When you apply a standard company analysis framework to a bank — look at debt/equity, operating margin, asset turns — you get meaningless or misleading numbers. Banks are fundamentally different: they borrow money (deposits) and lend it at a higher rate (loans), and the "raw material" they use is money itself. Every metric that matters for a bank is different from what matters for an FMCG, IT, or manufacturing company. This guide explains the complete set of banking metrics that BBS uses in every bank analysis — with real examples from HDFC Bank, SBI, ICICI Bank, IndusInd Bank, and Kotak — so you can read any Indian bank's quarterly results and annual report with full comprehension.
The Bank P&L: Net Interest Income Is the Core Revenue Line
A bank's primary revenue is Net Interest Income (NII) = Interest earned on loans − Interest paid on deposits. The ratio of NII to average assets is the Net Interest Margin (NIM). NIM is the bank equivalent of gross margin: it tells you how much the bank earns on every ₹100 of assets deployed. Indian NIM benchmarks: retail-heavy private banks (HDFC Bank, Kotak) typically earn 4-5% NIM. Public sector banks (SBI, PNB, Bank of Baroda) typically earn 2.5-3.5% NIM. NBFCs often earn 5-8% NIM but carry higher credit risk. A rising NIM is positive (bank is earning more on loans or paying less on deposits). A falling NIM signals either rising deposit costs, a shift to lower-yield assets, or competitive pressure in lending. Use the BBS Stock Scorecard on any bank — the NIM trend over 5 years is one of the primary quality indicators we track. Our detailed analyses of HDFC Bank, ICICI Bank, and SBI vs HDFC Bank each use NIM trend as a central analytical input. Beyond NII, banks earn Non-Interest Income (fee income): processing fees, wealth management fees, credit card fees, forex income, and treasury gains. For HDFC Bank and Kotak, fee income is 1-1.5% of assets — a significant and growing revenue stream that reduces dependence on interest spreads.
- NIM (Net Interest Margin) = NII ÷ Average Assets × 100
- HDFC Bank NIM: ~4.2-4.5% (consistently best-in-class private bank)
- SBI NIM: ~3.0-3.3% (improving from ~2.5% a decade ago)
- IndusInd Bank NIM: ~4.1-4.3% (historically strong, now under pressure due to microfinance)
- NIM above 4% for retail private banks = high quality; below 3% = likely PSU or stressed bank
CASA Ratio: The Most Important Funding Quality Metric
CASA stands for Current Account + Savings Account deposits. Current account deposits pay zero interest. Savings account deposits pay 2.5-4% interest (per RBI-mandated minimums). Together, CASA is the cheapest funding available to a bank. CASA ratio = (CASA deposits ÷ Total deposits) × 100. A high CASA ratio means the bank funds its loans primarily with cheap deposits — directly boosting NIM. A low CASA ratio means the bank is more dependent on Fixed Deposits (FDs) and bulk deposits, which cost 6-8%, compressing margins and making the bank vulnerable when interest rates rise. HDFC Bank consistently maintains a 40-45% CASA ratio — one of the highest among large banks — which is a structural funding advantage that directly explains its superior NIM. SBI's CASA ratio is also strong (~42-45%) because of its franchise reach across rural India. IndusInd Bank's CASA ratio has historically been 35-40% but has been under pressure during stress periods — a key risk flag. When a bank's CASA ratio falls sharply (2%+ in a quarter), investigate: it often signals that depositors are moving money to FDs for higher returns (rate cycle driven, usually temporary) or that the bank is losing brand trust (serious, investigate further). Read our Kotak Mahindra Bank analysis which discusses how Kotak rebuilt its CASA ratio from 25% to 50%+ over a decade as a core part of its business quality story.
NPA: Gross NPA vs Net NPA vs Slippage
Non-Performing Assets (NPAs) are loans where the borrower has not paid interest or principal for 90+ days. This is the most critical risk metric for any bank. Three related numbers to track: Gross NPA ratio = Gross NPAs ÷ Total Loans. This is the headline bad loan number before any provisions. Net NPA ratio = Net NPAs (after deducting provisions) ÷ Total Loans. This is the actual risk to the bank's equity after accounting for money already set aside to cover losses. Slippage ratio = Fresh NPAs added in the quarter ÷ Opening loan book. This is the leading indicator — it tells you whether the bank's asset quality is improving or deteriorating before the Gross NPA ratio moves. A bank where slippage is falling (fewer new bad loans) even if Gross NPA is still high is on a recovery path. A bank where slippage is rising (more new bad loans) even if Gross NPA looks stable is about to deteriorate. IndusInd Bank's microfinance stress in FY25 was visible in rising slippage several quarters before it hit the Gross NPA ratio and consensus analyst earnings estimates. Use the BBS Red Flag Detector on any bank — rising slippage is one of the primary signals we check, along with rising credit cost and falling PCR. Our IndusInd Bank stress analysis is a case study of how to read these signals from publicly available data 2-3 quarters before they become market-moving headlines.
PCR: Provision Coverage Ratio — How Conservatively Is the Bank Provisioning?
Provision Coverage Ratio (PCR) = Total provisions held against NPAs ÷ Gross NPAs × 100. PCR measures how much of its bad loans a bank has already set aside money to cover. A PCR of 70% means the bank has already provisioned for 70 paise of loss per ₹1 of bad loan — leaving only 30 paise of unexpected additional loss that could hit the P&L. PCR benchmark: RBI recommends 70%+. A PCR above 75% is conservative (bank has strong cushion). A PCR below 60% means the bank's future profits will need to absorb significant provisioning as bad loans resolve. PSU banks historically had very low PCR (45-55%) during the 2015-2020 NPA crisis — one reason their earnings remained depressed for years after the NPA recognition was done. HDFC Bank and Kotak typically maintain PCR above 70%, contributing to their earnings predictability. When evaluating a bank with rising NPA, check: is PCR rising (management proactively covering losses) or falling (management hoping loans recover, under-provisioning)? Rising NPAs + rising PCR = bank is being honest and conservative. Rising NPAs + falling PCR = significant earnings risk in coming quarters.
Credit Cost and Return on Assets: The Bottom-Line Metrics
Credit cost = Provisions made in the year ÷ Average loan book × 100. It is the annual "loss rate" on the loan book, expressed as a percentage. Normal credit cost for well-run Indian private banks: 0.5-1.0%. Stressed periods (post-COVID, post-NBFC crisis): 2-3% for weaker banks, 1.0-1.5% for strong ones. Credit cost is directly subtracted from NIM to arrive at risk-adjusted margins — a bank with 4.5% NIM but 2.5% credit cost is effectively earning only 2% on its assets, far worse than a bank with 3.5% NIM and 0.6% credit cost. Return on Assets (RoA) = PAT ÷ Average Assets × 100. This is the bank equivalent of ROCE: it tells you how efficiently the bank converts its asset base into profit. RoA benchmark for strong Indian banks: 1.5-2.0% (HDFC Bank, Kotak, Bajaj Finance-like NBFCs). PSU banks: 0.5-1.0%. Stressed banks: negative RoA during provisioning cycles. Return on Equity (RoE) = PAT ÷ Equity × 100. Banks are highly leveraged businesses (10-12x assets to equity), so RoE amplifies RoA significantly — a 1.5% RoA at 10x leverage produces 15% RoE, which is a strong outcome. Use the BBS PE Analyser on bank stocks and note: the correct valuation metric for banks is Price-to-Book (P/B), not PE. A bank trading at 3.5x book with 18% RoE is fairly priced; the same bank at 1x book is a deep value opportunity if the RoE can recover. Our Axis Bank analysis is the clearest case study of how credit cost normalisation (from 2.5% to 0.7%) drove the entire earnings recovery and re-rating story.
Capital Adequacy Ratio: The Regulatory Floor That Determines Growth Capacity
Capital Adequacy Ratio (CAR / CRAR) = (Tier 1 + Tier 2 Capital) ÷ Risk Weighted Assets × 100. RBI requires Indian banks to maintain a minimum CAR of 11.5% (including capital conservation buffer). Most well-run banks maintain 16-18% CAR — significantly above the minimum — which gives them the capacity to grow their loan book without needing to raise fresh capital. A bank with CAR near the minimum faces a capital constraint: it cannot grow the loan book without either raising equity (dilutive to existing shareholders) or reducing riskier assets (slows growth). When a bank announces a large QIP or rights issue, check its CAR — if it was near regulatory minimum, the capital raise was essential for survival, not growth. If it was already at 18% CAR, the capital raise is opportunistic and aggressive. Tier 1 Capital (mostly equity and retained earnings) is the highest quality capital. Tier 2 Capital (subordinated debt, revaluation reserves) is lower quality and has lower loss-absorption capacity. Strong banks like HDFC Bank and Kotak have Tier 1 CARs of 17-18% — meaning the buffer above the regulatory minimum is entirely equity-backed, the most conservative possible capital structure. Enrol in BBS banking analysis courses to build a complete bank valuation model using all these metrics — the course uses HDFC Bank and SBI as comparative case studies, walking through 5 years of quarterly data to show how NIM, NPA, PCR, credit cost, and RoA interact across different phases of the banking cycle.
🔍 BBS Insight
The single most important thing to check in any Indian bank's quarterly result is the slippage ratio trend — specifically, whether fresh NPA additions are rising or falling as a percentage of the opening loan book. This is the earliest warning signal available in public data. By the time Gross NPA ratio deteriorates significantly, the problem has been building for 3-6 quarters in the slippage data. The BBS bank analysis sequence: (1) Check NIM trend — is the bank maintaining spread? (2) Check CASA ratio — is funding quality stable? (3) Check slippage ratio — is asset quality improving or worsening? (4) Check credit cost — how much is the bank spending to cover losses? (5) Check PCR — is provisioning conservative or optimistic? (6) Compute implied RoA and compare to 5-year median. A bank where NIM is stable, CASA is growing, slippage is falling, and PCR is rising is almost always a buy at reasonable P/B. A bank where any two of these are deteriorating simultaneously warrants immediate investigation before adding or holding the position.