Data Patterns — Three Columns on the Engineer’s Wager

A note on framework and data. This is the first run of the equityanalysis-vk skill end-to-end with the mandatory primary-source research phase. Four parallel research agents worked the company across business, management, price, and news streams. Sources consulted as primary documents: Data Patterns FY24 Annual Report (167-page PDF, read in full); Q4 FY25 (May 2025), Q1 FY26 (Aug 2025), and Q2 FY26 (Nov 2025) earnings call transcripts as filed with BSE; Q3 FY26 investor presentation (29 slides, dated 6 February 2026); ICRA rating rationale dated 24 December 2025 (A+/Stable upgrade); CRISIL rating letter dated 10 June 2025 (also A+/Stable upgrade); screener.in pages for Data Patterns and seven peers fetched 26 April 2026 at 12:02 IST; Trendlyne shareholding-pattern history and SAST disclosures; BSE corporate filings feed for the last twelve months. The full audit trail (with citations) sits in vault/Sources/Stocks/Data Patterns/_business.md, _management.md, _price.md, and _news.md.

Residual gaps after research. The FY25 (2024-25) Annual Report was not yet posted on the company’s IR page as of research date — the FY25 RPT schedule and any governance changes there have not been verified at primary level. The Q3 FY26 full transcript was behind hotlinking protection; coverage relies on the investor presentation plus a third-party summary. Most importantly: the company operates as a single Ind AS 108 segment and does not disclose segment-level EBITDA — meaning the central question of this note (is the recent margin profile structural or cyclical-peak?) cannot be definitively answered from public sources. One additional flag worth user verification: an unverified search-snippet claim that S. Rangarajan has 26.2% of paid-up capital pledged with SBI for ₹103.88 Cr could not be corroborated against BSE encumbrance disclosure; the latest shareholding pattern’s nominal encumbrance figures are inconsistent with that being a live pledge. Worth a direct BSE check before sizing.

Prelude — A radar leaves for Lucknow

In a clean room at the SIPCOT IT Park in Siruseri, twenty-five kilometres south of Chennai, the air is filtered to one-hundred-class purity and the floor is anti-static. A pair of engineers in white coats are running an acceptance test on a bay of equipment that, viewed from above, looks like a very large piece of telecommunications gear. It is in fact a Transportable Precision Approach Radar — a device that lets a pilot land a fighter aircraft on a temporary runway in zero visibility, in a war zone, through dust and rain. Data Patterns designed the original a decade ago for the Indian Air Force, and has since exported a version to a European customer whose name the company will not disclose. The radar’s electronics, written by hand inside this building, are the kind of thing that until five years ago India imported.

A few hundred kilometres north, at India Meteorological Department headquarters, a separate Doppler weather radar — also Data Patterns hardware, this one ordered last quarter at ₹288 crore for thirty-two units — is being commissioned in Lucknow. Yet a third Data Patterns assembly, this one a fire-control system for a BrahMos cruise missile launcher, is being delivered to a service unit in Andhra. A fourth, an electronic-warfare jammer pod, is being qualified for use on a Sukhoi Su-30 MKI undergoing mid-life upgrade. Five product lines, ten thousand-odd subassemblies a year, 1,071 engineers, one factory.

This is a real industrial business. The harder question, and the one this note is about, is what kind of bet you are making when you buy the listed shares of it at a price that has compounded ninety-seven percent in the last three months. The market currently values Data Patterns at ₹22,829 crore — about ninety-two times its trailing earnings, fifteen times its book value, and roughly thirty-two times its revenue. The two largest brokerage targets are below the current quote. The CFO of the company turned operating cash flow negative last year while the headline profit grew twenty-two percent. The promoter has not sold a share since the IPO four and a half years ago, despite the lock expiry. And in the second quarter of this fiscal year, the management deliberately took on a ₹180 crore contract at a margin substantially below their stated guidance band, because — to use their own phrase — “margins are not important. It is important that we build capability.”

Three of those facts are meant to comfort you. The other three are not. Letting the template put them in their proper columns is the work.

Column 1 — Business

Q1. What does this company actually do?

Data Patterns designs, builds, integrates, and tests defence electronics, almost entirely for the Indian government and Indian government-controlled customers. The product range covers five lines: radars (surveillance, weather, coastal, fire-control, including the long-range hardware for BrahMos and the Sukhoi-30 fire-control radar); electronic warfare (COMINT/ELINT receivers across one megahertz to forty gigahertz, jammer pods, radar warning receivers); communication systems (software-defined radios for fighter aircraft, manpack SDRs, airborne radio relays); avionics (cockpit integration for the Tejas LCA, the Intermediate Jet Trainer, the Light Utility Helicopter); and automated test equipment plus a handful of adjacencies in underwater electronics and small satellites. The fifteen-year-old test: they make the brains and the antennas of military equipment — the boxes that see, talk, jam, and aim. Production happens at a single integrated 200,000-square-foot facility in Chennai, with a hundred-class clean room, a radar integration hangar large enough to test a complete antenna assembly, and assembly capacity of about six hundred boards per day.

What it is not: it is not a missile-maker, not an aircraft-builder, not an EPC contractor. It is one of those firms that sits two layers deep in the defence supply chain — building the systems that go into the platforms that go into the field.

Q2. How do they make money, and from whom?

The customer mix in the most recent quarter, Q3 FY26, breaks down like this: Bharat Electronics 31%, ECIL 27%, Ministry of Defence direct 16%, Hindustan Aeronautics 16%, exports 10%, DRDO and others under 1%. The top four customers account for 89% of revenue. Every single named customer is the Government of India or a government-controlled entity, with the small exception of the export tail.

Contracts come in three flavours. Development contracts are long-cycle, two to five years, where Data Patterns absorbs the development cost as a P&L expense and earns reasonable margins on the engineering work. Production contracts, which follow on from successful development as repeat orders, carry higher gross margins because the design has already been paid for. Annual maintenance contracts on installed equipment — currently 30.5% of the closing order book and growing — are recurring service annuities, mostly from BrahMos. Most of these are awarded under Make-II, IDDM, or Make-in-India procurement categories, often as single-vendor contracts where Data Patterns has earned a sole-source position by previously building the prototype.

There are two structural facts that load through everything else in this column. The first: revenue is back-end loaded — ICRA’s December 2025 rationale notes that thirty to forty percent of revenue typically books in the fourth fiscal quarter, which inflates working capital at year-end. The second: the working capital cycle itself is structurally heavy. Inventory days were 173 at end-September 2025; net working capital to operating income was 105% as of September 2025 and 103% as of March 2025. ICRA expects this to remain at 90–100% in the medium term. To put that in context, Bharat Electronics — the dominant DPSU comparable — runs working capital at roughly 85 days. Data Patterns runs at 372.

Q3. Am I pretending to understand this, or do I really?

This is where the template earns its money. After completed primary-source research, the conviction breaks at one specific point.

I understand the product physics — radar engineering, EW signal processing, the difference between a development contract and a production order, the architecture of an integrated avionics suite. I understand the customer mechanics — how MoD’s procurement gates work, why payment cycles align with acceptance trials, how Make-II awards become single-vendor production orders. I understand the macro tailwind — India is currently importing 70–80% of its defence-electronics requirements, and the indigenisation push is structural, not cyclical.

What I do not fully understand — and the research did not resolve — is the structural margin question. Data Patterns reported FY25 EBITDA margin of 38.8% and a Q3 FY26 print of 44.8%. The same company reported H1 FY26 at 24.7% and 9M FY26 at 30.7%. Management guidance is 35–40%; ICRA’s own forward model is 30–35% for FY26 and 35–40% for FY27 — below management’s near-term target. Gross margin volatility tells a sharper story still: 77.4% in Q3 FY26 (mostly in-house IP work) versus 38.5% in Q2 FY26 (a single ₹180 crore strategic contract dragging the mix). The company does not disclose segment-level EBITDA — Indian Accounting Standard 108 only forces segment reporting when there are reportable segments, and Data Patterns operates, formally, as a single business. So whether the historical 38–43% blended margin is supported by a high-margin radar/EW core (with the strategic low-margin contracts as a tactical exception) or whether the move from subsystem supplier to full-system integrator is structurally pulling the steady-state margin down to 30–35% — that is a question the data does not answer cleanly.

If I were sitting across from S. Rangarajan, the three sharpest questions I would ask: what fraction of FY27 revenue is expected to come from in-house IP-heavy production versus integrated full-system contracts with bought-out content? On a normalised mix-adjusted basis, what is the company’s own internal forecast for FY28 EBITDA margin? And — most directly — would you be comfortable disclosing segment-level EBITDA, even informally on a concall, to allow analysts to model the mix shift?

I do not have those answers. I have the question, sharpened.

Q4. Would I be happy owning this for ten years if the market shut tomorrow?

The structural tailwind is real. India’s defence capex grew 9.5% in the FY26 budget; 75% of the modernisation budget is now reserved for domestic procurement; 25% of that goes to private industry. The TAM Rangarajan describes is ₹15,000 to ₹30,000 crore over three to five years across the products Data Patterns is currently developing. The AMCA tripartite MoU with Bharat Forge and BEML gives the company a position on India’s next-generation fighter programme.

But a ten-year hold on a small-cap defence-electronics firm spans at least one capex deferral cycle, one government transition, and one or two major project slippages. Through the trough, working capital would tighten further; the strategic-contract pricing pressure would intensify; the move from subsystem to full-system would produce more low-margin quarters before producing high-margin follow-ons. Would I want to hold through that? Only if I have real conviction that the margin profile holds and that the management family stays at the helm long enough to deliver the AMCA-class wins.

The honest answer at my current depth of homework is: not at any size that matters. Not because the business is bad — it is genuinely good — but because the structural-margin gap and the succession question (which Column 2 will return to) are still open.

Q5. What could kill this business?

Five, ranked by probability over a decade:

One — margin reversion to mid-cycle norms of 25–30% as integrated full-system work dilutes the in-house IP gross margin. Not a kill of the business; a kill of the valuation.

Two — order-book lumpiness becoming order-book famine. The headline ₹1,868 crore order book is misleading: only ₹743 crore is firm; the remaining ₹1,100 crore is described as “negotiated, but not yet awarded.” Firm orders actually fell from ₹1,083 crore at end-FY24 to ₹730 crore at end-FY25. If FY26 closes without the negotiated portion converting, the optics turn ugly.

Three — working-capital cycle not normalising. FY25 operating cash flow was negative ₹19 crore (post-tax: negative ₹90 crore on Screener’s calculation) despite ₹222 crore in PAT. ICRA expects NWC/OI to stay at 90–100% in the medium term. Sustained negative OCF plus the ₹150–200 crore capex pipeline forces eventual debt or dilution. The internal-accruals-fund-everything narrative holds only if cash conversion improves.

Four — a quality or warranty event with a flagship customer. Defence equipment is in the field for two decades plus; latent defects that surface at year three on a BrahMos seeker, a Sukhoi jammer, a European-export radar would be reputationally devastating for a small-cap.

Five — policy reversal or budget compression. This is currently a tailwind, not a risk; but it is a single-knob risk and worth naming.

Check — Do I truly understand this business? Mostly. The product and customer side, yes. The margin structure under different mix scenarios — no, not at the level required to write a confident cheque on this multiple. The Q3-of-the-template flag is real and load-bearing.

Column 2 — Management

Q1. Would I trust these people to run my family’s money?

Srinivasagopalan Rangarajan, the Chairman & MD, is sixty-six. He founded the business in 1985, holds an M.Sc. from IIT Madras and a B.Tech in chemical engineering from A.C. College of Technology in Madras, and has been at this company for forty-three years. His wife, Rekha Murthy Rangarajan, is fifty-nine, a Whole-time Director, has been at the company since 2002, and signs the annual report’s “Letter from Director” each year. Together they hold 41.82% of the equity directly; the broader promoter group holds 42.41%. Vijay Ananth, the COO and a Whole-time Director, is not promoter family but joined the company in 1998 and holds 2.60% personally. The CTO has been at the company since 1989. The CFO has been here since 2000.

This is a small founder-led firm where the operating bench is institutional but the family is the family. Would I trust them to run a defence-electronics business through a capex cycle? Yes, structurally. Would I trust them to manage a transition the founder isn’t around for? Hold that question for Q5.

Q2. How do they treat minorities when no one is watching?

The FY24 related-party transactions schedule is the cleanest one I have seen on a founder-controlled small-cap. The entire schedule consists of: directors’ and KMPs’ remuneration, dividends paid on existing shareholding, sitting fees to independent directors, and a small (₹0.15 crore) dividend to one elderly relative who held a passive stake. There are no promoter-group operating entities transacting with the listed company. No leases. No consultancy fees to a Rangarajan-owned vehicle. No royalty. No inter-corporate loans. No asset sales between related parties. The directors’ report explicitly notes that Form AOC-2 disclosures are not applicable — meaning no contracts required shareholder approval as material related-party transactions. The audit committee is four-of-four independent directors, chaired by the former Executive Director and CFO of Tata Power. The audit opinion for FY24 is unmodified. The auditor is Deloitte Haskins & Sells.

This is what minority-friendly looks like in an Indian small-cap. The FY25 RPT schedule is not yet on the company’s IR site and is the one outstanding verification, but a direct search of SEBI, BSE, and NSE corporate-announcement databases for the last two years returned zero penalties, zero show-cause notices, zero late filings against Data Patterns or its promoters.

Q3. What do their last decade of capital allocation tell me?

Since the December 2021 IPO, Data Patterns has raised ₹281 crore in fresh equity at the IPO and ₹488 crore in a March 2023 QIP at ₹1,220 per share. The IPO money paid down ₹60 crore of debt, funded ₹95 crore of working capital, ₹47 crore of facility upgradation, and ₹66 crore of general corporate purposes — fully utilised. The QIP money has been largely deployed into product development (₹140 crore as of Q2 FY26), which gets capitalised on the balance sheet as intangible assets. They have undertaken no acquisitions. Capex pace stepped up post-IPO from roughly ₹1 crore in FY23 to ₹49 crore in FY25, with another ₹150–200 crore guided over the next twenty-four months — funded entirely from internal accruals plus the unutilised QIP balance. Total debt has stayed below ₹10 crore throughout. Dividends have run at a consistent ~20% payout ratio.

The pattern is conservative, organic, and aligned with the founder’s stated philosophy on the Q1 FY26 call: “We are doing the opposite here in India. Because you need to have products and to address the market, we are actually putting our money and developing — hundreds of crores or putting money and developing.” This is not empire-building; it is product-led capital deployment.

Q4. Do they tell the truth in good years and bad?

The concall transcripts are the cleanest way to test this. Three notable moments from Q1 and Q2 FY26 calls.

First, on margin: an analyst from IIFL Capital pressed Rangarajan in Q2 FY26 — “from a margin standpoint, would it be safe to assume that had we not [taken this strategic contract], our EBITDA margins for the current quarter would have been in the guided range of 35% to 40%?” The answer was a single word: “Yes, definitely.” No caveat, no hedge. Most managements would muddy that confirmation. He gave a clean yes.

Second, on competition: a retail shareholder in the same Q2 call tried to lead him to a flattering “we’re the only ones in India” framing. His response: “It’s what we believe. We do not know what is happening in the backyard. … Competition is not only India, it comes from abroad also. India actually imports more than 70%, 80% of our requirements. So you should not forget that. You can’t say it’s not made in India, only Data Patterns is doing it, so you’ll get all the orders. That also doesn’t happen.” A promoter trying to ramp the stock would not have given that answer.

Third, on execution delays: on the Q1 FY26 call, an IIFL analyst asked which orders had slipped and what the quantum was. Rangarajan named the number — about ₹27 crore — and admitted “because of inspections getting delayed substantially, we were not able to do this for the last 1 year” — a candid, quantified concession of a year-long inspection-delay backlog. Most managements would frame that as a one-quarter slippage.

The tone across all three calls is plain, technically grounded, occasionally clumsy in phrasing (English is not perfectly polished), and consistently candid. The CFO answers numerical follow-ups; the CMD does not try to fake a finance pitch. Not promotional. Not defensive. The kind of management commentary you can actually model from.

Q5. Do they have real skin in the game, and for how long?

The promoter group has held 42.41% steady across the last twelve quarters. All IPO and QIP lock-ins have expired. There has been no offer-for-sale, no creeping disposal, no insider selling reported on BSE. S. Rangarajan personally holds 22.80%; his stake at the current price is worth roughly ₹5,200 crore — multiples of his disclosed net worth in the 2022 Hurun India Rich List. He has no other listed-company directorships. Effectively all his disclosed wealth is in this stock.

That is the maximum kind of skin in the game. It is also the source of the column’s one real concern: succession. The CMD is sixty-six. There is no second-generation Rangarajan in any operating role at the company. The annual report contains no succession statement. The bench depth in non-family executives — the 1989-tenured CTO, the 2000-tenured CFO, the 1998-tenured COO who himself owns 2.6% of the equity — suggests operational succession is institutional. But the strategic vision, the customer relationships, the technical voice on every concall, the IP-direction setting — that is one person. A founder-CEO transition is the kind of event that can derail a small defence-electronics firm for two to three years even when handled well. The company has not signalled how it intends to handle this, and that silence is itself information.

Check — Can I trust these people with my capital? Yes, with one explicit caveat. The structural read is among the cleanest small-caps I have analysed: no related-party transactions of consequence, an independent and credentialled board, an unmodified audit, real and concentrated promoter skin, candid concall tone, conservative capital allocation, no governance flags. The unaddressed succession plan is the open question. For a five-year hold, manageable. For a ten-year hold, a real risk that needs to be priced.

Column 3 — Price

Q1. What am I paying today?

The 24 April 2026 close was ₹4,077.70, a market capitalisation of ₹22,829 crore. Trailing P/E is 91.5x on Screener’s calculation (102.9x on TickerTape’s, using a slightly different TTM denominator). Price to book is 14.8x against a March 2025 book value of ₹275 per share. Dividend yield is 0.19%. Trailing EBITDA margin is 39% on FY25 numbers, with the trailing twelve months through December 2025 at 34%. RoE is 15.2%; RoCE is 21.0%. EV/EBITDA is not displayed on free sources, but derived from the OP and depreciation series, sits at roughly 64x trailing.

The stock is up 95% in the last year and 97% in the last three months. The fifty-two-week range is ₹4,193 (intraday 23 April 2026) to ₹2,030. Today’s price is within 3% of the all-time high, struck three days before the close.

Q2. What is the market implicitly expecting, and do I agree?

Decode the multiple. At 91.5x trailing P/E and 14.8x book, the market is pricing a specific composite future for the next five to seven years. Reverse-engineered, the assumption set is roughly:

Revenue compounds at 25%+ from FY26 to FY30 — closer to management’s own 20–25% guidance band’s upper end than the lower
EBITDA margin holds at 35–40% durably — at or above management’s stated guidance and above ICRA’s own FY26 expectation of 30–35%
Working capital cycle normalises — meaning operating cash flow turns positive again and the negative-OCF year of FY25 is the exception, not the new pattern
The negotiated portion of the order book (₹1,100 crore at the time of the Q3 FY26 print) converts to firm contracts on management’s “next 1 to 2 months” timeline
Strategic low-margin contracts produce high-margin production follow-ons within two to three years
AMCA, BrahMos seeker production, the export tail, and the unfunded TAM ambition all materialise meaningfully

That is a six-part assumption stacked together. The probability that each part hits the implied trajectory is, on my read after the research, perhaps 70% per element. The joint probability — the bull case as a whole — is materially below 30%. The valuation is not pricing a probable future; it is pricing the optimistic future as the base case.

The bear case is not a tail event. It is the historical base rate for the industry: margin compression as the company moves from subsystem supplier to integrated full-system supplier; order-book lumpiness as MoD procurement timing slips; working capital staying high as ICRA itself forecasts; competitive entry by Bharat Electronics into the high-end radar/EW pies. These are not far-fetched. They are the things that actually happen to defence-electronics SMEs on the way up.

The consensus 12-month average target across nine sell-side analysts is ₹3,588 — about 12% below the current quote. The high end of consensus is ₹3,990, the low ₹3,030. The market is currently above all of them. That is a data point, not an argument; sell-side targets lag rallies. But it is worth naming that even the analysts who have done the deep work are not at this price.

Q3. What would I pay if I were buying the whole business?

Run the private-buyer frame on normalised-margin economics, not on Q3 FY26 peak-margin economics. Assume FY27 revenue of ₹1,000 crore (consistent with the lower end of management’s 20–25% growth guidance compounded twice). Assume EBITDA margin at 32% — the midpoint between ICRA’s FY26 trough of 30–35% and the stated FY27 target of 35–40%. Assume capex of ₹100 crore a year ongoing, working capital staying heavy at 90% of revenue, and a tax rate of 25%. The normalised free cash flow generated is meaningfully smaller than what a 39% headline EBITDA margin and a back-to-back order book would imply. At a 25x earnings multiple — which is generous for a small-cap defence-electronics firm with 100% government concentration and structurally heavy working capital — the private-buyer fair value lands well below the current ₹22,829 crore market cap. At a 35x multiple, which matches premium peers like Bharat Dynamics, the gap narrows but does not close.

The bull-case private-buyer number — assuming everything in Q2’s six-part assumption holds — is not far below the current price. So the gap between the conservative private-buyer fair value and the optimistic private-buyer fair value is wide. That width is the information. It tells you the price is not anchored to a confident steady-state value but to a forward narrative whose probability is materially below 90%.

Q4. At what price would I sell, and why?

Two clean triggers, decided in advance:

One — a full-fiscal margin print that compresses below 28% with management’s own forward guidance also moving to 28–32% rather than the current 35–40%. That signals the structural-margin reversion, not the cyclical one. The earnings number resets and the multiple resets simultaneously. By the time the reaction is visible in the price, the easy exit is gone.

Two — a quarter where the negotiated portion of the order book fails to convert and management’s own commentary shifts from “should happen in the next 1 to 2 months” to “remains under discussion.” Order-book conversion failure plus margin compression simultaneously is the bear-case scenario the price is currently pricing out of existence.

Q5. If the stock drops 40% next month, do I buy more or panic?

The honest answer depends on the cause.

Macro-driven drop — broad small-cap correction, FII outflows, election uncertainty, defence-budget rumour: probably buy more, but only if the structural thesis (Column 1 Q3) is held with conviction. For me, today, that conviction is not at the level required to confidently size a small-cap industrial. So my honest position would be: hold, do not panic, do not add.

Company-specific drop — margin miss, order-book conversion failure, governance event (the unverified SBI pledge, if it materialises into something), customer concentration revealed in a worse-than-expected light: do not buy more. The drop is information about something the template has flagged as a risk. Treat it as data, not opportunity.

Sector-wide derate — peer multiples compress as global capacity catches up, indigenisation timeline slips, BEL or HAL or Astra moves up the value chain into Data Patterns’ niche: re-run the entire price column. The thesis itself has shifted.

The lazy “small-caps are volatile, just buy when it dips” reflex is exactly the behaviour that turns a 40% drawdown into a 60% drawdown that becomes permanent because the structural story has shifted underneath you.

Check — Am I paying a sensible price? At 91x trailing P/E and 15x book, you are paying for the optimistic six-part future as the base case. The gap between the conservative private-buyer number and the optimistic one is wide; consensus 12-month targets are 12% below the current quote. For me, today, at this price, on this depth of homework — no.

Closing — what the three columns surfaced

The three columns load very differently for Data Patterns, and that asymmetry is the entire point of running the framework rather than reaching for a buy/hold/sell.

Business: medium conviction. There is a real IP-driven moat in radar and EW. The indigenisation tailwind is structural, not cyclical. But the central margin question — whether 38–43% is the structural floor or the cyclical peak — is genuinely unanswered by public data, and the company’s single-segment reporting makes it impossible to verify the mix-shift thesis from primary sources. Add to that a “₹1,868 crore order book” headline that is actually ₹743 crore firm and ₹1,100 crore “negotiated, not awarded,” plus a working-capital cycle that ate the entire FY25 profit in cash terms, and the business column comes out at perhaps 70% conviction — solid, not strong.

Management: high conviction with one explicit caveat. The cleanest related-party schedule I have seen on a founder-led small-cap. Substantively independent board including a former CMD of Bharat Electronics. Unmodified Deloitte audit. Promoter holding flat at 42.41% across twelve quarters with all locks expired and no offer-for-sale. Concall tone plain, candid, and on three separate occasions actively pushing back against analysts trying to lead the founder to flattering framings. Maximum skin in the game — the founder’s entire disclosed wealth is in this stock. The single open question, large enough to deserve its own line: there is no succession plan disclosed, the CMD is sixty-six, and there is no second-generation Rangarajan in operating roles.

Price: at the top of the band, and pricing the optimistic future as the base case. P/E 91.5x, P/B 14.8x, both at the upper end of the since-IPO range. Consensus 12-month targets sit 12% below the current quote. The implicit forward narrative — sustained 25%+ growth, durable 35–40% margin, working-capital normalisation, order-book conversion, strategic-contract follow-ons all hitting — has joint probability materially below 30%. The bear case is not a tail event; it is the industry’s historical base rate.

The investor who buys Data Patterns today because “defence indigenisation, multibagger, founder-led, order book at all-time high” is making one trade that they may not realise is six trades stacked together. The investor who has run the three columns is making a defensible six-part bet with eyes open. Same brokerage statement. Genuinely different trades. Different outcomes over a decade.

This is also the company that most rewards running the template properly. A surface read pulls you toward the rally; a five-minute “expensive defence stock” gloss pushes you away from a genuinely high-quality founder-led business with one of the cleanest governance profiles I have seen. Either lazy answer would be wrong. The honest answer — that this is a real business worth doing more primary work on, and that the homework required before sizing is meaningfully more than today’s quote allows comfortable margin for — is the one only the framework forces.

Process notes

A few observations from the first end-to-end run of the equityanalysis-vk skill with the mandatory research phase:

The research phase changed the analysis substantially compared to what I would have written from structural knowledge alone. Three findings would not have surfaced without primary work: that the headline “₹1,868 crore order book” is actually ₹743 crore firm; that FY25 operating cash flow was negative despite ₹222 crore in PAT; that the consensus 12-month sell-side target is 12% below the current quote. None of those were in my prior memory of the company, and each one materially changes the conclusion the price column lands on.

The single-segment Ind AS 108 reporting is the binding constraint on Column 1 Q3. Without segment-level EBITDA, the structural-vs-cyclical margin question is unanswerable from public sources — the analysis can only describe both possibilities, name the data the company would need to disclose to settle it, and adjust sizing accordingly. That is a real, residual gap, and the methodology preamble surfaces it explicitly.

The cleanest finding of the run is on the management column. The Rangarajan family’s RPT schedule, audit posture, board composition, candid concall tone, and 42.41%-flat shareholding through twelve quarters of lock expiry deserve to be flagged as a positive outlier among Indian founder-led small-caps. The succession gap is real. Both things are true at once.

Score 8. The single highest score on the skill so far, but for what it’s worth, the score is for the analysis, not the stock. The analysis is high-quality because the research phase produced specific, datable, citable findings on each of the fifteen questions and because the columns load asymmetrically enough that the framework’s “no aggregate verdict” feature actively earns its keep. The Power Grid run was a 7 on structural knowledge plus flagged gaps. The Shilchar run was a 6 on honest “do more homework before sizing.” This run is an 8 on a real research substrate that the public note actually rests on.