$168.80
$537.70
bear $168.80
base $360.30
bull $537.70
margin of safety $110.00
market price* $161.73
AutoNation is the largest franchised new-vehicle dealer group in the United States, with stores concentrated in Florida, Texas and California (about 65% of revenue). It makes money four ways: selling new vehicles, selling used vehicles, finance and insurance products, and parts and service. In 2025 gross profit was $4.95 billion: parts and service $2.36 billion (48%), finance and insurance $1.46 billion (30%), new vehicles $0.66 billion (13%) and used vehicles $0.46 billion (9%). Parts and service is the steady half. Its revenue grew from $3.57 billion in 2019 to $4.84 billion in 2025 (about 5% a year), including through the 2020 shutdown. HOW WE MEASURED THE EARNINGS. We start from operating cash flow and take out swings in inventory, receivables, payables and vehicle floorplan borrowing (the credit line that carries cars on the lot, which is matched almost exactly by the vehicles it finances: $3.8 billion of floorplan against $3.4 billion of inventory plus about $0.46 billion of loaner vehicles). We then subtract stock compensation and the capital spending needed to keep stores running. The captive lending arm, AutoNation Finance, is treated separately: its loan book ($2.1 billion at year-end 2025) is funded almost entirely by non-recourse securitization debt ($1.9 billion), so we neutralise the growth of that book, but we do charge its credit losses ($81 million in 2025). Floorplan interest is kept as a normal operating cost; interest on the company's own bonds, commercial paper and credit line is added back so we can value the business first and then subtract what it owes. Large non-cash write-downs ($197 million of goodwill and franchise-rights impairments in 2025) are added back because they are sunk acquisition costs. Our figures tie to net income plus non-cash charges to within rounding. WHAT THE EARNINGS HAVE DONE. Cash earnings available before paying lenders were about $390 million in 2016, $450 million in 2019, a pandemic and chip-shortage windfall of $730 million to $1.46 billion in 2020-2022 (profit per new vehicle reached about $5,900 versus roughly $1,750 before the pandemic), and then $993 million, $740 million and $962 million in 2023-2025. We set the normal level at about $850 million a year, the median of 2019 and 2023-2025. We left out 2020-2022 as distorted by store closures and the chip shortage, and 2016-2018 as a smaller business before the parts-and-service scale, used-vehicle stores and lending arm. 2024 is understated by the June 2024 CDK software outage, so $850 million is deliberately on the cautious side. Most recent actual growth: 2025 was 30% above 2024 ($962 million versus $740 million), which was a recovery from the outage year. For the next twelve months we assume only 2% growth from the normal level, which is below that trailing rate. Share count has fallen from 114 million (2015) to about 33 million through buybacks, so per-share figures rose faster than the business itself; we value the whole business and divide by today's shares (about 33.6 million diluted). WHAT IT OWES. At mid-2026 the company's own debt was about $4.4 billion ($3.45 billion of senior notes maturing from 2027 to 2035, about $0.35 billion of finance leases and other debt, and $635 million of commercial paper) less $53 million of cash. Commercial paper rose $435 million in the first half of 2026 as the company bought back $470 million of stock and paid $317 million for new dealerships. Much of the bond debt carries low fixed rates (1.95% to 4.75% on $2.95 billion of the notes), which is worth about $209 million to shareholders relative to refinancing at today's cost, so we reduce the debt figure by that amount. We exclude floorplan debt (matched to cars on the lots, with interest already charged to earnings) and non-recourse loan-book debt (matched to the loans). The dealerships bought in the first half of 2026 are credited at what was paid, since they are owned and operating but have no history in our earnings. BASE CASE ($360 per share). Business cash earnings start at $850 million and grow 2% a year for ten years. That growth comes from parts and service, which has compounded at about 5% a year since 2019 and is roughly half of gross profit, with the rest of the business held flat. We leave out share gains, more new stores, margin recovery and any benefit from the lending arm, none of which are locked in. Because auto retail has always cycled, the base path includes a mild downturn in years six to eight (earnings 15%, 25% and 10% below trend), and the long-run level afterwards is held 3% below the year-ten level and then kept flat, because franchised dealers have persisted for generations under state franchise-protection laws. The business is worth about $16.0 billion; it owes about $4.2 billion after the cost-of-debt adjustment, plus about $0.3 billion credit for the new stores; that leaves about $12.1 billion for shareholders. BEAR CASE ($169 per share). Three named problems. First, profit per new vehicle keeps normalising from about $2,450 toward the pre-pandemic $1,750 (new-vehicle gross profit already fell from $1,062 million in 2023 to $665 million in 2025), taking about $135 million a year off cash earnings and starting the case at $716 million. Second, a full downturn on the scale of 2008-2010, when U.S. vehicle sales fell about 35% from 16.1 million to 10.4 million: vehicle-related gross profit falls 35%, parts and service 5%, and expenses flex only partly, cutting about $395 million a year after tax at the trough, while AutoNation Finance suffers roughly twice normal credit losses for two years. Cash earnings run $716 million, $560 million, $280 million (trough), $330 million, $450 million, $580 million and $680 million before returning to $716 million by year eight. Third, a second, later cycle lowers the permanent level to $630 million. In the trough, earnings of roughly $280 million barely cover the roughly $190 million to $280 million interest bill and the covenant limiting debt to 3.75 times covenant earnings (2.44 times at year-end 2025) would be breached, so we charge $315 million for waivers and 2 percentage points of extra interest for four years. The business is worth about $10.0 billion; it owes about $4.5 billion; shareholders keep about $5.7 billion (including half credit for the new stores). The bear value is well above book assets only because parts and service and the owned real estate ($4.0 billion carrying value, roughly equal to the corporate debt) keep earning through a downturn. BULL CASE ($538 per share). The business keeps compounding parts and service at its demonstrated pace, producing 4% growth a year for ten years from the $962 million actually earned in 2025, with no downturn, then holds that level flat. Business worth about $21.9 billion, debt about $4.2 billion net of the cost adjustment, plus $0.3 billion of new-store credit, leaving about $18.1 billion for shareholders. WHAT CHANGED SINCE OUR LAST ANALYSIS. The earlier estimate was built on a different basis and its figures are not comparable. Business changes since then: first-half 2026 operating income of $633 million versus $554 million, $317 million of dealership acquisitions, $470 million of buybacks (shares outstanding down to 33.1 million), $435 million more commercial paper, and another $749 million of non-recourse securitizations in January 2026 as the lending arm keeps growing. TESTS WE RAN. The floorplan and loan-book exclusions hold because each is matched by a specific asset. Parts-and-service growth of 5% a year is confirmed from filed revenue ($3,572 million in 2019, $4,835 million in 2025). The covenant is currently comfortable (2.44 times against 3.75 times) but the buybacks funded with commercial paper move it the wrong way. What we could not reconcile: the exact long-term-debt balance at June 30, 2026 is not available in full, so we estimated it from the year-end balance plus the reported commercial paper increase. WHAT A MARGIN-OF-SAFETY PRICE OF $110 REQUIRES. We are fairly sure of the earnings level (eleven years of filings, a parts business that did not fall in 2020) but not of the profit per vehicle or of the next downturn, and the company is more indebted than a year ago, funded by commercial paper that must be rolled. The worst plausible outcome is the bear case combined with the earnings stream fading after year ten rather than holding, which would put the shares near $80, and the structure (investment-grade-style fixed notes, $4.0 billion of owned property, no maturity above $620 million before 2030) would survive it. At $110 a share, normal cash earnings after interest of about $21 per share would be roughly a 19% yield, and even the bear case, compounding its value over an assumed holding period of about three years, implies roughly 20% a year. The business pays no dividend; it returns cash by buying back shares, and the price assumes a three-to-five-year holding period for that to show up in the price.