When satire is no laughing matter
As this newsletter was being finalized, we learned that a bipartisan bill had been introduced into the U.S. House of Representatives that would “require Fannie Mae and Freddie Mac to establish procedures for considering certain credit scores in making a determination whether to purchase a residential mortgage, and for other purposes.” In other words, the bill would make Fannie Mae and Freddie Mac consider credit-scoring models besides the three outdated FICO 04 models they currently use when deciding what loans to purchase. We’ll have more to say about this news in the next issue of The Score, but until then, we hope you’ll read about the bill here.
For now, a few words on something we came across recently in our own reading: a recent opinion article in The Wall Street Journal that took a satirical look at the credit scoring industry. As implied by its headline, “My Tour of FICO Scores, Fido Loans, Whatever,” the article mainly poked fun at our chief competitor, but we didn’t find much in it to laugh about. The best satire is grounded in familiarity with its target and, unfortunately, what author Andy Kessler “knows” about credit scoring is largely tired old myths that we’ve worked for years to dispel.
Contrary to Kessler’s blanket statements about credit scoring models, some of today’s models actually do include such data as rent, utility, and telecom payments. Moreover, these models do not ding borrowers for positive behaviors such as paying off a mortgage or a collection account. These criteria are among the characteristics of the VantageScore® 3.0 scoring model we introduced in 2013, a model that is now used by thousands of lenders and has been emulated by our competitors.
Kessler’s observations better fit past decades, when there was little innovation of note and credit scoring models would receive occasional updates, but their inputs—the factors models consider when calculating scores—changed very little. You could also forget any modifications made with consumers’ benefit in mind. That lack of innovation arose because there wasn’t any meaningful competition. Models today are now more predictive, consistent, and inclusive—chiefly because competition has driven all of us in the industry to develop better approaches to measuring creditworthiness.
Although competition among model developers has created scoring alternatives for many types of lending, this is not true for mortgages. The reason for this anomaly is the current GSE lockout. The GSEs mandate the use of credit scores generated by older FICO models, built using consumer data from the late 1990s. Many mortgage lenders are reluctant to upgrade to newer credit-scoring models because their mortgage origination software is designed to accommodate that mandate.
Unbeknownst to many, FICO has enjoyed a government-sanctioned monopoly in the mortgage market for decades. To ensure that credit scoring models keep improving, competition and a fair playing field must be established and preserved in this lending sector.
Kessler is right about the importance of ensuring that consumers understand the connections between their financial habits and their credit scores. But Kessler’s suggested remedy—publishing the algorithms used to calculate scores—would be counterproductive. However, disclosing the complex computer code wouldn’t really help consumers understand credit scoring or better manage their finances. Exposing score modelers’ proprietary methods would severely hinder innovation and competition in model development.
To promote consumer awareness, we instead support educational efforts like CreditScoreQuiz.org, our venture with Consumer Federation of America, and other websites that offer free credit scores and content. Such websites provide personalized consumer credit education by coupling free credit scores and free credit reports with additional educational content tailored to individuals’ credit histories. This helps consumers better understand how their financial habits affect their credit scores.
All the best for the holiday season,