Credit Acceptance: The Subprime Auto Lender the Market Is Still Discounting

Credit Acceptance (NASDAQ: CACC) reported Q2 2026 GAAP EPS of $12.66, up 71% year-over-year, and adjusted EPS of $12.12, up 42% year-over-year. The stock trades at 13.6 times trailing earnings and 6.9 times EV/EBITDA, with a PEG ratio of 0.19. The valuation gap is not a subtle one: the market is still pricing this subprime auto lender as if the credit cycle has not turned. The earnings print and the balance sheet say otherwise.

Credit Acceptance buys auto installment contracts from dealers - typically subprime and near-prime borrowers - and services them over the life of the loan. Revenue comes from the spread between the interest earned on those contracts and the cost of funding and credit losses. The model is simple, but the credit cycle matters enormously. Q2 2025 was the trough: GAAP EPS was $7.42, and the company's own disclosure noted a decline in forecasted collection rates that reduced expected net cash flows by $55.8 million, or 0.5%. The market treated that as structural. It was cyclical.

Credit Losses Are the Pivot

The driver of the Q2 2026 recovery is lower credit losses. That is the entire earnings inflection in one sentence. The company's statistical underwriting model forecasts expected collection rates for each contract at the time of assignment, then updates those forecasts as loans age and actual performance data comes in. In Q2 2025, loans assigned in 2022 and 2023 showed significant negative variances from their initial forecasts - minus 7.8 percentage points and minus 3.4 percentage points, respectively. Those vintage cohorts loaded up provisions and suppressed earnings. By Q2 2026, the collection environment had improved enough for GAAP EPS to nearly double and for the market to take notice, though the stock has pulled back 9.1% over the past 20 days after a strong year-to-date run of 32.6%.

Revenue stayed flat at $587.4 million for Q2 2026, essentially in line with the $588.1 million consensus estimate and consistent with the $583.8 million reported a year ago. The average loan portfolio balance held at $8.0 billion, the same as Q2 2025. Net loans receivable sat at $7.96 billion. This is not a growth story on the top line. It is a credit-quality story on the bottom line. When collection rates improve on a stable portfolio, every basis point of improvement flows directly to earnings because cost structure is fixed and the loan book is not expanding.

Management flagged July unit volume as up over 20% year-over-year on the earnings call, and outlined an AI-enabled segmentation push in underwriting. That is the next lever: more originations at improved selectivity. If origination volumes reaccelerate while collection quality holds, the earnings base lifts without the portfolio balance needing to balloon.

The Debt Gate

Credit Acceptance carries $7.18 billion in total debt against $1.51 billion in total equity, for a debt-to-equity ratio of 86.4%. Net debt - after subtracting $25.7 million in cash - sits at $1.28 billion. The balance sheet structure is worth examining carefully because leverage is the single biggest risk in subprime lending.

The company's capital returns almost entirely to shareholders through share repurchases rather than dividends. Q2 2025 saw approximately 530,000 shares repurchased, or 4.5% of shares outstanding at the start of the quarter. There is no dividend. That means the equity base is constantly shrinking, which mechanically pushes the debt-to-equity ratio higher. The $7.18 billion debt figure is large in absolute terms, but it funds an $8.0 billion loan portfolio that generates $1.05 billion in trailing-twelve-month free cash flow at a 45.3% free cash flow margin.

ROIC sits at 32.4%, and ROE at 28.1%. Those return figures are exceptional for a lender in any credit environment and more than enough to cover the cost of debt. The debt gate is not breached. The concern is structural - the aggressive buyback program erodes the equity cushion over time, which amplifies leverage. That is a deliberate capital allocation choice, not a hidden risk, but it means the equity return profile is more sensitive to a credit reversal than the balance sheet headline might suggest.

Valuation: Where the Gap Lives

At 13.6x trailing earnings and 15.9x forward earnings, Credit Acceptance looks cheap relative to its own returns but not outrageously so. The PEG ratio of 0.19 - earnings growth normalized against the PE multiple - is the standout signal. It implies the market is pricing less than one-fifth of the earnings growth trajectory into the stock. That can only last if investors believe earnings are not durable.

EV/EBITDA at 6.9x is compelling for a company with 46.5% EBITDA margins, 45.4% operating margins, and a gross margin effectively at 100% (since interest income is revenue and funding costs are operating expenses in this model). For comparison, a direct peer like Autoliv trades at 14.2x trailing earnings, but Autoliv is an automotive parts manufacturer, not a lender - the comparison is imperfect. There is no direct-listed subprime auto finance peer at scale, which makes Credit Acceptance a unique-name play. The scarcity of comparables means the valuation has to stand on its own merits: the PEG, the margins, and the ROIC.

The market cap of $6.15 billion sits below the enterprise value of $7.43 billion, reflecting the debt load. But the underlying asset - the $8.0 billion loan portfolio - is generating free cash flow at a rate that would support a significantly larger equity valuation if credit quality normalizes. At 45.3% FCF margin and $1.05 billion in trailing FCF, the stock trades at just 5.8x price-to-free-cash-flow. That is the core of the valuation gap.

The Counterargument

The strongest bear case is straightforward: the credit cycle turns again. Subprime auto lending is inherently cyclical. If unemployment rises, interest rates stay elevated, or used car values decline sharply, collection rates will deteriorate. The 2022-2023 vintage cohorts already demonstrated how quickly forecasted collection rates can fall - the 2022 vintage showed a 7.8 percentage point miss versus initial expectations. The portfolio is concentrated in subprime borrowers by design, which means it is the first to feel economic stress.

The aggressive buyback program amplifies this risk. With equity constantly shrinking, a credit downturn hits a thinner capital base. The company has no dividend to fall back on as a capital cushion; the buybacks are the primary return mechanism, and they are discretionary. If management pulls back on repurchases during a downturn, the per-share earnings impact is more severe because the share count is lower.

This is a real risk, not a speculative one. The thesis requires that credit quality remains stable or continues to improve through the next 12-18 months.

The Verdict

Credit Acceptance is a turnaround play on credit quality, not a growth play on volume. The Q2 2026 results show that the earnings inflection is real: lower credit losses on a stable $8.0 billion portfolio drove GAAP EPS up 71% year-over-year, and adjusted EPS up 42%. The valuation at 5.8x price-to-free-cash-flow and 6.9x EV/EBITDA does not reflect the 32.4% ROIC or the 45.3% FCF margin. The debt-to-equity ratio of 86.4% is elevated, but the returns on capital more than justify the leverage.

The key risk is a credit reversal, and the aggressive buyback program makes the equity base more sensitive to one. But under a scenario where credit quality holds and July's 20% volume improvement marks the start of an origination reacceleration, the earnings base has significant upside from here.

Rating: Buy. The valuation gap is real. The credit-quality inflection is real. The thesis breaks if collection rates deteriorate meaningfully on the current portfolio - but the earnings call's AI underwriting push and improving July volumes suggest management is working to widen the margin of safety, not narrow it.