Should the Bank Lend? Credit Appraisal Case Framework for Interviews
The loan file looks solid until one line in the bank statement changes the room: cash collections are seasonal, but the proposed EMI is monthly and unforgiving. That is the real tension in lending - not “is this a good business?”, but “will cash arrive in time to repay debt even when things go wrong?”
- A bank lends to repayment capacity, not optimism. Collateral is the second way out, not the first.
- Use the 5Cs of credit: Character, Capacity, Capital, Collateral and Conditions.
- The core ratio is DSCR = cash flow available for debt service / scheduled principal plus interest.
- A good credit answer ends with a decision: approve, reject, or approve with structure.
- Stress the case: ask what happens if sales fall, receivables stretch, or input costs rise.
- For India, mention CIBIL or bureau record, GST trail, bank statements, RBI norms, collateral enforceability and 90-day NPA discipline.
Big Picture
A lending decision is a layered judgment. The bottom layers answer “is this borrower eligible and honest?” The middle layer answers “can cash flows service debt?” The top layers decide “what security, covenants and monitoring protect the bank if the base case slips?”
Core Framework: How to Decide If the Bank Should Lend
The cleanest answer is not a list of ratios. It is a sequence: understand the borrower, identify the repayment source, test the numbers, assess behaviour, structure the loan and then give a recommendation.
The 5Cs of Credit, Made Usable
The 5Cs are useful because they force you to examine both numbers and behaviour. A borrower can have strong collateral but weak character, or high growth but thin cash flow. Banks get hurt when they confuse one strength for the whole case.
The Ratios a Bank Actually Cares About
Ratios do not replace judgment, but they make judgment disciplined. Use these as interview-safe benchmarks for a non-financial operating borrower; sector, product and bank policy can change the exact threshold.
Worked Example: Approve, Reject or Restructure?
Suppose an SME asks for a ₹10 crore term loan to add capacity. The bank estimates the following from audited statements, bank statements and management projections. This is an illustrative case, not a real company.
Recommendation: do not approve the ₹10 crore loan as proposed. A stronger answer is “conditional approval” - reduce the loan amount, ask for promoter contribution, improve collateral cover, create an escrow of receivables and add covenants on DSCR and additional borrowing. The borrower may be lendable, but the proposed structure is not safe enough.
Definitions You Should Say Cleanly
Credit risk, Basel Committee: “the potential that a bank borrower or counterparty will fail to meet its obligations in accordance with agreed terms.”
DSCR: Cash flow available for debt service divided by scheduled principal plus interest due in the period.
NPA in India: A term loan generally becomes non-performing when interest or principal remains overdue for more than 90 days.
Collateral: An asset pledged to the lender as a secondary recovery source if the borrower defaults.
Case Study: IDFC FIRST Bank and the Shift to Granular Lending
IDFC FIRST Bank shows why lending quality improves when a bank moves from concentrated exposure to more granular, cash-flow-underwritten borrowers.
Situation: IDFC Bank historically had exposure to infrastructure and wholesale lending, where ticket sizes are large, projects are long-gestation and repayment often depends on execution, approvals and refinancing. After the 2018 merger with Capital First, the combined bank had an opportunity to reshape its lending model.
The move: IDFC FIRST Bank publicly emphasized building a more retail and granular loan book, supported by branch banking, liability franchise building and underwriting of individual and small-business borrowers. The primary driver was reducing concentration risk and improving the predictability of repayment. Supporting drivers included access to retail customer data, tighter credit processes, diversified deposits and a lending mix less dependent on a few large borrowers.
The lesson: In a “should the bank lend?” case, size is not safety. A large borrower with weak cash-flow visibility can be riskier than many smaller borrowers with verified income, bureau history and disciplined collections. The bank’s job is to avoid concentration, verify repayment capacity and structure exposure so that one borrower cannot damage the whole portfolio.
How AI Changes Should the Bank Lend?
1. Cash-flow underwriting becomes sharper. In India, lenders increasingly use digital trails such as bank statements, GST data, bureau records and Account Aggregator-enabled financial data to assess income stability. AI can detect seasonality, cash-flow volatility, circular transactions and sudden drops faster than manual statement review.
2. Early warning systems become more real-time. Models can flag changes in account conduct, delayed collections, falling GST filings, news risk, cheque returns or unusual utilization patterns before the account becomes overdue. The credit officer still owns the judgment, but AI improves the watchlist.
3. Credit memos become faster, but bias risk rises. LLMs can summarize annual reports, loan documents, covenants and site-visit notes. The risk is over-trusting a model that may miss context or embed bias. In regulated lending, explainability, audit trails, RBI expectations and data-privacy discipline matter.
Use NotebookLM with an annual report, credit-rating rationale and your own ratio calculations. Ask it: “Generate a one-page credit memo with repayment source, 5Cs, red flags, stress case and final lending recommendation.” Use only public or anonymized documents.
Interview Relevance
“A mid-sized manufacturing company wants a ₹50 crore term loan for capacity expansion. Revenues are growing, but receivables have stretched. Should the bank lend?”
Always separate business quality from credit quality. A fast-growing business may still be a poor borrower if working capital absorbs cash and DSCR is thin.
The mistake that costs candidates is approving a loan because the borrower has collateral or a famous promoter. That fails because the first source of repayment is cash flow, not asset sale. Fix: start with DSCR and cash-flow reliability, then use collateral and covenants as protection.
What to Revise Next
You now know how a bank thinks about lending risk. Next, revise how capital markets think about pricing and ownership value.