This guest post by Joanne Gaskin originally appeared on the FICO Banking Analytics Blog.
When you stop to think about it, the term “alternative credit data” is a catch-all phrase to describe data that is not currently reported on mainstream credit reports. But what, in reality, is alternative about it?
For millions of people, it may not be alternative data but the only credit record they have. For these consumers, it would seem it is primary data. Especially if that data can help them qualify for their first mainstream credit or loan. For everyone else, such data might be considered supplemental. It can be matched with an applicant’s traditional credit bureau file to provide a richer credit profile and lead to better lending decisions.
Take for example our newest risk score, the FICO® Mortgage Score Powered by CoreLogic®. Our scoring model bases its risk assessment on unique data from CoreLogic plus a traditional credit report from one or more of the national consumer credit reporting agencies. The result is stronger rank-ordering based on risk, a more dramatic separation of good risks from bad. In fact with this stronger score, 23.6% of consumers that score below 680 on classic FICO® Scores will score above 680 using the new scoring model. This can enable them to better qualify for a mortgage or home equity loan.
So what type of alternative data drives a stronger credit risk score? Based on our experience, data sources qualify as “best” if they are:
-Nationally representative and robust; not just the Mid-Atlantic states, for example.
-Consistent and dependably accurate.
-Objective and free from personal bias; not social media, for example.
-Unique new independent data that is supplemental or complements that which is in a traditional credit report.
-Predictive, so that matching it with performance data or other data elements produces a provable and useful correlation with target behavior.
-Operational and regulatory compliant, so it can be used by a lender to influence a credit decision.
FICO has long been a data-driven organization. Our scientists have examined hundreds of depersonalized data sources over the decades. Along the way, we have demonstrated a knack for extracting the most value for our clients from what is now described as Big Data. Having more data can be good, but understanding how to use it is even more important.