They sold off their prime mortgage and insurance divisions in 1998. Their CFO told Mortgage Banking magazine (4/98) "The margins were clearly superior...[subprime is] a different business that looks and feels a lot more like the consumer finance business." What I find interesting is what the President tells Forbes magazine (11/98): "This business isn't price sensitive...borrowers are far more concerned about just getting the loan."
There's a lot of talk about how efficient markets and the whole wave of freshwater Chicago thinking means that subprime mortgages should be perfectly priced. Here's one of the kings of subprime mortgages in the late 1990s telling us that this is wrong -- that consumers, for whatever reasons, are not price sensitive at the margins but instead just want to sign the paper to get the house. They'll compare prices, but they are constrained in how well they can search.
And here's what a writer for American Banker wrote (4/2004) when WMC was purchased by GE:
"Alt-A lending is often cited as a way to keep origination volumes growing in the face of higher rates, and many Alt-A practices have made their way into the prime market. One reason lenders like it: Higher rates and less competition come in exchange for bypassing normal requirements on the documentation of income or assets, credit scores, debt ratios or loan-to-value ratios. Prime lending, by contrast, is 'a low-margin business,' Ms. Brandtsaid. 'It's a widget-processing business.'"
And as opposed to the perfect model world, we can use the s-word here. Steering. There's going to be a lot of research into this over the next years, but already the evidence is conclusive -- a lot of perfectly prime borrowers were steered into subprime loans. These are loans that community banks could have been competitive on if the correct information got to the consumers.
What does that mean in practice? Fees. Lots of ways of hiding the numbers and making it harder to compare across venues. Take a moment to familiarize yourself with the way yield spread premiums (YSP) can scam mortgage borrowers who aren't experts and can't figure out how to compare mortgages across banks because of legal last-minute markups in the interest rate. Consumers aren't stupid, but they do have information problems. And having a vanilla loan product will give a baseline on how to compare across companies, leading to less YSP-esque last-minute markups and other ways to 'nudge' consumers into taking on a worse product that looked better.
Community banks must have some source of competitive advantage over Bank of America and Citigroup. There are two obvious possibilities. One is that customers prefer dealing with local banks, and based on my personal experiences, the level of customer service is far superior than what you get with a megabank. The other is that community banks, as Salmon said, are better at underwriting, because they actually know the characteristics of the neighborhoods they are underwriting in and, possibly, the borrowers they are underwriting.
Now, both of these are advantages that should be protected by the CFPA. That is, if your goal is to provide better customer service, you are probably not in the business of screwing your customers. And if your competitive advantage is in underwriting, then you have no need to confuse customers with unnecessarily complex products. You should be happy that Elizabeth Warren is keeping predatory lenders out of your backyard, because even if you refuse to match their mortgages, they are driving up housing prices and making it harder for you to find qualified borrowers.
I want to quantify James' point about community banks being better at underwriting. As I discussed a few months ago, there's a lot of research being done to exactly this point. Look at the following graph, taken from the excellent Did Securitization Lead to Lax Screening? Evidence From Subprime Loans (Benjamin J. Keys, Tanmoy Mukherjee, Amit Seru, Vikrant Vig): (opens .pdf)
This is the difference in default rates between FICO scores at 615-619 and 620-624. The big difference between the two groups is that scores above 620 could be sold to another party to be placed in a security, while those below it could not. So loans with a FICO score of 619 are held by the banks in question, and as such use "soft skills," local knowledge and information, in order to determine whether or not the loan goes through, and they do much better than the 621 ones. The 621 ones are taken and sold off onto the secondary market, and they do much worse.
Remember that the 615-619 ones have worse credit scores, so all else being equal they should have done worse.
I want to mention something Richard Serlin told me in a discussion:
There may be great opportunities to profit by deceiving consumers, and large scale advertising and other marketing can be very effective at that. Big companies are much more capable of doing this than small ones because of the great economies of scale involved. With regulation, you can take away these opportunities to deceive people into taking bad deals with large scale marketing, and this hurts the big companies' competitiveness much more than the small ones'.
This may be an important issue. With things more plain vanilla and clear, a lot of customers may have a natural preference to go with the local community bank. If the big and/or internet banks can't falsely make it seem like they have a substantially better deal through confusion, a lot of people may like to just go to their local bank, by their home, and sit down with someone.
We know that there was a lot of ways to deceive at the margins. And we also now have research that tells us the local underwriting adds value as long as the terms of the loans are comparable to the consumer end. It's about time everyone gets a fair shake at this, and that is what the Consumer Financial Protection Agency should allow to happen.
This article available online at: