CONSUMER PORTFOLIO SERVICES, INC. (CPSS) stock intrinsic value fair value estimate & owner earnings analysis

Consumer Portfolio Services buys car loans from dealers made to borrowers with weak credit. The typical 2025 contract was about $22,650 over 71 months on a seven-year-old vehicle. It funds these loans almost entirely with warehouse credit lines and quarterly securitizations, so the business is a spread lender: it earns the gap between what borrowers pay and what bond buyers charge, minus credit losses and operating costs. At June 30, 2026 the managed portfolio was $4.31 billion, up from $3.71 billion a year earlier, supported by $319 million of shareholders' equity, about $14.78 per share. For a lender, borrowing is the raw material, so we value net income to common shareholders and do not subtract the securitization and warehouse debt. We judge the earning level over the full cycle from 2016 to 2025. Per diluted share, earnings were $1.01, $0.14, $0.59, $0.22, $0.90, $1.84, $3.23, $1.80, $0.79 and $0.80. The 2017 figure includes a one-time $15.1 million tax charge. 2019 was the bottom of a credit downturn on the 2014-2016 loans, and 2021-2022 was an unusual peak when stimulus kept borrowers current. The median is about $0.85 a share. There is one accounting caveat. Since 2018 the company has carried new loans at fair value, and in 2022-2025 it booked upward revaluations of $12 million, $12 million, $21 million and $6.5 million before tax. Without them, 2024 earnings would have been about $0.15 a share. No revaluation has been booked in 2026, yet first-half earnings were $0.50 a share against $0.39 a year earlier, and quarterly earnings rose from $0.23 to $0.27. Net charge-offs were 7.28% of the average portfolio in the second quarter against 7.45% a year earlier, and 31-plus-day delinquencies fell to 9.97% from 10.50%. Earnings grew 1.6% in 2025. We start the base case near $0.95 in 2026, because two quarters of 2026 results without any revaluation support that pace. Base case, about $15.05 a share: $0.95 in 2026, then $0.90 a year, a return of roughly 6% on today's equity. This continues beyond ten years because the company has bought and serviced these loans since 1991, through the 2008-2010 collapse, and the dealer network and securitization program are proven, repeatable machinery. We credit no portfolio growth. The record second-quarter originations of $758 million, up 75%, depend on market share and credit conditions holding, and growing a lender consumes retained earnings, so the growth does not add owner cash. Bear case, about $3.38 a share: earnings soften to $0.80 and $0.60, then a credit downturn of the 2008-2010 kind arrives. In 2009-2011 this company lost $57.2 million, $33.8 million and $14.5 million, about 9% of its year-end 2009 portfolio, while the portfolio shrank from $1.19 billion to $756 million and funding markets closed. We charge after-tax losses of about $250 million over three years, roughly 6% of today's portfolio. We use less than the 2009 rate because loans are now carried at values that already assume charge-offs near 8%. Those losses would consume about four-fifths of current equity and would likely require outside capital, as last time. Earnings then recover to $0.85 a share. Bull case, about $17.50 a share: the 2026 pace of about $1.05 a share holds, with recovery rates above 32% and charge-offs below 8%. Debt is not subtracted in any case because it is the lender's working inventory, and the securitization debt is repaid from the loans pledged to each trust. Margin of safety price: $6.00. We are only moderately sure of the earning level. Reported profits have swung from $0.14 to $3.23 a share in ten years, and part of 2022-2025 rested on management's own revaluation estimates. The worst plausible event is a credit downturn coinciding with a freeze in the asset-backed bond market, which this company depends on every quarter. That happened in 2008-2009, and the company survived only through costly outside financing that diluted shareholders heavily. The equity cushion is thin against a $4.3 billion portfolio, and the chief executive has pledged about 1.7 million shares against personal loans. If we are wrong about the 2026 improvement and earnings settle near the revaluation-free 2024-2025 level of $0.15-$0.60, the shares are worth well under book value. The company pays no dividend and returns cash only through buybacks; diluted shares fell from 31.6 million in 2015 to 23.5 million. At $6.00 the price is about 40% of tangible book for a business that has earned around $0.85 a share across a full cycle. That is about 14% a year of earnings retained in the business relative to that price, even before any change in price, so the base case implies a 20% annual return within four to five years. If a downturn arrives first, the book provides a cushion at that price. Above $6, a bear-case downturn early in the holding period would leave too little room to earn 20%. Previous estimates for this company were built on a different basis and are not comparable.