Lending Standards Tightening

There was so much information given in the NY Fed’s household debt and credit report, I will add some more analysis of the report in this article. One of the most interesting findings of the report was that credit card loans which are considered seriously delinquent increased. The percentage of credit card loans considered to be seriously delinquent went up from 7.1% to 7.5% as you can see from the chart below. This corroborates with the default increase seen by Capital One. It pushed the average seriously delinquent rate of all loans up from 3.3% to 3.4%. I wouldn’t say this is a yellow flag that they economy is about to crash, but it is significant because the labor market has been strong. You would expect shrinking default rates when the labor market is strong.

An increase in the credit card default rate can either spring from job losses or the debt being too high. Anyone playing devil’s advocate, claiming that stagnant real wages are causing the increase in defaults, needs to realize that real wages fell at the beginning of the recovery and the credit card loan default rate fell. If it’s not weak wage growth or an increase in the unemployment rate that’s causing these delinquencies, it must be that the consumer is drowning in debt. This is a reasonable theory because credit card debt is excessive. The current $760 million in credit card debt isn’t at the $870 million peak, but when you consider the other debts, it’s high enough to cause trouble.

The credit cycle is important to understand when viewing these metrics because it can get confusing. When credit is expanding and lending standards are easing, it allows the economy to grow faster than the combined long run productivity and population growth rates. When the lending standards are easing and the debt is being accumulated from low amounts at a rate less than or equal to wage growth, the economy is healthy. Problems arise at the end of the cycle when standards get too low causing credit to spike. That’s what led to the subprime mortgage crisis. After being burned by defaults from risky borrowers, lenders tighten standards which makes it difficult to get a loan and halts economic growth. The lenders in the mortgage market were snake bitten from the housing collapse, so they tightened their standards.

As you can see from the chart below, the standards are the tightest ever. The percentage of loans given out to borrowers with a 760 or higher credit score has never been higher. This means there’s no risk of another mortgage crisis, but it’s also tough to get a loan. This trend is hard on millennials because now it’s tough for them to buy their first home. They’re also more hesitant to do so because they witnessed the collapse 9 years ago.

While housing prices have also been artificially inflated because of low interest rates, my point is that a similar sized decline in housing prices as the early 2000s collapse wouldn’t trigger close to the number of defaults seen in 2008. The fact that heightened lending standards have made defaults less likely is the reason why housing prices won’t fall like they did in the early 2000s. The consumers and the banks haven’t learned their lessons from the 2000s. They simply switched the area where sloppy lending standards take place. It moved to the auto loan market. The various percentiles of auto loans saw their average credit score fall to near the levels prior to the 2008 financial crisis. The average score has started to pick up lately, signaling a tightening of standards. That may be part of the reason why auto sales have declined for 4 straight months. This is not a good sign as it may mean we’re at the peak of the auto loan bubble. The total auto loan debt is $1.17 trillion. It’s burst could push the economy into a recession.

The auto loan bubble has been created by the expansion of leases and lowering of lending standards. Leases have lower monthly payments which encourage borrowers to get a car they otherwise couldn’t afford. If you remember from my last article, the median household income grew about 15% from 2008-2016. From the peak debt level of Q3 2007, the size of the auto loan market has increased about 43%. Clearly lending standards had to be lowered to get that much more debt issued given the slow earnings growth.

The chart below breaks down the long-term trends in auto sales. As you can see in the first rectangle, the percentage of sales which are leases increased from 0% in 1991 to 22.5% in 2016. At the peak of the last cycle in 2007, leases represented only 13.6% of sales. Leases are highly correlated to credit standards. If lending standards tighten, leases can fall quickly. The fact that they represent a greater proportion of sales makes the industry more vulnerable than ever to a crash.

The second rectangle shows the percentage of car loans which are subprime. As you can see, during the early 1990s barely any loans were given to subprime borrowers. As interest rates came down and the economy became financialized, the percentage of subprime borrowing increased. In 2006, the percentage of subprime borrowing peaked at 33.9%. Now the percentage is at 41.0% which signals there’s a high chance default rates will skyrocket above the prior peak in the next recession. Obviously, the final results will depend on how bad the recession gets. In relation to the size of the next recession, the auto market will feel more pain than ever before.

Conclusion

It’s a tale of two cities for housing loans and auto loans. Mortgage debt is down and lending standards were heightened for much of the recovery while auto lending repeated the same mistakes that were made in the housing market in the early 2000s. I think auto lending market blowup will happen this year as sales have started to decline and incentives reached records which can’t be sustained. That could push the economy into a recession. One of my base cases is for tax cuts to act as temporary prevention of a recession just like central bank action prevented a recession in 2016. In that case, the effects of the auto loan blow up would be seen in late-2018.

Comments are closed.