heading · body

Stock · PARAS · Defence

Paras Defence — a niche jewel at a king's ransom

Paras Defence and Space Technologies Ltd

period FY26 (year ended Mar 2026) + Q4 FY26 added 2026-06-20 score 5.5/10
wealth-lens buffett qglp india PARAS defence

Snapshot

Paras Defence makes the hard-to-make bits of India’s defence and space kit: submarine periscopes (it is the only maker in the Asia-Pacific), optical systems and hyperspectral cameras for ISRO satellites, electromagnetic-pulse (EMP) shielding (a number that protects gear from a nuclear-blast electrical surge), and a growing anti-drone business. It is a small company — about ₹477 crore of sales in the year to March 2026 — wearing a very large price tag: market cap ₹11,352 cr, CMP ₹1,409, sitting right at its 52-week high (range ₹580–₹1,445). Stock P/E 132×, price-to-book 15.6×, RoE 12.6%, RoCE 16.9%, almost no debt.

What kind of animal is it? A rare, technically gifted niche manufacturer — a genuine boat — but one that still earns ordinary returns on its money, gets paid painfully slowly, and trades at a price that assumes near-flawless growth for a decade. As of 2026-06-20, from screener snapshot.

The verdict in two boxes — the business first, the price second

Box 1 — The business (durable):

LensResult
Business-quality score13.5 / 23 (Quality 6.5/12 · Growth 4.5/6 · Longevity 2.5/5)
Buffett rubric4 / 10 PASS (≈4½ counting partials)
Business bucketGood, leaning toward Gruesome on capital intensity
Wealth-creator typeEmerging / Transitory-until-proven · Volatile
Economic Profit≈ +₹4 cr (net worth ₹725 cr × [RoE 12.6% − CoE 12%]) — barely creating value above the cost of owners’ money

A Good business — technically special, but capital-hungry and still earning only ordinary returns — that has not yet proven it is a durable wealth creator. This verdict would read the same if the share price doubled or halved tomorrow.

Box 2 — The price today (a current phenomenon):

ReadingResult
CMP₹1,409 (as of 2026-06-20)
Price pillar0 / 2 (PEG ≈ 3.2× · 5-yr payback ≈ 11×)
Margin-of-safety band₹350–₹520 (where PEG ≈ 1× / payback approaches sane)
Mr. Market’s mood nowGreedy — defence-indigenisation + drone-warfare + “monopoly moat” euphoria; every analyst with a calculator sits ₹700–₹820
CMP vs the bandDemanding — priced about 3× above the band

Today the market is pricing it rich — a mood driven by the post-Operation-Sindoor defence frenzy and a “monopoly” story, which can cool while the business above stays exactly what it is.

In plain English

Imagine a small, brilliant workshop that makes a handful of things almost nobody else in India can make. It grinds the precise mirrors and lenses that go into spy satellites. It builds the periscope a submarine captain looks through — and it’s the only firm in this corner of the world that does. It wraps military electronics in shielding so a nuclear-blast pulse can’t fry them. And lately it’s been building gear to shoot down enemy drones. That’s Paras. It is, genuinely, a clever little boat.

Here’s the catch, and it’s a big one. A wonderful workshop and a wonderful investment are two different things. Paras earns about ₹12.60 of profit for every ₹100 of owners’ money in it (RoE 12.6%). That’s roughly what a decent bank deposit-plus does — not the 20–25% you’d want from something this special. Worse, it gets paid agonisingly slowly. Its customers — mostly the government and big defence houses — take about 278 days to pay. Picture running a shop where every customer settles their bill nine months after they walk out with the goods. You have to fund all that inventory and all those unpaid invoices yourself. So even when profits look good on paper, the cash is stuck in receivables. That is the single most important fact about this business, and it is why a fast-growing Paras can still be a cash-thirsty Paras.

Now the price. The market is paying about 132 times one year’s earnings for it. That’s not a number you pay for a 12% RoE manufacturer that gets paid in nine months. That’s a number you pay when you’ve fallen in love with a story — and right now the story is irresistible: India is spending record sums on home-made weapons, drones are the new front line (India downed 500-plus Pakistani drones in May 2025’s Operation Sindoor), and Paras has “monopolies.” All true. But love is not a margin of safety. Every brokerage analyst who actually does the arithmetic has a fair-value somewhere around ₹700–₹820 — roughly half the market price.

The tension in one line: a special little business, but you are being asked to pay a king’s ransom for a workshop that still earns a clerk’s return on capital and waits the better part of a year to get paid. The quality is real; the price assumes a perfection the company has not yet delivered — it just missed its own FY26 growth guidance.

Sitting down with the management

Dear reader — if you sat across the table from the Shah family, you’d meet self-made engineers, not MBA empire-builders, and that’s to their credit. Sharad Virji Shah started a precision-components shop, M/s Paras Engineering, back in 1979 [HARD — IPO prospectus]. The filing is unusually frank: it says he completed school only to secondary level, and that the company couldn’t even locate records of his educational qualifications. This is a man who learned the craft on the floor over forty years. His son Munjal Shah runs it today and is the public face. (You’ll see third-party sites claim Munjal holds a US master’s degree — the SEBI prospectus says higher-secondary; trust the filing, not the legend.) Their stated creed is Atmanirbhar — make at home what India used to import. That’s a genuine engineering obsession, the kind Buffett’s “passion” test likes to see.

On spending the owners’ money, the record is mixed-to-fine. The 2021 IPO — famously oversubscribed about 304 times — raised modest money and listed at a 170% premium. A small, sensible ₹135 cr QIP in October 2024 at ₹1,045 diluted shareholders barely 3% [HARD]. They cleaned up a loss-making subsidiary (Ayatti) for a tiny gain in March 2026 rather than nursing it [HARD]. Their best capital move is the Controp-Paras electro-optics venture with Israel’s Controp, which is already winning real orders (a ₹305 cr order routed via L&T) [SOFT/MEDIUM]. The balance sheet is clean — debt has all but vanished, net cash. The one-dollar test (did each retained rupee make a rupee of value?) just barely passes: book value compounded, but at a 12% RoE the spread over the cost of capital is a sliver, so retained earnings are working only modestly hard.

Now the concern I can’t wave away. Promoter holding fell from 58.94% to 53.20% between mid-2024 and late-2025 — and the cause was the founders selling shares into the rally. On 19–20 May 2025, after the stock had doubled in ten weeks, three promoters dumped 13.34 lakh shares in bulk deals — the Chairman himself sold 9 lakh shares at ₹1,683 [HARD — BusinessToday, 2025-05-20]. No company reason was given; the honest read is profit-booking near the highs. Promoters selling aggressively into euphoria is exactly the tell Buffett warns about — they know the business best, and they were sellers, not buyers, at these levels. To be fair: no pledging (a clean April-2026 disclosure confirms zero encumbrance) [HARD], a clean auditor (Chaturvedi & Shah, unmodified) [HARD], and reasonable promoter pay (MD ~₹84 lakh). But it’s a family-dense outfit — several relatives in paid roles and a ₹293 cr related-party limit with their own associate [HARD] — which you’d want to keep an eye on for arm’s-length pricing.

Would Buffett and Agrawal shake hands on this management? A wary half-handshake. Admirable builders with a real swadeshi mission and clean books — but the founder selling a big block at the top, the thin RoE, and the chronic nine-month wait to get paid are things they’d press hard on. What would change their mind: a couple of years of promoters holding (or buying) rather than selling, and debtor days falling toward 150 while RoE climbs past 18%.

What’s on the horizon (live-issues tracker)

1) The crux — can a ₹477 cr business grow into an ₹11,350 cr price? (full interrogation below) — 🟡 Mixed/unproven

2) The receivables problem — getting paid. 🔴 Worrying. Debtor days have sat at 246 → 285 → 295 → 278 over the last four years — i.e. customers owe roughly ₹363 cr, about 76% of a full year’s sales. As revenue scales 30%+, the unpaid invoices scale with it and can swallow the cash that growth is supposed to produce. Free cash flow has been negative in three of the last five years (−₹71 cr in FY24, −₹10 cr in FY26) precisely for this reason. No concrete management plan to cut debtor days has surfaced — and the company doesn’t hold regular earnings calls, so we can’t press them on it. What to watch: debtor days in the FY26 annual report; operating cash flow turning reliably positive.

3) The optics & space ramp. 🟢→🟡 On track but mind the mix. The high-margin Optics & Optronics segment more than doubled its revenue in FY25 (₹69 cr → ₹177 cr) and carries ~54% segment margins versus ~22% for Defence Engineering. The bet is to push optics toward ~65% of revenue and lift blended margins above 28%. The risk showed up in Q4 FY26: a surge of low-margin engineering orders pulled blended EBITDA margin down to ~25%, below the ~27% guided. A ₹500 cr Optical Systems Development Park is real near-term capex; the headline “₹12,000 cr Optics Park” Davos MoU is intent with a 2028–2035 horizon, not committed money [SOFT].

4) Anti-drone — riding the right wave, but not the captain of the ship. 🟡 Early. Operation Sindoor (May 2025) made counter-drone the hottest theme in Indian defence, and Paras has won a string of orders (a ₹142 cr DRDO laser-dome, several smaller MoD orders). But the big named anti-drone contracts have gone to BEL (the Army’s landmark integrated drone-defence deal), Zen Technologies (a ₹404 cr AI-integrated system) and Adani Defence. Paras is a capable component and sub-system player here, not the market leader its narrative implies [HARD/SOFT].

The watch-list:

  • Debtor days falling below ~220 (working / breaking).
  • Operating cash flow positive and ≥ 60% of operating profit for two straight years.
  • Optics share of revenue climbing toward 60%+ (margin mix improving).
  • Order inflow disclosed and growing faster than the ~₹986 cr backlog burns off.
  • Promoter holding stabilising at ~53% or rising — no further selling into strength.
  • Revenue growth actually hitting the guided 35%+ (it missed at ~31% in FY26).

QGLP scorecard (the Motilal Oswal lens) — the receipts

The 25-point audit. Business-quality is the headline (/23); Price is reported separately (/2).

#QuestionScoreEvidence
Quality of Business (Q1–Q6)3.5/6
1Large opportunity?1Indian defence capex ₹6.81L cr (FY26) → ₹7.85L cr (FY27, +15%); 75% reserved for domestic industry — a vast, structural market
2Favourable industry structure?1Niche monopolies (only AsiaPac periscope maker, sole private hyperspectral camera) + high entry barriers; OPM held 25–31% over 8 yrs
3Clear, defensible moat?0.5Real niche moats, but RoE beat 15% in 0 of last ~7 years (RoCE row 10–17%) — moat shows in capability, not yet in returns; many “monopolies” rest on being DRDO’s chosen ToT partner
4Return ratios > 15% consistently?0RoE 12.6% (3-yr avg 10.9%); RoCE 16.9% latest but only crossed 15% in last 2 yrs — chronically sub-15% RoE
5Asset-light / low capital intensity?0.5Decent fixed-asset base but heavy working-capital intensity; FCF negative 3 of last 5 yrs — capital-hungry to grow
6Favourable terms of trade (neg. working capital)?0Debtor days 278 vs payable ~88; cash-conversion cycle ~435 days — it banks its customers (the opposite of favourable)
Quality of Management (Q7–Q12)3/6
7Unquestionable integrity?0.5Clean auditor, no pledging, no SEBI action; but founder sold a 9-lakh-share block near the top, family-dense RPTs
8Proven execution track record?0.5Grew sales 41% CAGR/5yr; but missed its own FY26 guidance (31% vs 35–40%) and margin target
9Growth mindset & vision?1Genuine R&D/indigenisation obsession; multiple JVs (Controp, HevenDrones, Green Optics) and new capacity
10Superior capital allocation?0.5Clean balance sheet, sensible small QIP & subsidiary cleanup; but retains at only ~12% RoE — one-dollar test barely passes
11Clear succession plan?0Policy exists on paper; control concentrated in Munjal Shah, no designated operating successor — real key-man risk
12Minority interests protected?0.5Small dividend started FY25; no value leakage found, but promoter selling and RPT density are watch items
Growth (Q13–Q18)4.5/6
13Structural tailwind?1Defence indigenisation + exports (₹38,424 cr FY26, +63%) growing well above GDP
14Volume-led growth?1Order-backed, project-volume driven, not price/commodity
15Operating leverage?0.5OPM rose to 25–27% but slipped in Q4 FY26 on mix; leverage real but uneven
16Manageable, accretive leverage?1Near debt-free (D/E ~3.6%), net cash — comfortable
17Market-share gain potential?0.5Gaining in optics; but a follower, not leader, in the headline anti-drone contracts
18Earnings growth > 15% CAGR?1PAT 41% CAGR/5yr, 35% CAGR/3yr (₹16cr→₹89cr) — clearly above the bar
Longevity (Q19–Q23)2.5/5
19Relevant for next 10–15 yrs?1Defence optics/electronics/anti-drone — durable, even rising, demand
20Can extend its moat (CAP)?0.5Possible if optics IP deepens; but DRDO-dependent monopolies can be re-tendered to others
21Can sustain growth runway (GAP)?0.5Large TAM, but execution + working capital cap how fast it can convert it
22Diversification headroom?0.5Optics → space → drones → exports gives optionality, all early-stage
23Adaptive, resilient culture?0No evidence through a real downturn yet; young as a listed, scaled entity
Business-quality total13.5 / 23Quality 6.5 · Growth 4.5 · Longevity 2.5
Price (Q24–Q25) — reported separately0/2
24Valuation reasonable (PEG)?0PEG ≈ 132 / 41 ≈ 3.2× (>2 = clear fail)
25Margin of safety (payback < 1×)?05-yr payback ≈ 11× market cap to projected earnings — no margin of safety

The pattern: Growth is the strength (4.5/6); Quality of returns and Longevity are the gaps. The business scores a respectable-but-not-elite 13.5/23 — a Good, emerging compounder with two real wounds (sub-15% RoE and a 278-day collection cycle). Price is a clean 0/2 — the cheapest thing in this report is the quality verdict; the price is the most expensive.

Buffett lens (the Berkshire-letters read)

#TestResultEvidence
1Good boat? (business > management)PARTIALA Good boat — special products, but capital-hungry and only ordinary RoCE/RoE
2Moat + franchise + pricing powerPARTIALNiche monopolies and stable 25%+ margins, but RoE beat its cost of capital in ~0 of 10 yrs — a franchise on paper, not yet in the returns
3See’s test (high returns, little capital)FAILNegative FCF in 3 of 5 yrs; growth eats cash via receivables — the opposite of See’s
4Capital allocation (one-dollar test)PARTIALBook compounded and no value-destroying M&A, but retained at only ~12% RoE — barely a rupee for a rupee
5Owner-oriented, candid managementPARTIALSelf-made, mission-driven, clean books; but no regular concalls and founder sold a big block at the top
6Integrity / forensic (no “credit P&L, debit B/S”)PARTIALNo pledging/auditor flags; but receivables balloon faster than sales and CWIP/other-assets grow — profit isn’t reliably becoming cash
7Circle of competence / predictabilityPARTIALDefence demand is durable, but the precise 5–10-yr earnings path is lumpy and order-dependent
8Mr. Market — gift or trap now?FAILPriced for perfection at 132× P/E, near all-time high, euphoric retail ownership — a trap, not a gift
9Patience / compounding runwayPASSLong runway in a structurally growing market if it can fund it
10The honest red flagSee below

The See’s test, spelled out. See’s Candies was magic because it threw off cash without needing much capital fed back in. Paras is the photographic negative: in FY24 it generated ₹30 cr of profit but −₹71 cr of free cash flow, because growth poured money into inventory and unpaid invoices. A business that consumes cash to grow is the “Good” company Buffett tolerates, not the “Great” one he treasures. FAIL.

The one-dollar test. Has every retained rupee created at least a rupee of market value? Book value did compound (reserves ₹177 cr in FY21 → ₹685 cr in FY26). But the engine ran at only ~12% RoE — barely above the ~12% it costs to keep owners’ money. So the rupees were retained and did grow the business, but at a spread so thin that the value created per rupee is marginal. PARTIAL — it clears the bar, but only just, and far less impressively than the share-price suggests.

Buffett count: ≈4 PASS (4 clean + many partials ≈ 4½). That lands it as a real business with real gaps — not in the temple, whatever the price.

The framework metrics

  • Economic Profit = Net Worth ₹725 cr × (RoE 12.6% − CoE 12%) = ≈ +₹4 cr — technically creating value, but a rounding error. The company sits right at the line between making and destroying economic value. [CoE assumed 12%, the Indian benchmark.]
  • Terms of Trade = Debtors ÷ Creditors ≈ 278 days ÷ ~88 days ≈ ~315% — deeply unfavourable. It funds its customers; cash-conversion cycle ~435 days.
  • 5-yr Payback = Mcap ₹11,352 cr ÷ projected cumulative 5-yr PAT ≈ ~11× (assuming a generous 30% PAT CAGR: FY26 ₹89 cr → cumulative ~₹1,047 cr) — vastly above the 1× multibagger signal.
  • PEG = P/E 132 ÷ PAT growth ~41% = ~3.2× (>1× = no price discipline).
  • RoE − CoE spread = ~0.6% ; RoE > 15% in ~0 of the last 7 years (RoCE crossed 15% only in FY25–FY26).
  • Consistent vs Volatile test = Volatile. PAT fell year-on-year in FY21 (₹20→₹16 cr) and FY24 (₹36→₹30 cr) and is lumpy quarter-to-quarter — a young, project-driven earnings stream. Value it on assets/track-record, not a smooth P/E.

Peer comparison

CompanyMkt cap (₹cr)CMP (₹)P/EP/BRoERoCEOPMSales (₹cr)
Paras Defence11,3521,40913215.612.6%16.9%25%477
Data Patterns26,9964,82298.715.616.9%23.3%40%925
Zen Technologies18,0311,99793.59.610.7%16.2%36%688
MTAR Technologies25,7608,37426631.412.5%15.1%20%876

Peer ratios from each company’s screener snapshot, 2026-06-20.

The relative read is sobering even before the absolute one. Data Patterns is the clear quality benchmark — bigger, faster-growing, 40% margins, and meaningfully better returns (RoE 16.9%, RoCE 23.3%) — yet trades cheaper than Paras on P/E. Zen Technologies has thinner returns but a 36% margin and, notably, the bigger anti-drone contracts, at a lower multiple. MTAR is the cautionary tale of what a sky-high multiple does when growth wobbles (266× P/E on 20% margins). So within its own asset class, Paras is neither the cheapest nor the highest-quality — it carries a premium P/B with the lowest margin and a middling RoE. The peer lens reinforces the absolute verdict here rather than flipping it: the whole defence-supplier shelf is priced for a glorious future, and Paras is priced like the best of them while operating like the middle of them.

Latest quarter & what’s happening now

Q4 FY26 (year ended March 2026, reported ~May 2026): revenue ₹171 cr (+58% YoY), net profit ₹39 cr (+87%) — the highest-ever quarter, ₹1 final dividend declared [HARD]. Full-year FY26: revenue ₹477 cr (+31%), PAT ₹89 cr (+46%). The quarter’s surge came largely from the lower-margin Defence Engineering segment, which is why blended margin (~25%) came in below the ~27% management had guided [MEDIUM].

Live catalysts, dated: a ₹72 cr DRDO submarine-periscope order (Nov 2025) [HARD]; a ₹142 cr DRDO “RayStrike” anti-drone laser-dome order (Mar 2025) [HARD]; the HevenDrones 51:49 JV (May 2025) and a Green Optics South Korea MoU (Mar 2026), both pre-revenue [SOFT]; a ₹500 cr Optical Systems Development Park (near-term) and a ₹12,000 cr “Optics Park” Davos MoU (2028–2035 horizon — intent, not committed) [SOFT]. The order book is roughly ₹986 cr firm (about 2× annual sales) [MEDIUM]. The company does not hold regular earnings calls, so most “guidance” reaches us second-hand through brokerage notes — a real disclosure gap.

Where the two lenses agree — and disagree

They agree on the headline: a Good-not-Great business at a Gruesome price. QGLP’s business score (13.5/23) and Buffett’s ~4/10 both say real but flawed, and both Price tests fail hard.

The most interesting divergence is on the moat. QGLP’s checklist hands Paras points for “niche monopolies” and a favourable industry structure (Q2, Q3) — the capability moat is real and rare. But Buffett’s tests 2–4 demote it, because a franchise is only proven when it shows up as high returns on capital sustained for years — and Paras’s RoE has never cleared 15% in the last seven. The lesson the divergence teaches: being the only maker of a thing is worth little if the customer takes nine months to pay and you earn 12% on the money. Capability monopoly ≠ economic moat. Trust the Buffett flag here.

The price as a current phenomenon

This section judges the price, not the business — the business verdict above is already settled.

The margin-of-safety band. For QGLP’s Price pillar to pass, you’d want PEG ≤ 1× and the 5-year payback heading toward sane. At even a generous 40% sustained growth, PEG ≤ 1× implies a P/E near 40× — which on FY26 EPS of ~₹10.9 puts the band around ₹350–₹520. That’s the price at which the quality on offer starts to be worth owning on these frameworks. CMP ₹1,409 is roughly 3× above the top of that band.

Mr. Market’s mood. Right now the crowd is plainly greedy on this name, and for an understandable reason: Operation Sindoor (May 2025) turned drone-warfare and home-made defence into the market’s favourite story, the Nifty Defence index ripped, and Paras — with its “monopoly” tags — became a momentum darling, up ~2.4× off its low and pinned to its all-time high. The tell that this is mood, not maths: every brokerage that actually models the company sits at ₹700–₹820 — about half the market price. The Buy-votes are sentiment; the spreadsheet-votes are cautious.

The tension, plainly: a special little business sitting at a thoroughly unspecial price for a buyer. A great business can be a poor purchase, and this is the textbook case — the quality is genuine but the price has run miles ahead of it. And remember: this price reading can flip next month — a defence-sentiment cooldown, a single weak quarter, or a budget disappointment — without one thing in the business above changing. Not a recommendation; just the arithmetic of what you’re being asked to pay.

Conviction texture

The bull case, at its strongest. India is in a once-in-a-generation defence-indigenisation supercycle, exports are compounding 60%+, and Paras owns truly scarce capabilities — periscopes, space optics, hyperspectral cameras, EMP shielding — that take decades to build and that the government wants made at home. The order book is ~2× sales, the balance sheet is net cash, and the optics segment (54% margins) is scaling fast. If optics becomes two-thirds of revenue and the anti-drone wave lifts inflows, earnings could compound 30–40% for years, and a 132× multiple compresses gracefully as profits catch up.

The bear case, at its strongest (Buffett’s red flag). This is a 12%-RoE manufacturer that funds its own customers for nine months and has burned free cash in three of five years — and it just missed its own growth guidance. The “monopolies” largely rest on being DRDO’s chosen partner, which can be re-tendered; in the hottest growth vertical, anti-drone, the marquee contracts went to BEL, Zen and Adani, not Paras. The founder-Chairman sold a 9-lakh-share block near the top. And you’re paying 132× earnings and 15.6× book for all of that. The single strongest reason this is not a wealth creator at this price: growth that consumes cash, earned at a return barely above the cost of capital, does not create economic value — and the price assumes it will create a great deal.

What the numbers actually support: a genuinely interesting, capable small company that is emerging — not yet proven — as a compounder, currently priced as if the proof were already in. The three things that would tip the verdict toward “wealth creator”: (1) debtor days falling toward 150 with positive operating cash flow; (2) RoE climbing durably past 18%; (3) promoters holding or buying instead of selling. Until then, the boat is good; the price of the seat is the whole argument.

No buy/sell/hold — the reader decides.

Sources

Assumptions: Cost of Equity 12%; PAT CAGR 41% (5-yr) used for PEG; 5-yr payback projected at 30% PAT CAGR. All prices/ratios as of 2026-06-20.