Power Grid — Three Columns on the Wire Monopoly

A note on framework: this is the first piece using Vishal Khandelwal’s 15-question template — three columns of five, business / management / price, with a check question at the bottom of each that can veto the trade. The point of the template is not to land on a verdict. It is to make visible where conviction is real and where you are pretending. Power Grid is a useful first test because the template surfaces three different textures of conviction in one company.

Prelude — Bhadla, 4:14 a.m.

The largest solar park in India sits in the Thar desert in Rajasthan, near a village called Bhadla. Two thousand two hundred and forty-five megawatts of installed capacity, spread across fourteen thousand acres of low scrub. At 4:14 in the morning the panels are still cold, the inverters are humming faintly, and every electron the park is about to generate over the next twelve hours has the same problem: nobody in Rajasthan is going to buy it. Rajasthan does not need that much power. The load — the demand for electricity, the AC compressors and the steel mills and the data centres — is in the north and the west, somewhere between Delhi and Mumbai, a thousand kilometres away.

For that power to reach a load centre, it has to travel along a wire. A specific wire, owned by a specific company, charged at a specific rate. The wire in question is the Northern Region Strengthening Scheme, a 765 kV high-voltage corridor commissioned over the past decade. The company in question is Power Grid Corporation of India Limited. The rate is set by the Central Electricity Regulatory Commission in Delhi, in a tariff order that runs to several hundred pages and is revised every five years. A regulated 15.5% return on equity, post-tax, on whatever the wire cost to build.

That is the business. India’s energy transition — the entire net-zero arithmetic, every gigawatt of solar that gets switched on between now and 2030 — runs on copper and steel and aluminium that someone has to put up between the desert and the city. The someone is, in roughly 85% of the inter-state transmission cases, Power Grid. The wires earn their tariff whether the panels in Bhadla are running flat out or cloudy-day idle, whether the buyer is a financially distressed state distribution company or a private cement plant. Tariff first, generation second.

That is what makes Power Grid an interesting test for the template. Khandelwal’s framework is built for active, contested businesses where management makes high-stakes capital allocation choices and the market is constantly trying to price uncertainty. Power Grid is almost the opposite — a regulated monopoly where the cash flows are statutory, the management is a state-owned career bureaucracy, and the price discount-or-premium has more to do with the ten-year G-sec yield than with anything Power Grid itself does. Holding it up to the template’s three columns shows you exactly that mismatch — and that is the useful part.

Column 1 — Business

Q1. What does this company actually do?

Power Grid owns and operates roughly 85% of India’s inter-state electricity transmission system. About 1,77,000 circuit-kilometres of high-voltage transmission lines (220 kV, 400 kV, 765 kV, and the HVDC corridors that run at ±800 kV). About 280 substations. A handful of telecom adjacencies riding on the same towers — over 75,000 km of optical fibre — and a smattering of consultancy contracts in Africa and South Asia. The 15-year-old test: it owns the wires that move electricity between states. Power producers and discoms pay it to use the wires, and the government tells it how much it can charge.

What it is not: it is not a generator. It does not own power plants, it does not buy or sell electricity, it does not have any view on whether coal or solar or nuclear wins. It is a tollbooth. Every electron in India that crosses a state border flows through, more often than not, a Power Grid wire — and Power Grid earns its tariff regardless of who generated the electron and who bought it.

Q2. How do they make money, and from whom?

The revenue model is the cleanest part of the business. CERC (Central Electricity Regulatory Commission) sets the tariff every five years using a cost-plus formula: capital cost recovery + 15.5% post-tax return on the regulated equity in each project + actual operations and maintenance expenses + depreciation. Roughly 90% of revenue is locked in this way. The bills are paid by the central transmission utility pool, which collects from state discoms and state transmission utilities under a uniform Point-of-Connection (PoC) tariff. The mechanism is mutualised — Power Grid does not chase individual discoms, it gets paid out of a pool — which is why even a heavily distressed state can default upstream without immediately starving Power Grid.

That said: counterparty risk is real and chronic. The discoms are perpetually thin on cash. The Late Payment Surcharge rules brought in around 2022 imposed real consequences on discoms that don’t pay on time, and Power Grid’s receivables days have improved from the 80–100 day range into the 50–60 day range. Chronic, not acute. The actual default risk is borne mostly by the central pool, not by Power Grid’s P&L.

The non-tariff revenue — telecom rentals, consultancy, smart metering through the Power Grid IntelliSmart joint venture, an interest in a few solar EPC and battery storage projects — is a rounding error today, perhaps 5–7% of revenue. It is interesting because it is the only place where Power Grid is actually competing for business rather than collecting a regulated tariff. Watch this number, but don’t size a thesis on it.

Q3. Am I pretending to understand this, or do I really?

This is where the template earns its money. I understand the regulated-utility frame — build asset, recover cost, earn 15.5% on equity, pay dividends. That much is genuinely solid.

What I half-understand: the slow structural shift from RTM (Regulated Tariff Mechanism — the cost-plus regime above) to TBCB (Tariff-Based Competitive Bidding — projects awarded by reverse auction to the lowest bidder, with the winning tariff fixed for 35 years and no cost-plus protection). Roughly two-thirds of new transmission capacity in India is now awarded under TBCB. Power Grid wins a chunk of these auctions, but it wins them by bidding lower than the 15.5% RoE the cost-plus regime would have given it. Which means the average return on new equity invested — the marginal RoE — is below the headline 15.5%. The blended RoE on the existing rate base stays in the mid-teens because the older RTM assets are still earning under the old rules. But every year, a bigger slice of the capex sits in TBCB-priced assets, and the long-run RoE gravitates downward toward whatever the auction market clears at, which the latest few cycles suggest is somewhere in the 11–13% range.

That is the structural question that takes more than a quarterly call to answer, and it is where I would not pretend to have full conviction. If you forced me to articulate the next three sharp questions I would ask the CMD, they would be: what fraction of FY26-29 capex is RTM vs TBCB, and is that mix moving towards TBCB faster than the official guidance suggests? At what blended RoE does the equity story stop working as a yield-plus-growth thesis? And how much of the recent TBCB win-rate is being subsidised by Power Grid’s lower cost of capital versus a private bidder, and is that gap durable?

I don’t know those answers cold. I know enough to ask the questions, which is short of conviction.

Q4. Would I be happy owning this for ten years if the market shut tomorrow?

This is where Power Grid actually shines. Two reasons. One: Indian electricity demand grows in a structural 5–7% band — air conditioning penetration, EV charging, data centres, the slow grind of urbanisation. The transmission grid has to grow with it. Two: the renewable buildout makes the grid grow faster than demand. A coal plant sits next to its load. A solar park sits in Rajasthan or Gujarat or Andhra and has to send its power somewhere else. Every gigawatt of renewable that comes on requires more transmission per megawatt than a coal gigawatt did. The Green Energy Corridors I and II and the next tranche announced for FY26 onwards are all Power Grid territory.

If you locked the market for ten years and made me hold this, I would expect the dividend (currently in the 4.5–6% yield range, payout 50%+ of net profit) to compound, the rate base to grow at perhaps 8–10% annually, and the RoE on existing equity to stay in the mid-teens. That is a low-volatility 12–14% total return. Not exciting. Not bond-like either. Boring is the word. Boring compounds.

Q5. What could kill this business?

Three things, ranked by probability over a decade.

One — and this is the slow killer rather than the dramatic one — distributed generation plus storage at a price point where long-distance transmission becomes economically inferior to local solar-plus-battery. Today the LCOE math still favours utility-scale solar wired through transmission. By 2035 or 2040, if battery costs fall another 70%, the math could flip for some load segments. Not a five-year risk. A fifteen-year question. Power Grid would adapt — it would lean into intra-state distribution work, smart meters, battery storage adjacencies — but the core moat would erode.

Two — a CERC tariff order that materially cuts the regulated RoE below the current 15.5%. The next regulatory cycle (2024–29 norms are now in effect; 2029–34 will be the next reset) is the moment to watch. The 2024 order trimmed some incentives but left the headline 15.5% intact. A future order that pushes it to 12% would re-rate the entire stock by 20–30% overnight. Not a base case. Not impossible.

Three — a concentrated DISCOM crisis that breaks the central pool mechanism. The pool has weathered every previous discom cycle, including the UDAY mess and the 2020 Covid receivables spike. But the mechanism assumes discoms eventually pay. A genuine cascading default in two or three large states could test that.

Notice that none of the three are short-term plot risks. The business does not have a quarter that breaks it. That is part of the appeal and part of the boredom.

Check — Do I truly understand this business? Mostly yes. The TBCB shift and the long-run blended RoE arithmetic is the gap. I would not size a position above ~5% of portfolio without doing the homework on that question first.

Column 2 — Management

Q1. Would I trust these people to run my family’s money?

The honest answer is: I would trust them to run a regulated utility. I would not trust them to run an unregulated business.

The current Chairman & MD is Ravindra Kumar Tyagi, a career Power Grid lifer who took the role in late 2023. Thirty-plus years in the company before being appointed. The pattern is consistent — the CMD almost always comes from inside, almost always an engineer, almost always someone who has run a regional load dispatch centre or a major project office before being elevated. The institutional memory is deep. The skills are operational, not strategic in the founder-CEO sense.

What that means for capital allocation: Power Grid does not choose its capex programme. The Ministry of Power does, in consultation with the Central Electricity Authority’s National Electricity Plan. Power Grid executes against that plan. The CMD’s actual job is to deliver the projects on time and on budget, not to decide whether the company should be in transmission at all or to debate whether the next ₹50,000 crore of capex should go into wires or batteries. That decision is made above the CMD’s pay grade.

For a regulated utility that is fine. For a compounder that needs to find its second and third growth vector, it would be disqualifying. So the answer to “would I trust them with family money” depends on what the family money is being used for. If the use is “park capital in a tollbooth and collect a 5% dividend that compounds at the rate base”, yes. If the use is “back a management team that finds the next adjacency and creates 5x value over a decade”, no.

Q2. How do they treat minority shareholders when no one is watching?

The Government of India holds 51.34% of Power Grid. Free float is around 49%. The “no one watching” question for a PSU is not really about fraud or related-party self-dealing — there is little of that, and Power Grid has a clean track record on disclosures. The question is whether the majority owner, which is the government, uses the company as an instrument for its own balance sheet at moments of fiscal stress.

The empirical record is mixed. Power Grid pays out 50–60% of net profit as dividends. Most years, this is genuine corporate finance — return cash you cannot deploy. In the disinvestment-heavy years (2017–18, 2021), the government has also forced special dividends or done OFS sales that pushed extra dividends into the centre’s coffers. The minority shareholder benefits because the dividend cheque hits everyone proportionally. But the underlying decision is being made by a 51% holder whose calculus is fiscal, not commercial.

Read the next ten years through that lens: every time the central deficit is wide and the government has a window, expect a Power Grid OFS or a special dividend. The minority is treated correctly within the rules. The minority is also along for whatever ride the majority is taking. Not malicious. Not pure either. Better than most PSUs. Worse than a clean private-sector promoter who treats the public float as a partner.

Q3. What do their last ten years of capital allocation tell me?

This is the column where Power Grid actually surprises. Read against the broad PSU comparison set, the capital allocation track record is genuinely good.

The numbers, roughly: gross block grew from about ₹1.2 lakh crore in FY15 to about ₹3 lakh crore plus by FY25, a 2.5x build. Capex peaked in FY16-FY18 at around ₹25,000 crore a year, eased off in FY19-FY22 as the central transmission build cycle slowed, and is now ramping back up to ₹15,000-25,000 crore a year on the renewable evacuation programme. Cumulative dividends paid out across FY15-FY25 are well above ₹50,000 crore. Debt-to-equity stayed in the 1.8-2.2x range — high in absolute terms, but appropriate for a regulated utility where the debt is matched against long-dated tariff revenue. Capital cost overruns on major projects have been low single digit, which by Indian PSU standards is remarkable.

What they did not do — and this is the part that is harder to evaluate — is build any meaningful non-transmission franchise. The telecom adjacency is real but small. Smart metering is in joint-venture form. The international consultancy is a vanity line. There is no Power Grid Energy Storage, no Power Grid Renewables, no platform business that would justify a re-rating beyond the regulated-utility multiple. Some of that is by mandate — the parent ministry has been clear that Power Grid stays in transmission. Some of it is institutional. Either way, the upside optionality that a private-sector counterpart might have created is missing.

A disciplined operator. Not an empire-builder, in either the good sense or the bad.

Q4. Do they tell the truth in good years and bad?

The concalls are dry, factual, engineering-heavy. The CFO walks through the regulated equity, the ROE on each tranche, the order book, the receivables. There is no narrative spin. There are no “exciting strategic announcements” that turn out to be slides. When the receivables stretch out, they say so. When a project gets delayed because a state government did not clear right-of-way for a substation, they say that too.

The annual report test — pull the chairman’s letter from the worst year in the last decade and read it cold — is unusual for Power Grid because there isn’t really a worst year. The growth has been monotonic. The closest thing to a stress year was FY21, the Covid year, when several discoms went into receivables crisis and the government had to intervene with the LPS rules. Power Grid’s FY21 chairman’s letter is plain, factual, owns the receivables stretch, and points at what the company is doing about it. No theatre.

The flip side: dry honesty does not equal strategic candour. Power Grid will not tell you what it thinks about the TBCB-RTM mix erosion, or about what happens if CERC trims the 15.5% in 2029. They will not editorialise on policy. The chairman’s letter is a compliance document with extra detail, not a thesis statement.

Q5. Do they have real skin in the game, and for how long?

This is the question PSUs structurally fail.

Tyagi as CMD has accumulated stock through ESOPs and standard officer compensation, and the figures are public. They are not the figures that would make him think like an owner. The actual stake of any individual in Power Grid is a rounding error against the company’s market cap. Skin in the game in the founder sense — the kind Munger had in Berkshire or Kothari Pioneer’s founders had in their fund — does not exist.

What exists instead is institutional and political accountability. The CMD’s tenure is short and his tenure-after-tenure depends on his record. The board includes independent directors with reputations to protect. The Comptroller and Auditor General audits the books. The Department of Public Enterprises evaluates Power Grid against KPIs and rates it accordingly. The central government’s political capital is tied to the lights staying on.

That is a different mechanism from owner-CEO skin in the game. It is real, but it is real the way a long-tenured surgeon’s reputation is real — it gets you good execution, not great strategic risk-taking. For a wire monopoly that is fine. For something that needed a Jeff Bezos, it would be an unbridgeable gap.

Check — Can I trust these people with my capital? For this specific business, yes. They will execute. They will pay dividends. They will not blow up. They will also not take the second-mile risk that creates compounding alpha. I trust the institution. I do not trust an institution to do something that requires founder-level conviction.

Column 3 — Price

A caveat before this column: I am writing from structural ranges, not today’s quote. The framework below holds regardless. If you want a verdict at the current price, paste in today’s P/E, P/B and dividend yield at the bottom and re-run the third column.

Q1. What am I paying today?

Power Grid has historically traded in a fairly tight valuation band. P/E in the 12–18x trailing range. P/B in the 1.8–3.0x range, mostly clustering around 2.0–2.5x. Dividend yield in the 4–6% band. EV/EBITDA in the 8–11x range. Those ranges have widened during specific moments — the 2022–23 PSU re-rating pushed the multiples to the top of the band, and the bond-yield spike of mid-2023 briefly compressed them back to the bottom — but the band itself is structurally narrow because the cash flows are statutory.

So the first honest question is: where in the band am I, today? That is the data I would ask you for before sizing anything.

Q2. What is the market implicitly expecting, and do I agree?

This is the column where the template earns the most. Decode the multiple instead of accepting it.

At the centre of the band — say 15x P/E, 2.0x P/B, 5% yield — the market is pricing: regulated rate base grows 6–8% per year, blended RoE stays in the mid-teens, dividend payout sustained at 50%+, no adverse CERC tariff cycle. That is a defensible expectation. The renewable evacuation programme alone supports the rate base growth. The structural elements all hold.

At the top of the band — 18–20x P/E, 2.5x+ P/B, sub-4% yield — the market is asking the stock to be more than a regulated utility. It is pricing in either the bond-yield trade (10-year G-sec yields fall meaningfully, lifting all duration-sensitive assets), or a re-rating of the adjacencies (telecom and smart metering monetised at multiples Power Grid does not yet earn), or a TBCB win-rate that materially exceeds peers. Two of those three are not within Power Grid’s control. The third is plausible but small.

At the bottom of the band — 12x P/E, 1.7x P/B, 6%+ yield — the market is pricing in something going wrong. A CERC haircut, a discom crisis, a long bear market in PSUs. None of these is a base case. All three are tail risks that occasionally get priced.

Decoded, the centre of the band is the honest valuation. The top is asking you to make a macro bet. The bottom is offering you a fear premium.

Q3. What would I pay if I were buying the whole business?

The replacement value frame works cleanly here. The regulated gross block is roughly ₹3 lakh crore. A buyer of the whole business at a 15.5% post-tax RoE on regulated equity, with regulated equity around ₹1 lakh crore plus, is buying a steady ~₹15,000–17,000 crore PAT stream growing at 7–9% with very low operational variance. At a 15x earnings multiple — the centre of the band — that is a fair private buyer’s price.

A strategic acquirer would arguably pay a small premium for the optionality on the telecom fibre footprint and the smart metering JV, perhaps 5–10%. A financial acquirer would pay marginally less because the LBO math does not work — leverage is already on the regulated balance sheet.

The number you get from this exercise is not very different from where the stock trades in the middle of its band. That is unusual — most listed companies trade at multiples that wildly diverge from the private-buyer number. Power Grid trades close to its private-buyer fair value because the cash flows are nearly statutory and the discount rate is what the discount rate is. There is not much room for narrative to move the price either way. A feature, not a bug.

Q4. At what price would I sell, and why?

Two clean triggers, decided in advance.

One — multiple expansion to sustained 2.7–3x P/B without a corresponding step-up in RoE or growth. At that point you are no longer pricing a regulated utility. You are pricing the bond-yield trade, and the bond-yield trade is mean-reverting. Trim aggressively above this band.

Two — a CERC tariff order that signals a meaningful cut to the regulated 15.5% RoE. This is the regulatory event you cannot recover from in the same business model. If the next cycle (FY29 onwards) prints a 12% RoE, the long-run earnings power of the existing rate base resets downward by roughly 20%, and the appropriate multiple resets too. That is a sell-the-business event, not a sell-the-dip.

Both triggers are decided before buying, written down, not negotiated after the fact. Khandelwal is right that this is the discipline most investors skip, and skipping it is what turns a 15% IRR thesis into a 6% IRR outcome.

Q5. If the stock drops 40% next month, do I buy more or panic?

Two scenarios.

One — the drop is macro-driven: bond yields spike, broader market correction, foreign portfolio outflows. Buy more. The cash flows have not changed. The discount rate has changed temporarily. This is the easy answer.

Two — the drop is regulation-driven: CERC issues a draft order trimming the RoE, or a major discom cascade triggers a re-pricing of counterparty risk. Do not buy more. The cash flows have changed, and what looked like a 5% yield at 1.8x P/B is now a 4% yield at 1.6x P/B with a smaller earnings stream behind it. Re-run the entire template.

The honest answer to the question depends on knowing why the price moved. That is the work. The lazy answer — “buy the dip, dividends are dividends” — is the way to lose money on a Power Grid drawdown. The template forces you to do the work before reaching for the lazy answer.

Check — Am I paying a sensible price? Depends on where in the band the stock trades today. Centre of the band (≈15x earnings, 2x book, 5% yield) is a reasonable buy for a low-variance dividend compounder. Top of the band is a macro trade dressed up as a value trade. Bottom of the band is an opportunity if you have done the work to confirm the fear premium is overstated.

Closing — what the three columns surfaced

The interesting thing about applying Khandelwal’s template to Power Grid is what it makes visible. Three columns, three different textures of conviction:

Business: high. This is a wire monopoly with a regulated return and a structural tailwind from the renewable buildout. The TBCB-RTM blended RoE arithmetic is the soft spot, and it deserves the deeper homework before sizing.

Management: institutional, not personal. The trust is in the system — the CERC framework, the operational culture, the LPS rules, the auditor and the political accountability — not in any individual. That is the right kind of trust for a regulated utility. It would be the wrong kind of trust for a compounder.

Price: depends on the day. The valuation band is structurally narrow, the centre of the band is a reasonable place to own it, the top of the band is a bond-yield trade in disguise, and the bottom of the band is the only place where it becomes interesting as a contrarian buy. A 40% drawdown is either a gift or a warning, and which one it is depends entirely on whether the move was macro or regulatory.

The template doesn’t produce a verdict, and that is the point. It surfaces that buying Power Grid is partly a business bet (the structural transmission monopoly), partly a macro bet (where the ten-year G-sec settles), and partly a regulatory bet (CERC keeps the 15.5% RoE intact through the next cycle). Three different things, three different ways to be wrong, three different ways to size.

The investor who buys Power Grid because “it’s a monopoly with dividends” is making one bet. The investor who buys it after running the three columns is making three. They look the same on a brokerage statement. They are not the same trade.

Process notes

A few observations from this first run that are worth keeping for the framework:

The template’s refusal to score is its load-bearing feature. I did not assign a number to any of the three columns. I noticed myself wanting to, in the way a checklist user wants the checkmarks. Resisting that instinct is the work.
The price column requires either current data or an explicit caveat. I wrote one. For future runs, fold in the live quote at the moment of analysis and treat the centre-of-band heuristic as a first pass.
The management column hits regulated PSUs at a different angle than it hits a founder-led private company. The “skin in the game” question is awkward; the right reframe is “what is the alternative accountability mechanism, and is it real?” For Power Grid, the answer is yes. For a non-strategic PSU, the answer would often be no.
The business column’s “what could kill this” question is the most useful for utilities. Three different time horizons (TBCB blend, regulatory reset, distributed generation), three different probability weights. Worth doing in writing rather than mentally.
The closing question — “if it drops 40% tomorrow” — is the test that retroactively confirms whether you really understood the first two columns. If you cannot answer it cleanly, you have not done the work.

Score 7. Solid framework, applied honestly. The output is not a buy/hold/sell. It is a clearer view of what you are actually buying when you buy Power Grid, and what you would have to be right about for the trade to work.