The stories hidden in RBI's annual report | A tricky quarter for tyres

ELI5/TLDR

The RBI’s 250-page annual report hides five quieter stories: small-business lending boomed but went to fewer, bigger borrowers; banks are scrambling for funding because they’re lending faster than deposits are growing; last year’s low inflation was mostly luck (cheap food, a gold rally) and tells you little about the RBI’s skill; the record cheque to the government was smaller than it looked because most of the RBI’s “profit” was unrealised paper gains on gold; and India is quietly building programmable money. Separately, India’s four big tyre makers had a strong FY26 on cheap old inventory, but rubber and crude costs spiked at the quarter’s end. The real test is next quarter, when the higher cost base actually hits the books.

The Full Story

Lending to small businesses boomed — for a shrinking club of borrowers

Between April and December, bank credit to micro, small and medium enterprises grew 23.5% year-on-year, jumping from 29.8 lakh crore to 36.8 lakh crore. Seven lakh crore in a single year. A boom.

The twist: the money grew, but the borrower count didn’t. Micro borrowers stayed roughly flat; small and medium accounts actually fell. The average borrower simply got a bigger cheque.

Why both at once? Lending to a brand-new small firm is hard work. A first-time micro borrower has no credit history, no audited accounts, nothing to pledge.

The effort of underwriting a 5 lakh rupee loan to an unknown kirana shop is not very different from that in giving a 5 crore rupee loan to an established company.

And the returns are smaller. On top of that, India’s priority-sector lending rules force banks to direct money to small firms — but the targets are measured in rupees lent, not borrowers served. So the path of least resistance is to write bigger cheques to firms already on the books. The RBI’s response: from April 2026 it doubled the ceiling on collateral-free small-firm loans from 10 lakh to 20 lakh. Whether banks actually have the appetite to lend on trust is next year’s question.

Banks are scrambling for funding

A bank’s fortunes are tied to how cheaply it can source money, and the cheapest source is your deposit. The problem: banks have been lending faster than deposits grow. Commercial-sector credit rose 15.9% (up from 10.9% the year before); overall bank credit grew 17.1%. Deposits grew a healthy 16.2% — quick, but still nearly a percentage point behind credit.

Banks don’t run out of money — almost all money sits somewhere in the banking system. But when current and savings balances won’t grow fast enough, banks reach for pricier sources: fixed deposits, then wholesale funding like certificates of deposit (short-term IOUs sold to mutual funds and insurers). CD issuance climbed to 13.5 lakh crore from 11.9 lakh crore a year earlier.

That wholesale money is the dangerous kind. It reprices the instant rates move, and in a crisis it’s the first to dry up — the channel through which a funding squeeze can curdle into a failure. Its cost rose over the year too.

One myth, debunked: the report says there’s nothing to the “banks are starved because everyone’s piling into mutual funds” story. If anything, bank fixed deposits and debt mutual funds move together, and there’s no relationship at all between deposits and equity funds.

Low inflation last year was mostly luck

Retail inflation averaged 2.1%, well below the prior year’s 4.6% and nearly half the RBI’s 4% target. Tempting to call that a win. Don’t.

Almost half the inflation basket is food, which barely responds to interest rates — it’s monsoon, harvest, and logistics. Food prices rose just 0.8% from April to December, against 7.6% a year earlier. October even saw the steepest food deflation in the index’s history, sitting on top of a very expensive base.

Core inflation (stripping food and fuel, the things the RBI can’t control) was higher at 3.7% — but half of that came from a historic gold and silver rally, as global investors bought safe havens. Strip those out and you’re back to 2.1%. September’s GST cuts then made durables cheaper still.

This was simply a free cure. It saw geopolitical upheaval, choked global energy lifelines, jumpy commodity markets, and an unusually benign time for agriculture.

None of it is information the RBI can steer policy by. With the inflation basket itself being revised, next year will look different regardless. The RBI expects inflation to rubber-band back to 4.6% as base effects wash out.

The record cheque to the government wasn’t what it looked like

The RBI handed the government 2.86 lakh crore — its largest ever, widely read as a windfall dividend. It isn’t quite a dividend. By law, once the RBI covers costs and sets aside reserves, it pays out what’s left. The tricky part is defining “what’s left.” A committee, after an ugly tug-of-war over the right buffer size, set a band for how much the RBI may keep. That band was re-pegged last year.

Here’s the tell: the RBI’s income jumped 26%, but the cheque grew only 6.7%. It kept 1.09 lakh crore as cushion — up sharply from 45,000 crore the year before. Why hoard? Because most of the extra income was paper profit. Its gold holdings leaped about 64% in value (it bought less than a ton), amplified by the rupee’s fall. Those gains aren’t real until the gold is sold, so they can’t be handed out — they sit in the revaluation account, a cushion against a swollen balance sheet.

India is quietly building programmable money

UPI carried over 200 billion transactions last year, up 30%, handling about 86% of every retail payment. The RBI’s next ambition is to make money itself programmable — money with rules baked in about where it can be spent.

It’s already tried this with its central bank digital currency. In Gujarat, Puducherry and Chandigarh, food subsidies were paid in a programmable rupee coded to be spendable only on eligible goods at ration shops.

This eliminated the need to set up monitoring mechanisms to police leakage and corruption. There was simply nowhere else that money could be spent.

Next up: tokenisation — issuing financial assets as digital tokens settled in the wholesale digital rupee, so they can change hands instantly the way UPI moves money. The first asset onto these rails was, fittingly, the certificate of deposit. A parallel project, the Unified Lending Interface (ULI), lets a lender pull together scattered records — land records, GST filings, business registration, even a dairy cooperative’s receipt — to underwrite faster. It crossed 9.6 crore data requests and 117 lenders. A reality check: not every project works. The retail digital currency in circulation actually shrank this year, from about 1,016 crore to 772 crore rupees.

A tricky quarter for tyres

Tyres are a grudge buy — nobody enjoys replacing them. India’s four big listed makers (CEAT, JK Tyre, Apollo, and BKT) closed FY26 with factories running hot after a demand surge that followed a GST cut on tyres. Then, in the last weeks of Q4, the raw-material picture turned sharply worse.

The headline numbers were good. CEAT’s standalone revenue crossed 4,000 crore for the first time (4,436 crore, up 18%), with a 14.5% EBITDA margin and PAT nearly tripling. But CEAT admitted it only held margins because it was still running on cheaper inventory bought earlier — raw-material costs are set to rise 15–20% in Q1 FY27. JK Tyre had its best year ever (Q4 revenue up 12% to 4,233 crore, PAT 188 crore vs 102 crore) and announced a 6,100 crore expansion. Apollo posted the biggest numbers (7,336 crore, 14.6% margin) but with a footnote: PAT included a one-time ~570 crore deferred-tax benefit; strip it out and the business is more modest. BKT had a flat revenue quarter (~2,900 crore, up 2%) but record off-highway volumes and a 22.9% margin — far above peers, because off-highway tyres (farm, construction, mining) carry more pricing power.

The shared villain is natural rubber, much of which India imports. So margins are exposed to global rubber prices and the rupee, which slid to 94 against the dollar in Q4. International natural rubber rose from about $1,800 to $2,100 a ton over the quarter. Q4 was insulated because companies buy ahead — but everyone expects the bill in Q1 FY27, with the raw-material basket guided up to the high teens.

The catch isn’t that companies can’t raise prices — it’s that different channels absorb hikes at different speeds. OEM pricing (what tyre makers charge vehicle manufacturers) is contractual and adjusts with a roughly three-month lag, so makers eat the higher costs first. Replacement demand — end consumers buying off the shelf — has no lag, but isn’t immune: fleet operators can stretch tyre life, retread, or delay non-urgent replacements when prices jump. CEAT warned truck and bus replacement demand may slow rather than collapse.

That matters because the whole industry is running near 90% utilisation and adding capacity. Apollo committed 3,500 crore of FY27 capex (80% toward growth); JK Tyre added a fresh 5,000 crore; BKT approved another 2,000 crore. A big part of the confidence is rural demand — JK Tyre’s farm volumes grew 58% year-on-year, tractor sales crossed 10.9 lakh units in 2025. But rural cash flows depend on rainfall; a poor monsoon across Maharashtra and central India could push out farm-equipment purchases.

That is why the next quarter matters more than the last one. If volumes hold despite higher prices, the sector’s margin recovery story remains intact. If replacement demand slows, even modestly, the recovery gets pushed out.

FY26 ended with factories running harder than they had in years. FY27 begins with the same factories and the same commitments, but a cost base that’s shifted against them.

Key Takeaways

MSME credit grew 23.5% YoY (29.8 → 36.8 lakh crore), but the borrower count was flat or falling — banks wrote bigger cheques to existing clients rather than finding new ones.
Priority-sector lending targets are measured in rupees lent, not borrowers served — a structural incentive to concentrate, not broaden, small-business credit.
Underwriting a tiny loan costs nearly as much effort as a large one, for a fraction of the return — the core reason new micro borrowers stay underserved.
From April 2026, the RBI doubled the collateral-free small-firm loan ceiling from 10 lakh to 20 lakh rupees.
Bank credit grew 17.1% vs deposits at 16.2% — a persistent gap forcing banks toward pricier wholesale funding (CD issuance up to 13.5 from 11.9 lakh crore).
Wholesale funding (certificates of deposit) reprices instantly with rates and dries up first in a crisis — the channel through which funding squeezes become failures.
Myth busted: bank deposits and equity mutual funds have no relationship; bank FDs and debt mutual funds actually move together. Equity inflows aren’t starving banks of deposits.
Judging the RBI by headline inflation is misleading — nearly half the basket is food, which responds to monsoons, not interest rates.
FY26’s low 2.1% inflation was largely luck: cheap food (up just 0.8% Apr–Dec vs 7.6% prior), a high base, and GST cuts on durables.
Even core inflation (3.7%) was distorted — half came from a gold/silver safe-haven rally; strip it out and you’re back at 2.1%.
The record 2.86 lakh crore transferred to the government understated the RBI’s income (up 26%); it retained 1.09 lakh crore (vs 45,000 crore prior) because most “profit” was unrealised paper gains.
Gold holdings rose ~64% in value despite almost no buying (<1 ton) — gains amplified by the falling rupee, parked in the revaluation account and not payable until sold.
UPI handled 200 billion+ transactions (up 30%), about 86% of all retail payments.
Programmable money in action: CBDC food subsidies in Gujarat, Puducherry and Chandigarh were coded to be spendable only on eligible goods at ration shops — eliminating leakage by design.
Next on the RBI roadmap: tokenising financial assets settled in the wholesale digital rupee (first asset: certificates of deposit) and the Unified Lending Interface (9.6 crore data requests, 117 lenders).
Reality check on CBDC: retail digital currency in circulation shrank this year, from ~1,016 crore to 772 crore rupees.
Tyre makers’ strong FY26 ran on cheap pre-bought inventory; rubber rose from ~$1,800 to $2,100/ton in Q4 and the rupee hit 94 — the cost hit lands in Q1 FY27, guided to high-teens inflation in the raw-material basket.
Off-highway tyres (BKT, 22.9% margin) carry more pricing power than truck or passenger tyres, which is why BKT’s margins ran far above peers.
OEM tyre pricing is contractual with a ~3-month lag, so makers absorb input-cost spikes before they can pass them on; replacement pricing has no lag but customers can stretch tyre life.
The sector runs near 90% utilisation while still committing large capex (Apollo 3,500 crore, JK Tyre +5,000 crore, BKT +2,000 crore) — if replacement demand slips, that new capacity takes far longer to pay off.

Claude’s Take

The RBI half of this is the genuinely good part — the kind of read that turns a dry annual report into five clean mental models. The standout is the reframing of the inflation number: the discipline to say “low inflation last year tells you almost nothing about the RBI’s competence” is the opposite of the usual narrative, and they back it with the food/base/gold breakdown rather than asserting it. The “record cheque was mostly paper gains” point is similarly the sort of thing that gets misreported everywhere, and they unpick it cleanly. The MSME-concentration story and the priority-sector incentive that drives it is the closest thing here to a durable structural insight.

The tyre half is competent earnings synthesis but less surprising — strong quarter on old inventory, cost spike coming, capex into a possibly-softening replacement market, watch Q1 FY27. All true, all well-organised, but it’s a setup with no payoff yet by construction; the interesting quarter hasn’t happened. The one transcript wobble worth flagging: the natural-rubber figure is garbled (“$1,800… to $250 to $2,100”), which I’ve read as $1,800 → $2,100. Treat the exact endpoint loosely.

Score 7. Above-average for the format — the RBI segment alone earns it — but it’s a daily-news digest, not a piece you’ll return to, and the tyre half is solid filler rather than signal.

The stories hidden in RBI's annual report | A tricky quarter for tyres | The Daily Brief #483