ALCO in a rate-cut cycle: 38% repo-linked assets against 30% savings liabilities
The asset-liability committee of a mid-sized bank meets a week after the Monetary Policy Committee cut the policy repo rate by 25 basis points. The balance sheet position is:
- Advances: ₹1,20,000 crore at an average yield of 9.20%. Of this, 38% is floating and linked to an external benchmark (repo), repricing immediately; the remaining 62% is on MCLR, and existing MCLR loans reset only at the annual reset date.
- Deposits: ₹1,45,000 crore - 30% savings at 2.75%, 20% current at nil, and 50% term deposits at an average 7.10%. Term deposits of ₹22,000 crore come up for renewal in the next quarter.
- The treasury expects a further 25 bps cut within six months.
The branch network's view is that cutting the savings rate will trigger deposits moving to competitors. The credit team argues that each 25 bps cut is "revenue-neutral" because the repo-linked book is only about a third of advances.
1. Assume the bank passes the full 25 bps cut to savings depositors and does nothing else. Compute the effect on net interest income and net interest margin, and evaluate the credit team's "revenue-neutral" claim.
Read the worked answer
Current position
- Interest income = 9.20% × ₹1,20,000 cr = ₹11,040 cr
- Interest expense = (2.75% × ₹43,500 cr) + (0% × ₹29,000 cr) + (7.10% × ₹72,500 cr) = ₹1,196.25 + 0 + ₹5,147.50 = ₹6,343.75 cr
- NII = 11,040 − 6,343.75 = ₹4,696.25 cr; NIM on advances = 4,696.25 / 1,20,000 = 3.91%
Effect of the 25 bps cut
- Repriced advances = 38% × ₹1,20,000 cr = ₹45,600 cr → income falls 0.25% × 45,600 = −₹114.00 cr
- Savings repriced = 30% × ₹1,45,000 cr = ₹43,500 cr → cost falls 0.25% × 43,500 = +₹108.75 cr
- Net change = −₹5.25 cr; new NII = ₹4,691 cr; new NIM = 3.909%
Verdict on "revenue-neutral": the arithmetic looks almost flat only because the savings pass-through happens to offset the asset repricing - and it leaves the bank marginally worse off, not neutral. The claim is also framing the wrong denominator: the exposure is not "a third of advances" versus the whole deposit base; it is ₹45,600 cr of immediately repricing assets against ₹43,500 cr of immediately repricing liabilities. A bank with a larger repo-linked book (say 60%) would lose ₹180 cr against the same ₹108.75 cr saving. Neutrality here is a coincidence of this specific mix, not a property of a 25 bps cut.
2. Assuming the ₹22,000 crore of term deposits is renewed at a rate 20 bps lower and the branch network's objection limits the savings cut to 15 bps, quantify the combined position and recommend what ALCO should do.
Read the worked answer
Revised arithmetic
- Asset repricing: −₹114.00 cr (unchanged, full pass-through is contractual on the external-benchmark book)
- Savings at 15 bps instead of 25 bps: 0.15% × ₹43,500 cr = +₹65.25 cr
- Term deposits: 20 bps on ₹22,000 cr, but only for the part of the year after renewal - assume renewals spread evenly and the benefit accrues for about 9 of the next 12 months: 0.20% × 22,000 × 9/12 = +₹33.00 cr
- Net = −114.00 + 65.25 + 33.00 = −₹15.75 cr; NII ≈ ₹4,680.5 cr; NIM ≈ 3.90%
So the partial savings cut is worse than the full one (−15.75 vs −5.25). The branch network's fear is expensive: withholding 10 bps of savings pass-through costs about ₹43.5 cr a year in foregone margin, far more than the deposit outflow risk being managed - unless it really triggers migration.
ALCO recommendation
- Cut savings by the full 25 bps, but protect small balances by keeping the rate on balances up to, say, ₹1 lakh higher than on larger balances - this addresses the branch's genuine concern (the small saver is rate-sensitive and vocal) while repricing the large, elastic balances.
- Push term-deposit rates down on the 1–2 year bucket now, ahead of the next cut, rather than waiting for the renewal calendar.
- Do not reprice the MCLR book downwards in a hurry; the 62% MCLR book is the only cushion against a further 25 bps cut, and resetting it early gives that cushion away.
- Extend the duration of funding selectively (3–5 year retail term deposits) while rates are still falling.
3. Explain why the external-benchmark book transmits the rate cut within days while the MCLR book does not, and state one structural reason why deposit rates remain sticky. Relate both to the ALCO decision.
Read the worked answer
Asset side. Loans priced off an external benchmark (repo, or the 91-day T-bill / 3-month MIBOR variants permitted by the RBI) are contractually benchmark + spread, with the benchmark reset at least once every three months - in practice immediately for repo-linked loans when the policy rate changes. The spread is fixed at sanction and cannot be widened to absorb the cut, so 100% of a 25 bps repo cut hits yield within days. MCLR is an internal cost-of-funds benchmark computed from the marginal cost of deposits plus negative carry on CRR/SLR and return on capital; it is revised by the bank at its own discretion, and even when MCLR moves, an existing loan reprices only at its next reset date. So a 25 bps cut reaches the MCLR book with a lag of months, and partially. This is precisely why the regulator moved a large share of retail and MSME floating loans to external benchmarking - the older MCLR regime was shown to transmit only a fraction of policy cuts.
Liability side. Savings deposits - 30% of this bank's funding - are the stickiest input. They are repayable on demand at a rate the bank administers, competitive pressure on the small-saver segment is high, and banks are reluctant to be the first mover down because deposit franchises are hard to rebuild. Current accounts (20%) pay nothing and cannot fall further. So on a 25 bps cut, roughly 62% of the deposit base either cannot reprice or reprices with a market-determined lag, against an asset book that reprices faster.
Consequence for ALCO. This asymmetry is directional: when rates fall, liabilities lag and NIM compresses; when rates rise, the same lag works in the bank's favour. The ALCO should therefore treat the current cut as a hit to NIM that must be managed on the deposit side - through the term-deposit calendar, small-balance tiering and funding duration - rather than pretending the asset side will absorb it. It should also model the second 25 bps cut now, because on this mix the cumulative margin loss roughly doubles and the cushion in the MCLR book shrinks with every reset.