Initiated 2026-09-29 · Price $52.20 (as of 2026-09-29 intraday, stockanalysis.com) · Mkt cap $6.10bn · Consumer discretionary / fast-casual restaurants · Model: verified

Rating: Sell — Conviction: High

CAVA is executing well: Q2 same-restaurant sales rose 9.0% on 5.3% traffic, far ahead of Chipotle, and it opens ~17% more units a year with no debt. The stock is already ~46% below its 52-week high of $98.79, but at 44.6x lease-excluded FY2025 GAAP EBITDA and ~79x forward earnings it still prices in more than the bull case. With a 1.74 beta, a 5.2% 10-year yield and guidance implying slower comps in H2, too much of the value sits beyond year five.

Conviction tests (3a-v-c): T1 pass (base −56.1% at +1pp, −51.8% at −1pp; Sell at both) · T2 pass (Gordon −82%, exit-multiple −26%; Sell on either) · T3 pass · T4 pass

The rule's High probably overstates certainty, because both methods are five-year DCFs that can't fully credit a unit runway the market is valuing out to ~15 years.

Model value range vs price
Bear $10.93Base $23.99Bull $45.01Price $52.20

Business overview

CAVA runs 476 company-owned Mediterranean fast-casual restaurants (2026-07-12). Guests build bowls and pitas priced around the fast-casual premium tier. The small CPG line (dips and spreads sold in grocery) accounts for ~$3mm of Q2's $368.4mm revenue. There are no franchisees, so revenue is restaurant sales. Three variables drive earnings. The first is unit growth: 72 net openings in FY2025 and 75–77 guided for 2026, on a 439 base. The second is same-restaurant sales, which were +4.0% in FY2025, +0.5% in Q4 2025, +9.0% in Q2 2026, and are guided to +4.5% to +6.5% for FY2026. The third is restaurant-level margin: 25.0% in FY2024, 24.4% in FY2025, 25.7% in Q2 2026, and guided to 23.7–24.3% for 2026. AUV is $3.1mm.

Competition

The main competitor is Chipotle, which fights for the same customer and lunch occasion and is where pressure on CAVA's traffic would show first. In Q2 2026 (to 2026-06-30), Chipotle's comparable sales rose 2.2% on 1.0% transaction growth, while CAVA's rose 9.0% on 5.3% traffic. CAVA is taking relative share for now. Sweetgreen and Shake Shack compete for the same premium lunch spend. Sector-wide softening shows up in traffic before price. August Placer.ai foot-traffic data softened across restaurants (Simply Wall St). A Cyclospora outbreak traced to Mexico-sourced iceberg lettuce weighed on fresh-produce chains; CAVA's own exposure is n/a (unverified). The closest listed comparables are all cheaper: Chipotle trades at 19.8x EV/EBITDA and 25.1x forward P/E, and Dutch Bros, the high-growth one, at 25.0x and 32.1x.

Bull case

  1. Unit runway — ~17% unit growth on a self-funded balance sheet, with AUVs at $3.1mm, gives a long compounding runway. Plays out if new units keep opening near 90% productivity and openings hold ~75+ a year, with comps of +3–4%. Model: rev_growth
  2. Margin leverage — restaurant-level margin reached 25.7% in Q2. G&A leverage on a 25%+ revenue growth base could lift the GAAP EBITDA margin from 10.9% to ~14%. Plays out if food and labour inflation stay contained and new units ramp on schedule. Model: ebitda_margin
  3. Growth premium holds — traffic-led share gains keep the multiple well above restaurant peers, and the beta normalises as the company matures. Plays out if comps beat Chipotle's by 4+ points for several more quarters. Model: exit_ev_ebitda, wacc

Bear case

  1. Duration at a 5% risk-free rate — at 44.6x EBITDA, most of the value lies beyond year five. CAPM gives a 13.9% cost of equity (5.21% + 1.74 × 5.0%), and a de-rating toward peers is the base path. Plays out if the 10Y stays ≥5% and growth multiples keep compressing. Model: wacc, exit_ev_ebitda
  2. Comp deceleration — the full-year guide of +4.5% to +6.5% after a +9.0% Q2 implies H2 comps near the low single digits. Q4 2025 was +0.5%, and August traffic data were soft. Plays out if comps go flat to negative for two years and openings slow. Model: rev_growth
  3. Unit economics dilute — restaurant margin fell 60bp in FY2025 and is guided lower again. Guidance assumes new units at ~90% productivity, so each opening needs the same capex for less revenue. Plays out if new-market AUVs undershoot. Model: ebitda_margin, capex_pct_rev

Valuation & balance sheet

Metric (definition) Current Own history (range or 5y avg) Peers Source, as-of
EV/EBITDA (model basis: mkt cap − cash & investments, no debt; GAAP EBITDA, leases excluded) 44.6x FY2025 · 37.2x LTM n/a (unverified) CMG 19.8x · SHAK 18.0x · BROS 25.0x (trailing, lease-inclusive, aggregator) Model; stockanalysis 2026-09-29
EV/EBITDA (aggregator, trailing, lease-inclusive) 38.9x n/a (unverified) same as above stockanalysis 2026-09-29
EV/Sales (aggregator, trailing) 4.55x n/a (unverified) CMG 3.63x · SHAK 2.00x · BROS 4.28x stockanalysis 2026-09-29
Forward P/E (consensus) 78.6x n/a (unverified) CMG 25.1x · SHAK 45.6x · BROS 32.1x stockanalysis 2026-09-29
FCF yield (TTM FCF / mkt cap) 0.8% ($49.0mm) FY2025: OCF $184.8mm − capex $158.7mm = $26.1mm n/a stockanalysis; EDGAR

Model-implied value range (from model-summary.json; generic module, Gordon-growth DCF and exit-EV/EBITDA DCF, midpoints): Bear $10.93 · Base $23.99 · Bull $45.01 per share, i.e. implied returns of −79.1% / −54.0% / −13.8% vs $52.20. These ranges show how the bull and bear drivers translate into value; they are not price targets. The price sits above the bull midpoint, so the market is paying for more than the bull case's five years of ~20% growth and margin expansion. The base exit multiple of 22x is a ~51% de-rating from today's 44.6x. The two methods differ widely (Gordon $9.40, exit $38.58). With year-five FCF still carrying growth capex (8% of revenue vs 6% D&A), the Gordon method embeds only ~3.6x year-five EBITDA. The exit method is the fairer read, and it is still 26% below the price. CAVA is a single brand, so I ran a tail sensitivity: FY2026 +22%, then revenue growth of +5% and +3% a year, EBITDA margin 8.5%, capex 6.5%, NWC −5%, a 12x exit and a 15.9% WACC. The bear midpoint falls to $7.47.

Balance sheet: not meaningful (net cash). $435.6mm of cash and current investments, no funded debt; net leverage −3.4x FY2025 EBITDA. $520.7mm of operating lease liabilities is excluded from EV because EBITDA is after rent. A $100mm repurchase authorisation was reported on 2026-09-18 (per Zacks, not tied to a filing this session). No ratings.

Model note: verified: LibreOffice recalculation matched all 3,206 formula cells. Unverified inputs: none. Assumptions without basis: none. No scenario CHECKs. The company doesn't guide EPS. Base Y1 GAAP EPS of $0.61 sits ~7% below the ~$0.66 implied by the consensus forward P/E, mainly because of SBC (GAAP vs adjusted) and the modelled tax path. Cash taxes are a judgment (15% rising to 25%) based on the $51.6mm deferred tax asset. Historicals start in FY2023 because FY2021–22 include Zoës Kitchen conversions. FY2023 NWC was not retrieved.

Scenario stress test

Reasoned from the bull/bear drivers above. The model column comes from the scenario overlays (Base case + shock).

Scenario Effect Mechanism Magnitude Model Δ value vs Base ($/sh)
S1 Fast equity crash − 1.74 beta and 12.6% short interest; multiple de-risks, no operating channel Med −$2.91
S2 Slow bear / recession − Discretionary ~$15 bowl: comps negative, new units under plan, fixed costs deleverage; transitory, so milder than the multi-year bear plateau High −$6.45
S3 Rapid rate shock − No debt; long-duration value re-prices on discount rate and growth multiple Med −$3.83
S4 Slow rate grind − Same channel, grinding Low −$1.50
S5 Soft-landing cuts + Lower discount rate plus consumer relief Med +$3.69
S6 Recession-driven cuts − Traffic loss and growth de-rating outweigh lower rates Med −$3.62
S7a Credit liquidity shock − Self-funded, so no funding channel; forced selling hits high-multiple equities Low −$1.94
S7b Slow default cycle − Developer and landlord financing slows some site deliveries Low −$1.46
S8 Stagflation − Protein, produce and wage inflation vs a consumer resisting more price Med −$3.22
S9a Dollar spike 0 No material effect, not modeled (all US) Low $0.00
S9b Dollar slide 0 No material effect, not modeled Low $0.00
S10 Melt-up + Heavily shorted high-beta growth name: covering and re-rating Med +$3.89
S11 Energy supply shock − Gasoline squeezes discretionary spend; distribution and packaging costs Low −$0.38
S12 Mega-cap/AI derating − Leadership unwind de-rates long-duration growth multiples broadly Low −$1.94

Dollar moves look small because the base value is already less than half the price; relative to that base, S2 is −27%. Currently active/on watch per the playbook: S3 partially active (price leg met); S8, S10 and S11 on watch. S3 is the live one here: CAVA has no debt, so the risk-free rate reaches it only through the discount rate, but that is the channel its valuation is most exposed to.

Model value change vs Base, by scenario
S2 Slow bear / recession−$6.45S3 Rapid rate shock−$3.83S6 Recession-driven cuts−$3.62S8 Stagflation−$3.22S1 Fast equity crash−$2.91S7a Credit liquidity shock−$1.94S12 Mega-cap/AI derating−$1.94S4 Slow rate grind−$1.50S7b Slow default cycle−$1.46S11 Energy supply shock−$0.38S9a Dollar spike$0.00S9b Dollar slide$0.00S5 Soft-landing cuts+$3.69S10 Melt-up+$3.89

What would change the call

Upgrades if: the price falls toward the base range. Or H2 comps hold ≥+5% on positive traffic while restaurant margin holds ≥24% and 2027 unit growth is guided ≥16%, which would move the base toward the bull case. Downgrades if: n/a at Sell. The thesis strengthens if Q3 comps come in below +3%, if 2027 openings are guided below ~75, or if restaurant margin falls below the 23.7% guide floor.

Watch items

Sources