Calculating Ability to Pay
Interest rates are at all-time lows in many global economies. This makes it more attractive for consumers to take out additional loans, with money being very, very cheap. However, this may create a situation of adverse selection, with many banks reluctant to lend to higher-risk individuals who need the credit—what my fellow blogger Daniel Melo calls the profitability paradox.
In addition to internal risk policies and income-based measures, lenders traditionally assess a customer’s ability to pay by using either application scores for new customers or behavior scores for existing customers. The problem is that two customers with similar risk scores may each handle a new loan very differently based on their credit profiles. Someone with mild delinquency and low utilization may receive the same score as someone else with no delinquency but high utilization; how each would handle new debt could vary.
Income estimators and income-based measures, like debt-to-income ratios, can augment lender strategies, however even these are limited. It's quite tricky to adjust for key factors like cost of living. Not to mention income is often self-reported, difficult to verify and subject to manipulation. Even honest consumers are often unclear how to calculate this accurately: Is the lender looking for income as gross or net? Do I account for bonuses or spousal support?
FICO has been working with clients to solve the credit capacity puzzle, using our patented analytic technology that more accurately measures a consumer's ability to take on new credit. The technology almost opens a third eye on risk prediction, allowing lenders to extend loans more safely and responsibly to those most capable of repaying them. Our credit capacity analytics can be leveraged by countries with mature bureau data; where there is lack of bureau data, sometimes internal behavior data can be used.
Lenders today are challenged with economic uncertainty and severe competition to attract and retain best customers. In this environment, yesterday's risk assessment tools may no longer be enough.