One way FICO clients are boosting profits in originations is by using analytics that provide critical insights into a wider range of factors affecting the performance of new accounts. Here are three key areas where we've found analytics can sharpen origination decisions:
1. Macro-economic impacts on credit risk
Analytics can help lenders understand and anticipate the impact of changing macro-economic conditions on the default rates historically associated with credit risk score ranges. By quantifying the relationships between forecasted economic factors, such as GDP, unemployment and housing prices, and expected default rates, they're able to predict the degree and direction (positive or negative) of likely default rate changes at each score range.
When such predictions are in the form of consistent metrics that can easily be added to origination strategies, lenders have an empirical means of adjusting score cutoffs. They can increase cutoffs to maintain their current default rate despite an economic downturn. They can also use these predictions to avoid "leaving money on the table" at point of origination by over-tightening their risk exposure or acting too late to readjust cutoffs in an upturn.
2. Consumer ability to pay
When assessing a customer's ability to pay for a new credit obligation, income and income estimators have proven to be significantly less reliable risk indicators than empirically developed analytic tools. While these approaches address regulatory requirements, they don't necessarily result in profitable new accounts.
Analytics that measure credit capacity, used in conjunction with credit scores, can improve decision accuracy. Credit capacity analytics analyse the same credit bureau data as the credit score, but in a new way that isolates consumer sensitivity to future conditions not present on the credit report. They answer: What is the ability of consumers to manage increments of new debt on top of what they already have?
This is even more critical considering how integrated the global economy is and the state it is in right now. We've seen how travails in the Eurozone affect lending confidence globally. Similarly, it would make sense to apply the same logic and examine how borrowers' and consumers' ability to repay under difficult economic conditions is impacted. Only by integrating prediction of future economic data into risk models, can this be gauged.
3. Application fraud
Profitability can also be increased by using business rules and analytics to detect fraud during the origination processes. It's a growing problem-based on FICO's work with leading banks, we estimate that gross losses attributable to first-party fraud/abuse are much larger than those from third-party fraud. Fraudulent activity artificially elevates bad debt levels and wastes collections resources.
While conventional application fraud analytics look for applicant behaviour indicating a potentially fraudulent or stolen identity, first-party models are tuned to detect fraudsters using their own identities and accounts to apply for additional credit. For fraud that slips through at point of origination, lenders can improve detection for early-life accounts by using the application fraud model with a transactional fraud model.
Andrew Jennings is chief analytics officer, FICO.
Sign up for Computerworld eNewsletters.