A held policy rate is generally read as a held cost of capital. In India through 2026 the two have moved apart, and the gap is not noise - it is the banking system’s balance sheet asserting itself over the announcement.

The RBI has kept the repo rate at 5.25% since the cutting cycle ended in late 2025, and is widely expected to hold there through the rest of the year. Over the same period the marginal cost of funds for banks and non-banks has risen. What follows concerns why, using system data to 31 July 2026, and what it implies for anyone who prices credit off the policy rate.

Credit Is Outrunning Deposits by Nearly Four Points

As of 31 July 2026, system deposits grew 15.4% year-on-year against advances at 19.1%, leaving the banking system’s loan-to-deposit ratio at 80.3%. Six months earlier the same series showed credit at 14.6% against deposits at 12.5%. The gap has not closed; it has widened, and it has widened while the policy rate sat still.

A bank that lends faster than it gathers deposits has to fund the difference somewhere. That somewhere is certificates of deposit, market borrowings and other wholesale sources, priced by the market rather than administered by the central bank. The marginal rupee lent is therefore priced off wholesale funding, not off the repo - and it is the marginal rupee that sets the offer rate to the next borrower.

The Loan-to-Deposit Ratio Is the Binding Constraint, Not the Repo

This is the ordinary mechanics of a loanable funds squeeze, and it has a specific consequence: the policy rate stops being the operative constraint once the deposit base cannot keep up. Below that point, a cut transmits. Above it, banks are rationed by their own liabilities, and further cuts transmit weakly or not at all because the binding constraint has moved.

This is why deposit rates have stayed competitive in a year of held policy. Banks are bidding for a deposit base that is growing four points slower than their loan book. That bidding is a cost, and it is passed on.

Transmission to Non-Banks Was Never Complete

The gap is wider still outside the banking system. An RBI working paper found transmission to NBFCs incomplete: a one-percentage-point change in the repo is associated with a 0.24 percentage point change in NBFC weighted average borrowing rates over three quarters, and 0.33 percentage points in their weighted average lending rates.

Read that carefully, because the asymmetry is the finding. Roughly a quarter of a policy move reaches non-bank funding costs, and a third reaches what non-banks charge. A sector financed at a quarter of the pass-through, lending at a third of it, does not experience an easing cycle the way the headline suggests - and it experiences a tightening one faster than the headline suggests. NBFCs have accordingly seen only marginal relief in borrowing costs despite the cuts already delivered.

Why the Announcement Is the Wrong Signal to Watch

The policy rate is a price the central bank sets. The cost of capital is a price the system clears at. They coincide when deposits are ample and diverge when they are not, and nothing in the announcement tells you which regime you are in.

The series that does tell you is the credit-deposit growth differential and the loan-to-deposit ratio it produces. Those are published fortnightly, they are not forecasts, and in the current data they are pointing the opposite way to the policy rate. A model that takes the repo as the cost-of-capital input is, at 80.3% LDR, reading the wrong number.

What This Means for Allocators

Price credit off the marginal funding rate, not the policy rate. For any lender running a high loan-to-deposit ratio, the wholesale curve is the input that matters, and it moves independently of the MPC calendar.

Treat bank and non-bank exposure as two different rate sensitivities rather than one. A 0.24 pass-through on borrowing costs is not a smaller version of full transmission; it is a different exposure profile, and it cuts both ways across a cycle.

And watch the differential rather than the decision. The announcement is a headline. The balance sheet is the rate.

Back to Writing