Larsen & Toubro — four quarters of “fifteen percent, that is baked in”

The State of Play

Larsen & Toubro closed FY26 with ₹2,85,874 crore in revenue — up 12% in a year it had guided 15% all the way through January — and a record order book of ₹7.4 trillion, up 28%. The miss has a specific address: a war in West Asia that froze supply chains in the quarter that was supposed to deliver the catch-up, plus a domestic water business that spent the whole year waiting for government cheques. Around that one soft number, almost everything else overshot: order inflows grew far past the 10% target, working capital collapsed to levels management never promised, and the company closed out its five-year Lakshya 2026 plan, sold off its last two legacy assets, and announced a new five-year plan that quietly lowers the return target to pay for bets on data centres, green hydrogen, and chips.

The Company

L&T is India’s largest engineering and construction house — per screener’s description, “primarily engaged in providing engineering, procurement and construction (EPC) solutions” across infrastructure, hydrocarbon, power, defence, IT and financial services. Think of it as a company that builds things too big for anyone else in India to build — metros, refineries, high-speed rail, missile systems — with an IT services arm (LTIMindtree, LTTS) and a lender (L&T Finance) bolted on. Infrastructure alone was 47% of FY26 revenue.

The business model has a quirk worth knowing: a good EPC contractor gets paid before it spends. Customer advances and supplier credit run ahead of its own outlays, which is why L&T’s cash conversion cycle is negative 151 days — customers effectively finance the work.

There is no promoter. The screener shareholding table has no promoter row at all — L&T is that rare large Indian company run by professional management and owned by institutions. The ownership story of the last three years is a quiet changing of the guard: foreign institutions sold down from 25.3% (June 2023) to 18.8% (March 2026), and domestic institutions absorbed nearly all of it, rising from 38% to 43.3%. The market values the company at ₹5.44 lakh crore, about 33 times earnings and — screener’s one auto-generated complaint — nearly 5 times book value.

The decade view: revenue has roughly tripled (₹91,929 crore in FY15 to ₹2,85,874 crore in FY26) while operating margin drifted from 17% to 12%. More volume, thinner spread — the deliberate trade of a contractor that decided growth was worth buying.

The Story So Far

The four calls of FY26 are best read as one long argument between a confident guidance line and an increasingly eventful world. The annual reports supply the run-up: FY24 was the year the Middle East took over the order book (Energy segment international orders jumped to 87% of inflow, “multiple ultra-mega international orders”), and FY25 was the year of deliberate diversification beyond it — infrastructure’s international order share leapt from 38% to 61%, including an airport-and-data-centre order in a CIS country, while the power business got relabelled “CarbonLite Solutions” and renewables was carved out as its own vertical. By the time FY26 opened, half of L&T’s work was outside India, and 82% of the international order book sat in one region: the Middle East. Keep that concentration in mind. It becomes the plot.

July 2025 — Q1: everything on plan, guidance unchanged

The year opened “on a strong note,” in IR head P. Ramakrishnan’s words. Order inflows of ₹945 billion were up 33%, the projects-and-manufacturing order book crossed ₹6 trillion for the first time, and profit rose 30% to ₹36 billion. The guidance was stated like a formula:

“Our guidance for FY26 remains unchanged. We expect our group order inflows and group revenues to grow at 10% and 15% respectively for FY26… With respect to the margin in the Projects and Manufacturing portfolio, we continue to target the 8.3% to 8.5% range for the full year.” — P. Ramakrishnan, Q1 FY26 call

Two soft spots were flagged early, and both would run all year. First, hydrocarbon margins: a set of competitively priced jobs won in FY21–22 had entered “peak execution,” dragging the Energy segment’s margin from 8.7% to 7.3%. Management insisted this was “along budgeted lines.” Second, water: the Jal Jeevan Mission pipeline had a funding drought, slowing execution and stretching working capital — excluding water, the infrastructure margin and working capital both looked meaningfully healthier.

The prospect pipeline was the eye-catching number: ₹14.8 trillion of identifiable opportunities for the remaining nine months, up 63% on the year, most of it international hydrocarbon work. An analyst from Goldman flagged that corporate treasury income had quietly tripled; Kotak’s analyst asked why working-capital guidance was still 12% when the actual number was already 10.1%. Management said it would revisit “at Q2 or Q3” — file that one away.

October 2025 — Q2: the bull quarter

This was the most confident call of the year, with deputy MD Subramanian Sarma joining to make the case. Order inflows hit ₹1,158 billion, up 45%. The 10% inflow guidance was upgraded to “exceeding,” and everything else was reaffirmed — 15% revenue (“we maintain”), 8.5% margin (“reasonably confident”), 12% working capital.

The thesis Sarma laid out was a multi-year one: L&T had crossed a size threshold where it could win $3–4 billion jobs that few competitors could even bid, and the Gulf’s capex cycle was structural, not an oil-price trade:

“This run will continue for some more time, at least for next 2, 3, 4 years.” — Subramanian Sarma, Q2 FY26 call

Two new storylines opened. Thermal power had “suddenly become very buoyant” — renewables couldn’t supply round-the-clock power and data centres were demanding it, so L&T had picked up ~13.5 GW of work and was “gearing up for another 10–15 GW over the next 2–3 years.” A decade after coal was left for dead, the AI boom was reviving it. And in Kuwait, L&T was lowest bidder on three of five tenders worth ~$4.5 billion, pending the customer finding extra budget — “should get done by this quarter or maybe latest by next quarter.”

The quarter’s biggest structural news was an exit: an in-principle deal for the Government of Telangana to take over Hyderabad Metro — L&T’s most famous problem child, carrying ~₹13,000 crore of debt against ~₹7,000 crore invested. L&T would get about ₹2,000 crore for its equity and took the impairment on its standalone books that quarter. After years of fare-hike arithmetic and ridership-breakeven math on every call, management was visibly relieved to walk analysts through the exit accounting instead.

The one darkening note: hydrocarbon margins fell further (Energy segment at 7.3%, against 8.9% a year before), now attributed to cost overruns on legacy projects “running even during the pre-COVID time.” Soft margins would “persist in the near term” — but were “baked” into the 8.5% guidance.

January 2026 — Q3: records, a provision, and a crack in Kuwait

Q3 produced the year’s highest-ever quarterly order inflow — ₹1.36 trillion — taking nine-month inflow growth to 30% against a full-year target of 10%. Working capital, guided at 12%, was now at 8.2%; management finally moved the official target to “around 10%.” Recurring profit rose 31%.

But the quarter carried three asterisks. Reported profit actually fell 4%, because new Labour Codes regulation forced a one-time ₹11.9 billion provision (the screener quarterly table shows the dent: net profit of ₹3,825 crore in the December quarter, the year’s lowest, with other income swinging negative). The Energy segment’s margin slid to 5.9% — the legacy hydrocarbon projects were now in “terminal execution,” with recovery promised “2 or 3 quarters from now.” And Kuwait cracked: the tenders where L&T was L1 were cancelled outright, the client’s budget having come in below everyone’s bids. Sarma’s defence was characteristic:

“These projects cannot be cancelled because these are strategically important… they will come back.” — Subramanian Sarma, Q3 FY26 call

He had a point of order, too — the Kuwait wins were never booked, so nothing came out of the order book. On the revenue guidance, with nine-month growth at 12% against the 15% target, ICICI Securities’ Mohit Kumar pressed directly. The answer left no wiggle room: “we continue to retain our guidance of 15% for the full year… That is baked in.”

Meanwhile the domestic order book was quietly transforming: private-sector share had jumped from 21% in March 2025 to 36% — data centres, semiconductor fabs, real estate, thermal power. The state was no longer the only customer that mattered.

May 2026 — Q4: the miss, explained, and a new five-year plan

The full year came in at 12% revenue growth against the 15% guided. Management quantified the gap precisely: about ₹50 billion of revenue slipped in Q4, across Middle East power transmission, Middle East renewables — both hit by the West Asia conflict’s supply-chain freeze — and the water business, patchy all year. Container freight had spiked to $8,000 a box during the disruption. The P&M margin, tracking 30 bps of improvement through nine months, gave most of it back in Q4. ROE closed at 15.5% against the Lakshya 2026 target of 18% — 110 bps of that gap being the Labour Codes provision, the rest the revenue slip and the metro exit sliding into FY27.

Everything else about the year overshot. Order inflows “significantly surpassed” guidance — roughly ₹4.4 trillion for the year. The order book reached ₹7.4 trillion. Working capital ended at 4.1% of revenue — against a target of 12% set a year earlier — flattered by customer advances, with management honest that ~10% is the normal level. Operating cash flow ex-financial-services hit ₹171 billion in Q4 alone.

On the conflict, Sarma refused the bearish frame entirely: no project cancelled, all sites functioning, payments on schedule, and reconstruction ahead —

“More opportunities than risk.” — Subramanian Sarma, Q4 FY26 call

— with the UAE announcing ~$55 billion of capital outlay and Sarma’s own estimate of $30–50 billion of reconstruction work over three years. The biggest risk, he said, was supply chain, not inflation, “but it is continuously getting better.”

The call’s real substance was the new plan. Lakshya 2031 (FY27–31) targets 10–12% order inflow CAGR, 12–15% revenue CAGR, and — the number analysts circled — an ROE target of 16–17%, down from the prior plan’s 18%. HSBC’s Puneet Gulati asked why directly. The answer: the new plan front-loads roughly ₹400+ billion of capex into businesses that won’t pay back inside the window — ~₹150 billion for green hydrogen, ~₹100 billion for data centres (a 200 MW ambition, with an Nvidia MoU for AI-ready capacity), ~₹50 billion for industrial electronics, ~₹30 billion for semiconductor design IP, ~₹50 billion for a hydrocarbon fabrication yard and shipbuilding upgrade. The core business, management agreed, should keep earning “18–20%+” on its own.

The closing plan got its scorecard: Lakshya 2026 delivered 20% order-inflow CAGR against a planned 14%, 16% revenue CAGR against 15%, and ROE from ~10% to 16.6% (ex-provision). The legacy cleanup completed on schedule — Nabha Power signed away to Torrent, the Hyderabad Metro SPA executed April 29, 2026, both closing in Q1 FY27, leaving “no residual legacy assets in the portfolio.” And a quiet succession note: Ramakrishnan was elevated to CFO, with R. Shankar Raman — the long-serving finance hand — staying on “two more years.”

FY27 guidance, when it came, wore the year’s lesson visibly: 10–12% revenue growth with a soft first half, margins merely “stable,” and an explicit caveat that the numbers assume “broad normalcy after Q1” — to be revisited in July.

The promise ledger, settled

Against the four-quarter record: order inflow — guided 10%, delivered far above (kept, with room). Revenue — guided 15%, reaffirmed in October and January, delivered 12% (missed; cause external but the January reaffirmation, with nine-month growth already at 12%, left no cushion for a bad quarter). P&M margin — guided 8.3–8.5%, pulled back by Q4 to roughly flat (missed narrowly). Working capital — guided 12%, delivered 4.1% (kept, embarrassingly so). Hydrocarbon margin recovery — promised “2–3 quarters” in October, again in January, and in May the legacy jobs were finally “handed over” with “no further cost creeps,” but FY27 margin guidance of merely “stable” concedes the recovery moved right. Hyderabad Metro exit by end-FY26 — SPA signed April 29, 2026, closing by June 30 (kept in substance, a quarter late in form). Kuwait re-tenders — still pending, one of three surviving.

Where Things Stand

L&T enters FY27 with a ₹7.4 trillion order book that is 92% infrastructure and energy, split evenly between India and abroad — with the Middle East alone at roughly ₹3 trillion, about 40% of everything. That concentration is now officially a two-sided coin: the FY26 revenue miss was the cost side; the reconstruction and Gulf-capex pipeline is the claimed opportunity side. Domestic demand has rotated visibly toward private capital — 39% of the domestic book by March 2026, from 21% a year earlier — riding data centres, fabs, thermal power and real estate.

Margins remain the stubborn variable. The P&M margin has sat in a 7.7–7.8% band for three years; the legacy hydrocarbon drag is declared over, but management would not commit to a higher number for FY27, only “stable,” with any customer cost-recoveries treated as unbooked upside. The growth engine is shifting from execution speed to capital deployment: Lakshya 2031’s heavy front-loaded capex into green hydrogen, data centres, electronics and chip design is explicitly expected to suppress group ROE to 16–17% while the core earns more — a conglomerate consciously paying current returns for future optionality. The next checkpoint is the July call, where the FY27 guidance gets its promised post-conflict review.

The Four Checks

1. Quality and moat. A genuinely good business with a real moat, of the unglamorous kind: scale and prequalification. L&T is the only Indian contractor that can bid — let alone execute — $3–4 billion jobs, a threshold management itself says few global competitors clear, and decades of delivered metros, refineries and missile systems are the entry ticket to every shortlist. The negative working capital (cash conversion cycle of minus 151 days — customers fund the work) is the financial signature of that trust. What the moat does not buy is pricing power: everything is won by competitive bid, the P&M margin has sat at 7.7–7.8% for three years, and Kuwait cancelled three tenders where L&T was lowest bidder without ceremony. Call it durable dominance of a niche where the niche is “anything too big for anyone else in India,” with margins that stay thin precisely because the work is contested.

2. Returns on incremental capital and runway. Improving, from a mediocre base to merely decent. ROCE spent most of the decade at 10–13% and has climbed to 15% in FY25 and FY26; ROE stands at 15.5% (16.6% excluding the one-time Labour Codes provision), up from roughly 10% when the Lakshya 2026 plan began. Management says P&M return ratios “more than doubled” between FY22 and FY26, and that the core business ex-new-ventures should earn 18–20%+. The runway is not in question — a ₹7.4 trillion order book, a ₹17.8 trillion prospect pipeline, and Gulf plus domestic private capex both running hot. The catch is where the next rupee goes: the Lakshya 2031 plan front-loads ₹400+ billion into green hydrogen, data centres (a “speculative” 13–14% target return, in the CFO’s own word), and chip design — bets explicitly expected to drag group ROE to 16–17%. Mid-teens returns on a very long runway, with the incremental rupee aimed at things that may earn less before they earn more.

3. Capital allocation for the stage. Disciplined on the cleanup, deliberate on the gamble. The legacy divestment programme actually finished: Hydel, IDPL, Nabha Power and the Hyderabad Metro — the famous ₹13,000-crore problem child — all signed away, “no residual legacy assets in the portfolio.” The prior plan included a buyback, dividends have run at a steady ~31–36% payout, and the outgoing CFO’s stated posture — “just because capital is available, we don’t want to just scatter it around” — has so far matched behaviour. The quibble is the new plan itself: ₹400+ billion of front-loaded capex into ventures that won’t pay back inside the five-year window, with the ROE target cut from 18% to 16–17% to accommodate them. That is a conscious, disclosed, partner-financed-where-possible bet rather than empire-building — but it is still mostly unproven, and the give-back of cash takes a back seat while it runs.

4. Price. Demanding. As of the June 2026 snapshot (fetched June 5; the June 10 refresh resolved the wrong ticker and is disregarded), the stock trades at ₹3,953 — a ₹5.44 lakh crore market cap, 33.2 times earnings, just under 5 times book value, and a 0.96% dividend yield. Set that against the economics: 15.9% ROE, FY26 revenue growth of 12% against a 15% guide, FY27 guidance of 10–12% with a soft first half, and margins promised only “stable.” Five times book for a sub-16% ROE business is screener’s own auto-generated complaint, and it is arithmetically fair — the multiple prices in the order book converting smoothly, the Middle East staying open for business, and the new-economy capex eventually earning its keep. Nothing here is cheap; the price asks the next five-year plan to go roughly as well as the last one did.

Sources

• Concall transcripts (4): July 29, 2025 (Q1 FY26); October 29, 2025 (Q2); January 28, 2026 (Q3); May 5, 2026 (Q4 + Lakshya 2031) — via screener.in/BSE filings
• Annual reports (3): FY24, FY25, FY26 — trimmed high-signal sections (MD&A, strategy, risks, segments); chairman’s letters did not survive the extraction in any year
• Screener.in consolidated snapshot: fetched 2026-06-05 (logged-out session; document list was nonetheless current through the May 2026 concall)
• Research dumps and per-document digests: vault/Sources/Earnings/Larsen & Toubro Ltd/ (not published)