BHEL — a thermal-cycle recovery priced like a structural compounder

The Pulse

BHEL is India’s state-owned maker of the big iron inside power stations — boilers, turbines, generators — and it is unmistakably mid-recovery after a brutal decade. A thermal-power ordering super-cycle has filled its order book past ₹1.3 lakh crore, FY26 delivered record-since-2015 revenue of ₹33,782 crore and ₹1,600 crore of profit, and — most importantly — operating cash flow swung firmly positive (+₹5,837 crore) after years of cash burn. That’s a genuine turn. The problem is everything below the headline: even in this good year, return on equity was just ~6% (a three-year average of 3.2%), roughly 40% of pre-tax profit was non-operating other income, and the stock trades at 86 times earnings and 5.3 times book. The market is pricing a low-return, lumpy, working-capital-heavy capital-goods PSU as if its recovery is already a structural, high-return story. The thermal cycle is real; the valuation has run far ahead of the economics. (Note: BHEL’s recent quarterly calls had no public transcripts, so the management voice below is from FY22–FY24 calls; the current financial picture is from the FY26 snapshot.)

The Business

BHEL builds and installs the heavy equipment that power plants are made of, and services it across multi-decade lives. It spans thermal power, transmission, industrial systems, rail transport, renewables, oil & gas and defence — the classic heavy-engineering arm of the Indian state, with the Government of India as promoter. The economics are long-cycle and project-based: it books a large order, then designs, manufactures, ships, erects and commissions over years, recognising revenue in lumps. The seasonality is extreme — the March quarter routinely lands at more than double a normal quarter, and individual quarters swing to operating losses inside profitable years.

The most distinctive thing about BHEL is its inverted balance sheet for a heavy-engineering firm. Hard fixed assets are a rounding error (₹3,094 crore) against a ₹76,186 crore balance sheet — the company’s “capital” is overwhelmingly receivables, inventory and contract assets, financed partly by customer advances. It’s a project-financing balance sheet, not a factory one, which means the entire returns story lives or dies on working-capital discipline. Management has pulled that lever hard over the decade (working-capital days from 192 to 79), and FY26’s cash inflection is the payoff. But two structural weaknesses temper the picture: BHEL makes only the core boiler-turbine-generator package in-house — the balance-of-plant is outsourced to a vendor base that hollowed out during the down-cycle, a vulnerability management openly conceded — and gross margins slid for seven straight years (from ~44% to ~30%) before stabilising.

The forward engine is the thermal super-cycle: the government plans 80+ GW of new coal capacity by 2032, of which ~40 GW was still to be tendered as of the last call, with BHEL guiding to ~10–12 GW of ordering a year. Alongside it sits a deliberate diversification push — defence (the SRGM naval guns story is the proudest, 38 guns ordered in 18 months versus 44 in the prior 30 years), HVDC transmission (the explicit “high-value, steady-margin” play), and the Vande Bharat trainset programme. The standout ownership event: the government cut its stake 5% (to 58%) in a single quarter — an OFS, cleanly absorbed by domestic institutions now at ~24%.

How Management Thinks

The defining trait across the calls is a near-doctrinaire refusal to give guidance — no revenue, margin, capex, cash-conversion or receivables targets, framed as principled (“hard fact versus speculative”) rather than evasive. The asymmetry is telling: management volunteers rich, specific detail on the order pipeline and new-business wins, but retreats to “we’ll share offline” or “it cannot be ascertained like this” the moment an analyst asks why a record ₹78,000-crore order-inflow year produced just ₹260 crore of profit and deeply negative operating cash flow. The clearest example: three different analysts pressing the EBITDA-to-cash-conversion question, and three executives in turn declining to answer it. This is a management that wants the conversation to be about the order book, not the economics of converting it.

Where it counts, though, the candour is real. Management openly admitted the balance-of-plant vendor base has thinned to a structural risk, that gross margins slid for seven years, that the Vande Bharat order is “challenging” after a 12% price cut with profit located in the 35-year maintenance tail rather than the build, and that money is stuck in disputed legacy projects. That willingness to name the hard spots lends some credibility to the rosier order-book claims.

The implicit business model is volume-and-timing-driven, not margin-driven — the CMD’s worldview in one line: “it is timely execution… automatically, profitability will increase and cash flow also will increase.” The bet is that improved customer payment terms (more price-variation clauses, milestone advances) plus execution discipline plus a revived vendor base mechanically convert the order book into profit and cash — without ever committing to a number. On capital allocation, the posture is conservative: low debt, a healthy ~31% dividend payout, capex confined to replacing worn machines and selective diversification JVs, and headcount managed down through re-skilling rather than hiring. One earnings-quality flag worth holding: historically, profit has been materially flattered by net provision write-backs, and ~40% of FY26 pre-tax profit was other (treasury) income — the core engineering engine is leaner than the headline.

Where It’s Going

The bull case is a multi-year thermal ordering cycle (40 GW still to tender at a ~30%-plus BHEL share, plus a large flue-gas-desulfurisation retrofit opportunity), a record order book converting into rising revenue, and the FY26 cash inflection proving the working-capital curse is finally easing as legacy projects commission. Diversification into defence, HVDC transmission and railways adds higher-quality, less-cyclical revenue over time, and management’s stated ambition is to move from ~70:30 power:industry toward 50:50.

The bear case is the returns and the valuation. Even at the top of this recovery, ROCE reached only ~9% and ROE ~6% — a business that has structurally struggled to earn its cost of capital, now priced at 5.3× book. The risks compound it: extreme execution lumpiness, the hollowed-out balance-of-plant vendor ecosystem, customer concentration (a handful of central/state utilities, with a large new Adani slug), thin and volatile margins, and a long history of cash getting trapped in contract assets. The thermal cycle can drive revenue and even the order book higher; what it has not historically done is generate the kind of returns that justify a compounder’s multiple. The gap between an improving business and an extraordinary valuation is the whole story here.

The Four Checks

Quality & moat (gate). Weak-to-moderate, and this is the binding constraint. BHEL has a genuine edge — near-monopoly domestic capability in 800 MW boiler-turbine-generator sets, proven high-ash-coal technology suited to Indian coal, deep engineering competence, and government alignment. But it’s a low-return, lumpy, working-capital-heavy business that outsources most of the plant and competes on big tenders; the seven-year margin slide and sub-cost-of-capital returns show the moat doesn’t translate into economics. Where there’s no durable high-return moat, the remaining checks are largely academic — and that’s the case here.
Returns on incremental capital & runway. Poor. The runway (thermal capex, defence, transmission) is genuinely large and open, but the returns on capital deployed into it are single-digit (ROCE ~9% at the peak of recovery, three-year average ROE 3.2%). A long runway at sub-cost-of-capital returns destroys value rather than creating it; the bet is that mix shift (HVDC, defence, services) and execution lift returns materially — unproven over any sustained period.
Capital allocation for the stage. Reasonable in posture. Low leverage, a sensible dividend, capex restricted to maintenance and selective diversification, and a focus on releasing cash from contract assets rather than chasing growth capex — all appropriate for a low-return cyclical. The critique is transparency (no framework disclosed) and the historical reliance on provision write-backs to flatter earnings. Given the weak returns, returning more cash rather than reinvesting heavily would arguably serve shareholders better.
Price. Extremely demanding — the clearest negative. 86× earnings and 5.3× book for a business earning ~6% ROE, with ~40% of profit from non-operating income, is the market pricing a structural high-return compounder onto a recovering low-return PSU. Even granting a strong multi-year thermal cycle, the earnings would have to multiply enormously to grow into this valuation. The recovery is real; the price assumes far more than a recovery.

Sources

Concall transcripts read: Q4 FY24 (21 May 2024), Q4 FY23 (26 May 2023), Q4 FY22 (May 2022), Q1 FY22 (Jul 2021). Important limitation: screener had no transcript links for BHEL’s recent quarters (Jan-2025 through May-2026), so the most recent management commentary available is the May-2024 call — over a year stale. The current financial state is taken from the FY26 snapshot; the strategic narrative (thermal super-cycle, diversification) is from FY23–FY24 calls and is directionally but not freshly confirmed.
Annual reports: FY25, FY24, FY23 — all three AR section extracts were largely financials/boilerplate (no Chairman’s letter or MD&A narrative survived trimming); the FY25 numbers (revenue +18.6%, Industry-segment profit turnaround, rising customer concentration) are grounded in the segment tables.
Snapshot: screener.in (logged-out), fetched 2026-06-09 — source for all FY26 financials.
Research dumps: vault/Sources/Earnings/Bharat Heavy Electricals Ltd/.