Data Patterns — the contract that’s always one to two months away

The State of Play

Data Patterns closed FY26 with revenue of ₹925 crore, up 31% — beating its own 20–25% guidance — at a 40% EBITDA margin that is, as a Goldman Sachs analyst put it on the February call, close to world-leading for a radar-electronics company. The quarterly path to that number was a rollercoaster (₹99 crore, then ₹308, then ₹173, then ₹345), and the order story ran on a peculiar clock: a bucket of “negotiated but not yet signed” contracts, around ₹1,100 crore, that was promised to convert “in the next 1 to 2 months” in November, again in February, and once more in May. The order book stands at ₹926 crore signed, ₹2,062 crore including that bucket, with a further ₹1,900 crore of single-tender repeat business expected during FY27.

The Company

A 35-year-old Chennai defence-electronics house, Data Patterns’ differentiator is vertical integration with in-house IP: it designs and builds its own radars, electronic-warfare systems, avionics, communication systems, satellite electronics and test equipment, rather than assembling imported sub-systems. Its hardware flies on the Tejas LCA (cockpit displays), the Light Utility Helicopter, and most importantly BrahMos — fire-control systems, launchers and test systems since 2006, now extended into a five-year comprehensive AMC and a missile seeker moving from development into production. Founder S. Rangarajan remains Chairman & MD; promoter holding has sat at exactly 42.41% for twelve straight quarters — no creep, no dilution — while DIIs accumulated over the past year and FIIs trimmed from their 2024 peak.

The economics are distinctive and management is frank about why: products are designed in-house down to the building blocks, development costs are written off in the year incurred (no capitalising), and the company refuses low-margin L1 tenders and buy-and-integrate work. The result is ~60–80% gross margins on product quarters — and a brutal working-capital cycle (365–428 days) because government development contracts pay slowly. The company is debt-free and net cash, runs ~1,200 engineers, and raised ₹500 crore via QIP to self-fund two big bets: a flight-control radar and a complete modern EW suite, against a stated ₹15,000–20,000 crore addressable market. The long arc is visible in the annual numbers: a ₹131 crore company in FY19, ₹925 crore in FY26, with operating margins that moved from the high teens to 40% once its own IP started shipping at scale.

The Story So Far

Q1 FY26 (call: August 8, 2025) — a stumble, and a ₹30,000 crore daydream

The year opened badly: revenue fell 4.6% to ₹99 crore because a customer’s inspection team showed up late — about ₹27 crore of finished production couldn’t be billed without acceptance. Gross margin actually rose to ~80%; the 32.3% EBITDA margin (below the guided 35–40%) was purely the fixed-cost base against a small quarter. Rangarajan’s framing of his own priorities was characteristically blunt: “My more worry is on the bottom line. We wanted to retain 20%, 25% bottom line growth.” The full-year guidance — 20–25% revenue growth, 35–40% EBITDA margin — was reaffirmed, with ₹1,000 crore-plus of orders expected within 6–8 months and the BrahMos seeker contract “in the next month or so.” The call’s expansive closing — scaling from a ₹700 crore company toward ₹5,000 crore in 4–5 years, with a ₹25,000 crore-plus addressable market across EW suites and Su-30 radars — set the tone for a year in which the long-term numbers were always more vivid than the near-term ones.

Q2 FY26 (call: November 13, 2025) — the strategic contract

Then revenue exploded: ₹308 crore, up 238% — but margins collapsed to 22% EBITDA, because ₹180 crore of the quarter was a deliberately low-priced “strategic” DRDO contract, the company’s first large electromechanical-systems job (power systems, construction, ~140 tons of movable payload — capability-building far outside its electronics comfort zone). Ex that contract, management confirmed margins sat inside the 35–40% band. The order machinery entered its holding pattern here: ~₹550 crore of negotiated orders “in the next 2, 3 months,” FY26 inflows to cross ₹1,500 crore (“Yes… we expect more than that”). The BrahMos seeker had advanced one notch: “negotiations completed; contract awaited.” Working capital stood at 343 days, with a medium-term path to 270 sketched out. A new thread: an AMCA fighter consortium with Bharat Forge and BEML, and the first fully self-developed radar export — a transportable precision-approach radar accepted by a European customer.

Q3 FY26 (call: February 6, 2026) — record book, repeating clock

The December quarter normalised: ₹173 crore (+48%) at a 44% EBITDA margin, and the order book hit an all-time high of ₹1,868 crore. But the negotiated bucket had grown rather than converted — now ~₹1,100 crore, with another ₹500–600 crore behind it, and the timeline restated:

“I think these contracts should happen in the next 1 to 2 months’ time because these are only document work… It should happen this quarter.” — S. Rangarajan, Q3 FY26 call

The seeker advanced again — trials complete, a development contract in hand, production-testing equipment ordered, series production “starting FY26-27.” AMCA moved to a real milestone: Data Patterns (with Bharat Forge) shortlisted as one of three to receive the RFP. The uncomfortable disclosure: cash on hand at its lowest in 8–9 quarters, with receivables stretched by development contracts in final acceptance — all of it, management said, collectible “in a maximum of 4 months.”

Q4 FY26 (call: May 15, 2026) — the year ends well; the bucket persists

The fourth quarter nearly doubled sequentially to ₹345 crore (though down 13% on a freak prior-year Q4 that had been 55% of FY25), at extraordinary margins — 73% gross, 56% EBITDA — because everything shipped was in-house IP with “no bought-out.” The full year: ₹925 crore (+31%), EBITDA ₹371 crore at 40%, PAT ₹271 crore (+22%) — revenue guidance beaten, margin and PAT guidance met. Order inflows tripled to ₹1,121 crore — a genuinely strong year, but short of the ₹1,500 crore the CMD had endorsed in November, because the negotiated bucket (still ~₹1,090 crore in May) never signed inside Q4 as promised in February. The timeline, once more: “in the next 1 to 2 months’ time… these are all government customers. So, I can’t predict for them.” The forward stack now reads: that ₹1,090 crore, plus ₹1,900 crore of single-tender repeat orders (including BrahMos seeker production, expected in ~4–5 months) during FY27, plus Tejas Mk1A avionics via HAL negotiation, plus the imminent AMCA RFP. Cash conversion improved from 428 to 365 days; guidance is 320–340. FY27 guidance: 20–25% growth, 38–40% margins — with management openly expecting to do “substantially higher.”

The ledger: said vs. delivered

Kept or beaten: FY26 revenue (+31% vs 20–25% guided); EBITDA margin (40%, top of band); PAT growth (+22%, in band); the Q2-onwards execution pickup promised in August; H2 margin recovery exactly as promised in November; working-capital improvement (428 → 365 days); debt-free status; the Q1 slippage (~₹27 crore) billed as promised. Slipped, repeatedly: order conversion — ₹550 crore “in 2–3 months” (November) grew into ₹1,100 crore “in 1–2 months” (February), promised to land “this quarter,” and was still ~₹1,090 crore pending in May with the same 1–2 month tag; FY26 inflows of ₹1,121 crore versus the ~₹1,500 crore endorsed mid-year. Rolling forward: the BrahMos seeker contract — “next month or so” in August became a development contract by February and a production order now due “in 4–5 months.” Management’s consistent defence is also consistently true: every delay sits with government customers’ paperwork, not with execution — deliveries, when inspections happen, take “2–3 months.” The pattern to internalise: this company’s P&L promises are reliable; its order-timing promises are weather forecasts.

Where Things Stand

FY27 opens with the most loaded pipeline in the company’s history: ~₹1,090 crore negotiated, ₹1,900 crore of expected single-tender repeats, seeker series production due to start, an EW tender (against German competition) with negotiations completed, Tejas Mk1A avionics in HAL’s negotiation cycle, the AMCA RFP “any time now,” and global OEM contracts teased for 2–4 months out. Capacity is being built ahead — nine floors of new factory space, 150 more engineers — explicitly to allow a jump to a “multiple thousand crore company rather than scale it 20% year-on-year.” The structural tensions are unchanged: extreme quarter-to-quarter lumpiness, a working-capital cycle near a full year, dependence on a handful of government programs whose timing nobody controls, and the gap — now familiar — between a record order book and the even larger bucket that is perpetually one to two months from signing.

The Four Checks

1. Quality and moat. A good business with a real moat of the design-in kind. In defence electronics, once your hardware is qualified onto a platform you tend to stay for the platform’s life — Data Patterns has been on BrahMos since 2006 (fire control, launchers, test systems, now a five-year AMC and a seeker heading into production), and on Tejas and the Light Utility Helicopter besides. The ₹1,900 crore of “single-tender repeat” orders expected in FY27 is that moat expressed as a procurement category: nobody else gets asked. Underneath sits in-house IP down to the building blocks, development costs expensed in the year incurred, and 60–80% gross margins on product quarters — margins a commodity assembler cannot print. The erosion risks are equally concrete: one customer (the Indian government) controls the clock entirely, and every new program is contested — BEL above, Astra Microwave moving into complete systems below. Strong moat on platforms already won; a knife fight for each one after.

2. Returns on incremental capital and runway. The snapshot puts ROCE at 23.3% and ROE at 16.9%. The gap between the two is mostly the ₹500 crore QIP sitting in the denominator as treasury cash (still largely undeployed as of November 2025, ~₹122 crore spent); ROCE itself has firmed from ~20% across FY23–FY25 to 23% this year. The underlying machine earns well — 40% EBITDA margin, PAT of ₹271 crore up 22%, debt-free and net cash. What dilutes a rupee retained is the working-capital toll: a cash-conversion cycle of 365 days by management’s count (down from 428, guided to 320–340; screener’s harsher all-in measure, counting inventory, says 499), meaning each rupee of revenue locks up roughly a rupee of working capital for a year. The runway, by contrast, is long and specific — a stated ₹15,000–20,000 crore addressable market in flight-control radar and EW suites, plus AMCA, Tejas Mk1A and exports. Low-twenties incremental returns with a genuinely long road to deploy on — gated by the government’s working-capital toll.

3. Capital allocation for the stage. Judged on actions: the cash-return side is steady and token — a ~20% dividend payout every year from FY21 through FY25 (a 0.17% yield at the current price), no buyback anywhere in the file set. The rest is reinvestment-first behaviour appropriate to the stage — the QIP raised to self-fund two large development bets rather than wait for funded contracts; nine floors of new factory space built ahead of demand, with management explicitly wary of stranded capex; a deliberately low-priced ₹180 crore DRDO contract taken to buy electromechanical capability; development costs expensed, not capitalised into the balance sheet; and promoter holding parked at exactly 42.41% for twelve straight quarters — no creep, no pledging signal, the only dilution the QIP itself. The quibble is the idle cash earning treasury returns while the orders it anticipates remain perpetually “one to two months away.” Rational for the stage, with a hoarded-cash asterisk.

4. Price. As of the June 2026 snapshot: market cap ₹24,613 crore, price ₹4,396 — near the top of a ₹2,131–4,695 yearly range — a P/E of 89.9, 14.2 times book value, and a 0.17% dividend yield. The denominator: ₹271 crore of FY26 profit growing 22%, guidance of 20–25% that management expects to beat “substantially,” a 37.6% five-year profit CAGR behind it. Run the simple arithmetic: at 25% compounding it takes roughly seven years of flawless execution for earnings to grow into a multiple a patient buyer would call ordinary — and the order clock that feeds those earnings has slipped three calls running. Ninety times earnings already prices in the bucket signing, the ₹1,900 crore of repeats landing, seeker production starting, and a good slice of the ₹5,000 crore daydream besides. Even by the standards of Indian defence electronics — never a cheap neighbourhood — this is demanding: the quality is real, but the market has paid for it well in advance.

Sources

• Concall transcripts (4): Q1 FY26 (Aug 8, 2025), Q2 FY26 (Nov 13, 2025), Q3 FY26 (Feb 6, 2026), Q4 FY26 + full-year (May 15, 2026) — BSE filings, converted to markdown.
• Annual reports (3): FY23, FY24, FY25 sections — all three were governance-heavy with little financial narrative; used mainly for sector framing and the single-segment structure.
• Screener.in snapshot: re-fetched 2026-06-10 (standalone page; the earlier consolidated fetch returned empty tables) — full ratios, valuation, quarterly/annual results and shareholding now in the file set, cross-checked against the concall transcripts.
• Research files: vault/Sources/Earnings/Data Patterns (India) Ltd/ — raw transcripts, AR sections, snapshot, per-document digests (not published).