Deck
A Xinjiang-based heavy-electrical group that builds power-grid transformers, cables and switchgear, mines coal and generates electricity, and — through its two-thirds-owned Xinte subsidiary — makes high-purity polysilicon for solar.
The FY2025 profit rebound is narrower than the headline number.
About half of TBEA's ¥5.95bn FY2025 profit came from one 85.78%-owned coal subsidiary whose margin is falling, and about a quarter was a single non-cash re-mark of one pre-IPO stake (Huadian, ¥1.00bn cost written up to ¥2.50bn), so roughly ¥4.3bn of the ¥5.95bn 'recovery' is one shrinking coal engine plus one paper gain. After those two legs, only about ¥1.6bn is recurring earnings from everything else. The counter-fact sits in the same numbers: the paper gain is genuine liquid value — ¥1.80bn unrealised, monetisable once Huadian's lock-up expires in July 2026 — and the coal engine is cushioned by a captive-power hedge whose gross profit (¥3.93bn) overtook coal's (¥3.80bn) in FY2025.
Capex has driven free cash flow to -¥12.75bn, part-funded from outside.
TBEA spent ¥22.08bn of capex against ¥9.33bn of operating cash in FY2025, pushing free cash flow to -¥12.75bn, a third straight year of decline; the spend has rotated from the completed polysilicon plant into a ¥17.03bn coal-to-gas project and a ¥6.78bn alumina build, part-funded by ¥4.59bn of fresh cash from subsidiaries' minority shareholders rather than TBEA's own. Depreciation is only about ¥6.51bn, so most of the spend is discretionary growth capital — a choice rather than distress, with free cash flow recovering the year the builds finish. Net debt at ~22% of equity and ¥25.69bn of cash leave room to carry it; the risk is duration, if capex stays above ¥20bn while coal and polysilicon margins stay depressed.
The ¥112bn valuation rests on the grid and coal engines.
- Grid — scale, not margin. TBEA is China's volume and UHV-capability leader: transformer-line revenue grew from ¥10.9bn to ¥26.8bn and orders reached ¥56.2bn signed in FY2025. But its ~20% segment margin sits below focused peers China XD, Pinggao and Sieyuan — scale is the moat, pricing power is not.
- Coal — the largest earner, cooling. The 85.78%-owned Tianchi arm netted ¥3.39bn (~¥2.91bn to TBEA) on captive pithead coal and 5,040MW of thermal power, the group's biggest single profit. Its product margin has fallen more than 10 points to 22.4%, and net profit has roughly halved since FY2023.
- Polysilicon — the swing, priced low. The 66.6%-owned Xinte arm lost ¥1.33bn as new-energy gross margin collapsed from ~58% to near zero; its separately-listed shares value the stake at only a small fraction of the market cap, so a recovery reads more as optionality than as what the price pays for.
A quarter of FY2025 profit was non-recurring, lifting the recurring multiple to ~25x.
- Reported vs recurring. At ¥22.01 and a ¥112bn cap, the stock trades near 19x reported FY2025 earnings and ~1.5x attributable book — but strip the ~¥1.40bn (24%) of non-recurring gain and it is closer to 25x recurring.
- Consensus is a revenue call. Five analysts carry ~10% revenue growth to roughly ¥107bn and ¥118bn, with a ¥32.96 target about 50% above spot — yet there is no forward EPS consensus, and reported EPS has missed estimates 23–50% for four straight prints.
- Bankable top line, un-underwritten recovery. The revenue line is the dependable part of the case; the earnings that turn it into a return are what the recent record will not let a buyer assume.
Whether the core is worth owning through the down-cycle turns on coal, not polysilicon.
- The bull. A share-gaining grid franchise and a cash-generative coal engine carry the company through a polysilicon trough the market already prices near zero, with free cash flow recovering as the capex build rolls off.
- The bear. The group's largest earner is cyclically shrinking, the grid earns thin margins, and a ¥22bn capex programme keeps free cash flow deeply negative — funded partly by debt and ¥4.59bn of outside minority cash.
- What settles it. The valuation already discounts polysilicon to a few percent of the cap, so the case is most sensitive to coal — the biggest attributable earner, whose margin has already fallen more than 10 points in a year.
Watchlist to re-rate: Coal's 22.4% product margin holding rather than falling further; polysilicon spot clearing Xinte's cash cost with utilisation lifting off the 37% floor; and free cash flow turning back toward zero as the coal-to-gas and alumina builds finish.