Business Models: How Banking & Lending Players Make Money
A kirana owner once waited weeks for a loan officer to inspect ledgers and collateral. Now the same business may receive a pre-approved credit line on an app - but behind that smooth screen, someone still has to fund the money, price the risk, collect repayments and absorb losses.
- Banks and lenders do not simply “sell money.” They earn by transforming funding into risk-priced loans and services.
- The core profit engine is: interest income + fees - funding cost - credit losses - operating cost.
- Banks usually have a lower funding cost because they accept deposits; NBFCs and fintech lenders often rely more on borrowings, co-lending, securitisation or distribution partnerships.
- Net interest income is the rupee spread between what the lender earns on loans and what it pays for funds.
- A high loan yield is meaningless unless you also check asset quality, credit cost, collections and cost of funds.
- Fee income - processing fees, interchange, insurance distribution, wealth products and servicing fees - makes the model less dependent on pure lending spreads.
- In interviews, explain the business model as a unit economics problem plus a risk management problem.
The Big Picture: Banking Is a Spread Business With a Risk Filter
At the simplest level, a banking or lending player raises funds, lends them to customers, earns a spread, charges some fees and manages the risk that borrowers may not repay. The mistake is to stop at “interest income.” The real model is risk-adjusted spread.
The Core Business Model: Where the Money Comes From
A banking and lending business has five linked money engines. A strong answer connects all five, because each one affects the others.
1. Interest Spread - The Main Engine
The lender earns interest income on loans and investments. It pays interest expense on deposits, bonds, bank borrowings or other funding. The difference is net interest income.
Net Interest Income = Interest Earned - Interest Paid.
For example, if a lender borrows at 9% and lends at 16%, the apparent spread is 7%. But that is not profit yet. From that 7%, it still has to pay for credit losses, collections, branches, technology, employees and capital.
2. Fee Income - The Less Capital-Heavy Layer
Lenders also earn fees that do not always require putting large loans on the balance sheet. Common examples include:
- Processing fees on loans
- Card and interchange income in payments-linked models
- Distribution commission from insurance, mutual funds or wealth products
- Loan servicing fees when the lender originates or manages loans for another institution
- Late payment and collection-related charges, subject to regulation and customer fairness norms
Fee income matters because it improves profitability without always increasing credit risk at the same speed.
3. Float and Transaction Banking
Commercial banks can earn from the movement and holding of money. Current accounts, salary accounts, merchant payments, cash management and trade services create low-cost balances and recurring fee income.
This is why a bank wants a company’s salary account, vendor payments and collections business - not just its term loan. The operating account can become a cheaper funding source and a data source for underwriting.
4. Treasury and Investment Income
Banks and large lenders hold investments for liquidity, regulatory requirements and yield. Treasury teams manage government securities, money-market instruments and interest-rate risk.
This income can help profitability, but it is not a substitute for a sound lending book. A lender with weak underwriting cannot “trade its way out” forever.
5. Partnerships and Balance-Sheet Light Models
Modern lending players may not always hold the full loan themselves. They can earn by originating, underwriting, servicing or co-lending.
- Co-lending: A bank and NBFC share a loan, often combining bank funding with NBFC reach.
- Loan marketplace: A fintech platform generates leads and passes borrowers to regulated lenders.
- Originate-to-distribute: A lender originates loans and sells down exposure through securitisation or assignment.
- Embedded lending: Credit is offered inside a commerce, payments or software journey.
Bank vs NBFC vs Fintech Lender: The Two-Sided Comparison
All lenders price risk, but their raw material differs. A bank’s raw material is usually deposits. An NBFC’s raw material is often wholesale borrowing, market borrowing or bank lines. A fintech lender may primarily provide acquisition, data, underwriting technology or servicing while partnering with a regulated lender.
The Lending Profit Formula: A Simple Worked Example
Use this example when you need to explain the economics quickly. The numbers are illustrative, not company-specific.
Suppose a lender gives a ₹100 loan for one year.
The apparent spread was 9 percentage points, but pre-tax profit was only ₹2 on ₹100. That is the heart of lending: small changes in defaults, funding cost or operating cost can wipe out the margin.
The Lending Lifecycle: Where Value Is Created or Destroyed
A lender’s business model is only as good as its credit process. Bad underwriting creates future NPAs. Weak collections turn small delays into losses. Poor funding discipline makes even good borrowers unprofitable.
Definitions You Should Be Able to Say Cleanly
Business model: “A business model describes the rationale of how an organization creates, delivers, and captures value.” - Alexander Osterwalder and Yves Pigneur, Business Model Generation
- Net Interest Income: Interest earned on loans and investments minus interest paid on deposits and borrowings.
- Net Interest Margin: Net interest income divided by average earning assets.
- Credit cost: Loan-loss provisions during a period divided by average loans or advances.
- Asset quality: The health of the loan book, usually tracked through overdue loans, NPAs, write-offs and collection efficiency.
- Cost of funds: The average interest rate a lender pays to raise money for lending.
- Capital adequacy: The cushion of regulatory capital available to absorb losses and support growth.
Key Metrics That Reveal the Business Model
When you analyse a bank or lender, do not jump straight to profit after tax. First check whether the profit is coming from durable economics or hidden risk.
A lender with high NIM but high credit cost may be taking risky loans. A lender with low NIM but very low cost of funds, strong fee income and clean asset quality may be more valuable. Always read the metrics as a system.
Case Study: Five Star Business Finance and the MSME Lending Model
Five Star Business Finance shows how an Indian NBFC can build a lending model around underserved small business borrowers by combining local underwriting, secured lending and high-touch collections.
Situation: Many micro-entrepreneurs and self-employed borrowers operate with informal income records. They may have a viable business, daily cash flows and local reputation, but not the kind of standard salary slip or audited financial trail that a traditional bank prefers.
The move: Five Star’s model focuses on secured small business lending to this borrower segment. The business logic is not “lend fast and hope.” It is to build underwriting around field assessment, property-backed security, borrower cash-flow understanding and close collections discipline.
The outcome and lesson: The strategic point is that a lender can profit in a difficult segment if its operating model is designed for that segment. The primary driver is local, cash-flow-aware underwriting backed by security. Supporting drivers include branch-level customer knowledge, repeatable credit processes, collection discipline and funding access appropriate to an NBFC model.
This is the interview-worthy insight: in lending, customer access alone is not a moat. The moat is access plus underwriting data, collection muscle, funding discipline and repeatable risk controls.
How AI Changes Banking & Lending Business Models
AI does not remove the basic lending formula. It changes how accurately and cheaply the lender can acquire, underwrite, monitor and collect.
- AI-assisted underwriting: Models can analyse bank statements, GST-style business flows, repayment history, device signals and behavioural data to estimate repayment capacity. The business impact is faster decisions and better risk segmentation, but lenders must manage bias, explainability and regulatory scrutiny.
- Early-warning and collections intelligence: Machine learning can flag borrowers likely to miss repayments, prioritise collection queues and suggest the next best action. This protects the spread by reducing slippages and collection cost.
- GenAI for relationship and credit teams: LLM tools can summarise borrower documents, draft credit memos, compare covenant breaches and prepare customer conversation notes. Human approval remains critical because credit accountability cannot be outsourced to a model.
Use NotebookLM: upload a lender’s annual report, investor presentation and this lesson. Ask it to create a one-page “revenue model map,” five likely interview questions and a risk checklist covering NIM, credit cost, GNPA, funding mix and fee income.
Interview Relevance
“Explain how a banking or lending company makes money. How would you compare a bank, an NBFC and a fintech lending platform?”
If the interviewer names a company, do not describe the entire banking sector. Identify that company’s dominant engine - deposits, secured lending, cards, gold loans, MSME lending, payments, co-lending or wealth distribution - and then explain the economics around that engine.
Common Mistake
The mistake: Saying “lenders make money by charging higher interest than they pay” and stopping there. Why it costs candidates: it ignores defaults, provisions, operating cost, capital and regulation - the exact things that decide whether lending is profitable. One-line fix: always say “spread after credit losses and costs,” not just “spread.”