In Defense of Subprime Auto Lending, But Not of GM

We learned today that General Motors will purchase subprime auto lender Americredit. I share my colleague Megan McArdle's confusion about just why GM feels that this is a smart strategy. It's particularly odd that the government would approve this move, which presumably it should have to since taxpayers retain 61% ownership in the automaker. But as strange as this strategy might appear, it's important to clear up a common misconception: subprime auto lending doesn't share many similarities with subprime mortgage lending.

So just what is subprime auto lending? It's providing auto loans to individuals with spotty credit histories. That translates to most target borrowers likely having sub-600 FICO scores. As you might expect, these companies are pretty used to a hefty number of defaults, and consequential losses. According to a company fact sheet (.pdf), Amercredit's loan portfolio had an 8.3% loss rate for its most recent portfolio. They make up those losses by charging relatively high interest rates. Americredit's average APR is around 17%, but can approach 30% for some borrowers.

This might sound a little bit like subprime mortgage lending, but there's a key difference. Subprime auto lenders' risk models have a big advantage over the failed models of subprime mortgage lenders. Auto loans are a depreciating asset, while housing had historically been an appreciating asset. So when an auto lender models defaults, it assumes that the car will be worth less at resale than it was at purchase. Thus, even in a deep recession, credit losses won't be as devastating if conservative stress scenarios were considered.

Subprime auto loans also tend to be more plain vanilla than the wacky subprime mortgages that became popular during the housing boom, like option adjustable-rate and negative-amortization loans. Auto loans are generally fixed rate. Americredit's average loan-to-value ratio at origination is just 109%, which is well below those accepted by many subprime mortgage lenders. Extending the term out to 72 months is about as crazy as subprime auto loans get, which is kind of a long time to pay back an auto loan.

Subprime auto loans can actually be a pretty good business when you know what you're doing. If you consider the math, it's pretty easy to see how you can make a profit by losing 8% if you charge 17%. Americredit had a net income of $0.45 per share last quarter. That's with near 10% unemployment and many Americans continuing to struggle to pay their bills.

The auto market also doesn't really have to worry about the kind of bubble that struck the mortgage market, specifically because autos are a depreciating asset. Millions of people are not going to hope to get rich quick by flipping their cars, for example. There's an old industry adage that most people will keep paying their auto loan even after they've defaulted on their mortgage, because they need their car to get to work. They can default on their mortgage and rent, but they probably don't want to have to walk if they lose their car. Moreover, if times really got tough, and they did lose their home, they could always live in their car temporarily.

Of course, none of this means to imply that it makes sense for GM to purchase Americredit. While most other auto companies, particularly foreign ones like Nissan and Honda, have found it sensible to keep a captive finance company in-house, none of those are subprime. They generally cater to people with very strong credit so they don't have to worry about strategy and can simply earn interest on loans that are a very safe bet. So it's puzzling that GM wouldn't just focus on building up a new prime borrower-driven captive unit instead. And it's even stranger that the government wouldn't raise its eyebrows when GM is making an acquisition rather than engaging in additional divestitures to try to pay back the billions it still owes Uncle Sam. Or for that matter, the $3.5 billion might have been a repayment to the taxpayers instead.