A senior vice president at FICO has blasted online lenders, saying that their “poorly conceived” lending models could have unintended consequences.
Dr. Andrew Jennings, senior vice president for scores at FICO, has written in a blog post that he is concerned about some of the new ways that fintech lenders are attempting to assess risk. Jennings writes that fintech business models, which in some cases circumvent regulation in the US surrounding credit processes, “could pose significant unintended dangers, including negative impacts for consumers, lenders, investors and the overall financial system”.
Jennings goes on to cite the 2008 financial crisis in reminding his readers that bad credit risk models “often lead to bad lending decisions”. He states that such decisions can in turn lead to “massive losses for investors, unsustainable business conditions for lenders, and a weakening of the financial system”.
Most telling, however, is Jennings’ implicit dig at student lending giant SoFi – which publicly ditched the pre-eminent credit reference agency in January, declaring itself a “FICO-Free Zone”. SoFi focuses its lending on young professionals – many of whom do not have established credit histories – thus diminishing the usefulness of a FICO score. SoFi CEO Mike Cagney declared at the time that the hallowed FICO score was “anything but transparent”.
But Jennings clearly disagrees, and appears to have taken aim at SoFi. “If, for example, a score uses data on a person's educational background, what can be done when the score is low?" Writes Jennings. "Tell that person that they should have attended a better college or read a different subject or attained a higher grade?"
"Such data is often predictive but would be problematic on many of the key requirements of a best-in-class risk model. Designers of alternative credit-risk scores may not intend to illegally harm anyone, but the unintended dangers of their models may result in unreliable lending decisions and/or widespread bias against certain groups of people."
In citing the key requirements of a stellar risk model, Jennings refers to “predictive power, broad coverage, fairness, transparency, and the ability for the consumer to dispute the process when mistakes may have been made”. If a prospective borrower receives a low FICO score, it will always come accompanied with a reason code, such as “balances too high”. But Jennings believes that this kind of treatment simply isn’t practicable for some online lending models.
Of course, the truth is that SoFi prides itself on being picky. The company took a 30 second advertising slot at Super Bowl 50 in February. The ad originally finished with the words: "Find out if you're great at SoFi.com. You're probably not." The words "you're probably not" were subsequently removed.