Another year tracking 2025's record charge-offs, with loan modifications at the highest level since they were first reported. The aggregate is bad; the dispersion is the useful part.
CUCollector's first-quarter read puts credit union new and used auto portfolios at $479.5 billion, down only $514 million on the quarter — the smallest decline since 2023. Sixty-day-plus delinquencies finished at $3.8 billion, a 0.80% ratio that mirrors the same quarter a year earlier. Charge-offs came in at $1.48 billion, above both the author's forecast and Q1 2025, on a trajectory he expects to finish near 2025's record $5.6 billion. Loan modifications reached $10.1 billion, the highest level since they began being reported in 2016.
Modifications lead the charge-off line
The modification figure is the one that repays attention. Modifications are a lender's response to stress rather than the stress itself, which means they lead the charge-off line and lag the underwriting decision that created the exposure. An institution running modifications well above its peer group is saying something about its book roughly two quarters before the charge-off number does.
None of which is evenly spread. Auto concentration varies enormously by institution — a credit union with 60% of its loan book in vehicles is in a categorically different position from one at 20%, in the same state, in the same quarter, behind the same national delinquency print. The industry line is the average of those two, and it describes neither.
For anyone selling into this — collections platforms, loss-mitigation tooling, indirect-lending analytics, deposit funding — that dispersion is the whole opportunity. A rising national delinquency rate is not a sales trigger, because it is true for everyone at once. An institution whose auto concentration and modification trajectory have moved against its peer group over two consecutive quarters is one.
Rank auto exposure and modification trends within peer group. A national print is context; a peer-relative move is a reason to call.
Sources
Everything above is commentary on these. Read them first if the two disagree.