Overseas rush installations drove sharp wafer and solar-cell price increases in August, but weak demand means the industry-chain profitability inflection remains structurally divergent

Overview

This report tracks profitability inflection points across China's solar industry chain through monthly supply and demand, inventories, spot prices, input costs, and implied cash gross margins. Goldman Sachs concludes that profitability in wafers, solar cells, and glass improved somewhat in August, but the drivers differed: wafers and solar cells were primarily stimulated by short-term overseas rush installations, while glass was supported by expectations of production cuts; the polysilicon market still shows a disconnect between quotations and actual transactions.

Core views

The report first notes that solar-cell and wafer prices have risen by 26% and 30%, respectively, since August. The main catalyst was overseas stocking and rush-shipment demand through OEM channels amid Section 232, rather than a broad and sustained improvement in end demand. Goldman Sachs therefore views this round of price increases as short-lived: prices may decline after rush shipments fade in October. According to InfoLink, Chinese module companies are also currently unwilling to fully accept upstream price increases, further limiting their pass-through downstream. Although the cash profitability implied by August spot prices improved for wafers and solar cells, the report does not regard this as an established long-term profitability inflection point. The improvement in solar-glass prices and profitability operates through a different mechanism. Glass prices have risen by 10% since the beginning of August, and although inventories remained broadly stable at 47 days over the same period, the market continued to be driven by expectations of further production cuts. According to Oilchem, a new round of approximately 10% supply cuts is scheduled from late August to early September. Goldman Sachs expects glass inventory days to fall to 35 days in September as production cuts and a seasonal demand recovery take effect together, supporting broadly stable glass prices in the second half of 2026. Based on this price-inflection logic, the report prefers Xinyi Solar among its covered companies. The polysilicon segment, by contrast, shows divergence among quotations, futures, and actual transactions. Following industry self-discipline initiatives in early August, polysilicon quotations and futures prices rose by 25%, but spot prices remained broadly flat during the month because trading was subdued. Goldman Sachs remains cautious about whether spot price increases can actually be implemented, given weak end demand for solar products and persistently high downstream inventories. Without effective transactions and inventory digestion, higher quotations may not translate into actual profitability improvements for producers. The report therefore remains cautious on rod polysilicon companies and assigns Sell ratings to Daqo New Energy ADR/A-shares and Tongwei. Overall supply-and-demand data across the industry chain also did not confirm a broad-based improvement. The ratio of output across subsectors to module output deteriorated from 105% in July to 110% in August, indicating increased pressure from upstream output relative to downstream demand. Meanwhile, producer inventory days calculated against demand improved slightly from 48 days to 46 days. The two indicators moved in different directions, reflecting a slight inventory decline but a still-loose balance between production and demand. This is also an important reason the report does not directly interpret August's price increases as a supply-and-demand reversal across the entire industry chain. Regarding end demand, the report estimates global module demand at 35GW in July 2026, down 5% month over month and 8% year over year. Cumulative demand in the first seven months of 2026 was 267GW, down 41% year over year, materially weaker than Goldman Sachs' forecast trajectory of a 12% full-year decline in 2026. This demand measure is derived by aggregating module demand implied by Chinese installations, Chinese module exports, and demand implied by US AC-side installations. China's newly installed capacity rose 28% year over year to 14.1GW in July, improving from 12.5GW in June, when it declined 13% year over year. However, cumulative installations in the first seven months of 2026 still fell 61% year over year to 86GW, below Goldman Sachs' forecast trajectory of a 26% full-year decline. China's module exports fell 32% year over year to 15GW in July, a steeper decline than the 24% drop in June. Cumulative exports in the first seven months were 138GW, down 9% year over year, also weaker than the 6% year-over-year decline in the first half of 2026. By region, demand in the Middle East, Asia-Pacific, and European Union fell 43%, 41%, and 30% year over year, respectively, while African demand grew 3%. These data indicate that localized overseas policy-driven rush installations have not changed the overall weakness in global demand. In company selection, Goldman Sachs prefers Maxwell Technologies, First Applied Material, Xinyi Solar, and LONGi Green Energy. The investment case for Maxwell Technologies centers on order opportunities from new applications; First Applied Material benefits from solar encapsulant-film price increases and the potential for unit-profit expansion; Xinyi Solar offers exposure to an inflection in glass prices and inventories; and LONGi Green Energy is considered to have potential in the energy-storage-system business, with lower upstream prices helping enhance EBITDA resilience and BC technology offering medium-cycle upside. In contrast, polysilicon companies face weak demand, high inventories, and insufficient implementation of spot price increases, leading the report to maintain a cautious stance on Daqo New Energy and Tongwei. Regarding valuation, Maxwell Technologies' 12-month target price of RMB278 is based on expected 2027 EV/EBITDA: the core business and new-application orders are assigned EBITDA weights of 90% and 10% and valuation multiples of 15x and 85x, respectively, producing a blended multiple of 22x, which is discounted to 2026 at an 8.3% cost of equity. First Applied Material's target price of RMB21 applies 17x expected 2027 EV/EBITDA and is discounted to 2026 at a 10.7% cost of equity. LONGi Green Energy's target price of RMB14.2 and Xinyi Solar's target price of HKD3.0 apply expected 2026 P/B multiples of 2.3x and 0.8x, respectively, determined using the regression relationship between each company's historical P/B and ROE.

Analysis framework

Goldman Sachs first tracks monthly output, downstream demand, and inventories for polysilicon, wafers, solar cells, modules, glass, and encapsulant film, and then combines spot selling prices with input costs to estimate implied cash gross margins and changes in unit gross profit for each segment. The report subsequently links August price changes to overseas policy-driven rush installations, industry self-discipline, production-cut plans, and seasonal demand, and uses Chinese installations, module exports, and US AC-side installations to estimate global module demand. Finally, the institution maps industry changes to covered companies and determines target prices using either EV/EBITDA or the P/B–ROE relationship.

Methodology notes

• Industry/Sector Analysis FrameworkSupply-demand framework

  Monthly industry-chain supply, demand, and inventory tracking

  The report compares output across subsectors, module output, and inventory days to determine whether price changes are supported by genuine supply-and-demand improvements.

• Industry/Sector Analysis Framework

  Cash gross margins implied by spot prices and input costs

  The report subtracts corresponding input costs from spot selling prices to estimate cash gross margins and unit gross profit for each segment and observe the direction of monthly profitability. This estimate excludes company-level price discounts or premiums and may also differ from actual operating results due to shutdowns or maintenance at specific plants.

• Industry/Sector Analysis Framework

  Aggregate estimation method for global module demand

  Global module demand is calculated as the sum of demand implied by Chinese installations, Chinese module exports, and demand implied by US AC-side installations, and is used to compare monthly and year-to-date data with the full-year forecast trajectory.

• Valuation MethodEV/EBITDA valuation

  Expected 2027 EV/EBITDA discounted to 2026

  Maxwell Technologies applies a blended 22x expected 2027 EV/EBITDA based on weighted valuations of its core business and new-application orders, discounted at an 8.3% cost of equity; First Applied Material applies 17x expected 2027 EV/EBITDA, discounted at a 10.7% cost of equity.

• Valuation MethodPB valuation

  Target P/B based on the historical regression relationship between P/B and ROE

  LONGi Green Energy and Xinyi Solar apply expected 2026 P/B multiples of 2.3x and 0.8x, respectively, with the target multiples determined by the regression relationship between each company's historical P/B and ROE.

Asset mapping & comparison

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

• Maxwell Technologies (300751.SZ)

  The report identifies it as a preferred name, primarily based on order opportunities from new applications.

  Strengths

  New-application orders are incorporated into the target price at a higher valuation multiple, reflecting potential incremental business.

  Weaknesses

  New applications remain dependent on customers' experimental capital expenditure and the pace of industry development.

  Comparison

  It is one of the four companies explicitly preferred in the report.

  Risks

  Weaker-than-expected demand for new applications could reduce experimental capital expenditure and lead to order cancellations; tighter export controls on solar equipment could hurt profitability; the semiconductor industry may develop more slowly than expected.

• First Applied Material (603806.SH)

  The report believes rising solar encapsulant-film prices and unit-profit expansion could improve the company's earnings.

  Strengths

  It has potential to benefit from higher encapsulant-film prices and a recovery in unit profit.

  Weaknesses

  Profitability remains affected by solar installations, capacity utilization, and average encapsulant-film selling prices.

  Comparison

  It is one of the four companies explicitly preferred in the report.

  Risks

  Solar installations may fall below expectations, capacity expansion may proceed more slowly than expected, high-end non-solar businesses may underperform expectations, and supply disruptions may increase procurement pressure for resins and other raw materials.

• Xinyi Solar (00968.HK)

  The report identifies it as a beneficiary of inflection points in solar-glass prices and inventories.

  Strengths

  Glass production cuts, inventory declines, and price stabilization could support an earnings recovery.

  Weaknesses

  Profitability is relatively sensitive to solar-glass selling prices and raw-material and energy prices.

  Comparison

  Compared with the polysilicon segment, the report considers the case for supply contraction and inventory improvement in glass to be clearer.

  Risks

  Average solar-glass selling prices may be lower than expected, raw-material and energy prices may be higher than expected, and overseas capacity expansion may proceed more slowly than expected.

• LONGi Green Energy (601012.SH)

  The report is positive on its energy-storage-system potential, EBITDA resilience from lower upstream prices, and medium-cycle upside from BC technology.

  Strengths

  The investment case encompasses improvements on the cost side as well as potential in energy-storage systems and BC technology.

  Weaknesses

  The medium-cycle earnings improvement depends on cost-reduction technologies, BC, and the energy-storage-system business progressing as planned.

  Comparison

  It is one of the four companies explicitly preferred in the report and is considered to have more resilient EBITDA when upstream prices decline.

  Risks

  Stronger-than-expected demand combined with restrictive supply policies could drive a rebound in polysilicon or glass prices; cost-reduction technologies, BC technology, or energy-storage systems may develop more slowly than expected.

• Daqo New Energy ADR/A-shares (DQ, 688303.SH)

  The report remains cautious on the rod polysilicon segment and assigns Sell ratings to Daqo New Energy's ADR and A-shares.

  Weaknesses

  Polysilicon spot trading is subdued, while weak demand and high downstream inventories make it difficult to implement higher quotations.

  Comparison

  Compared with glass and the report's preferred equipment, encapsulant-film, and module companies, visibility into a polysilicon price recovery is lower.

• Tongwei (600438.SH)

  The report maintains a Sell view because it remains cautious about a recovery in rod polysilicon prices.

  Weaknesses

  Quotations and futures rose following industry self-discipline initiatives, but spot prices did not follow, leaving actual profitability improvements constrained by transactions and demand.

  Comparison

  The report's stance on the company is less favorable than on Maxwell Technologies, First Applied Material, Xinyi Solar, and LONGi Green Energy.

Key data

• Solar-cell pricesUp 26% since the beginning of AugustMainly driven by overseas stocking and rush-shipment demand amid Section 232
• Wafer pricesUp 30% since the beginning of AugustThe report expects prices may decline in October after overseas rush shipments fade
• Solar-glass pricesUp 10% since the beginning of AugustDriven by expectations of further production cuts
• Planned solar-glass production cutsApproximately 10%Scheduled for implementation from late August to early September
• Solar-glass inventoryApproximately 47 days in August, expected to fall to 35 days in SeptemberProduction cuts and a seasonal demand recovery are expected to jointly drive inventory reduction
• Polysilicon quotations and futures pricesUp 25%Rose following industry self-discipline initiatives, but spot prices remained broadly flat during the month due to subdued trading
• Ratio of subsector output to module output110%Rose from 105% in July to 110% in August, indicating a deterioration in the supply-demand balance
• Producer inventory days46 daysImproved slightly from 48 days in July to 46 days in August
• Global module demand35GW in July 2026Down 5% month over month and 8% year over year
• Cumulative global module demand267GW in the first seven months of 2026Down 41% year over year, below the forecast trajectory of a 12% full-year decline
• New installations in China14.1GW in July 2026Up 28% year over year; cumulative installations in the first seven months were 86GW, down 61% year over year
• Chinese module exports15GW in July 2026Down 32% year over year; cumulative exports in the first seven months were 138GW, down 9% year over year
• Changes in regional export demandMiddle East -43%, Asia-Pacific -41%, European Union -30%, Africa +3%All figures are year-over-year changes for July 2026
• Maxwell Technologies target priceRMB27812-month target price, based on 22x expected 2027 EV/EBITDA and discounted at an 8.3% cost of equity
• First Applied Material target priceRMB2112-month target price, based on 17x expected 2027 EV/EBITDA and discounted at a 10.7% cost of equity
• LONGi Green Energy and Xinyi Solar target pricesRMB14.2; HKD3.0Based on expected 2026 P/B multiples of 2.3x and 0.8x, respectively

Impact & implications

The report believes the August profitability improvement across the industry chain does not represent a broad-based and sustainable reversal. Policy-driven overseas rush installations support wafer and solar-cell prices in the short term, but weak end demand, resistance to price increases from module manufacturers, and a deteriorating production-to-demand ratio imply that prices may retreat. The case for glass-price stabilization is relatively more robust due to explicit production cuts and expectations of declining inventories. At the company level, Goldman Sachs places greater emphasis on opportunities in new applications, unit profitability for encapsulant film, the glass inflection point, energy storage, and BC technology, while continuing to avoid polysilicon companies facing uncertainty over the implementation of spot price increases.

Risks

• Maxwell Technologies faces risks from weaker-than-expected demand for new applications, cooling experimental capital expenditure, and potential order cancellations; tighter export controls on solar equipment and slower semiconductor-industry development could also weaken profitability.
• First Applied Material faces risks from lower-than-expected solar installations, capacity utilization, and encapsulant-film selling prices, slower capacity expansion and development of high-end non-solar businesses, and greater raw-material procurement pressure from supply disruptions.
• LONGi Green Energy faces risks from a stronger-than-expected rebound in polysilicon or glass prices and slower-than-expected development of cost-reduction technologies, BC technology, and energy-storage systems.
• Xinyi Solar faces risks from lower-than-expected solar-glass selling prices, higher-than-expected raw-material and energy prices, and slower overseas capacity expansion.

What to watch

• Monitor whether wafer and solar-cell prices decline in October as the report expects after Section 232-related overseas rush shipments fade.
• Monitor whether the approximately 10% glass production cut planned for late August to early September is implemented and whether inventory can decline from 47 days to 35 days in September.
• Monitor whether increases in polysilicon quotations and futures translate into active spot transactions and actual price increases.
• Monitor whether global module demand, Chinese installations, and exports improve and narrow the gap with the full-year 2026 forecast trajectory.
• Monitor Maxwell Technologies' new-application orders, First Applied Material's encapsulant-film unit profitability, and the progress of LONGi Green Energy's energy-storage systems and BC technology.