← back to rankingKMX · CarMax, Inc.
Auto - Dealerships · mkt cap $6.2B · calls: Q1 FY2027 vs Q4 FY2026
31.0 conviction · conf-adj 31
conf 4/10 🚀 reported partial
enthusiasm:9.0 · trend:0 · quantifies:0 · impact:0 · under_radar:14 · credibility:5 · business_impact:4 · disruption:0 · commitment:-4 · confirmation:3
Enthusiasm latest 3 / prev 3 (flat)
CarMax management references AI and algorithmic tools at a low-to-moderate level across both calls, with no quantified business impact attached to any AI initiative. The most concrete disclosure in the latest call is the live deployment of AI assistants in digital channels and CECs, plus ongoing repricing of the algorithm stack to incorporate external market signals — but management explicitly defers specifics to a fall 2026 strategic update. Credibility is limited: the AI narrative is diffuse, vague, and entirely forward-leaning in terms of measurable outcomes, with CEO Keith Barr drawing primarily on hospitality-industry analogies rather than CarMax-specific AI proof points.
MODEL CONSENSUS (impact)
partial
Agree on adopter, soft Tier 2, thin-margin caveat. Sided with X's CAF-below-revenue accounting (rev~0); averaged the undisclosed EPS yield to ~2.9%.
Conflicts reconciled
- Claim1 rev_uplift_pct: X=0 vs Y=0.86 -> used 0 because CAF originations are loan principal recognized below net sales, not revenue (only the spread hits the top line)
- Claim1/est eps_uplift_pct: X=1.3 vs Y=4.5 -> used 2.9 (average) because the incremental CAF net yield (3% pretax vs 8% net) is undisclosed and both are defensible
- est_rev_uplift_pct: X=0.1 vs Y=0.86 -> used 0.1 because only the finance spread (~$6.7M), not origination volume, reaches revenue
- vs_analyst_expectations: X=inline vs Y=unclear -> used inline because consensus already models decline and the AI benefit is small/in-base
- priced_in: X=medium vs Y=high -> used high (more conservative) because the sized benefit is largely realized in the FY26 base
- confidence: X=4 vs Y=6 -> used 4, lowered for conservative category choices
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | 0.1 | – |
| EPS uplift % | 1.3 | – |
| Priced in | medium | – |
| vs analysts | inline | – |
| Confidence | 4 | – |
| Top line | Essentially negligible to the reported $25.88B revenue line. CAF income is recognized below net revenues, so a 100-150bps financing-penetration gain does not move the top line directly. The only genuine top-line channel is indirect: Tier 2 underwriting expansion lets marginal/near-prime customers obtain financing and complete a vehicle purchase they otherwise couldn't — a real unit-sales lever management is leaning on, but unquantified from the disclosed figures. Hence est_rev_uplift_pct ~0.1% (soft, negligible). | – |
| Bottom line | This is where the AI shows up. The 125bps midpoint penetration gain implies ~$222M of incremental CAF originations annually (125bps x ~$17.8B financeable sales), worth ~$6.7M pretax / ~$5.3M after-tax at an assumed ~3% net yield = ~1.3% of the $398M consensus net income base. Material caveat: net margin is 0.95% and operating income is negative, so any bottom-line item looks amplified against a thin earnings denominator — the absolute dollars are small. Most of the 100-150bps is already realized in the FY26 base, so forward incremental is even less. The bigger, genuinely-incremental lever is the Tier 2 share doubling (<10% -> 20%+) with a stated FY27 acceleration — but its Tier-2 dollar base is undisclosed, so it stays soft/unsized. | – |
| Reasoning | Consensus is modeling DECLINE, not AI-driven acceleration: revenue estimates stepped down 27.69B -> 26.37B -> 25.66B and EPS down 2.55 -> 3.28 -> 2.69 across the three snapshots. The quantifiable AI benefit (~1.3% EPS, ~$5.3M after-tax) is (a) small relative to consensus dispersion across 12 analysts and (b) largely already in the FY26 run-rate, so it is effectively in the numbers -> inline / priced-in. The one piece that could beat the soft consensus trajectory is the Tier 2 expansion accelerating into FY27, which is NOT obviously embedded in declining estimates — but it is unquantifiable from the disclosures, so it cannot be claimed as a clear gap. Net: the hard math is priced in; the upside is real but unsized. | – |
Rows highlighted where the two models disagreed.
PAST (realized)
- Part selection tool and tire selection tool already deployed in reconditioning operations (Q1 FY2027, Keith Barr).
- Pricing algorithms historically in place but were 'very margin-based'; had 'very complex pricing algorithms' already (Q1 FY2027, Keith Barr).
- Long-term investments made in digital platform cited as a strength at time of Keith Barr's arrival (Q4 FY2026, Keith Barr).
CURRENT (now)
- AI assistants integrated into CarMax's digital customer experience and in the Customer Experience Centers (CEC) as of Q1 FY2027 (Keith Barr).
- Pricing algorithms being updated to incorporate external competitive market data, local data points, and individual vehicle-type demand signals — actively shifting from margin-based to demand-maximizing pricing (Q1 FY2027, Keith Barr).
- Data sciences team actively investing in and expanding algorithmic pricing capabilities (Q1 FY2027, Keith Barr).
- Marketing team using real-time data analytics to make week-to-week ROI-driven marketing allocation decisions (Q1 FY2027, Keith Barr and Enrique Mayor-Mora).
- Digital experience improvements underway: shifted from sticker prices to monthly payment display, improved entry points from online ads, streamlined path to pre-qualification and vehicle reservation (Q1 FY2027, Keith Barr).
- Reconditioning process digitization in progress, with part and tire selection tools active (Q1 FY2027, Keith Barr).
FORWARD (guidance)
- Further evolution of pricing algorithms to maximize demand and profitability dynamically across vehicle types and markets (Q1 FY2027, Keith Barr).
- Continued and deeper digitization of reconditioning processes expected to yield cost and speed improvements — management characterizes this as the 'biggest opportunity' (Q1 FY2027, Enrique Mayor-Mora).
- More detail on technology and AI initiatives promised at a strategic update in fall 2026 (Q1 FY2027, Keith Barr).
- Ongoing investment in data sciences team and algorithmic capabilities (Q1 FY2027, Keith Barr).
- Planned further reduction of friction in digital-to-in-store experience, with technology enabling higher conversion (Q1 FY2027, Keith Barr).
- At Q4 FY2026 introduction, Keith Barr named 'use software, data and AI' as one of three explicit priorities — including personalization and inventory/pricing matching — with detail to follow (Q4 FY2026, Keith Barr).
TRACK RECORD — PROMISE vs DELIVERY
48/100 track record mixed 6 calls reviewed
CarMax reports AI/ML operating gains (Sky containment, CEC productivity) but rarely sets numbered AI deadlines; credit-model penetration guidance was partly met at best while the FY26 $125/unit COGS savings target was not reaffirmed.
Grow CAF sales penetration 100–150 bps near-term via new credit scoring/recapture — promised Q4 FY2025
partial Q1 FY26 penetration was 150 bps below prior year; Q2 rose only 60 bps YoY—well short of the guided uplift.
At least another $125 per unit logistics/reconditioning COGS savings in FY26 — promised Q4 FY2025
quietly-dropped Later calls cite ongoing COGS-efficiency work but never confirm the incremental $125/unit FY26 target after setting it.
Underwriting adjustments to add 100–200 bps of CAF penetration growth — promised Q2 FY2026
partial Management said realization is often offset by credit mix; Q2 penetration was up just 60 bps YoY.
Achieve full-year omni cost neutrality in FY26 (AI cited as efficiency driver) — promised Q4 FY2025
partial Q1 FY26 reported strong early progress on all three omni-cost metrics vs pre-omni and YoY; little follow-up quantification in later calls.
PRICED-IN (REFINED)
LOW (room left)Est. revisions flat · Fwd P/E 19.1 · EV/Sales 0.9x
AI claim maps to Used Vehicles, Wholesale Vehicles
Analyst ratings have been completely static for six consecutive months (2 buy / 14 hold / 4 strong sell every month), and price targets are actually being revised UP slightly (last month $43.29 vs. last quarter $41.50) but still sit well BELOW the current price of $48.72, indicating the street is not chasing the stock. Forward P/E of 19.1x on a low-margin used-car retailer is modest, and EV/Sales of ~0.95x is undemanding for any AI-efficiency narrative. Any AI upside — most plausibly flowing through operating cost reduction in used-vehicle retailing (the dominant ~$20.7B segment) — is not reflected in flat consensus revisions or stretched multiples, leaving meaningful room for the thesis to surprise to the upside.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
3Q3 FY20253Q4 FY20253Q1 FY20264Q2 FY20263Q3 FY20266Q4 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
Five quarters of digital-only ops talk; new CEO in Q4 FY26 named practical AI for friction, personalization, and pricing.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
5/10 qualitative impact moderate medium-term · soft evidence
Where AI matters: pricing algorithms, reconditioning efficiency, CEC productivity
CarMax has live AI deployments (demand-signal pricing algorithms, CEC AI assistants, part/tire selection tools) that can plausibly improve GPU and reduce reconditioning labor costs, but no disclosed metrics confirm realized savings; the largest cited opportunity (reconditioning digitization) is explicitly pre-quantification and deferred to a fall 2026 update.
Caveats: Management credibility score 48/100; prior AI-linked guidance (100-150bps CAF penetration lift, $125/unit COGS savings) was at best partially delivered and sometimes quietly dropped; All strategic AI detail explicitly deferred to fall 2026 — no near-term quantifiable catalyst from disclosures; New CEO draws on hospitality analogies rather than KMX-specific AI proof points, raising execution-translation risk; Reconditioning digitization described as the 'biggest opportunity' but remains entirely unquantified and multi-year in scope
AI DISRUPTION / CANNIBALIZATION RISK two-sided · 3/10
AI-powered consumer price-comparison tools incrementally compress used-car retail spreads, but CarMax's no-haggle, scale-driven model was already built on price transparency rather than information asymmetry, and the physical reconditioning/financing infrastructure is not automatable by software alone.
OPTIONS / MARKET STRUCTURE
option liquidity: fair
proxy inputs — dollar-ADV $171M · beta 1.202 · px $48.95
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.
CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders neutral, institutions adding, management language 3/10 hedged.
INSIDERS neutral no open-market buys/sells in last 6mo (routine: 13 awards, 9 tax-withholding)
INSTITUTIONS (13F) adding as of 2026-03-31: 100 new / 92 closed positions; 277 increased / 191 reduced; institutional ownership +1.86pp; +8 net 13F holders
MGMT LANGUAGE 3/10 hedged Barely discusses AI—one brief new-CEO priority line; future tense, no numbers, rollout proof, or AI Q&A.
commit “We'll use software, data and AI in practical ways that make it even easier for customers to buy and sell cars”
hedge “in practical ways that make it even easier for customers to buy and sell cars and easier for our associates to serve them”
hedge “I look forward to sharing more about our strategy and long-term objectives in due time”
VERBATIM AI QUOTES
“use software, data and AI in practical ways that make it even easier for customers to buy and sell cars and easier for our associates to serve them. That means reducing friction across the journey, personalizing the experience, improving how we match inventory and pricing to meet customer demand and ensuring a great experience both in our stores and online.”
— Keith Barr, Q4 FY2026
“we are incorporating competitive market insights within our pricing algorithms more granularly with a stronger emphasis on local data points. We're expanding comparison points across a broader set of vehicles to sharpen our individual unit pricing.”
— Keith Barr, Q1 FY2027
“We've integrated AI assistants both in our digital experience and in our CEC.”
— Keith Barr, Q1 FY2027
“our pricing algorithms are incorporating market demands and also unit demands specifically. We're understanding pricing by markets, pricing by vehicle types, so we can be more dynamic. What we're going to be focused on is how do we flex GPU to maximize sales and profitability rather than being tied to a fixed GPU quarter to quarter to quarter.”
— Keith Barr, Q1 FY2027
“we've got a fantastic data sciences team, and we're continuing to invest in that space and expand upon how we think about our algorithms and evolve them over time, because I think there's definitely opportunity in this space to sharpen up our pricing.”
— Keith Barr, Q1 FY2027
“We've got our part selection tool, which they continue to improve upon, which enables our teams to effectively find the right part for the right car at the best price. We've got our tire selection tool out there now, which has now been integrated into that as well. Similarly, make sure that we're looking at the entirety of the marketplace to get the right tire at the best price possible. That's just two examples, there's a lot more we can do with technology to leverage our efficiency in terms of labor and productivity and how we move our inventory from raw to WIP to being on the lot at the end of the day.”
— Keith Barr, Q1 FY2027
“the biggest opportunity that we have at CarMax is to really just digitize our reconditioning processes, update the processes as well within there. We think that there's a fair bit of upside when it comes to cost and speed just by leveraging technology more strongly in our reconditioning.”
— Enrique Mayor-Mora, Q1 FY2027
“I was really impressed with when I came into CarMax was the caliber of the marketing talent we have from a data analytics perspective and the way that they're focusing on high ROI marketing and in being real-time.”
— Keith Barr, Q1 FY2027
ANALYST QUESTIONS ON AI
Q (Q4 FY2026, David Bellinger (Mizuho)): as you implement some of these new tools, even some AI tools, is there an opportunity to operate the business with simply less inventory while still giving that core customer the breadth and depth that they need?
A: Enrique Mayor-Mora did not address the AI tools dimension directly; he responded only on inventory levels: 'that certainly is a key aspect that we need to consider, and we need to balance what is the right amount of inventory kind of by market, how quickly can we get it to customers. And I expect that will be a key component of our strategic deliberations as well, making sure we have the right amount of inventory. Is it less? Is it more? We'll end up seeing what that looks like.' The AI framing in the question was not picked up or addressed.
Q (Q1 FY2027, Jeff Lick (Stephens)): In the context of dynamic pricing / revenue management: 'you have a background in pricing and revenue management in my previous life and understanding the similarities and the differences between the two... I'm just wondering how you're thinking about that now in the context of the used car business, your business, and the data that you're seeing now.'
A: Keith Barr: 'we've got a fantastic data sciences team, and we're continuing to invest in that space and expand upon how we think about our algorithms and evolve them over time... bringing in that external market data into our pricing algorithms, understanding for individual types of vehicles into our pricing algorithms. Where do we have pricing flexibility, where we can maximize sales, and where should we actually hold firm in our pricing so we can maximize profitability?' No specific AI terminology used, but described an algorithmic, data-science-driven pricing evolution.