KMX’s Q2 Volume Rebound Has Not Yet Repaired Unit Economics

Summary: CarMax’s fiscal Q2 used-car volumes and reported earnings rebounded, but it bought that growth with lower per-unit vehicle profit; CAF income also benefited from a lower loan-loss provision and a loan-sale gain. The $59.23 September 29 close had already risen 4.7% on the report. A reference-only map through the November 3 strategy update has a flat base and 7.3% historical-low downside stress. The long fails 10/5: Reject / no trade.

Publication time: 2026-09-30 13:45 Asia/Singapore | Scope: U.S. market, long opportunities only | Classification: Reject / no trade

Market reference: KMX closed at $59.23 on September 29, up $2.68 or 4.74% from $56.55 the prior session. The vendor’s regular-session record shows a $60.41 open, $63.67 high, $58.03 low and 12.18 million shares. These are reference data, not an execution quote; spread, depth, venue, volume quality and exit capacity are unverified. [1]

Opportunity Ranking

Rank	Candidate	Direction	Mispricing under review	Evidence freshness	Catalyst window	Positioning	10/5 status	Tradeability	Main rejection risk
1	CarMax (`KMX`)	Long screen	Q2 used-unit growth accelerated, but retail and wholesale gross profit per unit fell; CAF income was boosted by lower loss provision and a $16.6M loan-sale gain. At $59.23 the stock trades around 1.33x August book value and about 17.8x a rough adjusted trailing EPS sum; the earnings bridge is improving but not yet clean. [1][2][3][4][5][6][7]	Sep. 29 Q2 release, latest quarterly balance sheet and regular close.	Nov. 3 strategic update; Q3 FY27 repurchases planned at a modest level, not yet completed.	No Q2 repurchases; broader ownership, short interest, options and fund flows not assessed.	Reject: flat base, 7.3% historical-low stress, 0:1 base/adverse ratio.	Liquid NYSE common in normal conditions; no live execution audit.	Volume recovery may be purchased through price cuts while credit, unit economics and capital returns remain unproven.
2	Vail Resorts (`MTN`)	Long screen	FY27 EBITDA guidance rebounds from a weather-hit year, but early season-pass units were down 12% and sales dollars down 6%; the forecast assumes normal weather. [8]	Sep. 28 results and Sep. 29 regular close.	2026/27 winter visitation, pass sales and weather.	Ownership and derivatives not assessed.	Reject pending valuation: seasonal rebound does not establish a +10% base with bounded downside.	Larger listed issuer; no execution audit.	Another poor snow season can overwhelm pricing and savings.
3	Micron (`MU`)	Long screen	Q4 FY26 results are scheduled for Sep. 30 after the close; the next earnings print is an information event, not yet an operating surprise. [9]	Latest Q3 FY26 results and Aug. 26 event notice.	Sep. 30 after close.	Current positioning not assessed.	Reject pending results: no post-event map yet.	Highly liquid common in normal conditions; no execution audit.	Memory pricing and customer mix can reverse quickly after an optimistic cycle.
4	SPDR S&P 500 ETF (`SPY`)	Long / market baseline	Broad-market exposure has no issuer-specific catalyst in this comparison. [10]	Sep. 29 market reference.	None specific to the fund.	Broad index positioning.	Reject as baseline: not a company-level 10/5 thesis.	Broad-market instrument; not a KMX implementation.	Macro repricing remains possible.

Selected opportunity: KMX ranks first for information value, not as the best executable opportunity. A Sep. 29 short-screen article had treated KMX’s Q2 results as pending. The release now resolves that catalyst; this article tests the new operating evidence, not the earlier pre-result setup. No prior KMX article or signal record was found. [9]

Why this one now: Q2 answers the demand question more favorably than Q1: comparable-store used units rose 13.0% after declining 0.8% in Q1. But retail gross profit per unit fell from $2,216 to $2,105 year over year, while wholesale gross profit per unit fell from $993 to $858. KMX’s regular close rose 4.74%, but the close was 6.97% below the day’s $63.67 high. Demand has recovered; the profitability of the pricing required to recover it remains the test. [1][2][3]

What should surprise the reader: The 81.3% diluted-EPS increase to $1.16 exceeded the 73.3% growth in net earnings, in part alongside 4.8% fewer weighted-average diluted shares. Yet CarMax repurchased no shares in Q2, and shares outstanding at August 31 were 141.934 million, nearly unchanged from 141.799 million at February 28. Management’s “modest” Q3 resumption is an intention, not buying already done. [2]

Why This Can Move More Than 5% Soon

KMX announced a November 3 strategic update and expects to resume modest repurchases in Q3 FY27. A concrete update on comp growth, per-unit gross profit, SG&A savings, CAF credit quality and actual repurchases could change the recovery case. The same facts can disappoint: pricing may continue to compress vehicle profit, loan-loss provisions can normalize upward, and the company can choose not to repurchase. That is a catalyst for information, not evidence that the long has favorable asymmetry. [2]

10/5 Asymmetry Gate

At $59.23, the observed post-report high of $63.67 represents 7.5% upside, short of the required 10%. A +10% base needs $65.15, near the secondary-source 52-week high of $65.28; the record does not support assigning that price a greater-than-50% probability by the November update. The Sep. 23 intraday low of $54.90 is a 7.3% downside stress. The highest-probability base case is flat, so gross base reward to adverse risk is 0:1. No fundamental valuation bridge proves that $54.90 bounds a downturn, and the bottom is not a stop or maximum-loss estimate. Reject before execution costs.

The Setup

CarMax reported $7.878 billion of net revenues for the quarter ended August 31, up 19.5%; combined retail and wholesale units rose 14.7%, retail comps rose 13.0%, and net earnings increased to $165.3 million from $95.4 million. The result follows Q1, when retail comps had been down 0.8% and diluted EPS was $1.31 versus $1.38 a year earlier. The sequential demand improvement is observable, but one quarter does not establish a new margin run rate. [2][3]

Gross profit rose only 11.4%, slower than revenue. Retail vehicle gross profit per unit fell $111 to $2,105, and wholesale gross profit per unit fell $135 to $858. EPP gross profit increased 33.1%, with margin per retail unit up $46 to $623; SG&A per total unit improved 8.8%. This mix explains how total gross profit and operating leverage improved while the core vehicle margin per unit narrowed. [2]

The Market Price

At the $59.23 close, multiplying by the 141.934 million shares outstanding reported on August 31 gives an illustrative equity value of about $8.41 billion. Against reported book equity of $6.305 billion, that is about 1.33x book. The share count and book value are dated to quarter-end, not a same-time execution or valuation snapshot. The 52-week low is $30.26 and high $65.28 in a secondary market summary, underscoring that historical levels do not define a loss boundary. [2][7]

The latest four reported quarterly diluted EPS figures were $0.43, $(0.85), $1.31 and $1.16, or $2.05 in aggregate GAAP trailing EPS. The FY26 fourth quarter included a $0.99 goodwill impairment and $0.20 of restructuring per diluted share; excluding those items, CarMax reported adjusted EPS of $0.34. Q3 FY26 included $0.08 of restructuring, implying roughly $0.51 adjusted EPS for that quarter. Adding the two unadjusted FY27 quarters gives an approximate $3.32 adjusted trailing sum and a 17.8x price/adjusted-EPS reference. That is a rough mixed-period multiple, not a forecast; CAF provision, loan-sale income and vehicle margins make normalization decisive. [2][3][4][5]

The Mispricing

Observed fact: The report shows a real recovery in retail comps, used-unit volume, EPP margin and SG&A per unit. A secondary preview published before the report estimated $0.72 EPS and $6.95 billion revenue; actual diluted EPS was $1.16 and revenue was $7.878 billion. The forecast was not issuer guidance, and other estimate providers differed, so the beat establishes a surprise against that preview—not fair value. [2][7]

Counterfact: Retail GPU has fallen for multiple quarters, wholesale GPU fell again, and CAF’s 32.1% income increase included a $28.8 million lower loan-loss provision and a $16.6 million gain on sale of auto loans. These effects are disclosed by the issuer; they are not proof that credit quality will worsen, but they mean the headline income growth is not a clean recurring run rate. [2][3]

The financial-services balance sheet needs separate treatment. At August 31, cash was $170.5 million and restricted cash from auto-loan collections was $604.9 million. Auto loans held for investment were $15.812 billion; non-recourse notes payable totaled about $15.694 billion. These largely correspond to CAF funding claims and assets and should not be naively added to corporate debt. Recourse long-term debt was $1.681 billion, operating lease liabilities $496 million, and book equity $6.305 billion. The capital structure is finance-heavy, not a simple net-cash retailer. [2]

Cash flow does not yet confirm earnings as a clean common-equity cash engine. First-half operating cash flow was $900.5 million versus $1.085 billion a year earlier, and included $579.9 million of proceeds from auto-loan sales; capital spending was $180.4 million. Auto loan balances and inventory also moved materially. Those flows can be normal for CAF and vehicle retail, but they make raw operating cash flow a poor substitute for a segmented, through-cycle owner-earnings bridge. [2]

Inference: The stock’s 4.74% report-day rise prices some validation of the demand-recovery strategy. It does not establish that lower per-unit margins have bottomed or that credit income is sustainable. A rough 17.8x adjusted trailing EPS multiple is not obviously distressed, and the reported 1.33x book multiple embeds confidence that the loan portfolio and retail franchise will earn through the cycle. This does not prove overvaluation; it shows why a cheap-asset claim alone is inadequate.

Positioning

The company reported no Q2 repurchases and $1.31 billion remaining under authorization, with an intention to resume modestly in Q3. Q2 diluted average shares fell year over year, but period-end shares were nearly flat versus February. This makes the rebound’s current flow support prospective rather than observed. No current short-interest, options, dealer, fund-flow or ownership evidence was assessed. [2]

Catalyst Ladder and Cheapest Falsification

Step	State and timing	Observable test	Failure path
Q2 demand recovery	Reported Sep. 29 for quarter ended Aug. 31.	Track comps and used/wholesale units separately from retail and wholesale GPU.	Price actions keep units moving but margins do not recover.
Capital returns	No Q2 repurchases; management intends modest Q3 FY27 resumption; $1.31B remains authorized.	Verify shares actually repurchased, dollars spent, average price and period-end denominator.	Authorization remains unused or repurchases compete with leverage and CAF capital requirements.
Strategic update	Scheduled Nov. 3 at 8:00 a.m. ET.	Test measurable milestones, margin targets, inventory and reconditioning throughput, CAF risk and buyback pace.	Update repeats aspirations without unit-economics, credit or cash evidence.

The cheapest falsification is the November update paired with Q3 reported per-unit gross profit and CAF delinquency/provision data. The key question is whether Q2 volume can hold without another step down in GPU or a favorable reserve/loan-sale tailwind.

Price Target and Probability Map

The horizon runs from the Sep. 29 regular close through the Nov. 3 strategic update. These are observed-price sensitivities, not fair-value targets. Top retests the Sep. 29 regular-session high; base holds the Sep. 29 close; bottom revisits the Sep. 23 intraday low. The 20%/55%/25% weights are subjective decision weights, not historical frequencies: the base dominates because the update may add information but no new earnings print is scheduled before it; the top requires confidence that unit growth and a strategic plan support another breakout; the bottom reflects margin and credit concerns if demand momentum stalls. [1][2]

Scenario	Probability	Target / Level	Return from $59.23 close	Horizon	Conditions	Evidence quality
Top Case	20%	$63.67	+7.5%	Through Nov. 3 strategic update	Retest the Sep. 29 high if unit momentum, margin plan and CAF normalization convince investors. Historical anchor, not fair value.	Low.
Base Case	55%	$59.23	0.0%	Same	Price holds while investors wait for implementation detail and actual buyback activity.	Low.
Bottom Case	25%	$54.90	-7.3%	Same	Retest the Sep. 23 low if vehicle GPU or credit quality disappoints. Historical stress, not a floor.	Low.
Invalidation	n/a	November update demonstrates durable positive GPU and normalized CAF earnings, or instead shows that unit growth requires further margin sacrifice and credit support.	Rebuild the operating and capital-return thesis after the disclosures; no stop is implied.	Nov. 3 update and subsequent quarter.	High that updates are observable; outcome uncertain.

Probability-weighted reference value: $59.04, or -0.3% before costs: 0.20×63.67 + 0.55×59.23 + 0.25×54.90.
Current market level and timestamp: $59.23 at Sep. 29, 4:00 p.m. EDT regular-session close.
Primary instrument: none; no trade.
10/5 favorable base move: 0.0%.
10/5 credible adverse move: 7.3% to an observed Sep. 23 low; not a maximum-loss estimate.
10/5 gross base reward / adverse risk: 0:1.
10/5 measurement basis: reference-only; entry is null.
10/5 status: Reject.
Confidence: Low; the operational release is primary, but price levels are historical and position/execution data are incomplete.

Sensitivity: If the bottom is raised to the Sep. 28 close of $56.55, adverse stress becomes 4.5% and weighted nominal value is about $59.45 (+0.4%); the base remains flat. A +10% base requires $65.15, which is not supported by the November 3 operating information currently disclosed. Even on the milder bottom, 10/5 fails because reward is zero and no costs or live entry are included.

The Kill Shot

Strongest counterargument: The Q2 demand recovery is real: retail comps grew 13%, total units rose 14.7%, revenue rose 19.5%, EPP margin improved, and SG&A per unit fell. CarMax says lower prices helped bring customers back; unit scale and a lower-cost platform could make the margin sacrifice temporary. CAF’s lower provision may also reflect better performance of older vintages, not merely a reserve release. If the strategic update shows credible GPU recovery, the market could be underestimating the earnings rebound. [2]

Load-bearing assumption: CarMax can retain the recovered unit volume while restoring retail/wholesale GPU and keeping CAF losses controlled. The latest release provides one quarter of faster growth but no completed post-report quarter to show that combination is durable.

The long can lose despite higher sales: promotional pricing can keep reducing gross profit per vehicle, used-car prices can change inventory values, auto delinquencies can reverse CAF provision trends, and securitization funding conditions can tighten. The buyback is discretionary and the share count has not yet contracted meaningfully in FY27. Macroeconomic, rate, credit, inventory and exit-liquidity risks remain. [2]

What Would Prove This Wrong

The no-trade call should be rebuilt if the November 3 update sets measurable milestones for sustained comparable-unit growth, stable or recovering GPU, specific inventory/reconditioning economics, normalized CAF credit losses and actual repurchases. Those facts would still need a per-share valuation bridge and downside stress to make a long. If growth relies on continued margin cuts, provisions normalize higher, or the Q3 buyback fails to occur, the recovery thesis weakens. Neither result can be inferred from the authorization or the Q2 headline.

Risk Audit

Cyclical and pricing risk: sales growth may be bought with reduced profit per unit; wholesale pricing and inventory values can turn quickly.
Credit risk: CAF yield growth, provision releases, loan-sale gains and non-prime expansion require vintage loss, delinquency and allowance tracking.
Funding risk: $15.7B of non-recourse notes finances a large auto-loan book; securitization access and loan-sale conditions affect CAF economics.
Capital allocation: $1.31B authorization is not cash spent; modest Q3 resumption is company intent only.
Execution risk: September 29 market close and historical range do not supply a live spread, depth, venue or exit audit. Overnight gaps and slippage can exceed the modeled reference range.

Best Trade Strategy

No trade. Keep entry.price null and execution.can_execute false. Common stock is the only plain instrument worth revisiting after the November 3 update, actual Q3 share repurchases and an updated GPU/CAF/loan-loss bridge. Do not use options, leverage, margin, market orders or price-floor language. A 4.74% close-to-close rally, a 52-week range and a large repurchase authorization do not establish a bounded 10/5 long.

Sources

KMX regular-session history — Sep. 29 close $59.23, prior close $56.55, daily range, volume and preceding Sep. 23 low.
CarMax Q2 FY2027 earnings release, SEC Exhibit 99.1 — report, unit and gross-profit metrics, CAF drivers, repurchase state, balance sheet and cash flows.
CarMax Q1 FY2027 earnings release, SEC Exhibit 99.1 — prior-quarter comp, GPU, CAF and share-count baseline.
CarMax Q4 and FY2026 results — Q4 impairment/restructuring and adjusted-EPS bridge.
CarMax Q3 FY2026 results — prior GAAP EPS and restructuring context.
CarMax second-quarter conference call / Nov. 3 update notice — issuer’s stated strategic-update timing and Q3 repurchase intention.
KMX pre-report estimate notice — a secondary preview published before the release cited $0.721 EPS and $6.947B revenue estimates. Estimates varied across providers; this is context, not fair value.
MTN FY2026 results and FY2027 outlook and MTN close history — comparative current results and price reference.
Micron Q4 FY2026 results notice — comparative scheduled event, not a result.
SPY finance reference — broad-market baseline.

Research Quality Scorecard

Criterion	Score	Evidence-based reason
Market disagreement	4/5	Unit and earnings rebound contrasts with lower vehicle GPU, CAF credit/provision sensitivity and a 4.74% market repricing.
Evidence base	4/5	Fresh SEC-filed exhibit and balance sheet; some comparison estimates and market prices are secondary.
Positioning and flows	2/5	No Q2 buyback and nearly flat period-end shares; broader holder, short and derivatives data are unknown.
Catalyst path	4/5	Nov. 3 strategy update and Q3 buyback plan are dated, but both are not operating proof.
Payoff architecture	1/5	Flat highest-probability case, 7.3% adverse reference stress and 0:1 ratio.
Invalidation discipline	4/5	GPU, comps, CAF provision/allowance, cash flows and completed shares provide measurable tests.
Differentiated insight	4/5	Separates volume recovery from per-unit profit, loan-sale income, loss provisioning and actual buybacks.
Client value	5/5	Clarifies that a strong EPS headline and planned repurchase do not demonstrate durable common-equity economics.
Total	28/40	Reject / no-trade screen; failed 10/5 overrides score.

Bottom Line

CarMax has made a real demand recovery visible: retail comps rose 13%, combined unit sales rose 14.7%, and reported net income rose 73%. But retail and wholesale profit per unit fell, CAF income included a lower loan-loss provision and loan-sale gain, and first-half operating cash flow fell year over year while including substantial loan-sale proceeds. The stock closed 4.74% higher, yet the highest-probability reference case through November 3 is flat and the observed-low stress exceeds 5%. A better quarter is not yet a 10/5 long. Reject / no trade.

AI Illustration Prompt

Create a restrained forensic editorial illustration for The Mispricing Desk: a used-car unit meter reads “Q2 FY27 UNITS +14.7% / COMPS +13.0%” while two price tags beneath show “RETAIL GPU $2,105, DOWN $111” and “WHOLESALE GPU $858, DOWN $135.” A CAF ledger reads “INCOME +32.1%” with separate amber lines “PROVISION -$28.8M” and “LOAN-SALE GAIN $16.6M.” To one side, show “Q2 BUYBACK $0 / $1.31B AUTHORIZED / Q3 MODEST RESUMPTION INTENDED,” and a finance book with “$15.812B AUTO LOANS / $15.694B NON-RECOURSE NOTES” clearly paired. A small market panel reads “SEP. 29 CLOSE $59.23 / FLAT BASE / 7.3% HISTORICAL-LOW STRESS.” Use cream paper, graphite and restrained navy with one amber risk mark; avoid a falling-arrow cliché or implied price floor. Add a subtle readable “The Mispricing Desk” watermark.