Why did the supply-demand balance point move forward from 2030 to 2029?

  • The previous base case assumed the entire data-center project pipeline would proceed. The new base case includes only projects under construction and announced projects with a high probability of proceeding, excluding those with low or medium probabilities. The Goldman Sachs team estimates that the excluded projects correspond to 213 GW of cumulative additional data-center capacity by 2030, making the demand scale smaller under the new base case.
  • Under the new base case, Goldman Sachs assumes non-data-center demand remains at 75–77 GW per year, while industry orders decline to 120 GW per year in 2028–2030. As alternatives ramp up, industry supply and demand are expected to balance in 2029, with demand slowing in 2030. If all projects proceed, the balance point would instead be 2030.

How would weaker demand affect Yingliu’s orders and earnings?

  • Goldman Sachs assumes that smaller gas-turbine manufacturers will stop expanding capacity from 2029; these manufacturers account for about 13%–19% of Yingliu’s revenue in 2026–2030. Easing supply-demand tightness would reduce Yingliu’s pricing power in 2028–2030, while customers would be more likely to stay with existing suppliers, limiting its ability to gain market share.
  • On this basis, Goldman Sachs cut its forecast for Yingliu’s earnings per share in 2026–2030 by 7%–29%, reflecting lower estimates for revenue, gross margin and operating leverage, as well as dilution from the conversion of convertible bonds. On the other hand, long-term contracts with Siemens Energy and Baker Hughes are expected to account for about 18% of total revenue, supporting revenue visibility through 2028. The company expects that contract renewals could extend this to 2030–2032.

Why focus on gross margin and growth despite the third-quarter beat?

  • Yingliu reported third-quarter 2026 revenue of 1.113 billion yuan, up 51% year on year, and net profit of 150 million yuan, up 42%. Both figures exceeded Goldman Sachs’ prior forecasts by 14% and 19%, respectively. Goldman Sachs said the results beat expectations on increased deliveries from the dual-engine business and growth in traditional businesses.
  • Third-quarter deliveries from the dual-engine business were valued at 530 million yuan. The company said output reached 180 million yuan per month in September, with a target of increasing it to 200 million yuan per month by the end of 2026. It plans to raise capacity to 300 million yuan per month by the end of 2027, corresponding to about 3 billion yuan in dual-engine revenue. However, the company’s overall revenue target for 2027 is more than 5 billion yuan, implying growth of about 22% year on year, below Goldman Sachs’ forecast of 41% revenue growth in 2026.
  • Growth has also come with gross-margin pressure: gross margin was 34.4% in the third quarter, down 4 percentage points year on year, mainly due to higher prices for raw materials including tungsten, cobalt and molybdenum. The company said gross margin in the dual-engine business remained above 40%, but efforts to expand market share left limited room to raise prices for customers.

Why does Goldman Sachs maintain its Buy rating after cutting the price target?

  • Goldman Sachs cut its 12-month price target from 72.2 yuan to 37.5 yuan, a reduction of about 48%. It now values the stock at 25 times forecast earnings for 2029, rather than the previously used 30 times forecast earnings for 2030, and discounts the valuation to 2027 using a 10% cost of equity. Based on the closing price of 27.8 yuan listed in the report for 17 November, the price target implies upside of about 34.9%; the Buy rating is maintained.
  • Another pillar of the Buy view is the business mix: Goldman Sachs expects domestic gas turbines and aerospace to together account for 24%–30% of revenue in 2026–2030, and is positive on steady growth from the aerospace expansion cycle and the localization of domestic gas turbines. Replacement demand for blades also makes that business less cyclical than complete gas-turbine units.

Further reading