Emmvee Photovoltaic — an integration bet on the right side of India's solar wall

The Pulse

Emmvee makes solar cells and the panels they go into, and for most of its life it was an unremarkable business — a 9% margin, single-digit-crore-profit assembler. Then, in the space of two years, it became something that looks spectacular on paper: FY26 sales of ₹5,050 crore (2.2x the prior year), net profit of ₹1,082 crore (3x), operating margins of 34%, and a return on capital near 45%. A late-2025 IPO recapitalised it — debt fell from ₹2,065 crore to ₹360 crore — and the promoter still owns 80%. The thing worth being clear-eyed about is that those returns are a peak, propped up by a tariff-and-policy wall keeping Chinese panels out of India and by Emmvee being early to the newer TopCon cell technology. Management is genuinely impressive — disciplined, technically fluent, candid about what it won’t disclose — and the runway is enormous. But this is still, underneath, a hardware-manufacturing business in an industry where prices have fallen 95% over twenty years and a wall of new capacity is coming.

The Business

Two things sit in a solar panel: the cell — the slice of silicon that actually turns sunlight into electricity — and the module, the framed, glassed-up panel that gets bolted to a roof or a field. Most Indian “panel makers” only do the second, easy half: they buy cells (usually from China) and assemble them. Emmvee does both. It is, by capacity, India’s second pure-play integrated cell-and-module manufacturer, and it makes its cells with TOPCon, the higher-efficiency technology that has displaced the older Mono PERC standard. In FY25 its sales were still 30% Mono PERC; today they are 100% TopCon, and it is already shipping the next format (G12R) before most of the market has moved.

Why does integration matter? Because the cell is where the value and the difficulty both live. Module assembly is comparatively trivial — anyone with a clean factory can do it, which is why the country has roughly 110 GW of module capacity chasing far less real demand. Cell-making is hard: semiconductor-grade clean rooms, high capex, fiddly process chemistry, and a long learning curve. Emmvee was one of the first in India to make TopCon cells at scale (it started in September 2024), and it feeds most of that cell output into its own modules. Management’s framing is blunt — “it is not about the size, but the conviction and the seriousness” — and the data backs the spirit of it: they cite a warranty-claim rate below 0.008% over three years, all lines running above 21% efficiency, and a roster of premium customers (Hero, CleanMax, Ayana, KP Green) who pay a little more for reliability.

The structural advantage, such as it is, comes from three things stacked together: the integration premium, the first-mover lead in TopCon, and — crucially — Indian policy. The government’s ALMM list (an approved-manufacturers register that effectively bars Chinese cells and modules from a growing share of demand) extends to cells from June 2026, and a chunk of orders carry “DCR” (domestic content) requirements that command a 5–7 cent-per-watt premium. Emmvee is on the right side of that wall. The order book is ₹-large and lengthening — about 9.3 GW at end-December including a 4.5 GW five-year cell contract — and management deliberately refuses to book beyond what it can ship in 12–18 months.

How Management Thinks

This is the strongest part of the story. The CEO, Suhas Manjunatha, runs the calls and does so like a teacher, not a promoter — explaining industry structure, contract mechanics and process chemistry in detail rather than reciting numbers. The single most-repeated message across both calls is a refusal to be judged on percentage margins: “Don’t look at percentage of margin… when we started manufacturing in 2006 the price of module was €3 per watt, but today the price is 15 cents — that doesn’t mean margins have gone down.” The metric they want you to watch is absolute EBITDA per watt, which they insist is stable through the price collapses because their large contracts pass raw-material costs — silver, aluminium, the dollar — straight through to customers.

That claim mostly holds up under their own evidence. When silver roughly doubled over four or five months, they neutralised it not by raising prices but by cutting silver consumption per cell 50–60% in three or four months through in-house process R&D, with another 40% reduction targeted. An analyst asked if that was really the mechanism; the chairman’s reply was simply “your understanding is 100% perfect.” Cell efficiency has crept from 24.5% to 25.2%. This is a company that competes partly on engineering, and the numbers back the words.

On capital allocation they have been rational for the stage. The IPO money went to killing debt, not a vanity acquisition — ₹1,621 crore of long-term loans repaid, finance costs already dropping quarter on quarter. They pay cash for land (the new 6 GW Devanahalli site), hold target leverage below 1x, and have signalled they may not even fully draw the ₹3,306 crore IREDA loan sanctioned for expansion, preferring internal accruals. The maiden dividend is token (6% payout), which is the correct call while the business can reinvest at these returns. The candour extends to what they won’t say: they openly decline to break out DCR-versus-non-DCR sales mix (“competitively sensitive”), won’t quantify exact silver consumption, and dodged a forward FY29–30 demand number twice. Guarded, but guarded out loud rather than evasive.

Where It’s Going

The trajectory is straightforwardly up-and-to-the-right, with the questions all about durability rather than direction. Capacity goes from roughly 10.3 GW of modules and ~2.9 GW of cells today toward a stated ~16.3 GW module and ~8.9 GW cell by FY28, anchored by the 6 GW integrated Devanahalli facility (land bought, financing closed). The next integration step — making the silicon wafers and ingots upstream of the cell — is announced as intent but deliberately left without a timeline, pending clarity on whether ALMM will extend to wafers too. That patience is characteristic.

The demand backdrop they describe is genuinely large: India installing toward 60 GW a year, rooftop running near 1 GW a month on the Suryaghar subsidy scheme, the Kusum agricultural programme and fast-growing commercial-and-industrial demand (data centres included) all feeding a domestic-content order pipe. They wave away the obvious bear points — China cutting export rebates (“really doesn’t matter much”), reported PPA cancellations (re-tendered, not killed) — with detailed, plausible answers.

The genuine tensions are three. First, the margins are a cyclical peak: a tight window in domestic cell supply plus the policy wall, and the pre-FY24 history (9% margins, 6–9% returns on capital) is a reminder of what this business looks like without those tailwinds. Second, a wave of capacity is coming — by management’s own account ~180 GW of module and ~140 GW of cell capacity is announced for post-2028 — and their defence (much of it won’t get built, and inexperienced cell-makers will fail) is reasonable but unproven. Third, and most concrete: the cash. FY26 turned only ₹200 crore of operating cash from ₹1,734 crore of operating profit, free cash flow has been negative for four straight years, and working-capital days have swung from -38 to +88 as inventory and receivables stretch. The profit-and-loss statement is gorgeous; the cash flow statement is the thing to keep watching.

The Four Checks

1. Quality & moat (gate) — 5/10. A real but contestable edge, and policy-dependent at its core. The genuine pieces: cell-making is hard and Emmvee is a credentialled early mover in TopCon, integration captures more of the chain, and process R&D gives a cost lead. The fragile pieces: the protective wall is regulatory (ALMM, DCR, import duties), not structural — it can be widened, narrowed or gamed by policy — and the product is ultimately a commoditising piece of hardware whose price has fallen 95% in twenty years. The pre-FY24 financials (9% margins, single-digit-crore profits) show what the business is without the up-cycle. Better than a pure price-taker; not a fortress. The checks below are therefore real but should be read through this lens.

2. Returns on incremental capital & runway — 7/10. Headline ROCE of 44.8% and ROE of 51.1% are exceptional, and the runway — India’s solar build-out, the company’s own path to 16+ GW — is long and open. Even if peak margins normalise toward the high-teens to mid-20s on return, that is still attractive reinvestment with years of demand to absorb it. The honest caveat is that today’s incremental returns are flattered by the cycle, and the reinvestment is currently being funded partly by stretched working capital rather than self-generated cash.

3. Capital allocation for the stage — 7/10. Textbook for a hyper-growth manufacturer: deleverage hard with IPO proceeds, reinvest aggressively while incremental returns are high, keep the dividend token, fund land with cash, hold leverage low, and stay disciplined on order intake. The restraint on drawing the full subsidised loan and the refusal to overbook the order pipe both read as genuine discipline rather than slogans. Minor demerit for the ballooning working-capital cycle, which is partly a growth artefact but bears watching.

4. Price — 5/10. Optically cheap, defensibly so. At ₹332 the stock trades on about 21x trailing earnings with 100%+ growth and a 50% ROE — a multiple that, taken at face value, looks undemanding. But the market is clearly already discounting the obvious: these are peak-cycle earnings in a commoditising industry facing a capacity glut, which is why it isn’t a 40x stock. Pay 21x and you are implicitly betting that the policy wall holds and margins don’t mean-revert hard. On reported numbers, full but reasonable; on through-cycle numbers, demanding. Call it fair — not the screaming bargain the trailing multiple suggests, not priced for perfection either.

Engine score: 19/30 (moat 5 + reinvestment 7 + allocation 7). Price 5.

Sources

• Concalls read: Q2 FY26 (call held 2 December 2025) and Q3 FY26 (call held 16 January 2026). Both via cleaned BSE transcripts.
• Annual reports: none available — Emmvee listed in late 2025 and has no annual report on screener.in yet. The Four Checks and trajectory lean more heavily on the two concalls and the financial snapshot as a result.
• Snapshot: screener.in (consolidated, logged-out) fetched 2026-06-11 09:43 IST.
• Gaps flagged: the most recent (May 2026) concall link returned an HTML intimation rather than a transcript PDF and could not be parsed, so the very latest quarter (Q4 FY26 / FY26 full-year call) is not captured here — though the FY26 annual figures are in the snapshot tables. Logged-out snapshot means the document list could in principle be stale. Promoter individuals beyond the Manjunatha family (named on the calls) are not detailed in the source. Quarterly EPS is distorted across the FY26 equity expansion and was not used as a clean series.
• Research dumps: vault/Sources/Earnings/Emmvee Photovoltaic Power Ltd/.