An average of 5,170 vehicles repossessed per day.
If you’ve been in the repossession industry long enough to remember The Great Recession, then you remember a boom of repo volume. Industry wide, an estimated 630,000 cars were hooked in Q1 2009 alone. Fast forward to 2024 when an estimated 1.88M were recovered, it begs the question, how high is that historically?
Repossessions are a front-row seat to America’s financial pulse. An end result of financial strain and failure. Looking back at 17 years of data, we’ve seen 2.1 million cars hauled away at its peak in 2009 and down to 1.1 million in the dark days of the pandemic. But since then, we’ve seen a steady climb from the abyss.
In this article, we will explore the rollercoaster of historical data, inciting events and seasonality that provides us insight into just how historical 2024’s 1.88M repossessions was.

From Meltdown to Momentum
Going way back in 2008, the housing market crumpled and the auto subprime bubble popped. The repossession industry was swamped. There were an estimated 1.9 million repossessions as unemployment hit double digits.

By 2009, it hit an all-time high: 2.1 million cars, with Q1 alone piling up 630,000 recoveries. “We couldn’t keep up,” recalls a veteran Manheim auction manager in a Cox Automotive recap. “Lots were overflowing.”
Recovery kicked in by 2010, dropping to 1.8 million, then bottoming out at 1.2 million by 2014 as lenders tightened up and the economy clawed back. The mid-2010s saw a steady climb. 1.5 million in 2015, leveling off at 1.7 million over the 2017–2019 period.
Q1 of 2020 started off slow with a drop from 2019’s Q1 of 476,000 to 390,000. Then came COVID in mid-March and the bottom dropped out. Q2 crashed to estimated 195,000 units thanks to repossession moratoriums, stimulus cash and forbearances, as noted in Cox’s 2021 Used Car Market Report. But, by Q4 things picked back up and 455,000 units were recovered, just below normal for the prior Q4’s steady volume and the year finished at 1.3 million units.
Still reeling from the 400 million unit drop in volume, 2021 was the real gut punch. With a quarterly average of 275,000 units recovered per quarter, it ended at an estimated 1.1 million, the lowest over the reported period. An estimated 20% to 30% of the industry had closed, retired or shifted to more reliable income channels.
But 2022 bounced back and provided some relief. The momentum had shifted and an estimated 1.35 million vehicles were recovered. This trend carried into 2023 with an improved 1.5 million. While still below the pre-pandemic 1.7M a year, it was progress.
By mid-year 2024, Cox Automotive had reported that there was 23% jump in repossessions. They reported that 950,000 more cars were repossessed then than were recovered by the end of Q2 23’ (Bloomberg, July 16, 2024).

With Q’s 3 and 4 historically counting for 52% of all annual repossession volume, as we had previously reported, a total of 1.85 million vehicles were repossessed by year end 2025.
That’s an average of:
- 5,170 vehicles per day.
- 215 per hour.
- 3.6 a minute.
These are the third highest numbers on record. These are based on a 365 day years, 24 hours a day and 60 minutes and hour.
The Repo Rollercoaster: 2008–2024
Looking back at the data, we can see six pivotal years in particular.

- 2008: 1.9M – Recession kicks off.
- 2009: 2.1M – Peak chaos, 630K in Q1.
- 2014: 1.2M – Lending tightened, a Quiet year.
- 2020: 1.3M – Pandemic – Q2 low at 195K.
- 2021: 1.1M – Year end. Rock Bottom.
- 2024: 1.8M – Full recovery. Overheated car market, inflation.
So, what made 2024 such a busy year? Car prices averaging $46,000+ (per Kelley Blue Book, June 2024). Record high car payments, 19% over $1,000. Average 7.3% interest rates (Edmunds, June 2024), record negative equity ($6,700) and borrowers stretched thin by record high inflation.
Where the Numbers Come From – Ratios and Reasons
I’ve said it before and I’ll say it again, there is no centralized data set maintained that collects all repossession data. The best that could be created would probably require all here major credit reporting bureau data to be merged and track new repossessions as lenders report them with the “8” status code.
In that absence, we do have some reasonable data to rely on.
Seasonality: Repossessions, like delinquency, aren’t random, they follow a pattern. If you look at the rollercoaster chart, you’ll notice it is almost identical to an auto loan delinquency chart for the same period. They go hand in glove.
Most everyone knows these cycles either numerically or intuitively. Data backs it up and provides us the following ratios and reasons:
- Winter (Q1): 28% of the haul. Post-holiday defaults. January’s a killer before tax refunds hit. Think of 2009’s all-time high 603,000 repos in Q1.
- Spring (Q2): 20%, the slow season. Tax checks averaging $3,000 (IRS, 2023) ease the borrower’s pain. 2015’s 300,000 units was moderate breather from Q1.
- Summer (Q3): 22%, a middle grind. Back-to-school costs nudge it up. In 2019, 391,000 cars were recovered followed by 493,000 in Q4.
- Fall (Q4): 30%, Showtime! Holiday spending plus year-end lender pressure to close the books strong. Consistently the busiest quarter of the year.
This rhythm’s held steady from 2008’s pre- housing meltdown chaos to today, bending only for shocks like 2009’s Q1 spike or 2020’s Q2 drop. These same patterns can be seen in delinquency patterns, and for good reason.
Cox Automotive Data
I didn’t develop this formula, Cox Automotive did. They are arguably the best auto data aggregator in the industry and have been tracking repossession data informally for years using Manheim auction volume, as the denominator of seasonality in the repo pipeline.

The next denominator is auto loan delinquency. They estimate that 80% of loans 120+ days late turn into repossessions, per their 2018 Used Car Market Report.
Of course, there are many cars that don’t end up at the auction, redemptions, cars that go to the insurance auctions and small dealer repossessions that usually go back to the originating lot for resell. But that is a data point completely unavailable by any source except the lenders themselves. Good luck getting that.
The 2009 peak of 2.1 million was cited by Fast Title Lenders, 2023). This trended down until 2014 provided an abysmal 1.2 million repossessions, the lowest over the available data window until the pandemic. These volume numbers very closely replicate the pattens in auto loan delinquency in the 120-day delinquency tranches used to develop the volume estimates.
Using reliable seasonal trends applied year by year to available data sets and applying that same Q1-Q4 seasonal trend, the consistent data is derived. No rocket science, just auction flows, default trends, and your seasonal reality, crunched into numbers. Absent a better methodology, this is about as good as it gets.
2025?
Looking back at the chart, we can see that the 2024 numbers look almost identical to the 2008 numbers. What followed was a year with the highest repossession volume in recorded repo history.
So, this begs the question, is this trend going to continue as it did in 2009, or is this the peak?
We’re only two months into the year. I’m going to hold back that projection a little while longer. With potential $5,000 DOGE Refunds, tariffs, war and Federal Government layoffs in the mix, things could go either way real quick.
But before I get to that, I’ll be reporting to the breakdown of repossessions by state.
Kevin Armstrong
Publisher
Facebook Comments