KMX · CarMax Inc. · Stock research

Current Thesis

The read from early July — spent positioning, no catalyst until September — has been overtaken by tape. KMX closed $57.33 on 2026-07-17 after a five-session run of +17.2% off the $50–$53 shelf, and the character of the move changed along the way. What powered May and June was short covering into a binary. The remarkable number is the consensus target, $41.73, sitting some 27% below spot. When a Street that is structurally short a name starts marking targets up toward a price it already trades at, upgrades chase price for months rather than the reverse. Behind it sits an actual operating print — Q1 FY2027 (reported 2026-06-17, quarter ended 2026-05-31) delivered revenue of $8.01B, up 6.2% against a $7.42–$7.43B bar, EPS of $1.31 against a $0.94–$0.96 consensus, total units +3.3%, and SG&A per unit down 6.8%. Keith Barr, CEO since 2026-03-16, put a four-pillar plan behind it: sharper pricing, omni/digital, finance and protection attachment, structural cost. Director Mark F. O'Neil bought alongside him, over $1M combined. This is a legacy retailer pivoting under new management with insiders funding it personally and the analyst community still positioned wrong. The theme is ACCELERATING again. It is also extended, and the peer group is not confirming.

Bullish and bearish views on CarMax Inc.

The model's bull view on CarMax Inc. (KMX), in brief: Q1 FY2027 (2026-06-17) beat by a wide margin, not a rounding error: EPS $1.31 vs $0.94–$0.96 consensus, revenue $8.01B (+6.2%) vs ~$7.43B expected, total units +3.3% on strong wholesale demand. The bear view: The beat was bought with margin. Retail gross profit per used unit compressed because management priced aggressively to move volume; unit growth is being purchased, and a second quarter of that trade without accelerating comps reads as discounting rather than share capture.… Both cases follow in full.

Bull Case

• Q1 FY2027 (2026-06-17) beat by a wide margin, not a rounding error: EPS $1.31 vs $0.94–$0.96 consensus, revenue $8.01B (+6.2%) vs ~$7.43B expected, total units +3.3% on strong wholesale demand.
• SG&A per unit fell 6.8% in the quarter, against a management commitment to remove ≥$150M of SG&A over 18 months with most landing inside FY2027 — an EPS lever that works regardless of used-car demand.
• Stephens upgraded to Overweight with a $66 target, raised from $43, in mid-July; the stock jumped ~9% that session from ~$51.78. First real rating conversion since the print.
• Consensus price target is $41.73 versus a $57.33 close on 2026-07-17. Barclays sits Underweight $37, JP Morgan Underweight $38 (both 2026-06-18/22). Every desk below spot is a forced marker-upward as the tape holds.
• Insider conviction is dated and open-market, not vesting: CEO Keith Barr 9,400 shares at $53.005 on 2026-06-22 (Form 4 filed 06-24); director Mark F. O'Neil also purchasing, >$1M combined.
• 2026-07-15 printed $58.94 (+5.8%) on 4.71M shares, roughly double the surrounding sessions' volume — participation confirmed the breakout rather than drifting it higher on air.
• 52-week range $30.26–$71.99 leaves the prior high 26% overhead if the turnaround converts the neutral rating cluster.

Bear Case

• The beat was bought with margin. Retail gross profit per used unit compressed because management priced aggressively to move volume; unit growth is being purchased, and a second quarter of that trade without accelerating comps reads as discounting rather than share capture.
• Trailing fundamentals do not support the multiple: LTM revenue -1.8%, operating margin -1.8%, P/E 33.8 against an S&P 500 median near 24.2 (Trefis, 2026-07-16). Price has run ahead of the business.
• Peer tape is not confirming. Carvana fell ~8% on 2026-06-17 on used-car margin and credit concerns the same session KMX reported. A used-vehicle re-rate with the highest-beta comparable under credit pressure is a narrow move.
• Macro is hostile to the financing arm. The 2026-06-17 Fed decision — Kevin Warsh's first as chair — signaled future tightening and the Dow fell 500+ on 06-18. CAF income was already -9.8% YoY at $143.7M in the prior quarter; higher-for-longer compresses spreads and consumer affordability together.
• The rating distribution remains defensive: BofA Underperform $45, RBC Sector Perform $45, Mizuho Neutral $43, Morgan Stanley Equal-Weight $44, Truist Hold $50. Dispersion from $37 to $66 marks a contested turnaround.
• 2026-07-17 closed -1.95% at $57.33 after tagging $58.94 — the first distribution day of the leg.

Setup & Price Structure

• Base: 2026-07-06 to 2026-07-10 held $50.21–$53.49 with 07-01 at $51.82. That congestion is the shelf.
• Breakout: 07-13 $54.87 → 07-14 $55.73 → 07-15 $58.94 (+5.8%, 4.71M shares). Gap through the shelf on expanding volume.
• Follow-through: 07-16 $58.47, 07-17 $57.33 (-1.95%). Holding above the breakout, cooling off the high.
• $53 is the operative structure — it caps the pre-breakout range, sits at the rising short-term average, and coincides with the $53.005 print where the CEO transacted on 2026-06-22.
• Extension: roughly 12–14% above the July base. Entering here is buying a fifth-day move, and the standard consumer-discretionary behaviour is a retest of the gap before the next leg.
• Overhead: the $66 Stephens target and the $71.99 52-week high. No dense supply between $59 and $66 from the recent structure — the air pocket is the reason the move ran as fast as it did.

Catalyst Calendar (next 30 days)

• 2026-07-20 to 2026-08-18 — no scheduled company event. The window is driven entirely by rating migration and Form 4 flow.
• Ongoing, undated — further target revisions from the $37–$50 cluster (Barclays, JPM, Mizuho, Morgan Stanley, Truist). Each mark-up toward spot is narrative confirmation; a downgrade held at $37–$38 through a $57 tape is a stance nobody defends for long.
• Ongoing, undated — additional insider Form 4s. Continued open-market buying above $53 would be the cleanest confirmation available before the print.
• 2026-08-31 — Q2 FY2027 quarter ends. No disclosure, but the pricing-versus-GPU trade is being decided inside this window.
• 2026-09-29 (confirmed) — Q2 FY2027 results pre-market, conference call 8:00 a.m. ET. Binary. This is where "traded margin for units" either shows up as comp acceleration or as a broken plan.
• Macro — Fed commentary under Warsh through August. Any hardening of the tightening path pressures CAF spreads and the affordability case directly.

What Would Change Our Mind

• A weekly close below $53 fills the 2026-07-13/15 gap, loses the July shelf, and returns the name to the $50–$53 chop it broke out of. That ends the leg.
• The $41.73 consensus target failing to migrate upward over the next four to six weeks while price holds $57+ would indicate the Stephens move was isolated rather than the front of an upgrade cycle.
• A second consecutive quarter of retail GPU compression at the 2026-09-29 print without comparable-store unit acceleration converts the four-pillar plan from share capture into margin destruction.
• Insider selling by Barr or O'Neil after the July run would remove the strongest non-price signal in the file.
• Deterioration in used-vehicle credit — rising subprime delinquencies, a CAF loss-provision step-up, or peer credit blowups — would break the financing leg independently of the retail story.
• A push above $62 without volume expansion, followed by a sharp reversal, would mark the move as exhaustion rather than accumulation.

Correlation Notes

• Direct comps: CVNA (highest beta, currently the weak link on margin and credit), AN, LAD, PAG, GPI, ABG, SAH. KMX outperforming this group is the turnaround thesis working; KMX rallying while the group sags is a single-stock re-rate with less durability.
• Auto credit complex: ALLY, CACC, OMF, CRMT. CarMax Auto Finance ties KMX to subprime consumer credit; widening spreads in this group hit KMX earnings with a lag the equity usually ignores until the print.
• Rates: inverse sensitivity to the front end. The used-vehicle affordability equation is a monthly-payment calculation, so KMX trades against 2-year yields more tightly than most retail.
• Consumer discretionary tape (XLY, XRT) sets the regime. A broad discretionary rotation supports the leg; a defensive rotation strands it.
• Wholesale pricing (Manheim Used Vehicle Value Index, monthly) is the fundamental input that leads GPU by roughly a quarter and is the earliest read on whether the aggressive-pricing strategy is compressing or expanding spread.