Asset Classification, Bad Loans & Provisioning Norms in India - Interview Revision Guide
A bad loan is not a loan the bank “feels nervous about.” It becomes a bad loan when time, cash flows and RBI rules say so - usually after the borrower crosses a defined overdue threshold and the bank must stop pretending the income is normal.
- Asset classification is the RBI-driven process of tagging bank loans as Standard, Sub-standard, Doubtful or Loss based on repayment performance and recoverability.
- A loan generally becomes a Non-Performing Asset when interest or principal remains overdue for more than 90 days.
- SMA buckets are early-warning stages before NPA: SMA-0 is 1-30 days overdue, SMA-1 is 31-60 days, and SMA-2 is 61-90 days.
- Once an account becomes NPA, interest is usually recognized on cash basis, not accrual basis.
- Provisioning is the expense banks set aside against expected loss: higher ageing and lower security cover mean higher provision.
- Key interview metrics are GNPA ratio, NNPA ratio, Provision Coverage Ratio, Slippage Ratio, Credit Cost and Write-off Ratio.
- The biggest trap: treating “provision,” “write-off,” “loss” and “NPA” as the same thing. They are related, not identical.
Think of a bank loan as moving through a disciplined risk-recognition loop. The same borrower can start as healthy, show early stress, become NPA, attract provisions, and then either recover, upgrade after sustained repayment, or be written off while recovery efforts continue.
The Core Idea: Banks Must Recognize Credit Deterioration Early
A bank’s biggest accounting risk is simple: showing interest income today on a loan that may never pay tomorrow. RBI asset classification norms force banks to recognize that deterioration through standard categories, income-recognition rules and provisioning.
There are three linked concepts:
- Asset classification: the loan’s status based on repayment and recoverability.
- Bad loans or NPAs: loans that have stopped generating income normally for the bank.
- Provisioning: the hit taken in the profit and loss account to absorb expected loss.
Definitions You Must Be Able to Say Clearly
RBI: “An asset, including a leased asset, becomes non-performing when it ceases to generate income for the bank.”
An NPA is a loan or advance where interest and/or instalment of principal remains overdue for a period of more than 90 days.
The RBI Classification Ladder: From Early Stress to Loss
The most interview-friendly way to remember the system is by time. A borrower may move from regular repayment to SMA, then NPA, then deeper NPA categories depending on ageing and recovery prospects.
Provisioning Norms: What the Bank Must Set Aside
Provisioning is conservative banking discipline. If a loan is riskier, older in default, unsecured or unlikely to be recovered, the bank must absorb more loss through provisions.
For interview revision, remember the broad RBI structure below. Actual regulatory treatment can vary by asset type, security, restructuring status and RBI updates, so state the core norms and add “subject to RBI category-specific rules.”
A provision is an accounting charge against profit. A write-off removes the asset from the balance sheet, but recovery action can still continue.
Income Recognition: Why NPA Status Hits Profit Twice
Once a loan becomes NPA, the bank usually cannot keep booking interest income on accrual basis. Interest on NPAs is recognized when actually received, not merely when due.
This creates a double hit:
- Revenue hit: interest income stops being accrued normally.
- Cost hit: provisions increase in the profit and loss account.
- Capital hit: lower profit can reduce internal capital generation.
Key Metrics to Track in Asset Quality
When analyzing a bank, do not stop at “NPAs are high or low.” Use a small dashboard that separates stock, net exposure, provisioning strength and fresh deterioration.
Worked Example: How Provisioning Changes the Reported Picture
Assume a bank has gross advances of ₹10,000 crore. Out of this, ₹300 crore are gross NPAs. The bank has already made ₹210 crore of provisions against these NPAs.
The learning: GNPA tells you the stock of bad loans; NNPA and PCR tell you how much loss has already been absorbed.
Case Study: IDFC FIRST Bank and the Clean-up Logic of Provisioning
IDFC FIRST Bank shows how a lender can move from legacy wholesale stress toward a more granular retail-led book by recognizing, providing and rebuilding asset quality.
IDFC FIRST Bank was formed after IDFC Bank merged with Capital First. The combined institution inherited a mix of infrastructure, wholesale and retail exposures. Some legacy wholesale accounts carried higher concentration and stress risk - exactly the kind of book where asset classification and provisioning discipline matter.
The strategic move was not one magic action. The primary driver was balance-sheet clean-up through recognition, provisioning and reduction of legacy stressed exposure. Supporting drivers included shifting the loan book toward granular retail and commercial banking, building a stronger deposit franchise, improving risk analytics, and tightening credit processes.
The outcome lesson is powerful: a bank does not “solve” NPAs only by recovering old loans. It improves asset quality by combining upfront recognition, adequate provisions, portfolio redesign, better underwriting and disciplined collections.
How AI Changes Asset Classification, Bad Loans & Provisioning Norms in India
AI does not replace RBI classification rules. A loan overdue for more than 90 days is still governed by the rulebook. What AI changes is how early and how accurately a bank can detect stress before the account formally becomes NPA.
- Early Warning Systems become sharper: ML models can flag behavioural signals such as delayed GST payments, declining account credits, cheque bounces, bureau deterioration, EMI bounce patterns and sector-level stress.
- Collections become segmented: AI can prioritize borrowers by probability of cure, expected recovery value and best contact strategy, improving effort allocation without treating all overdue accounts alike.
- Provisioning analytics become more forward-looking: scenario models can estimate likely slippages under stress conditions, although regulatory provisioning must still follow RBI norms and bank-approved models.
Use NotebookLM: upload a bank's latest annual report, RBI asset-quality disclosures and this lesson, then ask: “Create 10 interview questions on this bank's GNPA, NNPA, PCR, slippages and provisioning quality, with model answers.”
Interview Relevance
“Explain how a bank classifies loans as NPAs in India. How do provisioning norms affect the bank's profitability and balance sheet?”
If asked to analyze a bank, never quote GNPA alone. Add NNPA and PCR immediately - that shows you understand both stress and loss absorption.
Common Mistake
The single biggest mistake is saying “the bank wrote off the NPA, so the loan is forgiven.” That costs candidates because write-off is an accounting action, not automatic borrower waiver. One-line fix: say, “A write-off removes the asset from books after provisioning, but recovery proceedings can continue.”
What to Revise Next
Now move from recognizing bad loans to resolving them and capitalizing banks against risk. Revise these next: