The Repo Economy: Subprime Auto and the $1.7 Trillion Car Debt Machine

In a country built around the car, the auto loan is the credit product people default on last — you can sleep in a car, but you can't drive a house to work. Which is why the current data should command attention: subprime auto delinquency has broken a 32-year record, above Great Recession levels; the average new-car payment crossed $770 with one in five buyers north of $1,000; and nearly a third of trade-ins arrive owing more than the car is worth, rolling an average $7,183 of old debt into new loans. The most defended payment in American life is being missed at record rates. This report is the machine, the conveyor, and the casualty list.

The core thesis

Auto credit is where America's transportation policy, labor market, and consumer finance collide: for most working households, the car is not an asset — it's the admission ticket to income, which makes the car loan simultaneously the most essential and most extractable credit product in the file. That's why the "payment priority" folklore — Americans pay the car first — has been true for decades, and why its erosion now is the single most informative signal in the consumer cycle: when the payment people protect above all others hits a 32-year delinquency record, the stress isn't preferential — it's arithmetic. The budget doesn't close.

Our thesis is that the auto market has built a negative-amortization culture on a depreciating asset: prices reset upward, rates followed, and the industry's answer at every step was duration and rollover — stretch to 72, 84 months; roll the old loan's stub into the new one; make the payment fit and let the balance sheet be tomorrow's problem. Tomorrow arrived: record underwater trade-ins feeding record loan sizes feeding record payments feeding record subprime delinquency, while the prime tier — bifurcation again, the cycle report's signature — sails on undisturbed. The system's shock absorbers (used-car values, ABS investor appetite, lender equity at the bottom of the market) are each thinner than last cycle, and the casualty list among subprime specialists is already several names long. This isn't 2008 — auto is smaller than housing, loans amortize faster, and repossession is swift, brutal loss mitigation — but it is the sharpest sector expression of the same divide running through every report in this series: the economy of the scored and the economy of the squeezed.

You can sleep in a car; you can't drive a house to work. When the payment people defend hardest sets delinquency records, the budget isn't choosing — it's failing.

The machine: how a car loan actually gets made

The structure matters because it explains the incentives. Most auto credit is indirect: the dealer arranges financing, shopping the deal to captive finance arms, banks, credit unions, and subprime specialists — and earning a markup or flat fee on the financing itself, which makes the F&I office a profit center with a structural preference for bigger, longer loans (financing is where dealer margin lives, along with the add-ons folded into it). Underwriting spans the spectrum we mapped in the underwriting report: prime runs on score and payment-to-income; subprime adds down payment, LTV, and job verification; the deepest tier barely underwrites the borrower at all — it underwrites the collateral and the collection method (GPS locators and remote disablement devices as standard equipment). Pricing follows tier: roughly 7% average APRs for new cars and ~11% for used at the market's center, climbing steeply through subprime and into the buy-here-pay-here lots where rates brush state maximums and the small-dollar logic — price the desperation, secure the paycheck's enabler — completes itself.

The stress readings

Two readings deserve the analyst's asterisk. The subprime record is a share-of-a-shrinking-pool statistic in part: lenders tightened after 2022's bad vintages, so today's subprime borrowers are somewhat better-selected — and still defaulting at records, which makes the reading worse, not better. And the 30-day flow rate actually eased slightly year-over-year while 90-day rates climbed — the signature of borrowers who fall behind staying behind: cure rates, not stumble rates, are what's deteriorating.

The negative equity conveyor

The market's core pathology is a conveyor belt with three stations. Station one: the payment shock. Post-pandemic vehicle prices reset upward and never fully retreated; rates doubled; insurance surged. The monthly number stopped fitting median budgets. Station two: duration as anesthesia. The industry's fix for an unaffordable payment is never a cheaper car — it's a longer loan: 72 months normalized, 84 spreading, because term is the only dial that lowers the payment without lowering anyone's revenue. But term on a depreciating asset manufactures underwater time: the car sheds value faster than a stretched loan amortizes, so the borrower spends years owing more than the collateral is worth. Station three: the rollover. Life doesn't wait for amortization — the average trade cycle arrives while 31% of loans are underwater, and the $7,183 gap gets financed into the next vehicle, which now starts deeper underwater on a longer term with a higher payment. Each pass through the conveyor raises the temperature. The record delinquency at the subprime end and the record negative equity at the trade-in desk aren't two problems; they're the same borrower, a few years apart. And the exit — repossession — is uniquely unforgiving: swift self-help seizure, auction sale, and a deficiency balance for the difference, meaning the underwater borrower loses the car and keeps thousands of the debt, plus a derogatory file entry that reprices the replacement vehicle they still need to get to work. The conveyor doesn't just carry debt forward; it compounds the borrower's position at every station.

The bottom of the market: deep subprime's own physics

Below securitized subprime sits the buy-here-pay-here economy — dealer-lenders financing their own inventory to the unscoreable, at maximum rates, with weekly payments, ignition-interrupt devices, and a business model candid about its math: default rates north of 30% are priced in, the car is recovered, reconditioned, and re-financed to the next customer — the same physical vehicle sometimes generating three or four loan lifecycles. It's the auto market's answer to the invisibility problem, and it shares payday lending's closed loop: many BHPH lots historically don't furnish positive payment history, so years of completed weekly payments build nothing — transportation secured, file unimproved. The strategic observation for the credit-building world: the deep-subprime auto borrower is the clearest case in consumer finance where file construction is the exit ramp — every reported tradeline moves the next car loan out of the ignition-interrupt tier and toward rates that don't presume default. That arbitrage — between what the borrower actually is and what the file says — is the entire distance between an 8% loan and a 24% one on identical transportation.

The funding layer: securitization and the casualty list

Subprime auto runs on the asset-based finance machine: originators pool loans into ABS, tranche the risk, and recycle the proceeds into new originations — the architecture that lets thinly capitalized specialists lend at scale, and that transmits their performance to insurers, pensions, and private credit funds holding the paper. The stress readings are flowing through: issuance down roughly 6%, subordinate tranches pricing wider, structural protections (overcollateralization, excess spread) doing their intended work but with less headroom than the benign years. And the equity layer beneath the structures is where the bodies are: American Car Center and U.S. Auto Sales failed in 2023; Tricolor collapsed in 2025 amid fraud allegations; a major buy-here-pay-here chain spent mid-2026 negotiating with its lenders to survive. The pattern is the sector's oldest: securitization disciplines gradually (worse pools price worse) but punishes suddenly (warehouse lines pull, and the originator — whose whole model is recycling — stops mid-cycle). None of this is systemic in the 2008 sense; auto ABS structures have absorbed every cycle since the 1990s. But it is exactly how a credit problem becomes a supply problem: each failed originator is a tier of borrowers whose next loan doesn't exist, at precisely the moment the conveyor is delivering them to the trade-in desk underwater.

Scenarios and what we're watching

What we're watching: the Fitch subprime index (the cycle's cleanest thermometer); used-car values (the collateral variable that decides whether repossession is loss mitigation or loss realization); the 84-month term share and underwater trade-in share (the conveyor's speed gauges); further originator distress; and the labor market above all — because auto credit's entire stress test compresses to one question: can the borrower keep getting to work? The car loan sits exactly where this series keeps finding the American consumer: essential on one side, extractable on the other, with the credit file as the toll gate between. The repo economy is what it looks like when the toll outruns the wage — one defended payment at a time.

Frequently asked questions

This report is for general information only and does not constitute financial or investment advice. Figures are drawn from publicly reported sources including Federal Reserve, Fitch Ratings, Experian, and Edmunds data, and change with each reporting cycle.