China Solar Sector

See important disclosures, including any required research certifications, beginning on page 163

China Utilities

29 October 2015

China Solar Sector

Initiation: a more powerful future for solar

China’s solar installations are set to accelerate from 10.5GW in 9M15 to 16GW by end-2015; we see rush installations continuing into 2016

Solar distributed generation (DG) should benefit from PV system cost reduction; we expect retail parity in 2017

Prefer downstream solar; top pick is Singyes (Buy [1]); pecking order within up- to mid-stream solar: JinkoSolar, GCL-Poly and Xinyi Solar

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

China Solar Sector: 29 October 2015

Table of contents

Investment thesis ..................................................................................................... 7

4Q15-2016: rush installation period ........................................................................ 9

2017-18: beginning of the retail parity revolution ................................................ 22

Key stock calls ........................................................................................................ 38

Valuation ................................................................................................................. 43

Risks to our call ...................................................................................................... 47

Appendix I: an overview of the global solar industry .......................................... 49

Appendix II: major solar power policies in China ................................................ 60

Appendix III: solar DG local government subsidies ............................................ 63

Appendix IV: IRR model (solar utility-scale projects).......................................... 65

Appendix V: IRR model (solar DG projects) ......................................................... 66

Appendix VI: Renewable Energy Fund forecasts ................................................ 67

Appendix VII: yieldco structure to unlock long-term volume growth potential 68

Company Section China Singyes Solar Technologies ..................................................................................73

JinkoSolar ........................................................................................................................77

GCL-Poly Energy ........................................................................................................... 107

Xinyi Solar Holdings ....................................................................................................... 143

GCL New Energy ........................................................................................................... 147


China Utilities

Investment case: We see solar installations accelerating starting from 4Q15, with a rush of installations occurring in 1H16 ahead of a potential on-grid tariff cut starting in 2H16. We expect solar DG to be a major capacity growth driver from 2017 when solar costs achieve retail parity, ie, projects become economical for end-users without government subsidies. Strong capacity growth in 4Q15-2016E: We forecast China’s solar PV installations to reach 5.5GW in 4Q15, up 100% QoQ from 2.8GW in 3Q15. We also expect a solar on-grid tariff cut in 2H16, leading to a rush of installations in 2016 to achieve 18GW installations, on our forecasts. We see module and polysilicon prices remaining stable in 2016E vs. prices in 2015, at USD0.64/W and USD17.4/tonne, respectively, given our bullish view on solar installations in China. The import restriction on polysilicon in China, and the retirement of low-efficiency module capacity, should ensure high margins for the tier-1 manufacturers. China’s solar projects to reach retail parity by 2017E. Our forecasts indicate that 27 out of 36 key cities in China should achieve retail parity by end-2017E, compared with 15 cities by end-2015E. With this, we believe solar DG’s reliance on subsidies will gradually subside, which may trigger even faster capacity additions in the future. We forecast solar new DG capacity to surge from 2GW in 2014 to 3.7GW in 2017E, or a 22% CAGR. Catalysts: The major catalysts for the China Solar Sector include a bullish 13th FYP solar-installation target, which we expect to be increased from 100GW to 150GW by 2020, with a potential increase in the renewable-energy surcharge from CNY0.015/kWh currently to CNY0.025/kWh in order to reverse the CNY50bn renewable-fund deficit by end-2015. Recommendation: In the long term, we prefer the downstream solar operators over the upstream and midstream manufacturers, in the absence of overcapacity, plus a more stable net profit stream. We reiterate our Buy (1) rating on Singyes (750 HK, HKD6.58) for its exposure to solar DG and downstream operations, with an appealing valuation of 4.9x 2016E PER. We reiterate our Outperform (2) on Xinyi Solar (XYS) (968 HK, HKD3.28), on improving solar glass demand despite a rich-looking valuation. And we initiate coverage of GCL-Poly (3800 HK, HKD1.66) and JinkoSolar (JKS; JKS US, USD25.28) with Outperform (2) ratings, and like their leadership positions in polysilicon and solar modules, respectively. Risks: The main risks: worse-than-expected grid curtailment, international disputes and interest-rate hikes.

29 October 2015

China Solar Sector

Initiation: a more powerful future for solar

China’s solar installations are set to accelerate from 10.5GW in 9M15 to 16GW by end-2015; we see rush installations continuing into 2016

Solar distributed generation (DG) should benefit from PV system cost reduction; we expect retail parity in 2017

Prefer downstream solar; top pick is Singyes (Buy [1]); pecking order within up- to mid-stream solar: JinkoSolar, GCL-Poly and Xinyi Solar

Key stock calls

Source: Daiwa forecasts

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

New Prev.China Singyes Solar Technologies (750 HK)Rating Buy BuyTarget 8.50 8.50Upside 29.2%

JinkoSolar (JKS US)Rating OutperformTarget 28.50Upside 12.7%

GCL-Poly Energy (3800 HK)Rating OutperformTarget 1.85Upside 11.4%

Xinyi Solar Holdings (968 HK)Rating Outperform OutperformTarget 3.50 3.60Upside 6.7%

4


How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook China: solar installations (2012-20E)

We foresee sustained growth in China’s solar installations through 2020E (with a slight decline in growth in 2017E due to a potential solar power tariff cut), and forecast capacity to rise at a CAGR of 12% for 2014-20E. We also expect accelerated growth in solar DG installations, which should gradually take market share away from utility-scale solar projects starting in 2017E. The main driver of this process would be retail parity being reached (our forecasts call for 15 of 36 key cities in China to achieve retail parity by end-2015E, increasing to 27 cities by end-2017E), whereby we assume solar PV system costs fall by 4.7-5.5% YoY each year.

Source: NEA, Daiwa

Valuation Solar sector: 2016E PBR vs. 2016E ROE

In terms of PBR vs. ROE, Singyes’ valuation looks the most attractive in our coverage universe, trading at a 2016E PBR of 0.8x with an 18% ROE. We think investors’ concerns over the company’s conventional curtain wall business are overdone, and believe the stock’s prevailing valuation offers a good risk-reward profile. Among the solar module manufacturers, JKS looks undervalued, trading at a 0.9x 2016E PBR with a 21% ROE. We believe the company’s superior gross margin merits an above-average valuation, whereas the stock is trading currently at a discount to its peers in terms of 2016E PBR vs. ROE.

Source: Daiwa forecasts for Singyes, JKS, GCL-Poly, and XYS; Bloomberg forecasts for others

Earnings revisions Solar sector: consensus 2015E EPS revisions

There has been a divergence in earnings revisions in the solar space since 2H14, with JKS and GCL-Poly being derated amid disappointing solar installations in China, which have affected sales volumes and ASPs. By contrast, XYS’s sales have boomed, driven by its new 2x900tpd raw glass production line. In 2015, XYS has seen upward revisions to the consensus 2015E EPS forecasts given decent solar glass volumes and margins. However, Singyes has been derated on the back of 2 sets of results that fell short of market expectations. Meanwhile, a fall in poly prices has triggered earnings cuts for GCL-Poly, while JKS has seen positive revisions as investors look for a good year for solar installations in China and the US.

Source: Bloomberg, Daiwa

-10%

0%

10%

20%

30%

40%

50%

60%

0

5

10

15

20

25

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)

Utility-scale (LHS) Solar DG (LHS) YoY growth (%) (RHS)

Xinyi Solar

GCLNEFuxinSingyes

DQ

Huaneng REUnited PVJA Solar

GCL-PolyTrina

Comtec

JKS

Longyuan

CSIQ

0%

5%

10%

15%

20%

25%

- 0.5 1.0 1.5 2.0 2.5 3.0

2016

E R

OE

(%)

2016E PBR

5060708090

100110120130140

Jun-

14

Jul-1

4

Aug-

14

Sep-

14

Oct

-14

Nov

-14

Dec

-14

Jan-

15

Feb-

15

Mar

-15

Apr-1

5

May

-15

Jun-

15

Jul-1

5

Aug-

15

Sep-

15

Oct

-15

JKS XYS Singyes GCL-Poly

(Rebased to 100)

5


Sector stocks: key indicators

Source: Bloomberg, Daiwa forecasts

China solar sector: key assumptions 2013 2014 2015E 2016E 2017E 2014 2015E 2016E 2017E

(YoY) Solar farm capacity addition (MW) - XYS - 273 784 800 800 n.a. 187% 2% 0%- Singyes - 21 100 100 100 n.a. 381% 0% 0%- GCL-Poly 288 721 1,500 1,800 2,000 150% 108% 20% 11%- JKS 350 310 600 600 600 -11% 94% 0% 0%Solar farm installed capacity (MW) - XYS 35 308 1,092 1,892 2,692 780% 254% 73% 42%- Singyes 180 171 265 365 465 -5% 55% 38% 27%- GCL-Poly 318 1,039 2,539 4,339 6,339 227% 144% 71% 46%- JKS 360 670 1,270 1,870 2,470 86% 90% 47% 32%Solar farm gross profit exposure (%) - XYS 0% 4% 15% 35% 43% 3.5pp 11.1pp 20.1pp 7.9pp- Singyes 0% 2% 3% 7% 7% 2.5pp 0.5pp 4.3pp 0pp- GCL-Poly 25% 8% 16% 29% 37% -17.1pp 8.3pp 12.9pp 7.8pp- JKS 1% 6% 16% 19% 26% 4.7pp 9.7pp 3.4pp 6.3ppGross margin (%) - XYS 30% 32% 38% 43% 46% 1.5pp 6.5pp 5.2pp 3.1pp- Singyes 24% 21% 21% 22% 22% -2.9pp 0.5pp 0.2pp 0.7pp- GCL-Poly 12% 21% 24% 30% 32% 8.6pp 3pp 6.3pp 2.3pp- JKS 20% 22% 23% 25% 27% 2.1pp 1pp 1.1pp 2.3ppNet-debt-to-equity gearing (%) - XYS -12% 23% 13% 58% 81% 35pp -10.1pp 45.5pp 23pp- Singyes 25% 58% 71% 69% 59% 33.3pp 12.1pp -2.1pp -9.2pp- GCL-Poly 164% 145% 112% 170% 189% -19.7pp -32.5pp 57.5pp 19.1pp- JKS 175% 161% 205% 213% 203% -14.5pp 44.4pp 8.2pp -9.6pp

Source: Companies, Daiwa forecasts

China: solar installation forecasts China: renewable energy surcharge forecasts

Source: NEA, Daiwa forecasts Source: Ministry of Finance, Daiwa forecasts

Share

Company Name Stock code Price New Prev. New Prev. % chg New Prev. % chg New Prev. % chg

China Singyes Solar Technologies 750 HK 6.58 Buy Buy 8.50 8.50 0.0% 0.858 0.858 0.0% 1.108 1.108 0.0%

GCL-Poly Energy 3800 HK 1.66 Outperform 1.85 0.131 0.192

JinkoSolar JKS US 25.28 Outperform 28.50 22.035 27.936

Xinyi Solar Holdings 968 HK 3.28 Outperform Outperform 3.50 3.60 (2.8%) 0.190 0.189 0.7% 0.246 0.243 1.4%

Rating Target price (local curr.) FY1

EPS (local curr.)

FY2

-10%

0%

10%

20%

30%

40%

50%

60%

0

5

10

15

20

25

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)


0.000

0.005

0.010

0.015

0.020

0.025

0.030

Jul-0

6

Jul-0

7

Jul-0

8

Jul-0

9

Jul-1

0

Jul-1

1

Jul-1

2

Jul-1

3

Jul-1

4

Jul-1

5

Jul-1

6

Jul-1

7

(CNY per kWh)

6


Polysilicon price trend Tier-1 manufacturers’ capacity vs. PV demand

Source: Bloomberg, WIND, Daiwa forecasts Source: Company data, Daiwa forecasts

China: retail parity (2015) China: retail parity (2018-20)

Source: Daiwa Source: Daiwa forecasts

10

15

20

25

Jan-13Apr-13 Jul-13 Oct-13 Jan-14Apr-14 Jul-14 Oct-14 Jan-15Apr-15 Jul-15 Oct-15

(USD/kg)

China local China ImportedInternational China local (forecast)China Imported (forecast) International (forecast)

0%

50%

100%

150%

0

20,000

40,000

60,000

80,000

2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)

PV demand (LHS)Tier-1 production capacity (LHS)% satisfied by Tier-1 manufacturers (RHS)

Reached retail parity

Not yet retail parity



7


Investment thesis

We initiate coverage on the China Solar Sector with a Positive stance, backed by 2 key themes that we expect to play out in the sector from 4Q15 to 2018: 4Q15 to 2016: We believe solar installations in China will see strong growth starting in 4Q15, and forecast 5.5GW capacity additions for the quarter (up 100% QoQ from 2.8GW in 3Q15). Heading into 2016, we expect rush installations ahead of a potential on-grid tariff cut starting in 2H16, as well as a hike in the renewable energy surcharge. Combined, these factors would boost China’s solar capacity addition to 18GW for the year, on our forecasts, vs. 16GW for 2015E. 2017 to 2018: We believe China’s retail parity revolution will likely begin in this period, triggering a new scenario in which solar DG replaces utility-scale solar projects as the major capacity growth driver once solar costs reach retail parity, ie, solar projects become economically viable for end users without government subsidies. Given this thematic underpinning, we expect China’s solar stocks to experience a cyclical upturn until the end of 2016. For 4Q15 to 2016, we prefer the mid-stream manufacturers (our preferred pick: JKS) and downstream EPC names (our preferred pick: Singyes), backed by our forecast for improving sales volume and margins. We also like utility-scale solar farm operators, such as GCL-Poly through its 62.28% stake in GCL New Energy (GCLNE) (not rated), as they should experience rapid capacity expansion during this period. From 2017, we expect China to experience a sudden drop in solar installations post our forecast tariff cut, especially the utility-scale solar farms, which is likely to worsen the supply-demand picture across the solar PV supply chain. We expect the upstream and mid-stream manufacturers to face a challenging year, with contracting sales volumes and deteriorating margins. The downstream utility-scale operators should be less affected, given their more stable cash flow from power generation. Hence, for 2017-19, we prefer solar DG (our preferred pick: Singyes), which we think will be the major beneficiary of our expected retail parity revolution. China Solar Sector: Daiwa’s preferred exposure from 4Q15 to 2018

Source: Daiwa

4Q15-2016: rush installation period We anticipate a solar installation boom in China, given our expectation for a solar on-grid tariff cut starting in 2H16, which would lead to a rush of installations in 2016. The Renewable Energy Development Fund, which collects surcharges from electricity users in order to provide subsidies to developers of renewable-energy projects, is currently operating at a shortfall, which we expect to expand to CNY50bn by the end of 2015. To alleviate this shortfall, we expect the NDRC to increase renewable energy surcharges for electricity consumers in 2016, which would also shorten the subsidy payback period for solar farms, improve their cash flow and trigger faster solar power investments.

For 4Q15, we prefer the midstream module manufacturers and downstream EPC names

But starting in 2017, we prefer solar DG related names

We expect a rush of installations driven by a likely tariff cut in 2H16, as well as an increase in renewable surcharges in China

8


Considering the above factors, we expect the polysilicon manufacturers (our preferred pick: GCL-Poly) to benefit from China’s existing polysilicon import restrictions, and foresee polysilicon prices gradually recovering from a trough of USD15.0/kg in mid-2015 to ~USD18.0/kg by end-2015. For the mid-stream segment, China’s policy on retiring low-efficiency solar module (with conversion efficiency rates of less than 16.5%) production lines should ease the overcapacity within the industry, benefiting tier-one manufacturers in the form of better sales volume and margins.

2017-18: beginning of the retail parity revolution For 2017-18, we believe the decline in utility-scale solar projects will be partly offset by an acceleration in solar DG, triggered by our expectation that solar DG will achieve retail parity. On our forecasts, 27 out of 36 key cities in China will reach retail parity for solar DG by 2017E, up from just 15 in 2015E. Our analysis indicates that PV system costs for solar DG would decline by 4.7-5.5% YoY over the period, benefiting from a 4.8-5.6% annual reduction in equipment costs, as well as a 4.6-5.3% annual reduction in balance-of-system (BOS) costs upon improving conversion efficiency. Hence, we look for solar DG installations to see a 22% capacity CAGR, from 2GW in 2014 to 3.7GW in 2017E. In turn, we forecast annual solar installations in China to expand from 16GW in 2015 to 18GW in 2016, before easing slightly to 17GW in 2017. By extension, given our bullish view on China’s solar installation growth, we expect this development to improve the supply-demand situation along the solar value chain, including polysilicon and solar modules, which should provide the tier-1 manufacturers with stable margins.

China: solar installation forecasts China: cumulative solar capacity vs. NEA target

Source: NEA, Daiwa Source: NEA, Daiwa

Corporate restructuring another focus Alongside the sector trends we have highlighted, we identify some corporate restructuring stories that we think will create buying opportunities for investors. Downstream solar IPP spin-offs: We expect JKS to spin off its downstream solar IPP business in mid-2016, and for XYS to follow suit in 2017. In both cases, we foresee such spin-offs bringing at least 3 major benefits: 1) value enhancement by unlocking the value of the downstream assets, 2) improvement in net gearing by deconsolidating the capital-intensive solar IPP businesses, and 3) one-off disposal gains from the spin-off. Yieldco listing: GCLNE (a 62.28%-owned subsidiary by GCL-Poly) expects to list its yieldco, possibly in 2016-17. Management sees 4 major benefits in a yieldco listing: 1) value enhancement for the yieldco and parent company through faster cash recycling, 2) a reduced cost of capital, 3) benefits from tax shelters, and 4) deleveraging the parent company for better gearing. Refer to the individual company section for further details.

-10%

0%

10%

20%

30%

40%

50%

60%

0

5

10

15

20

25

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)


6 17

28 44

62 79

97 116

137

0

20

40

60

80

100

120

140

160

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)

Insta lled capacityNEA 12th FYP target for 2020Potential target uplif t for 13th FYP plan for 2020

China’s solar DG should achieve retail parity starting 2017

Investors should also watch for corporate restructuring themes for individual companies

9


4Q15-2016: rush installation period

We forecast solar installations in China to pick up rapidly and total 5.5GW in 4Q15, up 100% QoQ. Heading into 2016, we see solar installations accelerating (we forecast a 2016E capacity addition of 18GW vs.16GW in 2015E), ahead of a potential solar on-grid tariff cut in 2H16, which would likely prompt a rush of installations in 2016E. Another potential positive factor we see would be an increase in the renewable energy surcharge for electricity consumers in 2016, which would help gradually reduce the shortfall the renewable energy fund is currently experiencing and shorten the subsidy payback period for solar farms. In turn, we expect the improving cash flow of the solar farm operators to trigger faster solar power investments. In the same timeframe, we see the polysilicon manufacturers (our preferred pick: GCL-Poly) benefiting from China’s existing polysilicon import restrictions, and foresee polysilicon prices gradually recovering from a trough of USD15.0/kg in mid-2015 to ~USD18.0/kg by end-2015. For the mid-stream segment, China’s policy of retiring low-efficiency solar module production lines (with conversion efficiency rates of less than 16.5%) should ease the overcapacity within the industry, benefiting tier-1 manufacturers in the form of better sales volume and margins.

Feed-in tariffs could be cut starting in 2H16 We forecast utility-scale solar installations to surge by 13% YoY to 18GW in 2016, reflecting the impact of rush installations ahead of a potential cut in the solar tariff in 2H16. A similar rush installation situation occurred in 2013, when capacity additions increased by 243% YoY ahead of a cut in the solar on-grid tariff starting in 2014. Feed-in tariff programme for solar PV in China China’s feed-in tariff (FiT) policy came into effect in July 2011, with the NDRC announcing that solar projects approved prior to July 2011 (and completed by end-2011) would enjoy a new benchmark tariff of CNY1.15/kWh, while solar projects approved after July 2011 (or approved before July 2011 but completed after end-2011) would be entitled to a FiT of CNY1.0/kWh. Following a decline in solar project construction costs, as well as the asymmetric development of solar power between northern and southern China, the NDRC announced updates to the country’s FiT policy in August 2013. The new policy divided China into 3 zones, depending on the solar irradiation of the respective regions, with the FiT of Zone I being lowered from CNY1.0/kWh to CNY0.9/kWh, that for Zone II being lowered to CNY0.95/kWh, and that for Zone III remaining at CNY1.0/kWh then. The table below shows the FiT rates in detail by region.

We look for a solar power tariff cut starting in 2H16

10


China: solar feed-in tariffs (September 2013)

Utility zone Benchmark on-grid tariff

(CNY/KWh, VAT incl.) Regions

Zone I 0.9 Ningxia Qinghai: Haixi Gansu: Jiayuguan, Wuwei, Zhangye, Jiuquan, Dunhuang, Jinchang Xinjiang: Hami, Tacheng, Altay, Karamay Inner Mongolia: All except those included in Type II

Zone II 0.95 Beijing Tianjin Heilongjiang Jilin Liaoning Sichuan Yunnan Inner Mongolia: Chifeng, Tongliao, Xing'anmeng, Hulunbeier Hebei: Chengde, Zhangjiakou, Tangshan, Qinhuangdao Shanxi: Datong, Shuozhou, Xinzhou Shaanxi: Yulin, Yan'an Qinghai: All except those included in Type I Gansu: All except those included in Type I Xinjiang: All except those included in Type I

Zone III 1.0 All except those included in Type I or Type II Distributed - Subsidy of CNY0.42/KWh for self-consumption

Extra electricity shall be acquired at local thermal on-grid tariff level by the grid companies

Source: NDRC

Indeed, it is not unusual for China to adjust the renewable energy FiT. Apart from the FiT cut for solar energy during 2014, the FiT has also been adjusted down for wind power (to be implemented in 2016). According to our checks with various wind operators, most companies have pursued aggressive capacity-addition plans in 2015 in order to avoid the FiT cut in 2016, which has led to rush installations this year.

China: wind tariff cuts starting January 2016 Benchmark on-grid tariff (CNY/KWh, VAT incl.) Utility zone Regions New policy Current policy Changes (%)

Type I Inner Mongolia: All except those included in Type II

0.49 0.51 -3.9% Xinjiang: Urumqi, Yili, Changji, Klamyi, Shihezi

Type II Hebei: Zhangjiakou, Chengde

0.52 0.54 -3.7% Inner Mongolia: Chifeng, Tongliao, Xing'anmeng, Hulun Buir Gansu: Zhangye, Jiayuguan, Jiuquan

Type III

Jilin: Baicheng, Songyuan

0.56 0.58 -3.4% Heilongjiang: Jixi, Shuangya, Qitaihe, Suihua, Yichun, Daxing'an moutain Gansu: All except those included in Type II Xinjiang: All except those included in Type I Ningjia: All

Type IV All except those included in Type I, Type II or Type III 0.61 0.61 0.0%

Source: NDRC, Daiwa

One trigger for a FiT cut, in our view, is when there is increased growth in installations in a particular year, which can be viewed as a sign of improving economics for renewable power projects, and which in turn prompts the government to look for a FiT adjustment in the next year. Indeed, the 20GW-plus addition in wind power’s connected capacity during 2014 (up 35% YoY) led to the regulator deciding to cut wind power tariffs. We expect a similar scenario to play out for solar in 2H16, ie, the 50% YoY growth in solar installations that we forecast for 2015E will likely prompt the regulator to lower the solar FiT. Our checks with the solar IPPs suggest that utility-scale solar projects have become very profitable, with equity IRRs of more than 15%, given rapidly declining costs in recent years.

Wind operators have accelerated capacity additions to avoid coming wind tariff cuts

11


China wind: grid-connected capacity additions China solar: capacity additions

Source: CEIC, Daiwa Source: NEA, Daiwa

Potential rise in renewable energy surcharge would be a positive catalyst As mentioned, the Renewable Energy Development Fund is currently suffering a shortfall of around CNY14bn as at end-2014, which we estimate to grow to CNY50bn by the end of 2015. This shortfall has affected the solar farm operators, in that the subsidy payments (accounting for 58% of their revenue) due to them have been delayed by 9-12 months, which has impacted their working capital and cash flow. To make up for the shortfall, we expect the NDRC to increase the renewable energy surcharge levied on customers in 2016 to ~CNY0.025/kWh, from CNY0.015/kWh currently. An increase in the surcharge should accelerate subsidy payments from the government to the operators, and in turn improve the solar-farm operators’ cash flow and ultimately lead to better sentiment toward solar project investment. Before solar power generation achieves grid parity in China, solar development in the country will still rely on government subsidies. Given that subsidies still account for more than half of the income for solar PV projects, the ability and timeliness of the government paying the subsidies are crucial to the profitability and cash flow of the solar farm projects. Against this backdrop, we analyse whether the current subsidy system can support the rapid development of the solar industry. The Renewable Energy Development Fund In China, the renewable energy subsidy system is operated through the Renewable Energy Development Fund (renewable fund). The renewable fund was established in 2006 by the Ministry of Finance. According to the “Temporary Measures on the Collection of Renewable Energy Development Fund” (Chinese-only website: http://www.nea.gov.cn/2011-12/20/c_131316289.htm) published in November 2011, the renewable fund would aggregate the 2 main funding channels into 1 pool: 1) an annual special purpose fund, or the Renewable Energy Development Special Fund, arranged by the Ministry of Finance, mainly supporting the technology side of the renewable energy development, and 2) the surcharges collected from retail electricity sales, mainly supporting the renewable energy power generation FiT. According to the policy, the 2 sources of funds complement each other, so that if one source is depleted to support its purposes, the other source could help support the shortfall.

(20%)

0%

20%

40%

60%

80%

100%

120%

140%

0

5,000

10 ,000

15 ,000

20 ,000

25 ,000

2006 2007 2008 2009 2010 2011 2012 2013 2014

(MW)

Connected capacity additions (MW) (LHS) YoY growth (RHS)

2014 high growth triggered FiT cut starting in 2016

-50%

0%

50%

100%

150%

200%

250%

300%

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

2010 2011 2012 2013 2014 2015E

(MW)

Capacity additions (MW) (LHS) YoY growth (RHS)

50% YoY growth in 2015E should trigger FiT cut in 2H16

http://www.nea.gov.cn/2011-12/20/c_131316289.htm

12


Renewable Energy Development Fund structure

Source: Daiwa

A few points to note about the renewable fund: • Renewable Energy Surcharge (surcharge): Since 2006, the Ministry of Finance has

imposed a surcharge on retail electricity sales. The surcharge is mainly used to pay for the subsidy portion of the on-grid tariff. Given the rapid renewable energy development in China, the surcharge has been raised several times from CNY0.2¢ in 2006 to CNY1.5¢ in 2013.

Renewable energy surcharge

Source: NDRC

• FiT components: The renewable energy FiT can be broken down into 2 parts. The first

part is the local coal-fired on-grid tariff, with the second part being the subsidy, which is the difference between the FiT and the local coal-fired on-grid tariff. The grid companies settle the coal-fired on-grid tariff with the renewable energy project owners, whereas the Ministry of Finance pays for the subsidy portion of the tariff.

CNY50bn shortfall expected for the renewable fund However, we believe the renewable fund has not been able to support new project development since it was established in 2013, even in 2015, with the cumulative shortfall expected to grow, from CNY10.4bn in 2014 to CNY50bn by end-2015E, assuming the current renewable energy surcharge is unchanged.

Electricity price + renewable energy surcharge

Renewable energy surcharge

Subsidy Subsidy Subsidy Subsidy

Desulfurized and denitrated coal-fired tariff

Retail electricity customers

Grid companies

Renewable Energy Development Fund

Solar farms Wind farmsBiomass power

plantsOther renewable energy projects

Ministry of Finance

Annual special-purpose funds

0.000

0.005

0.010

0.015

0.020

0.025

0.030

Jul-0

6

Jan-

07

Jul-0

7

Jan-

08

Jul-0

8

Jan-

09

Jul-0

9

Jan-

10

Jul-1

0

Jan-

11

Jul-1

1

Jan-

12

Jul-1

2

Jan-

13

Jul-1

3

Jan-

14

Jul-1

4

Jan-

15

Jul-1

5

Jan-

16

Jul-1

6

Jan-

17

Jul-1

7

(CNY per kWh)

We expect renewable fund shortfall to reach CNY50bn by end-2015

13


We attribute this shortfall to 2 factors: 1) renewable energy capacity expansion has far outpaced power consumption growth in China over the past 2 years, leading to demand for subsidies growing much faster than the surcharges sourced from the electricity customers, and 2) the surcharge collection rate is very low (at only around 65% on our estimates), meaning that many electricity customers have not been paying the surcharges, or the surcharges have not been able to reach the fund for various reasons such as bureaucratic inefficiencies. The impact on the downstream renewable energy operators is such that many of the projects have only been able to receive the coal-fired on-grid tariff portion of the subsidy. Actually, if we look at the renewable energy operators’ accounts receivable from 2010 onwards, it is clear that most have seen a surge in their accounts receivable days, indicating a worsening cash flow situation for most companies.

Renewable energy operators: A/R days trends Components of FiT for renewable energy power generations (as of 2014)

Source: Companies, Daiwa Source: NDRC, Daiwa

The situation is especially serious for solar projects, given that 58% of their cash flow comes from subsidies. Several options for the regulator We believe the current situation is not sustainable, as we have already seen the pace of solar installations slow in 2014, even amid strong government support, which we attribute mainly to poor cash flow for the solar projects. We expect the government to implement solutions soon to lower the fund deficit, and see several possible measures as outlined below. 1. Increasing the renewable energy surcharge We believe raising the renewable energy surcharge is the most straightforward solution. The surcharge has been raised about every 1-2 years since 2006 (see table below), and as such, we see a good chance that another hike will be introduced in the near term. Renewable energy surcharge change Date of announcement Surcharge (CNY per kWh)Jul-06 0.001July-08 0.002Nov-09 0.004Jan-12 0.008Sep-13 0.015Source: NDRC

Our analysis shows that the surcharge would need to be raised to ~CNY0.025/kWh in 2016 in order to stop the fund deficit increasing. We think the surcharge could be raised further to ~CNY0.03/kWh by 2017E and stay at that elevated level until 2020E, against the backdrop of an increasing renewable energy contribution within China’s power generation portfolio.

0

50

100

150

200

250

300

1H10 2H10 1H11 2H11 1H12 2H12 1H13 2H13 1H14 2H14

Longyuan Huaneng RE

Datang RE China Power New Energy

58% 51%

28% 33%

42% 49%

72% 67%

0%

20%

40%

60%

80%

100%

Solar (utility-scale) Solar (DG) Wind Biomass

Subsidy Coal-fired on-grid tariff

Many solar farm operators are suffering cash flow problems due to delayed subsidy payments

14


Average renewable energy surcharge forecasts (assuming no special funds from MoF)

Source: NEA, MoF, NDRC, Daiwa

Refer also to our detailed workings in Appendix VI. We believe the regulator’s only concern would be that a hike in the surcharge might further impact the profitability of power-intensive industries (such as basic materials and manufacturing industries), while most are already suffering from thin margins amid overcapacity. That said, we believe the recent coal-fired on-grid tariff cut should provide room for a renewable energy surcharge hike. China: coal-fired on grid tariff

Source: NDRC, Daiwa

2. Lowering feed-in tariffs A FiT cut actually correlates more with the reduction in solar installation costs, as the government cannot cut the tariffs simply to reduce the fund deficit. Given that we expect to see a continuous reduction in renewable energy installation costs from 2015-20, we believe FiT cuts are inevitable going forward. For the wind power sector, the NDRC has already announced a FiT cut from January 2016 onwards, with the FiT for Type I to Type III regions declining by ~CNY0.02 per kWh, while the FiT for Type IV regions will remain constant (see the following wind tariff table).

Wind tariff cuts starting January 2016 Benchmark on-grid tariff (CNY/KWh, VAT incl.) Utility zone Regions New policy Current policy Changes (%)

Type I Inner Mongolia: All except those included in Type II

0.49 0.51 -3.9% Xinjiang: Urumqi, Yili, Changji, Klamyi, Shihezi

Type II Hebei: Zhangjiakou, Chengde

0.52 0.54 -3.7% Inner Mongolia: Chifeng, Tongliao, Xing'anmeng, Hulun Buir Gansu: Zhangye, Jiayuguan, Jiuquan

Type III

Jilin: Baicheng, Songyuan

0.56 0.58 -3.4% Heilongjiang: Jixi, Shuangya, Qitaihe, Suihua, Yichun, Daxing'an moutain Gansu: All except those included in Type II Xinjiang: All except those included in Type I Ningjia: All

Type IV All except those included in Type I, Type II or Type III 0.61 0.61 0.0%

Source: NDRC, Daiwa

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

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

(CNY cents per kWh)

Average renewable energy surcharge

0.320.330.340.350.360.370.380.390.400.410.420.43

Jun-08 Aug-08 Nov-09 Jun-11 Dec-11 Dec-12 Sep-13 Aug-14 May-15

(CNY/kWh)

Renewable surcharge hike could impact the profitability of the power-intensive industries

15


As for the solar PV industry, we have observed a decline in installation costs, from ~CNY9-10 per W in 2013 to ~CNY8-9 per W in 2015, driven mainly by a decline in solar module ASPs. As such, we anticipate a FiT cut for solar power generation effective by 2H16. 3. Increasing the collection ratio As explained above, we believe the current issues surrounding the renewable energy surcharge collection ratio is that many electricity customers have not been paying the surcharges, or the surcharges paid have not been able to reach the fund. For example, based on China’s power consumption in 2014, excluding some exemption from agricultural electricity usage, the government should be able to collect ~CNY71bn in renewable energy surcharges, while in reality we estimate that the surcharge fees collected amounted to only ~CNY46bn, implying a collection ratio of just 65%. Renewable energy surcharge collection ratio

Source: Daiwa

Given that neither the balance of the renewable fund nor the collection status of the surcharges is transparent, we refrain from speculating on the reasons behind the low collection ratio, beyond bureaucratic inefficiencies. We also believe that it might take a relatively long time for any “inefficient process” to be improved. As such, we believe a collection ratio improvement could only happen gradually over an extended period of time, even if any improvement policy were introduced today. 4. One-off subsidy from MoF Another option open to the government to help reduce the renewable fund deficit is for the Ministry of Finance (MoF) to grant a one-off subsidy to the renewable fund to narrow the gap between the amount of subsidies it offers and the surcharges it collects. In fact, this is not a new measure, as the MoF has granted 2 special subsidies previously, namely ~CNY8.6bn in 2012 and ~CNY14.8bn in 2013, in order to reduce the shortfall. We would not be surprised if the government made another grant to the renewable fund. However, we would point out that under renewable funding policy, the regulator states that the additional cost of renewable energy power generation should be evenly borne by the electricity users, and as such “consistent” one-off subsidies is not a sustainable solution.

Solar supply chain set to ride the sector upcycle An installation boom in China from 4Q15 to 2016 would augur well for the solar supply chain, ranging from polysilicon in the upstream sector to solar modules and solar glass in the mid-stream. Polysilicon: ASP should recover to USD18.0/kg by end-2015E We anticipate a diverging trend between the international and Chinese markets. While the international market continues to suffer from worsening supply-demand, for which international polysilicon prices should remain depressed at ~USD15.0/kg, we believe

0%

10%

20%

30%

40%

50%

60%

70%

80%

0

10

20

30

40

50

60

70

80

2009 2010 2011 2012 2013 2014

(CNY bn)

Surcharge eligible for collection (LHS) Actual surcharge collected (LHS) Collection ratio (RHS)

China’s renewable surcharge collection ratio was only ~65% in 2014

We expect polysilicon prices to recover to USD18/KG by end-2015

16


China’s foreign import restrictions will protect the local market from oversupply. Combining this with a likely installation boom from 4Q15, we forecast polysilicon prices to gradually recover, with the ASP trending toward USD18.0/kg by end-2015. Polysilicon price trend forecasts

Source: Solarzoom, Bloomberg, Daiwa

Worsening supply-demand situation globally, driven by rapid capacity expansion … The main driver of the international polysilicon market, in our view, is the supply-demand situation. While solar demand is in a booming stage given the increased awareness of environmental protection, as well as the falling PV construction costs, we are concerned that the aggressive capacity expansion plans of the tier-1 polysilicon manufacturers will outpace PV demand growth. On the demand side, as explained in the previous section, we look for c.53GW/60GW of PV installations in 2015/16E, respectively, representing 35%/13% YoY growth from 39.4GW in 2014. We see China, with rush installations in 2016 before a potential on-grid tariff cut in 2H16, and the US with rush installations in 2016 to avoid the Investor Tax Credit (ITC) cut from 30% to 10% starting 2017, continuing to be the growth engines until end-2016, with India and other developing countries catching up further ahead. Global solar installation estimates

Source: BP, Daiwa

On the supply side, however, we expect the production volume to outpace PV demand growth, given the polysilicon manufacturers’ aggressive capacity expansion schedules. We expect the polysilicon production volume to increase by 47,000 tonnes/68,450 tonnes for 2015 and 2016, respectively, representing 16%/19% YoY growth. We believe this surge in production capacity is due to the relatively balanced supply-demand situation in 2013, when loss-making companies declared bankruptcy, hence limiting supply, with healthy PV demand growth of over 20% leading to stabilised and slightly increased polysilicon pricing. Given the 2-year lead time for polysilicon manufacturing plant construction, most of the tier-1 polysilicon manufacturers who announced capacity expansion plans in 2013 will see their new plants commissioned over

10

12

14

16

18

20

22

24

Jan-13 Apr-13 Jul-13 Oct-13 Jan-14 Apr-14 Jul-14 Oct-14 Jan-15 Apr-15 Jul-15 Oct-15

(USD/kg)

China local China Imported International

China local (forecast) China Imported (forecast) International (forecast)

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

2010 2011 2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(MW)

China United States Japan Europe India Rest of world

We look for a surge in polysilicon production capacity in 2015-16 caused by the balanced supply-demand situation in 2013

17


2015-16.

Global: polysilicon supply-demand Tier-1 manufacturers’ capacity vs. PV demand

Source: Company data, Daiwa Source: Company data, Daiwa

As a result, we expect a huge amount of capacity to come online in 2015-16. GCL-Poly plans to add 25,000 tonnes through its fluidized bed reactor (FBR) plant in 2016, and Germany’s Wacker is also adding 20,000 tonnes through its new plant in Tennessee. Besides, Korea’s OCI’s P3.9 debottleneck project should unlock around 10,000 tonnes of capacity, with the US’s REC expanding mainly through debottlenecking and its joint venture, with 19,000 tonnes of capacity in China.

Capacity expansion plans for global polysilicon manufacturers Company Capacity(tonnes) Timeline Details OCI 10,000 1Q15 P3.9 debottleneck expected to be completed in 3Q15 Qstec 8,000 1H15 JV with Solarworld with 8,000 tonnes started in 1H15 Daqo 6,000 2Q15 New plant in Xinjiang with 6,000 tonnes capacity fully ramped up in 3Q15; Phase III with 5,850 tonnes is scheduled to be commissioned by

early-2017 Wacker 20,000 2H15 The new plant in Charleston, Tennessee, will commence operations in 2H15 Tokuyama 13,800 1Q16 PS-2 commenced production in 4Q14 and will fully ramp up in 1Q16 Hanwha 3,000 1H16 Debottlenecking expected to be completed in 2H15 SunEdison (MEMC) 13,500 1H16 JV with Samsung with 10,000 tonnes production capacity starting in 2H15, will ramp up to 13,500 tonnes by 1H16 IDEA Polysilicon 10,000 2Q16 Located in Saudi Arabia with 10,000 tonnes production capacity. Expected to completed by 2Q16 GCL-Poly 25,000 2H16 FBR plant in commercial production since 1H15; will ramp up to full capacity into 2016 REC 19,000 1H17 JV company with Shaanxi Non-ferrous Metals established CSG Holdings 6,000 1H17 Capacity expansion plan expected to complete in 2017, with 2,500 tonnes capacity for the semiconductor grade Source: Daiwa

Polysilicon manufacturing capacity expansion by company (tonnes) 2015E 2016E 2017E

OCI 10,000 - - Qstec 8,000 - - Daqo 6,000 6,150 5,850 Wacker 10,000 10,000 - Tokuyama - 13,800 - Hanwha - 3,000 - SunEdison (MEMC) 10,000 3,500 - IDEA Polysilicon - 10,000 - GCL-Poly 3,000 22,000 - REC - - 19,000 CSG Holdings - - 6,000 Total 47,000 68,450 30,850 % production volume in previous year 16% 19% 8%

Source: Company, Daiwa Note: 2016 capacity expansion will have full impact on 2017

… but China should be more immune to a supply shock In contrast to our expectation for continued pricing pressure on polysilicon globally, we believe the Chinese poly market will be relatively less affected due to: 1) the booming Chinese PV market lifting demand from 10.3GW in 2014 to 16-18GW in 2015-16E, 2) Chinese anti-dumping and countervailing tariffs on foreign polysilicon imports, as well as the ban on processing trade, should limit polysilicon supply in China.

(40%)

(30%)

(20%)

(10%)

0%

10%

20%

30%

40%

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)

Polysilicon production (LHS) PV demand (LHS) % oversupply

Capacity expansion decision made

0%

50%

100%

150%

0

20,000

40,000

60,000

80,000

2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)


China announced anti-dumping and countervailing tariffs for polysilicon imports in January 2014

18


China announced anti-dumping and countervailing tariffs on US and Korean polysilicon manufacturers in January 2014, and European polysilicon manufacturers in April 2014, with US/Korea/EU manufacturers subject to 53.6-57%/ 2.4-48.7%/14.3-42% AD duties, respectively. US and EU manufacturers are also subject to countervailing duties ranging from 0-2.1% and 1.2%, respectively. Anti-dumping and countervailing duties imposed on major polysilicon manufacturers Anti-dumping duties Countervailing dutiesUnited States REC Silicon 57.00% 0.20%Hemlock 53.30% 2.10%MEMC 53.60% 0.00%South Korea OCI 2.40% 0.00%Hankook 2.80% 0.00%KCC 48.70% 0.00%Others 12.30% 0.00%Europe Wacker 14.30% 1.20%MEMC SpA 42.00% 1.20%Others 14.30% 1.20%

Source: Ministry of Commerce

Since the announcement of these tariffs, many foreign manufacturers have avoided anti-dumping and countervailing duties by importing polysilicon through “processing trades”. Subsequently, on 14 August 2014, China’s Ministry of Commerce and the Customs announced a policy (58th article) to temporarily suspend the import of solar grade polysilicon through processing trade starting in September 2014. The policy gave the importers a 2-week grace period such that the processing trades approved before September 2014 could still be executed until the contract term was complete. This policy triggered a surge in processing trade approvals during the 2-week period, with industry associations estimating that the approvals during that 2 weeks amounted to 100,000 tonnes, nearly on a par with the total import amount of 102,000 tonnes in 2014. We believe overseas rush orders was the main reason leading to the slump in polysilicon prices in 1H15. To our understanding, most of the processing trade contracts have a term of 1 year, which implies that the majority of the orders will expire by August 2015. We estimate that more than half of the rush orders have been executed already, as import volume between September 2014 and April 2015 already amounted to 66,203 tonnes. Import volume from September 2014 to April 2015 Import volume (MT)Korea 30,700US 14,320Germany 21,183Total 66,203

Source: The Customs

We believe that once those rush orders have been executed, imports from the US and South Korea should decline, releasing 20,000 tonnes sales volume to Chinese local manufacturers (with Wacker and OCI being less affected given their low anti-dumping tariffs).

China’s polysilicon prices were impacted by “processing trades” imports from foreign countries

19


Polysilicon supply in China by country in 2014

Source: The Customs, Company data, Daiwa

Scenario analysis on utilisation rate of tier-1 manufacturers in China (tonnes)

China’s annual PV installation 16GW 17GW 18GW

Local poly production in 2014 121,500 121,500 121,500 Poly imports in 2014 102,177 102,177 102,177 Total poly demand in 2014 223,677 223,677 223,677 2014 Local production capacity 132,500 132,500 132,500 % utilisation 92% 92% 92%Incremental PV demand (GW) 5.7 6.7 7.7 Poly demand per MW 5.3 5.3 5.3 Implied additional poly demand from 2014 30,210 35,510 40,810 Capacity released from imports 30,000 30,000 30,000 Local poly production in 2015 181,710 187,010 192,310 Capacity addition by tier-1 manufacturers 47,000 47,000 47,000 2015E Local production capacity 179,500 179,500 179,500 % utilisation 101% 104% 107%Source: Daiwa Note: 2014 solar installation was 10.3GW Note: temporary shortage in 2015E

As we expect China’s PV installation to surge from 10.3GW in 2014 to 16-18GW in 2015-16, this should increase polysilicon demand by 30,210-40,810 tonnes, on our estimates. This, coupled with the capacity released from import decline and partially offset by capacity expansion plans locally, lead us to expect all tier-1 manufacturers to operate at full capacity by end-2015. Please find our scenario analysis above. With our utilisation rate forecast of around 100% close to the end of 2015E, we expect poly pricing to trend upward eventually until more tier-2 capacities come in to make up for the excess demand. That said, we do not expect poly prices to experience a sharp recovery to USD21 per kg or above, as witnessed in late 2013 to 2014. In China, there is already a huge amount of idle capacity which was shut down in 2012 due to sustained depressed poly prices. As such, any price recovery to a sustained USD21 per kg or above could trigger resumption of those idle capacities, which would act as a price ceiling for poly prices in China. Hence, we estimate a more reasonable price range would be USD17.5-21 per kg.

121,500

35,743

30,235 21,133 8,599 6,466

0

50,000

100,000

150,000

200,000

250,000

Poly supply

(MW)

China Korea Germany US Taiwan Others

We expect tier-1 polysilicon manufacturers to operate at full utilisation by end-2015

20


Global polysilicon cash cost curve Global polysilicon production cost curve

Source: Daiwa estimates Source: Daiwa estimates

Solar modules: shipment volume likely to surge with expanded margins Similar to polysilicon, solar module manufacturers should also benefit from a surge in sales volume and margin expansion. For China’s tier-1 module manufacturers, we forecast aggregate shipment volume growth of 40%/16% for 2015-16E, mainly benefiting from the solar installation boom in China and the US.

Global solar installations Module manufacturers: shipment volume

Source: BP, NEA, Daiwa Source: Company, Daiwa

Also, if we look at the gross margin and net margin trend of the solar module manufacturers, we can see improving margins across the board since 2015. We expect continued margin improvement for the rest of 2015, as well as going into 2016. We look for solar module prices to remain relatively stable at ~USD0.64/W, while further margin expansion should come from production cost reductions.

Module manufacturers: gross margin trend Module manufacturers: net margin trend

Source: Companies, Daiwa Source: Companies, Daiwa

0

5

10

15

20

25

- 50 100 150 200 250 300 350

(US$/kg)

MT (‘000)

Total demand in 2015E: 294,404MT

Implied poly price: $17.5/kg

0

5

10

15

20

25

- 50 100 150 200 250 300 350

(US$/kg)

MT ('000)

Total demand in 2015E: 294,404MT

Implied poly price: $21.0/kg

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

2010 2011 2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(MW)


-20%

0%

20%

40%

60%

0

10,000

20,000

30,000

40,000

2011 2012 2013 2014 2015E 2016E 2017E

(MW)

Yingli Trina Canadian Solar

Jinko Renesola Hanwha SolarOne

JA Solar YoY growth (RHS)

(30%)

(10%)

10%

30%

50%

1Q10

2Q10

3Q10

4Q10

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

Yingli Trina JinkoCanadian Solar JA RenesolaHanwha Q Cell

(40%)

(30%)

(20%)

(10%)

0%

10%

20%

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15


Solar module ASPs should remain stable going into 2016

21


Solar module ASP Solar module production costs

Source: Bloomberg, Daiwa Source: Company, Daiwa

0.40

0.50

0.60

0.70

0.80

0.90

1.00

1.10

Jan-

13

Apr-1

3

Jul-1

3

Oct

-13

Jan-

14

Apr-1

4

Jul-1

4

Oct

-14

Jan-

15

Apr-1

5

Jul-1

5

Oct

-15

(USD per W)

Solar module (Mono) Solar module (Multi)

Solar module (China)

(45%)(40%)(35%)(30%)(25%)(20%)(15%)(10%)(5%)0%

0.00

0.20

0.40

0.60

0.80

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

(USD per W)

PV module cost (LHS) YoY change (RHS)

22


2017-18: beginning of the retail parity revolution

Having mentioned all of the above favourable changes for the solar DG business, we expect China’s solar industry to still focus on utility-scale solar farms in the short term, where we expect to see a utility-scale solar installation boom in 2015-16E. For solar DG to prevail in China, we believe retail parity is one of the most important criteria. Our analysis below shows that China should attain retail parity starting 2017E, at which time we should see increasing solar DG installations. In terms of financing, we believe a yieldco structure is the most appropriate solution for solar companies, given its advantage of lowering the cost of capital as well as relieving cash flow tightness. As such, we are positive on the growth of yieldcos in the future. However, possible CNY depreciation risks may hinder the growth of yieldcos in the short term, where we might see asset-backed securities (ABS) used as a temporary solution.

Levelised cost of electricity analysis In our view, an appropriate way to measure the cost of solar power generation is to calculate its levelised cost of electricity (LCOE). We would call solar energy reaching grid parity if the LCOE is below the on-grid tariff (without subsidy). However, before reaching grid parity, which we believe is more relevant to utility-scale solar projects than to solar DG projects, as they are paid the on-grid tariff for their electricity sold, we would compare the LCOE of a DG project with retail electricity prices, ie, whether the distributed solar has reached “retail parity” yet. Details of our calculations In calculating the LCOE of solar DG projects in China, we first determine the system costs of the project. We can break down the costs of a DG project into 3 categories: 1) module costs, 2) inverter costs, and 3) balance of system costs. 1) PV module costs Following the steep decline in module costs during 2011-12, module costs seem to have moderated recently.

Polysilicon solar module price trend Polysilicon solar module price trend in China

Source: PVinsights Source: pvXchange, Bloomberg, Daiwa

Especially for China, prices rebounded slightly in 1H15 (refer to the above chart). Given the likely scarcity of PV modules during 2H15, we believe module prices should be sustained at ~CNY3.8-4.0 per W for 2015. 2) Inverter costs Similar to PV module prices, inverter prices are also on a downward trend (refer to the below chart). Inverter prices in China are pretty much in line with the global average, which is ~CNY1 per W, as of 2015.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15

(USD per W)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Aug-11 Feb-12 Aug-12 Feb-13 Aug-13 Feb-14 Aug-14 Feb-15

(USD per W)

To assess whether China has reached retail parity, we compared retail electricity prices with the LCOE for solar energy in China

23


Inverter price trend

Source: PVinsights

3) Balance of system costs The balance of system cost includes a range of construction costs other than solar modules and inverters, such as mounting, cabling and grid connections, and other soft costs such as development costs and EPC margins. Balance of system costs (BOS) vary widely across the globe, and even across different regions in China, depending on a number of factors, such as location of the project, distance from the grid, labour wage level, EPC competitive landscape. In China, we estimate that the average balance of system cost for solar DG projects ranges between CNY2.7-3.0 per W. As described above, we estimate that the construction cost of a solar DG project in China is around CNY7.7-8.0 per W. Refer to the cost breakdown table below. Solar DG: construction costs (CNY per W) Cost breakdown %Equipment Solar module 4.0 52%Inverter 1.0 13%Equipment total 5.0 65% BOS Mounting 0.3 4%Development costs 0.4 5%Cable and grid connection 0.5 6%EPC 1.5 19%BOS total 2.7 35%Total 7.7 100%

Source: Daiwa

We also make the following assumptions in our LCOE analysis: Solar DG project: assumptions Assumption Unit ValueCapacity (MW) 5Operation (Years) 20Unit-capex (including VAT) (CNY/W) 7.7Unit-capex (excluding VAT) (CNY/W) 6.6Total investments (including VAT) (CNYm) 39Total investments (excluding VAT) (CNYm) 33Tariff (including VAT) (CNY/kWh) 0.78Tariff (excluding VAT) (CNY/kWh) 0.67Insolation (hours) 1,400System efficiency (%) 78%Utilization hours (hours) 1,092Degradation (1st year) (%) 2.00%

Source: Daiwa

0.0

0.1

0.2

0.3

0.4

Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15

(USD per W)

Inverter - 1-10kW Inverter - 10-30kW Inverter - >30kW

BOS cost reduction should mainly be driven by efficiency improvements

24


Retail tariff assumption Electricity tariffs differ widely between cities in China, with cities adjacent to coal mines and those with lower electricity demand, such as Urumqi, Hohhot and Xining, enjoying electricity tariffs as low as CNY0.51-0.55 per kWh. Conversely, industrial customers in southern China cities such as Chongqing, Wuhan and Shanghai, are charged electricity tariffs as high as CNY0.9+ per kWh. The tariff assumption we use is CNY0.79 per kWh, which is the average industrial electricity tariff of the 36 cities surveyed by the Price Monitoring Centre. Cities with highest and lowest electricity tariff City Province (CNY per kWh)Top-5 electricity tariffs Chongqing n.a. 0.95Wuhan Hubei 0.93Shanghai n.a. 0.91Changchun Jilin 0.9Hefei Anhui 0.89Bottom-5 electricity tariffs Urumqi Xinjiang 0.51Hohhot Inner Mongolia 0.54Xining Qinghai 0.55Haikou Hainan 0.67Lhasa Tibet 0.68

Source: CEIC

Cost of equity assumption We assume the cost of equity of a solar DG project is ~10%, which is the required rate of return of China equities. We believe the market required rate of return is a good proxy given the solar DG projects are owned by a wide array of companies, ranging from large coal miners, manufacturers, financial institutions, to logistics centres. We also cross-check our cost of equity assumption with the average cost of equity of the downstream renewable energy names listed on the Hong Kong Stock Exchange, which is broadly in line with our assumption of 10% (refer to the below table). Renewable energy operators: cost of equity BBG Ticker Beta Risk free rate Country risk premium Cost of equity

686 HK 1.3 1.90% 9.00% 13.20%1165 HK 0.6 1.90% 9.00% 7.50%3800 HK 0.9 1.90% 9.00% 10.10%916 HK 1.0 1.90% 9.00% 11.10%958 HK 0.8 1.90% 9.00% 9.50%1798 HK 0.9 1.90% 9.00% 9.80%816 HK 0.9 1.90% 9.00% 10.00%Average 10.20%


Insolation We acknowledge that insolation in China differs substantially, with the northwestern provinces such as Xinjiang, Gansu, Qinghai, or Inner Mongolia having very strong sunlight (due to fewer rainy days and less pollution) whereas provinces in southern China such as Hunan, Hubei, Guizhou, and Sichuan receive little sunlight (because of more pollution and more rain). In taking the average, we consider mostly the coastal provinces given their more imminent need for renewable energies. For provinces like Jiangsu, Zhejiang, Guangdong and Shandong, insolation is similar at around 1,300-1,400kWh/m2. We conservatively assume an overall system efficiency of ~78%, and arrive at an average utilisation of ~1,100 hours per year.

Retail tariffs diverge to a great extent between cities

25


Financing assumptions The higher portion of debt financing a solar project can get, usually the higher the equity IRR a project can achieve. In China, the maximum portion of debt financing for a solar project is 80%. Given the increasing support from banks for the solar PV industry, we believe it is not difficult for the corporates to be granted an 80% bank loan for a project, which is also our assumption in our LCOE model. In terms of debt tenure, it’s still difficult for a project to get 20 years of financing (which is the most ideal, in our view). For a normal solar PV project, we believe the longest loan a solar project can obtain is 10-12 years. In our model, we assume a project is financed by a 12-year 7% bank loan. Our results By making the above assumptions, we estimate that a 20-year solar DG project has a LCOE of CNY0.85 per kWh, assuming no subsidy at all. Comparing this with China’s average industrial electricity tariff of CNY0.79 per kWh, on average China’s solar DG has not reached retail parity yet.

Commercial electricity tariffs vs. LCOE by province

Source: Daiwa Note: Orange-highlighted ones are the cities that have achieved grid-parity already.

Sensitivity of LCOE on insolation and finance costs China irradiation map

Finance cost (%)

6.0% 6.5% 7.0% 7.5% 8.0%

Inso

latio

n (h

ours

)

1800 0.65 0.67 0.68 0.69 0.70

1600 0.73 0.74 0.75 0.77 0.78

1400 0.82 0.84 0.85 0.87 0.88

1200 0.95 0.96 0.98 1.00 1.02

1000 1.13 1.15 1.17 1.19 1.21

Source: CEIC, Daiwa Source: SolarGIS

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Cho

ngqi

ng

Wuh

an

Shan

ghai

Cha

ngch

un

Hef

ei

Tian

jin

Shen

zhen

Nan

ning

Gua

ngzh

ou

Shen

yang

Nan

jing

Dal

ian

Han

gzho

u

Nin

gbo

Qin

gdao

Beijin

g

Har

bin

Kunm

ing

Shijia

zhua

ng

Fuzh

ou

Xiam

en

Nan

chan

g

Lanz

hou

Jina

n

Zhen

gzho

u

Cha

ngsh

a

Xian

Che

ngdu

Taiy

uan

Yinc

huan

Gui

yang

Lhas

a

Hai

kou

Xini

ng

Hoh

hot

Uru

mqi

(CNY per kWh)

Commercial electricity tariff LCOE

China has not reached retail parity in general yet

26


However, the above chart also tells us that some cities’ solar DG projects have already achieved retail parity, ie, their LCOE is below industrial electricity tariffs even without subsidies. We believe there are 2 major factors leading to the variations between cities. Retail electricity tariff: the higher the retail electricity tariff, the more probable it is that a city can reach retail parity. In the above chart, nearly all cities with retail tariffs above the average LCOE line of CNY0.85/kWh have reached retail parity. The only exceptions, in our view, are Chongqing and Wuhan, which have relatively poor solar insolation. Solar insolation: the higher the solar insolation, the more probable it is that a city can reach retail parity. However, we believe this factor is less important than the retail electricity tariff, as usually those cities have very low retail electricity tariffs as well. Our conclusion From the current costs and retail electricity tariffs, we identify that some cities have already reached retail parity. These cities are mainly characterised by having: 1) high retail electricity tariffs, such as Shanghai, Guangdong, Changchun, Hefei, etc., or 2) very good sunshine with relatively high retail electricity tariffs such as Kunming, Lanzhou. With continuous cost reductions on solar project construction, we expect more and more cities to achieve grid parity over the next few years. We believe that once the majority of the cities have achieved retail parity, China could lift its subsidies for solar DG projects, or at least reduce tariffs. We believe 2017 is possible timing for a subsidy reduction.

Some cities in China have already achieved retail parity

27


China: current retail parity status for 36 cities

Source: Daiwa

Still supported by government policies before retail parity We expect solar DG development to remain slow in China, before retail parity is achieved. Given the government’s increasing emphasis on developing solar DG, we see the country’s supportive policies remaining in place until solar DG projects attract more investment. NEA 15 guidelines for DG in September 2014 In September 2014, the NEA released new supporting guidelines for distributed solar power projects (link to policy in Chinese language only: http://zfxxgk.nea.gov.cn/auto87/201409/t20140904_1837.htm). The NEA reiterated its support for distributed photovoltaic (PV) development in China, and addressed the concerns of some solar farm developers by: 1) allowing distributed PV systems to sell electricity to the grid at the benchmark solar FiT if electricity self-consumption declines or the customer can no longer pay the electricity price, and 2) encouraging solar-financing innovation and urging banks to provide financing to distributed solar projects at more favourable lending rates.

Dalian

Ningbo

Beijing

Changsha

Wuhan

Urumqi

Xining

Yinchuan

Nanchang

NanjingShanghai

Shijiazhuang

Hohhot

Taiyuan

Tianjin

Qingdao

LhasaChengdu

Kunming

Guiyang

Zhengzhou

Harbin

Changchun

Shenyang

HefeiHangzhou

Fuzhou

Lanzhou

Haikou

Xi’an


Almost reached retail parity

Not yet retail parity×

×

×

×

×

×

×

×

×

×

×

×

×

×

×

ShenzhenGuangzhouNanning

Chongqing

Xiamen

×

Jinan

×

China’s government has been very supportive of solar DG since 2014

http://zfxxgk.nea.gov.cn/auto87/201409/t20140904_1837.htm

28


In the table below we summarise the policy details and our take on the impact on China’s distributed solar market.

NEA 15 guidelines for distributed solar systems Policy details Our takes Encourages local government to provide subsidies to DG on top of the national subsidy We expect more local governments to announce additional subsidies to boost DG demand. DG projects can choose between: 1) self-consumption with remaining power sold back to the grid, or 2) selling all electricity to the grid, with tariffs sold to the grid equivalent to the solar FiT for that region.

We believe the concerns about tariff rates sold to the grid are addressed here, which should increase the willingness of companies with lower power consumption to invest in DG.

The solar project can switch from self-consumption to selling full electricity to the grid, should electricity consumption be reduced or electricity buyers default.

We believe this policy should remove DG developers’ concerns about counterparty risk, one of the major concerns previously.

DG projects can be connected to the grid via substations of 35kV or below. This should remove the unnecessary voltage step-up process which reduces investment costs and power losses by solar developers.

The grid should provide electricity usage breakdown such as the amount of electricity consumed locally and amount sold to the grid.

Clearer statistical systems should assure more transparent electricity billing calculations.

Electricity tariffs and subsidies should be paid by the grid on a monthly basis. We believe this shows the NEA’s determination to ensure timely subsidy payments to the solar developers, reducing developers' cash flow risks.

Encourages financial institutions to offer attractive financing rates and innovative financing means to solar developers.

We believe another major concern of DG developers is financing, and this policy goes some ways to resolving this issue. However, we are waiting for more concrete financing policies to be released.

Source: NEA, Daiwa research Strong local government subsidies fuelling capacity growth Local governments’ supportive stance towards distributed solar projects should nurture the development of distributed solar in provinces with subsidies. As such, provinces offering generous local government subsidies should become the growth engine for national capacity additions. See the following tables for a list of local governments currently offering subsidies on top of the national subsidy of CNY0.42/kWh. As many of the above local governments offer generous subsidies to distributed solar projects, we contend that these projects could be at least as profitable as the utility-scale solar farms built in northwestern China (a region characterised by high solar irradiation), if not more profitable. To quantify, we compare the equity IRRs of distributed solar projects in various provinces. We assume a standard unit investment cost of CNY8 per watt, 20 years of operation, a financing cost of 7%, and 70% leverage. Distributed solar: estimated equity IRR

Source: Daiwa research Note: Refer to Appendix III for subsidy details As shown in the chart above, although cities in coastal provinces have lower utilisation rates than those in northwestern provinces such as Xinjiang, and yield lower equity IRRs (around 13% compared with 15% in Xinjiang), local subsidies can significantly boost the investment returns. For example, we estimate that the DG project in Tongxiang City, Zhejiang Province, one of the cities with the most generous subsidies, could yield an equity IRR of 29% over its operating life. Even in Shanghai, where the local government offers a CNY0.25 per kWh subsidy for a period of 5 years, the equity IRR could be as high as 19%, on our estimates.

29%

19% 17% 17% 15% 15% 13%

0%5%

10%15%20%25%30%35%

Tong

xiang

, Zhe

jiang

Shan

ghai

Guan

gzho

u,Gu

angd

ong

Qing

dao,

Shan

dong

Xinji

ang

Jiang

su

Citie

s in S

E Ch

inawi

thout

local

subs

idies

Equity IRR

Some local governments offer very attractive subsidies to solar DG projects

29


IRR analysis: key assumptions Tongxiang Shanghai Guangzhou Qingdao Xinjiang Jiangsu Other SE cities Capacity (MW) 30 30 30 30 30 30 30 Operating years 20 20 20 20 20 20 20 Unit-capex (incl. subsidies & VAT) 6.5 8 8 8 8 8 8 Total investments 195 240 240 240 240 240 240 Tariff (including VAT) 1.26 1.29 1.39 1.31 0.93 1.27 1.22 Utilisation hours 1,050 1,050 1,050 1,100 1,500 1,050 1,050 Degradation (1st year) 2.0% 2.0% 2.0% 2.0% 2.0% 2.0% 2.0% Degradation (2nd year and after) 0.8% 0.8% 0.8% 0.8% 0.8% 0.8% 0.8% Self-use rate 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% Debt 70% 70% 70% 70% 70% 70% 70% Financing cost 7% 7% 7% 7% 7% 7% 7% Operating expenses 15% 15% 15% 15% 15% 15% 15%

Source: Daiwa research In conclusion, we believe that, given sufficient local government subsidies, attractive investment returns from distributed solar projects should drive solar capacity growth to a great extent. When will solar DG projects reach retail parity? In the previous section, we concluded that China’s solar DG projects have yet to achieve retail parity on average (although some cities have already reached parity given their high retail electricity tariff or insolation). So, the next question is, when do we expect China to achieve retail parity? Solar DG project costs should decline by 4-5% YoY from 2016-20E To answer this question, we estimate the future construction-cost trend of a solar DG project in China. We also break down our cost forecasts into 3 parts: 1) solar module costs, 2) inverter costs, and 3) balance of system costs. 1) PV module costs We forecast China’s PV module prices to decline by 5% per year over 2016-17E, and 4% per year over 2018-20E. During 2013-14, PV module prices fell by ~10% YoY (refer to the chart below), due we believe mainly to overcapacity in the industry, as several module manufacturers expanded their production capacity during that period. However, our research suggests that many manufacturers have become more prudent in terms of capacity expansion in 2015. As such, we do not believe the declining ASP of modules will continue to outpace its cost reduction (which we forecast at 6-8% for 2015-17E). As such, we expect the declining ASP to moderate to 4-5 % YoY from 2016E.

PV module price trend Learning curve for solar module

Source: PVinsights Source: Fraunhofer ISE

-20%

-15%

-10%

-5%

0%

5%

10%

0.00.10.20.30.40.50.60.70.8

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

(CNY per W)

PV module ASP (LHS) YoY growth (%) (RHS)

We forecast China’s PV module prices to decline by 5% per year over 2016-17E

30


We cross-checked our estimates with the PV module learning curve (defined as how much production cost reduction can be achieved by the manufacturers when cumulative production volume has doubled). According to industry expert Fraunhofer ISE, from 1980 until now, the learning curve for PV modules has been around 20%, which suggests that PV module prices would decline by ~20% every time their cumulative capacity doubles (the learning curve chart is shown above). From the above, another phenomenon we can spot is that starting from 2012, the PV module ASP actually fell more sharply than the learning curve (S-line falling below the straight line). This price decline exceeding the learning curve was, in our view, due to overcapacity in the PV module sector. That said, from a longer-tem perspective, we do not believe PV module prices will continue diverging from the historical learning curve, ie, we expect a mean reversion ahead. This is suggested by the relatively small changes in PV module prices in 2014 vs. 2013, suggesting that once the market restore its supply-demand equilibrium, prices should again follow the historical learning rate. 2) Inverter costs We project inverter costs to contract by 8% YoY from 2016-18E, and moderate to a 5% reduction for 2019-20E. As of 2Q15, inverter ASPs had declined by 8-11% YoY from 2Q14. We believe the trend will continue for the next 3 years, given that the competition for the global inverter market is still intensifying, especially the fierce competition between the US and the European manufacturers. Further ahead, we also see some Chinese manufacturers, such as Sungrow (300274 CH, CNY28.7, not rated) and KSTAR (002518 CH, CNY28.52, not rated), gradually gaining scale to compete in the global market, and as such expect to see sustained pricing pressure in the inverter sector. 3) Balance of system costs We believe a BOS cost reduction is mainly driven by an increase in efficiency. With solar cell efficiency increasing, the same amount of BOS costs can support more capacity, such as mounting and development costs, can support a large capacity of solar farm, thereby decreasing the per Watt cost. We project BOS cost components, such as mounting development costs, and cable and grid connection, to decline by 2-4% YoY over 2015-20E due to efficiency improvements. Engineering, procurement, and construction (EPC) costs, which benefit from both efficiency improvements and EPC margin contraction, will likely see a larger cost reduction, which we estimate to be ~6-7% YoY for 2016-20E. Please refer to the below table for our solar farm construction cost estimates from 2015-20E.

We project inverter costs to contract by 8% YoY from 2016-18E

We project BOS cost components to decline by 2-4% YoY from 2016-20E, and EPC costs by 6-7% YoY

31


China: solar farm construction cost trend 2015E 2016E 2017E 2018E 2019E 2020E Equipment Solar module 4.0 3.8 3.6 3.5 3.3 3.2 Inverter 1.0 0.9 0.8 0.8 0.7 0.7 Equipment total 5.0 4.7 4.5 4.2 4.1 3.9 Balance of system Mounting 0.3 0.3 0.3 0.3 0.3 0.3 Development costs 0.4 0.4 0.4 0.4 0.3 0.3 Cable and grid connection 0.5 0.5 0.5 0.5 0.5 0.5 EPC 1.5 1.4 1.3 1.2 1.1 1.1 Balance of system total 2.7 2.6 2.4 2.3 2.2 2.1 Total 7.7 7.3 6.9 6.6 6.3 6.0 YoY growth (%) Equipment Solar module -5.00% -5.00% -4.00% -4.00% -4.00% Inverter -8.00% -8.00% -8.00% -5.00% -5.00% Equipment total -5.60% -5.60% -4.80% -4.20% -4.20% Balance of system Mounting -4.00% -4.00% -3.00% -3.00% -3.00% Development costs -4.00% -4.00% -3.00% -3.00% -3.00% Cable and grid connection -2.00% -2.00% -2.00% -2.00% -2.00% EPC -7.10% -7.10% -6.40% -6.00% -6.00% Balance of system total -5.30% -5.30% -4.60% -4.40% -4.30% Total -5.50% -5.50% -4.70% -4.20% -4.20%

Source: Daiwa Retail parity in 2017E From the above solar DG project cost reduction schedule, we ran a scenario analysis on our LCOE model. Keeping our other assumptions, such as retail electricity prices, financing costs, and operating costs, constant, we estimate that China’s DG projects could achieve a LCOE of CNY0.78 per kWh, below the average retail electricity tariff of CNY0.79 per kWh, implying retail parity by 2017E. China: LCOE vs. industrial power tariff (2015-20E)

Source: Daiwa Note: assume industry power tariff remains unchanged

0.60

0.65

0.70

0.75

0.80

0.85

0.90

2015E 2016E 2017E 2018E 2019E 2020E

(CNY per kWh)

LCOE - solar DG Average industrial electricity tariff

We expect solar DG to achieve retail parity in China by 2017E

32


Timeline of 36 cities reaching retail parity 2015E 2016E 2017E 2018E 2019E 2020E Chongqing Not yet Parity Parity Parity Parity Parity Wuhan Not yet Not yet Parity Parity Parity Parity Shanghai Parity Parity Parity Parity Parity Parity Changchun Parity Parity Parity Parity Parity Parity Hefei Parity Parity Parity Parity Parity Parity Tianjin Parity Parity Parity Parity Parity Parity Shenzhen Parity Parity Parity Parity Parity Parity Nanning Parity Parity Parity Parity Parity Parity Guangzhou Parity Parity Parity Parity Parity Parity Shenyang Parity Parity Parity Parity Parity Parity Nanjing Not yet Parity Parity Parity Parity Parity Dalian Parity Parity Parity Parity Parity Parity Hangzhou Not yet Parity Parity Parity Parity Parity Ningbo Not yet Parity Parity Parity Parity Parity Qingdao Parity Parity Parity Parity Parity Parity Beijing Not yet Not yet Parity Parity Parity Parity Harbin Not yet Not yet Parity Parity Parity Parity Kunming Parity Parity Parity Parity Parity Parity Shijiazhuang Not yet Not yet Parity Parity Parity Parity Fuzhou Not yet Not yet Parity Parity Parity Parity Xiamen Not yet Not yet Parity Parity Parity Parity Nanchang Not yet Not yet Parity Parity Parity Parity Lanzhou Parity Parity Parity Parity Parity Parity Jinan Not yet Not yet Not yet Parity Parity Parity Zhengzhou Not yet Not yet Not yet Parity Parity Parity Changsha Not yet Not yet Not yet Not yet Not yet Not yet Xian Not yet Parity Parity Parity Parity Parity Chengdu Not yet Not yet Not yet Not yet Not yet Not yet Taiyuan Parity Parity Parity Parity Parity Parity Yinchuan Parity Parity Parity Parity Parity Parity Guiyang Not yet Not yet Not yet Not yet Not yet Not yet Lhasa Parity Parity Parity Parity Parity Parity Haikou Not yet Not yet Not yet Not yet Not yet Not yet Xining Not yet Not yet Not yet Not yet Not yet Not yet Hohhot Not yet Not yet Not yet Not yet Not yet Not yet Urumqi Not yet Not yet Not yet Not yet Not yet Not yet

Source: Daiwa

2015 retail parity chart 2016 retail parity chart

Source: Daiwa Source: Daiwa





33


2017 retail parity chart 2018-20 retail parity chart


Of course, the above results are highly dependent on our assumptions, and any changes to these assumptions could bring forward or delay the timing of retail parity. Below we present a few possible changes to China’s solar industry which might lead to these changes. Changes in retail electricity tariff In our view, the change in retail electricity tariffs is one of the most important factors determining whether China can achieve retail parity by 2017E. The higher the retail tariff, the more willing the retail customers would be to change from traditional electricity purchases from the grid to solar power generation. According to a survey undertaken by Price Monitoring Centre, the average industrial electricity tariff has only increased slightly over the past 3 years, probably due to the weak industrial production in China.

China: average industrial electricity tariff China: average industrial electricity tariff (sudden surge in retail tariff)

Source: CEIC Source: Daiwa

Going forward, we expect to see persistently low increments in retail tariffs over the next few years. However, in the case of a sudden surge in retail tariffs by, say, 3-4% next year, retail parity could be achieved one year earlier, ie, in 2016E. However, with the prevailing weak coal prices, we think there could be a drop in on-grid coal-fired tariffs, which in turn would drag down industrial tariffs; under this scenario, retail parity could take 1-2 more years (ie, achieved in 2018/19E).





0%

1%

2%

3%

4%

5%

6%

0.55

0.60

0.65

0.70

0.75

0.80

2007 2008 2009 2010 2011 2012 2013 2014 2015

(CNY per kWh)

Average commercial and industrial power tariff (LHS)YoY changes (RHS)

0.60

0.70

0.80

0.90

2015E 2016E 2017E 2018E 2019E 2020E

(CNY per kWh)

LCOE - solar DGAverage industrial electricity tariff (base)Average industrial electricity tariff (sudden surge)Average industrial electricity tariff (5% drop)

We should mind that changes in retail tariff could affect the timing retail parity

34


China: Qinhuangdao coal (5,500kcal) spot price China: coal-fired on-grid tariff curve

Source: Bloomberg, Daiwa Source: Daiwa

Change in borrowing costs and tenure In our analysis, we assume that borrowing terms remain constant in the future. Currently, the borrowing cost for most solar projects is ~7% on average, with a maximum loan period of ~12 years. That said, we believe the relatively high interest rate and short tenure (12 years versus ~20 years project life) still reflects that commercial banks are comparatively sceptical towards the solar industry. However, after they see the long-term stability of the solar projects, we believe they may start to offer loans to those projects on better terms. As such, we see a chance that the loan terms will be improved in the future, which could further lower the LCOE. We therefore carried out a sensitivity analysis to assess the impact of changes in financing terms on changes in LCOE (as shown in the below tables).

DG project: LCOE changes to changes in financing costs and loan tenure (2016E)

DG project: LCOE changes to changes in financing costs and loan tenure (2017E)

Borrowing costs (%) 6.00% 6.50% 7.00% 7.50% 8.00%

Loan

tenu

re

(yea

rs)

12 0.79 0.80 0.82 0.83 0.84 13 0.78 0.79 0.81 0.82 0.84 14 0.77 0.79 0.80 0.82 0.83 15 0.76 0.78 0.79 0.81 0.82 16 0.75 0.77 0.79 0.80 0.82

Borrowing costs (%) 6.00% 6.50% 7.00% 7.50% 8.00%

Loan

tenu

re

(yea

rs)

12 0.76 0.77 0.78 0.80 0.81 13 0.75 0.76 0.78 0.79 0.80 14 0.74 0.75 0.77 0.78 0.80 15 0.73 0.75 0.76 0.78 0.79 16 0.73 0.74 0.76 0.77 0.79

Source: Daiwa Note: Green-highlighted ones are those scenarios achieving retail parity

Source: Daiwa Note: Green-highlighted ones are those scenarios achieving retail parity

Conclusion In conclusion, we have relatively prudent assumptions for our retail parity analysis (no retail electricity tariff hike, as well as no improvements in financing terms for solar project bank loans), which still lead us to the conclusion that China’s DG projects will reach retail parity by 2017E. If any of the factors (ie, retail electricity tariff hike, improvements in financing terms) change (which is usually a favourable change), the timing of retail parity could even be earlier. Case study: California When talking about retail parity, the most obvious example that investors should refer to is California. The state’s solar energy has been seeing rapid solar capacity growth since 2011, and today it has more installed solar capacity than any other state in the US. In 2014, its annual solar installations amounted to 3.5GW, accounting for 57% of the US’s total solar installation.

0

200

400

600

800

1,000

1,200

Jun-08 Jun-09 Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 Jun-15

(Rmb/tn)

Monthly avg. coal price (LHS) AverageAvg-1SD Avg-2SD

0.2800

0.3300

0.3800

0.4300

0.4800

0

200

400

600

800

1,000

1,200

Jun-08 Jun-09 Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 Jun-15

(CNY/kWh)(Rmb/tn)

Benchmark on-grid tariff (RHS) Monthly avg. coal price (LHS)

If retail electricity tariffs increase in the future, China may reach retail parity even earlier

35


US: solar installation by state US: solar installation proportion by state

Source: SEIA, Daiwa Source: SEIA, Daiwa

Since 2007, California’s solar installations were supported by the state government incentives, namely the California Solar Initiative (CSI), with a generous budget of USD2.167bn until 2016. The CSI offers funding to customers of publicly owned utilities including Pacific Gas and Electric (PG&E), Southern California Edison (SCE), and San Diego Gas & Electric (SDG&E), in addition to federal incentives, to build their own solar PV systems on existing homes and buildings. However, even though the budget for the CSI has nearly been depleted since 2013, California’s solar installations actually experienced explosive growth during that year and the next. The reason for this is because the state had just reached retail parity by 2013. If we refer to the following chart on California’s LCOE vs. retail tariffs, we note that the declining LCOE from 2010-13, which was due mainly to gradually declining solar module costs and BOS system costs, had already fallen below the retail electricity price. This explains why California has experienced such explosive growth since 2013. LCOE vs. retail tariff for selected regions

Source: CleanTechnica We therefore believe that if China’s solar DG LCOE fell below its retail electricity tariff (which means retail parity in China is reached), this would trigger an accelerated pace of growth for China’s solar DG installations. This is also the main theme of our retail parity revolution thesis.

259 577 1,046 2,621

3,549

01,0002,0003,0004,0005,0006,0007,000

2010 2011 2012 2013 2014

(MW)

California North Carolina Nevada MassachusettsArizona New Jersey Others

30% 30% 31%55% 57%

15% 16%12%

5% 4%36% 33% 22% 18% 18%

0%

20%

40%

60%

80%

100%

2010 2011 2012 2013 2014

California North Carolina Nevada MassachusettsArizona New Jersey Others

California’s solar DG installations accelerated even after local subsidies were terminated

36


Power storage could also help in the long run In the future, we expect a sharp decline in the cost of lithium batteries and possible government subsidies for energy storage projects, which would greatly increase the popularity of solar installations, especially for distributed projects. The development of solar power storage could greatly alleviate the problems brought about by the variability and intermittent nature of solar power, enabling the stations to smooth out their output and stabilise their power supply with lower costs and higher efficiency. This would be particularly favourable for distributed solar projects, which are more able to meet the consumption pattern of end users. China’s government has been supportive of the development of energy storage technology. For example, the National Wind and Solar Power Storage Demonstration Project was set up in Zhangjiakou, Hebei during the Golden Sun project period, where a 20MW and 50MW storage station was completed in 2011 and 2014. A storage station was also built in Shenzhen Baoqing as a demonstration project. In various power sector reform documents and the national strategic plan of “Made in China 2025”, power storage is mentioned multiple times as the major focus issue when developing renewable energy and home-grown high-end equipment manufacturing in China. In the future, we expect the government to continue to support the technological development and more popular application of energy storage.

China: major supporting policies for energy storage Time Govt body Policy Details Dec-09 NPCSC Announced "PRC Renewable Energy Law (Amendment)" Stated that energy storage is one of the focus development areas 2011 Multiple Solar and wind power storage demonstration projects in Zhangjiakou (20MW)

and Shenzhen (10MW) First batch of energy storage demonstration projects, mainly for grid supply to industrial users

Mar-11 NDRC Published "Guiding Manual for Industrial Restructuring" Suggested strongly supporting energy storage industries, including large battery technologies and application

Mar-11 NEA Published consultation draft for "Management Guideline for Distributed Power Generation"

Emphasised the importance of energy storage in developing distributed power

Jul-11 MoST Published 12th FYP for "National Scientific and Technological Development" Promised strong support for new energy, smart grid and new energy vehicles Oct-11 Multiple Published the 2011 Guideline for "Supporting Prioritised High-tech Industries" Stated to strongly support development of new energy storage technologies and

their commercialisation May-12 MIIT Published "Market Entry Requirements for Lead-acid Battery Industry" Regulated the lead-acid battery industry in terms of environmental effects Sep-13 State Grid Established the "Technological Standards for Chemical Power Storage System

Application" Set up the national standards for general application of power storage technologies

Nov-14 State Council Published "Action Plan for Energy Development Strategies 2014-2020" For the first time included energy storage in the national energy strategic plan Feb-15 CIAPS / NEA Published "Notice about Major Issues in Energy Storage 13th FYP Planning" Started formulation of the 13th FYP on energy storage Jul-15 NEA Published "Guiding Opinion for Progressing Construction of New Energy Micro-

grid Demonstration Projects" Re-emphasised the importance of energy storage in developing micro-grid

Source: NPCSC, NDRC, NEA, MoST, MITT, State Grid, CIAPS, Daiwa Looking for subsidies to the storage sector Furthermore, we see government subsidies playing a bigger role in accelerating battery installation growth in China, given the government’s enthusiastic support for distributed power shown in previous policies. This could be one important catalyst to encourage more installation of distributed solar power stations. For example, in 2013 the German Government announced a EUR600-660 energy storage installation subsidy given to PV facilities smaller than 30kW, which reduced prices for solar power storage systems by 25% within few months, and greatly shortened the payback periods of distributed solar power stations. More than 4,000 new solar battery systems were installed in the first year of the policy (>40% of growth in total number). Thus, we think future energy storage installation subsidies provided by the Chinese Government could greatly enhance the economics for distributed solar projects. Costs still on a steep downward trend … We also expect the cost of energy storage to drop significantly in the next 5 years. In the past, the cost of energy storage has been a major hindrance to its popular application. For example, in the 2 demonstration projects in Zhangjiakou and Shenzhen, which mainly adopted lithium battery technology, fixed costs of the storage stations are around

We expect China to offer subsidies to power storage in the future

We expect power storage costs to decline rapidly over 2015-20

37


CNY5,000/kWh. On the contrary, the possible tariff income forgone by curtailment is only around half of the storage costs, making curtailment much cheaper than storage today. Nonetheless, according to the New Energy and Industrial Technology Development Organization (NEDO), the cost of lithium batteries could go down by 50% from 2015 to 2020; and from a study done by the Nature magazine, lithium battery costs are decreasing at an unexpectedly rapid rate (-14% CAGR of 2007-14), which has already met the 2020 target set by IEA in 2013. As we see it, when storage costs reach parity with curtailment losses in 2020E, this would trigger explosive installation growth. Lithium battery costs (2005-30E)

Source: www.nature.com

38


Key stock calls

Prefer solar downstream developers We prefer solar downstream developers over upstream manufacturers, given our expectation that China’s policy environment will nurture healthy development of solar projects. The declining polysilicon and solar module prices should also reduce costs for solar PV systems, providing better economics for solar projects. Top pick: China Singyes (750 HK, HKD6.58, Buy) Singyes is involved in both the solar EPC business and solar downstream development business. The company has a core advantage in solar DG projects, and it has a long history of providing EPC services to customers since China’s Golden Sun Programme was implemented in 2009. We reiterate our Buy (1) rating on Singyes. In our view, a solar installation boom in China would benefit Singyes in 2 ways: 1) an accelerating pace of installations would increase Singyes’ EPC volume, and 2) Singyes could also develop more solar farms which would increase its power generation revenue. In longer run, we also see the company benefiting from the solar DG boom, once China approaches and achieves retail parity. We also believe the company will be able to maintain its EPC gross margin at 25-26% over 2015-17E, given that it has applied for permits for most of its EPC projects that yield higher margins. Furthermore, a likely rising solar DG proportion as a percentage of its total EPC volume should offer Singyes a higher margin than that earned on its utility-scale solar projects. Singyes is trading at a 4.9x 2016E PER, which we find appealing, given our 23% EPS CAGR forecast for 2015-17E. The valuation is also at the bottom of its trading cycle (at 1SD below its past-5-year average, and has stayed above that valuation for 80% of the past 5 years). We believe the company should trade at a 6.5x 2016E PER, given the accelerating pace of solar installations that we see in 2016.

Singyes: share price vs. 1-year forward PER bands Singyes: 1-year forward PER bands

Source: Bloomberg, Daiwa forecasts: Source: Bloomberg, Daiwa forecasts

Singyes: share price vs. 1-year forward PBR bands Singyes: 1-year forward PBR bands

Source: Bloomberg, Daiwa forecasts: Source: Bloomberg, Daiwa forecasts

0

5

10

15

20

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(HKD)

Price 4x PER 6x PER8x PER 10x PER 12x PER

2

4

6

8

10

12

14

16

Jan-11 Jan-12 Jan-13 Jan-14 Jan-15

(x)

14.3x Avg+2SD

11.4x Avg+1SD

8.4x Avg

5.4x Avg-1SD

2.5x Avg-2SD

02468

10121416

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(HKD)

Price 0.6x PBR 1x PBR1.4x PBR 1.8x PBR 2.2x PBR

0.50.70.91.11.31.51.71.92.12.32.5

Jan-

11Ma

y-11

Sep-

11Ja

n-12

May-1

2Se

p-12

Jan-

13Ma

y-13

Sep-

13Ja

n-14

May-1

4Se

p-14

Jan-

15Ma

y-15

Sep-

15

(x)

2.3x Avg+2SD

1.9x Avg+1SD

1.4x Avg

1.0x Avg-1SD

0.5x Avg-2SD

We prefer downstream solar operators over upstream and midstream manufacturers, given the absence of overcapacity in the downstream sector. We also like the sector’s more stable net profit stream than the cyclical nature of the manufacturing business.

Singyes is our top pick within China solar sector, on recovering solar EPC volume, as well as its exposure in the solar DG sector

39


JinkoSolar (JKS US, USD25.28, Outperform (2): our preferred pick among the module manufacturers We initiate coverage of solar-module manufacturer JinkoSolar (JKS) with an Outperform (2) rating. JKS is our preferred pick within the solar module manufacturing sector, as: 1) we believe, that as a cost leader, JKS will benefit the most among the China module makers, from the upcoming solar installation boom over 4Q15-2016, and 2) the listing of its downstream solar subsidiary, Jinko Power (JKP), slated for mid-2016, would unlock the value of JKS’s downstream solar business. Improving solar module gross margins: For the reasons mentioned above, we expect JKS’s solar module sales volume to increase from 2.4GW for 2014 to 3.5GW for 2015E and 4.1GW for 2016E, up 44% and 18% YoY, respectively. The strong global demand we expect in 2016 should also lead to JKS’s solar module gross margin expanding slightly from 19.5% for 2015E to 20.3% for 2016E (vs. 15-17% for its peers). Accelerated downstream growth, with limited equity dilution risk: We expect JKS’s downstream solar development to accelerate over the next few years (we forecast ~600MW capacity additions per year over 2015-17), and see minimal risk of equity dilution. This could be achieved through the company’s strategy of developing solar projects through JV with JKP. Potential listing of JKP in mid-2016 should unlock value: We expect JKP’s IPO listing to be completed in mid-2016, with 3 major benefits: 1) value enhancement for JKS, whose fair value we expect to rise by 34% to USD38.1, 2) an improvement in net gearing by end-2016, from 213% pre-IPO to 78% post-IPO, due to the company deconsolidating its capital-intensive solar IPP business, and 3) an immediate non-cash disposal gain of ~CNY86m, on our estimates.

JKS: share price vs. 1-year forward PER bands JKS: 1-year forward PER bands

Source: Bloomberg, Daiwa forecasts Source: Bloomberg, Daiwa forecasts

JKS: share price vs. 1-year forward PBR bands JKS: 1-year forward PBR bands


0

10

20

30

40

50

Dec-1

2

Feb-

13

Apr-1

3

Jun-

13

Aug-

13

Oct-1

3

Dec-1

3

Feb-

14

Apr-1

4

Jun-

14

Aug-

14

Oct-1

4

Dec-1

4

Feb-

15

Apr-1

5

Jun-

15

(USD)

Price 3.5x PER 6.5x PER9.5x PER 12.5x PER 15x PER

02468

1012141618

Dec-1

2

Mar-1

3

Jun-

13

Sep-

13

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

(x)

16.2x Avg+2SD

12.0x Avg+1SD

7.7x Avg

3.4x Avg-1SD

05

10152025303540

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(USD)

PX_LAST 0.4x PBR 0.8x PBR1.2x PBR 1.6x PBR 2x PBR

0.0

0.5

1.0

1.5

2.0

Jan-

11

May-1

1

Sep-

11

Jan-

12

May-1

2

Sep-

12

Jan-

13

May-1

3

Sep-

13

Jan-

14

May-1

4

Sep-

14

Jan-

15

May-1

5

Sep-

15

(X)1.9x Avg+2SD

1.5x Avg+1SD

1.0x Avg

0.5x Avg-1SD

0.0x Avg-2SD

JKS is our preferred top pick among the solar module manufacturers

40


GCL-Poly (3800 HKD, HKD1.66, Outperform (2): a good proxy for the solar sector We initiate coverage of GCL-Poly with an Outperform (2) rating. We like the company’s ample cost-cutting initiatives, which should drive down its polysilicon production costs, and in turn allow the company to enjoy a margin improvement over our forecast period. We also expect its solar farm revenue to see rapid growth, fuelled by the aggressive capacity expansion plan of its subsidiary, GCLNE. Gross margin expansion in sight: Followed by GCL-Poly’s commissioning of its captive power plant in July 2015, we anticipate its production costs to fall to USD12.9/kg for 2017E, down 18% from USD15.9/kg for 2014. This, coupled with the more moderate ASP decline trend we see (0%/3% YoY declines for polysilicon, and 7%/5% YoY declines for wafers for 2016-17E) given the expected strong demand in China (18.2GW/18.0GW solar installations for 2016 and 17E, respectively), leads us to look for a gross margin improvement for its solar materials business, to 26% for 2017E, from 23% for 2014. Disposal of the traditional IPP business by end-2015 should improve its net gearing: We are positive on GCL-Poly’s disposal of its traditional IPP business, and see 3 major benefits: 1) it will give it a clearer business focus on the upstream and downstream solar sectors, 2) improve its net gearing from 145% for 2014 to 112% for 2015E, and 3) provide an immediate after-tax disposal gain of HKD225m. GCLNE likely to be the next earnings growth driver: We see GCL-Poly’s earnings growth from 2017 being fuelled by its downstream solar business through its listed subsidiary GCLNE. With GCLNE’s aggressive capacity addition plans of 2GW, 2.5GW and 3GW for 2015-17E (versus our forecasts for just 1.5GW/1.8GW/2.0GW for 2015-17E), we expect GCLNE to account for HKD1.8bn of GCL-Poly’s net profit for 2017E (55% of total), from nil in 2014. In terms of financing, we expect GCL-Poly’s capex needs to be satisfied by GCLNE. For its short-term capital needs, GCLNE is likely to rely on the issuance of asset-backed securities (ABS), while in the longer term, it has already set up a yieldco with Goldman Sachs Investment Holdings (ownership split 55% GCLNE, 45% others), which management expects to be listed sometime in 2016 or 2017. Thus, we see minimal risk of share dilution for GCL-Poly.

GCL-Poly: share price vs. 1-year-forward PER bands GCL-Poly: 1-year-forward PER bands


0123456

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(HKD)


0

5

10

15

20

25

30

Jan-

14

Mar-1

4

May-1

4

Jul-1

4

Sep-

14

Nov-1

4

Jan-

15

Mar-1

5

May-1

5

Jul-1

5

Sep-

15

(x)

25.9x Avg+2SD

21.0x Avg+1SD

16.3x Avg

11.6x Avg-1SD

6.8x Avg-2SD

We see GCL-Poly as a good proxy for China’s solar sector

41


GCL-Poly: share price vs. 1-year-forward PBR bands GCL-Poly: 1-year-forward PBR bands


0

1

2

3

4

5

6

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(HKD)

Price 0.6x PBR 1x PBR1.5x PBR 2x PBR 2.8x PBR

0.51.01.52.02.53.03.54.04.5

Jan-

11Ma

y-11

Sep-

11Ja

n-12

May-1

2Se

p-12

Jan-

13Ma

y-13

Sep-

13Ja

n-14

May-1

4Se

p-14

Jan-

15Ma

y-15

Sep-

15

(x)

3.3x Avg+2SD

2.6x Avg+1SD

2.0x Avg

1.3x Avg-1SD

0.6x Avg-2SD

42


Xinyi Solar Holdings (968 HK, HKD3.28, Outperform): solar glass giant going downstream Xinyi Solar Holdings (XYS) is the leading solar glass manufacturer, both in China and globally. Its core advantage is its cost leadership, as well as it downstream solar development business which provides the company with long-term stable cash flow. We reiterate our Outperform (2) rating on XYS. We are positive about its solar-glass volume growth, which would be backed by the solar installation boom we expect in 2015-16 We expect the solar-glass gross margin to improve from 31% in 2014 to 37% in 2015E, due to: 1) cost reductions from 2x900tpd production lines, 2) raw material and gas cost declines, and 3) a favourable product mix shift. We look for further cost reductions, such as technical upgrades of its remaining 2x500 production lines in 2016 (increasing capacity for each from 500tpd to 600tpd), and the commissioning of the Malaysia plant, with lower energy and labour costs, targeted to be completed by end-2016. Regarding the company’s solar farm development business, after Xinyi Solar’s (XYS) top-up share placement on 27 October (168.8mn shares at HKD3.22/share), we believe the company’s capital-raising plan for its 2015-16 solar farm development is on track. We are therefore optimistic that XYS can deliver 800MW of new solar farm capacity in 2016.

XYS: share price vs. 1-year-forward PER bands XYS: 1-year-forward PER bands


XYS: share price vs. 1-year-forward PBR bands XYS: 1-year-forward PBR bands


1.01.52.02.53.03.54.04.55.0

Dec-

13Ja

n-14

Feb-

14Ma

r-14

Apr-1

4Ma

y-14

Jun-

14Ju

l-14

Aug-

14Se

p-14

Oct-1

4No

v-14

Dec-

14Ja

n-15

Feb-

15Ma

r-15

Apr-1

5Ma

y-15

Jun-

15Ju

l-15

Aug-

15Se

p-15

Oct-1

5

(HKD)


6.0

11.0

16.0

21.0

26.0

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(x)

22.0x Avg+2SD

18.6x Avg+1SD

15.3x Avg

11.9x Avg-1SD

8.5x Avg-2SD

0.81.31.82.32.83.33.8

Dec-1

3Ja

n-14

Feb-

14Ma

r-14

Apr-1

4Ma

y-14

Jun-

14Ju

l-14

Aug-

14Se

p-14

Oct-1

4No

v-14

Dec-1

4Ja

n-15

Feb-

15Ma

r-15

Apr-1

5Ma

y-15

Jun-

15Ju

l-15

Aug-

15Se

p-15

Oct-1

5

(HKD)

Price 1.5x PBR 2.0x PBR2.5x PBR 3.0x PBR 3.5x PBR

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(x)

3.5x Avg+2SD

3.1x Avg+1SD

2.6x Avg

2.1x Avg-1SD

1.5x Avg-2SD

We also like XYS for its dominating position within solar glass sector, while we expect increasing solar glass demand in 2016

43


Valuation

Current valuations at historical lows; recent reratings still have legs Looking at the following forward PER and PBR charts, we believe the solar sector just embarked on an industry upcycle in September 2015. The sector has gone through 2 deratings since 2014, the first during 2H14 when solar installations in China disappointed investors at a mere 10.6GW, missing the government’s previous target of 14GW by 24%. The second derating happened during June 2015 when China’s A-share market bubble burst, affecting both the H-share market as well as Chinese stocks listed on the NYSE. We believe the sector’s valuation is very attractive at the moment: JKS has returned to 2013 lows, whereas XYS, Singyes and GCL-Poly are all trading at historical trough valuations. These valuations contrast with our expectation that the solar sector is entering another upcycle from 2015-16, triggered by strong China and US solar installations. As such, we believe the recent solar stock rally is just the beginning of the hike, and expect valuations to gradually return to 2014 levels.

Solar sector: forward PER trend Solar sector: forward PBR trend

Source: Company, Daiwa Source: Company, Daiwa

Solar sector: share-price performance chart


0

5

10

15

20

25

30

Jan-

13

Mar-1

3

May-1

3

Jul-1

3

Sep-

13

Nov-1

3

Jan-

14

Mar-1

4

May-1

4

Jul-1

4

Sep-

14

Nov-1

4

Jan-

15

Mar-1

5

May-1

5

Jul-1

5

Sep-

15

(x)


0

1

2

3

4

5

Jan-

13

Mar-1

3

May-1

3

Jul-1

3

Sep-

13

Nov-1

3

Jan-

14

Mar-1

4

May-1

4

Jul-1

4

Sep-

14

Nov-1

4

Jan-

15

Mar-1

5

May-1

5

Jul-1

5

Sep-

15

(x)


0

50

100

150

200

250

300

Dec-1

3

Jan-

14

Feb-

14

Mar-1

4

Apr-1

4

May-1

4

Jun-

14

Jul-1

4

Aug-

14

Sep-

14

Oct-1

4

Nov-1

4

Dec-1

4

Jan-

15

Feb-

15

Mar-1

5

Apr-1

5

May-1

5

Jun-

15

Jul-1

5

Aug-

15

Sep-

15

Oct-1

5

(Rebased to 100)


We think valuations in the China Solar Sector are very attractive at the moment

44


Singyes and JKS’s valuations seem the most attractive Among the 4 solar stocks under our coverage, Singyes has the most undemanding valuations, with both its forward PER and PBR ratios being the lowest among our covered stocks; whereas XYS has the richest valuations, both in terms of forward PER and PBR. When comparing PBR valuation to ROE, Singyes’ valuation is still the most attractive, as it is trading currently at a 0.8x 2016E PBR, while its 2016E ROE remains at elevated at 18%. We think investor concerns about its conventional curtain wall business, as well as the company meeting its solar EPC volume target look overdone, and believe the current valuation provides a decent risk-reward. Among the China solar module manufacturers listed in the US, JKS looks the most undervalued, given that the stock is currently trading at a 0.9x 2016E PBR while offering a 21% 2016E ROE to investors. We believe the company’s superior margins warrant an above-average valuation, and that the stock provides a decent investment opportunity to investors, given it is trading currently at around the average valuation of its peers (ie, JKS’s 2016E PER of 6x vs. peers at 5-8x; JKS’s 2016E PBR of 0.9x vs. peers at 0.4-1.4x).

Solar sector: 2016E PER (x) vs. 2015-17E EPS CAGR (%) Solar sector: 2015E PBR (x) vs. 2015E ROE (%)

Source: Daiwa forecasts for JKS, GCl-Poly, Singyes, XYS; Bloomberg forecasts for the rest Source: Daiwa forecasts for JKS, GCl-Poly, Singyes, XYS; Bloomberg forecasts for the rest

Xinyi Solar

SingyesTrina

Canadian Solar

JKS JA Solar

GCL-PolyDaqoHuadian Fuxin Longyuan

Huaneng RE

-10%-5%0%5%

10%15%20%25%30%35%40%

- 2.0 4.0 6.0 8.0 10.0 12.0 14.0

2015

-17E

EPS

CAG

R (%

)

2016E PER

Xinyi Solar

GCLNEFuxinSingyes

DQ

Huaneng REUnited PVJA Solar

GCL-PolyTrina

Comtec

JKS

Longyuan

CSIQ

0%

5%

10%

15%

20%

25%

- 0.5 1.0 1.5 2.0 2.5 3.0

2016

E RO

E (%

)

2016E PBR

Singyes is trading at the most attractive valuation in our universe, in terms of both PER and PBR; JKS also looks cheap among the solar module manufacturers

45


Solar sector: valuation comparisons (1) Company name Bloomberg code Rating Share price Target price Upside Market cap Daily trade vol (3M) Free float PER (x) PBR (x) ROE (%) EPS CAGR (%)

(local curr.) (local curr.) (%) (USDm) (USDm) (%) 2015E 2016E 2017E 2015E 2016E 2017E 2015E 2016E 2017E 2015-17E

Solar glass manufacturers Xinyi Solar Holdings 968 HK Outperform 3.28 3.50 7% 2,861 12.3 36% 17.2 13.3 10.3 2.9 2.6 2.2 22% 19% 21% 29.1%

CSG HOLDING CO-A 000012 CH NR 9.9 n.a. n.a. 2,616 44.4 81% 27.1 20.5 15.7 2.3 2.2 2.0 8% 11% 13% 0.6% ASAHI GLASS CO 5201 JP NR 745.0 n.a. n.a. 7,262 37.4 90% 19.4 20.9 19.2 0.7 0.7 0.7 4% 4% 4% 47.7% NIPPON SHEET GLA 5202 JP NR 115.0 n.a. n.a. 848 13.9 100% 242.1 62.2 16.1 0.6 0.6 0.5 1% 1% 4% -171.3% SAINT GOBAIN SGO FP NR 38.6 n.a. n.a. 24,049 101.7 81% 15.9 13.7 11.4 1.2 1.1 1.1 7% 8% 9% 14.1% XINYI GLASS 868 HK NR 4.19 n.a. n.a. 2,120 4.8 46% 8.9 7.4 6.0 1.2 1.1 1.0 14% 15% 17% 29.3% Simple average

6,626 55.1 23.0 13.1 1.5 1.4 1.3 9% 10% 11% -8.4%

Weighted average

21.8 16.1 12.9 1.3 1.2 1.1 8% 8% 10% 17.3% Solar farm developers

China Singyes Solar Technologies 750 HK Buy 6.58 8.50 29% 590 5.3 64% 6.3 4.9 4.2 1.0 0.8 0.7 15% 18% 18% 22.0% SOLARCITY CORP SCTY US NR 38.0 n.a. n.a. 3,690 134.7 63% n.a. n.a. n.a. 3.5 2.7 2.6 -86% 395% 63% 15.8% GCL NEWENERGY 451 HK NR 0.6 n.a. n.a. 1,038 1.2 28% 17.6 8.7 5.3 2.1 1.5 n.a. 10% 16% 38% n.a. UNITED PHOTOVOLT 686 HK NR 0.9 n.a. n.a. 526 1.7 60% 12.4 6.8 6.8 0.9 0.6 0.6 9% 10% 9% 115.4% CONCORD NE 182 HK NR 0.6 n.a. n.a. 658 2.0 58% 10.2 7.4 6.1 0.8 0.7 0.7 8% 9% 12% 25.5% Simple average

1,300 11.6 6.9 5.6 1.7 1.3 1.1 -9% 89% 28% 44.7% Weighted average

5.4 3.1 2.4 2.6 2.0 1.7 -44% 230% 45% 22.9% PV module manufacturers

JinkoSolar JKS US Outperform 25.3 28.50 13% 777 22.6 68% 7.3 5.7 5.1 1.2 0.9 0.8 20% 21% 20% 18.2% YINGLI GREEN-ADR YGE US NR 0.7 n.a. n.a. 136 1.1 n.a. n.a. n.a. n.a. n.a. n.a. n.a. -461% 119% n.a. -35.5% TRINA SOLAR-ADR TSL US NR 10.3 n.a. n.a. 947 22.1 n.a. 9.2 7.5 7.3 0.8 0.7 0.6 9% 11% 10% 20.6% CANADIAN SOLAR I CSIQ US NR 20.7 n.a. n.a. 1,134 49.9 75% 8.7 6.5 6.1 1.4 1.1 1.1 17% 20% 10% -6.3% JA SOLAR HOL-ADR JASO US NR 8.3 n.a. n.a. 417 7.2 n.a. 6.7 6.9 5.2 0.4 0.4 0.4 6% 6% 6% 19.9% RENESOLA LTD-ADR SOL US NR 1.3 n.a. n.a. 134 0.7 n.a. n.a. n.a. 132.0 1.2 1.4 1.4 -65% -159% n.a. -129.6% HANWHA Q CEL-ADR HQCL US NR 17.3 n.a. n.a. 1,442 0.5 n.a. 36.9 27.5 n.a. n.a. n.a. n.a. 14% 9% n.a. n.a. SOLARWORLD AG SWVK GR NR 14.8 n.a. n.a. 243 0.9 47% n.a. 28.6 16.8 0.8 0.8 0.6 -9% 4% 5% -160.2% SUNPOWER CORP SPWR US NR 23.8 n.a. n.a. 3,244 35.5 42% 15.0 13.2 12.9 2.1 2.0 2.1 10% 15% 17% 12.1% FIRST SOLAR INC FSLR US NR 49.2 n.a. n.a. 4,960 97.6 73% 16.3 12.7 18.1 0.9 0.9 0.8 7% 6% 3% 1.3% Simple average

1,343 14.3 13.6 25.4 1.1 1.0 1.0 -45% 5% 10% -28.8%

Weighted average

15.6 13.0 12.9 1.1 1.0 1.0 4% 11% 8% 1.4% Poly and wafer manufacturers

GCL-Poly Energy 3800 HK Outperform 1.66 1.85 11% 3,278 15.9 62% 12.7 8.6 7.8 1.2 1.1 1.0 10% 13% 14% 27.5% DAQO NEW ENE-ADR DQ US NR 16.2 n.a. n.a. 170 1.3 n.a. 8.3 4.4 3.3 0.7 0.6 0.5 6% 12% 13% 23.8% OCI CO LTD 010060 KS NR 84,600.0 n.a. n.a. 1,779 15.7 67% 10.3 95.5 30.1 0.7 0.7 0.6 8% 1% 3% -228.2% REC SILICON ASA REC NO NR 1.5 n.a. n.a. 439 5.5 70% n.a. n.a. n.a. 0.4 0.4 0.4 -3% -2% 2% -100.0% TBEA CO LTD-A 600089 CH NR 12.0 n.a. n.a. 6,173 137.0 79% 15.5 12.0 11.6 1.8 1.7 1.5 11% 12% 13% 19.9% COMTEC SOLAR 712 HK NR 0.8 n.a. n.a. 133 0.4 55% n.a. 18.1 13.4 0.7 0.6 0.6 -6% 5% 5% 16.6% Simple average

1,995 11.7 27.7 13.2 0.9 0.8 0.8 4% 7% 8% -40.1%

Weighted average

13.1 23.0 12.8 1.4 1.3 1.2 10% 10% 11% -19.3%

Source: Bloomberg forecasts for non-rated stocks, Daiwa forecasts for rated stocks Note: Pricing as of 27 Oct 2015

46


Solar sector: valuation comparisons (2) Company name Bloomberg code Rating Div yield (%) EV/EBITDA (x) Net debt-to-equity (%) EBITDA margin (%) Net margin (%)

2015E 2016E 2017E 2015E 2016E 2017E 2015E 2016E 2017E 2015E 2016E 2017E 2015E 2016E 2017E

Solar glass manufacturers Xinyi Solar Holdings 968 HK Outperform 2.4% 3.1% 4.0% 14.4 11.6 9.5 13% 60% 85% 35% 38% 42% 26% 25% 26%

CSG HOLDING CO-A 000012 CH NR n.a. n.a. n.a. 13.7 11.2 9.1 n.a. n.a. n.a. 22% 23% 24% 11% 11% 14% ASAHI GLASS CO 5201 JP NR 2.4% 2.4% 2.4% 6.3 6.0 5.9 36% 35% 30% 16% 16% 16% 3% 3% 3% NIPPON SHEET GLA 5202 JP NR n.a. n.a. 0.8% 8.2 7.9 7.3 210% 198% 183% 10% 10% 10% 0% 0% 1% SAINT GOBAIN SGO FP NR 3.3% 3.4% 3.7% 7.7 6.7 6.0 22% 23% 19% 10% 11% 11% 3% 4% 4% XINYI GLASS 868 HK NR 5.1% 6.0% 7.4% 8.6 7.6 6.5 37% 29% 22% 21% 22% 23% 16% 17% 19% Simple average

3.3% 3.7% 3.7% 9.8 8.5 7.4 64% 69% 68% 19% 20% 21% 10% 10% 11%

Weighted average

2.9% 3.0% 3.4% 8.4 7.3 6.5 27% 31% 28% 14% 15% 16% 6% 6% 7% Solar farm developers

China Singyes Solar Technologies 750 HK Buy 1.6% 1.7% 2.0% 5.7 4.9 4.2 71% 69% 59% 19% 19% 20% 9% 10% 11% SOLARCITY CORP SCTY US NR n.a. n.a. n.a. n.a. n.a. 82.7 186% 238% 375% -108% -50% 6% -163% -89% -40% GCL NEWENERGY 451 HK NR n.a. n.a. 0.9% 11.0 5.0 3.1 304% 290% n.a. 48% 57% 62% 14% 18% 21% UNITED PHOTOVOLT 686 HK NR n.a. n.a. 4.3% 5.2 2.9 2.0 134% 154% 213% 79% 80% 81% 18% 23% 23% CONCORD NE 182 HK NR 2.3% 1.8% 3.5% 8.8 6.2 4.0 11% 9% 7% 18% 23% 30% 10% 12% 13% Simple average

2.0% 1.7% 2.7% 7.6 4.7 19.2 141% 152% 164% 11% 26% 40% -23% -5% 5%

Weighted average

0.4% 0.3% 1.0% 3.6 2.1 48.3 172% 201% 236% -44% -9% 25% -87% -43% -15% PV module manufacturers

JinkoSolar JKS US Outperform 0.0% 0.0% 0.0% 4.5 4.2 4.0 205% 213% 203% 15% 17% 20% 6% 7% 7% YINGLI GREEN-ADR YGE US NR n.a. n.a. n.a. 9.0 6.2 9.5 n.a. n.a. n.a. 14% 21% 13% -15% -6% -3% TRINA SOLAR-ADR TSL US NR n.a. n.a. n.a. 6.3 4.9 4.1 81% 111% 156% 10% 12% 14% 4% 4% 4% CANADIAN SOLAR I CSIQ US NR n.a. n.a. n.a. 6.5 4.4 2.7 129% 47% 66% 12% 16% 29% 5% 7% 7% JA SOLAR HOL-ADR JASO US NR n.a. n.a. n.a. 3.8 3.4 3.4 72% 72% 70% 11% 12% 10% 3% 3% 3% RENESOLA LTD-ADR SOL US NR n.a. n.a. n.a. n.a. n.a. n.a. 473% 509% 427% 7% 9% 13% -3% -2% 0% HANWHA Q CEL-ADR HQCL US NR n.a. n.a. n.a. 10.3 9.1 n.a. n.a. n.a. n.a. 11% 10% n.a. 1% 2% n.a. SOLARWORLD AG SWVK GR NR n.a. n.a. n.a. 8.5 5.4 4.4 108% 87% 71% 8% 10% 11% -3% 1% 2% SUNPOWER CORP SPWR US NR n.a. n.a. n.a. 8.4 8.2 7.0 21% 12% 14% 18% 16% 17% 10% 10% 10% FIRST SOLAR INC FSLR US NR n.a. n.a. n.a. 5.9 5.2 6.8 -24% -32% -37% 17% 17% 14% 9% 10% 8% Simple average

0.0% 0.0% 0.0% 7.0 5.7 5.2 133% 127% 121% 12% 14% 16% 2% 4% 4%

Weighted average

0.0% 0.0% 0.0% 6.9 6.1 5.2 34% 24% 26% 15% 15% 15% 7% 8% 7% Poly and wafer manufacturers

GCL-Poly Energy 3800 HK Outperform 5.3% 1.9% 2.6% 4.3 5.1 5.0 112% 170% 189% 33% 43% 46% 9% 8% 10% DAQO NEW ENE-ADR DQ US NR n.a. n.a. n.a. 7.4 5.0 4.1 119% 96% 66% 32% 35% 36% 9% 13% 14% OCI CO LTD 010060 KS NR 0.3% 0.3% 0.3% n.a. n.a. n.a. 66% 69% 70% 15% 18% 19% 8% 1% 2% REC SILICON ASA REC NO NR n.a. n.a. n.a. 13.7 7.9 4.1 8% 3% 9% 11% 15% 27% -11% -7% -2% TBEA CO LTD-A 600089 CH NR 1.7% 2.1% 2.6% 12.5 10.8 9.3 11% 9% -6% 11% 11% 12% 6% 6% 7% COMTEC SOLAR 712 HK NR n.a. n.a. n.a. 8.6 5.5 4.2 13% 20% 35% 13% 16% 16% -2% 2% 3% Simple average

2.4% 1.4% 1.8% 9.3 6.9 5.3 55% 61% 60% 19% 23% 26% 3% 4% 6%

Weighted average

2.4% 1.7% 2.1% 8.3 7.3 6.4 48% 63% 61% 18% 22% 23% 6% 5% 7%

Source: Bloomberg forecasts for non-rated stocks, Daiwa forecasts for rated stocks Note: Pricing as of 27 Oct 2015

47


Risks to our call Grid curtailment the main risk With the rapid growth in installation capacity since 2013, curtailment problems may worsen in the future and reduce margins for the solar farm developers. Currently in some provinces, curtailment rates remain high (Gansu: 40%; Xinjiang: 8%; Yunnan: 4%), hurting the profitability of solar projects in these areas. In the case of lower-than-expected demand growth or slower-than-expected development of solar power transmission infrastructure and utilisation mechanisms, curtailment problems may deteriorate and thus pull down the unit margins of the solar project developers. China: solar curtailment map

Source: Daiwa China: curtailment rate for solar projects (1H15)

Source: China5e, Daiwa Note: PCC: Production and Construction Corps

Curtailment rate

0%0% - 2%2% – 5%5% – 10%> 10%

Beijing

GuangdongGuangxi

Hunan

Hubei

Hainan

Xinjiang

Qinghai Ningxia

Shaanxi

Jiangxi

Jiangsu

Shanghai

Hebei

Inner Mongolia

Shanxi

Tianjin

Shandong

Tibet Sichuan

Yunnan

Guizhou

Henan

Heilongjiang

Jilin

Liaoning

Anhui

Zhejiang

Taiwan

Fujian

Hong KongMacau

Gansu

Western Route

0.7% 1.1% 1.1% 1.7% 2.4% 3.7%7.8% 8.1%

40.0%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

Shaanxi Inner Mongolia Ningxia Qinghai Tibet Yunnan Xinjiang Xinjiang PCC Gansu

Worsening grid curtailment of solar power is our primary risks

48


International disputes a secondary risk Currently in China there are anti-dumping and countervailing tariffs on US, Korean and European polysilicon imports, making locally manufactured solar cells and modules more competitive to meet the strongly growing Chinese market. However, these policies may trigger retaliation from foreign countries and in turn hurt the competitiveness of Chinese manufacturers in overseas markets. With an increasing percentage of solar panels attributable to overseas markets, especially the EU and US, these duties could impact the sales volume of the Chinese manufacturers, by increasing the cost of selling into the overseas markets and thus affecting their export sales. US anti-dumping duties and countervailing duties to Chinese manufacturers

Company Anti-dumping

tariff (2015) Anti-dumping

tariff (2012)

Countervailing duties

(2015)

Countervailing duties

(2012) Anti-dumping tariff

(incl. Taiwan)

Countervailing duties (incl.

Taiwan) Jinko 9.67% 24.48% 20.94% 15.24% 78.42% 38.43% Yingli 0.79% 24.48% 20.94% 15.24% 52.13% 38.43% Trina 9.67% 18.32% 20.94% 15.97% 26.71% 49.21% Canadian Solar 9.67% 24.48% 20.94% 15.24% 52.13% 38.43% ReneSola 9.67% - 20.94% 15.24% 78.42% 38.43% Suntech 33.08% 29.14% 20.94% 14.78% 52.13% 27.64% Other companies on the list 9.67% 24.48% 20.94% 15.24% 52.13% 38.43% Rest of PRC 238.95% 249.96% 20.94% 15.24% 165.04% 38.43%

Source: USITC, Daiwa Interest-rate hikes a secondary risk Expanding at a fast pace, most solar cell and module manufacturers as well as solar farm developers have high net debt-to-equity ratios, leading to a high proportion of finance costs in the operating cost structure. As such, changes in costs of borrowing for these companies have a considerable impact on their earnings. Against the backdrop of a slightly easing environment in China, a cut in the PBOC’s benchmark rate would mean a decrease in borrowing costs in the future, and thus higher profitability for the solar companies. Conversely, if the PBOC increases its benchmark rate unexpectedly, or if the industry’s credit rating as a whole deteriorates with higher gearing ratios in the future, borrowing costs could increase, adding pressure to the sector’s profitability, or even solvency. Net debt/equity gearing for Chinese solar companies (2014)

Source: Companies, Bloomberg, Daiwa

0%

50%

100%

150%

200%

250%

300%

Daqo GCL-Poly Trina Jinko JA Solar Xinyi Singyes GCL-NE Shunfeng United PV

Worsening trade disputes between China and the foreign countries may lead to higher export tariffs for Chinese manufacturers

High interest rates will increase the financing costs and therefore lower the return of a solar project

49


Appendix I: an overview of the global solar industry After going through the challenging period of 2014, the solar industry is showing signs of rebounding in 2015, which we believe should be sustained throughout 2015-16, with volume growth likely picking up to a double-digit pace again (37%/12% YoY growth for 2015/16E, respectively). We believe the disappointing solar demand in 2014 was mainly due to China missing its solar installation target, which accounted for 26% of new capacity additions in 2014. For 2015-16, we expect growth to remain focused on a small number of countries, including China and the US, which are mainly driven by strong policy support and improving solar project economics, while we think the golden days are over for Europe and Japan, where growth will likely be negative or stagnant at best. We see more uncertainties for solar installation post-2016, with the US’s ITC policy coming to an end, which could temporarily dampen solar demand in 2017, coupled with China’s potential FiT cut (which could be delayed from end-2015 to end-2016), after which an interim slowdown in solar demand could ensue. That said, some emerging markets should start to mature, such as Chile, South Africa, and India, where we may see promising demand growth partially offsetting the demand shock from China and the US. Germany Germany’s solar industry went through 3 years of elevated capacity growth, from 2010 to 2012, with over 7GW installed each year. But in 2013 and 2014, solar capacity growth decelerated to around 3.3GW, mainly driven by a rapid decline in FiTs, which has hurt the profitability of the solar operators. The FiT system in Germany is automatically adjusted according to the pace of solar installations. The more solar panels the country installs against its national target, the faster the FiT will come down. As such, we believe the slump in solar installations in 2013 was just what the government had planned. In fact, the German Government planned to restrain its solar installation to within 2.5-3.5GW for the year.

Germany has encountered a decline in solar installation due to rapid decline in solar FiTs

50


Germany: solar FiT adjustment system (EEG 2014)

Source: EEG, Daiwa research Solar PV demand in 2014 was, however, disappointing at only 1.9GW, which we believe was due mainly to the revision of the Renewable Energy Sources Act (Erneuerbare Energien Gesetz, or EEG) effective August 2014. Germany’s solar development is supported by the EEG, which ensures long-term sustainable growth of renewable energies within the country. A few highlights of the newest version of the EEG, or so-called EEG 2014, are listed below:

• Annual solar installation target was set at 2,500MW.

• The FiT model, where the solar operator receives a predetermined remuneration from the grid, was changed to a direct selling model where operators are obliged to market to electricity suppliers themselves.

• The financial support of the freestanding PV plants will be determined by tendering system, as opposed to a predetermined rate from the government.

• Self-consumed power generation is subject to a 40% EEG surcharge as opposed to full exemption under previous EEGs.

We believe the EEG surcharge for self-consumed power generation will weigh on new solar installations going forward. Germany’s solar industry features a high self-consumption rate from the industries that generate electricity themselves to reduce its energy costs and were previously exempted from the tax surcharge. In our view, the inclusion of a 40% tax surcharge on self-consumption would lower the return on investment (ROI) for solar energy, and with it consumers’ willingness to install PVs on their rooftops. That said, the FiT system should help to alleviate the depressed solar installation as time goes on. For example, in 1Q15, the FiT was automatically adjusted downward by only 0.25% per month after the slump in solar installations in 2014. This, coupled with the continued decline in PV prices, leads us to believe that the IRR on solar energy should gradually improve, and in the long term we expect the country’s average annual new capacity to be not far from its target of 2,500MW.

2,600 MW

3,500 MW

4,500 MW

5,500 MW

6,500 MW

-2.8%

-2.2%

-1.8%

-1.4%

-1.0%

2,400 MW

1,500 MW

-0.50%

7,500 MW-2.5%

-0.25%

-0%1,000 MW

+1.5%0 MW

The FiT reduction rate is applied monthly, referencing to the previous month’s FiT

The FiT reduction rate will be determined quarterly, on 1 Jan, 1 Apr, 1 July, and 1 Oct each year

The capacity addition is defined as 12-month total capacity additions before the last month (e.g. capacity additions between Nov 2013 to Nov 2014 for 1 Jan 2015 determination)

The target capacity addition is between 2.4 GW to 2.6 GW over 12 months (base case), within which a FiT reduction of 0.5% monthly will be applied

We expect the revision of EEG to weigh on Germany’s solar installations going forward

51


United Kingdom Solar installation growth in the UK took off in 2011, mainly as a result of the release of the FiT scheme by the UK Government in April 2010.

UK: solar annual capacity additions (2007-16E) UK: cumulative solar photovoltaics deployment (monthly)

Source: BP, DECC, Daiwa Source: DECC

Solar energy investments in the UK are mainly supported by 2 incentive schemes: 1) the FiT scheme, and 2) renewables obligations. As shown in the chart above, around 55% of solar projects are supported by the FiT scheme, with renewables obligation applications accounting for another 36%. FiT scheme The FiT scheme in the UK was introduced for the purpose of promoting small-scale renewable electricity generation, such as solar PV, wind turbines, hydro, micro CHP, etc. For the case of solar PV, the maximum capacity eligible for the FiT scheme is 5MW. The tariff rate varies across different sizes of solar projects, with larger projects usually receiving lower rates. In terms of the tariff trend, similar to other countries, FiT rates have decreased rapidly in the UK, from 34.7-42.6 p/kWh in 2010 to 6.1-13.4 p/kWh in April 2015, due to declining solar installation costs. See following graph for the FiT rate trend. UK: FiT rate trend

Source: Ofgem, Daiwa Renewables obligations The Renewables Obligation (RO) scheme was introduced in 2002. This is essentially a system providing incentives for renewable energy developers in order to promote renewable energy in the country. How does it work? Under the RO system, the UK regulator, Ofgem, assigns the amount of renewable energy (denoted by Renewables Obligation Certificates or ROCs) each electricity supplier needs to purchase. Ofgem will then issue the ROCs to the renewable generators for the amount

0

500

1,000

1,500

2,000

2,500

3,000

3,500

2007 2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)

0

1,000

2,000

3,000

4,000

5,000

6,000

2010 2011 2012 2013 2014 2015

(MW)

Feed in Tariffs eligible Renewables Obligation accredited Other Solar

0

10

20

30

40

50

Apr-1

0Ju

n-10

Aug-

10Oc

t-10

Dec-1

0Fe

b-11

Apr-1

1Ju

n-11

Aug-

11Oc

t-11

Dec-1

1Fe

b-12

Apr-1

2Ju

n-12

Aug-

12Oc

t-12

Dec-1

2Fe

b-13

Apr-1

3Ju

n-13

Aug-

13Oc

t-13

Dec-1

3Fe

b-14

Apr-1

4Ju

n-14

Aug-

14Oc

t-14

Dec-1

4Fe

b-15

Apr-1

5

(p/kWh)

< 4kW 4kW to 10kW 10kW to 50 kW 50kW to 100kW

100kW to 150kW 150kW to 250kW > 250kW

The FiT scheme in the UK was introduced for the purpose of promoting small-scale renewable electricity generation

52


of electricity they generate. In turn, the generators can sell the ROCs to the electricity suppliers in the open market so that the electricity suppliers will present enough ROCs to Ofgem in order to claim compliance. While the introduction of the FiT scheme triggered the rapid growth of the UK’s solar installations in 2011 to 2012, the effects from renewables obligation was leading the growth starting 2013. This was mainly due to the fact that many solar farm projects had been delaying construction and equipment purchases due to the rapidly declining PV module prices. However, the cut in RO subsidies in April 2013 prompted many developers to complete their solar farm construction before the cut came into effect. The same happened in 2014, when solar farm developers were again rushing for completion before the April 2014 subsidy cut in order to enjoy higher subsidy levels. What will happen next? The Renewable Obligation scheme was originally planned to close in April 2017, but UK Prime Minister Cameron decided to bring forward the deadline to April 2015. As a result, large-scale solar farms can now only engage in competitive bidding for renewable energies, namely through the Contract for Difference (CfD), which essentially puts solar PV into direct competition with other established renewable energies, such as onshore wind, hydro, etc. What is the Contract for Difference (CfD)? The CfD is like a futures contract in the electricity market, a new idea introduced during the UK electricity reform in 2013. It offers a 15-year contract for renewable developers with a set electricity price (called the strike price). When the market price for electricity falls below the strike price, the renewable energy generators get paid the difference. Conversely, if the market price for electricity exceeds the strike price, the generator needs to pay back the difference. An illustration of how the CfD works

Source: European Energy Centre While we expect solar PV system costs in the UK to continue to decline for the next few years, we believe solar PV is still not in the same league as other established renewable energies, such as onshore wind, in terms of costs. The first auction for supplying electricity in February 2015 provided strong evidence of the cost disadvantage of PV, where only 5 solar projects with a total capacity of 100MW made winning bids out of a total of 2.1GW capacity auctioned.

Renewable Obligation was leading the growth of the UK’s solar installation starting 2013

53


We believe the solar developers had expected the results soon after the CfD policy was released in 2014. As a result, solar installations for the first 3 months of 2015 surged by 69% YoY to 1,541MW as players rushed to enjoy the RO subsidy. Looking ahead, while we believe small-scale solar PV systems will continue to experience sustainable growth given the FiT scheme and falling solar PV system costs, a slump in large-scale solar project installations was widely expected after March 2015. As such, we forecast 1.8GW/ 1.6GW to be installed for 2015-16E, respectively. Japan The solar market in Japan started its boom in July 2012, when its FiT system was initiated to promote the country’s renewable energy development, as enacted by the Act on Special Measures Concerning the Procurement of Renewable Energy (ASM), a policy released to promote renewable energy and offset the impact caused by the nuclear shutdown after the Fukushima accident. Japan: feed-in tariff system

Source: METI, Daiwa Generous FiT system fuelling capacity growth … According to the ASM, Japan’s FiT system will remain in place at least until 2021. The FiT rate is very generous compared with the FiTs in other countries and has attracted a large number of solar farm developers to invest in solar farms. The rate was initially set at JPY40/kWh plus tax (or JPY42/kWh for residential) in 2012, and was lowered to JPY36/kWh plus tax (or JPY38/kWh for residential) starting April 2013 and to JPY32/kWh plus tax (or JPY37/kWh for residential) starting April 2014. Japan remains one of the countries with the highest solar tariff. As a consequence, Japan’s solar installations rose from 1.3GW in 2011 to 1.8GW in 2012, and to 6.9GW in 2013. Besides, solar projects applications have kept flooding in. As of November 2014, total approved projects had reached 70GW, over 3 times the amount of completed solar projects at 20GW.

Government

Solar PVWindHydro

BiomassGeothermal

ElectricityCustomers

ElectricityUtilities

Sale of electricity

Electricity purchased at fixed price for a guaranteed period

Electricity supply

Collection of surcharge and

electricity chargeSurcharge adjustment

organization (to collect and distribute the

surcharge)

Deciding tariffs and durations respecting the opinion of the committeeApproval

of facilities

Decision of surcharge unit price per kWh (every fiscal year)

Special committee for determination of tariff’s and durations

Ministry of Economy, Trade and Industry (METI)

Japan started its FiT system for solar energy in July 2012

54


Japan: solar capacity additions (2004-17E)

Source: BP Statistics, Daiwa Before the FiT system was introduced in Japan, the solar PV market was mainly dominated by homeowners with capacities usually below 10kW, driven by the country’s subsidy policy towards residential PV systems starting in 1994. According to METI, in July 2012, 83% of the solar projects were below 10kW. The situation has changed since the FiT system was introduced. Given its generous nature, the FiT system has attracted a large amount of corporate investment in large-scale solar farm development. As such, the proportion of installed capacity for residential PV system fell below 40% in late 2014.

Japan: installed capacity breakdown by size (Nov 2014) Japan: approved project breakdown by size (Nov 2014)

Source: METI, Daiwa Source: METI, Daiwa

We expect capacity from large-scale solar farms to continue to increase in the future, given the huge amount of projects already approved and pending construction (ie, only <5% coming from projects below 10kW). However, this also suggests that Japan’s future solar installations may face more fluctuations, considering corporate investors’ high sensitivity to profitability and returns. … but more uncertainties in the long run We believe its ample project pipeline will likely sustain Japan’s solar installations over the next few years. Nevertheless, a few factors are adding to the uncertainties: • Following METI’s efforts to cancel the projects approved since 2012 that have not

made any progress, it was announced in September 2014 that solar projects with a total capacity of 1.8GW had been cancelled. We believe if more similar projects are cancelled in the future, solar installations might not be as high as expected.

• METI proposed the next round of FiT cuts in April 2015, from JPY32/kWh to JPY29/kWh, and schedule another cut in July 2015, to JPY27/kWh, squeezing the profitability of solar power projects in Japan. We believe it will continue to lower the FiT in the future as solar project construction costs are reduced.

0

2,000

4,000

6,000

8,000

10,000

12,000

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)

38.4%

25.2%

8.5%

8.8%

14.7%

4.5%

below 10kW

10kW to 50kW

50kW to 500kW

500kW to 1000kW

1000kW to 2000kW

> 2000kW

4.8%

32.9%

4.6%5.6%

16.4%

35.8%

below 10kW

10kW to 50kW

50kW to 500kW

500kW to 1000kW

1000kW to 2000kW

> 2000kW

Japan’s solar installation shifted towards utility-scale solar farm vs. solar DG since the FiT introduction

55


• In December 2014, METI announced that the grids would be permitted to limit the output of renewable energy projects for up to 360 hours per year, from 30 days previously, effectively expanding the curtailment powers of the grids.

• The Japan Government approved the “basic energy plan” in April 2014 which designated nuclear energy as an important source of electricity in Japan, implying a possible restart of nuclear operation in the future and threatening renewable energy development.

Japan: solar FiT rate revision since July 2012 (below 10kW) Japan: solar FiT rate revision since July 2012 (>10kW and excl. tax)

Source: METI, Daiwa research Source: METI, Daiwa research

Considering the rising uncertainties associated with the solar market, it is not easy to accurately forecast long-term solar PV demand in Japan. In addition, following the release of the Basic Energy Plan in April 2014, we expect a more quantitative energy mix target to be announced by the Japanese government by end-2015. Solar rush looks set to continue until March 2016 That said, we believe the solar rush in Japan will continue at least until March 2016, mainly driven by the expiry of the country’s Green Investment Tax Credit. Apart from the FiT system, the Ministry of Finance under the Japanese Government also announced the Green Investment Tax Credit, which will expire by March 2016. According the tax credit, solar project developers can choose to enjoy one of the following:

• 1) 30% special depreciation above the ordinary depreciation

• 2) 100% upfront depreciation

• 3) 7% of the acquisition costs (only applicable to small and medium sized enterprises) As such, we expect solar farm developers to rush to complete their solar projects before end-March 2016 to enjoy the favourable tax credit policy, similar to what happened in the US when its investment tax credit for wind power expired at end-2012. The rush should create strong momentum, enough to offset the potential negative impacts from the above uncertainties, in our view, implying that Japan will experience another strong solar installation year in 2015. United States Solar energy in the US has been developing rapidly over the last 10 years, with solar capacity seeing a 57% CAGR over the past 9 years (2006-14). In 2014, the US installed 6.2GW of solar PV for the year, mainly driven by strong utility scale solar installation.

42

3837

35

30

32

34

36

38

40

42

44

July 2012 - March2013

April 2013 - March2014

April 2014 - March2015

Since April 2015

(JPY/kWh)

4036

3229 27

05

1015202530354045

July 2012 -March 2013

April 2013 -March 2014

April 2014 -March 2015

April 2015 - June2015

Since July 2015

(JPY/kWh)

We see Japan’s solar installation to remain rapid before March 2016

56


US: cumulative solar capacity

Source: BP Statistics, Daiwa Going forward, we expect solar installation in the US to accelerate further and eventually peak in 2016, mainly as the 30% Federal Investment Tax Credit (ITC) policy is due to expire at end-2016. We believe several solar farm developers will rush to complete their solar projects before end-2016 to be eligible for the ITC, and therefore utility-scale installation should slump post-2016, but will subsequently recover on improving economics of solar projects, in our view. California is the single-largest contributor to solar installation in the US, and accounted for 57% of capacity additions in 2014. We attribute the large solar market to the state’s solar rebate programme, the California Solar Initiative (CSI). The CSI was introduced in 2007, with a total budget of USD2.167bn, until 2016. It offers funding to customers of publicly-owned utilities, including Pacific Gas and Electric (PG&E), Southern California Edison (SCE), and San Diego Gas & Electric (SDG&E), in addition to federal incentives, to build their own solar PV systems on existing homes and buildings. The CSI programme triggered widespread solar installation across California, given its generous initial subsidies (the subsidy amount declines gradually upon the increase of total solar installations). The CSI program was nearly terminated in 2014, given the early accomplishment of its installation target (1.94GW by 2017), but we still see sustainable installation driven by falling solar installation costs, and higher electricity bills for higher consumption being an incentive for customers to install solar systems for net metering, and sophisticated financing solutions. US: solar capacity additions by state (2014)

Source: SEIA, Daiwa That said, we expect the planned termination of the CSI to narrow the advantage California has over other states. We have seen some other states, such as Massachusetts and Nevada, rapidly catching up in terms of the number of installations since 2012, and believe California’s share should gradually decrease in the longer term. We especially expect

0

5,000

10,000

15,000

20,000

2006 2007 2008 2009 2010 2011 2012 2013 2014

(MW)

Residential Non-residential Utility

57%

6%

5%

5%

4%

4%

18% California

North Carolina

Nevada

Massachusetts

Arizona

New Jersey

Others

The Federal Investment Tax Credit cut starting 2017 should also lead to rush of installations during 2016, in our view

57


increasing resistance to solar installation from the utilities sector in California as more residential customers are using net metering, damaging the utilities industries’ profitability (eg, the AB327, which allows the California Commission to determine a net metering structure). Australia Australia’s solar development is mainly supported by two policies: 1) the Renewable Energy Target (RET) at the federal government level, and 2) FiT systems usually enacted by individual state governments. Although we see that solar PV developed rapidly over 2011-13, the overall power generation share from solar was still low (~1.6% as of 2013). Australia: power generation breakdown (2013)

Source: Clean Energy Australia, Daiwa Renewable Energy Target (RET) The RET is a scheme initiated by the Australia Government in 2001 to promote renewable energy generation. The latest target was set in 2009, which specified that 20% of Australia’s power generation will come from renewable energy by 2020. The RET scheme essentially works through the trading of renewable energy certificates. Both large-scale renewable power stations and small-scale systems are allowed to create renewable energy certificates for the amount of electricity they generate, and electricity retailers are obliged to generate a pre-determined proportion of power from renewable energy each year and to purchase renewable energy certificates if they have a shortfall. Given the financial incentives for renewable energy operators, as well as the mandatory fulfilment of renewable energy proportion from electricity retailers, the RET scheme has driven robust installation of renewable energy investment over the past 14 years, and the government has estimated that renewable energy may account for 27% of Australia’s total electricity by 2020. In August 2014, following a request from then Prime Minister Tony Abbott, the government requested a review of the RET scheme and suggested the scheme be abolished or downsized in anticipation of a decline in total power demand. The proposal triggered widespread discontent among the opposition party and solar groups. Negotiations were held between parties, but a final decision has yet to be made. Feed-in tariff systems The FiT systems are operated separately by different states. Many of the FiT systems were introduced in 2008-09 with the goal of boosting solar power generation in the relevant region. In general, there are 2 types of tariffs: 1) gross FiT, and 2) net FiT. Gross FiT refers to power consumption and generation separately recorded and paid, whereas net FiT refers to consumption and generation first being offset, with the excess electricity exported then compensated by the net FiT rate.

85.2%

8.1%

4.0%1.6%1.0%

Fossil fuels

Hydro

Wind

Solar PV

Others

Australia’s solar installation is supported by RET (federal) and FiT systems (local)

58


Initially, states such as Victoria and Canberra offered generous rates to customers (AUD0.6/kWh). But after the solar boom in 2011-13 (up from 383MW installed to over 800MW each year), these rates were significantly dropped. See the table below for the current FiT rates. Australia: FiT rates for solar State FiT (AUD per kWh) FiT type VIC 8c Net FiT SA Closed n.a. Canberra (ACT) 7.5c Net FiT TAS 5.55c Gross FiT NT 27.13c Net FiT WA 8.9c Net FiT QLD 6-8c Net FiT NSW 5-12.9c Gross FiT

Source: regulators in Australia, Daiwa Going forward, we expect Australia’s solar capacity growth to be driven by: 1) the mandatory 20% renewable consumption as stipulated by the RET, 2) consistently increasing electricity tariffs, driving customers to generate electricity themselves, and 3) additional income generated from renewable energy certificates for homeowners and solar energy operators (usually around AUD40 per MWh, though fluctuating depending on market supply and demand). India Characterised by high solar insolation and electricity supply shortage, India is well-positioned to experience accelerated growth in solar installation over 2015-20, in our view. In 2010, former Prime Minister Manmohan Singh set out a long-term plan for the solar power industry, the Jawaharlal Nehru National Solar Mission. He aimed to develop the solar industry to both supply additional electricity and create a multiplier effect for local economies by way of jobs and industrial production. The plan also provided generous subsidies to solar energy developers, including up to a 90% upfront subsidy for remote areas without electricity supply, and up to a 30% upfront subsidy (with a progressively declining rate depending on the installed base) to other urban areas in India. The table below illustrates the capacity targets announced at the time. India: installed capacity target (MW) up to 2012-13 2013-17 2017-22 Utility Grid Power including rooftop 1,000 - 2,000 4,000 - 10,000 20,000 Off-grid Solar applications 200 1,000 2,000

Source: SECI, Daiwa In India, due to rising electricity and declining solar module costs, some states are already enjoying grid parity at the user level, especially for the commercial and industrial segments. For the residential segment, given the low electricity tariff currently, solar power generation costs are still not low enough for investors. Therefore, the government is currently providing subsidies to residential users. That said, as the solar industry in India is entering the second phase (2013-17) of the National Solar Mission, the Indian Government is aiming to develop the solar industry rapidly, and distributed solar will not be its priority for the next 3 years. Instead, we note that the Ministry of New & Renewable Energy (MNRE) has laid out plans to develop a number of solar parks, or so-called Ultra Mega Solar Power Projects, which are solar power stations with at least 500MW capacity each.

Local government subsidies have shrunk significantly after 2013

Jawaharlal Nehru National Solar Mission is the high-level framework for India’s solar energy development

59


3 phases of developing Ultra Mega Solar Power Projects According to the solar industry expert group Bridge to India, the India Government has planned a 3-phase development for solar power industry until 2019 (end of the existing government’s tenure), with a total capacity addition of 15GW. The first phase has a total capacity of 3GW, and will begin with the state-owned Solar Energy Corporation of India (SECI) and NTPC Vidyut Vyapar Nigam (NVVN) developing 3 solar parks of 1GW each. The Ultra Mega Solar Power Projects will be bundled with thermal power, with electricity sold to the national grid at a blended price. The second phase has a total capacity of 5GW, in which the electricity bundling plan will be removed, and instead the solar power projects will likely be subsidised through interest-rate reduction. The third phase, focusing on the last 2 years of the development plan, will have a total capacity addition of around 7GW. Subsidies for solar power will gradually be removed, but the low-enough costs should attract large corporations or even foreign investors to the market. All in all, we believe India’s solar power will grow rapidly over the next few years, making the country one of the major solar powers globally.

The India Government planned to add 15GW of solar farms before 2019

60


Appendix II: major solar power policies in China In December 2014, the NEA announced that it was starting to conduct research and plan for solar power development for the 13th FYP period. We expect the draft plan to be released in early-2016 and then be subject to fine-tuning by the provincial regulatory units. According to the 12th FYP target, solar capacity should reach 35GW by 2015 and 100GW by 2020. The full-year 2015 target was exceeded by end-1H15, and we estimate actual solar installed capacity tol reach 44GW by end-2015. China: Daiwa’s solar installation forecasts versus NEA target

Source: NEA, Daiwa Also, according to our forecasts, China’s solar installed capacity for 2020 should exceed the NEA’s 12th FYP target of 100GW by early-2019, which suggests that the 12th FYP’s 2020 target could be too conservative currently. As such, we expect the new target to exceed the target set during 12th FYP. Should the target be raised to 150GW (according to industry sources), our solar-installation forecasts for 2016-20 may face further upside. For the shorter term (in 2015), the NEA issued a previously proposed target of 15GW at the beginning of 2015, which was revised up in March 2015 to 17.8GW, up 27% YoY from last year’s target and an implied 68% increase from the capacity addition in 2014 (10.6GW). The table below shows the breakdown by province for the NEA targets.

China: the NEA’s 2015 solar installation target Province Solar capacity target (MW) vs. 2014 target (%) 2014 additions (MW) actual vs. target 2014 target (MW) Hebei 1,200 20% 970 -3% 1,000 Shanxi 650 44% 230 -49% 450 Inner Mongolia 800 45% 1,640 198% 550 Liaoning 300 20% 50 -80% 250 Jilin 300 100% 50 -67% 150 Heilongjiang 300 200% 0 -100% 100 Jiangsu 1,000 -17% 1,520 27% 1,200 Zhejiang 1,000 -17% 300 -75% 1,200 Anhui 1,000 82% 430 -22% 550 Fujian 400 14% 40 -89% 350 Jiangxi 600 58% 260 -32% 380 Shandong 800 -33% 320 -73% 1,200 Henan 600 -20% 160 -79% 750 Hubei 500 25% 90 -78% 400 Hunan 400 60% 50 -80% 250 Guangdong 900 -10% 220 -78% 1,000 Guangxi 350 133% 40 -73% 150 Hainan 200 82% 70 -36% 110 Sichuan 600 500% 30 -70% 100 Guizhou 200 233% 0 -100% 60 Yunnan 600 445% 150 36% 110 Shaanxi 800 60% 420 -16% 500 Gansu 500 -9% 970 76% 550 Qinghai 1,000 82% 1,020 85% 550 Ningxia 1,000 100% 820 64% 500 Xinjiang 1,800 177% 590 -9% 650 Others 0 n.a. 170 -83% 990 Total 17,800 27% 10,610 -24% 14,050

Source: NEA, Daiwa research

6 17

28 44

62 79

97 116

137

020406080

100120140160

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)

Installed capacity NEA 12th FYP target for 2020 Potential target uplift for 13th FYP plan for 2020

The NEA target for total solar capacity by 2020 could be higher than 100GW

61


We also note that the 17.8GW target only includes utility-scale solar farms, and quasi-distributed solar projects (solar farms with less than 20MW in capacity), mostly built above farmlands or fish ponds. China: NEA 2015 target (MW)

Source: NEA, Daiwa Note: Xinjiang figures include capacity target from Xinjiang Production and Construction Corps We believe the aggressive target shows the regulator’s determination to develop solar power in the country. In addition, we note that the NEA has excluded rooftop solar systems from the target this year, implying there is still a long road for rooftop DG projects to ramp up, in our view. With such an ambitious target for 2015, we think the natural question from investors would be – is it achievable? In our view, the target will be challenging to meet but not altogether unrealistic. We see 2 main drags: 1) first-tier solar-module manufacturers are already operating at full capacity amid current demand, and 2) some regions (Gansu and Qinghai) are already experiencing, or are highly likely to experience, grid curtailment issues, constraining their ability to add capacity too aggressively. This leads us to our forecast of about 16GW for 2015E. Renewable Portfolio Standard (RPS) to be the longer-term catalyst To our understanding, the Renewable Portfolio Standard policy has already been submitted to the State Council for approval, after 3 years of drafting. The RPS will provide a target for each province regarding how much renewable energy must be purchased by the grid, with a relevant reward or punishment should the province exceed or miss the target. A major benefit from the RPS is an increasing number of provinces likely providing stronger subsidies for renewable power generation, including solar PV, in order to meet the RPS target, which in turn should increase the economics of solar projects, leading to higher capacity additions. To a lesser extent, we also believe the grid curtailment issues may improve as the relevant grid infrastructure will be built to accommodate the increasing portion of renewable energy input; but we believe the improvement can only be realised in the longer term (by 2017).

1,800

1,200

1,000

1,000

1,000

1,0001,000

900800800800

6,500

Xinjiang Hebei

Jiangsu Zhejiang

Anhui Qinghai

Ningxia Guangdong

Inner Mon Shandong

Shaanxi Others

Solar DG has been excluded from the NEA target set for 2015

RPS could be a longer-term capacity growth driver for renewable energies, including solar

62


China’s major policies for the solar power industry Date Regulator Policy Policy (English) Jul-13 State Council 国务院关于促进光伏产业健康发展的若干意见 Several opinions of the State Council on promoting the healthy development of

photovoltaic industry Aug-13 NDRC 关于发挥价格杠杆作用促进光伏产业健康发展的通知 Notice about using price leverage to facilitate healthy development of PV industry Aug-13 NDRC 关于调整可再生能源电价附加标准与环保电价有关事项的通知 Notice about additional standards of renewable power price adjustment and green

electricity price Sep-13 MOF 关于光伏发电增值税政策的通知 Notice about solar power generation VAT policy Mar-14 MEP 能源行业加强大气污染防治工作方案 Work plan of reinforcing air pollution prevention Aug-13 NEA 关于印发<光伏电站项目管理暂行办法>的通知 Notice about issuing "Temporary Management Method for PV Station Project" Nov-13 NEA 关于印发<光伏发电运营监管暂行办法>的通知 Notice about Issuing "Temporary Management Method for Operation of PV Power

Generation" Jan-14 NEA 关于下达 2014 年光伏发电年度新增建设规模的通知 Notice about assignment of annual construction target of PV power generation Feb-14 NEA 新建电源接入电网监管暂行办法 Temporary supervision methods for grid connection of newly installed power supply Mar-14 NEA 关于印发加强光伏产业信息检测工作方案的通知 Notice about issuing work plan of enhancing information detection of PV industry Dec-13 MOF 关于可再生能源电价附加补助资金管理有关意见的通知 Notice about relevant opinions for management of additional subsidy fund for

renewable electricity price Oct-14 NEA 国家能源局关于进一步加强光伏电站建设与运行管理工作的通知 Oct-14 NEA 国家能源局关于开展新建电源项目投资开发秩序专项监管工作的通

知

Oct-14 NEA 国家能源局关于规范光伏电站投资开发秩序的通知 Mar-15 NEA 国家能源局关于下达 2015 年光伏发电建设实施方案的通知 Notice about solar power generation project construction plan in 2015 Mar-15 NEA 光伏扶贫试点实施方案编制大纲(修订稿)

Source: State Council, NDRC, MOF, MEP

China’s major policies for solar DG Date Regulator Policy Policy (English) Jul-13

分布式发电管理暂行办法 Temporary Methods of Distributed PV Power Generation Management

Aug-13

国家能源局关于开展分布式光伏发电应用示范区建设的通知 Notice of National Energy Administration about Development of Distributed PV Demonstration District

Aug-13

关于支持分布式光伏发电金融服务的意见 Opinions about Financial service to Support Distributed PV Power Generation

Nov-13

国家能源局关于印发分布式光伏发电项目管理暂行办法的通知 Notice of National Energy Administration about Issuing Temporary Management Method for PV Station Project

Jul-13

关于分布式光伏发电实行按照电量补贴政策等有关问题的通知 Notice about Matters Relevant to Subsidy Policy of Distributed PV Projects

Feb-14

国家认监委、能源局关于加强光伏产品检测认证工作的实施意见 Execution Opinions of CNCA and NEA about Enhancing Certification Testing of PV Products

Dec-13

关于印发分布式电源并网相关意见和规范（修订版）的通知 Notice about Issuing Relevant Opinions (Modified) for Grid Connection of PV Power Supply

Jun-14

关于国家电网公司购买分布式光伏发电项目电力产品发票开具等有

关问题的公告 Announcement about Invoice Issuance on Distributed PV Products Purchased by State Grid

Sep-14

国家能源局关于进一步落实分布式光伏发电有关政策的通知 Sep-14

关于加快培育分布式光伏发电应用示范区有关要求的通知

Nov-14

关于《低碳社区试点建设指南》征求意见稿公开征求意见的公告 Source: State Council, NDRC, MOF, MEP

63


Appendix III: solar DG local government subsidies China: local government subsidies for solar DG Province City Subsidy Anhui Haozhou CNY 0.25 per kWh for 10 years for newly completed distribution PV projects which connected to national grid

Hefei CNY 0.25 per kWh for 15 years + national and provincial subsidy CNY 3 per W, not exceed 15 thousand per station for family PV CNY 2 per W for distributed PV installation, not exceed 1m per project

Hefei CNY 3-4 per W for solar installation for residential solar installation completed between Mar 2014 and Dec 2015 Beijing Beijing CNY 0.3 per kWh for 5 years Guangdong Guangzhou CNY 0.1 per kWh for 10 years for DG projects approved before 2020

Dongguan CNY 0.1 per kWh for 5 years for projects approved in 2014-2016 Dongguan CNY 0.25 per W one-off subsidy for distributed PV projects of installed capacity

CNY 0.1 per kWh for 5 years for non-residential distributed PV projects CNY 0.3 per kWh for 5 years for residential distributed PV projects

Foshan CNY 0.15 per kWh for projects completed between 2014-2015 Yangjiang CNY 0.42 per kWh for family PV projects

Guangxi Provincial CNY 4 per W for residential rooftop installation CNY 3 per W for public utilities rooftop installation

Hebei Langfang CNY 0.42 per kWh Provincial CNY 0.42 per kWh for distributed PV

CNY 0.3/0.2/0.1 per kWh for PV stations completed before end 2014/2015/2017 for 3 years since projects put into use Henan Luoyang CNY 0.1 per W for 3 years for DG projects completed before 2015 Hubei Provincial CNY0.25 per kWh for 5 years, applied to PV projects completed by end-2015

Huangshi CNY 0.1 + national subsidy Additional CNY 0.1 based on feed in tariff for 10 years

Hunan Provincial CNY 0.2 per kWh for 10 years Changsha CNY 0.1 per kWh for 5 years

Jiangsu Suzhou CNY 0.1 kWh Zhenjiang CNY0.1 per kWh for 3 years Jurong CNY0.3 per kWh for 5 years Suzhou CNY0.1 per kWh for 3 years Wuxi CNY 200,000 per MW one-off subsidy for distributed PV station Provincial CNY 1.3/1.25/1.2/1.15 per kWh for 2012/13/14/15 for projects completed between 2012-2015 without national subsidy

Jiangxi Provincial CNY 4 per W for first phase rooftop installation CNY 3 per W for second phase rooftop installation

Shangrao CNY 0.15 per kWh for 5 years Provincial Feed in tariff subsidy: CNY 0.2 for projects completed before end 2015

Feed in tariff subsidy: CNY 0.1 for projects completed before end 2017 Provincial CNY 0.2 per kWh for 20 years for projects listed in annual construction plan Provincial CNY 0.15 per kWh for 5 years Nanchang CNY 3 per W for family roof-top projects

Jilin Provincial CNY 0.15 per kWh Shaanxi Provincial CNY 1 per W for solar installation between 2014-17

Provincial CNY 1 per W one-off subsidy for projects completed between 2015-17 Shangnan CNY 0.006 per W for ground PV station

CNY 0.03 per W for distributed PV Xi'an CNY 1 per W one-off subsidy for distributed PV station in addition to national and provincial subsidy

CNY >1 per W one-off subsidy for ground PV station Shangluo CNY 0.01 for ground PV station

CNY 0.05 for distributed PV station Shandong Provincial CNY 0.05 per kWh for DG projects approved in 2013-15 Shanghai Provincial For projected completed between 2013-15, CNY 0.25 per kWh for C&I solar installations, and CNY 0.4 per kWh for residential solar installations for 5 years

Provincial CNY 0.3 per W for PV stations for 5 years Shanxi Provincial 50% VAT refund before 2015

Taiyuan Fixed subsidy for investors with -50m/50-100m /+100m investment Jincheng CNY 0.2 per kWh

CNY 3 per W one-off subsidy for installation capacity Zhejiang Provincial CNY 0.1 per kWh provincial subsidy, applicable to all cities in Zhejiang

Wenzhou CNY 0.2/0.15/0.1 per kWh for 2014/15/16 projects for 5 years, respectively CNY 0.3 per kWh for roof-top projects for 5 years

Tongxiang For projects approved before 2014, CNY 1.5 per W for solar installations CNY 0.3 per kWh for the first 2 years CNY 0.2 per kWh for 3rd to 5th years

Hangzhou CNY 0.1 per kWh for 5 years, provided to projects completed in 2014-15 Fuyang For DG projects completed in during 2014-16

CNY 0.3 per kWh for the first 2 years CNY 0.2 per kWh for 3rd to 5th years

Jiaxing CNY 2.8 per kWh for first 3 years subsidy reduced by CNY 0.05 kWh every year thereafter

Dongtou DG above 50 kW: CNY 0.4 per kWh for first 5 years. Additional CNY 0.1 per kWh for demonstration project. Rooftop: CNY 2 per W for solar installation and CNY 0.2 per kWh for first 5 years

Xinchang CNY0.1 per kWh until end of 2017 Haining CNY0.2 per W for solar installations on rooftops

CNY0.2 per kWh subsidy during 2015-18, for projects approved in 2014 & 15 Hangzhou xiaoshan CNY 0.2 per kWh for installation capacity not less than 30KW

64


China: local government subsidies for solar DG (cont’d) Province City Subsidy Zhejiang Hangzhou Jiande CNY 0.2 per kWh for manufacturer's self-use purposes

CNY 0.1 per kWh for 上大网 Wenzhou Yongjia CNY 0.4 per kWh for 5 years

CNY 2 per W one-off subsidy for family roof-top projects; CNY 0.3 per kWh for 5 years after projects completion and put into use Wenzhou Dongtou CNY 0.4 per kWh for 5 years

CNY 0.1 per kWh additional subsidy for projects listed in the "光电旅一体化"plan Wenzhou Ruian CNY 0.3 per kWh for 5 years

CNY 0.05 per kWh for 5 years for roof-top lease of distributed PV Wenzhou Leqing CNY 0.3 per kWh for 5 years

CNY 0.05 per kWh for 5 years for roof-top lease of distributed PV Jiaxing Xiuzhou CNY 1 per W one-off subsidy Jiaxing Haining CNY 0.35 per kWh for projects completed before end 2014

CNY 0.3 per W for roof-top resources provider Quzhou CNY 0.3 per kWh + CNY1 per kWh provincial subsidy for distributed PV power stations Quzhou Longyou CNY 0.3 per kWh

CNY 0.3 per W one-off subsidy for installation capacity Quzhou Kaihua CNY 0.1 per kWh for purchase local PV product Quzhou Jiangshan CNY 0.3 per W for PV stations installation capacity

CNY 0.2 per kWh + national and provincial subsidy CNY 0.3 per W one-off subsidy for distributed PV projects CNY 0.15 per kWh + national and provincial subsidy for electricity self-produced and self-use

Ningbo CNY 0.1 per kWh for 5 years for projects completed before 31 Dec 2015 Huzhou CNY 0.18 per kWh Huzhou Deqing CNY 0.1 per kWh for 3 years + national and provincial subsidy Huzhou Anji CNY 0.1 per kWh for 2 years + national and provincial subsidy Shaoxing CNY 0.2 per kWh for 5 years + CNY0.42 per kWh national subsidy and CNY0.1 per kWh provincial subsidy Yiwu CNY 2 per W one-off for households joining PV power projects

CNY 0.3 per W for corporations providing the rooftop CNY 0.1 per kWh for 3 years for corporations

Dongyang CNY 0.2 per kWh for 3 years for new distributed PV projects Lishui CNY 0.15 per kWh + national and provincial subsidy for projects completed between 2014-end 2016

Source: NDRC, Daiwa

65


Appendix IV: IRR model (solar utility-scale projects) China: solar utility-scale project IRR Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 Year 11 Year 12 Year 13 Year 14 Year 15 Year 19 Year 20 Income statement Net electricity generated (mn kWh)

59 59 58 58 57 57 56 56 56 55 55 54 54 53 53 51 51

Gross revenue (incl. VAT)

59 59 58 58 57 57 56 56 56 55 55 54 54 53 53 51 51

Net revenue (excl. VAT)

51 50 50 49 49 49 48 48 47 47 47 46 46 46 45 44 43 Depreciation

(18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18)

Operating expenses

(5) (5) (5) (5) (6) (6) (6) (6) (6) (7) (7) (7) (7) (7) (8) (9) (9) VAT rebate

9 9 8 8 8 8 1 0 0 0 0 0 0 0 0 0 0

Government subsidies

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 EBIT

37 36 35 35 34 33 25 24 23 23 22 22 21 21 20 18 17

EBIT margin (%)

72% 71% 71% 70% 69% 69% 52% 50% 49% 48% 48% 47% 46% 45% 44% 40% 39%

Interest expense

(22) (21) (20) (18) (17) (15) (13) (11) (9) (7) (5) (3) 0 0 0 0 0 Profit before tax

14 15 16 16 17 18 12 13 14 16 17 19 21 21 20 18 17

Income tax

0 0 0 (2) (2) (2) (3) (3) (4) (4) (4) (5) (5) (5) (5) (4) (4) Net profit

14 15 16 14 15 16 9 9 11 12 13 14 16 15 15 13 13

Net margin (%)

28% 30% 31% 29% 31% 33% 19% 20% 22% 25% 28% 31% 34% 34% 33% 30% 29%

IRR calculations Profit before tax

14 15 16 14 15 16 9 9 11 12 13 14 16 15 15 13 13

Depreciation

18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 Interest expense

22 21 20 16 15 13 10 9 7 5 4 2 0 0 0 0 0

Capex (415) FCFF (415) 54 54 53 48 47 47 37 36 35 35 34 34 34 33 33 31 30

Interest expense

(22) (21) (20) (16) (15) (13) (10) (9) (7) (5) (4) (2) 0 0 0 0 0 Net repayment in debt

(19) (20) (22) (23) (24) (26) (28) (30) (32) (34) (36) (39) 0 0 0 0 0

New debt 332 FCFE (83) 13 12 12 9 8 8 (1) (3) (3) (4) (5) (6) 34 33 33 31 30

Project IRR 7.9%

Equity IRR 11.3% Source: Daiwa

IRR assumptions Assumptions Unit Capacity (MW) 50 Operation (Years) 20 Unit-capex (including VAT) (CNY/W) 8.3 Unit-capex (excluding VAT) (CNY/W) 7.1 Total investments (including VAT) (CNYm) 415 Total investments (excluding VAT) (CNYm) 355 Tariff (including VAT) (CNY/kWh) 1.00 Tariff (excluding VAT) (CNY/kWh) 0.85 Insolation (hours) 1,550 System efficiency (%) 78% Utilization hours (hours) 1,209 Degradation (1st year) (%) 2.0% Degradation (2nd year and after) (%) 0.8% Debt (%) 80% Debt repayment years (years) 12 Finance cost (%) 6.7% Initial operating expenses (CNY/W) 0.10 Inflation (%) 3%

Source: Daiwa

66


Appendix V: IRR model (solar DG projects) China: solar DG projects IRR Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 Year 11 Year 12 Year 13 Year 14 Year 15 Year 19 Year 20 Income statement Net electricity generated (mn kWh)

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5


6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

Net revenue (excl. VAT)

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Depreciation

(2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2)

Operating expenses

(1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) VAT rebate

1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0

Government subsidies

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 EBIT

4 4 4 4 4 3 3 3 3 3 3 3 3 3 3 2 2

EBIT margin (%)

78% 77% 77% 76% 75% 61% 57% 56% 56% 55% 54% 53% 53% 52% 51% 47% 46%

Interest expense

(2) (2) (2) (2) (2) (1) (1) (1) (1) (1) (0) (0) 0 0 0 0 0 Profit before tax

2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 2 2

Income tax

0 0 0 (0) (0) (0) (0) (0) (0) (1) (1) (1) (1) (1) (1) (1) (1) Net profit

2 2 2 2 2 2 1 1 1 2 2 2 2 2 2 2 2

Net margin (%)

39% 40% 41% 38% 39% 29% 24% 26% 28% 31% 33% 36% 40% 39% 38% 36% 35%

IRR calculations Profit before tax

2 2 2 2 2 2 1 1 1 2 2 2 2 2 2 2 2

Depreciation

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Interest expense

2 2 2 2 1 1 1 1 1 1 0 0 0 0 0 0 0

Capex (39) FCFF (39) 6 6 6 5 5 4 4 4 4 4 4 4 4 4 4 3 3

Interest expense

(2) (2) (2) (2) (1) (1) (1) (1) (1) (1) (0) (0) 0 0 0 0 0 Net repayment in debt

(2) (2) (2) (2) (2) (2) (3) (3) (3) (3) (3) (4) 0 0 0 0 0

New debt 31 FCFE (8) 2 2 2 2 1 1 0 0 0 0 (0) (0) 4 4 4 3 3

Project IRR 10.2%

Equity IRR 19.3% Source: Daiwa

IRR assumptions Assumptions Capacity (MW) 5 Operation (Years) 20 Unit-capex (including VAT) (CNY/W) 7.7 Unit-capex (excluding VAT) (CNY/W) 6.6 Total investments (including VAT) (CNYm) 39 Total investments (excluding VAT) (CNYm) 33 Tariff (including VAT) (CNY/kWh) 1.20 Tariff (excluding VAT) (CNY/kWh) 1.03 Insolation (hours) 1,400 System efficiency (%) 78% Utilization hours (hours) 1,092 Degradation (1st year) (%) 2.0% Degradation (2nd year and after) (%) 0.8% Debt (%) 80% Debt repayment years (years) 12 Finance cost (%) 7.0% Initial operating expenses (CNY/W) 0.10 Inflation (%) 3%

Source: Daiwa

67


Appendix VI: Renewable Energy Fund forecasts China: Renewable Energy Fund forecasts Units 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source of fund Renewable energy surcharge (CNY bn) 0.6 1.4 2.2 4.2 9.6 8.1 16.9 23.5 46.1 72.2 86.8 107.2 120.5 134.5 149.3 Total power consumption (bn kWh) 2,837 3,253 3,438 3,660 4,200 4,703 4,959 5,322 5,523

- Primary industry consumption (bn kWh) 83 84 88 94 98 101 101 101 99 - residential consumption (bn kWh) 324 358 408 458 509 562 622 679 693 Consumption entitled to surcharge (bn kWh) 2,430 2,811 2,942 3,108 3,593 4,039 4,236 4,542 4,731 4,968 5,206 5,446 5,685 5,924 6,161

% YoY growth

16% 5% 6% 16% 12% 5% 7% 4% 5% 5% 5% 4% 4% 4% RE Surcharge (CNY per kWh) 0.0005 0.0010 0.0015 0.0023 0.0040 0.0040 0.0080 0.0103 0.0150 0.0224 0.0256 0.0303 0.0303 0.0303 0.0303 % RE surcharge received

50% 50% 50% 58% 67% 50% 50% 50% 65% 65% 65% 65% 70% 75% 80%

Subsidy from MOF (CNY bn)

8.6 14.8 Total source of fund (CNY bn) 0.6 1.4 2.2 4.2 9.6 8.1 25.5 38.3 46.1 72.2 86.8 107.2 120.5 134.5 149.3

Use of fund Wind power (CNY bn) 0.2

5.0 8.2 11.1 13.0 18.2 24.5 35.5 43.8 51.0 51.9 58.4 56.8

- Power generation (bn kWh) - - 13.1 27.6 49.4 74.1 103.0 140.1 156.3 200.74 240.74 280.74 320.74 360.74 400.74 - Wind FiT (CNY per kWh)

0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.55 0.55 0.53 0.53 0.51

- Thermal on-grid tariff (CNY per kWh)

0.35 0.35 0.38 0.39 0.41 0.43 0.43 0.40 0.38 0.36 0.36 0.36 0.36 0.36 Installation (MW) 2,070 4,200 8,390 17,600 29,580 46,230 61,420 76,520 96,371 116,371 136,371 156,371 176,371 196,371 216,371 Utilisation hours

2,596 2,699 2,470 2,250 2,092 2,174 1,942 2,000 2,000 2,000 2,000 2,000 2,000

Solar power (CNY bn) 0.0

4.7 12.1 19.9 29.1 39.3 42.9 50.9 54.4 - Power generation (bn kWh)

8.5 23.1 37.0 56.8 77.1 93.8 111.7 130.2

- Solar utility FiT (CNY per kWh)

1.00 0.95 0.95 0.90 0.90 0.85 0.85 0.80 - Thermal on-grid tariff (CNY per kWh)

0.43 0.40 0.38 0.36 0.36 0.36 0.36 0.36

- Solar utility subsidy (CNY per kWh)

0.57 0.55 0.57 0.54 0.54 0.49 0.49 0.44 - Solar DG subsidy (CNY per kWh)

0.42 0.42 0.42 0.42 0.42 0.37 0.37 0.37

Installation (MW) 80 100 140 300 800 3,300 6,498 17,460 28,070 44,570 62,570 76,470 91,570 107,170 123,170 - Distributed (MW)

2,300 2,630 4,680 8,680 13,680 17,680 22,680 28,680 35,680

- Utility scale (MW)

4,198 14,830 23,390 35,890 48,890 58,790 68,890 78,490 87,490 Utilisation hours

1,180 1,180 1,180 1,180 1,180 1,180 1,180

Other renewable energies (CNY bn) 0.0

0.5 2.2 5.7 14.5 6.3 19.8 16.9 14.3 17.2 20.4 23.6 26.8 Biomass (bn kWh)

23.5

WTE (bn kWh)

13.5 Total other renewable energies (bn kWh)

20.0

37.0

78.0 60.3 72.5 86.1 99.7 113.3

- Average FiT (CNY per kWh)

0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60 - Thermal on-grid tariff (CNY per kWh)

0.38 0.39 0.41 0.43 0.43 0.40 0.38 0.36 0.36 0.36 0.36 0.36

Installation (MW) 2,000 2,800 3,600 3,200 5,500 7,184 8,000 8,500 10,750 13,000 16,400 19,800 23,200 26,600 30,000

4,598

4,405 4,305 4,205 4,105 4,005 4,005 4,005 4,005

Total use of fund (CNY bn) 0.3 - - 5.5 10.3 16.8 27.5 29.2 56.5 72.3 87.1 107.5 115.2 132.9 138.0 Cumulative shortage (CNY bn)

1.3 2.0 10.7 12.7 4 14.0 14 14 15 9 8 (4)

Yearly shortage (CNY bn)

1.3 0.7 8.7 2.0 (9.1) 10.4 0 0 0 (5) (2) (11) Source: Daiwa

68


Appendix VII: yieldco structure to unlock long-term volume growth potential We believe China’s long-term solar-capacity growth will also be driven by the emergence of innovative financing channels. With increasing examples of successful innovative energy financing in developed markets (especially the US), we believe alternative financing channels, especially the “yieldco” structure, will provide significant business-growth opportunities to renewable energy developers in China. Difficulties in financing a major hindrance We believe that one of the major obstacles for downstream solar operators in China is funding difficulties. Many of the solar companies are either integrated solar companies with both upstream and downstream assets, or pure solar companies with ultra-high leverage. The main drawback of these structures is that those companies may find it hard to expand their solar capacity rapidly due to funding difficulties. For solar equipment manufacturers, many of China’s commercial banks are unwilling to lend too aggressively to them, given the nightmare of widespread bankruptcies across the solar industry during 2011-12 (eg, Wuxi Suntech, LDK, Chaori, etc.). As such, once the upstream manufacturers expand to a certain extent and achieve certain gearing levels, they may not be able to borrow further or only at very high interest costs. As such, we note that many module manufacturers, such as Trina, Canadian Solar, and Jinko Solar, have relatively slow expansion plans after they have developed and owned several hundred megawatts of solar farms. For pure downstream solar farm developers such as United PV and GCL New Energy, the main problem is their high gearing level (usually above 300% net gearing). With stretched balance sheets, their financing rate could be as high as 10%, substantially reducing their equity IRRs, as well as limiting their choice of project acquisitions given that many solar projects become value destructive at such high interest costs. What is a “yieldco”? A yieldco is essentially a company created by renewable energy operators to hold long-term contracted operating power generation assets. It offers the characteristics of infrastructure investment (predictable and growing dividends) to its shareholders. Usually a yieldco is a dual-class share structure, with Class A issuing to external shareholders, and Class B remained held by the parent company for controlling voting interests. See the simplified illustration below:

Solar farm developers may encounter difficulties in getting enough financings

69


Illustration of a yieldco structure

Source: Daiwa research The idea of a yieldco is not new; it came from master limited partnerships (MLP) structure, a capital formation mechanism funding energy infrastructure investments in North America. The reason why renewable energy companies cannot set up an MLP structure is that the criteria to be legally classified as an MLP is that over 90% of cash flow must come from real estate, natural resources and commodities. As such, a yieldco structure is also known as a “synthetic MLP”. Benefits of setting up a yieldco The main purpose of creating a yieldco is to segregate the more risky assets, such as projects in construction or research and development units, from the more stable operating power-generation units, in order to unlock the value of those low-risks assets. Compared with other financing channels, using a yieldco as a form of financing has multiple benefits for renewable energy developers, including the following: • Value enhancement for the yieldco and parent company: projects eligible to be

dropped down to the yieldco should fulfil a few requirements: 1) an operating track record of at least 3 months, and 2) signed long-term power purchase agreements (PPAs). From these requirements, the assets under the yieldco are deemed low-risk with predictable cash flows, which usually command a higher valuation than the original company with both risky and low-risk assets combined.

• Lower cost of capital. Due to the high valuation, yieldcos in the US usually pay 4-6% dividends in the current market situation (refer to table below), which is considered a cheap source of funding for the parent company when compared with similar assets. Also, a growth premium is embedded in the valuation, resulting in a reduced average cost of capital for the parent company.

• Tax shelters: Both MLP and yieldco enjoy a single layer of taxation, meaning that profits are only taxed at the shareholder but not corporate level. While the MLP is legally exempt from taxation at the corporate level, the yieldco “synthesises” it. Given that the yieldco usually has a high depreciation versus its revenue during the initial years, a yieldco is usually in a tax loss position initially. Combined with possible tax credits, accelerated depreciation, and continuous acquisitions, a yieldco could be tax-sheltered for years.

• Deleveraging the parent company: downstream power generation operators suffer from high leverage, which in turn limits the company’s ability to source cheap financing. Spinning off the company’s assets in the form of yieldco effectively increases its equity base without diluting existing shareholders’ interests.

100% Class A UnitsX% Economic Interest

100% Class B Units100-X% Economic

Interest

Class B Common Stock100-X% Voting Interest0% Economic Interest

Class A Common StockX% Voting InterestX% Economic Interest

Parent company

Operatingsubsidiaries

Operating assets

Public shareholders

100%

YIELDCO

We identify at least 4 benefits of setting up and listing a yieldco

70


Listed yieldcos in the US and Canada Market cap Total capacity (MW) Dividend yield (%) Stock code Listing date (USD m) Type of assets (end-2014) Parent company 2014 2015E 2016E NRG Yield, Inc. NYLD US 7/17/2013 3,931 Solar/Wind/Thermal 2,985 NRG Energy Inc. 3.01 3.4 4.05 Pattern Energy Group, Inc. PEGI US 9/27/2013 2,023 Wind 1,636 Pattern Development 5.27 4.94 5.63 Abengoa Yield Plc. ABY US 6/13/2014 2,731 Solar/Thermal 2,246 Abengoa (ABGB US) - 4.87 6.28 TransAlta Renewables, Inc. RNW CN 8/9/2013 1,148 Hydro/Wind 1,376 TransAlta Corp. 5.59 6.12 6.17 NextEra Energy Partners, LP NEP US 6/27/2014 4,114 Solar/Wind 10,351 NextEra Energy (NEE US) 0.58 2.35 2.86 TerraForm Power, Inc. TERP US 7/18/2014 4,763 Solar/Wind 1,507 SunEdison (SUNE US) - 3.4 4.19 8point3 Energy Partners Pending listing - - Solar 432 First Solar / Sunpower - - -

Source: Bloomberg, Company data, Daiwa Opportunity comes only to those who are prepared Currently, some of China’s solar companies are considering separate listings of yieldcos in 2016, including, we believe, GCL New Energy and Canadian Solar. We expect an increasing number of solar companies to come up with similar plans in the future. Despite all the attendant benefits, we believe that not all solar companies are qualified for spinning off a yieldco, with the differentiating factors highlighted below: Capacity threshold should be met. Before any company can establish a yieldco for a separate listing, it should have sufficient operating assets in order that the listed shares have enough liquidity. According to Martin, K. in "Yield Cos Compared", a yieldco should hold a project value of at least USD500m for an IPO. Large project pipeline. A yieldco structure has a snowball effect which can lead to accelerated capacity growth. However, we should note that the growth rate could be capped by limited solar pipeline projects for the company itself, significantly dampening the growth premium of the yieldco. Quality assets. Only quality assets should be injected to the yieldco, given that a yieldco is usually characterised by having lower-risk and predictable cash flows. As such, solar farms with low-quality assets, such as substandard modules or serious grid curtailment issues, may not fit the bill. As such, we expect a gradual divergence in the performances of downstream developers in the future, with the ones qualifying for a yieldco more likely to experience accelerated capacity growth, and the ones failing to fulfil the above factors being weeded out by the industry eventually. In this context, we prefer solar developers with quality assets and ample project pipelines. ABS might be a short-term alternative However, we still see challenges for setting up a yieldco in the short term, such as the possible CNY depreciation and the potential US rate hikes. As such, we believe a short-term alternative might be ABS products. An introduction to ABS An asset-backed security (ABS) is an investment product which is bond-like and whose payments are derived from a specified pool of underlying assets. Typically, an ABS consists of receivables, loans, manufactured-housing contracts, etc. ABS differs from other bond-like structures because its creditworthiness is dependent on underlying assets, rather than on the financial originator. The ABS is usually structured in a way that a special-purpose vehicle (SPV) is set up to contain all the assets (pools of loans and receivables), and it subsequently sells them to a trust, which in turn repackages the assets as interest-bearing securities and issues them to the investors. This completes the securitisation process by which the ABS is formed.

To spin off the yieldco, the companies need to have enough operating capacity, ample project pipeline, as well as quality solar farm assets

We see yieldco listing facing challenging over the short term, which we believe companies may opt for ABS as an alternative financing channel

71


We believe ABS issuances will become more popular in China over the next few years, as we see increasing needs for the renewable-energy operators to obtain financing. We also see rising attention from China’s government on developing “green financing” (ie, financing instruments specifically for the renewable energy sector, such as “green bonds”), as evidenced by the PBOC’s plan to issue guidelines for this in the near term, according to a special report by Xinhua Net during the Fifth Plenum.

Solar securitisation process ABS tranching

Source: NREL Source: NREL The table below summarises the major benefits of securitising solar farms.

ABS: major benefits Benefit Explanation Risk mitigation Assets are removed from their originator’s balance sheet and thus insulated from the parent’s corporate risk. The assets are then said to be made

“bankruptcy remote.” The process of pooling assets serves as diversification to mitigate credit risk, geographic concentration risks. Securitisation improves the resulting credit rating and thus cost of capital.

Access to Broader Capital Pool Securitisation can open up the investment assets to investors that might otherwise be out of reach by other financing alternatives. Improvements in Financing Terms Asset originators can bargain for more favourable financing terms than they may otherwise be able to obtain from traditional sources, including longer

tenors and lower cost of capital. Opportunities for Market Growth Originators have a means to free up balance sheet capacity with the option to offload their assets into an SPV. Also, originators could monetise assets that

were previously illiquid. Investor demand for a particular securitised asset may create demand pull, which can incentivise the upstream market to originate more assets.

Source: NREL, Daiwa

72



China Industrials

What's new: We expect the pace of solar installations to pick up in China in 4Q15, with the country likely to achieve 16GW of installed solar capacity (9M15: 10.5GW) for the full year, implying 5.5GW for 4Q15. As such, Singyes should be the major beneficiary of both this boom over 4Q15-2016 and the retail parity revolution that we see for China starting in 2017. What's the impact: Accelerated solar EPC volume growth: The coming boom should benefit Singyes in 2 ways: 1) it should increase its EPC volume, and 2) it should lead to it developing more solar farms, which should increase its power-generation revenue. In the longer run, we also see it benefiting from the solar DG boom starting from 2017, once China achieves retail parity. In our model, we forecast Singyes to deliver solar EPC volume of 600MW in 2016 and 750MW in 2017. Solar EPC gross margin should remain resilient: Singyes has demonstrated its ability to obtain solar project approval from the government, as evidenced by the 470MW in solar permits it was granted in 2015. We expect it to maintain this core advantage; in other words, it should be able to maintain its utility-scale solar EPC gross margin over 2015-17. This, combined with our expectation that it will increase its revenue contribution from the higher-margin solar DG EPC business, leads us to expect an overall EPC gross margin of 25-26% for 2015-17. What we recommend: We reiterate our Buy (1) rating on the stock. Singyes is our top pick within the China solar sector for the reasons stated above. Our 12-month TP of HKD8.50 is based on a 6.5x 2016E PER, which we see as attractive (6.5x was the stock’s average one-year forward PER in 2012-13, when Singyes’ valuation bottomed out). The stock is trading at a 4.9x 2016E PER, and a 0.8x 2016E PBR, its past-5-year valuation troughs (1SD below its past-5-year PER mean; its 1-year forward PER has been above its current 1-year forward PER for 76% of the time since 2011). This low valuation is in contrast to our view of the company’s improving fundamentals if the solar installation boom occurs. The key risk would be slower-than-expected solar EPC volume growth. How we differ: Our 2016-17E EPS are 3-6% above consensus due to our more positive outlook over the period, as we expect Singyes to recover solar EPC volume, and also backed by the accelerating pace of solar DG installations in the country in 2017E.

29 October 2015

China Sing yes Sol ar Technologies

Major beneficiary of retail parity revolution

Singyes’ solar EPC volume should accelerate in 2016 Getting permits remains its core advantage, sustaining EPC margins Reiterate Buy (1) rating; 5x 2015E PER is appealing; top sector pick


Source: FactSet, Daiwa forecasts

China Singyes Solar Technologies (750 HK)

Target price: HKD8.50 (from HKD8.50)

Share price (27 Oct): HKD6.58 | Up/downside: +29.1%

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

Forecast revisions (%)Year to 31 Dec 15E 16E 17ERevenue change - - -Net profit change - - -Core EPS (FD) change - - -

30

49

68

86

105

4

7

10

12

15

Oct-14 Jan-15 Apr-15 Jul-15 Oct-15

Share price performance

CSST (LHS) Relative to HSI (RHS)

(HKD) (%)

12-month range 4.33-14.86Market cap (USDbn) 0.593m avg daily turnover (USDm) 5.24Shares outstanding (m) 696Major shareholder Strong Eagle Holdings (35.3%)

Financial summary (CNY)Year to 31 Dec 15E 16E 17ERevenue (m) 5,968 7,415 8,299Operating profit (m) 996 1,234 1,433Net profit (m) 505 734 871Core EPS (fully-diluted) 0.858 1.108 1.297EPS change (%) 24.9 29.1 17.1Daiwa vs Cons. EPS (%) (6.0) 3.0 6.4PER (x) 6.3 4.9 4.2Dividend yield (%) 1.6 1.7 2.0DPS 0.088 0.092 0.109PBR (x) 1.0 0.8 0.7EV/EBITDA (x) 5.7 4.9 4.2ROE (%) 14.5 17.6 17.7

74

China Singyes Solar Technologies (750 HK): 29 October 2015



Growth outlook Singyes: revenue forecast by segment

We forecast a 23% EPS CAGR over 2015-17, mainly driven by average annual revenue growth of 18% over the same period, and offset partially by its likely slightly contracting solar EPC margin. We see Singyes’ solar EPC business picking up during 4Q15, and expect it to deliver solar EPC volume of 400MW in 2015, 600MW in 2016 and 750MW in 2017.

Source: Company, Daiwa forecasts

Valuation Singyes: Rolling 1-year forward PER

The stock is trading currently at a 4.9x 2016E PER, and a 0.8x 2016E PBR, both 5-year troughs (1SD below its past-5-year PER mean – its 1-year forward PER has been above its current 1-year forward PER for 76% of the time since 2011). This low valuation is not in line with our view of the company’s improving fundamentals, due to the solar installation boom that we expect during 2015-16. We believe this upcoming upcycle should warrant a higher PER multiple for Singyes. Our target price of HKD8.5 implies a 2016E PER of 6.5x, which is still 0.6SD below its past 5-year PER average.


Earnings revisions Singyes: Bloomberg consensus EPS forecast revisions

The main reasons for the downward revisions since 4Q14 to the Bloomberg consensus 2015 EPS forecasts for Singyes have been the slower-than-expected pace of solar project installations in China in 2015 YTD, as well as the country’s poor real estate market this year. While we believe there could be further downside to the consensus numbers for 2015, we see Singyes’ fundamentals bottoming out in 4Q15, which leads us to our above-consensus EPS forecasts for 2016-17 (our EPS are above consensus by 3% for 2016E and 6% for 2017E).

Source: Bloomberg

0%

10%

20%

30%

40%

0

2,000

4,000

6,000

8,000

10,000

2012 2013 2014 2015E 2016E 2017E

(CNYm)

Solar EPC Conventional curtain wallPower generation OthersYoY growth - RHS

2

4

6

8

10

12

14

16

Jan-11 Jan-12 Jan-13 Jan-14 Jan-15

(x)

14.3x Avg+2SD

11.4x Avg+1SD

8.4x Avg

5.4x Avg-1SD

2.5x Avg-2SD

0.91.01.11.21.31.41.51.6

Mar-1

3

Jun-

13

Sep-

13

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(CNY)

2015E EPS 2016E EPS 2017E EPS

75


Financial summary Key assumptions

Profit and loss (CNYm)

Cash flow (CNYm)


Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar EPC volume (MW) 20 50 83 220 240 400 600 750Solar EPC gross margin (%) 35.8 33.7 35.5 29.5 26.7 26.0 25.4 26.2Installed solar capacity (MW) 0 0 0 180 171 265 365 465Conventional curtain walls revenue growth (%) (0.3) 28.4 24.6 12.1 21.9 (17.2) (5.0) 2.0

Conventional curtain walls gross margin (%) 15.8 15.0 15.4 16.1 15.3 14.8 14.3 14.3

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar EPC 561 826 1,072 1,889 2,064 3,350 4,827 5,668Conventional curtain walls business 764 981 1,223 1,371 1,671 1,383 1,314 1,340Other Revenue 458 599 803 891 1,276 1,235 1,274 1,290Total Revenue 1,782 2,406 3,098 4,151 5,011 5,968 7,415 8,299Other income 7 40 18 71 223 187 218 246COGS (1,345) (1,826) (2,348) (3,162) (3,962) (4,686) (5,810) (6,447)SG&A (132) (185) (204) (227) (246) (308) (395) (442)Other op.expenses (17) (30) (58) (101) (216) (164) (194) (224)Operating profit 294 405 505 731 810 996 1,234 1,433Net-interest inc./(exp.) (13) (34) (85) (104) (227) (282) (326) (360)Assoc/forex/extraord./others 0 0 0 0 116 150 0 0Pre-tax profit 281 372 420 627 699 864 909 1,073Tax (65) (80) (93) (136) (114) (141) (148) (175)Min. int./pref. div./others (0) (1) 2 (0) (1) (1) (1) (1)Net profit (reported) 216 291 329 491 584 722 759 897Net profit (adjusted) 223 297 329 471 415 505 734 871EPS (reported)(CNY) 0.441 0.472 0.521 0.752 0.842 1.038 1.091 1.288EPS (adjusted)(CNY) 0.454 0.482 0.521 0.722 0.597 0.726 1.054 1.251EPS (adjusted fully-diluted)(CNY) 0.441 0.471 0.521 0.737 0.687 0.858 1.108 1.297DPS (CNY) 0.030 0.032 0.057 0.071 0.071 0.088 0.092 0.109EBIT 294 405 505 731 810 996 1,234 1,433EBITDA 305 429 560 798 924 1,140 1,409 1,636

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017EProfit before tax 281 372 420 627 699 864 909 1,073Depreciation and amortisation 11 24 54 67 114 144 174 204Tax paid (45) (23) (124) (104) (141) (141) (148) (175)Change in working capital (260) 34 (78) 76 (766) (667) (327) (213)Other operational CF items 36 159 206 463 185 195 280 314Cash flow from operations 23 565 479 1,129 91 395 887 1,203Capex (225) (974) (369) (1,332) (1,185) (920) (900) (858)Net (acquisitions)/disposals (13) 3 (0) 0 18 0 0 0Other investing CF items (24) (21) (170) (105) (43) 32 29 29Cash flow from investing (262) (992) (540) (1,437) (1,209) (888) (871) (829)Change in debt 344 189 541 344 (212) 354 337 543Net share issues/(repurchases) (0) 195 0 256 0 0 0 0Dividends paid (15) (15) (17) (37) (50) (49) (61) (64)Other financing CF items (10) (42) (80) (57) 1,490 (74) (288) (853)Cash flow from financing 319 328 444 507 1,229 231 (12) (374)Forex effect/others (3) 1 (1) (1) (0) 0 0 0Change in cash 77 (99) 383 197 111 (261) 4 0Free cash flow (202) (409) 110 (203) (1,093) (525) (13) 345

76


Financial summary continued … Balance sheet (CNYm)

Key ratios (%)


As at 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ECash & short-term investment 414 315 698 895 901 486 490 490Inventory 19 64 63 77 115 136 168 186Accounts receivable 795 992 1,281 1,801 2,192 2,774 3,345 3,744Other current assets 149 139 323 694 1,208 1,437 1,662 1,795Total current assets 1,378 1,510 2,366 3,467 4,417 4,832 5,665 6,215Fixed assets 267 1,261 1,545 2,680 3,455 4,234 4,962 5,618Goodwill & intangibles 1 5 3 2 2 1 1 1Other non-current assets 92 174 149 152 336 333 331 329Total assets 1,737 2,950 4,064 6,301 8,209 9,400 10,959 12,162Short-term debt 300 569 917 870 988 800 853 947Accounts payable 74 298 515 1,293 1,284 1,391 1,803 2,090Other current liabilities 103 306 201 216 519 577 666 717Total current liabilities 477 1,173 1,634 2,378 2,791 2,768 3,322 3,753Long-term debt 177 96 289 681 1,751 2,379 2,730 2,715Other non-current liabilities 20 153 289 642 526 439 393 347Total liabilities 674 1,422 2,212 3,702 5,068 5,586 6,446 6,815Share capital 34 36 43 46 46 46 46 46Reserves/R.E./others 1,010 1,488 1,802 2,553 3,094 3,766 4,465 5,297Shareholders' equity 1,044 1,524 1,845 2,599 3,140 3,813 4,511 5,344Minority interests 19 5 7 0 1 1 2 3Total equity & liabilities 1,737 2,950 4,064 6,301 8,209 9,400 10,959 12,162EV 3,838 4,111 4,271 4,413 5,594 6,451 6,851 6,931Net debt/(cash) 63 350 508 656 1,837 2,693 3,093 3,172BVPS (CNY) 2.126 2.899 2.915 3.757 4.515 5.475 6.478 7.674

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESales (YoY) n.a. 35.0 28.7 34.0 20.7 19.1 24.3 11.9EBITDA (YoY) n.a. 40.7 30.4 42.6 15.8 23.5 23.5 16.2Operating profit (YoY) n.a. 37.7 24.6 44.6 10.8 23.0 23.9 16.1Net profit (YoY) n.a. 33.1 10.7 43.4 (12.0) 21.8 45.2 18.7Core EPS (fully-diluted) (YoY) n.a. 6.9 10.5 41.4 (6.7) 24.9 29.1 17.1Gross-profit margin 24.5 24.1 24.2 23.8 20.9 21.5 21.7 22.3EBITDA margin 17.1 17.8 18.1 19.2 18.4 19.1 19.0 19.7Operating-profit margin 16.5 16.8 16.3 17.6 16.2 16.7 16.6 17.3Net profit margin 12.5 12.3 10.6 11.4 8.3 8.5 9.9 10.5ROAE n.a. 23.1 19.5 21.2 14.5 14.5 17.6 17.7ROAA n.a. 12.7 9.4 9.1 5.7 5.7 7.2 7.5ROCE n.a. 21.7 19.2 20.3 16.2 15.5 16.4 16.8ROIC 20.1 21.2 18.6 20.4 16.5 14.5 14.6 14.9Net debt to equity 6.0 23.0 27.6 25.2 58.5 70.6 68.6 59.4Effective tax rate 23.1 21.5 22.2 21.7 16.3 16.3 16.3 16.3Accounts receivable (days) n.a. 135.5 133.9 135.5 145.4 151.9 150.6 155.9Current ratio (x) 2.9 1.3 1.4 1.5 1.6 1.7 1.7 1.7Net interest cover (x) 22.7 12.0 5.9 7.0 3.6 3.5 3.8 4.0Net dividend payout 6.8 6.9 10.9 9.4 8.4 8.4 8.5 8.5Free cash flow yield n.a. n.a. 2.9 n.a. n.a. n.a. n.a. 9.2

Company profile

China Singyes Solar Technologies (Singyes) specialises in building integrated photovoltaic (BIPV) systems and green buildings, such as installing conventional curtain walls. It has 3 production bases: in Zhuhai Nanping (R&D), Hunan Xiangtan (manufacturing), and Zhuhai Hi-tech Park (exports).

China Information Technology

Investment case: We initiate coverage of solar-module manufacturer JinkoSolar (JKS) with an Outperform (2) rating. JKS is our preferred pick within the solar module manufacturing sector, as: 1) we believe, that as a cost leader, JKS will benefit the most among the China module makers, from the upcoming solar installation boom over 4Q15-2016, and 2) the listing of its downstream solar subsidiary, Jinko Power (JKP), slated for mid-2016, would unlock the value of JKS’s downstream solar business. Improving solar module gross margins: For the reasons mentioned above, we expect JKS’s solar module sales volume to increase from 2.4GW for 2014 to 3.5GW for 2015E and 4.1GW for 2016E, up 44% and 18% YoY, respectively. The strong global demand we expect in 2016 should also lead to JKS’s solar module gross margin expanding slightly from 19.5% for 2015E to 20.3% for 2016E (vs. 15-17% for its peers). Accelerated downstream growth, with limited equity dilution risk: We expect JKS’s downstream solar development to accelerate over the next few years (we forecast ~600MW capacity additions per year over 2015-17), and see minimal risk of equity dilution. This could be achieved through the company’s strategy of developing solar projects through its JV, JKP. Potential listing of JKP in mid-2016 should unlock value: We expect JKP’s IPO listing to be completed in mid-2016, with 3 major benefits: 1) value enhancement for JKS, whose fair value we expect to rise by 34% to USD38.1, 2) an improvement in net gearing by end-2016, from 213% pre-IPO to 78% post-IPO, due to the company deconsolidating its capital-intensive solar IPP business, and 3) an immediate non-cash disposal gain of ~CNY86m,on our estimates. Catalysts: Major catalysts: 1) a rush of solar installations in China and the US in 2016 ahead of subsidy cuts, and 2) the listing of JKP in mid-2016. Valuation: We have a SOTP-derived 12-month target price of USD28.5. The stock is trading currently at a 5.7x 2016E PER, which we see as attractive. In our view, JKS stands to benefit from a solar industry upcycle from 4Q15, with improving demand and margins from 4Q15-16E. Our TP of USD28.5 implies a 2016E PER of 6.5x, which we see as undemanding given the company’s 2016E ROE of 21.5x and EPS CAGR of 19% over 2015-17E. Risks: 1) a slower-than-expected pace of solar farm installations, 2) a worsening of solar equipment international disputes, 3) slower-than-expected solar project development.

29 October 2015

JinkoSolar

Initiation: a rising star with more value to be unlocked

Cost leader stands to benefit from the 4Q15-16 solar installation boom

Accelerated downstream volume growth; JKP listing could unlock value

Initiate with Outperform (2) rating; rerating likely on industry upcycle


JinkoSolar (JKS US)

Target price: USD28.50Share price (27 Oct): USD25.28 | Up/downside: +12.7%

Scott Chui(852) 2848 4443

[email protected]

Dennis Ip, CFA(852) 2848 4068

[email protected]

60

79

98

116

135

14

19

23

28

32



JinkoSolar (LHS)Relative to S&P 500 Index (RHS)

(USD) (%)

12-month range 15.68-31.85Market cap (USDbn) 0.783m avg daily turnover (USDm) 2.22Shares outstanding (m) 31Major shareholder Li Xiande (16.2%)

Financial summary (CNY)Year to 31 Dec 15E 16E 17ERevenue (m) 13,522 15,726 16,346Operating profit (m) 1,405 1,814 2,268Net profit (m) 786 1,033 1,174Core EPS (fully-diluted) 22.035 27.936 31.258EPS change (%) 1.9 26.8 11.9Daiwa vs Cons. EPS (%) 1.0 9.2 (6.1)PER (x) 7.3 5.7 5.1Dividend yield (%) 0.0 0.0 0.0DPS 0.000 0.000 0.000PBR (x) 1.2 0.9 0.8EV/EBITDA (x) 4.5 4.2 4.0ROE (%) 20.2 21.5 19.8


78

JinkoSolar (JKS US): 29 October 2015



Growth outlook JKS: revenue

With a potential boom in solar module demand in China and the US from 4Q15, on the back of policy tailwinds and continual declines in PV prices, JKS should continue to see robust solar module revenue growth. As such, we forecast its solar module revenue to rise by 33% and 14% YoY for 2015-16E, respectively. We also expect its profitability to improve on: 1) a rise in margins for its solar module business on the aforementioned solar installation boom in China in 2016, and 2) an increasing contribution from its higher-margin solar farm business. These factors, combined with our revenue growth forecasts, should lead to diluted EPS growth of 2% and 27% YoY for 2015-16E.


Valuation JKS: one-year forward PER bands

The stock is trading currently at a 5.7x 2016E PER and 0.9x 2016E PBR, both below its past-3-year trading averages. We think the current PER valuation is appealing given: 1) our 2015-17E EPS CAGR forecast of 19%, and 2) our expectation of a solar installation boom in China and the US over 4Q15-2016. Our target price of USD28.5 per ADS implies a 2016E PER of 6.5x, and 1.3x 2016E PBR, which we see as undemanding, given that we forecast a 2016E ROE of 21.5x and EPS CAGR of 19% over 2015-17E.


Earnings revisions JKS: Bloomberg consensus earnings revisions

The Bloomberg-consensus EPS forecasts for 2015-16 were relatively stable in 1H15, after declining expectations were factored in during 2H14, mainly driven by the weaker-than-expected pace of solar installations in China. From June 2015, however, we have started to see upward earnings revisions, possibly due to the market’s increasing expectations from the upcoming solar installation boom that we expect. That said, we believe there could be further upward revisions for 2016E, as we believe the company’s gross margin could increase by more than the market expects. Our 2016E EPS is now 9% above consensus.

Source: Bloomberg

02,0004,0006,0008,000

10,00012,00014,00016,00018,000

2012 2013 2014 2015E 2016E 2017E

(CNYm)

Solar module Silicon wafer Solar cell Power generation Others

02468

1012141618

Dec-1

2

Mar-1

3

Jun-

13

Sep-

13

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(x)

16.2x Avg+2SD

12.0x Avg+1SD

7.7x Avg

3.4x Avg-1SD

15

20

25

30

35

40

45

Jan-14 Apr-14 Jul-14 Oct-14 Jan-15 Apr-15 Jul-15 Oct-15

(CNY)


79



Profit and loss (CNYm)

Cash flow (CNYm)


Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar module sales volume (MW) 0 0 912 1,765 2,423 3,500 4,133 4,409Solar module ASP (CNY per W) 12.2 8.7 4.3 3.8 3.8 3.5 3.4 3.2Solar module non-silicon costs (CNY per W) 5.3 4.4 3.1 2.5 2.2 2.1 2.0 1.8

Solar module silicon costs (CNY per W) 3.4 2.2 0.8 0.6 0.6 0.5 0.5 0.5

Solar module unit gross profit (CNY per W) 3.6 2.1 0.4 0.7 0.9 0.9 0.9 0.9

Solar farm attributable capacity (MW) 0 9 9 213 503 1,103 1,703 2,303

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar module 3,248 6,647 3,897 6,660 9,155 12,166 13,935 13,975Solar power generation 0 0 2 77 244 776 1,211 1,791Other Revenue 1,407 738 896 342 579 579 579 579Total Revenue 4,655 7,385 4,795 7,079 9,979 13,522 15,726 16,346Other income (1) 28 4 7 (2) 0 0 0COGS (3,297) (6,235) (4,563) (5,641) (7,738) (10,356) (11,864) (11,948)SG&A (330) (758) (1,104) (723) (1,193) (1,616) (1,880) (1,954)Other op.expenses (37) (76) (362) (69) (116) (144) (168) (175)Operating profit 989 344 (1,229) 652 930 1,405 1,814 2,268Net-interest inc./(exp.) (64) (183) (222) (223) (288) (399) (533) (687)Assoc/forex/extraord./others 104 193 (102) (221) (50) 66 66 66Pre-tax profit 1,028 354 (1,553) 208 592 1,072 1,346 1,648Tax (146) (81) 9 (19) 134 (106) (130) (153)Min. int./pref. div./others (46) 0 1 (1) (53) (180) (184) (321)Net profit (reported) 836 273 (1,542) 188 673 786 1,033 1,174Net profit (adjusted) 784 397 (268) 360 767 786 1,033 1,174EPS (reported)(CNY) 44.638 11.636 (69.515) 7.999 21.890 25.295 33.235 37.767EPS (adjusted)(CNY) 41.882 16.899 (12.085) 15.311 24.937 25.295 33.235 37.767EPS (adjusted fully-diluted)(CNY) 38.851 15.464 (12.085) 14.989 21.632 22.035 27.936 31.258DPS (CNY) n.a. 0.000 0.000 0.000 0.000 0.000 0.000 0.000EBIT 989 344 (1,229) 652 930 1,405 1,814 2,268EBITDA 1,089 614 (894) 1,016 1,372 2,014 2,625 3,273

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017EProfit before tax 1,028 354 (1,553) 208 592 1,072 1,346 1,648Depreciation and amortisation 101 270 336 363 442 610 811 1,005Tax paid (146) (81) 9 (19) 134 (106) (130) (153)Change in working capital (768) (1,617) 554 (205) (1,100) (241) (173) (162)Other operational CF items 80 491 1,593 501 406 322 479 660Cash flow from operations 295 (584) 938 848 475 1,657 2,334 2,998Capex (1,308) (2,107) (358) (993) (3,002) (4,382) (4,206) (4,086)Net (acquisitions)/disposals 0 (459) (1,387) (1,824) (2,521) (576) (371) (104)Other investing CF items (245) 133 1,244 1,530 1,364 (129) (80) (23)Cash flow from investing (1,553) (2,434) (501) (1,286) (4,159) (5,087) (4,657) (4,213)Change in debt 516 2,051 (636) 430 1,236 3,068 2,944 2,860Net share issues/(repurchases) 816 728 0 431 3,057 0 0 0Dividends paid 0 0 0 0 0 0 0 0Other financing CF items 300 159 44 (243) 723 431 (55) (660)Cash flow from financing 1,633 2,938 (592) 618 5,016 3,498 2,888 2,200Forex effect/others 0 0 0 0 0 0 0 0Change in cash 369 (87) (155) 177 1,321 68 565 985Free cash flow (1,013) (2,691) 580 (144) (2,527) (2,726) (1,872) (1,088)

80


Financial summary continued … Balance sheet (CNYm)

Key ratios (%)


As at 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ECash & short-term investment 973 1,074 1,142 1,589 4,026 4,864 5,922 7,045Inventory 820 798 528 712 1,891 2,531 2,899 2,920Accounts receivable 577 1,631 1,818 1,933 3,293 4,462 5,189 5,394Other current assets 825 1,105 497 1,109 1,197 1,315 1,371 1,383Total current assets 3,194 4,608 3,986 5,343 10,407 13,173 15,381 16,742Fixed assets 1,939 3,841 3,866 4,546 7,455 11,236 14,639 17,727Goodwill & intangibles 308 372 372 366 382 373 365 357Other non-current assets 438 355 148 357 844 1,115 1,290 1,351Total assets 5,880 9,176 8,372 10,611 19,088 25,897 31,675 36,177Short-term debt 1,172 3,240 2,559 2,812 2,674 2,067 2,788 2,067Accounts payable 355 377 1,377 1,768 3,149 4,214 4,828 4,862Other current liabilities 1,415 2,026 2,302 2,658 4,701 6,305 7,243 7,338Total current liabilities 2,942 5,643 6,238 7,237 10,523 12,586 14,859 14,267Long-term debt 269 543 651 1,162 3,297 6,971 9,194 12,775Other non-current liabilities 4 85 109 191 1,735 2,010 2,250 2,269Total liabilities 3,215 6,271 6,999 8,591 15,555 21,567 26,303 29,311Share capital 0 0 0 0 0 0 0 0Reserves/R.E./others 2,665 2,895 1,365 2,010 3,507 4,293 5,326 6,499Shareholders' equity 2,665 2,895 1,365 2,010 3,507 4,293 5,326 6,499Minority interests 0 10 9 11 26 37 46 367Total equity & liabilities 5,880 9,176 8,372 10,611 19,088 25,897 31,675 36,177EV 5,459 7,709 7,032 7,293 6,857 9,082 10,961 13,002Net debt/(cash) 468 2,709 2,068 2,385 1,944 4,174 6,060 7,797BVPS (CNY) 112.127 129.488 61.521 74.399 112.866 138.162 171.397 209.164

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESales (YoY) n.a. 58.7 (35.1) 47.6 41.0 35.5 16.3 3.9EBITDA (YoY) n.a. (43.7) n.a. n.a. 35.1 46.9 30.3 24.7Operating profit (YoY) n.a. (65.2) n.a. n.a. 42.5 51.0 29.1 25.1Net profit (YoY) n.a. (49.4) n.a. n.a. 113.0 2.5 31.4 13.6Core EPS (fully-diluted) (YoY) n.a. (60.2) n.a. n.a. 44.3 1.9 26.8 11.9Gross-profit margin 29.2 15.6 4.8 20.3 22.4 23.4 24.6 26.9EBITDA margin 23.4 8.3 n.a. 14.3 13.7 14.9 16.7 20.0Operating-profit margin 21.2 4.7 n.a. 9.2 9.3 10.4 11.5 13.9Net profit margin 16.8 5.4 (5.6) 5.1 7.7 5.8 6.6 7.2ROAE n.a. 14.3 n.a. 21.3 27.8 20.2 21.5 19.8ROAA n.a. 5.3 n.a. 3.8 5.2 3.5 3.6 3.5ROCE n.a. 6.4 n.a. 12.3 12.0 12.3 11.8 11.6ROIC 27.1 6.1 (27.1) 15.1 18.8 18.1 16.4 15.8Net debt to equity 34.5 115.7 214.7 175.0 160.5 204.9 213.1 203.5Effective tax rate 14.2 22.9 n.a. 8.9 n.a. 9.9 9.6 9.3Accounts receivable (days) n.a. 54.6 131.3 96.7 95.6 104.7 112.0 118.2Current ratio (x) 1.1 0.8 0.6 0.7 1.0 1.0 1.0 1.2Net interest cover (x) 15.4 1.9 n.a. 2.9 3.2 3.5 3.4 3.3Net dividend payout 0.0 0.0 n.a. 0.0 0.0 0.0 0.0 0.0Free cash flow yield n.a. n.a. 11.6 n.a. n.a. n.a. n.a. n.a.

Company profile

JinkoSolar (JKS) is one of the largest solar module manufacturers globally. Headquartered in China, the company has a global manufacturing and sales footprint, with Japan and the US its key markets outside China. JinkoSolar also develops downstream solar projects through its subsidiary Jinko Power (JKP), and the company targets for capacity to reach 1.1GW by the end of 2015.

81


Solar module business: riding on the upcoming solar installation boom

China and the US markets should drive sales volume growth We expect the pace of global solar installations to continue to be driven by the US and China markets in 2015-16. Given the respective policy tailwinds, as well as the continuing decline in solar PV pricing, we expect these 2 markets to experience a solar installation boom over 4Q15-16, for which we forecast capacity addition growth to reach 31% and 46% YoY for the US, and 50% and 14% YoY for China, respectively. Given that both China and the US are still the key markets for JKS (for 1H15, the company shipped 1.5GW of solar modules to third-parties, of which 41% of the volume was shipped to China, and 19% to the US), we expect robust installations for both markets to remain the key drivers for JKS’s solar module sales over 2015-16.

Global solar PV installations JKS: module sales by market (1H15)

Source: BP, EPIA, Daiwa forecasts Source: Company

As such, we forecast JKS’s solar module sales volume to increase by 44% and 18% YoY for 2015-16, respectively. JKS: solar module sales volume

Source: Company, Daiwa forecasts The US: another important market for JKS apart from China For 2014, revenue from the North American region accounted for 17.7% of JKS’s total revenue. We believe the revenue contribution from this region will remain in the range of 18-20% over 2015-17E, given the likely strong pace of solar installations in the US over that period.

010,00020,00030,00040,00050,00060,00070,00080,00090,000

2010 2011 2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(MW)

China US Europe Japan Others

41.4%

15.7%

17.6%

6.7%

18.6% China

APAC

Europe

Emerging market

North America

0500

1,0001,5002,0002,5003,0003,5004,0004,5005,000

2012 2013 2014 2015E 2016E 2017E

(MW)

China APAC Europe Emerging market North America

We look for the pace of solar installations to accelerate over 4Q15-16, which should buoy JKS’s solar module sales volume for the same period. Its cost leadership should also help it maintain a stable gross margin of 19-21% over our forecast period

82


JKS: revenue contribution by country US: solar installations

Source: Company, Daiwa forecasts Source: SEIA, Daiwa forecasts

We expect the pace of solar installations in the US to accelerate further going into 2016E, due mainly to the expiration of the Investment Tax Credit (ITC) policy by the end of 2016. According to the US Department of Energy, the ITC for commercial and utility-scale PV systems will be cut from 30% to 10% starting 2017, whereas the ITC for residential solar systems will be abolished completely. Given the high residential electricity tariffs in the US currently, and that many states have achieved retail parity, such as California and North Carolina, we do not expect much impact on residential systems following the expiration of the ITC. However, utility-scale solar systems will encounter a reduction in IRRs once the ITC is cut from 30% to 10%. As such, we expect a rush of installations in the US before the end of 2016, mainly driven by the utility-scale solar segment, and then a slump in utility-scale installations post-2016, followed by a gradual recovery on the improving economics of solar projects starting in 2018. US: solar capacity additions by state (2014)

Source: SEIA, Daiwa As such, we currently expect JKS’s module sales to the North America region to rise by 30% YoY for 2016E, but subsequently moderate to 5% YoY growth for 2017E. We assume JKS will outperform the industry in terms of sales volume growth (5% YoY for 2017E vs. a 49.2% decline for the US solar installation), which is based on: 1) JKS’s focus on the residential market vs. utility-scale market, as evidenced by its cooperation with Vivint Solar (VSLR US, not rated), the leading residential solar installation service provider in the US, and 2) an incremental volume contribution from other countries in North America, ie, Canada and Mexico.

45% 49% 44% 48% 50% 55%

3%6% 17.7% 18.0% 20% 19%3%

13% 11% 9% 8%51%31%

10% 9% 8% 7%

0%

20%

40%

60%

80%

100%

2012 2013 2014 2015E 2016E 2017E

China America UK Chile South Africa Japan Rest of the world

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

E

2016

E

2017

E

(MW)

Residential Non-residential Utility

57%

6%5%

5%4%

4%

18%California

North Carolina

Nevada

Massachusetts

Arizona

New Jersey

Others

We expect US solar module demand to be supported by a rush of installations, due to the Investment Tax Credit being reduced from 30% to 10% starting 2017

83


JKS should maintain cost leadership vs. its peers We like JKS’s cost leadership compared to its peers, mainly attributed to: 1) its vertically integrated business model, capturing higher margins through the value chain, and 2) its major capex occurring in 2010-11, when construction costs had fallen and depreciation expenses were much lower following the GFC, compared with its competitors’ capex period in 2008-09 when costs were much higher (JKS’s depreciation expenses accounted for 3.6% of COGS in 2014 vs. peers’ 6.3%). On the back of its cost advantages, JKS has enjoyed better margins than its peers since 2010. Its gross margin has been around 6pp higher than the industry average since 2010, at 18% vs. peers’ 12%, while its net margin has been 7pp higher at 8% vs. peers’ 1% (excluding 4Q11 to 4Q12). Also, it was the first company to return to the black after the serious overcapacity situation in 2011-12 when solar module production capacity growth outpaced demand growth significantly, and many solar equipment manufacturers went bankrupt. JKS also started commercial operations at its Malaysia manufacturing base in May 2015, with an annual capacity of 450MW modules and 500MW cells. Given that there are no anti-dumping and countervailing duties imposed by the US on Malaysian imports, while its unit cost of production is comparable with that for its China business, according to management, its solar modules manufactured in Malaysia yield a higher margin than its China-made modules (gross margin: 25-30%, versus China modules: ~18-20%).

JKS: gross margin vs. peers JKS: net margin trend for the solar module manufacturers


Note: JKS was the first company to break even in 2013, followed by Canadian Solar and Trina JKS: gross margin (USD per W)

Source: Company, Daiwa

(30%)

(10%)

10%

30%

50%

1Q10

2Q10

3Q10

4Q10

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15


(40%)

(30%)

(20%)

(10%)

0%

10%

20%

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15


0.40

0.50

0.60

0.70

0.80

0.90

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

0%

5%

10%

15%

20%

25%

30%(USD per W)

Gross margin - RHS Module ASP Module unit cost

JKS’s cost advantage comes from: 1) its vertically integrated business model, and 2) capex only being incurred in 2010-11, when production plant construction costs had declined significantly

84


Highly integrated model capturing full margin across the value chain The solar module manufacturers are usually vertically integrated to some extent, but we think what distinguishes JKS among its peers in terms of manufacturing capacity is that its wafer and cell capacities are well matched to its module capacity, as compared to that of its peers. At the end of June 2015, JKS had an annual wafer manufacturing capacity of 3,000MW, a cell capacity of 2,500MW, and module capacity of 4,000MW, implying that most of its solar module sales volume should be able to capture the full margin across the value chain, thereby enabling the company to enjoy a higher gross margin than its peers.

JKS: manufacturing capacity vs. peers (end-June 2015) JKS: manufacturing capacity proportion vs. peers (end-June 2015)


JKS vs. peers: Solar module unit costs

Source: Companies, Daiwa Lower unit capex than peers Established in 2006, JKS only produced recoverable silicon materials initially. The company expanded into ingots and wafers manufacturing in 2007-08. It wasn’t until 2009 that JKS entered the solar cell and module industry. As a late market entrant, the company has essentially avoided the expensive capex, given: 1) the rapid decline in plant construction costs after the GFC in 2008, 2) the rapid improvement in technology since 2006 leading to declining equipment costs, and 3) the acquisition of high-quality production equipment from players exiting the industry at low costs (see the following cumulative capex chart, which shows JKS’s lagging capex over 2008-10).

0

2,000

4,000

6,000

8,000

10,000

12,000

Jinko CSIQ Trina Renesola JA Solar Yingli Hanwha

MW

Wafer Cell Module

32%

5%19%

58%

15%29%

14%

26%

32%

34%

7%

42%31%

52%

42%64%

47%35% 42% 39% 35%

0%10%20%30%40%50%60%70%80%90%

100%

Jinko CSIQ Trina Renesola JA Solar Yingli HanwhaWafer Cell Module

0.35

0.45

0.55

0.65

0.75

0.85

1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15

USD per W

Yingli Trina Jinko Canadian SolarJA Renesola Hanwha Q Cell

JKS had annual wafer manufacturing capacity of 3,000MW, cell capacity of 2,500MW, and module capacity of 4,000MW as at end-June 2015

Production plant capex went down steeply due to sluggish solar module demand, and improving technology

85


Solar module manufacturers: cumulative capex (since 2008) JKS: depreciation as % of revenue and % of COGS vs. that of its peers (2014)

Source: Companies, Daiwa Source: Companies

As a result, we believe JKS is relatively better off in terms of depreciation expenses for its manufacturing capex. The previous chart shows that the company had the lowest depreciation expenses as a percentage of revenue and COGS among its China peers in 2014. Slightly expanding margin trend on weak polysilicon prices As we explained in our initiation report on GCL-Poly, we foresee a slightly declining trend in polysilicon and wafer pricing over the long term, albeit a pick-up as a result of seasonal factors in 4Q15. In addition, we expect a diverging pricing trend between China and the international market, given that the China’s 58th Article has closed the loopholes in the processing trades. We forecast China’s polysilicon prices to reach about USD18-18.5/kg by the end of 2015, while international polysilicon prices remain weak at ~USD15/kg over the same period. Daiwa’s polysilicon price forecasts

Source: Solarzoom, Bloomberg, Daiwa forecasts Despite our view that polysilicon prices will stabilise in China in 2016E, we forecast JKS’s silicon costs to decline by 6%/4% YoY for 2016-17E, given the increasing overseas exposure of its manufacturing plants, which should benefit from weaker international polysilicon prices. This, coupled with our slightly declining solar module ASP trend (3%/6% YoY decline for 2016-17E), leads us to expect JKS’s solar module gross margin to increase slightly, from 19.4% for 2015E to 20.3% for 2016E (vs. 15-17% for its peers).

0%

20%

40%

60%

80%

100%

2008 2009 2010 2011 2012 2013 2014Yingli Trina JinkoCanadian Solar JA RenesolaHanwha Q Cell

0%

2%

4%

6%

8%

10%

12%

14%

Jinko CanadianSolar

Trina JA Solar Renesola Hanwha Yingli

Dep % rev Dep % COGS

1012141618202224


(USD/kg)

China local China Imported InternationalChina local (forecast) China Imported (forecast) International (forecast)

Polysilicon prices will likely remain weak for 2016-17E, especially in international markets

We expect a stable solar module gross margin trend for JKS over 2015-17E

86


JKS: solar module key assumptions

2012 2013 2014 2015E 2016E 2017E

Solar module sales volume (MW) 912 1,765 2,423 3,500 4,133 4,409 - China 345 871 1,045 1,567 1,881 2,163 - North America 52 120 422 632 822 863 - Europe 485 316 361 469 469 422 - APAC 30 292 291 465 558 558 - Others n.a. 165 306 367 403 403 Sales volume growth (YoY)

93% 37% 44% 18% 7%

- China

153% 20% 50% 20% 15% - North America

131% 250% 50% 30% 5%

- Europe

-35% 14% 30% 0% -10% - APAC

870% -1% 60% 20% 0%

- Others

85% 20% 10% 0% ASP (USD per W) 0.68 0.61 0.61 0.56 0.53 0.49 YoY change (%) -51% -12% 0% -8% -3% -6% Unit cost (USD per W) 0.62 0.50 0.46 0.42 0.39 0.36 YoY change (%) -40% -20% -7% -9% -8% -7% - Non-silicon cost 0.50 0.41 0.36 0.34 0.31 0.29 - Silicon cost 0.12 0.09 0.10 0.08 0.08 0.07 Unit gross margin (USD per W) 0.06 0.12 0.15 0.14 0.14 0.13 Gross margin (%) 7% 17% 20% 19% 20% 21%


87


Robust solar project development through JKP We are positive on JKS’s future development of its downstream solar projects, given the company’s ample project pipeline and its plan for a separate listing of its downstream solar IPP, Jinko Power (JKP). Such a set-up should limit the equity dilution risk for JKS’s shareholders, as equity financing would be done at the JKP level rather than by JKS itself. In our current model, we forecast JKS to add 600MW in solar project capacity over 2015-17E. For 2015 especially, JKS is on track to achieve its target capacity additions of 600-800MW, on the back of its 223MW in solar projects already connected to the grid during 1H15, with about 500MW already under construction as at the end of June 2015. As such, we forecast JKS’s gross profit contribution from its solar farm business to increase from 6% for 2014 to 26% for 2017.

JKS: attributable solar capacity JKS: gross profit breakdown

Source: Company, Daiwa forecasts Source: Company, Daiwa forecasts

Note: We expect JKS’s solar farm proportion as a percentage of total gross profit to increase from 6% for 2014 to 27% for 2017E

5.6GW of pipeline projects, adequate for development over the next 3-5 years As at end-May 2015, JKS had a pipeline of 5.6GW of solar projects spread over 11 provinces in China. While its project pipeline usually only factors in preliminary investment agreements with local governments that may not necessarily turn into actual projects for various reasons, we believe JKS’s pipeline of 5.6GW should be adequate for its project development over the next 3-5 years. JKS: project pipeline

Source: Company

0%20%40%60%80%100%120%140%160%

0

500

1,000

1,500

2,000

2,500

2012 2013 2014 2015E 2016E 2017E

(MW)

Attributable capacity (MW) YoY growth (%) - RHS

02,0004,0006,0008,000

10,00012,00014,00016,00018,000

2012 2013 2014 2015E 2016E 2017E

(CNYm)

Solar module Silicon wafer Solar cell Power generation Others

We forecast JKS to add 600MW in solar project capacity over 2015-17E

88


We like JKS’s project quality in terms of geographical location. At the end of June 2015, most of JKS’s pipeline of projects was located in eastern China where grid curtailment is not an issue. Among its 5.6GW of pipeline projects, only 11% are located in Xinjiang, Jilin, and Ningxia provinces, where we are cautious about grid curtailment issues. We would note in particular that none of its pipeline projects are in Gansu Province, which had a 40.2% grid curtailment rate between July 2014 and May 2015, according to various media sources, including China5e and Energy Trend. JKS: project pipeline by province (end-June 2015)

Source: Company Downstream listing We also like JKS’s strategy of growing its downstream solar project business using a separate entity, which is its subsidiary, JKP. JKS was among the first solar module manufacturers in China to enter the downstream solar industry in 2011. However, as a manufacturer, JKS was unable to gear up enough to fund downstream solar farm projects, which limited its development pace before 2014. This was also why JKS decided to invite private equity funds to invest in its downstream business. In July 2014, it signed a private equity investment agreement with China Development Bank International (CDBI), Macquarie Greater China Infrastructure Fund (MGCIF) and New Horizon Capital (New Horizon), with the 3 investors investing USD225m into JKS’s subsidiary, JKP, and JKP issuing redeemable convertible bonds to them. Upon the full conversion of the convertible bonds, JKS will own around 55% of JKP, and the 3 investors will jointly hold the remaining 45%. JKP: corporate structure chart before the IPO listing

Source: Daiwa, Company

Jiangxi34.5%

Zhejiang 13.7%Henan

11.2%

Shanxi 9.4%

Anhui8.9%

Shandong 8.9%

Xinjiang 6.4%

Ningxia 2.7%

Others4.3%

Jinko Power

Jinko Solar (55%) China Development Bank International (21%)

Macquarie Greater China Infrastructure Fund (20%) New Horizon Capital (4%)

Most of JKS’s pipeline of solar projects is located in provinces that are not subject to grid curtailment

JKS signed a private equity investment agreement with CDBI, MGCIF and New Horizon to jointly invest in JKP

89


JKP is JKS’s IPP platform. After the private equity investment was signed in July 2014, JKS will not incur the capex directly, but through the capital of JKP. JKS also plans to spin off JKP under a separate listing in either the H-share market or the US market, and we expect this to happen in the middle of 2016. Upon the successful listing, JKS’s stake in JKP will decline from 55% to about 44%, according to our estimates, assuming the IPO will sell a 20% stake of JKP. We see at least 3 benefits of doing so: • JKS has invited 3 private equity funds to invest in JKP, with JKS and the total of 3

private equity (PE) funds holding 55% and 45% of the joint venture (JKP), respectively. This reduces JKS’s capital requirement when funding its solar project capex, but JKS can still maintain a controlling stake in JKP.

• Utility companies can usually tolerate a net gearing level of over 300% given their stable

cash flows, while manufacturing companies usually have difficulty borrowing if their gearing level is above 100%. As such, separating the stable solar IPP business from JKS’s solar module manufacturing business should help JKP to gear up more for its solar projects and thus achieve a better IRR.

• The potential spin-off of JKP, which has been planned and disclosed by the company

since the beginning of 2015, should reduce JKS’s holding of JKP to the sub-50 level, and JKS could then deconsolidate the solar farm business from JKS’s balance sheet. Thus, this should return JKS into an asset-light model, while receiving a stable cash flow from its solar projects.

JKP IPO should raise additional capital of around CNY850m We carried out a scenario analysis on the impact of a JKP IPO listing. Given that we expect JKP’s IPO to happen in mid-2016, we assume that JKP will be able to grid-connect 1.4GW of solar projects before the IPO, which is JKS’s average 2015/16E capacity, according to our forecasts (versus JKS’s targeted installed capacity 1.3-1.5GW). We further analysed the IPO market capitalisation of JKP, using PER and PBR valuation methodologies, assuming JKS could complete 1.4GW of solar farms before the IPO. PBR valuation methodology: We assign a 2016E PBR of 1.2x for its 1.4GW of projects, and arrive at a market capitalisation of CNY4.3bn, with the IPO raising ~CNY856m for 20% additional shares in JKP. PER valuation methodology: We assign a 2016E PER to value JKP. We estimate that post-IPO, JKP would be able to complete 400MW more solar projects versus the case if there were no IPO, and thus the effective capacity for 2016E would increase by 200MW, given that the IPO should happen in 2H16. We then estimate that, in aggregate, those assets should generate a net profit of ~CNY429m for 2016E, assuming average utilisation hours of 1,100, an average solar tariff of CNY0.95/kWh and a net profit margin of 30%. Using an average 2016E PER of 10x that we assign for clean-energy IPPs, we estimate that JKP’s market capitalisation should be about CNY4.3bn, and the IPO should raise about CNY859bn for 20% additional shares.

We see at least 3 benefits of spinning off JKP: 1) reduced capex needs, 2) easier solar project financing, and 3) net gearing improvement

JKS targets to complete 1.3-1.5GW of solar farm capacity before spinning off JKP

90


JKP: Daiwa’s estimates of the capital raised through an IPO

PBR method PER method

Average capacity for 2015/16E (MW) 1,403 Cost per W (CNY) 7.5 Projects assets (CNY) 10,520 Gearing 200% Estimated book value (CNY) 3,507 Estimated average 2015/16E PBR 1.2 Effective capacity for 2016E (MW)

1,603

2016E net profit (CNYm)

429 Estimated 2016E PER

10.0

Estimated market cap (CNY) 4,278 4,294 IPO % 20% 20% Capital raised (CNY) 856 859

Source: Daiwa

JKS: calculation of JKP disposal gain

Pro forma 2016E – PBR Pro forma 2016E – PER

Estimated market cap (CNYm) 4,278 4,294 Estimated book value (CNYm) 3,507 3,507 JKS's stake before IPO 55% 55% JKS's stake after IPO 44% 44% Gains from JKP disposal through IPO 85 87

Source: Daiwa The above PER and PBR analysis arrive at very similar market capitalisation estimates for JKP upon listing, at CNY4.3bn, and, as such, we base any further analysis on JKP’s listing using this number. Combining additional capital with existing capital raised since 2014, we expect JKP to have enough capital to add 600MW in capacity in 2016E We expect IPO proceeds of about CNY850m to support JKP’s additional planned capacity of 380MW, assuming CNY7.5/W construction cost and a 70:30 debt-to-capital ratio. Also, we estimate that the capital raised since 2014, amounting to CNY3.1bn (~CNY1.4bn through private equity investments, ~CNY800m through share placements, and ~CNY908m through CB issuance), could support ~1.3GW of solar farm additions. Subtracting the 290MW capacity added in 2014, as well as the 600MW capacity expected to be added during 2015E, our analysis shows that JKS’s remaining capital could support a further 446MW of projects in 2016E. JKS: solar farm capital utilisation schedule for solar projects Date Type of financing Capital raised (USDm) Capital raised (CNYm) Capacity addition for the year (MW) Jan-14 Follow-on offering 132.2 799.9 - Jan-14 CB offering 150.0 907.7 - Jul-14 PE agreement 225.0 1,395.4 - Total capital raised 507.2 3,103.0

Capital utilised (CNYm) Capacity addition for the year (MW) Capacity addition in 2014

696.1 290.0

Capacity addition in 2015E

1,404.0 600.0 Capacity addition able to support for 2016E 1,002.9 445.8

Source: Company, Daiwa Combining the 446MW supported by JKS’s existing capital with the 380MW supported by JKP’s IPO proceeds of ~CNY850m, JKP should have enough capital to add 600MW in 2016E. For 2017E, JKP may need to rely on its operating cash flow, and maybe another placement.

Both PER and PBR valuation methodologies return similar results: we estimate JKP’s market capitalisation would be ~CNY4.3bn after an IPO

After IPO, JKP should have enough capital to add 600MW in 2016E

91


600-800MW capacity addition target on track For 1H15, JKS connected 222.5MW of solar farm capacity on an attributable basis, bringing its total attributable solar capacity to 725MW, as at the end of June 2015. We are therefore positive on the company’s ability to achieve its annual capacity addition target of 600-800MW for 2015, backed by ~500MW of solar projects already under construction as at the end of June 2015. If we refer to the following chart, JKS was actually outpacing the other downstream developers in 1H15, in terms of capacity installed, which we believe was the result of better capital conditions given the private equity capital it raised in 2H14. China Solar Sector: solar farm capacity installed

Source: Companies, Daiwa

290

668

384

890

280223

158 65 88

35 0

100200300400500600700800900

1,000

JKS GCLNE United PV Shunfeng XYS

(MW)

2014 1H15

JKS’s solar farm capacity additions target of 600-800MW in 2015 seems well on track

92


Financials

Net gearing: waiting for deleveraging through JKP IPO Given the capital-intensive nature of the solar farm development business, JKS’s net gearing level has remained elevated since it started to develop solar projects. Apart from 2012, when the China solar sector faced substantial difficulties and solar stocks were trading at depressed levels, JKS has relied on issuing convertible bonds and follow-on public offerings to replenish its capital needs every year, similar to other solar companies. JKS: financing history Date Type of financing Capital raised (USDm) No. of ADS (m) 10-May IPO 64.2 5,835 10-Nov Follow-on offering 72 2,000 11-May CB offering 125 n.a. 13-Sep Follow-on offering 71 4,370 14-Jan Follow-on offering 132.2 3,750 14-Jan CB offering 150 n.a. 14-Jul PE agreement 225 n.a.

Source: Company, Daiwa Soaring net gearing due to the consolidation of JKP Even after the joint venture was formed in July 2014, we estimate that JKS’s gearing should still rise, given that JKP is still consolidated under JKS (unless the IPO occurs), with its net debt-to-equity level rising from 161% for 2014, to 205% for 2015E and 213% in 2016E. For 2017E, we expect a slight improvement as JKS’s cash flow from its solar farm should increase as the capacity base gets larger. JKS: net gearing

Source: Company, Daiwa forecasts IPO timing: most probably in the middle of 2016 We estimate that JKS’s capital on hand could support its solar farm development to a maximum capacity of 1.6GW, if no more capital is raised after the July 2014 private equity agreement. We arrive at the 1.6GW capacity by taking into account the 213MW capacity it had as at the end of 2013, as well as the capital raised in 2014 (including USD132m from a follow-on offering, USD150m from a CB issuance, and USD225m from a private equity agreement). From the current solar farm addition schedule (~600MW new capacity a year from 2015-17), we expect the company to run out of capital in 2H16. As such, we expect the most possible timing for JKP to go for an IPO to be the middle of 2016. Should JKP fail to undertake an IPO, it might need to raise more capital through capital injections, a CB issuance at the JKP level, or by introducing extra investors to JKP.

0%

50%

100%

150%

200%

250%

2012 2013 2014 2015E 2016E 2017E

Net debt to equity - assuming IPO Incremental gearing impact if no IPO

JKS has relied on CB issuances and share placements for equity financing since listing

We see JKS’s net gearing level to surge unless JKP’s IPO occurs

93


Nevertheless, we have estimated the balance sheet and net gearing impact on JKS should there be an IPO for JKP. Our pro-forma calculations below show that JKS’s net gearing would fall back to around 78% after JKP is spun off from JKS. We believe this is a more reasonable level for a pure manufacturing company. JKS: pro forma net gearing impact of a JKP IPO (CNYm) 2016E Pro forma 2016E Net debt

Net debt with the IPO, assuming CB redemption 11,351 11,351 Adjustments:

- Debt deconsolidated (60% of project assets) - (6,881) - Redemption of convertible bonds - (1,779) Adjusted net debt 11,351 2,691 Shareholders' equity

Shareholders' equity without an IPO 5,326 5,326 Adjustments:

- Equity deconsolidated (55% of book value) - (1,929) - Gain from IPO - 85 - Loss of profit contribution - (141) - Convertible bond interest savings - 87 Adjusted shareholders' equity 5,326 3,428 Net gearing (%) 213% 78%

Source: Daiwa After the spin-off of JKP, we believe the major impact on JKS’s P&L would be the loss of revenue due to changes from JKP no longer being consolidated under JKS, and becoming an associate company, which could also result in a loss of 11% of net profit that had been previously been contributed by JKP. In considering this, we estimate that JKS’s net profit would be impacted by 6% and 8% for 2017 and 2018E, respectively. But despite all this, we believe the good should far outweigh the bad, as we see huge advantages, such as a much better balance sheet, as well as a disposal premium being realised. JKS: net profit impact in the case of an IPO (CNYm) 2017E 2018E Net profit from JKP – assuming no IPO

Profit after tax 702 884 Minority interests (321) (403) Net profit from JKP 381 481 Net profit from JKP – assuming IPO

Associate profit from JKP (44%) 309 389 Loss of net profit 72 92 Total net profit 1,174 1,179 Loss of net profit to JKS 6.2% 7.8%

Source: Daiwa forecasts Working capital: well within control Inventory: JKS maintained relatively low inventory days relative to its peers until 2H14, implying that the company has efficient inventory management. For 2010-12, the inventory level increased slightly due to over-supply in the China solar equipment industry, but JKS managed to control its level to 80 days. Since 2H14, JKS has deployed a more aggressive inventory strategy as it expected the solar installation boom to start earlier in 2015. This led to an increase in its current inventory days to 87 days for 2Q15, slightly above the average of its peers at 82 days. That said, we are not worried about the rise in inventory level, given that the company says its module shipments are fully filled for the rest of the year, meaning that we should see a declining inventory level amid the shipment peak in 3Q15 and 4Q15.

After the spin-off, JKS’s net gearing level would fall to ~78% in 2016E

JKS stocked up for an anticipated solar installation peak during 2015, and we see inventory level coming down during 4Q15

94


JKS: inventory days vs. peers JKS: accounts receivables days vs. peers

Source: Companies Source: Companies

Accounts receivable: JKS has a more aggressive credit sales strategy versus that of its peers, which can be illustrated by its slightly above average receivable days since 2011. That said, JKS has managed to maintain its accounts receivable days within its controlled range of 60-120 days. As at the end of June 2015, JKS’s accounts receivable turnover was 96 days, which was still well within the peer range of 75-100, excluding the outliers. In addition, JKS’s accounts receivable profile was healthy in both 2013 and 2014, and we note declining bad debt provisions within its accounts receivable (falling from 28% for 2012, to 21% for 2013 and 12% for 2014). The company has collected the then-bad debts from its customers during those years, with 39% of the bad debt collected in 2013 (or CNY308m), as well as a further 25% of bad debt collected in 2014 (or CNY161m).

JKS: bad debt allowance as a % of accounts receivable JKS: bad debt expenses


20

40

60

80

100

120

1Q10

2Q10

3Q10

4Q10

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

(days)

Yingli Trina Canadian SolarJinko Renesola Hanwha SolarOneJA Solar

020406080

100120140160180

1Q10

2Q10

3Q10

4Q10

1Q11

2Q11

3Q11

4Q11

1Q12

2Q12

3Q12

4Q12

1Q13

2Q13

3Q13

4Q13

1Q14

2Q14

3Q14

4Q14

1Q15

2Q15

(days)

Yingli Trina Canadian SolarJinko Renesola Hanwha SolarOneJA Solar

0%

5%

10%

15%

20%

25%

30%

0500

1,0001,5002,0002,5003,0003,5004,000

2009 2010 2011 2012 2013 2014

(CNYm)

Allowance for doubtful accounts Account receivables, Net% bad debt allowance to A/R

0%

10%

20%

30%

40%

50%

0

100

200

300

400

500

600

2009 2010 2011 2012 2013 2014

(CNYm)

Bad debt expense % bad debt written off% bad debt reversal % bad debt expense to A/R

JKS’s bad debt seems well within control

95


Valuation

SOTP-derived target price of USD28.5 We initiate coverage of JKS with an Outperform (2) rating and 12-month target price of USD28.5 per ADS, representing 13% upside potential from its current share price. We derive our target price using SOTP valuation methodology, which we consider to be the most appropriate given the different potential earnings growth and risk profiles of the company’s 2 main businesses – solar manufacturing and downstream solar project development. JKS: SOTP valuation (CNYm) Methodology Valuation % of value Manufacturing business EV/EBITDA 11,976 59% Solar farm business DCF 8,385 41% Corporate value

20,361

Net cash / (net debt)

(11,351) Minority interests

(1,925)

Equity value

7,085 No. of shares (m)

155

USD:CNY FX

6.41 Fair value per share (USD)

7.1

Fair value per ADS (USD)

28.5

Source: Daiwa forecasts We value the company’s solar manufacturing business at a 2016E EV/EBITDA multiple of 7.5x, which is on a par with the average of its US-listed solar module manufacturing peers. We believe the solar module business is becoming commoditised globally and, in this regard, JKS’s solar module business should only command an average trading multiple. For the downstream solar IPP business, we use a DCF-derived valuation, assuming a WACC of 8.9% and a terminal growth rate of 1%. Within the WACC, we assume a China risk-free rate of 3.4%, a market risk premium of 9.6% and beta of 1.0.

JKS: Daiwa’s free-cash flow calculations for solar farm business 12 months to 31 Dec, all figures in CNYm

2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E … 2030E

Valuation Date 27-Oct-15

26-Nov-15 30-Jun-16 30-Jun-17 30-Jun-18 30-Jun-19 30-Jun-20 30-Jun-21 30-Jun-22 30-Jun-23 30-Jun-30

Next Balance Date 31-Dec-15

First Year Cash Flow Adjustment 0.19

Free Cash Flow

EBITDA

690 1,078 1,610 2,045 2,045 2,045 2,027 2,009 1,992 … 1,873

Less: Other non-cash

5 7 (305) (387) (404) (375) (354) (353) (347) (325)

Less: Cash tax payable on EBIT

(1) (4) (20) (59) (96) (157) (226) (256) (296) (299)

Plus: Decrease in working capital

0 0 0 0 0 0 0 0 0 0

Less: Capital expenditure

(3,926) (3,806) (3,686) 0 0 0 0 0 0 0

Free Cash Flow

(3,231) (2,725) (2,401) 1,599 1,545 1,512 1,447 1,401 1,349 1,249

Free cash flow for valuation purposes

(3,231) (2,725) (2,401) 1,599 1,545 1,512 1,447 1,401 1,349 … 1,249

WACC 8.9%

8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9%

NPV of free cash flow

(3,205) (2,569) (2,078) 1,271 1,127 1,012 889 790 698 … 355 Source: Daiwa forecasts

We use SOTP valuation for JKS, arriving at our target price of USD28.5

We value manufacturing business using 7.5x 2016E EV/EBITDA

We value solar IPP business using DCF model

96


JKS: DCF valuation for solar-farm business Target gearing (debt/capital) (%) 70.0 Market risk premium (%) 9.6 Risk-free rate (%) 3.4 Cost of debt (%) 9.0 Cost of equity (%) 13.0 WACC (%) 8.9 Terminal Value Terminal Growth Rate 1.00% Terminal WACC 8.94% Estimated Terminal Free Cash Flow 1,297 NPV of Terminal Value (as at 30 Jun 2025) 16,333 NPV of Terminal Value (as at 27 Oct 2015) 7,122 DCF Valuation NPV of Forecasts (CNYm) 1,263 NPV of Terminal Value (CNYm) 7,122 Add: Market value of stakes in Associates #1 - Associate #2 - Associate #3 - Enterprise Value (CNYm) 8,385

Source: Daiwa JKS: DCF sensitivity analysis for sola-farm business Discount Rate NPV of FCF Term Val. at 2020 PV of Term Val. Enterprise Value Equity Value Equity Value Per Share (HK$) 6.4% 2,830.1 23,839.7 13,018.7 15,848.9 14,548.8 58.5 6.9% 2,484.1 21,833.0 11,393.4 13,877.5 12,577.5 50.6 7.4% 2,155.3 20,137.8 10,044.2 12,199.5 10,899.5 43.8 7.9% 1,842.7 18,687.0 8,910.4 10,753.1 9,453.0 38.0 8.4% 1,545.3 17,431.1 7,947.5 9,492.8 8,192.7 33.0 8.9% 1,262.3 16,333.5 7,122.2 8,384.5 7,084.5 28.5 9.4% 992.9 15,365.8 6,409.4 7,402.3 6,102.3 24.5 9.9% 736.4 14,506.5 5,789.4 6,525.8 5,225.7 21.0 10.4% 492.0 13,738.1 5,246.8 5,738.8 4,438.8 17.9 10.9% 259.1 13,047.1 4,769.4 5,028.5 3,728.4 15.0 11.4% 37.0 12,422.2 4,347.3 4,384.3 3,084.3 12.4

Source: Daiwa forecasts Current valuation close to the 2012-13 level; solar sector set to boom We believe JKS’s current valuation is appealing, as we expect the solar sector to enter an upcycle from 4Q15, during which we expect JKS to experience improving margins and cash flow, amid booming solar installation demand in the US and China. We see a close relationship between JKS’s historical valuation and its ROE, and note that there is a widening divergence between its gross margin trend and its PBR, which we believe is due to 2 factors: 1) in 2H14, China’s pace of solar installations turned out to be much slower than expected, with JKS’s downstream solar development missing its target despite the private equity agreement signed with CDBI and Macquarie, which was a disappointment to investors, and 2) despite the share price recovering in 1H15, the shares dipped again in July and August, due mainly due to the worsening global financial markets and US-listed China stocks getting hit heavily on the back of the slowing China economy. JKS: PBR vs. in-house gross margin

Source: Company, Bloomberg

0%

5%

10%

15%

20%

25%

30%

0.0

0.5

1.0

1.5

2.0

Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15

(x)

Gross margin (%) - RHS PBR - LHS

JKS’s valuation looks appealing if comparing its ROE or gross margin with its current 1-year forward PBR

97


The stock is now trading at a 0.9x 2016E PBR, which we believe is attractive given: 1) its current gross margin of ~20% should warrant a 1-year forward PBR of 1.5x, and 2) that we expect the solar sector to boom in 4Q15-2016, which should drive the company’s gross margin even higher. In terms of PER valuation, the stock is now trading at a 5.7x 2016E PER, which is also below its past 3 year trading average of 8.5x. We forecast an EPS CAGR of 19% over 2015-17E, implying a PEG ratio of only 0.3. Considering the above, we look for a mean reversion in the valuation ahead of a sector rally when the pace of solar installations accelerates starting in 4Q15. Our 12-month target price implies a 2016E PER of 6.5x and 2016E PBR of 1.3x, which we believe are attractive. We forecast a 2016E ROE of 21.5x and EPS CAGR of 19% over 2015-17E for the company.

JKS: 1-year forward PER JKS: 1-year forward PER bands

Source: Bloomberg, Daiwa Source: Bloomberg, Daiwa

JKS: 1-year forward PBR JKS: 1-year forward PBR bands


JKS’s peers: 2016E PBR (x) vs. ROE (%) JKS’s peers: 2016E PER (x) vs. 2015-17E EPS CAGR

Source: Company, Daiwa forecasts for Jinko, GCL-Poly, Singyes, and Xinyi Source: Company, Daiwa forecasts for Jinko, GCL-Poly, Singyes, and Xinyi

02468

1012141618

Dec-1

2

Mar-1

3

Jun-

13

Sep-

13

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(x)

16.2x Avg+2SD

12.0x Avg+1SD

7.7x Avg

3.4x Avg-1SD 0

10

20

30

40

50

Dec-1

2

Feb-

13

Apr-1

3

Jun-

13

Aug-

13

Oct-1

3

Dec-1

3

Feb-

14

Apr-1

4

Jun-

14

Aug-

14

Oct-1

4

Dec-1

4

Feb-

15

Apr-1

5

Jun-

15

Aug-

15

Oct-1

5

(USD)

Price 3.5x PER 6.5x PER9.5x PER 12.5x PER 15x PER

0.0

0.5

1.0

1.5

2.0

Jan-

11

May-1

1

Sep-

11

Jan-

12

May-1

2

Sep-

12

Jan-

13

May-1

3

Sep-

13

Jan-

14

May-1

4

Sep-

14

Jan-

15

May-1

5

Sep-

15

(x)1.9x Avg+2SD

1.5x Avg+1SD

1.0x Avg

0.5x Avg-1SD

0.0x Avg-2SD

05

10152025303540

Jan-

11Ap

r-11

Jul-1

1Oc

t-11

Jan-

12Ap

r-12

Jul-1

2Oc

t-12

Jan-

13Ap

r-13

Jul-1

3Oc

t-13

Jan-

14Ap

r-14

Jul-1

4Oc

t-14

Jan-

15Ap

r-15

Jul-1

5Oc

t-15

(USD)

PX_LAST 0.4x PBR 0.8x PBR1.2x PBR 1.6x PBR 2x PBR

JKS

Trina

Canadian Solar

JA SolarFirst Solar

SunPower

SolarWorld

GCL-Poly

Xinyi Singyes

0%

5%

10%

15%

20%

25%

0.0 0.5 1.0 1.5 2.0 2.5 3.0

2016

E RO

E(%

)

2016E PBR (X)

Jinko Solar Trina

Canadian Solar

JA Solar

SunPower

GCL-PolyXinyi

Singyes

-10%-5%0%5%

10%15%20%25%30%35%

- 5.0 10.0 15.0 20.0 25.0 30.0

2015

-17E

EPS

CAG

R (%

)

2016E PER (X)

We expect a slightly improving gross margin trend for JKS

98


Additional 34% TP upside potential from JKP IPO In our model, we have yet to take into account the value enhancement to JKS of a potential JKP IPO, given the uncertainties of whether an IPO, or other capital market financing, would be successful. That said, management seems determined to push ahead with the IPO, and as such we have estimated the potential value enhancement an IPO would bring to JKS. Under a scenario whereby the IPO goes ahead, we would stick with our SOTP-derived valuation for JKS, given the varied risks and return profiles of its solar manufacturing and downstream solar project businesses. For the solar manufacturing business, we would assign a 7.5x 2016E EV/EBITDA multiple (on a par with the peer average), which is the same as our valuation for this part of the business currently. For JKS’s downstream business, we would use our estimated market capitalisation for JKP, and use same PER and PBR (PER 10x; PBR 1.2x), and multiply the market capitalisation by JKS’s equity stake in JKP after the IPO, which we estimate would be 44%. Our analysis in the table below shows that both the PER and PBR methodologies would result in valuation enhancement for JKS, with 34% upside potential to our target price for JKS from current share price levels. JKS: value enhancement after IPO listing of JKP (CNYm) Valuation PBR PER Solar manufacturing 12,024 12,024 Net cash / (net debt) (4,390) (4,390) Minority interest (56) (56) Equity value 7,579 7,579 Share in JKP 1,882 1,889 Total equity value 9,461 9,468 No. of shares (m) 155 155 USD:CNY FX 6.4 6.4 Fair value per share (USD) 9.51 9.52 Fair value per ADS (USD) 38.1 38.1 Potential upside % 34% 34%

Source: Daiwa forecasts JKS: SOTP valuation (CNYm) Methodology Valuation % of value Manufacturing business EV/EBITDA 11,976 59% Solar farm business DCF 8,385 41% Corporate value

20,361


(11,351) Minority interests

(1,925)

Equity value

7,085 No. of shares (m)

155

USD:CNY FX

6.41 Fair value per share (USD)

7.1

Fair value per ADS (USD)

28.5


In case of an IPO, we would expect JKS’s fair value to rise to USD38.1

99


Major investment risks

Slowing global solar installations Our estimates assume JKS posts an increase in sales volume across all its regions, with growth particularly strong in China and the US. Our assumptions are based on the policy tailwinds in both countries at least in 2015-16, as well as a continuously declining solar power installation costs. However, we do see uncertainties relating to JKS’s sales volume growth over the next year, should solar installations in these countries turn out to be slower than expected. China: slow subsidy payment and lack of financing channels the major risks For China, we see 2 main factors that could drag down solar installation growth. First, although the subsidies provided by the central government and local governments are considered generous, the payment of such subsidies can be very slow, especially those from the central government (ie, Ministry of Finance, NDRC, and NEA). As of June 2015, we note that the authorities were still reviewing solar projects completed in 2H13 for their eligibility for subsidies. For JKS in particular, only those projects completed before August 2013 have been approved for subsidies, while the rest, ie, those connected after August 2013, are still awaiting approval. We believe the slow subsidy approval is due mainly to the lack of renewable energy funds. As explained in our sector report, currently electricity users are obliged to pay a CNY0.015 per kWh renewable energy surcharge, but this amount was announced back in September 2013 and is now insufficient to cover the subsidy level, considering the rapid renewable energy development under way currently. While we believe an increase in the renewable energy surcharge is imminent, we are unsure about the exact timing, and any further delays in an increase in surcharge would further delay subsidy approvals. An extended period of delays would ultimately impact solar installations in China given that many projects are already suffering from delays in subsidy payment, implying a worsening cash flow for those projects. The project developers in turn will have less incentive to develop new projects in this case. China: renewable energy surcharge adjustment Surcharge (CNY per kWh) 1-Jul-06 0.001 1-Jul-08 0.002 1-Nov-09 0.004 1-Jan-12 0.008 25-Sep-13 0.015

Source: NDRC

Second, some solar project developers may face financing difficulties due to: 1) banks becoming cautious on lending to solar projects, or offering very high financing costs if they do lend, or 2) a lack of equity financing opportunities from solar project developers amid the depressed stock market. Many solar IPPs rely heavily on external financing, with a debt/capital ratio as high as 80%. As such, the ability to obtain project loans from banks is crucial to these developers. As a result, should capital conditions in China worsen, banks might become even more reluctant to approve loans to solar developers, especially loans with tenures of over 10 years, or at an elevated interest rate. Under such a scenario, solar project developers might need to slow their capacity expansion, which in turn would affect JKS’s solar module demand. It is the same for capital financing: in the case of a depressed equity market, many solar IPPs might not be able to replenish their capital through equity offerings, which would also slow down their expansion pace.

We believe the main risks to JKS include slowing solar installation globally, worsening trade disputes between China and overseas countries, and slower-than-expected solar project development

100


Our sensitivity analysis shows that should module sales growth to China drop by 10pp, JKS’s revenue would be negatively impacted by 2.7%/3.4% for 2015E/16E, respectively, and its net profit would decline by 5.1%/6.2% for 2015E/16E, respectively. JKS: revenue and earnings sensitivity to module sales to China Sales volume growth Revenue Net profit 2015E 2016E 2015E 2016E -20pp -5.40% -6.70% -10.10% -12.40% -10pp -2.70% -3.40% -5.10% -6.20% Base case 0.00% 0.00% 0.00% 0.00% +10pp 2.70% 3.40% 5.10% 6.20% +20pp 5.40% 6.70% 10.10% 12.40%

Source: Daiwa

US: policy risks is the major factor We believe the risks to JKS from the US are much less than those from China, given the price competitiveness of solar energy in the US versus other types of electricity generation. There is also a higher proportion of residential projects versus commercial and utility-scale projects in the US, which are less susceptible to changes in subsidies. US: total solar installed capacity by segment

Source: SEIA To name a few risks relating to the growth of the US solar industry, the first and foremost is the net metering debates over a number of states. As the utilities recognise the threat of solar development to their business, many of them are urging the government to change the current net metering policy, including limiting the net-metered distributed solar capacity, or increasing fixed charges on residential customers or solar systems. According to PV news in August 2015 (website: http://www.pv-magazine.com/news/details/beitrag/16-us-states-consider-changes-to-net-metering_100020629/#axzz3lrdcSjLc), 16 states are already considering changes to net metering policies, and if the changes are negative overall, it might hurt solar development in the US. Our sensitivity analysis shows that should module sales growth to the US drop by 10pp, JKS’s revenue would be negatively impacted by 1.1%/1.4% for 2015E/16E, respectively, and the impact on its net profit would be 3.3%/3.4% for 2015E/16E, respectively. JKS: revenue and earnings sensitivity to module sales to the US Sales volume growth Revenue impact Net profit impact 2015E 2016E 2015E 2016E -20pp -2.20% -2.70% 6.70% 6.80% -10pp -1.10% -1.40% 3.30% 3.40% Base case 0.00% 0.00% 0.00% 0.00% +10pp 1.10% 1.40% -3.30% -3.40% +20pp 2.20% 2.70% -6.70% -6.80%

Source: Daiwa

28%43% 29%

16% 15% 17% 20%

67% 56%

40%43%

32% 23% 17%

5% 15%31%

41%54% 60% 63%

0%

20%

40%

60%

80%

100%

2008 2009 2010 2011 2012 2013 2014Residential Non-residential Utility

We expect subsidy cuts in the US should lead to less serious decline than in China, given the higher proportion of residential projects in the US

http://www.pv-magazine.com/news/details/beitrag/16-us-states-consider-changes-to-net-metering_100020629/#axzz3lrdcSjLc



101


Trade disputes between China and the US/EU Given JKS’s high proportion of solar module sales outside China, the company is exposed to any negative changes in anti-dumping and countervailing duties imposed by the US and European Commission. We consider the policy stance of these countries towards China as difficult to forecast given the complicated political relationship between the countries. US anti-dumping and countervailing duties Since 2012, the US Department of Commerce has conducted 2 rounds of anti-dumping and countervailing investigations into Chinese solar module manufacturers. The final ruling for the first case took place in December 2012, under which an anti-dumping duty of between 18.32% and 249.96%, and a countervailing duty of between 14.78% and 15.97% were imposed on solar cells manufactured in China. Since then, many Chinese manufacturers have used solar cells manufactured from Taiwan and assembled in China to avoid the tariffs, and that has led to the second round of investigation by the US. The final ruling resulted in an anti-dumping duty of between 26.71% and 165.04%, and countervailing duty of between 27.64% and 49.21% being imposed on all Chinese solar products (excluding thin-film and products included in the first case) manufactured in China. The first case was reviewed in July 2015, during which both tariffs were investigated and adjusted. Winners such as Yingli and JKS had their imposed tariffs adjusted downward, from 24.48% to 0.79%, and from 24.48% to 9.67%, respectively. Some manufacturers were worse off, however, such as Suntech, which had its anti-dumping duties adjusted upward from 29.14% to 33.08%. All manufacturers had similar upward adjustments, with countervailing duties increasing by ~5pp across the board. US anti-dumping duties and countervailing duties to Chinese manufacturers Company AD tariff (2015) AD tariff (2012) CVD (2015) CVD (2012) AD tariff (incl. Taiwan) CVD (incl. Taiwan) Jinko 9.67% 24.48% 20.94% 15.24% 78.42% 38.43% Yingli 0.79% 24.48% 20.94% 15.24% 52.13% 38.43% Trina 9.67% 18.32% 20.94% 15.97% 26.71% 49.21% Canadian Solar 9.67% 24.48% 20.94% 15.24% 52.13% 38.43% ReneSola 9.67% - 20.94% 15.24% 78.42% 38.43% Suntech 33.08% 29.14% 20.94% 14.78% 52.13% 27.64% Other companies on the list 9.67% 24.48% 20.94% 15.24% 52.13% 38.43% Rest of PRC 238.95% 249.96% 20.94% 15.24% 165.04% 38.43%

Source: USITC, Daiwa

EU anti-dumping and countervailing duties In December 2013, the European Commission imposed anti-dumping and anti-subsidy duties on China-imported solar cells and modules for 2 years, which is due to expire in December 2015. The combined anti-dumping and countervailing duties imposed were on average 47.7% for Chinese exporters. Below are the tariffs for major solar manufacturers. Anti-dumping and countervailing duties imposed by the EU AD tariff CV tariff Combined rate Jinko 41.20% 6.50% 47.70% JA Solar 51.50% 5.00% 56.50% Trina 44.70% 3.50% 48.20% ReneSola 43.10% 4.60% 47.70% Yingli 35.50% 6.30% 41.80% Suntech 41.40% 4.90% 46.30% Other cooperating companies 41.30% 6.40% 47.70% Other companies 53.40% 11.50% 64.90%

Source: European Commission On the same day, the European Commission also accepted the price undertaking jointly requested by the Chinese solar module manufacturers and China Chamber of Commerce for Import and Export of Machinery and Electronic products (CCCME), in which the

JKS was subject to 9.67% AD tariff and 20.94% CVD as of 2015

102


manufacturers would limit their exports of solar modules to the EU and impose the minimum import price, in exchange for the European Commission forgoing the imposition of anti-dumping duties on those modules within the price and quantity limits. According to the EU, ~75% of the Chinese module manufacturers were covered by the undertakings and were thus exempt from the anti-dumping and countervailing duties. As the current anti-dumping and countervailing duties, as well as the price undertaking agreement will expire by December 2015, we are unable to forecast what the duties and price undertaking agreements will be going forward. To our knowledge, the European Commission has already begun a new round of anti-dumping and countervailing investigations against Chinese solar module manufacturers in July 2015. Also, in March 2015, the European Commission proposed to remove Canadian Solar, Renesola, and ET Solar from the price undertaking, and this was subsequently confirmed in June 2015. The removal was mainly based on these companies’ breaches of the minimum price agreement. Slower-than-expected solar project development As explained in the previous section, we believe JKS’s solar farm development might be impacted by a number of factors, including: 1) payment timing of the FiT subsidy by the central government, as well as capital conditions in China, especially the availability of solar project financing for JKS. Our current estimates for JKS’s solar capacity additions are 600MW per year from 2015-17E, based on our analysis that JKS’s capital raised during 2014 should be adequate for its total capacity up to ~1.6GW. Given the company’s year-end attributable capacity of 503MW as of 2014, we expect JKS to run out of capital in late-2016E. This leads to our conclusion that the most probable timing of JKP’s IPO would be around mid-2016, if the market then is favourable. However, should JKS suffer difficulties in financing, such as banks becoming less willing to lend, or JKP not able to get listed separately due to market conditions, JKS’s capacity additions could miss its target. We estimate a 100MW miss in JKS’s solar farm capacity additions in 2015E and 2016E could lead to a reduction in net profit of 2.75%/1.04% for the following year, respectively.

JKS: earnings sensitivity to 2015E capacity additions JKS: earnings sensitivity to 2016E capacity additions Capacity additions (MW) Net profit impact 2015E 2016E -200MW -5.59% -5.50% -100MW -2.79% -2.75% Base case 0.00% 0.00% +100MW 2.79% 2.75% +200MW 5.59% 5.49%

Capacity additions (MW) Net profit impact 2016E 2017E -200MW -0.49% -2.09% -100MW -0.25% -1.04% Base case 0.00% 0.00% +100MW 0.25% 1.04% +200MW 0.49% 2.09%


Current anti-dumping and countervailing duties in the EU against Chinese solar module manufacturers will expire in December 2015

103


Appendix I: company background

Company introduction JKS was founded in 2006 by Xiande Li (chairman), Kangping Chen (CEO), and Xianhua Li (vice president). At first, the company commenced operations in June 2006 through its then consolidated subsidiary, Jiangxi Desun Energy. Following a series of equity transactions, JKS established a holding company structure with it being the ultimate holding company in 2009. It conducts its business principally through its wholly owned operating subsidiaries in China, Jiangxi Jinko and Zhejiang Jinko. In 2007, 2008 and 2009, JinkoSolar launched its ingot, wafer, and cell and module manufacturing businesses, respectively, creating a vertically integrated solar product value chain. In May 2010, it completed its IPO on the NYSE under the symbol “JKS”, raising USD64.2m in proceeds before expenses. Two follow-on public offerings took place in 2013 and 2014, respectively. JKS: financing history Date Type of financing Capital raised (USDm) No. of ADS (m) 10-May IPO 64.2 5,835 10-Nov Follow-on offering 72 2,000 11-May CB offering 125 n.a. 13-Sep Follow-on offering 71 4,370 14-Jan Follow-on offering 132.2 3,750 14-Jan CB offering 150 n.a. 14-Jul PE agreement 225 n.a.


As of end-2014, JKS’s founders, Xiande Li (chairman), Kangping Chen (CEO), and Xianhua Li (vice president), beneficially owned around 16.2%, 9.7% and 4.9%, respectively (or 30.7% in the aggregate), of JKS’s outstanding ordinary shares. Please refer to the following shareholding structure of JKS. JKS: shareholding structure

Source: Daiwa

JKS’s business segments JKS has 2 major business segments, which are solar product manufacturing business including recoverable silicon, wafers, solar cells, and solar modules, as well as the solar IPP business including power generation and EPC services, etc. As of 2014, solar module sales were still by far the largest revenue stream for JKS, accounting for ~92% of its revenue for that year. That said, we expect revenue from solar power generation to increase gradually over the next few years, due mainly to the ambitious capacity addition plans of the company starting in 2013.

JinkoSolar (JKS)

Brilliant Win Holding Limited Yale Pride Limited Peaky Investment

Limited

16.2% 9.7% 4.9%

Public shareholders

67.2%

Xiande Li Kangping Chen Xianhua Li

Other directors and executive officers

2%

JKS is a leading solar module manufacturer globally. In 2014, it was ranked the fourth largest module manufacturer by shipments globally

104


JKS: revenue breakdown (2012-17E)


Solar product business The company’s manufacturing business is based in both Jiangxi and Zhejiang provinces in China, and utilises a vertically integrated solar power product value chain, from recovering silicon materials to manufacturing wafers, solar cells, and solar modules. Apart from China, the company has manufacturing facilities overseas, including 120MW capacity of module manufacturing plant in South Africa, 40MW of solar cell manufacturing plant in Portugal, as well as the Malaysian plant just completed in May 2015, with a module capacity of 450MW and solar cell capacity of 500MW. Adding up all this manufacturing capacity, JKS’s integrated annual capacity reached 3GW for silicon ingots and wafers, 2.5GW for solar cells, and 4GW for solar modules, as of end-June 2015.

JKS: manufacturing footprint JKS: solar products global sales network

Source: Company Source: Company JKS also has a wide sales and distribution network around the globe, with subsidiaries in the US, the UK, Chile, South Africa, Japan, India, Australia, etc. At end-2014, the company had over 710 customers for its solar modules globally, including distributors, project developers, and system integrators. That said, China and the US remain the key markets for JKS. As of 1H15, 41.4% of module sales were from China, with 18.6% of module sales coming from the US. We expect these 2 markets to continue to be JKS’s key markets given the accelerating solar installations in both countries.

81%94% 92% 90% 89% 85%

2% 6% 8% 11%

0%

20%

40%

60%

80%

100%

2012 2013 2014 2015E 2016E 2017ESolar module Silicon wafer Solar cell Power generation Others

Despite still concentrating in China, JKS has expanded its manufacturing footprint to overseas countries such as South Africa and Malaysia

105


JKS: solar module sales volume by region (1H15) JKS: module products

Source: Company Source: Company Solar IPP business Starting from 2011, JKS also developed downstream solar power projects in China, and thereby selling electricity to the State Grid for the Feed-in Tariffs. As of end-June 2015, the company had 725MW attributable capacity of solar projects completed, 500MW capacity of solar projects under construction, as well as a 5.6GW project pipeline. Among the 725MW capacity of completed solar projects, more than half are located in traditional regions with high solar radiation, such as Xinjiang, Gansu, and Inner Mongolia. That said, JKS also has considerable capacity of projects located in coastal regions such as Jiangsu and Zhejiang provinces.

JKS: solar module sales volume by region (1H15) Gansu Longchang PV

Source: Company, Daiwa Source: Company

Xinjiang Shaya PV Xinyi Songshan PV

Source: Company Source: Company

China41.4%

APAC15.7%

Europe17.6%

Emerging market6.7%

North America18.6%

Xinjiang23.4%

Jiangsu19.2%

Inner Mongolia17.9%

Gansu10.5%

Zhejiang9.8%

Jiangxi6.9%

Qinghai5.0%

Others7.2%

JKS started its solar IPP business in 2011

106


JKS: downstream solar project list

Project name Region Commercial operation date Installed capacity

Ownership Attributable capacity

(MW) FiT

(CNY per kWh) Installation type (MW) Delingha Phase I Qinghai 11-Dec 10 88.70% 8.9 1.2 Utility-scale Delingha Phase II Qinghai 13-Jan 20 88.70% 17.7 1 Utility-scale Longchang Gansu 13-Feb 20 100.00% 20 1 Utility-scale Jinchang Gansu 13-Aug 100 28.00% 28 1 Utility-scale Gansu Jintai Electric II Gansu 13-Oct 100 28.00% 28 1 Utility-scale Gonghe Qinghai 13-Sep 10 100.00% 10 1 Utility-scale Shaya Phase 1 Xinjiang 13-Oct 20 100.00% 20 1 Utility-scale Shaya Phase 2 Xinjiang 13-Dec 20 100.00% 20 1 Utility-scale Ala’er Phase 1 Xinjiang 13-Dec 20 100.00% 20 1 Utility-scale Wusu Xinjiang 13-Dec 20 100.00% 20 1 Utility-scale Bohu Xinjiang 13-Dec 20 100.00% 20 1 Utility-scale Erlongshan Jiangsu 14-May 15 100.00% 15 1.1 Utility-scale Songshan Jiangsu 14-Jun 24 100.00% 24 1.1 Utility-scale Xiangshui Jiangsu 14-Sep 100 100.00% 100 1 Utility-scale Alxa Left Banner Inner Mongolia 14-May 40 86.80% 34.7 0.9 Utility-scale Dengkou Inner Mongolia 14-Jun 50 86.80% 43.4 0.9 Utility-scale Tumed Right Banner Inner Mongolia 14-Sep 60 86.80% 52.1 0.9 Utility-scale Jiaxing Zhejiang 14-Dec 20.9 100.00% 20.9 1.08-1.33 Rooftop DG Jiaxing Zhejiang 15-Mar 12 100.00% 12 1.08-1.33 Rooftop DG Hengfeng Jiangxi 15-Feb 50 100.00% 50 1 Utility-scale Feicheng Shandong 15-Mar 20 95.00% 19 1.2 Utility-scale Ala’er Phase 2 Xinjiang 15-Mar 30 100.00% 30 0.9 Utility-scale Huizhou TCL Guangdong 15-Mar 3 100.00% 3 1.08-1.33 Rooftop DG Shizuishan Ningxia 15-May 10 100% 10 1 Utility-scale Jiaxing Zhejiang 15-May 20 93% 18.5 1 Utility-scale Jiande Zhejiang 15-May 20 100% 20 1 Utility-scale Chabei Hebei 15-May 20 100% 20 1 Utility-scale Atushi Xinjiang 15-May 20 100% 20 0.9 Utility-scale Shufu Xinjiang 15-May 20 100% 20 0.9 Utility-scale


Management profile JKS: management profile Name Age Position Principal roles and responsibilities

Mr. Xiande LI 40 Chairman, Director

Mr. Li is one of the co-founders of JinkoSolar, with over 16 years of PV entrepreneurial experience. His major responsibilities include strategic development of the company, client management, and management over the solar IPP business segment. Prior to founding the company, he served as the marketing manager at Zhejiang Yuhuan Solar Energy Source Co., Ltd. from 2003 to 2004, where his responsibilities included overseeing and optimizing day-to-day operations. From 2005 to 2006, he was the chief operations supervisor of ReneSola.

Mr. Kangping CHEN 42

Chief Executive

Officer, Director

Mr. Chen is the co-founder of the company, and has 18 years of management experience, including 7 years of PV industry experience. He is responsible for strategic development and operation management in JinkoSolar. Prior to founding the company, he was the CFO of Zhejiang Supor Cookware Company Ltd., an A-share listed company, from October 2003 to February 2008, where his major responsibilities included establishing and implementing its overall strategy and annual business plans. Mr. Chen is the brother-in-law of Mr. Xiande Li.

Xianhua Li 41 Director, Vice President

Mr. Li is a founder, director and vice president of JinkoSolar as well as deputy general manager of Jiangxi Jinko. Prior to founding the company, Mr. Li served as the chief engineer of Yuhuan Automobile Company, where his major responsibilities included conducting and managing technology research and development activities and supervising production activities, from 1995 to 2000. From 2000 to 2006, he was the factory director of Zhejiang Yuhuan Solar Energy Source Co., Ltd., where he was responsible for managing its research and development activities. Mr. Li is a brother of Mr. Xiande Li.

Mr. Charlie CAO 38 Chief Financial

Officer Mr. Cao joined JinkoSolar as the financial controller in February 2012, and became the CFO from September 2014. He has over 13 years of experience in accounting, auditing and finance, and is responsible for overall financial management and reporting in the company. Prior to join the company, Mr. Cao was a senior audit manager at PricewaterhouseCoopers from 2002 to 2012. He holds professional accounting qualification including AICPA, CICPA, CTA.

Mr. Arturo HERRERO 43 Chief Sales

Officer Mr. Herrero joined JinkoSolar as the Chief Strategy Officer in March 2010. He has 13 years of PV industry experience, and is responsible for global sales and marketing. Prior to that, he was a Vice President of Sales and Marketing at Trina Solar from 2007 to 2010 and director of Trina Solar from 2006 to 2007, and global procurement manager of BP Solar from 2002 to 2006.



Hong Kong Information Technology

Investment case: We initiate coverage of GCL-Poly with an Outperform (2) rating. With GCL-Poly being the largest polysilicon manufacturer in China as well as the cost leader, we look for its margins to improve meaningfully, especially as we expect a solar installation boom in China over 4Q15 to 2016. We also expect its solar farm revenue to see rapid growth, fuelled by the aggressive capacity expansion plans of solar subsidiary GCL New Energy (GCLNE) (451 HK, not rated). China’s strong downstream demand fuelling growth: We expect GCL-Poly’s polysilicon and wafer sales volumes to rise by 10% YoY each for 2016E, on strong solar demand in China. This, coupled with moderating ASP-decline trend that we see for 2016E (polysilicon: flat vs. 20% YoY fall for 2015E; wafers: 7% YoY decline for 2016E vs. 8% YoY fall for 2015E), as well as GCL-Poly’s polysilicon production costs falling by 16% YoY to USD13.3/kg for 2016E (from USD15.9/kg for 2014), due to the commissioning of its captive power plant and FBR plant, leads us to look for an increase in its solar materials gross margin to 26% for 2016E, from 23% for 2014. GCLNE the next earnings growth driver: We expect GCL-Poly’s longer-term earnings growth to be fuelled by its downstream solar business through GCLNE, for which we expect 1.5GW/1.8GW/2.0GW solar capacity additions for 2015-17E. With GCL-Poly’s cash flow needs satisfied by GCLNE, and GCLNE’s ultimate goal to list its yieldco, as well GCL-Poly’s practice of using asset-backed securities (ABS) for financing, we see minimal risk of share dilution for GCL-Poly. Disposing of its traditional IPP business would benefit GCL-Poly, as it would: 1) give it a clearer business focus on the upstream and downstream solar sectors, 2) improve its net gearing from 145% for 2014 to 112% for 2015E, and 3) provide an immediate after-tax disposal gain of HKD225m. Catalysts: We see 2 major catalysts: 1) the injection of 353MW of solar farms into GCLNE, and 2) the listing of GCLNE’s yieldco. Valuation: We initiate with an SOTP-derived 12-month target price of HKD1.85, which implies a 2016E PER of 10x, on a par with the industry average. If GCLNE succeeds in listing its yieldco, we would expect the TP to rise by 15%, to HKD2.12. We believe GCL-Poly is a good proxy for the China solar sector given its market leadership within the solar chain. Risks: 1) weaker-than-expected solar demand in China, 2) slower-than-expected solar capacity additions by GCLNE, 3) further delayed commissioning of GCL-Poly’s FBR plant, and 4) further CNY depreciation.

29 October 2015

GCL-Pol y Energ y

Initiation: A growing solar giant

Solar installation boom bodes well for GCL-Poly’s volume/margins Incremental profit contribution from GCLNE, with minimal equity dilution Initiate with Outperform (2) rating; EPS CAGR of 28% over 2015-17E


GCL-Poly Energy (3800 HK)

Target price: HKD1.85Share price (27 Oct): HKD1.66 | Up/downside: +11.4%

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

50

64

78

91

105

1.0

1.4

1.9

2.3

2.7



GCL-Poly E (LHS) Relative to HSI (RHS)

(HKD) (%)

12-month range 1.18-2.66Market cap (USDbn) 3.313m avg daily turnover (USDm) 15.65Shares outstanding (m) 15,490Major shareholder Zhu Gongshan (32.4%)

Financial summary (HKD)Year to 31 Dec 15E 16E 17ERevenue (m) 36,868 31,177 33,821Operating profit (m) 7,802 8,677 10,101Net profit (m) 2,021 2,971 3,296Core EPS (fully-diluted) 0.131 0.192 0.213EPS change (%) 4.0 46.7 10.9Daiwa vs Cons. EPS (%) 1.6 3.9 (1.8)PER (x) 12.7 8.6 7.8Dividend yield (%) 5.3 1.9 2.6DPS 0.089 0.032 0.043PBR (x) 1.2 1.1 1.0EV/EBITDA (x) 4.3 5.1 5.0ROE (%) 10.1 13.4 13.7

108

GCL-Poly Energy (3800 HK): 29 October 2015



Growth outlook GCL-Poly: net profit breakdown

We forecast GCL-Poly’s EBITDA to rise at a CAGR of 15% over 2014-17E, leading to its EPS increasing by 18% over the same period. We expect the company’s solar materials business to remain its key earnings driver, accounting for 70% of its overall EBITDA by 2016E. Apart from this, we see an incremental profit contribution from solar subsidiary GCLNE. We forecast GCLNE’s EBITDA to grow from HKD1.2bn for 2015E to HKD5.1bn over 2015-17E, or an EBITDA CAGR of 105% over the same period, as a result of GCLNE’s aggressive capacity addition plans.


Valuation GCL-Poly: one-year forward PER bands

The stock is trading currently at a 9x 2016E PER, or 1.5SD below its past-2-year average of 17.5x. In terms of PBR, the stock is currently trading at a 1.1x one-year forward PBR, which is below its PBR range of 1.3-2.0x in 2012-13, where the whole solar industry entered a recessionary phase (GCL-Poly’s ROAE fell to -19% and -4% for 2012/13, respectively). Our target price of HKD1.85 implies a 2016E PER of 10x, which we believe is still attractive given that we forecast an EPS CAGR of 28% over 2015-17E.


Earnings revisions GCL-Poly: 2015-17E Bloomberg consensus earnings revisions

The Bloomberg-consensus EPS forecasts for GCL-Poly have been revised down since the middle of 2014, due we believe to: 1) China’s disappointing solar installation progress in 2014, and 2) delays in the group’s cost reduction initiatives, especially the commissioning of its captive power plant, which was delayed from March 2014 to July 2015, and 3) the weak polysilicon price trend due to rising polysilicon imports since late-2014.That said, we believe earnings revisions should stabilise from here, given the more stable polysilicon pricing outlook since the middle of 2015, increasing contribution from GCLNE, and cost reduction initiatives that are now on track.

Source: Bloomberg

0

5,000

10,000

15,000

2012 2013 2014 2015E 2016E 2017E

(HKDm)

Solar materials business Solar farm businessSolar farm business - GCLNE Traditional IPP businessCorporate and others

0

5

10

15

20

25

30

Jan-

14

Mar-1

4

May-1

4

Jul-1

4

Sep-

14

Nov-1

4

Jan-

15

Mar-1

5

May-1

5

Jul-1

5

Sep-

15

(x)

25.9x Avg+2SD

21.0x Avg+1SD

16.3x Avg

11.6x Avg-1SD

6.8x Avg-2SD

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Jan-

14Fe

b-14

Mar-1

4Ap

r-14

May-1

4Ju

n-14

Jul-1

4Au

g-14

Sep-

14Oc

t-14

Nov-1

4De

c-14

Jan-

15Fe

b-15

Mar-1

5Ap

r-15

May-1

5Ju

n-15

Jul-1

5Au

g-15

Sep-

15Oc

t-15

(HKD)

2015E EPS 2016E EPS 2017 EPS

109



Profit and loss (HKDm)

Cash flow (HKDm)


Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017EWafer sales (MW) 1,451 4,451 5,594 9,436 12,909 14,071 15,478 16,716Wafer ASP (USD per W) 0.81 0.54 0.25 0.21 0.22 0.20 0.19 0.18Wafer unit cost (USD per W) 0.57 0.33 0.25 0.18 0.17 0.16 0.15 0.14Wafer unit dollar margin (USD per W) 0.24 0.21 0.00 0.03 0.05 0.05 0.04 0.04Polysilicon sales (MT) 10,507 2,812 12,593 16,329 15,443 10,810 11,891 13,080Polysilicon ASP (USD per kg) 52.6 47.9 20.7 17.4 21.7 17.4 17.4 16.5Polysilicon unit cost (USD per kg) n.a. 20.8 19.7 17.0 15.9 14.8 13.3 12.5Polysilicon unit dollar margin (USD per kg) n.a. 27.13 1.03 0.37 5.84 2.61 4.06 3.98

Solar farm installed capacity (MW) 0 0 33 321 1,039 2,539 4,339 6,339

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar material business 14,043 20,477 13,889 18,121 25,455 25,392 26,047 26,708Power business 4,429 4,989 5,807 6,579 10,057 8,126 0 0Other Revenue 0 40 2,652 830 1,712 3,350 5,130 7,113Total Revenue 18,472 25,506 22,348 25,530 37,225 36,868 31,177 33,821Other income 575 613 784 965 1,316 1,653 2,141 2,352COGS (11,661) (17,039) (20,599) (22,490) (29,574) (28,173) (21,856) (22,924)SG&A (1,055) (1,756) (1,995) (1,828) (2,585) (2,546) (2,786) (3,148)Other op.expenses 0 0 0 0 0 0 0 0Operating profit 6,331 7,323 538 2,177 6,382 7,802 8,677 10,101Net-interest inc./(exp.) (559) (1,070) (2,120) (2,204) (2,710) (3,871) (4,355) (5,090)Assoc/forex/extraord./others (225) (414) (1,679) (229) (877) 533 (1,013) (468)Pre-tax profit 5,547 5,839 (3,261) (256) 2,795 4,464 3,310 4,543Tax (1,159) (1,269) (124) (190) (639) (858) (362) (532)Min. int./pref. div./others (364) (295) (130) (218) (200) (421) (491) (716)Net profit (reported) 4,024 4,275 (3,516) (664) 1,955 3,185 2,456 3,296Net profit (adjusted) 4,024 4,275 (3,516) (664) 1,955 2,021 2,971 3,296EPS (reported)(HKD) 0.260 0.276 (0.227) (0.043) 0.126 0.206 0.159 0.213EPS (adjusted)(HKD) 0.260 0.276 (0.227) (0.043) 0.126 0.130 0.192 0.213EPS (adjusted fully-diluted)(HKD) 0.260 0.276 (0.227) (0.043) 0.126 0.131 0.192 0.213DPS (HKD) 0.051 0.055 0.000 0.000 0.000 0.089 0.032 0.043EBIT 6,331 7,323 538 2,177 6,382 7,802 8,677 10,101EBITDA 7,641 9,220 3,280 5,455 10,005 12,204 13,336 15,635

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017EProfit before tax 5,547 5,839 (3,261) (256) 2,795 4,464 3,310 4,543Depreciation and amortisation 1,351 1,958 2,829 3,387 3,780 4,402 4,659 5,534Tax paid (512) (1,857) (191) (78) (533) (862) (428) (501)Change in working capital 942 (4,744) (844) 3,458 (147) 123 (2,233) (180)Other operational CF items 512 1,549 3,828 2,042 3,156 4,171 4,344 5,079Cash flow from operations 7,840 2,745 2,361 8,554 9,051 12,297 9,652 14,476Capex (7,876) (14,987) (4,439) (3,480) (6,398) (16,431) (18,215) (18,494)Net (acquisitions)/disposals (808) (28) (476) 50 749 3,492 0 0Other investing CF items (2,104) (3,211) (430) (3,285) (3,811) 287 1,442 (1,156)Cash flow from investing (10,788) (18,226) (5,346) (6,714) (9,460) (12,651) (16,773) (19,651)Change in debt 4,656 17,445 3,953 491 2,174 12,734 8,623 13,196Net share issues/(repurchases) 2 (36) 3 1,556 744 1 0 0Dividends paid (93) (883) (1,427) (86) (281) 0 (1,372) (491)Other financing CF items (626) (992) (1,913) (2,290) (2,866) (3,871) (4,355) (5,090)Cash flow from financing 3,938 15,534 617 (329) (228) 8,864 2,897 7,614Forex effect/others 0 0 0 0 0 0 0 0Change in cash 1,194 378 (2,387) 1,673 (640) 8,510 (4,224) 2,440Free cash flow (36) (12,242) (2,078) 5,074 2,654 (4,133) (8,563) (4,018)

110


Financial summary continued … Balance sheet (HKDm)

Key ratios (%)


As at 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ECash & short-term investment 8,466 10,954 9,526 14,262 17,360 25,544 19,859 23,425Inventory 1,647 2,473 2,248 1,657 2,696 2,547 1,976 2,073Accounts receivable 2,599 7,065 8,681 11,057 13,658 13,535 11,446 12,416Other current assets 161 1,523 1,714 1,082 605 581 566 573Total current assets 12,873 22,015 22,169 28,057 34,319 42,208 33,846 38,487Fixed assets 23,662 41,181 42,233 43,995 50,612 50,445 64,061 77,082Goodwill & intangibles 1,146 1,062 899 853 806 780 754 727Other non-current assets 2,899 3,230 2,518 3,737 4,266 4,306 4,288 4,301Total assets 40,581 67,488 67,818 76,643 90,004 97,739 102,949 120,597Short-term debt 6,522 12,016 20,170 26,331 25,174 27,238 32,412 42,118Accounts payable 4,384 8,207 9,128 13,737 20,833 19,837 15,389 16,141Other current liabilities 1,724 1,229 1,142 1,977 2,065 2,022 1,717 1,813Total current liabilities 12,630 21,452 30,439 42,045 48,072 49,097 49,517 60,072Long-term debt 7,821 20,800 16,741 14,460 18,794 22,506 25,900 29,390Other non-current liabilities 2,751 3,102 2,867 2,132 1,775 1,165 985 1,068Total liabilities 23,202 45,354 50,048 58,638 68,641 72,768 76,403 90,530Share capital 1,547 1,547 1,548 1,548 1,549 1,549 1,549 1,549Reserves/R.E./others 14,605 19,020 14,662 14,598 16,857 20,043 21,128 23,932Shareholders' equity 16,152 20,567 16,210 16,146 18,406 21,592 22,677 25,481Minority interests 1,227 1,567 1,560 1,859 2,958 3,378 3,869 4,585Total equity & liabilities 40,581 67,488 67,818 76,643 90,004 97,739 102,949 120,597EV 32,472 48,752 54,207 53,566 54,689 52,660 67,394 77,728Net debt/(cash) 5,877 21,861 27,385 26,530 26,608 24,200 38,454 48,083BVPS (HKD) 1.044 1.329 1.047 1.043 1.188 1.394 1.464 1.645

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESales (YoY) n.a. 38.1 (12.4) 14.2 45.8 (1.0) (15.4) 8.5EBITDA (YoY) n.a. 20.7 (64.4) 66.3 83.4 22.0 9.3 17.2Operating profit (YoY) n.a. 15.7 (92.7) 304.8 193.1 22.3 11.2 16.4Net profit (YoY) n.a. 6.2 n.a. n.a. n.a. 3.4 47.0 10.9Core EPS (fully-diluted) (YoY) n.a. 6.2 n.a. n.a. n.a. 4.0 46.7 10.9Gross-profit margin 36.9 33.2 7.8 11.9 20.6 23.6 29.9 32.2EBITDA margin 41.4 36.1 14.7 21.4 26.9 33.1 42.8 46.2Operating-profit margin 34.3 28.7 2.4 8.5 17.1 21.2 27.8 29.9Net profit margin 21.8 16.8 (15.7) (2.6) 5.3 5.5 9.5 9.7ROAE n.a. 23.3 n.a. n.a. 11.3 10.1 13.4 13.7ROAA n.a. 7.9 n.a. n.a. 2.3 2.2 3.0 2.9ROCE n.a. 16.9 1.0 3.8 10.3 11.1 10.9 10.8ROIC 21.5 17.0 1.2 4.9 10.6 13.0 13.5 12.5Net debt to equity 36.4 106.3 168.9 164.3 144.6 112.1 169.6 188.7Effective tax rate 20.9 21.7 n.a. n.a. 22.9 19.2 10.9 11.7Accounts receivable (days) n.a. 69.1 128.6 141.1 121.2 134.6 146.2 128.8Current ratio (x) 1.0 1.0 0.7 0.7 0.7 0.9 0.7 0.6Net interest cover (x) 11.3 6.8 0.3 1.0 2.4 2.0 2.0 2.0Net dividend payout 19.6 19.9 n.a. n.a. 0.0 43.1 20.0 20.0Free cash flow yield n.a. n.a. n.a. 19.7 10.3 n.a. n.a. n.a.

Company profile

GCL-Poly is the largest polysilicon and wafer manufacturer in China, with polysilicon production capacity of 65,000 tonnes and wafer manufacturing capacity of 13GW, as at end-June 2015. In addition, the company operates downstream solar projects, mainly through its subsidiary GCLNE, which had a capacity of 1.1GW (GCL-Poly: 353MW, GCLNE: 773MW) as at end-June 2015.

111


Polysilicon: margin expansion amid steady production volume growth in 2016

Sales volume: surge on expected rush of installations in 2016 As detailed in the sector section of this report, we expect the pace of solar installations in China to pick up rapidly in 4Q15, to 5.5GW, which is a 100% increase QoQ compared to 3Q15. Going into 2016, we expect an even higher rate of solar installations in China (2016E capacity additions of 18GW vs.16GW for 2015E), as we see a high likelihood of a solar on-grid tariff cut starting in 2H16, which would likely to lead to a spate of rush installations in 1H16. Another positive factor we expect is the increase in the renewable energy surcharge for electricity consumers, which should gradually reduce the shortfall in China’s renewable energy fund, and shorten the subsidy payback period for solar farms. We expect the improving cash flow of the solar-farm operators to trigger faster solar-power investment as well over our forecast period.

China: solar installation forecasts China: cumulative solar capacity vs. NEA target

Source: NEA, Daiwa forecasts Source: NEA, Daiwa forecasts

Given our expectation of a solar installation boom in China over the 4Q15-2016 period, we see the solar upstream sectors (ie, polysilicon and wafer) as being the key beneficiaries, on higher demand from the downstream players. In particular, we expect GCL-Poly’s polysilicon and wafer sales to register healthy growth for 2016E, at 10% YoY for each.

GCL-Poly: polysilicon sales GCL-Poly: wafer sales

Source: Company, Daiwa forecasts Source: Company, Daiwa forecasts

-10%

0%

10%

20%

30%

40%

50%

60%

0

5

10

15

20

25

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)


6 17

28 44

62 79

97

116

137

0

20

40

60

80

100

120

140

160

2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(GW)

Installed capacity NEA 12th FYP target

-50%

0%

50%

100%

150%

200%

250%

300%

350%

400%

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

2012 2013 2014 2015E 2016E 2017E

(MT)

Polysilicon sales volume (MT) YoY growth (RHS)

0%

10%

20%

30%

40%

50%

60%

70%

80%

0

2

4

6

8

10

12

14

16

18

2012 2013 2014 2015E 2016E 2017E

(GW)

Wafer sales volume (GW) YoY growth (RHS)

We expect a rush of installation for China’s solar sector in 2016

112


Stable polysilicon pricing in China amid import restrictions Over 2015-17, we also expect the polysilicon manufacturers to benefit from China’s polysilicon import restrictions, and see polysilicon prices recovering gradually from the trough of USD15.0/kg in mid-2015, to around USD18.0/kg by the end of 2015. Diverging trend between prices in international and China markets We anticipate a diverging trend between the international and China markets:

• International market: we forecast international PV grade polysilicon pricing to stay at the depressed level of ~USD15/KG throughout 2015, driven by the worsening supply-demand situation, and

• China market: we expect China’s polysilicon prices to gradually stabilise and pick up slightly in 2H15, following the expiration of foreign import orders in August 2015. We see the spot price recovering from USD15/KG in August 2015 to ~USD18/KG by the end of 2015.

Polysilicon prices

Source: Solarzoom, Bloomberg, Daiwa Globally, demand to grow by 35% and 13% YoY for 2015-16E … In our view, solar demand in China is booming, given the increased awareness of environmental protection in the country, as well as falling PV construction costs. However, we are concerned that the aggressive capacity expansion plans for the tier-1 polysilicon manufacturers will outpace PV demand growth globally in 2016-17. On the demand side, we look for c.53GW/60GW of PV to be installed in 2015/16E, respectively, representing 35% and 13% YoY growth, from 39.4GW for 2014. We see China, with the rush of installations that we expect in 2016 before a potential on-grid tariff cut, and the US, with a rush of installations in 2016 to avoid the Investor Tax Credit (ITC) cut from 30% to 10% starting 2017, to remain the demand growth engines over 2015-17, while India and other developing countries increase in significance further ahead. Global solar installation estimates

Source: BP Statistics, Daiwa forecasts

1012141618202224


(USD/kg)

China local China Imported InternationalChina local (forecast) China Imported (forecast) International (forecast)

010,00020,00030,00040,00050,00060,00070,00080,00090,000

2010 2011 2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E

(MW)


Global polysilicon capacity should outpace demand in 2016-17

113


… slightly lower than capacity growth of 16%/19% for 2015-16E On the supply side, we expect polysilicon production volume to increase by 47,000 tonnes and 68,450 tonnes for 2015 and 2016, respectively, representing respective 16% and 19% YoY growth, outpacing PV demand growth, given the more aggressive capacity expansion schedule for polysilicon. We believe the higher polysilicon capacity growth is due to the relatively balanced supply-demand situation in 2013, when numerous loss-making manufacturers went into bankruptcy, limiting supply resulting in healthy PV demand growth of over 20% during 2013, leading to a slightly increasing polysilicon pricing trend (see the polysilicon price chart above). Given the usual 2-year lead time for polysilicon manufacturing plant construction, most of the tier-1 polysilicon manufacturers that announced capacity expansion plans in 2013 are scheduled to commission their new plants over 2015-16.

Global: polysilicon supply/demand Tier-1 manufacturers capacity vs. PV demand

Source: Company, Daiwa forecasts Source: Company data, Daiwa forecasts

As a result, our market research shows that a significant number of new plants should come online in 2015-16. For example, GCL-Poly plans to add 25,000 tonnes through its FBR plant in 2016, and Wacker will add 20,000 tonnes through its new plant in Tennessee. Also, OCI’s (010060 KR, not rated) P3.9 debottleneck project should unlock around 10,000 tonnes in capacity, with REC Silicon (REC NO, not rated) expanding mainly through debottlenecking and its joint venture in China, which has a capacity of 19,000 tonnes.

Capacity expansion plans for global polysilicon manufacturers Company Capacity(tonnes) Timeline Details OCI 10,000 1Q15 P3.9 debottleneck expected to be completed in 3Q15 Qstec 8,000 1H15 JV with Solarworld with 8,000 tonnes started in 1H15 Daqo 6,000 2Q15 New plant in Xinjiang with 6,000 tonnes capacity fully ramped up in 3Q15; Phase III with 5,850 tonnes is scheduled to be commissioned by

early-2017 Wacker 20,000 2H15 The new plant in Charleston, Tennessee, will commence operations in 2H15 Tokuyama 13,800 1Q16 PS-2 commenced production in 4Q14 and will fully ramp up in 1Q16 Hanwha 3,000 1H16 Debottlenecking expected to be completed in 2H15 SunEdison (MEMC) 13,500 1H16 JV with Samsung with 10,000 tonnes production capacity starting in 2H15, will ramp up to 13,500 tonnes by 1H16 IDEA Polysilicon 10,000 2Q16 Located in Saudi Arabia with 10,000 tonnes production capacity. Expected to completed by 2Q16 GCL-Poly 25,000 2H16 FBR plant in commercial production since 1H15; will ramp up to full capacity into 2016 REC 19,000 1H17 JV company with Shaanxi Non-ferrous Metals established CSG Holdings 6,000 1H17 Capacity expansion plan expected to complete in 2017, with 2,500 tonnes capacity for the semiconductor grade Source: Daiwa

(40%)(30%)(20%)(10%)0%10%20%30%40%

010,00020,00030,00040,00050,00060,00070,00080,000

2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)

Polysilicon production (LHS) PV demand (LHS) % oversupply

Capacity expansion decision made

0%

50%

100%

150%

0

20,000

40,000

60,000

80,000

2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E

(MW)


We believe the strong capacity growth for 2016-17 should be due to the balanced supply-demand situation in 2013

114


Polysilicon manufacturing capacity expansion by company (tonnes)

2015E 2016E 2017E

OCI 10,000 - - Qstec 8,000 - - Daqo 6,000 6,150 5,850 Wacker 10,000 10,000 - Tokuyama - 13,800 - Hanwha - 3,000 - SunEdison (MEMC) 10,000 3,500 - IDEA Polysilicon - 10,000 - GCL-Poly* 3,000 22,000 - REC Silicon - - 19,000 CSG Holdings - - 6,000 Total 47,000 68,450 30,850 % production volume in previous year 16% 19% 8%

Source: Companies, *Daiwa forecasts for GCL-Poly Note: 2016 capacity expansion will have a full impact on 2017 China should be more immune to a supply shock In contrast to our forecasts for sustained pricing pressure for the global polysilicon market, China’s polysilicon market should be relatively better off, due to: 1) the booming Chinese PV market lifting demand, from 10.3GW for 2014 to 16-18GW for 2015-16, on our forecasts, and 2) China’s anti-dumping and countervailing tariffs on foreign polysilicon imports, as well as the forbidden processing trade, which should limit polysilicon supply in China. China announced its anti-dumping and countervailing tariffs for US and Korean polysilicon manufacturers in January 2014, and European polysilicon manufacturers in April 2014, with the US, Korea and EU manufacturers subject to 53.6-57%, 2.4-48.7%, and 14.3-42% anti-dumping duties, respectively. These US and EU manufacturers are also subject to countervailing duties ranging from 0-2.1% and 1.2%, respectively. Anti-dumping and countervailing duties for the major polysilicon manufacturers Anti-dumping duties Countervailing duties US REC Silicon 57.00% 0.20% Hemlock 53.30% 2.10% MEMC 53.60% 0.00% Korea OCI 2.40% 0.00% Hankook 2.80% 0.00% KCC 48.70% 0.00% Others 12.30% 0.00% Europe Wacker 14.30% 1.20% MEMC SpA 42.00% 1.20% Others 14.30% 1.20%

Source: Ministry of Commerce China Pain from the 58th Article should be over soon Since April 2014, many foreign manufacturers have avoided the anti-dumping and countervailing duties by importing polysilicon through “processing trades”. On 14 August 2014, China’s Ministry of Commerce and General Administration of Customs announced a policy (the 58th article) to temporarily suspend the import of solar grade polysilicon through processing trade starting September 2014. The policy gave importers a 2-week grace period so that processing trades approved before September 2014 could still be executed until the contract term was completed. This policy triggered a huge surge in processing trade approvals during the 2-week period, and the industry associations have estimated that approvals during those 2 weeks amounted to 100,000 tonnes, nearly at par with the total import amount of 102,000 tonnes for 2014. We believe China’s overseas dumping order for processing trade from 2014 was the main reason for the slump in polysilicon prices in 1H15. To our understanding, most processing trade contracts have a 1-year term, which implies that the majority of the orders will have expired by the end of August 2015.

China’s strong local demand plus the import restrictions should help sustain local polysilicon prices

Importing polysilicon through “processing trade” hurt China’s polysilicon prices in 2014 and early-2015

115


We believe once those rush orders have been executed, imports from the US and Korea will decline, releasing 20,000+ tonnes of sales volume to China’s local manufacturers (with Wacker and OCI being less affected given their low anti-dumping tariffs). Polysilicon supply in China by country in 2014

Source: Companies, Ministry of Commerce and the General Administration of Customs, compiled by Daiwa We expect China’s poly prices to recover to USD18.0 per kg A surge in China’s PV installations from 10.3GW for 2014 to 16-18GW for 2015-16E would increase demand for polysilicon by 30,210-40,810 tonnes over the period, on our estimates. This, coupled with the capacity released from decline in imports and partially offset by capacity expansion plans locally, leads us to expect all the tier-1 manufacturers will be operating at full capacity by the end of 2015. Please find our scenario analysis below. Scenario analysis on the utilisation rates of the tier-1 manufacturers in China

China's annual PV installation

16GW 17GW 18GW

Local poly production in 2014 (tonnes) 121,500 121,500 121,500 Poly imports in 2014 (tonnes) 102,177 102,177 102,177 Global poly demand in 2014 (tonnes) 223,677 223,677 223,677 2014 local production capacity (tonnes) 132,500 132,500 132,500 % utilisation 92% 92% 92% Incremental PV demand (GW) 5.7 6.7 7.7 Poly demand per MW 5.3 5.3 5.3 Implied additional poly demand from 2014 (tonnes) 30,210 35,510 40,810 Capacity released from imports (tonnes) 30,000 30,000 30,000 Local poly production in 2015 (tonnes) 181,710 187,010 192,310 Capacity addition by Tier-1 manufacturers (tonnes) 47,000 47,000 47,000 2015E Production capacity (tonnes) 179,500 179,500 179,500 % utilisation 101% 104% 107%

Source: Daiwa Our cost curve charts below show that the average breakeven point for the polysilicon manufacturers should be around USD17.5/kg, on a cash-cost basis, which implies that many of the tier-2 manufacturers are struggling currently with the market price of around USD15/kg. However, as we forecast an average utilisation rate of around 100% close to the end of 2015E in China, we expect poly pricing to trend upward eventually, until more tier-2 capacity comes in to make up for the excess demand. That said, we do not expect polysilicon prices to experience a sharp recovery, to USD21/KG or above, as witnessed in late-2013 to 2014. In China, there is a huge amount of idle capacity, which was shut down in 2012 due to sustained depressed polysilicon prices. As such, any price recovery to a sustained USD21/KG or above could trigger a resumption of idle capacity, which could act as a price ceiling for polysilicon prices in China. Hence, we forecast polysilicon to climb back to USD18.0/kg in China by end-2015, which is well within the reasonable price range of USD17.5-21 per kg.

121,500

35,743 30,235 21,133

0

50,000

100,000

150,000

200,000

250,000

Poly supply

(MW)

China Korea Germany US Taiwan Others

End of processing trade imports into China should release polysilicon capacity to local manufacturers

We expect China’s polysilicon price to fluctuate at USD17.5-21/kg

116


Global polysilicon cash cost curve (2015E) Global polysilicon production cost curve (2015E)

Source: Daiwa Source: Daiwa Margins could improve given ample cost-reduction potential We expect GCL-Poly’s polysilicon unit cost to fall by 7% and 10%, reaching USD14.8/KG and USD13.3/KG for 2015-16E, respectively, from USD15.9/KG for 2014, mainly driven by: 1) the commissioning of the company’s captive power plant in July 2015, and 2) GCL-Poly’s self-developed FBR production line, which the company expects to ramp up to a capacity of 25,000 tonnes by the end of 2016E. Given its lower polysilicon cost, GCL-Poly’s wafer production cost are set to decline by 7% and 10% for 2015-16, based on our forecasts, with its wafer unit costs reaching USD0.16/W and USD0.15/W, respectively. This, combined with the stable pricing outlook that we see for both polysilicon and wafers, leads us to project GCL-Poly’s margins to gradually trend up over 2014-16E, with the gross margin for the company’s overall solar materials business improving from 23% for 2014, to 26% for 2016E. GCL-Poly: gross margin


0%

10%

20%

30%

40%

50%

60%

2010 2011 2012 2013 2014 2015E 2016E 2017E

Polysilicon Wafer Overall - solar materials business

GCL-Poly is one of the most cost-effective manufacturers in the polysilicon industry, and we see more cost-cutting ahead, driving down its operating costs further

117


GCL-Poly: polysilicon gross margin GCL-Poly: wafer gross margin

Source: Company, Daiwa forecasts Source: Source: Company, Daiwa forecasts

GCL-Poly: operating figures

1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15

Polysilicon capacity (tonnes) 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 16,250 Polysilicon utilisation rate 75% 80% 47% 26% 53% 82% 83% 92% 99% 100% 106% 106% 110% 117% Polysilicon production (tonnes) 12,235 12,998 7,631 4,191 8,653 13,327 13,550 14,910 16,022 16,319 17,241 17,294 17,800 18,968 Polysilicon price (USD/kg) 26.0 21.8 20.1 13.8 16.6 17.2 17.4 18.0 21.7 22.5 21.8 21.2 19.1 15.4 Polysilicon production cost (USD/kg) 19.7 18.2 20.1 24.1 19.0 16.9 16.7 16.1 16.0 16.1 15.9 15.8 15.5 15.0 Polysilicon unit gross profit (USD/kg) 6.3 3.6 0.0 -10.3 -2.4 0.3 0.7 1.9 5.7 6.4 5.9 5.4 3.6 0.4 Polysilicon gross margin 24.1% 16.3% 0.2% -74.9% -14.2% 1.5% 4.2% 10.4% 26.3% 28.3% 27.0% 25.3% 19.0% 2.3% Wafer capacity (MW) 2,000 2,000 2,000 2,000 2,150 2,475 2,838 2,963 2,963 2,963 3,188 3,250 3,475 3,475 Wafer utilisation rate 58% 94% 84% 45% 57% 90% 90% 89% 97% 102% 115% 108% 100% 104% Wafer production (MW) 1,164 1,878 1,689 891 1,231 2,221 2,546 2,636 2,884 3,019 3,681 3,514 3,478 3,624 Wafer ASP (USD/W) 0.28 0.26 0.25 0.25 0.21 0.21 0.21 0.21 0.23 0.22 0.21 0.21 0.20 0.19 Wafer production cost (USD/W) 0.25 0.23 0.25 0.26 0.22 0.19 0.18 0.17 0.18 0.18 0.17 0.17 0.17 0.16 Wafer unit gross profit (USD/W) 0.03 0.03 0.00 -0.02 -0.01 0.02 0.03 0.04 0.05 0.04 0.04 0.04 0.03 0.03 Wafer gross margin 12.2% 13.3% -2.0% -8.1% -4.5% 10.5% 14.8% 18.8% 22.2% 19.6% 18.0% 18.0% 14.4% 17.9% Source: Company, Daiwa

GCL-Poly: operating assumptions

2012 2013 2014 2015E 2016E 2017E

Polysilicon capacity (tonnes) 65,000 65,000 65,000 70,000 80,000 90,000 Polysilicon utilisation rate 57% 78% 103% 115% 109% 103% Polysilicon production (tonnes) 37,055 50,440 66,876 80,467 86,981 92,556 Polysilicon price (USD/kg) 20.7 17.4 21.7 17.4 17.4 16.5 Polysilicon production cost (USD/kg) 19.7 17.0 15.9 14.8 13.3 12.5 Polysilicon unit gross profit (USD/kg) 1.0 0.4 5.8 2.6 4.1 4.0 Polysilicon gross margin 5.0% 2.1% 26.9% 15.0% 21.5% 21.5% Wafer capacity (MW) 8,000 11,850 13,000 14,000 15,500 16,000 Wafer utilisation rate 70% 74% 101% 101% 100% 104% Wafer production (MW) 5,622 8,827 13,098 14,071 15,478 16,716 Wafer ASP (USD/W) 0.25 0.21 0.22 0.20 0.19 0.18 Wafer production cost (USD/W) 0.25 0.18 0.17 0.16 0.15 0.14 Wafer unit gross profit (USD/W) 0.00 0.03 0.05 0.05 0.04 0.04 Wafer gross margin 0.6% 10.4% 18.8% 22.5% 22.0% 22.6%

Source: Company, Daiwa forecasts Captive power plant should boost margins in 2016E Given that GCL-Poly started out in the coal-fired power-generation business, it has a lot of expertise in power-plant operations. Compared with its peers, which wanted to lower their electricity costs by moving to areas with lower electricity tariffs, GCL-Poly decided to build a captive power plant, and thereby lower its electricity costs itself. Given that the cost of electricity accounted for about 38% of GCL-Poly’s cost of polysilicon manufacturing in 2014, the largest proportion of all its operating costs, we believe that any cost reductions from its captive power plant, which was commissioned in July 2015, would be large. With an installed capacity of 350MW, the captive power plant could generate 2,380m kWh a year, assuming a utilisation rate of 6,800 hours a year, which we estimate to account for about 65% of GCL-Poly’s polysilicon electricity needs.

0

5

10

15

20

25

0

2

4

6

8

2012 2013 2014 2015E 2016E 2017E

(USD/kg)(USD/kg)

Polysilicon unit gross profit (LHS) Polysilicon price (RHS)Polysilicon production cost (RHS)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.00

0.01

0.02

0.03

0.04

0.05

0.06

2012 2013 2014 2015E 2016E 2017E

(USD/W)(USD/W)

Wafer unit gross profit (LHS) Wafer ASP (RHS)Wafer production cost (RHS)

GCL-Poly’s captive power plant commenced operations in July 2015

118


Referring back to GCL-Poly’s unit cost of electricity for its co-generation plants, we estimate that the captive power plant could run at a unit cost of about CNY0.4/kWh, or 45% lower than Xuzhou’s C&I retail tariff of about CNY0.73/kWh. As such, we estimate that the captive power plant could reduce GCL-Poly’s unit electricity cost from about USD6.0/kg to about USD4.3/kg Daiwa’s captive power plant savings calculations Installed capacity (MW) 350 Utilisation (hours) 6,800 Electricity generated (m kWh) 2,380 GCL-Poly's electricity needs for 2015E (m kWh) 3,629 % electricity supported by captive power plant 66% Unit cost of electricity - external (CNY/kWh) 0.73 Unit cost of electricity - captive (CNY/kWh) 0.40 Unit electricity cost - original (USD/kg) 6.0 Unit electricity cost - after captive power plant (USD/kg) 4.3 Electricity cost savings (USD/kg) 1.7

Source: Company, Daiwa estimates GCL-Poly: polysilicon production cost

Source: Company, Daiwa forecasts We also did a sensitivity analysis on the impact of changes in GCL-Poly’s production costs on its gross margin and net profit. We estimate that if the production cost were 2% higher than what we expect, then the company’s gross margin and net profit would be impacted by 0.4pp and 4.3%, respectively for 2015E. For 2016E, if the production cost were 2% higher than what we expect, GCL-Poly’s gross margin and net profit would be affected by 0.5pp and 2.9%, respectively. Gross margin and net profit sensitivity towards changes in production costs Production cost (USD/kg) Gross margin Net profit 2015E 2016E 2015E 2016E +6% -1.1pp -1.6pp -12.8% -8.7% +4% -0.7pp -1.1pp -8.5% -5.8% +2% -0.4pp -0.5pp -4.3% -2.9% base case 0.0pp 0.0pp 0.0% 0.0% -2% 0.4pp 0.5pp 4.3% 2.9% -4% 0.7pp 1.1pp 8.5% 5.8% -6% 1.1pp 1.6pp 12.8% 8.7%

Source: Daiwa FBR technology is the next step GCL-Poly is developing a 25,000 tonnes polysilicon production plant using FBR technology, versus the traditional modified Siemens technology. By improving efficiency and electricity savings, GCL-Poly estimates that the FBR cash cost could be as low as USD8/KG. As at the end of 2014, GCL-Poly had completed the first phase of the production plant for a trial run, with a capacity of 3,000 tonnes. During 2014, it produced around 500 tonnes of polysilicon using the FBR technology. Although, there were general comments that the FBR method resulted in inferior quality versus the modified Siemen method, the company said that the 500 tonnes of polysilicon was of sufficient quality to meet the electronic grade quality requirement.

02468

1012141618

2014 2015E 2016E 2017E

(USD/kg)

Electricity cost Steam Raw materials Labour Depreciation Others

GCL-Poly’s plans to commission 22,000 tonnes of polysilicon capacity through its FBR plant in 2016

119


GCL-Poly plans to commission the remaining 22,000 tonnes of its 25,000 tonnes production lines by the end of 2015, so that the plant will be ready for mass production starting in 2016. However, given that the company is still doing further trial operations and testing, we think it is unlikely that the FBR plant will meet the commission schedule (ie, by the end of 2015). As such, we doubt that its FBR plant will be fully ramped up to 25,000 tonnes capacity in 2016E, and as such we only assume that the company will produce 10,000 tonnes and 15,000 tonnes of polysilicon using the FBR method in 2016 and 17E, respectively. Despite numerous challenges, potential earnings enhancement from FBR could be substantial We analyse the potential cost savings and earnings impact if GCL-Poly were to successfully commission its FBR production plant by the end of 2015, making it ready for mass production starting in 2016E. We estimate that GCL-Poly could achieve a FBR cash cost of USD8/KG, which management previously guided for. Depreciation should be about USD2.9/KG, based on competitor REC Silicon’s estimates of its FBR capex savings in China versus in the US. Considering this, we estimate that GCL-Poly’s FBR total cost could reach as low as USD10.9/KG, or a 20% cost saving to Siemens’ method. Assuming the above cost structure, we estimate that GCL-Poly could lower its overall polysilicon production cost by as much as 6.2% for 2016E, from USD13.6/KG to USD12.8/Kg. This could lead to a gross margin improvement of 1.9pp for 2016E from 26.9% to 28.8%, as well as an EPS enhancement for 2016E of 11.1%, from HKD0.21/share to HKD0.24/share.

GCL-Poly: sensitivity of costs and margins to changes in the FBR production volume

GCL-Poly: sensitivity of earnings and ROE to changes in the FBR production volume

Volume (tonnes) Unit cost Gross margin

2016E 2017E 2016E 2017E 0 2.4% 3.4% -0.7pp -1.0pp 5,000 1.2% 2.2% -0.4pp -0.7pp 10,000 0.0% 1.1% 0.0pp -0.3pp 15,000 -1.2% 0.0% 0.4pp 0.0pp 20,000 -2.4% -1.1% 0.7pp 0.3pp 25,000 -3.5% -2.2% 1.1pp 0.7pp

Volume (tonnes) EPS ROE

2016E 2017E 2016E 2017E 0 -4.9% -6.2% -0.6pp -0.8pp 5,000 -2.4% -4.1% -0.3pp -0.5pp 10,000 0.0% -2.1% 0.0pp -0.3pp 15,000 2.4% 0.0% 0.3pp 0.0pp 20,000 4.9% 2.1% 0.6pp 0.3pp 25,000 7.3% 4.1% 0.9pp 0.5pp

Source: Daiwa Note: Blue highlighted cells are our base-case scenarios

Source: Daiwa Note: Blue highlighted cells are our base-case scenarios

GCL-Poly: cost-reduction strategy

Source: Daiwa Note: Middle bar denotes after the captive power plant is commissioned, and right-hand bar is after the FBR plant is fully ramped up

15.0

13.6

10.9

8

9

10

11

12

13

14

15

16

2Q15 With captive power plant FBR cost

(USD per kg)

We see the FBR cash cost hitting as low as USD8/KG

120


Downstream solar the next earnings driver, through GCLNE While we expect GCL-Poly’s polysilicon and wafer businesses to be relatively stable in the future, driven mainly by cost-reduction initiatives, we see the company’s next earnings growth driver coming from its 62%-owned listed subsidiary, GCLNE. On our forecasts, the subsidiary will contribute 30% of GCL-Poly’s net profit in 2017E, from a zero contribution in 2014. GCL-Poly: net profit breakdown (2017E)

Source: Daiwa GCLNE, the clean energy IPP arm of GCL-Poly, listed on the Hong Kong Stock Exchange through a backdoor listing in May 2014, after the original listed company, Same Time Holdings Limited, issued 360m new ordinary shares to GCL-Poly and placed 50m shares to other places at HKD4.0 per share. The subsidiary is part of GCL-Poly’s plan to expand into the downstream solar business. In our view, GCL-Poly effectively utilises its listed platforms to raise the funds needed to execute its project development plans, which involve: 1) a clean energy IPP, which could raise funds itself, as well as through its yieldco in the future, and 2) GCL System Integration Technology (GCLSIT; 002506 CH, not rated) offering better payment terms to GCLNE for its solar projects. Accelerated capacity expansion over 2015-17E As at end-2014, GCLNE had a solar capacity of 668MW. It plans to add 2.0GW/2.5GW/3.0GW for 2015-17E, respectively, representing a 2014-17E capacity CAGR of 108%. GCLNE’s capacity addition plans seem aggressive, and we believe that annual additions of 1.5GW/1.8GW/2.0GW for 2015-17E, respectively, would be more realistic. For 2015E, GCLNE completed 158MW of solar farms in 1H15 alone, and it has around 1GW of solar-farm capacity already under construction. Hence, we believe the company will be able to complete at least 1.5GW of solar projects this year.

Solar material business68.3%

Power business0.0%

Overseas solar power plant business

1.0%

New energy business30.4%

We view GCL-Poly as more than just a wafer manufacturer, and forecast its new energy business to contribute one-third of GCL-Poly’s net profit for 2017

121


GCLNE: installed capacity forecasts

Source: Daiwa forecasts However, as we have established, we believe that a lack of financing channels will remain a bottleneck to solar farm development in China in the next few years. By extension, companies coming up with more innovative financing measures, such as GCLNE, should continue to outpace their peers in capacity expansion, in our view. We are also confident of GCLNE’s project-execution abilities, considering that most of its project development personnel were formerly with GCL-Poly, which has a long history of solar farm development (the company commenced construction of its first utility-scale solar farm project in back 2009). We note that in just 4 months, GCLNE completed 668MW of solar farms for 2014, despite the construction of its first solar farm only commencing in late-August 2014. Big project pipeline and we expect 1.5GW capacity additions for 2015E We expect GCLNE to meet its 1.5GW capacity target for 2015E, given the following factors:

• As at end-June 2015, GCLNE had completed 158MW of solar projects.

• It has 1GW of solar projects under construction and another 900MW in development. GCLNE: 2015E capacity target breakdown

Source: Company data, Daiwa forecasts Cost leadership, both in construction and operation We like GCL Group’s strong track record in solar development, which we believe will help GCLNE to maintain its cost leadership in both the construction and development phases. We estimate GCLNE’s solar farm construction costs to be as low as CNY8 per W, vs. an average industry level of ~CNY8.5-9 per W and an average M&A cost of ~CNY10 per W. In our view, central to GCLNE’s cost leadership are its relationships with GCL-Poly and GCLSIT, or the legacy Chaori Solar, which allow GCLNE to differentiate itself through payment terms, and quality control.

1.5

1.8

2.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

2014 2015E additions 2016E additions 2017E additions end-2017E capacity

Installed capacity (GW)

158

985

357

(100)

100

300

500

700

900

1,100

1,300

1,500

Capacity (MW)

Projects to be developed Projects under construction Completed projects

Solar farm developers should compete in terms of financing in the next few years

122


By utilising its cost advantage, we estimate that GCLNE could enhance a solar project’s equity IRR from 9.2% to 14.2% (or the project IRR from 7.2% to 8.8%). Hence, through this ability to enhance IRRs, we believe GCLNE stands to capture more solar projects than its peers and thus develop more rapidly. IRR comparison: GCLNE vs. industry average

Source: Daiwa research

Project IRR comparison Equity IRR comparison Unit-capex (CNY per W) 8 8.5 9 9.5 10

Oper

atin

g ex

pens

es

(CNY

per

W) 0.08 8.80% 7.90% 7.20% 6.50% 5.90%

0.085 8.70% 7.80% 7.10% 6.40% 5.80% 0.09 8.60% 7.70% 7.00% 6.30% 5.70%

0.095 8.50% 7.70% 6.90% 6.20% 5.60% 0.1 8.40% 7.60% 6.80% 6.10% 5.50%

Unit-capex (CNY per W) 8 8.5 9 9.5 10

Oper

atin

g ex

pens

es

(CNY

per

W) 0.08 14.20% 11.40% 9.10% 7.20% 5.60%

0.085 13.90% 11.10% 8.90% 7.00% 5.40% 0.09 13.50% 10.80% 8.60% 6.80% 5.20%

0.095 13.20% 10.60% 8.40% 6.50% 5.00% 0.1 12.90% 10.30% 8.10% 6.30% 4.80%

Source: Daiwa research Note: Orange highlight is what GCLNE can achieve; Grey highlight is the industry average

Source: Daiwa research Note: Orange highlight is what GCLNE can achieve; Grey highlight is the industry average

Cost savings in development phase GCLNE prefers to develop solar farms itself rather than acquiring completed solar farms from third parties, an approach which in our view can lower its investment costs by ~CNY1/W relative to a solar farm acquired through M&A. Harnessing the strong record of its development team, GCLNE can undertake the engineering and procurement itself, though it still outsources the construction part to other EPC developers. We estimate that this approach could save the company another ~CNY1/W relative to other developers that outsource the whole EPC process to outside parties. GCLNE: cost savings

Source: Company data, Daiwa research

8.8%

14.2%

7.2%

9.2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

Project IRR Equity IRR

GCLNE industry average

6.0

7.0

8.0

9.0

10.0

Acquisition base casetotal investment cost

Self-developmentcost savings

EPC cost savings Total investment withcombined cost savings

(CNY/W)

10.00

8.00

-1.00

-1.00

GCLNE’s solar farm construction cost is ~CNY8/ W

123


Purchase arrangement with GCLSIT preserves capex According to management, GCLSIT is unable to offer GCLNE deep discounts on GCLNE’s solar panel purchases, given both GCLNE and GCLSIT are listed companies controlled by different boards of directors, and any unreasonable pricing would likely be rejected by GCLSIT’s directors.

Instead, we believe the advantage of GCLNE purchasing solar modules from GCLSIT is the favourable payment terms it can obtain, as GCLSIT urgently needs to sell more modules (as it under “special treatment” status on the A-share market currently – ie, if it continues to make a net loss for 2015, it could get be delisted), which means GCLNE has strong bargaining power in terms of payment terms. As disclosed by GCLNE, for its connected transactions with GCLSIT, GCLNE only needs to pay 20% upfront payment to GCLSIT for its module purchases, with 70% of GCLNE’s payment to GCLSIT only occurring 6 months after the module delivery. This is opposed to the usual industry practice of around a 3-month payment period. The remaining 10% is paid 2 weeks after the 1-year warranty period. GCLNE: payment schedule for solar modules

Source: Daiwa research The prolonged payment period effectively allows GCLNE to preserve its capex, according to management, leading to better cash flow and financing conditions for the company.

Within 1 week of signing Module Sales

Agreement

20%0% 90% 100%Total Purchase Price

Within 6 months of solar modules delivery

Within 2 weeks following the expiration

of 1-year warranty period

GCLNE enjoys better payment terms from GCLSIT than its peers

124


Financials

Minimal share dilution risks from GCLNE expansion We project GCL-Poly’s capex to surge from CNY11.1bn in 2014 to HKD16.4bn in 2015, HKD18.2bn in 2016, and further to HKD18.5bn in 2017, driven mainly by its expansion into the downstream solar business through GCLNE.

GCL-Poly: capex trend GCL-Poly: capex breakdown


On our forecasts, GCLNE’s capex will surge from HKD6.7bn in 2014 to HKD13.0bn in 2015, HKD15.2bn in 2016, and further to HKD16.5bn in 2017, given the company’s 1.5-2.0GW annual capacity expansion plan. That said, we believe the share dilution risks to GCL-Poly should be minimal, given: 1) GCL-Poly’s solar materials business should require limited capex, and 2) GCLNE is a separately listed company with ample financing channels. There is a clear distinction between GCL-Poly and GCLNE in terms of business scope, with GCL-Poly focusing on polysilicon and wafer manufacturing, and GCLNE focusing on downstream solar farm development. Given that both companies are listed, they plan to finance their own expansion at their own listco level. Although GCL-Poly could potentially fund GCLNE’s solar farm development through GCLNE placing equities into GCL-Poly, GCL-Poly has no plans to increase or decrease its 62.28% stake in GCLNE, which suggests that the solar farm capex will be financed by GCLNE itself. As such, we expect GCL-Poly to focus on expanding its solar materials manufacturing, with capex of only ~HKD2-3bn per year over 2015-17E, on our estimates. According to the company, GCL-Poly’s future development will focus mainly on its FBR plant ramp-up in 2016, as well as a mono wafer manufacturing plant in Ningxia. Assuming HKD3bn of capex per year, relative to the company’s 2015-17E EBITDA of HKD8-10bn per year, we see no immediate equity financing need for the company. Indeed, GCL-Poly has not done an equity placement since 2009, which strengthens our belief that it may not need an equity placement at the GCL-Poly level for the next few years. The chart below shows that GCL-Poly’s capex as a percentage of EBITDA (excluding GCLNE) has declined steadily from 174% and 79% in 2012-13 (which we consider to be quite high) to 44% in 2014, and should ease further to ~30% for 2015-16E, on our forecasts.

0

5,000

10,000

15,000

20,000

2011 2012 2013 2014 2015E 2016E 2017E

(HKDm)

Solar materials Power business Solar farm business GCLNE

98%

72%

42%24% 18% 16% 11%

28%

16%

42%

11%

60%79% 84% 89%

0%

20%

40%

60%

80%

100%

2011 2012 2013 2014 2015E 2016E 2017E

Solar materials Power business Solar farm business GCLNE

We expect GCL-Poly’s capex to surge over 2015-17E, but mostly at GCLNE’s level

We see no immediate equity financing need for GCL-Poly

125


GCL-Poly: capex vs. EBITDA

Source: Company, Daiwa Despite our view that there will be no equity placements at the GCL-Poly level, we do not rule out the possibility of GCLNE doing an equity placement in 2016E, given it has high net debt to equity gearing and large initial capital needs. How will GCLNE fund its projects ahead of a yieldco listing? Apart from an equity placement, which we believe would be GCLNE’s choice of last resort, GCLNE plans to utilise a variety of financing channels. For example, it plans to set up various asset-backed securities (ABS) to fund its capex for the coming 2 to 3 years, and in the longer term it plans to list its yieldco for further fundraising. That said, a yieldco listing could be less probable at this stage given: 1) GCLNE’s installed solar capacity is still not large enough for a yieldco (we believe at least 1-2GW of solar projects would need to be injected), 2) the issue of delayed subsidy payment to all solar farm operators in China remains unresolved, and 3) rate hikes in the US would increase the required dividend yield and thus the cost of capital for yieldcos. We estimate that the required return for GCLNE’s ABS should be around 10%, which is still lower than for its solar projects (yielding 12-16% equity IRRs) and should therefore provide the company with a residual return after payments are made to the ABS. After 2 to 3 years of ABS funding, we believe GCLNE may tap into a yieldco listing, should capital-market conditions allow it. The chart below shows our take on GCLNE’s capital funding plans. GCLNE: capital funding before yieldco listing


0%

50%

100%

150%

200%

0

2,000

4,000

6,000

8,000

10,000

12,000

2012 2013 2014 2015E 2016E 2017E

(HKDm)

Capex excl. GCLNE Segment EBITDA excl. GCLNE % of capex to EBITDA (RHS)

GCLNE

YIELDCO

ABS

ABS

GCLNE

20151.5 GW

additionalcapacity

20161.8 GW

additionalcapacity

Capex funding

Capex funding

Capex funding

Capex funding

~10% fixed return

~10% fixed return

2016-17listing

Buy back

residualreturn

residualreturn

For the short term, GCLNE may need to rely on asset-backed securities for financing

126


Potential IPO listing of the yieldco in 2016-17E We believe GCLNE will need to complete the IPO of the yieldco so that GCL-Poly can maintain at least a 50% controlling stake while limiting GCLNE’s net debt-to-equity (our forecasts suggest that 2017E net debt to equity will exceed 500% if no additional equity capital is injected). In the scenario analysis below, we assume the yieldco IPO will be done in 2017 at the latest, with the initial injection of 2GW of capacity into the yieldco. We assume as our base case that yieldco investors require an 8% yield (we show a sensitivity analysis covering yields of 7-9%, as we think GCLNE is unlikely to list its yieldco beyond such rates), which we derive thus: the US 30-year government bond treasury of c.2.8% + BBB- yield spread of 2% + country yield spread of 1.5% + another 1.7% for reasons of conservatism given the limited track records of China yieldcos. On this basis, we estimate that GCLNE could raise some HKD1.7bn cash from a yieldco IPO by selling a 20% stake. Per our analysis shown below, GCLNE could fund its capex independently, ie, without needing GCL-Poly to inject capital into it. As such, even though we expect GCL-Poly’s net debt to equity to surge from 112% in 2015, to 169%/189% in 2016/17, respectively, we do not see any need for equity financing at the GCL-Poly level, as most of the net gearing increase should be due solely to GCLNE’s capacity expansion.

GCLNE: capex funding estimates and GCL-Poly’s net gearing GCLNE: our estimated amount of cash raised from yieldco IPO at 2017 (assuming 20% stake IPO)

2015E 2016E 2017E 2018E 2019E Capacity data

Total capacity (MW) 2,168 3,968 5,968 7,968 9,968 Capacity additions (MW) 1,500 1,800 2,000 2,000 2,000 Per W investment (HKD per W) 8.0 7.8 7.6 7.5 7.4 Cash outflow (13,372) (16,579) (18,819) (19,304) (19,781) - Capex (12,996) (15,215) (16,494) (16,294) (16,094) - Debt financing costs (262) (994) (1,814) (2,755) (3,686) - Dividends from wealth management products (115) (370) (510) (255) 0 Cash inflow 14,383 18,485 16,678 14,961 19,488 - Equity financing 1,332 1,438 0 0 0 - Wealth management products 2,300 2,800 0 (5,100) 0 - Debt financing 10,396 12,172 13,196 14,836 12,876 - Solar farm disposal to yieldco 0 1,000 1,200 1,440 1,600 - Yieldco distribution to GCLNE 355 1,075 2,282 3,786 5,013 GCL-Poly net debt to equity gearing 112% 170% 189% 195% 193%

(HKDm) Calculations Dividend yield assumptions 7% yield 8% yield 9% yield Total capacity in 2016 (MW) 2,000 2,000 2,000 Investment costs 19,651 19,651 19,651 Average utilisation rate (hr) 1,170 1,170 1,170 Net power generation (mn kWh) 2,340 2,340 2,340 Average tariff excl VAT (HKD per W) 1.0 1.0 1.0 Revenue 2,314 2,314 2,314 Depreciation (840) (840) (840) Operating costs (252) (252) (252) Finance costs (1,100) (1,100) (1,100) Net profit 122 122 122 Distributable income 962 962 962 Dividend distribution (assuming 70% payout) 673 673 673 Required dividend yield 7.0% 8.0% 9.0% Market capitalisation 9,620 8,417 7,482 Proceeds from IPO (assuming 20% stake) 1,924 1,683 1,496

Source: Daiwa forecasts Source: Daiwa forecasts Disposal of IPP business to reduce its net gearing In September 2015, GCL-Poly proposed a disposal of the company’s non-solar power business to its principal shareholder, the Zhu family. The business consists of 17 cogeneration power plants, 2 incineration power plants, and a wind power plant, with a total installed capacity of 1,414.5MW, as well as under-construction cogeneration power plants and waste-to-energy incineration power plants with installed capacities of 360MW and 36MW, respectively. The consideration amounts to CNY3.2bn (or HKD4.1bn). After the IPP business disposal, GCL-Poly will cease to operate any non-solar power business. GCL-Poly: proposed IPP business disposal details

Quantity Installed capacity (MW) Attributable capacity (MW)

Coal-fired cogeneration plants 12 399.0 307.2 Gas-fired cogeneration plants 3 870.0 391.1 Biomass cogeneration plants 2 60.0 60.0 Solid-waste incineration plants 2 36.0 36.0 Wind power plant 1 49.5 49.5 Total 20 1,414.5 843.8


We believe GCLNE will fund its capex independently

GCL-Poly plans to dispose of its traditional IPP business by end-2015

127


Apart from the IPP business disposal, GCL-Poly will also get a compensation payment of CNY1.16bn (HKD1.47bn) from the Zhu family, for the cancellation of the Jinshanqiao acquisition, which was suggested by Mr. Zhu in March 2015. Upon the disposal of the IPP business and when the compensation is received for the Jinshanqiao power plant, GCL-Poly has promised to pay out a special dividend amounting to HKD1.37bn, representing 35% of the aggregate of the disposal gain and compensation gain. GCL-Poly: disposal transaction details (HKDm) 2013 2014 1H14 1H15 YoY Revenue 6,217 9,823 4,599 4,941 7% PBT 664 710 268 257 -4% PAT 456 440 151 143 -5% Net income 312 278 114 79 -31% Transaction value (CNYm) 3,200 3,200 3,200 3,200

Transaction value (HKDm) 4,058 4,058 4,058 4,058 PER (x) 13.0 14.6 17.8 25.7 Net asset value - 1H15 (HKDm) 3,492

PBR (x) 1.16

Source: Company, Daiwa We believe the proposed deal will be value-accretive given it will be transacted at a PER of 14.6x versus GCL-Poly’s 2014 PER of 13.3x. Apart from this, we believe the disposal transaction has 3 more major benefits to GCL-Poly: 1) it will give it a clearer business focus on the solar upstream and downstream sectors, 2) an immediate after-tax gain of around HKD225m from the disposal (calculated from the HKD4.1bn transaction price minus HKD3.5bn net asset value, and transaction costs and taxes of HKD341m), and 3) a reduction in the net debt, which we estimate to fall from 145% as at the end of 2014 to 112% as at the end of 2015E. GCL-Poly: net gearing

Source: Company, Daiwa forecasts Note: we assume no asset disposal of 353MW of solar farm capacity to GCLNE

100%110%120%130%140%150%160%170%180%190%200%

2011 2012 2013 2014 2015E 2016E 2017E

GCL-Poly expects to receive CNY1.16bn as a compensation payment from the Zhu family on the cancellation of the Jinshanqiao acquisition

128


CNY devaluation risks GCL-Poly’s business is principally derived from China, with the exception of its wafer business, for which 22% of its sales were derived overseas (or 13% of total revenue) in 2014. Its overseas sales are usually denominated in USD.

GCL-Poly: revenue geographical breakdown GCL-Poly: revenue geographical breakdowns (2014)

2014 2013

Revenue (HKDm) Proportion Revenue (HKDm) Proportion

China 32,567 87% 20,731 81% Taiwan 3,734 10% 2,435 10% Japan 197 1% 289 1% Hong Kong 125 0% - 0% Netherlands 113 0% - 0% US 110 0% 761 3% Germany 102 0% 1,244 5% Others 277 1% 69 0% Total 37,225 100% 25,530 100%


Although most of its revenue is derived from China, the company has about HKD12bn of bank borrowing in USD, as well as 3 outstanding convertible bonds (2 issued by GCLNE) that are USD or HKD denominated and amount to about HKD2.7bn. GCL-Poly: debt currency breakdown

Total debt balance (HKDm) Proportion (%)

Dec-15 Dec-16 Dec-15 Dec-16

Foreign currency borrowing - USD 12,142 8,542 24.4% 14.7% Foreign currency CBs - HKD 200 200 0.4% 0.3% - USD 2,519 2,519 5.1% 4.3% Local currency debt 34,870 47,038 70.1% 80.7% Total debt 49,732 58,300 100.0% 100.0%

Source: Company, Daiwa We see an imbalance between its revenue source currency and borrowing currency, as its overseas revenue is only able to provide a limited natural hedge to its foreign currency borrowing, thus leading to currency risks. Also, as management thinks the company’s current hedging costs are too expensive (about 2-4% currently), it does not hedge its foreign currency risks – our Daiwa economics team forecasts the CNY to depreciate from 6.21 as at the end of 2014, to 6.6 as at the end of 2015, and further to 7.5 as at the end of 2016. GCL-Poly is well aware of the risks to the company of further CNY devaluation, and it has a plan to repay or early retire part of its USD borrowing. The company plans to use the proceeds from the potential disposal of its traditional power business, and the compensation for its Jianshanqiao transaction, after paying for its special dividend and working capital requirements. We expect it to repay HKD3.6bn in 2016, and reduce its USD-denominated debt to HKD11.3bn in 2016, from HKD14.9bn in 2015E. By assuming all of the above, we expect GCL-Poly to incur non-cash FX losses of HKD473m for 2015E and HKD511m for 2016E. However, we would emphasise that these FX losses will be non-cash in nature, and that the FX losses that are cash in nature only amount to HKD17m and HKD30m for 2015 and 2016E, respectively.

China87%

Taiwan10%

Japan1%

Others2%

We estimate that most of the FX loss for GCL-Poly will be non-cash in nature

129


GCL-Poly: FX losses Debt exposure (HKDm) Currency 2015E 2016E USD debts (1H15) USD 12,142 8,542 2018 Zero-coupon convertible bonds (GCLNE) USD 775 775 2018 Convertible bonds (GCLNE) HKD 200 200 2019 Convertible bonds USD 1,744 1,744 Total

14,861 11,261

- minus revenue in USD USD 4,658 4,714 - minus assets in USD / HKD USD / HKD 2,290 2,290 Net USD/HKD liability exposure

7,913 4,257

Assumptions

2015E 2016E USD debt interest rate

3.5% 3.5%

Daiwa USD/CNY forecasts (year-end) 6.6 7.5 Average USDCNY

6.4 7.1

FX loss on: - Interest expense

(17) (30) - Fair value

(473) (511)

Total FX loss

(490) (541) Source: Daiwa forecasts

We also did a sensitivity analysis on the impact of changes in the USD:CNY exchange rate on our forecast FX losses for GCL-Poly. For 2015E, a further 0.1 depreciation of the CNY from our forecast of 6.6 would lead to the FX loss increasing by HKD113m (including a HKD3m cash interest expense and a HKD110m FX adjustment loss). For 2016E, if CNY depreciation turns out to be 0.25 less than our forecast CNY:USD exchange rate of 7.5, then our forecast FX loss would shrink by HKD134m (including a HKD5m lower cash interest expense and a HKD129m lower FX adjustment loss).

GCL-Poly: FX sensitivity analysis 2015E USD:CNY rate FX loss 2016E USD:CNY rate FX loss

cash non-cash

cash non-cash

6.40 (10) (237) 6.50 (7) 65 6.50 (14) (354) 6.75 (13) (95) 6.60 (17) (467) 7.00 (19) (243) 6.70 (20) (576) 7.25 (24) (381) 6.80 (24) (682) 7.50 (30) (511) Source: Daiwa estimates

130


Valuation Target price of HKD1.85; current valuation attractive We initiate coverage of GCL-Poly with an Outperform (2) rating and 12-month target price of HKD1.85, representing 11% upside potential from its current share price. We derive our target price using SOTP methodology, given the diverse nature of its businesses. GCL-Poly: SOTP valuation (HKDm) Methodology Valuation % of value Solar materials EV/EBITDA 54,796 77% Solar farm DCF 4,472 6% New energy DCF 9,374 13% Corporate and others DCF 2,356 3% Corporate value

70,999


(38,476) (Minority interests)

(3,869)

Equity value

28,654 No. of shares (mn)

15,514

Fair value per share (HKD)

1.85

Source: Daiwa Within our SOTP model, we use a 6x 2015E EV/EBITDA multiple for GCL-Poly’s solar material business, which is on a par with the average trading multiple of its solar manufacturing business in China. For its solar farm business (excluding GCLNE), we use a DCF-based valuation as well, with a WACC of 11.4%, assuming beta of 1.2, and a terminal growth rate of 0%.

GCL-Poly: solar farms free cash flow valuation (HKDm) 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E … 2030E Valuation Date 27-Oct-15 15-Nov-15 30-Jun-16 30-Jun-17 30-Jun-18 30-Jun-19 30-Jun-20 30-Jun-21 30-Jun-22 30-Jun-23 30-Jun-30 Next Balance Date 31-Dec-15


Free Cash Flow

EBITDA

597 594 587 579 572 564 557 550 542 494

Less: Other Non Cash

(98) (98) (97) (97) (96) (96) (95) (95) (94) (90)

Less: Cash Tax Payable on EBIT

(42) (42) (41) (39) (38) (37) (36) (35) (34) (27)

Plus: Decrease in Working Capital

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Less: Capital Expenditure

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Free Cash Flow

456.6 454.6 448.7 442.8 437.0 431.3 425.6 419.9 414.3 … 376.8

Free Cash Flow for Valuation Purposes

456.6 454.6 448.7 442.8 437.0 431.3 425.6 419.9 414.3 … 376.8

WACC 10.6% 10.6% 10.6% 10.6% 10.6% 10.6% 10.6% 10.6% 10.6% 10.6% 10.6%

NPV of Free Cash Flow

450.9 421.6 376.4 336.0 299.9 267.6 238.9 213.2 190.3 … 85.6 Source: Daiwa forecasts

For the new energy business under GCLNE, we use a DCF-based valuation for its solar farm assuming a WACC of 8.9%, a beta of 1.2, target gearing of 70%, and a terminal growth rate of 0%.

We use an SOTP methodology to value GCL-Poly

131


GCL-Poly: new energy business free cash flow calculation (HKDm) 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E … 2034E Valuation Date 27-Oct-15 21-Nov-15 30-Jun-16 30-Jun-17 30-Jun-18 30-Jun-19 30-Jun-20 30-Jun-21 30-Jun-22 30-Jun-23 30-Jun-34 Next Balance Date 31-Dec-15


Free Cash Flow

EBITDA

1,215 3,081 5,135 7,359 9,419 11,428 13,192 11,179 10,730 8,899


(139) (290) (425) (663) (935) (1,120) (1,303) (564) (423) (75)


(26) (37) (89) (200) (343) (562) (991) (1,065) (1,269) (1,080)


(301) (589) (569) (547) (613) (605) (636) 45 45 46


(12,996) (15,215) (16,494) (16,294) (16,094) (15,894) (99) (99) (99) (119)

Free Cash Flow

(12,247) (13,050) (12,443) (10,345) (8,568) (6,753) 10,164 9,496 8,984 … 7,671


(12,247) (13,050) (12,443) (10,345) (8,568) (6,753) 10,164 9,496 8,984 … 7,671

WACC 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9%


(12,133) (12,273) (10,742) (8,199) (6,234) (4,510) 6,231 5,345 4,642 … 1,546 Source: Daiwa forecasts

GCL-Poly: new energy business DCF valuation Target gearing (debt/capital) (%) 70.0 Market risk premium (%) 9.6 Risk-free rate (%) 3.4 Cost of debt (%) 8.3 Cost of equity (%) 15.3 WACC (%) 8.9 Terminal Value Terminal Growth Rate 0.00% Terminal WACC 8.93% Estimated Terminal Free Cash Flow 8,048 NPV of Terminal Value (as at 30 Jun 2034) 90,131 NPV of Terminal Value (as at 27 Oct 2015) 18,163 DCF Valuation NPV of Forecasts (HKDm) (8,789) NPV of Terminal Value (HKDm) 18,163 Add: Market value of stakes in Associates #1 - Associate #2 - Associate #3 - Enterprise Value (HKDm) 9,374

Source: Daiwa forecasts We cross-check our target price with the implied PERs. Our target price implies a 9.6x 2016E PER, which is on a par with GCL-Poly’s China solar manufacturing peers, which we see as appropriate given GCL-Poly’s large market capitalisation, leading to many investors using the stock as a proxy for the solar sector value chain. GCL-Poly: price performance vs. peers

Source: Bloomberg

(100%)

(50%)

0%

50%

100%

150%

Oct-1

0

Jan-

11

Apr-1

1

Jul-1

1

Oct-1

1

Jan-

12

Apr-1

2

Jul-1

2

Oct-1

2

Jan-

13

Apr-1

3

Jul-1

3

Oct-1

3

Jan-

14

Apr-1

4

Jul-1

4

Oct-1

4

Jan-

15

Apr-1

5

Jul-1

5

Oct-1

5

3800 HK Equity DQ US Equity JKS US Equity HSCEI Index HSI INDEX

We see GCL-Poly as the proxy for the solar sector value chain

132


GCL-Poly: PER chart GCL-Poly: PBR chart


Solar manufacturers: 2015E PBR vs. ROE Solar manufacturers: 2015E PER vs 2015-17E EPS CAGR


Further upside if GCLNE successfully lists its yieldco We further estimate the potential value enhancement to GCL-Poly should GCLNE list its yieldco set up earlier this year, such that it could dispose of its solar farm assets earlier for better cash recycling and also additional one-off disposal gains. We created a model illustrating how a yieldco can enhance the value of the solar farm operators (refer to Appendix 3). From our estimates, we believe listing its yieldco could enhance GCLNE’s NPV by 20%, assuming a 100% dividend payout and GCLNE selling a 50% stake in its solar farms to 3rd party investors when completed; or 58% if GCLNE only distributes 70% of its cash flow and reinvests the remaining 30% in new solar farms. Value enhancement from yieldco structure

Source: Daiwa research To incorporate the yieldco value enhancement into our forecasts, we assume a disposal premium of ~15% to the yieldco (as guided by the company), as well as GCLNE’s owning 55% of the yieldco.

0

5

10

15

20

25

30

Jan-

14

Mar-1

4

May-1

4

Jul-1

4

Sep-

14

Nov-1

4

Jan-

15

Mar-1

5

May-1

5

Jul-1

5

Sep-

15

(x)

25.9x Avg+2SD

21.0x Avg+1SD

16.3x Avg

11.6x Avg-1SD

6.8x Avg-2SD

0.51.01.52.02.53.03.54.04.5

Jan-

11Ma

y-11

Sep-

11Ja

n-12

May-1

2Se

p-12

Jan-

13Ma

y-13

Sep-

13Ja

n-14

May-1

4Se

p-14

Jan-

15Ma

y-15

Sep-

15

(X)

3.3x Avg+2SD

2.6x Avg+1SD

2.0x Avg

1.3x Avg-1SD

0.6x Avg-2SD

GCL-PolyDaqo

OCIREC

TBEA

Comtec

Trina Solar

Canadian SolarJinkoSolar

JA Solar

-5%

0%

5%

10%

15%

20%

25%

- 0.5 1.0 1.5 2.0

2016

E RO

E(%

)

2016E PBR (X)

GCL-PolyDaqo

OCIREC

TBEA

Comtec

Trina Solar

Canadian SolarJinkoSolar

JA Solar

-5%

0%

5%

10%

15%

20%

25%

- 0.5 1.0 1.5 2.0

2016

E RO

E(%

)

2016E PBR (X)

121

192

25

46

0

50

100

150

200

Assuming holding 100% stake tomaturity

Assuming selling 51% stake forreinvestment

Assuming 70% dividend payout Value enhancement potential

NPV (capex = 100)

GCL-Poly’s fair value could rise if GCLNE were to spin off its yieldco in the future

133


The results below show that using the listed-yieldco strategy could enhance GCLNE’s NPV by ~HKD1bn through to 2034E, given the resulting faster cash flow recycling as well as disposal premium from injecting the farms into the yieldco. The terminal value would also improve due to the greater capacity GCLNE could support under a yieldco. As such, we estimate that GCLNE’s enterprise value could be enhanced to HKD14.7bn, up 41% from HKD10.4bn, but well within our value enhancement range of 20-58%. Consequently, we estimate that our target price for GCL-Poly could have 15% upside potential, from HKD1.85 to HKD2.12. GCL-Poly: SOTP valuation (under a listed yieldco) (HKD mn) Methodology Valuation % of value Solar materials EV/EBITDA 54,796 72% Solar farm DCF 4,472 6% New energy DCF 13,615 19% Corporate and others DCF 2,347 3% Corporate value 75,231 100% Net cash / (net debt) (38,340) (Minority interests) (4,011) Equity value 32,881

No. of shares (mn) 15,514

Fair value per share (HKD) 2.12

Source: Daiwa

GCL-Poly: pro-forma free cash flow calculation for new energy business assuming a yieldco 12 mths to 31 Dec, All figures in HK$millions 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E … 2034E

Valuation Date 27-Oct-15 21-Nov-15 30-Jun-16 30-Jun-17 30-Jun-18 30-Jun-19 30-Jun-20 30-Jun-21 30-Jun-22 30-Jun-23 30-Jun-34

Next Balance Date 31-Dec-15


Free Cash Flow

EBITDA

1,215 3,081 5,135 7,359 9,419 11,428 13,223 12,326 12,473 10,515


(139) 746 698 516 187 (167) (494) 353 (561) (287)


(26) (31) (85) (197) (340) (558) (920) (1,036) (1,280) (1,196)


(301) (589) (569) (547) (613) (605) (603) (181) (196) 53


(12,996) (15,215) (16,494) (16,294) (16,094) (15,894) (6,030) (5,880) (2,241) (119)

Free Cash Flow

(12,247) (12,009) (11,315) (9,163) (7,442) (5,796) 5,176 5,584 8,195 … 8,965


(12,247) (12,009) (11,315) (9,163) (7,442) (5,796) 5,176 5,584 8,195 … 8,965

WACC 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9% 8.9%


(12,133) (11,294) (9,769) (7,262) (5,415) (3,871) 3,173 3,143 4,234 … 1,807 Source: Daiwa forecasts

GCL-Poly: pro forma enterprise value calculations assuming a yieldco Target gearing (debt/capital) (%) 70.0 Market risk premium (%) 9.6 Risk-free rate (%) 3.4 Cost of debt (%) 8.3 Cost of equity (%) 15.3 WACC (%) 8.9 Terminal Value Terminal Growth Rate 0.00% Terminal WACC 8.93% Estimated Terminal Free Cash Flow 9,232 NPV of Terminal Value (as at 30 Jun 2034) 103,386 NPV of Terminal Value (as at 27 Oct 2015) 20,834 DCF Valuation NPV of Forecasts (HKDm) (7,218) NPV of Terminal Value (HKDm) 20,834 Add: Market value of stakes in Associates #1 - Associate #2 - Associate #3 - Enterprise Value (HKDm) 13,615


Potential yieldco listing could enhance GCL-Poly’s value by 15%

134


Key investment risks

Higher-than-expected decline in wafer pricing In our model, we forecast GCL-Poly’s wafer ASPs to fall by 8%/8%/5% for 2015-17E, respectively, which we would regard as healthy given that the price decline would be in line with the continuous production cost decline for the industry. However, investors should note that wafer pricing could fall faster than the unit cost decline if the supply-demand situation changes abruptly. For instance, wafer ASPs dropped by 54% YoY in 2012 while unit costs only fell by 11% YoY over the same period, given the oversupply situation in the solar supply chain, which led to various solar module manufacturers going bankrupt, in turn hampering wafer demand. In that year, GCL-Poly made nearly a 0% gross margin on its wafer sales. GCL-Poly: wafer ASP vs. unit cost change

Source: Company, Daiwa We thus also assessed the sensitivity of GCL-Poly’s wafer gross margin and EPS towards changes in wafer pricing, assuming no change in our unit cost assumptions. We found that each 2pp incremental decline in wafer ASP could lead to its wafer margin declining by 1.7pp/1.6pp for 2015/16E, respectively, and could lead to its EPS declining by 15%/20% for 2015/16E, respectively. GCL-Poly: sensitivity of margin and EPS towards changes in wafer pricing Wafer ASP YoY changes Wafer margin EPS 2015E 2016E 2015E 2016E +4pp 3.2pp 3.0pp 30% 20% +2pp 1.7pp 1.5pp 15% 10% base case 0.0pp 0.0pp 0% 0% -2pp -1.7pp -1.6pp -15% -10% -4pp -3.5pp -3.3pp -30% -20%

Source: Daiwa Slower-than-expected capacity expansion by GCLNE Given the increasing exposure of GCL-Poly to the downstream solar business through GCLNE, if GCLNE misses its capacity addition targets this could drag on GCL-Poly’s earnings in subsequent years. Currently, we assume GCLNE will add capacity of 1.5GW/1.8GW/2.0GW for 2015-17E, respectively. However, we should note that whether GCLNE meets its targets depends on various factors, including: 1) any changes in government policies towards solar power generation, such as solar feed-in tariffs and local government subsidies, 2) GCLNE’s cash flow conditions, which could be affected by further delays to the subsidies to be paid by the central government as well as successful listing of the yieldco.

0%5%10%15%20%25%30%35%40%45%

2010 2011 2012 2013 2014 2015 2016 2017(60%)

(50%)

(40%)

(30%)

(20%)

(10%)

0%

10%

Wafer margin (RHS) Wafer ASP changes Wafer unit cost changes

Wafer margin contraction is the biggest risk for GCL-Poly, in our view

Slower-than-expected capacity additions by GCLNE could hurt GCL-Poly’s long-term value

135


Our sensitivity analysis shows that should GCLNE miss its targets by 500MW in 2015/16E, this would affect GCL-Poly earnings in subsequent years (ie, 2016/17E) by 3.4%/2.2%, respectively. GCL-Poly: EPS sensitivity to GCLNE’s capacity additions in previous years Capacity addition (MW) EPS changes (%) 2016E 2017E -1GW capacity -6.80% -4.30% -500MW capacity -3.40% -2.20% base case 0.00% 0.00% +500MW capacity 3.50% 2.20% +1GW capacity 7.10% 4.50%

Source: Daiwa

136


Appendix 1: company background Founded in 1996, and originally as a China Poly Group company which engaged in green energy businesses, GCL-Poly is now the largest polysilicon and wafer manufacturer in China, with polysilicon production capacity of 70,000 tonnes and wafer manufacturing capacity of 14GW as at June 2015. In addition, the company operates downstream solar projects, mainly through its subsidiary, GCLNE. Excluding the traditional IPP business, which the company plans to dispose of by the end of 2015, GCL-Poly’s operations can be divided into 2 business segments: Solar materials: GCL-Poly mainly manufactures upstream solar materials, ie, polysilicon and wafers. The company is one of the largest polysilicon manufacturers globally, with a production capacity of 70,000 tonnes per year in 2014, or a market share of ~23%. Apart from this, the company also has 14GW of production capacity in wafer manufacturing, which accounts for a ~60% market share in China. Downstream solar farm business: GCL-Poly began its downstream solar farm development in 2009, when it completed the construction and connection of its first 20MW solar project in Xuzhou, Jiangsu Province. In 2014, GCL-Poly acquired Same Time Holdings Limited, which was later renamed GCL New Energy (GCLNE), and started developing its solar farms through the new GCLNE platform. As of 1H15, GCL-Poly owned and operated 1.1GW of solar farms in China and globally, including 353MW of solar farms directly held by GCL-Poly, and a further 773MW of solar projects though GCLNE.

GCL-Poly: revenue breakdown by business segment GCL-Poly: EBITDA breakdown by business segment


Two major restructuring exercises done in 2014 and 2015 GCL-Poly’s major revenue drivers are the solar materials business (polysilicon and wafer manufacturing) and traditional IPP business. In 2014-15, however, the company underwent 2 major restructurings of its business, namely, the acquisition of GCLNE in 2014 and the disposal of its traditional IPP business in 2015, which shifted GCL-Poly’s business nature towards a pure solar company, with a dual focus on both upstream solar materials and downstream solar farm development businesses. In February 2014, GCL-Poly entered into a subscription agreement with Same Time Holdings to subscribe for 360m new shares of Same Time at a price of HKD4.00 per share. The acquisition was completed in May 2014 and Same Time was renamed GCL New Energy thereafter. This acquisition marked GCL-Poly’s commitment to develop its downstream solar farm business, with an equity stake in GCLNE of 62.28%. In September 2015, the company announced a proposed disposal of its traditional IPP business to its major shareholder, the Zhu family. The transaction involves 17 cogeneration power plants, 2 incineration power plants, and a wind power plant, with total installed capacity of 1,414.5MW, as well as the under-constructed cogeneration power plants and

62% 71% 68% 69%83% 79%

12% 3%2% 2%

2% 2%3% 7%

14% 19%26% 26% 27% 22%

0%

20%

40%

60%

80%

100%

2012 2013 2014 2015E 2016E 2017E

Solar materials business Solar farm businessSolar farm business - GCLNE Traditional IPP business

53%67% 79% 72% 68% 61%

3%1%

3% 5% 4%4%

10% 23% 33%36%

29%16% 11%

(20%)

0%

20%

40%

60%

80%

100%

2012 2013 2014 2015E 2016E 2017ESolar materials business Solar farm businessSolar farm business - GCLNE Traditional IPP businessCorporate and others

GCL-Poly’s 2 major businesses will be: 1) solar materials (poly-si and wafer) manufacturing, and 2) its downstream solar farm operations, after the disposal of its traditional IPP business at end-2015

137


waste-to-energy incineration power plants, with installed capacity of 360MW and 36MW, respectively. The consideration will amount to CNY3.2bn (or HKD4.1bn). After the disposal of the IPP business, GCL-Poly will cease to operate non-solar power business. GCL-Poly: proposed IPP business disposal details

Quantity Installed capacity (MW) Attributable capacity (MW)

Coal-fired cogeneration plants 12 399.0 307.2 Gas-fired cogeneration plants 3 870.0 391.1 Biomass cogeneration plants 2 60.0 60.0 Solid-waste incineration plants 2 36.0 36.0 Wind power plant 1 49.5 49.5 Total 20 1,414.5 843.8

Source: Company, Daiwa Apart from the IPP business disposal, the GCL-Poly will also receive a compensation payment of CNY1.16bn (HKD1.47bn) from the Zhu family upon the cancellation of the Jinshanqiao acquisition, which had been promised by Mr. Zhu during the annual results in March 2015. GCL-Poly: disposal transaction details (HKDm) 2013 2014 1H14 1H15 YoY Revenue 6,217 9,823 4,599 4,941 7% PBT 664 710 268 257 -4% PAT 456 440 151 143 -5% Net income 312 278 114 79 -31% Transaction value (CNYm) 3,200 3,200 3,200 3,200

Transaction value (HKDm) 4,058 4,058 4,058 4,058 PER (x) 13.0 14.6 17.8 25.7 Net asset value - 1H15 (HKDm) 3,492

PBR (x) 1.16

Source: Company, Daiwa As such, from 2015E onwards, we should start to see solar farm development (including GCLNE) gain significance within GCL-Poly’s business portfolio, with its revenue contribution up from 5% in 2014 to 21% in 2017E and EBITDA contribution up from 3% in 2014 to 37% in 2017E, based on our forecasts. GCL-Poly: shareholding structure (as of 1H15)


32.4%

Asia Pacific Energy Fund Limited

5.3%

PAG Holdings Limited

10.1%

JP Morgan Chase & Co

Zhu Gongshan

GCL-Poly

Templeton Global Advisors Limited

Templeton Investment

Counsel, LLC

5.0% 41.1%

Public shareholders

6.0%

We should see revenue from its solar farm operations (including GCLNE) become more important from 2015

138


GCL-Poly: management profile Name Age Position Details Mr. ZHU Gongshan 57 Chairman, CEO Mr. Zhu is the founder of GCL-Poly, and has been the company's Executive Director since

2006. He is currently the member of the 12th National Committee of the CPPCC, and also holds multiple leadership roles in various professional associations and social groups, such as Chinese Society for Electrical Engineering , China Industrial Overseas Development & Planning Association, etc. In addition, he is also the Honorary Chairman and an executive director of GCL New Energy Holdings Limited.

Mr. JI Jun 67 Executive Director Mr. Ji has been an Executive Director of the Company since November 2006. He is also a member of the Strategic Planning Committee of the Company. Mr. Ji focuses on strategic planning and business development of the Group. He has extensive experience in the power industry and has experience in handling corporate finance projects.

Mr. ZHU Yufeng 33 Executive Director Mr. Zhu has been an Executive Director of the Company since September 2009. He is also a member of the Remuneration Committee of the Company. He is responsible for internal control, human resources, administration and project tender of the Group. Mr. Zhu is also the Vice Chairman and a nonexecutive director of GCL New Energy Holdings Limited.

Mr. YEUNG Man Chung, Charles

47 Executive Director & Executive President

– Solar Business

Mr. Yeung has been an Executive Director of the Company since 2014, and is also a member of the Nomination Committee, Corporate Governance Committee and Strategic Planning Committee of the Company. He is a former partner of Deloitte Touche Tohmatsu and, with over 20 years of experience in accounting, auditing and financial management. He is responsible for the financial control and reporting, corporate finance, tax and risk management of the Company and its subsidiaries.

Mr. SHU Hua 45 Non-executive Director

Mr. Shu was the company's Executive Director for the period from 2007 to 2015 and has been re-designated as a non-executive director of the Company since January 2015. He has over 20 years of experience in the energy industry, and has obtained a Master’s degree in Business Administration for Senior Management from the Tongji University in the PRC. Mr. Shu is also the chairman of GCL System Integration Technology Co., Ltd, a company with its shares listed in the Shenzhen Stock Exchange (Stock Code: 2506) with its majority board members deemed to be controlled by Mr. Zhu Gongshan’s and Mr. Zhu Yufeng’s family.


139


Appendix 2: FBR technology The Fluidized Bed Reactor (FBR) method is a relatively new polysilicon process through which polysilicon is deposited when monosilane gas mixes with silicon seed particles in a reactor. REC Silicon is one of the earliest pioneers in this technology, beginning the development of this technology in 2002. Currently, over 80% of REC Silicon’s polysilicon production is from FBR technology. The company also has a plan to add an 18,000-tonne FBR production plant in China, which is targeted to be completed by the end of 2016. Apart from REC Silicon, both GCL-Poly and SunEdison have plans to develop a FBR production plant each, with capacity of 25,000 tonnes and 16,000 tonnes, respectively. FBR technology is a less costly method compared with the traditional Siemens method for producing high purity solar-grade silicon. According to REC Silicon, FBR only needs 9% of the electricity a conventional Siemens reactor consumes when producing the same amount of polysilicon (4.1kWh/kg vs. 45kWh/kg). The better energy efficiency is due to 3 reasons. • First, it consumes less energy during the production process. In the traditional Siemens

method, a significant amount of unproductive energy is consumed for the purpose of superheating the silicon gas and seed rods, and at the same time cooling the chamber walls. On the contrary, the FBR method fluidizes silicon granules with heated silane gas inside a heated-wall chamber, and thus does not waste energy by placing heated gas and silicon in contact with cold surfaces.

Reactor structure: Siemens vs. FBR

Source: REC Silicon • Second, it can achieve better utilisation of the reactor space, producing more silicon per

cubic metre. As the fluidised silicon granules have a larger total surface area than the silicon rods in a Siemens process, the silane gas can deposit a larger amount of silicon layers on the granules than on the rods.

• Third, unlike the batch process in the Siemens method, an FBR silicon refining process

is continuous, which leads to less wasted downtime for replacing seed rods and setting up the electrical contacts.

REC Silicon is one of the earliest pioneers in FBR technology

140


REC Silicon: FBR cost breakdown REC Silicon: FBR cost trend


REC Silicon’s FBR cash costs have been stable since 3Q14 at ~USD10-11/kg, and its total production cost ~USD18-19/kg, which might seem average compared to other polysilicon manufacturers. However, REC Silicon estimates that the FBR plants it is planning to build in China (19,000 tonnes by the end of 2017) could enjoy further cost reductions, mainly due to lower raw-material, manufacturing and labour costs, which it estimates should bring down its cash cost from ~USD12-13/kg (in the US) to ~USD8/kg (in China). This, coupled with its more advanced FBR plants in China vs. those in the US and lower land costs in China, could mean that REC Silicon’s initial capex for its new Chinese plants is reduced by half, and therefore its depreciation cost lowered by ~50% from ~USD6/kg (in the US) to only USD3-4/kg (in China) by 2017, according to company estimates. As such, we believe GCL-Poly’s cost reduction from FBR plant development could be substantial, based on REC Silicon’s initial estimates.

Electricity - FBR0.6%

Electricity -Silane2.3%

Raw materials21.3%

Manufacturing & labour18.4%

Depreciation32.5%

SG&A9.9%

Others15.0%

12.5 12.1 10.513.6 14.3

10.5 10.8 10.7 11

5.9 5.95.7

7.6 6.8

5.3 5.9 6.3 6.2

2.4 2.52.2

3 2.5

2.2 2 1.8 1.9

0

5

10

15

20

25

2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15

(USD per kg)

FBR cash cost Depreciation SG&A

REC expects its cash costs to fall from USD10-11/KG in the US to USD8/KG in China

141


Appendix 3: value enhancement through the yieldco model The benefits to a company of early solar farm disposal via a yieldco include: 1) moving forward the company’s cash inflow and realising the profits earlier, and 2) enabling the company to use the cash proceeds for reinvestment. The following yieldco cash flow model illustrates the NPV effects of early disposal (at a ~19% premium) and further reinvestment, versus the NPV of a hold-to-maturity strategy.

Yieldco cash flow model Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 Year 20 Assuming holding to maturity Capex 100 FCFE -100 14 14 14 14 14 14 14 14 14 14 14 IRR 13.0% NPV 121 Assuming reinvesting from cash recycled through 50% project sales Projects sold 60 35 21 13 7 4 3 2 1 1 - Project invested / reinvested 100 60 35 21 13 7 4 3 2 1 1 -

Return by projects

Year 0 14.2 14.2 14.2 14.2 14.2 14.2 14.2 14.2 14.2 14.2 14.2 Year 1 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 Year 2 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Year 3 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 Year 4 1.8 1.8 1.8 1.8 1.8 1.8 1.8 Year 5 1.1 1.1 1.1 1.1 1.1 1.1 Year 6 0.6 0.6 0.6 0.6 0.6 Year 7 0.4 0.4 0.4 0.4 Year 8 0.2 0.2 0.2 Year 9 0.1 0.1

Year 10 0.1

Project returns (assuming 50% sales after one year)

Year 0 14.2 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 Year 1 - 8.5 4.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2 Year 2 - - 5.0 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 Year 3 - - - 3.0 1.5 1.5 1.5 1.5 1.5 1.5 1.5 Year 4 - - - - 1.8 0.9 0.9 0.9 0.9 0.9 0.9 Year 5 - - - - - 1.1 0.5 0.5 0.5 0.5 0.5 Year 6 - - - - - - 0.6 0.3 0.3 0.3 0.3 Year 7 - - - - - - - 0.4 0.2 0.2 0.2 Year 8 - - - - - - - - 0.2 0.1 0.1 Year 9 - - - - - - - - - 0.1 0.1

Year 10 - - - - - - - - - - 0.0 Total return 14.2 22.7 27.7 30.7 32.5 33.6 34.2 34.6 34.8 34.9 35.0 Dividend payout 14.2 22.7 27.7 30.7 32.5 33.6 34.2 34.6 34.8 34.9 35.0 Residual - - - - - - - - - - - Return to company 14.2 15.6 16.4 16.9 17.2 17.3 17.4 17.5 17.5 17.5 17.5 Dividend payout 14.2 15.6 16.4 16.9 17.2 17.3 17.4 17.5 17.5 17.5 17.5 FCFE -100 14.2 15.6 16.4 16.9 17.2 17.3 17.4 17.5 17.5 17.5 17.5 IRR 15.8% NPV 146 Multiplier 1.20 Source: Daiwa forecasts

142



China Industrials

What's new: After Xinyi Solar’s (XYS) top-up share placement on 27 October (168.8m shares at HKD3.22/share), we believe the company’s capital-raising plan for its 2015-16 solar-farm development is on track. We are therefore optimistic that XYS can deliver 800MW of new solar-farm capacity in 2016. What's the impact: 800MW in new solar-farm capacity in 2016E: Under the equity investment agreement between Xinyi Energy, XYS, and Xinyi Glass’s director, both shareholders promised to invest an additional HKD3bn into Xinyi Energy for solar farm development. This, coupled with our understanding that XYS is also in the process of inviting more third-party investors into Xinyi Energy, reinforces our view that XYS should have enough capital to develop more solar farms in 2016 vs. 2015. Solar-glass margin improvement should continue: We expect the solar glass gross margin to improve from 31% in 2014 to 37% in 2015, due to: 1) cost reductions from 2x900tpd production lines, 2) raw material and gas cost declines, and 3) a favourable product mix shift. We look for further cost reductions, such as technical upgrades of its remaining 2x500 production lines in 2016 (increasing capacity for each from 500tpd to 600tpd), and the commissioning of the Malaysia plant, with lower energy and labour costs, targeted to be completed by end-2016. For XYS’s plan to expand its raw glass production capacity, from 3,900tpd to 5,900tpd, by building two 1,000tpd lines in Wuhu, Anhui, we expect revenue contribution from these lines to kick in by 2018 at the earliest, given the around 2-year construction period for the production lines. What we recommend: We reiterate our Outperform (2) rating for XYS, and slightly tweak our EPS forecasts by 0-1% following the share placement on 27 October, reflecting lower debt financing needs. Nonetheless, we slightly lower our SOTP-derived 12-month target price to HKD3.50 from HKD3.60. We are positive about its solar-glass volume growth, which would be backed by the solar installation boom we expect in 2015-16. XYS shares are trading at a 13x 2016E PER, which we believe is undemanding vs. our forecast 2015-17E EPS CAGR of 29%. Near-term share-price catalysts: 1) the announcement of new supportive solar policies by the government; and 2) Xinyi Energy securing third-party investors. Key risks: a slowdown in solar installation globally and XYS being unable to secure third-party investors for Xinyi Energy. How we differ: Our 2015-17E EPS are 2-10% below consensus, as it has yet to adjust for the EPS impact on Xinyi Energy’s 25% stake disposal.

29 October 2015

Xinyi Solar H oldi ngs

Solar glass giant going downstream

Solar glass margin should continue to improve in 2016 We expect 800MW solar farm additions in 2016, outpacing 2015 Reiterate Outperform (2); valuation looks undemanding



Xinyi Solar Holdings (968 HK)

Target price: HKD3.50 (from HKD3.60)

Share price (27 Oct): HKD3.28 | Up/downside: +6.7%

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

Forecast revisions (%)Year to 31 Dec 15E 16E 17ERevenue change - - 1.4Net profit change 3.8 4.2 3.4Core EPS (FD) change 0.7 1.4 0.6

70

86

103

119

135

1.8

2.3

2.9

3.4

3.9



Xinyi Sola (LHS) Relative to HSI (RHS)

(HKD) (%)

12-month range 1.99-3.90Market cap (USDbn) 2.853m avg daily turnover (USDm) 12.33Shares outstanding (m) 6,749Major shareholder Xinyi Glass (29.3%)

Financial summary (HKD)Year to 31 Dec 15E 16E 17ERevenue (m) 4,617 6,226 7,689Operating profit (m) 1,397 1,996 2,686Net profit (m) 1,194 1,540 1,985Core EPS (fully-diluted) 0.190 0.246 0.318EPS change (%) 126.1 29.4 28.9Daiwa vs Cons. EPS (%) (2.3) (10.1) (5.8)PER (x) 17.2 13.3 10.3Dividend yield (%) 2.4 3.1 4.0DPS 0.080 0.103 0.132PBR (x) 2.9 2.6 2.2EV/EBITDA (x) 14.4 11.6 9.5ROE (%) 21.8 18.9 21.4

144

Xinyi Solar Holdings (968 HK): 29 October 2015



Growth outlook XYS: net-profit growth forecasts by segment

We expect XYS’s net profit to record a 28% CAGR from 2015-17 (EPS CAGR: 29%), with net profit reaching HKD2.0bn in 2017, from only HKD493m in 2014. We expect solar glass revenue to grow by 17% in 2016 and 13% in 2017, driven by growing solar installation globally, as well as XYS’s market-share gains within the solar glass sector. For the solar farm business, we expect XYS to add solar-farm capacity of 754MW in 2015 and 800MW in 2016, a major earnings-growth driver for the segment.


Valuation XYS: rolling 1-year-forward PER chart

XYS is currently trading at a 1-year-forward PER of 14.6x (or a 2016E PER of 13.3x), which is slightly below (-0.2SD) its average 20-month historical PER of 15.3x. Our target price of XYS is HKD3.50 per share, implying a 2016E PER of 14.2x, which is around its historical average. We believe the current valuation of a 13.3x 2016E PER is undemanding, as we anticipate a solar installation boom in 4Q15-16, which should boost investment sentiment towards the solar sector.


Earnings revisions XYS: Bloomberg EPS consensus revisions

The Bloomberg consensus EPS for 2015-17 have risen by 21-35% YTD, which suggests that the outlook for the company has been increasingly positive. Our 2015-17E EPS are 2-10% below consensus, as the consensus has yet to adjust for the EPS impact on Xinyi Energy’s 25% stake disposal.

Source: Bloomberg

(100%)

(50%)

0%

50%

100%

150%

200%

0

500

1,000

1,500

2,000

2,500

2011 2012 2013 2014 2015E 2016E 2017E

(HKDm)

Solar glass Solar farmCorporate & others YoY net profit growth (RHS)

6.0

11.0

16.0

21.0

26.0

Dec-1

3

Mar-1

4

Jun-

14

Sep-

14

Dec-1

4

Mar-1

5

Jun-

15

Sep-

15

(x)

22.0x Avg+2SD

18.6x Avg+1SD

15.3x Avg

11.9x Avg-1SD

8.5x Avg-2SD

0.050.100.150.200.250.300.350.40

Feb-

14Ma

r-14

Apr-1

4Ma

y-14

Jun-

14Ju

l-14

Aug-

14Se

p-14

Oct-1

4No

v-14

Dec-1

4Ja

n-15

Feb-

15Ma

r-15

Apr-1

5Ma

y-15

Jun-

15Ju

l-15

Aug-

15Se

p-15

Oct-1

5

(HKD)


145



Profit and loss (HKDm)

Cash flow (HKDm)


Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017E

Total solar glass sales volume (k tonne) 179 242 360 536 557 906 1,073 1,303

Overall solar glass gross margin (%) 51.5 35.9 17.2 30.1 30.9 36.5 37.3 38.2Total solar capacity (MW) 0 0 35 35 308 1,092 1,892 2,692Average utilisation hours n.a. n.a. n.a. n.a. n.a. 1,100 1,100 1,100

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESolar glass 1,077 1,234 1,533 1,968 2,379 4,287 4,997 5,644Solar farm 0 0 0 0 31 330 1,229 2,044Other Revenue 0 0 0 0 0 0 0 0Total Revenue 1,077 1,234 1,533 1,968 2,410 4,617 6,226 7,689Other income 39 66 40 63 87 79 20 20COGS (523) (791) (1,269) (1,375) (1,649) (2,857) (3,531) (4,120)SG&A (75) (65) (54) (187) (160) (246) (213) (168)Other op.expenses (38) (53) (108) (100) (111) (196) (506) (736)Operating profit 480 391 143 369 577 1,397 1,996 2,686Net-interest inc./(exp.) 0 0 1 2 (5) (35) (81) (194)Assoc/forex/extraord./others 0 0 0 0 0 0 0 0Pre-tax profit 480 391 144 370 572 1,362 1,916 2,491Tax (72) (70) (24) (67) (79) (168) (177) (203)Min. int./pref. div./others 0 0 0 0 0 0 (199) (303)Net profit (reported) 408 321 120 304 493 1,194 1,540 1,985Net profit (adjusted) 408 321 120 304 493 1,194 1,540 1,985EPS (reported)(HKD) 0.117 0.092 0.034 0.073 0.084 0.183 0.228 0.294EPS (adjusted)(HKD) 0.117 0.092 0.034 0.073 0.084 0.190 0.246 0.318EPS (adjusted fully-diluted)(HKD) 0.117 0.092 0.034 0.073 0.084 0.190 0.246 0.318DPS (HKD) 0.000 0.000 0.000 0.018 0.040 0.080 0.103 0.132EBIT 480 391 143 369 577 1,397 1,996 2,686EBITDA 520 439 253 468 675 1,608 2,378 3,262

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017EProfit before tax 480 391 144 370 572 1,362 1,916 2,491Depreciation and amortisation 40 49 110 100 98 211 382 576Tax paid (10) (45) (61) (44) (67) (168) (177) (203)Change in working capital (45) (405) (110) 224 117 (452) 115 217Other operational CF items 9 (1) 3 3 6 35 81 194Cash flow from operations 474 (10) 87 654 725 988 2,316 3,276Capex (550) (819) (348) (240) (2,322) (4,298) (5,871) (5,718)Net (acquisitions)/disposals (1) 0 0 2 0 0 0 0Other investing CF items 53 (33) 39 113 (6) 16 25 12Cash flow from investing (498) (852) (309) (125) (2,328) (4,282) (5,847) (5,706)Change in debt 0 0 0 0 1,300 2,149 2,936 2,859Net share issues/(repurchases) 0 0 0 0 778 1,694 0 0Dividends paid 0 0 0 0 (200) (146) (537) (693)Other financing CF items 22 874 244 (305) 0 1,529 (105) (207)Cash flow from financing 22 874 244 (305) 1,878 5,226 2,293 1,959Forex effect/others 1 1 0 1 (2) 0 0 0Change in cash (2) 12 22 225 274 1,931 (1,238) (471)Free cash flow (76) (830) (261) 414 (1,596) (3,311) (3,556) (2,442)

146


Financial summary continued … Balance sheet (HKDm)

Key ratios (%)


As at 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ECash & short-term investment 20 32 54 279 543 2,474 1,236 766Inventory 70 415 217 91 299 506 589 654Accounts receivable 241 432 641 705 760 1,826 2,087 2,205Other current assets 8 4 18 0 0 0 0 0Total current assets 340 883 931 1,075 1,602 4,807 3,912 3,625Fixed assets 1,170 1,832 1,655 1,368 3,685 7,777 13,270 18,415Goodwill & intangibles 308 190 188 189 180 176 173 169Other non-current assets 154 22 22 55 296 296 296 296Total assets 1,972 2,927 2,796 2,687 5,764 13,056 17,651 22,505Short-term debt 0 0 0 0 143 0 0 0Accounts payable 134 381 190 350 1,121 1,942 2,400 2,800Other current liabilities 499 1,536 1,512 17 28 28 28 28Total current liabilities 633 1,917 1,701 367 1,291 1,970 2,428 2,828Long-term debt 0 0 0 0 1,157 3,449 6,385 9,244Other non-current liabilities 16 89 29 10 10 10 10 10Total liabilities 649 2,006 1,731 377 2,458 5,428 8,822 12,081Share capital 900 900 900 1,706 2,241 5,515 5,515 5,515Reserves/R.E./others 423 21 165 604 1,065 2,113 3,115 4,407Shareholders' equity 1,323 921 1,065 2,310 3,306 7,628 8,630 9,922Minority interests 0 0 0 0 0 0 199 502Total equity & liabilities 1,972 2,927 2,796 2,687 5,764 13,056 17,651 22,505EV 22,115 22,104 22,082 21,857 22,893 23,111 27,484 31,116Net debt/(cash) (20) (32) (54) (279) 757 975 5,149 8,478BVPS (HKD) 0.379 0.263 0.305 0.405 0.544 1.130 1.279 1.470

Year to 31 Dec 2010 2011 2012 2013 2014 2015E 2016E 2017ESales (YoY) n.a. 14.5 24.3 28.3 22.5 91.6 34.9 23.5EBITDA (YoY) n.a. (15.6) (42.5) 85.4 44.1 138.1 47.9 37.2Operating profit (YoY) n.a. (18.6) (63.4) 158.2 56.4 142.2 42.9 34.5Net profit (YoY) n.a. (21.3) (62.6) 153.5 62.3 142.2 28.9 28.9Core EPS (fully-diluted) (YoY) n.a. (21.3) (62.6) 113.9 14.8 126.1 29.4 28.9Gross-profit margin 51.5 35.9 17.2 30.1 31.6 38.1 43.3 46.4EBITDA margin 48.3 35.6 16.5 23.8 28.0 34.8 38.2 42.4Operating-profit margin 44.6 31.7 9.3 18.7 23.9 30.3 32.1 34.9Net profit margin 37.9 26.0 7.8 15.4 20.5 25.9 24.7 25.8ROAE n.a. 28.6 12.1 18.0 17.6 21.8 18.9 21.4ROAA n.a. 13.1 4.2 11.1 11.7 12.7 10.0 9.9ROCE n.a. 34.8 14.4 21.8 16.7 17.8 15.2 15.4ROIC 31.3 29.2 12.5 19.9 16.3 19.3 16.0 15.0Net debt to equity n.a. n.a. n.a. n.a. 22.9 12.8 59.7 85.5Effective tax rate 15.1 18.0 16.6 18.0 13.8 12.4 9.3 8.2Accounts receivable (days) n.a. 99.6 127.7 124.8 110.9 102.2 114.7 101.9Current ratio (x) 0.5 0.5 0.5 2.9 1.2 2.4 1.6 1.3Net interest cover (x) n.a. n.a. n.a. n.a. 113.5 40.3 24.7 13.8Net dividend payout 0.0 0.0 0.0 24.5 49.3 43.4 45.0 45.0Free cash flow yield n.a. n.a. n.a. 1.9 n.a. n.a. n.a. n.a.

Company profile

Xinyi Solar (XYS) manufactures and sells solar PV glass products. Its customers include downstream PV module manufacturers and other solar glass processing companies. As of 1H15, the company was the largest solar glass manufacturer globally by production capacity (3,800tpa). XYS also develops downstream solar farms in Central and Eastern China.


Hong Kong Utilities

Background: Backed by its innovative financing channels, as well as GCL Group’s extensive experience in solar farm development, GCLNE expects to significantly outpace its peers in terms of capacity expansion. Highlights: Innovative financing channels: Unlike its peers, which depend heavily on equity financing or CB issuance to fulfil the capex needs for solar farm development, GCLNE explores the use of wealth management products, asset-backed securities (ABS), and a yieldco, giving it access to a wider range of funding. The company has established a yieldco with Goldman Sachs Investment Holdings (ownership split 55% GCLNE, 45% others), and management is targeting a listing in 2016-17. Once this is done, it would potentially lower the cost of equity on a long-term horizon from over 10% currently to c.7-8%, enabling it to capture more opportunities that may be deemed unprofitable currently. As such, management expects GCLNE to achieve capacity additions of 2.0GW/2.5GW/3.0GW for 2015-17E. Plans to rely on ABS for financing short term: A yieldco listing is seemingly still some way off according to management, given 1) GCLNE’s installed solar capacity is still not big enough for a yieldco (we believe at least 1-2GW of solar projects are required to be injected), 2) China’s delayed subsidy payment remains unresolved, and 3) expected interest-rate hikes in the US would increase the required dividend yield and thus the cost of capital for yieldcos. Hence, GCLNE plans to issue various asset-backed securities (ABS) to fund its capex over the next 2-3 years. When the market allows, may turn to a yieldco listing: By implementing a yieldco listing strategy, GCLNE could enjoy some major benefits, according to management: 1) faster cash recycling, unleashing value, 2) reduced cost of capital, 3) deleveraging the parent company without diluting equity, and 4) capturing extra investment opportunities after the cost of capital is reduced. According to management, faster cash recycling could enhance a project’s value by c.20% in terms of NPV. Cost advantage over peers. GCLNE inherited solar farm development personnel from GCL-Poly (3800 HK, HKD1.66, Outperform [2]), which has a long track record in solar farm engineering, procurement and construction (EPC). GCLNE’s management believes its relationship with affiliate GCL System Integration Technology (GCLSIT; 002506 CH, not rated) is also beneficial in terms of the payment terms for its solar panel purchases. Valuation: The stock is trading currently at a 17.6x 2015E PER, according to the Bloomberg consensus.

29 October 2015

GCL N ew Energ y

Solar IPP growth engine for GCL-Poly

Aggressive capacity addition plan backed by innovative financing Depending on ABS short term; may turn to yieldco listing longer term Cost leadership due to relationship with GCL-Poly and GCLSIT


GCL New Energy (451 HK)

Target price: n.a.Share price (27 Oct): HKD0.58 | Up/downside: -

Dennis Ip, CFA(852) 2848 4068

[email protected]

Scott Chui(852) 2848 4443

[email protected]

30

50

70

90

110

0.3

0.6

0.8

1.1

1.4



GCL New En (LHS) Relative to HSI (RHS)

(HKD) (%)

12-month range 0.39-1.33Market cap (USDbn) 1.033m avg daily turnover (USDm) 1.16

148

GCL New Energy (451 HK): 29 October 2015

Financing channels – the key to success

Lack of financing channels the main obstacles in China Many solar farm developers in China are either integrated solar companies with both upstream and downstream assets, or pure solar companies with substantial leverage. The main drawback of these structures is funding difficulties, which may hinder their capacity expansion plans. For integrated solar companies which include both upstream and downstream operations, many Chinese commercial banks are unwilling to lend aggressively to them, given the widespread bankruptcies seen across the solar industry during 2011-12 (examples being Wuxi Suntech, LDK and Chaori). As such, these companies expand to a certain extent and after reaching certain gearing levels, they may not be able to borrow any more or only at very high interest costs. As a result, many module manufacturers, such as Trina Solar, Canadian Solar, and Jinko Solar, have relatively slow expansion plans after developing and owning several hundred megawatts of solar farms. For pure downstream solar farm developers, such as United PV and GCL New Energy, the main problem is high gearing level (which is usually above 300% net debt-to-equity ratio). With stretched balance sheets, their financing rates can be as high as 10%, substantially reducing their equity IRRs, as well as limiting their choice of project acquisitions given that many solar projects become value-destructive at such high interest costs and WACCs. Equity placements may not be adequate Solar farm development is usually very capital intensive. For example, a 100MW solar farm already needs around CNY800m in capex, or CNY160m equity if we assume an 80% debt/asset ratio. As such, we believe equity placements cannot fully satisfy solar farm development capital needs, considering that under the general mandate, companies are usually permitted to raise a maximum of 20% additional equity a year. Assuming a placement discount of ~5%, we estimate most solar farm developers are unable to, or can barely, meet their capacity expansion needs via equity placements, even taking into account the recent share-price rallies. By extension, exhausting the 20% general mandate may also pose negative implications for the company’s share price as well. Solar farm developers: maximum capital raised by equity placements vs. their capacity expansion plans

Market cap (USDm)

Market cap (CNYm)

Max fund raised (CNYm)

Implied capacity (MW)

2015 capacity plan (MW)

% surplus / (shortfall)

GCLNE 1,826 11,316 2,150 1,344 2,000 -33% United PV 734 4,514 858 536 1,000 -46% Shunfeng 2,354 14,569 2,768 1,730 1,500 15% XYS 2,471 15,318 2,910 1,819 1,000 82% Singyes 1,184 7,339 1,394 871 300 190%

Source: Daiwa research Hence, while equity placements can be a way to quickly replenish capital, alternative or more innovative capital financing ways would be needed to fulfil solar farm developers’ aggressive capacity expansion plans. So here’s the plan … In contrast to other solar farm developers in China, which usually finance their solar farm capex by equity placements and CB issuances, GCLNE has established a yieldco with Goldman Sachs Investment Holdings (ownership split 55% GCLNE, 45% others) for which management targets a separate listing in 2016-17. Once this is done, management expects the yieldco would give the company a long-term competitive advantage by: 1) unleashing the value of the IRR/cost of equity gap through faster cash

By listing the yieldco, GCLNE could recycle its cash much faster, lower its cost of equity to enhance value, and avoid serious share dilution due to its huge capex needs

149


recycling, 2) lowering the cost of equity, 3) allowing for minimum equity dilution, and 4) acting as a synthetic tax shelter. As shown below, GCLNE plans to separate the company by segment. Under these plans, the service business will be grouped under the management company, which mainly provides development and EPC services, as well as O&M services to its under-construction and completed solar farms, respectively. An asset holding company will also be set up, which will hold the solar farms under construction. Once the solar farms are completed, they will be sold to the yieldco, owned by GCLNE and other third-party investors (ie, Goldman Sachs Investment Holdings initially), with the ownership split initially set at 55:45.

GCLNE: financing structure

Source: Company

What is a “yieldco”? A yieldco is essentially a company created by renewable energy operators to hold long-term contracted operating power generation assets. It offers the characteristics of infrastructure investment, which are predictable and with growing dividends, to its shareholders. Usually a yieldco is established in the form of a dual-class share structure, with Class A issuing to external shareholders, whereas Class B remained held by the parent company for controlling voting interests. See the simplified illustration below.

GCLNE

Yield Co.(established in

2015)

Asset Holding Company

Management Company

Development services EPC services

SolarFarm A

SolarFarm B

SolarFarm C

SolarFarm D

SolarFarm E

SolarFarm N

3rd PartyInvestors

55%45%

Solar farms to be sold to Yield Co. once completed

Provides development and EPC services

Provides operation and maintenance services

O&M services

GCLNE plans to separate the company into management company, asset holding company, and the yieldco

150


Illustration of a yieldco structure

Source: Daiwa research The idea of a yieldco is actually nothing new, and was originated from the master limited partnerships (MLP) structure, a capital formation mechanism funding energy infrastructure investments in North America. The reason a renewable energy company cannot be set up under an MLP structure is because the criteria to be legally classified as an MLP calls for over 90% of its cash flow to come from real estate, natural resources and commodities. As such, a yieldco structure is also called a “synthetic MLP”. What are the benefits? The main purpose of creating a yieldco is to segregate the more risky assets, such as projects under construction or research and development units, from the more stable operating power generation units, in order to unlock the value of those low-risks assets. Compared with other financing channels, GCLNE believes that using a yieldco as a form of financing has multiple benefits for renewable developers, including the followings:

• Unleashing value through faster cash recycling. By disposing of its completed solar farms to the yieldco, a company essentially sell its minority stakes to the 3rd party investors at a premium (at an IRR above their cost of equity). As such, the company can capture the value enhancement once it sells the solar farm, which is much earlier than realising the value over the lifetime of the solar farm. By cashing out, the company can then use the recycled cash to invest in new solar farms, and again at a later point of time, dispose of the new solar farm to the yieldco at a premium.

• Lower cost of capital. Projects eligible to be disposed to a yieldco need to fulfil a few requirements: 1) have an operating track record of least 3 months, and 2) have long-term power purchase agreements (PPA) signed. If these requirements are fulfilled, the assets under a yieldco are deemed low-risk with predictable cash flow, which usually command a higher valuation than the original company with both risky and low-risk assets combined. Due to their high valuations, yieldcos in the US usually pay 4-6% dividends in the current market situation (refer to the table below), which is considered a relatively cheap source of equity funding for the parent company compared with other similar asset classes. Also, a growth premium is embedded in the valuation, resulting in a reduced average cost of capital for the parent company.

100% Class A UnitsX% Economic Interest

100% Class B Units100-X% Economic

Interest

Class B Common Stock100-X% Voting Interest0% Economic Interest

Class A Common StockX% Voting InterestX% Economic Interest

Parent company

Operatingsubsidiaries

Operating assets

Public shareholders

100%

YIELDCO

The yieldco structure resembles the master limited partnership (MLP) structure in North America

151


• Deleveraging the parent company without diluting equity. Downstream power generation operators often suffer from high gearing levels, which in turn limit the company’s ability to source cheap financing. The spin-off of the company’s assets in the form of yieldco effectively increases its equity base without diluting existing shareholders’ interests.

• Capturing extra investment opportunities. Given that solar farm developers usually suffer from a high cost of equity (usually above 10%), many solar farms are deemed not investable given solar farms’ general equity IRR range of 8-15%, depending on the region, tariff and subsidy. By lowering its cost of equity, many of the previously unattractive opportunities will become investable, and as such, in GCLNE’s case, management believes that it would be able to capture a much wider spectrum of solar farms than its peers.

GCLNE: cost of capital and amount of investable projects vs. peers

Source: Company • Tax shelters: Both MLPs and yieldcos enjoy a single layer of taxation, meaning that

profits are only taxed at the shareholder level and not at the corporate level. Whereas the MLP is legally exempted from taxation at the corporate level, the yieldco “synthesises” it. Given that the yieldco usually has a high depreciation versus its revenue during the beginning years, a yieldco is usually in a tax loss position initially. Combined with possible tax credits, accelerated depreciation, and continuous acquisitions, a yieldco can be tax-sheltered for many years.

According to management’s estimates, setting up a yieldco could enhance GCLNE’s NPV by 20%, assuming a 100% dividend payout and GCLNE selling a 50% stake in its solar farms to 3rd party investors when completed; or 58% if GCLNE only distributes 70% of its cash flow and reinvests the remaining 30% in new solar farms. GCLNE: value enhancement from a yieldco structure

Source: Company

7% 9% Solar farm IRRGCLNE’s

WACCNormal developers’ WACC

GCLNE’s Monopoly

Volume

0

50

100

150

200

Assuming holding 100% stake tomaturity

Assuming selling 51% stake forreinvestment

Assuming 70% dividend payout Value enhancement potential

NPV (capex = 100)

GCLNE’s management expects the yieldco to enhance GCLNE’s NPV

152


Opportunities only come to those who are prepared Despite all these benefits, not all solar companies may be qualified to spin off a yieldco, with the below differentiating factors. Capacity threshold should be met: Before any company can establish a yieldco for a separate listing, it must have enough operating assets such that the listed shares have sufficient liquidity to increase the valuation. According to Martin, K. in "Yield Cos Compared", a yieldco should hold at least USD500m project value for an IPO. Ample project pipeline: A yieldco structure has a snowball effect which can lead to accelerated capacity growth. However, we should note that, if the company does not have enough project pipelines, the growth rate could only be as high as its pipeline growth, significantly dampening the growth premium of the yieldco.

Overseas listed yieldcos Market cap Total capacity (MW) Dividend yield (%) Stock code Listing date (USDm) Type of assets (end-2014) Parent company 2014 2015E 2016E NRG Yield, Inc. NYLD US 7/17/2013 3,931 Solar/Wind/Thermal 2,985 NRG Energy Inc. 3.01 3.40 4.05 Pattern Energy Group, Inc. PEGI US 9/27/2013 2,023 Wind 1,636 Pattern Development 5.27 4.94 5.63 Abengoa Yield Plc. ABY US 6/13/2014 2,731 Solar/Thermal 2,246 Abengoa (ABGB US) - 4.87 6.28 TransAlta Renewables, Inc. RNW CN 8/9/2013 1,148 Hydro/Wind 1,376 TransAlta Corp. 5.59 6.12 6.17 NextEra Energy Partners, LP NEP US 6/27/2014 4,114 Solar/Wind 10,351 NextEra Energy (NEE US) 0.58 2.35 2.86 TerraForm Power, Inc. TERP US 7/18/2014 4,763 Solar/Wind 1,507 SunEdison (SUNE US) - 3.40 4.19 8point3 Energy Partners Pending listing - - Solar 432 First Solar / Sunpower - - -

Source: Bloomberg, compiled by Daiwa Quality assets: Only quality assets could be injected to the yieldco, given that a yieldco is usually characterised by having lower risk and predictable cash flows. As such, solar farms with low quality, such as substandard modules or serious grid curtailment issues may not be qualified. As such, we expect a gradual divergence of the performance of downstream developers in the future, with only the qualifying ones to experience accelerated growth. In contrast, the ones who fail to fulfil the above factors will likely be eventually weeded out by the industry. How would GCLNE fund its projects before a yieldco listing? A yieldco listing still seems some way off according to management, given: 1) GCLNE’s installed solar capacity is still not big enough for a yieldco (we believe at least 1-2GW of solar projects are required to be injected), 2) China’s delayed subsidy payment remains unresolved, and 3) an expected interest-rate hike in the US would increase the required dividend yield and thus the cost of capital for yieldcos. As a result, GCLNE plans to issue various ABS to fund its capex over the next 2-3 years. Management estimates that the required return for GCLNE’s ABS is ~10%, which is still lower than its solar projects’ 12-16% equity IRR, and therefore should provide the company with a residual return after interest payments are made to ABS investors. After 2-3 years of ABS funding, GCLNE said it may proceed with a yieldco listing should capital market conditions allow.

To spin off a yieldco, a company needs to have enough operating capacity, an ample project pipeline, and quality solar farm assets

GCLNE expects to issue asset-backed securities before spinning off the yieldco

153


Capacity expansion to drive earnings growth

Accelerated capacity expansion over 2015-17E At the end of 2014, GCLNE had solar capacity of 668MW, and management is guiding for the company to add 2.0GW/2.5 GW/3.0GW for 2015-17E, respectively, representing a 2014-17E capacity CAGR of 131%. GCLNE completed 158MW of solar farms in 1H15, with ~1GW capacity of solar farms already under construction. However, as we have discussed, a lack of financing channels will likely continue to create a bottleneck for solar farm development in China over the next few years, suggesting that companies coming up with more innovative financing measures, such as GCLNE, may continue to outpace peers in terms of capacity expansion. GCLNE’s project execution ability is backed by the fact that most of its project development personnel came from its parent, GCL-Poly (3800 HK, HKD1.66, Outperform [2]), which has a long history of solar farm development (the company commenced construction of its first utility-scale solar farm project in 2009). In just 4 months, GCLNE completed 668MW of solar farms in 2014 despite its first solar farm only commencing construction in late-August 2014.

Cost leadership – both in construction and operation GCL Group has a strong track record in solar development, which management believes may help GCLNE maintain its cost leadership in both the construction phase and development phase. According to GCLNE, its solar farm construction costs are as low as CNY8 per W, versus the average industry level of ~CNY8.5-9 per W, and average M&A cost of ~CNY10 per W. The core strengths of GCLNE’s cost leadership come from its relationships with GCL-Poly and GCLSIT, or the legacy Chaori Solar, which allow GCLNE to differentiate itself by having GCLSIT offer superior pricing, payment terms, and quality control, according to management. By utilising this cost advantage, management estimates that GCLNE could enhance a solar project’s equity IRR from 9.2% to 14.2% (or project IRR from 7.2% to 8.8%). Through its ability to enhance IRR, GCLNE believes it will be able to capture more solar projects than its peers. IRR comparison: GCLNE vs. industry average

Source: Company

GCLNE: project IRR comparison GCLNE: equity IRR comparison Unit-capex (CNY per W) 8 8.5 9 9.5 10

Oper

atin

g ex

pens

es

(CNY

per

W) 0.08 8.80% 7.90% 7.20% 6.50% 5.90%

0.085 8.70% 7.80% 7.10% 6.40% 5.80% 0.09 8.60% 7.70% 7.00% 6.30% 5.70%

0.095 8.50% 7.70% 6.90% 6.20% 5.60% 0.1 8.40% 7.60% 6.80% 6.10% 5.50%

Unit-capex (CNY per W) 8 8.5 9 9.5 10

Oper

atin

g ex

pens

es (

CNY

per W

) 0.08 14.20% 11.40% 9.10% 7.20% 5.60% 0.085 13.90% 11.10% 8.90% 7.00% 5.40%

0.09 13.50% 10.80% 8.60% 6.80% 5.20% 0.095 13.20% 10.60% 8.40% 6.50% 5.00%

0.1 12.90% 10.30% 8.10% 6.30% 4.80%

Source: Company Note: Orange highlight denotes what GCLNE can achieve; grey highlights show the industry average

Source: Company Note: Orange highlight denotes what GCLNE can achieve; grey highlights show the industry average

8.8%

14.2%

7.2%

9.2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

Project IRR Equity IRR

GCLNE industry average

GCLNE plans to accelerate capacity expansion in 2015-17E, mainly backed by relatively cheap financing from its yieldco, as well as its cost advantage relative to peers

GCLNE has lower solar farm construction costs than its peers, and therefore can achieve higher IRRs

154


Cost savings in the development phase GCLNE prefers to develop solar farms itself, instead of acquiring completed solar farms from 3rd parties, which according to the company can lower its investment costs by ~CNY1/W versus solar farms acquired via M&A. Due to the strong record of its development team (which mainly came from GCL-Poly), GCLNE can engage in engineering and procurement itself, despite still outsourcing the construction part to other EPC developers. According to management, this can save the company ~CNY1/W compared to other developers which might outsource the whole EPC process to outside parties. GCLNE: cost savings

Source: Company data, Daiwa research Purchase arrangement with GCLSIT preserves capex According to management, GCLSIT is not able to offer GCLNE deep discounts on GCLNE’s solar panel purchases, given both GCLNE and GCLSIT are listed companies controlled by different boards of directors, and unreasonable pricing would likely be rejected by GCLSIT’s directors.

However, one of the main advantages of GCLNE purchasing solar modules from GCLSIT is the payment terms, as GCLSIT’s urgency to sell more modules at reasonable prices means GCLNE should have strong bargaining power in terms of payment terms. As disclosed by GCLNE, for its connected transactions with GCLSIT, GCLNE only needs to make a 20% upfront payment to GCLSIT for its module purchases, with 70% of GCLNE’s payment to GCLSIT occurring only 6 months after the module delivery. This is opposed to the usual industry practice of around a 3-month payment period. The remaining 10% is paid 2 weeks after the 1-year warranty period. GCLNE: payment schedule for solar modules

Source: Company The prolonged payment period effectively allows GCLNE to preserve its cash, according to management, leading to better cash flow and financing conditions for the company.

6.0

7.0

8.0

9.0

10.0

Acquisition base casetotal investment cost

Self-developmentcost savings

EPC cost savings Total investment withcombined cost savings

(CNY/W)

10.00

8.00

-1.00

-1.00

Within 1 week of signing Module Sales

Agreement

20%0% 90% 100%Total Purchase Price

Within 6 months of solar modules delivery

Within 2 weeks following the expiration

of 1-year warranty period

GCLNE’s solar farm development team comes from GCL-Poly.

Although GCLSIT is unable to offer big discounts to GCLNE, its payment terms for GCLNE are favourable

155


Valuation The stock is trading at a 17.6x 2015E PER, based on Bloomberg consensus forecasts. Valuation charts

GCLNE: share price vs. 1-year forward PER bands GCLNE: 1-year forward PER bands

Source: Bloomberg Source: Bloomberg

GCLNE: share price vs. 1-year forward PBR bands GCLNE: 1-year forward PBR bands


GCLNE: share price performance vs. peers and indices

Source: Bloomberg

8

28

48

68

88

108

128

148

168

Jul-14 Oct-14 Jan-15 Apr-15 Jul-15 Oct-15

(x)

107.9x Avg+2SD

3.6x Avg-2SD

29.7x Avg-1SD

55.7x Avg

81.8x Avg+1SD

0.00.20.40.60.81.01.21.41.6


(HKD/sh)


0.0

1.0

2.0

3.0

4.0

5.0

6.0


(x)

4.6x Avg+2SD

0.8x Avg-2SD

1.7x Avg-1SD

2.7x Avg

3.6x Avg+1SD

0.00.20.40.60.81.01.21.41.6


(HKD/sh)

Price 1x PBR 2x PBR2.5x PBR 3x PBR 3.5x PBR

0.00.20.40.60.81.01.21.41.6

Apr-14 Jun-14 Aug-14 Oct-14 Dec-14 Feb-15 Apr-15 Jun-15 Aug-15 Oct-15

HKD per share

GCL New Energy Shunfeng Int'l United Photovoltaics MSCI China HSI Index

156


Downstream solar peers: 2016E PBR (x) vs. 2016E ROE (%) Downstream solar peers: 2016E PER (x) vs. 2014-17E EPS CAGR (%)


GCL New Energy

Xinyi Solar

United PV

Concord New Energy

Singyes Solar

NRG Yield

Pattern Energy Abengoa

TransAltaNextEra Energy

TerraForm Power 0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

- 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

2016

E RO

E (%

)

2016E PBR (x)

GCL New EnergyXinyi Solar

United PV

Concord

Singyes Solar

NRG Yield

Abengoa

TransAlta

NextEra Energy

0%

20%

40%

60%

80%

100%

120%

140%

0.0 10.0 20.0 30.0 40.0 50.0

2014

-17E

EPS

CAG

R

2016E PER (x)

157


Appendix I – company background Major downstream developers of GCL-Poly GCLNE is a downstream solar farm development platform for its parent company GCL-Poly. The company was established and listed on the Hong Kong Stock Exchange via a backdoor listing in May 2014, when the original company, Same Time Holdings Limited, issued 360m new ordinary shares to GCL-Poly and placed 50m shares to other placees at HKD4.0 per share. The company subsequently changed its name to GCL New Energy. After 2 share subdivisions and an equity placement in 2014, at the end of 2014 GCL-Poly remained the single largest shareholder of GCLNE, holding a 62.28% equity stake in the company. GCLNE: shareholding structures Shareholders Shares % of TSO GCL-Poly 8,640,000,000 62.28% Yip Sum Yin, Former Chairman and Executive Director 847,055,352 6.11% CHUNG Chi Shing, Former Executive Director 839,252,568 6.05% Public Shareholders 3,545,486,128 25.56% Total 13,871,794,048 100.00%

Source: Company data

GCLNE: company history since listing

Source: Company, Daiwa research

Rapid solar farm development since its establishment In terms of operations, GCLNE started construction of its first solar farm, a 100MW utility-scale solar farm in Jinhu, Jiangsu Province, in late August 2014 (see exhibits below). Since then, the company has developed its solar farms rapidly, with projects located in various provinces in China. In just 4 months, from late August to end-2014, the company achieved an installed capacity of 668MW. In terms of geographical distribution, at the end of 2014, the company had completed and under-construction solar farm projects mainly in Jiangsu (21%), Inner Mongolia (19%), Qinghai (12%), and Shaanxi (12%). According to management, GCLNE has good risk control over its solar farm development, and aims to avoid regions with existing and potential grid curtailment issues.

9 May 2014 22 Aug 26 Aug 16 Oct 12 Nov 29 Dec 31 Dec 2015

Became a listed subsidiary of GCL-Poly and raised a total of HKD1,640mn gross proceeds

• Previously known as Same Time Holdings Ltd, Company was in the business of manufacturing and selling Printed Circuit Boards

• Company then issued 360mn new ordinary shares on 9 May 2014 to GCL-Poly at HKD4.00/share

• Company also placed an additional 50mn new ordinary shares at HKD4.00/share on the same date to other placees

Operation Service Agreement

• Company signed an operation service agreement with GCL-Poly on 22 Aug 2014 to provide operation and management services to its solar power plants with installed capacity of 353MW

Top-up placement for 291mn primary shares

• Company issued 291mn new ordinary shares on 16 Oct 2014 via a top-up placement at HKD2.55/share

• The HKD742mn gross proceeds have since been used for investments in solar projects and for general working capital purposes

Secured USD 80mn revolving facility from BOCI

• Company secured USD 80mn revolving facility from BOCI Leveraged and Structured Finance Ltd.

First jointly-developed project• Company started construction

of its first jointly-developed solar farm, Jiangsu Jinhu(100MW) on 26 Aug 2014

First grid-connection• Company grid-connected its

first solar farm Zhejiang Xiaoshan (17.5MW) on 12 Nov 2014

Capacity today• Company plans to have

2,600MW of total operational installed capacity by 31 Dec 2015

GCLNE was listed on HKEx through backdoor listing in May 2014

GCLNE has developed 668MW of solar farms from August 2014 to end-2014

158


GCLNE: solar farm geographical distribution

Source: Company GCLNE: solar projects by province as of January 2015

Source: Company data Note: solar projects including grid-connected and under-construction capacity GCLNE: geographical distribution vs. wind curtailment Province Geographical proportion Wind curtailment (2014) Jiangsu 21% 0% Inner Mongolia 19% 9% Qinghai 12% 0% Shaanxi 12% 2% Xinjiang 9% 15% Shanxi 9% 0% Sichuan 4% 0% Yunnan 4% 4% Hainan 4% 0% Hebei 4% 12% Others 2% 8%

Source: Company, NEA

Jiangsu21.0%

Inner Mongolia19.0%

Qinghai12.0%

Shaanxi12.0%

Xinjiang9.0%

Shanxi9.0%

Sichuan4.0%

Yunnan4.0%

Hainan4.0%

Hebei4.0%

Others2.0%

159


Aqua Solar Project: Jiangsu Jinhu (100MW) Desert Solar Project: Ningxia Zhongwei (100MW)


GCLNE: management profile Name Title Background ZHU Gongshan Honorary Chairman Founder of Golden Concord Holdings and GCL-Poly

Committee member of the 12th National Committee CPPCC Chairman of Asian PV Industry Association, Co-chairman of China PV Industry Alliance More than 25 years in IPP industry

TANG Cheng Chairman Former president of Golden Concord Former executive director and executive vice president of large-scale state-owned IPP More than 20 years management experience in the IPP industry Skilled in industry insights and strategy management

ZHANG Guoxin President Former director and general manager of Guodain Nanjing Automation Co., Ltd. Former executive manager of China Water Property Group Ltd. More than 30 years’ experience in the IPP industry Skilled in IPP investment and operation

HU Xiaoyan Executive Director Vice president of Golden Concord Vice president of GCL-Poly More than 20 years management experience in large-scale enterprises Skilled in financial management and risk control for large-scale enterprises

YIP Sum Yin Executive Director Co-founder and Chairman of Same Time Holding Ltd. More than 40 years’ experience in electronics industry

Source: Company

160


Appendix II – IRR model (industry) Industry: IRR assumptions Assumptions Capacity (MW) 50 Operation (Years) 20 Unit-capex (including VAT) (CNY/W) 8.5 Unit-capex (excluding VAT) (CNY/W) 7.3 Total investments (including VAT) (CNY mn) 425 Total investments (excluding VAT) (CNY mn) 363 Tariff (including VAT) (CNY/kWh) 1 Tariff (excluding VAT) (CNY/kWh) 0.85 Insolation (hours) 1,500 System efficiency (%) 78% Utilization hours (hours) 1,170 Degradation (1st year) (%) 2.00% Degradation (2nd year and after) (%) 0.80% Debt (%) 80% Debt repayment years (years) 15 Finance cost (%) 7.00% Initial operating expenses (CNY/W) 0.1 Inflation (%) 3%


Industry: IRR analysis

Year

0 Year

1 Year

2 Year

3 Year

4 Year

5 Year

6 Year

7 Year

8 Year

9 Year

10 Year

11 Year

12 Year

13 Year

14 Year

15 Year

19 Year

20 Year

21 Year

22 Year

23 Year

24 Year

25 Year

26 Year

27 Year

28 Year

29 Year

30 Income statement Net electricity generated (mn kWh)

57 57 56 56 56 55 55 54 54 53 53 52 52 52 51 50 49 0 0 0 0 0 0 0 0 0 0


57 57 56 56 56 55 55 54 54 53 53 52 52 52 51 50 49 0 0 0 0 0 0 0 0 0 0 Net revenue (excl. VAT)

49 49 48 48 47 47 47 46 46 46 45 45 44 44 44 42 42 0 0 0 0 0 0 0 0 0 0

Depreciation

(18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) 0 0 0 0 0 0 0 0 0 0 Operating expenses

(5) (5) (5) (5) (6) (6) (6) (6) (6) (7) (7) (7) (7) (7) (8) (9) (9) 0 0 0 0 0 0 0 0 0 0

VAT rebate

8 8 8 8 8 8 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Government subsidies

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

EBIT

34 34 33 32 32 31 26 22 21 21 20 20 19 19 18 16 15 0 0 0 0 0 0 0 0 0 0 EBIT margin (%)

70% 69% 68% 68% 67% 66% 56% 48% 47% 46% 45% 44% 43% 42% 41% 37% 36%

Interest expense

(24) (23) (22) (21) (20) (18) (17) (16) (14) (12) (11) (9) (7) (5) (2) 0 0 0 0 0 0 0 0 0 0 0 0 Profit before tax

10 11 11 12 12 13 9 6 7 8 10 11 12 14 16 16 15 0 0 0 0 0 0 0 0 0 0

Income tax

0 0 0 (1) (2) (2) (2) (2) (2) (2) (2) (3) (3) (3) (4) (4) (4) 0 0 0 0 0 0 0 0 0 0 Net profit

10 11 11 10 11 11 7 5 6 6 7 8 9 10 12 12 11 0 0 0 0 0 0 0 0 0 0

Net margin (%)

21% 22% 23% 21% 22% 24% 15% 10% 12% 14% 16% 18% 21% 24% 27% 28% 27% IRR calculations

Profit before tax

10 11 11 10 11 11 7 5 6 6 7 8 9 10 12 12 11 0 0 0 0 0 0 0 0 0 0 Depreciation

18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 0 0 0 0 0 0 0 0 0 0

Interest expense

24 23 22 18 17 16 13 12 11 9 8 7 5 4 2 0 0 0 0 0 0 0 0 0 0 0 0 Capex (425)

FCFF (425) 52 52 51 46 46 45 38 35 34 34 33 33 33 32 32 30 30 0 0 0 0 0 0 0 0 0 0 Interest expense

(24) (23) (22) (18) (17) (16) (13) (12) (11) (9) (8) (7) (5) (4) (2) 0 0 0 0 0 0 0 0 0 0 0 0

Net repayment in debt

(14) (14) (15) (17) (18) (19) (20) (22) (23) (25) (27) (28) (30) (33) (35) 0 0 0 0 0 0 0 0 0 0 0 0 New debt 340

FCFE (85) 15 14 14 12 11 10 5 1 0 (0) (1) (2) (3) (4) (5) 30 30 0 0 0 0 0 0 0 0 0 0 Project IRR 7.1%

Equity IRR 9.7% Source: Daiwa research

161


Daiwa’s Asia Pacific Research Directory HONG KONG Takashi FUJIKURA (852) 2848 4051 [email protected] Regional Research Head Kosuke MIZUNO (852) 2848 4949 /

(852) 2773 8273 [email protected]

Regional Research Co-head John HETHERINGTON (852) 2773 8787 [email protected]

Regional Deputy Head of Asia Pacific Research Rohan DALZIELL (852) 2848 4938 [email protected]

Regional Head of Product Management Kevin LAI (852) 2848 4926 [email protected]

Chief Economist for Asia ex-Japan; Macro Economics (Regional) Christie CHIEN (852) 2848 4482 [email protected]

Macro Economics (Regional); Banking; Insurance (Taiwan) Junjie TANG (852) 2773 8736 [email protected]

Macro Economics (China) Jonas KAN (852) 2848 4439 [email protected]

Head of Hong Kong and China Property Cynthia CHAN (852) 2773 8243 [email protected]

Property (China) Leon QI (852) 2532 4381 [email protected]

Banking (Hong Kong/China); Broker (China); Insurance (China) Anson CHAN (852) 2532 4350 [email protected]

Consumer (Hong Kong/China) Jamie SOO (852) 2773 8529 [email protected]

Gaming and Leisure (Hong Kong/China) Dennis IP (852) 2848 4068 [email protected]

Power; Utilities; Renewables and Environment (Hong Kong/China) John CHOI (852) 2773 8730 [email protected]

Head of Hong Kong and China Internet; Regional Head of Small/Mid Cap Kelvin LAU (852) 2848 4467 [email protected]

Head of Automobiles; Transportation and Industrial (Hong Kong/China) Brian LAM (852) 2532 4341 [email protected]

Transportation – Railway; Construction and Engineering (China) Jibo MA (852) 2848 4489 [email protected]

Head of Custom Products Group Thomas HO (852) 2773 8716 [email protected]

Custom Products Group PHILIPPINES Bianca SOLEMA (63) 2 737 3023 [email protected]

Utilities and Energy

SOUTH KOREA Sung Yop CHUNG (82) 2 787 9157 [email protected]

Pan-Asia Co-head/Regional Head of Automobiles and Components; Automobiles; Shipbuilding; Steel Mike OH (82) 2 787 9179 [email protected]

Banking; Capital Goods (Construction and Machinery) Iris PARK (82) 2 787 9165 [email protected]

Consumer/Retail SK KIM (82) 2 787 9173 [email protected]

IT/Electronics – Semiconductor/Display and Tech Hardware Thomas Y KWON (82) 2 787 9181 [email protected]

Pan-Asia Head of Internet & Telecommunications; Software – Internet/On-line Game TAIWAN Rick HSU (886) 2 8758 6261 [email protected]

Head of Regional Technology; Head of Taiwan Research; Semiconductor/IC Design (Regional) Steven TSENG (886) 2 8758 6252 [email protected]

IT/Technology Hardware (PC Hardware) Christine WANG (886) 2 8758 6249 [email protected]

IT/Technology Hardware (Automation); Pharmaceuticals and Healthcare; Consumer Kylie HUANG (886) 2 8758 6248 [email protected]

IT/Technology Hardware (Handsets and Components) Helen CHIEN (886) 2 8758 6254 [email protected]

Small/Mid Cap INDIA Punit SRIVASTAVA (91) 22 6622 1013 [email protected]

Head of India Research; Strategy; Banking/Finance Saurabh MEHTA (91) 22 6622 1009 [email protected]

Capital Goods; Utilities SINGAPORE Ramakrishna MARUVADA (65) 6499 6543 [email protected]

Head of Singapore Research; Telecommunications (China/ASEAN/India) Royston TAN (65) 6321 3086 [email protected]

Oil and Gas; Capital Goods David LUM (65) 6329 2102 [email protected]

Property and REITs Shane GOH (65) 64996546 [email protected]

Small/Mid Cap (Singapore) Jame OSMAN (65) 6321 3092 [email protected]

Telecommunications (ASEAN/India); Pharmaceuticals and Healthcare; Consumer (Singapore)

162


Daiwa’s Offices

Office / Branch / Affiliate Address Tel Fax

DAIWA SECURITIES GROUP INC

HEAD OFFICE Gran Tokyo North Tower, 1-9-1, Marunouchi, Chiyoda-ku, Tokyo, 100-6753 (81) 3 5555 3111 (81) 3 5555 0661

Daiwa Securities Trust Company One Evertrust Plaza, Jersey City, NJ 07302, U.S.A. (1) 201 333 7300 (1) 201 333 7726

Daiwa Securities Trust and Banking (Europe) PLC (Head Office) 5 King William Street, London EC4N 7JB, United Kingdom (44) 207 320 8000 (44) 207 410 0129

Daiwa Europe Trustees (Ireland) Ltd Level 3, Block 5, Harcourt Centre, Harcourt Road, Dublin 2, Ireland (353) 1 603 9900 (353) 1 478 3469

Daiwa Capital Markets America Inc. New York Head Office Financial Square, 32 Old Slip, New York, NY10005, U.S.A. (1) 212 612 7000 (1) 212 612 7100

Daiwa Capital Markets America Inc. San Francisco Branch 555 California Street, Suite 3360, San Francisco, CA 94104, U.S.A. (1) 415 955 8100 (1) 415 956 1935

Daiwa Capital Markets Europe Limited, London Head Office 5 King William Street, London EC4N 7AX, United Kingdom (44) 20 7597 8000 (44) 20 7597 8600

Daiwa Capital Markets Europe Limited, Frankfurt Branch Neue Mainzer Str. 1, 60311 Frankfurt/Main, Germany (49) 69 717 080 (49) 69 723 340

Daiwa Capital Markets Europe Limited, Paris Representative Office 17, rue de Surène 75008 Paris, France (33) 1 56 262 200 (33) 1 47 550 808

Daiwa Capital Markets Europe Limited, Geneva Branch 50 rue du Rhône, P.O.Box 3198, 1211 Geneva 3, Switzerland (41) 22 818 7400 (41) 22 818 7441

Daiwa Capital Markets Europe Limited, Moscow Representative Office

Midland Plaza 7th Floor, 10 Arbat Street, Moscow 119002, Russian Federation

(7) 495 641 3416 (7) 495 775 6238

Daiwa Capital Markets Europe Limited, Bahrain Branch 7th Floor, The Tower, Bahrain Commercial Complex, P.O. Box 30069, Manama, Bahrain

(973) 17 534 452 (973) 17 535 113

Daiwa Capital Markets Hong Kong Limited Level 28, One Pacific Place, 88 Queensway, Hong Kong (852) 2525 0121 (852) 2845 1621

Daiwa Capital Markets Singapore Limited 6 Shenton Way #26-08, OUE Downtown 2, Singapore 068809, Republic of Singapore

(65) 6220 3666 (65) 6223 6198

Daiwa Capital Markets Australia Limited Level 34, Rialto North Tower, 525 Collins Street, Melbourne, Victoria 3000, Australia

(61) 3 9916 1300 (61) 3 9916 1330

DBP-Daiwa Capital Markets Philippines, Inc 18th Floor, Citibank Tower, 8741 Paseo de Roxas, Salcedo Village, Makati City, Republic of the Philippines

(632) 813 7344 (632) 848 0105

Daiwa-Cathay Capital Markets Co Ltd 14/F, 200, Keelung Road, Sec 1, Taipei, Taiwan, R.O.C. (886) 2 2723 9698 (886) 2 2345 3638

Daiwa Securities Capital Markets Korea Co., Ltd. 20 Fl.& 21Fl. One IFC, 10 Gukjegeumyung-Ro, Yeongdeungpo-gu, Seoul, Korea

(82) 2 787 9100 (82) 2 787 9191

Daiwa Securities Co. Ltd., Beijing Representative Office Room 301/302，Kerry Center，1 Guanghua Road，Chaoyang District， Beijing 100020, People’s Republic of China

(86) 10 6500 6688 (86) 10 6500 3594

Daiwa (Shanghai) Corporate Strategic Advisory Co. Ltd. 44/F, Hang Seng Bank Tower, 1000 Lujiazui Ring Road, Pudong, Shanghai China 200120 , People’s Republic of China

(86) 21 3858 2000 (86) 21 3858 2111

Daiwa Securities Co. Ltd., Bangkok Representative Office 18th Floor, M Thai Tower, All Seasons Place, 87 Wireless Road, Lumpini, Pathumwan, Bangkok 10330, Thailand

(66) 2 252 5650 (66) 2 252 5665

Daiwa Capital Markets India Private Ltd 10th Floor, 3 North Avenue, Maker Maxity, Bandra Kurla Complex, Bandra East, Mumbai – 400051, India

(91) 22 6622 1000 (91) 22 6622 1019

Daiwa Securities Co. Ltd., Hanoi Representative Office Suite 405, Pacific Palace Building, 83B, Ly Thuong Kiet Street, Hoan Kiem Dist. Hanoi, Vietnam

(84) 4 3946 0460 (84) 4 3946 0461

DAIWA INSTITUTE OF RESEARCH LTD

HEAD OFFICE 15-6, Fuyuki, Koto-ku, Tokyo, 135-8460, Japan (81) 3 5620 5100 (81) 3 5620 5603

MARUNOUCHI OFFICE Gran Tokyo North Tower, 1-9-1, Marunouchi, Chiyoda-ku, Tokyo, 100-6756 (81) 3 5555 7011 (81) 3 5202 2021

New York Research Center 11th Floor, Financial Square, 32 Old Slip, NY, NY 10005-3504, U.S.A. (1) 212 612 6100 (1) 212 612 8417

London Research Centre 3/F, 5 King William Street, London, EC4N 7AX, United Kingdom (44) 207 597 8000 (44) 207 597 8550

163


Important Disclosures and Disclaimer

This publication is produced by Daiwa Securities Group Inc. and/or its non-U.S. affiliates, and distributed by Daiwa Securities Group Inc. and/or its non-U.S. affiliates, except to the extent expressly provided herein. This publication and the contents hereof are intended for information purposes only, and may be subject to change without further notice. Any use, disclosure, distribution, dissemination, copying, printing or reliance on this publication for any other purpose without our prior consent or approval is strictly prohibited. Neither Daiwa Securities Group Inc. nor any of its respective parent, holding, subsidiaries or affiliates, nor any of its respective directors, officers, servants and employees, represent nor warrant the accuracy or completeness of the information contained herein or as to the existence of other facts which might be significant, and will not accept any responsibility or liability whatsoever for any use of or reliance upon this publication or any of the contents hereof. Neither this publication, nor any content hereof, constitute, or are to be construed as, an offer or solicitation of an offer to buy or sell any of the securities or investments mentioned herein in any country or jurisdiction nor, unless expressly provided, any recommendation or investment opinion or advice. Any view, recommendation, opinion or advice expressed in this publication may not necessarily reflect those of Daiwa Securities Group Inc., and/or its affiliates nor any of its respective directors, officers, servants and employees except where the publication states otherwise. This research report is not to be relied upon by any person in making any investment decision or otherwise advising with respect to, or dealing in, the securities mentioned, as it does not take into account the specific investment objectives, financial situation and particular needs of any person. Daiwa Securities Group Inc., its subsidiaries or affiliates, or its or their respective directors, officers and employees from time to time have trades as principals, or have positions in, or have other interests in the securities of the company under research including market making activities, derivatives in respect of such securities or may have also performed investment banking and other services for the issuer of such securities. The following are additional disclosures. Ownership of Securities For “Ownership of Securities” information, please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. Investment Banking Relationship For “Investment Banking Relationship”, please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. Japan Daiwa Securities Co. Ltd. and Daiwa Securities Group Inc. Daiwa Securities Co. Ltd. is a subsidiary of Daiwa Securities Group Inc. Investment Banking Relationship Within the preceding 12 months, the subsidiaries and/or affiliates of Daiwa Securities Group Inc. * has lead-managed public offerings and/or secondary offerings (excluding straight bonds) of the securities of the following companies: Modern Land (China) Co. Ltd (1107 HK); econtext Asia Ltd (1390 HK); Accordia Golf Trust (AGT SP); Hua Hong Semiconductor Ltd (1347 HK); GF Securities Co Ltd (1776 HK); Mirae Asset Life Insurance Co Ltd (085620 KS). *Subsidiaries of Daiwa Securities Group Inc. for the purposes of this section shall mean any one or more of: Daiwa Capital Markets Hong Kong Limited (大和資本市場香港有限公司), Daiwa Capital Markets Singapore Limited, Daiwa Capital Markets Australia Limited, Daiwa Capital Markets India Private Limited, Daiwa-Cathay Capital Markets Co., Ltd., Daiwa Securities Capital Markets Korea Co., Ltd. Hong Kong This research is distributed in Hong Kong by Daiwa Capital Markets Hong Kong Limited (大和資本市場香港有限公司) (“DHK”) which is regulated by the Hong Kong Securities and Futures Commission. Recipients of this research in Hong Kong may contact DHK in respect of any matter arising from or in connection with this research. Relevant Relationship (DHK) DHK may from time to time have an individual employed by or associated with it serves as an officer of any of the companies under its research coverage.

Singapore This research is distributed in Singapore by Daiwa Capital Markets Singapore Limited and it may only be distributed in Singapore to accredited investors, expert investors and institutional investors as defined in the Financial Advisers Regulations and the Securities and Futures Act (Chapter 289), as amended from time to time. By virtue of distribution to these category of investors, Daiwa Capital Markets Singapore Limited and its representatives are not required to comply with Section 36 of the Financial Advisers Act (Chapter 110) (Section 36 relates to disclosure of Daiwa Capital Markets Singapore Limited’s interest and/or its representative’s interest in securities). Recipients of this research in Singapore may contact Daiwa Capital Markets Singapore Limited in respect of any matter arising from or in connection with the research. Australia This research is distributed in Australia by Daiwa Capital Markets Australia Limited and it may only be distributed in Australia to wholesale investors within the meaning of the Corporations Act. Recipients of this research in Australia may contact Daiwa Capital Markets Stockbroking Limited in respect of any matter arising from or in connection with the research. India This research is distributed in India to Institutional Clients only by Daiwa Capital Markets India Private Limited (Daiwa India) which is an intermediary registered with Securities & Exchange Board of India as a Stock Broker, Merchant Bank and Research Analyst. Daiwa India, its Research Analyst and their family members and its associates do not have any financial interest save as disclosed or other undisclosed material conflict of interest in the securities or derivatives of any companies under coverage. Daiwa India and its associates may have received compensation for any products other than Investment Banking (as disclosed) or brokerage services from the subject company in this report during the past 12 months. Unless otherwise stated in BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action, Daiwa India and its associates do not hold more than 1% of any companies covered in this research report. There is no material disciplinary action against Daiwa India by any regulatory authority impacting equity research analysis activities as of the date of this report. Taiwan This research is distributed in Taiwan by Daiwa-Cathay Capital Markets Co., Ltd and it may only be distributed in Taiwan to institutional investors or specific investors who have signed recommendation contracts with Daiwa-Cathay Capital Markets Co., Ltd in accordance with the Operational Regulations Governing Securities Firms Recommending Trades in Securities to Customers. Recipients of this research in Taiwan may contact Daiwa-Cathay Capital Markets Co., Ltd in respect of any matter arising from or in connection with the research. Philippines This research is distributed in the Philippines by DBP-Daiwa Capital Markets Philippines, Inc. which is regulated by the Philippines Securities and Exchange Commission and the Philippines Stock Exchange, Inc. Recipients of this research in the Philippines may contact DBP-Daiwa Capital Markets Philippines, Inc. in respect of any matter arising from or in connection with the research. DBP-Daiwa Capital Markets Philippines, Inc. recommends that investors independently assess, with a professional advisor, the specific financial risks as well as the legal, regulatory, tax, accounting, and other consequences of a proposed transaction. DBP-Daiwa Capital Markets Philippines, Inc. may have positions or may be materially interested in the securities in any of the markets mentioned in the publication or may have performed other services for the issuers of such securities. For relevant securities and trading rules please visit SEC and PSE links at http://www.sec.gov.ph/irr/AmendedIRRfinalversion.pdf and http://www.pse.com.ph/ respectively. Thailand This research is distributed to only institutional investors in Thailand primarily by Thanachart Securities Public Company Limited (“TNS”). This report is prepared by analysts who are employed by Daiwa Securities Group Inc. and/or its non-U.S. affiliates. This report is provided to you for informational purposes only and it is not, and is not to be construed as, an offer or an invitation to make an offer to sell or buy any securities. Neither Thanachart Securities Public Company Limited, Daiwa Securities Group Inc. nor any of their respective parent, holding, subsidiaries or affiliates, nor any of their respective directors, officers, servants and employees accept any liability whatsoever for any direct or consequential loss arising from any use of this research or its contents. The information and opinions contained herein have been compiled or arrived at from sources believed to be reliable. However, Thanachart Securities Public Company Limited, Daiwa Securities Group Inc. nor any of their respective parent, holding, subsidiaries or affiliates, nor any of their respective directors, officers, servants and employees make no representation or warranty, express or implied, as to their accuracy or completeness. Expressions of opinion herein are subject to change without notice. The use of any information, forecasts and opinions contained in this report shall be at the sole discretion and risk of the user. Daiwa Securities Group Inc. and/or its non-U.S. affiliates perform and seek to perform business with companies covered in this research. Thanachart Securities Public Company Limited, Daiwa Securities Group Inc., their respective parent, holding, subsidiaries or affiliates, their respective directors, officers, servants and employees may have positions and financial interest in securities mentioned in this research. Thanachart Securities Public Company Limited, Daiwa Securities Group Inc., their respective parent, holding, subsidiaries or affiliates may from time to time perform investment banking or other services for, or solicit investment banking or other business from, any entity mentioned in this research. Therefore, investors should be aware of conflict of interest that may affect the objectivity of this research. United Kingdom This research report is produced by Daiwa Securities Co. Ltd. and/or its affiliates and is distributed in the European Union, Iceland, Liechtenstein, Norway and Switzerland. Daiwa Capital Markets Europe Limited is authorised and regulated by The Financial Conduct Authority (“FCA”) and is a member of the London Stock Exchange and Eurex. This publication is intended for investors who are not Retail Clients in the United Kingdom within the meaning of the Rules of the FCA and should not therefore be distributed to such Retail Clients in the United Kingdom. Should you enter into investment business with Daiwa Capital Markets Europe’s affiliates outside the United Kingdom, we are obliged to advise that the protection afforded by the United Kingdom regulatory system may not apply; in particular, the benefits of the Financial Services Compensation Scheme may not be available.

https://daiwa3.bluematrix.com/sellside/Disclosures.action


http://www.sec.gov.ph/irr/AmendedIRRfinalversion.pdf%20and%20http:/www.pse.com.ph/

164


Daiwa Capital Markets Europe Limited has in place organisational arrangements for the prevention and avoidance of conflicts of interest. Our conflict management policy is available at http://www.uk.daiwacm.com/about-us/corporate-governance-regulatory. Germany This document is distributed in Germany by Daiwa Capital Markets Europe Limited, Niederlassung Frankfurt which is regulated by BaFin (Bundesanstalt fuer Finanzdienstleistungsaufsicht) for the conduct of business in Germany. Bahrain This research material is distributed in Bahrain by Daiwa Capital Markets Europe Limited, Bahrain Branch, regulated by The Central Bank of Bahrain and holds Investment Business Firm – Category 2 license and having its official place of business at the Bahrain World Trade Centre, South Tower, 7th floor, P.O. Box 30069, Manama, Kingdom of Bahrain. Tel No. +973 17534452 Fax No. +973 535113 United States This report is distributed in the U.S. by Daiwa Capital Markets America Inc. (DCMA). It may not be accurate or complete and should not be relied upon as such. It reflects the preparer’s views at the time of its preparation, but may not reflect events occurring after its preparation; nor does it reflect DCMA’s views at any time. Neither DCMA nor the preparer has any obligation to update this report or to continue to prepare research on this subject. This report is not an offer to sell or the solicitation of any offer to buy securities. Unless this report says otherwise, any recommendation it makes is risky and appropriate only for sophisticated speculative investors able to incur significant losses. Readers should consult their financial advisors to determine whether any such recommendation is consistent with their own investment objectives, financial situation and needs. This report does not recommend to U.S. recipients the use of any of DCMA’s non-U.S. affiliates to effect trades in any security and is not supplied with any understanding that U.S. recipients of this report will direct commission business to such non-U.S. entities. Unless applicable law permits otherwise, non-U.S. customers wishing to effect a transaction in any securities referenced in this material should contact a Daiwa entity in their local jurisdiction. Most countries throughout the world have their own laws regulating the types of securities and other investment products which may be offered to their residents, as well as a process for doing so. As a result, the securities discussed in this report may not be eligible for sales in some jurisdictions. Customers wishing to obtain further information about this report should contact DCMA: Daiwa Capital Markets America Inc., Financial Square, 32 Old Slip, New York, New York 10005 (Tel no. 212-612-7000). Ownership of Securities For “Ownership of Securities” information please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. Investment Banking Relationships For “Investment Banking Relationships” please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. DCMA Market Making For “DCMA Market Making” please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. Research Analyst Conflicts For updates on “Research Analyst Conflicts” please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. The principal research analysts who prepared this report have no financial interest in securities of the issuers covered in the report, are not (nor are any members of their household) an officer, director or advisory board member of the issuer(s) covered in the report, and are not aware of any material relevant conflict of interest involving the analyst or DCMA, and did not receive any compensation from the issuer during the past 12 months except as noted: no exceptions. Research Analyst Certification For updates on “Research Analyst Certification” and “Rating System” please visit BlueMatrix disclosure link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. The views about any and all of the subject securities and issuers expressed in this Research Report accurately reflect the personal views of the research analyst(s) primarily responsible for this report (or the views of the firm producing the report if no individual analysts[s] is named on the report); and no part of the compensation of such analyst(s) (or no part of the compensation of the firm if no individual analyst[s)] is named on the report) was, is, or will be directly or indirectly related to the specific recommendations or views contained in this Research Report.

The following explains the rating system in the report as compared to relevant local indices, unless otherwise stated, based on the beliefs of the author of the report. "1": the security could outperform the local index by more than 15% over the next 12 months. "2": the security is expected to outperform the local index by 5-15% over the next 12 months. "3": the security is expected to perform within 5% of the local index (better or worse) over the next 12 months. "4": the security is expected to underperform the local index by 5-15% over the next 12 months. "5": the security could underperform the local index by more than 15% over the next 12 months. Disclosure of investment ratings

Rating Percentage of total Buy* 63.8% Hold** 22.2% Sell*** 14.0%

Source: Daiwa Notes: data is for single-branded Daiwa research in Asia (ex Japan) and correct as of 30 September 2015. * comprised of Daiwa’s Buy and Outperform ratings. ** comprised of Daiwa’s Hold ratings. *** comprised of Daiwa’s Underperform and Sell ratings. Additional information may be available upon request. Japan - additional notification items pursuant to Article 37 of the Financial Instruments and Exchange Law (This Notification is only applicable where report is distributed by Daiwa Securities Co. Ltd.) If you decide to enter into a business arrangement with us based on the information described in materials presented along with this document, we ask you to pay close attention to the following items. • In addition to the purchase price of a financial instrument, we will collect a trading commission* for each transaction as agreed beforehand with you. Since commissions may be included in

the purchase price or may not be charged for certain transactions, we recommend that you confirm the commission for each transaction. • In some cases, we may also charge a maximum of ¥ 2 million (including tax) per year as a standing proxy fee for our deposit of your securities, if you are a non-resident of Japan. • For derivative and margin transactions etc., we may require collateral or margin requirements in accordance with an agreement made beforehand with you. Ordinarily in such cases, the

amount of the transaction will be in excess of the required collateral or margin requirements. • There is a risk that you will incur losses on your transactions due to changes in the market price of financial instruments based on fluctuations in interest rates, exchange rates, stock prices,

real estate prices, commodity prices, and others. In addition, depending on the content of the transaction, the loss could exceed the amount of the collateral or margin requirements. • There may be a difference between bid price etc. and ask price etc. of OTC derivatives handled by us. • Before engaging in any trading, please thoroughly confirm accounting and tax treatments regarding your trading in financial instruments with such experts as certified public accountants.

*The amount of the trading commission cannot be stated here in advance because it will be determined between our company and you based on current market conditions and the content of each transaction etc.

When making an actual transaction, please be sure to carefully read the materials presented to you prior to the execution of agreement, and to take responsibility for your own decisions regarding the signing of the agreement with us. Corporate Name: Daiwa Securities Co. Ltd. Financial instruments firm: chief of Kanto Local Finance Bureau (Kin-sho) No.108 Memberships: Japan Securities Dealers Association, The Financial Futures Association of Japan Japan Securities Investment Advisers Association Type II Financial Instruments Firms Association

http://www.uk.daiwacm.com/about-us/corporate-governance-regulatory






China Solar Sector

Documents

Transcript of China Solar Sector