Americans Owe $1.68 Trillion On Car Loans — More Than Credit Card Debt And As Much As All Federal Student Loans

Americans now owe about 1.68 trillion dollars on auto loans and leases, the highest level ever recorded.

What that $1.68T actually represents

• The figure comes from a new analysis by The Century Foundation and Protect Borrowers, which used national credit panel data and aligns with New York Fed estimates around 1.67 trillion in auto balances at the end of 2025.

• It includes both traditional auto installment loans and leases on cars and light trucks.

• Roughly 86 million Americans (about one in four people) are carrying auto loan or lease debt.

How we got here

• Total auto debt is up about 37–40% since 2018, when balances were closer to 1.2 trillion.

• The average amount financed on a vehicle has risen to the low–mid 30‑thousand‑dollar range, about 10,000 dollars higher than in 2018.

• Average monthly payments are now around 680 dollars, almost 40% higher than in 2018, driven by higher vehicle prices, elevated interest rates, and longer loan terms.

How big is this compared with other debt?

• Auto loans are now one of the largest categories of consumer debt after mortgages, comparable in size to federal student loans and bigger than total credit card balances.

• New York Fed data puts auto loan balances around 1.67 trillion versus about 1.28 trillion in credit card balances and around 1.66–1.69 trillion in student loan balances, as of late 2025.

Snapshot of major consumer debts

Stress signals and risk

• Delinquencies have been rising, particularly among subprime borrowers, with record shares of lower‑credit borrowers at least 60 days behind on payments.

• Advocates warn that auto debt now makes up more than a third of all non‑mortgage consumer debt and is increasingly a driver of financial distress, especially for lower‑income households facing higher car prices, higher rates, and gas costs.

Will the $1.68 trillion in auto debt signal an impending consumer credit crisis or sustainable household leverage?

1. Will the $1.68 Trillion in Auto Debt Signal an impeding consumer crisis or sustainable household leverage? Where the Models Agree

2. Where Models Disagree

3. Unique Discoveries

4. Comprehensive Analysis

High-Confidence Findings. All three models converge on the key interpretation: $1.68T is a warning light, not proof of an imminent, system-wide consumer credit crisis. The level is record-high, but the macro context matters—especially whether households can service payments and whether distress is spreading beyond the weakest borrowers. On the “serviceability” side, the Federal Reserve’s household debt-service ratio sits around 11.32% in Q4 2025, well below pre-2008 peaks, which supports the view that broad household leverage is not yet at a breaking point.

Second, there’s strong agreement that the auto market is bifurcated: prime borrowers are generally fine while subprime is under acute strain. The subprime signal is hard to ignore—multiple sources show subprime 60+ delinquency near record highs (~6–7%) in the ABS performance data, while prime delinquency remains very low. This pattern is consistent with a “contained deterioration” cycle: painful for marginal borrowers and certain lenders/investors, but not automatically contagious to the entire household sector.

Areas of Divergence. The models mainly differ in how close today’s conditions are to “crisis” language. GPT-5.5 Thinking puts more weight on system-level stabilizers—DSR levels and New York Fed metrics showing the flow into serious auto delinquency roughly flat year-over-year—and therefore leans toward “macro sustainable unless unemployment rises.” Gemini 3.1 Pro Thinking reaches a similar macro conclusion, but via a different lens: auto debt burden relative to disposable income and the continued cleanliness of prime performance.

Claude Opus 4.7 Thinking is more alarmed about the pipeline mechanics: negative equity and ultra-long terms can keep borrowers paying for a while but increase the time spent “underwater,” raising the odds of eventual default if anything else goes wrong (job loss, repair shock, insurance spike). That view treats the current episode as already a “crisis” for the bottom of the distribution—even if it’s not a systemic financial-stability event. Evidence supporting that concern includes the rise in long terms noted in the TCF/Protect Borrowers report and the Fed’s own warning that longer maturities increase default risk partly by increasing negative equity risk.

Unique insights worth noting. GPT-5.5 Thinking’s point about post-repossession deficiency balances is important because it highlights why “secured” doesn’t mean “clean exit.” The CFPB has documented that consumers often still owe substantial balances after repossession and sale, and that repossession assignment rates rose above pre-pandemic levels in the CFPB’s studied period. That can sustain losses and consumer distress even after collateral recovery, and it can show up as collections pressure rather than just delinquency stats.

Recommendations. Treat $1.68T as macro-manageable but micro-fragile: watch (1) labor-market deterioration, (2) subprime ABS delinquencies/loss severity and recoveries, and (3) negative-equity/term length trends as leading indicators of spillover risk. If you want a single decision rule: absent a jobs shock, this looks like a contained subprime/affordability crisis; with a jobs shock, the long-term/negative-equity structure makes the downside convex.