Interarch is a mid-cycle, asset-light, net-cash, capital-efficient proxy for the structural shift of Indian construction from concrete to engineered steel — specifically in the pockets where that shift is most economic (data centres, fabs, gigafactories, large warehouses). It’s the #2 player in an industry where the top seven are taking share from an unorganised tail that used to dominate. The business compounds earnings not through margin expansion (capped by steel pass-through at ~10%) but through volume, capacity additions, and repeat customers. At ~21x FY26E / ~18x FY27E it’s the cheapest of the organised quality cohort and the only one that’s genuinely debt-free. The bull case pays you for data-centre optionality you’re getting free today; the bear case is a normal cyclical down-leg plus sharper competition from EPack. Both live inside the range of expected outcomes.
Valuation — what you’re being asked to pay
At ~₹1,990, market cap is ~₹3,340 Cr.
| Multiple | Current (TTM) | FY26E | FY27E | Comment |
|---|---|---|---|---|
| P/E | ~24x | ~21x | ~18x | FY27E EPS ~₹110 assuming 12–15% growth, stable margin |
| EV/EBITDA | ~16x | ~14x | ~12x | Below 3Y average; EV adjusted for net cash |
| P/B | ~4.2x | ~3.8x | — | Justified by 25% ROCE via Gordon Growth math |
| Market cap / sales | ~1.8x | ~1.75x | ~1.6x | Cheap vs capital goods median |
Peer context: EPack Prefab trades at ~34x P/E, Pennar at ~27x, Interarch at ~24x. On a PEG basis, with 20%+ 3-year earnings CAGR, Interarch is the cheapest of the high-quality cohort.
The ROCE–WACC spread is where the value lives. With ROCE at ~25% and WACC probably 12–13% for a debt-free mid-cap, every incremental rupee of invested capital creates ~₹0.13 of value per year. The capex programme — adding 65k MTPA at ~₹150 Cr — should generate ~₹500 Cr of incremental revenue at 9% EBITDA margin, i.e. ₹45 Cr of EBITDA for ₹150 Cr capex. That’s a ~30% incremental ROCE before tax. As long as utilisation comes through, the spread stays intact.
Investment thesis
India is quietly doing something unusual: it’s pouring concrete — or rather, bolting steel — at a pace the country has never seen. Semiconductor fabs in Assam. Battery gigafactories in Gujarat. Hyperscale data centres in Greater Noida. Amazon-scale warehouses off every expressway. Every one of these projects has one thing in common: it doesn’t want to be built with brick and RCC. It wants pre-engineered steel — faster, cheaper, lighter, and delivered in six months instead of eighteen.
Interarch Building Solutions sells exactly that. It’s the second-largest integrated pre-engineered building (PEB) company in India, it calls itself the “Mercedes of the PEB industry”.
This post is about whether the stock deserves the premium, what could rerate it, and what could break it.
What Interarch actually does
Think of Interarch as a factory-born building contractor. A client — say, Tata Projects building a semiconductor fab — needs a 50,000 sqm column-free shed. Interarch’s engineers design every beam in-house, its plants fabricate the steel frames, its trucks ship them, and its crews bolt them together on site. Design → engineer → manufacture → erect. All under one roof. Turnkey.
The revenue splits into three buckets (FY25):
- PEB contracts (~84%) — end-to-end turnkey projects. The bread and butter.
- PEB sales (~15%) — supplying the steel structures while the customer handles erection. Higher margin, lower ticket.
- Other (~1%) — metal ceilings (TRAC®), roofing (TRACDEK®), light-gauge framing.
By end-use: 77% industrial/manufacturing, 21% infrastructure, 2% buildings. This is a near-pure play on India’s private-sector capex cycle.
Scale snapshot: ~200,000 MTPA installed capacity across five plants (Pantnagar and Kichha in Uttarakhand, two in Tamil Nadu, Athivaram in Andhra Pradesh). Going to ~265,000 MTPA by Q2 FY27 with a new Gujarat plant and an AP heavy-structures line.
The industry map
The Indian PEB market is roughly ₹21,000 crore in FY25, projected to hit ₹34,500 crore by FY30 — ~10–12% CAGR. But the more interesting number is the mix shift: the organised segment has gone from 32% of the market in FY18 to 44–45% in FY25, and the top seven players hoover up 80–85% of that organised pie.
Here’s how the focal firm sits inside its neighbourhood:
| Company | Installed Cap. (MTPA) | Revenue (FY25, ₹ Cr) | ROE (FY24) | EBITDA Margin | Notes |
|---|---|---|---|---|---|
| Kirby | 300,000 | ~2,400 | 46% | ~10% | Market leader; Kuwait-owned |
| Interarch | 200,000 (→265k) | 1,454 | 18% | 10% | Net cash, turnkey |
| EPack Prefab | 134,000 | 1,134 | 23% | 10.4% | Fastest grower, just IPO’d |
| M&B Engineering | 104,000 | ~900 | — | — | 75% share in self-supported roofing |
| Zamil Steel | 100,000 | — | — | — | Saudi-owned |
| Pennar Industries | 90,000 | 3,226 (conglomerate) | 13% | 9.6% | Diversified steel |
| Everest Industries | 72,000 | ~1,400 | negative | 1.7% | Turnaround story |
Two observations that matter:
- The top three are pulling away. Interarch and EPack both sport 20%+ ROCE; Pennar and Everest don’t. When large industrial clients place a ₹100+ crore order, the cost of a delay far exceeds any price saving from the unorganised tail.
- There is no patent moat. PEB engineering is applied — not frontier — science. The moat is soft: brand recall, a 677-contract track record, and the reputational capital that gets you invited to the shortlist. That’s why Interarch has 80–85% repeat-order ratio
The moat
Three durable reasons:
1. Switching costs hide in project timelines. Once a client trusts Interarch to deliver on a Tata Motors JLR plant, giving the next one to a cheaper unknown is a firing offence. Hence 80–85% repeat.
2. Capital efficiency is genuinely differentiated. Interarch generates ~₹450–500 crore of revenue off just ₹50–60 crore of capex. That’s ~8–10x asset turn at the capacity level — possible only because the business is really engineering + coordination, not heavy steel-making. Steel is bought on a pass-through basis from Tata, JSW, SAIL, AMNS.
3. Vertical integration compresses the bid-to-bolt cycle. Design, fab, and erect under one corporate roof shaves weeks off execution. In a market where clients pay premium for speed (data centres literally race AI demand curves), the turnkey model is the product.
What isn’t a moat: the core PEB steel structure itself. Ten unorganised fabricators in Tier-3 cities can cut I-beams. The moat is everything around the beam.
The numbers that matter (9M FY26 and the trajectory)
9M FY26 (Q1+Q2+Q3 to Dec 2025):
| Metric | 9M FY26 | 9M FY25 | YoY |
|---|---|---|---|
| Revenue | ₹1,395 Cr | ₹990 Cr | +41% |
| EBITDA | ₹124 Cr | ₹87 Cr | +42% |
| EBITDA margin | 8.9% | 8.8% | flat |
| PAT | ₹97 Cr | ₹69 Cr | +41% |
Q3 FY26 alone was the best quarter in company history — ₹523 Cr revenue (+44% YoY), ₹50 Cr EBITDA (+43%), ₹37 Cr PAT (+32%). The PAT growth lagged EBITDA because of a one-time ₹3.24 Cr statutory hit from the new Labour Codes and modest tax-rate drift. Scrubbed, the quality of the quarter is real.
Historical trajectory (₹ Cr):
| FY21 | FY22 | FY23 | FY24 | FY25 | TTM (to Dec’25) | FY26E (guidance) | |
|---|---|---|---|---|---|---|---|
| Revenue | 576 | 835 | 1,124 | 1,293 | 1,454 | 1,858 | ~1,900 |
| EBITDA | 14 | 37 | 111 | 120 | 144 | 172 | ~175 |
| EBITDA % | 2% | 4% | 10% | 9% | 10% | 9% | ~9% |
| PAT | 6 | 17 | 81 | 86 | 108 | 137 | ~140 |
| ROCE | 5% | 10% | 31% | 29% | 25% | ~25% | — |
Source: Screener.in, company filings
Two things jump out:
- The FY21 → FY25 inflection is dramatic: revenue 2.5x, PAT 18x. The business genuinely repriced after COVID as industrial capex returned and organised players took share.
- Margins are range-bound around 8–10%. This is a fabricator, not a software business. Operating leverage is real but capped by steel pass-through. The 10% target management keeps nudging toward is the ceiling, not the floor.
Order book: ₹1,685 Cr as of Jan 31, 2026 — the highest ever, up from ~₹1,300 Cr in Aug 2024. Near-term pipeline (P1) is ₹1,200 Cr at a 20–22% hit rate. That’s 12–18 months of revenue visibility.
Balance sheet: near-zero debt (₹11 Cr borrowings vs ₹800 Cr reserves), ₹51 Cr in investments, current ratio 2.1x. A ₹100 Cr QIP is in the works — growth capital, not rescue capital.
Why it could rerate: five triggers
- Data centres are an asymmetric call option. India’s hyperscale DC capacity is projected to double to ~2 GW by 2026 — and every MW of DC is ~10,000 sqft of column-free, airtight, seismic-tolerant steel shed. Interarch has already won work from a Greater Noida facility and management flagged data centres as the primary growth driver. This isn’t in FY26 numbers yet.
- Semiconductor + battery + EV capex is just starting. Interarch has ₹220 Cr of orders from Tata Projects for the Assam fab and the Agratas Gujarat gigafactory. If India’s semi/battery localisation plays out at anything like Korea or Taiwan’s intensity, this is a 5-year tailwind.
- Capacity ramp is ahead of the market. Going from 161k → 265k MTPA (+65%) over FY25–FY27, with revenue per MT roughly ₹80–90k. That’s ₹2,000+ Cr of revenue capacity being installed. Management openly talks about building toward a ₹3,500 Cr revenue base. Classic early-cycle capex bet.
- Exports are a zero-in-the-model optionality. Q3 FY26 was the first meaningful export quarter — ₹13 Cr of orders. Management is opening international sales offices. Kirby (the leader) already exports meaningfully; if Interarch captures even 10–15% export mix over 3–4 years, the margin mix upgrades materially.
- The JSW / Jindal Steel partnership for high-rise steel. Still nascent, but the structural pitch — making steel the preferred material for multi-storey commercial, data centres, heavy structures — is a genuine market-creation play. India’s multi-storey steel penetration is <5% vs 60%+ in the US.
Why it could disappoint: the inverse case
A good thesis is only as strong as its adversarial version. Here’s what breaks it:
- Building capacity for ₹3,500 Cr revenue while doing ₹1,900 Cr is a bet. If the private capex cycle rolls over — and capex cycles always roll over — Interarch is left with fixed overheads eating margins. Management itself has guided FY27 growth at 12–15% (down from FY26’s 30%+). That decel is already telegraphed.
- Fixed-price contracts + volatile HR steel = margin whiplash. Steel is 55–60% of COGS. Interarch has hedged better than peers (witness stable 8–10% margins through the FY22–FY25 steel cycle), but a sharp HR coil spike with a locked-in order book can erase a quarter.
- EPack Prefab is growing faster. EPack has clocked a 46% revenue CAGR FY22–FY25 vs Interarch’s 15%, with 22.7% ROE vs Interarch’s 18%. It’s now post-IPO, well-capitalised, and openly gunning for 10% market share. This is the sharpest competitive threat — and it’s not priced as a threat.
- Margin pressure is explicit in the guidance. Management has flat-out said near-term margins won’t expand — higher employee costs, new plant overheads, and export/heavy-structure investments offset operating leverage.
- Working capital is quietly elongating. WC days went from 21 (FY24) to 51 (FY25). Cash from operations / operating profit fell to 60% in FY25 from 93% in FY24. Larger orders mean longer receivables. This is normal but not free — it’s why free cash flow turned negative in FY25 despite record profits.
