| Origin | What it is | Who captures it | This report's evidence |
|---|---|---|---|
| 1. Engineered power equipment | Transformers, switchgear and grid technology sold against genuine, if diffuse, global manufacturing scarcity | Voltamp, ABB India and Hitachi Energy India show the clearest evidence; GE Vernova T&D and Schneider Electric Infrastructure show rich valuations without matching evidence | Real and quantified at some names, aspirational at others — §9's leftmost/middle columns |
| 2. Differentiated cabling | Certified, purpose-engineered fibre and copper products for AI-data-centre connectivity | Sterlite Technologies most concentrated; Polycab and KEI Industries more diversified but management-quantified at the sector level | The single most concentrated DC-revenue story in this report sits here, not in power equipment (§9) |
| 3. Engineered cooling | Purpose-built chiller and precision electro-mechanical systems for data-hall thermal management | Blue Star, Voltas and Thermax all hold real capability; this report's own evidence on margin durability here is genuinely mixed, not uniformly positive | Real capability, inconsistent financial results — see each company's own margin trend |
| 4. Commodity civil/EPC construction | Building the shell and executing competitively-bid transmission and construction contracts | Nobody durably — this is this report's clearest, most repeated finding of structural margin compression (KEC, Ahluwalia) | The mirror image of the sister report's assembly-layer finding: commodity execution compresses, regardless of how large the order book grows |
| 5. The invisible layer | Precision CRAC/UPS systems — the most visible brand names in data-centre power (APC, Vertiv) | Nobody listed on an Indian exchange — see §5's structural finding | A genuine gap in India's public markets, not a research oversight |
India's data-centre capacity is genuinely scaling — operational capacity crossed ~1,700MW in 2025 with hyperscale pre-committed capacity running well ahead of what gets counted as "operational" (§6, §8) — and this demand sits on top of, not separate from, an even larger national grid and renewable-integration buildout that several of this report's power-equipment companies are already benefiting from regardless of the data-centre cycle specifically. At least three companies in this report (Voltamp, ABB India, Sterlite Technologies) have management-confirmed, quantified evidence of real, current data-centre-driven demand — not analyst speculation. Several names carry stable-to-improving credit ratings (Polycab and KEI Industries both hold top-tier, multi-agency ratings) reflecting genuine balance-sheet strength behind the growth story, and at least one name (Kalpataru Projects International) shows margin actually improving against this report's broader compression finding.
This report's clearest, most repeated finding is structural margin compression in commodity civil/EPC construction — the sub-category with the most companies in this report's universe — regardless of order- book size, and management's own guidance at more than one name (KEC International) pushes any real recovery out to FY28 or later. The single most concentrated, best-quantified data-centre revenue story in the report (Sterlite Technologies) sits on top of financial disclosures this report's own research found genuinely unreliable, a combination that should give any reader pause rather than excitement. At least one name (Schneider Electric Infrastructure) trades at a valuation this report's own arithmetic cannot connect to either its current fundamentals or to which entity actually owns the brand equity investors likely associate with it. And the scarcity this report's bull case rests on is, by this report's own §7 finding, diffuse and slower-moving than the sister report's Nvidia-centric story — meaning it is also slower to resolve in either direction, for better or worse.
