This article is adapted from one first published at CreditSesame.com
In late 2012, the Consumer Financial Protection Bureau (CFPB) published a study analyzing differences between credit scores sold to lenders and those sold to consumers. The study found that the different scores provided “similar information about the relative creditworthiness of consumers,” but that for a small percentage of consumers, different scoring models gave “meaningfully different results.”
The CFPB is also now encouraging lenders to provide free credit scores to their customers. Inevitably, this brings up a sensitive issue among consumers and their advocates and a question the 2012 CFPB study left unanswered: Are consumers better off with so many credit scores?
As the credit scoring marketplace continues to evolve, consumers are benefiting in ways they might not recognize. Competition certainly has its upsides.
If you looked exclusively at credit bureau-based risk-scoring systems, you’d find over five dozen options that lenders can choose from. Those models are referred to as “generic” scoring systems, which means they’re sitting on a shelf and any lender in the United States can buy and use them for risk assessment for any variety of credit obligation. The next time you walk down the soft drink aisle at your local supermarket, count the number of options available and you’ll have a pretty good idea of how many credit scoring model options are commercially available for lenders.
And while there are dozens of scoring systems available for use by lenders, those models were developed by only a small handful of companies, including VantageScore Solutions, FICO and the credit bureaus themselves. Each of the credit scoring models available to lenders essentially competes against the others, as they’re all designed to do pretty much the same things. What’s also a given is that every one of those models will perform differently in its efforts to predict the likelihood that a consumer will become delinquent on his or her credit obligations.
When lenders choose which scoring model to use, they normally test several options against each other to determine which does the best job of separating future bads, or consumers who will default, or go 90-days or more overdue on their loan payments, from consumers who make payments on time. This testing generally entails a process called validation: Models are applied to a set of recent consumer data for which the actual number of defaults is known. (These historical files are chosen to be representative of the U.S. population overall, and are stripped of personal information that could be used to identify any consumers involved). The models’ predictions of default are compared to actual consumer behavior, and whichever model identifies more bad payers than the others is the more predictive model.
Predictive prowess is just one of many significant factors that determine a lender’s decision whether to implement a new model. Other considerations include compliance and governance guidelines, universe expansion and myriad other factors.
The ability to perform validations with competing models is the considerable value of having many different scoring models to choose from. If only one model were available, lenders wouldn’t have the ability to test multiple options to identify the best-performing tool for their particular product mix and target customer. That would leave all lenders at a disadvantage because their loss rates would be much higher than necessary. But, that disadvantage wouldn’t fall solely on the shoulders of lenders. Consumers would likely be the ones who would be asked to subsidize lenders’ additional risk.
Whenever lenders determine there is financial risk of doing business with certain groups of consumers, they have a choice to make. They can either avoid doing business with those consumers or they can ask those consumers to subsidize their risk by paying higher interest rates and fees. There’s no doubt we would all be paying higher interest rates if there were limited credit scoring model options, or only one.
So whether it’s free from a lender, through any number of websites or through the purchase of a credit-monitoring service, any time we can expose consumers to credit scores and the context around how that score was calculated, consumers benefit. They benefit because credit score management is not taught at any level of academia, and because, until the early 2000s at least, only consumers who also happened to work in the financial services environment knew anything about credit scores.
Today we live in an environment where most people are familiar with the concept of credit scoring. Interest rates are relatively low and credit can be extended within a very short period of time. Innovations in credit scoring and the competitive pressure to produce the most accurate and innovative credit risk management system are in large part responsible.
Regardless of how many scores are sold to consumers or given away freely, or how many models are available for use by lenders, all of them are based on the same three credit reports, compiled by the major credit reporting companies (CRCs), Equifax, Experian, and TransUnion. That makes credit score management as easy as ensuring that the information on your three credit reports is accurate and speaks glowingly of your credit management practices.