Bajaj Finance: the moneylender who remembers everything
If you owned the whole company
Suppose you bought every share of Bajaj Finance today. The bill would be about ₹6,40,620 crore — the market capitalisation, which is simply the price of one share multiplied by all the shares in existence. In return, the business earned ₹19,332 crore of net profit in the year ended March 2026.
Think of it as a lemonade stand. You pay ₹6,40,620 for the stand; it hands you about ₹19,332 a year. That is an "earnings yield" of roughly 3 paise per rupee invested — the flip side of the stock's price-to-earnings ratio of 33.3. A P/E of 33 means the market charges you thirty-three years of today's profit for the business. Nobody sane pays that for a stand whose lemonade sales stay flat. You pay it only if you believe the profit will keep growing fast. So the whole question of this chapter is: will it?
What does this business actually do?
Bajaj Finance is a moneylender — the modern, computerised descendant of the man in the bazaar who lent against your word and knew your family for three generations. It is an NBFC, a non-banking financial company: it lends like a bank but cannot offer you a chequing account, though it does take public deposits.
The money comes from one simple act, repeated tens of millions of times: borrow money wholesale at one price, lend it retail at a higher price, and keep the difference — minus whatever the borrowers fail to repay. The company lends for televisions and phones bought on instalment, for small businesses, for homes, in cities and villages across India. Every EMI on a fridge is a tiny toll paid to this bridge.
The science underneath
Charlie here. There is no chemistry in lending. The science is arithmetic and probability, and both are unforgiving.
A lender's profit per rupee lent is: (interest charged − interest paid − operating cost − loans that go bad). Each piece is small — a few paise per rupee. To turn paise into a living, the lender uses leverage: it lends out many rupees for every rupee of its own. Look at the balance sheet: the owners' money (equity of ₹622 crore plus reserves of ₹1,13,377 crore) supports a vastly larger loan book funded with borrowed money. Leverage multiplies the thin spread into an 18.2% return on equity — for every ₹100 the owners left in the business, ₹18.2 came back as profit last year.
But leverage is a two-edged sword of exactly the same sharpness on both edges. If the spread is 3 paise per rupee and loans going bad rise by 3 paise, the profit is zero. Rise by 6, and the owners' capital starts burning. The whole game, therefore, reduces to one number: the probability of default — out of 100 borrowers, how many don't pay you back?
Here is where Bajaj Finance built something real. When you finance ten million televisions, you learn — statistically, not anecdotally — which pin codes repay, which occupations repay, which repayment histories predict the next default. Every loan is an experiment, and the company has run more consumer-credit experiments than almost anyone in India. Better prediction → fewer bad loans → fatter spread at the same interest rate → more profit → more data. That loop is the moat, such as it is. The raw material of a lender is money, and its refining process is information.
One more piece of arithmetic: the raw-material price. A lender's cost of funds is what it pays to borrow. A trusted, highly-rated borrower pays less, and every fraction it saves drops straight into the spread. Trust, painstakingly earned, is a cost advantage — the lending trade's version of owning the cheapest coal mine.
The moat test
Hand a capable rival ₹6,40,620 crore in cash and ten years of patience. Can they take the castle?
They can certainly build the branches and the app. Money is a perfect commodity — my rupee lends exactly like yours. So the moat cannot be the product. What the rival cannot buy quickly:
- The data. A decade of repayment behaviour on tens of millions of small loans. You can only gather it by lending, and lending without it means eating losses while you learn. Tuition is payable in advance.
- Cheap funds. Lenders earn trust slowly; a newcomer borrows dearer, so its lemonade costs more before it sells a glass.
- Distribution habit. Millions of customers already have a repayment relationship; the next loan is one tap away. A switching cost of convenience, not of contract.
What the rival can do: banks have even cheaper funds (deposits), and a determined giant — a large bank or a conglomerate with deep pockets — can subsidise losses for years. The castle is defensible but not impregnable. We call this moat narrow: real, earned, but resting on execution rather than on any law of nature.
The numbers Warren would check
| What we check | What it means | Bajaj Finance |
|---|---|---|
| Sales growth, 10 yrs | Is the toll bridge getting more traffic? | 27% a year |
| Profit growth, 10 yrs | Does traffic turn into money? | 31% a year |
| Return on equity | Profit per ₹100 of owners' money | 18.2% (10-yr average 19%) |
| Financing margin | Paise kept per rupee of revenue | 33% (FY26); peaked at 39% in FY23 |
| Stock P/E | Years of profit you pay upfront | 33.3 |
| Price to book value | Price vs owners' capital per share | 5.62× (book value ₹183, price ~₹1,029) |
| Dividend payout | Share of profit posted home | 20% |
| Promoter holding | Skin in the game | 54.71%, down from 55.87% in Jun 2023 |
Revenue grew from ₹5,392 crore in March 2015 to ₹81,982 crore in March 2026; profit from ₹898 crore to ₹19,332 crore. Only one stumble in eleven years — profit dipped to ₹4,420 crore in the pandemic year of March 2021 — and that stumble is the most informative data point on the page, because it shows what one bad year does to a leveraged lender.
Two oddities a beginner should not misread. First, cash from operations is hugely negative — minus ₹65,790 crore in FY26. For a shopkeeper that would be a hearse at the door; for a fast-growing lender it is normal, because every new loan pushed out the door counts as cash going out. The cash comes back over years, with interest — if the borrowers pay. Second, equity capital jumped from ₹124 crore to ₹622 crore in FY26. That reflects bonus shares — an accounting reshuffle that hands existing owners more pieces of the same pie — not new shares sold for cash. Per-share earnings still grew, from ₹26.77 to ₹30.56.
What could go wrong
Invert: what would kill this business?
- A credit cycle. Consumer lending in India has grown magnificently for a decade. Cycles have not been abolished; they have merely been on holiday. When unsecured borrowers hit a bad patch together, the default probability jumps for everyone at once, and leverage does the rest.
- The price. Screener's own cautions: the stock trades at 5.62 times book value, and profit growth has already slowed — 31% over ten years, 19% over three, 15% in the trailing twelve months. Pay 33 times earnings for 15% growth and your margin of safety is the width of a cigarette paper.
- Accounting soft spots. The screen flags a low interest-coverage ratio (routine for lenders, since interest is their raw material) and notes the company might be capitalising interest cost — parking interest on the balance sheet instead of the profit line. Worth watching, not ignoring.
- Regulation and rivals. The RBI can reprice the rules of unsecured lending overnight, and every bank and tech giant in India covets exactly this customer.
What management must do to keep the castle
- Guard underwriting discipline above growth. The data moat dies the day loan targets outrank default forecasts.
- Keep the credit rating pristine; cheap funds are the cost advantage.
- Don't stray into lending you have no data edge in — big-ticket corporate loans have humbled better-dressed lenders.
- Keep dilution rare and bonus issues honest; per-share value is the scoreboard.
- Tell shareholders bad news fast. Leveraged institutions run on trust, and trust compounds — in both directions.
The verdict
Moat: narrow. Bajaj Finance is a genuinely superior lending machine — the compounding record from ₹898 crore to ₹19,332 crore of profit in eleven years is among the finest in Indian finance, and the data-and-distribution loop should keep it ahead of most rivals for five to ten years. But it is a leveraged lender in a commodity product, and its moat requires flawless execution every single year, which is a moat with a mortgage on it. At 33 times earnings and 5.6 times book, the market has already prepaid for another decade of brilliance, while growth has visibly downshifted. Wonderful business; demanding price. The dinner-table summary: this is the moneylender who remembers everything — just remember that you're being asked to pay for a memory of the future too.
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.