Borderline, landed KILL. Re-check at first results.
Why:
- Amtech takes whatever price the market gives it, and the one advantage it might have had turns out not to exist. Last year it sold 16.8% more resin at a 3.66% lower price per tonne (T1). The obvious edge, sitting next door to its Delhi-area customers, is not real: this resin ships as a liquid in drums at about ₹120 a kilo, so makers in Gujarat and Maharashtra serve the north without difficulty, and one competitor headquartered in New Delhi, Revex Plasticisers, already runs about ten times Amtech's revenue across more grades (T3). The plant this IPO is really about needs 2.6 times last year's tonnage to run as full as the old one ever has, and it opens with no order book and no named customer (T1).
- Last year's margin gain was cheap raw materials, and they have already turned. All five petrochemicals Amtech buys fell through the year to March 2026 (T2 and T3), by more than enough to cover the whole 2.1 point gross margin gain on their own; styrene is now about 36% above its December 2025 low (T3). Strip that out and margins went backwards, roughly 16.3% against 17.6% the year before, on 10.3% sales growth (derived from T1). In the same year, money owed by customers stretched from 68 days to 92 and money owed to suppliers stretched with it (T1).
- ₹10.0 cr of company cash went to the managing director for his own pigment business, at a price he set himself. No independent valuer appears anywhere in the 680 pages of this document and the earlier draft of it (T1). The audited related-party note, the schedule whose job is to list exactly this, leaves the transaction out. It cannot be traced in that year's cash flow either: the line meant for it is nil and the group's cash went up, not down (T1). The company had no audit committee when it happened. What was bought is real and now earns about a quarter of revenue, but nothing in the file lets a reader test the price.
Valuation at the band
| Floor ₹71 (T1) | Cap ₹75 (T1) | |
|---|---|---|
| Bid window | 9 to 11 September 2026 | |
| Bid lot | 1,600 shares | |
| Shares after the offer | 88,29,168 | 88,29,168 |
| Market capitalisation | ₹63 cr | ₹66 cr |
| P/E on FY26 profit | 14.8x | 15.7x |
| EV/EBITDA (reported net debt) | 8.6x | 9.1x |
| EV/EBITDA (illustrative, net of the ₹4.0 cr earmarked repayment) | 8.1x | 8.5x |
| Promoter holding after | 36.3% | 36.3% |
The filing states there is no listed peer in this industry and no unlisted one of comparable size (T1), so the document offers nothing to read these multiples against. The price does not move the verdict either way; the six ratings judge the business and the offer, not what it costs.
The story
Resin is the engine. It is 63% of revenue, three quarters of last year's revenue increase, and the whole point of the raise, since the new plant makes resin. Tonnage grew 75% in two years on a plant whose certified size never changed, which is the best fact in this filing. The other quarter of the business, colour paste, was bought from the promoter in December 2023 and stopped growing last year.
What this business is
Amtech cooks five bought petrochemicals in a reactor at 210 degrees for eleven or twelve hours and blends in styrene to make unsaturated polyester resin: the liquid that sets hard when a fabricator mixes it with glass fibre. It sells 49 grades of it from one plant in Bahadurgarh, Haryana, to small workshops that make fibreglass sheet, cooling towers, switchgear housings, buttons, statues and car body panels. 86% of sales go to Delhi, Uttar Pradesh and Haryana (T1). A wholly owned subsidiary, Croda Pigments, makes colour paste in a shed next door rented from the promoter, and the company also buys in and resells hardeners, silicone and fibre resin.
The money is the gap between the petrochemical price it pays and the resin price it charges. Nothing is locked in at either end: no long-term supply contracts, no long-term customer agreements, only purchase orders either side can drop (T1). Customers are numerous and small, the largest at 9.8% of sales and the top ten at 40.8% (T1). Revenue was ₹40.7 cr in the year to March 2026, profit ₹4.2 cr, and sixty people work there, four of them in quality control (T1).
Easy or difficult business? Run-of-the-mill. This is one batch reactor, one blender and one boiler running a known recipe, on 68 KVA of sanctioned power at the main plant, which is a workshop rather than a chemical works (T1). There is no product approval to win, no customer qualification to pass, no patent, and no research spend or research employee anywhere in the company. The hard parts are commercial: buying five volatile petrochemicals well, mixing small batches to order, and getting paid.
Key numbers
| ₹ cr | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Revenue | 24.60 | 36.89 | 40.67 |
| Revenue growth % | 10.7 | 49.9 | 10.3 |
| Gross margin % | 23.2 | 27.6 | 29.7 |
| EBITDA | 1.65 | 6.50 | 7.49 |
| EBITDA margin % | 6.7 | 17.6 | 18.4 |
| PAT | 2.84 | 3.72 | 4.22 |
| PAT growth % | n/a | 31.2 | 13.4 |
Profit outran sales in two of the three years, for two different reasons. FY2025 was volume filling a half-empty plant, helped by two cost lines that fell in rupees while sales rose 49.9%, several of them one-off costs left over from the acquisition year (T1). FY2026 was input prices: cost per tonne of output fell about 6.4% while the selling price fell 3.66% (derived from T1 and T2). FY2024 profit also carries a ₹2.39 cr one-off gain on selling investments, so that year is flattered and the FY2025 profit growth shown here is understated (T1).
| Segment, ₹ cr | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Resin (UPR) | 14.96 | 22.71 | 25.56 |
| Resin growth % | n/a | 51.8 | 12.5 |
| Colour paste (Croda Pigments) | 5.25 | 9.80 | 9.82 |
| Colour paste growth % | n/a | 86.7 | 0.2 |
| Hardeners, silicone, fibre resin | 4.40 | 4.38 | 5.30 |
| Hardeners, silicone, fibre resin growth % | n/a | -0.4 | 20.9 |
The filing's two revenue splits do not agree in the latest year. The third line above totals ₹5.30 cr on the product table while the bought-and-resold total on the same page is ₹4.16 cr, a ₹1.14 cr gap the document does not explain (T1). On that second basis the bought-in basket has been flat for three years, which matters because the company presents it as the reason customers come to it.
Scorecard
| Block | Rating | Why |
|---|---|---|
| Right to win | MARGINAL | Amtech is a price taker: volume up 16.8%, price per tonne down 3.66% (T1). None of the four strengths the filing claims survives checking. The bundle of extra products it calls an integrated offer is bought in and resold, earns nothing, and is flat over three years (T1). The quality claim rests on a management-system certificate, with no product approval, no customer qualification and no rejection rate behind it. There is no patent and no research spend. The strongest unstated candidate, being the local supplier, is refuted rather than unproven: the product ships as a drummed liquid at about ₹120 a kilo, so distant makers serve the north, and a New Delhi rival sits in the same catchment on about ten times the revenue (T3). What keeps this off a FAIL is the one number nothing has dented: resin output of 1,255, 1,879 and 2,195 tonnes on a plant certified at the same 2,960 tonnes a year throughout (T1), about 32% a year against a market growing about 6.7% (T3), lifting share from roughly 0.95% to 1.46%. Customers are buying a lot more from this company and nobody, the company included, can say why. |
| Industry and TAM | PASS | India uses roughly 150,000 tonnes of this resin a year, growing about 6.7% by volume (T3). Amtech makes about 1.5% of it, the new plant adds about 3.2% of the market, and imports run at 41,155 tonnes a year (T1). There is room. |
| Financial momentum | MARGINAL | One burst year between two 10% years, and last year's margin gain was cheaper inputs rather than better operations, while customer credit stretched from 68 days to 92 (T1). |
| Risks, governance, RPTs | MARGINAL | A clean legal file and a clean audit opinion, but the largest related-party deal in the company's history has no valuation, no entry in the audited note and no cash trail (T1). |
| Promoter and cap table | MARGINAL | Promoters sell nothing and pledge nothing, and they held steady through the pre-offer churn; but 14.13% of the company changed hands nine days before the offer at prices that exist nowhere, and no outside investor has ever priced this equity (T1). |
| Offer structure | MARGINAL | All fresh money, nothing to selling shareholders and the plant capex is quoted and certified; but half the working-capital ask has no stated basis, the biggest object expands colour-paste capacity 2.6 times in the year that line grew 0.2%, and up to 35% of the raise is unnamed (T1). |
Watch out for
- The statutory auditors' CARO report could not be read for any year. It is not public. That is the one document that would show whether the related-party approvals required by law were followed for the ₹10.0 cr payment, so the question stays open rather than answered either way.
- The new plant is due to run from September 2026 without three approvals it needs. The factory licence, the stability certificate and the fire clearance have not been obtained, and they are not on the filing's own list of pending approvals (T1). The site also has no disclosed grid connection, only a 58.5 KVA diesel generator.
- Lists in this document should not be read as complete. Four of them are provably not: the audited related-party note misses the ₹10.0 cr transaction, the cash flow's dedicated line for it is nil, the statutory list of shareholders above 1% dated 28 August leaves out four holders who are each above 1% on the company's own table, and the document cross-references a financial annexure it does not contain (T1). Take the numbers at face value; do not take any "these are all of them" claim at face value.
- Nobody knows what the pre-offer buyers paid. About 14% of the company moved between private holders in the week before the offer document, in a filing that prices every other transfer going back years (T1). Off-market transfers in an unlisted company are not registered publicly anywhere, so this cannot be found out. Outside checks on the buyers who can be identified found no link to the promoters, with one company's record unobtainable.
- A promoter-group company in the same business, Bawa Resins, holds 9.1% after the offer and has reported no revenue since FY2023. The non-compete with it was signed only in May 2026, after all three reported years. Its own accounts are not on the public record, so its earlier size is UNVERIFIED; the related-party schedule shows no dealings with Amtech in the reported years (T1).
The offer
- Raising ₹16.9 cr to ₹17.9 cr at the band, all of it fresh money into the company, with nothing going to any selling shareholder (T1).
- For lending ₹8.8 cr to the colour-paste subsidiary (₹3.4 cr of plant, ₹5.4 cr of working capital) and repaying ₹4.0 cr of borrowings, with the rest, up to 35% of the raise, going to unnamed acquisitions and general purposes. There is no monitoring agency and no bank appraisal, and the document lets management move money between the objects at its own discretion (T1).
- Implied valuation is in the table above: ₹63 cr to ₹66 cr.
- Promoters hold 49.7% before the offer and 36.3% after, and sell none of it; with the wider promoter group the holding is 45.5% after (T1). No promoter shares are pledged.
Assessment from the RHP dated 31 August 2026, which supersedes the DRHP of 30 May 2026, with outside checking of the market size, the raw-material prices, the named competitors and the subsidiary's corporate record. The price band and offer dates come from the NSE and BSE public records. Numbers carry source tiers: (T1) the filing's audited and certified sections, (T2) exchange or registry data and listed-company results, (T3) trade research and press, UNVERIFIED where nothing supports them. Not a recommendation. The valuation section states what the announced band implies and is not a view on whether that price is right.