Stagflation, Bank Margins and the Policy Mix
In a mid-sized open economy, real GDP is growing at 4.1% against estimated potential growth of 6.5%; the output gap was about −1.0% a year ago. Headline CPI is 6.9%, core 5.2%, food 9.8% and fuel 8.4%. Unemployment has risen from 7.4% to 8.2%. The policy rate is 6.50% and the currency has depreciated 7% in six months. Bank credit is growing 14% against deposit growth of 11%; the system's CASA ratio has fallen from 44% to 38%. The fiscal deficit is 5.6% of GDP and public debt is 82% of GDP. The central bank's flexible inflation target is 4% ±2%.
At the next policy meeting the committee must choose between a 50 bp hike, a 25 bp hawkish hold and a cut. The finance ministry is pressing for lower rates to support growth and has already announced additional fuel subsidies. A large bank has asked how the decision affects its net interest margin, given that 30% of its loan book is repo-linked, 40% is MCLR-linked and 30% is fixed-rate.
1. Diagnose the inflation, compute the output gap and the real policy rate, and say whether a rate cut is defensible.
Read the worked answer
Output gap. Potential growth 6.5% against actual 4.1% widens the gap by 2.4 percentage points a year. From −1.0% last year, the gap is now approximately −3.3% - a large demand shortfall, not an overheating economy.
Real policy rate. Ex post: 6.50 − 6.90 = −0.4%. Using expected inflation of about 6.2%, roughly +0.3%. Policy is accommodative-to-neutral, not restrictive.
What kind of inflation is this?
- Food 9.8% and fuel 8.4% are supply shocks: a rate rise does not grow grain or refine diesel.
- A 7% depreciation adds imported inflation; with roughly 20% import content in the CPI basket, that is about 1.4 percentage points.
- Core at 5.2% is the problem. Core is where second-round effects - wage claims, margin restoration, expectations - show up, and it is 1.2 points above target even with a −3.3% output gap. That combination says expectations are at risk of drifting.
- Adding headline 6.9% and unemployment 8.2% gives the textbook stagflation signature.
Is a cut defensible? Weakly. The output gap and rising unemployment argue for easing, but with inflation above the 6% upper tolerance for several quarters, a negative real rate, a depreciating currency and sticky core, a cut would signal that the target is negotiable. It would probably weaken the currency further, importing more inflation - the classic case where easing is self-defeating.
Verdict. A cut is not defensible on these numbers. A 25 bp hike or a hawkish hold with a clear conditional path is the defensible middle; 50 bp into a −3.3% gap risks a sharp slowdown for little food-price relief.
2. Estimate the impact of a 50 bp policy rate increase on the bank's net interest margin, quarter by quarter, and identify the vulnerable balance-sheet profile.
Read the worked answer
Asset side.
- Repo-linked (30%): reprices immediately → +0.30 × 0.50% = +0.15%
- MCLR-linked (40%): passes ~25 bp over two quarters → +0.40 × 0.25% = +0.10%
- Fixed-rate (30%): 0% within the year
Liability side. CASA 38% reprices little; term deposits 62% reprice about 40 bp of the 50 bp move → 0.62 × 0.40% = +0.25%
Quarter 1. Asset yield +0.15%, cost of funds +0.25% → NIM compresses by about 10 bp. Deposits reprice before the loan book does; this is the classic transmission asymmetry.
Quarters 2–3. MCLR reset adds 0.10% to yields; asset +0.25% against cost +0.25% → roughly neutral. Full-year NIM is close to flat unless funding costs overshoot.
The vulnerable profile. Three features make a bank fragile in a hiking cycle: (i) a low and falling CASA ratio (44% → 38% here) because savers shift into term deposits and market instruments as rates rise - if CASA falls to 34% and term deposits must pay 60 bp to defend volume, cost of funds rises another 10–15 bp and the NIM gain disappears; (ii) a large fixed-rate book, which delays repricing and exposes the bank to margin squeeze if funding costs keep climbing; (iii) heavy reliance on wholesale funding, where rates reprice instantly.
Also watch. A 50 bp rise lifts bond yields, producing mark-to-market losses on the AFS/HFT portfolio - a hit to capital, not to NIM, but often bigger in the short run. Management actions: reprice EBLR-linked retail loans, defend CASA with transaction banking rather than rate, and hedge the fixed-rate book.
3. The finance ministry wants lower rates while expanding fuel subsidies. Assess the policy mix, including the cost of disinflation.
Read the worked answer
The mix is internally inconsistent. A fiscal deficit of 5.6% of GDP with debt at 82% of GDP, plus untargeted fuel subsidies, adds to demand at precisely the moment the central bank is being asked to hold or raise rates. Untargeted fuel subsidies are also regressive - they transfer most to large fuel users - and they blunt the price signal that would reduce consumption. If the central bank is pressed to cut into this, the result is fiscal dominance: the inflation target is subordinated to the borrowing programme, risk premia rise, and the currency weakens further.
What fiscal policy should do instead.
- Target transfers to vulnerable households rather than universal price subsidies.
- Shift spending toward supply capacity - logistics, cold chains, power - which raises potential growth and reduces the −3.3% output gap without adding inflation.
- Publish a credible consolidation path; the debt path, not the current deficit alone, drives the risk premium.
What monetary policy should do. Hold or hike modestly, communicate the reaction function clearly, and use liquidity tools separately from the rate. There is a case for tolerating a slower return to 4% given the supply shock, but not for abandoning the anchor.
The cost of disinflation. With a sacrifice ratio of about 1.5, cutting inflation from 6.9% to 4% (2.9 points) costs roughly 1.5 × 2.9 ≈ 4.4% of a year's GDP in lost output. If nominal GDP is ₹330 lakh crore, that is about ₹14 lakh crore. The judgement is whether that cost is better paid now, with expectations still near target, or later, after expectations have reset upward and the sacrifice ratio itself has risen. Credibility is cheaper to keep than to rebuild.