I started reading the Focus Report [Oct. 13 issue] guest opinion article entitled "How to Use the Best Predictive Tools to Revive Mortgage Lending" by Joanne Gaskin of FICO. I hoped I would learn more about FICO's tools and how my credit union ($3 billion in assets, $900 million mortgage portfolio, approximately the 30th largest CU in the country) can improve mortgage lending.
I then realized just how little FICO understands the credit union business. When Gaskin wrote, "We found that a typical credit union originating 500,000 mortgages a year using a score cutoff of about 620 could save $62 million more than they could be using the general-risk score," it illustrated how much FICO is out of touch with our industry.
Five-hundred thousand mortgages a year? Did she mean the typical mega-bank? As I said, we're about the 30th largest credit union in the country. In our primary market, we're the second largest mortgage lender on the basis of number of loans closed per month. We funded $450 million in mortgage loans last year-a record for us-which equated to about 3,000 loans.
Her example of saving $62 million a year is also confounding and misleading. That equates to $124 per loan in terms of future losses. On an average of $150,000 per loan, that translates to about eight basis points of improved loss ratios. For many credit unions, including ours, that's probably close to the current loss ratio for the portfolio. For credit unions in the sand states, they'd love to have that kind of performance. However, her example assumed an industry standard of $50,000 for each loan that goes bad. I have seen no statistics to refute this; I can tell you our typical loss on a first mortgage is about $20,000 when we foreclose. Booking a loan today at lower home prices, lower interest rates and improved underwriting should not result in a $50,000 loss at foreclosure. At least I hope not.
It's no wonder that consumers question the accuracy of FICO scores. It also bears mentioning that many of my peers are scratching their heads, wondering why their portfolios are doing so poorly having utilized a FICO score along with custom scores and bankruptcy scores.
We don't use the auto loan or mortgage loan-enhanced scores as mentioned in Gaskin's article because quite frankly, we don't believe in them. Our auto portfolio, while performing fairly well, has had more than its share of losses on loans when our FICO score was 650, but the dealer's fancy auto-enhanced FICO score indicated the borrower to be more like a 720. In the end, we find the auto-enhanced score to be worth no more than the paper on which it's printed. I also question the value and applicability of the mortgage-enhanced score. FICO, along with the credit bureaus, is going to have to do a better job of educating and selling lenders on the value of scoring without pie in the sky claims.
Senior VP/Chief Lending Officer