HOW BANKING WORKS · CREDIT DECISIONS 17
The central credit question is not “Does this borrower look good?” It is “Which future cash flow repays this exact obligation?”
Every loan is a claim on the future. The bank gives purchasing power now and expects interest and principal to return later. That means credit underwriting is fundamentally an exercise in representing an uncertain future without pretending that uncertainty has disappeared.
A borrower can be wealthy and still be unable to service a badly structured loan. A profitable company can still run out of cash. A household with a high salary can still carry too many obligations. Good underwriting therefore begins with repayment mechanics rather than appearances.
This article is part of the How Banking Works authority spine. Batch 05 moves from payments back to the point where banking decides which future promises are strong enough to become bank assets.
Step 1: identify the primary repayment source
The first question is simple enough to sound obvious: where will the repayment money come from?
- For a salaried household, it may be employment income after normal living costs and existing debt.
- For a small business, it may be operating cash flow generated by sales and margins.
- For a property investment, it may be rental income plus the owner’s other resources.
- For project finance, it may be cash generated by the project itself.
- For working capital, it may be the conversion of inventory and receivables back into cash.
The bank should be able to point to a primary cash-flow engine. Collateral is normally a secondary recovery route, not the first repayment plan.
Step 2: measure income differently from cash flow
Accounting profit, salary and cash flow are related but not interchangeable. A business can report profit while cash is trapped in unpaid invoices. A household can have a high gross salary while mandatory expenses and other loans consume most of it.
The bank therefore tries to move from headline income toward cash available for debt service.
income or operating cash generation − unavoidable outflows − existing debt obligations − reasonable buffers = capacity available for the proposed debt.
This is conceptual rather than a universal lending formula. Different products, institutions and jurisdictions use different measures. The discipline is the important part: repayment must come from usable cash, not from a flattering headline number.
Step 3: inspect existing obligations
A borrower does not repay the new bank in isolation. Rent, taxes, salaries, family costs, supplier payments and existing debts may already claim the same cash flow.
This is why debt-service ratios and leverage measures matter. The bank asks how much room remains after prior obligations are recognised.
A borrower earning S$12,000 a month with S$9,000 of fixed obligations can be less resilient than one earning S$7,000 with S$2,000 of obligations. Income alone is not capacity.
Step 4: test the stability of the repayment source
The next question is not merely how much cash exists today, but how likely that cash flow is to survive.
- How stable is employment?
- How concentrated is business revenue?
- Does one customer generate most of the company’s sales?
- How cyclical is the industry?
- Can margins fall quickly?
- Does the borrower depend on one property tenant?
- Is revenue denominated in one currency while debt is in another?
- Can interest rates reprice before income adjusts?
A strong current cash flow can still support a weak credit if the underlying engine is fragile.
Step 5: stress the future rather than approving the present
Good underwriting asks what happens when assumptions become less favourable.
| Stress | Question |
|---|---|
| Interest rate rises | Can the borrower still service a floating-rate loan? |
| Income falls | How much decline can the household or business absorb? |
| Customer loss | Does one lost contract break the repayment plan? |
| Property value falls | Does repayment still work without assuming a sale at yesterday’s price? |
| Currency moves | Does debt service rise relative to the borrower’s income currency? |
| Cost inflation | Can the business preserve enough operating margin? |
A stress test is not a prophecy. It is a way of asking whether the loan survives plausible disappointment.
Step 6: examine credit history without treating history as destiny
Past payment behaviour can reveal how the borrower has handled obligations before. Missed payments, high utilisation, restructurings or defaults may signal elevated risk. Consistent repayment can support confidence.
But credit history is backward-looking. A borrower can have a perfect record because conditions were easy and still face a weaker future. Conversely, a past problem may have been temporary and fully resolved.
The bank therefore uses history as evidence, not fate.
Step 7: understand the purpose of the loan
Purpose matters because it tells the bank what economic event the debt is financing.
A loan used to buy productive equipment can create future cash flow. A working-capital facility can bridge a timing gap. A mortgage finances a long-lived asset. Debt used repeatedly to cover an underlying operating loss may merely postpone failure.
The same borrower can therefore be creditworthy for one purpose and unsuitable for another.
Step 8: make the loan structure fit the cash-flow structure
A good borrower can still fail under the wrong repayment schedule. Structure matters.
- A seasonal business may need repayments aligned with seasonal cash generation.
- A long-lived asset may justify longer amortisation than short-lived inventory.
- A bullet repayment can create refinancing risk at maturity.
- A floating rate can create payment volatility.
- A currency mismatch can turn an affordable loan into an unaffordable one.
The bank is not merely deciding whether to lend. It is designing a contract that the projected cash flow can plausibly carry.
Step 9: use collateral as a second exit
Collateral can reduce expected loss if the borrower defaults, but the bank should ask whether the loan works before forced sale.
The value of collateral can fall, enforcement can take time and transaction costs can reduce recovery. Collateral can also suffer from wrong-way risk: the same shock that weakens the borrower can weaken the collateral value.
The better sequence is:
primary cash flow first → structure second → collateral and guarantees as secondary protection.
Step 10: price the risk rather than merely detect it
A borrower can be lendable but not at every price. The bank must decide whether expected interest and fees compensate for funding, capital, expected loss, operating cost and uncertainty.
Higher risk does not always mean “decline.” It can mean a smaller amount, shorter tenor, stronger covenant, additional security, different amortisation or higher price. At some point, however, the risk may exceed what responsible structure can repair.
Step 11: check concentration at the bank level
Credit quality is not only about the borrower. A bank can decline an individually strong loan because it already has too much exposure to the same sector, country, property type, sponsor or risk factor.
This is portfolio logic. One excellent property loan may be harmless. Ten thousand loans depending on the same property cycle can become systemic to the bank.
The bank therefore asks two separate questions: “Is this borrower good?” and “Do we need more of this risk?”
Step 12: preserve independent challenge
Lending businesses want to serve customers and grow assets. Risk functions are designed to challenge whether the proposed exposure fits policy and risk appetite. Good governance prevents enthusiasm from becoming evidence.
That can involve delegated approval limits, independent credit review, exceptions processes, documentation standards and escalation for larger or unusual exposures.
The quality of underwriting often depends less on whether a model exists than on whether someone can say no when the model and commercial pressure disagree.
The loan is not finished at approval
A credit decision is provisional in one important sense: the bank must continue learning after disbursement.
- Are payments arriving on time?
- Has income changed?
- Has utilisation increased?
- Has collateral value weakened?
- Has the borrower breached a covenant?
- Has industry risk increased?
- Does the borrower ask repeatedly for short-term relief?
Monitoring compares the represented future used at approval with the future that is actually arriving.
A simplified household example
Two households each earn S$10,000 a month.
| Household A | Household B | |
|---|---|---|
| Gross monthly income | S$10,000 | S$10,000 |
| Existing fixed debt service | S$1,000 | S$4,500 |
| Income concentration | Two earners | One earner |
| Emergency buffer | Strong | Thin |
| Proposed new payment | S$2,000 | S$2,000 |
The headline salary is identical. The resilience is not. Household B has less room for unemployment, illness, rate increases or ordinary financial shocks.
A simplified business example
A company reports S$1 million of annual profit, but customers take 150 days to pay invoices while suppliers require payment in 30 days. The business can be profitable on paper and still face severe liquidity pressure.
The bank therefore reconstructs the cash-conversion cycle rather than stopping at profit. It asks whether working capital, receivables quality and operating cash generation can support the proposed debt.
This is why Why Cash Flow Matters More Than a Good Story in Bank Lending is the next article in this batch.
Four misconceptions to remove
| Misconception | Better model |
|---|---|
| “High income means easy approval.” | Capacity depends on usable cash flow, existing obligations and resilience. |
| “Collateral makes repayment capacity less important.” | Collateral is usually a secondary recovery route, not the primary repayment source. |
| “A good credit score proves the new loan is affordable.” | History helps, but underwriting must test the proposed obligation and future conditions. |
| “Approval is the end of credit analysis.” | Monitoring continues after the loan is disbursed. |
A mastery test
- What is the primary repayment source?
- Why can profit and cash flow give different credit signals?
- Why does existing debt matter even when income is high?
- Why should a bank stress interest rates and income?
- Why can an individually strong borrower still be declined because of portfolio concentration?
If those answers connect, credit underwriting stops looking like a mysterious score. It becomes a disciplined attempt to decide whether a future promise deserves to become a bank asset.