Built by Grok 4.6 High

Reports Thursday, October 22, 2026. Prepared Thursday, September 24, 2026. This is the Q3 2026 print (quarter ending September 30). It has not been reported. The date is the vendor calendar, not an AutoNation scheduling release. Time of day is not company-confirmed. The historical pattern is a morning call. Last reported quarter is Q2 2026, released the morning of July 31, 2026. AutoNation is a calendar-year company. The homepage calendar labels this date 2026Q4. That label is the announcement quarter. This page is 2026Q3.

AN | Earnings Preview

AutoNation reports Q3 on Thursday, October 22, if the estimated date holds. There is still no company release locking the hour. On July 31 the company printed revenue of $6,929.8 million, down 0.6% versus $6,974.4 million, and adjusted diluted EPS of $5.56, up 1.8% versus $5.46. Adjusted operating income fell 7.1%, to $343.1 million from $369.3 million. The EPS line rose because the share count fell. Diluted weighted-average shares were 33.8 million, versus 38.3 million a year earlier.

Growth trajectory. Annual revenue has been stuck. It went from $25,844 million in 2021 to $26,985 million in 2022, then $26,948.9 million, $26,765.4 million, and $27,631.4 million. That is +4.4%, −0.1%, −0.7%, and +3.2%. Five years, one real up year. Adjusted EPS peaked at $24.57 in 2022, fell to $17.46 in 2024, and only partly recovered to $20.22 in 2025, +15.8% and still 18% under the peak. Quarterly, the top line is clawing back toward zero and is not yet growing: +6.9% in Q3 2025, −3.9% in Q4 2025, −2.1% in Q1 2026, −0.6% in Q2 2026. Adjusted EPS growth did the opposite. It collapsed from +24.6% in Q3 2025 ($5.01 versus $4.02) to +2.2%, +0.2%, and +1.8%, and has sat there. The street's Q3 bar of +13.4% is a break from that path, not a continuation of it.

The profit engine is decelerating, which is why the EPS streak is not the setup. Same-store parts-and-service gross profit went from +13.0% in Q2 2025 to +7.0%, +3.9%, +2.6%, and +0.1%. That is a 1,290 bp slowdown in four quarters, to flat. Total parts-and-service gross profit in Q2 was $607.1 million, only +1.4% versus $598.6 million, while revenue grew 3.4% because wholesale mix is lower margin. The margin fell about 95 bps, to about 48.1%. New-vehicle GPU is still down hard year over year, $2,381 versus $2,785, −14.5%. Used units are getting worse, not better: same-store −8.0%. Finance income and free cash flow are the lines still moving the right way. Quarterly AutoNation Finance income was $10.7 million versus $2.0 million. Quarterly adjusted free cash flow was $183.6 million, +17.1% versus $156.8 million. The finance line is real, and it is about $11 million against a roughly $190 million earnings pool.

What consensus is underwriting. FMP for the unreported October 22 row is adjusted EPS $5.68 and revenue $7,086.7 million. Versus Q3 2025 that is +13.4% on EPS against $5.01 and +0.7% on revenue against $7,037.4 million. At the Q2 share count of 33.8 million, $5.68 is about $192 million of earnings, against adjusted net income of $190.9 million in Q3 2025. The entire EPS step-up is the buyback, not a higher earnings pool. The last two quarters of adjusted net income were down about 10% year over year. Flat net income would already be a 10-point improvement from that run-rate. GAAP diluted EPS in the year-ago quarter was $5.65. Do not put the street number against GAAP. FMP scores the adjusted line. Internal models, Visible Alpha, and Bloomberg were not available.

What September changed. July's language was second-half adjusted EPS growth, customer pay still mid-single-digit, unit profit stable, and SG&A back to a 66–67% run-rate by year-end. On September 17, at Morgan Stanley's Laguna Conference, that operating bridge was qualified. Q3 service growth was called more modest, with customers more selective on large maintenance. New-vehicle GPU was guided about 10% lower sequentially on model-year changeover and more price given away. CFS unit profit was described as down $50 to $100 sequentially after weaker July and August attachment. Ten percent below Q2's $2,381 is about $2,140, roughly −6% versus Q3 2025's $2,281. That is a milder year-over-year GPU hole than Q2's −14.5%, and it is still a break from "stable." The harder comp is units. Q3 2025 same-store new units were +4.5%, the EV-credit rush. The easy volume comp is Q4, not Q3. Published EPS estimates checked after the conference were still about $5.67 to $5.73. The street has not followed the walk-down.

What to watch on October 22

• Same-store parts-and-service gross profit versus +0.1% and versus the +7.0% being lapped. Customer pay can still be up while the consolidated gross-profit dollar is flat. The consolidated line is what hits earnings. Aftersales was about half of Q2 gross profit.
• New GPU versus both the Q2 band and $2,281. A print near $2,140 confirms Laguna. A print that only dips low-single-digits sequentially means the conference comment was caution. New units versus 66,189, split import versus domestic, are the volume read. Import was the only brand group up in Q2.
• F&I per vehicle versus $2,775. Growth has already cooled from +7.6% to +3.2%. A $50 to $100 sequential step-down turns that into roughly flat to down about 3% year over year, the first decline in this window.
• The SG&A ratio, not the EPS streak. Reported SG&A was $856.3 million on gross profit of $1,231.1 million, 69.6%, versus 67.0% a year earlier. Management's adjusted ratio was 68.2%. A lower GPU and only modest service shrink the denominator. Year-end 66–67% is a Q4 hope unless advertising actually rolls off 6.1% of gross profit.
• Whether $5.68 was earned with flat adjusted net income. A seventh quarter of adjusted-EPS growth only has to clear $5.01. That is the uninformative hurdle. The model hurdle is the earnings pool. Read same-store revenue, not the headline. Four acquired stores, described as about $600 million of annual revenue, are in this print for a full quarter.

CarMax on Tuesday, September 29, before the open, is the only auto retail print before AutoNation. It is a fiscal quarter ended August 31, so it speaks to used cars and cost, not to franchise new-vehicle GPU. Sonic is estimated the same morning. Asbury, Group 1, Lithia, Penske, and Carvana are estimated for the following week. AutoNation sets the franchise tone. It does not receive one.

AutoNation does not guide quarterly or full-year revenue, EPS, gross margin, or operating margin. The 2026 outlook slide came off on the May 1 call and was not replaced with numbers on July 31. What is still standing is directional, plus two quantified items that are not a Q3 P&L range: an adjusted SG&A-to-gross-profit run-rate of 66% to 67% by year-end, and full-year CapEx of about $325 million. Customer pay was "continued mid-single-digit growth." Units were "in line with the markets we serve." New and used GPU were "a very narrow bandwidth." Second-half adjusted EPS was expected to grow, with no rate and no dollar.

Classification: conservative to the point of being uninformative on EPS, aggressive on the SG&A exit and on holding aftersales at mid-single digits, and the operative Q3 commentary is a guide-down. The company has cleared "EPS growth" for six straight quarters while the year-over-year gain collapsed from +24.6% to +0.2% and +1.8%. A print above $5.01 lets management claim a seventh quarter and still miss $5.68 by a wide margin. That gap is the setup. Same-store aftersales gross profit has missed a 4–6% band for three straight quarters and is now +0.1%. Adjusted SG&A has been above 67% in seven of the last eight quarters.

The street, not the company, is asking for a reacceleration. Q3 revenue of $7,086.7 million is +0.7%, the first positive revenue compare since Q3 2025, and it only extends a less-negative trend. Q4 street revenue of $7,208.2 million is +4.0% versus $6,929.2 million. The revenue reacceleration in the model is a Q4 story. Q4 street EPS of $5.91 is +16.4% versus $5.08. First-half adjusted EPS is $4.69 plus $5.56, or $10.25, only +1.1% versus $4.68 plus $5.46. FY2026 street EPS of $21.75 is +7.6% versus $20.22. Almost all of that is a second-half assumption of about $11.50, +14% versus last year's second half. Management has not endorsed any of those numbers. Quarterly and annual analyst counts do not foot. H1 plus the Q3 and Q4 averages is about $21.84, versus the FY average of $21.75.

FY2026 street revenue of $27,791 million is +0.6% versus $27,631.4 million. The year stays flat. CapEx of about $325 million is +5% versus FY2025 $309.4 million. First-half cash CapEx was $56.4 million plus $69.6 million, or $126 million, 39% of the bogey. The residual is a second-half spending step-up, which is the plan, not a demand signal. FY2027 is not guided. FMP's $28,668 million and $24.93 are consensus only, about +3.2% and +14.6% versus the FY2026 averages. $24.93 would roughly retake the 2022 adjusted-EPS peak of $24.57. That is a street destination, not a company framework.

Q3 2026 has not been reported. Guide columns are July 31 language unless September replaced them. Consensus is FMP. FMP does not publish new, used, parts, F&I, GPU, or SG&A estimates. Internal is n/a. There is no company range to difference against the street.

Current quarter — Q3 2026

Next quarter — Q4 2026 guide, expected

Nothing on July 31 was a Q4 revenue or EPS range. The year-end SG&A run-rate is an exit rate, so it belongs here. The 30% to 40% off-lease comment is second-half supply versus the first half, not a Q4 year-over-year guide. Q4 is the easier new-vehicle lap. Q4 2025 new-vehicle revenue was down 8.8%, $3,441.3 million versus $3,775.2 million, and same-store new units were −10.2%. Management did not endorse the street's +4% revenue or +16% EPS.

FY2027 — no old guide

No FY2027 revenue, EPS, margin, CapEx, or GPU framework was given on July 31 or on the two calls before it. The only 2027 remark was qualitative: incremental opportunity with outside finance providers through 2027, with no dollar. The 66–67% SG&A band was restated as the ongoing target, not as an FY2027 range.

Eight reported quarters — trajectory, not the open print

Columns are calendar quarters. Growth is versus the year-ago quarter. Q3 2026 is omitted because it has not been reported. Same-store rates are already year-over-year.

Read the slope, not the level. Revenue YoY has improved by about 330 bps from the Q4 2025 trough and is still negative. Adjusted EPS YoY has compressed by about 2,280 bps from the Q3 2025 rate and is stuck near zero. Same-store service gross profit has decelerated in every comparable quarter since the Q2 2025 spike. Parts-and-service gross-profit dollar YoY is blank before Q3 2025 because that year-ago dollar is not in this table. From the cited levels it then steps +7.0%, +6.0%, +4.5%, and +1.4%. New GPU YoY is negative in every column, from −30.3% to −14.5%, and it is not stabilizing. New-unit declines are narrowing into a harder comp. Used-unit declines are widening. F&I per unit is still growing, and the YoY rate has cooled from +7.6% to +3.2%.

Five years, annual

July 31 was a confidence restoration after May took the outlook slide down. The restoration was about the EPS streak continuing and about SG&A timing, not about vehicle gross profit reaccelerating. Mike Manley closed expecting adjusted EPS growth in the second half, with units tracking the markets served and affordability stable into Q3 and Q4 absent a shock. Tom Szlosek put adjusted SG&A at 68.2% of gross profit, 200 bps worse than 66.2% a year earlier, and pulled the 66–67% run-rate forward to year-end. Customer pay was still described as mid-single-digit. New and used unit profit were "a very narrow bandwidth, and we expect that to continue."

September 17 caught the adjectives up to the numbers. The two sentences that were supposed to carry Q3, service growth and new-GPU stability, were both qualified. Service growth "a little bit more modest," a mid-single-digit grower "a little bit pressured here in the near term," with customer-pay traffic still fine and the consumer more selective on big-ticket maintenance. New GPU "down about 10% or so sequentially," model-year changeover plus more price given away, with a typical Q4 pickup after that. CFS attachment softer in July and August, September "back to normal," and a $50 to $100 sequential hit to unit profit for the quarter. EV mix down from about 8–9% to the low single digits. AutoNation Finance was the offset he wanted remembered: book approaching $3 billion, about 18% of financed units, lifetime profit two to three times third-party paper. The 80% mix of profit from service and financial products was repeated, not revised. No 8-K replaced the SG&A band, the $325 million CapEx plan, or the directional EPS comment.

The assumptions under the July guide, and what September did to them:

1. Second-half volume comps ease because 2025's pull-forward is over. That is a Q4 statement. Q3 2025 same-store new units were +4.5%, with BEV sales up sharply into the September 30, 2025 credit expiry. Q4 2025 was −10.2%. August industry SAAR held near 16.8 million while unadjusted volume was down year over year. The industry did not roll over. The units and the tickets did not match the SAAR sentence.

2. Unit profit stays inside the recent band. The four-quarter new band is $2,281, $2,398, $2,514, $2,381. A 10% sequential decline puts Q3 under the bottom of that band. Management has also said they will give up GPU to keep the customer. The band was a preference, not a floor. September is the preference in action.

3. Customer pay keeps growing mid-single digits while the 3- to 10-year park shrinks. The same-store customer-pay gross-profit increment was $11.6 million in Q2 2026 versus $23.6 million in Q3 2025. The increment has been cut about in half. Internal reconditioning has flipped from a tailwind to a drag because unit volume is down. Wholesale fills revenue at a thinner margin. September added the nearer-term version: traffic can be fine and the ticket can still slow if owners defer large maintenance.

4. SG&A reaches the target because gross profit improves and advertising moderates. Advertising was 6.1% of gross profit versus 5.1% a year ago, +100 bps. If Q3 gross profit is lower because new GPU steps down and service is only modest, the ratio gets worse before it gets better.

5. Off-lease supply and AutoNation Finance fill the used and F&I gap without spoiling CFS per unit. Lease returns were called up 30% to 40% in the second half versus the first half, still modest. CFS per unit of $2,799 was +3.2% and already absorbed about a 2-point drag from routing loans into the captive. The September sequential cut says the attachment did not hold through July and August.

6. Buybacks continue on residual free cash flow. They already slowed inside the first half. Cash repurchases were $301.0 million in Q1 and $168.7 million in Q2. Manley would not pin a second-half dollar, and a large deal can pause them. Nothing in the window is a new deal.

Tone across the last four calls was not a straight line. Q3 2025 opened "very pleased" on 25% adjusted EPS growth. Q4 2025 was still constructive on the year and already warned that early 2026 SG&A would sit above the band. Q1 2026 removed the outlook slide and named the middle-income buyer as treading water. Q2 2026 put a forward slide back without putting numbers back, and called affordability stable. September put the operating caution back in one morning. The part of the tone that did not crack is the captive: delinquencies still described inside the roughly 3% migration they have been reserving for, and no credit scare at Laguna.

What confirms July, and treats Laguna as one-quarter mix, is same-store service gross profit reaccelerating toward mid-single digits, new GPU holding near $2,381 rather than printing near $2,140, and the adjusted SG&A ratio moving from 68.2% toward 66–67% because advertising actually moderates. What confirms Laguna is service gross profit flat to down against the +7.0% comp, new GPU about 10% below the Q2 level, and the explanation still price given away and deferred maintenance. Adjusted EPS up year over year on roughly flat revenue does not decide it. A share count about 11% lower than Q3 2025's 38.1 million, plus finance income against a $1.5 million base, can do that while gross profit is down.

Consensus is not asking for a good operating quarter. It is asking for adjusted net income to stop falling. First-half adjusted net income was down about 10% year over year. Q3 consensus, once the share count is stripped out, is about flat versus $190.9 million. The September 17 comments cut against that inflection on the three lines that produce the earnings pool.

Brand mix is the relative, not the absolute, positive. In Q2, import units were 30,060 versus 29,748, +1.0%. Domestic units were 17,080 versus 19,354, −11.7%. Premium luxury was 16,100 versus 16,745, −3.9%. Cox's September 24 industry note still had 2026 U.S. new-vehicle sales down about 1.2% versus 2025, with Asian brands taking share from Detroit. That mix favors AutoNation's import stores. It is not a reason for total units to grow against the EV-credit quarter.

The September 16 funds-rate increase of 25 bps covers roughly the last two weeks of the quarter. It is a Q4 outlook item, not the Q3 profit pool. Floorplan interest was already a small year-over-year benefit in Q2, −$42.0 million versus −$43.6 million. Do not build the quarter on the rate.

Window is August 1 through September 24, after the July 31 print. No new store acquisition, divestiture, or C-suite change. The June deals were already in the Q2 release and are not repeated. The pattern is a softer in-quarter operating tone, not a capital-structure event. The revolver adds capacity. It does not change Q3 revenue or margin.

Left out on purpose. August price-target increases were reactions to the July 31 print. A director sale of 900 shares on August 5 and the CFO's August 7 RSU vesting were not a cluster. No post-print litigation, labor action, or franchise termination turned up in the sources checked. Nothing in the window is a new guidance number.

Mixed. Over the last 12 reported quarters AutoNation beat FMP adjusted-EPS consensus 10 times and missed twice, 83%. The last four quarters are 4/4, and the streak is now seven, from Q4 2024 through Q2 2026. Revenue beat in 5 of 12 quarters, 42%, and in 0 of the last 4. The two lines have split. The street has been light on EPS and, since Q3 2025, heavy on revenue. What is deteriorating is the EPS beat. The four-quarter average EPS surprise went from about −0.1% to +7.6% to +3.4%, and the latest print was only +1.5%, or +$0.08. Under that shrinking beat, adjusted net income has fallen year over year for three straight quarters. Q3 2026 is not on the grid.

FMP scores adjusted diluted EPS, not GAAP. Q2 GAAP diluted EPS was $5.39 and adjusted EPS was $5.56. FMP's actual is $5.56. The Q2 2024 and Q3 2024 EPS misses, −8.1% and −8.2%, are the CDK quarters, and the Q2 miss is after the CDK add-back. Q1 2024 revenue is a miss only on the strict sign rule, about −0.1%. The Q4 2024 revenue surprise of +11.7% uses an FMP estimate that looks light. The EPS beat that quarter, +16.7%, is the robust one. The 5/12 and 0/4 revenue counts do not depend on that cell.

Green is a beat. Solid green is a surprise of about 3% or more on revenue and about 8% or more on EPS. Yellow is a narrow beat, or the Q1 2024 revenue miss that is a miss in name only. Red is a miss. Solid red is a miss of about 3% or more on revenue and the two high-single-digit EPS misses. Read left to right. Revenue opens green, turns red through the CDK year, spikes in Q4 2024, and then stays red for four quarters. The red is fading. It has not flipped. The EPS row is two dark red cells, two dark green cells, then a fade from mid-green to the palest green of the streak. Seven beats in a row is not the same thing as a beat that is still widening.

The quality of the beat deteriorated with the size. Adjusted net income went from +17.7% in Q3 2025, $190.9 million versus $162.2 million, to −6.6%, −10.6%, and −10.2%. Q2 adjusted net income was $187.8 million versus $209.2 million. Shares were 33.8 million versus 38.3 million, about −12%. A roughly 10% drop in adjusted earnings and a roughly 12% drop in the share count is how the company printed +1.8% EPS growth and an eight-cent beat. The street has learned the buyback. The residual surprise is now cents.

Management does not bridge the print to a consensus number, because it does not guide one. The variance versus last year, in their words, is new-vehicle dollars: affordability, the tariff pull-forward, EV-credit timing, and premium-luxury mix, not a share loss. The offset they want scored is aftersales and CFS, plus the cash used to retire shares. That offset was large enough to beat EPS when the year-ago base was depressed. It is no longer large enough to grow adjusted net income.

Read-through for October 22. Do not carry the seven-quarter streak forward as if the magnitude were stable. A repeat of the last-four average EPS surprise, about +3.4%, on a $5.68 bar is about +$0.19. A repeat of the Q2 surprise is about eight cents. Either one is a beat of a number the company never published. Neither one is the +16% surprises of Q4 2024 and Q2 2025. The path inside the last four is down. The +13.4% growth rate inside $5.68 does not match the last three year-over-year prints of +2.2%, +0.2%, and +1.8%. An 11% share reduction on a flat $190.9 million lands near $5.68. Flat net income is not what the last two quarters did. If the earnings decline persists, the buyback the street can already see produces a small beat or a miss, not a 13% earnings inflection.

Revenue is the cleaner miss risk, and the bar is already low. The company has missed it four times in a row, by 2.7%, 3.8%, 1.4%, and 1.0%. The last two misses were about $70–$90 million. That is the miss to underwrite unless new-vehicle units break again against the EV-credit quarter. What would actually change the pattern is a revenue beat with adjusted net income up year over year, not another eight-cent EPS beat on a smaller share count.

The July 31 reaction is context, not a multiple. The print was in the market before the open. The open was about 7.7% below the prior close. The close was about 1.0% below that prior close. The next session bounced about 1.4%. The gap was the tape's first read of an EPS beat that was not an operating beat. The close mostly filled it. It did not upgrade the trajectory. Those prices are Daloopa adjusted closes and the July 31 open. They have no fundamental id and are not a valuation.