Banks pass on rate hikes, but go slow when rates fall; RBI study finds

The RBI study found that banks respond differently depending on the direction of monetary policy. During tightening, when the repo rate rises, contracted lending rates adjust upward more completely. During easing, when the repo rate falls, banks pass through a smaller share of the reduction. This matters because the timing and size of borrowers’ interest-rate relief can differ from the central bank’s action.

Between March 2022 and January 2025, a 250-basis-point repo increase led to a 191-basis-point rise in contracted lending rates. That equals 76.4% pass-through. Between February 2025 and March 2026, a 125-basis-point cut produced only an 80-basis-point fall, or 64% pass-through.

The difference narrowed when the study used headline WALR, which includes loan-book composition changes. WALR rose 181 basis points during tightening and fell 93 basis points during easing. Banks can therefore soften cuts by changing spreads or shifting toward higher-yielding loans.

The repo rate is the rate at which a central bank lends short-term funds to commercial banks against eligible securities. It is a key monetary-policy signal. When it changes, banks’ funding conditions and the broader cost of money can change. Banks may then adjust interest rates on new loans or reprice existing loans linked to benchmarks.

The RBI study tracks this connection through lending rates and WALR. A 250-basis-point repo-rate increase from March 2022 to January 2025 produced a 191-basis-point increase in contracted lending rates. A later 125-basis-point cut produced an 80-basis-point decline during February 2025 to March 2026.

The link is not automatic or complete. Banks may adjust spreads, protect their net interest margins, or shift their loan books toward higher-yielding credit. That is why borrowers can experience a smaller or slower change than the repo-rate movement.

The study measured pass-through as the share of a repo-rate change reflected in banks’ contracted lending rates. During tightening, banks passed through 76.4% of the increase. During easing, they passed through 64% of the reduction. This comparison shows that rate cuts reached loan pricing less fully than rate hikes.

The tightening period ran from March 2022 to January 2025. The repo rate rose 250 basis points, while lending rates increased 191 basis points. The easing period ran from February 2025 to March 2026. The repo rate fell 125 basis points, while lending rates declined 80 basis points.

Using headline WALR, the results looked different because it includes loan-book composition. The 250-basis-point hike increased WALR by 181 basis points, while the 125-basis-point cut reduced it by 93 basis points. That produced pass-through rates of 72.4% and 74.4%, respectively.

A bank’s average lending rate is not simply a fixed list of individual loan prices. The weighted average lending rate, or WALR, also depends on the size and mix of loans in the portfolio. If the share of high-rate loans increases, the average can rise even when existing loan rates do not change. If lower-rate loans gain weight, the average can fall.

The RBI study separates price effects from composition effects. During easing, contracted lending rates fell 80 basis points, but headline WALR fell 93 basis points. During tightening, contracted rates rose 191 basis points, while WALR rose 181 basis points. The differences reflect shifts in the credit mix.

This measure matters because it shows what the average loan book actually costs, not only how individual contracts are repriced. In March 2026, fresh lending rates rose 12 basis points despite an easing stance, illustrating how pricing and composition can produce a different average result.

Net interest margin, or NIM, is the gap between what banks earn on loans and what they pay for funding. When market rates fall, banks may try to preserve that gap rather than pass the entire reduction to borrowers. The RBI study identifies two methods: raising lending spreads and shifting the loan book toward higher-yielding loans.

These actions can reduce or offset the effect of a repo-rate cut. During the February 2025 to March 2026 easing cycle, the repo rate fell 125 basis points, but contracted lending rates fell only 80 basis points. A positive weight effect also partly offset the decline in large-industry rates, reflecting movement toward higher-spread loans or lenders.

The result was uneven transmission. Headline WALR still fell 93 basis points, but fresh lending rates rose 12 basis points in March 2026 despite the easing stance. Banks’ effort to protect NIMs therefore remains important for how quickly borrowers feel rate cuts.

When banks pass on rate cuts slowly, borrowers do not immediately receive the full benefit of cheaper policy funding. Loan rates may decline by less than the central bank’s reduction, leaving interest costs higher than they would be under complete pass-through. This is especially relevant for borrowers whose loans are repriced during an easing cycle.

The RBI study gives a clear comparison. A 125-basis-point repo-rate cut between February 2025 and March 2026 led to only an 80-basis-point fall in contracted lending rates. That represented 64% pass-through. By contrast, a 250-basis-point hike earlier led to a 191-basis-point increase, or 76.4% pass-through.

Borrowers may therefore experience asymmetric changes: costs rise more fully during tightening, while relief arrives less fully during easing. The effect can also vary by loan type. MSME rates fell 122 basis points, while education rates fell 162 basis points, but their overall influence differed because of credit shares.

A loan category affects the overall average not only through the size of its rate change, but also through its share of total credit. A large category can move system-wide WALR substantially even after a moderate rate decline. A small category can show a bigger decline yet barely change the average.

MSME loan rates fell 122 basis points during the easing cycle, almost matching the 125-basis-point repo cut. Benchmark-linked repricing accounted for 118 basis points of that movement and reduced system-wide WALR by 24 basis points. Education loan rates fell more, by 162 basis points, but contributed virtually nothing because education loans represent a small share of credit.

The same pattern appeared elsewhere. Rupee export credit rates fell 161 basis points but contributed only five basis points. Large-industry rates fell 67 basis points and contributed nine basis points. The results show why composition is central to understanding average lending-rate movements.