Disparate impact and dealer compensation were at the crux of the debate at yesterday’s Consumer Financial Protection Bureau forum on auto finance.
CFPB Director Richard Cordray kicked off the event, reiterating that lenders must police their dealers to ensure compliance with fair lending regulations.
Industry representatives called out regulators for using unproven proxy methodology to determine biased lending, and argued that the shift to a flat-fee dealer compensation structure would still result in discretionary lending practices. Ultimately, “eliminating dealer discretion would do more harm than good,” argued Damon Lester, president of the National Association of Minority Automobile Dealers. Consumer advocates volleyed back with concerns about consumers being overcharged for loans and for lack of transparency on how dealers are being compensated.
In the final session, a pair of CFPB officials noted that a flat-fee structure is just one possible solution for alternate compensation methods.
As someone who starting an auto leasing company (Vanguard Credit dba Decisive) especially for used cars back in 1997, I can relate to your article. In fact, our company has become more of a consultant to lessors in recent years because of the expertise we delevoped in modeling our own residual values. When we started, ALG didn’t even want to sell use the residual values we were interested in. Now we have a very sophisticated model for auto lease portfolios. ALG does a fine job of estimating future values, but no one is perfect. Furthermore, if a bank, finance company or credit union uses ALG data to begin with, having ALG re-estimate or re-evaluate a portfolio later has little or no value because their methodology is same in both cases – and after all, why would their second “guess” be better than their first “guess?” The mathematics of risk relies on a branch of calculus called Stochastic Calculus, which is, in essence, a combination of regular calculus and probability theory. The real power in this branch of mathematics is that random variables which define risk in the auto lease world can be represented by a Monte Carlo process. Our model uses this math to help us define and manage risk for our clients/partners.