The 50 bps Repo Cut That Stops at the Deposit Counter
The ALCO of a scheduled commercial bank with a loan book of ₹38,000 crore meets a fortnight after the MPC cuts the repo rate by 50 bps to 6.00%, citing CPI inflation easing to 3.9%. The bank's own position is awkward: it raised one-year retail term deposit rates to 7.40% only last quarter to defend a 78% credit-deposit ratio, and ₹4,200 crore of those deposits reprice only next year. On the asset side, 55% of loans are linked to the one-year MCLR, 20% to the external benchmark (repo-linked retail and MSME), 15% are fixed rate and 10% are legacy base-rate loans. Credit growth is 13.5% against deposit growth of 9.8%. Treasury expects another 25 bps cut at the next policy. The CFO wants the one-year MCLR cut by 40 bps at once to protect market share; the risk head warns that net interest margin, now 3.15%, will fall through the 3.00% floor the board has set.
1. Estimate how much of the 50 bps repo cut can realistically reach this bank's borrowers this quarter, and state whether a 40 bps cut in the one-year MCLR is prudent.
Read the worked answer
Pass-through arithmetic. Only two buckets respond to the policy signal:
- EBLR (20% of book): a repo cut passes almost one-for-one, so 0.20 × 50 = 10 bps off asset yield immediately.
- MCLR (55% of book): a 40 bps MCLR cut gives 0.55 × 40 = 22 bps.
- Fixed-rate (15%) and legacy base-rate (10%) loans: no change this quarter.
Total asset-yield give-up = 10 + 22 = 32 bps.
Cost side. Deposits are roughly 70% of liabilities, and the ₹4,200 crore raised at 7.40% does not reprice until next year. Only new and renewing deposits respond, so the cost of funds falls maybe 5–10 bps this quarter.
NIM impact. Income lost ≈ 32 bps × 0.65 (loans/assets) ≈ 21 bps. Cost saved ≈ 8 bps × 0.70 ≈ 6 bps. Net ≈ 15 bps of NIM, taking 3.15% to about 3.00–3.01% - straight through the board floor.
With a 25 bps MCLR cut instead: 0.55 × 25 = 13.75 + 10 = ~24 bps of yield give-up, and about 9–10 bps of NIM compression, leaving headroom. On these numbers a 40 bps MCLR cut breaches the mandate; the defensible move is 20–25 bps.
2. The ALCO must now decide a package. What should it approve, and in what sequence, given that credit is growing faster than deposits?
Read the worked answer
The bank faces a trilemma: it cannot simultaneously cut lending rates, cut deposit rates and fund 13.5% credit growth with 9.8% deposit growth. Something must give, and the sequencing matters more than the size of any single cut.
Step 1 - Protect the liability base first. Do not cut the flagship one-year retail TD rate that was raised last quarter; cutting it now would accelerate the deposit gap and force costlier wholesale funding. Instead cut 20–25 bps only in the 2–3 year buckets and on bulk deposits, where the bank is over-priced relative to peers.
Step 2 - Cut MCLR by 25 bps, not 40 bps, effective from the next reset date, and communicate it as a full pass-through of the 50 bps EBLR signal already delivered. This keeps the NIM around 3.05–3.06%.
Step 3 - Fill the funding gap without price wars: refinance lines (NABARD, SIDBI, MUDRA), certificates of deposit for tenor-matched needs, and deposit mobilisation through salary/transaction accounts rather than rate.
Step 4 - Defend margin through mix, not price: fee income, tighter spreads on unsecured EBLR-linked personal and MSME loans, and repricing of the fixed-rate book on renewal.
A 40 bps MCLR cut would buy market share the bank cannot fund - it would have to re-raise deposits at a higher rate within two quarters, reversing the gain.
3. Why is the one-year MCLR a weak transmission benchmark, and what would structurally improve pass-through for this bank?
Read the worked answer
MCLR is an internal, bank-specific benchmark computed from the marginal cost of funds, with a minimum 92-day reset. Three consequences follow:
- It transmits deposit costs, not policy. When retail deposit competition is fierce - as now, with mutual funds and small savings competing - MCLR stays sticky even as the repo falls.
- It is not comparable across banks, so a customer cannot see whose loan is genuinely cheaper, and competition on the benchmark is muted.
- It lags by design. A 50 bps repo cut on 1 April may not appear in the one-year MCLR until July, and in existing borrowers' EMIs later still.
That is precisely why the external benchmark regime (repo-linked pricing for retail and MSME floating loans) was introduced - pass-through there is near-immediate, which is why the 20% EBLR bucket moved 10 bps on the day of the announcement while the other 80% did not.
Structural fixes for this bank: raise the EBLR share on new floating loans and migrate base-rate borrowers on renewal; build a genuinely transactional CASA franchise so deposit pricing is less rate-sensitive; use interest rate swaps to hedge the 15% fixed-rate book rather than pricing in a risk premium; and fund asset growth with tenor-matched refinance so the marginal cost of funds stops being set by the most expensive retail bucket.