Craftsman Automation — betting the balance sheet on becoming an aluminium giant

The Pulse

Craftsman Automation is a Coimbatore precision-engineering house that has, in five years, deliberately turned itself into something five times bigger and quite different — buying its way (DR Axion, Sunbeam, a German foundry) and building its way into aluminium die-casting, which has now overtaken its original truck-engine machining business. FY26 revenue jumped 42% to ₹8,069 crore and profit recovered to a record ₹384 crore after an FY25 air-pocket. But the scale came at a visible price: borrowings nearly tripled to ₹3,623 crore, interest costs hit ₹309 crore, return on capital halved from 21% to the low-teens, operating margins reset structurally from ~27% to ~15%, and free cash flow has been deeply negative two years running. The founder is admirably blunt that he is choosing growth over deleveraging. So this is a high-conviction, debt-funded transformation in mid-flight — and at 55× earnings, the market is paying handsomely for the bet that the new capacity earns its keep.

The Business

Craftsman is fundamentally a build-to-print precision manufacturer — it doesn’t own the end product, it owns the machines, the process know-how and the tight tolerances, and makes components to a customer’s drawing. Three legs: Powertrain (machining critical engine and transmission parts for trucks and tractors — the original franchise, where it claims to be India’s No.1 in heavy-CV machining), Aluminium die-casting (now the larger leg, scaled hard via the DR Axion and Sunbeam acquisitions), and Industrial & Engineering (warehouse storage and automated retrieval systems, including a genuinely in-house product, V-Store). It’s a deeply capital-heavy model — ₹4,676 crore of fixed assets, ₹444 crore of annual depreciation — where everything rides on capacity utilisation and the spread over metal, machine-time and interest.

The distinctive thing about the last three years is the deliberate, front-loaded scale-up. Founder Srinivasan Ravi’s argument is that Indian aluminium tier-1 suppliers are sub-scale (sub-$0.5bn versus global peers at $3–8bn), and you simply cannot bid for $100–200 million multinational orders without first building the capacity — so he is building ahead, funded by debt and acquisitions, and consolidating DR Axion, Sunbeam and Craftsman’s own aluminium into one entity targeting $1 billion in aluminium within 2–3 years. DR Axion is the jewel (~20% margins, full plants); Sunbeam was bought distressed and is a slow turnaround; the German Fronberg foundry is the entry ticket to the marquee bet: large stationary engine blocks for data-centre backup power, riding the AI energy boom, where only 8–9 firms worldwide compete and Craftsman supplies three of the four leaders. That order book ($100 million) is now “finalised” — but revenue doesn’t start until FY29.

The cost of all this is written across the balance sheet: debt near-tripled, free cash flow of roughly minus ₹700 crore two years running, margins re-rated down as lower-margin aluminium volume swamped high-margin powertrain. Ownership shifted too — the promoter sold down once (to ~49%) in 2024, and domestic institutions have steadily mopped up the float to 28%.

How Management Thinks

Ravi runs the calls solo — no CFO voice — and the defining trait is candour over polish. He openly admits Sunbeam’s margins are “lacking behind,” that the Gurgaon land he’s trying to sell is “elusive,” that new capex “spoils” reported margins for a couple of years before maturing, and — most strikingly — that growth and deleveraging are a genuine trade-off he is consciously choosing: “if we grow at 5%, the debt will be gone in four years and there will be no new customers.” That’s an unusually honest framing of why he’s keeping the leverage on. He volunteers disclaimers (“please negate that if taken as guidance”) and leads with the uncomfortable facts.

The flip side: he’s persistently evasive on absolute numbers — declining to read out the debt figure or specific plant expenses, steering everything to ratios (net-debt-to-EBITDA, which he manages to rather than absolute debt; a 20% pre-tax ROCE hurdle on new capex). His most-repeated talking point is a margin-defence argument: aluminium is “100% pass-through,” so when metal prices jumped ~16% this year the optical margin fell even though the value-added profit was intact — read gross value-addition, not the inflated top line. It’s a fair point, repeated to the edge of over-use.

On the scorecard of dated promises, the picture is mixed and he doesn’t hide it. Delivered: the Gurgaon plant closure, the $100-million stationary-engine order intake, the alloy-wheel plant ramp (broadly). Slipped: the ₹350-crore Gurgaon land sale (unsold a full year on — though he’s deliberately holding out for price), Sunbeam’s promised double-digit margin (pushed to FY27, still single-digit), and the net-debt/EBITDA he kept saying would “keep improving” but which stayed roughly flat at ~2.4. By the last call his stated top worry had shifted from debt to inflationary manpower cost — minimum-wage resets and the new Labour Code — which he says is hard to pass on to customers.

Where It’s Going

The road ahead is the aluminium build-out maturing and the data-centre engine bet ripening. Management guides FY27 to “mid-teens” revenue growth — powertrain double-digit (helped by India’s emergence as a global hub for higher-horsepower trucks, and OEMs outsourcing in-house machining), aluminium high-teens, industrial high-single-digit. The structural tailwinds Ravi leans on are real: China+1 manufacturing migration, rising aluminium content per vehicle, the data-centre power boom, and Western aluminium ancillaries going insolvent and ceding export share. If utilisation climbs and the aluminium entity crosses $1 billion with capex slowing, the operating leverage on this asset base is significant.

The risks are the leverage and the timing. Margins won’t expand near-term because fresh FY27 capex will again depress them; the data-centre revenue is 3+ years out; Sunbeam’s turnaround keeps slipping; the CV cycle has been muted; and free cash flow stays negative while the build-out runs. The whole thesis rests on the new capacity earning its 20% hurdle — if utilisation and margins climb, the debt flatters returns; if they stall, the ₹309 crore of interest and ₹444 crore of depreciation bite hard. FY26’s profit recovery is encouraging, but FY25’s air-pocket showed how quickly the leverage can turn against reported earnings when capacity and interest land before the volumes mature.

The Four Checks

Quality & moat (gate). Qualified pass. The edge is genuine — precision-machining and engineering depth (Ravi insists “95% is machined,” not low-value raw casting), sticky qualification-heavy OEM relationships, an in-house storage product, and a rare validated position in stationary-engine blocks (3 of the top 4 makers). But the FY23→FY25 margin slide (powertrain EBIT 25%→15%) shows the moat isn’t pricing-power-proof; the edge is being partly spent to buy scale and reduce cyclicality rather than to defend returns. Good engineering franchise, moderate and softening moat.
Returns on incremental capital & runway. Currently weak, deliberately so. ROCE has halved to the low-teens as enormous capex and acquisitions landed ahead of the volumes; recent incremental capital has generated negative free cash flow. The bet is that maturing aluminium capacity and the data-centre franchise lift returns back toward the 20% hurdle. The runway (aluminium scale, data-centre power, China+1) is real and large; the current return on the capital being poured in is poor and unproven.
Capital allocation for the stage. Aggressive, coherent, and honestly communicated — but high-risk. Building capacity ahead of demand to bid for large multinational orders is a defensible strategy if the orders come, and the 20%-ROCE hurdle and net-debt discipline are the right guardrails. Ravi’s transparency about consciously choosing growth over deleveraging is a credit. The marks against: debt near-tripled, capex guidance kept creeping up, deleveraging promises went unmet, and a distressed acquisition (Sunbeam) is still a drag. This is a founder making a big, conviction-led bet with shareholders’ balance sheet.
Price. Very demanding. 55× earnings and 6.7× book for a capital-heavy auto-components manufacturer earning ~14% ROCE with negative free cash flow and rising leverage prices in the transformation succeeding handsomely. The valuation assumes the aluminium build-out, the data-centre franchise, and margin normalisation all come through. Even granting the growth, this is a premium multiple for a business mid-bet with the cash flow still going out the door.

Sources

Concall transcripts read: Q1 FY26 (30 Jul 2025), Q2 FY26 (10 Nov 2025), Q3 FY26 (29 Jan 2026), Q4/FY26 (8 May 2026) — all run solo by CMD Srinivasan Ravi.
Annual reports: FY23 (richest — chairman’s statement + engagement roster), FY24 (capacity/greenfield strategy + DR Axion), FY25 (Sunbeam + Fronberg acquisitions, margin/PBT deterioration). Risk and segment-definition sections were largely boilerplate; financials grounded in the segment tables.
Snapshot: screener.in consolidated, fetched 2026-06-09 (logged-out).
Research dumps: vault/Sources/Earnings/Craftsman Automation Ltd/.