Most lenders think about loan requirements primarily as filters. Set the bar high enough, and the risky borrowers won’t get through. The problem with that framing is that it treats approval as the finish line. The way you structure your requirements does more than screen for creditworthiness as it also actively shapes whether the borrowers you approve will actually repay. There is a meaningful difference between requirements designed to protect the lender at the point of approval and requirements designed to produce healthy repayment behaviour over the life of the loan. Getting that distinction right is where a lot of lenders leave performance on the table.
This piece walks through the factors that matter most when designing loan requirements, with a focus on what the evidence shows actually moves repayment outcomes.
Debt-to-income ratio, the number lenders underuse
The debt-to-income (DTI) ratio measures the share of a borrower’s gross monthly income that goes toward servicing existing debt. It is widely used in mortgage lending and far less consistently applied in personal and business credit and that inconsistency costs lenders.
Research on mortgage lending across multiple euro area countries found that the loan-to-income ratio at origination, alongside employment status, is among the strongest predictors of eventual default. The same principle holds across credit products. When a borrower is already spending a significant portion of income on existing obligations, adding another repayment obligation increases the probability of a missed payment, regardless of how clean their credit score looks.
The standard threshold most lenders work with is a back-end DTI of 43% or lower. Borrowers with DTI above 50% have very little financial slack to absorb income disruption. When you approve someone at 55% DTI, you are extending credit to someone with almost no buffer. The next unexpected expense, salary delay, or family obligation can tip them into delinquency even if their payment intent was strong at origination.
The pragmatic implication is that a DTI threshold applied as a single uniform number misses too much. A borrower with a 45% DTI and a stable government salary carries very different risk from a borrower with a 45% DTI and seasonal commission income. Separating DTI thresholds by income type and employment stability gives you a more accurate risk picture and lets you approve more of the right borrowers rather than simply fewer borrowers.
Loan size relative to repayment capacity
Over-lending to a creditworthy borrower is one of the more common and least discussed sources of repayment problems. When a borrower’s monthly repayment obligation exceeds what their income realistically supports, default risk rises regardless of their intent.
Research on informal lending in Tanzania found that clients with multiple simultaneous loans were consistently associated with poor repayment outcomes, while progressive lending, where loan sizes increase gradually as a relationship develops, contributed to positive repayment performance. The mechanism is straightforward enough. Progressive lending keeps the repayment burden at a level the borrower can manage at each stage, rather than loading them with obligations they have to grow into.
This has a direct implication for how lenders should structure their maximum loan amounts. The ceiling should be anchored to demonstrated repayment capacity, not only to collateral value or credit score. A borrower with a strong credit history who currently earns N180,000 monthly can service a very different loan than they could five years ago, and neither figure should be extrapolated naively.
A repayment-to-income (RTI) ratio is the cleaner metric here. Most lenders who use it cap the monthly repayment at 30 to 35% of verifiable monthly income. This leaves enough room for the borrower to meet living expenses and absorb variability without defaulting. Lenders who let this ratio creep to 50% or higher, even for supposedly low-risk borrowers, tend to see elevated late payments and restructuring requests down the line.
Repayment structure and how it maps to cash flow
One of the most underappreciated levers in loan design is how repayment is structured relative to when the borrower actually has money. A borrower with a business that receives client payments at the end of every month has a genuinely different cash flow profile than a salaried employee paid on the 25th. Treating them with the same fixed repayment schedule introduces friction that did not need to exist.
Cash flow-based lending addresses this directly. Real-time visibility into a borrower’s income patterns makes it possible to structure repayment dates and frequencies in ways that align with how money actually flows into their account. For seasonal businesses, aligning repayment peaks with revenue peaks reduces the probability that a borrower faces an obligation during a cash-dry period. For salaried borrowers, scheduling the debit two days after payroll is a small operational decision that produces measurably fewer missed payments.
Income-contingent repayment takes this logic further. It links monthly payment amounts to verified income levels and has been embedded into national lending systems in both the UK and Australia. The underlying principle is that borrowers pay more when they earn more and less when they earn less, which aligns repayment obligation with actual capacity at each point in time. The challenge for most commercial lenders is the administrative overhead of income verification and payment adjustment, though in markets where transaction data is increasingly available through open banking or bank statement APIs, this is becoming more workable.
At minimum, lenders should evaluate whether their standard repayment schedule is compatible with how borrowers in a particular segment actually receive income. A one-size-fits-all monthly schedule may work adequately for formal salary earners and create unnecessary default risk for traders, farmers, and gig economy workers.
Employment and income stability as quality cues
Most lenders check employment status as a binary, employed or not employed. A more useful frame is income stability, which is a different question entirely.
A borrower who has been self-employed for seven years with consistent revenue is more creditworthy than a borrower who has been salaried for three months. Yet many lenders apply higher scrutiny to the self-employed borrower almost automatically. The result is both missed lending opportunities and inaccurate risk pricing.
Income stability matters because it determines whether the borrower will still have the capacity to repay twelve or twenty-four months from now. For salaried borrowers, length of tenure in the current role and the nature of the employment, permanent versus contract, are stronger signals than employment type on its own. For business borrowers, revenue trend over at least two years, combined with the debt service coverage ratio (DSCR), provides a more reliable view of future repayment capacity than a single period’s income statement.
Lenders who capture these dimensions in their underwriting criteria, rather than relying solely on current income level, build approval processes that are both more inclusive and more predictive of actual repayment behaviour.
Loan tenure and the risk of over-extension
Longer tenures make loans more affordable on a monthly basis, which is why borrowers often prefer them. They also extend the period over which repayment can go wrong. A borrower’s circumstances at month three are considerably more predictable than their circumstances at month thirty-six.
The ECB’s research on residential real estate lending found that original loan maturity at origination was a significant predictor of default rates, independent of the borrower’s initial creditworthiness. Longer-maturity loans carry more time for income disruption, relationship changes, and economic shocks to intervene.
Tenure should be calibrated against the purpose and stability of the underlying income, not set primarily to reach an affordable monthly payment. A thirty-six month personal loan to a public sector employee carries different tenure risk than a thirty-six month loan to a contract worker whose current engagement ends in ten months. These two borrowers may present similar DTI ratios at origination while their tenure risk profiles are meaningfully different.
A practical approach is to model the repayment burden not just at current income, but at a moderately stressed income level, say 15 to 20% below current income, and to set the maximum tenure at the point where repayment remains manageable under that stressed scenario. This prevents the common pattern where a loan is approved with a comfortable DTI at origination but becomes unmanageable when the borrower experiences any income reduction.
When requirements discourage good borrowers
There is a risk on the other side of rigorous requirements. Requirements that are technically sound but poorly calibrated to the reality of the borrower population end up screening out people who would have repaid. Research on microcredit programmes found that lower transaction costs for accessing loans were associated with better repayment outcomes. Borrowers who find the application process manageable are less likely to feel overwhelmed by the overall lending relationship.
For lenders operating in markets with significant informal employment, requiring formal payslips and tax returns from every applicant eliminates a large share of creditworthy borrowers. The better approach is to accept alternative evidence of income where formal documentation is unavailable. Bank statement analysis, mobile money transaction history, and business revenue records can all serve as income verification mechanisms when structured properly.
The goal is to require enough documentation to make an accurate credit decision, and no more. Requirements that exist primarily as friction reduce access without improving portfolio quality. Periodically reviewing which requirements are actually predictive of repayment versus which are inherited conventions is a worthwhile exercise for any lending team.
What separates a performing portfolio from a problem one
Loan requirements that improve repayment outcomes share a few characteristics. They are calibrated to repayment capacity, not just to creditworthiness at a point in time. They account for income type and stability alongside income level. They match repayment structure to cash flow patterns. And they are designed with the full loan lifecycle in mind, not just the approval decision.
The lenders who manage this well tend to have lower default rates because the borrowers they approve are matched to products they can realistically service. That matching is a design problem, and like most design problems, the quality of the solution shows up months later in the data. If you’re looking for a well-rounded loan management software designed your full loan lifecycle in mind talk to our team.
