Why Auto Loans Are Not the Next Debt Crisis

Last week, Fitch Ratings reported that subprime auto loan delinquencies are now higher than 2009 levels and indeed, the highest since 1996. It has been reported for a while now that auto lending standards are getting weaker. Some have even claimed that this could be the next debt crisis. This is a space where I am involved through asset-backed securities (ABS), so I thought I would give some intelligence on what is going on and what might develop. Let's start with how subprime auto lending in 2016 compares subprime mortgage lending during the housing bubble. There are some superficial parallels between the surge in questionable lending during the housing bubble and what is happening in autos today. It is certainly true that auto-lending standards on average are weaker now than three and four years ago. The most obvious signs would be the increase in six-year amortizing loans and the rise in lending to deep subprime (i.e., borrowers with FICO scores below 500). This has been driven by new entrants into the market, most of which are non-traditional lenders. The scrambling for new customers is exactly what caused home lending standards to decline and eventually led to the housing crisis. There are some fundamental differences, however, between home lending and auto lending that are important to consider. 1. An auto loan is typically recourse whereas mortgages often are not: In many states, home mortgages are non-recourse, meaning once the bank takes your house, the debt is extinguished. Car loans are recourse. You owe the money no matter what happens. 2. Home mortgages amortize very slowly, auto loans amortize quickly: A typical home mortgage will only repay 3.6% of principal in the first two years. An auto loan will repay 36.5% (even the six-year loans repay 29% in the first two years). That means that even if losses start to pile up in auto ABS, the value of the securitization will be much clearer. Mortgages are such long duration assets that the price can drop in a huge way even without commensurate losses. 3. Autos are not purchased as an investment: One thing that really exacerbated the housing bust was the fact that people who were underwater on their home loans became much more likely to default. More defaults led to more supply which led to lower prices, which led to more defaults. Autos are known to be a depreciating asset, so they would not experience this type of spiral. No one buys a car thinking they can flip it at a higher price later, so the fact that the price depreciates doesn't influence borrower behavior. It could be that businesses relying on car sales could suffer if an auto lending bust caused a big increase in used car supply. But borrowers will not default on their car just because its price has declined below their purchase price. 4. An auto loan default plays out very differently than a home default: With a car, the repossession process is straight forward. Indeed, the repo man can be at your house within hours of you hitting the severe delinquency level. In housing, the foreclosure process has substantial legal hurdles. Even once the bank finally does foreclose, the sale process can take many months. From start to finish, a foreclosure usually takes well over a year. During the housing crisis, this resulted in huge shadow supply of homes that had yet to be fully foreclosed but would someday hit the market. This was a key reason why the housing market saw depressed prices linger long after the worst of the crisis had passed. To me, this means that any increase in auto-loan defaults will be easier to contain. For example, we will know what kinds of loans are more at risk, because the impetus for defaults to spread is limited. If a deep subprime borrower defaults, it doesn't influence a higher quality prime borrower's situation. That wasn't really true during the housing crisis. The path to wider economic contagion is also very different. Banks hold $389 billion in auto loans but $2.1 trillion in home loans. More than that, houses are used as collateral for other types of spending, whereas car loans are just car loans. So a big increase in auto losses would hurt banks, but it wouldn't bring the whole banking system to its knees like it did during the crisis. Put another way, autos are about 23% of bank capital. Housing is 121%. So, in my opinion, not only would an auto bust be more concentrated in bad lenders, it would be unlikely to cause a wide contagion. Indeed, if auto lending suffers through a bust, it will likely be because unemployment starts increasing, which would happen for some exogenous reason, and auto losses would be the fallout. If that occurs, perhaps the weaker underwriting standards will cause more auto-loan defaults than happened in the past. Interestingly, auto-related asset-backed securities are trading relatively wide, even very high-quality bonds. Yet bonds for auto manufacturers, such as General Motors (GM) and Ford (F), are trading at year-to-date highs. The same goes for parts manufacturers such as American Axle & Manufacturing Holdings (AXL), Goodyear Tire & Rubber (GT), etc. I find this bizarre. If auto lending gets hit, there will be materially fewer autos sold. These companies are operating on thin margins. I really think people are looking in the wrong places for trouble. Easy credit almost always ends badly, and it never just takes down the bad borrowers. But for now, the bond market only seems to be looking at those bad borrowers. I think that's a huge mistake.

Why Auto Loans Are Not the Next Debt Crisis

Let's compare subprime auto lending in 2016 to subprime mortgage lending during the housing bubble.

Columnist Conversation

Implied Volatility on Crude Exploding

Ulcer Index

The Market Doesn't Believe Musk

COLUMNIST TWEETS

Please Join or Log In to manage and receive alerts.