Teardown

Retail / Used Cars · Deep dive

CarMax

The largest US used-car retailer — 256 superstores, 780,684 retail used units and ~$26B revenue in the fiscal year ended February 28, 2026, only ~3.6% share of the 0-10-year vehicle market, a $16.4B captive auto-finance book showing rising loss provisions on 2022-23 vintages, and a brand-new outsider CEO (Keith Barr, ex-IHG, started March 16, 2026) inheriting a share-losing incumbent whose stock nonetheless ripped ~50% in 2026 on hopes his four-pillar cost-and-pricing reset can defend the franchise against Carvana's ~43% unit growth.

at risk

CarMax retails ~780k used cars a year yet controls only ~3.6% of the 0-10-year market, retail comp used units went negative (-0.8%) in Q1 FY2027, the CAF captive book is absorbing rising loss provisions on 2022-23 vintages (Q2 FY2026 provision $142.2M vs $112.6M PY), and Carvana just posted 596,641 units (+43%) on $20.3B FY2025 revenue while CarMax posted ~1% comp unit growth — a share transfer, not a competitive draw.

My take

HQ
Richmond, VA
Founded
1993 (first Richmond, VA superstore opened October 1993 as a Circuit City subsidiary)
Ownership
Public (NYSE: KMX); institution-dominated float, no PE sponsor or founding-family control block
Funding
No venture capital. Bootstrapped inside Circuit City Stores from 1991 concept development; IPO'd as a Circuit City tracking stock in February 1997; fully spun off from Circuit City into an independent public company on October 1, 2002. Market cap ~$8.9B as of Aug 25, 2026 at ~$63.02/share; ~140M diluted shares outstanding; long-term debt ~$18.5B (including $16.4B of non-recourse notes securitizing the CAF portfolio).
Valuation
Market cap ~$8.9B and enterprise value roughly $27B as of August 25, 2026 at ~$63.02/share; shares up ~50% YTD in 2026 on the Barr-led turnaround narrative even as absolute market cap sits far below the ~$20B peak of 2021.
Revenue
FY2026 (year ended February 28, 2026): 780,684 retail used units, 538,203 wholesale units; used-vehicle sales ~80% of revenue, wholesale ~17%, ESP/other ~3%; Q4 FY2026 combined units +0.7%, CAF income -9.8%; Q1 FY2027 (ended May 31, 2026) revenue $8.0B (+6.2%), EPS $1.31 vs $0.96 consensus, combined units 392,357 (+3.3%), comp used units -0.8%, GPU $2,177 (down $230 YoY from record); SG&A -3.7%; $200M exit-rate SG&A cost-out targeted by year-end FY2027.
Headcount
Approximately 28,000 associates as of 2026 (down from ~30,000+ at the pandemic peak); 22 consecutive years on Fortune's 100 Best Companies to Work For; Glassdoor 3.5/5 across ~8,200 reviews, ~60% would recommend.
Screen
Public incumbent (bucket 5) — ~$26B FY2026 revenue, ~$8.9B market cap, ~$27B EV; largest US used-car retailer by revenue and unit volume; 256 superstores; ~28,000 associates.
Published
2026-08-25
Web
www.carmax.com
Elsewhere
LinkedIn · Crunchbase

Founders and leadership

  • Keith Barr President, CEO & Director (from March 16, 2026)

    Hospitality operator, not a car guy. Barr ran InterContinental Hotels Group as CEO from July 2017 to June 2023, overseeing ~6,000 properties across ~100 countries and ~345,000 employees, and sat on the MGM Resorts board afterward. His appointment is the CarMax board's admission that the operating problem is not automotive — it is consumer-experience, digital-in-store integration, and cost discipline in a mature-format retailer. His Q1 FY2027 four-pillar framework (competitive pricing and inventory access; connected digital + in-store; grow financing and ESP profits; cut costs through technology and ops efficiency) is textbook hospitality-turnaround language. He inherited an $8.0B-quarter business, a $16.4B captive-finance book, and a new-CEO honeymoon that has priced roughly ~$3B of equity value ahead of any evidence the plan works.

  • William D. (Bill) Nash CEO 2016-2025 (departed)

    A CarMax lifer and JMU accounting graduate who joined the company in 1997 as auction manager after starting his career as a CPA and then as an accountant at Circuit City. Nash rotated through auction services, merchandising SVP, HR and administrative services, and stepped up to President in February 2016 before becoming CEO later that year. He led the pivot to an omnichannel platform — by FY2025 >60% of retail sales involved an online interaction, >40% of customers finished most steps online — and rolled out Instant Offer online-car-buying against Carvana. His tenure ended in 2025 with the board installing David McCreight as Interim CEO, then Barr full-time, an outcome that reads as a shareholder-driven acknowledgement that the omnichannel spend had not defended units share.

  • Austin Ligon Co-founder and first standalone CEO (2002-2006)

    The retail strategist inside Circuit City who, with then-CEO Richard Sharp, developed the CarMax concept in 1991 as a big-box, no-haggle transplant of Circuit City's consumer-electronics playbook to used cars. Ligon was named SVP of Automotive at Circuit City in 1992, President of CarMax in 1995, and CEO on the October 2002 spinoff. He retired in 2006. Ligon's original insight — that used-car dealer opacity was a retail-experience problem, not an inventory-cost problem — is the founding truth of the franchise and the reason the moat has thinned: every online-first entrant since 2013 has copied the no-haggle experience without the ~$100M-per-store real-estate footprint.

Snapshot

CarMax is the largest US used-car retailer by unit volume — 256 superstores, ~780,684 retail used vehicles sold in the fiscal year ended February 28, 2026, plus 538,203 wholesale units, on roughly $26B of total revenue. It invented no-haggle, fixed-price, big-box used-car retail in 1993 as a Circuit City incubation, still owns the brand-defining used-car experience for a generation of American buyers, and runs a $16.4B captive auto-finance book (CAF) that funds 42.4% of retail units. But every operating trend it reported through FY2026 is going the wrong way relative to its main challenger: Carvana just posted 596,641 units and $20.3B revenue for calendar 2025 at 43% unit growth, while CarMax retail comp used units went slightly negative (-0.8%) in Q1 fiscal 2027. In February 2026 the board named Keith Barr, ex-CEO of InterContinental Hotels Group with no prior automotive experience, as CEO effective March 16, 2026 — an outsider hire that reads as an admission that the incumbent playbook needs rewiring.

Founding story

The company was a specialty-retail experiment inside a consumer-electronics chain. Richard Sharp, CEO of Circuit City Stores in Richmond, Virginia, hired Austin Ligon in the late 1980s and by 1991 the two were prototyping what the Circuit City fixed-price, big-box, professional-sales-associate model would look like applied to used cars — the most opaque retail category left in America. Ligon became SVP of Automotive at Circuit City in 1992. The first CarMax superstore opened in Richmond in October 1993 as “The Auto Superstore” — a large lot with hundreds of inspected, reconditioned, non-negotiable-priced vehicles, an internal finance arm, and salespeople paid a flat commission per car rather than a percentage of markup so their incentive was to move units, not fleece the buyer.

Ligon was named President of CarMax in 1995. Circuit City took CarMax public as a tracking stock in February 1997, and on October 1, 2002 spun it off tax-free into a fully independent NYSE-listed company with Ligon as its first standalone CEO. Circuit City itself would collapse into liquidation in 2008-09, precisely as CarMax was proving out at scale. The parent’s death and the subsidiary’s compounding are the founding irony: the format Circuit City incubated survived because used cars are less deflationary than televisions.

How it works

A CarMax superstore is a ~10-15-acre lot with 300-500 late-model used vehicles on display, a large showroom, a reconditioning bay, and a sales floor where associates are paid a flat commission (roughly $150-$200 per unit before bonuses) regardless of the vehicle sold. Inventory sourcing is three-legged. First, on-the-lot Instant Offer purchases: a walk-in seller gets an appraisal in ~30 minutes, and the offer stands for seven days whether or not the customer buys. Second, online Instant Offer, launched in 2020 to compete directly with Carvana. Third, wholesale auctions, where CarMax bought 538,203 units in FY2026 that it later resold to franchised dealers.

Cars that clear CarMax’s ~125-point inspection are reconditioned centrally and shipped to superstores; those that don’t clear are pushed into CarMax’s own dealer-only wholesale auctions, which now run online. Financing is the second business: CAF funds 42.4% of retail units on ~$16.4B of loans held for investment as of February 28, 2026, split across prime (Tier 1) and increasingly non-prime (Tier 2 and 3) buckets. Third-party lenders (Ally, Chase, Capital One, Santander) sit alongside CAF on the credit-app waterfall. Roughly 60% of retail sales involve some online interaction; the omnichannel plumbing was built under Nash between 2019 and 2024.

Product and business overview

Three revenue lines. Retail used-vehicle sales — ~780k retail units at an average selling price around $26,000, ~80% of consolidated revenue. Wholesale sales — ~538k units through dealer-only auctions, ~17% of revenue. Other (~3%) is dominated by CAF interest and fees plus extended service plans and GAP insurance, sold with most retail units and material to gross profit dollars — other-gross-profit contribution per retail unit runs $500-$600. Geographic footprint is 256 superstores across ~110 US media markets. Reconditioning centres sit in Kingman, AZ; Macon, GA; and elsewhere. CAF is a separate reportable segment; its income fell 9.8% in Q4 FY2026 to $143.7M as provisions rose and margin compressed.

Business model and pricing

Fixed-price, no-haggle retail — the founding proposition — is now table stakes across online-first competitors. What CarMax charges is a spread: retail sold price minus acquisition cost minus reconditioning minus overhead. Retail gross profit per used unit was $2,177 in Q1 FY2027 (down $230 from the prior-year record of $2,407 as management priced more aggressively to defend volumes); wholesale GPU was $1,046. Ancillary and CAF income are the layered margin, and in the current cycle they are getting squeezed simultaneously — retail-GPU cuts to compete on price, and CAF provisions absorb losses on prime-adjacent 2022-23 vintages originated at high prices and rates. FY2027 SG&A is guided to fall $200M on an exit-rate basis via tech and operational efficiency — a rare cost-out for a retailer that had grown its expense base every year through 2024. The tension is straightforward: cutting price to defend units drops GPU; growing CAF Tier-2 penetration drops CAF income; and cost-cutting is a one-time lever.

Traction over time

FYRetail used unitsRevenueNotable
FY2020~832k~$20.3BPre-COVID peak; the last “normal” year
FY2021~751k~$18.9BCOVID trough, then price surge
FY2022~924k~$31.9BAll-time revenue high on used-car price bubble
FY2023~808k~$29.7BSharp margin compression as prices normalised
FY2024~709k~$26.5BTrough of the freight/consumer-credit cycle
FY2025~750k~$26.4BNash’s last full year; omnichannel scale-up
FY2026780,684~$26.0B+4% units, roughly flat revenue as ASPs eased
Q1 FY2027234k retail$8.0B (+6.2%)Barr’s first quarter; comp units -0.8%

The pattern is a franchise that has spent five years failing to grow past its FY2020 unit peak. Total US used-retail volumes are up over the same window (Cox 20.3M forecast for 2026). Every point CarMax hasn’t recaptured is a share loss.

Market analysis

Cox Automotive forecasts 20.3M US retail used-vehicle sales in calendar 2026, down ~0.7% from 2025 — a slowing-but-stable market with average used prices still elevated versus pre-COVID. CarMax pegs its own share at ~3.6% of vehicles zero-to-10 years old. The category is one of the last genuinely fragmented mass-market US retail sectors: >40,000 franchised and independent used-vehicle dealers, mostly single-lot operators using Manheim (Cox Automotive) for wholesale supply, vAuto for pricing, and CDK/Dealertrack for DMS. Structural forces are ambivalent. Fleet age is at a record 12.6 years, expanding the aftermarket pool. New-vehicle affordability is stretched, pushing more first-time buyers into used. Interest rates are still elevated, which hurts monthly-payment affordability and elevates loss content in captive finance. And online-first entrants (Carvana; earlier Vroom, Shift, Fair, EchoPark) have proved that the no-haggle experience does not require CarMax’s real-estate footprint — a permanent structural threat to superstore economics.

Competitive intel

Carvana is the frame. FY2025 revenue $20.3B (+49%), retail units 596,641 (+43%), net income $1.9B, adjusted EBITDA $2.2B; Q1 2026 revenue $6.43B, 187,393 units (+40%). It has recovered from its 2023 near-bankruptcy to record profitability and is compounding at a rate CarMax has not seen since 2021. Its cost structure — vending machines and central inspection/reconditioning centres, no traditional lots — is structurally lower-overhead than CarMax’s 256 superstores. AutoNation is the franchised-dealer version, ~350 stores and ~$27B revenue, competing on off-lease supply and F&I depth. Vroom is the cautionary comp — exited retail in January 2024 — but Carvana’s rebound has retired the “online used doesn’t work” argument. Group 1 and Sonic/EchoPark compete at the metro level on trade-in supply and used-only formats respectively; Sonic pulled back EchoPark hard in 2023-24. Copart is the wholesale-auction platform that trades at ~$45B market cap on ~$4.4B revenue — a reference multiple for what the market pays for asset-light, digital, dealer-facing auto infrastructure vs CarMax’s ~$27B EV. CarMax’s honest edges are three: brand and physical presence in ~110 metros; CAF’s $16.4B captive book funding 42.4% of retail units; and a decades-old wholesale-auction supply pipeline. Its honest deficit is unit growth — the only KPI that matters in a share game.

History and evolution

What people say

The case for. Bulls (Insider Monkey, Market Inference, several 2026 Seeking Alpha long theses) frame CarMax as the highest-quality asset in used-car retail — the only national brand, the only integrated captive-finance book at scale, the only operator with a proven wholesale-auction supply pipeline. The Q1 FY2027 beat (EPS $1.31 vs $0.96 consensus) triggered JPMorgan’s Rajat Gupta and Barclays’ John Babcock to upgrade to Neutral / Equal-Weight with $60-61 targets (up from $37-38), and the stock is up ~50% YTD in 2026. Employee reviews on Glassdoor average 3.5/5 across ~8,200 reviews; 60% would recommend; CarMax has landed on Fortune’s 100 Best Companies to Work For for 22 consecutive years. Bulls argue Barr’s outsider status is the point — he brings cost discipline (the $200M SG&A cut) and consumer-experience rigour without the inherited assumptions of a 30-year CarMax lifer.

The complaints. The bear case is quiet, dense, and persistent. Wedbush’s Scott Devitt cut to Hold and lowered target 35% to $54 on missed revenue, units, ASP, EPS, and GPU. Needham cut to Hold with the thesis that digital-only competitors are pushing CarMax’s unit growth to negative — the Q1 FY2027 -0.8% comp confirmed this. Complaint boards and BBB threads (>800 complaints filed with the Richmond BBB across the 2010s alone) cite reconditioning-quality failures — undisclosed flood damage, unreported collision history, premature drivetrain failures — and the 2016 FTC settlement over failure to disclose open safety recalls is the tail risk that never fully goes away. Negative Glassdoor threads flag pay compression after a 2025 bonus overhaul (some call it “corporate gaslighting”), inconsistent management, and silos between merchandising and store ops. And the CAF loss provision — $142.2M in Q2 FY2026, $101.7M in Q1, $74M in Q4, with $71.3M of lifetime-loss estimate added on 2022-23 vintages — is the sell-side’s most durable near-term worry.

Outlook: well positioned or at risk?

At-risk — a genuinely great American retail franchise being outrun by a lower-cost online competitor while its captive finance book absorbs credit losses and its board resorts to an outsider CEO from hospitality to fix it. CarMax’s problem is not that the format is broken; it is that the format is no longer scarce. Every proposition it originated in 1993 — no-haggle, inspected inventory, in-house financing, seven-day returns — Carvana and hundreds of independent used-only lots have copied without the ~10-15-acre-per-store overhead. The math shows through: Carvana grew retail units 43% in 2025 to 596,641 while CarMax grew ~4% to 780,684, and Carvana’s Q1 2026 was +40%. On current trajectories Carvana overtakes CarMax on unit volume inside 2-3 years with a materially lighter physical footprint. Meanwhile the ~$16.4B CAF book — historically a differentiator — is a source of loss-provision volatility, and Q4 FY2026 CAF income fell 9.8%. Barr’s four-pillar plan (competitive pricing, connected digital+in-store, grow financing/ESP, cut $200M of SG&A) is coherent and near-term executable — Q1 FY2027 EPS beat and the stock’s ~50% YTD move mean the market has priced roughly two quarters of benefit of the doubt. But the strategic constraint is real: cutting price to defend units drops GPU, Tier-2 growth drops CAF income, and cost-cutting is a one-time lever. Barr does not have a growth engine. Three things to watch through FY2028: (1) whether retail comp used units turn sustainably positive or the -0.8% Q1 print becomes trend; (2) whether CAF loss provisions on FY2027 vintages moderate as underwriting tightens; (3) whether Carvana ships >800k units in calendar 2026 — the moment its retail volume crosses CarMax’s, the “largest used-car retailer” tagline becomes historical.

How to attack it

Attack the physical footprint, the reconditioning-quality perception, and the CAF cost of capital simultaneously. CarMax’s exploitable weaknesses are all in the file. First, its ~10-15-acre superstore economics can’t compete with a delivery-plus-vending-machine model — Carvana has proved it, and the next attacker’s wedge is to do the same in the ~$15-20k ASP segment CarMax historically underweighted, chasing the price-conscious first-time buyer with lower reconditioning overhead and no test-drive real estate. Second, the reconditioning-quality perception in complaint forums and the 2016 FTC recall-disclosure settlement is an opening for a challenger to publish per-VIN inspection reports as verifiable data (dyno results, cold-start metrics, undercarriage photos with timestamps) rather than a “125-point checklist” bullet — the same trick Carfax pulled on the ownership-history side of the market, applied to the mechanical-condition side. Third, CAF’s ~$16.4B book funds only 42.4% of retail units and is now absorbing provisions on Tier-2 vintages; a challenger that partners with a diversified consumer-credit platform (Upstart, Affirm, or a bank consortium) can offer materially cheaper subprime auto rates than CAF’s lifetime-loss math currently supports, poaching the exact buyers CarMax needs to keep to grow units. Fourth, the sales-associate flat-commission model — historically a trust builder — becomes irrelevant when the transaction happens on a phone; the attacker replaces it with fitment/finance/insurance/warranty bundling in-app. Fifth, CarMax’s wholesale-auction business (~538k units) is a defensive asset; a challenger that plugs into Manheim, ACV, and OPENLANE and cross-lists CarMax’s own reject inventory on those platforms shrinks CarMax’s wholesale spread. The pattern: don’t out-superstore CarMax; render superstores irrelevant.

Adjacent-segment play

Package the CAF capability and the reconditioning capability as B2B services to independent dealers, not consumers. CarMax’s genuinely novel capabilities are three: a national reconditioning pipeline that inspects and remediates ~1.3M vehicles a year; a captive auto-finance underwriting engine with 30+ years of loss data across prime and non-prime segments; and a dealer-only wholesale-auction platform running online at scale. The buyer it under-serves for all three is the ~40,000+ independent US used-car dealers who currently rely on Manheim (Cox Automotive) for wholesale, vAuto for pricing, and a rotating cast of subprime lenders for F&I. A CarMax that spun the reconditioning centres into a fee-per-vehicle service, licensed CAF’s underwriting-as-a-service to independents, and merchandised the wholesale-auction inventory to dealers under a Copart-style membership model would monetize the infrastructure without depending on CarMax retail growth to justify its capex. Copart’s ~$45B market cap on ~$4.4B of revenue is the reference multiple for what the market pays for asset-light, digital, dealer-facing auto infrastructure — a re-rating any CarMax CFO could plausibly chase if the four-pillar retail plan stalls. The reason CarMax hasn’t gone there is cultural: the company was built as a consumer retailer and its incentives, KPIs, and executive experience all point at units-per-store, not API-calls-per-dealer. Barr’s outsider status makes the pivot more likely than under a lifer, but it is still a two-to-three-year investment cycle before it would show up in the P&L.

Sources and further reading

Capital history

DateRoundAmountValuationLead(s)
1991 Concept incubation inside Circuit City Funded from Circuit City parent balance sheet n/a Richard Sharp (Circuit City CEO), Austin Ligon
October 1993 First superstore opens (Richmond, VA) Parent-funded capex n/a — subsidiary of Circuit City Stores, Inc. Circuit City
February 1997 Tracking-stock IPO Public listing as Circuit City Stores - CarMax Group tracking stock IPO at $20/share; Circuit City retained majority economic interest via tracked equity Public markets
October 1, 2002 Spinoff from Circuit City Tax-free distribution of CarMax shares to Circuit City holders CarMax begins trading as independent NYSE:KMX; opens as a standalone ~$1.5B market-cap company Public markets
2004-present Recurring CAF non-recourse securitizations Hundreds of millions to >$1B per issuance Ongoing funding of the CAF receivables book (~$16.4B outstanding as of Feb 28, 2026); $900M non-prime issuance flagged in Q3 FY2026 Structured-credit desks (bank syndicate)
2020-2024 Buyback and revolver activity Multiple share-repurchase authorisations; balance sheet capacity ~$2B undrawn at various points Buyback pace slowed materially through 2024-25 as CAF provisions rose CarMax board

Investors / owners: Institutional index and active funds dominate the float — Vanguard, BlackRock, State Street, plus long-time value shops such as Ariel Investments (a long-time top-10 KMX holder)., No PE sponsor, no founding-family block; Circuit City parent overhang ended definitively at the October 2002 spinoff., 19-analyst median price target ~$55 as of August 2026, below the ~$63 tape — a Hold consensus dressed as recovery; JPMorgan and Barclays upgraded to Neutral/Equal-Weight in mid-2026 with $60-61 targets, up from ~$37-38; Wedbush and Needham cut to Hold in 2025 on missed metrics.

Competitive set

  • Carvana (NYSE: CVNA) — The one that has actually broken CarMax's growth model. FY2025 revenue $20.3B (+49%), retail units 596,641 (+43%), net income $1.9B, adjusted EBITDA $2.2B. Q1 2026 revenue $6.43B (+52%), units 187,393 (+40%). Says it is at ~1.5% of the US auto market and gunning for 3M units/year by 2030-2035. Carvana is now growing units ~40x faster than CarMax and did it while returning to record profitability post-2023 near-bankruptcy — vindicating the online-first, no-superstore cost model.
  • AutoNation (NYSE: AN) — Franchised new-car dealer group with a large used-car second act. ~350 stores, FY2024 revenue ~$27B, used retail ~300k units. Different economics — new-car franchise protections plus F&I and service — but competes with CarMax on off-lease and trade-in supply, and its AutoNation USA used-only stores are the direct format competitor.
  • Vroom (formerly NASDAQ: VRM) — The cautionary online comp — exited retail used-car sales in January 2024 after burning through its SPAC proceeds and pivoted to an AI/analytics vendor and a captive-finance subsidiary. Its collapse is the reason CarMax bulls argue online-first doesn't work at scale — but Carvana's 2024-26 recovery has retired that argument.
  • Group 1 Automotive (NYSE: GPI) — Franchised dealer group (~200 US stores plus UK/Brazil) with a growing used business. FY2024 revenue ~$19B. Competes for the same trade-in and off-lease supply; a share taker at the metro level, especially where its Ford / Toyota / BMW franchises give it first look at high-quality trade-ins.
  • EchoPark (Sonic Automotive: NYSE: SAH) — Sonic's used-only superstore format, launched 2013 as a direct CarMax knockoff. Sonic retrenched EchoPark heavily in 2023-24 after losses, closing more than half of its standalone stores. Proof that the CarMax playbook itself is hard to replicate — but also proof it doesn't compound at scale.
  • Copart (NASDAQ: CPRT) — Not a direct retail competitor but a structurally adjacent salvage/wholesale auction platform with a ~$45B market cap on FY2024 revenue ~$4.4B. Its digital auction dominance sets the wholesale-price benchmark CarMax's ~538k-unit wholesale business is measured against, and shows what a purely digital, asset-light comp trades at.
  • Franchised and independent dealers (~40,000+ US used-car retailers) — The real market. Cox Automotive puts total US used-vehicle retail sales at ~20.3M in 2026. CarMax's 780k retail units are ~3.6% of the 0-10-year segment — meaning the other ~96% of the market sits with dealer groups and single-lot independents, most of whom now use vAuto, CDK, and Manheim as their tech stack.