Analytics Can Predict Risk, If Used Well
NEW YORK-As the mortgage industry becomes more risk averse, lenders and servicers alike are finding that analytics can be a big help in ferreting out bad loans.
"In the past lenders looked at credit score and appraisal," said Chaten Patel, an executive vice president in ISGN's mortgage division. "The new regulation has improved this process. Now lenders are looking at future trending. You need to have historical data like the borrower's FICO, information on the geographical area, local economy analysis, the nature of the borrower's position at work and the likely future the borrower has at that job. If the borrower works in a car factory and car factories in that area are closing down that needs to be considered."
Many borrowers are complaining that it's too tough to get a loan even when lenders don't take into account the added details that Mr. Patel suggests should be considered. And for lenders, it's also tough for them to get all of these added data elements about a borrower. Nonetheless, Mr. Patel believes use of these types of analytics would make for a more risk-averse lending process. "Yes, the criteria would be tougher and analytics from various new sources would be used to make a better loan. Lenders need to know that the borrower won't go into default. There is money out there but lenders are being more risk averse.
"To make this happen now," continued Mr. Patel, "you need to go to the credit agencies, the census, compile market research to get local job trending, etc. So, there's an opportunity for technology vendors to come up with a new risk model. The data is public record, but if you don't have a centralized system and method the score will be different each time. We're working on such a proprietary score."
However, even the best analytics can't predict major personal life events in a borrower's life that may impact their ability to pay back the loan like divorce, for example. In these cases Vladimir Bien-Aime, CEO at Global DMS, warns lenders that using analytics to determine the current and future value of the property is just as important as getting to know the borrower better by using analytics. "The key to getting to a good valuation includes getting the right people involved. If the appraiser is not familiar with the market, the risk goes up. Technology can automate the process and make the assignment less risky," he said.
"Now, FHA is calling for geographic competency. Technology can help there and make the right assignments to the right appraisers that have that competency. Second, having technology to extract data from a report and run rules around it is important today. Lenders can have their own rules that match their individual guidelines. In terms of predicting the future value of the property, you have to look at historic and current values to guess at future values. We're working with a company to bring in predictive methods."
On the servicing side, Jeff Taylor, managing partner at Digital Risk, adds that analytics can help overworked servicers complete better workouts that will help borrowers avoid foreclosure. "The cascading volume of residential mortgages being considered for modification is forcing loan servicers to make tough business decisions. The sooner a servicer can accurately organize their portfolio, the sooner they can apply the appropriate resources to resolve each borrower's business challenge.