According to an estimate from Cox Automotive and Experian, 1.73 million cars were repossessed in America in 2024. In 2009, at the bottom of the financial crisis, the same estimate peaked at 1.77 million. So the headlines almost wrote themselves: repossessions, back at crisis levels.
Except that 1.73 million was never a count of cars. It is the output of a formula. In 2022, Cox wrote that no government or third-party source officially and accurately tallies repossession volumes.
How the number is built:
Step one: the base. In 2024, about 74 million car loans.
Step two: an estimate of how many defaulted. Cox defines a default as a loan more than 120 days late, without a bankruptcy. For 2024, that estimate is 2.33 million.
Step three: an assumed share of defaults that ends with a tow truck. It comes out at 80% in almost every year from 2006 to 2020. In 2024, it was 74%.
The same data, as a rate:
In 2009, about 3.29 cars were repossessed for every 100 loans. In 2024, it was 2.33. That is 29% lower, because the base of car loans grew 38% in between.
In 2019, the rate was 2.31. Measured as a rate, 2024 is not 2009. It is 2019. Cox says so itself: the repossession rate "returned to 2019 levels."
Where the damage actually sits:
Ford Motor Credit files its U.S. repossessions with the SEC. In 2019 it repossessed about 27 thousand vehicles. In 2025, about 24 thousand. Its repossession ratio was 1.24% in 2019 and 1.23% in 2025. Flat.
But the expected loss on each repossession went from $10,700 in 2019 to $17,300 in 2025. That is 62% more. Adjusted for inflation, the 2019 figure is about $13,500, so the real increase is 28%.
Ford's own filing names the cause: an increase in new vehicle prices and the higher average amount financed.
Ford's net losses on U.S. retail loans went from $28 million in 2021 to $362 million in 2025.
The chain behind the number:
In January 2025, the Consumer Financial Protection Bureau wrote that the most widely cited data on repossessions are an annual estimate drawn from auction sales of repossessed vehicles.
In its data from nine lenders, the share of accounts sent out for repossession each month rose from about 0.54% before 2020 to 0.75% by December 2022.
Forwarders, the companies that manage repossession agents for the lender, went from 31% to 66% of completed repossessions between January 2018 and December 2022. The bureau found average repossession costs charged to consumers were higher when a forwarder was used.
Cox, the source of the famous estimate, also owns Manheim, an auction where repossessed cars are sold. Cox's own explanation for Manheim's growth in repossessions: "much of that growth is market share gains at Manheim." That is a conflict worth knowing about. It is not proof of bad faith.
The one stretch where the clock helps: between the day an account is assigned for repossession and the day the car is gone, the average was 15.9 days for borrowers with the best credit and 23.1 days for subprime borrowers. Catch-up rights are state law and differ in every state.
So, is this a repossession crisis? Not in the count. The count everyone repeats is an estimate, and as a rate it sits where it sat in 2019. The crisis, if there is one, is in the subtraction: fewer cars, bigger holes, and a chain built to process the cars, not to shrink the holes.
No advice here. No tips, no recommendations. Just the numbers, and where each one is published. The full source list is in the video description.
— The Cost of Living
Book mentioned in this channel: The Psychology of Money, by Morgan Housel — https://amzn.to/47KvNx3
As an Amazon Associate I earn from qualifying purchases. This is an affiliate link: it costs you nothing extra.