Torrent Power — The Discom Crown Jewel Goes On a Capex Spree
Torrent Power Limited
Torrent Power — The Discom Crown Jewel Goes On a Capex Spree
1. Snapshot
Torrent Power is one of India’s few genuinely integrated private utilities — it generates power (gas, coal, renewables), transmits it, and crucially distributes it through licensed and franchised discoms that are among the best-run in the country. Market cap ₹72,481 Cr, current price ₹1,438, 52-week range ₹1,188–1,824, trading at a P/E of ~30 and ~3.8x book. This is not a regulated annuity bond and not a pure merchant developer — it’s a stable distribution annuity (the crown jewel) wrapped around a generation portfolio that is now being aggressively expanded into an ~₹80,000 Cr, five-year renewables-thermal-storage capex animal. As of 2026-06-20, from screener snapshot.
2. Business & position in the value chain
Torrent sits across the whole electricity value chain, but its identity is distribution. It is a licensed distributor in Ahmedabad, Surat, Dahej SEZ, Dholera, Dadra & Nagar Haveli, and Diu & Daman, and a franchised distributor in Agra, Bhiwandi, Shil, Mumbra and Kalwa. These discoms are the reason Torrent is interesting: Indian power distribution is where almost every other utility bleeds — state discoms run aggregate technical & commercial (AT&C) losses in the high teens and survive on subsidies. Torrent’s licensed networks run AT&C losses around 2.5% (per the concall exchange), world-class numbers that translate into a regulated, low-risk, growing annuity stream.
The rest of the mix as of 31 March 2026:
- Thermal generation: 2.7 GW gas + 362 MW coal operational. A 1,600 MW coal plant in MP is under construction; the 1,400 MW Nabha Power (coal) acquisition closes Q1 FY27, taking total capacity from ~5.1 GW to ~6.4 GW.
- Renewables: ~2 GW operational (solar + wind, roughly balanced), with ~4 GW under implementation (~50/50 wind/solar, ~₹28,000 Cr capex).
- Pumped storage (PSP): 3 GW under construction (~₹14,000 Cr), part of a much larger multi-GW pipeline — the grid-balancing play as renewables scale.
- Transmission: ~952 ckm at 400 kV; Khavda transmission line commissioned in Q4 FY26; Solapur under construction.
- Gas + green hydrogen: Torrent Gas (group CGD) shares the LNG procurement book; 18 KTPA green-hydrogen/ammonia awarded under PLI — early-stage optionality.
Revenue is geographically concentrated in Gujarat (the distribution heartland) plus Maharashtra/UP for the new generation projects. Roughly half the gas plants are tied to its own discoms on long-term PPAs (fixed-cost recovery on availability), which de-risks the otherwise-ugly economics of gas generation. The strategic logic is elegant: a cash-generative regulated distribution base funds a growth pivot into renewables and storage.
3. Management & promoter quality
Promoter is the Mehta family (Sudhir and Samir Mehta) via the Torrent Group / Torrent Investments structure, the same family behind Torrent Pharmaceuticals. This is, by Indian standards, a high-quality, low-drama promoter group — Ahmedabad-based, conservatively run, professional CEOs under family chairmanship, with a multi-decade track record of building rather than financial engineering. (Torrent Group — Wikipedia; Business Standard)
Promoter holding is 51.09% (down from 53.56% — the family trimmed ~2.5% in late 2024, partly to fund group initiatives). That is a mild signal to keep an eye on, but it is still majority control, and the sale was disclosed, orderly, and modest. There is no pledging flagged in the snapshot, and the family has historically funded the power business with group resources rather than over-leveraging it. DII holding has steadily risen to 22.8% while FIIs sit at 8.4% — domestic institutions are voting with their feet for the integrated story.
Capital-allocation track record is the key judgment here, and it is genuinely good: Torrent has compounded book value (₹379/share now), maintained a healthy dividend (38–42% payout, 1.39% yield), and historically built distribution and generation assets that earn their cost of capital. The concern, stated plainly, is that the company is now stepping decisively out of its comfort zone — an ~₹80,000 Cr capex plan across coal, renewables, PSP and an acquisition is a different animal from incrementally upgrading the Ahmedabad grid. Management on the call was confident the under-leveraged balance sheet (net D/E 0.67, net debt/EBITDA ~2.06) has ample room, and that committed capex runs to FY31–32 with EBITDA “falling off significantly” as projects commission. Governance and disclosure on the call were straight — they distinguished one-offs cleanly and refused to give capex guidance they couldn’t stand behind. Even-handed verdict: strong promoter, proven allocators, now being tested on a much larger and more cyclical canvas (coal + merchant gas + auction-tariff RE).
4. Financial trends
| Metric | Reading | Note |
|---|---|---|
| ROE | 13.2% | Below the ~15% regulated cap — diluted by under-utilised gas plants & growth capex not yet earning |
| ROCE | 14.0% | Healthy for a utility; should hold as distribution incentives kick in |
| OPM | ~18–19% | Recovered from the FY23–24 dip (19→17→18→19%) caused by the merchant-gas/trading mix |
| Revenue (FY26) | ₹28,966 Cr | Roughly flat YoY (₹29,165 Cr FY25); the GST-era jump was FY22→23 (₹14k→₹26k) on trading/franchise consolidation |
| Net Profit (FY26) | ₹2,469 Cr | Down from ₹3,059 Cr FY25 — but FY25 was flattered by a ₹637 Cr non-cash deferred-tax reversal |
| Debt | Net D/E 0.67, net debt/EBITDA 2.06 | Under-leveraged for a utility entering a capex supercycle |
| Dividend payout | 42% (FY26) | Consistent; the screener “pro” |
The honest read: ROE at 13.2% is the weak spot, and it’s structural, not accidental — the gas fleet runs at 30–35% PLF (you can’t make a gas plant in India hum at LNG prices), and a large slug of growth capex is in-flight but not yet generating. The distribution business, by contrast, is quietly getting more profitable: a new tariff regime (effective 1 April 2025) shifted post-FY25 assets to a return-on-capital-employed basis with base ROE 13% rising to 15% on incentive milestones, and — crucially — extended performance incentives to pre-FY25 assets too, lifting effective discom ROE from 14% toward 15%. Management quantified this as roughly a +1% ROE uplift annually on the distribution book if they hit milestones (smart meters, network availability, collection efficiency, T&D losses). They also cut the regulatory gap by ~₹800 Cr this year on better efficiency and lower power-purchase costs.
Screener’s two flags: the “pro” (38.6% dividend payout) is real and reflects discipline. The “con” (3.8x book) is the whole valuation debate — see §7. Don’t read 3.8x as a red flag; read it as the price of the distribution quality.
4-note (lens).
Torrent is a hybrid, so judge it in layers: the distribution book on a regulated-utility lens (RoE capped ~15%, so the levers are capex growth, AT&C losses already near-best-in-class, receivable cycle, smart-meter/collection incentives — not multiple expansion); the renewables build on a developer lens (capacity pipeline, PPA tariffs, funding cost, project IRRs — and Torrent is leaning toward complex hybrid/RTC bids for better tariffs); the gas fleet on a merchant-plus-PPA lens (fixed-cost recovery on availability for the ~50% tied to own discoms, optionality on merchant peak sales for the rest). The crown jewel deserves the regulated lens; the growth engine deserves the developer lens; don’t blend them into one number.
5. Latest quarter
Q4 FY26, reported 12–13 May 2026. Reported PBT ₹547 Cr vs ₹619 Cr a year ago — but that headline is noise. It includes a ₹171 Cr non-recurring provision for a capping of UNOSUGEN power-purchase cost in the FY24-25 true-up order. Management noted an identical cap was contested and reversed in their favour last year (a ₹273 Cr credit), so they’re “reasonably certain” this reverses too. Adjusted PBT ₹718 Cr, +16% YoY — the cleaner number. (TradingView/Quartr summary)
What the segments did:
- Distribution was the engine — ₹186 Cr of carrying-cost income on regulatory assets (management insists this is a recurring feature, not a one-off), plus the new tariff-regime ROE/ROCE uplift and a ₹58 Cr solar-rooftop incentive.
- Thermal EBITDA fell ~₹90 Cr (ex the one-off) on scheduled gas-plant maintenance and a prior-year provision reversal not repeated.
- Renewables EBITDA was flattish-to-down (~₹186 Cr) despite capacity additions — because a generation-based incentive (GBI) on one project expired. A genuine, if mechanical, drag.
Capex actually spent in FY26: ~₹1,600 Cr distribution, ₹550 Cr transmission, ₹700 Cr coal, and a ₹6,500 Cr chunk in renewables — the giveaway that the pivot is real and funded. On FY27 capex, management would only say “much, much higher.” Notable quote on the discom incentive math, when an analyst pointed out AT&C losses are already 2.5%: “It’s not only T&D losses… it’s network availability, smart meters, collection efficiencies.” — i.e., there’s still incentive runway even on a best-in-class book.
6. What’s happening now
The live wire is the capex supercycle plus the Nabha acquisition:
- HARD — Nabha Power: CCI cleared the ₹6,889 Cr (EV ~₹6,800 Cr; ~₹3,400 Cr debt + ~₹3,400 Cr equity) acquisition of L&T’s 1,400 MW Nabha coal plant. Consolidates from Q1 FY27, lifting capacity ~5 GW → 6.4 GW. Two-part tariff, variable cost pass-through. The stock hit a 52-week high on the approval. (Whalesbook — CCI nod)
- HARD — generation milestones: 1,600 MW MP coal project has PPA executed with MPPMCL, BTG letter of award issued, environmental clearance received. 3 GW PSP has the MSEDCL storage agreement signed, LoAs out, EC received. Khavda transmission line commissioned. 367 MW MSEDCL solar progressively commissioned (some delayed on a government land-replacement issue).
- SOFT — the ₹80,000 Cr / 5-year plan: ~₹28,000 Cr renewables (4 GW), ~₹23,000 Cr coal (1.6 GW), ~₹14,000 Cr PSP (3 GW), plus ~₹2,000 Cr/yr distribution. Management confirmed this framing on the call. (SolarQuarter)
- Renewables phasing: 1.2–1.4 GW expected to commission in FY27 after a lacklustre FY26 (only ~250 MW added) — execution is the thing to watch.
- Gas / geopolitics: the US–Iran flare-up briefly spooked LNG, but Torrent sourced its 3 summer cargoes from non-Strait-of-Hormuz origins; the BP/JERA long-term contracts (~10 cargoes/year split with Torrent Gas, ~5 for power) start calendar 2027. Management is relaxed on availability, cautious on price.
Sector tailwinds it rides: the discom-reform / smart-metering / RDSS-era push to single-digit AT&C losses (Torrent is already there and earns incentives for it — Mercom), India’s ~6% power-demand CAGR, and the renewables-plus-storage build-out. Headwind: a global gas glut is coming (pushed out slightly by the Iran war), but until it lands, the gas fleet stays a low-PLF drag.
7. Expectations baked in
At ~₹1,438, Torrent trades at ~30x trailing P/E (TTM ~24x on cleaner estimates) and ~3.8x book — a clear premium to regulated PSU peers and roughly in line with the other private growth utilities. Peer frame (TTM, approximate): NTPC ~14.7x P/E / 1.85x P/B; Tata Power ~29x / 3.15x; JSW Energy ~36x / 2.83x; Torrent ~24x / 3.76x. (Goodreturns peer set; BusinessToday — Elara)
So the market is not pricing Torrent as a sleepy regulated utility (that’s NTPC’s 14x). It is pricing it as a growth compounder — paying ~3.8x book against a current ROE of only 13.2%, which is mathematically demanding. The implied bet is that ROE inflects upward (distribution incentives + the ₹80,000 Cr capex commissioning at decent IRRs by FY28–31) and that the renewables/PSP book gets the market to value those assets like a developer rather than a utility. Note the sell-side split: JP Morgan rates it neutral and Elara has flagged Torrent (with JSW Energy) as overvalued, preferring regulated PSU names. (Business Standard)
Translation: at 3.8x book on 13% ROE, you are paying today for execution that hasn’t happened yet. The distribution crown jewel justifies a premium; whether it justifies this premium depends entirely on the capex landing on time and on return.
8. Rerating signals — up vs down
| Could re-rate UP if… | Could re-rate DOWN if… |
|---|---|
| The ₹80,000 Cr capex commissions on time and at promised IRRs, and ROE inflects from 13% toward 15%+ as projects start earning | Execution slips — RE phasing slides again (FY26 added only ~250 MW), PSP/coal milestones delay, and capex earns no return for years while interest & depreciation bite |
| Distribution incentives (smart meters, collection, availability) deliver the +1%/yr ROE uplift and the regulated book keeps compounding | The market de-rates the multiple toward NTPC-style 1.8–2x book, judging it a utility not a compounder (Elara’s “overvalued” thesis plays out) |
| Nabha + new thermal add cheap, two-part-tariff EBITDA that lifts blended returns; market starts valuing the RE+PSP book on developer multiples | Gas stays expensive (Iran-driven LNG spike persists), gas PLFs stay stuck at 30–35%, and the merchant-gas optionality keeps disappointing |
| UNOSUGEN ₹171 Cr provision reverses (as the prior one did), and regulatory carrying-cost orders keep flowing | A capex supercycle this size strains the (currently comfortable) balance sheet if a project underperforms — net debt/EBITDA 2.06 has headroom but not infinite |
| Power-demand CAGR holds ~6%, merchant/peak power tightens, and India’s RE+storage build re-rates the whole node | A new tariff order tightens distribution returns, or parallel-licensing competition (currently dormant) opens Torrent’s protected discom turf |
9. Conviction texture
The bull case, stated at its strongest: Torrent owns the single best thing in Indian power — a distribution business that actually works, with AT&C losses a tenth of the national average, a regulated annuity that compounds, and a tariff regime that’s just been tweaked to pay it more for being good. On top of that annuity sits a well-funded, under-leveraged balance sheet now pivoting hard into renewables and pumped storage — exactly where India’s electricity system has to go. Run by the Mehtas, who are about as clean and patient as Indian promoters get. If the ₹80,000 Cr lands, you have a 13% ROE business becoming a 15%+ ROE business at scale, and 3.8x book will look cheap in hindsight.
The bear case, equally strong: you are paying 3.8x book today for a 13.2% ROE — that math only works if the future is bright, and the future here is a coal-and-gas-heavy capex plan in a world drifting away from thermal, plus a renewables book that underbuilt this year and whose EBITDA went sideways despite additions. The gas fleet is a structural drag no one can fix (you can’t make Indian LNG cheap). Promoters trimmed their stake. And a capex supercycle is exactly when good allocators occasionally turn into empire-builders — the call’s repeated “there’s room for more projects” is either confidence or a tell, and we won’t know which for three years.
What the evidence actually supports: the distribution quality is real and verifiable (2.5% AT&C losses, the tariff-incentive mechanics, the ₹800 Cr regulatory-gap reduction). The capex ambition is real and disclosed (₹6,500 Cr already spent on RE in FY26 alone). The ROE inflection is promised, not yet delivered — and the renewables phasing miss this year is the first small crack to watch. The thing that breaks the tie, in either direction, is execution on FY27–FY28 commissioning: 1.2–1.4 GW of RE next year, the Nabha integration, and whether ROE actually starts climbing off 13%. Watch the segmental EBITDA on renewables (does new capacity finally show up in profit?) and the distribution incentive accrual (does the +1% materialise?). Get those two right and the premium is earned; miss them and 3.8x book is a long way to fall.
Sources: local screener snapshot (fetched 2026-06-20) and Q4 FY26 concall transcript (13 May 2026); web — TradingView/Quartr, SolarQuarter, Whalesbook (CCI/Nabha), Goodreturns (peer valuation), BusinessToday (Elara), Wikipedia (Torrent Group), Business Standard (promoter stake), Mercom (discom reform). No buy/sell/hold verdict.