Lending Perspectives: There’s No Such Thing as a Free (Car) Ride, Part Two

Last month I shared some basic methodology using “back-of-the-napkin math” to address what your credit union is likely facing: falling used car values. In reality, what all auto lenders are experiencing now is a double whammy. Not only are prices falling, lowering the amount realized at sale when you repossess a car, but the extremely strong values of late 2020 into 2022 mean your members likely financed a much larger loan than historical trends, regardless of whether your credit union experienced higher loan-to-values. Translation: Two factors drive higher losses. But there’s more to forecasting your future losses.

My previous napkin math only addresses magnitude of loss—how much we’ll lose on every repossession. It doesn’t address what happens if we experience more repossessions. That said, watch for these leading indicators which might drive up repossessions:

• Gas prices
• Job losses
• Voluntary repossessions

As an amateur economist, I know this: History has a way of repeating itself. I was rather worried about auto loan performance in 2022 when I saw gas prices cross over $5 per gallon because I remember observing an increase in repossessions in 2007 and 2008 as gas prices surged to $4 per gallon before the financial crisis. Fifteen years ago, each increase of $.25 per gallon saw another spike in cars coming back.

Job losses will ultimately do more than any other economic factor to drive defaults. It’s a funny thing, if you have a job, you most likely need a car to get there. If you don’t have a job, you don’t need a car, especially if you owe more than its value. 

Being seriously upside-down on a car tends to do funny things to consumers who have a job as well; if their older car starts having problems and they can’t trade it in, it often goes back to the lender; I’ve also seen repossessions increase when values decline. Fifteen years ago, Ent started seeing voluntary repossessions increase; members were turning in seriously upside-down cars, often after buying another, newer car because they couldn’t trade in their old car.

The repossession rate measures the frequency of loss as opposed to the magnitude of loss. If I forecast our Q1 repossessions will increase 25% year-over-year, losses will likely increase close to 25% as well. This is not an exact measure because my portfolio may have grown 10%, and a 10% increase may just be keeping up with overall growth. 

Here’s the bottom line: Ultimately, loss ratios are a function of the magnitude of loss and the frequency of loss. If I forecast our magnitude of loss will double in 2023 from falling values, and our frequency of loss goes up 25%, my credit union is likely to see loss ratios go up 125% (100% from magnitude, increased by 25% from the frequency of loss). 

A Thought on Your Portfolio Mix

For 30 years, I’ve written and spoken about the virtues of “the fix is in the mix.” I recall doing a two-day session for the New York Credit Union League in 1993 on how to fix a broken loan portfolio, and I haven’t shut up since. Having made a career of fixing and rebuilding a couple of broken portfolios, I know improving the credit mix—the mix of A, B, C and D credit loans—is part of the fix. The weaker your portfolio, the more problematic 2023 and likely 2024 will be for your credit union for two additional reasons:

1. You have more C and D loans, and higher losses on those loans will have a greater financial impact. Let’s just say the magnitude of loss for your portfolio doubles across the board. If your A+ tier loss ratio for new auto loans is .10%, and it doubles to .20%, that’s not going to materially change the profitability of your A+ loans. But let’s say you have a lot of C and D loans, and your recent history shows the loss ratio to be 1.5% and 2.5% for those credit tiers. Doubling that loss ratio is going to have a big impact on the profitability of those loans, overall losses and your return on assets.
2. You’ll probably see a much bigger increase in the frequency of loss (repossessions per $1 million in loans) on those lower credit tiers. Most recessions don’t tend to be thought of as “equal opportunity.” Borrowers with weaker jobs and/or lower incomes, often with lower FICO scores as well, will be hit harder. Combine this with the likelihood these lower-tier borrowers bought an older, higher-mileage car that may get hit harder from a value standpoint, you’ll have another layer of risk to navigate.

I Could Be Wrong

I like to joke that I’ve predicted seven of the last three recessions. In addition, just four or five months ago, I felt the used car market probably had another strong year or two in it due to the chip supply chain impacting new car sales. Yet we’re seeing the market shift pretty rapidly at a time when we’re still not in a recession, for a combination of reasons.  

It’s really up to your economic forecasting skills as to what to do in response to a normalizing auto market. If you think 2023 could be a bad year for auto values, and you’ve excessively loosened your credit policy over the last few years, there’s not a lot you can do other than tighten things up. Take a look at your credit policy, loss mitigation strategies, remarketing plans, and overall collection tactics. Hold on tight.

CUES member Bill Vogeney is chief revenue officer and self-professed lending geek at $9.8 billion Ent Credit Unionin Colorado Springs.