Goldman Sachs lowers its medium- to long-term earnings outlook for China solar, but turns selectively constructive on early-cycle inflection segments such as glass at trough valuations.

Overview

This report reassesses the cyclical inflection point of China’s solar materials industry. Goldman Sachs still expects a relatively mild cyclical bottom for leading companies in 2H26E, but believes the industry will continue to be affected by weak demand, slow capacity exits, and structurally low utilization rates. The report therefore cuts China solar installation forecasts for 2027E-2030E, industry chain prices for 2026E-2030E, and earnings forecasts for covered companies, while also pointing out that valuations in some segments have entered trough territory and are suitable for selective positioning.

Core views

The core view is that the overall outlook is weak, but structural opportunities exist. On the demand side, China’s solar installations for 2027E-2030E are cut by an average of 17%, with utility-scale projects affected by declining market-based electricity prices and a delayed recovery in self-consumption C&I projects; on the supply side, capacity exits are slower than previously expected, and nominal capacity is expected to broadly remain at current levels over 2026E-2030E, resulting in low utilization rates persisting for longer. On the earnings side, industry chain prices are cut by an average of 9%, EBITDA by an average of 38%, and target prices by an average of 33%. However, among subsectors, glass is showing an earlier inflection point due to inventory destocking, more cold repairs, improving supply-demand dynamics, and leading companies’ overseas capacity and cost advantages, leading to rating upgrades for Xinyi Solar and Flat A/H.

Analysis framework

The report uses a bottom-up China solar installation model, a supply-demand balance model, industry chain ASP and cost assumptions, company EBITDA and ROE forecasts, as well as historical regression valuation methods based on P/B and ROE. For film companies, it continues to use a discounted mid-cycle EV/EBITDA method, while for upstream materials and segments such as glass with weaker earnings, it shifts to 2026E P/B valuation to better reflect a slower earnings recovery.

Methodology notes

• Industry supply and demandS/D balance model

  Supply-demand balance model

  The report uses module demand, capacity, utilization rates, inventories, and price changes to determine the cyclical position of the solar industry chain, and notes that the model mainly references the module perspective.

• Valuation methodsP/B-ROE valuation

  P/B and ROE regression valuation

  For solar upstream, module, and glass companies, the report shifts to 2026E P/B combined with historical regression of 2027E-2028E ROE to determine target valuation.

• Valuation methodsEV/EBITDA valuation

  Enterprise value/EBITDA valuation

  For film company Hangzhou First, the report continues to use a discounted mid-cycle EV/EBITDA method because its valuation depends more on sustainable through-cycle growth.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Xinyi Solar (0968.HK)

  Core upgraded name, leading photovoltaic glass company

  Strengths

  Valuation is at a low level, overseas capacity mix is increasing, industry supply-demand is improving, and ROE is expected to recover from -3% in 2026E to 5% in 2027E-2028E.

  Weaknesses

  The target price is still cut from HK$3.5 to HK$2.8, reflecting weaker long-term earnings and price assumptions.

  Comparison

  Compared with non-leading glass makers, leading companies have higher overseas ASPs and lower unit production costs.

  Risks

  Glass ASP below expectations, slower-than-expected industry capacity cuts, higher-than-expected raw material and energy prices, and slower-than-expected overseas expansion.

• Flat A/H (601865.SS/6865.HK)

  Photovoltaic glass name with upgraded rating

  Strengths

  Improving glass supply-demand and better LNG usage efficiency from larger furnaces support its cost advantage.

  Weaknesses

  The target price still implies -13%/-15% downside, so the rating is only upgraded to Neutral.

  Comparison

  Like Xinyi, it is a leading glass company, but its risk-reward is less compelling than Xinyi’s.

  Risks

  Delayed glass price inflection, insufficient cold repair execution, rising energy costs.

• Hangzhou First (603806.SS)

  Film name in the preferred segment

  Strengths

  The film segment is listed as a relative preference, and the report continues to use mid-cycle EV/EBITDA valuation while raising the target multiple.

  Weaknesses

  Film prices are still cut by about 3% due to falling resin prices.

  Comparison

  Compared with upstream materials, film has better through-cycle growth characteristics.

  Risks

  Resin prices, module demand, and changes in the competitive landscape.

• Longi (601012.SS)

  Preferred high-efficiency module name

  Strengths

  High-efficiency modules are listed as an earlier inflection point and a preferred exposure direction.

  Weaknesses

  The downstream module segment as a whole still faces pressure from declining prices and low utilization rates.

  Comparison

  Better than traditional modules and some upstream material exposure.

  Risks

  Competition from new cell technologies and cost/performance progress falling short of expectations.

• Daqo ADR/A (DQ/688303.SS)

  Cautious polysilicon name

  Strengths

  Cash flow and funding gap risks may be manageable assuming short-term debt can be refinanced.

  Weaknesses

  Polysilicon price forecasts are sharply cut, and industry inventory and supply pressure remain high.

  Comparison

  Weaker than film, glass, and high-efficiency modules.

  Risks

  Further declines in polysilicon prices, inventory buildup, and slow capacity exits.

• Tongwei (600438.SS)

  Cautious polysilicon and cell/module name

  Strengths

  If industry supply clears faster, the earnings inflection point may come earlier.

  Weaknesses

  One of the largest ROE downgrades, with cuts to polysilicon and cell/module prices combined with rigid operating expenses.

  Comparison

  The report is most cautious on the polysilicon segment within coverage.

  Risks

  Price declines, operating leverage, fixed cost pressure, and weaker-than-expected demand.

Key data

• YTD share price performance of covered namesAverage decline of 37% YTDThe report says the decline occurred against a backdrop of an average 16% drop in industry chain prices, global demand challenges, and insufficient supply response.
• Revision to China solar installation forecast for 2027E-2030EAverage cut of 17% to 263GWMainly reflects lower-than-expected market-based spot electricity prices and a delayed recovery in self-consumption C&I demand.
• 2026E China solar installation forecast235GW, -26% yoyThe 2026E forecast is unchanged.
• 2026E-2030E utilization rate assumption29%-68%Goldman previously forecast 60%-80%, but now believes low utilization rates will persist longer.
• Revision to industry chain ASP forecastsAverage cut of 9% for 2026E-2030EPolysilicon is cut by about 20%, glass by about 10%, and film by about 3%.
• Revision to EBITDA forecastsAverage cut of 38% for 2026E-2030E (excluding Daqo)About 13 percentage points come from demand cuts, 10 percentage points from price cuts, and 16 percentage points from higher costs caused by operating leverage.
• Revision to 12-month target pricesAverage cut of 33%The latest target prices for the coverage basket imply an average downside of about 3%.
• Glass price inflection point2H26E up 9% versus 2Q26, with a further 16% increase in 2027EThis assumes supply discipline is well executed and cold repairs and inventory destocking continue.
• Xinyi Solar rating and target priceBuy, HK$2.8, implying +38% upsideUpgraded from Neutral to Buy, valued at 0.8X target 2026E P/B.
• Flat A/H rating and target priceNeutral, Rmb8.0/HK$5.5Upgraded from Sell to Neutral, implying about -13%/-15% share price downside.

Impact & implications

The investment implication is that overall earnings recovery in China’s solar industry will be slower, so investors should not simply bet on a rebound across the entire industry chain; instead, they should selectively choose subsectors with earlier supply-demand inflection points, cost-curve advantages, or exposure to higher-priced overseas markets. The report prefers film, glass, and high-efficiency modules, believing leading glass companies benefit from cold repairs, falling inventories, rising overseas capacity mix, and lower fuel costs; it is relatively cautious on polysilicon, as prices may approach the full cash cost of efficient leading producers while inventory pressure is greater.

Risks

• China solar demand is lower than expected, especially for utility-scale projects affected by low electricity prices and declining IRRs.
• Capacity exits or cold repair execution are slower than expected, leading to weaker-than-expected recovery in utilization rates and prices.
• Power curtailment, base project construction, overseas installations, and changes in module exports may affect China’s addressable demand.
• Changes in new cell technologies may alter profit distribution across the industry chain and the competitiveness of existing technology routes.
• Rising raw material, sand, soda ash, oil, and natural gas energy costs may compress glass companies’ margins.
• Slower-than-expected overseas expansion may weaken the high-ASP advantage of companies such as Xinyi Solar.

What to watch

• Whether glass prices show an inflection point in 3Q26 and 2H26 as expected.
• The scale of cold repairs in the glass industry, inventory days, and execution of supply discipline.
• China’s monthly solar installation and monthly module export data.
• Changes in market-based spot electricity prices, utility-scale project IRRs, and regional power pricing mechanisms.
• Construction progress of large base projects and project news from state-owned power generation companies.
• Approval progress and payback period changes for self-consumption C&I projects.
• Company news, earnings calls, bidding volume/pricing, and industry exhibitions such as EU Intersolar and Shanghai SNEC.