← all companies · 2026-07-07 · moat: narrow · raw .md

Shriram Finance: The Lender Who Reads Tyres

Four in the morning on the highway

It is not yet light on the national highway when the driver finishes his tea and swings up into the cab. The truck is nine years old. It has carried cement, onions, ceramic tiles, and once, memorably, a wedding band's equipment. Its first owner was a fleet company that sold it at five years, as fleets do. Its second and current owner is the man now turning the key — a driver for fifteen years, an owner for four, since the day a finance company agreed to trust him with a loan that no bank would touch.

Not because banks are lazy. Because the man is, in a banker's vocabulary, invisible. He has no salary slip — his income is freight payments, irregular, partly in cash. No income-tax return worth reading. No credit history to score, because to have a credit history you must first have had credit. Every instrument of modern formal lending — the payslip, the bank statement, the credit-bureau file — returns a blank when pointed at him. He is one of tens of millions of Indians who earn real, durable incomes that leave no paper trail.

Yet somebody lends to him, has lent to millions like him for over four decades, and earns a 16% return on equity doing it. Understanding how — why this is a solvable problem, and why the solution resists every attempt by larger, cheaper, more technological competitors to copy it — is the subject of this chapter.

The physics of a market for lemons

Start with the underlying science, which won a Nobel prize. In 1970 the economist George Akerlof described what happens to a market when one side knows more than the other. His example was used cars: the seller knows if the car is a lemon; the buyer does not; so the buyer offers only an average price; so owners of good cars refuse to sell; so the average quality falls, the price falls further, and the market can collapse entirely. This is information asymmetry, and it is not a metaphor here — used trucks and their invisible-income buyers are the literal textbook case, squared.

A bank facing our driver confronts two asymmetries stacked together. It cannot verify his income (he knows it; the bank doesn't), and it cannot judge the second-hand truck's true condition and worth (he knows; the bank doesn't). Formal lending solves asymmetry with documents; where documents end, banks end. That is not regulation or snobbery — it is arithmetic. A bank's model needs verifiable inputs the way a thermometer needs contact.

But notice: the information exists. It just isn't written down. The driver's income is knowable — from his routes, the freight rates on those routes this season, how many trips a month the truck can make, what diesel and tolls eat. The truck's worth is knowable — from its engine's sound, its tyre wear, its accident scars, its resale price in the local market this quarter. What's missing is not data; it is a sensor that can read data stored in the physical world and in community knowledge rather than in databases. Shriram's branch network — 1,758 branches and 831 rural centres, by its own accounting, plus working partnerships with some 500 informal local financiers — is exactly that: a human sensor array, pointed at the cash economy.

The field officer who prices our driver's loan often comes from the trucking world himself. He knows the freight rates because he asks weekly; he knows the truck's worth because he has seen a thousand like it; he knows the driver's reputation because the transport yard is a village. Lending here does not scale like software. It scales like craft — slowly, by apprenticeship, one trusted officer at a time. Which is precisely why it has not been commoditised.

The asset with a second and third life

The collateral deserves its own physics. A commercial truck is a machine whose earning power declines far more slowly than its price. Depreciation runs fastest in the early years — the first owner absorbs the steepest slide — yet a well-kept truck at year eight still hauls nearly the loads it hauled at year two. So the used truck occupies an economic sweet spot: cheap to buy, still fully able to earn. It is the natural first rung for a driver becoming an owner, which is the great ladder of the Indian road-transport economy.

For the lender this creates an unusually honest form of security. The loan is not against the borrower's promise alone but against a mobile, income-producing, re-sellable machine whose secondary market the lender knows intimately. If the loan fails, the truck can work for someone else — and here the human network matters again, because repossession in the informal economy is a social process, done through relationships and mediation in yards and markets, not through lawyers. A rival cannot download that capability.

From a Chennai chit fund to the Nifty 50

The firm's history is the strategy in narrative form. It began in Chennai in the late 1970s, inside the old South Indian institution of the chit fund — a community savings circle, finance built on neighbourhood trust. From that soil grew a lender that made pre-owned commercial vehicles its life's work at a time when organised finance considered the segment untouchable. The group's insight, held stubbornly for four decades, was that the bottom of the trucking pyramid was not a charity case but an underpriced market — that information asymmetry was a cost to be engineered down, not a wall.

The record in our data carries a visible seam: revenue jumps from ₹19,255 crore in Mar 2022 to ₹30,492 crore in Mar 2023, with profit leaping from ₹2,721 crore to ₹6,020 crore. That is the great merger — the group folded its consumer and small-business lender and its holding company into the transport financier, creating today's Shriram Finance: one balance sheet lending against trucks, two-wheelers, gold, and small enterprises. The diversification logic is sound — the customer is the same invisible-income Indian; only the collateral changes.

Across the full decade, revenue compounded 17% (₹9,177 crore in Mar 2015 to ₹48,135 crore in Mar 2026) and profit 24% (₹1,028 crore to ₹10,024 crore), with earnings per share rising from ₹8.13 to ₹53.28. Financing margins widened from 18% to 29% along the way — partly the merger's mix, partly pricing power doing its quiet work. Return on equity has been a metronome: 16% over ten years, 16% over five, 16% over three, 16.4% last year. In a cyclical industry, that flatness is the accomplishment.

Why the giants stay on the paved road

What stops a large bank — with funding costs several points cheaper — from taking this market? It has been tried, repeatedly, and the failures are instructive.

The bank's advantage is cheap money; its handicap is that its entire credit process is built on documents this customer does not have. To lend here it must either lower its standards (and bleed) or rebuild Shriram's sensor network (and spend a decade doing it, at branch economics banks find unattractive for small-ticket loans). The fintech's advantage is reach; its handicap is that its algorithms eat digital exhaust, and this customer's economic life is analogue. There is no server anywhere holding the data an app would need — the data lives in freight yards and in the heads of men who attend truck auctions. Meanwhile the informal moneylender has the knowledge but not the balance sheet, which is why five hundred of them found it more profitable to partner with Shriram than to fight it.

So the niche is defended on both flanks: too information-dark for the formal giants, too capital-hungry for the informal locals. The moat is the four-decade-old apparatus for turning street knowledge into credit judgment — narrow, because it protects a segment rather than the whole territory of finance, but deep within its segment.

The margins of safety Charlie would inspect

Munger's checklist, applied honestly.

Incentives: field officers who both originate and collect — the Shriram tradition — carry the healthiest incentive in lending, because the person who approves the loan personally lives with its consequences. Preserving that loop as the company digitises is culture-critical work.

Inversion: how would this franchise die? Route one: funding. Shriram, like every non-bank, borrows wholesale and holds deposits without a central-bank backstop; its borrowers are the first to wobble in a downturn, and its lenders know it — Screener's standard warning of low interest coverage attaches, as it does across the industry. The 2018–19 NBFC funding freeze was the dress rehearsal; survival then is the credential now. Route two: the moat's own foundation erodes — see the risks section — as India formalises and the invisible become visible to everyone.

Opportunity cost: the 16% return on equity, held steady through a merger, a pandemic and rate cycles, sets a demanding bar for any diversification. Adjacent lending to the same customer clears it; anything requiring a new customer or new craft likely does not.

What should management do this decade to preserve the moat? Three things. First, convert craft into institution — the sensor network lives in people, so apprenticeship pipelines, retention of field talent, and careful codification of local knowledge (without pretending a database can replace a yard visit) are the moat's maintenance capex. Second, keep buying funding durability — longer-term borrowings, a broader deposit base, insurance against the sector's recurring liquidity winters; margin traded for stability is well sold. Third, meet formalisation halfway — as the customer acquires digital footprints, fuse the new data with the old craft rather than surrendering the segment to bureau-score lenders; the firm that knows both the freight yard and the FASTag data is stronger than either kind of rival.

One governance note an owner must file: promoter holding is just 20.3%, and slipped about five points in the recent record; foreign institutions hold 56.1%. This is a professionally governed, institutionally owned firm — the founder generation's philosophy must now persist as culture rather than as shareholding, which is a real, if unquantifiable, long-term risk. The shareholder register has meanwhile tripled, from about 97,000 to 3.3 lakh names in under three years.

Pricing the whole yard

The Buffett arithmetic. The market asks ₹2,50,938 crore for the entire company — 25 times last year's ₹10,024 crore of profit (an earnings yield of 4%), and about 3 times the ₹350 book value per share against a price near ₹1,066. A 16% return-on-equity business at three times book implies the buyer's underlying yield on carried equity is around 5%, growing as the book compounds. Dividends add a real if modest stream — a payout around 20% of profit, a yield of about 1% — with the rest retained at that same 16%.

Quality of earnings: this is a spread business on a seasoned book, about as predictable as lending to the cash economy can be — which is to say, cyclical but mean-reverting, as the flat decade of ROE attests. Cash from operations is negative in every year of our record (−₹13,281 crore in Mar 2026), which in a growing lender is disbursal outrunning collection — the cost of growth, funded in the market, carrying the funding risks already named. Pricing power is genuine but bounded: the customer has few alternatives, yet decency and regulation both cap what may be charged to a truck driver; the margin story (18% → 29%) has likely travelled most of its road.

Could it compound for decades? The engine is proven and the market is vast — but the owner's return will track book value growth, not multiple expansion, and book compounds at roughly ROE minus payout: call it low-to-mid teens. Honest, durable, unglamorous.

Where the road could wash out

Structural risks only:

The funding architecture. The permanent one. A wholesale-funded lender to cyclical borrowers is exposed at both ends of its balance sheet in the same storms. Each freeze so far has been survived; each was also a coin the company did not flip itself.
Formalisation eats the asymmetry. The moat exists because the customer is illegible to banks. UPI trails, FASTag records, GST-registered small fleets — every year, more of the invisible become visible, and visible customers get bank pricing. The moat's erosion is slow, demographic, and one-directional; Shriram's counter is to ride the data across rather than drown under it.
The powertrain transition. Electric trucks would rewrite the depreciation curves, resale markets and mechanic-knowledge on which collateral valuation rests. Diesel's long Indian twilight gives decades of runway, but a lender of nine-year-old vehicles must think about what vehicles look like nine years hence.
Regulatory tightening. Non-bank deposit-takers of systemic size drift steadily toward bank-like regulation — capital, provisioning, perhaps rate scrutiny on the very spreads that fund the model.
Culture dilution. The sensor network is people. Scale, digitisation and institutional ownership all press, gently and permanently, toward making Shriram more like the banks it out-lends — which would be efficient, tidy, and fatal to the edge.

The cab in 2050

Twenty-five years out, India will still move overwhelmingly by road, and the ladder — driver to owner, one used truck at a time — will still be the way families climb into the transport economy. The trucks will change powertrains; the freight data will live in the cloud; the driver will have a credit score. The question for Shriram is whether its craft survives its customers' legibility — whether an institution built to read tyres can learn to read telemetry without losing the field officer's judgment that no telemetry captures: is this man good for it? Firms that industrialise a human skill before rivals can either match the skill or obsolete it tend to enjoy long, quiet reigns. That is the wager here, and four decades of evidence say it is not a naive one.

An Omaha Investments chapter. Educational material, not investment advice. Figures from Screener.in and NSE data via Angel One as of the date above.