Borderline, landed KILL. Re-check at first results.
This is a DRHP-stage read of the public draft prospectus, done long after the fact as a held-back test of the process. The draft carried no price band, no lot and no dates, and its audited numbers stop at March 2025. The six ratings are frozen on what was knowable on 30 September 2025; what actually happened afterwards is reported at the end and moved no rating.
Why:
- Twenty-five years in, the business does not make money. Reported FY2025 EBITDA of ₹108 crore rests on ₹35 crore of other income and ₹80 crore of operating costs booked into the new plant's construction account instead of the P&L; strip both and operating profit is roughly zero (T1, DERIVED). Revenue fell 4% in FY2025, the loss was ₹102 crore, cash flow never covered interest and lease payments in any of the three years, and a 2019 press interview puts revenue then at about ₹920 crore, the same as FY2025 (T3): six years of no growth.
- The aerospace edge is real but it is not converting. Airbus and Boeing approvals of this depth take a decade and are genuinely scarce, and the segment did grow 41% over two years with utilisation rising to 67%. But in FY2025 it grew 8.9% while every listed Indian aerospace peer grew 16% to 33%, including Dynamatic, the closest comparable, whose aerospace arm grew 19% at a 26% margin against Aequs's 19% (T1, T2). The largest capacity in the cohort is earning the weakest returns.
- The offer mostly pays for the past. Of the ₹720 crore fresh issue, 9% buys machines, 58% repays debt of which over half is the failed consumer plant's, and 32% has no named use, pressed against the regulatory ceiling. The repayment also releases up to ₹515 crore of the chairman's personal loan guarantees, 95% of his exposure (T1).
The story
A buyer is buying India's most qualified aerospace machining cluster, and paying for a consumer detour. The Belagavi ecosystem makes precision parts for Airbus, Boeing and their big suppliers: 89% of revenue, all of the profit, six Airbus best-supplier awards, and a market where those two customers already buy about ₹19,000 crore a year from India. Around it sits a consumer business (toys, cookware, and a new ₹480 crore laptop-shell plant) that holds 46% of the group's operating assets, produced 11% of revenue, halved last year, and had its goodwill written off. The IPO's main job is to clean up the balance sheet that build left behind.
What this business is
Aequs machines, forges and surface-treats metal aircraft parts to its customers' drawings: structural pieces, interiors, landing and engine components, about 4,700 approved part numbers across the A220, A320, A350, B737 and B787 programmes. Almost everything is exported. The work is done in one special economic zone at Belagavi in Karnataka, with smaller plants in France and Texas, and two joint ventures fill in forging and chemical treatment so a part never leaves the campus. Customers approve every step, switching a qualified part to another supplier takes years, and the three biggest customer groups have been buying for 15 years on average.
The consumer side applies the same machines to different buyers: toys for Hasbro and Spin Master at Koppal, cookware now moved into a joint venture with Tramontina, and, newest, machined laptop bottom covers and watch casings at Hubballi for one unnamed global electronics customer, in production only since July 2025.
Easy or difficult business? The aerospace work is genuinely difficult: four-micron tolerances, titanium and inconel, accreditations that take years and audits that never stop. That difficulty is the company's real asset. The consumer work is not difficult in the same way, which is why it has struggled against Asian contract manufacturers.
Key numbers
| ₹ cr | FY2023 | FY2024 | FY2025 |
|---|---|---|---|
| Revenue | 812 | 965 | 925 |
| Revenue growth % | n/a | 18.8 | -4.2 |
| EBITDA | 63 | 146 | 108 |
| EBITDA margin % | 7.8 | 15.1 | 11.7 |
| PAT | -109 | -14 | -102 |
| Cash from operations | 10 | -19 | 26 |
Profit swung harder than revenue in both directions and neither headline year is clean: FY2024's near break-even includes a ₹19 crore gain on selling a French building, and FY2025's loss includes a ₹48 crore write-off of the toy subsidiary's goodwill after management's own growth forecast for it failed within a year (T1). EBITDA here follows the company's definition, which includes other income.
| Segment, ₹ cr | FY2023 | FY2024 | FY2025 |
|---|---|---|---|
| Aerospace | 585 | 757 | 825 |
| Aerospace growth % | n/a | 29.4 | 8.9 |
| Aerospace EBITDA | 83 | 174 | 160 |
| Consumer | 227 | 208 | 100 |
| Consumer growth % | n/a | -8.3 | -52.0 |
| Consumer EBITDA | -16 | -16 | -29 |
Scorecard
| Block | Rating | Why |
|---|---|---|
| Right to win | MARGINAL | The moat is qualification: six-time Airbus detailed-parts partner, Boeing approved since 2017, accredited chemical processing, an integrated campus that matches how the big buyers say they want to buy, and 15-year customer tenures. What the moat has not produced is winning: no market share stated anywhere in the filing, FY2025 growth below every listed peer, contracts that commit customers to nothing, prices cut annually by the customers as standard practice, and 25 years without a profitable year on record. A real edge that has yet to pay, and a management that spent ₹370 crore of it entering laptop shells against Foxconn. |
| Industry and TAM | PASS | Airbus and Boeing already buy about ₹19,000 crore a year from India, growing fast, against Aequs's ₹825 crore; the order backlog behind it is 11 to 14 years of production; and a third of the Indian aerospace capacity is still idle, so growth needs no new market (T1, T3). The runway is not the problem. |
| Financial momentum | FAIL | Revenue shrank in FY2025, comparable operating profit is roughly zero, returns on capital are about 1%, working capital absorbed 96 more days than two years ago, and three years of operations generated ₹17 crore of cash against ₹533 crore of capex. The aerospace segment alone is decent but earns below the company's own hurdle rate, and ₹395 crore of nearly finished plant is about to start depreciating into the weak segment (T1). |
| Risks, governance, RPTs | PASS | Zero criminal cases, zero pending regulatory action, zero SEBI history, every approval in hand, and all special investor rights die at listing. The one recurring weakness is foreign-exchange compliance paperwork: an RBI penalty in 2020, a since-withdrawn exporter caution-listing in 2022, and one unresolved query on ₹12 crore of guarantee payments (T1). |
| Promoter and cap table | PASS | The promoter funded the entire COVID down round alone, put ₹85 crore of new cash in at ₹74.64 five months before filing, personally guarantees ₹542 crore of group debt, pledges nothing, and sells 0.36% of his holding. The caution: the land, buildings, utilities and even the brand sit with promoter-owned entities that collect about 14% of revenue a year, and control runs through a Cayman foundation with no named beneficiaries (T1). |
| Offer structure | MARGINAL | Nobody who matters is leaving: promoters sell almost nothing and the largest outside investors sell nothing. But the fresh money mostly settles the past. The biggest single use recapitalises the loss-making consumer subsidiary, over 40% of the repaid debt is working-capital lines the company reserves the right to redraw, a third of the issue has no named purpose, and the most durable single effect is releasing the chairman's personal guarantees (T1). |
Watch out for
- The estate is not in the company. Every Indian plant stands on land leased from promoter-owned entities, which also supply power, water and the "Aequs" name. About ₹130 crore a year, 14% of revenue, flows to them, with ₹46 crore of company cash parked as deposits. Disclosed, priced, and shrinking, but a buyer owns the operator, not the campus (T1).
- A depreciation wave is scheduled. ₹395 crore of plant under construction starts depreciating as the Hubballi lines come on stream, roughly ₹40 to 45 crore a year of new charges plus interest that stops being capitalised, more than the interest the IPO saves (T1, DERIVED).
- Customer concentration with no committed volumes. Top three customer groups are 54% of revenue, top ten are 89%, all on requirement-based contracts, and one named customer (Hasbro) fell 57% in two years without any qualification event (T1).
- The consumer plant's customer is unnamed and its economics undisclosed. The filing names only "one of the largest consumer electronics companies in the world"; pre-filing press reports say Apple (T3, unconfirmed). No order value, no term, no unit price appears anywhere in the document.
- Tax benefits worth checking, not fearing. The scary-looking "23% of revenue" of SEZ benefits is mostly refundable duty for an exporter; the real income-tax holiday was worth ₹7 crore in FY2025 (T1).
The offer
- Raising up to ₹720 crore of fresh money plus up to 3.18 crore shares sold by existing holders, no price band yet (T1).
- For repaying ₹419 crore of debt (over half of it the consumer plant's), ₹67 crore of aerospace machines, and up to ₹233 crore of unnamed acquisitions and general purposes (T1).
- Implied valuation from the last price actually paid: ₹74.64 per share in the May 2025 rights issue (1.7 crore shares, about 3% of the base), which puts the company at about ₹4,516 crore before the IPO (T1).
- Promoters hold 65.7% before the offer and sell 0.36% of their own holding; the selling is a 2023-vintage private equity fund taking half to two-thirds off the table while Amansa, Steadview and Catamaran sit tight (T1).
What happened next
Everything below post-dates 30 September 2025, moved no rating, and is the answer key this held-back run gets marked against.
The offer priced in December 2025 at ₹118 to 124, well above the ₹74.64 last private price, valuing the company at about ₹8,300 crore. The fresh issue was cut to ₹670 crore, the offer for sale to 2.03 crore shares, and the red herring carried half-year numbers to September 2025 the draft did not have: revenue up 17% and the loss nearly gone. The issue was subscribed 102 times, listed on 10 December 2025 at ₹140, up 13%, and after dipping to ₹113 in March 2026 stood at ₹256 in August 2026, roughly double the issue price. FY2026 results showed aerospace revenue up 27%, consumer up 84%, an aerospace order book past $1 billion and a 15-year Safran wheel contract, but also a consolidated loss slightly wider than FY2025's, as the plant costs and depreciation this assessment flagged arrived on schedule (T2, T3).
Where the frozen read was right: the accounting mechanics played out as written. The capitalised costs and the depreciation wave hit the P&L and kept the company loss-making through FY2026 and into FY2027's first quarter, and the offer was reshaped at the red herring stage exactly where this read called it soft, with the sale shrunk and fresh half-year numbers doing the selling. The borderline label was also honest: this note says one grade separates KILL from TRACK, and the two blocks named as the hinge, conversion and the offer, are precisely what the later record moved.
Where it was wrong: on the hinge itself. The FY2025 share-loss reading, the strongest single fact behind the kill, reversed immediately: FY2026 aerospace growth of 27% and a $1 billion order book are what conversion looks like, and with them right_to_win reads PASS and this note reads TRACK. The half-year numbers that proved it existed six weeks after this document's date and appeared in the very next filing. A process that waits for the red herring before killing a borderline draft would have caught it; the market, at 102 times subscribed, did.
A DRHP-stage assessment of the Updated Draft Red Herring Prospectus - I dated 30 September 2025, run as a held-back test: ratings were frozen on the filing plus outside checks dated on or before that day, marked where used, and the closing section reports the later record without moving any rating. No price band existed in the draft; audited numbers stop at 31 March 2025. Numbers carry source tiers: (T1) the filing's audited and certified sections, (T2) exchange, regulator or company primary records, (T3) the issuer-commissioned industry chapter or press, DERIVED where computed from cited inputs, UNVERIFIED where nothing supports them. Not a recommendation.