Analytics have long been essential to the credit business thought pre-bust they may have been most used to find every possible person to whom credit could be extended. Now, in these tighter times, analytics are being used to more accurately measure and mitigate risk.
The Los Angeles Times reports that lenders and the firms that help them build “better underwriting mousetraps” are looking at “non-credit proprietary databases maintained by Papa John’s or Victoria’s Secret” to get a better picture of the people who are applying for mortgages. Lenders are relying less on the information provided by applicants or traditional credit scores and more on that which can be verified through or compared to some objective, third-party data. For example, one line of thinking is that you’ll most often have deliveries – of everything from a large pepperoni pizza to a new nightie – to the address where you actually live (which might differ from what’s on the mortgage application if an applicant is trying to pull a fast one) and that where you shop may give some indication of your true income.
It isn’t perfect, of course but the Times quotes Alex Santos, president of Digital Risk as saying, “We’re looking for any type of data source that you can plug into a computer. It takes only a month of trial and error to determine whether the information can help [determine credit risk] or not. We have a hypothesis, push a button, and the computer tells us whether the data is predictive or not.”
This article, of course, is written for civilians (non-analytics profesionals) who have little view into all of the ways that data can, and is, being used to manage risk, hone offers, determine pricing, and much, much more. What is intriguing about this story is the use of non-traditional databases: these firms are looking beyond the usual industry sources to try to find meaningful new indicators. It also reflects how much more cautious mortgage providers have become.
Have you come across any unusual or counterintuitive data sources that you are able to share?