AutoNation Inc (NYSE: AN) — Cheaper Buybacks Meet a Thinner Profit Pool

Published: 2026-10-31 · Verdict: Hold · Entry price: $165 · Price target: $210 · Research confidence: High (82%)

Executive conclusion

Analyst Take

Hold at the verified October 30, 2026 close of $179.64. My preferred entry is approximately $165, conditional on intact funding and credit performance, and my 12-month value estimate is $210. Investment conviction is medium. The value estimate uses roughly ten times $21 of normalized earnings per share, slightly below the operating base case developed below. These are analyst judgments, not company guidance. The quotation implies approximately 8.8 times estimated trailing adjusted EPS of $20.41. The lower valuation improves prospective returns, but it arrives alongside weaker operating evidence. Company Financials and the public daily-price table agree on the closing reference. S6

The central tension is that AutoNation is concentrating a smaller dealer profit pool into fewer shares while building a promising, incompletely seasoned lender. Third-quarter adjusted net income fell 10.8% to $170.3 million, although adjusted EPS increased 2.0% to $5.11. Comparable gross profit declined 4.7%; new-vehicle gross profit per unit, or GPU, fell to $2,176; and adjusted selling, general and administrative expense absorbed 69.6% of gross profit. Share retirement creates a real larger ownership claim for remaining holders, but it does not establish that the enterprise is growing. S2

The operating defenses remain substantial. Parts and service produced $616.6 million of quarterly gross profit, and service plus finance-and-insurance commissions supplied approximately 80% of dealer gross profit. AutoNation Finance earned $15.4 million, with annualized quarterly net losses of 1.8% and delinquency above 30 days of approximately 2.5%. Those facts contradict an immediate credit-break narrative. However, approximately 82% of September loan principal originated in 2025 and 2026. Favorable current losses on a growing portfolio do not establish six-year lifetime returns, particularly when the filing itself expects delinquency to rise with seasoning. S1

My variant perception concerns the quality and availability of cash. Nine-month GAAP operating cash flow was positive $160.4 million; the earnings release’s negative narrative wording conflicts with its own reconciliation and the filed statement. Management’s $739.0 million adjusted free cash flow adds back retained-loan investment. Matching that investment against actual net nonrecourse borrowing produces an intermediate $677.7 million cash bridge before acquisitions and corporate distributions. That figure includes favorable tax timing and financing cash, so it is not a sustainable owner-cash measure. Both the negative-cash caricature and an unrestricted 17% owner yield overstate what the evidence establishes. [S1][S2]

Capital allocation has also changed meaningfully. Repurchases averaged $198.68 during the first nine months but $157.28 from October 1 through October 28. The newer vintage buys substantially more normalized earnings per dollar if those earnings endure. Meanwhile, liquidity increased and the revolver was extended to 2031. The counterweight is that commercial paper rose $355 million during the first nine months, acquisitions absorbed $316.5 million, and actual cash repurchase payments exceeded the post-acquisition matched cash bridge. Lower purchase prices improve prospective economics without proving that every earlier purchase or acquisition was attractive. [S1][S10]

The strongest counter-case is that this caution misses an inexpensive recovery requiring only modest stabilization. Easier comparisons, repaired used inventory, growing service traffic, finance operating leverage and cheaper retirement could improve per-share cash without another scarcity-margin windfall. The strongest objection is that comparable service gross profit rose only 0.8%, repair content weakened in management’s discussion, F&I cancellations increased, and dealer costs remain elevated. The case therefore depends on operational improvement that has identifiable tests, rather than automatically crediting customer lifetime value or technology spending.

Evidence confidence is higher for filed financial values than for normalized earnings, acquisition returns or lending economics. Investment conviction is lower than accounting confidence because modest changes in gross profit and expense absorption create large equity-value changes. The decision sequence is Q4 vehicle and wholesale correction, the year-end cost outcome, cash-tax normalization, and 2027 loan seasoning. Two quarters of growing absolute adjusted income, comparable service gross-profit growth of at least 4%, and leverage below three times would materially strengthen the case. Persistent SG&A at or above 68.5% through mid-2027, continued vehicle-margin deterioration, or seasoned losses above the prior 4% credit threshold would weaken it. The prior thresholds remain visible; weaker results do not justify quietly loosening them. [S1][S2][S8]

Changes since 2026-09-10

The September analysis correctly identified share-count support and weak organic contribution. Its new-GPU stability premise broke, its treatment of standardized ROIC was too categorical, and its cash-yield discussion needed a better financing and insurance bridge. The latest audit also identifies an actual software-accounting adoption, rather than treating accounting policies as wholly unchanged.

The former accumulation price condition was first reached after publication on September 17: the shares traded below $185 and closed at $174.88. That establishes that the condition occurred, not that the investment thesis succeeded. The October 30 close is 13.3% below the subsequently verified September 10 close of $207.22, or approximately 13.0% below the baseline’s explicitly intraday observation. [S6]

Intervening disclosure matters to causality. At the September 17 conference, management warned of approximately 10% sequential new-GPU pressure and a $50–$100 reduction in customer-financial-services PVR. Actual new GPU fell 8.6% from Q2 and F&I PVR fell $101. The results largely confirmed a prior warning, rather than constituting wholly unexpected October information. S9

Stock Price Action — Five-Year Event Map

Company Financials supplied 1,255 split-adjusted daily observations from November 1, 2021 through October 30, 2026. The first close was $123.38 and the latest $179.64, a 45.6% price increase. The five-year intraday low was $94.92 on October 20, 2022; the high was $235.81 on July 29, 2026. The trailing 52-week range was $151.82–$235.81. The current close is approximately one-third of the way up that range, 18.3% above its low and 23.8% below its high. These are price observations; the extrema are not closing-price returns. [S6]

• November 2021–October 2022: the close fell from $123.38 to $95.60. Interpretation: investors anticipated financing pressure and eventual normalization while reported dealer earnings remained unusually strong. The annual filing confirms the earnings context; it cannot allocate the stock decline among competing explanations. [S12]
• October 2022–July 2023: the close recovered to $181.47, approximately 90% above the October closing trough. Cash generation and substantial retirement offered a plausible explanation for confidence that normalization was survivable. That interpretation does not convert the entire price gain into operating growth. [S12][S13]
• June 2024: the close fell from $161.78 on June 20 to $155.25 on June 25 during the CDK disruption. The filing documents interrupted dealership functions and estimated lost contribution. The precise market-adjusted price impact was not isolated. [S29]
• April–September 2025: the shares reached $148.33 intraday on April 7 and closed at $226.82 on September 5. Tariff uncertainty, transaction timing, improving adjusted results and repurchases provide operating context. None alone explains the complete move. [S3][S15]
• July 2026: the shares reached the five-year intraday high on July 29. On July 31, after Q2 results, they opened at $198.04 and closed at $212.40. The recovery within that session demonstrates changing expectations, not a verified earnings trough. [S5][S8]
• September–October 2026: September 17’s close fell 10.49% to $174.88 on conference day. The stock subsequently reached $151.82 intraday on October 14. Management’s margin warning was relevant, but dealer-sector weakness also mattered. [S9]
• October 30: the close rose 6.08% from $169.34 to $179.64 following Q3 results, after trading between $176.41 and $192.99. Favorable repricing relative to expectations coexisted with a retreat from the session high. Neither movement identifies individual buyers’ motives. [S2][S6]

A matched-date comparison improves the attribution discipline. From September 10 through October 30, AN declined 13.3%, Lithia 18.1%, Asbury 5.2% and Group 1 7.5%, using consistent closing prices from Company Financials. AN underperformed two selected peers and outperformed one. Their simple equal-weight price return was approximately negative 10.3%. This is descriptive context, not a regression-derived abnormal return; dividends and differing event exposures also matter. It refutes treating the entire decline as uniquely company-specific. No verified current borrow or short-covering census supports a squeeze explanation. [S6]

Verdict: The valuation reset is substantial and partly shared with peers. The tape supplies evidence about expectations and volatility, while the operating filings must establish whether earnings have stabilized.

Business Overview

Customer value and the connected transaction

The business is understandable: sell a vehicle, arrange the transaction and financing, provide ongoing care, and compete for the replacement purchase. Customers buy selection, inspection, trade-in liquidity, documentation, registration, credit access and physical execution. They later buy technician time, parts, diagnosis and repair reliability. The analytical complexity lies in allocating shared costs and capital among those activities, especially when financing moves from an immediate commission into a retained loan. S3

At September 30, AutoNation operated 319 new-vehicle franchises through 242 U.S. stores representing 30 brands, plus 51 collision centers, 24 AutoNation USA stores, four auctions and three parts-distribution centers. A franchise is not a rooftop: one store can house several authorizations. The smaller store and franchise counts relative to year-end coexist with acquisitions because divestitures, closures and franchise changes affect the perimeter. Counting acquisition announcements cannot reproduce the estate. [S1] (Note 1)

The reportable dealer segments are Domestic, Import and Premium Luxury, with AutoNation Finance separately reported. Used vehicles and service are departments spanning those segments. Corporate and other contains unallocated costs and several nonfranchised businesses. It should not be valued as a single coherent fourth dealership segment. Departmental gross profit also precedes substantial shared expense, so separately capitalizing each attractive department can double-count the same premises, customers and administrative infrastructure.

These are filed values and simple percentage calculations. The lender’s $15.4 million income sits separately below dealer gross profit. Service and F&I generated approximately 80% of that gross profit on 23% of revenue. The mix supplies a useful defense, but its share also increases when vehicle earnings fall. A larger defensive percentage does not independently establish faster growth in the defensive activities. [S1]

F&I’s near-100% accounting gross margin reflects net commission presentation. Selling compensation, customer acquisition, administration, compliance and cancellations remain economic costs. AutoNation Finance adds interest income but also funding, servicing and expected losses. A transaction routed internally must be assessed across both locations in the statements; adding current F&I to future lending income without subtracting the foregone third-party fee overstates the economics.

Revenue stability and recurrence

Nominal revenue has been relatively stable, but the business remains predominantly transactional and its margins are considerably less stable than sales. Revenue moved from $25.84 billion in 2021 to $27.63 billion in 2025, approximately 1.7% annualized growth. Vehicle prices, mix, service and acquisitions contributed alongside units. That stability does not establish organic volume growth or contracted future demand. [S11][S12][S13]

Service demand recurs because vehicles require care. Customers can nevertheless defer work, drive less, select an independent provider or replace the vehicle. Warranty and recall activity is more directed toward authorized dealers; customer-pay work is more contestable. Loan interest persists only while balances remain outstanding and performing, with amortization, prepayment and losses reducing duration. Protection-product revenue usually requires another vehicle sale or renewal. These are different forms of continuity, not interchangeable subscriptions.

Internal reconditioning creates another interpretive complication. A service department repairs a vehicle owned by another department before retail sale. Departmental contribution and consolidation eliminations determine presentation. Lower internal work can depress service gross profit while externally sourced traffic improves. Conversely, higher reconditioning can look like service expansion without representing additional independent customer demand. This distinction explains why reported service records and weak comparable contribution can coexist.

Geography, manufacturers and assets

All operating stores are in the United States. Florida, Texas and California supplied approximately 65% of 2025 revenue. Density can improve inventory sharing, local advertising, parts routes and resource utilization. It also concentrates employment, housing, insurance and catastrophe exposures. A geographically dense estate is advantageous when it reduces required capital or raises retained contribution; density alone does not prove either outcome. [S3] (Business)

Eight major manufacturer groups represented approximately 88% of new vehicles sold during the latest nine months. Brand breadth reduces dependence on one model cycle, but OEMs still determine product design, allocation, incentives, facility requirements and warranty terms. The retailer can execute better within the assortment available to it. It cannot independently create a desirable product or reproduce a manufacturer’s subsidized financing program at no economic cost. [S1]

Economically valuable assets incompletely captured by book value include customer relationships, technician capacity, operating data, local reputation and possible appreciation in owned property. Acquired franchise rights and goodwill are already recognized and cannot simply be added again as hidden value. September property and equipment was $4.04 billion at depreciated cost. No independently verified appraisal establishes a realizable property surplus. Selling property can release capital while creating rent, taxes and reduced operating flexibility. [S1][S3]

Private investments require the same discipline. Investments without readily determinable fair values carried $123.3 million in September. Observable financing-round marks can affect earnings without generating cash. Ownership percentages, dilution, liquidation preferences, restrictions and taxes determine realizable value. The report gives no unsupported premium for these investments and does not treat their carrying amount as unrestricted parent liquidity. [S1] (Note 15)

AN is ordinary common equity in a Delaware corporation listed on the NYSE, not an ADR, MLP or partnership interest requiring issuer-level partnership K-1 reporting. Its public security wrapper is conventional. The important complexity is the combined retailer and lender, rather than an unusual pass-through structure. [S3] (cover)

Verdict: AutoNation offers a coherent, useful customer service with an attractive gross-profit mix. Transaction dependence, manufacturer control and shared capital prevent treating the whole operation as a simple recurring-income franchise.

Industry Dynamics

Domestic scale and demand constraints

The addressable market is large, domestic and mature; individual operators can grow while the industry’s profit per transaction contracts. NADA’s midyear 2026 data reports 16,991 franchised light-vehicle dealers, 7.9 million vehicles sold, more than $640 billion of dealer sales and over 136 million repair orders generating nearly $83 billion in service and parts sales. Fragmentation creates potential acquisition targets, but those targets have sellers who capitalize expected future profits. Market size does not establish attractive acquisition returns. S16

September’s U.S. new-light-vehicle sales pace was 16.0 million annualized, down 3.6% year over year; the year-to-date pace was 16.1 million, down 1.9%. SAAR is an annualized selling rate, not the number of vehicles sold in one month. NADA highlights EV-credit timing and holiday-reporting differences. Those comparison effects limit inference from one month and support a normalization framework rather than a structural demand boom. S33

Affordability depends on the total payment burden: purchase price, interest, term, trade equity, insurance and operating costs. NADA cited an estimated September new-vehicle payment of $821, up 3.2% year over year, using JD Power’s forecast. That is an industry estimate, not AutoNation’s realized customer payment. Longer terms can reduce installments while extending negative-equity exposure. Higher transaction prices can preserve nominal dealer revenue while shrinking the population able to replace vehicles.

Used demand is related but not mechanically countercyclical. An unaffordable new vehicle can redirect a customer to used, but also reduce trade-in supply and raise used acquisition costs. A customer unable to finance either purchase may defer replacement entirely. Retailers therefore need the right condition, mileage and price assortment, rather than simply more inventory. The latest AutoNation disclosures of inadequate lower-priced sourcing illustrate the operational consequence of this industry constraint. [S1]

Profit pools, competitors and barriers

Vehicle retail has thin percentage margins; service and F&I offer richer contribution, while thousands of local incumbents remain relevant competitors. OEM consent, franchise restrictions, suitable property, technician capability, licenses and inventory funding constrain entry. Those protections support approved local franchises, but neighboring authorized dealers frequently possess similar rights. The market is fragmented nationally and competitive locally, rather than a national oligopoly with coordinated pricing. [S3] (Markets and Competition)

Service is differentiated by authorization, diagnostics, training, parts access, appointment availability and repair quality. Warranty and recall work benefits from the OEM relationship. Customer-pay work faces independents, tire chains, quick-service providers and mobile operators. Skilled technicians are a binding resource: empty bays do not establish saleable capacity if staffing and parts are missing. More repair orders can also increase scheduling and advisory costs without proportional labor hours or contribution.

F&I combines protection distribution and credit arrangement. Attractive reported gross profit is offset by selling expense, compliance and cancellation risk. Captive lending adds a different capital cycle: cash leaves at origination and returns over years, while expected losses are recognized upfront and unexpected losses emerge later. A consumer-credit portfolio can therefore increase both future earnings potential and the resources required before those earnings are realized.

Competition is intensifying in digital discovery, used sourcing, transaction transparency and service convenience, despite constrained new-franchise entry. Carvana reported 39% used retail unit growth to 217,385 in Q3. CarMax reported 13.8% used retail growth in its quarter ended August. These figures do not establish which customers left AutoNation, but they contradict an explanation that macro conditions alone prevent every retailer from growing. Business mix and reporting periods must remain explicit. [S22][S30]

The supply-side capital cycle

Franchise consolidation mostly transfers existing capacity and relationships. It can avoid the oversupply associated with unrestricted greenfield expansion, yet acquisitions often capitalize the very scarcity rents the buyer hopes to earn. Synergies create shareholder value only after purchase price, inventory, facilities, integration and overhead. If every bidder anticipates the same local savings, much of the value accrues to the seller.

Used retail has fewer legal franchise barriers and more flexible geographic competition. It still requires inventory, reconditioning, logistics, funding and customer acquisition. National digital capacity can improve selection and delivery while increasing competitive supply. AutoNation USA’s limited 24-store footprint shows that a recognized brand and capital access did not establish an automatically replicable growth format. Capital discipline can be more valuable than maximizing store openings. [S1]

The pandemic shortage made the capital cycle unusually visible. AutoNation’s new-vehicle days supply was eight days in March 2022 and 52 days in September 2026. Annual new GPU declined from $5,942 in 2022 to $3,045 in 2024, followed by $2,176 in the latest quarter. Product availability improved while dealer bargaining power weakened. Scarcity earnings should not be treated as permanent simply because facilities and franchise agreements endured. Equally, normalization does not imply margins must decline indefinitely. [S34][S11][S1]

Regulation, trade and technological change

The FTC withdrew the CARS Rule effective February 12, 2026, following the court decision. Existing prohibitions on deceptive pricing remain. The pricing-transparency FAQs state staff views about existing law and are not binding new rules. Required dealer charges belong in the advertised offer; optional products should not be presented as mandatory. The investment effects include compliance cost and changes in price presentation, rather than an automatic reduction in every lawful final transaction price. [S18][S17]

Transparent advertising can help a retailer whose competitors previously used unattainable headline prices. It can also make genuine price competition more intense. Attributing a particular company’s growth to enforcement requires evidence on actual competitors, customer conversion and final prices. Coincidence between guidance and unit growth is insufficient. Conduct standards also reach financing, add-on products, privacy and title execution.

Foreign low-cost labor cannot directly replace local repairs, warranty execution, title processing or physical delivery; global production and parts costs can materially change the merchandise economics. Tariffs can increase costs, shift brand demand or constrain allocation, depending on OEM responses. Electric vehicles may reduce some routine maintenance while increasing demand for specialized diagnostics and repair. The net service outcome depends on the installed fleet and work mix, not a single assumption that every technological change either destroys or enlarges the whole profit pool. [S3]

Verdict: Local franchise protection and authorized service support durable participation. Shared barriers, transparent shopping, acquisition pricing and delayed credit losses keep excess returns contestable, particularly after extraordinary scarcity margins disappear.

Competitive Position

The credible advantages and their financial tests

AutoNation’s strongest advantage is approved local access joined to service capability. Losing a desirable franchise would reduce new inventory access, OEM-linked traffic, warranty reimbursement and customer confidence, potentially stranding facilities. That is a real economic mechanism. It belongs partly to the authorized channel and location, however, rather than exclusively to the corporate parent. Peers can own similarly valuable local positions. [S3]

Density offers more company-specific potential. Inventory sharing improves matching, shared reconditioning can reduce duplicated investment, and parts routes spread delivery infrastructure across brands and customers. Management reports double-digit wholesale-parts growth with limited incremental capital. That supports the mechanism as a hypothesis. Its financial test is incremental contribution after labor, fuel, inventory and overhead, not simply route revenue or the number of stores within a metropolitan area. [S7][S9]

The new-vehicle operation also supplies trade-ins, product attachment and future service customers. This can improve used sourcing relative to a retailer that must buy nearly everything separately. It does not make trade inventory free: appraisal errors, reconditioning, aging and wholesale disposal remain costs. Q3’s used-inventory correction is disconfirming evidence against assuming that access alone ensures superior execution. The advantage must appear in complete transaction returns and inventory velocity.

Brands, competition and customer switching

OEM brands matter through product demand, incentives, residual values and authorized service; AutoNation’s incremental corporate-brand return is unquantified. A shopper may select Toyota, BMW or Ford before choosing the retailer. The national banner can support trust, direct sourcing and repeat visits, but advertising expense itself does not establish a causal premium or lower acquisition cost. A useful test would compare contribution and retention across matched markets after marketing expense. [S3][S9]

Competition centers on the complete transaction: vehicle availability, final price, trade allowance, payment, credit terms, execution and subsequent convenience. A low advertised vehicle price can be offset by an unfavorable trade or financing offer. Broad lender access and physical delivery help AutoNation reduce complexity, but competitors can reproduce many digital features. A website and online application are necessary capabilities; their advantage must appear in conversion, customer economics or required capital.

Purchase switching costs are low; service history, warranty authorization, prepaid products, location and trust create stronger but incomplete retention. Customers can choose another same-brand dealer, refinance a loan or cancel eligible protection products. After warranty, independents become more relevant. Process friction after assembling a trade appraisal and approval is real, but it is weaker than contractual captivity. Lifetime-value forecasts must incorporate attrition instead of treating the first purchase as permanent ownership of the customer. [S3]

Current peer evidence

The relevant financial peers are Lithia, Asbury and Group 1, with Penske and Sonic providing additional franchise and format context. Used-only CarMax and Carvana are direct competitors in important transactions, but their service architecture and financing recognition differ. Independent repair networks contest another part of the relationship. A single peer average cannot represent all these exposures.

The issuer releases verify the numbers. Geography, brand mix, same-store populations, protection underwriting and cost allocation differ; these are not controlled efficiency rankings. Nevertheless, AutoNation’s current new GPU is lower than this selected group and its expense ratio is not best in the group. Broad sector pressure is real, while the differences leave room for company execution. S19 S20 S21

Asbury’s completed systems conversion supplies a different operating catalyst and its protection-product operation changes F&I economics. Lithia adds U.K., Canadian and financing complexity. Group 1 combines U.S. and U.K. operations and is reshaping its acquired portfolio. Penske’s trucks and distribution activities dilute its usefulness as a pure dealership reference. Sonic’s separate used format makes capital-allocation and stand-alone-store economics particularly relevant.

Carvana sold more than three times AutoNation’s 65,410 quarterly used retail units and grew much faster. Its broad digital reach challenges inventory selection and convenience, while the absence of AutoNation’s authorized service network changes retention economics. CarMax’s fiscal-quarter growth likewise challenges a purely macro explanation for AutoNation’s used-unit decline. Neither comparison proves direct share transfer or permits importing competitors’ margins into AutoNation’s valuation. [S22][S30][S1]

Evidence limiting the moat claim

The most important disconfirming evidence is financial. Comparable service growth remains weak, adjusted costs absorb more contribution, and current used inventory required correction. In 2025, Mobile Service goodwill impairment was $65.3 million and franchise-right impairment $93.7 million. These are evidence that some committed capital failed to support its carrying value, even within activities that fit the broader customer strategy. [S3]

Technology could improve that outcome through sourcing forecasts, customer contact, collections and process automation. The tools themselves are widely available. Durable advantage would come from execution, useful data and customer response, measured after implementation expense. Announcing an AI agenda does not establish cost leadership, and selected productivity metrics do not substitute for enterprise contribution.

Verdict: AutoNation has valuable local franchises and credible density opportunities. The evidence supports a competitive franchise operator with execution-dependent advantages; it does not establish an exclusive corporate moat or immunity from margin pressure.

Growth History and Forward Opportunities

Growth quality and the service opportunity

The best outlook is for service retention, selective finance growth and density improvements; each needs evidence of contribution after required capital. Revenue increased only 6.9% between 2021 and 2025, while net income fell from $1.373 billion to $649 million. Diluted shares declined from 75.0 million to 38.1 million. The resulting per-share resilience reflects substantial retirement alongside operating cash generation, rather than broad expansion of the underlying earnings pool. [S12][S3]

Departmental history explains the transformation. Service gross profit increased from $1.900 billion in 2022 to $2.355 billion in 2025, while new-vehicle gross profit fell from $1.367 billion to $665 million. Used contribution also declined over that interval. The company improved the relative importance of more durable activities while losing a substantial temporary vehicle-margin windfall. Extrapolating consolidated EPS without separating these movements obscures both progress and dependence on repurchases. [S11][S13][S3]

The attractive organic opportunity is better use of existing facilities and technician capacity. Management reported Q3 customer-pay and warranty gross-profit growth of 5% each, supported by traffic growth of 5% and 13%, respectively. The filed comparable-service bridge is narrower: customer-pay GP increased $5.6 million, wholesale-parts GP $4.3 million, and internal preparation GP declined $7.0 million. Total same-store GP consequently rose only $4.5 million. These disclosures describe different perimeters and should not be combined into a single organic growth rate. [S7][S1]

There is a further apparent tension in repair content. The call acknowledged more deferred work and fewer labor hours per repair order, while the filing described customer-pay growth supported partly by higher-value repair orders. Those statements can coexist through pricing, parts content, mix or period differences. They do not establish uniformly improving or uniformly deteriorating repair economics. The missing bridge is matched repair orders, labor hours, effective rates, parts mix and contribution after staffing.

Management expects the pandemic-related gap in vehicles entering preferred dealership-service ages to peak in 2027. That is a management outlook, not an independently modeled fleet forecast. It challenges the broad assumption that an older national vehicle fleet automatically accelerates franchise service. AutoNation must retain owners longer, win older vehicles back from independents and convert deferred work into actual contribution. Tracking deferred recommendations is useful, but does not prove where or whether customers ultimately complete them. [S7][S8]

Wholesale parts offers a different route. The company can use existing franchise inventory and local delivery routes to serve external repairers. Lower percentage margin can still accompany attractive incremental returns when new capital needs are modest. The correct evaluation combines dollar contribution and inventory turns. Rejecting every distribution-mix margin decline misses useful growth; accepting every such explanation risks overlooking delivery, labor and stocking costs.

F&I and retained lending

F&I remains large but its immediate outlook weakened. Q3 PVR fell from $2,799 in Q2 to $2,698, and total gross profit fell to $349 million. The filing attributes the comparable decline mainly to lower units, with the remainder from PVR. Cancellations and retained lending were headwinds, partly offset by protection-product margins and financing-arrangement economics. Retained lending therefore explains only part of the weakness; it cannot absorb every unfavorable commission trend into a strategic timing story. [S1][S2]

AutoNation Finance originated $488.7 million across 14,373 units in Q3. Weighted-average contract rate was 10.6%, FICO 711, front-end loan-to-value 100.9% and term 72 months. Penetration was 11.1% of all retail new and used units. That denominator must remain explicit: percentages of financed transactions are higher and answer a different question. The filing also defines front-end collateral value using selling price plus specified taxes and fees; the reported LTV is not a liquidation-value coverage ratio. [S1]

Management says retained loans can earn two to three times the lifetime profit of third-party placement. That is a management estimate without a public matched-cohort cash-return schedule. A valid comparison deducts foregone commission, funding, expected and unexpected losses, servicing, overhead, taxes, required equity and time value. More nominal profit spread over six years can be economically inferior to a smaller immediate fee. Better retention or incremental retail sales could improve the result, but require separate evidence. [S9][S1]

The latest call also described approximately $93 million of year-to-date finance operating cash and a path to at least $150 million in 2027. These are management-defined claims requiring reconciliation to consolidated loan growth, debt issuance and restricted cash. They should not be added to adjusted group FCF as another independent stream. The lender’s value is its future cash after necessary funding and equity, rather than the sum of every cash label used in presentations. [S7]

Acquisitions and technology

The four 2026 acquisitions add about $600 million of annual revenue in existing markets, according to management. Density offers a credible synergy mechanism, but reported acquired revenue is not acquired cash return. The purchase-price allocations are preliminary and public disclosures do not isolate earnings, integration, working capital and facility commitments sufficiently to prove the investments outperform repurchases. [S1][S7]

Ravi Simhambhatla became Chief Technology and AI Officer on October 1. Management describes gradual 2027 deployment of a used-vehicle model incorporating replenishment, velocity and lifetime customer value. The opportunity is to improve existing capital productivity rather than assume a separate technology revenue stream. Forecast benefits should follow measurable sourcing, turnover, conversion and cost outcomes after implementation spending. [S26][S7]

Verdict: The opportunities are credible and mostly connected to existing assets. Their quality depends on organic contribution, disciplined cohort returns and allocated capital; traffic, portfolio balances and acquired revenue alone do not establish growth in economic value.

Financial Quality

Five-year statements and the cycle

Earnings are well below the scarcity-era peak, but a cyclical trough has not been demonstrated. Vehicle margins and comparable dealer contribution remain under pressure; service growth is modest; and finance earnings are rising on a young portfolio. A trough requires stabilization across the combined contribution bridge, not merely a large decline from 2022. [S1][S3]

The table uses filed GAAP amounts, including corrected 2022 operating income of $2,024.5 million. Company Financials’ standardized operating-income series differs materially in several years: for example, $1,309.9 million versus filed $1,239.9 million in 2025. The standardized series remains useful for discovery, but the numerator used for margins and returns must be reconciled to the filing. Its standardized 7.7% 2025 ROIC is consequently a diagnostic with a separate methodology, not an audited economic-return conclusion. [S3][S11][S12][S13][S14][S32]

The income history contains substantial mix change. Between 2022 and 2025, revenue rose slightly while net income fell approximately 53%. Diluted EPS fell approximately 30% because the average share count contracted. That is economically meaningful ownership concentration, but it cannot be forecast indefinitely without a source of cash and an acceptable purchase price. Conversely, comparing only absolute income understates the benefit to continuing holders of shares retired below intrinsic value.

Loans, acquisition assets, inventory normalization and distributions explain why asset growth did not produce a comparable increase in common-equity protection. The table is not evidence that every added asset earned an inadequate return; it establishes the increasing importance of capital productivity. Repurchases also change book equity, making ROE a poor independent quality measure. [S1][S3][S12][S13]

Segment and quarterly bridges

The annual segment history provides a useful countercheck against treating finance growth as the principal earnings engine.

Dealer segment income includes floorplan interest and precedes unallocated corporate items and other interest. Quarter and annual columns are not directly comparable run rates. The table shows genuine lender progress, some dealer recovery in 2025, and renewed current pressure. It also shows why $15 million of quarterly finance income cannot by itself neutralize weakness in a dealer profit pool exceeding $300 million. [S3] (segment note)[S1]

Q3 revenue declined 0.4%, gross profit 2.5%, GAAP operating income 15.5% and GAAP net income 24.5%. Adjusted operating income fell 7.3% to $322.7 million and adjusted net income 10.8% to $170.3 million. Different GAAP and adjusted comparisons reflect insurance recoveries, market-linked compensation and current designated weather and transformation items. Neither comparison should be selected solely because it supports a desired conclusion. [S2]

At the prior-year diluted share count of 38.1 million, current adjusted income would produce approximately $4.47 per share, versus reported $5.11. This is an analyst counterfactual demonstrating the denominator’s importance. Management’s EPS bridge likewise assigns substantial benefit to retirement. It does not prove value destruction: a shrinking enterprise can still create attractive per-share returns if cash exceeds necessary reinvestment and stock is purchased sufficiently cheaply.

The comparable-store bridge is weaker than consolidated revenue suggests. New GP declined $16.6 million, used including wholesale $11.2 million and F&I $32.8 million, while service added only $4.5 million. Comparable total GP fell $58.0 million. Acquisitions therefore softened the consolidated comparison without curing the inherited estate’s decline. Combined dealer-segment income fell 9.4%; Domestic was down 19.4%, Import 6.8% and Premium Luxury 6.4%. [S1][S2]

Adjusted SG&A dollars increased from $834.4 million to $841.0 million while the ratio rose from 67.4% to 69.6%. Management says dollars were approximately flat excluding acquisitions. That explanation does not remove operating deleverage: lower gross profit still leaves less cash after a largely unchanged expense base. The filed GAAP ratio was higher, 70.3%, versus 69.6% adjusted. Expense efficiency must be judged on matched perimeters and consistent adjustments.

Normalization and returns on capital

For 2025, GAAP net income was $649.1 million and adjusted net income $770.3 million; GAAP EPS was $17.04 and adjusted EPS $20.22. A simple trailing adjusted-EPS estimate is $20.22 plus current nine-month $15.33 less prior nine-month $15.14, or $20.41. Discrete-period EPS uses changing share counts and rounding, so this is an approximation. Trailing adjusted net income is $708.7 million, a more direct enterprise-earnings bridge. [S15][S2]

Impairments are noncash when recognized but reveal earlier capital committed without the expected support. Insurance recoveries reimburse a disruption rather than supply recurring operating profit. Weather losses may be unusual in a quarter while catastrophe exposure recurs over time. Repeated transformation costs should not become permanently invisible. Normalized earning power therefore requires judgment beyond accepting every management exclusion.

A matched, acquisition-inclusive 2025 capital-return sensitivity is approximately 11.1%; it is definition-dependent rather than a unique economic ROIC. Start with $1,239.9 million operating income, subtract $188.8 million floorplan interest, add back $76.3 million finance funding interest, and apply an assumed 25% tax rate. The numerator is $845.6 million. Average capital is $7,596.1 million, calculated below. [S3]

This convention treats floorplan as inventory financing with its interest retained as operating cost, and lender funding as financing with its debt included in capital. Finance leases embedded in non-vehicle debt remain included. Operating leases are excluded while rent remains expensed; adding their liabilities alone would mismatch numerator and denominator. Goodwill and recognized intangibles remain. Endpoint averaging, assumed taxes and required operating cash limit precision.

Adding back the two major impairments totaling $159.0 million while removing all $54.2 million of operating other income increases the defined return to approximately 12.2%. This is a specified sensitivity, not management’s adjustment schedule. Historical write-downs and amortization still mean book capital does not reconstruct every dollar originally committed. Returns on original acquisitions can therefore be weaker than this normalized accounting return.

Management’s compensation return uses a different numerator and quarterly averaged capital over three years. Its reported 16.48% three-year result cannot be presented as a current consolidated single-year return. Likewise, management’s approximately 24% finance ROE annualizes Q3 and depends on allocated equity. Loan principal less debt is not a complete finance-equity calculation because reserves, restricted cash, accrued items and other liabilities intervene. [S4][S7]

Credit quality and accounting judgments

September loan principal was $2,864.9 million, allowance $114.6 million and net receivables $2,783.6 million after accrued interest, deferred origination costs and discounts. Delinquent principal above 30 days was $71.1 million. Quarterly annualized losses were 1.8%, while the nine-month rate was 2.0%. Principal, managed balances and net accounting assets must not be mixed indiscriminately in coverage or funding ratios. [S1] (Note 6)

The allowance reconciliation is unambiguous: $95.4 million opening reserve plus $57.0 million provision, less $90.1 million write-offs, plus $52.3 million recoveries equals $114.6 million. A lower coverage ratio did not represent a dollar reserve release, but improved expected-loss assumptions still benefited earnings. The filing expressly attributes lower modeled loss rates to better originations. Both observations can hold simultaneously. [S1][S5]

Quarterly finance profit improved $13.9 million. Net interest margin dollars added $9.9 million, lower provision added $4.8 million and direct expense increased $0.8 million. Approximately one-third of the improvement therefore came from the provision comparison. Meanwhile, annualized net interest margin declined from 7.2% to 6.3% of average managed receivables. Dollar growth should not be confused with improving spread economics.

Approximately $2.357 billion of principal came from 2025 and 2026 originations. Their full loss curves remain incomplete. Higher FICO and lower front-end LTV support underwriting improvement, but the six-year term extends consumer and collateral exposure. Aggregate loss ratios may also improve when newer, better-credit loans dilute older balances. Matched-age vintage outcomes are more informative than either the most favorable aggregate ratio or an assumed inevitable credit collapse.

Accounting uses conventional frameworks but contains consequential estimates in credit losses, franchise values, cancellations, self-insurance and private investments. The 2025 audited statements and controls received unqualified opinions. That supports reliability without proving each valuation assumption conservative. The January 2026 software adoption, discussed below, also demonstrates why policy changes must be checked directly rather than inferred from stable headline presentation. [S3][S1]

Cash, funding and economic obligations

Net income diverges from operating cash mainly because retained-loan investment consumes operating cash while much of the associated borrowing enters financing cash flow. Nine-month operating cash of positive $160.4 million compared with $549.8 million net income includes $708.4 million of loan investment and $195.4 million inventory use. Cash taxes were $104.4 million versus $238.0 million a year earlier, a $133.6 million timing benefit. The release’s negative wording must yield to the filed statement and reconciliation. [S1][S2]

September unrestricted cash was $51.2 million; restricted cash was another $91.4 million. Corporate and finance-lease debt plus commercial paper totaled approximately $4.344 billion; floorplan was $4.019 billion; nonrecourse loan funding had a $2.587 billion carrying amount. Each should be matched to its purpose, assets and interest costs. Summing them into one retail leverage ratio obscures risk, while ignoring either matched debt category understates economic dependence on funding.

Covenant leverage was 2.71 times against a 3.75 maximum and interest coverage 4.61 times against a 3.00 minimum. The September amendment allows unrestricted-cash netting, qualifying comparisons with older calculations. Reported liquidity was approximately $1.5 billion. Nominal commitments, leverage-limited capacity, commercial-paper backup and unrestricted cash are distinct resources; recent facility extensions improve access without making all committed dollars immediately spendable. [S1][S10]

The business is moderately capital-intensive in facilities and substantially capital-intensive after inventory and lending are included. Physical cash capex was $309.4 million in 2025, but September inventory and net loans exceeded $6.4 billion. Tangible equity was approximately negative $405 million after goodwill and other intangibles. That is not insolvency, but it reduces the accounting cushion against losses and impairments. [S1][S3]

Material economic obligations include leases, self-insurance, OEM facility requirements, F&I chargebacks, legal contingencies and lender collateral or liquidity requirements. Most leases and securitization debt are recognized; rigidity and contingent cash demand matter more than a claim of hidden conventional debt. Floorplan exceeds reported inventory partly because financing also supports loaner and rental vehicles and related balances. The difference is not evidence of freely distributable excess financing. [S3][S1]

Other interest expense rose to $51.7 million from $45.1 million in Q3, reflecting higher average corporate debt. Floorplan expense instead fell to $45.4 million from $47.5 million as lower rates offset higher balances. Broad commentary about higher interest should not erase this distinction. [S1]

Verdict: Current cash generation and near-term credit are stronger than the most bearish narrative. Organic contribution and cost absorption remain weaker, and both capital returns and lender economics require matched definitions and continued seasoning.

Capital Allocation

Allocation philosophy and the cash bridge

Management prioritizes long-term value per share through reinvestment, selective acquisitions and repurchases, with no recurring common dividend. It explicitly compares opportunities with the repurchase alternative. The framework can create value, but EPS accretion alone does not establish an adequate return after duration, debt and required equity. [S1][S9]

Net nonrecourse borrowing was $647.1 million, versus $708.4 million loan use. This constructs an observed financing-dependent cash bridge, not a permanent equity requirement or owner FCF. It does not fully adjust restricted cash, legacy investing-classified collections, debt issuance costs, tax timing or every working-capital movement. The $133.6 million lower cash-tax payment is particularly important when assessing repeatability. [S1][S2]

Actual repurchase cash payments were $631.4 million, compared with the $378.6 million intermediate remainder: a $252.8 million gap before other flows. The commitment-style repurchase amount was lower, $619.9 million excluding excise tax. Settlement timing and tax explain why these bases should not be interchanged. Commercial paper increased $355 million, while other investing and financing items also affected cash. It is fair to conclude allocation exceeded this intermediate internally generated post-acquisition measure; it is not fair to label the entire difference unexplained cash burn.

Repurchases, dilution and funding

Net share retirement is substantial, but purchase price and funding differ across vintages. The company repurchased approximately $2.30 billion in 2021 near $103 per share, $1.71 billion in 2022 at $109.86, $864 million in 2023 near $135, $460 million in 2024 near $161 and $785 million in 2025 near $193. The early program bought much more subsequent earnings capacity per dollar than the latest high-price vintages, although a complete investment return depends on future cash and funding costs. [S12][S13][S3]

First-nine-month 2026 purchases averaged $198.68; Q3 averaged $193.49. October 1–28 purchases averaged $157.28 on $47.9 million. The October price was approximately 21% below the nine-month average. At unchanged normalized earnings, that materially improves the earnings yield purchased. A lower price is favorable evidence, not an unconditional value-creation certificate if earning power deteriorates. [S1]

September shares outstanding excluding treasury were 32.288 million, down from 35.200 million at year-end; October 28 shares were 31.984 million. The nine-month net reduction was approximately 8.3%, despite employee issuance. Shares bought into treasury still reduce outside ownership and the economic denominator; formal cancellation of every certificate is not necessary for that benefit.

Employee equity issuance is meaningful but currently much smaller than repurchases: nine-month stock compensation was $35.3 million versus $619.9 million committed to purchases. The proxy says 2025 grants represented potential issuance of approximately 1.5% of beginning shares. September’s new plan permitted up to 2.2 million shares, roughly 6.9% of the October outstanding base, but authorization is not a forecast of immediate issuance. Realized vesting, withholding, future grants and net shares matter more than equating expense, capacity and issuance. [S1][S4]

Another $1 billion authorization left approximately $1.4 billion available in October. Authorization supplies neither cash nor an obligation to buy. Full use would require operating cash, lender funding and covenant flexibility. The strongest capital-allocation evidence would combine continued net retirement with stable absolute contribution and declining corporate debt, rather than maximizing authorization usage.

M&A outcomes and compulsory reinvestment

Acquired revenue is clearer than acquired returns. Acquisitions absorbed $459.1 million in 2025 and $316.5 million in the latest nine months. The current four-store rationale is local density, but public disclosures do not provide a complete purchase-price-to-cash-return bridge. Management’s underwriting claim that the investments exceed repurchase returns remains a claim, not observed ROIC. [S3][S1][S7]

Older transactions show why original capital remains relevant. CIG Financial cost $83 million plus repayment of $21 million of obligations in October 2022. RepairSmith cost approximately $190 million in January 2023. The finance operation has progressed from losses to current profitability; Mobile Service later suffered goodwill impairment. These are different outcomes and should not be combined into a generic success or failure label. A lifetime acquisition return includes subsequent capital, overhead and any exit proceeds. [S12]

Facility expenditure is partly required to preserve franchise standards. At the September conference, the CFO characterized compulsory annual facility spending around $250–$300 million. That is a management estimate rather than an audited maintenance-capex schedule. It argues against treating depreciation as distributable cash or assuming reinvestment can be cut without affecting OEM relationships and future productivity. [S9]

Governance, compensation and insider evidence

The annual management incentive uses adjusted operating income per basic share, with floorplan interest charged and an explicit capital charge on repurchases. The 2025 target was $28.17, attainment $34.58 and payout 163% of target. Long-term awards use relative TSR and defined ROIC. These safeguards make the plan more thoughtful than pure EPS compensation, but discretion over unusual items and book-capital definitions still warrants scrutiny. S4

CEO Michael Manley’s reported 2025 compensation was $34.39 million, including $28.06 million in grant-date stock-award value. A special award can earn up to 134,365 shares through 2029, requiring continued service and a 30-business-day average price meeting specified hurdles. The first hurdle is $313.51 and the maximum $374.21. These are compensation conditions, not valuation forecasts. A qualifying average can establish the performance condition before the final year, so alignment includes retention and price incentives rather than a requirement that every intermediate operating measure improve.

Directors receive a $50,000 annual retainer and ordinarily $250,000 of annual RSUs, with additional chair and committee retainers. The independent chairman’s additional retainer is $200,000. Incentives and behavior show a strong preference for per-share outcomes, tempered by capital charges and return measures. The investment consequence is to track absolute profit, debt and investment returns alongside per-share results. [S4]

The inspected CFO August Form 4 records RSU conversion of 3,194 shares and withholding of 1,257 for tax, codes M and F. Neither is an open-market purchase. Director Lisa Lutoff-Perlo’s August sale of 900 shares at $220 was code S. The September filing says no officer or director adopted or terminated a trading arrangement during Q3; that does not mean no pre-existing plan operated. An exhaustive five-year transaction census was not established, so no blanket absence-of-buying claim is made. [S23][S24][S1]

Cascade’s May amendment reported 7,062,629 shares, attributed the percentage increase to fewer issuer shares and reported no transactions in Item 5©. This rejects the old contemporaneous-exit narrative for that filing window. It does not establish the holder’s indefinite intentions or certify its position months later. Concentrated ownership, denominator changes and actual purchases are separate observations. [S25]

There is no recurring common dividend; dividend coverage is therefore not applicable and shareholder distributions depend principally on discretionary repurchases. That increases the importance of valuation, liquidity and funding discipline in the distribution policy. [S3][S9]

Verdict: Net retirement is real and October’s pricing improved. Acquisitions and distributions still consume flexibility, acquired returns are insufficiently disclosed, and compensation warrants a broader scorecard than adjusted EPS.

Changes and Headwinds — Last Two Years

Recent results reflect both external normalization and internal actions: affordability and supply pressured vehicle economics, while inventory, spending, finance growth, acquisitions and repurchases determined the shareholder outcome. Peer differences in service and used performance prevent attributing every weak result to the macro environment. Conversely, broad dealer pressure prevents treating the entire decline as uniquely poor execution. [S1][S19][S20][S21]

The largest external change is the disappearance of scarcity economics. Greater availability benefits customers but reduces bargaining power. Higher merchandise costs and competitive concessions can compress GPU even while selling prices and revenue remain substantial. The current 52 days of new supply and 43 days of used supply compare with 47 and 37 a year earlier. Used inventory also shifted toward more expensive vehicles because lower-priced supply was constrained. Higher inventory at the wrong price points is a different problem from insufficient total demand. [S1]

The CDK incident in June 2024 interrupted sales, service, inventory, customer management and accounting. The filing estimated an approximately 5% adverse gross-profit effect in Q2 2024 and disclosed roughly $43 million compensation support to commission employees. Subsequent insurance recoveries arrived in different periods. Economic disruption and reimbursement should therefore be separated in both earnings and cash comparisons; reimbursement does not make the original disruption costless. [S29]

Tariff announcements and the September 2025 EV-credit expiration changed transaction timing. NADA’s latest data show BEV sales down 29.2% year to date and hybrids up 22.4%. AutoNation management described much sharper Q3 BEV weakness within its own mix. These observations support a demand-composition explanation, but not an assumption that every earlier purchase returns as future catch-up volume. Manufacturer allocation and customer affordability remain consequential. [S33][S7]

Internal used-inventory decisions also hurt results. Management acknowledged retaining some vehicles too long and holding the wrong mix, leading to incremental wholesale losses. Inventory and organizational changes followed. Because the September conference already discussed correction and sourcing constraints, the October explanation represents follow-through on an identified issue. The next evidence should be shorter aging and better combined retail-plus-wholesale contribution, rather than a higher retail GPU achieved by deferring disposals. [S9][S7]

Important facility and management changes include four 2026 dealership acquisitions, two Domestic divestitures, a smaller stand-alone used footprint, a new technology leader and an additional independent director. Ravi Simhambhatla’s appointment took effect October 1; Mark Stewart joined the board October 13, increasing membership to ten. These changes add capability and oversight without establishing a financial benefit before execution. [S1][S26][S27]

Funding improved. The revolver increased to $2 billion and maturity extended to September 2031, with a conditional additional $1 billion accordion. The October warehouse consolidation created a $1.5 billion facility expiring in October 2027. Corporate maturity extension and shorter lender funding remain separate risks. The October 27 commercial-paper ceiling increase to $2 billion matched the revolver’s nominal commitment; the two facilities cannot be counted as independent unrestricted liquidity pools. [S10][S1]

Accounting policy did change: AutoNation early-adopted ASU 2025-06 on internal-use software on January 1, 2026, recording a $10.5 million after-tax reduction to opening retained earnings. The modified transition derecognized in-process software that no longer qualified for capitalization. Management said the adoption was not material to the consolidated statements. It does not explain Q3 dealer-margin deterioration, but omitting it would incorrectly suggest no relevant policy adoption occurred. Changes in CECL expectations remain estimation changes under an existing framework. [S1] (Note 1)

Conduct risk also has a company-specific example. California AutoNation dealerships agreed in February 2025 to a $650,000 settlement over alleged delayed ownership transfers. The resolution included $450,000 civil penalties, $150,000 investigative costs and $50,000 for consumer protection, plus operating controls. The dealerships resolved the case without admitting the allegations. The financial amount was small relative to group earnings; required title checks, staffing and commission deferral demonstrate that transaction execution has compliance costs beyond vehicle GPU. [S31]

The CARS withdrawal did not remove existing federal and state conduct obligations. The report establishes no new AutoNation-specific FTC liability from industry guidance. Equally, a search that finds no newly material group-level contingency is not proof that no litigation exists. The conclusion is bounded by disclosed filings and verified proceedings rather than an exhaustive court-docket census. [S17][S18][S3]

Verdict: Funding access improved and leadership capacity changed, while affordability, inventory execution and cost absorption worsened. The software adoption and title-transfer settlement are concrete corrections to an account that otherwise risks describing the period only through macro narratives.

Risk Analysis

Likelihoods are analyst judgments, not measured probabilities. The matrix follows current operating exposure and documented contractual or historical risks. Insurance and current funding reduce severity without eliminating interruption, deductibles or residual losses. [S1][S3][S29][S31]

The most plausible substantial stock decline combines weaker vehicle and F&I contribution with inadequate cost adjustment and continued distributions. On approximately half a million annual retail transactions, a $200 reduction in F&I per unit removes about $100 million of gross profit before offsets. A further $200 decline in new GPU on roughly a quarter-million units removes another $50 million. These are rounded analyst sensitivities; actual volume, mix and variable compensation determine the earnings effect. [S1][S3]

Facilities, management, systems and advertising do not all adjust as quickly as transaction gross profit. Expense dollars can be stable while conversion deteriorates sharply. Aggressive cuts can also impair service capacity or customer acquisition, so a lower expense ratio is insufficient if it sacrifices the future earnings floor. The useful downside bridge models both delayed cost removal and the contribution lost through indiscriminate cuts.

Credit supplies a delayed second channel. One additional percentage point of annual losses on a $3 billion portfolio represents approximately $30 million of annual credit cost before balance and pricing changes. That is material relative to annualized Q3 finance income of roughly $62 million. CECL timing, recoveries and funding alter the accounting path. The sensitivity nevertheless shows that a small profitable lender can lose its contribution without a dramatic headline portfolio default event. [S1]

Nonrecourse status limits contractual parent liability, but the parent can lose residual equity, receive less securitization cash or choose strategic support. Funding can remain legally available while becoming economically unattractive through lower advances or higher spreads. A sound stress analysis preserves legal isolation while recognizing residual capital and customer-franchise effects. It should not assume unlimited parent recourse or complete economic insulation.

A catastrophic investment loss would require correlated deterioration in dealer cash, consumer credit, funding, asset values and capital-allocation discipline. A recession could depress vehicle contribution and repair content while increasing losses and reducing collateral recoveries. Falling EBITDA raises leverage even before additional borrowing. Continued repurchases would then consume flexibility precisely when franchise or property impairments and refinancing pressures increase. [S1][S3]

A literal total loss appears remote on current evidence, but a plausible path is sustained operating losses, lender-equity impairment, restricted refinancing, covenant acceleration and forced dilution or restructuring. Property, franchise access, positive matched cash and present covenant headroom are meaningful defenses. They are not guaranteed liquidation proceeds to common shareholders after secured claims, landlords, employees, taxes and costs. No numerical total-loss probability is justified. [S1]

There are risks to excessive pessimism as well. Corrected inventory can improve cash and contribution quickly; service traffic can support the existing facility base; and modest lender growth can spread direct expenses. Retirement at lower prices can improve ownership economics without a broad industry boom. The bear therefore needs deterioration across multiple profit pools or capital defenses, rather than extrapolating one weak GPU observation into insolvency.

Verdict: The principal downside is a combined earnings-and-allocation feedback loop. Current liquidity and credit are defenses, while retained lending introduces delayed losses and financing conditions that can amplify a conventional dealer downturn.

Valuation Discussion

Current earnings and the capital structure

At $179.64 and 31,983,552 shares outstanding on October 28, equity value is approximately $5.75 billion. The quotation and share count are separately dated observations. Estimated trailing adjusted EPS of $20.41 implies 8.8 times earnings; discrete-period GAAP EPS totals approximately $20.80 but includes investment and adjustment effects. Neither number is forward guidance. [S6][S1][S2][S15]

Share retirement creates a denominator trap. Trailing adjusted net income of $708.7 million divided by the October outstanding base equals about $22.16 per share, above historical trailing EPS because earlier quarters used more shares. That is not newly generated earnings or a fully diluted forecast. A forward model must choose operating income and an average diluted denominator explicitly. Applying a current share count to old income and then adding another full repurchase benefit double-counts retirement.

Adding approximately $4.344 billion of non-vehicle debt and commercial paper, then subtracting $51 million unrestricted cash, produces a hybrid enterprise claim near $10.04 billion. It excludes floorplan and lender funding, with their costs retained in corresponding earnings. Adding those funding categories yields approximately $16.64 billion, but that larger claim requires a differently adjusted earnings denominator. Floorplan and nonrecourse debt cannot simply be included or excluded for whichever result appears cheapest. [S1]

A trailing adjusted operating-income bridge is approximately $1.313 billion. Adding roughly $252 million depreciation and subtracting about $180 million floorplan interest gives approximately $1.384 billion, retaining finance contribution after its funding expense. The hybrid claim is therefore around 7.3 times this hybrid earnings measure. It is not a pure dealer EBITDA or bank enterprise-value multiple. Operating-lease capitalization would require a matched rent adjustment. [S1][S3][S15][S2]

Cash normalization and peers

Trailing management-adjusted FCF is approximately $998.5 million: $1,045.2 million plus $739.0 million less $785.7 million. Removing roughly $40 million remaining cyber-insurance proceeds gives $958.5 million, approximately 16.7% of equity value. This is an analyst calculation using the company’s definition. It remains above sustainable freely distributable cash if lender equity, tax timing or necessary reinvestment are inadequately recognized. [S15][S2]

The comparison with nine-month 2025 is less weak after insurance normalization: $785.7 million less $40 million equals $745.7 million, close to current $739.0 million. That corrects an exaggerated cash-decline narrative without establishing growth. The separate $677.7 million financing-matched bridge also remains dependent on period debt flows. It should not be annualized or capitalized as if financing availability and tax timing repeat indefinitely.

This deliberately limited comparison uses Company Financials prices and issuer EPS. It is not forward consensus valuation: quarterly seasonality, acquisitions, continuing-operations definitions, share changes and adjustments differ. It shows that AN’s lower valuation does not establish the largest peer discount. A premium requires support from durable composition, capital discipline or eventual returns, rather than the label of a U.S.-only dealer. [S6][S19][S20][S21]

Own-history context also needs matched earnings. Company Financials returned approximately 11.9 times GAAP earnings at December 2025 and 10.4 times in March 2026; its June fields were incomplete. The current adjusted multiple is directionally lower, but no exact historical percentile is asserted. In particular, a low multiple of peak 2022 earnings is not equivalent to the same multiple of normalized income. [S32]

Explicit 2027 scenarios

The following are analyst assumptions. Dealer gross profit excludes separately reported finance income. SG&A is an adjusted ratio; depreciation, floorplan and corporate interest are charged afterward. Share counts include expected net employee issuance. Dollars are millions unless stated otherwise.

The scenarios are anchored to reported departmental contribution, expenses and funding, but their future combinations are assumptions. The finance-equity allowances are sensitivities because allocated permanent capital is not disclosed. These values are scenario outputs, not additional recommendations or price targets. [S1][S2][S3]

The bear combines new GPU near or below $2,100, F&I around $2,450–$2,550, little organic service contribution and inadequate cost flexibility. Lending remains profitable but cannot offset weaker dealerships. Repurchases slow to preserve resources, preventing the share count from supplying the recent pace of EPS support. This requires several weak outcomes together, but each has a plausible operating mechanism.

The base assumes new GPU around $2,150–$2,300, F&I $2,650–$2,750, modest organic service growth and improved inventory execution. Gross profit stays slightly below the recent trailing level. Importantly, a 69% expense ratio can describe a valuation stabilization scenario while still failing the prior 68.5% operational falsifier. A fair-value scenario and a successful original thesis are different tests; the old threshold is not quietly reset.

The bull requires new GPU around $2,400–$2,600, better used sourcing, mid-single-digit service growth and an expense ratio approaching 67%. Finance expands without unfavorable seasoning or substantially higher capital requirements. The assumptions can reinforce one another through operating leverage, but must not be counted twice as both full incremental GP and independent percentage-based savings.

At $4.8 billion gross profit, one percentage point of SG&A absorption equals approximately $48 million pretax or $1.14 per share after assumed tax and 31.5 million shares. Financing and volume effects can therefore overwhelm a small difference in terminal revenue growth. Likewise, a $100 million repurchase at $180 retires about 0.56 million shares before issuance, but any resulting EPS benefit must be charged for financing cost and lost flexibility.

Terminal economics and embedded expectations

A steady-state equity-cash sensitivity supplies another lens. Annual cash of $700 million growing 1%, discounted at 12%, implies approximately $6.43 billion, or about $201 per current outstanding share. With $600–$800 million starting cash, 11%–13% required returns and 0%–2% growth, the range becomes very wide. The example assumes cash after necessary reinvestment and debt service; current adjusted FCF is not automatically an eligible input.

Terminal growth requires retained capital for facilities, inventory, systems and lending. The lender’s equity requirement rises if advance terms worsen or losses increase. Existing service capacity can permit some growth without proportionate property spending, but that does not justify perpetual zero-reinvestment growth. A DCF that simultaneously distributes all adjusted cash, expands loans and reduces debt would violate its own financing logic.

The market appears to recognize substantial service and F&I continuity while discounting weak dealer contribution. The favorable possible error is underestimating modest stabilization combined with cheaper retirement. The unfavorable possible error is capitalizing adjusted cash as unrestricted and assuming annualized finance returns survive seasoning. No current consensus census permits presenting that interpretation as a measured universal market belief.

Verdict: The earnings burden has eased, but peer cheapness and a trough are unproven. Gross-profit conversion, lender capital and the funding of retirement determine the valuation range more than modest changes in headline revenue.

Variant Perception

The actual debate and competing cases

Recent investor questions test service pricing and repair content, the new-GPU floor, retained-loan returns, used competition, pricing enforcement and SG&A accountability. The October call and September conference directly document those concerns. They are stronger evidence of the debate than an invented uniform sell-side consensus. A comprehensive current estimate and recommendation census was not established. [S7][S9]

Service questions elicited management’s acknowledgment of higher traffic alongside fewer accepted labor hours. Margin questions challenged the earlier stability framing. Other investors asked whether Carvana’s growth was affecting demand and when technology would change economics. These exchanges focus on whether customer-lifetime management improves current and future contribution, rather than merely supplying a reason to tolerate weaker present margins.

The strongest bull case is that the earnings floor is better than vehicle retail alone suggests. Service and F&I supply approximately 80% of dealer GP; current credit remains favorable; facilities and local density support retention; and October purchases retired stock much more cheaply. Modest stabilization can therefore produce acceptable per-share growth without restoring pandemic margins or accelerating store construction. [S1][S2]

The strongest bear case is that lifetime-value language delays recognition of deteriorating transaction economics. Comparable GP fell 4.7%, F&I dollars weakened, SG&A absorption rose and part of finance’s improvement came from lower provision. Required facilities, portfolio capital and acquisitions reduce the cash available for distributions. A low multiple is not compelling if normalized income and balance-sheet flexibility continue shrinking.

Four assumptions carry most of the disagreement: external service traffic converts into contribution; vehicle and F&I margins stop eroding together; retained loans earn superior discounted returns after full capital costs; and repurchases remain subordinate to liquidity. Each has current supporting and disconfirming evidence. Their simultaneous success is more demanding than any one favorable quarterly statistic.

The factor model and positioning

The factor model dated October 30 reports Market 0.968, SmallSize 0.814, Growth -0.203, NewValue 0.117 and Quality -0.010. Its larger sector coefficients include Consumer Discretionary +0.314, Energy -0.277 and Materials +0.243. OilPrice is +0.100, GoldPrice -0.104, Utilities -0.117, Financials -0.041 and the dividend-related exposure +0.110. These are internal statistical diagnostics, not fundamental evidence, industry classifications or causal sensitivities. The dividend coefficient does not imply a cash dividend. [S28]

R-squared is 0.392 and adjusted R-squared 0.381. Approximately 60.8% of sample return variation is unexplained by the included factors, including omitted systematic effects and model error as well as company-specific information. Residual momentum is -0.0705, residual Sharpe -0.714 and residual volatility 0.2503 in unspecified units. Alpha is -0.000143 with an unspecified frequency. Without the estimation window, uncertainty estimates and unit conventions, these values support limited descriptive interpretation.

Near-zero Quality exposure does not establish a low-quality business, and weak residual performance does not establish persistent future alpha. The diagnostics help characterize historical return behavior while filings determine business quality. No verified short-interest, borrow or fund-position census supports a stronger crowding claim. Relative peer price moves likewise do not establish customer market-share changes.

Prior hypotheses and reusable analytical lessons

The directly relevant reserve hypothesis survives: a declining allowance ratio in a growing captive does not establish a dollar release. The current qualification is equally important: increased allowance dollars do not exclude favorable modeled-loss effects on provision. The June and September filings establish both the reserve-stock increase and the expected-loss-rate benefit. This is a narrowly scoped accounting conclusion, not proof of conservative or aggressive lifetime underwriting. [S5][S1]

The pharmaceutical and software research-capital observations do not transfer mechanically. AutoNation’s main denominator questions concern inventory funding, retained loans, leases and acquisition capital. Historical research-life assumptions from unrelated businesses would add precision without a matched economic mechanism. Similarly, banking securities marks in OCI do not establish an AutoNation capital problem.

Several older analytical shortcuts should be discarded. Share-price outperformance is not proof of operating market-share gains. A near-zero factor quality loading is not a business-quality verdict. Reaching a prior price condition does not establish a successful investment. Those propositions require customer, contribution, cash and capital evidence respectively.

The supported new hypotheses are narrower: match retained-loan investment to actual funding before discussing corporate cash availability, and compare retained lending with immediate fees using discounted fully allocated cohort economics. Neither establishes an industry-wide return advantage. Their value is to identify the information needed to accept or reject a specific application. [S1][S2][S9]

Verdict: The useful variant separates improving current finance performance from contracting dealer contribution and distinguishes cash generation from cash available for distribution. Neither a broken-franchise narrative nor an already-proven recovery fits all the evidence.

Fact vs. Interpretation

Management statements can be plausible and financially important without becoming independently verified outcomes. Deferred service returning, superior lifetime loans, acquisition synergies and technology productivity belong in questions and scenarios until supported by actual customer and capital evidence. Reported non-GAAP facts also remain subject to judgment about which exclusions describe recurring economics.

Verdict: The largest mistakes are interpretive: confusing denominator growth with enterprise growth, coverage with reserve dollars, financing-dependent cash with owner cash, and strategic opportunities with demonstrated returns.

Open Questions

The most valuable missing disclosures concern allocated capital and contribution rather than more consolidated revenue detail. Current loan schedules, departmental bridges and funding notes substantially improve analysis, but do not establish lifetime customer economics. [S1][S7]

1. What equity and restricted liquidity support finance after shared overhead, original acquisition capital and customer-acquisition costs?
2. What are cumulative losses and discounted returns for 2024 and 2025 vintages at matched ages and borrower risk?
3. How much of the expected-loss improvement comes from mix, model assumptions, recoveries and macro forecasts?
4. How does management’s $93 million finance cash claim reconcile to portfolio investment, securitizations and group adjusted cash?
5. What fraction of deferred service returns to AutoNation, after how long and with what incremental contribution?
6. How do labor hours, effective rates, parts mix and productive technician capacity reconcile the different repair-content observations?
7. What cash have the newest acquired stores produced after purchase consideration, facilities, inventory and integration?
8. When does the cash-tax timing benefit reverse, and what is the ordinary cash-tax requirement?
9. What measurable sourcing, conversion or cost benefit emerges from technology after implementation expense and software capitalization?
10. What would trigger slower repurchases if leverage rises while dealer income continues contracting?

The evidence base includes all five 2021–2025 annual filings and available 2022–2026 quarterly filings, with targeted financial and operating passages, plus current and historical proxy materials and selected material-event and ownership filings. An exhaustive 60-month census of every material 8-K and relevant Form 3, 4 or 5 was not established. Historical proxy extraction was also uneven. Accordingly, no absence-of-litigation, absence-of-buying or complete contractual-history conclusion is inferred from this review.

Those limits particularly affect broad governance and insider claims. They do not overturn directly reconciled current financial values. A lack of disclosed cohort or acquisition returns is also not evidence that management lacks private information; it limits what outside investors can responsibly capitalize. The report’s scenarios therefore retain uncertainty instead of assigning undisclosed synergies or private-asset gains to the central earnings bridge. [S3][S4][S23][S24]

Verdict: The decisive information gaps are finance equity and cohort returns, acquired-store cash, service conversion and tax normalization. More optimistic strategic language would not resolve them.

What Must Be True

Bull tests

1. Dealer profit must stabilize under the original registered test. Monitor new GPU, retail units and combined Domestic, Import and Premium Luxury income. The prior constructive condition remains GPU above $2,300 with dealer income avoiding double-digit contraction. Its falsifier remains two consecutive quarters with dealer income down more than 10% and GPU below $2,250. Q3’s $2,176 GPU fails the constructive condition, but its 9.4% dealer-income decline does not complete the joint falsifier. Do not replace these thresholds with a lower floor after the result. [S1][S5]

2. Service must demonstrate organic economics. Retain the prior constructive condition of at least 4% comparable GP growth and margin near or above 48%. Q3’s 0.8% and 47.5% comparable margin do not pass. The falsifier remains flat or negative comparable GP in three of four quarters, particularly with weaker customer-pay activity. Monitor repair orders, hours, rates, technician utilization, external work and mix; one distribution-driven margin decline alone cannot establish failure. [S1][S7]

3. F&I must hold dollars as well as PVR. The constructive condition remains PVR above $2,650 and stable aggregate GP. Q3 passes the former and fails the latter. The falsifier remains PVR below $2,500 while retail units decline, unless independently demonstrated discounted finance contribution supplies the offset. Monitor cancellations, attachment, financing fees and retained-loan penetration with explicit denominators. [S1][S2]

4. Finance must earn through seasoning. Retain losses below 3% and delinquency below 3.5% as near-term constructive conditions; current 1.8% and 2.5% pass. The prior severe falsifier remains losses above 4%, impaired funding, inadequately explained credit costs or required parent capital that eliminates economic returns. Earlier deterioration can justify reassessment before the severe threshold. Monitor vintage losses, recoveries, model changes, restricted cash and fully allocated equity, rather than the reserve ratio alone. [S1][S5]

5. Cost initiatives must pay back. The dated 66%–67% year-end run-rate objective remains a separate management test. Preserve the prior falsifier of SG&A at or above 68.5% through mid-2027 without measurable conversion, sourcing, retention or contribution improvement. Q3’s 69.6% is adverse evidence, not a completed future-period test. Expense dollars and service capacity must accompany the percentage. [S2][S8]

6. Repurchases must preserve financial flexibility. Monitor actual cash payments, acquisition spending, loan funding, corporate debt, net shares, leverage and coverage. Continued net retirement with leverage below three times and stable absolute earnings would strengthen the case. The prior concern intensifies above 3.25 times with ongoing substantial purchases and contracting contribution. October’s lower price improves expected returns but does not suspend the liquidity condition. [S1][S10]

Bear tests

1. The bear requires weakness across multiple contribution pools. It weakens materially if two successive quarters show growing absolute adjusted income and comparable GP, supported by service and F&I dollars rather than merely a smaller denominator. The original stronger counter-test—new GPU near $2,400, F&I above $2,750 and service growth at least 4%—remains unmet. Monitor the complete dealer-plus-finance bridge. [S1][S2]

2. Credit skepticism requires actual seasoned loss and capital evidence. Rapid growth alone is not a credit event. The strongest bear is falsified if mature cohorts deliver adequately priced cumulative losses, positive fully allocated equity returns and uninterrupted funding through weaker conditions. Aggregate loss ratios can be diluted by new originations; matched-age cohorts and parent support are the monitoring signals. [S1]

3. The buyback critique requires value and funding evidence. A record nominal purchase price is insufficient. It weakens if normalized earnings remain above $20 per share, cash per remaining share grows after necessary capital and corporate debt declines. October purchases must be scored separately from earlier vintages. Net retirement without adequate investment returns is incomplete; adequate returns without a funding bridge are also incomplete. [S1][S3]

4. The service-moat critique must survive actual operating improvement. It weakens if customer retention, repair content, productive capacity and comparable contribution improve despite the vehicle-cohort gap. A larger service share caused only by falling vehicle GP does not qualify. Monitor external contribution and capital productivity, not appointment counts in isolation. [S7][S1]

5. A solvency argument needs covenant and liquidity deterioration. Negative tangible equity alone is insufficient. Sustained positive cash after matched lending investment, leverage below three times, coverage comfortably above four times and accessible liquidity contradict an imminent-distress thesis. Simultaneous funding restrictions, lender-equity losses, dealer contraction and continued distributions would change that assessment. [S1][S10]

Q4’s inventory correction, the old year-end expense objective and final tax payments are immediate catalysts. During 2027, service-cohort performance, technology deployment and mature loan data become more informative. Each catalyst can deliver unfavorable as well as favorable evidence; an event’s occurrence is not itself a positive outcome.

Verdict: Current evidence passes near-term credit and net-retirement tests, fails the old constructive vehicle and service conditions, and leaves the longer combined falsifiers open. The decisive next evidence is stable absolute contribution with adequate cash and returns on the capital supporting it. Core monitoring sources: latest quarterly filing, quarterly results, annual filing, and latest management discussion.

Public source appendix

• S1: AutoNation Q3 2026 Form 10-Q — primary_quarterly_filing; published 2026-10-30; Cover; statements pp.1–4; Note 1 software adoption; Note 6 principal, vintages and allowance pp.11–13; Notes 9–16; MD&A pp.32–63; Part II trading arrangements.
• S2: AutoNation Q3 2026 issuer earnings release — issuer_earnings_release; published 2026-10-30; GAAP and adjusted results; operating and same-store tables; segment income; SG&A and FCF reconciliations. Cash-flow narrative has incorrect negative wording; reconciliation shows positive $160.4m.
• S3: AutoNation 2025 Form 10-K — primary_annual_filing; published 2026-02-12; Business, Competition, geographic concentration, risks; statements pp.61–65; accounting, debt, impairment and finance notes; segment schedules pp.105–109.
• S4: AutoNation 2026 proxy statement — primary_proxy; published 2026-03-17; Board compensation p.10; annual incentives pp.19–20; long-term ROIC and one-time award pp.21–24; Summary Compensation Table p.29; award dilution.
• S5: AutoNation Q2 2026 Form 10-Q — primary_quarterly_filing; published 2026-07-31; Financial statements, allowance rollforward and expected-loss discussion pp.11–13, operating tables, liquidity, capital allocation.
• S6: Company Financials daily prices with public price-history cross-check — market_data; published 2026-10-30; NYSE:AN split-adjusted 1,255 daily observations 2021-11-01–2026-10-30; NYSE:LAD, NYSE:ABG and NYSE:GPI matched closes 2026-09-10–2026-10-30. Retrieved October 31; public page corroborates current AN close and recent daily prices.
• S7: Company Financials Q3 2026 management transcript, publicly cross-checked — management_transcript; published 2026-10-30; Prepared remarks slides 3–14 and analyst Q&A: service traffic/content, finance cash and ROE, SG&A, used correction, cancellations, competition and technology.
• S8: Company Financials Q2 2026 management transcript, publicly cross-checked — management_transcript; published 2026-07-31; Prepared remarks and Q&A on 66%–67% year-end run rate, sequential GPU, service, lending penetration, portfolio and acquisitions.
• S9: AutoNation September 2026 Laguna Conference management transcript — management_event_transcript; published 2026-09-17; CFO/IR discussion: used sourcing, CFS moderation and lifetime lending profit; facility capex; acquisition and repurchase framework; approximately 10% sequential new-GPU warning.
• S10: AutoNation September 2026 credit-agreement Form 8-K — primary_material_event_filing; published 2026-09-15; Item 8.01 and Exhibit 10.1: September 14 amendment, $2bn revolver, conditional accordion, 2031 maturity and covenants.
• S11: AutoNation 2024 Form 10-K — primary_annual_filing; published 2025-02-14; 2022–2024 audited statements and operating tables; 2022 GAAP operating income $2,024.5m; new GPU history.
• S12: AutoNation 2022 Form 10-K — primary_annual_filing; published 2023-02-17; 2020–2022 audited statements; 2021–2022 cash, shares and repurchases; CIG consideration and RepairSmith subsequent acquisition; Note 6 allowance.
• S13: AutoNation 2023 Form 10-K — primary_annual_filing; published 2024-02-16; 2021–2023 income and cash-flow statements, balance sheets, service GP, repurchases and segment note.
• S14: AutoNation 2021 Form 10-K — primary_annual_filing; published 2022-02-17; Audited statements; $1,627.7m operating cash, $215.7m physical capex, SBC, repurchases and share count.
• S15: AutoNation full-year 2025 results, SEC exhibit — primary_earnings_exhibit; published 2026-02-06; Adjusted income $770.3m, EPS $20.22, operating income $1,386.9m; FCF $1,045.2m; insurance-recovery conversion footnote.
• S16: NADA midyear 2026 industry data — industry_association; publication date unavailable; Midyear report: 16,991 dealers, 7.9m vehicles, $640bn+ sales and 136m+ repair orders.
• S17: FTC automobile-industry pricing-transparency FAQs — regulator_guidance; publication date unavailable; Staff-view and nonbinding status; required charges, misleading advertisements, optional products and all-in pricing.
• S18: Federal Register action withdrawing the CARS Rule — regulatory_primary_document; published 2026-02-12; Summary, effective date and section I.B: withdrawal effective February 12, 2026 following court vacatur.
• S19: Lithia Q3 2026 issuer results — peer_primary_release; published 2026-10-21; Adjusted EPS, same-store new GPU and service growth, SG&A tables and finance operations.
• S20: Asbury Q3 2026 results, SEC exhibit — peer_primary_release; published 2026-10-27; EPS $6.16, adjusted SG&A 67.4%, same-store new GPU $2,901, service GP $352.6m/$361.0m; Tekion and TCA.
• S21: Group 1 Q3 2026 results, SEC exhibit — peer_primary_release; published 2026-10-28; Continuing adjusted EPS $9.36, same-store new GPU $3,088, consolidated service GP +0.1%, adjusted SG&A 71.4%; geography.
• S22: Carvana Q3 2026 results, SEC exhibit — competitor_primary_release; published 2026-10-28; 217,385 retail units, 39% growth, $8.04bn revenue and reported operating results.
• S23: Thomas Szlosek August 2026 Form 4 — primary_ownership_filing; published 2026-08-10; August 7 Table I codes M and F; RSU conversion 3,194, tax withholding 1,257; footnotes.
• S24: Lisa Lutoff-Perlo August 2026 Form 4 — primary_ownership_filing; published 2026-08-07; August 5 code-S sale of 900 shares at $220; Table I.
• S25: Cascade and William Gates May 2026 Schedule 13D amendment — primary_beneficial_ownership_filing; published 2026-05-05; 7,062,629 shares; Item 1 denominator-change explanation; Item 5© no transactions.
• S26: AutoNation appoints Ravi Simhambhatla — issuer_management_announcement; published 2026-10-01; Appointment effective October 1; technology, data science and AI responsibilities.
• S27: AutoNation director appointment Form 8-K — primary_governance_filing; published 2026-10-16; Item 5.02: Mark Stewart joined October 13, ten-member board, independence, retainer and equity eligibility.
• S28: the factor model — supplied_internal_diagnostic; published 2026-10-30; Supplied dated exposures, residual statistics and diagnostics; 13 factors; R-squared 0.39208. Window, frequency and full unit conventions unavailable.
• S29: AutoNation Q2 2024 Form 10-Q — primary_quarterly_filing; published 2024-07-31; MD&A CDK incident, disrupted core functions, estimated GP effect and approximately $43m employee compensation support.
• S30: CarMax Q2 fiscal 2027 results, SEC exhibit — competitor_primary_release; published 2026-09-29; Quarter ended August 31, 2026; retail unit growth 13.8%, comparable 13.0%; pricing actions and finance.
• S31: Los Angeles County DA: AutoNation title-transfer settlement — enforcement_authority; published 2025-02-26; $650,000 settlement, allocation, title-transfer controls and no admission of allegations; linked judgment.
• S32: Company Financials statements, ratios and valuation; issuer reconciliation — configured_financial_data; publication date unavailable; Retrieved October 31 after resolving NYSE:AN: profile, 2021–2025 statements and profitability, enterprise value and valuation. Standardized operating income differs from GAAP; June valuation/debt fields incomplete. Linked issuer page supplies public reconciliation rather than the underlying data-service response.
• S33: NADA September 2026 Market Beat — industry_association; published 2026-10-05; September and YTD SAAR, EV/hybrid comparison and explicitly estimated JD Power payment data.
• S34: AutoNation Q1 2022 Form 10-Q — primary_quarterly_filing; published 2022-04-21; MD&A inventory days supply: eight days new, 30 days used at March 31, 2022; statements and operating tables.