China Wind and Solar

China Utilities 11 July 2019

Prefer upstream product makers to downstream project operators

See important disclosures, including any required research certifications, beginning on page 105

China Wind and Solar: 11 July 2019

Table of contents

2019-21 summary: prefer renewable upstream product makers to downstream project operators ...... 5 Grid and retail parity in sight ...... 10 Wind equipment: to benefit from rush installations in 2019-21 ...... 17 Wind IPP: worsening cash flows in 2019-21E ...... 21 Solar product: be selective among segments ...... 25 Solar farm operators: distressed cash flow with liquidity risk ...... 54 Stock calls ...... 59 Risks to our Neutral sector view ...... 61

Company Section Xinjiang Goldwind Science & Technology ...... 63 China Longyuan Power ...... 67 Huaneng Renewables ...... 71 China Datang Corp Renewable Power ...... 75 China Resources Power ...... 79 Huaneng Power International ...... 83 Huadian Fuxin Energy ...... 87 China Suntien Green Energy ...... 91 GCL-Poly Energy ...... 95 Xinyi Solar Holdings ...... 99

China Utilities 11 July 2019

China Wind and Solar

Prefer upstream product makers to downstream project operators

 We expect wind and solar product demand to be supported by robust installations over 2019-21E Dennis Ip, CFA  Downstream project operators continue to be impacted by worsening (852) 2848 4068 cash flow resulting from delays in collecting subsidies [email protected] Prev. Xinjiang Goldwind ScienceAnna &Lu, Technology CFA (2208 HK)  We prefer the manufacturers of upstream wind and solar products Rating (852) Buy2848 4465 Buy (positive) to downstream project operators (negative) Target [email protected] 11.90 Upside p 36.9% China Longyuan Power (916 HK) Rating Hold Outperform What's new: We see 2019-21 as a special period prior to grid and retail KeyTarget stock calls 5.40 5.75 parity for China’s Wind and Solar Sector. In our view, robust wind and solar Upside p 4.7%New Prev. installations over 2019-21 will support upstream product demand. TongweiXinjiang Goldwind(600438 CH) Science & Technology (2208 HK) Meanwhile, downstream project operators face worsening cash flows due Rating OutperformBuy OutperformBuy to subsidy delays. Thus, we prefer upstream wind and solar product Target 15.2011.90 15.2011.90 Upside p 14.4%36.9% makers over downstream project operators, and suggest being further LONGiChina Longyuan Green Energy Power Technology (916 HK) (601012 CH) selective by choosing those better positioned in terms of demand/supply of Rating OutperformHold Outperform sub-segments, as well as company-level competitive advantages. Target 24.505.40 25.505.75 Upside p 13.1%4.7% What's the impact: As approved wind projects with fixed tariffs and those XinyiTongwei Solar (600438 Holdings CH) (968 HK) Rating UnderperformOutperform UnderperformOutperform gained via competitive bidding should start operation before end-2020/21E, Target 15.203.50 15.203.80 we expect rush installations in the wind segment over 2019-21E (we DownsideUpside qp 14.4%6.2%

forecast 25-33GW vs. 21GW in 2018). With its solar policy finalised in Source:LONGi Daiwa Green forecasts Energy Technology (601012 CH) 2Q19, we expect China’s solar installations to kick off in 2H19, with likely Rating Outperform Outperform stable full-year demand over 2019-20 (40-42GW) before recovering in 2021 Target 24.50 25.50 (50GW on retail parity). With increasing overseas solar installations, we Upside p 13.1% Xinyi Solar Holdings (968 HK) expect global solar installations to rise from 107GW in 2018 to 125-150GW Rating Underperform Underperform over 2019-21E. Meanwhile, with a large deficit in the Renewable Energy Target 3.50 3.80 (RE) Fund, subsidy collection delays will likely persist in 2019-21, weighing Downside q 6.2% on project operators’ cash flows. We prefer upstream makers to downstream operators, and are positive on wind turbine generator (WTG) makers in the wind sector. For solar, we are selective with the following criteria: 1) supply-demand conditions (prefer polysilicon and solar glass), 2) competitive landscape (prefer leaders in more consolidated segments like wafer, solar glass and polysilicon), 3) low downstream solar project exposure, 4) technology shift to high-efficiency products (mono wafers and PERC cells), 5) healthy balance sheets, and 6) undemanding valuations.

What we recommend: We downgrade China Wind to Neutral from Positive, and upgrade China Solar to Neutral from Negative. For Wind, we are positive on Goldwind (2208 HK, HKD8.69, Buy [1]), a leading WTG maker vs. other wind IPPs, and downgrade Longyuan (916 HK, HKD5.16) and DTR (1798 HK, HKD0.78) to Hold (3). In Solar, we downgrade GCL- Poly (3800 HK, HKD0.43) to Hold (3) on short-term refinancing risk, despite its appealing valuation. We have an Underperform (4) on Xinyi Solar (968 HK, HKD3.73) due to its rich valuation. Upside/downside sector risks: higher-/ lower-than-expected installations; better-/worse-than-expected subsidy payment.

How we differ: We hold different views on different sub-segments in the renewable sector and prefer upstream makers to downstream operators.

See important disclosures, including any required research certifications, beginning on page 105

China Wind and Solar: 11 July 2019

How do we justify our view? Growth outlook Valuation Earnings revisions

Growth outlook China Wind and Solar: revenue, EBIT, and net profit growth We expect Goldwind to see a 13% net profit CAGR in 2018-21E CAGR 2018-21E 2018-21E, with FCF yield expansion, mainly led by rising Recurring Yield gap FCF yield Company Ticker Rating Revenue EBIT net profit change change WTG sales. While wind IPPs should see net profit CAGRs Goldwind 2208 HK Buy 17% 12% 13% 1.5pp 22.3pp of 5-19% in 2019-21E amid capacity rises, we expect their Longyuan 916 HK Hold 4% 10% 10% 1.5pp -15.9pp cash flows to worsen, with declining FCF yields. While HNR 958 HK Hold 11% 13% 10% 1.5pp -44.8pp DTR 1798 HK Hold 10% 13% 13% 2.6pp -135.4pp GCL-Poly’s net profit may turn positive in 2019E, we are Huadian Fuxin 816 HK Buy 6% 19% 19% 5.6pp -61.0pp concerned about its short-term refinancing risk. XYS Suntien 956 HK Buy 20% 16% 5% 1.6pp -5.1pp should see a 19% net profit CAGR in 2018-21E, but with a GCL-Poly 3800 HK Hold -5% 13% n.a. 0.8pp 96.1pp Xinyi Solar 968 HK U/P 22% 22% 19% 2.8pp 6.0pp limited FCF boost given capex on capacity expansion and CRP 836 HK Hold -2% 12% 11% 3.2pp -10.7pp solar farm exposure. CRP faces worsening FCF given HNP-H 902 HK Buy 2% 37% 53% 5.7pp -22.1pp aggressive plans to add wind capacity, in our view. Source: Daiwa forecasts

Valuation China Wind and Solar: PBR vs. OCF/Invested capital China’s Wind and Solar Sector is trading at a wide PBR 2020E OCF/Invested capital 25% range of 0.3-4.0x for 2019E. The wind IPPs are trading at Tongwei 0.4-0.7x 2019E PBRs, c.1SD below their averages since 20% listing, likely due to cash-flow concerns from subsidy delays and rising capex. Goldwind is trading at 1.0x 15% LONGi Longyuan 2019E PBR, 0.7SD below its historical mean, likely due to HDFX 10% HNR Goldwind market concerns over a GPM decline in 2019, though we CRP XYS GCL-Poly Suntien expect a pick-up in 2020. As for solar, Tongwei and LONGi HNP-A 5% HNP-H are trading at higher 3.0x and 4.0x 2019E PBRs, likely on DTR rapid market-share gains. XYS is trading at a 2.2x 2019E 0% 0.0 1.0 2.0 3.0 4.0 PBR, 0.5SD above its past-3-year mean and higher than 2019E PBR peers’. GCL-Poly is only at a 0.3x 2019E PBR amid short- Source: Bloomberg, Daiwa estimates term refinancing risks. CRP and HNP-H are trading at 0.7x 2019E PBR, 1-1.7SD below their historical averages.

Earnings revisions China Wind and Solar: Bloomberg 2019E EPS consensus trend The market has cut its 2019E EPS for most wind IPPs by (Rebased at 100) 6-11% YTD, with robust wind capacity installation 130 120 guidance offset by limited utilisation growth and likely 110 100 higher expenses. The market’s forecast of Goldwind’s 90 EPS is down 13% YTD, on lower GPM guidance for 2019; 80 70 but this looks set to recover in 2020 on rising ASPs. XYS’s 60 50 consensus EPS has been stable YTD, likely with higher 40 solar glass ASPs offset by a share placement and higher 30 Jan-19 Feb-19 Mar-19 Apr-19 May-19 Jun-19 Jul-19 minority interests. GCL-Poly’s consensus EPS is down Goldwind Longyuan HNR 61% YTD after a net loss in 2018. CRP’s consensus DTR HDFX Suntien Tongwei LONGi GCL-Poly 2019E EPS is down 6% since CRP released disappointing XYS CRP HNP

2018 results in March. HNP’s consensus 2019E EPS Source: Bloomberg ticked up since May 2019 after it reported a strong 1Q19.

China Wind and Solar: 11 July 2019

EPS (local curr.) Sector stocks: key ShareindicatorsRating Target price (local curr.) FY1 FY2 Company Name Stock code Price New Prev. New Prev. % chg New Prev. EPS% chg (local curr.)New Prev. % chg China Datang Corp Renewable Power 1798 HK 0.78 Hold Buy 0.80 1.75 (54.3%) 0.185 0.217 (14.7%) 0.205 0.242 (15.1%) Share Rating Target price (local curr.) FY1 EPS (local curr.) FY2 CompanyChina Longyuan Name Power Stock916 HK code SharePrice5.16 NewHoldRating OutperformPrev. TargetNew5.40 price (localPrev.5.75 curr.)%(6.1%) chg 0.538New FY10.572Prev. %(5.9%) chg 0.582New FY20.629Prev. %(7.5%) chg CompanyChina DatangResources Name Corp Power Renewable Power Stock1798836 HK HK code 11.30Price0.78 NewHold Prev.HoldBuy 11.50New0.80 11.50Prev.1.75 (54.3%)%0.0% chg 0.1851.468New 0.2171.701Prev. (14.7%)(13.7%)% chg 0.2051.861New 0.2422.104Prev. (15.1%)(11.6%)% chg China DatangLongyuanSuntien CorpGreen Power Renewable Energy Power 1798916956 HK HK 0.785.162.14 HoldBuy OutperformBuy 0.805.403.50 1.755.753.50 (54.3%)(6.1%)0.0% 0.1850.5380.406 0.2170.5720.406 (14.7%)(5.9%)0.0% 0.2050.5820.432 0.2420.6290.435 (15.1%)(7.5%)(0.6%) ChinaGCL-Poly LongyuanResources Energy Power Power 9168363800 HK HK 11.305.160.43 Hold OutperformHold 11.505.400.45 11.505.750.52 (13.5%)(6.1%)0.0% 0.5381.4680.015 0.5721.7010.018 (13.7%)(17.7%)(5.9%) 0.5821.8610.039 0.6292.1040.046 (11.6%)(15.5%)(7.5%) ChinaHuadian ResourcesSuntien Fuxin GreenEnergy Power Energy 836956816 HK 11.302.141.43 HoldBuy HoldBuy 11.503.502.40 11.503.502.40 0.0% 1.4680.4060.332 1.7010.4060.365 (13.7%)(9.1%)0.0% 1.8610.4320.407 2.1040.4350.423 (11.6%)(0.6%)(3.9%) ChinaGCL-PolyHuaneng Suntien Power Energy Green International Energy 9563800902 HK HK 2.140.434.70 HoldBuy OutperformBuy 3.500.455.95 3.500.525.95 (13.5%)0.0% 0.4060.0150.299 0.4060.0180.239 (17.7%)25.0%0.0% 0.4320.0390.369 0.4350.0460.314 (15.5%)(0.6%)17.5% GCL-PolyHuadianHuaneng Fuxin Power Energy Energy International-A 3800816600011 HK HK CH 0.431.436.28 UnderperformHoldBuy UnderperformOutperformBuy 0.452.405.80 0.522.406.00 (13.5%)(3.3%)0.0% 0.0150.3320.299 0.0180.3650.239 (17.7%)(9.1%)25.0% 0.0390.4070.369 0.0460.4230.314 (15.5%)(3.9%)17.5% HuadianHuaneng Fuxin PowerRenewables Energy International 816902958 HK 1.434.702.22 HoldBuy HoldBuy 2.405.952.30 2.405.952.45 (6.1%)0.0% 0.3320.2990.358 0.3650.2390.366 (9.1%)(2.3%)25.0% 0.4070.3690.363 0.4230.3140.385 (3.9%)(5.5%)17.5% HuanengLONGi Green Power Energy InternationalInternational-A Technology 902600011601012 HK CH 21.674.706.28 UnderperformOutperformBuy UnderperformOutperformBuy 24.505.955.80 25.505.956.00 (3.3%)(3.9%)0.0% 0.2991.027 0.2391.033 (0.6%)25.0% 0.3691.298 0.3141.298 17.5%0.0% HuanengTongwei PowerRenewables International-A 600011958600438 HK CH 13.296.282.22 UnderperformOutperformHold UnderperformOutperformHold 15.205.802.30 15.206.002.45 (3.3%)(6.1%)0.0% 0.2990.3580.879 0.2390.3660.879 (2.3%)25.0%0.0% 0.3690.3631.136 0.3140.3851.136 (5.5%)17.5%0.0% HuanengXinjiangLONGi Green Goldwind Renewables Energy Science Technology & Technology 9582208601012 HK HK CH 21.672.228.69 OutperformHoldBuy OutperformHoldBuy 11.9024.502.30 11.9025.502.45 (6.1%)(3.9%)0.0% 0.3580.7671.027 0.3660.7581.033 (2.3%)(0.6%)1.2% 0.3631.0881.298 0.3851.0191.298 (5.5%)6.7%0.0% LONGiXinyiTongwei Solar Green Holdings Energy Technology 601012968600438 HK CH 21.6713.293.73 UnderperformOutperform UnderperformOutperform 24.5015.203.50 25.5015.203.80 (3.9%)(7.9%)0.0% 1.0270.2820.879 1.0330.2890.879 (0.6%)(2.6%)0.0% 1.2980.3321.136 1.2980.3391.136 (2.0%)0.0% Source:TongweiXinjiang Bloomberg, Goldwind Science Daiwa &forecasts Technology 6004382208 HK CH 13.298.69 OutperformBuy OutperformBuy 15.2011.90 15.2011.90 0.0% 0.8790.767 0.8790.758 0.0%1.2% 1.1361.088 1.1361.019 0.0%6.7% XinjiangXinyi Solar Goldwind Holdings Science & Technology 2208968 HK HK 8.693.73 UnderperformBuy UnderperformBuy 11.903.50 11.903.80 (7.9%)0.0% 0.7670.282 0.7580.289 (2.6%)1.2% 1.0880.332 1.0190.339 (2.0%)6.7% China:Xinyi Solar wind Holdings capacity installations968 (GW) HK 3.73 Underperform Underperform Global:3.50 solar capacity3.80 (7.9%) installations0.282 0.289 (MW)(2.6%) 0.332 0.339 (2.0%) 33 33 200,000 60% 35 80% 55% 29 30 60% 25 150,000 40% 25 40% 21 30% 19 24% 20 18 20% 100,000 20% 20% 16 16% 12% 10% 15 0% 9% 9% 50,000 0% 10 (20%) -7%

5 (40%) 0 (20%) 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E 0 (60%) China US Japan 2014 2015 2016 2017 2018 2019E 2020E 2021E India Europe Rest of the world Newly connected capacity YoY YoY growth (RHS)

Source: NEA, Daiwa estimates Source: BP, NEA, BNEF, Daiwa

China Wind and Solar Sector: key calls Company Stock code Rating Rationale Goldwind 2208 HK Buy (1) Beneficiary of China wind installation upcycle in 2019-21E Longyuan 916 HK Hold (3) Distressed balance sheet impacted by rush installation of wind farms; slow wind power generation growth likely in 9M19; upcoming RPS/GC policy unlikely to slow subsidy receivables growth HNR 958 HK Hold (3) Balance sheet may remain tight with significant rise in finance costs; monthly data and 2H19 results unlikely to be positive catalysts; upcoming RPS/GC not likely to improve subsidy collection DTR 1798 HK Hold (3) Surging finance cost on rush installation despite utilisation growth on curtailment improvement Huadian Fuxin 816 HK Buy (1) Continuous recovery in hydro power from 1Q19; coal-fired power to also benefit from a YoY downtrend in coal prices; distressed 0.4x PBR Suntien 956 HK Buy (1) Attractive valuation and yield; set for further rerating on A-share IPO GCL-Poly 3800 HK Hold (3) Short-term refinancing risk is still high with a large amount of debt due in 2019, despite distressed valuation Xinyi Solar 968 HK Underperform (4) Valuation seems rich compared with its peers, however solar glass may see recovery in 2019E while XYE spin-off has been completed CRP 836 HK Hold (3) Coal-fired power utilisation YTD is slightly lower than expected; aggressively adding wind capacity, which would result in worsening cash flow HNP-H 902 HK Buy (1) 1H19 earnings benefitted from YoY coal price drop; large capex with aggressive renewable capacity addition in 2019-21E Source: Daiwa

China Wind and Solar: other calls China Wind and Solar: long/short pair trade Small-cap high growth Long/short pair trade Rationale Suntien Buy (1) 0.6x 2019E PBR, 2018-21E reported net profit CAGR of 11% from Long Goldwind, Short Benefiting from a wind installation upcycle over 2019-21E wind/gas segment HNR

Fuxin Buy (1) 0.4x 2019E PBR, 19% net profit CAGR in 2018-21E Counter consensus XYS U/P (4) Rich valuation compared with its peer HNR Hold (3) High gearing and further large capex for rush installation of wind farms

Source: Daiwa Source: Daiwa

China Wind and Solar: 11 July 2019

Please also see: China Power Sector: China Power and Coal: China Power and Coal: Power struggle: all fighting a losing Initiation: power struggle II – all eyes Power struggle III – keeps chasing battle? on coal prices lower costs 3 November 2016 29 December 2017 28 January 2019

Dennis Ip, CFA (852) 2848 4068 Dennis Ip, CFA (852) 2848 4068 Dennis Ip, CFA (852) 2848 4068 ([email protected]) ([email protected]) ([email protected]) Anna Lu, CFA (852) 2848 4465 ([email protected])

China Solar: Downgrading: China Coal-Fired Power: Why ROEs China Wind: Offshore Wind: a new downstream thirsty for cash inflows; still have a mountain to climb market for growth upstream drowning in oversupply 10 August 2017 19 July 2017 7 June 2017

Dennis Ip, CFA (852) 2848 4068 Dennis Ip, CFA (852) 2848 4068 Dennis Ip, CFA (852) 2848 4068 ([email protected]) ([email protected]) ([email protected])

China Wind and Solar: 11 July 2019

2019-21 summary: prefer renewable upstream product makers to downstream project operators

China’s Wind and Solar Sector is approaching the era of grid and retail parity, which marks a milestone for the 2 sub-segments and should see the sector achieve cost competitiveness and eliminate its reliance on subsidies. We regard the next 2-3 years (2019-21) as a special period before grid and retail parity, in which our preference differs between upstream product makers and downstream project operators. In our view, wind and solar installations will remain robust over 2019-21, giving support to upstream product demand (wind: 25-33GW in China; solar: 40-50GW in China and 125-150GW globally).

Meanwhile, downstream project operators are still suffering from worsening cash flows resulting from delays in subsidy collection, with the vast RE Fund deficit (c.CNY285bn, on our estimates) unlikely to be resolved in the near term. In this regard, compared with downstream project operators, we prefer the manufacturers of upstream wind and solar products, where we suggest being further selective by choosing those in a better position in terms of technology competitiveness, supply-demand situation, and market share.

Renewable product demand supported by robust installations Wind: rush installations in China over 2019-21 to support WTG demand We expect rush According to government guidelines, competitive bidding will not affect pipeline projects installations in China to that have already been approved on or before 2018 and are entitled to fixed tariffs if they support WTG demand are connected on-grid before 2020. Large wind IPPs still have sufficient approved project over 2019-21 pipelines. According to BNEF, there are still 88GW of approved projects that are unaffected by the competitive bidding policy. Thus, we expect wind IPPs to mainly focus on developing approved project reserves with fixed tariffs over 2019-20E. Also, project approvals gained through competitive bidding over 2019-20 should be connected on-grid before end-2021. Thus, we think there will be a rush of installations, and forecast new installations in China to rise from 21GW in 2018 to 25GW in 2019, 33GW in 2020, and 29GW in 2021.

China Wind: tariffs for new onshore wind projects Approved Regions Approved before 2018 Approved in 2018 Approved in 2019 Approved in 2020 in 2021 Zone IV 0.60 (CNY/kWh) 0.57 (CNY/kWh) 0.52 (CNY/kWh) 0.47 (CNY/kWh) Zone III 0.54 (CNY/kWh) 0.49 (CNY/kWh) 0.43 (CNY/kWh) 0.38 (CNY/kWh) Zone II 0.50 (CNY/kWh) 0.45 (CNY/kWh) 0.39 (CNY/kWh) 0.34 (CNY/kWh) No subsidy anymore Zone I 0.47 (CNY/kWh) 0.40 (CNY/kWh) 0.34 (CNY/kWh) 0.29 (CNY/kWh) Note Start construction by end-2019, and Complete grid connection by end- Need to complete grid connection by end- Need to complete grid connection by end-2021, complete grid connection by end- 2020, otherwise no subsidy 2021, otherwise no subsidy. otherwise no subsidy. 2020, otherwise no subsidy Tariff is the cap of the competitive bidding Tariff is the cap of the competitive bidding Source: NDRC, Daiwa

China: annual installations for new wind farms (GW) 35 33 33 80% 29 30 60% 25 25 40% 21 19 20 18 20% 16 15 0%

10 (20%)

5 (40%)

0 (60%) 2014 2015 2016 2017 2018 2019E 2020E 2021E Newly connected capacity YoY

Source: NEA, Daiwa forecasts

China Wind and Solar: 11 July 2019

Solar: product demand supported by both China and overseas demand According to the latest government policy, utility-scale and commercial and industrial distributed generation (C&I DG) solar projects will go to auction in 2019 under a total subsidy amount of CNY2.25bn, and the NEA approved a total of 23GW in competitive projects on 11 July 2019, including 18GW in utility-scale projects and 5GW in C&I DG solar projects. The National Energy Administration (NEA) also allocated CNY750m for 3.5-4GW in subsidised household-rooftop DG project construction. In addition, there will be a 2-3GW poverty-alleviation project that will not need to go through auction, a 2GW top-runner project expected to be installed in mid-2019, and 3-5GW of subsidy-free projects that are set to be installed during 2019. We see potentially 36-40GW of solar capacity to be installed in 2019, in line with the NEA’s guidance for a similar installation scale to that of 2018 (44GW).

2019 China solar installation estimate (GW) Utility-scale and C&I DG projects 23 Household-rooftop DG project 3.5-4 Poverty-alleviation project 2-3 Top-runner project 2 Subsidy-free project 3-5 Others 2-3 Total 36-40 Source: Daiwa

We see a potential surge Based on our previous estimate, we think DG projects across Zone I-III will reach retail in DG installations in parity and cease relying on the government subsidies in 2021, while utility-scale projects in 2021 with the coming of Zones II-III would likely reach grid parity later than 2021. We expect utility-scale solar retail parity project installations to be largely stable over 2019-21, as they are still largely subject to the amount of government subsidy. Besides, according to a notice released by the NEA on 9 January 2019, subsidy-free projects can be approved by provincial-level governments and no longer require NEA quotas. Thus, DG installations may see slight increase in 2019-20, with more areas reaching retail parity and a potential surge in DG installations in 2021 with the coming of retail parity in most regions in China.

Overall, on our estimates, solar installations will be relatively stable at 40GW for 2019 and 42GW for 2020, compared with 44GW in 2018, before surging to 50GW in 2021E.

China: annual solar installations China: solar installations 60 140% 35 26-30 120% 50 30 100% 24.3 20 25 40 80% 19.7 30 20 4 21 60% 30 21 23 40% 15 c.10 20 20% 34 10 1 30 23 0% 10 2 19 19 20 5 14 (20%) 9 0 (40%) 0 2014 2015 2016 2017 2018 2019E 2020E 2021E 1H 2H Utility-scale DG YoY growth (RHS) 2018 2019

Source: NEA, Daiwa estimates and forecasts Source: NEA, Daiwa estimates and forecasts

With the finalisation of China’s solar policy, we expect China’s solar installation demand to kick off in 2H19. Full-year demand is likely to remain stable over 2019-20 before a recovery in 2021. In the meantime, the overseas market is steadily increasing solar installations. Thus, we forecast global solar installations to increase from 107GW for 2018 to 125GW for 2019, 136GW for 2020, and 150GW for 2021E, representing growth of 16% YoY, 9% YoY and 10% YoY for the respective years, and a CAGR of 12% over 2018-21E.

China Wind and Solar: 11 July 2019

Global: solar annual installations (MW) 160,000 60% 140,000 50% 120,000 40% 100,000 30% 80,000 20% 60,000 10% 40,000 0% 20,000 (10%) 0 (20%) 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E China US Japan India Europe Rest of the world YoY growth (RHS)

Source: NEA, BP, BNEF, Daiwa estimates

Large RE Fund deficit unlikely to be solved near term Continued subsidy As the final RPS policy states that Green Certificates (GC) are only subject to voluntary collection delay would trading, we expect subsidy collections going forward still to largely rely on the RE Fund, continue to put pressure which has a large deficit currently (of over CNY200bn, on our estimates). This means that on the cash flows and subsidy collection delays are likely to continue, putting pressure on the cash flows and balance sheets of wind balance sheets of wind and solar project operators. and solar project operators We think the situation will worsen going forward, because: 1) the wind capacity ramp-up on rush installations puts a greater burden on the RE Fund and increasing subsidy receivables coupled with rising cash outflow for capex for the wind IPPs, and 2) the NEA may reallocate CNY3bn for the competitive bidding of new solar power capacities in 2019 and pay subsidies to new solar projects ahead of existing wind and solar projects starting from this year.

RE Fund movement (CNYbn) CNYbn 700

600 28 24 74 119 84 114 500 31 51 84 47 400 105 27 79 85 300 89 24 497 497 72 423 464 464 200 66 71 21 285 43 65 65 100 175 55 91 0 36 Shortfall Subsidy Fund Shortfall Subsidy Fund Shortfall Subsidy Fund Shortfall Subsidy Fund Shortfall Subsidy Fund Shortfall Subsidy Fund Shortfall payable inflow payable inflow payable inflow payable inflow payable inflow payable inflow 2015 2016 2017 2018 2019E 2020E 2021E

Total fund shortfall Wind Solar Others Fund inflow

Source: MoF, NEA, Daiwa Note 1: assuming GC price of CNY0.1/kWh starting from 2020, leading to lower new subsidy payable in 2020E Note 2: we think the RE Fund has not received most of the settlement of the unpaid RE surcharges of previous years from captive coal-fired power plants, according to the fund inflow data released by MoF

China Wind and Solar: 11 July 2019

Accounts receivable days for wind farm operators China solar segment: receivables days 300 700 600 250 500 200 400 150 300

100 200 100 50 0 0 GCLNE Xinyi United Kong Longi Tongwei ShunfengGCL-Poly Singyes Longyuan HNR DTR PV Sun 1H14 2H14 1H15 2H15 1H16 2H16 1H17 2H17 1H18 2H18 2016 2017 2018

Source: Company, Daiwa Source: Company, Daiwa

Prefer renewable upstream equipment makers over downstream operators We prefer the Based on the above analysis, we prefer the manufacturers of upstream wind and solar manufacturers of products to downstream project operators. We recommend that investors be further upstream wind and solar selective by choosing those companies in better positions in terms of their technology products to downstream edge, supply-demand situation, and market share. project operators For the Wind sector, we are positive on Goldwind, the leading WTG manufacturer with the largest market share (32%) in China, which we expect to benefit the most from rush installations over 2019-21E.

For the Solar sector, we suggest being selective with the following criteria: 1) supply- demand situation (we prefer polysilicon and solar glass in this respect), 2) competitive landscape (we prefer companies with relatively high market shares in more consolidated segments, including wafer, solar glass and polysilicon), 3) low downstream solar project exposure, 4) technology shift to higher-efficiency products (the market is shifting to mono wafers and PERC cells), 5) healthy balance sheets, and 6) undemanding valuations. We have a Hold (3) rating on GCL-Poly on its short-term refinancing risk, despite its cheap valuations and its plan to dispose of its solar farm subsidiary. And we have an Underperform (4) rating on Xinyi Solar on its rich valuation (trading at 2.2x 2019E PBR) and some downstream exposure, despite it being the market leader in the solar glass segment.

Capacity vs. installation growth Solar glass supply and demand vs. global solar installation growth 30% 20% 25% 15% 20% 10% 15% 5% 10% 0% 5% 2019E 2020E 2021E

0% Solar glass supply growth Solar galss demand growth 2019E 2020E 2021E Polysilicon Wafer Cell Module Global solar installation Global solar installation growth

Source: BNEF, Daiwa estimates Source: BNEF, company, Daiwa estimates

China Wind and Solar: 11 July 2019

Mono wafer demand vs. global solar installation growth PERC cell demand vs. global solar installation growth 45% 90% 40% 80% 35% 70% 30% 60% 25% 50% 20% 40% 15% 30% 10% 20% 5% 10% 0% 0% 2019E 2020E 2021E 2019E 2020E 2021E Mono wafer demand growth Global installation growth PERC cell demand growth Global installation growth

Source: BNEF, Daiwa Source: BNEF, Daiwa

China: comparison of solar companies Supply-demand Competitive Healthy balance Undemanding Company situation landscape Low downstream exposure Technology shift sheet valuation Tongwei      - LONGi      - GCL-Poly   - -   (depends on whether GNE disposal is (depends on whether quasi mono is successful) successful) Xinyi Solar   - -  

Source: Daiwa

China Wind and Solar: 11 July 2019

Grid and retail parity in sight

Cost parity is on the development roadmap for renewable energy, like wind and solar power, in the next few years. The renewable energy industry believes wind and solar projects will be as competitive as conventional power sources and therefore not reliant on government subsidies, and are likely to see surging installation demand once they reach cost parity. In general, there are 2 stages of cost parity: 1) retail parity – where the levelised cost of electricity (LCOE) of renewable energy projects reaches a level similar to the electricity tariff to end users, and 2) grid parity – where the LCOE reaches a level similar to the on-grid tariff of conventional power projects.

In our view, major factors in the cost parity of renewable energy include: 1. Cost: installation and operation cost of the project 2. Power generation: utilisation and equipment efficiency 3. Coal-fired and retail tariffs: vary by province and fluctuate according to coal prices and progress on power reform

Where are we now? Wind: likely to achieve grid parity in 2023, on our estimates We expect wind projects We expect the LCOE for wind power to increase in 2020 from the levels in 2019 and to reach grid parity in decline marginally in 2021, before seeing a sharper decline after 2021. The major 2023 assumptions in our forecasts include: 1) a 4% decline in wind project construction costs in 2019 (as WTGs at lower than previous year’s bidding price in 2018 are delivered in 2019), but a 5% rebound in 2020 and relatively stable unit capex in 2021 against the background of rush installations during 2019-21E; followed by a significant drop of 15-20% for 2022- 23E and then a slower 10-15% pa decline over 2024-25E, and 2) a decline in the grid curtailment rate in China from an average of 7% for 2018 to 5% for 2020E and thereafter.

In this scenario, we estimate the wind LCOE to fall to CNY0.23-0.32/kWh by 2023E for Zones I-IV, down 32-41% in the next 5 years (2019-23) compared with the levels in 2018. On our estimates, the LCOE will be the same or lower than the coal-fired benchmark tariff by 2023, meaning that wind projects in China should be able to reach grid parity in 2023.

China wind: LCOE forecast (CNY/kWh) 0.50 0.460.45 0.46 0.45 0.45 0.41 0.41 0.40 0.39 0.38 0.39 0.39 0.38 0.40 0.37 0.37 0.340.34 0.33 0.35 0.33 0.32 0.32 0.28 0.27 0.27 0.30 0.26 0.25 0.23 0.23 0.24 0.25 0.20 0.21 0.21 0.18 0.20 0.15 0.10 0.05 0.00 Zone I Zone II Zone III Zone IV 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E

Source: Daiwa forecasts

China Wind and Solar: 11 July 2019

China wind: cost parity outlook (CNY/kWh) 0.50

0.45

0.40

0.35

0.30

0.25

0.20

0.15

2018 2018 2018 2018

2023E 2025E 2019E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2019E 2020E 2021E 2022E 2024E 2020E 2021E 2022E 2023E 2024E 2025E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Zone I Zone II Zone III Zone IV

Wind LCOE Coal-fired benchmark tariff (grid parity) Coal-fired market-based tariff Coal-fired blended tariff

Source: Daiwa forecasts

The major risks to our forecasts for China’s wind sector include the WTG price, grid curtailment rate and coal prices. Higher/lower-than-expected WTG prices would cause a slower/faster-than-expected decline in wind construction costs. If the improvement in the curtailment rate is slower than expected, or if coal prices fall significantly, there could be delays in the cost parity of wind power in China, in our view.

Solar: continuous decline in the LCOE going forward We estimate solar C&I DG projects in Zone I-III will achieve retail parity over 2019-21E We estimate that the C&I Unlike wind power, the LCOEs for solar power in China are mainly driven by fast-declining DG LCOE for Zone II will equipment costs and increasing solar product efficiency. We forecast installation costs of fall below the C&I user DG projects to fall at a CAGR of 10% to CNY3.0/watt during 2018-25E, due to technology tariff in 2019E, followed improvements and price cuts in the upstream solar segment. by Zone I in 2020E and Zone III in 2021E Retail parity is more relevant for DG projects. We estimate that the DG LCOE for Zone II will fall below the C&I user tariff in 2019E, followed by Zone I in 2020E and Zone III in 2021E. Thus, we expect DG projects to reach retail parity in all regions by 2021E. Given that the investment cost of a DG project per kWh is lower than the electricity savings after reaching retail parity, such projects will become economically viable for electricity end users, even without the government subsidy, which should lead to a rise in installation demand for DG projects going forward, in our view.

In our above analysis, we only factor in a 10% decline in end-user electricity tariffs in 2019 and 2-5% annual decline over 2020-25E. However, we note that with the growing portion of market-based power wholesale in China and the ongoing T&D reform, the effective electricity tariff burden of end users could be significantly eased in the next few years, potentially delaying the schedule for solar power retail parity.

We expect solar utility-scale projects in most regions to achieve grid parity in 2024E We expect solar utility- According to our estimates, it will take longer for utility-scale projects to achieve grid parity scale projects of Zone II- than for DG projects to reach retail parity. We assume around a 9% annual decline in III to achieve grid parity installation costs, driven by falls in solar module costs and increase in solar product in 2024E and Zone I in efficiency. According to our estimates, utility-scale LCOE in Zone II and III will drop to 2025E below the local coal-fired power benchmark tariff by 2024E. Meanwhile, as the local coal- fired benchmark tariff is relatively low in Zone I, we expect it to take even longer for solar farms in Zone I to reach grid parity than other zones. We estimate the utility-scale LCOE in Zone I to drop to the level of local coal-fired benchmark tariff and reach grid parity in 2025E.

China Wind and Solar: 11 July 2019

China solar: installation cost for utility-scale and DG solar project (CNY/w) 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E Utility scale Solar module 2.2 1.7 1.5 1.3 1.3 1.2 1.1 1.1 Other equipment 0.9 0.9 0.8 0.7 0.7 0.7 0.6 0.5 BOS system 2.5 2.4 2.2 2.0 1.9 1.9 1.6 1.2 Total installation cost 5.6 4.9 4.4 4.0 3.9 3.8 3.3 2.9 C&I DG project Solar module 2.2 1.7 1.5 1.3 1.3 1.2 1.1 1.1 Other equipment 0.9 0.9 0.8 0.7 0.7 0.7 0.6 0.5 BOS system 3.0 2.8 2.5 2.2 2.2 2.1 1.7 1.4 Total installation cost 6.1 5.3 4.8 4.3 4.1 3.9 3.5 3.0

Source: Daiwa forecasts

China solar: LCOE forecast for utility scale (CNY/kWh) China solar: LCOE forecast for C&I DG (CNY/kWh) 0.7 0.9 0.61 0.80 0.8 0.6 0.55 0.71 0.69 0.51 0.49 0.7 0.61 0.61 0.63 0.5 0.45 0.47 0.58 0.40 0.430.42 0.55 0.56 0.41 0.39 0.6 0.52 0.51 0.36 0.380.37 0.47 0.490.48 0.540.49 0.4 0.33 0.34 0.35 0.5 0.430.42 0.43 0.320.30 0.31 0.41 0.45 0.31 0.37 0.39 0.3 0.27 0.28 0.4 0.24 0.34 0.3 0.2 0.2 0.1 0.1 0.0 0.0 Zone I Zone II Zone III Zone I Zone II Zone III 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E

Source: Daiwa forecasts Source: Daiwa forecasts

China solar: cost parity outlook (CNY/kWh) 0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2018 2019E 2020E 2021E 2022E 2023E 2024E 2025E Zone I Zone II Zone III

Solar utility scale LCOE Solar DG LCOE Coal-fired benchmark tariff (grid parity)

Electricity tariff for C&I user (retail parity) Coal-fired market-based tariff Coal-fired blended tariff

Source: Daiwa forecasts

2019-21: transition period before grid and retail parity 2019-21 will mark a As we believe it is inevitable the wind and solar power industry will step into an era of grid transition period before and retail parity starting from 2022, the next 3 years (2019-21) will mark a transition period grid and retail parity before grid and retail parity.

Wind: characterised by rush installations over 2019-21 In December 2016, the NDRC finalised the tariff adjustments for new wind farms: 1) for wind farms approved before 2018, they will receive tariffs of CNY0.47-0.60/kWh (the same as over 2016-17) as long as a project commences construction before 31 December 2019 and is connected to the grid before end-2020, and 2) for wind farms approved after 2018 and commencing operation before end-2020, they will receive tariffs of CNY0.40-0.57/kWh (a 7-15% cut from 2016-17).

China Wind and Solar: 11 July 2019

Tariffs for the approved wind projects starting from 2019 will no longer be decided by the government and will be decided by competitive bidding. On 24 May 2018, the NEA issued a document on wind farm installations, in which it outlined plans to introduce a competitive bidding mechanism for granting installation quotas from 2019, which was finalised by the 2019 wind power management method released on 30 May 2019. We regard this move as a step towards transiting to grid parity.

We expect the wind According to government guidelines, competitive bidding will not affect pipeline projects sector to see rush that have already been approved in or before 2018 and are entitled to fixed tariffs if they installations during are connected on grid before 2020. In our view, large wind IPPs still have sufficient project 2019-21 reserves. According to BNEF, there are still 88GW of approved projects that are unaffected by the competitive bidding policy. Thus, we expect wind IPPs to mainly focus on developing approved project reserves with fixed tariffs over 2019-20E. Also, project approvals gained through competitive bidding in 2019-20 should be connected on grid before end-2021. Thus, we expect a period of rush installations over 2019-21.

China wind: tariffs for new onshore wind projects Approved Regions Approved before 2018 Approved in 2018 Approved in 2019 Approved in 2020 in 2021 Zone IV 0.60 (CNY/kWh) 0.57 (CNY/kWh) 0.52 (CNY/kWh) 0.47 (CNY/kWh) Zone III 0.54 (CNY/kWh) 0.49 (CNY/kWh) 0.43 (CNY/kWh) 0.38 (CNY/kWh) No subsidy Zone II 0.50 (CNY/kWh) 0.45 (CNY/kWh) 0.39 (CNY/kWh) 0.34 (CNY/kWh) anymore Zone I 0.47 (CNY/kWh) 0.40 (CNY/kWh) 0.34 (CNY/kWh) 0.29 (CNY/kWh) Note Start construction by end-2019, and Complete grid connection by end- Need to complete grid connection by end- Need to complete grid connection by end-2021, complete grid connection by end- 2020, otherwise no subsidy. 2021, otherwise no subsidy. otherwise no subsidy. 2020, otherwise no subsidy. Tariff is the cap of the competitive bidding. Tariff is the cap of the competitive bidding. Source: NDRC, Daiwa

China Wind and Solar: 11 July 2019

Competitive bidding methods announced by some provinces Fujian Offshore Fujian Offshore (Investment (Investment Points Guangdong Offshore Guangdong Onshore Ningxia Shaanxi Tianjin entities not entities confirmed) confirmed) Corporate ability 30 20 20 24 20 15 8 Investment ability 6 4 4 6 4 4 3 Performance of established investments 20 12 14 16 8 9 3 Technology ability 2 2 - - 4 - - Business credibility 2 2 2 2 4 2 2 Advance of equipment 8 8 5 6 5 25 20 Unit capacity 4 4 1 2 - 22 17 Type approval 2 2 2 2 2 Utilisation rate 2 2 2 2 3 3 3 Technology solutions 6 7 - 6 5 20 20 Smart wind farm design 2 2 - 2 2 1 1 Economic reasonability 2 2 - 2 2 18 18 Removal plan 1 1 - - 1 1 1 Measured utilisation hours 1 2 - 2 - - - Completion of preliminary work 10 18 25 24 23 - 8 Overall plan 2 2 - 3 - - Resources assessment 2 3 3 9 - 4 3 Feasibility research 3 3 - - 2 Gained supporting documents 3 10 22 18 14 - 2 Connection and consumption conditions 6 7 10 - 7 - 4 Declared tariff >40, no maximum Maximum 40 Maximum 40 Maximum 40 Maximum 40 Maximum 40 On-grid tariff decrease CNY0.01/kWh and government. government. On-grid tariff On-grid tariff On-grid tariff CNY0.01/kWh and CNY0.02/kWh: For weighted-average weighted-average On-grid tariff On-grid tariff On-grid tariff portions above CNY0.02/kWh, 1 pts for every tariff as the tariff as the decrease decrease > decrease CNY0.02/kWh decrease benchmark price, benchmark price, >CNY0.03/kWh: For CNY0.02/kWh: >CNY0.02/kWh: with a point of 30. with a point of 30. portions above For portions For portions CNY0.03/kWh, 1 above above 3. For every 3. For every pts for every CNY CNY0.02/kWh, CNY0.02/kWh, CNY0.01/kWh CNY0.01/kWh 0.01/kWh decrease 0.2 pts for every 0.2 pts for every below benchmark below benchmark CNY 0.01/kWh CNY 0.01/kWh price, 1 pts is price, 1 pts is decrease decrease awarded, awarded, maximum is 40 maximum is 40 points. For every points. For every CNY0.01/kWh CNY0.01/kWh above benchmark above benchmark price, 2 pts is price, 2 pts is deducted. deducted.

Source: Various provincial governments Note: 1) For Guangdong onshore wind projects (Zone IV) to achieve 40 pts for tariff, on-grid tariff needs to be cut by CNY0.22/kWh (39% cut) from current CNY0.57/kWh to CNY0.35/kWh, while current coal-fired benchmark on-grid tariff in Guangdong Province is CNY0.4530/kWh. 2) For Tianjin onshore wind projects (Zone IV) to achieve 40 pts for tariff, on-grid tariff needs to be cut by CNY0.13/kWh (23% cut) from current CNY0.57/kWh to CNY0.44/kWh, while current coal- fired benchmark on-grid tariff in Tianjin is CNY0.3655/kWh. 3) Ningxia has announced the competitive bid results. The average declared tariff of the 28 projects announced is CNY0.45/kWh, which is CNY0.04/kWh lower than the benchmark wind tariff of CNY0.49/kWh in Ningxia (Zone III), while current coal-fired benchmark on-grid tariff in Ningxia is CNY0.2595/kWh. 4) Current offshore wind on-grid tariffs are CNY0.85/kWh and CNY0.75/kWh for nearshore and intertidal projects, respectively. For Guangdong offshore wind projects to achieve 40 pts for tariff, on-grid tariff needs to be cut by CNY0.22/kWh (26-29% cut). For Fujian offshore wind projects to achieve 40 pts for tariff, on-grid tariff needs to be cut by CNY0.32/kWh (38-43% cut) for projects with investment entities not confirmed and CNY0.12/kWh (14-16% cut) for projects with investment entities confirmed, while current Fujian coal-fired benchmark tariff is CNY0.3932/kWh.

Solar: competitive bidding starts as a transition period in search of parity Subsidy quota for solar Unlike the wind power industry, the solar power industry does not have a large leftover projects needs to be project pipeline that is entitled to fixed benchmark tariffs or subsidies. Apart from a small gained through amount of subsidy-free projects, household DG projects, poverty-alleviation projects and competitive bidding in top-runner projects, we expect installations starting from 2019 to be mainly contributed by 2019 projects that gain subsidy quota from competitive bidding, before complete elimination of the subsidy when the solar sector reaches grid and retail parity.

China Wind and Solar: 11 July 2019

According to 2019 solar power management methods released by the NEA, the government has introduced a competitive bidding mechanism for utility-scale and C&I DG solar projects.

The NEA has set different subsidy plans for the 5 types of newly built solar projects in 2019:

1) Utility-scale projects: will go to auction with a total subsidy amount of CNY2.25bn 2) C&I DG projects: will go to auction with a total subsidy amount of CNY2.25bn 3) Household-rooftop DG projects: no auction; independent subsidy of CNY750m for 3.5GW of installations (3.5GW is conservative in our view, as the unit subsidy will reach CNY0.2/kWh compared to fixed household DG subsidy of CNY0.18/kWh set by the government. We think household-rooftop installations will reach 4GW in 2019 given the CNY750m subsidy) 4) Poverty-alleviation projects: no auction 5) Special projects (ie, top-runner): no auction

We estimate a total of According to the latest government policy, utility-scale and C&I DG solar projects will go to 36-40GW of solar auction under a total subsidy amount of CNY2.25bn, and the NEA approved total 23GW capacity will be installed competitive projects on 11 Jul 2019, including 18GW utility-scale projects and 5GW C&I in 2019E DG solar projects. The NEA also allocated CNY750m for 3.5-4GW of subsidised household-rooftop DG project construction. In addition, there will be 2-3GW in poverty- alleviation projects that do not need to go through auction, a 2GW top-runner project expected to be installed in mid-2019, and 3-5GW of subsidy-free projects that are set to be installed during 2019. We see a total of 36-40GW of solar capacity that could be installed in 2019, in line with the NEA’s guidance for an installation scale similar to that of 2018 (44GW).

The NEA sets strict construction schedules for new-build projects. If a solar project fails to complete construction within the applied timeframe, tariffs and subsidies are to be cut accordingly. If a project fails to complete construction within 2 quarters after the applied timeline, the project’s subsidy is to be cancelled. The NEA required the provincial Energy Administrations to submit the first batch of wind/solar subsidy-free projects before 25 April and submit competitive bidding proposals before 31 May 2019. Before the subsidy-free projects lists are submitted, each province is not allowed to organise any subsidy-required projects, per the NEA’s statement. Given the current schedule, the first batch of subsidy- free projects were announced in May 2019 and various provinces finished registration of the subsidy-required projects planned to participate in competitive bidding process before end-June 2019, and kick off the competitive bidding process after this. We see potential rush installations in 2H19 as there are only 1-2 quarters left in 2019 to complete new-built project construction.

The NDRC also released a notice on 2019 solar power tariffs on 30 April 2019. According to the NDRC’s notice, the benchmark on-grid tariff for utility-scale solar projects will be changed into a guidance tariff (upper limit of competitive bidding) in 2019, which will be cut to CNY0.4/0.45/0.55 per kWh for Zone I-III, compared with CNY0.5/0.6/0.7 per kWh in 2018. Poverty alleviation solar plants are still entitled to the original tariff of CNY0.65- 0.85/kWh for Zone I-III. For DG projects, the subsidy for C&I DG projects (self-generation- self-consumption) is cut to CNY0.1/kWh (upper limit in competitive bidding), while tariffs for

China Wind and Solar: 11 July 2019

C&I DG projects (under the “all on grid” model) are also subject to the guidance tariff of utility-scale projects (CNY0.4-0.55/kWh). Meanwhile, subsidies for all household DG projects are set at CNY0.18/kWh, which is not subject to competitive bidding. Therefore, subsidies for DG also face significant cuts, compared with CNY0.32/kWh for all DG The government sets the projects previously. The new tariff policy is in line with market expectations and the upper limit tariff for discussions between the government and the industry earlier this year. competitive bidding in

2019 China: on-grid tariff for new solar projects (CNY/kWh) 2015 2016 2017 2018 2019 Zone I 0.95 0.80 0.65 0.50 0.40 Zone II 0.98 0.88 0.75 0.60 0.45 Zone III 1.00 0.98 0.85 0.70 0.55 DG subsidy 0.42 0.42 0.42 0.37 (commissioned before 1 June 2018) 0.1 (C&I) 0.32 (commissioned after 1 June 2018) 0.18 (Household)

Source: NEA, NDRC, Daiwa Note: tariff of CNY0.40-0.55/kWh for Zone I-III and subsidy of CNY0.1/kWh for C&I DG projects are upper limit of competitive bidding in 2019

China Wind and Solar: 11 July 2019

Wind equipment: to benefit from rush installations in 2019-21 Demand: to see support from rush installations We expect rush According to government guidelines, competitive bidding will not affect pipeline projects installations and large that have already been approved on or before 2018 and are entitled to fixed tariffs if they capex to occur in the are connected on grid before 2020. Large wind IPPs still have sufficient project reserves, in wind sector over 2019-21 our view. According to BNEF, there are still 88GW in approved projects that are unaffected by the competitive bidding policy. Thus, we expect wind IPPs to mainly focus on developing approved project reserves with fixed tariffs over 2019-20E. Also, project approvals gained through competitive bidding over 2019-20 should be connected on grid before end-2021. Thus, we anticipate a rush of installations over 2019-21.

China wind: tariffs for new onshore wind projects Approved Regions Approved before 2018 Approved in 2018 Approved in 2019 Approved in 2020 in 2021 Zone IV 0.60 (CNY/kWh) 0.57 (CNY/kWh) 0.52 (CNY/kWh) 0.47 (CNY/kWh) Zone III 0.54 (CNY/kWh) 0.49 (CNY/kWh) 0.43 (CNY/kWh) 0.38 (CNY/kWh) Zone II 0.50 (CNY/kWh) 0.45 (CNY/kWh) 0.39 (CNY/kWh) 0.34 (CNY/kWh) No subsidy anymore Zone I 0.47 (CNY/kWh) 0.40 (CNY/kWh) 0.34 (CNY/kWh) 0.29 (CNY/kWh) Note Start construction by end-2019, and Complete grid connection by end- Need to complete grid connection by end- Need to complete grid connection by end-2021, complete grid connection by end- 2020, otherwise no subsidy 2021, otherwise no subsidy. otherwise no subsidy. 2020, otherwise no subsidy Tariff is the cap of the competitive bidding Tariff is the cap of the competitive bidding Source: NDRC, Daiwa

We note that most major wind operators are hastening the progress of wind capacity installations with heavy capex in the next few years. The most notable is HNP, which introduced aggressive plans for 2019 aiming for a 3.4x YoY rise in wind capex from CNY7bn to CNY24bn, for over 2GW of installations. Also, many major wind IPPs have set ambitious installation targets for 2019, including Longyuan (1.2-1.5GW), HNR (1.6GW), DTR (1GW), and CRP (1.7GW).

Wind capacity installations (MW) Capex for wind installations (CNY bn) 2,500 30

25 2,000 20 1,500 15 1,000 10 500 5

0 0 Longyuan HNR DTR Suntien HDFX CRP HNP Longyuan HNR DTR Suntien HDFX CRP HNP 2018 2019E 2018 2019E

Source: Company, Daiwa Source: Company, Daiwa

China wind: management guidance for 2019E Wind capacity installations (MW) Wind capex (CNY bn) Wind capacity approved but not yet in operation (GW) 2018 2019E 2018 2019E As of end-2018 Longyuan 524 1,200-1,500 7.8 12.3 7.0 HNR 347 1,600 4.2 12.0 4.2 DTR 189 1,000 2.3 7.5 4.3 Suntien 510 500 4.5 6.0-7.0 3.2 Huadian Fuxin 295 500 5.1 8.6 3.4 CRP 1,187 1,700 9.7 13.7 9.0 HNP 482 2,300 7.0 24.0 n.a.

Source: Company, Daiwa Note: For Huadian Fuxin, capex refers to the whole company including wind power and other power sources

China Wind and Solar: 11 July 2019

We expect more new wind projects to start construction before end-2019/20, and believe 2020-21E will see the peak for grid connections. Given that most wind farms will take 1-2 years to complete, we expect new installations in China to gradually rise from 21GW in 2018 to 25GW in 2019E, 33GW in 2020E and 29GW in 2021E.

China: annual installations for new wind farms (GW) 35 33 33 80% 29 30 60% 25 25 40% 21 19 20 18 20% 16 15 0%

10 (20%)

5 (40%)

0 (60%) 2014 2015 2016 2017 2018 2019E 2020E 2021E Newly connected capacity YoY

Source: NEA, Daiwa forecasts

Supply: market consolidation ongoing Market consolidation of We note that market consolidation of the wind turbine generator (WTG) manufacturers is WTG manufacturers is ongoing, as the market share of the leading players continues to rise with few newcomers ongoing entering the industry. In 2018, China added 21GW of wind capacity, 42.4% of global new installations, illustrating that China is the largest incremental wind power market in the world. The top-5 WTG manufacturing companies in China earned nearly 73% market share in total in 2018, up from 62% in 2017, showing further consolidation of the Chinese WTG market.

China: market share of WTG manufacturers in 2018 China: market share of WTG manufacturers in 2017 Siemens Gamesa, 1.3%

Vestas, 2.8% Others, Xiangtan 11.4% Electric, 3.3% Goldwind, CSIC, 3.8% 31.9% Others, 32% Goldwind, 29% China Windey, 4.5% Shanghai Electric, Envision, Envision, 15% 5.3% Mingyang, United Power, 17.5% CSIC, 5% 12.0% 6.2% Shanghai Electric, 6% Mingyang, 6% United Power, 7% Source: BNEF Source: BNEF

This is also the trend globally. In 2018, the top-5 WTG manufacturers together installed 63% of the world’s total installations, up from 59% in 2017, indicating further consolidation of the global WTG market.

China Wind and Solar: 11 July 2019

Global: market share of WTG manufacturers in 2018 Global: market share of WTG manufacturers in 2017

MHI Vestas, 2.6% Others, Vestas, 20.3% 16.8% Vestas, 18% United Power, Others, 29% 2.6% Siemens Nordex, 4.9% Goldwind, Gamesa, 14.2% 15% Mingyang, 5.1% Siemens Nordex, 6% Enercon, 5.1% GE, Gamesa, 11.0% GE, 9% 10.0% Envision, 6% Goldwind, 11% Envision, 7.4% Enercon, 6%

Source: BNEF Source: BNEF

We expect market leaders like Goldwind to benefit in the coming years, with their WTG sales volume driven by both WTG demand growth resulting from rush installations as well as increasing market share. Goldwind added 6.7GW in onshore and offshore wind in China in 2018, accounting for 31.9% of China’s total new installations, up from 29% in 2017, and has been ranked the leader in terms of market share in the Chinese market for the past 8 consecutive years. Goldwind expects its WTG external sales to rise from 5.9GW in 2018 to 8GW in 2019 and even higher in 2020. We forecast 8-10GW of WTG sales volume for Goldwind in 2019-21E.

Goldwind: market share in China Goldwind: WTG external sales (MW) 40% 12,000 80% 34% 34% 10,257 35% 32% 33% 29% 10,000 9,231 60% 27% 30% 25% 8,205 23% 25% 8,000 7,051 40% 20% 20% 20% 18% 20% 19% 20% 5,883 5,861 6,000 5,082 20% 15% 4,190 4,000 0% 10% 2,584 2,923 5% 2,000 (20%) 0%

0 (40%)

2011 2017 2008 2009 2010 2012 2013 2014 2015 2016 2018

2020E 2021E 2019E 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E Goldwind's market share in China WTG shipment YoY growth

Source: Company Source: Company, Daiwa

WTG: ASP started to recover from 4Q18 We note the WTG According to Goldwind, average bidding prices have rebounded since 4Q18, stabilising at bidding price has started a level above CNY3,400/kW as of end-1Q19, and are likely to increase further in the near to recover since 4Q18 term. The average bidding price for 2.0MW-unit WTGs was CNY3,410/kW in March 2019, and is likely to continue which was 6.7% higher than the trough in September 2018. The average bidding price for to do so in 2019 2.5MW-unit WTGs was CNY3,464/kW in March 2019, 4.0% higher than the trough in August 2018.

China Wind and Solar: 11 July 2019

China: WTG monthly average bidding price (CNY/kW) 3,900 3,800 3,700 3,600 3,500 3,400 3,300 3,200

3,100

Jul-18

Apr-18 Oct-17 Oct-18

Jan-18 Jun-18 Jan-19

Mar-18 Mar-19

Feb-18 Feb-19

Dec-17 Nov-17 Aug-18 Sep-18 Nov-18 Dec-18 May-18 2.5MW 2.0MW

Source: Goldwind

For Goldwind, the gross margin for the WTG segment declined from 25% in 2017 to 19% in 2018, while the company expects it to decline further this year, as most of the low-ASP orders in 2018 will be delivered in 2019. Thus, its WTG gross margin may see a bottom in 2019 before recovering in 2020 and 2021, as the bidding price started to rebound in 4Q18, and should rise further in 2019 and 2020.

Goldwind: WTG gross margin 4,500 30% 3,807 3,855 3,854 4,000 3,782 3,641 25% 25% 3,500 3,000 20% 19% 19% 18% 2,500 15% 2,000 14% 1,500 10% 948 1,000 707 721 709 499 5% 500 0 0% 2017 2018 2019E 2020E 2021E ASP (CNY/kW) GP (CNY/kW) GPM (RHS)

Source: Company, Daiwa estimates

China Wind and Solar: 11 July 2019

Wind IPP: worsening cash flows in 2019-21E Limited utilisation improvement We see limited upside We believe wind IPPs will face headwinds in 2019. With a high base in 2018 and the for wind utilisation curtailment rate already low, we see limited upside for utilisation growth in 2019 and going growth in 2019 and forward. Most of the wind farm operators recorded a decline in wind power utilisation in going forward early 2019, with Longyuan’s wind utilisation hours down by 0.2% YoY for 5M19, DTR down by 4% YoY for 1Q19, Suntien’s down 0.2% YoY for 5M19, and Huadian Fuxin down by 4% YoY for 5M19, according to our estimates.

China wind: management guidance for 2019E Utilisation hours Curtailment rate 2018 2019E YoY change 5M18 5M19 YoY change 2018 2019E YoY change Longyuan 2,209 2,230 1.0% 985 983 -0.2% 5.7% 5.0% -0.7pp HNR 2,234 2,234 0.0% 1,077 1,143 6.1% 6.0% 4.0% -2pp DTR 2,096 2,100 0.2% 586 566 -3.4% 8.1% 7.0% -1.1pp Suntien 2,482 2,300 -7.3% 1,225 1,223 -0.2% 5.4% n.a. n.a. Huadian Fuxin 2,128 2,150 1.0% 1,005 962 -4.3% 9.0% 7.0% -2pp

Source: Company, Daiwa Note: Utilisation hours of 5M19 and 5M18 are estimated by Daiwa

Less supportive RPS policy On 15 May 2019, the NDRC and NEA published the final official version (Chinese- language link) of the RPS for each province in China, after issuing 3 draft versions in 2018. The details in the final version are largely similar to those in the third draft issued in November 2018, which specified 2 RPS targets (ie, total RPS and non-hydro RPS) for each province, with the parties that need to comply with the RPS remaining on the consumer side (not including coal-fired IPPs). The targets have been further divided into obligatory and incentive targets, while those for some provinces were adjusted (mainly hydro consumption target) compared with the third draft. While the 2018 targets are just for self-review and 2019 targets are for trial assessment, various provinces should complete preparatory works before end-2019 as the RPS is to be formally implemented starting from 2020, vs. commencement in 2019 in the third draft. The 2020 RPS targets outlined currently are for guidance of renewable energy development in each province, while actual targets in 2020 will be released before end-1Q20. Similar to the 3rd draft, GCs are subject to voluntary trading, while parties that need to comply with the RPS can directly buy the excess renewable energy consumption volumes of other parties.

Voluntary trading of GCs The final version states that the completion of RPS targets should be mainly through actual is one way to achieve consumption of renewable energy, while there are 2 complementary ways: 1) buying the RPS compliance excessive renewable energy volumes of other parties, with the transfer price determined by both sides independently, and 2) buying GCs from renewable energy producers on a voluntary basis. We think the first way is more suitable for a regional grid company, with the trading of excess RPS completion volumes similar to a bilateral wholesale transaction, while the second way is more attractive for smaller RPS-compliant players, eg, power-retail companies, power end-users and captive power-plant owners, since there is a public trading platform for GCs offering more transparent prices.

GCs were introduced in China on 1 July 2017 but so far they have only been subject to voluntary trading and are not traded actively. Under the GC compulsory trading proposed in the second draft, it is likely that GC trading can partially replace the subsidies paid by the RE Fund and accelerate collection of part of the subsidies. Meanwhile, under voluntary trading (which is the current trading method for GCs) as stated in the third draft and final version, we expect the subsidy collection from renewable energy companies to still rely mainly on the RE Fund (which currently has a large deficit) and may continue to face delays.

China Wind and Solar: 11 July 2019

China: provincial total RPS targets for provincial grids, and power retailers/end-users/captive power plant owners Total RPS target Change compared with the 3rd draft 2018 2018 2019 2019 2020 2020 2018 2018 2020 2020 2017 obligatory incentive obligatory incentive obligatory incentive obligatory incentive obligatory incentive actual target target target target target target target target target target Beijing 12.1% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Tianjin 11.0% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Hebei 11.6% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Shanxi 14.1% 15.0% 16.5% 15.5% 17.1% 16.5% 18.1% 0.0pp 0.2pp 0.0pp 0.1pp Inner Mongolia 19.2% 18.5% 20.4% 18.5% 20.4% 18.5% 20.4% 0.0pp 0.1pp 0.0pp 0.1pp Liaoning 12.2% 12.0% 13.2% 12.0% 13.2% 12.5% 13.7% 0.0pp 0.2pp 0.0pp 0.1pp Jilin 22.2% 20.0% 22.0% 21.5% 23.7% 22.0% 24.2% 0.0pp 0.5pp 0.0pp 0.5pp Heilongjiang 20.2% 19.5% 21.5% 21.5% 23.7% 26.0% 28.6% 0.0pp 0.5pp 0.0pp 0.5pp Shanghai 33.3% 31.5% 34.9% 32.0% 35.2% 33.0% 36.5% 0.0pp 2.9pp 0.0pp 3.0pp Jiangsu 14.7% 12.5% 13.7% 13.5% 14.9% 14.0% 15.5% -2.0pp -1.4pp -1.0pp -0.3pp Zhejiang 19.3% 18.0% 19.8% 17.5% 19.3% 18.5% 20.4% 0.0pp 1.3pp -0.5pp 0.6pp Anhui 14.3% 13.0% 14.3% 13.5% 14.9% 14.5% 15.9% 0.0pp 0.3pp 0.0pp 0.2pp Fujian 24.2% 17.0% 18.7% 18.5% 20.4% 19.5% 21.4% 0.0pp 1.2pp -2.5pp -1.2pp Jiangxi 25.4% 23.0% 25.1% 25.5% 28.1% 29.0% 32.1% 0.0pp 1.6pp 0.0pp 2.1pp Shandong 7.3% 9.5% 10.4% 10.0% 11.0% 10.0% 11.0% 0.0pp 0.0pp -0.5pp -0.6pp Henan 14.6% 13.5% 14.9% 13.5% 14.9% 16.0% 17.6% 0.0pp 0.4pp 0.0pp 0.5pp Hubei 43.0% 39.0% 43.0% 37.5% 41.3% 40.0% 44.0% 0.0pp 3.1pp 0.0pp 3.0pp Hunan 50.4% 46.0% 50.5% 47.0% 51.7% 49.0% 53.9% -5.5pp -1.9pp -2.5pp 1.5pp Guangdong 32.4% 31.0% 34.2% 28.5% 31.4% 29.5% 32.5% 0.0pp 2.8pp 0.0pp 2.5pp Guangxi 51.6% 51.0% 56.2% 45.5% 50.1% 50.0% 55.0% 0.0pp 4.8pp 0.0pp 4.5pp Hainan 13.3% 11.0% 12.1% 11.0% 12.1% 11.5% 12.6% 0.0pp 0.6pp 0.0pp 0.6pp Chongqing 49.2% 47.5% 52.1% 42.5% 46.8% 45.0% 49.5% 0.0pp 4.6pp 0.0pp 4.2pp Sichuan 83.5% 80.0% 88.0% 80.0% 88.0% 80.0% 88.0% 0.0pp 7.6pp 0.0pp 7.6pp Guizhou 35.6% 33.5% 36.9% 31.5% 34.7% 31.5% 34.7% 0.0pp 2.9pp 0.0pp 2.7pp Yunnan 85.6% 80.0% 88.0% 80.0% 88.0% 80.0% 88.0% 0.0pp 7.0pp 0.0pp 6.8pp Tibet 83.8% NA NA NA NA NA NA NA NA NA NA Shaanxi 16.0% 17.5% 19.2% 18.5% 20.4% 21.5% 23.7% 0.0pp 0.8pp 0.0pp 1.0pp Gansu 46.9% 44.0% 48.4% 44.0% 48.4% 47.0% 51.1% 0.0pp 2.8pp 0.0pp 2.2pp Qinghai 64.9% 70.0% 77.0% 69.5% 76.5% 70.0% 77.0% 0.0pp 5.1pp 0.0pp 4.5pp Ningxia 23.0% 20.0% 22.2% 20.0% 22.0% 22.0% 24.2% 0.0pp 0.2pp -3.0pp -2.8pp Xinjiang 26.0% 21.0% 23.1% 21.0% 23.1% 22.5% 24.5% -4.0pp -3.4pp -3.5pp -2.8pp

Source: NDRC, NEA, Daiwa Note: the 3rd draft did not give targets for 2019

China: provincial non-hydro RPS targets for provincial grids, and power retailers/end-users/captive power plant owners Non-hydro RPS target Change compared with the 3rd draft 2018 2018 2019 2019 2020 2020 2018 2018 2020 2020 2017 obligatory incentive obligatory incentive obligatory incentive obligatory incentive obligatory incentive actual target target target target target target target target target target Beijing 10.4% 10.5% 11.6% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Tianjin 10.4% 10.5% 11.6% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Hebei 10.4% 10.5% 11.6% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Shanxi 12.0% 12.5% 13.8% 13.5% 14.9% 14.5% 16.0% 0.0pp 0.0pp 0.0pp 0.0pp Inner Mongolia 18.3% 18.0% 19.8% 18.0% 19.8% 18.0% 19.8% 0.0pp 0.0pp 0.0pp 0.0pp Liaoning 9.2% 10.0% 11.0% 10.0% 11.0% 10.5% 11.6% 0.0pp 0.0pp 0.0pp 0.0pp Jilin 16.4% 15.0% 16.5% 15.5% 17.1% 16.5% 18.2% 0.0pp 0.5pp 0.0pp 0.0pp Heilongjiang 15.8% 15.0% 16.5% 17.5% 19.3% 20.5% 22.6% 0.0pp 0.0pp 0.0pp 0.0pp Shanghai 2.7% 2.5% 2.8% 3.0% 3.3% 3.0% 3.3% 0.0pp 0.0pp 0.0pp 0.0pp Jiangsu 5.4% 5.5% 6.1% 6.5% 7.2% 7.5% 8.3% 0.0pp 0.0pp 0.0pp 0.0pp Zhejiang 4.2% 5.0% 5.5% 6.5% 7.2% 7.5% 8.3% 0.0pp 0.0pp 0.0pp 0.0pp Anhui 8.8% 9.5% 10.5% 10.5% 11.6% 11.5% 12.7% 0.0pp 0.0pp 0.0pp 0.0pp Fujian 4.5% 4.5% 5.0% 5.0% 5.5% 6.0% 6.6% 0.0pp 0.0pp 0.0pp 0.0pp Jiangxi 6.5% 6.5% 7.2% 7.0% 7.7% 8.0% 8.8% 0.0pp 0.0pp 0.0pp 0.0pp Shandong 6.9% 9.0% 9.9% 10.0% 11.0% 10.0% 11.0% 0.0pp 0.0pp -0.5pp -0.6pp Henan 8.1% 9.0% 9.9% 9.5% 10.5% 10.5% 11.6% 0.0pp 0.0pp 0.0pp 0.0pp Hubei 6.8% 7.5% 8.3% 9.0% 9.9% 10.0% 11.0% 0.0pp 0.0pp 0.0pp 0.0pp Hunan 7.2% 9.0% 9.9% 11.5% 12.7% 13.0% 14.3% 0.0pp 0.0pp 0.0pp 0.0pp Guangdong 3.2% 3.5% 3.9% 3.5% 3.9% 4.0% 4.4% 0.0pp 0.0pp 0.0pp 0.0pp Guangxi 3.0% 4.0% 4.4% 4.5% 5.0% 5.0% 5.5% 0.0pp 0.0pp 0.0pp 0.0pp Hainan 4.7% 4.5% 5.0% 5.0% 5.5% 5.0% 5.5% 0.0pp 0.0pp 0.0pp 0.0pp Chongqing 2.4% 2.0% 2.2% 2.5% 2.8% 2.5% 2.8% 0.0pp 0.0pp 0.0pp 0.0pp Sichuan 3.3% 3.5% 3.9% 3.5% 3.9% 3.5% 3.9% 0.0pp 0.0pp 0.0pp 0.0pp Guizhou 4.3% 4.5% 5.0% 5.0% 5.5% 5.0% 5.5% 0.0pp 0.0pp 0.0pp 0.0pp Yunnan 14.2% 11.5% 12.7% 11.5% 12.7% 11.5% 12.7% 0.0pp 0.0pp 0.0pp 0.0pp Tibet 14.0% NA NA NA NA NA NA NA NA NA NA Shaanxi 7.7% 9.0% 9.9% 10.5% 11.6% 12.0% 13.2% 0.0pp 0.0pp 0.0pp 0.0pp Gansu 13.8% 14.5% 16.0% 17.0% 18.7% 19.0% 20.9% -1.0pp -1.1pp 0.0pp 0.0pp Qinghai 18.5% 19.0% 20.9% 23.0% 25.3% 25.0% 27.5% 0.0pp 0.0pp 0.0pp 0.0pp Ningxia 21.0% 18.0% 19.8% 18.0% 19.8% 20.0% 22.0% 0.0pp 0.0pp 0.0pp 0.0pp Xinjiang 13.1% 11.5% 12.7% 12.0% 13.2% 13.0% 14.3% -3.0pp -3.3pp -3.0pp -3.3pp

Source: NDRC, NEA, Daiwa Note: the 3rd draft did not give targets for 2019

China Wind and Solar: 11 July 2019

Subsidy collection delays continue Subsidy collection As discussed previously, we expect subsidy collection going forward to still mainly rely on delays will still likely the Renewable Energy Fund, which has a large deficit currently (over CNY200bn deficit continue, pressuring the according to our estimate). This means that a subsidy collection delay is likely to continue, cash flows and balance pressuring cash flows and balance sheets of the wind farm operators. We think the sheets of the wind farm situation may even worsen going forward, because: 1) capacity ramp-up on rush operators installations will place a greater burden on the RE Fund and increase subsidy receivables for the wind IPPs, and 2) the NEA may reallocate CNY3bn for the competitive bidding of new solar power capacities in 2019 and pay subsidies to new solar projects ahead of existing wind and solar projects starting from this year.

RE Fund movement (CNYbn) CNYbn 700

Total fund shortfall Wind Solar Others Fund inflow

Accounts receivables for wind farm operators (CNYm) Accounts receivable days for wind farm operators 12,000 300

10,000 250

8,000 200

6,000 150

4,000 100

2,000 50

0 0 Longyuan HNR DTR Longyuan HNR DTR 2016 2017 2018 2016 2017 2018

Source: Company, Daiwa Source: Company, Daiwa

The government has not Under the current policy, renewable energy projects need to be admitted into a subsidy launched registration for catalogue before they are entitled to subsidy payments. So far, the government has the 8th batch of RE organised registration of 7 batches of subsidy catalogues, covering the renewable energy subsidy catalogues projects that started operation before end-March 2016. Meanwhile, as of now, the government has not launched registration for the 8th batch of RE subsidy catalogues, meaning that the projects that started operation after March 2016 are not yet entitled to subsidy collection and may face even longer waiting periods.

China Wind and Solar: 11 July 2019

China wind: exposure to editions of subsidy catalogue as of end-2018 Capacity (MW) Consolidated installed Proportion in Proportion not in 1-5th 6th 7th capacity (MW) 1-7th batch 1-7th batch Longyuan 11,300 2,268 2,646 18,919 86% 14% Huaneng Renewables* 5,579 1,947 2,762 11,033 93% 7% Datang Renewable 4,716 1,324 640 8,835 76% 24% Huadian Fuxin* 3,000 1,997 2,078 7,993 89% 11% Goldwind 1,304 696 739 4,429 62% 38% Suntien 1,346 351 541 3,858 58% 42% China Resources Power 2,874 700 654 6,816 62% 38%

Source: NEA, company, Daiwa Note1: Data for companies with asterisks (*) based on our estimate. Note2: Covering period: 1-5th batch (before end-August 2013), 6th batch (Sep 2013 to Feb 2015), 7th batch (Mar 2015 to Mar 2016)

Rush installations causing even tighter cash flow Wind IPPs’ free cash As mentioned previously, we expect to see rush installations for wind IPPs in 2019-21, flow will be even tighter meaning that wind IPPs need to have a large amount of capex to install their existing and more stretched with project pipelines before the deadline set by the government, so as not to lose the higher significant capex benchmark on-grid tariff. In this case, wind IPPs’ free cash flow will be even tighter and more stretched with significant capex, on top of operating cash flow impacted by subsidy collection delays and the increase in accounts receivables. Also, more debt will result in an increase in finance cost, which may offset EBIT growth for some wind IPPs.

China wind: capex of listed companies (CNYbn) 30

0 Longyuan HNR DTR Suntien HDFX CRP HNP 2018 2019E

Source: Company, Daiwa

China Wind and Solar: 11 July 2019

Solar product: be selective among segments

As the solar industry has a very long value chain and each segment has different characteristics, we recommend taking a selective approach to segments.

PV value chain

Source: svmi.com, silfex.com, company, Daiwa

Industry demand in China and overseas China: solar installations remain relatively stable before retail parity China’s new solar In a document released on 31 May 2018, the NEA announced that it would stop granting capacity was 44GW in installation quotas for ordinary utility solar farms in 2018 on the grounds that capacity 2018, down 16% expansion in previous years was too aggressive. Only 10GW of installation quotas for compared with distributed generation (DG) projects were to be granted. installations in 2017 According to the NEA, China installed 44GW of solar power capacity in 2018, down 16% YoY, which included 23GW of utility-scale projects (down 30% YoY) and 20GW of DG projects (up 7% YoY). We think the installations in 2018 mainly came from ordinary utility- scale projects from the previous year’s quota, top-runner projects, poverty alleviation projects and C&I/household DG projects (some DG projects were completed without a subsidy quota). As at end-2018, China’s solar capacity stood at 174.63GW, likely comprising 125GW of utility-scale projects and 50GW of DG projects.

We think solar The “May 31” policy in 2018 hit China’s solar industry hard. Meanwhile, as mentioned in installation will be the previous chapter, the government has introduced a competitive bidding mechanism for relatively stable in 2019 utility-scale and C&I DG solar projects starting this year, according to the 2019 solar power compared with 2018 management method released by the NEA on 30 May 2019. Apart from a small amount of subsidy-free projects, household DG projects, poverty-alleviation projects and top-runner projects, we expect the installation starting from 2019 will come mainly from projects that gain subsidy quotas through competitive bidding, before subsidies are removed altogether.

According to the latest government policy, utility-scale and C&I DG solar projects will go to auction under a total subsidy amount of CNY2.25bn, and the NEA approved a total of 23GW in competitive projects on 11 July 2019, including 18GW in utility-scale projects and 5GW in C&I DG solar projects. The NEA also allocated CNY750m for 3.5-4GW of subsidised household-rooftop DG project construction. In addition, there will be 2-3GW in poverty-alleviation projects that do not need to go through auction, a 2GW top-runner project expected to be installed in mid-2019, and 3-5GW of subsidy-free projects that are set to be installed during 2019. We see a total of 36-40GW of solar capacity that could be installed in 2019, in line with the NEA’s guidance for an installation scale similar to that of 2018 (44GW).

China Wind and Solar: 11 July 2019

2H19 may see rush According to the current policy schedule, the competitive bidding process will commence installations in China’s after various provinces finish registering the subsidy-required projects planning to solar industry participate in competitive bidding by end-June 2019. The implication is that demand was weak in 1H19 but is likely to recover in 2H19, as many solar projects may need to finish construction in 3Q19 and 4Q19 under more stringent deadlines required by the new policy. We expect these developments to support solar product ASPs in 2H19.

We see a potential surge Based on our previous estimate, we think DG projects across Zones I-III could reach retail in DG installation in 2021 parity and stop relying on government subsidies by 2021, while utility-scale projects in with the coming of retail Zones II-III would reach grid parity beyond that time frame. We expect utility-scale solar parity project installation to be largely stable in 2020-21, as it is still largely subject to the amount of government subsidies. Meanwhile, according to a notice released by the NEA on 9 January 2019, subsidy-free projects can be approved by provincial-level governments and no longer require NEA quotas. Thus, DG installation may see an increase in 2020, with more areas reaching retail parity, and potentially surge in 2021 with the coming of retail parity in most regions in China.

Overall, based on our estimates, we expect solar installation to be relatively stable, at c.40GW in 2019 and 42GW in 2020, compared with 44GW in 2018, before surging to 50GW in 2021E.

China: annual solar installation 60 140% 120% 50 100% 20 40 80% 30 4 21 60% 30 21 23 40% 20 20% 34 1 30 23 0% 10 2 19 19 20 14 (20%) 9 0 (40%) 2014 2015 2016 2017 2018 2019E 2020E 2021E Utility-scale DG YoY growth (RHS)

Source: NEA, Daiwa estimates and forecasts

Global: solar installations to see a steady increase While China is the largest solar market in the world and is likely to remain so in the coming years, the overseas market is steadily expanding. While major overseas markets such as the US, India, Europe, and Japan are likely to maintain relatively robust solar installation going forward, new markets are emerging, too.

US: decline in Section 201 tariffs (while bifacial modules are excluded from the tariffs for now) and continued ITC subsidies in the next few years. On 23 January 2018, US President Trump signed a proclamation that placed tariffs on imported solar cells and modules for a period of 4 years, known as Section 201 Solar Tariffs, effective from 7 February 2018. The tariff level was set at 30%, with a 5pp declining rate per year over the 4-year term of the tariff. While the imposition of the tariff would have increased solar installation costs in the US, the decline in Section 201 tariffs would likely have a positive

China Wind and Solar: 11 July 2019

impact on demand in the US. Furthermore, on 13 June 2019, US Federal trade authorities ruled that bifacial solar modules were no longer subject to the Section 201 ruling. The tariff waiver will help boost shipments to the US and spur the rise in market share of bifacial modules, in our view.

Another important factor in the US is the Solar Investment Tax Credit (ITC), a dollar-for- dollar reduction in the income taxes that a person or company would otherwise pay to the federal government, based on the amount of investment in solar property. Both the residential and commercial ITC are equal to 30% of the investment in eligible solar property upon which construction has begun through 2019. The ITC then drops to 26% and 22% for projects upon which construction begins in 2020 and 2021, respectively. After 2021, the residential credit drops to zero while the commercial credit drops to a permanent rate of 10%. Commercial and utility-scale projects that commenced construction before 31 December 2021 may still qualify for the 30%, 26% or 22% ITC if they start operation before 31 December 2023. We maintain our view that the ITC policy will support solar installation in the US over 2019-21, though its effect is likely to wane considerably beyond 2021.

India: decline in protective tariffs. India was the third-largest solar market in the world in 2018. The Indian government imposed a safeguard duty on solar imports from China and Malaysia (which accounted for nearly 90% of India’s solar cell imports) for 2 years. The safeguard duty is 25% from 30 July 2018 to 29 July 2019, and will gradually come down to 20% from 30 July 2019 to 29 January 2020, and 15% from 30 January 2020 to 29 July 2020. Lower protective tariff duties would likely have a positive impact on installation demand in India, in our view.

Europe: MIP ended in September 2018. With the minimum import price (MIP) measures imposed on solar PV cells and modules from China since 2013 officially ending in September 2018, the installation cost of solar projects is likely to drop, with demand likely to recover starting in 2019. BNEF expects that European unsubsidised and tendered utility- scale markets will grow strongly going forward.

Japan: rush installations in 2019-20. In December 2018, the government of Japan announced a reduction in Feed-In-Tariffs (FIT) for large-scale PV projects (over 2MW) approved between 2012 and 2014 that do not submit grid-connection applications by September 2019. The deadline for the start of operations was the end of September 2020. According to Japanese government estimates, around 23.5GW of solar projects that are entitled to FITs of JPY40/kWh (USD0.36/kWh), JPY36/kWh (USD0.32/kWh) and JPY32/kWh (USD0.29/kWh) may be affected and assigned a new rate of JPY21/kWh (USD0.19/kWh). The Japanese government is unsatisfied with the number of unfulfilled permits and has said that the cuts are necessary to reduce the public burden of the FIT subsidies, which are added to consumers’ bills. According to Power Technology, about 23% of the total capacity approved in 2012 is not operating, while 49% approved in 2013 and 59% approved in 2014 are also not yet operating. Developers that can finalise their projects by end-3Q20 will receive FITs set before the cuts were proposed. Thus, we expect rush installation in Japan in 2019-20.

We expect global solar Besides traditionally large solar markets, new markets for solar are picking up speed, such installation to continue as the Middle East and North Africa. We believe countries such as Vietnam and Mexico will to increase steadily also see significant increases in their solar installation over 2019-21E.

We maintain our view that the global solar installation will rise from 107GW in 2018 to 125W in 2019E, 136GW in 2020E, and 150GW in 2021E, representing a CAGR of 12% over 2018-21E.

China Wind and Solar: 11 July 2019

Global: solar annual installation (MW) 160,000 60% 140,000 50% 120,000 40% 100,000 30% 80,000 20% 60,000 10% 40,000 0% 20,000 (10%) 0 (20%) 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E China US Japan India Europe Rest of the world YoY growth (RHS)

Source: NEA, BP, BNEF, Daiwa estimates

Global: solar annual installations (MW) 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E China 3,198 10,962 10,610 15,130 34,540 53,000 44,260 40,000 41,995 49,654 US 3,369 4,751 6,238 7,357 14,726 10,600 11,659 12,301 14,888 16,292 Japan 1,718 6,968 9,740 10,811 7,850 7,000 6,785 8,557 7,712 2,030 India 361 360 2,007 1,878 4,250 9,629 11,107 12,150 14,310 14,880 Europe 17,984 10,073 7,214 8,512 7,101 8,970 11,080 17,577 19,087 19,991 Rest of the world 2,796 3,470 5,021 5,129 7,408 9,801 23,270 34,473 38,365 47,154 Total 29,426 36,583 40,830 48,817 75,875 99,000 108,159 125,058 136,356 150,000 YoY growth (RHS) -7% 24% 12% 20% 55% 30% 9% 16% 9% 10% Source: NEA, BP, BNEF, Daiwa estimates

Global: proportion of solar installations by regions (2018) Global: proportion of solar installations by regions (2019E)

Rest of the world Rest of the China 22% China world 32% 41% 27% Europe 10%

Europe India US 14% 10% 10% US India Japan 11% Japan 10% 7% 6%

Source: NEA, BNEF, Daiwa estimates Source: BNEF, Daiwa estimates

Supply: analysis of different segments Polysilicon: ASP likely to rebound starting from 2H19 Many tier-1 companies The polysilicon ASP fell by 44% YoY in 2018, most notably after the release of the “31 are building polysilicon May” policy. Besides, many tier-1 companies are building new production facilities, driven production facilities in by intensifying competition and pressure to cut production costs by investing in more provinces with relatively advanced capacity. Most of these new polysilicon production facilities are scheduled to be low power tariffs commissioned in 2H18-1H19, which we estimate will result in an 21% CAGR in total polysilicon capacity over 2017-19E. Furthermore, much of this new capacity will have relatively low production costs, as it will be in provinces where power tariffs are relatively low (eg, Xinjiang, Inner Mongolia and Sichuan) will be equipped with advanced technology, in our view. Therefore, polysilicon pricing saw a continued downtrend in 1H19.

China Wind and Solar: 11 July 2019

Global: new polysilicon capacity commission schedule of major players Capacity (tonne) Location Commission/Resumption REC 7,000 Shaanxi 1Q18 Wacker (resume operation) 20,000 Tennessee 2H18 East Hope (resume operation) 15,000 Xinjiang 2H18 OCI 3,200 Malaysia 3Q18 GCL-Poly 50,000* Xinjiang Oct-18 Tongwei 25,000* Inner Mongolia (Phase I) Oct-18 Tongwei 25,000* Sichuan (Phase I) Dec-18 Daqo 12,000 Xinjiang 4Q18 OCI 10,000 Malaysia 1Q19 Xinte (TBEA) 36,000 Xinjiang 1H19 Daqo 5,000 Xinjiang 1H19 Daqo 35,000 Xinjiang 2020 Tongwei 50,000 Inner Mongolia and Sichuan (Phase II) TBC

Source: Company, Daiwa Note: GCL-Poly’s plant in Xinjiang will reach 60ktpa capacity by mid-2019. Tongwei’s plants in Inner Mongolia and Sichuan (Phase I) can reach 30ktpa respectively with 1-3 months’ production ramp-up after operation commencement

Tie 1-3 effective polysilicon capacities in 2019E Polysilicon effective capacities (k MT) Tier 1-low cost capacities 303 Tongwei (new plant) 40 DAQO 35 East Hope 30 Tongwei (old plant) 20 Xinte 50 GCL 110 Asia Silicon 18 Tier 2-middle cost capacities 216 OCI 70 REC Silicon & Youser JV 14 Wacker 69 China Silicon 18 REC Solar Norway (Elkem/Bluestar) 7 Hankook Silicon 14 DunAn 10 Hanwha Chemical 14 Tier 3-high cost capacities 93 Other Chinese companies 40 REC Silicon FBR 4 Hemlock 30 Wacker (semiconductor) 8 REC Silicon Siemens 2 Tokuyama Shunan, Mitsubishi, Osaka Titanium 10 Total capacities (k MT) 612 Source: Companies, BNEF, Daiwa estimate Note: this list excludes some idle capacities

Global polysilicon production capacity (’000 tonnes) Polysilicon ASP trend (USD/tonne) 900 831 841 40% 25 2019-21E CAGR: 5% 756 800 35% 23 700 21 599 30% 600 515 19 475 25% 500 17 20% 400 356 15 275 15% 300 13 200 10% 11 100 5% 9 -1% YoY -32% YoY +16% YoY 0 0% 7 +18% YoY -48% YoY -12% YTD 2014 2015 2016 2017 2018 2019E 2020E 2021E 5 Capacity YoY (RHS) Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Jan-17 Jul-17 Jan-18 Jul-18 Jan-19 Jul-19 Source: Company, Daiwa forecasts Source: Bloomberg, Daiwa

We think further capacity Most of the new lower-cost capacity was largely commissioned in 2H18 and 1H19. We expansion in the think further capacity expansion in the industry will be limited in 2H19-21, with capacity industry will be limited in growth slowing from an 21% CAGR in 2017-19E to a 5% CAGR in 2019-21E, as new 2H19-21 capacity entails large capex. We believe polysilicon prices are likely to bottom out in 1H19. As the price of polysilicon currently is lower than the production cost of many polysilicon

China Wind and Solar: 11 July 2019

manufacturers, we think further downside risk is limited and polysilicon prices could stage a robust rebound starting from 2H19, once demand in China kicks off in the second half of the year, as we expect.

China: capex for polysilicon plant construction Total investment Unit investment Commission Period from announcement Construction Location Capacity (tonnes) (CNYbn) (CNY per tonne) date to commission period GCL -Poly Xinjiang 60,000 6.0 99,967 28-Oct-2018 1.6 years 8 months Tongwei Inner Mongolia (Phase I) 30,000 3.2 107,617 30-Oct-2018 1.3 years 12 months Tongwei Sichuan (Phase I) 30,000 3.2 106,140 28-Dec-2018 1.8 years 12 months

Source: Company, Daiwa

Polysilicon cost curve (2019E) 2021E demand:499k2021E 18 demand:435k2019E

) 12

MT MT 10 Polysilicon spot price: US$6.9/KG 8

6 Cash Cash cost (US$/kg

0 - 50 100 150 200 250 300 350 400 450 500 550 600 2019 effective capacity (k MT) Tongwei (new plant) Daqo New Energy East Hope Xinte (TBEA) GCL (average cost of Xuzhou/Xinjiang plant) Asia Silicon Tongwei (old plant) REC Solar Norway (Elkem/Bluestar) OCI REC & Youser JV Wacker Hemlock Semiconductor Source: Company, Daiwa estimates Note: assume 125GW global solar installation in 2019E and 3.6 tonnes polysilicon use per MW

Wafer: competition between multi and mono Demand-supply looks Looking forward, we forecast China’s solar installations to expand by 5% YoY to 42GW in relatively balanced in the 2020E, while global installation growth slows to 9% YoY for 2020E to 136GW after a wafer segment rebound to 16% YoY for 2019E. For 2021E, we see 2021E solar installations rebounding in China by 18% YoY to 50GW and globally by 10% YoY to 150GW, thanks to China’s DG retail parity in 2021E.

On the supply side, global wafer capacity looks set to rise rapidly from 168GW in 2018 to 184GW for 2019E, 205GW for 2020E, and 228GW for 2021E, on our forecasts, representing a CAGR of 11% over 2019-21E, which would largely meet our forecast for a global solar installation demand CAGR of 12% over the same period. Given this balanced demand-supply structure, we see limited upside for wafer prices. Currently, GCL-Poly (which focuses on multi-wafers) and LONGi and Zhonghuan (which focus on mono-wafers) are the 3 largest players in the wafer segment.

Global wafer capacity (GW, 2015-21E) 250 40% 228 205 35% 200 184 30% 168 150 147 25% 30 20% 107 100 90 30 50 15% 30 30 40 10% 50 30 30 25 50 65 5% 14 18.5 11 28 36 0 52 83 15 0% 2015 2016 2017 2018 2019E 2020E 2021E LONGi Zhonghuan GCL-Poly Other YoY

Source: Company, Daiwa estimates

As a highly pure form of silicon used as the raw material in the solar upstream manufacturing sector, polysilicon can be directly cast into multi silicon rods or submitted to a recrystallisation process to grow mono-silicon ingots. Of the 2 approaches, silicon

China Wind and Solar: 11 July 2019

casting has lower entry barriers, and the market is currently highly competitive. The recrystallisation process is more complex and requires polysilicon of greater purity (currently, the price of mono-use polysilicon is CNY17/kg [USD2.1/kg] higher than the price of multi-use polysilicon). The mono ingot produced can be further sliced to mono wafers and used to produce mono cells, which have better electricity conversion efficiency than multi cells.

Market preference has Due to the recrystallisation technology and use of higher-purity polysilicon, mono silicon been shifting from multi has a more organised structure than multi silicon, and thus greater conversion efficiency wafer to mono wafer (solar power to electricity). On the flip side, mono silicon costs more. The market used to be dominated by multi-silicon products because of their low cost. However, the cost difference between multi and mono silicon products has been steadily narrowing on the back of technological developments. The price-discount percentage ([mono-wafer price- multi wafer price]/mono-wafer price) declined from 30% in 1Q17 to 10% in 1Q18. The market has gradually shifted to mono wafers as mono’s cost-to-performance ratio was higher than multi wafer’s over the same period. Meanwhile, we note that the multi-wafer price discount increased from a 2-year average of 20% to a 1-year average of 25%, with the price slump of wafer products after 31 May 2018. That said, the mono-wafer price decline was less severe than that of multi wafer when facing policy shocks like the 31 May Policy.

Multi and mono solar-wafer ASP (USD/piece) Multi wafer price discount % as of mono wafer 40% 0.80 35% 0.70 30% 0.60 25% 0.50 One year average of 25% 20% 0.40 15% 0.30 0.20 10%

Jul-17 Jul-18 Jul-19

Jan-19 Jan-18

Mar-18 Mar-19

Nov-17 Nov-18

Sep-17 Sep-18 May-19 May-18 Apr-17 Oct-17 Apr-18 Oct-18 Apr-19 Mono Multi Multi-wafer price discount% as of mono-wafer

Source: Bloomberg, Daiwa Source: Bloomberg, Daiwa

Besides the narrowing price gap, China Top Runner Projects (government-organised projects which aim to accelerate solar products’ efficiency improvement and achieve solar power grid parity) have helped promote mono technology, as mono solar products can meet the high efficiency requirements much more readily than multi ones. With DG retail parity approaching, cost reductions and efficiency gains are the main ways of achieving solar power parity. We view the market’s shift from multi to mono silicon as irreversible, with the market-share gains of mono products likely to accelerate over the next few years.

China Top Runner Projects efficiency requirements for modules Top Runner Projects Batch 1 Batch 2 Batch 3 Multi module ≥16.5% ≥16.5% ≥17.0% Mono module ≥17.0% ≥17.0% ≥17.8% Source: bjx.com

According to the China Photovoltaic Industry Association (CPIA), the market share of mono silicon likely exceeded 45% in 2018 and could reach 55% in 2019. Given limited scope for cost reductions and multi-silicon’s comparatively low efficiency, its market share will likely fall to 27% by 2025E, according to the CPIA. We note that GCL-Poly has been actively developing mono-like technology (multi silicon in nature) in an apparent bid to avoid losing multi market share. However, the CPIA forecasts already include a mono-like portion in the multi-silicon share, and it is hard to see multi-silicon’s market-share decline coming to an end.

China Wind and Solar: 11 July 2019

Global multi- and mono-silicon market share 100% 90% 27.0% 35.5% 31.0% 29.0% 80% 45.5% 70% 55.0% 77.0% 71.0% 60% 85.0% 50% 95.0% 40% 73.0% 64.5% 69.0% 71.0% 30% 54.5% 20% 45.0% 23.0% 29.0% 10% 15.0% 0% 5.0% 2014 2015 2016 2017 2018 2019E 2020E 2021E 2023E 2025E Mono Multi Source: CPIA, Company Note: CPIA skipped 2022 and 2024 forecasts

Demand-supply We expect the increasing mono market share to bring 68GW of mono-wafer demand in structure for mono- 2019E and 88GW in 2020E. As at end-2018, there was only 44GW of effective mono-wafer wafers should be largely capacity, on our estimates. To fill the capacity shortage, mono-wafer producers are balanced, while multi- spending an additional CNY20bn-plus in capex to expand capacity over 2019-20E. The wafer manufacturers world’s largest mono-wafer producer, LONGi, added 13GW (up 87% YoY) of mono-wafer look likely to succumb to capacity last year, which took its total wafer capacity to 28GW as at end-2018. oversupply pressure LONGi plans to further expand its mono-wafer capacity to 36GW by 2019E, 50GW by 2020E, and 65GW by 2021E, representing a 32% CAGR from 2019-21E, which would secure its position as the world’s largest mono-wafer producer with a mono-wafer market share of 47% and total wafer market share of 29% in 2021E. Thanks to the aggressive capacity expansion plans of the mono-wafer producers, global wafer capacity looks set to expand to 205GW in 2020E and 228GW in 2021E, of which mono-wafer capacity will amount to 114GW in 2020E and 139GW in 2021E, and effective mono capacity will amount to 84GW in 2020E and 102GW in 2021E, after considering utilisation rates and new capacity installation time.

Global mono-wafer demand and supply (2015-21E) Global multi-wafer demand and supply (2015-21E) 120 100 92 104102 88 86 79 82 100 88 84 80 70 66 68 80 68 58 59 57 65 60 48 48 47 60 49 44 40 40 29 25 14 17 16 20 8 20

0 0 2015 2016 2017 2018 2019E 2020E 2021E 2015 2016 2017 2018 2019E 2020E 2021E

Mono-wafer demand (GW) Mono effective capacities (GW) Multi-wafer demand (GW) Multi effective capacities (GW)

Source: Daiwa estimate and forecasts Source: Daiwa estimate and forecasts Note: The effective mono wafer capacity is calculated based on an assumption of an 85% Note: The effective multi wafer capacity is calculated based on an assumption of a 90% utilisation utilisation rate and adjusted by new capacity installation time rate and adjusted for new capacity installation time

The demand-supply structure for mono-wafers should be largely balanced, on our estimates, assuming a 33% CAGR expansion in mono-wafer effective capacity over 2019- 21E. We forecast total mono-wafer demand to expand to 104GW for 2021E given 150GW of global solar installations and a 69% mono-wafer market share. On the supply side, we expect total mono-wafer capacity to expand to 139GW by the end of 2021E after the substantial capex spend, with an effective mono-wafer capacity addition to 102GW (based on an 85% utilisation rate and adjusted by new capacity installations) for 2021E, which would largely meet total demand of 104GW. As such, we estimate the global mono-wafer shortage will narrow from 12% in 2017 to 2% in 2021E.

China Wind and Solar: 11 July 2019

On the flip side, multi-wafer manufacturers are likely to succumb to oversupply pressure as all new added solar demand would naturally flow to mono products, and the rapid loss in multi market share continues to exacerbate the oversupply situation. Multi-wafer producers would only be able to compete for the extra demand in the absence of adequate mono- wafer supply, in our view.

Mono- and multi-wafer supply-demand structure (2015-21E) 2015 2016 2017 2018 2019E 2020E 2021E Global solar installation (GW) 56 75 99 108 125 136 150 Mono market share 15% 23% 29% 45% 55% 65% 69% Multi market share 85% 77% 71% 55% 46% 36% 31% Mono wafer demand (GW) 8 17 29 49 68 88 104 Multi wafer demand (GW) 48 58 70 59 57 48 47 Global mono wafer capacity (GW) 16 24 44 70 87 114 139 Global multi wafer capacity (GW) 73 83 103 98 97 91 89 Mono wafer effective capacity (GW) 14 16 25 44 65 84 102 Multi wafer effective capacities (GW) 66 68 79 92 88 86 82 Mono wafer (shortage)/oversupply 66% -8% -12% -10% -4% -4% -2% Multi wafer (shortage)/oversupply 39% 18% 13% 54% 55% 77% 76% Source: CPIA, Company, Daiwa estimates Note: The effective capacity is based on an assumption of 85% and 90% utilisation rates for mono and multi-capacities, respectively, and adjusted by new capacity installation time.

We expect stable wafer We saw severe price competition between multi and mono wafers in 2018. Strict cost ASPs in 2019 control and declining unit costs allowed LONGi to make 11 rounds of wafer price cuts (from CNY5.4/piece to CNY3.05/piece, down 44% YoY) in 2018 to maintain its market share. As at end-2018, LONGi claimed 40% of the global mono wafer market, up 35pp from its share in 2014.

LONGi: mono wafer 11 rounds of price cuts in 2018 and price hike in February 2019 6.0 5.4 5.5 5.2 4.8 5.0 4.55 4.5 4.45 4.5 4.25

4.0 3.65 3.35 3.5 3.15 3.05 3.15 3.07 3.0

2.5 Dec-17 Jan-18 Feb-18 Mar-18 Apr-18 May-18 Jun-18 Jul-18 Aug-18 Sep-18 Oct-18 Nov-18 Dec-18 Jan-19 Feb-19 LONGi mono-wafer prices changes

Source: Company Note: The latest wafer price cut on 22 March 2019 was triggered by China's Value-Added Tax (VAT) cut from 16% to 13%. LONGi chose to pass through the benefit to downstream manufacturers and keep the net VAT price unchanged at CNY2.72/piece

On 30 January 2019, the world’s largest multi-wafer producer, GCL-Poly, raised its multi- wafer price by 4.8% to CNY2.15/piece. Around the same time, the world’s second-largest mono-wafer producer, Zhonghuan (002129 CH, not rated), raised the price of its mono wafers by 4.8% to CNY3.25/piece. After 11 rounds of price cuts, LONGi followed suit and raised the price of its mono wafers by 3.3% to CNY3.15/piece. We see these moves as a short-term stabilisation of supply-demand dynamics in the solar wafer market and believe the price slump experienced after 31 May 2018 is unlikely to be repeated.

Still, with more new capacity coming on stream in 2H19E from LONGi and Zhonghuan, price competition may intensify and wafer prices are likely to come under pressure, in our opinion. We expect stable wafer ASPs in 2019 despite a potential rise in polysilicon prices in 2H19, with mono-wafer prices seen trading at a CNY0.7-1.0/piece premium to multi wafers. Multi-wafer prices are moving towards cash cost and therefore the downside seems limited; mono-wafer ASPs should also be supported, in our view.

China Wind and Solar: 11 July 2019

Compared with multi, According to PVinfoLink, the polysilicon (for multi) price breached its past-10-year low of mono wafers likely face CNY58/kg on 10 April 2019, and is down by 50% since 31 May 2018. We believe the less margin-squeeze recent price slump was driven mainly by new lower-cost polysilicon capacity commissioned pressure given a by GCL-Poly, Tongwei and Xinte. As the current polysilicon price is lower than major plants’ potential price rally for cash costs, we see limited further downside for polysilicon prices and consider a price rally polysilicon in 2H19 to be very likely in 2H19E.

Polysilicon (for multi) ASP (USD/kg) Multi- and mono-wafer ASP (US D/piece)

18.5 0.80 17.5 16.5 0.70 15.5 0.60 14.5 13.5 0.50 12.5 0.40 11.5 10.5 0.30 9.5 8.5 0.20

7.5

Jul-17 Jul-18

Jan-18 Jan-19

Mar-18 Mar-19

Nov-18 Sep-17 Nov-17 Sep-18

May-17 May-18 May-19

Jul-17 Jul-18

Apr-18 Oct-17 Oct-18 Apr-19

Jun-17 Jan-18 Jun-18 Jan-19

Mar-18 Feb-18 Feb-19 Mar-19

Nov-18 Nov-17 Dec-17 Dec-18

Sep-17 Aug-18 Sep-18

Aug-17 Mono Multi

May-18 May-19 May-17 Source: Bloomberg Source: Bloomberg

The average price gap between mono- and multi-use polysilicon was CNY8.2/kg in 2018. The current price gap is 1.1x larger, at CNY17/kg (based on mono/multi-wafer polysilicon prices of CNY76 and CNY59 per kg, respectively / USD9.0 and USD6.9 per kg, respectively on 3 July 2019). We think a potential recovery in multi-use polysilicon prices in 2H19E would first put pressure on multi-wafer gross margins, while the mono-wafer margin squeeze would be less than for multi products as the price gap between mono and multi polysilicon should narrow first to the 2018 average of CNY8/kg from CNY17/kg currently. That said, mono-wafer gross margins would be affected only mildly, in our view, as the upside potential for polysilicon (mono) prices would not be as great as that for polysilicon (multi) prices.

We prefer wafer Among the 3 key players in the wafer segment, GCL-Poly, LONGi and Zhonghuan, we manufacturers with note that LONGi and Zhonghuan have much lower gearing and thus stronger balance limited solar farm sheets than GCL-Poly, thanks to LONGi and Zhonghuan’s limited exposure to solar farms. exposure and strong Besides GCL-Poly has a significant cash flow burden from its solar farm assets (most of balance sheets which are under GCL New Energy) due to subsidy collection delays, in our view.

China: wafer manufacturers’ net-debt-to-equity ratio (2014-18) 250% 223.3% 192.0% 200% 174.4% 144.6% 146.6% 150% 84.2% 100% 69.5% 67.3% 41.7% 50% 32.7% 11.9% 8.2% 0% -6.5% -16.5% -23.9% (50%) 2014 2015 2016 2017 2018 LONGi Zhonghuan GCL-Poly Source: Wind, Daiwa estimates Note: Bloomberg consensus for Zhonghuan

A comparatively strong balance sheet can enable the company to undertake rapid capacity expansion without putting its liquidity under too much pressure before retail parity kicks in in 2021E. Both LONGi and Zhonghuan have aggressive expansion plans in the coming years, with LONGi’s wafer production capacity expected to increase from 28GW as at end- 2018 to 65GW in 2021 and Zhonghuan’s wafer production capacity from 25GW as at end-

China Wind and Solar: 11 July 2019

2018 to 50GW in 2021. Meanwhile, GCL-Poly has no expansion plans in the wafer segment, which we think is due to its relatively weak balance sheet as well as the industry shift from multi wafers to mono wafers.

Global: market share of leading players in the wafer segment Global: market share of leading players in the wafer segment (2018) (2021E)

LONGi 17% LONGi Others 29% 36% Zhonghuan Others 15% 50% Zhonghuan GCL-Poly GCL-Poly 22% 18% 13%

Source: Company, Daiwa estimates Source: Company, Daiwa estimates

Cells: focus on improving cell efficiency The cell market is China is the largest solar-cell production base globally, with its production volume fragmented accounting for over 60% of the global production volume. According to the CPIA, China produced 85GW of solar cells in 2018, a rise of 18% YoY.

Global: solar cell production volume (GW) 120

100 29 32 80

60 24 21 40 17 85 15 72 15 17 51 20 14 33 41 20 21 25 11 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 China Others

Source: CPIA, Daiwa estimates

Leading cell manufacturers are mainly from China and Taiwan. We note that the market for solar cells is highly fragmented, with the top player, Tongwei, accounting for less than 10% of the market as of end-2018.

Global: market share of solar cell production capacity (2018)

Tongwei Hanwha 6% 6% JA Solar 5%

Canadian Solar 4% Others Trina Solar 4% 48% Aiko Solar 5% URE 4% Jinko Solar 5% Shunfeng Uniex Solar (including Longi Motech 3% Suntech) 3% 3% 4%

Source: Company, Daiwa estimates

China Wind and Solar: 11 July 2019

Material cost reductions In order for solar power to become economically appealing compared with traditional and efficiency power sources like coal-fired power, retail and grid parity are needed. These conditions enhancement are 2 key largely depend on a decline in unit investment for solar projects. Material cost reductions factors in bringing down and efficiency gains are 2 ways to get there. unit investment costs for solar projects Material cost reductions effectively means lowering costs in various segments along the whole solar product industry chain (from polysilicon and wafers to cells and modules), which together brings down unit investment costs for solar project construction.

Forecasts for mono PERC module cost decline 1.6 1.4 1.2 0.68 1.0 0.63 0.61 0.59 0.57 0.8 0.54 0.6 0.34 0.31 0.3 0.29 0.4 0.27 0.25 0.22 0.19 0.17 0.2 0.16 0.14 0.13 0.21 0.19 0.17 0.15 0.13 0.12 0.0 2018 2019E 2020E 2021E 2023E 2025E Polysilicon cost (CNY/w) Wafer non-silicon cost (CNY/w) Cell non-silicon cost (CNY/w) Module non-silicon cost (CNY/w)

Source: CPIA Note1: CPIA does not give forecasts for 2022E and 2024E. Note2: Daiwa thinks polysilicon cost for 2020E would maintain stable

Efficiency enhancements effectively mean increasing the IPCE (monochromatic incident photon-to-electron conversion efficiency) of the solar cell and thus increasing capacity of the solar module through technological improvements. In this way, the unit investment of solar projects can be indirectly reduced. Solar cell efficiency refers to the portion of energy in the form of sunlight that can be converted by the solar cell via photovoltaics into electricity. For example, a solar cell with 20% efficiency and an area of 1 m2 will produce 200W under standard test conditions (1000W/m2 insolation). Thus, for a solar cell with a size of 156mm*156mm and 20% efficiency, its capacity is 4.86W. A solar module usually consists of 60 solar cells, and its capacity is largely determined by the efficiency of the solar cells.

The reason efficiency gains can reduce solar plant unit investment is that a large proportion of the total solar project investment is related to the area of the module, such as the cost of cables, mounting systems, land, instalment fee, and maintenance. Therefore, a higher-capacity module can effectively dilute the area-related cost and reduce the unit investment of solar projects per watt. For example, if we assume the area-related construction cost is CNY500 per piece of module, the area-related cost will be CNY1.82/W (500/275=1.82) for a 275W multi-silicon module, and CNY1.61/W (500/310=1.61) for a 310W mono-PERC module (an 11% drop).

If we further assume total installation cost is CNY5/w for a solar project using a 275W multi module and do not consider the price difference between multi and mono-PERC modules, the above-mentioned area-related cost cut could lead to a 4% reduction in total installation cost. The dilution effect will become greater when land and staff costs continuously increase in the future. Furthermore, a higher-efficiency mono-PERC module can produce 3% more electricity than a traditional multi-module, even with the same capacity, due to better absorption of weak light and lower working temperature, meaning utilisation is higher and is positive for a lower LCOE. Therefore, we can see that efficiency enhancement is also an irreversible future development trend for solar cells, apart from direct solar material cost reduction.

China Wind and Solar: 11 July 2019

Sensitivity analysis: area-related cost drop by using higher Sensitivity analysis: total installation cost drop by using higher efficiency module vs. 275W multi module efficiency module vs. 275W multi module Area-related cost per piece of 60-cell module (CNY) Area-related cost per piece of 60-cell module (CNY) 500 600 700 800 900 1000 500 600 700 800 900 1000 275 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 275 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 280 -1.8% -1.8% -1.8% -1.8% -1.8% -1.8% 280 -0.6% -0.7% -0.8% -0.9% -0.9% -1.0% 285 -3.5% -3.5% -3.5% -3.5% -3.5% -3.5% 285 -1.3% -1.4% -1.6% -1.7% -1.8% -1.9% Module power 290 -5.2% -5.2% -5.2% -5.2% -5.2% -5.2% Module power 290 -1.9% -2.1% -2.3% -2.5% -2.6% -2.8% output (W) 295 -6.8% -6.8% -6.8% -6.8% -6.8% -6.8% output (W) 295 -2.5% -2.8% -3.0% -3.2% -3.4% -3.6% 300 -8.3% -8.3% -8.3% -8.3% -8.3% -8.3% 300 -3.0% -3.4% -3.7% -4.0% -4.2% -4.4% 305 -9.8% -9.8% -9.8% -9.8% -9.8% -9.8% 305 -3.6% -4.0% -4.4% -4.7% -5.0% -5.2% 310 -11.3% -11.3% -11.3% -11.3% -11.3% -11.3% 310 -4.1% -4.6% -5.0% -5.4% -5.7% -6.0%

Source: Daiwa estimates Source: Daiwa estimates Note: assume total installation cost is CNY5/w for a solar project using 275W multi module with area-related construction cost CNY500 per piece of module and do not consider the price difference between multi and higher efficiency module

There are many Solar cell technologies are being developed very quickly. There are many technologies technologies today for today for manufacturing and structuring the silicon solar cell, each with different manufacturing solar characteristics and conversion efficiency. BSF cells are considered 1st generation cells, each with different traditional technology, and PERC cells are gradually becoming the focus of the market (the characteristics and 2nd generation). PERC cells are being improved through developments such as PERT and conversion efficiency TOPCON (2.5G), and there are newer technologies such as HIT (considered 3rd generation) and IBC (still largely at the laboratory stage) on the way.

Solar cell technologies Conversion efficiency Generation of Abbreviation Name Description Advantage (as of 2018) technology A traditional solar cell composed of a two-sided silicon Aluminium layer with a single P-N junction, doped with phosphorus Traditional multi: 18.7% Al-BSF back surface and boron. In order to increase solar cell efficiency, after Traditional mature technology Multi black-silicon: 19.2% 1G field making P-N junction, an aluminium film is deposited at Mono: 20.4% the rear surface of the wafer to make a P+ layer The PERC technology can be applied to either multi or The upfront side of the cell is identical to an Al-BSF cell, mono wafers. Manufacturing procedures are not much but a passivation film (or dielectric layer) with tiny holes different from traditional solar cells as only the rear Passivated is added at the rear surface that allows the transmitted Multi black-silicon: 20.3% surface is modified by simply adding a dielectric layer PERC emitter and photons to be reflected back to the silicon layer for a Quasi mono: 21.6% 2G and using lasers to open the holes. This means the rear contact second chance to get absorbed. This increases the Mono: 21.8% solar panels have a higher energy output with little number of photons that release electrons in the solar investment and few risks. Operating cell voltages cell and generates more electricity increase and therefore power outputs increase as well Passivated Improvement on PERC technology. Unlike standard Improvement on PERC technology with limited PERT emitter rear cells and PERC, which both use an aluminium-alloy 2.5G additional cost. Easy to achieve large-scale production totally-diffused BSF, PERT cells have a diffused rear surface In order to reduce recombination losses from the metal N-PERT+TOPCON: 21.5% Tunnel oxide contacts and the silicon surface of cells, a thin oxide TOPCON passivated Improvement on N-PERT cell 2.5G layer is introduced, in addition to heavily doped contact polycrystalline silicon between metal contacts and wafer An HJT solar cell is composed of a mono thin crystalline Hetero-junction silicon (c-Si) wafer surrounded by ultra-thin amorphous Totally new technology with very good passivation HJT (or HIT) with intrinsic silicon (a-Si) layers, which act as an effective surface 22.5% 3G effect and high conversion efficiency thin-layer passivation layer for c-Si wafer. The p+/n+ doped a-Si functions as an effective emitter/BSF for the cell Instead of placing the contacts in the front of the cell, an IBC cell places them on its rear side. This allows it to Interdigitated achieve higher efficiency due to reduced shading on the Higher effective power generation area, which is IBC 23.4% Laboratory stage back contact front of the cell, while at the same time electron-hole beneficial for efficiency increase pairs generated by the absorbed light can still be collected on the rear side of the cell.

Source: CPIA, Instyle Solar, pv magazine, Wikipedia, Daiwa

China Wind and Solar: 11 July 2019

Different cell technologies

Silicon solar cell

Multi cell Mono cell

BSF multi cell P-type mono cell N-type mono cell (IPCE>18.7%)

P-type BSF multi N-PERT + TOPCON P-type BSF mono black-silicon cell mono cell (front cell (IPCE>20.4%) (IPCE>19.2%) IPCE>21.5%)

P-type multi P-type mono N-type HJT mono black-silicon PERC PERC cell cell (IPCE>22.5%) cell (IPCE>20.3%) (IPCE>21.8%)

P-type quasi- N-type IBC mono mono PERC cell cell (IPCE>23.4%) (IPCE>21.6%)

Source: CPIA, Daiwa Note1: P-type quasi-mono PERC cell and N-type IBC mono cell are currently at the pilot stage. Note2: Colours in the chart refers to generations of cell: yellow (1G), green (2G), orange (2.5G), red (3G), and purple (lab stage).

Each technology has its own potential to further improving its conversion efficiency. In 2018, the conversion efficiency of multi black-silicon cells reached 19.2%, while that of mono PERC cells and multi PERC cells increased to 21.8% and 20.3%, up by 0.5pp and 0.3pp compared with 2017. The conversion efficiency of N-type PERT mono cells and HIT mono cells, which had just entered the production stage, reached 21.5% and 22.5%, respectively. According to the CPIA, the respective conversion efficiency of each technology can be expected to increase further in the coming years.

Conversion efficiency forecast for different cell technologies 2018 2019E 2020E 2021E 2023E 2025E BSF P-type multi black-silicon cell (1G) 19.2% 19.4% 19.7% 19.9% 20.2% 20.5% Multi cell PERC P-type multi black-silicon cell (2G) 20.3% 20.5% 20.8% 21.1% 21.3% 21.6% PERC P-type quasi-mono cell (2G) 21.6% 21.8% 22.2% 22.4% 22.6% 22.8% P-type mono cell PERC P-type mono cell (2G) 21.8% 22.1% 22.4% 22.6% 22.8% 23.0% N-PERT+TOPCON mono cell (front) (2.5G) 21.5% 22.0% 22.5% 23.0% 23.5% 24.0% N-type mono cell HJT N-type mono cell (3G) 22.5% 23.0% 23.5% 24.0% 24.5% 25.0% IBC N-type mono cell 23.4% 23.6% 23.8% 24.3% 24.6% 25.0%

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E. P-type quasi-mono PERC cell and N-type IBC mono cell are currently at the pilot stage.

Currently, mainstream In 2018, BSF cells accounted for 60% of the market, a decline of 23pp YoY, and their solar cell production share is likely to decline further as new technologies come to the fore. Among all the high- technology shifting from efficiency cell technologies (such as PERC, PERT, HJT, and IBC), PERC cells currently BSF to PERC cells have the largest production capacity. The market share of PERC ballooned from 15% in 2017 to 33.5% in 2018, and the CPIA expects PERC’s market share to exceed 50% in 2019 as the technology surpasses BSF to become the most popular cell technology. The solar cell segment is in the throes of shifting from BSF (1G) to PERC (2G), as PERC is more efficient and it is relatively straightforward to shift production lines from BSF to PERC. Besides, the CPIA expects the market share of other high-efficiency cell technologies, such as N-PERT, HJT and IBC to increase gradually.

China Wind and Solar: 11 July 2019

Forecast market share of different cell technologies

100% 0.2%0.0% 0.1%0.7% 0.2%0.8% 0.8% 0.9% 1.0% 1.0% 1.5% 5.0%0.7% 1.0% 3.0%0.5% 5.0%1.0% 3.0% 5.0% 15.0% 8.0% 10.0% 7.5% 15.0% 10.0% 80% 17.0% 33.5% 18.0% 50.6% 60% 55.7% 60.8% 40% 83.3% 60.9% 61.0% 60.0% 20% 39.4% 30.0% 17.3% 10.6% 0% 5.0% 2017 2018 2019E 2020E 2021E 2023E 2025E BSF (1G) PERC (2G) N-PERT (2.5G) HJT (3G) IBC Others

Source: CPIA Note: CPIA does not provide forecasts for 2022E and 2024.

Demand for cells of different technologies (GW) 160 140 120 100 80 60 40 20 0 2017 2018 2019E 2020E 2021E Total cell demand BSF (1G) PERC (2G) N-PERT (2.5G) HJT (3G) IBC Others

Source: CPIA, BNEF, Daiwa estimates

China Wind and Solar: 11 July 2019

Solar cell segment: shifting from BSF (1G) to PERC (2G)

BSF (1G) PERC (2G)

PERT (2.5G) HJT (3G)

Source: ISFH, Semantic Scholar (Wei Wen Luo et al., Investigation of Potential-Induced Degradation in n-PERT Bifacial Silicon Photovoltaic Modules with a Glass/Glass Structure [2018]), ResearchGate (Atteq ur Rehman and Soo Hong Lee, Advancements in n-Type Base Crystalline Silicon Solar Cells and Their Emergence in the Photovoltaic Industry [2013]), Daiwa

We prefer cell Within the cell segment, we see technology upgrades and product efficiency manufacturing enhancements as a clear trend. Currently, the cell market is fragmented. To gain market equipment producers share, a cell manufacturing company needs to continuously step up capex and invest in new capacity and high-efficiency technology, in our view. As the mainstream cell technology is shifting from BSF to PERC, many cell companies have been significantly increasing PERC capacity. According to the CPIA, newly-built cell capacity in 2018 was exclusively PERC-related, while existing BSF production lines are accelerating their upgrades to PERC production lines. If in the longer term, another technology becomes mainstream technology (likely HJT), we see the potential for another round of capacity investment and upgrades.

PERC cell capacity of leading cell manufacturers (GW)

18 16.5 16 14 12 11.0 10.0 9.8 10 8.4 7.5 8 7.0 5.0 5.0 4.6 6 4.0 4.0 4 2 0 Tongwei Runyang LONGi Aiko Solar JA Solar Lu'an Jinko Solar Suming Jiayue Risen Solar Canadian Trina Solar Yueda Solar 2016 2017 2018 2019E

Source: Company, Daiwa

China Wind and Solar: 11 July 2019

Meanwhile, for different generations of cell technology, the basic production equipment is pretty similar. We think the cell manufacturing equipment segment stands to benefit from the continuous capex of cell manufacturers, driven by upgrades to existing production lines and the addition of new lines with the coming of retail and grid parity, coupled with increasing localisation of cell manufacturing equipment. In our opinion, these developments should be positive for cell manufacturing equipment producers like Shenzhen SC New Energy (300724 CH), which produces equipment for saw damage removal, surface texturisation, diffusion, edge isolation, PECVD, and screen printing.

Cell technologies: major techniques BSF PERC PERT HIT Major Surface texturisation, Surface texturisation, diffusion, Surface texturisation, diffusion, Surface texturisation, techniques diffusion, edge isolation, edge isolation, laser, PECVD edge isolation, laser, PECVD diffusion, edge isolation, TCO PECVD, screen printing (bifacial), screen printing (bifacial), screen printing making, screen printing

Source: ISFH, Daiwa

Module: another fragmented market The market for modules As with cells, China is the largest solar module production base in the world. According to is also fragmented the CPIA, China produced 84GW of solar modules in 2018, up by 12.4% YoY, and accounted for over 72% of global production volume.

Global: solar module production volume (GW) 140

120

100 32 31 80 24 60 18 40 16 84 14 75 14 14 54 20 46 10 36 21 23 27 11 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 China Others

Source: CPIA, Daiwa

The market for solar modules is highly fragmented, with the leading players like Trina Solar, Jinko Solar and Hanwha each claiming market shares of less than 10%.

Global: market share of solar module production capacity (2018)

Trina Solar, 8% Jinko Solar, 7%

Hanwha, 7% Others, 41% Longi, 6%

Canadian Solar, 6%

JA Solar, 6% Suntech, 3% Yingli, 3% Risen Solar, 4% Talesun Solar, 3% HT Solar, 4% GCL System, 4%

Source: Company, Daiwa estimates

Module power output The power output (capacity) of a module depends on the conversion efficiency of the cells depends on cell as well as the cell-to-module (CTM) ratio (which is the ratio of module capacity to the sum efficiency and the cell- of cell capacity). Thus, besides increasing cell efficiency, using new advanced module to-module ratio technologies (such as half-cell configuration, shingled layering, MBB, and bifacial module) can also increase the CTM ratio and module capacity.

China Wind and Solar: 11 July 2019

60-cell module power output from different types of cells (W) 2018 2019E 2020E 2021E 2023E 2025E BSF multi module 275 280 285 290 295 300 Multi module Black-silicon multi module 280 285 290 295 300 305 Black-silicon PERC multi module 295 300 305 315 320 325 BSF mono module 285 290 295 300 303 305 Mono module P-PERC mono module 305 310 315 325 330 340 N-PERT mono module (front) 310 320 325 330 340 345 N-type mono module HJT N-type mono module 320 330 335 340 350 355 IBC N-type mono module 340 345 350 355 360 365 MWT black-silicon multi module 305 310 315 325 330 340 MWT (Metal Wrap Through) packaging method (with PERC cell) MWT mono module 315 320 325 335 340 350

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E.

Half-cell modules and Half-cell modules feature solar cells that are cut in half. Traditional 60-cell panels will have shingled modules are 120 half-cut cells. When solar cells are halved, their current is also halved, so resistive new advanced module losses are lowered and the cells can produce a little more power. Also, half-cell module technologies technology allows the panel to be separated into 2. Having independent upper and lower module halves helps reduce the shading effect (ie, if the bottom half of a module is shaded, the top half will still perform). Currently, half-cell technology is relatively mature compared with other new module production technologies, meaning half-cell technology can increase module capacity by 5-10W (2-4%) compared with traditional modules.

Half-cell module Half-cell module vs. full-cell module

Half cell module Full cell module

Traditional cell Half-cut cell

Half-cell module

Source: SolarBe, Daiwa Source: SolarBe, Daiwa

Shingled modules are another technology that aims to increase the CTM ratio. In traditional modules, conductive ribbons or busbars are used to connect solar cells together into strings, and there is a slight gap between solar cells to allow for these connections. Shingled technology omits ribbons and instead directly connects stripes of solar cells by “tiling” the solar cells (similar to how shingles are installed on a roof). The result: the 2mm gap between cells is eliminated, which allows for more cells to be used in a module. As the cell spacing area is avoided and this leads to a larger active cell area within the module, the efficiency of the module increases.

Shingled modules can increase module power by 15-20W (c.7%) compared with traditional modules by virtue of having more cells within a module (66-cell vs. 60-cell conventional module) and being less prone to the shading effect by using conductive adhesive to replace ribbons.

Traditional module vs. shingled module

Source: ResearchGate (Max Mittag et al., Cell-to-Module [CTM] Analysis for Photovoltaic Modules with Shingled Solar Cells [2017])

China Wind and Solar: 11 July 2019

Market share of half-cell Shingled technology is complex and not yet mature. Moreover, it is incompatible with and shingled modules existing module production capacity, making it difficult to upgrade capacity to shingled should gradually technology. Notwithstanding these hurdles, the CPIA forecasts that the market share of increase shingled modules will gradually increase. In the meantime, CPIA forecasts that half-cell module will become the mainstream by 2025.

Market share of different module production technologies

100% 1.0%0.5% 0.8% 3.8% 7.5% 8.8% 13.8% 90% 16.7% 19.5% 23.8% 80% 25.5% 70% 35.0% 60% 39.5% 42.5% 50% 98.5% 91.7% 40% 79.5% 30% 65.7% 51.3% 20% 41.0% 33.8% 10% 0% 2017 2018 2019E 2020E 2021E 2023E 2025E Full-cell module Half-cell module Shingled module

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E.

MBB (multi-busbar) technology involves changes to both cells and modules. With MBB, a solar cell is equipped with over 10 busbars instead of 4 or 5. Increasing the number of busbars not only reduces the distance travelled by the current, but reduces the current carried by each busbar, resulting in smaller resistance losses and greater conversion efficiency. The width of the busbar can be narrower, which not only reduces the consumption of the silver paste (5-10% lower) but lowers the shading area of the busbar. MBB cells also have an advantage of low cracking, while absorption of light increases due to the reduction of shading based on the round wire design. With MBB technology, module capacity could be increased by 2-3W according to the CPIA.

Market share of modules with different numbers of busbars 100% 2.4% 2.4% 8.0% 3.8% 7.2% 13.4% 2.0% 15.8% 16.7% 23.3% 80% 28.8%

60% 60.0% 53.3% 85.0% 60.0% 40% 80.8% 66.7% 63.9% 20% 33.3% 30.0% 23.3% 8.8% 10.0% 0% 1.0% 0.1% 2017 2018 2019E 2020E 2021E 2023E 2025E 4BB 5BB MBB Others

Source: CPIA Note: CPIA does not provide forecasts for 2022E and 2024E

China Wind and Solar: 11 July 2019

MBB vs. 5BB module MBB vs. 5BB module

Source: Yingli Source: IBC Solar

The CTM ratio could According to the CPIA, the CTM ratio could further increase going forward on the back of further increase going further improvements to module production technologies. forward CTM ratio 102% 101% 100% 99% 98% 97% 96% 95% 94% 2018 2019E 2020E 2021E 2023E 2025E Mono CTM Multi CTM

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E

Another important development in the module segment is the bifacial module, which can realise bifacial power generation without significantly increasing cost, and achieve a power generation gain of 10-25%. The CPIA forecasts that the market share of bifacial modules will expand from 10% in 2018 to 20% in 2019E. Furthermore, on 13 June 2019, the US Federal trade authorities ruled that bifacial solar modules are no longer subject to the Section 201 ruling. The tariff waiver should help boost shipments to the US and accelerate the rise in the market share of bifacial modules, in our view.

Market share of single-sided modules vs. bifacial modules 100% 2% 10% 20% 30% 80% 40% 48% 60% 60% 98% 90% 40% 80% 70% 60% 52% 20% 40%

0% 2017 2018 2019E 2020E 2021E 2023E 2025E Single-sided module Bifacial module

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E.

China Wind and Solar: 11 July 2019

From the above analysis, it is clear that technology upgrades and product efficiency enhancements will be trends in the module segment. Similar to the cell segment, the module market is very fragmented. In order for a module manufacturing company to expand its market share, it needs to continuously introduce more capex and invest in new capacity and high-efficiency technology, which in turn creates demand for module manufacturing equipment.

Solar glass: likely to see volume and ASP recovery in 2019 The penetration rate of As discussed in the module segment, the penetration rate of bifacial modules looks set to double-glass bifacial increase, due mainly to: 1) the promise of 5-30% increases in power output, and 2) the modules looks set to compatibility of bifacial technology with high-efficiency cells, including mainstream P-PERC increase cells and emerging N-type mono cells and HJT cells.

Market share of single-sided modules vs. bifacial modules 100% 2% 10% 20% 30% 80% 40% 48% 60% 60% 98% 90% 40% 80% 70% 60% 52% 20% 40%

0% 2017 2018 2019E 2020E 2021E 2023E 2025E Single-sided module Bifacial module

Source: CPIA Note: CPIA does not give forecast for 2022E and 2024E.

Double-glass bifacial Solar glass, a key component of a solar module, protects the solar cells from outside modules need more forces like moisture. One of the biggest differences between double-glass bifacial modules solar glass and traditional modules is that double-glass modules use 2 pieces of 2.5mm solar glass for module packaging, whereas traditional single-sided modules use only 1 piece of 3.2mm solar glass for the front side and a back sheet for the rear.

Traditional one-glass module Double-glass module

3.2 mm solar glass 2.5 mm solar glass

EVA film EVA film Solar cell Solar cell

EVA film EVA film

Backsheet 2.5 mm solar glass

Source: pv.ally, Daiwa Source: pv.ally, Daiwa

According to our estimates, solar glass demand for double-glass modules (91kt/GW) is 56% higher than for single-glass modules (58kt/GW).

China Wind and Solar: 11 July 2019

Solar glass demand: traditional modules vs. double-glass modules (300W) Traditional solar module Bifacial double-glass module Glass thickness (mm) 3.2 2.5 Processed glass capacity per tonne of raw glass capacity (m2/tonne) 125 160 Yield rate of processed solar glass 75% 75% Processed glass production volume per tonne of raw glass capacity (m2/tonne) 94 120

Glass number per piece of solar module 1 2 Area of solar module (60-piece type) (m2) 1.635 1.635 Glass area per piece of solar module (m2) 1.635 3.270 Module production volume corresponding to 1 tonne of raw glass capacity (piece) 57 37 Solar module capacity (60-piece type) (W) 300 300

Solar installation corresponding to one tonne of raw glass capacity (kW) 17.2 11.0 Raw glass demand per 1GW solar module (tonnes) 58,133 90,833

Source: Company, Daiwa estimates

Solar glass demand We forecast global solar installation to increase from 107GW in 2018 to 125GW in 2019E, should be supported by 136GW in 2020E, and 150GW in 2021E, representing YoY growth of 16% in 2019E, 9% in global solar installation 2020E and 10% in 2021E. This expansion, coupled with rising bifacial module penetration growth and higher from 10% in 2018 to 40% in 2021E, should lead to solar glass demand rising from 6.7m penetration of double- tonnes in 2018 to 8.0m tonnes in 2019E, 8.8m tonnes in 2020E, and 10.0m tonnes in glass modules 2021E, representing YoY growth of 18% in 2019E, 11% in 2020E and 12% in 2021E, which is c.2pp higher than global solar installation growth in the respective years, on our forecasts.

Solar glass demand analysis 2017 2018 2019E 2020E 2021E Solar new installation (GW) 99 108 125 136 150 YoY 9.1% 15.7% 8.8% 10.3%

Penetration rate of bifacial double-glass module 6.6% 10.0% 20.0% 30.0% 40.0% New installation with traditional solar module (GW) 92 97 100 95 90 New installation with bifacial double-glass module (GW) 7 11 25 41 60

Module capacity (W) 285 295 305 315 325 Unit demand of solar glass - traditional module (tonnes/GW) 61,193 59,119 57,180 55,365 53,662 Unit demand of solar glass - double-glass module (tonnes/GW) 95,614 92,373 89,344 86,508 83,846

Solar glass demand - traditional module (m tonnes) 5.66 5.75 5.72 5.27 4.83 Solar glass demand - double-glass module (m tonnes) 0.62 1.00 2.23 3.53 5.03 Total solar raw glass demand (m tonnes) 6.28 6.74 7.95 8.80 9.86 YoY 7.3% 17.9% 10.7% 12.0%

Source: BNEF, CPIA, Daiwa estimates Note: We have considered the negative impact to solar glass demand resulting from increasing solar module power capacity

Xinyi Solar and Flat Market share in the solar glass segment is concentrated among the leading players, with Glass account for 50% of Xinyi Solar and Flat Glass, the top 2 players, accounting for 50% of the market in 2018. the market currently The players in the solar glass market have been in the industry for many years; there have been few newcomers in recent years. Hence, capacity expansion in the solar glass industry mostly depends on the top players. Thus, the combined market share of the 2 biggest players is likely to expand further, in our view.

Solar glass market share for operating capacity (2018) Solar glass market share for operating capacity (2020E) AVIC Sanxin, AVIC Sanxin, 5% Others, 5% 3% Others, 7% Ancai Hi-tech, Ancai Hi-tech, 4% 3% IRICO, 7% IRICO, 9% Xinyi Solar, 32% Xinyi Solar, 39% CSG, 9% CSG, 6%

CNBM, CNBM, 5% 10% Jinxin Flat Glass, Solar, 18% Flat Glass, 19% 11% Jinxin Solar, 8%

Source: Company, Daiwa Source: Company, Daiwa

China Wind and Solar: 11 July 2019

We expect solar glass As at end-2018, some 21kt/d of solar glass production capacity was in operation. Taking demand growth to into account 3.7kt/d of new capacity (including two 1,000t/d production lines from Flat exceed supply growth in Glass and 850t/d from IRICO), 2.4kt/d of capacity on which maintenance work is being 2019 completed (including 1,100t/d from Xinyi Solar and 600t/d from Flat Glass), coupled with 2.1ktpd of capacity that will go into maintenance during 2019, we estimate that operating capacity will increase to 25ktpd by end-2019. Besides, effective capacity should be 23ktpd for full-year 2019, a rise of 12% YoY — lower than our forecast for demand growth of 18% YoY in 2019.

Solar glass supply vs. demand 2017 2018 2019E 2020E 2021E Year -end nominal solar glass capacity (t/d) 21,290 23,290 26,940 30,940 30,940 YoY 9.4% 15.7% 14.8% 0.0% Year-end solar glass capacity in operation (t/d) 20,090 20,890 24,870 29,190 30,940 YoY 4.0% 19.1% 17.4% 6.0%

Newly constructed capacity (t/d) 3,360 2,000 3,650 4,000 - Capacity into maintenance (t/d) 1,200 2,400 2,070 2,350 - Capacity out of maintenance (t/d) 1,200 2,400 2,070 2,350

Utilisation rate assumption for new capacity or 50% 50% 50% 50% 50% capacity into/out of maintenance Effective solar glass capacity (t/d) 19,010 20,490 22,880 26,730 29,765

Solar raw glass supply (m tonnes) 6.94 7.48 8.35 9.76 10.86 YoY 7.8% 11.7% 16.8% 11.4% Solar raw glass demand (m tonnes) 6.28 6.74 7.95 8.80 9.86 YoY 7.3% 17.9% 10.7% 12.0% Supply-demand difference (m tonnes) 10% 11% 5% 11% 10%

Source: Daiwa estimates Note: new capacity is based on the capacity addition schedule that have already announced by various solar glass companies; capacity maintenance is our own estimate based on construction time of existing capacity

We expect the relatively tight supply-demand situation to support an ASP recovery from the low level of 2018. The ASP for processed 3.2mm solar glass has risen from CNY24/m2 at end-2018 to CNY26.3/m2 currently, and we believe it will increase further once China’s demand kicks off in 2H19E.

3.2mm processed solar glass ASP (CNY/m2) 34

Oct-15 Apr-16 Oct-16 Apr-17 Oct-17 Apr-18 Oct-18 Apr-19

Jun-15 Jun-16 Jun-17 Jun-18 Jun-19

Feb-16 Feb-17 Feb-18 Feb-19

Dec-15 Dec-16 Dec-17 Dec-18

Aug-16 Aug-17 Aug-18 Aug-15 Source: SCI99, Daiwa

We think solar glass Meanwhile, as the market leaders aggressively add new capacity (Xinyi Solar intends to supply may loosen in add 4 1,000t/d production lines in 2020), we expect effective capacity to increase to 2020, which could see 2.7ktpd for full-year 2020, a rise of 16.8% YoY — higher than our forecast demand growth solar glass prices come of 10.7% YoY. Hence, we believe solar glass supply may become loose again in 2020, under pressure which could put pressure on solar glass prices in 2020.

China Wind and Solar: 11 July 2019

Summary for solar product makers: factors to consider As the solar sector has a long value chain, we suggest being selective on solar companies based on the following criteria: 1) supply-demand situation, 2) competitive landscape, 3) low downstream solar project exposure, 4) technology shift to high-efficiency products, 5) healthy balance sheets, and 6) undemanding valuations. Apart from the financial situations and valuations of individual companies, we favour the following product segments:

Product segments that we prefer Product segment Description Polysilicon Oversupply situation will start to ease with slower capacity expansion in 2H19-21. Market is more consolidated compared with cell and module. Mono wafer Oligopolistic market with high market power for leading players. Future trend of chasing high-efficiency products with increasing market share of mono products. PERC cell Future trend of chasing high-efficiency products with increasing market share of PERC cell technology. Solar glass Supply growth is likely to be lower than demand growth in 2019. Market is relatively consolidated with high market share for leading players.

Source: Daiwa a. Supply vs. demand With the finalisation of China’s solar policy, China’s solar installation demand looks set to kick off in 2H19, and we expect solar installation to remain stable for 2019-20 before recovering in 2021. In the meantime, overseas markets are steadily expanding. Thus, we forecast global solar installation to increase from 107GW in 2018 to 125GW in 2019, 136GW in 2020, and 150GW in 2021, representing YoY growth of 16% in 2019, 9% in 2020 and 10% in 2021 and a CAGR of 10% over 2019-21E.

Source: NEA, Daiwa estimates and forecasts

Global: solar annual installations (MW) 160,000 60% 55% 2019-21E CAGR: 10% 140,000 50%

120,000 40%

100,000 30% 28% 24% 80,000 20% 20% 17% 12% 60,000 10% 9% 10% 10%

40,000 0% -7% 20,000 (10%)

0 (20%) 2012 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E China US Japan India Europe Rest of the world YoY growth (RHS)

Source: NEA, BP, BNEF, Daiwa estimates

China Wind and Solar: 11 July 2019

We think the polysilicon Meanwhile, supply conditions differ in each segment of the solar product value chain. In segment will be in a the polysilicon segment, most of the new, lower-cost capacity was commissioned in 2H18 better position and 1H19. We think further capacity expansion in the segment will be limited in 2H19-21, compared with other with capacity growth seen slowing from an 21% CAGR in 2017-19E to a 5% CAGR in solar products in the 2019-21E, as new capacity entails large amounts of capex. near term Our forecast polysilicon capacity CAGR of 5% for 2019-21E is lower than our forecast global solar installation CAGR of 10% for the same period, whereas our forecast wafer capacity CAGR of 11% for 2019-21E is broadly in line with our forecast global solar installation CAGR. Finally, our forecast cell/module capacity CAGR of 15% over 2019-21E is higher than our forecast global solar installation CAGR.

Hence, we believe that the polysilicon segment, with easing oversupply and modest capacity expansion, will be better positioned compared with other solar products in 2H19- 2021E.

Global polysilicon production capacity (‘000 tonnes) Global wafer production capacity (GW) 900 831 841 40% 250 228 40% 2019-21E CAGR: 5% 2019-21E CAGR: 11% 756 205 35% 800 35% 200 184 700 168 30% 599 30% 147 25% 600 150 515 25% 475 107 20% 500 90 20% 100 15% 400 356 68 275 15% 10% 300 50 5% 200 10% 0 0% 100 5% 2014 2015 2016 2017 2018 2019E 2020E 2021E Capacity YoY 0 0% 2014 2015 2016 2017 2018 2019E 2020E 2021E Capacity YoY (RHS)

Source: Company, Daiwa estimates Source: Company, Daiwa estimates

Global cell production capacity (GW) Global module production capacity (GW) 300 40% 300 45% 2019-21E CAGR: 16% 253 2019-21E CAGR: 15% 35% 244 40% 250 236 250 212 35% 30% 200 188 200 184 30% 25% 147 153 25% 150 20% 150 125 115 20% 15% 84 90 100 80 86 100 80 15% 65 10% 10% 50 50 5% 5%

0 0% 0 0% 2014 2015 2016 2017 2018 2019E 2020E 2021E 2014 2015 2016 2017 2018 2019E 2020E 2021E Capacity YoY Capacity YoY Source: Company, Daiwa estimates Source: Company, Daiwa estimates

Capacity vs. installation growth 30%

25%

20%

15%

10%

0% 2019E 2020E 2021E Polysilicon Wafer Cell Module Global solar installation

Source: BNEF, Daiwa estimates

China Wind and Solar: 11 July 2019

As for solar glass, we expect the supply-demand situation to be a bit tighter in 2019 before loosening again in 2020.

Solar glass demand-supply 12 20%

10 15% 8

6 10%

4 5% 2

0 0% 2018 2019E 2020E 2021E Solar raw glass supply (mn tonnes) Total solar raw glass demand (mn tonnes) Supply YoY Demand YoY

Source: BNEF, Daiwa estimates

It is hard to increase From a capex and construction duration point of view, it appears that it is more difficult to capacity in the increase capacity in the polysilicon segment within a short period as compared with other polysilicon segment segments. Currently, the capacity of a new polysilicon plant tends to exceed 10,000 tonnes within a short period pa, meaning that the investment threshold for a polysilicon plant is around CNY1bn. By comparison, the investment thresholds for other segments are CNY68m-420m/GW pa for ingots, wafers, cells and modules, and CNY600-700m/ktpd for solar glass. The construction period for a polysilicon plant is also longer than for other segments. Therefore, we think that most of the leading players in the polysilicon segment are unlikely to initiate another round of large-scale capacity expansion in the near term, given their capacity expansion undertaken in 2H18-1H19.

Capex and construction period along the solar product chain as of 2018 Capex Construction period Polysilicon CNY1,150m per 10,000 tonnes pa (equivalent to CNY414m/GW pa) 1 year Ingot Multi ingot: CNY28m/ktpa (c.CNY100m/GW pa); mono ingot: CNY65m/ktpa (c.CNY250m/GW pa) < 1 year Wafer (Diamond-wired) CNY80-90m/GW pa < 1 year Cell (PERC) CNY250-420m/GW pa 6-9 months Module CNY68m/GW pa 6-9 months Solar glass CNY600-700m/ktpd (equivalent to CNY96-111m for 1GW pa one-glass module and CNY149- 9-12 months 174m for 1GW pa double-glass module)

Source: CPIA, company, Daiwa

b. Competitive landscape Competitive landscapes We believe that leading companies operating in segments where market fragmentation is in the polysilicon, wafer low are the best positioned in terms of competition. And we contend that the competitive and solar glass landscapes of the polysilicon, wafer and solar glass segments are more favourable than for segments look more the cell and module segments. appealing than for cells and modules

Polysilicon market share (2019E) Wafer market share (2019E)

GCL-Poly, 16% Others, 22% LONGi , 20% Hemlock, 1% REC, 2% Wacker, 11% Others, 48% Asia Silicon, 3% Zhonghuan , Tongwei, 16% China Silicon, 11% 3% Daqo, 5% OCI, 10% GCL-Poly, 16% East Hope, 8% Xinte (TBEA), 9%

Source: Company, Daiwa estimates Source: Company, Daiwa estimates

China Wind and Solar: 11 July 2019

Cell market share (2019E) Module market share (2019E) Tongwei 10% Trina Solar, 8% Hanwha 5% Jinko Solar, 8%

JA Solar 6% Hanwha, 5% Others Canadian Solar Others, 38% 42% 5% Longi, 9% Trina Solar 4% Canadian Aiko Solar Solar, 6% 6% URE Suntech, 4% Shunfeng 3% JA Solar, 5% Yingli, 2% Risen Solar, Longi (including Jinko 5% Motech Uniex Solar Suntech) Solar Talesun Solar, GCL System, 6% 3% 3% 3% 5% 3% HT Solar, 3% 3%

Source: Company, Daiwa estimates Source: Company, Daiwa estimates

Solar glass market share (2019E)

AVIC Sanxin, 1%

Ancai Hi-tech, 4% Others, 8%

IRICO, 11% Xinyi Solar, 31% CSG, 5% CNBM, 8% Flat Glass, 23%

Jinxin Solar, 8%

Source: Company, Daiwa estimates

c. Low downstream solar project exposure We prefer companies Given the large deficit in China’s Renewable Energy Fund (we estimate over CNY200bn), with low downstream companies with large exposure to downstream solar projects are at risk of subsidy solar project exposure collection delays, increasing accounts receivable days, worsening cash flow, rising and strong balance gearing, and high finance expenses. As a result, we prefer companies with low exposure to sheets downstream solar projects and strong balance sheets.

RE Fund movement (CNYbn) CNYbn 700

Total fund shortfall Wind Solar Others Fund inflow

China Wind and Solar: 11 July 2019

China solar segment: receivables days China solar segment: net gearing ratios 700 600% 600 500% 500 400% 400 300% 200% 300 100% 200 0% 100 (100%) 0 GCLNE Xinyi United Kong Longi Tongwei ShunfengGCL-Poly Singyes PV Sun 1H14 2H14 1H15 2H15 1H16 2H16 1H17 2H17 1H18 2H18 1H14 2014 1H15 2015 1H16 2016 1H17 2017 1H18 2018

Source: Companies, Daiwa Source: Companies, Daiwa

China solar segment: OCF/IC China solar segment: OCF/interest expense cover (x) 30% 20

20% 15

10% 10

0% 5

(10%) 0

(20%) (5) 2014 2015 2016 2017 2018 2014 2015 2016 2017 2018 GCLNE Xinyi United PV GCLNE United PV Shunfeng Shunfeng GCL-Poly Singyes GCL Poly Singyes Longi Longi Tongwei Tongwei

Source: Companies, Daiwa Source: Companies, Daiwa

d. Technology shift to higher-efficiency products Apart from declining material costs, such as for polysilicon, the focus for cutting costs in solar installation going forward is likely to be on higher-efficiency solar products, as most low-cost polysilicon capacity has already ramped up and hence there is less room for cost reductions on the materials side.

With higher power generation efficiency, the industry is shifting from multi to mono products. Thus, we prefer mono product producers over multi product producers, given the continuously increasing market share of mono products.

Global mono wafer demand and supply (2015-2021E) Global multi wafer demand and supply (2015-2021E) 120 100 92 104102 88 86 79 82 100 88 84 80 70 66 68 80 68 58 59 57 65 60 48 48 47 60 49 44 40 40 29 25 14 17 16 20 8 20

0 0 2015 2016 2017 2018 2019E 2020E 2021E 2015 2016 2017 2018 2019E 2020E 2021E

Mono-wafer demand (GW) Mono effective capacities (GW) Multi-wafer demand (GW) Multi effective capacities (GW)

Source: Daiwa estimate and forecasts Source: Daiwa estimate and forecasts Note: The effective mono wafer capacity is calculated based on an assumption of an 85% Note: The effective multi wafer capacity is calculated based on an assumption of a 90% utilization utilization rate and adjusted by new capacity installation time rate and adjusted for new capacity installation time

China Wind and Solar: 11 July 2019

We prefer companies Higher-efficiency cell and module technologies such as PERC cells are likely to be a trend. that focus on high- Right now, the cell and module markets are very fragmented. In order for a cell or module efficiency products manufacturing company to achieve a larger market share, it has to continuously introduce more capex and invest in new capacity and new high-efficiency technology. We think the segment of cell and module manufacturing equipment will benefit from the continuous capex of cell and module manufacturers, which is driven by upgrades to existing production lines and the addition of new lines with the coming of retail and grid parity, coupled with increasing localisation of manufacturing equipment.

PERC cell capacity of leading cell manufacturers (GW)

Source: Company, Daiwa

e. Healthy balance sheets We prefer companies with healthy balance sheets, most notably low net-debt-to-equity ratios. Typically, companies with low solar farm exposure and more of a focus on upstream solar product production have relatively low gearing ratios.

Source: Companies, Daiwa Source: Companies, Daiwa

f. Undemanding valuation We prefer stocks where valuations appear undemanding. For example, for Xinyi Solar (which on our forecasts is trading currently at 2.2x 2019E PBR and 13x 2019E PER), if we assume a valuation for its solar farm segment of 1.0x 2019E PBR (better than GNE’s 0.6x 2019E PBR) and 8x 2019E PER (similar to Longyuan’s), we estimate its solar glass segment is trading at 3.9x 2019E PBR and 16x 2019E PER — significantly higher than its peer Flat Glass’s (6865 HK, NR) 1.5x 2019E PBR and 10x 2019E PER (based on consensus estimates). Thus, we think the valuation of Xinyi Solar is rich, even allowing for the fact it is the market leader in the solar glass segment.

China Wind and Solar: 11 July 2019

Solar farm operators: distressed cash flow with liquidity risk

Worsening cash flows and balance sheets As with wind power projects, we expect subsidy collection of solar projects going forward to rely mainly on China’s Renewable Energy Fund means. By extension, we believe that delays in subsidy collection are likely to continue, which would result in cashflow and balance-sheet pressures for solar farm operators. Indeed, the situation could worsen as the NEA intends to re-allocate CNY3bn for competitive bidding for new solar power capacity in 2019 and to pay subsidies to new solar projects ahead of existing projects starting this year.

RE Fund movement (CNYbn) CNYbn 700

Total fund shortfall Wind Solar Others Fund inflow

Under the current policy, renewable energy projects need to be admitted to the subsidy catalogue before they are entitled to a subsidy payment. So far, the government has organised registration of 7 batches of subsidy catalogues, covering renewable energy projects that entered operation before end-March 2016. The government has not launched the registration for the 8th batch of the RE subsidy catalogue, which means that the projects that entered operation after March 2016 are not yet entitled to subsidy collection and may face even longer waiting periods. We note too that solar farm operators have a significantly larger proportion of installed capacity not listed in the RE Fund subsidy catalogue (57-70%) as compared with wind farm operators (7-42%), while the subsidy per kWh of solar power could be higher than that of wind power.

China solar: exposure to editions of subsidy catalogue as of end-2018 Capacity (MW) Installed capacity Proportion in Proportion not in 1-5th 6th 7th (MW) 1-7th batch 1-7th batch Xinyi Solar - 250 824 2,500 43% 57% GCL-Poly 20 774 1,416 7,258 30% 70%

Source: NEA, company, Daiwa Note: Covering period: 1-5th batch (before end-August 2013), 6th batch (Sep 2013 to Feb 2015), 7th batch (Mar 2015 to Mar 2016)

Deteriorating financial While we have seen an impressive expansion of China’s solar power capacity in recent conditions and rising years, solar farm operators’ balance sheets and cash flows have continued to deteriorate. default risks in the solar- The receivables days of most Hong Kong-listed solar companies with exposure to solar- farm operation segment farm operations have been on a clear uptrend since 2014, from 28-177 days in 1H14 to 74- 554 days in 2H18, due to the longstanding back-payment issue of China’s FiT programme for solar power — significantly higher than the 27-68 days for solar companies without large exposure to solar farms.

China Wind and Solar: 11 July 2019

With these rising receivables days and ongoing capacity expansion, solar-farm operators have had to add to their debt in order to fund their capex, foreshadowing increasing net gearing ratios and finance costs. We note that the net gearing of many of the solar companies has surged to 175-396% as at end-2018, compared with net cash to 214% in 2014, while upstream solar companies without large exposure to solar farms had lower gearing of 51-82% as of end-2018.

On 18 October 2018, Singyes Solar (750 HK, not rated) announced it had defaulted on its USD160m 6.75% senior note due 17 October 2018. Singyes’ default automatically led to cross-defaults of its other bonds such as its USD260m 7.95% notes due February 2019, CNY930m 5%-coupon convertible bonds (CBs) due August 2019, and some other bank facilities. As of 5 June 2019, Singyes’ default amounts of offshore and onshore debt totalled CNY207.9m and CNY782.4m, respectively. We expect other privately-owned, highly-geared solar companies to face similar risks.

Source: Companies, Daiwa Source: Companies, Daiwa

China solar segment: OCF/IC China solar segment: OCF/interest expense cover (x) 30% 20

20% 15

10% 10

0% 5

(10%) 0

(20%) (5) 2014 2015 2016 2017 2018 2014 2015 2016 2017 2018 GCLNE Xinyi United PV GCLNE United PV Shunfeng Shunfeng GCL-Poly Singyes GCL Poly Singyes Longi Longi Tongwei Tongwei

Source: Companies, Daiwa Source: Companies, Daiwa

Solar farm asset disposal A wave of disposals of In the face of high gearing and tight cash flows at solar-farm operators, disposals of solar solar farm assets by farm assets by privately-owned solar companies are under way, either through selling privately-owned solar individual solar power projects or directly selling stakes in solar farm operating companies, companies is under way mainly to SOEs.

GCL New Energy (GNE, 451 HK, NR) has sold solar plant projects with a total capacity of 1.6GW since late-2018, bringing in some CNY2.8bn.

China Wind and Solar: 11 July 2019

GCL New Energy: disposal of solar power projects Capacity Net cash Remaining sold Percentage proceeds Transaction capacity Time Seller (MW) Buyer Buyer's parent company of stakes (CNYm) (CNY/w) (MW) Oct-2018 GCL New Energy 160 CGN Solar CGN Group 80% 306 8.4 7,469 Dec-2018 GCL New Energy 140 Three Gorges Three Gorges Group 100% 164 3.6 7,309 New Energy Mar-2019 GCL New Energy 280 Wuling Power State Power Investment 55% 246 n.a. 7,029 Corporation May-2019 GCL New Energy 977 Shanghai Yunnan Provincial Energy 70% 2,060 8.5 6,052 Rongyao Investment Group (a limited partner of Ningbo Rongshang Investment Fund, which holds the majority stake of Shanghai Rongyao) Total 1,557 2,776 6,052

Source: Company, Daiwa Note: Transaction price per watt is estimated by Daiwa

Solar farm operators are Furthermore, GCL-Poly announced on 4 June 2019 that it may sell 9.7bn shares (a c.51% selling individual solar stake) of GNE to Huaneng Group, 1 of the 5 largest power-generating SOEs in China, with power projects or the timing and consideration yet to be decided. GCL-Poly currently has a 62.3% stake in directly selling stakes of GNE. We think the disposal would, if completed, lessen GCL-Poly’s financial burden, with in solar farm operating GNE’s debt being deconsolidated and GCL-Poly receiving cash from the deal. Besides, we companies, mainly to believe a significant valuation premium is unlikely as the deal would need to be approved SOEs by the State-owned Assets Supervision and Administration Commission (SASAC). If the deal can be completed in 2020, we estimate GCL-Poly’s 2020E net gearing would decline to 105-111%, from our existing 2020E forecast of 219%, assuming GNE were sold at 0.7- 0.9x 2020E PBR (currently trading at 0.6x 2020E PBR).

We think a deal is highly likely, given: 1) Huaneng Group is not a listed company and thus there are fewer short-term concerns on the impact of a large acquisition, and 2) GNE has CNY9bn in debt due this year, the cost of which will be high if refinanced. Also, we note that GCL-Poly’s short-term refinancing risk is still high, with a large amount of debt (c.CNY15bn) due in 2019 even when excluding GNE’s debt, which is the main reason why GCL-Poly further announced a share placement of 7.6% of enlarged shares outstanding to raise HKD680m after its announcement of plans to dispose some of its stake in GNE to the Huaneng Group.

GCL-Poly: estimate of financial impact of GNE disposal for 2020E Valuation of GNE disposal (PBR) Year 2020 (CNYm) Base case (without disposal) 0.7x 0.8x 0.9x Disposal gain/ (loss) (1,174) (783) (391) Cash received 2,739 3,131 3,522 Net debt 48,188 21,853 21,522 21,192 Change vs. base case -55% -55% -56% Net gearing 194% 93% 91% 88% Change vs. base case -101pp -103pp -106pp Net finance cost (including interest income) (2,995) (1,201) (1,194) (1,187) Change vs. base case -60% -60% -60% Reported net profit 776 (551) (249) 53 Recurring net profit (excluding disposal loss) 776 339 345 350 Change vs. base case -56% -56% -55%

Source: Daiwa estimates Note: We think loss of earnings from GNE due to GNE’s disposal would not concern us, as the company can save almost CNY2bn in finance cost, which is more cash-based

It is a similar story at Singyes Solar. On 5 June 2019, the company announced a placement of 1,687m shares to Water Development (HK) at HKD0.92 per share, or a 7% discount to the last closing price, accounting for 67% of its enlarged share capital, for a total consideration of HKD1,552m (net proceeds: HKD1,529m). Water Development (HK) is a subsidiary of Shuifa Group, an SOE whose ultimate controlling shareholder is the Shandong provincial government. This means that the controlling shareholder of Singyes Solar would also become an SOE after the deal.

China Wind and Solar: 11 July 2019

Besides, Xinyi Solar (968 HK) completed the spin-off of Xinyi Energy (XYE, 3868 HK) on 28 May 2019, though XYE remained a subsidiary of Xinyi Solar after listing. The spin-off resulted in the issuance of 1.88bn new shares, accounting for 28.41% of its enlarged share capital. The issue price was HKD1.94/share, close to the low end of the offering range of HKD1.89-2.35, with a market cap of HKD12.9bn after listing. About 90% of net proceeds, or HKD3,526m, will be used to buy 540MW of solar farms from XYS. Along with its existing 954MW, XYE will have a total solar capacity of 1,494MW by end-2019. After the spin-off, XYE will focus on operating solar farms, while XYS will focus on solar glass manufacturing and solar farm EPC. Going forward, XYE plans to acquire solar farms from XYS (which still has 850MW in reserves) and third parties, and aims to become a high-yield play with a policy of distributing more than 90% of distributable income each year (likely 100% for 2019-20). We view the XYE spin-off as beneficial to XYS in terms of lowering its gearing, capex burden, and equity financing risk, as XYE will take the cash-flow risk on solar farms despite still being consolidated within XYS.

China: recent actions of solar farm operators in the capital market Current/ future Original major major Enlarged stake Time Target company shareholder shareholder disposal Channel Net proceeds Jun-2019 (not GCL New Energy GCL-Poly Huaneng Group 51.0% Stake transfer CNY2.7- 3.5bn finalized yet) (estimated by Daiwa) Jun-2019 Singyes Solar Company Shuifa Group (a 66.9% Share placement HKD1,529m management Shandong SOE) May-2019 Xinyi Energy Xinyi Solar Xinyi Solar 28.4% Spin-off HKD3,526m

Source: Company, Daiwa estimates

Less supportive RPS policy Voluntary trading of GCs On 15 May 2019, the NDRC and NEA published the final official version (Chinese-language is one way to ensure link) of the Renewable Portfolio Standards (RPS) for each province in China, having issued 3 RPS compliance draft versions in 2018. The details in the final version are largely similar to those in the third draft issued in November 2018, which specified 2 RPS targets (ie, total RPS and non-hydro RPS) for each province, with the parties that need to comply with the RPS remaining on the consumer side (not including coal-fired IPPs). The targets have been further divided into obligatory targets and incentive ones, while those for some provinces were adjusted (mainly hydro consumption target) relative to the third draft. While the 2018 targets are just for self- review and the 2019 targets are for trial assessment, various provinces should complete preparatory work before end-2019 as the RPS is to be formally implemented starting from 2020, vs. commencement in 2019 in the third draft. The 2020 RPS targets outlined currently are for guidance of renewable energy development in each province, while actual targets for 2020 will be released before end-1Q20. Similar to the third draft, Green Certificates (GC) are subject to voluntary trading, while parties that need to comply with the RPS can directly buy the excess renewable energy consumption volumes of other parties.

The final version states that completion of RPS targets should be mainly through actual consumption of renewable energy, but there are 2 complementary approaches: 1) buying the excessive renewable energy volumes of other parties, with the transfer price to be determined by both sides independently, and 2) buying GCs from renewable energy producers on a voluntary basis. We think the first way is more suitable for a regional grid company, with the trading of excess RPS completion volumes similar to a bilateral wholesale transaction, while the second way is more attractive to smaller, RPS-compliant players such as power-retail companies, power end-users and captive power-plant owners, as there is a public trading platform for GCs offering more transparent prices.

GCs were introduced in China on 1 July 2017, but so far they have only been subject to voluntary trading and are not traded actively. Under the GC compulsory trading proposed in the second draft, we expect that GC trading can partly replace the subsidies paid by the Renewable Energy Fund and accelerate collection of part of the subsidies. Besides, under voluntary trading (which is the current trading method of GCs) as stated in the third draft and final version, we maintain our view that the subsidy collection from renewable energy companies will rely mainly on the RE Fund (which currently has a large deficit) and continue to face delays.

China Wind and Solar: 11 July 2019

China: provincial total RPS targets for provincial grids, and power retailers/end users/captive power plant owners Total RPS target Change compared with the 3rd draft 2018 2018 2019 2019 2020 2020 2018 2018 2020 2020 2017 obligatory incentive obligatory incentive obligatory incentive obligatory incentive obligatory incentive actual target target target target target target target target target target Beijing 12.1% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Tianjin 11.0% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Hebei 11.6% 11.0% 12.1% 13.5% 14.9% 15.0% 16.5% 0.0pp 0.0pp 0.0pp 0.0pp Shanxi 14.1% 15.0% 16.5% 15.5% 17.1% 16.5% 18.1% 0.0pp 0.2pp 0.0pp 0.1pp Inner Mongolia 19.2% 18.5% 20.4% 18.5% 20.4% 18.5% 20.4% 0.0pp 0.1pp 0.0pp 0.1pp Liaoning 12.2% 12.0% 13.2% 12.0% 13.2% 12.5% 13.7% 0.0pp 0.2pp 0.0pp 0.1pp Jilin 22.2% 20.0% 22.0% 21.5% 23.7% 22.0% 24.2% 0.0pp 0.5pp 0.0pp 0.5pp Heilongjiang 20.2% 19.5% 21.5% 21.5% 23.7% 26.0% 28.6% 0.0pp 0.5pp 0.0pp 0.5pp Shanghai 33.3% 31.5% 34.9% 32.0% 35.2% 33.0% 36.5% 0.0pp 2.9pp 0.0pp 3.0pp Jiangsu 14.7% 12.5% 13.7% 13.5% 14.9% 14.0% 15.5% -2.0pp -1.4pp -1.0pp -0.3pp Zhejiang 19.3% 18.0% 19.8% 17.5% 19.3% 18.5% 20.4% 0.0pp 1.3pp -0.5pp 0.6pp Anhui 14.3% 13.0% 14.3% 13.5% 14.9% 14.5% 15.9% 0.0pp 0.3pp 0.0pp 0.2pp Fujian 24.2% 17.0% 18.7% 18.5% 20.4% 19.5% 21.4% 0.0pp 1.2pp -2.5pp -1.2pp Jiangxi 25.4% 23.0% 25.1% 25.5% 28.1% 29.0% 32.1% 0.0pp 1.6pp 0.0pp 2.1pp Shandong 7.3% 9.5% 10.4% 10.0% 11.0% 10.0% 11.0% 0.0pp 0.0pp -0.5pp -0.6pp Henan 14.6% 13.5% 14.9% 13.5% 14.9% 16.0% 17.6% 0.0pp 0.4pp 0.0pp 0.5pp Hubei 43.0% 39.0% 43.0% 37.5% 41.3% 40.0% 44.0% 0.0pp 3.1pp 0.0pp 3.0pp Hunan 50.4% 46.0% 50.5% 47.0% 51.7% 49.0% 53.9% -5.5pp -1.9pp -2.5pp 1.5pp Guangdong 32.4% 31.0% 34.2% 28.5% 31.4% 29.5% 32.5% 0.0pp 2.8pp 0.0pp 2.5pp Guangxi 51.6% 51.0% 56.2% 45.5% 50.1% 50.0% 55.0% 0.0pp 4.8pp 0.0pp 4.5pp Hainan 13.3% 11.0% 12.1% 11.0% 12.1% 11.5% 12.6% 0.0pp 0.6pp 0.0pp 0.6pp Chongqing 49.2% 47.5% 52.1% 42.5% 46.8% 45.0% 49.5% 0.0pp 4.6pp 0.0pp 4.2pp Sichuan 83.5% 80.0% 88.0% 80.0% 88.0% 80.0% 88.0% 0.0pp 7.6pp 0.0pp 7.6pp Guizhou 35.6% 33.5% 36.9% 31.5% 34.7% 31.5% 34.7% 0.0pp 2.9pp 0.0pp 2.7pp Yunnan 85.6% 80.0% 88.0% 80.0% 88.0% 80.0% 88.0% 0.0pp 7.0pp 0.0pp 6.8pp Tibet 83.8% NA NA NA NA NA NA NA NA NA NA Shaanxi 16.0% 17.5% 19.2% 18.5% 20.4% 21.5% 23.7% 0.0pp 0.8pp 0.0pp 1.0pp Gansu 46.9% 44.0% 48.4% 44.0% 48.4% 47.0% 51.1% 0.0pp 2.8pp 0.0pp 2.2pp Qinghai 64.9% 70.0% 77.0% 69.5% 76.5% 70.0% 77.0% 0.0pp 5.1pp 0.0pp 4.5pp Ningxia 23.0% 20.0% 22.2% 20.0% 22.0% 22.0% 24.2% 0.0pp 0.2pp -3.0pp -2.8pp Xinjiang 26.0% 21.0% 23.1% 21.0% 23.1% 22.5% 24.5% -4.0pp -3.4pp -3.5pp -2.8pp

Source: NDRC, NEA, Daiwa Note: the 3rd draft did not give targets for 2019

China Wind and Solar: 11 July 2019

Stock calls

Xinjiang Goldwind (2208 HK): reiterating Buy (1) with TP of HKD11.90 We prefer upstream Goldwind’s share price has declined by 18% since end-March, likely due to the market’s product makers concerns over the company’s lower WTG GPM in 2019. We believe the rush installation including Goldwind, story in the wind sector is intact, with the NDRC’s notice of 24 May stating that onshore Tongwei and LONGi, and projects on FiT approved before 2018 have to be connected on-grid before 2020, and small-cap names those on the competitive bidding tariff approved in 2019-20 have to be connected on-grid including Huadian Fuxin before 2021 if they are to get subsidies. We expect Goldwind’s WTG sales to rise to 8- and Suntien 10GW in 2019-21E — much higher than the 5.9GW in 2018. We believe Goldwin’s WTG GPM is likely to bottom in 2019E but recover in 2020E, as bidding prices started rebounding 4Q18. As such, we reaffirm our Buy (1) rating with a TP of HKD11.9, on a lower 11x 2019-20E average PER (from 13x 2019E PER) considering the WTG GPM squeeze in 2019.

Huadian Fuxin (816 HK): reiterating Buy (1) and TP of HKD2.40 Fuxin’s share price is down by 25% YTD, underperforming its wind-IPP peers by 9-28% and the HSI by 34%. We expect Suntien to report c.30% YoY net profit growth on a recover in hydro power generation from 67.7% YoY in 1Q19 to c.130% YoY in 1H19. For 2Q-3Q19, we expect Fuxin’s hydro and coal-fired power assets to record a decent YoY recovery in profitability, led by either utilisation gains or fuel costs. As with other wind IPPs, we expect Fuxin’s capex to rise, from CNY4.4bn in 2018 to CNY11-13bn pa in 2019-21E before the expiry of subsidy tariffs. Hence, Fuxin’s FCF is likely to remain stretched. However, we note that Fuxin’s 0.4x PBR for 2019E (peers: 0.4-0.7x) is the lowest in the sector, and therefore we reiterate our Buy (1) rating, backed also by its attractive 13-14% ROE for 2019-21E (peers: 8-14%). Our 12-month TP is unchanged at HKD2.4 on an unchanged 0.7x 2019E PBR. We recommend investors accumulate the stock for a visible output pick-up in 2Q-3Q19E.

China Suntien (956 HK): reiterating Buy (1) with TP of HKD3.50 Suntien’s share price is flat YTD, meaning the stock has outperformed its wind-IPP peers’ 1-22% YTD falls but underperformed city-gas peers’ 7-25% performance (HSI: +9% YTD). We expect Suntien to report a c.5-10% YoY net profit growth for 1H19 on a recovery in wind power generation from +6% YoY in 1Q19 to +12.4% YoY in 5M19. As with other wind IPPs, we expect Suntien’s capex to rise, from CNY4.5bn in 2018 to CNY5.0-6.2bn pa in 2019-21E before the expiry of subsidy tariffs. Hence, Suntien’s FCF is likely to remain stretched. However, we note that Suntien’s ex-gas 0.4x PBR for 2020E is the lowest in the sector, with an attractive 7% yield for the same year (peers: 1-4%), and therefore we reiterate our Buy (1) rating with an unchanged DCF-based 12-month TP of HKD3.5. We would see a potential A-share IPO as a further rerating catalyst.

Huaneng Power International (902 HK): reiterating Buy (1) with TP of HKD5.95 Due to a high base in 1H18, China’s thermal coal price declined by 3-10% YoY in 1H19, which we believe will contribute to decent growth in HNP’s 1H19 earnings. Besides, HNP plans to step up the progress of wind capacity installations, with heavy capex in the coming few years, which we believe may result in rising capex, worsening cash flow, and higher gearing. HNP is currently trading at 0.7x 2019E PBR, 0.9SD below its average 1-year forward PBR since 2007. We reiterate our Buy (1) rating and TP of HKD5.95, based on a 0.9x 2019-20E average PBR — still 0.6SD lower than its average 1-year forward PBR of since 2007.

China Longyuan (916 HK): downgrading to Hold (3) with new TP of HKD5.40 Longyuan’s share price is down 6% YTD, underperforming the HSI by 15%. Management has guided for only a 0.9% YoY rise in 2019 utilisation hours to 2,230 (5M19: 0.2% YoY

China Wind and Solar: 11 July 2019

drop). As with other wind IPPs, we forecast Longyuan’s capex to rise, from CNY7.8bn in 2018 to CNY12.5-13.5bn pa in 2019-21E before the expiry of subsidy tariffs. Hence, Longyuan’s FCF will likely be stretched. Therefore, we downgrade Longyuan to Hold (3) from Outperform (2), based on a lower 8.5x 2019E PER (previously: 8.9x 2019E PER) and a resulting TP of HKD5.40.

Huaneng Renewables (958 HK): maintaining Hold (3) with new TP of HKD2.30 HNR’s share price is up 1% YTD, underperforming the HSI by 9%. We expect HNR to report c.13% YoY net profit growth with a 9% YoY increase in wind power generation in 1H19 (1Q19: +11% YoY). As with other wind IPPs, we expect HNR’s capex to rise, from CNY3.8bn in 2018 to CNY11.9-13.6bn pa. in 2019-21E before the expiry of subsidy tariffs. Hence, HNR’s FCF will likely be stretched. As a result, we maintain our Hold (3) rating on HNR, based on a lower 5.6x 2019-20E average PER (previously: 6.0x 2019E PER) for a new TP of HKD2.30.

Datang Renewables (1798 HK): downgrading to Hold (3) with new TP of HKD0.80 DTR’s share price is off by 15% YTD, underperforming the HSI by 26%. Management has guided for almost flat utilisation hours in 2019 compared with 2018. As with other wind- IPPs, we expect Longyuan’s capex to rise, from CNY2.3bn in 2018 to CNY7.3-9.4bn pa in 2019-21E before the expiry of subsidy tariffs. Hence, DTR’s FCF will likely be stretched. As a result, we downgrade DTR from Buy (1) to Hold (3), based on a lower target 0.4x PBR (previous: 0.8x PBR) for a resulting TP of HKD0.80.

GCL-Poly (3800 HK): downgrading to Hold (3) with new TP of HKD0.45 Despite the fact that GCL-Poly announced its plan to dispose of a 51% stake in GCL New Energy (GNE; 451 HK, NR) to Huaneng Group on 4 June, the company made a further share placement (7.6% of the expanded base) for HKD680m on 18 June and announced a move to dispose of a 31.5% holding in Xinjiang GCL for CNY2.5bn on 26 June, mainly for debt repayment. We think these recent actions are a sign that GCL-Poly’s cash shortage is larger than we had expected, and short-term refinancing risk seems high given the large amount of debt (c.CNY15 bn) due in 2019, even if we exclude GNE’s debt. Hence, we downgrade our rating to Hold (3), and lower our TP to HKD0.45 (from HKD0.52) on 0.3x 2019-20E average PBR (previously 0.4x PBR) to reflect share dilution and a lower ROE.

China Resources Power (836 HK): reiterating Hold (3) and TP of HKD11.50 CRP’s share price dropped significantly by 19% in the 2 days after it unexpectedly announced a dividend policy change on 22 March, and has remained at a similar level since then. We think its coal-fired power utilisation YTD is slightly lower than expected. It is also aggressively adding wind capacity, which we believe would result in rising accounts receivable and capex, and worsening cash flow. As such, we maintain our Hold (3) rating and SOTP-based 12-month TP of HKD11.5, representing a 0.7x 2019E PBR — similar to the current valuations of HNP and Longyuan.

Xinyi Solar (968 HK): reiterating Underperform (4) with new TP of HKD3.50 With demand supported by global solar installation growth, and the rising penetration of bifacial double-glass modules, we expect the solar glass segment to see a recovery this year. Also, XYS completed the spin-off of Xinyi Energy (XYE; 3868 HK, not rated) on 28 May 2019, which is beneficial for XYS as it implies less equity financing risk. However, its valuation still looks rich compared with its historical multiples and peer valuations, in our opinion. We trim our SOTP-based 12-month TP to HKD3.5 (from HKD3.8) after fine-tuning the valuation of the solar glass and solar farm segments, implying a 2.1x 2019E PBR — slightly higher than its past-3-year mean of 2.0x PBR. Underperform (4) reaffirmed.

China Wind and Solar: 11 July 2019

Risks to our Neutral sector view

Downside risk to WTG makers: lower-than-expected wind installation The primary risks to our We expect the wind IPPs to focus on developing approved project reserves, with fixed sector stance are lower- tariffs over 2019-20E, while those gained through competitive bidding in 2019-20 should be than-expected wind and connected on-grid before end-2021. The likely result: rush installations in 2019-21E, with solar installation and installations in China forecast to rise from 21GW in 2018 to 25-33GW in 2019-21E. If better-than-expected China’s wind installation falls short of expectations, there would be negative implications subsidy collection for WTG makers’ shipping volume in 2019-21E and a WTG bidding price recovery in 2019, in our view.

Upside risk to wind IPPs: better-than-expected subsidy collection As the final RPS policy states that Green Certificates are only subject to voluntary trading, we expect subsidy collection going forward will still largely rely on the Renewable Energy Fund, which has a huge deficit currently (over CNY200bn, on our estimates). This means that subsidy collection delay would still likely continue, giving pressure on the cash flows and balance sheets of the wind IPPs. If subsidy collection is better than expected, there would be an upside risk to the wind IPPs. Besides, better-than-expected Green Certificate sales or higher-than-expected utilisation hours would also be upside risks to the wind IPPs.

Downside risk to solar product makers: lower-than-expected solar installation With the finalisation of China’s solar policy, China’s solar installation demand is likely to kick off in 2H19, while we expect full-year demand to be stable for 2019-20 before recovering in 2021. We also see a steady expansion in the overseas market. Thus, we forecast global solar installation to increase from 107GW in 2018 to 125GW in 2019, 136GW in 2020, and 150GW in 2021, representing YoY growth of 16% in 2019, 9% in 2020 and 10% in 2021 and a CAGR of 12% over 2018-21E. If solar installation in China or overseas falls short of these expectations, there would be a negative impact on the demand for solar products in 2019-21E and likely pressure on solar ASPs.

Upside risk to solar project operators: better-than-expected subsidy collection We maintain our view that subsidy collection will largely rely on the Renewable Energy Fund. This means that subsidy collection delays could still continue, putting pressure on the cash flows and balance sheets of the solar project operators. If subsidy collection exceeds expectations, there would be an upside risk to solar project operators, in our view.

China Wind and Solar: 11 July 2019

China Industrials 11 July 2019

(2208 HK) Xinjiang Goldwind Science & Technology Xinjiang Gol dwind Science & T echnolog y

Target price: HKD11.90 (from HKD11.90) Share price (9 Jul): HKD8.69 | Up/downside: +36.9%

Set to tap upcycle in 2019-21E on rush installations Dennis Ip, CFA (852) 2848 4068  WTG demand supported by wind IPPs’ capacity additions in 2019-21E [email protected]  WTG GPM likely to bottom out in 2019E before recovery in 2020E Anna Lu, CFA (852) 2848 4465  Reiterating our Buy (1) rating and TP of HKD11.9 [email protected]

What's new: Goldwind’s share price is down c.20% since end-March, likely Forecast revisions (%) due to the concern that its WTG GPM will decline in 2019. Nonetheless, we Year to 31 Dec 19E 20E 21E believe the wind-sector rush installation story still applies given the NDRC’s Revenue change - 12.3 (6.6) 24 May notice stating that onshore projects that were FiT approved in or Net profit change 1.1 6.6 (15.6) Core EPS (FD) change 1.2 6.7 (15.5) before 2018 must be connected on-grid before 2020, and those on the Source: Daiwa forecasts competitive bidding tariff approved in 2019-20 must be connected on-grid before 2021 if they are to get subsidies. We expect Goldwind’s WTG GPM Share price performance to bottom in 2019E and recover in 2020E as bidding prices had started to (HKD) (%) rebound in 4Q18. As such, we reiterate our Buy (1) rating. 12 120 10 105 What's the impact: WTG demand likely to see support from rush 8 90 installation in 2019-21E. According to the NDRC’s latest policy, projects 7 75 5 60 approved in or before 2018 are still entitled to fixed tariffs if they are Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 connected on-grid before 2020. Large wind IPPs still have sufficient project Xinj Gold (LHS) Relative to HSI (RHS) reserves, and most wind operators are stepping up installation through heavy capex in the coming 2-3 years. Thus, we expect wind IPPs to focus 12-month range 5.44-11.30 mainly on developing approved project reserves with fixed tariffs over Market cap (USDbn) 4.70 2019-20E. Also, projects gained through competitive bidding in 2019-20 3m avg daily turnover (USDm) 6.70 should be connected on-grid before end-2021, according to the policy. Due Shares outstanding (m) 4,225 Major shareholder Xinjiang Wind Energy (13.7%) to policy adjustments (in which projects approved before 2018 change from construction commencement before end-2020 to grid connection before Financial summary (CNY) end-2020), our new estimates for China wind installations are 25GW for Year to 31 Dec 19E 20E 21E 2019E, 33GW for 2020E, and 29GW for 2021E (from 24GW/ 27GW/ Revenue (m) 37,739 49,093 46,155 32GW). Hence, we adjust our forecasts for Goldwind’s WTG sales to 8GW Operating profit (m) 4,272 6,234 6,233 for 2019E, 10GW for 2020E, and 9GW for 2021E (from 8GW/9GW/10GW) Net profit (m) 2,984 4,597 4,487 Core EPS (fully-diluted) 0.767 1.088 1.062 (2018: 5.9GW). EPS change (%) (13.3) 41.9 (2.4) Daiwa vs Cons. EPS (%) (2.1) 12.4 0.6 WTG GPM likely to bottom out in 2019E. WTG bidding prices have PER (x) 10.0 7.0 7.2 rebounded since 4Q18; those for 2.0MW and 2.5MW in March 2019 were Dividend yield (%) 2.8 4.0 3.9 DPS 0.217 0.308 0.300 CNY3.41/W and CNY3.46/W, respectively, 6.7% and 4.0% above PBR (x) 1.0 0.9 0.8 September 2018 levels, and should increase further in the near term, in our EV/EBITDA (x) 8.1 6.3 5.8 view. The WTG segment GPM fell 6pp YoY to 19% in 2018. We expect it ROE (%) 10.5 13.8 12.2 fall to 14% in 2019 as most low-ASP orders in 2018 will be delivered in Source: FactSet, Daiwa forecasts 2019, before rising to 19% in 2020.

What we recommend: We raise our 2020E EPS by 7% but lower our 2021E EPS by 16% after adjusting WTG sales estimates in accordance with the policy change. We reiterate our Buy (1) call and TP of HKD11.9, on a lower 11x 2019-20E average PER (from 13x 2019E PER) considering the forecast WTG GPM squeeze in 2019. Risk: lower WTG ASPs.

How we differ: Our 2019-21E EPS differ from the consensus by -2% to 12%, likely due to our different forecasts of WTG sales in 2019-21E.

See important disclosures, including any required research certifications, beginning on page 105

Xinjiang Goldwind Science & Technology (2208 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook Goldwind: net profit forecasts (CNYm) While we expect Goldwind’s wind farm segment to see 5,000 60% steady growth over 2019-21E (wind farm gross profit likely 4,500 50% to show a 27% CAGR), we think net profit growth in the 4,000 40% coming years will to a large extent be affected by the WTG 3,500 30% segment. We expect Goldwind’s net profit to decline by 5% 3,000 20% YoY in 2019, mainly due to a lower WTG GPM, with most 2,500 10% low-ASP orders in 2018 delivered in 2019. However, we 2,000 0% forecast the company’s net profit to rebound by 54% YoY 1,500 (10%) on rising WTG sales volume and a WTG GPM recovery in 1,000 (20%) 2020E. On our forecasts, 2021E net profit will decline by 2016 2017 2018 2019E 2020E 2021E 2% YoY on the high 2020E base, with slightly lower WTG Net profit YoY growth (RHS) sales after the end-2020 grid connection deadline for wind Source: Company, Daiwa forecasts projects approved in or before 2018.

Valuation Goldwind: 1-year forward PER Goldwind’s share price is off 23% since end-March, likely PER (x) due to the market’s concerns that the company’s WTG 21 GPM will be lower in 2019. The stock is now trading at an 19 Last rush 8x 12-month forward PER, 0.8SD below its past-7-year 17 installation mean. Nonetheless, we see upside to the share price with 15 14.8x Avg+2SD 13 the recovery of WTG demand and ASPs in the coming 12.4x Avg+1SD years. Our TP of HKD11.9 is based on a 2019-20E 11 10.0x Avg average PER of 11x, 0.4 SD above its past-7-year mean, in 9 7.6x Avg-1SD view of the coming upcycle for rush installations in the wind 7 sector. 5 5.2x Avg-2SD 3 Aug-12 Aug-13 Aug-14 Aug-15 Aug-16 Aug-17 Aug-18

Source: Bloomberg, Daiwa forecasts

Earnings revisions Goldwind: Bloomberg-consensus EPS forecast revisions (CNY) The market has lowered its 2019 EPS forecast for 1.6 Goldwind by 13% YTD, on lower WTG GPM guidance for 1.5 1.4 2019 as well as the dilution impact of the 1.9-for-10 rights 1.3 issue completed in 1H19. The consensus forecast for 2020 1.2 has been relatively stable YTD, with the WTG GPM seen 1.1 as recovering in 2020E on rising tender prices backed by 1.0 robust WTG demand from wind sector rush installations. 0.9 0.8 0.7

Our 2019-21E EPS differ from the consensus forecasts by

Jul-18

Apr-18 Oct-18 Apr-19

Jun-18 Jan-19 Jun-19

-2% to 12%, likely due to our different forecasts of WTG Jan-18

Feb-18 Mar-18 Feb-19 Mar-19

Aug-18 Sep-18 Nov-18 Dec-18

May-18 May-19 sales in 2019-21E. 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

Xinjiang Goldwind Science & Technology (2208 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E ASP of WTG - ex-VAT (CNY/kW) 3,670 3,773 3,717 3,682 3,586 3,483 3,715 3,715 Sales volume of WTG (MW) 4,190 7,051 5,883 5,082 5,861 8,205 10,257 9,231 Year-end wind farm consolidated 1,352 2,594 3,737 3,669 4,429 5,929 7,429 7,929 capacity (MW)

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E WTG manufacturing 15,704 26,858 22,264 19,346 22,169 29,877 39,537 35,578 Wind farm development 1,169 1,553 2,414 3,247 3,904 5,092 6,696 7,622 Other Revenue 700 1,435 1,496 2,377 2,518 2,769 2,860 2,955 Total Revenue 17,573 29,846 26,174 24,971 28,590 37,739 49,093 46,155 Other income 794 773 1,085 1,631 2,073 2,303 1,684 1,434 COGS (13,006) (22,069) (18,616) (17,505) (21,247) (29,509) (36,486) (33,764) SG&A (2,222) (3,959) (3,227) (3,497) (3,282) (4,471) (5,880) (5,026) Other op.expenses (533) (950) (1,370) (1,501) (1,648) (1,791) (2,176) (2,566) Operating profit 2,606 3,641 4,047 4,098 4,487 4,272 6,234 6,233 Net-interest inc./(exp.) (584) (556) (687) (818) (1,071) (1,082) (1,179) (1,276) Assoc/forex/extraord./others 87 162 192 210 266 327 380 412 Pre-tax profit 2,109 3,247 3,552 3,491 3,682 3,517 5,435 5,369 Tax (255) (371) (446) (342) (400) (399) (672) (717) Min. int./pref. div./others (24) (26) (140) (165) (137) (134) (167) (165) Net profit (reported) 1,830 2,849 2,966 2,984 3,145 2,984 4,597 4,487 Net profit (adjusted) 1,350 2,625 2,500 2,984 3,145 2,984 4,597 4,487 EPS (reported)(CNY) 0.522 0.809 0.834 0.839 0.884 0.767 1.088 1.062 EPS (adjusted)(CNY) 0.385 0.746 0.703 0.839 0.884 0.767 1.088 1.062 EPS (adjusted fully-diluted)(CNY) 0.385 0.746 0.703 0.839 0.884 0.767 1.088 1.062 DPS (CNY) 0.400 0.480 0.200 0.200 0.250 0.217 0.308 0.300 EBIT 2,606 3,641 4,047 4,098 4,487 4,272 6,234 6,233 EBITDA 3,040 4,186 4,971 5,222 5,840 5,767 8,116 8,504

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 2,109 3,247 3,552 3,491 3,682 3,517 5,435 5,369 Depreciation and amortisation 434 545 924 1,124 1,353 1,495 1,881 2,271 Tax paid (255) (371) (446) (342) (400) (399) (672) (717) Change in working capital 1,920 1,020 (1,077) 252 2,119 (146) (1,881) 504 Other operational CF items (1,378) 336 150 (1,502) (3,629) 117 799 864 Cash flow from operations 2,829 4,776 3,103 3,023 3,125 4,584 5,563 8,291 Capex (2,089) (7,006) (5,039) (4,720) (5,751) (7,689) (7,589) (4,189) Net (acquisitions)/disposals (7) (341) (95) (2,001) (1,074) 931 0 0 Other investing CF items 421 101 (1,877) (377) 712 0 0 0 Cash flow from investing (1,675) (7,245) (7,011) (7,098) (6,114) (6,758) (7,589) (4,189) Change in debt 5,022 407 5,763 4,812 1,609 2,000 2,000 2,000 Net share issues/(repurchases) 0 0 0 0 0 4,743 0 0 Dividends paid (216) (1,078) (1,313) (547) (711) (1,056) (916) (1,299) Other financing CF items (672) (299) 796 (883) 304 (783) (1,347) (943) Cash flow from financing 4,135 (970) 5,247 3,382 1,201 4,904 (263) (242) Forex effect/others (42) 56 47 (88) 53 0 0 0 Change in cash 5,248 (3,382) 1,385 (780) (1,734) 2,730 (2,289) 3,859 Free cash flow 741 (2,229) (1,936) (1,697) (2,626) (3,105) (2,026) 4,102 Source: FactSet, Daiwa forecasts

Xinjiang Goldwind Science & Technology (2208 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 10,108 6,306 8,274 7,740 5,066 7,796 5,507 9,366 Inventory 3,650 3,037 3,192 4,083 4,997 7,614 7,978 6,451 Accounts receivable 11,294 14,526 16,746 17,048 16,895 27,910 28,849 24,513 Other current assets 3,043 1,417 4,884 4,211 5,960 6,381 8,810 5,643 Total current assets 28,095 25,287 33,097 33,081 32,918 49,700 51,144 45,973 Fixed assets 10,482 17,015 19,479 22,838 25,509 31,576 37,444 39,515 Goodwill & intangibles 265 535 776 2,255 3,477 3,324 3,177 3,037 Other non-current assets 6,935 9,736 11,086 14,613 19,461 19,545 19,625 19,651 Total assets 45,777 52,572 64,437 72,788 81,364 104,144 111,390 108,176 Short-term debt 5,858 1,734 2,672 5,999 3,470 3,470 3,470 3,470 Accounts payable 10,839 14,275 14,473 15,257 20,000 28,967 31,578 23,924 Other current liabilities 5,623 4,950 7,518 8,344 8,131 13,070 12,312 11,439 Total current liabilities 22,320 20,959 24,663 29,600 31,601 45,507 47,360 38,833 Long-term debt 6,023 10,761 15,419 15,886 18,865 20,865 22,865 24,865 Other non-current liabilities 2,208 3,462 3,656 3,827 4,424 4,424 4,424 4,424 Total liabilities 30,550 35,182 43,738 49,313 54,889 70,795 74,648 68,121 Share capital 2,695 2,736 2,736 3,556 3,556 4,232 4,232 4,232 Reserves/R.E./others 12,073 14,026 17,241 19,130 21,405 27,540 30,838 34,057 Shareholders' equity 14,768 16,761 19,976 22,687 24,961 31,772 35,070 38,289 Minority interests 459 629 723 788 1,514 1,577 1,672 1,766 Total equity & liabilities 45,777 52,572 64,437 72,788 81,364 104,144 111,390 108,176 EV 33,678 38,134 41,595 44,906 47,486 46,819 51,203 49,437 Net debt/(cash) 1,773 6,188 9,817 14,145 17,269 16,539 20,828 18,969 BVPS (CNY) 4.216 4.713 5.617 6.379 7.019 7.520 8.300 9.062

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 44.1 69.8 (12.3) (4.6) 14.5 32.0 30.1 (6.0) EBITDA (YoY) 234.4 37.7 18.8 5.0 11.8 (1.2) 40.7 4.8 Operating profit (YoY) 284.8 39.7 11.1 1.3 9.5 (4.8) 45.9 (0.0) Net profit (YoY) 628.5 94.4 (4.8) 19.4 5.4 (5.1) 54.1 (2.4) Core EPS (fully-diluted) (YoY) 628.5 93.4 (5.7) 19.4 5.4 (13.3) 41.9 (2.4) Gross-profit margin 26.0 26.1 28.9 29.9 25.7 21.8 25.7 26.8 EBITDA margin 17.3 14.0 19.0 20.9 20.4 15.3 16.5 18.4 Operating-profit margin 14.8 12.2 15.5 16.4 15.7 11.3 12.7 13.5 Net profit margin 7.7 8.8 9.6 11.9 11.0 7.9 9.4 9.7 ROAE 9.6 16.7 13.6 14.0 13.2 10.5 13.8 12.2 ROAA 3.3 5.3 4.3 4.3 4.1 3.2 4.3 4.1 ROCE 10.7 12.8 11.8 9.7 9.5 8.0 10.3 9.5 ROIC 13.4 15.9 13.1 10.9 9.8 8.1 10.2 9.3 Net debt to equity 12.0 36.9 49.1 62.4 69.2 52.1 59.4 49.5 Effective tax rate 12.1 11.4 12.6 9.8 10.9 11.4 12.4 13.4 Accounts receivable (days) 219.1 157.9 218.1 247.0 216.7 216.7 211.0 211.0 Current ratio (x) 1.3 1.2 1.3 1.1 1.0 1.1 1.1 1.2 Net interest cover (x) 4.5 6.6 5.9 5.0 4.2 3.9 5.3 4.9 Net dividend payout 76.6 59.3 24.0 23.8 28.3 28.3 28.3 28.3 Free cash flow yield 2.3 n.a. n.a. n.a. n.a. n.a. n.a. 12.7 Source: FactSet, Daiwa forecasts

Company profile

Founded in 1998, Xinjiang Goldwind is the largest wind turbine supplier in China in terms of WTG installed volume. The company is also engaged in the wind-farm operation business with an attributable capacity of 4.7GW as at end-2018. Goldwind was listed on the Shenzhen Stock Exchange in 2007 and on the Hong Kong Stock Exchange in 2010.

China Utilities 11 July 2019

(916 HK) China Longyuan Power China Long yuan Power

Target price: HKD5.40 (from HKD5.75) Share price (9 Jul): HKD5.16 | Up/downside: +4.7%

Downgrading: slow 2019E; tightening CF for 2020-21E Dennis Ip, CFA (852) 2848 4068  Slow wind power generation likely in 9M19E [email protected]  Upcoming RPS/GC policy unlikely to slow subsidy receivables growth Anna Lu, CFA (852) 2848 4465  Downgrading to Hold (3) from Outperform (2); lowering TP to HKD5.40 [email protected]

What's new: After correcting by 16% between mid-March and mid-June, Forecast revisions (%) Longyuan’s share price has rebounded by 3% (see our 19 June Flash). Year to 31 Dec 19E 20E 21E However, we believe subsidy receivables are still a concern as we expect Revenue change (0.5) (1.1) (0.6) Net profit change (5.9) (7.5) (6.3) Longyuan’s capex to rise from CNY7.8bn in 2018 to CNY12.5-13.5bn to Core EPS (FD) change (5.9) (7.5) (6.3) add 1.2-1.6GW of new wind power capacity pa in 2019-21 due to rush Source: Daiwa forecasts installations before the expiry of subsidy tariffs. We see Longyuan’s FCF falling rapidly in 2019-21E given the need for significant working capital Share price performance amid the upcoming GC policy in 2020. Therefore, we downgrade our rating (HKD) (%) to Hold (3), and prefer wind power equipment maker Goldwind in 2019-21. 7.5 115 6.8 105 What's the impact: distressed balance sheet likely on rush installation 6.0 95 5.3 85 of wind farms. We expect Longyuan’s capex to rise by CNY4.8-5.7bn pa 4.5 75 in 2019-21, vs. 2018, leading to a decline of CNY5.5-6.5bn pa in FCF in the Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 same period. As such, we expect its finance cost to increase by 14% Longyuan P (LHS) Relative to HSI (RHS)

(peers: +25-55%), representing an 18% rise (peers: +23-52%) in EBIT. 12-month range 4.77-7.40 Monthly data and 3Q19 results unlikely to be positive catalysts. In Market cap (USDbn) 5.31 2018, Longyuan added only 523.5MW of wind capacity, vs. 1,026- 3m avg daily turnover (USDm) 6.13 Shares outstanding (m) 8,036 2,222MW in 2015-17. Also, Longyuan’s utilisation for 3Q18 was up 14% Major shareholder Guodian Group (58.4%) YoY, which we see as a downside risk to utilisation for 3Q19, given management is only guiding for a 0.9% YoY rise in 2019 utilisation hours to Financial summary (CNY) 2,230 (5M19: -0.2% YoY). We see upside in 4Q19 with a curtailment Year to 31 Dec 19E 20E 21E decline, from 5.7% in 2018 to 5.2% in 2019E. Revenue (m) 27,013 28,115 29,504 Operating profit (m) 10,194 11,050 12,163 Net profit (m) 4,327 4,675 5,180 RPS/GC not likely to improve subsidy collection. As per the final draft Core EPS (fully-diluted) 0.538 0.582 0.645 of the RPS, besides actual renewable consumption, there are 2 ways to EPS change (%) 10.3 8.1 10.8 remain compliant: 1) buy excess renewable energy volumes from other Daiwa vs Cons. EPS (%) (10.0) (12.4) (11.2) parties, or 2) buy GCs from renewable energy producers on a voluntary PER (x) 8.4 7.8 7.1 Dividend yield (%) 2.4 2.6 2.8 basis. We cut our 2019-21E EPS by 6-8% due to higher finance costs as DPS 0.108 0.116 0.129 GCs won’t be mandatory to improve working capital, and also given higher PBR (x) 0.7 0.6 0.6 operating costs based on the company’s 1Q19 results. EV/EBITDA (x) 6.7 6.6 6.4 ROE (%) 8.5 8.6 8.9

Source: FactSet, Daiwa forecasts What we recommend: We downgrade Longyuan to Hold (3) from Outperform (2) and cut our 12-month TP to HKD5.40 from HKD5.75, based on a lower 8.5x 2019-20E average PER (previously: 8.9x 2019E PER), as we expect the balance sheet to remain stretched, leading to slower EPS growth on rising finance costs over 2019-21. Risks: better-/worse-than- expected collection of subsidy receivables.

How we differ: Our 2019-21E EPS are 10-12% below the consensus, likely due to our higher finance costs and updated utilisation assumptions.

See important disclosures, including any required research certifications, beginning on page 105

China Longyuan Power (916 HK): 11 July 2019

How do we justify our view? Growth outlook Valuation Earnings revisions

Growth outlook Longyuan: net profit growth and forecasts (CNYm) We forecast Longyuan to deliver a 10% net profit CAGR 6,000 30% over 2018-21, driven mainly by: 1) a 0.2% CAGR in its 5,000 25% wind utilisation rate for the same period, and 2) a 7.4% wind capacity CAGR, despite a 4.4% CAGR in finance 4,000 20% costs over the same period. 3,000 15%

2,000 10%

1,000 5%

0 0% 2014 2015 2016 2017 2018 2019E 2020E 2021E Net profit (LHS) YoY growth (RHS)

Source: Company, Daiwa forecasts

Valuation Longyuan: 1-year forward PER

Longyuan shares are currently trading at 8.1x and 0.7x 12- PER (x) month forward PER and PBR, respectively, both at their 29 Last rush historical troughs, or more than 1.4SD below their historical installation 26.6x Avg+2SD mean since 2010. However, we would expect Longyuan to 24 stay at a relatively distressed valuation level until the end 21.2x Avg+1SD 19 of rush installations at end-2021; thereafter, we see FCF 15.8x Avg starting to improve on lower capex and a stabilising of 14 surges in subsidy receivables. 10.5x Avg-1SD 9 Longyuan’s PBR was de-rated from 20x PER to 7x PER 5.1x Avg-2SD (or 1.7x PBR to 0.7x PBR) in 2015 (last rush installation), 4 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 Jan-18 Jan-19 due to a significant 11% drop in utilisation hours over 2013- Source: Bloomberg 15 on surging curtailment. We see finance costs surging during this round of installation on worsening receivables.

Earnings revisions Longyuan: Bloomberg-consensus EPS forecast revisions (CNY) The 2019-20 Bloomberg-consensus EPS for Longyuan 0.9 have come down by 9-10% since March 2019, when the company reported disappointing 2018 results on rising 0.8 operating costs and announced plans to install over 1GW of new wind farms, which we believe could result in 0.7 declining FCF, rising debt, and higher finance expenses. 0.6 Our 2019-21E EPS are 10-12% below the consensus, likely due to our higher finance and operating-cost 0.5

assumptions. Jul-18

Oct-18 Oct-18 Apr-19

Jun-18 Jan-19 Jan-19 Jun-19

Mar-19 Mar-19

Feb-19

Aug-18 Dec-18 Aug-18 Sep-18 Nov-18

May-19 May-19 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

China Longyuan Power (916 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Consolidated capacity - Total (MW) 15,698 17,950 19,494 20,520 21,044 22,244 23,944 25,544 Consolidated capacity - wind (MW) 13,543 15,765 17,369 18,395 18,919 20,119 21,819 23,419 Wind power utilization hours (hours) 1,980 1,888 1,901 2,035 2,209 2,216 2,220 2,222 Consolidated wind power output (m 22,131 24,414 28,497 32,764 37,609 39,910 42,571 46,263 Kwh) Average on-grid wind farm tariff 0.585 0.590 0.570 0.566 0.564 0.561 0.556 0.546

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of electricity 14,934 15,891 17,367 19,662 21,895 22,540 23,617 24,972 Service concession construction 228 662 541 74 14 14 14 14 revenue Other Revenue 3,066 3,130 4,396 4,856 4,478 4,458 4,484 4,517 Total Revenue 18,228 19,683 22,304 24,592 26,388 27,013 28,115 29,504 Other income 437 450 647 712 917 1,060 1,060 1,060 COGS (5,010) (5,044) (6,511) (7,166) (6,641) (6,412) (6,310) (6,261) SG&A (1,606) (1,794) (2,032) (2,218) (2,663) (2,661) (2,703) (2,768) Other op.expenses (5,489) (6,162) (6,860) (7,582) (8,759) (8,805) (9,113) (9,371) Operating profit 6,561 7,133 7,548 8,337 9,242 10,194 11,050 12,163 Net-interest inc./(exp.) (2,960) (3,039) (2,774) (3,215) (3,513) (3,612) (3,805) (4,029) Assoc/forex/extraord./others 454 576 376 344 167 200 290 290 Pre-tax profit 4,055 4,670 5,150 5,465 5,897 6,782 7,534 8,424 Tax (510) (601) (660) (916) (976) (1,210) (1,364) (1,531) Min. int./pref. div./others (990) (1,189) (1,074) (862) (997) (1,245) (1,495) (1,713) Net profit (reported) 2,554 2,879 3,416 3,688 3,924 4,327 4,675 5,180 Net profit (adjusted) 2,554 2,879 3,416 3,688 3,924 4,327 4,675 5,180 EPS (reported)(CNY) 0.318 0.358 0.425 0.459 0.488 0.538 0.582 0.645 EPS (adjusted)(CNY) 0.318 0.358 0.425 0.459 0.488 0.538 0.582 0.645 EPS (adjusted fully-diluted)(CNY) 0.318 0.358 0.425 0.459 0.488 0.538 0.582 0.645 DPS (CNY) 0.060 0.072 0.085 0.092 0.098 0.108 0.116 0.129 EBIT 6,561 7,133 7,548 8,337 9,242 10,194 11,050 12,163 EBITDA 11,572 12,711 13,890 15,135 16,528 17,573 18,763 20,154

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 4,055 4,670 5,150 5,465 5,897 6,782 7,534 8,424 Depreciation and amortisation 5,011 5,578 6,342 6,798 7,286 7,379 7,713 7,991 Tax paid (510) (601) (660) (916) (1,000) (1,210) (1,364) (1,531) Change in working capital (2,794) 4,023 (1,544) (1,811) (2,307) (3,038) (4,122) (5,295) Other operational CF items 7,120 2,655 4,245 2,594 4,379 3,596 3,743 3,953 Cash flow from operations 12,881 16,325 13,533 12,131 14,255 13,510 13,503 13,542 Capex (13,635) (16,193) (11,637) (10,040) (7,751) (12,529) (13,459) (12,849) Net (acquisitions)/disposals (867) (745) 0 0 0 0 0 0 Other investing CF items (6,362) 619 428 1,426 (882) 0 0 0 Cash flow from investing (20,864) (16,319) (11,209) (8,614) (8,633) (12,529) (13,459) (12,849) Change in debt 12,035 4,405 1,322 1,564 (2,000) 5,000 5,000 5,000 Net share issues/(repurchases) 0 0 0 0 0 0 0 0 Dividends paid (382) (480) (576) (683) (738) (785) (865) (935) Other financing CF items (4,000) (3,418) (4,030) (1,233) (5,064) (4,466) (4,406) (4,683) Cash flow from financing 7,653 507 (3,284) (352) (7,802) (251) (271) (618) Forex effect/others (6) (22) (21) 2 (30) 0 0 0 Change in cash (336) 490 (981) 3,166 (2,210) 730 (227) 76 Free cash flow (754) 132 1,896 2,091 6,504 981 44 693 Source: FactSet, Daiwa forecasts

China Longyuan Power (916 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 2,835 3,273 1,933 5,105 3,114 3,844 3,617 3,693 Inventory 1,017 1,081 1,040 953 852 858 831 810 Accounts receivable 6,416 4,243 5,901 7,155 10,542 12,891 16,823 21,893 Other current assets 4,526 4,106 4,458 3,909 3,278 3,779 3,937 4,137 Total current assets 14,795 12,703 13,333 17,122 17,786 21,371 25,209 30,533 Fixed assets 88,555 98,997 105,598 109,473 110,001 116,309 122,567 127,937 Goodwill & intangibles 8,542 8,699 8,860 8,754 8,171 7,906 7,482 7,057 Other non-current assets 11,921 13,468 10,870 10,286 10,547 10,583 10,784 10,986 Total assets 123,813 133,867 138,661 145,635 146,504 156,168 166,042 176,513 Short-term debt 36,114 44,977 44,511 35,820 28,390 37,215 39,526 41,837 Accounts payable 1,021 1,902 2,550 1,891 2,059 1,877 1,819 1,772 Other current liabilities 9,194 9,121 8,747 9,448 9,332 9,332 9,332 9,332 Total current liabilities 46,328 56,000 55,807 47,159 39,780 48,423 50,676 52,941 Long-term debt 33,922 29,970 31,788 42,035 47,006 43,595 46,284 48,972 Other non-current liabilities 2,658 3,322 3,279 3,141 3,152 3,152 3,152 3,152 Total liabilities 82,909 89,292 90,874 92,336 89,939 95,170 100,112 105,065 Share capital 8,036 8,036 8,036 8,036 8,036 8,036 8,036 8,036 Reserves/R.E./others 25,071 30,099 32,853 38,089 41,200 44,662 48,402 52,546 Shareholders' equity 33,107 38,136 40,890 46,126 49,236 52,698 56,438 60,582 Minority interests 7,797 6,439 6,896 7,173 7,329 8,300 9,491 10,866 Total equity & liabilities 123,813 133,867 138,661 145,635 146,504 156,168 166,042 176,513 EV 107,947 109,842 113,331 112,004 111,614 117,068 123,196 129,205 Net debt/(cash) 67,200 71,673 74,366 72,750 72,282 76,966 82,192 87,117 BVPS (CNY) 4.120 4.745 5.088 5.740 6.127 6.557 7.023 7.539

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) (4.8) 8.0 13.3 10.3 7.3 2.4 4.1 4.9 EBITDA (YoY) 11.1 9.8 9.3 9.0 9.2 6.3 6.8 7.4 Operating profit (YoY) 9.1 8.7 5.8 10.5 10.9 10.3 8.4 10.1 Net profit (YoY) 24.5 12.7 18.6 8.0 6.4 10.3 8.1 10.8 Core EPS (fully-diluted) (YoY) 24.5 12.7 18.6 8.0 6.4 10.3 8.1 10.8 Gross-profit margin 72.5 74.4 70.8 70.9 74.8 76.3 77.6 78.8 EBITDA margin 63.5 64.6 62.3 61.5 62.6 65.1 66.7 68.3 Operating-profit margin 36.0 36.2 33.8 33.9 35.0 37.7 39.3 41.2 Net profit margin 14.0 14.6 15.3 15.0 14.9 16.0 16.6 17.6 ROAE 8.0 8.1 8.6 8.5 8.2 8.5 8.6 8.9 ROAA 2.2 2.2 2.5 2.6 2.7 2.9 2.9 3.0 ROCE 6.3 6.2 6.2 6.5 7.0 7.4 7.5 7.7 ROIC 5.7 5.5 5.5 5.6 6.1 6.3 6.3 6.5 Net debt to equity 164.3 160.8 155.6 136.5 127.8 126.2 124.7 121.9 Effective tax rate 12.6 12.9 12.8 16.8 16.5 17.8 18.1 18.2 Accounts receivable (days) 131.4 98.8 83.0 96.9 122.4 158.3 192.9 239.5 Current ratio (x) 0.3 0.2 0.2 0.4 0.4 0.4 0.5 0.6 Net interest cover (x) 2.2 2.3 2.7 2.6 2.6 2.8 2.9 3.0 Net dividend payout 18.8 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Free cash flow yield n.a. 0.4 5.2 5.7 17.8 2.7 0.1 1.9 Source: FactSet, Daiwa forecasts

Company profile

Listed in December 2009, China Longyuan Power is Asia's largest and the world's second-largest wind-farm operator, with total operational experience over 20 years. Total attributable capacity of wind farms reached 18,919MW as of 2018. The group has developed projects across 29 provinces and autonomous regions in China and overseas.

Hong Kong Utilities 11 July 2019

Huaneng Renewables (958 HK)

Huaneng R enewabl es

Target price: HKD2.30 (from HKD2.45) Share price (9 Jul): HKD2.22 | Up/downside: +3.6%

Tightening cash flow for 2020-21E Anna Lu, CFA (852) 2848 4465  Finalised RPS leaves subsidy collection reliant on the RE fund [email protected]  Balance sheet likely to remain tight on a significant rise in finance costs Dennis Ip, CFA (852) 2848 4068  Maintaining our Hold (3) rating; lowering TP to HKD2.30 [email protected]

What's new: HNR’s share price has risen 6% YTD, underperforming the Forecast revisions (%) HSI by 3%. We expect HNR to report c.13% YoY net profit growth with a Year to 31 Dec 19E 20E 21E 9% YoY increase in wind power generation in 1H19 (1Q19: +11% YoY). Revenue change 11.5 19.3 n.a. Net profit change 0.7 (2.6) n.a. Similar to other wind-IPPs, HNR’s capex should rise from CNY3.8bn in Core EPS (FD) change (2.3) (5.5) n.a. 2018 to CNY11.9-13.6bn on plans to add 1.6-1.9GW of new wind power Source: Daiwa forecasts capacity pa in 2019-21E before the expiry of subsidy tariff, in our view. Therefore, HNR’s FCF is likely to turn negative. As such, we maintain our Share price performance

Hold (3) rating on HNR and prefer wind power equipment maker Goldwind (HKD) (%) in 2019-21E. 3.1 115 2.7 104 What's the impact: distressed balance sheet likely on rush installation 2.4 93 2.1 81 of wind farms. We expect HNR’s capex to rise by CNY8.1-9.8bn pa in 1.8 70 2019-21 vs. 2018, leading to a decline of CNY7.5-9.5bn pa in FCF pa Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 during the same period; as such, we see its finance costs rising by 55% H'neng Ren (LHS) Relative to HSI (RHS)

(peers: +14-46%), implying a 48% increase (peers: +18-52%) in EBIT. 12-month range 1.90-3.03 Monthly data and 2H19 results unlikely to be positive catalysts. In Market cap (USDbn) 3.00 2018, HNR only added 347MW wind capacity, vs. 434-2,194MW in 2015- 3m avg daily turnover (USDm) 4.17 Shares outstanding (m) 10,567 17. Also, HNR’s utilisation for 3Q18 was up 2% YoY, its peak level since Major shareholder Huaneng Group (56.9%) 2015, which we see more as a downside risk to utilisation for 3Q19 due to the high base in 2018. Financial summary (CNY) Year to 31 Dec 19E 20E 21E Upcoming RPS/GC not likely to improve subsidy collection. As per the Revenue (m) 13,058 14,637 16,109 final draft of the RPS, besides actual renewable consumption, there are 2 Operating profit (m) 6,822 7,692 8,462 Net profit (m) 3,779 3,839 4,095 ways to remain compliant: 1) buy excess renewable energy volumes from Core EPS (fully-diluted) 0.358 0.363 0.388 other parties, or 2) buy GCs from renewable energy producers on a EPS change (%) 22.4 1.6 6.7 voluntary basis. Thus, we cut our 2020E EPS by 6% due to higher finance Daiwa vs Cons. EPS (%) 4.3 (4.9) (6.6) costs given the GC will not be mandatory to improve working capital. PER (x) 5.5 5.4 5.0 Dividend yield (%) 2.7 2.8 3.0 DPS 0.054 0.054 0.058 What we recommend: We maintain our Hold (3) rating on HNR, but cut PBR (x) 0.7 0.6 0.6 our TP to HKD2.30 (from HKD2.45), now based on a target PER of 5.6x EV/EBITDA (x) 6.7 6.6 6.7 ROE (%) 13.0 11.9 11.5 applied to our average 2019-20E EPS (previously: 6.0x 2019E PER), as we Source: FactSet, Daiwa forecasts expect HNR’s balance sheet to remain stretched with the untimely receipt of subsidy receivables during the rush installation period over 2019-21E. Upside/downside risks: better-/worse-than-expected subsidy receivables.

How we differ: Our 2019E EPS is 4% above the consensus, likely as we have factored in the strong 1H19 wind power generation data. However, our 2020-21E EPS are 5-7% below the consensus, likely as we assume higher finance costs on surging capex and subsidy receivables, and also a lower wind tariff.

See important disclosures, including any required research certifications, beginning on page 105

Huaneng Renewables (958 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook HNR: net profit forecasts (CNYm)

We forecast HNR’s earnings growth to slow to a 10% 5,000 70% CAGR over 2018-21, from an 18% CAGR for 2015-18, 4,095 3,779 3,839 60% mainly due to a contraction in new capacity additions in 4,000 3,012 3,086 50% 2016-18 and rising finance costs in 2019-21E, in our view. 2,659 3,000 40% 1,860 2,000 30% 1,121 888 20% 1,000 10% 0 0% 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E

Net profit - LHS Net prfit (YoY) - RHS

Source: Company, Daiwa forecasts

Valuation HNR: 1-year forward PER HNR is trading at 5.5x and 0.7x 2019E PER and PBR, PER (x) Last rush respectively, 1.0SD and 1.4SD below the stock’s past-8- 25 year averages, which on the face of it we consider installation 20 20.7x Avg+2SD undemanding. However, with the prospect of a slowdown in its earnings growth on rising finance costs, we believe 15 15.6x Avg+1SD the prevailing valuation is fair. 10 10.5x Avg Our new TP of HKD2.3 is based on a PER of 5.6x (1SD 5.3x Avg-1SD below past-8-year average) applied to our average 2019- 5

20E EPS. HNR’s PBR was de-rated from 12x PER to 5x 0 0.2x Avg-2SD PER (or 1.6x PBR to 0.5x PBR) in 2015 (last rush Jun-11 Jun-12 Jun-13 Jun-14 Jun-15 Jun-16 Jun-17 Jun-18 Jun-19 installation), leading to a significant 10% drop in utilisation Source: Bloomberg, Daiwa forecasts hours, vs. 2013, on a surging curtailment rate. We believe finance costs will surge during this installation round on worsening receivables.

Earnings revisions HNR: Bloomberg-consensus EPS forecast revisions (CNY)

The 2019-20 Bloomberg-consensus EPS forecasts have 0.50 been revised down by 5% since March, when the company reported lower-than-expected 2018 results and announced 0.45 plans to install over 1GW of new wind farms, which we believe could result in declining FCF, rising debt and higher 0.40 finance expenses. 0.35

Our 2019E EPS is 4% above the consensus, likely as we 0.30

have factored in the strong 1H19 wind power generation

Jul-18

Oct-18 Oct-18 Apr-19

Jun-18 Jan-19 Jan-19 Jun-19

Mar-19 Mar-19

Feb-19

Aug-18 Aug-18 Sep-18 Nov-18 Dec-18 May-19 data, but our 2020-21E EPS are 5-7% below the May-19 consensus, likely as we see higher finance costs and 2019E EPS 2020E EPS 2021E EPS subsidy receivables during the rush installation period in Source: Bloomberg 2019-21E.

Huaneng Renewables (958 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Consolidated capacity - Total (MW) 8,011 10,345 11,087 11,567 11,965 13,595 15,525 17,355 Consolidated capacity - wind (MW) 7,526 9,720 10,252 10,687 11,033 12,633 14,533 16,333 Wind power utilization hours (hours) 1,875 1,882 1,966 2,082 2,234 2,346 2,346 2,346 Solar power utilization hours (hours) 1,420 1,591 1,373 1,449 1,557 1,588 1,588 1,588 Consolidated wind power output (Million 11,170 13,232 17,684 20,344 24,541 25,878 29,884 34,003 Kwh) Average on-grid wind farm tariff 0.598 0.598 0.562 0.553 0.541 0.538 0.527 0.514 (CNY/kWh) Average on-grid solar farm tariff 0.99 0.89 0.89 0.86 0.96 0.92 0.90 0.88 (CNY/kWh)

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of electricity 6,114 7,354 9,233 10,548 11,645 13,053 14,632 16,104 Service concession construction 33 0 0 0 0 0 0 0 revenue Other Revenue 4 3 6 6 5 5 5 5 Total Revenue 6,151 7,357 9,239 10,554 11,650 13,058 14,637 16,109 Other income 168 435 322 469 388 399 442 481 COGS (120) (109) (189) (191) (265) (266) (307) (349) SG&A (447) (559) (626) (762) (925) (997) (1,074) (1,134) Other op.expenses (2,406) (3,008) (3,837) (4,419) (4,953) (5,372) (6,006) (6,645) Operating profit 3,346 4,116 4,908 5,651 5,896 6,822 7,692 8,462 Net-interest inc./(exp.) (2,112) (2,073) (1,995) (2,238) (2,189) (2,236) (2,983) (3,383) Assoc/forex/extraord./others (2) (3) (3) (5) (11) (11) (11) (11) Pre-tax profit 1,232 2,041 2,910 3,408 3,695 4,574 4,698 5,068 Tax (86) (141) (202) (346) (567) (747) (814) (929) Min. int./pref. div./others (26) (40) (49) (50) (42) (48) (44) (43) Net profit (reported) 1,121 1,860 2,659 3,012 3,086 3,779 3,839 4,095 Net profit (adjusted) 1,121 1,860 2,659 3,012 3,086 3,779 3,839 4,095 EPS (reported)(CNY) 0.124 0.191 0.273 0.294 0.292 0.358 0.363 0.388 EPS (adjusted)(CNY) 0.124 0.191 0.273 0.294 0.292 0.358 0.363 0.388 EPS (adjusted fully-diluted)(CNY) 0.124 0.191 0.273 0.294 0.292 0.358 0.363 0.388 DPS (CNY) 0.020 0.030 0.041 0.043 0.044 0.054 0.054 0.058 EBIT 3,346 4,116 4,908 5,651 5,896 6,822 7,692 8,462 EBITDA 5,616 6,851 8,370 9,449 9,869 11,135 12,555 13,898

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 1,232 2,041 2,910 3,408 3,695 4,574 4,698 5,068 Depreciation and amortisation 2,270 2,735 3,462 3,798 3,973 4,313 4,863 5,436 Tax paid (86) (141) (202) (348) (567) (747) (814) (929) Change in working capital 13 347 (3,079) (2,664) (2,297) (1,921) (3,616) (5,271) Other operational CF items 2,695 2,701 4,248 3,486 3,127 2,324 3,121 3,509 Cash flow from operations 6,125 7,682 7,338 7,681 7,931 8,543 8,251 7,813 Capex (13,827) (13,549) (6,909) (3,188) (3,769) (11,868) (13,630) (12,913) Net (acquisitions)/disposals (513) (468) (16) 0 (135) 0 0 0 Other investing CF items 1,379 1,503 (444) (1,003) (313) 0 0 0 Cash flow from investing (12,961) (12,514) (7,370) (4,191) (4,217) (11,868) (13,630) (12,913) Change in debt 10,489 5,246 1,277 (1,157) 423 8,500 8,500 8,500 Net share issues/(repurchases) 1,358 0 0 1,909 0 0 0 0 Dividends paid (382) (480) (292) (433) (454) (567) (576) (614) Other financing CF items (2,577) (2,989) (2,843) (2,846) (2,791) (2,313) (3,110) (3,498) Cash flow from financing 8,889 1,778 (1,858) (2,527) (2,822) 5,620 4,814 4,388 Forex effect/others 10 139 85 (126) (14) 0 0 0 Change in cash 2,062 (2,915) (1,804) 837 878 2,296 (564) (712) Free cash flow (7,702) (5,867) 429 4,493 4,162 (3,324) (5,378) (5,100) Source: FactSet, Daiwa forecasts

Huaneng Renewables (958 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 7,788 4,531 2,605 2,555 3,482 5,778 5,214 4,502 Inventory 13 43 33 55 44 54 60 66 Accounts receivable 3,160 2,900 4,635 7,214 9,968 11,202 15,256 19,865 Other current assets 180 178 1,423 1,418 1,088 1,712 1,422 2,023 Total current assets 11,141 7,652 8,696 11,242 14,582 18,746 21,952 26,455 Fixed assets 57,873 68,658 72,107 71,406 70,812 78,417 87,234 94,763 Goodwill & intangibles 696 682 654 621 589 548 507 467 Other non-current assets 4,427 5,561 3,989 3,080 2,689 2,667 2,646 2,625 Total assets 74,138 82,553 85,445 86,349 88,671 100,378 112,340 124,309 Short-term debt 17,306 18,727 22,563 20,353 21,614 24,685 28,258 31,831 Accounts payable 434 496 435 352 472 419 573 518 Other current liabilities 8,001 9,311 7,520 6,381 6,945 6,945 6,945 6,945 Total current liabilities 25,741 28,533 30,517 27,086 29,031 32,048 35,776 39,294 Long-term debt 26,842 30,677 28,372 29,451 28,611 34,040 38,967 43,894 Other non-current liabilities 4,670 4,718 5,561 4,369 2,816 2,816 2,816 2,816 Total liabilities 57,253 63,928 64,451 60,907 60,458 68,905 77,559 86,004 Share capital 9,728 9,728 9,728 10,567 10,567 10,567 10,567 10,567 Reserves/R.E./others 6,373 8,070 10,409 14,035 16,842 20,054 23,317 26,798 Shareholders' equity 16,101 17,798 20,137 24,602 27,408 30,621 33,884 37,365 Minority interests 784 827 857 840 805 853 897 940 Total equity & liabilities 74,138 82,553 85,445 86,349 88,671 100,378 112,340 124,309 EV 57,709 66,268 69,758 68,513 68,024 74,287 83,406 92,673 Net debt/(cash) 36,360 44,873 48,330 47,249 46,743 52,947 62,011 71,224 BVPS (CNY) 1.783 1.830 2.070 2.400 2.594 2.898 3.207 3.536

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 6.1 19.6 25.6 14.2 10.4 12.1 12.1 10.1 EBITDA (YoY) 20.5 22.0 22.2 12.9 4.4 12.8 12.8 10.7 Operating profit (YoY) 22.7 23.0 19.2 15.1 4.3 15.7 12.8 10.0 Net profit (YoY) 26.3 65.9 43.0 13.3 2.5 22.4 1.6 6.7 Core EPS (fully-diluted) (YoY) 19.6 54.2 43.0 7.5 (0.6) 22.4 1.6 6.7 Gross-profit margin 98.1 98.5 98.0 98.2 97.7 98.0 97.9 97.8 EBITDA margin 91.3 93.1 90.6 89.5 84.7 85.3 85.8 86.3 Operating-profit margin 54.4 56.0 53.1 53.5 50.6 52.2 52.5 52.5 Net profit margin 18.2 25.3 28.8 28.5 26.5 28.9 26.2 25.4 ROAE 7.5 11.0 14.0 13.5 11.9 13.0 11.9 11.5 ROAA 1.7 2.4 3.2 3.5 3.5 4.0 3.6 3.5 ROCE 6.2 6.4 7.0 7.7 7.7 8.1 8.0 7.8 ROIC 6.6 6.6 6.9 7.1 6.8 7.2 7.0 6.7 Net debt to equity 215.3 240.9 230.2 185.7 165.7 168.2 178.3 185.9 Effective tax rate 6.9 6.9 6.9 10.2 15.3 16.3 17.3 18.3 Accounts receivable (days) 180.0 150.3 148.9 204.9 269.2 295.9 329.9 397.9 Current ratio (x) 0.4 0.3 0.3 0.4 0.5 0.6 0.6 0.7 Net interest cover (x) 1.6 2.0 2.5 2.5 2.7 3.1 2.6 2.5 Net dividend payout 16.1 15.7 15.0 14.6 15.1 15.0 15.0 15.0 Free cash flow yield n.a. n.a. 2.1 21.7 20.1 n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

Listed in July 2011, Huaneng Renewables is one of the big-5 wind farm operators in China. Its total consolidated capacity of wind farms and solar farms reached 11,033MW and 932MW, respectively, by the end of 2018. The group has developed wind farm and solar farm projects across 21 provinces and autonomous regions in China. Huaneng Group, one of the big-5 independent power producer groups, is the company’s largest shareholder.

China Utilities 11 July 2019

China Datang Corp Renewable Power (1798 HK)

China D atang C orp Renewable Power

Target price: HKD0.80 (from HKD1.75) Share price (9 Jul): HKD0.78 | Up/downside: +2.6%

Downgrading: tightening CF despite robust utilisation Anna Lu, CFA (852) 2848 4465  We expect DTR to see a rise in finance costs on rush installations… [email protected]  … despite strong utilisation growth on curtailment improvement Dennis Ip, CFA (852) 2848 4068  Downgrading to Hold (3) from Buy (1); lowering TP to HKD0.8 [email protected]

What's new: DTR’s share price has fallen by 17% YTD, underperforming Forecast revisions (%) the HSI by 26%. We expect DTR to report flat EPS with a 9% YoY increase Year to 31 Dec 19E 20E 21E in wind power generation in 1H19 (1Q19: +11% YoY). As with other wind Revenue change (2.2) 3.7 n.a. Net profit change (14.7) (15.1) n.a. IPPs, we see DTR’s capex rising, from CNY2.3bn in 2018 to CNY7.3-9.4bn Core EPS (FD) change (14.7) (15.1) n.a. to add 1.0-1.3GW of new wind power capacity pa in 2019-21E before the Source: Daiwa forecasts expiry of subsidy tariffs. Hence, DTR’s FCF is likely to face pressure. Hence, we downgrade DTR to Hold (3), and prefer wind power equipment Share price performance maker Goldwind in 2019-21. (HKD) (%) 1.6 115 What's the impact: distressed balance sheet likely on rush 1.3 99 installations of wind farms. We expect DTR’s capex to rise by CNY5.0- 1.1 83 0.9 66 7.1bn pa in 2019-21 vs. 2018, leading to a decline of CNY5.0-8.3bn pa in 0.7 50 FCF during the same period; as such, we see its finance costs rising by Jul-18 Oct-18 Jan-19 Apr-19 Jul-19

39% (peers: +14-55%), implying a 52% increase (peers: +18-48%) in EBIT. China Data (LHS) Relative to HSI (RHS) Hence, we expect DTR’s share price to remain subdued on a worsening balance sheet (from 306% net debt to equity at end-2018 to 342% at end- 12-month range 0.76-1.51 2021E) and accounts receivables (from 275 days at end-2018 to over 600 Market cap (USDbn) 0.73 days at end-2021E) given prolonged delays in subsidy payments, 3m avg daily turnover (USDm) 0.23 especially for the new 3GW of capacity commissioned in 2019-21E. Shares outstanding (m) 7,274 Major shareholder Datang Corporation (65.6%) Accounts receivable days rose from 202 at end-2017 to 275 in 2018, with only 500MW of new capacity installed in 2017. Financial summary (CNY) Year to 31 Dec 19E 20E 21E Likely to see strongest EPS growth on falling wind curtailment rate. Revenue (m) 8,914 10,113 11,016 The grid curtailment rate on wind power in China dropped to 7% in 2018 Operating profit (m) 4,216 4,951 5,400 from 17% in 2016 and 12% in 2017, and should continue to drop to meet Net profit (m) 1,345 1,493 1,577 Core EPS (fully-diluted) 0.185 0.205 0.217 the government’s targeted 5% in 2020-21E, in our view. We expect DTR, EPS change (%) 23.1 11.0 5.6 with the largest exposure to curtailed provinces among Hong Kong-listed Daiwa vs Cons. EPS (%) (5.6) (7.5) (8.5) wind IPPs, to see its curtailment rate drop from 8.1% (peers: 5-9%) in 2018 PER (x) 3.7 3.3 3.2 to 5-7% in 2019-21. Hence, we expect it to record a 13% EPS CAGR over Dividend yield (%) 4.0 4.5 4.7 DPS 0.028 0.031 0.033 2018-21, outpacing the 10% CAGR of peers during the same period. PBR (x) 0.4 0.3 0.3 EV/EBITDA (x) 7.5 7.5 7.6 What we recommend: We cut our 2019-20E EPS by 15% to reflect lower ROE (%) 10.5 10.7 10.4 2019E power generation volume, and higher finance costs and working Source: FactSet, Daiwa forecasts capital on rising subsidy receivables for 2019-20E. We downgrade our rating to Hold (3) and cut our TP to HKD0.80 from HKD1.75, based on a lower 0.4x PBR (previously 0.8x PBR) as we see a more stretched balance sheet eroding cash-based returns, which argues for a bigger discount with reference to the Chinese solar farm companies in 2017-18 given DTR’s SOE status. Key upside/downside risks: declining/rising interest rates, and stronger-/weaker-than-expected curtailment improvement.

How we differ: Our 2019E EPS is 6% below the consensus, likely as we have factored in lower 2019E power generation. Our 2020-21E EPS are 8- 9% below due to our higher finance cost assumptions.

See important disclosures, including any required research certifications, beginning on page 105

China Datang Corp Renewable Power (1798 HK): 11 July 2019

How do we justify our view? Growth outlook Valuation Earnings revisions

Growth outlook DTR: net profit forecasts (CNYm) 1,800 1600% We forecast DTR to see a 1% CAGR in wind utilisation 1,577 over 2018-21E vs. a 0-1% CAGR for peers. In turn, we 1,600 1,493 1400% 1,345 expect DTR to realise the strongest earnings growth 1,400 1200% among peers, with a 13% net profit CAGR for 2018-21 1,200 1,093 1000% compared with its peers’ 10% CAGR over the same period. 1,000 800% 800 612 600% 600 400%

400 198 200% 200 14 0% 0 (200%) 2015 2016 2017 2018 2019E 2020E 2021E Net profit YoY

Source: Company, Daiwa forecasts

Valuation DTR: 1-year forward PBR

DTR is trading currently at a 0.4x 2019E PBR, 1.2SD PBR (x) below its average since its IPO in 2010. By contrast, we 1.8 find that major pure wind-IPPs are trading currently at 0.7x 1.6 1.4 2019E PBRs. Despite DTR’s low valuation, we downgrade Last rush 1.2x Avg+2SD 1.2 our rating to Hold (3) in view of its stretched balance sheet. installation 1.0 1.0x Avg+1SD 0.8 0.7x Avg DTR’s PBR was de-rated from 0.7x to 0.4x in 2015 (last 0.6 rush installation), due to a significant 14% drop in utilisation 0.4 0.4x Avg-1SD 0.2 hours over 2013-15 on surging curtailments. We see 0.1x Avg-2SD 0.0

finance costs surging during this round of installations on

Jun-13 Jun-19 Jun-11 Jun-12 Jun-14 Jun-15 Jun-16 Jun-17 Jun-18

Dec-17 Dec-11 Dec-12 Dec-13 Dec-14 Dec-15 Dec-16 Dec-18 worsening receivables. Dec-10

Source: Bloomberg, Daiwa

Earnings revisions DTR: Bloomberg-consensus EPS forecast revisions (CNY) The 2019-20 Bloomberg-consensus EPS for DTR has 0.30 fallen by 1% YTD, likely due to DTR’s weak 1Q19 results.

0.25 Our 2019-21E EPS are 6-9% below the consensus, likely as we have factored in lower power generation in 2019E, and higher finance cost and working capital assumptions 0.20 on rising subsidy receivables and capex after the company revised up its installation target to 1,000-1,300MW in 2019-

20 before the expiry of the benchmark subsidy tariff. 0.15

Jul-18

Oct-18 Apr-19

Jun-19 Jun-18 Jan-19

Feb-19 Mar-19

Nov-18 Dec-18

Sep-18 Aug-18 May-19 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

China Datang Corp Renewable Power (1798 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Consolidated capacity - Total (MW) 6,033 7,146 8,498 8,826 9,015 10,015 11,315 12,515 Consolidated capacity - wind (MW) 5,916 7,029 8,345 8,647 8,835 9,835 11,135 12,335 Wind power utilization hours (hours) 1,803 1,745 1,755 1,905 2,096 2,117 2,180 2,180 Solar power utilization hours (hours) 1,653 1,544 1,476 1,474 1,473 1,488 1,503 1,503 Consolidated wind power output (Million 9,859 10,175 11,401 14,299 16,988 18,471 21,383 24,066 Kwh) Average on-grid wind farm tariff 0.589 0.587 0.570 0.562 0.553 0.544 0.535 0.519 (CNY/kWh) Average on-grid solar farm tariff 0.96 0.96 0.94 0.88 0.84 0.81 0.78 0.77 (CNY/kWh)

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of electricity 5,132 5,264 5,718 7,045 8,237 8,831 10,030 10,933 Services under energy performance 44 37 41 41 79 79 79 79 contracts Other Revenue 11 288 27 19 4 4 4 4 Total Revenue 5,186 5,588 5,786 7,104 8,319 8,914 10,113 11,016 Other income 231 117 189 204 270 391 442 481 COGS (104) (385) (210) (193) (253) (252) (290) (327) SG&A (476) (455) (489) (522) (622) (649) (717) (760) Other op.expenses (2,713) (2,780) (3,161) (3,726) (3,929) (4,187) (4,597) (5,010) Operating profit 2,124 2,084 2,115 2,868 3,785 4,216 4,951 5,400 Net-interest inc./(exp.) (2,234) (1,939) (1,735) (1,870) (2,112) (2,158) (2,648) (2,945) Assoc/forex/extraord./others 49 10 22 61 56 56 56 56 Pre-tax profit (62) 155 401 1,059 1,729 2,115 2,358 2,511 Tax (66) (92) (108) (156) (303) (391) (460) (515) Min. int./pref. div./others (23) (49) (95) (291) (333) (378) (405) (420) Net profit (reported) (150) 14 198 612 1,093 1,345 1,493 1,577 Net profit (adjusted) (150) 14 198 612 1,093 1,345 1,493 1,577 EPS (reported)(CNY) (0.021) 0.002 0.027 0.084 0.150 0.185 0.205 0.217 EPS (adjusted)(CNY) (0.021) 0.002 0.027 0.084 0.150 0.185 0.205 0.217 EPS (adjusted fully-diluted)(CNY) (0.021) 0.002 0.027 0.084 0.150 0.185 0.205 0.217 DPS (CNY) 0.000 0.000 0.012 0.018 0.020 0.028 0.031 0.033 EBIT 2,124 2,084 2,115 2,868 3,785 4,216 4,951 5,400 EBITDA 4,485 4,605 4,944 6,027 7,182 7,860 8,961 9,803

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax (62) 155 401 1,059 1,729 2,115 2,358 2,511 Depreciation and amortisation 2,362 2,521 2,830 3,159 3,397 3,643 4,010 4,403 Tax paid (66) (92) (108) (156) (303) (391) (460) (515) Change in working capital 640 3,392 1,904 (4,424) (4,299) (2,262) (4,550) (4,610) Other operational CF items 2,320 1,384 (1,144) 3,566 4,782 2,153 2,663 2,925 Cash flow from operations 5,193 7,360 3,883 3,205 5,306 5,258 4,021 4,714 Capex (5,520) (6,397) (8,495) (3,686) (2,299) (7,275) (9,363) (8,470) Net (acquisitions)/disposals 0 0 0 0 0 0 0 0 Other investing CF items (215) 1,332 2,878 (827) (2,781) (20) (20) (20) Cash flow from investing (5,735) (5,064) (5,616) (4,513) (5,080) (7,295) (9,383) (8,490) Change in debt 2,006 (954) 3,973 3,446 4,578 6,000 6,000 6,000 Net share issues/(repurchases) 1,979 0 0 0 0 0 0 0 Dividends paid (56) (312) (253) (242) (318) (261) (318) (340) Other financing CF items (2,198) (2,140) (1,896) (1,838) (2,078) (2,325) (2,835) (3,097) Cash flow from financing 1,732 (3,407) 1,824 1,366 2,182 3,414 2,847 2,563 Forex effect/others (1) (1) (2) 0 1 0 0 0 Change in cash 1,189 (1,112) 88 58 2,409 1,377 (2,514) (1,213) Free cash flow (327) 963 (4,612) (482) 3,007 (2,017) (5,342) (3,756) Source: FactSet, Daiwa forecasts

China Datang Corp Renewable Power (1798 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 2,540 1,083 1,179 1,279 3,674 5,051 2,536 1,323 Inventory 35 48 114 138 168 156 200 187 Accounts receivable 3,279 1,337 2,801 5,042 7,473 12,222 15,368 22,493 Other current assets 828 1,141 1,537 1,262 1,486 1,461 1,887 1,762 Total current assets 6,683 3,609 5,631 7,722 12,801 18,889 19,992 25,766 Fixed assets 48,783 52,523 57,914 58,088 56,430 60,170 65,633 69,812 Goodwill & intangibles 582 812 807 635 564 492 417 341 Other non-current assets 4,061 3,870 4,440 4,103 4,622 4,662 4,703 4,743 Total assets 60,110 60,815 68,792 70,548 74,417 84,213 90,744 100,662 Short-term debt 6,223 10,574 10,166 13,315 14,627 15,804 17,485 19,166 Accounts payable 434 1,318 2,362 1,429 364 1,535 557 1,719 Other current liabilities 4,294 5,170 7,982 6,518 5,978 7,257 7,302 8,518 Total current liabilities 10,951 17,062 20,511 21,262 20,969 24,596 25,344 29,402 Long-term debt 35,020 29,724 34,160 34,507 35,781 40,604 44,923 49,242 Other non-current liabilities 491 449 416 410 2,385 2,385 2,385 2,385 Total liabilities 46,462 47,235 55,086 56,179 59,135 67,585 72,652 81,029 Share capital 7,274 7,274 7,274 7,274 7,274 7,274 7,274 7,274 Reserves/R.E./others 3,645 3,492 3,606 4,120 5,018 6,102 7,278 8,514 Shareholders' equity 10,918 10,765 10,879 11,394 12,292 13,376 14,551 15,788 Minority interests 2,730 2,814 2,826 2,975 2,990 3,252 3,541 3,845 Total equity & liabilities 60,110 60,815 68,792 70,548 74,417 84,213 90,744 100,662 EV 45,767 46,370 50,289 53,779 53,922 58,752 67,499 74,960 Net debt/(cash) 38,703 39,215 43,148 46,542 46,733 51,357 59,871 67,084 BVPS (CNY) 1.501 1.480 1.496 1.566 1.690 1.839 2.001 2.171

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) (7.9) 7.8 3.5 22.8 17.1 7.1 13.5 8.9 EBITDA (YoY) (3.7) 2.7 7.4 21.9 19.2 9.4 14.0 9.4 Operating profit (YoY) (13.6) (1.8) 1.5 35.6 32.0 11.4 17.4 9.1 Net profit (YoY) n.a. n.a. 1,345.5 208.6 78.7 23.1 11.0 5.6 Core EPS (fully-diluted) (YoY) n.a. n.a. 1,345.5 208.6 78.7 23.1 11.0 5.6 Gross-profit margin 98.0 93.1 96.4 97.3 97.0 97.2 97.1 97.0 EBITDA margin 86.5 82.4 85.5 84.8 86.3 88.2 88.6 89.0 Operating-profit margin 41.0 37.3 36.6 40.4 45.5 47.3 49.0 49.0 Net profit margin (2.9) 0.2 3.4 8.6 13.1 15.1 14.8 14.3 ROAE n.a. 0.1 1.8 5.5 9.2 10.5 10.7 10.4 ROAA n.a. 0.0 0.3 0.9 1.5 1.7 1.7 1.6 ROCE 4.0 3.8 3.8 4.8 5.9 6.1 6.4 6.4 ROIC 4.1 1.6 2.8 4.2 5.1 5.3 5.5 5.2 Net debt to equity 283.6 288.8 314.8 323.9 305.8 308.9 330.9 341.7 Effective tax rate 0.0 59.4 27.0 14.8 17.5 18.5 19.5 20.5 Accounts receivable (days) 249.3 150.7 130.5 201.5 274.5 403.2 497.9 627.2 Current ratio (x) 0.6 0.2 0.3 0.4 0.6 0.8 0.8 0.9 Net interest cover (x) 1.0 1.1 1.2 1.5 1.8 2.0 1.9 1.8 Net dividend payout n.a. 0.0 44.0 21.4 13.3 15.0 15.0 15.0 Free cash flow yield n.a. 19.3 n.a. n.a. 60.1 n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

Listed in December 2010, Datang Renewables is one of the leading wind farm operators in China. Its total consolidated capacity of wind farms and solar farms reached 8,835MW and 174MW, respectively, by the end of 2018. Datang Corporation, one of the big-5 independent power producer groups, is the company’s largest shareholder.

China Utilities 11 July 2019

(836 HK) China Resources Power China R esources Power

Target price: HKD11.50 (from HKD11.50) Share price (9 Jul): HKD11.30 | Up/downside: +1.8%

Aggressive plan to add wind power capacity Dennis Ip, CFA (852) 2848 4068  Coal-fired power utilisation hours dropped by 6% YoY in 5M19 [email protected]  Rising accounts receivables and capex likely to worsen cash flow Anna Lu, CFA (852) 2848 4465  Maintaining our Hold (3) rating and TP of HKD11.50 [email protected]

What's new: CRP’s share price dropped significantly by 19% in two days Forecast revisions (%) after it unexpectedly announced a dividend policy change on 22 March, Year to 31 Dec 19E 20E 21E and has remained at a similar level since then. We think its coal-fired power Revenue change (2.2) (4.1) (5.5) utilisation YTD is slightly lower than expected. It is also aggressively adding Net profit change (13.7) (11.6) (15.5) Core EPS (FD) change (13.7) (11.6) (15.5) wind capacity, which we believe would result in rising accounts receivables Source: Daiwa forecasts and capex, and worsening cash flow. As such, we maintain our Hold (3). Share price performance What's the impact: coal-fired power utilisation hours dropped by 6% (HKD) (%) YoY in 5M19. CRP’s consolidated power generation was down 3.9% YoY 17 125 in 5M19, and net power generation of the coal-fired segment was down 15 111 5.9% YoY in 5M19, with utilisation down 6.1% YoY, likely due to 13 98 maintenance of power plants, lower temperatures compared with the same 12 84 10 70 period last year, and lower market share in direct power sales in Henan Jul-18 Oct-18 Jan-19 Apr-19 Jul-19

Province (27% YoY drop in coal-fired utilisation in Henan) amid hydropower Ch Res Pow (LHS) Relative to HSI (RHS) utilisation recovery. Management previously expected 4,800 utilisation hours in 2019, down 3.5% YoY from 4,976 in 2018. Thus, we think coal- 12-month range 10.44-16.14 fired utilisation in 5M19 is slightly lower than our expectation. Market cap (USDbn) 6.94 3m avg daily turnover (USDm) 13.58 Plans to aggressively add wind power capacity in coming years. Shares outstanding (m) 4,797 Major shareholder China Resources Group (63.0%) According to the latest policy, projects approved in or before 2018 are still entitled to fixed tariffs if they are connected on the grid before 2020. Financial summary (HKD) According to CRP, it still has 9GW of wind capacity that has gained Year to 31 Dec 19E 20E 21E approval, but is not yet in operation. As such, it is hastening the progress of Revenue (m) 71,495 70,855 71,678 installation via heavy capex in the coming years and plans to add wind Operating profit (m) 12,517 13,608 14,987 capacity of 1.7GW and 2.3GW in 2019-20, respectively, vs. 1.2GW in 2018. Net profit (m) 7,017 8,896 9,600 Core EPS (fully-diluted) 1.468 1.861 2.008 EPS change (%) 1.0 26.8 7.9 Likely worsening cash flow on upcoming rush installation. With Daiwa vs Cons. EPS (%) (6.6) (0.4) (1.9) management guiding for 1.7GW/2.3GW of wind installation in 2019-20, PER (x) 7.7 6.1 5.6 CRP expects capex of HKD19bn in 2019 (including HKD13.7bn for wind). Dividend yield (%) 3.9 4.9 5.3 DPS 0.440 0.558 0.602 With 2019-21 being the construction peak, we expect capex to remain high PBR (x) 0.7 0.7 0.6 at HKD19-24bn in 2019-21, which, coupled with accounts receivables EV/EBITDA (x) 6.6 6.4 6.3 building up due to rising subsidy receivables (we estimate accounts ROE (%) 9.7 11.4 11.3 receivables days to rise from 80 days in 2018 to 190 days in 2021), would Source: FactSet, Daiwa forecasts result in a worsening cash flow (negative FCF in 2019-21E).

What we recommend: We lower our 2019-21E EPS by 12-16% on lower coal-fired power utilisation hours and HKD:CNY rate, and maintain our SOTP-based 12-month TP of HKD11.5 and Hold (3) rating on CRP. Our TP represents a 0.7x 2019E PBR, similar to the current valuation of HNP and Longyuan. Upside/downside risks: higher-/lower-than-expected coal price.

How we differ: Our 2019-21E EPS are 0-7% below the consensus, likely on our lower coal-fired utilisation hours and higher finance cost estimates.

See important disclosures, including any required research certifications, beginning on page 105

China Resources Power (836 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook CRP: recurring profit forecasts (HKDm)

We forecast CRP to record an 11% recurring profit CAGR 12,000 60% over 2018-21 on lowering fuel cost for the coal-power 50% 10,000 segment and strong capacity growth for the wind-power 40% 30% segment. Also, we expect a higher wind-IPP profit 8,000 20% contribution with a 2018-21E sales CAGR of c.19%, mainly 10% 6,000 driven by 1.7-2.3GW pa capacity additions. 0% 4,000 (10%) (20%) However, we expect CRP’s capex to rise by CNY1.4-6.3bn 2,000 (30%) pa in 2019-21 vs. 2018, leading to a decline of CNY5.3- (40%) 7.8bn pa in FCF in the same period, also on a surge in 0 (50%) 2014 2015 2016 2017 2018 2019E 2020E 2021E subsidy receivables. As such, we expect its finance costs Net profit (LHS) YoY growth (RHS) to increase by 27% (pure wind-IPPs peers: +14-55%), implying a 14% rise (peers: +18-52%) in EBIT. Source: Company, Daiwa forecasts

Valuation CRP: 1-year forward PBR

CRP’s shares are currently trading at a 0.7x 2019E PBR, PBR (x) 1.7SD below the stock’s past-8-year average. Although we 2.0 30% 1.8 view its current valuation as undemanding, we expect CRP 1.7x 25% 1.6 to stay at a distressed valuation level until the end of rush 1.4 1.5x 20% installations of wind farms by end-2021E; thereafter, we 1.2 1.2x 15% see FCF starting to improve on lower capex and a 1.0x 1.0 10% stabilising of surges in subsidy receivables. 0.8 0.8x 0.6 5% Our 12-month TP of HKD11.5 is based on 0.7x 2019E 0.4 0%

PBR, similar to the current valuation of HNP and

Jul-14

Apr-13 Oct-15 Apr-18

Jan-12 Jun-12 Jan-17 Jun-17

Feb-14 Mar-16 Feb-19

Dec-14 Sep-18 Nov-12 Sep-13 Aug-16 Nov-17 May-15 Longyuan, which are the proxies of coal-IPPs and wind- PB (LHS) ROE (RHS)

IPPs listed in Hong Kong. Source: Bloomberg , Daiwa forecasts

Earnings revisions CRP: Bloomberg-consensus EPS forecast revisions (HKD)

The 2019-20 Bloomberg-consensus EPS forecasts for (HKD) CRP have dropped by 7-8% since its disappointing 2018 2.2 results announcement in March 2019. Our 2019-21E EPS 2.1 are 0-7% below the consensus likely on our more 2.0 conservative coal-IPP utilisation hour assumption and 1.9 lower HKD:CNY FX rate, with rising finance costs likely 1.8 due to increasing subsidy receivables and capex from the 1.7 wind farm segment. 1.6

1.5

Jul-19

Apr-19 Apr-19

Jan-19 Jan-19 Jan-19 Jun-19 Jun-19

Mar-19 Feb-19 Feb-19 Mar-19

May-19 May-19 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

China Resources Power (836 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Attributable capacity - year end (MW) 31,331 34,731 36,184 36,077 37,438 39,416 42,832 44,503 Utilization hours - consolidated coal- 5,334 4,989 4,922 4,964 4,976 4,777 4,634 4,541 fired plants Consolidated tariff ex-VAT (CNY/MWh) 374 353 327 328 333 336 326 316 Unit fuel cost - consolidated coal-fired 186 146 156 214 221 208 205 204 plants (CNY/MWh)

Profit and loss (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of electricty 64,319 64,961 59,485 64,014 66,001 65,081 64,430 65,141 Heat supply 2,861 3,183 3,128 3,764 4,808 4,956 4,995 5,126 Other Revenue 3,500 3,292 3,600 5,534 6,131 1,458 1,430 1,410 Total Revenue 70,681 71,436 66,213 73,312 76,940 71,495 70,855 71,678 Other income 0 0 0 0 0 0 0 0 COGS (31,044) (25,954) (27,199) (38,072) (39,437) (36,419) (34,996) (34,541) SG&A 0 0 0 0 0 0 0 0 Other op.expenses (27,387) (27,418) (24,617) (24,743) (26,738) (22,559) (22,252) (22,151) Operating profit 12,249 18,063 14,397 10,497 10,765 12,517 13,608 14,987 Net-interest inc./(exp.) (3,325) (3,216) (3,557) (3,991) (4,315) (4,549) (5,018) (5,486) Assoc/forex/extraord./others 2,261 1,474 2,202 2,015 861 2,915 2,908 2,900 Pre-tax profit 11,184 16,321 13,042 8,520 7,311 10,883 11,498 12,400 Tax (4,291) (5,809) (4,398) (2,902) (2,286) (3,474) (2,187) (2,398) Min. int./pref. div./others 2,321 (487) (935) (995) (1,074) (392) (416) (402) Net profit (reported) 9,215 10,025 7,708 4,623 3,950 7,017 8,896 9,600 Net profit (adjusted) 9,215 10,025 7,708 4,623 6,949 7,017 8,896 9,600 EPS (reported)(HKD) 1.937 2.104 1.615 0.968 0.826 1.468 1.861 2.008 EPS (adjusted)(HKD) 1.937 2.104 1.615 0.968 1.454 1.468 1.861 2.008 EPS (adjusted fully-diluted)(HKD) 1.932 2.101 1.615 0.968 1.454 1.468 1.861 2.008 DPS (HKD) 0.748 0.799 0.875 0.876 0.328 0.440 0.558 0.602 EBIT 12,249 18,063 14,397 10,497 10,765 12,517 13,608 14,987 EBITDA 20,849 27,851 24,279 21,005 21,918 23,153 25,000 27,579

Cash flow (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 11,184 16,321 13,042 8,520 7,311 10,883 11,498 12,400 Depreciation and amortisation 8,600 9,787 9,882 10,507 11,153 10,636 11,392 12,593 Tax paid (4,291) (5,809) (4,398) (2,902) (2,286) (3,474) (2,187) (2,398) Change in working capital 837 2,761 (1,280) 2,522 (5,911) (6,115) (6,970) (8,120) Other operational CF items 10,170 8,929 5,049 (86) 7,830 2,324 2,800 3,277 Cash flow from operations 26,500 31,989 22,295 18,562 18,097 14,255 16,534 17,752 Capex (27,582) (17,150) (15,751) (16,000) (17,563) (19,000) (23,835) (23,000) Net (acquisitions)/disposals 1,173 2,171 0 0 0 0 0 0 Other investing CF items 4,391 1,317 483 1,279 9,973 762 822 896 Cash flow from investing (22,018) (13,662) (15,268) (14,721) (7,590) (18,238) (23,014) (22,104) Change in debt 13,123 (8,542) (1,966) 5,878 5,575 10,500 10,500 10,500 Net share issues/(repurchases) 54 20 3 34 0 0 0 0 Dividends paid (5,537) (6,030) (6,509) (5,956) (5,305) (2,105) (2,669) (2,880) Other financing CF items (9,847) (4,319) (1,109) (3,188) (7,716) (5,196) 649 (4,410) Cash flow from financing (2,207) (18,871) (9,580) (3,231) (7,445) 3,199 8,481 3,210 Forex effect/others (24) (467) (373) 425 (466) 0 0 0 Change in cash 2,250 (1,011) (2,927) 1,035 2,595 (784) 2,001 (1,142) Free cash flow (1,082) 14,839 6,543 2,562 534 (4,745) (7,302) (5,248) Source: FactSet, Daiwa forecasts

China Resources Power (836 HK): 11 July 2019

Financial summary continued … Balance sheet (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 8,285 7,274 4,347 5,382 7,977 7,193 9,194 8,052 Inventory 3,309 2,307 3,124 3,205 3,296 3,708 3,769 3,625 Accounts receivable 19,685 14,587 14,766 16,007 17,899 24,485 29,895 37,704 Other current assets 1,556 1,665 1,648 3,103 777 434 434 434 Total current assets 32,835 25,833 23,886 27,696 29,949 35,821 43,292 49,816 Fixed assets 146,908 143,098 138,528 144,546 151,542 159,425 171,395 181,457 Goodwill & intangibles 21,434 17,739 16,622 17,091 1,984 6,684 2,277 2,290 Other non-current assets 24,471 21,416 21,075 31,639 24,748 21,853 22,462 23,031 Total assets 225,648 208,086 200,111 220,972 208,223 223,783 239,427 256,594 Short-term debt 20,897 27,404 21,297 32,161 27,465 27,465 27,465 27,465 Accounts payable 31,961 28,622 28,338 32,182 28,254 29,138 27,639 27,184 Other current liabilities 3,198 3,640 2,027 3,265 3,863 2,735 2,735 2,735 Total current liabilities 56,056 59,665 51,661 67,607 59,582 59,338 57,839 57,384 Long-term debt 79,536 61,114 70,167 68,762 69,210 79,710 90,210 100,710 Other non-current liabilities 4,264 3,568 3,718 3,861 2,146 2,146 2,146 2,146 Total liabilities 139,856 124,347 125,547 140,231 130,939 141,194 150,195 160,241 Share capital 28,000 28,150 22,258 22,342 22,317 22,317 22,317 22,317 Reserves/R.E./others 48,692 48,665 46,763 52,872 47,819 52,731 58,958 65,678 Shareholders' equity 76,692 76,815 69,021 75,214 70,135 75,047 81,274 87,995 Minority interests 9,101 6,925 5,544 5,528 7,149 7,541 7,957 8,359 Total equity & liabilities 225,648 208,086 200,111 220,972 208,223 223,783 239,427 256,594 EV 144,236 132,894 138,411 146,532 140,577 151,993 160,659 172,466 Net debt/(cash) 92,148 81,243 87,116 95,541 88,698 99,982 108,481 120,123 BVPS (HKD) 14.757 14.752 14.356 15.645 14.588 15.610 16.905 18.303

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 1.6 1.1 (7.3) 10.7 4.9 (7.1) (0.9) 1.2 EBITDA (YoY) (5.9) 33.6 (12.8) (13.5) 4.3 5.6 8.0 10.3 Operating profit (YoY) (17.2) 47.5 (20.3) (27.1) 2.6 16.3 8.7 10.1 Net profit (YoY) (16.3) 8.8 (23.1) (40.0) 50.3 1.0 26.8 7.9 Core EPS (fully-diluted) (YoY) (16.4) 8.7 (23.1) (40.1) 50.2 1.0 26.8 7.9 Gross-profit margin 56.1 63.7 58.9 48.1 48.7 49.1 50.6 51.8 EBITDA margin 29.5 39.0 36.7 28.7 28.5 32.4 35.3 38.5 Operating-profit margin 17.3 25.3 21.7 14.3 14.0 17.5 19.2 20.9 Net profit margin 13.0 14.0 11.6 6.3 9.0 9.8 12.6 13.4 ROAE 13.6 14.1 11.0 6.4 9.6 9.7 11.4 11.3 ROAA 4.2 4.6 3.8 2.2 3.2 3.2 3.8 3.9 ROCE 6.8 10.1 8.5 6.0 6.1 6.9 6.9 6.9 ROIC 4.4 6.8 5.8 4.1 4.3 4.9 5.8 5.8 Net debt to equity 120.2 105.8 126.2 127.0 126.5 133.2 133.5 136.5 Effective tax rate 38.4 35.6 33.7 34.1 31.3 31.9 19.0 19.3 Accounts receivable (days) 98.2 87.6 80.9 76.6 80.4 108.2 140.1 172.1 Current ratio (x) 0.6 0.4 0.5 0.4 0.5 0.6 0.7 0.9 Net interest cover (x) 3.7 5.6 4.0 2.6 2.5 2.8 2.7 2.7 Net dividend payout 38.6 38.0 54.2 90.5 39.7 30.0 30.0 30.0 Free cash flow yield n.a. 27.4 12.1 4.7 1.0 n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

China Resources Power is one of the big-5 Hong Kong-listed independent power producers (IPP) in China, with a total attributable coal-fired power capacity of 29.8GW at the end of 2018. It also owns coal mines (2018 production: 15m tonnes) and a portfolio of wind power farms in China (6.8GW at the end of 2018).

China Utilities 11 July 2019

(902 HK) Huaneng Power International Huaneng Power Internati onal

Target price: HKD5.95 (from HKD5.95) Share price (9 Jul): HKD4.70 | Up/downside: +26.6%

1H19 likely to show benefits of coal-price drop Anna Lu, CFA (852) 2848 4465  1H19 earnings likely benefitted from YoY coal-price drop [email protected]  Large capex with aggressive renewable capacity additions in 2019-21E Dennis Ip, CFA (852) 2848 4068  Reiterating our Buy (1) rating and TP of HKD5.95 [email protected]

What's new: Due to a high base in 1H18, China’s thermal coal price Forecast revisions (%) recorded a 3-10% YoY decline in 1H19, which we believe will contribute to Year to 31 Dec 19E 20E 21E strong growth in HNP’s 1H19 earnings. Besides, HNP is stepping up the Revenue change - 0.2 0.8 pace of wind capacity installations, with heavy capex in the coming few Net profit change 25.0 17.5 10.3 Core EPS (FD) change 25.0 17.5 10.3 years, which we believe may lead to rising capex, worsening cash flow, and Source: Daiwa forecasts higher gearing. HNP is trading at 0.7x 2019E PBR, 0.9SD below its average 1-year forward PBR since 2007. We reaffirm our Buy (1) call and Share price performance TP of HKD5.95, based on a PBR of 0.9x, still 0.6SD below its average 1- (HKD) (%) year forward PBR of 1.2x since 2007, on our average 2019-20E BVPS. 6.5 115 5.9 106 What's the impact: HNP’s 1H19 results likely benefited from YoY coal- 5.3 98 price decline. The China Thermal Coal Price Index rose by 2-9% YoY to 4.6 89 4.0 80 CNY515-567/tonne in 1H18. With less extreme weather and sufficient coal Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 inventory at IPPs, the thermal coal price fell by 3-10% YoY to CNY495- Huaneng P (LHS) Relative to HSI (RHS) 511/tonne in January-May 2019 (average 7% YoY decline). The average

CCI5500 coal spot price also declined, by 9% YoY, to CNY609/tonne in 12-month range 4.07-6.10 1H19. We believe the YoY decline in unit fuel cost, despite a slight drop in Market cap (USDbn) 9.20 utilisation hours (coal-fired utilisation hours down 3% YoY in 5M19: NEA) 3m avg daily turnover (USDm) 19.21 and a slight decline in tariff due to a higher market-based sales proportion, Shares outstanding (m) 15,283 Major shareholder Huaneng Group (47.2%) will lead to strong YoY growth in 1H19 earnings.

Financial summary (CNY) Large capex with aggressive renewable capacity additions in 2019-21. Year to 31 Dec 19E 20E 21E Based on the latest NDRC policy, wind projects approved in or before 2018 Revenue (m) 174,691 176,214 177,782 are still entitled to fixed tariffs if they are connected on-grid before 2020. Operating profit (m) 19,625 22,628 25,288 We see that HNP is stepping up the progress of wind capacity installations Net profit (m) 4,693 5,795 6,768 Core EPS (fully-diluted) 0.299 0.369 0.431 starting from 2019, with heavy capex in the next few years. For 2019E, EPS change (%) 140.5 23.5 16.8 HNP looks for capex of CNY35bn (+66% YoY), with wind-power capex Daiwa vs Cons. EPS (%) (13.8) (14.2) (13.6) rising by 3.4x YoY to CNY24bn in 2019E to add 2.35GW in wind capacity. PER (x) 13.9 11.2 9.6 Besides, NHP also guided for 2.4GW/2.0GW in renewable capacity Dividend yield (%) 5.1 6.2 7.3 DPS 0.209 0.258 0.302 (including wind and solar) with capex of CNY22.8bn/CNY19bn in 2020/21E. PBR (x) 0.7 0.7 0.7 We expect this high capex to lead to negative FCF in 2019-21E, with net EV/EBITDA (x) 7.7 7.4 7.1 gearing rising further from 204% in 2018 to 208-221% in 2019-21E. ROE (%) 5.2 6.5 7.4 Source: FactSet, Daiwa forecasts What we recommend: We raise our 2019-21E EPS by 10-25% on our lower coal-price assumptions, partly offset by higher finance expenses with higher debt to support large capex ahead; but we reiterate our 12-month TP of HKD5.95, based on a target PBR of 0.9x applied to our average BVPS for 2019-20E (previously 0.9x 2019E PBR) — 0.6SD below its average 1-year forward PBR of 1.2x since 2007. We reiterate our Buy (1) call. Downside risk: higher-than-expected coal prices.

How we differ: Our 2019-21E EPS are 14% below consensus, likely on our more conservative tariff, fuel cost, and utilisation-hour assumptions.

See important disclosures, including any required research certifications, beginning on page 105

Huaneng Power International (902 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook HNP: dark spread and ROE We forecast HNP’s recurring earnings to recover from (CNY/MWh) CNY1.9bn for 2018 to CNY3.8-5.9bn for 2019-21E on a 250 20% 18% gradual decline in coal prices and rising wind power 200 16% capacity. Therefore, we look for HNP’s ROE to recover 14% from 2% in 2018 to 5-7% for 2019-21E. 150 12% 10% 100 8% 6% 50 4% 2% 0 0% 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019E2020E2021E Dark spread (LHS) ROE (RHS)

Source: Company, Daiwa forecast

Valuation HNP-H: 1-year forward PBR HNP H-shares are trading currently at a 0.7x 2019E PBR, (x) or around 0.9SD below the mean since 2007. We see 3.5 20% upside potential for the company’s valuation, with thermal 3.0 15% 2.5 10% coal prices likely to decline in the future, benefiting HNP in 2.1x 2.0 5% terms of lowering its unit fuel cost and thus leading to a 1.6x potential recovery in its earnings and ROE. 1.5 1.2x 0% 1.0 0.7x -5% 0.5 0.3x -10%

0.0 -15%

Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 Jan-18 Jan-19 PB (LHS) ROE (RHS)

Source: Bloomberg, Daiwa forecasts

Earnings revisions HNP: Bloomberg-consensus EPS (CNY) The Bloomberg consensus 2019-21E EPS for HNP saw a 0.6 sharp correction in March-April 2019 as a result of the company’s weak 2018 results due to high coal prices, 0.5 higher-than-expected operating expenses, and large impairments. The consensus forecasts have been adjusted 0.4 up slightly since May 2019 due to the strong 1Q19 results and YoY weaker coal prices YTD. 0.3

0.2

Our 2019-21E EPS are 14% below consensus, likely on

Jul-19

Apr-19 Jan-19

our more conservative tariff, fuel cost, and utilisation-hour Jun-19

Feb-19 Mar-19 May-19 assumptions. 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

Huaneng Power International (902 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Consolidated capacity - year end (MW) 78,622 82,331 83,878 104,321 105,991 109,231 112,525 115,427 Utilization hours - consolidated coal- 4,618 4,147 4,107 4,194 4,495 4,473 4,428 4,428 fired plants Consolidated tariff ex-VAT (CNY/MWh) 389 379 339 354 359 364 363 359 Unit fuel cost - consolidated coal-fired 201 174 171 226 237 220 208 196 plants (CNY/MWh) Dark spread - consolidated coal-fired 218 235 192 140 144 163 173 180 plants (CNY/MWh)

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of power 124,562 127,849 112,795 148,925 166,307 170,854 172,418 174,024 Port and transportation services 340 316 343 306 198 198 198 198 Other Revenue 505 739 677 3,228 3,046 3,638 3,598 3,559 Total Revenue 125,407 128,905 113,814 152,459 169,551 174,691 176,214 177,782 Other income 81 115 1,070 1,742 (279) 0 0 0 COGS (99,200) (98,604) (94,259) (141,900) (157,647) (153,223) (151,727) (150,618) SG&A 0 0 0 0 0 0 0 0 Other op.expenses (932) (1,158) (1,178) (1,376) (1,789) (1,843) (1,859) (1,876) Operating profit 25,355 29,258 19,448 10,925 9,836 19,625 22,628 25,288 Net-interest inc./(exp.) (7,824) (7,970) (7,068) (9,605) (10,647) (10,647) (10,858) (11,364) Assoc/forex/extraord./others 1,518 1,670 1,433 1,481 2,785 3,180 3,048 3,243 Pre-tax profit 19,050 22,958 13,813 2,802 1,973 12,158 14,818 17,168 Tax (5,487) (5,699) (3,465) (1,218) (643) (3,039) (3,704) (4,292) Min. int./pref. div./others (2,805) (3,607) (1,828) (73) (938) (4,425) (5,319) (6,108) Net profit (reported) 10,757 13,652 8,520 1,511 392 4,693 5,795 6,768 Net profit (adjusted) 10,757 13,652 9,379 572 1,900 4,693 5,795 6,768 EPS (reported)(CNY) 0.764 0.942 0.561 0.099 0.026 0.299 0.369 0.431 EPS (adjusted)(CNY) 0.764 0.942 0.617 0.038 0.124 0.299 0.369 0.431 EPS (adjusted fully-diluted)(CNY) 0.764 0.942 0.617 0.038 0.124 0.299 0.369 0.431 DPS (CNY) 0.389 0.493 0.290 0.100 0.100 0.209 0.258 0.302 EBIT 25,355 29,258 19,448 10,925 9,836 19,625 22,628 25,288 EBITDA 37,002 43,670 34,263 31,106 30,302 41,233 45,078 48,252

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 19,050 22,958 13,813 2,802 1,973 12,158 14,818 17,168 Depreciation and amortisation 11,780 14,719 15,164 20,636 20,917 21,608 22,451 22,964 Tax paid (5,487) (5,699) (3,465) (1,218) (643) (3,039) (3,704) (4,292) Change in working capital 1,281 (1,030) (867) 419 (10,421) (4,264) (5,728) (5,661) Other operational CF items 6,697 11,415 6,866 6,558 16,902 (2,513) (2,724) (2,912) Cash flow from operations 33,320 42,363 31,511 29,197 28,728 23,949 25,111 27,267 Capex (20,969) (22,740) (19,658) (22,457) (17,079) (35,412) (34,000) (30,000) Net (acquisitions)/disposals (11) (26,462) (320) (63,781) (3,469) 0 0 0 Other investing CF items 1,509 16,187 2,328 54,489 172 295 360 433 Cash flow from investing (19,471) (33,015) (17,650) (31,749) (20,376) (35,117) (33,640) (29,567) Change in debt 4,690 (1,179) 3,743 14,990 3,749 5,000 12,000 7,000 Net share issues/(repurchases) 0 4,684 0 5,000 8,245 0 0 0 Dividends paid (6,815) (8,490) (9,902) (6,537) (2,785) (1,570) (3,285) (4,056) Other financing CF items (8,769) (9,156) (7,444) (9,440) (11,452) 0 0 0 Cash flow from financing (10,894) (14,141) (13,602) 4,013 (2,243) 3,430 8,715 2,944 Forex effect/others (58) 33 73 10 26 0 0 0 Change in cash 2,897 (4,760) 332 1,472 6,135 (7,738) 186 644 Free cash flow 12,351 19,623 11,852 6,740 11,649 (11,463) (8,889) (2,733) Source: FactSet, Daiwa forecasts

Huaneng Power International (902 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 12,608 7,538 7,882 9,365 15,833 7,680 7,866 8,510 Inventory 6,702 5,423 6,879 7,385 9,544 8,121 7,977 7,877 Accounts receivable 14,882 16,377 16,393 25,448 29,279 35,372 40,508 45,739 Other current assets 3,673 4,227 5,812 6,340 7,144 7,147 7,203 7,261 Total current assets 37,865 33,565 36,967 48,538 61,799 58,319 63,554 69,386 Fixed assets 193,999 226,709 230,909 295,749 294,545 354,371 332,610 372,992 Goodwill & intangibles 22,323 25,316 26,088 32,311 32,548 32,674 32,802 32,915 Other non-current assets 20,984 23,276 20,877 19,991 31,011 (13,132) 22,433 (8,529) Total assets 275,172 308,866 314,840 396,590 419,903 432,233 451,399 466,764 Short-term debt 74,187 93,063 97,836 113,415 97,195 97,195 97,195 97,195 Accounts payable 27,036 26,186 28,325 38,900 35,139 35,517 34,886 34,448 Other current liabilities 3,623 4,588 4,035 3,635 5,873 3,896 3,896 3,896 Total current liabilities 104,846 123,837 130,196 155,950 138,206 136,608 135,977 135,539 Long-term debt 80,364 77,289 77,173 123,025 155,533 160,533 172,533 179,533 Other non-current liabilities 5,179 6,047 5,283 10,000 10,043 21,842 22,153 22,478 Total liabilities 190,389 207,173 212,653 288,975 303,782 318,983 330,663 337,550 Share capital 14,420 15,200 15,200 15,200 15,698 15,698 15,698 15,698 Reserves/R.E./others 55,709 68,942 70,803 72,441 78,737 71,783 74,292 77,004 Shareholders' equity 70,129 84,142 86,003 87,642 94,435 87,481 89,991 92,702 Minority interests 14,653 17,552 16,184 19,973 21,686 25,769 30,746 36,511 Total equity & liabilities 275,172 308,867 314,840 396,590 419,903 432,233 451,399 466,764 EV 202,285 223,937 226,995 290,847 302,343 317,487 332,041 341,797 Net debt/(cash) 141,943 162,815 167,127 227,075 236,895 250,048 261,861 268,218 BVPS (CNY) 4.863 5.536 5.658 5.766 6.016 5.573 5.733 5.905

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) (6.3) 2.8 (11.7) 34.0 11.2 3.0 0.9 0.9 EBITDA (YoY) 3.9 18.0 (21.5) (9.2) (2.6) 36.1 9.3 7.0 Operating profit (YoY) 4.2 15.4 (33.5) (43.8) (10.0) 99.5 15.3 11.8 Net profit (YoY) 3.2 26.9 (31.3) (93.9) 232.3 147.0 23.5 16.8 Core EPS (fully-diluted) (YoY) 3.0 23.4 (34.5) (93.9) 230.5 140.5 23.5 16.8 Gross-profit margin 20.9 23.5 17.2 6.9 7.0 12.3 13.9 15.3 EBITDA margin 29.5 33.9 30.1 20.4 17.9 23.6 25.6 27.1 Operating-profit margin 20.2 22.7 17.1 7.2 5.8 11.2 12.8 14.2 Net profit margin 8.6 10.6 8.2 0.4 1.1 2.7 3.3 3.8 ROAE 16.2 17.7 11.0 0.7 2.1 5.2 6.5 7.4 ROAA 4.0 4.7 3.0 0.2 0.5 1.1 1.3 1.5 ROCE 10.8 11.4 7.1 3.5 2.8 5.3 5.9 6.4 ROIC 8.1 9.0 5.5 2.0 1.9 4.1 4.6 4.9 Net debt to equity 167.4 160.1 163.6 211.0 204.0 220.8 216.9 207.6 Effective tax rate 28.8 24.8 25.1 43.5 32.6 25.0 25.0 25.0 Accounts receivable (days) 44.3 44.3 52.5 50.1 58.9 67.5 78.6 88.5 Current ratio (x) 0.4 0.3 0.3 0.3 0.4 0.4 0.5 0.5 Net interest cover (x) 3.2 3.7 2.8 1.1 0.9 1.8 2.1 2.2 Net dividend payout 50.9 52.3 51.7 100.6 389.8 70.0 70.0 70.0 Free cash flow yield 19.5 31.0 18.7 10.6 18.4 n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

Huaneng Power International is one of the big 5 Hong Kong listed independent power producers in China, with a consolidated power capacity of 106GW by the end of 2018.

China Utilities 11 July 2019

(816 HK) Huadian Fuxin Energy Huadian F uxin Energy

Target price: HKD2.40 (from HKD2.40) Share price (9 Jul): HKD1.43 | Up/downside: +67.8%

Distress play; recovery looks set to continue until 3Q19 Dennis Ip, CFA (852) 2848 4068  We look for continuous recovery in hydro power from 1Q19 [email protected]  Coal-fired power should benefit from YoY downtrend in coal prices Anna Lu, CFA (852) 2848 4465  Reiterating Buy (1) rating and TP of HKD2.4; distressed 0.4x PBR [email protected]

What's new: Fuxin’s share price is down 25% YTD, underperforming its Forecast revisions (%) wind-IPP peers’ by 9-28% and the HSI by 34%. We expect Fuxin to report Year to 31 Dec 19E 20E 21E c.30% YoY net profit growth in 1H19 on a recovery in hydro power Revenue change (1.2) 4.5 n.a. Net profit change (9.1) (3.9) n.a. generation, from a rise of 67.7% YoY in 1Q19 to a rise of c.130% YoY in Core EPS (FD) change (9.1) (3.9) n.a. 1H19. For 2Q-3Q19, we expect Fuxin’s hydro and coal-fired power assets Source: Daiwa forecasts to record a decent YoY recovery in profitability, led by both hydro utilisation gains and lowering fuel costs. As with other wind IPPs, Fuxin’s capex Share price performance should rise from CNY4.4bn in 2018 to CNY11-13bn pa in 2019-21E before (HKD) (%) the expiry of subsidy tariffs, in our view. Therefore, Fuxin’s FCF is likely to 2.0 120 remain stretched. However, Fuxin’s 2019E PBR of 0.4x (peers: 0.5-0.7x) is 1.8 109 the lowest in the sector, and hence we reiterate our Buy (1) rating, noting 1.6 98 1.5 86 also its attractive forecast ROE of 13-14% (peers: 8-14%). 1.3 75 Jul-18 Oct-18 Jan-19 Apr-19 Jul-19

What's the impact: balance-sheet impact on wind-farm rush Huadian Fu (LHS) Relative to HSI (RHS) installation better than peers due to Fuxin’s hydro exposure. We expect capex to increase by CNY6.4-8.4bn pa in 2019-21 vs. 2018, paving 12-month range 1.38-1.92 the way for further negative FCF of CNY5.8-7.6bn pa during the same Market cap (USDbn) 1.54 period. We expect a 24% increase in its finance costs (peers: 14-55%), 3m avg daily turnover (USDm) 0.72 Shares outstanding (m) 8,408 equivalent to a 23% (peers: 18-52%) increase in EBIT. Major shareholder Huadian Group (62.8%)

Continuous recovery in hydro output in 2Q19-3Q19E. From a low base, Financial summary (CNY) Fuxin’s hydro power generation was up by 131% YoY in 5M19. We expect Year to 31 Dec 19E 20E 21E Fujian to continue to post YoY improvement in hydropower output over Revenue (m) 19,418 20,806 21,768 June-August 2019 due to a low base (a 33-67% YoY drop the prior year). Operating profit (m) 6,307 7,435 8,057 Net profit (m) 2,788 3,420 3,612 Wind-power generation also improved from a 5.4% YoY drop in 1Q19 to a Core EPS (fully-diluted) 0.332 0.407 0.430 0.6% YoY drop in 5M19 amid better wind speeds. EPS change (%) 30.4 22.6 5.6 Daiwa vs Cons. EPS (%) 6.3 11.1 14.2 Coal-price correction likely to last until August. Led by the recovery in PER (x) 3.8 3.1 2.9 Dividend yield (%) 7.5 8.9 9.3 hydro output, Fuxin’s coal-fired power generation declined by 11.9% YoY DPS 0.094 0.112 0.118 for 5M19 but benefited from lower fuel costs on less coal demand; thus we PBR (x) 0.4 0.4 0.4 assume thermal coal prices in Fujian will continue to fall YoY for 8M19E, EV/EBITDA (x) 6.2 6.0 6.0 ROE (%) 12.5 13.6 12.8 following the 3-10% YoY drop (CNY749-789/tonne) in February-March. Source: FactSet, Daiwa forecasts

What we recommend: We trim our core 2019-20E EPS by 4-9% after incorporating the 1Q19 results, where wind power generation was weaker than expected, and also an increase in finance costs on accelerating wind- farm capex and rising accounts receivable. Our 12-month TP is unchanged at HKD2.4, based on an unchanged 0.7x 2019E PBR. We recommend investors accumulate the stock for a visible hydro output pick-up in 2Q- 3Q19E. Downside risk: weaker-than-expected water inflows.

How we differ: Unlike the market, we see Fuxin’s share price being driven by a 1Q-3Q19E profit recovery rather than its weak earnings in 2018.

See important disclosures, including any required research certifications, beginning on page 105

Huadian Fuxin Energy (816 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook Fuxin: reported net profit forecast (CNYm) We look for Fuxin’s reported net profit growth to return to 4,000 50% 3,500 the fast track in 2019E, with a 22% CAGR over 2018-21E 40% vs. a 2% CAGR over 2014-18. The main earnings drivers, 3,000 30% in our opinion, will be: 1) the bottoming-out of hydro output 2,500 in 2018 (down by 36% YoY), 2) a continued wind utilisation 2,000 20% recovery with the curtailment rate dropping to c.5% in 1,500 10% 2021E (from 13% in 2017), and 3) the declining trend in 1,000 0% coal prices for its coal-fired power units, despite the low 500 return for the newly-commissioned Sanmen nuclear units 0 (10%) (Gen III). 2013 2014 2015 2016 2017 2018 2019E 2020E Net profit (LHS) YoY (RHS)

Source: Company, Daiwa forecasts

Valuation Fuxin: 1-year forward PBR

On our forecasts, Fuxin shares are trading currently at a PBR (x) 3.8x PER and 0.x PBR for 2019E — troughs since 2012. 2.2 Last rush installation 1.7 Our TP of HKD2.4 is based on a 2019E PBR of 0.7x, which 1.4x Avg+1SD we consider to be fair given our view of a potential rerating 1.2 upon a hydropower recovery for Fuxin, as well as our 0.9x Avg 0.7 forecast earnings CAGR of 22% over 2018-21E. 0.5x Avg-1SD

0.2 0.1x Avg-2SD Fuxin’s PBR was derated from 1.9x to 0.5x in 2015 (last rush installation), as a result of a significant 14% drop in -0.3

utilisation hours vs. 2013, on surging curtailment. We

Jul-14

Apr-13 Oct-15 Apr-18

Jun-17 Jun-12 Jan-17

Mar-16

Feb-14 Feb-19

Sep-18 Nov-12 Sep-13 Dec-14 Aug-16 Nov-17 May-15 believe finance costs will surge on this round of Source: Bloomberg, Daiwa forecasts installations on worsening receivables.

Earnings revisions Fuxin: Bloomberg-consensus forecasts (CNY) The 2019-20E Bloomberg-consensus EPS forecasts for 0.40 Fuxin fell by 5% after the company announced disappointing 2018 results at end-March 2019, which we attribute to weak hydropower output. 0.35 Our 2019-21E EPS are 6-14% higher than the consensus, likely as we are more positive on the growth of various power segments.

0.30

Jul-18

Oct-18 Oct-18 Apr-19

Jun-18 Jan-19 Jan-19 Jun-19

Mar-19 Feb-19 Mar-19

Aug-18 Aug-18 Sep-18 Nov-18 Dec-18

May-19 May-19 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

Huadian Fuxin Energy (816 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Consolidated installed coal-fired power 3,850 3,600 3,600 3,600 3,600 3,600 3,600 3,600 capacity at year end (MW) Coal-fired power average utilization 5,018 4,011 3,255 4,171 4,865 4,330 4,070 3,948 hours Consolidated installed hydropower 2,457 2,508 2,508 2,508 2,608 2,608 2,608 2,608 capacity at year end (MW) Hydropower average utilization hours 3,649 3,860 5,622 3,700 2,311 3,004 3,455 3,628 Consolidated installed wind power 4,889 6,417 7,341 7,698 7,993 8,493 9,293 9,993 capacity at year end (MW) Wind power average utilization hours 1,888 1,745 1,765 1,940 2,128 2,130 2,132 2,134 Total electricity sales (GWh) 35,239 35,634 40,191 41,822 44,476 46,146 49,758 53,568 Nuclear installed capacity 1,000 2,000 3,000 4,000 4,000 4,000 4,000 4,000

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales of electricity 13,544 14,091 15,133 16,069 17,345 18,437 19,825 20,791 Service concession construction 27 40 3 133 80 77 76 73 revenue Other Revenue 325 1,262 861 611 904 904 904 904 Total Revenue 13,895 15,393 15,998 16,813 18,330 19,418 20,806 21,768 Other income 237 155 180 210 343 331 331 331 COGS (5,750) (6,460) (5,315) (6,464) (8,004) (7,368) (7,195) (7,181) SG&A (405) (456) (513) (512) (422) (449) (482) (506) Other op.expenses (3,036) (3,846) (4,942) (5,186) (5,438) (5,625) (6,025) (6,356) Operating profit 4,942 4,786 5,407 4,861 4,809 6,307 7,435 8,057 Net-interest inc./(exp.) (2,344) (2,589) (2,835) (2,800) (2,991) (3,142) (3,385) (3,719) Assoc/forex/extraord./others 84 428 630 742 940 1,075 1,096 1,097 Pre-tax profit 2,681 2,626 3,202 2,802 2,758 4,240 5,147 5,434 Tax (534) (406) (534) (337) (285) (551) (705) (746) Min. int./pref. div./others (280) (397) (715) (463) (334) (901) (1,022) (1,077) Net profit (reported) 1,867 1,823 1,953 2,002 2,138 2,788 3,420 3,612 Net profit (adjusted) 1,867 1,823 1,953 2,002 2,138 2,788 3,420 3,612 EPS (reported)(CNY) 0.234 0.217 0.232 0.238 0.254 0.332 0.407 0.430 EPS (adjusted)(CNY) 0.234 0.217 0.232 0.238 0.254 0.332 0.407 0.430 EPS (adjusted fully-diluted)(CNY) 0.234 0.217 0.232 0.238 0.254 0.332 0.407 0.430 DPS (CNY) 0.046 0.040 0.051 0.056 0.057 0.094 0.112 0.118 EBIT 4,942 4,786 5,407 4,861 4,809 6,307 7,435 8,057 EBITDA 7,609 8,042 9,450 9,215 9,379 11,010 12,464 13,365

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 2,681 2,626 3,202 2,802 2,758 4,240 5,147 5,434 Depreciation and amortisation 2,667 3,256 4,043 4,354 4,570 4,703 5,029 5,309 Tax paid (534) (580) (462) (415) (300) (551) (705) (746) Change in working capital 300 1,088 322 (768) 1,707 (772) (1,926) (2,406) Other operational CF items 1,887 2,145 2,718 2,066 1,967 1,983 2,204 2,539 Cash flow from operations 7,001 8,535 9,823 8,038 10,703 9,604 9,749 10,130 Capex (15,734) (14,756) (10,735) (9,133) (6,176) (10,906) (12,831) (12,056) Net (acquisitions)/disposals (378) (402) (65) (173) 0 0 0 0 Other investing CF items 2,107 (1,288) 565 (44) 127 319 463 545 Cash flow from investing (14,005) (16,446) (10,235) (9,350) (6,049) (10,587) (12,368) (11,512) Change in debt 9,322 8,641 5,888 2,448 (4,125) 6,000 6,000 6,000 Net share issues/(repurchases) 2,244 0 0 0 0 0 0 0 Dividends paid (498) (535) (925) (803) (822) (792) (943) (989) Other financing CF items (2,732) (1,455) (3,232) (1,126) 1,770 (4,452) (4,860) (5,277) Cash flow from financing 8,337 6,651 1,730 520 (3,177) 757 197 (266) Forex effect/others 0 0 0 0 0 0 0 0 Change in cash 1,334 (1,261) 1,318 (792) 1,476 (226) (2,422) (1,647) Free cash flow (8,732) (6,221) (912) (1,095) 4,527 (1,302) (3,082) (1,926) Source: FactSet, Daiwa forecasts

Huadian Fuxin Energy (816 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 5,145 3,798 4,415 4,040 5,751 6,791 5,836 5,418 Inventory 427 398 249 277 414 298 274 377 Accounts receivable 3,492 3,606 5,140 6,992 5,970 6,728 8,419 10,538 Other current assets 686 448 102 15 22 22 22 22 Total current assets 9,749 8,250 9,907 11,325 12,157 13,839 14,552 16,355 Fixed assets 66,074 76,255 78,932 81,191 80,930 87,220 95,110 101,946 Goodwill & intangibles 1,100 1,120 1,130 1,299 1,359 1,307 1,255 1,203 Other non-current assets 9,012 12,666 13,311 13,425 13,858 14,897 15,957 17,018 Total assets 85,935 98,291 103,280 107,240 108,304 117,263 126,874 136,522 Short-term debt 12,061 12,956 14,294 15,408 10,894 10,894 10,894 10,894 Accounts payable 2,915 2,528 1,569 1,081 1,104 1,165 1,278 1,226 Other current liabilities 11,602 12,376 10,355 8,877 7,347 7,347 7,347 7,347 Total current liabilities 26,579 27,860 26,218 25,366 19,345 19,406 19,519 19,467 Long-term debt 40,508 48,026 52,723 54,005 54,387 60,387 66,387 72,387 Other non-current liabilities 1,190 1,254 1,343 1,401 1,470 1,470 1,470 1,470 Total liabilities 68,277 77,140 80,284 80,773 75,202 81,263 87,376 93,324 Share capital 8,408 10,402 10,402 12,396 17,378 17,378 17,378 17,378 Reserves/R.E./others 6,605 8,076 9,699 11,028 12,678 15,190 18,182 21,321 Shareholders' equity 15,013 18,477 20,101 23,424 30,055 32,568 35,560 38,699 Minority interests 2,646 2,673 2,895 3,043 3,047 3,432 3,938 4,499 Total equity & liabilities 85,935 98,291 103,280 107,240 108,304 117,263 126,874 136,522 EV 56,234 63,713 68,589 70,823 64,189 68,460 74,825 80,707 Net debt/(cash) 47,424 57,184 62,602 65,373 59,529 64,490 71,445 77,863 BVPS (CNY) 1.881 1.960 2.154 2.312 2.508 2.807 3.163 3.536

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 4.9 10.8 3.9 5.1 9.0 5.9 7.1 4.6 EBITDA (YoY) 15.8 5.7 17.5 (2.5) 1.8 17.4 13.2 7.2 Operating profit (YoY) 12.4 (3.1) 13.0 (10.1) (1.1) 31.2 17.9 8.4 Net profit (YoY) 27.2 (2.4) 7.2 2.5 6.8 30.4 22.6 5.6 Core EPS (fully-diluted) (YoY) 21.5 (7.4) 7.2 2.5 6.8 30.4 22.6 5.6 Gross-profit margin 58.6 58.0 66.8 61.6 56.3 62.1 65.4 67.0 EBITDA margin 54.8 52.2 59.1 54.8 51.2 56.7 59.9 61.4 Operating-profit margin 35.6 31.1 33.8 28.9 26.2 32.5 35.7 37.0 Net profit margin 13.4 11.8 12.2 11.9 11.7 14.4 16.4 16.6 ROAE 14.2 11.6 11.3 10.7 10.6 12.5 13.6 12.8 ROAA 2.4 2.0 1.9 1.9 2.0 2.5 2.8 2.7 ROCE 7.8 6.3 6.3 5.2 5.0 6.1 6.6 6.6 ROIC 6.7 5.6 5.5 4.8 4.7 5.7 6.1 6.0 Net debt to equity 315.9 309.5 311.4 279.1 198.1 198.0 200.9 201.2 Effective tax rate 19.9 15.5 16.7 12.0 10.3 13.0 13.7 13.7 Accounts receivable (days) 85.9 84.1 99.8 131.7 129.1 119.3 132.9 158.9 Current ratio (x) 0.4 0.3 0.4 0.4 0.6 0.7 0.7 0.8 Net interest cover (x) 2.1 1.8 1.9 1.7 1.6 2.0 2.2 2.2 Net dividend payout 19.6 18.6 22.0 23.4 22.3 28.4 27.6 27.4 Free cash flow yield n.a. n.a. n.a. n.a. 42.7 n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

Huadian Fuxin Energy Corporation Limited (Fuxin) is a diversified clean-energy company in China. It engages in the development, management and operation of hydropower and coal-fired power plants in Fujian Province, and wind power and other clean-energy projects throughout China. The company also owns a 39% stake in the Fuqing nuclear project and 10% stake in the Sanmen nuclear project in Fujian.

China Utilities 11 July 2019

China Suntien Green Energy (956 HK)

China Sunti en Gr een Energ y

Target price: HKD3.50 (from HKD3.50) Share price (9 Jul): HKD2.14 | Up/downside: +63.6%

Distress play; A-share IPO as positive catalyst Dennis Ip, CFA (852) 2848 4068  Sector’s cheapest stub value, 0.4x PBR for wind, 7% dividend yield [email protected]  Attractive valuation and yield; set for further rerating on A-share IPO Anna Lu, CFA (852) 2848 4465  Reiterating our Buy (1) rating and TP of HKD3.5 [email protected]

What's new: Suntien’s share price rose 6% YTD, outperforming wind-IPP Forecast revisions (%) peers’ 1-23% YTD drop but underperforming city-gas peers’ 7-32% rise Year to 31 Dec 19E 20E 21E (HSI: +9% YTD). We expect Suntien to see c.10% YoY net profit on Revenue change (4.9) (9.6) (13.4) Net profit change - (0.6) (0.8) recovering wind power generation from +6% YoY in 1Q19 to +14.4% YoY in Core EPS (FD) change - (0.6) (0.8) 1H19. Similar to other wind-IPPs, its capex should rise from CNY4.5bn in Source: Daiwa forecasts 2018 to CNY5.0-6.2bn pa in 2019-21E before the expiry of its subsidy tariff, on our estimate. Thus, Suntien’s FCF remains stretched. But, Suntien’s ex- Share price performance gas 0.4x PBR is the lowest in the sector, and hence we reiterate our Buy (HKD) (%) (1) rating, also on attractive 7% yield (peers: 1-4%). 2.6 110 2.4 104 What's the impact: balance sheet impact on rush installation of wind 2.2 98 2.0 91 farms better than peers on city-gas exposure. We expect an increase of 1.8 85 CNY0.5-1.7bn pa of capex in 2019-21E vs. 2018, introducing a further Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 negative CNY0.4-1.4bn of FCF pa during the same period, and thus Ch Suntien (LHS) Relative to HSI (RHS) leading to a 46% (peers: 14-55%) increase in finance cost, representing a 28% (peers: 18-52%) EBIT increase. 12-month range 1.85-2.53 Market cap (USDbn) 1.02 Attractive valuation and yield. Suntien’s 0.6x 2020E PBR is at the higher 3m avg daily turnover (USDm) 1.22 Shares outstanding (m) 3,715 end of the China wind IPP’s 0.4-0.6x range. However, we think that Major shareholder Hebei Const & Inv (50.5%) Suntien’s wind farms could actually trade at 0.4x if we assume its gas business (representing c.20% of company’s book value) trades at 0.9x Financial summary (CNY) (pure city-gas peers: 0.9-3.0x). Executive Director Mei Chun Xiao recently Year to 31 Dec 19E 20E 21E accumulated 50k shares. Our 2020-21E EPS are broadly unchanged. Revenue (m) 12,256 14,687 17,379 Operating profit (m) 2,830 3,140 3,457 Net profit (m) 1,508 1,607 1,727 Set for further rerating, along with upward revisions to consensus Core EPS (fully-diluted) 0.406 0.432 0.465 EPS. Suntien submitted its A-share IPO application to the CSRC in early EPS change (%) 0.9 6.5 7.5 November 2018. The new 134.74m A-shares, representing 3.5% of Daiwa vs Cons. EPS (%) 4.9 (2.4) (4.9) Suntien’s 3,850m H-share base, are likely to be completed by end-2019, in PER (x) 4.6 4.4 4.1 Dividend yield (%) 6.9 7.1 7.4 our view. The listing price can be: 1) no higher than 23x PER (equivalent to DPS 0.130 0.134 0.139 c.3x PBR), and 2) no lower than Suntien’s BVPS (1x PBR). In our view, the PBR (x) 0.6 0.6 0.5 A-share IPO will not only ease concerns on gearing (>210% in 2019E vs. EV/EBITDA (x) 7.8 7.9 7.9 ROE (%) 14.0 13.3 13.0 c.185% upon completion) but enhance its H-share valuation with Source: FactSet, Daiwa forecasts southbound buying interest in the stock (H-share: 0.5x PBR vs. A-share: 1.0-3.0x PBR), after H-shares are included in the HK-SZ Stock Connect upon completion of the A-share IPO.

What we recommend: We reiterate our Buy (1) rating and our DCF-based 12-month TP of HKD3.5. We view Suntien’s potential A-share IPO as a further rerating catalyst. Key risks: weaker-than-expected wind speeds and delay of A-share IPO.

How we differ: Unlike some in the market, we expect Suntien to rerate on its A-share listing, potential reversal of impairments and attractive yield.

See important disclosures, including any required research certifications, beginning on page 105

China Suntien Green Energy (956 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook Suntien: net profit forecasts (CNYm) We forecast Suntien’s 2018-21 recurring net profit to rise at 2,000 140% a CAGR of 5% to CNY1,727m (or +11% YoY in reported 1,800 120% 1,600 profit), on the back of a 14% net wind power generation 100% 1,400 80% CAGR and 25% gas sales volume CAGR. 1,200 60% 1,000 40% 800 600 20% 400 0% 200 (20%)

0 (40%)

2013 2018E 2009 2010 2011 2012 2014 2015 2016 2017 2019E 2020E 2021E

Net profit YoY growth (RHS)

Source: Company, Daiwa forecasts

Valuation Suntien: 1-year-forward PBR

Riding on a continued recovery in the wind and gas PBR (x) segments, Suntien’s share price has rerated from 0.3x 1- 1.6 year forward PBR in mid-2016 to 0.55x currently, closer to 1.4 1.3x Avg+2SD its historical mean of 0.7x since 2010. Our TP implies a 1.2 0.9x 2020E PBR, as we see further upside in the 1.0 1.0x Avg+1SD company’s current valuation of 0.5x 2020E PBR from: 1) 0.8 attractive 2020E dividend yield of 7%, 2) potential 0.7x Avg 0.6 catalysts, including an A-share IPO (priced at least at 1x 0.5x Avg-1SD 0.4 PBR), and 3) inclusion in the Hong Kong-Shenzhen Stock 0.2x Avg-2SD Connect, while its gas segment’s earnings (at a higher 0.2 return) are likely to grow faster than wind’s over 2019-21E, 0.0 Oct-10 Oct-12 Oct-14 Oct-16 Oct-18 despite a slow 5% EPS CAGR for 2018-21E. Hence, we Source: Bloomberg, Daiwa forecasts believe Suntien could see a valuation rerating on its A- share IPO, should it be successful.

Earnings revisions Suntien: 2019-21E Bloomberg EPS consensus The Bloomberg-consensus 2019-20E EPS for Suntien rose 0.50 by 1-2% YTD, as the company’s utilisation recovery has mostly finished in 2018, with a 9.6% CAGR from 1,887 0.45 hours in 2015 to 2,482 hours in 2018. Its gas sales volume growth is also likely to slow from a 52% CAGR for 2016-18 0.40 to a 25% CAGR for 2018-21E.

Our 2020-21E EPS are 2-5% below the consensus, likely 0.35

due to our more prudent FCF estimates with higher finance

2/5/2019 4/9/2019

5/21/2019 6/11/2019 7/10/2018 7/31/2018 8/21/2018 9/11/2018 10/2/2018 12/4/2018 1/15/2019 2/26/2019 3/19/2019 4/30/2019

cost compared with the street. 6/19/2018

10/23/2018 11/13/2018 12/25/2018 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

China Suntien Green Energy (956 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Natural gas sales volume (mn m3) 1,523 1,127 1,111 1,879 2,632 3,329 4,155 5,115 Wholesales gas proportion (%) 55 70 71 58 62 63 65 66 Gas ASP, excl tax (CNY/m3) 2.44 2.32 2.03 2.02 2.41 2.40 2.39 2.38 Gas cost, excl tax (CNY/m3) 2.00 1.89 1.67 1.80 2.15 2.15 2.15 2.15 Consolidated wind capacity (MW) 1,697 2,094 2,796 3,348 3,858 4,358 4,858 5,358 Wind utilisation (hours) 1,996 1,887 2,195 2,392 2,482 2,393 2,391 2,390

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Natural gas 3,903 2,792 2,392 3,946 6,548 8,194 10,139 12,345 Wind power 1,246 1,427 1,983 3,096 3,417 4,062 4,548 5,033 Other Revenue 0 6 9 15 10 0 0 0 Total Revenue 5,149 4,224 4,384 7,058 9,975 12,256 14,687 17,379 Other income 56 77 97 80 100 128 125 130 COGS (3,256) (2,409) (2,166) (3,625) (5,934) (7,502) (9,157) (11,125) SG&A (256) (273) (302) (453) (502) (705) (892) (1,022) Other op.expenses (601) (910) (784) (1,296) (1,401) (1,347) (1,623) (1,904) Operating profit 1,093 710 1,228 1,764 2,238 2,830 3,140 3,457 Net-interest inc./(exp.) (487) (572) (549) (774) (785) (886) (1,030) (1,144) Assoc/forex/extraord./others 69 63 65 214 290 366 399 435 Pre-tax profit 675 200 744 1,203 1,743 2,310 2,510 2,748 Tax (176) (11) (96) (99) (168) (369) (436) (516) Min. int./pref. div./others (163) (21) (106) (165) (306) (433) (467) (505) Net profit (reported) 335 168 542 940 1,268 1,508 1,607 1,727 Net profit (adjusted) 363 356 542 1,077 1,495 1,508 1,607 1,727 EPS (reported)(CNY) 0.091 0.045 0.146 0.253 0.341 0.406 0.432 0.465 EPS (adjusted)(CNY) 0.099 0.096 0.146 0.290 0.402 0.406 0.432 0.465 EPS (adjusted fully-diluted)(CNY) 0.099 0.096 0.146 0.290 0.402 0.406 0.432 0.465 DPS (CNY) 0.031 0.015 0.063 0.103 0.125 0.130 0.134 0.139 EBIT 1,121 898 1,228 1,764 2,238 2,830 3,140 3,457 EBITDA 1,718 1,591 2,005 2,889 3,420 4,176 4,764 5,361

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 675 200 744 1,203 1,743 2,310 2,510 2,748 Depreciation and amortisation 598 694 776 1,125 1,182 1,347 1,623 1,904 Tax paid (176) (117) (117) (126) (168) (369) (436) (516) Change in working capital (197) 386 73 (44) (617) (788) (835) (948) Other operational CF items (13) 176 118 484 495 519 631 708 Cash flow from operations 885 1,339 1,593 2,642 2,635 3,020 3,493 3,897 Capex (2,446) (5,687) (4,221) (3,921) (5,123) (6,860) (7,030) (6,740) Net (acquisitions)/disposals (3) 0 0 0 0 0 0 0 Other investing CF items 116 458 612 351 34 34 34 34 Cash flow from investing (2,333) (5,228) (3,609) (3,570) (5,089) (6,826) (6,996) (6,706) Change in debt 2,067 4,797 1,217 2,640 6,000 5,000 4,500 4,000 Net share issues/(repurchases) 1,256 0 0 0 0 0 0 0 Dividends paid (248) (167) (81) (317) (844) (877) (906) (942) Other financing CF items (465) (579) (672) (760) (2,572) (931) (241) (298) Cash flow from financing 2,610 4,051 464 1,563 2,584 3,192 3,353 2,760 Forex effect/others 0 0 0 0 0 0 0 0 Change in cash 1,162 162 (1,551) 634 130 (615) (150) (50) Free cash flow (1,561) (4,348) (2,628) (1,280) (2,487) (3,840) (3,537) (2,843) Source: FactSet, Daiwa forecasts

China Suntien Green Energy (956 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 3,198 3,139 1,491 2,128 2,240 1,625 1,475 1,425 Inventory 43 48 45 40 46 70 113 170 Accounts receivable 1,402 1,241 1,597 2,564 3,296 3,624 4,429 5,315 Other current assets 689 804 736 801 836 1,229 1,598 2,032 Total current assets 5,331 5,232 3,869 5,533 6,418 6,549 7,615 8,942 Fixed assets 11,731 14,971 19,668 22,466 26,584 32,321 37,827 42,756 Goodwill & intangibles 2,198 2,101 2,021 1,918 1,796 1,713 1,624 1,539 Other non-current assets 2,350 4,620 3,816 4,371 4,363 4,110 4,101 4,091 Total assets 21,610 26,924 29,374 34,288 39,161 44,693 51,166 57,329 Short-term debt 1,730 2,440 5,113 5,708 4,644 4,644 4,644 4,644 Accounts payable 437 553 465 576 148 494 715 997 Other current liabilities 1,364 1,561 2,240 3,190 3,810 3,350 3,512 3,658 Total current liabilities 3,531 4,555 7,818 9,473 8,602 8,488 8,870 9,299 Long-term debt 9,296 13,386 11,933 13,217 16,683 21,683 26,183 30,183 Other non-current liabilities 21 82 90 1,097 1,479 184 184 184 Total liabilities 12,848 18,023 19,840 23,787 26,764 30,355 35,237 39,666 Share capital 3,715 3,715 3,715 3,715 3,715 3,715 3,715 3,715 Reserves/R.E./others 3,644 3,698 4,185 4,890 6,321 7,829 8,953 10,182 Shareholders' equity 7,360 7,413 7,900 8,605 10,036 11,545 12,669 13,897 Minority interests 1,403 1,487 1,634 1,896 2,360 2,793 3,260 3,765 Total equity & liabilities 21,610 26,924 29,374 34,288 39,161 44,693 51,166 57,329 EV 15,315 20,109 23,042 24,075 26,624 32,672 37,789 42,344 Net debt/(cash) 7,828 12,687 15,554 16,797 19,087 24,702 29,352 33,401 BVPS (CNY) 2.001 1.995 2.127 2.316 2.701 3.107 3.410 3.741

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 10.5 (18.0) 3.8 61.0 41.3 22.9 19.8 18.3 EBITDA (YoY) (2.8) (7.4) 26.0 44.1 18.4 22.1 14.1 12.5 Operating profit (YoY) (7.9) (19.9) 36.8 43.6 26.9 26.4 11.0 10.1 Net profit (YoY) (21.0) (1.8) 52.2 98.6 38.8 0.9 6.5 7.5 Core EPS (fully-diluted) (YoY) (30.5) (2.8) 52.2 98.6 38.8 0.9 6.5 7.5 Gross-profit margin 36.8 43.0 50.6 48.6 40.5 38.8 37.7 36.0 EBITDA margin 33.4 37.7 45.7 40.9 34.3 34.1 32.4 30.8 Operating-profit margin 21.8 21.3 28.0 25.0 22.4 23.1 21.4 19.9 Net profit margin 7.1 8.4 12.4 15.3 15.0 12.3 10.9 9.9 ROAE 5.4 4.8 7.1 13.0 16.0 14.0 13.3 13.0 ROAA 1.9 1.5 1.9 3.4 4.1 3.6 3.4 3.2 ROCE 6.3 4.0 4.8 6.3 7.1 7.6 7.2 7.0 ROIC 5.2 3.5 4.6 6.2 6.9 6.7 6.2 5.8 Net debt to equity 106.4 171.1 196.9 195.2 190.2 214.0 231.7 240.3 Effective tax rate 26.1 5.7 12.9 8.2 9.6 16.0 17.4 18.8 Accounts receivable (days) 79.6 114.2 118.1 107.6 107.2 103.0 100.1 102.3 Current ratio (x) 1.5 1.1 0.5 0.6 0.7 0.8 0.9 1.0 Net interest cover (x) 2.3 1.6 2.2 2.3 2.9 3.2 3.0 3.0 Net dividend payout 34.4 33.1 43.2 40.7 36.6 32.0 31.0 30.0 Free cash flow yield n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Source: FactSet, Daiwa forecasts

Company profile

China Suntien Green Energy is a clean-energy enterprise controlled by the Hebei Construction & Investment Group. Its main businesses include the investment, construction and operation of wind farms, as well as the transmission and distribution of natural gas and compressed natural gas (CNG).

Hong Kong Information Technology 11 July 2019

(3800 HK) GCL-Poly Energy GCL-Pol y Energ y

Target price: HKD0.45 (from HKD0.52) Share price (9 Jul): HKD0.43 | Up/downside: +4.7%

Downgrading: short-term refinancing risk is still high Dennis Ip, CFA (852) 2848 4068  Recent actions signal GCL-Poly is in need of cash for debt repayment [email protected]  Short-term refinancing risk seems high, with large debt due in 2019 Anna Lu, CFA (852) 2848 4465  Downgrading to Hold (3); lowering our 12-month TP to HKD0.45 [email protected]

What's new: After GCL-Poly announced its plan to dispose of a 51% stake Forecast revisions (%) in GCL New Energy (GNE; 451 HK, not rated) to Huaneng Group on 4 Year to 31 Dec 19E 20E 21E June, it placed shares for disposal in upstream polysilicon company Revenue change (12.1) (14.0) (15.5) Net profit change (14.3) (8.6) (14.4) Xinjiang GCL to raise a net CNY1.7bn in cash, to partly finance its debt of Core EPS (FD) change (17.7) (15.5) (20.9) CNY7.6bn due in one year. Given GNE’s still-distressed 0.6x 2020E PBR Source: Daiwa forecasts valuation, we expect the disposal to recycle only c.CNY2.7-3.1bn, at 0.7- 0.8x PBR, and not CNY3.5bn at 0.9x PBR. Hence, we expect GCL-Poly’s Share price performance refinancing pressure to continue. (HKD) (%) 0.80 115 What's the impact: recent actions to raise cash for debt repayment. 0.70 101 GCL-Poly completed a share placement (7.6% of its expanded base) for 0.60 88 0.50 74

HKD680m (net proceeds: HKD669m) on 18 June. On 26 June, it 0.40 60 announced plans to sell a 31.5% stake in Xinjiang GCL (currently 70%- Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 owned by GCL-Poly) for CNY2.5bn (net proceeds: CNY2.4bn) to Xuzhou GCL-Poly E (LHS) Relative to HSI (RHS)

Zhongping GCL Industrial Upgrading Equity Investment Fund, in which GCL-Poly is a limited partner and will invest CNY1.35bn for a 40.26% stake 12-month range 0.41-0.78 (but with no control over the fund, according to GCL-Poly). Subsequently, Market cap (USDbn) 1.09 Xinjiang GCL will no longer be consolidated in GCL-Poly (but ultimately 3m avg daily turnover (USDm) 5.40 Shares outstanding (m) 19,841 51%-held by GCL-Poly and providing associate earnings). GCL-Poly Major shareholder Zhu Gongshan (32.4%) currently expects to record a gain of CNY1.6bn and a net cash inflow (after considering investment in Xuzhou Zhongping Fund) of CNY1.1bn after the Financial summary (CNY) disposal. Cash from both actions will likely go mainly towards debt Year to 31 Dec 19E 20E 21E repayment. While this helps address GCL-Poly’s high gearing (237% as at Revenue (m) 19,307 18,034 17,569 end-2018), we think the moves signal that GCL-Poly is short of cash. Operating profit (m) 5,151 5,709 5,424 Net profit (m) 281 776 852 Core EPS (fully-diluted) 0.015 0.039 0.043 Short-term refinancing risk appears high. GCL-Poly had CNY22.6bn of EPS change (%) n.a. 165.8 9.7 debt due in one year as at end-2018 (adjusted after some bank waivers), Daiwa vs Cons. EPS (%) (8.0) (11.1) (14.1) including CNY7.6bn in GNE. This implies that GCL-Poly itself has PER (x) 25.8 9.7 8.8 Dividend yield (%) 0.0 0.0 0.0 CNY15.1bn of debt due in 2019 even after excluding GNE, while it has DPS 0.000 0.000 0.000 CNY7.5bn cash in hand (excluding GNE). As such, GCL-Poly (excluding PBR (x) 0.3 0.3 0.3 GNE) has net debt of CNY7.6bn to be repaid or refinanced in 2019, while EV/EBITDA (x) 5.2 4.9 4.8 ROE (%) 1.2 3.2 3.4 the actions mentioned above can only raise cash of c.CNY1.7bn. Thus, the Source: FactSet, Daiwa forecasts company’s short-term refinancing risk seems high, in our view.

What we recommend: We cut our 2019-21E EPS by 16-21% to reflect the share placement and lower earnings from Xinjiang GCL after the stake sale, partly offset by lower finance costs. We cut our TP to HKD0.45 (from HKD0.52), based on a PBR of 0.3x (from 0.4x) on our average 2019-20E BVPS on dilution and lower ROEs, and downgrade to Hold (3) despite low multiples. Risks: higher/lower-than-expected solar product ASPs.

How we differ: Our 2019-21E EPS are 8-14% below the consensus, likely as we factor in the share placement and Xinjiang GCL stake disposal.

See important disclosures, including any required research certifications, beginning on page 105

GCL-Poly Energy (3800 HK): 11 July 2019

How do we justify our view? Growth outlook Valuation Earnings revisions

Growth outlook GCL-Poly: recurring net profit forecast (CNYm) We forecast GCL-Poly to report a net profit of CNY1,482m 2,500 in 2019, mainly driven by a CNY1.6bn gain from the 2,000 disposal of its 31.5% stake in Xinjiang GCL. Excluding this 1,500 gain, we expect GCL-Poly to post a recurring net profit of 1,000 CNY281m in 2019, compared with a net loss of CNY693m in 2018, primarily led by the earnings contribution from 500 Xinjiang GCL commissioned in October 2018 and lower 0 finance expenses. We forecast its 2020-21E earnings to (500) further recover to CNY776m and CNY852m, respectively, (1,000) on the further ramp-up of Xinjiang GCL. 2015 2016 2017 2018 2019E 2020E 2021E

Source: Company, Daiwa forecasts

Valuation GCL-Poly: 1-year forward PBR GCL-Poly is currently trading at a past-6-year low 1-year PBR (x) forward PBR of 0.3x, which is 1.3SD below its past-6-year 2.5 average. 2.0 2.0x Avg+2SD

1.5x Avg+1SD Despite the distressed valuation, we expect GCL-Poly’s 1.5 short-term refinancing risk to weigh against any potential 1.0 1.0x Avg for a rerating. We lower our 12-month TP to HKD0.45 (from 0.5 0.5x Avg-1SD HKD0.52), now based on a PBR of 0.3x, or c.1.3SD below its mean PBR since 2013, applied to our average 2019- 0.0 0.0x Avg-2SD 20E BVPS. Despite the very low valuation multiples, we -0.5 downgrade our rating on the stock to Hold (3), from Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 Jan-18 Jan-19

Outperform (2), on refinancing risks. Source: Bloomberg, Daiwa estimates

Earnings revisions GCL-Poly: Bloomberg-consensus EPS forecast revisions (CNY) The 2019-20 Bloomberg-consensus EPS forecasts for 0.20 0.18 GCL-Poly have declined by 64% and 18% YTD, 0.16 respectively, as GCL-Poly announced a net loss of 0.14 CNY693m in 2018 and given the decline in polysilicon and 0.12 multi wafer prices in 1H19. 0.10 0.08 0.06 Our 2019-21E EPS are 8-14% below the consensus, likely 0.04 as we have factored in the share placement and Xinjiang 0.02 0.00

GCL stake disposal announcement in June 2019.

Jul-18

Apr-18 Oct-18 Apr-19

Jan-18 Jan-19 Jun-19 Jun-18

Mar-18 Mar-19

Feb-18 Feb-19

Aug-18 Sep-18 Nov-18 Dec-18

May-18 May-19 2019E EPS 2020E EPS 2021E EPS

Source: Bloomberg

GCL-Poly Energy (3800 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Wafer sales (MW) 12,909 15,178 17,518 23,417 24,761 25,009 25,259 25,511 Wafer ASP (USD per W) 0.22 0.19 0.16 0.14 0.09 0.07 0.07 0.06 Wafer unit cost (USD per W) 0.17 0.13 0.10 0.09 0.08 0.06 0.05 0.05 Wafer unit dollar margin (USD per W) 0.05 0.06 0.06 0.05 0.01 0.01 0.01 0.01 Polysilicon sales (MT) 15,443 18,023 9,951 7,316 20,041 40,082 48,098 48,098 Polysilicon ASP (USD per kg) 21.7 15.1 15.0 15.5 11.7 9.9 9.5 9.2 Polysilicon unit cost (USD per kg) 15.2 13.5 11.5 10.6 9.6 7.3 6.2 6.2 Polysilicon unit dollar margin (USD per 6.48 1.61 3.51 4.88 2.10 2.59 3.29 3.00 kg) Solar farm installed capacity (MW) 1,039 2,539 3,509 5,874 7,328 6,123 6,123 6,123

Profit and loss (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Solar material business 25,455 19,243 19,270 19,355 14,436 13,293 12,558 12,097 Power business 10,057 0 0 0 0 0 0 0 Other Revenue 1,712 1,242 2,755 4,439 6,130 6,014 5,476 5,472 Total Revenue 37,225 20,484 22,025 23,794 20,565 19,307 18,034 17,569 Other income 1,316 909 926 843 891 805 813 800 COGS (29,574) (14,744) (14,980) (15,596) (15,533) (12,946) (11,245) (11,101) SG&A (2,585) (1,565) (1,920) (2,308) (2,133) (2,015) (1,893) (1,844) Other op.expenses 0 0 0 0 0 0 0 0 Operating profit 6,382 5,085 6,051 6,734 3,791 5,151 5,709 5,424 Net-interest inc./(exp.) (2,710) (1,881) (1,954) (2,312) (3,154) (2,969) (2,995) (3,012) Assoc/forex/extraord./others (877) (428) (1,253) (1,510) (1,148) 140 (1,375) (745) Pre-tax profit 2,795 2,775 2,844 2,912 (511) 2,322 1,339 1,667 Tax (639) (484) (537) (638) 52 (398) (229) (285) Min. int./pref. div./others (200) 134 (278) (300) (235) (444) (334) (529) Net profit (reported) 1,955 2,425 2,029 1,974 (693) 1,481 776 852 Net profit (adjusted) 1,955 2,425 2,029 1,974 (693) 281 776 852 EPS (reported)(CNY) 0.126 0.155 0.110 0.107 (0.038) 0.078 0.039 0.043 EPS (adjusted)(CNY) 0.126 0.155 0.110 0.107 (0.038) 0.015 0.039 0.043 EPS (adjusted fully-diluted)(CNY) 0.126 0.152 0.111 0.105 (0.038) 0.015 0.039 0.043 DPS (CNY) 0.000 0.072 0.000 0.000 0.000 0.000 0.000 0.000 EBIT 6,382 5,085 6,051 6,734 3,791 5,151 5,709 5,424 EBITDA 10,005 8,281 9,452 10,483 7,978 10,857 11,697 12,283

Cash flow (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 2,795 2,775 2,844 2,912 (511) 2,322 1,339 1,667 Depreciation and amortisation 3,780 3,298 3,548 3,807 4,284 5,706 5,988 6,859 Tax paid (533) (578) (598) (619) (493) (387) (218) (281) Change in working capital (147) (6,358) (960) 29 (273) (5,194) (4,255) (1,252) Other operational CF items 3,156 4,127 2,994 2,855 3,995 2,572 2,468 2,519 Cash flow from operations 9,051 3,264 7,827 8,984 7,003 5,020 5,322 9,512 Capex (6,398) (9,144) (9,407) (13,994) (12,300) (6,030) (4,030) (8,030) Net (acquisitions)/disposals 749 (94) (143) (1,746) (255) 5,055 0 0 Other investing CF items (3,811) 1,787 1,354 (3,035) (114) 415 (64) (78) Cash flow from investing (9,460) (7,451) (8,197) (18,775) (12,669) (560) (4,094) (8,108) Change in debt 2,174 11,472 (2,510) 14,396 630 (5,000) 1,000 1,000 Net share issues/(repurchases) 744 2,166 3,565 3 0 669 0 0 Dividends paid (281) (1,269) (106) (106) (56) 0 0 0 Other financing CF items (2,866) (2,428) (2,181) (2,595) (3,817) (3,147) (3,182) (3,186) Cash flow from financing (228) 9,941 (1,232) 11,698 (3,244) (7,478) (2,182) (2,186) Forex effect/others (4) 225 241 (189) 101 0 0 0 Change in cash (640) 5,979 (1,360) 1,717 (8,809) (3,018) (954) (782) Free cash flow 2,654 (5,880) (1,580) (5,010) (5,297) (1,010) 1,292 1,482 Source: FactSet, Daiwa forecasts

GCL-Poly Energy (3800 HK): 11 July 2019

Financial summary continued … Balance sheet (CNYm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 17,360 16,929 12,413 14,494 9,823 10,289 9,553 8,989 Inventory 2,696 1,387 966 991 992 993 924 973 Accounts receivable 13,658 14,368 12,285 14,537 13,309 14,282 16,305 17,329 Other current assets 605 372 1,449 1,089 2,751 1,306 1,248 1,227 Total current assets 34,319 33,055 27,112 31,111 26,875 26,870 28,030 28,518 Fixed assets 50,612 41,650 52,462 63,780 71,000 63,159 61,328 62,624 Goodwill & intangibles 806 231 302 1,030 801 701 601 500 Other non-current assets 4,266 4,756 7,144 11,359 13,818 14,134 14,645 15,138 Total assets 90,004 79,691 87,019 107,280 112,494 104,864 104,603 106,780 Short-term debt 25,174 24,271 14,545 22,598 27,084 25,182 26,159 27,136 Accounts payable 20,833 15,698 17,860 19,592 20,959 17,470 15,173 14,979 Other current liabilities 2,065 1,024 1,086 1,234 1,889 857 794 789 Total current liabilities 48,072 40,993 33,491 43,424 49,933 43,509 42,126 42,904 Long-term debt 18,794 20,310 28,399 35,614 34,657 31,559 31,582 31,605 Other non-current liabilities 1,775 829 1,736 934 1,072 370 358 354 Total liabilities 68,641 62,132 63,625 79,972 85,661 75,438 74,067 74,863 Share capital 1,549 1,372 1,632 1,632 1,610 1,743 1,743 1,743 Reserves/R.E./others 16,857 14,482 19,189 21,143 20,256 22,273 23,049 23,901 Shareholders' equity 18,406 15,854 20,821 22,775 21,866 24,016 24,792 25,644 Minority interests 2,958 1,705 2,573 4,532 4,967 5,411 5,744 6,274 Total equity & liabilities 90,004 79,691 87,019 107,280 112,494 104,864 104,603 106,780 EV 36,501 36,594 39,824 53,774 61,516 56,097 57,640 59,240 Net debt/(cash) 26,608 27,651 30,530 43,718 51,919 46,452 48,188 49,752 BVPS (CNY) 1.188 1.024 1.120 1.225 1.193 1.210 1.250 1.292

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 45.8 (45.0) 7.5 8.0 (13.6) (6.1) (6.6) (2.6) EBITDA (YoY) 83.4 (17.2) 14.1 10.9 (23.9) 36.1 7.7 5.0 Operating profit (YoY) 193.1 (20.3) 19.0 11.3 (43.7) 35.9 10.8 (5.0) Net profit (YoY) n.a. 24.0 (16.3) (2.7) n.a. n.a. 176.3 9.7 Core EPS (fully-diluted) (YoY) n.a. 20.8 (27.1) (5.4) n.a. n.a. 165.8 9.7 Gross-profit margin 20.6 28.0 32.0 34.5 24.5 32.9 37.6 36.8 EBITDA margin 26.9 40.4 42.9 44.1 38.8 56.2 64.9 69.9 Operating-profit margin 17.1 24.8 27.5 28.3 18.4 26.7 31.7 30.9 Net profit margin 5.3 11.8 9.2 8.3 (3.4) 1.5 4.3 4.8 ROAE 11.3 14.2 11.1 9.1 n.a. 1.2 3.2 3.4 ROAA 2.3 2.9 2.4 2.0 n.a. 0.3 0.7 0.8 ROCE 10.3 8.0 9.4 8.9 4.4 5.9 6.5 6.1 ROIC 10.6 9.0 9.9 8.4 4.5 5.5 6.1 5.6 Net debt to equity 144.6 174.4 146.6 192.0 237.4 193.4 194.4 194.0 Effective tax rate 22.9 17.4 18.9 21.9 10.3 17.1 17.1 17.1 Accounts receivable (days) 121.2 249.7 220.8 205.7 247.1 260.8 309.5 349.4 Current ratio (x) 0.7 0.8 0.8 0.7 0.5 0.6 0.7 0.7 Net interest cover (x) 2.4 2.7 3.1 2.9 1.2 1.7 1.9 1.8 Net dividend payout 0.0 46.6 0.0 0.0 n.a. 0.0 0.0 0.0 Free cash flow yield 35.3 n.a. n.a. n.a. n.a. n.a. 17.2 19.7 Source: FactSet, Daiwa forecasts

Company profile

GCL-Poly is the largest polysilicon and wafer manufacturer in China, with polysilicon production capacity of 115,000 tonnes and wafer manufacturing capacity of 30GW, as at end-2018.

China Information Technology 11 July 2019

(968 HK) Xinyi Solar Holdings Xinyi Solar H oldi ngs

Target price: HKD3.50 (from HKD3.80) Share price (9 Jul): HKD3.73 | Up/downside: -6.2%

Market leader in solar glass segment, but rich valuation Anna Lu, CFA (852) 2848 4465  Potential solar glass recovery in 2019E with XYE spin-off done [email protected]  Valuation seems rich compared with peers’ Dennis Ip, CFA (852) 2848 4068  Reiterating our Underperform (4) rating; trimming our TP to HKD3.5 [email protected]

What's new: With demand supported by global solar installation growth Forecast revisions (%) and the rising penetration of bifacial double-glass modules, we expect the Year to 31 Dec 19E 20E 21E solar glass segment to see a recovery this year. Also, XYS completed the Revenue change (1.7) (1.7) (1.7) spin-off of Xinyi Energy (XYE; 3868 HK, not rated) on 28 May, which is Net profit change (2.5) (1.9) (1.5) Core EPS (FD) change (2.6) (2.0) (1.7) beneficial for XYS as it implies less equity financing risk, in our view. We Source: Daiwa forecasts reiterate our Underperform (4) rating as valuations strike us as rich. Share price performance What's the impact: solar glass segment likely to see recovery in (HKD) (%) 2019E; XYE spin-off completed. Currently, solar glass demand is 4.6 210 supported by global solar installation growth and increasing penetration 3.9 178 rate of double-glass modules, which should support solar glass sales 3.3 145 volume and an ASP recovery from the low levels of 2H18. The processed 2.6 113 2.0 80 solar glass ASP has risen from CNY24/m2 at end-2018 to CNY26.3/m2 Jul-18 Oct-18 Jan-19 Apr-19 Jul-19 currently, and could increase further once China’s demand kicks off in Xinyi Sol (LHS) Relative to HSI (RHS) 2H19E. But we expect XYS’ 1H19 earnings to fall YoY due to the high base in 1H18. Separately, XYS completed the XYE spin-off on 28 May with the 12-month range 2.09-4.51 issuance equivalent to 28.41% of XYE’s enlarged share capital. About 90% Market cap (USDbn) 3.85 of the net proceeds, or HKD3,526m, was used to buy 540MW of solar 3m avg daily turnover (USDm) 19.74 farms from XYS (3 June). Going forward, XYE will focus on operating solar Shares outstanding (m) 8,058 Major shareholder Xinyi Glass (29.0%) farms (1,494MW currently), and XYS on solar glass production and solar farm EPC. We view the spin-off as beneficial to XYS, as XYE will take the Financial summary (HKD) cash-flow risk on solar farms despite still being consolidated within XYS. Year to 31 Dec 19E 20E 21E Revenue (m) 9,289 11,287 13,795 Valuation looks rich. XYS is trading at 2.2x 2019E PBR and 13x 2019E Operating profit (m) 3,156 3,691 4,393 PER, 0.5SD and 1.5SD above the past-3-year means, respectively. If we Net profit (m) 2,241 2,674 3,175 Core EPS (fully-diluted) 0.282 0.332 0.394 value the solar farm segment at 1.0x 2019E PBR (vs. GCL New Energy’s EPS change (%) 13.2 17.9 18.7 [GNE; 451 HK, not rated] 0.6x 2019E PBR) and 8x 2019E PER (similar to Daiwa vs Cons. EPS (%) (8.0) (9.1) (8.2) Longyuan’s [916 HK] and higher than GNE’s 6x), we estimate its solar PER (x) 13.2 11.2 9.5 glass business is trading at 3.9x 2019E PBR and 16x 2019E PER — Dividend yield (%) 3.7 4.4 5.2 DPS 0.138 0.163 0.193 significantly higher than the multiples of its peer, Flat Glass (6865 HK, not PBR (x) 2.2 2.0 1.8 rated) (1.5x 2019E PBR and 10x 2019E PER, based on Bloomberg). On 18 EV/EBITDA (x) 12.3 11.1 9.5 June, parent Xinyi Glass (868 HK, not rated) announced it would sell 3.9% ROE (%) 18.6 18.4 19.6 of XYS’ issued shares at a 9.2% discount to the last close, which, coupled Source: FactSet, Daiwa forecasts with XYS’ chairman’s sale of a 0.5% stake between 31 May and 5 June, would seem to support our view of XYS’ valuation as rich.

What we recommend: We reiterate our Underperform (4) rating but trim our SOTP-based 12-month TP to HKD3.5 (from HKD3.8) after fine-tuning the valuation of the solar glass and solar farm segments, implying a 2.1x 2019E PBR — slightly higher than past-3-year mean of 2.0x PBR. Key upside risk: higher-than-expected solar glass ASPs.

How we differ: Our 2019-21E EPS are 8-9% below the consensus, likely on our more conservative assumptions for solar glass ASPs and volumes.

See important disclosures, including any required research certifications, beginning on page 105

Xinyi Solar Holdings (968 HK): 11 July 2019

How do we justify our view?

Growth outlook Valuation Earnings revisions

Growth outlook XYS: gross profit forecast by segment (CNYm) We forecast XYS’ solar glass ASP to recover by 5% YoY in 6,000 140% 120% 2019E, driven by an improved supply-demand situation on 5,000 global solar installation growth and increasing penetration 100% 4,000 of double-glass modules. Besides, we expect solar glass 80% ASPs to be under pressure again given the increasing 3,000 60% capacity of major players in the solar glass industry and we 40% 2,000 assume stable ASPs for 2020-21E. In the meantime, XYS 20% 1,000 is aggressively adding capacity, from 6,700t/d as at end- 0% 2018 to 11,800t/d at end-2021E. 0 (20%) 2013 2014 2015 2016 2017 2018 2019E 2020E 2021E

We expect XYS to see a gross profit CAGR of 21% and net Solar glass Solar farm EPC & others YoY (RHS) profit CAGR of 19% over 2018-21E, driven by a 29% gross Source: Company, Daiwa forecasts profit CAGR in the solar glass segment and a 12% CAGR in the solar farm segment.

Valuation XYS: 1-year forward PBR XYS is trading at 2.2x 2019E PBR and 13x 2019E PER, PBR (x) 0.5SD and 1.5SD above past-3-year means, respectively 3.0 2.7x Avg+2SD — higher than the multiples of its peer, Flat Glass (1.5x 2.5 2.4x Avg+1SD 2019E PBR and 10x 2019E PER). 2.0 2.0x Avg 1.7x Avg-1SD 1.5 Our new SOTP-based 12-month TP of HKD3.5 implies a 1.3x Avg-2SD 2019E PBR of 2.1x, slightly higher than past-3-year mean 1.0 of 2.0x PBR. 0.5

0.0

Jul-18 Jul-16 Jul-17

Jan-18 Jan-17 Jan-19

Mar-17 Mar-18 Mar-19

Nov-17 Sep-16 Nov-16 Sep-17 Sep-18 Nov-18

May-18 May-19 May-17 Source: Bloomberg , Daiwa forecasts

Earnings revisions XYS: Bloomberg-consensus EPS forecast revisions (HKD) The consensus forecast of XYS’ 2019E EPS has been 0.50 relatively stable YTD, likely because higher solar glass 0.45 ASPs and GPM are seen as being offset by the share placement and higher minority interest resulting from the 0.40 spin-off of XYE. Meanwhile, the market has revised its 2020 and 2021 earnings forecasts by 5% and 23% YTD, 0.35 respectively, likely in view of the company’s aggressive capacity addition plans. 0.30

0.25 Our 2019-21E EPS are 8-9% below the consensus, which Jan-18 Mar-18 May-18 Jul-18 Sep-18 Nov-18 Jan-19 Mar-19 May-19 we attribute to our more conservative assumptions for solar 2019E EPS 2020E EPS 2021E EPS glass ASPs and volumes. Source: Bloomberg

100

Xinyi Solar Holdings (968 HK): 11 July 2019

Financial summary Key assumptions Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E

Total solar glass sales volume (k tonne) 557 1,065 1,210 1,710 1,626 1,867 2,331 2,910

Overall solar glass gross margin (%) 30.9 34.9 42.2 30.2 26.2 29.7 29.3 29.0 Total solar capacity (MW) 308 687 1,647 1,972 2,500 2,800 3,100 3,400 Average solar farm utilisation hours n.a. 1,136 1,078 981 1,008 1,004 1,002 1,004 Average solar tariff (CNY/kWh) 1.00 1.00 1.00 0.94 0.92 0.90 0.88 0.87

Profit and loss (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Solar glass 2,379 4,437 4,276 5,746 5,562 6,750 8,556 10,844 Solar farm 31 313 1,050 1,474 1,920 2,191 2,395 2,626 Other Revenue 0 0 681 2,307 189 348 336 325 Total Revenue 2,410 4,750 6,007 9,527 7,672 9,289 11,287 13,795 Other income 87 143 176 181 176 176 176 176 COGS (1,649) (3,040) (3,257) (6,122) (4,711) (5,563) (6,906) (8,553) SG&A (160) (278) (466) (649) (656) (716) (832) (988) Other op.expenses (111) (164) (11) (23) (38) (30) (33) (36) Operating profit 577 1,410 2,448 2,914 2,443 3,156 3,691 4,393 Net-interest inc./(exp.) (5) (16) (95) (165) (246) (282) (265) (308) Assoc/forex/extraord./others 0 0 37 40 49 43 43 43 Pre-tax profit 572 1,394 2,390 2,789 2,246 2,917 3,470 4,129 Tax (79) (188) (241) (265) (205) (296) (381) (484) Min. int./pref. div./others 0 0 (164) (192) (179) (380) (415) (469) Net profit (reported) 493 1,206 1,986 2,332 1,863 2,241 2,674 3,175 Net profit (adjusted) 493 1,206 1,986 2,332 1,863 2,241 2,674 3,175 EPS (reported)(HKD) 0.084 0.185 0.294 0.326 0.249 0.282 0.332 0.394 EPS (adjusted)(HKD) 0.084 0.185 0.294 0.326 0.249 0.282 0.332 0.394 EPS (adjusted fully-diluted)(HKD) 0.084 0.185 0.294 0.326 0.249 0.282 0.332 0.394 DPS (HKD) 0.042 0.086 0.140 0.150 0.120 0.138 0.163 0.193 EBIT 577 1,410 2,448 2,914 2,443 3,156 3,691 4,393 EBITDA 675 1,636 2,468 2,937 2,472 3,186 3,724 4,429

Cash flow (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Profit before tax 572 1,394 2,390 2,789 2,246 2,917 3,470 4,129 Depreciation and amortisation 98 226 19 23 28 30 33 36 Tax paid (67) (130) (265) (248) (233) (296) (381) (484) Change in working capital 117 (538) (1,771) (1,740) (358) (1,846) (1,828) (1,890) Other operational CF items (5) (100) 307 473 627 1,005 1,084 1,221 Cash flow from operations 715 851 681 1,298 2,310 1,809 2,378 3,011 Capex (2,381) (3,583) (4,810) (2,511) (3,059) (2,160) (2,669) (2,009) Net (acquisitions)/disposals 0 (54) (57) 1 (14) 0 0 0 Other investing CF items 53 24 (30) 20 37 7 41 31 Cash flow from investing (2,328) (3,613) (4,897) (2,490) (3,035) (2,153) (2,628) (1,978) Change in debt 1,300 2,290 3,076 1,235 806 0 1,000 1,000 Net share issues/(repurchases) 778 1,674 0 1,507 9 1,366 0 0 Dividends paid (200) (434) (844) (1,039) (641) (322) (1,112) (1,310) Other financing CF items 0 1,580 0 0 0 3,123 (797) (830) Cash flow from financing 1,878 5,110 2,232 1,703 174 4,168 (909) (1,141) Forex effect/others (2) (23) (42) 26 (46) 0 0 0 Change in cash 264 2,326 (2,025) 537 (597) 3,824 (1,159) (107) Free cash flow (1,666) (2,732) (4,129) (1,213) (748) (351) (291) 1,002 Source: FactSet, Daiwa forecasts

101

Xinyi Solar Holdings (968 HK): 11 July 2019

Financial summary continued … Balance sheet (HKDm) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Cash & short-term investment 543 2,869 843 1,381 784 4,651 3,535 3,471 Inventory 299 199 288 374 430 562 715 911 Accounts receivable 760 1,634 3,261 5,684 5,524 7,436 9,641 11,937 Other current assets 0 192 212 28 66 66 66 66 Total current assets 1,602 4,893 4,604 7,467 6,803 12,714 13,956 16,385 Fixed assets 3,685 7,104 11,079 14,240 15,804 17,176 18,956 19,979 Goodwill & intangibles 180 180 328 346 324 317 311 304 Other non-current assets 296 557 776 715 961 961 961 961 Total assets 5,764 12,735 16,786 22,767 23,893 31,168 34,184 37,628 Short-term debt 143 474 1,952 3,145 3,773 3,773 3,773 3,773 Accounts payable 1,121 2,156 2,539 2,941 2,780 2,978 3,508 4,111 Other current liabilities 28 79 100 155 184 184 184 184 Total current liabilities 1,291 2,710 4,591 6,241 6,737 6,935 7,465 8,067 Long-term debt 1,157 3,116 4,767 4,787 4,997 4,997 5,997 6,997 Other non-current liabilities 10 17 0 59 100 100 100 100 Total liabilities 2,458 5,843 9,359 11,087 11,834 12,031 13,561 15,164 Share capital 608 675 675 742 766 806 806 806 Reserves/R.E./others 2,698 5,070 5,541 9,379 9,668 12,913 14,476 16,340 Shareholders' equity 3,306 5,745 6,216 10,121 10,434 13,719 15,281 17,146 Minority interests 0 1,146 1,212 1,559 1,625 5,417 5,341 5,319 Total equity & liabilities 5,764 12,735 16,786 22,767 23,893 31,168 34,184 37,628 EV 30,812 31,747 36,752 37,713 39,232 39,157 41,196 42,238 Net debt/(cash) 757 722 5,877 6,552 7,986 4,119 6,235 7,299 BVPS (HKD) 0.544 0.851 0.921 1.363 1.362 1.703 1.897 2.128

Key ratios (%) Year to 31 Dec 2014 2015 2016 2017 2018 2019E 2020E 2021E Sales (YoY) 22.5 97.1 26.5 58.6 (19.5) 21.1 21.5 22.2 EBITDA (YoY) 44.1 142.3 50.8 19.0 (15.9) 28.9 16.9 18.9 Operating profit (YoY) 56.4 144.6 73.6 19.0 (16.1) 29.1 17.0 19.0 Net profit (YoY) 62.3 144.6 64.7 17.4 (20.1) 20.3 19.3 18.7 Core EPS (fully-diluted) (YoY) 14.8 120.0 58.8 10.8 (23.7) 13.2 17.9 18.7 Gross-profit margin 31.6 36.0 45.8 35.7 38.6 40.1 38.8 38.0 EBITDA margin 28.0 34.4 41.1 30.8 32.2 34.3 33.0 32.1 Operating-profit margin 23.9 29.7 40.8 30.6 31.9 34.0 32.7 31.8 Net profit margin 20.5 25.4 33.1 24.5 24.3 24.1 23.7 23.0 ROAE 17.6 26.6 33.2 28.5 18.1 18.6 18.4 19.6 ROAA 11.7 13.0 13.5 11.8 8.0 8.1 8.2 8.8 ROCE 16.7 18.7 19.9 17.3 12.1 13.0 12.7 13.8 ROIC 16.3 20.9 21.1 16.7 11.6 13.1 13.1 13.7 Net debt to equity 22.9 10.5 79.1 56.1 66.2 21.5 30.2 32.5 Effective tax rate 13.8 13.5 10.1 9.5 9.1 10.2 11.0 11.7 Accounts receivable (days) 110.9 92.0 148.7 171.3 266.6 254.6 276.1 285.5 Current ratio (x) 1.2 1.8 1.0 1.2 1.0 1.8 1.9 2.0 Net interest cover (x) 113.5 85.5 25.9 17.7 9.9 11.2 13.9 14.3 Net dividend payout 51.8 46.4 47.6 46.0 48.1 49.0 49.0 49.0 Free cash flow yield n.a. n.a. n.a. n.a. n.a. n.a. n.a. 3.3 Source: FactSet, Daiwa forecasts

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 end- 2018, the company was the largest solar glass manufacturer globally by production capacity (7,800tpd). XYS also develops downstream solar farms in Central and Eastern China.

102

China Wind and Solar: 11 July 2019

Daiwa’s Asia Pacific Research Directory

HONG KONG SOUTH KOREA Takashi FUJIKURA (852) 2848 4051 [email protected] Sung Yop CHUNG (82) 2 787 9157 [email protected] Regional Research Head Pan-Asia Co-head/Regional Head of Automobiles and Components; Automobiles; Jiro IOKIBE (852) 2773 8702 [email protected] Shipbuilding; Machinery Co-head of Asia Pacific Research Mike OH (82) 2 787 9179 [email protected] John HETHERINGTON (852) 2773 8787 [email protected] Banking; Capital Goods (Construction and Defence); Utilities; Steel Co-head of Asia Pacific Research Josh RHEE (82) 2 787 9124 [email protected] Craig CORK (852) 2848 4463 [email protected] Chemicals Regional Head of Asia Pacific Product Management SK KIM (82) 2 787 9173 [email protected] Paul M. KITNEY (852) 2848 4947 [email protected] IT/Electronics – Semiconductor/Display and Tech Hardware Chief Strategist for Asia Pacific; Strategy (Regional) Henny JUNG (82) 2 787 9182 [email protected] Kevin LAI (852) 2848 4926 [email protected] IT/Electronics – Semiconductor/Display and Tech Hardware (Small/Mid Cap) Chief Economist for Asia ex-Japan; Macro Economics (Regional) Thomas Y KWON (82) 2 787 9181 [email protected] Kelvin LAU (852) 2848 4467 [email protected] Pan-Asia Head of Internet & Telecommunications; Software – Internet/On-line Games Head of Automobiles; Transportation and Industrials (Hong Kong/China) Fiona LIANG (852) 2532 4341 [email protected] TAIWAN Industrials (Hong Kong/China) Rick HSU (886) 2 8758 6261 [email protected] Jay LU (852) 2848 4970 [email protected] Head of Regional Technology; Head of Taiwan Research; Semiconductor/IC Design (Regional) Automobiles and Components (Hong Kong/China) Nora HOU (886) 2 8758 6249 [email protected] Leon QI (852) 2532 4381 [email protected] Banking; Diversified financials; Insurance; Strategy Regional Head of Financials; Banking; Diversified financials; Insurance (Hong Kong/China) Steven TSENG (886) 2 8758 6252 [email protected] Kevin JIANG (852) 2532 4383 [email protected] IT/Technology Hardware (Automation & PC Hardware) Kylie HUANG (886) 2 8758 6248 [email protected] Banking (China) Anson CHAN (852) 2532 4350 [email protected] IT/Technology Hardware (Handsets and Components) Consumer (Hong Kong/China) Helen CHIEN (886) 2 8758 6254 [email protected] Adrian CHAN (852) 2848 4427 [email protected] Small/Mid Cap Consumer (Hong Kong/China) Andrew CHUNG (852) 2773 8529 [email protected] INDIA Punit SRIVASTAVA (91) 22 6622 1013 [email protected] Head of Gaming (Hong Kong/China) John CHOI (852) 2773 8730 [email protected] Head of India Research; Strategy; Banking/Finance Saurabh MEHTA (91) 22 6622 1009 [email protected] Head of Hong Kong and China Internet; Regional Head of Small/Mid Cap Capital Goods; Utilities Carlton LAI (852) 2532 4349 [email protected]

Small/Mid Cap (Hong Kong/China) SINGAPORE Dennis IP (852) 2848 4068 [email protected] Ramakrishna MARUVADA (65) 6228 6742 [email protected] Regional Head of Power, Utilities, Renewable and Environment (PURE); PURE (Hong Kong/China) Head of Singapore Research; Telecommunications (China/ASEAN/India) Anna LU (852) 2848 4465 [email protected] David LUM (65) 6228 6740 [email protected] Power, Utilities, Renewable and Environment (PURE) – IPP, Wind & Nuclear (China) Banking; Property and REITs Jonas KAN (852) 2848 4439 [email protected] Royston TAN (65) 6228 6745 [email protected] Head of Hong Kong and China Property Oil and Gas; Capital Goods Cynthia CHAN (852) 2773 8243 [email protected] Jame OSMAN (65) 6228 6744 [email protected] Property (China) Transportation – Road and Rail; Pharmaceuticals and Healthcare; Consumer Selwyn CHENG (852) 2773 8716 [email protected] Custom Products Group JAPAN Jack CHAN (852) 2773 8731 [email protected] Yukino YAMADA (81) 3 5555 7295 [email protected] Strategy (Regional) Custom Products Group

PHILIPPINES Renzo CANDANO (63) 2 737 3022 [email protected] Consumer Micaela ABAQUITA (63) 2 737 3021 [email protected] Property Gregg ILAG (63) 2 737 3023 [email protected] Utilities; Energy

103

China Wind and Solar: 11 July 2019

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 4325, 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, Midland Plaza 7th Floor, 10 Arbat Street, Moscow 119002, (7) 495 641 3416 (7) 495 775 6238 Moscow Representative Office Russian Federation Daiwa Capital Markets Europe Limited, Bahrain Branch 7th Floor, The Tower, Bahrain Commercial Complex, P.O. Box 30069, (973) 17 534 452 (973) 17 535 113 Manama, Bahrain 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 7 Straits View, Marina One East Tower, #16-05 & #16-06, (65) 6387 8888 (65) 6282 8030 Singapore 018936, Republic of Singapore Daiwa Capital Markets Australia Limited Level 34, Rialto North Tower, 525 Collins Street, Melbourne, (61) 3 9916 1300 (61) 3 9916 1330 Victoria 3000, Australia DBP-Daiwa Capital Markets Philippines, Inc 18th Floor, Citibank Tower, 8741 Paseo de Roxas, Salcedo Village, (632) 813 7344 (632) 848 0105 Makati City, Republic of the Philippines 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, (82) 2 787 9100 (82) 2 787 9191 Seoul, Korea Daiwa Securities Co. Ltd., Beijing Representative Office Room 301/302，Kerry Center，1 Guanghua Road，Chaoyang District， (86) 10 6500 6688 (86) 10 6500 3594 Beijing 100020, People’s Republic of China Daiwa (Shanghai) Corporate Strategic Advisory Co. Ltd. 44/F, Hang Seng Bank Tower, 1000 Lujiazui Ring Road, Pudong, (86) 21 3858 2000 (86) 21 3858 2111 Shanghai China 200120 , People’s Republic of China Daiwa Securities Co. Ltd., Bangkok Representative Office 18th Floor, M Thai Tower, All Seasons Place, 87 Wireless Road, (66) 2 252 5650 (66) 2 252 5665 Lumpini, Pathumwan, Bangkok 10330, Thailand Daiwa Capital Markets India Private Ltd 10th Floor, 3 North Avenue, Maker Maxity, Bandra Kurla Complex, (91) 22 6622 1000 (91) 22 6622 1019 Bandra East, Mumbai – 400051, India Daiwa Securities Co. Ltd., Hanoi Representative Office Suite 405, Pacific Palace Building, 83B, Ly Thuong Kiet Street, (84) 4 3946 0460 (84) 4 3946 0461 Hoan Kiem Dist. Hanoi, Vietnam

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

104

China Wind and Solar: 11 July 2019

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. Daiwa Securities Group Inc., its subsidiaries or affiliates do and seek to do business with the company(s) covered in this research report. Therefore, investors should be aware that a conflict of interest may exist. 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: Cromwell European REIT (CERT SP), Shenwan Hongyuan Group Co Ltd (6806 HK).

*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 or from any third party during the past 12 months. Daiwa India and its associates may have debt holdings in the subject company. For information on ownership of equity, please visit BlueMatrix disclosure Link at https://daiwa3.bluematrix.com/sellside/Disclosures.action. 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. Associates of Daiwa India, registered with Indian regulators, include Daiwa Capital Markets Singapore Limited and Daiwa Portfolio Advisory (India) Private Limited and Associates of Daiwa India having office in India but not registered with any Indian regulators include Daiwa Corporate Advisory India Private Limited.

Taiwan This research is solely for reference and not intended to provide tailored investment recommendations. This research is distributed in Taiwan by Daiwa-Cathay Capital Markets Co., Ltd. and it may only be distributed in Taiwan to specific customers who have signed recommendation contracts with Daiwa-Cathay Capital Markets Co., Ltd. and non-customers including (i) professional institutional investors, (ii) TWSE or TPEx listed companies, upstream and downstream vendors, and specialists that offer or seek advice, and (iii) potential customers with an actual need for business development in accordance with the Operational Regulations Governing Securities Firms Recommending Trades in Securities to Customers. Recipients of this research including non-customer recipients of this research shall not provide it to others or engage in any activities in connection with this research which may involve conflicts of interests. Neither Daiwa-Cathay Capital Markets Co., Ltd. nor its personnel who writes or reviews the research report has any conflict of interest in this research. Since Daiwa-Cathay Capital Markets Co., Ltd. does not operate brokerage trading business in foreign markets, this research is prepared on a “without recommendation” to any foreign securities basis and Daiwa-Cathay Capital Markets Co., Ltd. does not accept orders from customers to trade in such foreign securities. Recipients of this research shall carefully judge their own investment risk and take full responsibility for the results of any resulting investments in the companies and/or sectors featured in this research. Without the prior written permission of Daiwa-Cathay Capital Markets Co., Ltd., recipients of this research are prohibited from disclosing the research to the media, reprinting the research, or quoting from the research to other parties. 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 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 TNS, 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, TNS, 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. TNS, 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

105

China Wind and Solar: 11 July 2019

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.

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 research is distributed into the United States directly by Daiwa Capital Markets Hong Kong Limited and indirectly by Daiwa Capital Markets America Inc. (DCMA), a U.S. Securities and Exchange Commission registered broker-dealer and FINRA member firm, exclusively to “major U.S. institutional investors”, as defined under Rule 15a-6 promulgated under the U.S. Securities Exchange Act of 1934, as amended, and as interpreted by the staff of the U.S. Securities and Exchange Commission (SEC). This report is not an offer to sell or the solicitation of any offer to buy securities. U.S. customers wishing to effect transactions in any designated investment discussed in this report should do so through a qualified salesperson of DCMA. Non-U.S. customers wishing to effect transactions in any designated investment discussed in this report should contact a Daiwa entity in their local jurisdiction. The securities or other investment products discussed in this report may not be eligible for sale in some jurisdictions. Analysts employed outside the U.S., as specifically indicated elsewhere in this report, are not registered as research analysts with FINRA. These analysts may not be associated persons of DCMA, and therefore may not be subject to FINRA Rule 2241 restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account. ADDITIONAL IMPORTANT DISCLOSURES CAN BE FOUND AT: https://daiwa3.bluematrix.com/sellside/Disclosures.action

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 analyst is named on the report); and no part of the compensation of such analyst (or no part of the compensation of the firm if no individual analyst 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* 67.99% Hold** 22.52% Sell*** 9.49% Source: Daiwa Notes: data is for single-branded Daiwa research in Asia (ex Japan) and correct as of 30 June 2019. * 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

106