover - EBIT (X) 6.3 0.6 0.9 1.2

Cash flow statement (HK$ mn) 12/11 12/12E 12/13E 12/14E Valuation 12/11 12/12E 12/13E 12/14E

Net income pre-preferred dividends 4,274.9 (984.4) 56.4 578.9

D&A add-back 2,088.0 2,897.9 3,050.7 2,979.0 P/E (analyst) (X) 4.3 NM NM 31.8

Minorities interests add-back 295.1 103.1 50.0 50.0 P/B (X) 0.9 0.9 0.9 0.9

Net (inc)/dec working capital (3,176.7) (2,608.3) 78.1 (202.0) EV/EBITDA (X) 8.6 10.9 8.8 7.8

Other operating cash flow (717.4) (103.1) (50.0) (50.0) EV/GCI (X) 1.5 0.8 0.8 0.7

Cash flow from operations 2,763.8 (694.8) 3,185.1 3,355.9 Dividend yield (%) 0.0 0.0 0.0 0.0

Capital expenditures (14,792.7) (5,100.0) (2,000.0) (2,000.0)

Acquisitions 0.0 0.0 0.0 0.0

Divestitures 0.0 0.0 0.0 0.0

Others (3,452.6) 0.0 500.0 500.0

Cash flow from investments (18,245.3) (5,100.0) (1,500.0) (1,500.0)

Dividends paid (common & pref) 0.0 0.0 0.0 0.0

Inc/(dec) in debt 15,496.1 6,800.0 (2,000.0) (2,000.0)

Common stock issuance (repurchase) 0.0 0.0 0.0 0.0

Other financing cash flows 37.9 0.0 0.0 0.0

Cash flow from financing 15,534.0 6,800.0 (2,000.0) (2,000.0)

Total cash flow 377.5 1,005.2 (314.9) (144.1) Note: Last actual year may include reported and estimated data.

Source: Company data, Goldman Sachs Research estimates.

-20%

-10%

0%

10%

20%

30%

40%

50%

0.0X

0.5X

1.0X

1.5X

2.0X

2.5X

3.0X

3.5X

4.0X

4.5X 12-m forward P/B ROE

GCL-Poly (3800.HK) P/B vs. ROE



Exhibit 82: GCL-Poly’s earnings matrix


Solar operational data 2012E 2013E 2014E 1H12 2H12E 1H13E 2H13E

Capacity - period-end annualized

Polysilicon (MT) 65,000 65,000 65,000 65,000 65,000 65,000 65,000

Wafer (MW) 8,000 8,000 8,000 8,000 8,000 8,000 8,000

Production

Polysilicon (MT) 50,233 54,000 60,000 25,233 25,000 27,000 27,000

Wafer (MW) 5,842 6,400 6,400 3,042 2,800 3,200 3,200

Shipment

Polysilicon (MT) 18,462 19,120 26,080 9,012 9,450 9,720 9,400

Wafer (MW) 5,842 6,400 6,400 3,042 2,800 3,200 3,200

ASP

Polysilicon (US$/kg) 21.3 18.0 15.0 23.2 19.6 18.0 18.0

Wafer (US$/W) 0.26 0.22 0.20 0.27 0.25 0.22 0.22

Unit cost

Polysilicon (US$/kg) 18.4 17.0 14.0 18.9 18.0 17.0 16.0

Wafer processing (US$/W) 0.12 0.10 0.09 0.13 0.12 0.10 0.09

P&L snapshot 2012E 2013E 2014E 1H12 2H12E 1H13E 2H13E

Revenue (HK$ mn) 25,624 27,901 28,073 11,782 13,842 13,356 14,545

Solar business 20,032 22,299 21,913 8,991 11,041 10,555 11,744

Silicon 3,058 2,684 3,051 1,624 1,434 1,365 1,320

Wafer 12,043 10,982 9,984 6,677 5,366 5,491 5,491

Solar farm & system integration 4,931 8,632 8,878 690 4,241 3,700 4,933

Power business 5,592 5,602 6,160 2,791 2,801 2,801 2,801

Gross margin (%) 14% 14% 15% 14% 11% 11% 14%

Solar business 13% 13% 14% 15% 11% 11% 15%

Silicon 13% 8% 7% 18% 7% 6% 11%

Wafer 16% 16% 18% 18% 13% 12% 19%

Solar farm & system integration 6% 12% 12% (18%) 10% 12% 12%

Power business 19% 19% 19% 11% 11% 11% 11%

Opex (HK$ mn) 1,844 1,529 1,684 944 900 801 727

EBIT (HK$ mn) 1,362 2,057 2,536 745 618 689 1,369

EBIT margin (%) 5% 7% 9% 6% 4% 5% 9%

Interest expense (HK$ mn) 2,526 2,456 2,316 1,093 1,332 1,170 1,170

Net income (HK$ mn) (984) 56 579 (330) (654) (361) 417

Net margin (%) (4%) 0% 2% (3%) (5%) (3%) 3%

Dilluted EPS (HK$) (0.06) 0.00 0.04 (0.02) (0.04) (0.02) 0.03

Balance sheet snapshot 2012E 2013E 2014E 1H12 2H12E 1H13E 2H13E

A/R day 178 160 160 175 180 N.A. N.A.

Inventory day 88 90 90 97 80 N.A. N.A.

A/P day 199 190 190 199 200 N.A. N.A.

Cash (HK$ mn) 7,888 7,573 7,429 8,836 7,888 N.A. N.A.

Short term debt (HK$ mn) 15,382 13,382 11,382 16,359 15,382 N.A. N.A.

Long term debt (HK$ mn) 20,704 20,704 20,704 23,042 20,704 N.A. N.A.

Net Debt to equity (%) 133% 125% 113% 145% 133% N.A. N.A.

ROE (%) (4.5%) 0.3% 2.7% (3.1%) (6.3%) N.A. N.A.

Cash flow snapshot 2012E 2013E 2014E 1H12 2H12E 1H13E 2H13E

Cash flow from operations (695) 3,185 3,356 (1,432) 840 N.A. N.A.

Change of working capital (2,608) 78 (202) (2,640) 32 N.A. N.A.

Cash flow from investing (5,100) (1,500) (1,500) (3,950) (1,150) N.A. N.A.

Capex (5,100) (2,000) (2,000) (3,000) (2,100) N.A. N.A.

Cash flow from financing 6,800 (2,000) (2,000) 7,335 (535) N.A. N.A.

Net change of ST debt 3,800 (2,000) (2,000) 4,776 (976) N.A. N.A.

Net change of ST debt 3,000 0 0 3,507 (507) N.A. N.A.



信息披露附录

申明我们，宋赟波, CFA、胡玲玲，在此申明，本报告所表述的所有观点准确反映了我们对上述公司或其证券的个人看法。此外，我们的薪金的任何部分不曾与，不与，也将不会与本报告中的具体推荐意见或观点直接

或间接相关。

投资摘要投资摘要部分通过将一只股票的主要指标与其行业和市场相比较来评价该股的投资环境。所描述的四个主要指标包括增长、回报、估值倍数和波动性。增长、回报和估值倍数都是运用数种方法综合计算而成，以确

定该股在地区研究行业内所处的百分位排名。

每项指标的准确计算方式可能随着财务年度、行业和所属地区的不同而有所变化，但标准方法如下：

增长是下一年预测与当前年度预测的综合比较，如每股盈利、EBITDA 和收入等。回报是各项资本回报指标一年预测的加总，如 CROCI、平均运用资本回报率和净资产回报率。估值倍数根据一年预期估值比率综

合计算，如市盈率、股息收益率、EV/FCF、EV/EBITDA、EV/DACF、市净率。波动性根据 12 个月的历史波动性计算并经股息调整。

Quantum Quantum 是提供具体财务报表数据历史、预测和比率的高盛专有数据库，它可以用于对单一公司的深入分析，或在不同行业和市场的公司之间进行比较。

GS SUSTAIN GS SUSTAIN是侧重于长期做多建议的相对稳定的全球投资策略。GS SUSTAIN 关注名单涵盖了我们认为相对于全球同业具有持续竞争优势和出色的资本回报、因而有望在长期内表现出色的行业领军企业。我们

对领军企业的筛选基于对以下三方面的量化分析：现金投资的现金回报、行业地位和管理水平（公司管理层对行业面临的环境、社会和企业治理方面管理的有效性）。

信息披露

相关的股票研究范围宋赟波, CFA：中国新能源。胡玲玲：中国新能源、中国电信行业、中国科技行业。

中国新能源：京运通、佛山照明、保利协鑫、晶澳太阳能、雷士照明、阳光电源、天合光能、英利绿色能源、精功科技、阳光照明。

中国科技行业：光迅科技、科大讯飞、亚信、高德软件、北纬通信、神州泰岳、京信通信、烽火通信、歌尔声学、拓维信息、MStar Semiconductor、联发科、国民技术、四维图新、Parade Technologies Ltd、大

富科技、日海通讯、台积电、台积电 (ADR)、联电 (ADR)、联电、中兴通讯(A)、中兴通讯(H)。

中国电信行业：中国通信服务、中国移动、中国移动(ADR)、中电信、中电信(ADR)、中联通(H)、中联通(ADS)、中联通(A)。



与公司有关的法定披露以下信息披露了高盛高华证券有限责任公司（“高盛高华”）与北京高华证券有限责任公司（“高华证券”）投资研究部所研究的并在本研究报告中提及的公司之间的关系。

高盛高华在今后 3个月中预计将从下述公司获得或寻求获得投资银行服务报酬：天合光能 ($4.23)

没有对下述公司的具体信息披露：京运通 (Rmb6.90)、保利协鑫 (HK$1.17)、晶澳太阳能 ($0.85)、阳光电源 (Rmb9.48)、英利绿色能源 ($1.82)、精功科技 (Rmb9.40)

公司评级、研究行业及评级和相关定义买入、中性、卖出：分析师建议将评为买入或卖出的股票纳入地区投资名单。一只股票在投资名单中评为买入或卖出由其相对于所属研究行业的潜在回报决定。任何未获得买入或卖出评级的股票均被视为中性评

级。每个地区投资评估委员会根据 25-35%的股票评级为买入、10-15%的股票评级为卖出的全球指导原则来管理该地区的投资名单；但是，在某一特定行业买入和卖出评级的分布可能根据地区投资评估委员会的决

定而有所不同。地区强力买入或卖出名单是以潜在回报规模或实现回报的可能性为主要依据的投资建议。

潜在回报：代表当前股价与一定时间范围内预测目标价格之差。分析师被要求对研究范围内的所有股票给出目标价格。潜在回报、目标价格及相关时间范围在每份加入投资名单或重申维持在投资名单的研究报告中

都有注明。

研究行业及评级：分析师给出下列评级中的其中一项代表其根据行业历史基本面及／或估值对研究对象的投资前景的看法。具吸引力(A)：未来 12 个月内投资前景优于研究范围的历史基本面及／或估值。中性(N)：未来 12 个月内投资前景相对研究范围的历史基本面及／或估值持平。谨慎(C)：未来 12 个月内投资前景劣于研究范围的历史基本面及／或估值。

暂无评级(NR)：在高盛高华于涉及该公司的一项合并交易或战略性交易中担任咨询顾问时并在某些其他情况下，投资评级和目标价格已经根据高华证券的政策予以除去。暂停评级(RS)：由于缺乏足够的基础去确定

投资评级或价格目标，或在发表报告方面存在法律、监管或政策的限制，我们已经暂停对这种股票给予投资评级和价格目标。此前对这种股票作出的投资评级和价格目标(如有的话)将不再有效，因此投资者不应依

赖该等资料。暂停研究(CS)：我们已经暂停对该公司的研究。没有研究(NC)：我们没有对该公司进行研究。不存在或不适用(NA)：此资料不存在或不适用。无意义(NM)：此资料无意义，因此不包括在报告内。

一般披露本报告在中国由高华证券分发。高华证券具备证券投资咨询业务资格。

本研究报告仅供我们的客户使用。本研究报告是基于我们认为可靠的目前已公开的信息，但我们不保证该信息的准确性和完整性，客户也不应该依赖该信息是准确和完整的。我们会适时地更新我们的研究，但各种

规定可能会阻止我们这样做。除了一些定期出版的行业报告之外，绝大多数报告是在分析师认为适当的时候不定期地出版。

高盛高华为高华证券的关联机构，从事投资银行业务。高华证券、高盛高华及它们的关联机构与本报告中涉及的大部分公司保持着投资银行业务和其它业务关系。

我们的销售人员、交易员和其它专业人员可能会向我们的客户及我们的自营交易部提供与本研究报告中的观点截然相反的口头或书面市场评论或交易策略。我们的自营交易部和投资业务部可能会做出与本报告的建

议或表达的意见不一致的投资决策。

本报告中署名的分析师可能已经与包括高华证券销售人员和交易员在内的我们的客户讨论，或在本报告中讨论交易策略，其中提及可能会对本报告讨论的证券市场价格产生短期影响的推动因素或事件，该影响在方

向上可能与分析师发布的股票目标价格相反。任何此类交易策略都区别于且不影响分析师对于该股的基本评级，此类评级反映了某只股票相对于报告中描述的研究范围内股票的回报潜力。

高华证券及其关联机构、高级职员、董事和雇员，不包括股票分析师和信贷分析师，将不时地对本研究报告所涉及的证券或衍生工具持有多头或空头头寸，担任上述证券或衍生工具的交易对手，或买卖上述证券或

衍生工具。

在任何要约出售股票或征求购买股票要约的行为为非法的地区，本报告不构成该等出售要约或征求购买要约。本报告不构成个人投资建议，也没有考虑到个别客户特殊的投资目标、财务状况或需求。客户应考虑本

报告中的任何意见或建议是否符合其特定状况，以及(若有必要)寻求专家的意见，包括税务意见。本报告中提及的投资价格和价值以及这些投资带来的收入可能会波动。过去的表现并不代表未来的表现，未来的回

报也无法保证，投资者可能会损失本金。外汇汇率波动有可能对某些投资的价值或价格或来自这一投资的收入产生不良影响。

某些交易，包括牵涉期货、期权和其它衍生工具的交易，有很大的风险，因此并不适合所有投资者。投资者可以向高华销售代表取得或通过 http://www.theocc.com/about/publications/character-risks.jsp 取得当前的

期权披露文件。对于包含多重期权买卖的期权策略结构产品，例如，期权差价结构产品，其交易成本可能较高。与交易相关的文件将根据要求提供。

北京高华证券有限责任公司版权所有 © 2012 年

未经北京高华证券有限责任公司事先书面同意，本材料的任何部分均不得(i)以任何方式制作任何形式的拷贝、复印件或复制品，或(ii)再次分发。

goldman sachs-转型中国：探索太阳能行业的理性化之路－展望，下一个十年

Documents