Q3 US Consumer Loan Data: Housing Good, Autos Not

By in
Q3 US Consumer Loan Data: Housing Good, Autos Not

The New York Fed was out with its quarterly Household Debt and Credit Report today, which tracks both aggregate consumer debt levels and delinquency rates. A quick summary of the former:

  • Housing related debt is up 2.8% over the last year, to $9.83 trillion. This grew by $31 billion over Q3, all due to incremental mortgage balances (now $9.44 trillion). Home equity lines of credit continue to shrink (down $3 billion in Q3).
  • Non-housing debt balances also rose in Q3, up 4.3% over the last year to $4.12 trillion. Q3 2019 saw increases across the board here, from student debt (+$20 billion) to auto loans (+$18 billion) and credit card balances (+$13 billion).

We focus most intently on the 90-day delinquency data in this report because it gives us a decent real-world look into the health of the US consumer. Loan growth is fine, but it only represents marginal lenders. Non-payment data shows how the consumer economy as a whole is performing.

Here is a chart with that data:

Starting from the top of that graph and working down, here is what we see:

Student loan debt (red line) continues to be a problem. Delinquency rates ticked higher in Q3 2019, to 10.9% from 10.8% last quarter. The Fed also notes that these rates are almost certainly understated due to deferrals. The fact that these are relatively stable over the last few years is therefore slim comfort.

Credit card delinquencies (blue line) were stable last quarter at 8.3%, the same as Q2 2019. Worth noting: these are better than the last cycle, when they ran between 9% – 10%. Given better labor market conditions, they should be.

Auto loan delinquencies (green line) are up slightly, to 4.7% in Q3 from 4.6% in Q2. Unlike credit card debt, non-payment here is actually higher than the last cycle and actually approaching Great Recession levels.

Having spent +10 years as an auto analyst, let us share a few observations:

  • First, auto finance companies always take marginal customers late in a cycle but have the ability to manage credit exposure by increasing lending standards. We wouldn’t be surprised to see this happen in 2020.
  • Second, the auto loan market is not the same as the housing finance market so fears of the “next subprime crisis” are off the mark. Cars and trucks can be repossessed and sold wherever there is demand (houses obviously can not), and used car prices remain strong. And do not forget that the auto loan market is only 13% of the mortgage/HELOC market.
  • Lastly, and on a more cautious note, we think vehicle affordability and consumer preferences are two underappreciated factors in higher auto loan delinquencies. If everyone leased a Nissan compact car for $200/month, auto loan data would likely look just fine. But America’s love affair with high priced SUVs alters the equation substantially…

Mortgage (orange line) and HELOC (purple line) delinquencies were generally stable in Q3, at 1.0%/1.1% respectively versus Q2’s 0.9%/1.1%.

Our takeaway is that the 2 broadest measures of US consumer health (housing and credit cards) reveal a solid picture but 2 others (student loans and autos) remain worrisome. Student loan delinquencies have been high for years, but at least the latest data is no worse. Auto loans are another matter, but as bad as the data might be it does not ring any alarm bells in our book.