This Could Be A Problem: Losses On "Deep" Subprime Auto Double Industry Average

On Saturday, we highlighted a rather disturbing statistic.

60+ day delinquencies for subprime auto ABS have now risen above crisis levels to 5.16% - levels we haven't seen since 1996.

That won’t surprise regular readers. The writing has been on the wall for quite a while. More auto loan originations are going to borrowers with shoddy credit and loan terms are looking more and more stretched by the quarter. Just ask the NY Fed. Or Experian, where even permabull Melinda Zabritski will tell you that underwriting standards are getting looser (although she likely won’t say that’s a bad thing).

While Citi and others are quick to point out that the originate to sell model isn’t prevalent in the auto loan industry, the inability for lenders to securitize subprime loans may well put the brakes on US auto sales. After all, the pool of creditworthy borrowers is finite. That means that at a certain point, incremental sales must be engineered by making ineligible borrowers eligible by resorting to looser underwriting. But that only works if you can offload that credit risk. No lender wants to be sitting on a book full of used car loans to deep subprime borrowers with sub-600 FICOs, and so, if demand for subprime auto ABS dries up, so too will credit to the subset of borrowers who are driving (no pun intended) incremental sales growth.

Here’s a look at the share of total auto loans that have been securitized:

As you can see, and as we noted last week, the share has remained range-bound for at least 15 years. But again, the question is what happens to auto sales if lenders can no longer expand the pool of eligible borrowers by relaxing their standards? That would likely occur if subprime and deep suprime start to underperform, leading investors to pull back from paper backed by loans to less creditworthy borrowers. And guess what? Deep subprime is underperforming. Massively. Here's Goldman: 

Many of the deep subprime pools are experiencing losses above the industry average. Exhibit 6, for example, shows the losses on three large 2015 vintage deep subprime deals vs. the subprime industry aggregate. The deeper subprime deals have losses over two times the industry average.



Remember, somebody owns this paper and they sure as hell won't be buying into any more deep subprime deals if the collateral pools continue to perform like that. And if investors aren't buying, then lenders can't offload their credit risk which in turn means they will simply stop lending to shoddy borrowers. Once that happens, you can kiss the US auto sales "miracle" goodbye - for good. 

If you need proof of just how important the ability to securitize truly is for subprime auto lenders, simply review the first-hand account we published last week that details what's happening "under the industry's hood," so to speak.

For now, we'll close with one last chart from Goldman which shows what happens when a pair of trillion dollar bubbles collide head-on with Americans' inability to free themselves from the shackles of debt: