The Three Musketeers of Gas Turbines

$Triangle Defense (SZ300775)$ $Yingliu Co. (SH603308)$ $Hangyu Technology (SH688239)$

The initial reason for paying attention to the engine sector was because Boeing couldn’t produce aircraft. Part of the reason was its own design and management issues, but another part came from the aviation supply chain. Assembling a plane requires tens of thousands of components, each a specialized, precision-processed part. Compared to general mechanical components, these are produced in much smaller quantities, have low substitutability, and require complex certification processes. During the pandemic, combined with air crashes, the supply chain shut down, and many skilled workers were lost (most of them were older, trained during the golden age of aviation; younger workers rarely enter this field now). When production gradually tried to resume, it was discovered that many small component suppliers had gone bankrupt, exited the business, or couldn’t ramp up capacity, even though demand was surging — both from new aircraft orders and maintenance needs for existing fleets. As a result, the supply-demand balance collapsed: engine prices soared, maintenance times were delayed, and component shortages became even worse.

At that time, the focus was mainly on GE and Rolls-Royce, since they basically monopolize the global aircraft engine market. There wasn’t much consideration for Chinese industrial substitution, as this is not an industry that can be broken into in the short term.

Then came AI computing power — a much stronger demand driver than aviation. To meet the massive power requirements of data centers while also satisfying national carbon peak targets, gas turbines became the only feasible option. Coal power is impossible to expand; solar and wind are too volatile, geographically limited, and land-intensive (and where there’s sun, there isn’t always water for cooling). Nuclear power plants take at least five to ten years to build, and after the major accidents in the U.S., related R&D mostly stopped — only China, South Korea, and Russia continue to invest. Now, large U.S. nuclear plants mostly use Korean equipment and are operated domestically by companies like Constellation Energy. SMR technology is immature and severely over budget. Thus, gas turbines — which are fast to build, technologically mature, powerful, and clean — remain the best option. Over 60% of new power orders now come from gas turbines, with total order value in the hundreds of billions, stretching beyond 2030. As Sun once said: short-term chip shortage, long-term power shortage, and always a shortage of storage. The gas turbine industry therefore enjoys a highly sustained boom, with strong certainty, virtually no upper limit, and high barriers to entry due to component shortages and technological difficulty — making it an excellent investment sector. Of the two giants, one (PCC) was acquired by Buffett, and the other, Howmet Aerospace (HWM), remains — this is its stock performance.

This is an industry that breeds multi-bagger stocks. Upstream, the U.S. also has a specialty alloy producer called Carpenter Technology, which shows a similar stock trajectory.

What’s rare and valuable about this sector is its stability: no matter which chips or technical routes AI adopts, electricity will always be needed. The sector’s long-term growth trend looks like a steady 45-degree line upward — that’s why I like it so much. It suits value investors: companies you can sleep well owning.

Recently, Siemens and other major turbine companies have come to China to sign supply agreements, bringing more Chinese firms into the global value chain. The main reason is that U.S. and European component suppliers lack enough capacity, so OEMs urgently need to expand their supplier base to meet growing demand. Only a few Chinese firms have the industrial capability to step in — I believe supplier integration is happening faster than expected, within six months after sample delivery, most can enter mass production. As cooperation deepens, the value of supplied components will rise, since Chinese manufacturers respond quickly and will be prioritized for future contracts.

Yingliu Co.
In terms of component value, the top choices are turbine blades and disks, as they have the highest technical barriers and value. Next come components like casings and cylinders, which have lower gross margins and are less scarce.

Regardless of valuation, purely in terms of business quality, the ranking would be Yingliu > Triangle > Hangyu. Yingliu basically has a domestic monopoly on turbine blades, with high gross margins near 40%. Most importantly, it’s highly focused: 60% of its business comes from high-temperature alloys (for the “two engines” — aviation and gas turbines), and over 20% comes from nuclear power (and potentially future nuclear fusion applications). It’s the only truly scarce A-share target in this field. Its valuation is high — around RMB 25 billion market cap, with expected RMB 600 million profit next year, about 40x PE — roughly in line with Howmet Aerospace. Given the strong industry outlook, this valuation seems fair, with earnings growth potentially over 30%.

Triangle Defense & Hangyu Technology
Triangle and Hangyu are quite similar. Both originated from the military sector, with a “military-to-civilian” transformation theme. Hangyu’s biggest customer accounts for roughly 35–40% of revenue from Aero Engine Corporation of China, while Triangle’s exposure is close to 80%. Triangle mainly supplies the defense sector, while Hangyu has more exposure to civil aviation. Both suffered from declining military orders since late 2023 through mid-2025, but have recently shown signs of stabilization and renewed growth — possibly due to aircraft model adjustments or post-pandemic demand shifts.

Military business still offers higher gross margins and stability. Triangle’s gross margin is roughly 40%, similar to Yingliu, while Hangyu’s is around 28%. Triangle focuses on disk forgings, while Hangyu makes ring forgings. I asked ChatGPT — it said disks are technically harder and carry higher value, which explains Triangle’s better margins. Triangle has also started small-scale production of forged blades. Siemens is confirmed to source disk parts, but it’s unclear whether they’ll also source blades — if they do, that’s a bonus, since blades have higher unit value than Yingliu’s.

Their investment logic mainly revolves around: (1) recovery in military orders, marking a cycle rebound; (2) growth in the new gas turbine business as a second driver, lifting valuation via military-to-civil diversification. There are also other potential growth drivers — C919 aircraft, commercial spaceflight (Blue Arrow’s Zhuque-3 rocket’s upcoming first reusable launch is a milestone, echoing SpaceX’s model), nuclear fusion, and deep-sea equipment. These are all related to high-end forging, making future growth highly promising.

For Hangyu, based on brokerage estimates, next year’s profit could reach RMB 330 million, and by 2030 between RMB 800 million and 1 billion, implying a 25% CAGR (potentially higher if turbine orders expand or Aero Engine Power orders surge). Assuming RMB 1.5 billion turbine revenue and RMB 4 billion aviation revenue by 2030, a 30–35x PE seems fair — lower than Yingliu’s due to less concentration and slightly weaker customer profile. That gives roughly RMB 11.5 billion valuation (35x * 330m profit), which looks attractive given its current RMB 10 billion market cap — suitable for medium- to long-term holding.

I personally favor Triangle Defense more. Its chart shows a rebound from the bottom, and its valuation is lower. Military companies used to trade at around 40x PE. With an expected RMB 600 million profit next year (excluding turbines) and a RMB 16 billion market cap (~26x PE), adding gas turbine upside gives more room. It’s hard to estimate total value yet, but gross margin should approach Yingliu’s or slightly lower. The company reports over 40 alloy grades currently under certification, expected to complete by April 2026. These products have high value and 30–40% margins. I believe it’s worth at least RMB 5 billion more, suggesting a fair value around RMB 29 billion (40x6 + 5). That’s roughly 100% upside within a year.

Feedback and discussion welcome.