Adani Enterprises: An Incubator Running on Borrowed Fuel
If you owned the whole company
Buy every share of Adani Enterprises today and the bill is about ₹4,17,103 crore. What do you get for it? In the year ended March 2026 the company reported ₹9,951 crore of net profit — about ₹2.40 of profit for every ₹100 you paid. But before you pocket even that thin coupon, read one line from the accountants: earnings include "other income" of ₹11,688 crore. Other income is money that did not come from running the businesses — gains from selling stakes, interest, one-off items. It is larger than the entire reported profit. Strip it out and the operations, after interest and tax, made less than nothing last year. Screener doesn't even print a P/E for this stock, and the return on equity — profit per ₹100 of the owners' own money — was minus 3.38%.
So the honest opening answer is: you would pay ₹4,17,103 crore for a collection of ambitious projects that, taken together, currently earns its owners a negative return. Everything else in this chapter is an explanation of why intelligent people pay that anyway, and why we won't.
What does this business actually do?
Adani Enterprises is the family's incubator — the nursery of the Adani group. It digs and trades coal, runs airports, builds roads and data centres, manufactures solar modules, and dabbles in defence and agriculture. The pattern repeats: hatch a business inside this company, feed it capital until it can walk, then spin it out or list it separately. Adani Ports, Adani Power, Adani Green — several of today's separately listed Adani companies were once eggs in this nursery.
So when you buy ADANIENT you are not buying a business. You are buying a machine that makes businesses, plus whatever eggs are currently in the nest. That is a perfectly legitimate thing to be — America has seen conglomerate-incubators come and go for a century. The question is only ever the same one: what does the machine pay for its raw material, and what is the raw material? Here the answer is: debt.
The science underneath
Charlie speaking. There is no chemistry in a holding company, so the first principles here are pure arithmetic — the arithmetic of leverage, which is the most dangerous simple mathematics in commerce.
A business earns a return on the capital inside it. If you can borrow at 9% and deploy at 14%, every borrowed rupee mints 5 paise a year for the owners, free. Beautiful. Now invert it, as we always must: if the project returns 7% and the loan costs 9%, every borrowed rupee destroys 2 paise a year, forever, until the debt is repaid or the equity is gone. Leverage is a lever in the strict physics sense — it multiplies force in whichever direction the force already points. It has no opinion about which direction that is.
Now apply the arithmetic to the file in front of us. Borrowings were ₹12,419 crore in March 2020. Six years later, March 2026: ₹1,06,622 crore. The debt grew more than eight-fold. Against that, the company's return on capital employed — what all the money inside the business earned, before dividing it between lenders and owners — is 5.8%. Ask a child: if the whole pile of capital earns ₹5.80 per ₹100, and a growing share of that pile is borrowed at Indian corporate interest rates, who is getting paid? The lender is. The owner is holding the lever's other end. Screener's red flags say the same thing in accountant's language: "low interest coverage ratio" (profits barely cover interest) and "company might be capitalizing the interest cost" — meaning some interest may be parked on the balance sheet as an "asset" rather than shown as an expense, which makes today's profits look better and tomorrow's problem bigger.
The incubator defence is real, and fairness demands we state it: airports and data centres lose money for years by design, then earn for decades. Today's 5.8% may be tomorrow's 15%. But that is a forecast, not a fact, and the interest bill is a fact with a due date.
The moat test
Hand a rival ₹4,17,103 crore and ten years. Could they replicate Adani Enterprises? Notice the question almost answers itself, because what they would replicate is not a castle but a building programme. Some individual assets in the nest are genuinely moated — an airport is a licensed local monopoly; you cannot build a competing one across the road. But those moats belong to the assets, and the assets are pledged to an ever-larger stack of lenders. The holding company itself has no brand a customer pays extra for, no switching costs, no network effect, no low-cost advantage in coal trading — trading is a commodity business where the science is: a tonne of coal is a tonne of coal.
What ADANIENT undeniably possesses is execution speed and political-economy skill — the ability to win concessions and build infrastructure faster than almost anyone in India. That is a real organisational talent. But a talent is not a moat; it lives in people and relationships and can sour, and it is already priced as if permanent. Moat rating: none — with moated children inside it that you can, notably, buy separately.
The numbers Warren would check
| What we check | Why it matters | ADANIENT |
|---|---|---|
| Return on equity, last yr | Owners' return on their money | −3.38% |
| ROCE | Return on all capital used | 5.8% |
| Borrowings Mar 2020 → Mar 2026 | The fuel | ₹12,419 cr → ₹1,06,622 cr |
| Net profit FY26 | Reported earnings | ₹9,951 crore |
| "Other income" in earnings | Non-operating portion | ₹11,688 crore |
| Cash from operations FY26 | Profit you can touch | ₹2,357 crore |
| Sales growth, 3 yr | Direction of the core | −8% a year |
| Dividend payout FY26 | Cash to owners | 2% |
| Promoter holding | Family stake | 74.67% (up from 67.65% in Jun 2023) |
Three of these deserve prose. First, the cash line: reported profit ₹9,951 crore, cash from operations ₹2,357 crore. When profit and cash disagree year after year, believe the cash. Second, sales have shrunk 8% a year for three years while borrowings and reserves ballooned — the balance sheet is growing much faster than the business. Reserves jumped from ₹50,199 crore to ₹80,797 crore in one year and equity capital ticked up from ₹115 to ₹129 crore, which tells you fresh shares were sold to somebody at high prices; that's how incubators eat — equity raises plus debt. Third, the promoters raised their stake to 74.67% while foreign institutions cut theirs from 19.34% to 10.80%. The family's conviction is genuine. Whether conviction at 5.16 times book value is wisdom is a separate question.
What could go wrong
Invert: what kills this company? Not a competitor — a refinancing. When your return on capital is 5.8% and your debt has grown eight-fold in six years, the business model is, functionally, continuous access to cheap capital. Anything that interrupts that access — a global rate spike, a credit-rating cut, a governance scandal that spooks lenders (this group has met one before), a stumble at a flagship project — and the arithmetic runs backwards fast. Add the Screener cons in plain words: profits lean on one-off income; interest cover is thin; interest may be being capitalised; three-year ROE is 2.41%; and the stock sells at over five times book value with a dividend payout of 2%. You are paying a rich price for a stressed balance sheet and being paid almost nothing to wait. The bull case is that the eggs hatch into monopolies worth multiples of today's debt. It may even happen. But "it may even happen" is a speculation, and we promised you honesty: this is a leveraged bet on execution and continued access to capital, not an investment in a demonstrated earnings stream.
What management must do to keep the castle
There is no castle yet — there is a construction site with a large mortgage. The board memo writes itself, and it is long because the moat is absent:
- Halt the debt growth. Set a hard ceiling on borrowings relative to operating cash flow — not reported profit — and publish it.
- Prove the incubation model with cash: sell or list a matured asset and use proceeds to retire debt, not to start three new ventures.
- Stop capitalising costs that belong in the profit-and-loss account; adopt the most conservative accounting available and let the numbers look worse and be truer.
- Get ROCE above the cost of borrowing before adding any new business line. Until then, every new venture is subsidised by lenders' patience.
- Report each incubated business's own profit, capital employed, and debt separately, so owners can do the arithmetic we just did — per egg.
- Raise the dividend only after the above; 2% payout is correct today. At least they're not pretending.
The verdict
No moat at the company level. Adani Enterprises is a collection of infrastructure bets — some potentially excellent — financed by borrowings that grew from ₹12,419 crore to ₹1,06,622 crore in six years, against a business earning 5.8% on capital and a negative return on equity last year. The reported profit leans on ₹11,688 crore of other income and converts poorly to cash. The lever is long, the family is pushing on it with real skill and real conviction, and levers multiply outcomes in both directions. A prudent owner who admires the Adani machine can buy its finished products — the listed ports or utilities — separately. Paying ₹4,17,103 crore for the factory floor, mid-construction, at five times book? We'll watch this one from the riverbank. Charlie adds: it takes a lifetime to learn that "exciting" and "leveraged" are the same word pronounced differently.
Written in the style of Buffett & Munger for the Omaha Investments book project. Educational material, not investment advice. Numbers from Screener.in and live NSE data via Angel One as of the date above.