SEBI Registered Research Analyst INH000027478 support@skygrowthwealth.com +91 98765 43210
Equity Selective BanksNIMFinancials

Private Banks: Where the Margin Reset Actually Lands

Deposit repricing is not uniform. We decompose the NIM impact across eight lenders and identify which balance sheets absorb it best.

R. DeshpandeSenior Analyst — Financials 11 Aug 2026 11 min read 24 pages
Executive summary

Deposit repricing is not landing evenly across lenders. We decompose the net interest margin impact across eight private banks and identify which balance sheets absorb the reset with the least damage to return on assets.

Lenders modelled
8
NIM spread
29bps
Rate path
Common
Model horizon
7 qtrs

Key findings

  1. CASA composition explains most of the variance

    The spread in modelled NIM compression across the group is wide, and the dominant explanatory variable is the current and savings account share rather than loan mix.

  2. Fee income is the underrated offset

    Two lenders in the group generate enough non-interest income to hold return on assets roughly flat through the reset. That is not reflected in their relative valuation.

  3. Credit cost normalisation compounds the pressure

    The reset arrives at the same time as provisioning returns toward through-cycle levels. Modelling the two independently understates the combined effect on earnings.

Modelled NIM path by lender cohort Sample data
-54-39-24-96Q0Q2Q4Q6Q7
High CASA cohort Low CASA cohort

Basis points relative to the starting quarter. Sample data.

Decomposing the margin effect

We rebuild each bank's interest-earning asset and interest-bearing liability schedule by repricing bucket, then apply a common deposit cost path across all eight. Holding the rate path constant isolates balance-sheet structure as the only differentiating variable, which is the point of the exercise.

The output is a modelled NIM path per bank rather than a single point estimate. We publish the full schedule in the attached model so the rate path can be changed and the answer recomputed.

Where the offsets are real

Non-interest income quality varies considerably. Fee streams tied to transaction volume behave differently from those tied to distribution of third-party products, and only the former holds up reliably when balance-sheet growth slows.

We separate the two in the model, because treating fee income as a single line materially overstates the resilience of two names in the group.

Positioning and invalidation

This is a selective view rather than a sector call. The dispersion in outcomes is wide enough that a basket approach captures the compression without the offsets.

The thesis breaks if deposit competition intensifies beyond the rate path modelled, or if credit costs normalise faster than the provisioning trajectory we assume. Both are observable quarterly.

How to read this note. Views expressed are the analyst's own, formed from the evidence cited and the model attached. They are research, not personalised advice, and do not account for your circumstances, horizon or risk capacity. Investments in securities markets are subject to market risks. Read all related documents carefully before investing. Past performance is not indicative of future results.
Report package

Download the full report and model

24-page PDF plus the three-statement model as an editable spreadsheet. Change an assumption and see what the answer becomes.

Downloads are available to subscribers on the Professional and Premium tiers. Demo links on this build are inactive.

R. Deshpande

Senior Analyst — Financials · SkyGrowthWealth Research

Maintains the model behind this note and publishes every revision to it.

Share

Subscribe

Every note, with the model attached

Read the reasoning, change an assumption, rerun the model. That is what a research subscription should give you.

Research is educational and informational. It is not personalised investment advice.