Bharat Electronics — the PSU that earns like a private champion, waiting on one big order

The Pulse

Bharat Electronics is the rarest of things: a government-owned company that earns like a private-sector champion. The defence-electronics PSU posted record FY26 numbers — revenue up 16% to ₹27,480 crore, profit ₹6,048 crore, operating margins around 30%, and a 37% return on capital with effectively zero debt. Those returns would be enviable for any private firm and are almost unheard of for a state enterprise. The business is funded by its own customers’ advances, sits on ₹23,000 crore of reserves, and has widened margins structurally over a decade as it climbed from making components to integrating whole systems. The two things to watch are both about timing, not quality: an order book that stayed flat all year (because new orders roughly matched revenue), and a marquee ₹30,000-crore missile contract (QRSAM) that management promised “by March” and which slipped past year-end. At ~50× earnings, the market pays a premium for the franchise and the defence-capex tailwind — pricing the optionality, not the trailing numbers.

The Business

BEL is a build-to-order systems integrator for the Indian armed forces — it designs and builds large, complex electronic systems (radars, fire-control, electronic warfare, communications, network-centric C4I, increasingly weapon systems and seekers) on long government contracts, and is also the firm that makes India’s electronic voting machines. It runs 29 strategic business units, four of them recently created in fast-growing domains: cyber security, unmanned systems, seekers, and arms & ammunition. The promoter is the Government of India at a steady 51%.

What’s genuinely distinctive — and unusual for a PSU — is the quality of the economics. A 37% ROCE and 28% ROE on a near-debt-free balance sheet is franchise-grade, the product of a near-monopoly position in indigenous defence electronics, pricing discipline, and a working-capital base funded largely by customer advances rather than BEL’s own capital. Margins have widened from 17% a decade ago to a stable ~30%, which is the opposite of what competitive tendering usually does to a contract manufacturer — a sign of real pricing power and a richer, more system-level product mix as the new SBUs climb the value chain. The one blemish is cash timing: debtor days crept to 170 and the working-capital cycle ballooned to 342 days in FY26, dragging operating cash conversion well below reported profit (management said a year-end receivables spike was collected in April–May, but it’s worth watching).

The strategic shift underway is moving up the value chain — from supplying electronic modules to integrating entire platforms. BEL is now a lead on the AMCA fifth-generation fighter (in a consortium with L&T), is set to take ~50–60% of the electronics on the ₹90,000-crore P-75I submarine programme, and is the system integrator for ~70% of the QRSAM air-defence system. Management is candid about the trade-off: integrator roles carry lower BEL value-add (10–15%) than ground-up subsystem work (40–45%), so bigger absolute scale comes at thinner margins on that slice.

How Management Thinks

CMD Manoj Jain (appointed mid-2024) runs the calls — voluble, confident, repetitive, and notable for one consistent habit: under-promising on margin and over-delivering. All year, with reported EBITDA running ~30%, he held guidance at “27%” and deflected every analyst who asked why, calling it “a complex product-mix equation… a big AI algorithm decides margins.” Only at year-end did he lift FY27 guidance to “>28%,” tacitly conceding the structural gains were real. The honest subtext, surfaced in a pointed exchange: high public margins are awkward for a company that sells to the government on nomination pricing — and Jain’s revealing answer was that the MoD deliberately lets the indigenisation profit flow back to fund BEL’s R&D and capex, and “should not cut the root itself.” That’s a genuinely useful window into why the returns are durable: the government is a patient, profit-tolerant customer by design.

On candour, the pattern is good on operations, guarded on economics. He openly attributed a soft Q1 to ~₹200 crore of Israel-Iran-conflict component delays, openly flagged semiconductor dependence (~17–19% of material cost, all imported), and openly admitted the slippages on QRSAM, the next-gen corvette, and the Shatrughat EW programme. He was less forthcoming on QRSAM’s economics, the weak cash conversion (which he simply declined to engage when an analyst pressed the ~19% operating-cash-to-EBITDA ratio), and the capital-allocation framework (“we generally don’t disclose”). The clearest credibility check: BEL met or beat every FY26 guidance parameter — revenue >15% (delivered 16%), margin (beat), order inflow (₹30,045 crore, beat the ex-QRSAM target) — except the QRSAM order itself.

Capital allocation runs along three rails: heavy in-house R&D (6–7% of sales, rising toward ₹2,200 crore), JVs to bridge specific technology gaps with global OEMs (Thales, IAI, SAFRAN — the Thales JV is now a real ₹118-crore business, not a paper one), and shared national testing infrastructure. The dividend payout has run a steady ~34%. The bias is partnership-leveraged growth over pure organic capex.

Where It’s Going

The demand model management repeats is a steady base of mid-size orders (a list of “30-plus items each above ₹1,000 crore”) plus one big-ticket programme (>₹20,000 crore) every three to four years. The current big-ticket is QRSAM (~₹30,000 crore), repeatedly promised for FY26 and now expected “before June-end,” with FY27 inflow guidance of >₹55,000 crore built around it landing. The next is Project Kusha, the indigenous S-400 equivalent, around 2028-29. Beyond that sits a deep pipeline — the corvette programme, P-75I submarine electronics, Akash-NG, directed-energy weapons with DRDO, and a fast-growing non-defence push (railway KAVACH signalling, airports, and an ambitious fully-indigenous secure data-centre stack built with C-DAC, targeted at ₹2,000–10,000 crore for government customers).

The risks are real but bounded. The flat order book is the central investor worry — at some point inflow has to outpace revenue for the 15% growth to continue, and management’s answer is the big-ticket-every-few-years model plus the ~₹6 lakh crore of defence approvals it claims a ~₹50,000-crore-plus share of (much of it routed indirectly through shipyards and integrators). Other risks: persistent order slippage (QRSAM, corvette, Shatrughat all drifted right), the weak cash conversion, semiconductor import dependence, a modest wage revision from January 2027, and — most strategically — the candid admission in BEL’s own risk disclosure that the same Atmanirbhar policy feeding it is inviting private-sector competition that “may impact our market position in certain segments.” The moat rests on policy preference and incumbency, not an unassailable cost or technology lead.

The Four Checks

Quality & moat (gate). Clear pass — the best business in this batch. A near-monopoly incumbent in indigenous defence electronics, with 133 of its items on government positive-indigenisation lists that effectively embargo imports and channel demand to BEL, a tri-service/DRDO relationship that converts into nomination orders (90% of orders are nominated, not competed), and franchise-grade returns to prove the moat is real. The honest caveat, which management itself flags, is that the moat is policy-anchored and private competition is being deliberately invited in — durable, but not unassailable.
Returns on incremental capital & runway. Excellent and the standout strength. ROCE has climbed almost monotonically to ~37%, the business is funded by customer advances rather than its own capital (so incremental capital needs are low), and the runway — India’s multi-year defence-capex super-cycle, indigenisation, exports, and a non-defence diversification — is long and open. The one watch-item is the move up to system integration, which is lower-margin per rupee of revenue, so blended returns may ease as that mix grows.
Capital allocation for the stage. Rational and shareholder-friendly. Heavy reinvestment into R&D and capacity while returns are exceptional is exactly right, the JV route to access foreign technology is sensible, and a steady ~34% dividend returns cash without starving growth. The marks against are transparency, not substance: management explicitly declines to share the allocation framework, and sidestepped the weak cash-conversion question. For a 37%-ROCE business, retaining and reinvesting is clearly the right call.
Price. Demanding. ~50× earnings and 12.6× book is a rich multiple even for a business this good — it prices in the defence super-cycle, the QRSAM/Kusha big-ticket pipeline, export and data-centre optionality, and continued margin strength, with little cushion if order slippage persists or growth normalises. This is the rare case where the business clearly merits a premium and the question is purely whether this premium leaves any margin of safety; on the trailing numbers, not much.

Sources

Concall transcripts read: Q1 FY26 (30 Jul 2025), Q2 FY26 (31 Oct 2025), Q3 FY26 (28 Jan 2026), Q4/FY26 (20 May 2026) — all with CMD Manoj Jain and CFO Damodar Bhattad.
Annual reports: FY23 (richest — indigenisation/cost narrative, FX risk), FY24 (diversification/SBU and order-mix headlines), FY25 (JV maturation — Thales JV numbers, IAI JV incorporated, SAFRAN MoU). Note: the Chairman’s/MD’s letters were not recoverable in any year (Hindi-font corruption); thesis evolution inferred from MD&A. Segment P&L is exempted for defence PSUs, so only headline mix is available.
Snapshot: screener.in consolidated, fetched 2026-06-09 (logged-out).
Note on the fetch: an initial run on the ticker “bel” resolved to the wrong company (Belrise Industries) and was discarded; this digest is the correct Bharat Electronics.
Research dumps: vault/Sources/Earnings/Bharat Electronics Ltd/.