CAVA Is Buying Its Next 500 Restaurants With Today’s Margin

CAVA Is Buying Its Next 500 Restaurants With Today’s Margin

CAVA’s third-quarter report lands with the stock already in a deep drawdown: shares are down 21% over the last 90 trading days and 59% below the December 2024 peak. At a $7.2 billion market cap, the company is no longer priced like a flawless category killer. The market has taken 3.8x revenue as a verdict on decelerating growth. The quarter itself was not a demand collapse: revenue rose 22.1% to $353.7 million, 24 net new restaurants brought the base to 500, and same-restaurant sales grew 3.1%. But 3.1% is a different CAVA than the 9% comp discussed on the August call, and that is the tension the tape is trading.

The comp miss has a specific cause, and management is careful with the language. CAVA booked a 150-basis-point hit from cyclospora and broader concerns around fresh-produce consumption early in the quarter, followed by a gas-price spike after Labor Day. The CEO framed it as category-wide, not company-specific:

The impact of broader concerns around the consumption of fresh produce at the beginning of the quarter followed by the reemergence of high gas prices impacted traffic across the category.

The important part is what happened next. CAVA says traffic turned positive again as the quarter closed and those pressures abated: “We saw that bounce back, as we noted at our Q2 earnings call, it had rebounded. And then post-Labor Day, we did see some choppiness as gas prices spiked, but that abated as we exited the quarter.” — Unidentified Speaker · 2026-11-04 That is a more constructive exit-rate message than the headline comp suggests. It also contrasts with the prior call, when the company explicitly quantified the duration risk of the outbreak: “our guidance incorporates what we're seeing to date as well as macroeconomic and geopolitical uncertainty, but it's a very prudent assumption regarding the duration of the Cyclospora impact itself and what that will mean for the business.” — Tricia Tolivar, Chief Financial Officer (CFO) · 2026-08-11 The named crisis has become a generic consumer concern. That is a rhetorical shift, not a denial.

The investment line is the real story

Restaurant-level margin was 23.4%, down from 24.6% a year ago. Labor rose 50 basis points on a 3% wage investment, the expanded AGM role, and extra hours in new restaurants; other operating expenses rose 70 basis points on delivery mix and repairs and maintenance. Pre-opening expense jumped to $8.4 million from $4.9 million. On the call, an analyst noted that full-year pre-opening guidance moved roughly $7 million higher while the unit-opening guide rose by only one restaurant. That is the tell: CAVA is spending on the quality of each opening, not just the quantity.

CFO Tricia Tolivar said the company is adding roughly $50,000 to $60,000 per restaurant, mostly by bringing general managers and team leads in earlier and sending them to other openings before their own. “We are bringing in general managers earlier, team leads earlier, traveling to other openings to experience a new restaurant opening ahead of what the restaurants that they're going to open themselves.” — Unidentified Speaker · 2026-11-04 This is the pre-opening spend that management believes buys lower GM turnover and better guest experience. Brett Schulman said turnover among trained opening GMs has dropped to almost zero, and Academy GMs run low-to-mid-single-digit turnover. The payback is visible in people metrics; it is not yet visible in comps.

That matters because the market is no longer paying for unit growth alone. Operating margin has still improved to 8.3%, but free cash flow margin is only 6.6%, and capex was $41 million in the latest filed quarter. The company is trading current margin for future unit economics, a legitimate strategy only if the new stores keep earning their keep.

Project Sunkist and the dining-room bet

The freshest company-specific theme is not value-menu discounting. It is physical experience. CAVA’s Project SOL prototype is now being extended to older restaurants through a new initiative called Project Sunkist. Management says one urban remodel produced an 18% increase in the dinner daypart. That is early, but it is a real sales lever that does not require a new unit.

This is a deliberate contrast with peers. McDonald’s is talking about entry level value; YUM is leaning on its value platform. CAVA has value too — its entry bowl is $10.65, and lower-income cohorts continue to post the strongest same-restaurant sales, with premium-protein attachments indexing higher than one might expect. But Schulman is explicitly investing in dining rooms while others shrink them: “I've been on record for years saying the demise of the dining room is greatly exaggerated.” — Unidentified Speaker · 2026-11-04 That is a differentiated bet, and it explains why restaurant level margin is being asked to carry remodel and pre-opening costs now.

The operating backbone is quietly new

CAVA also introduced a set of operational tools that barely registered in prior quarters: a CAVA Command Center that gives general managers a single view of restaurant vitals, an operator portal that lets them control out-of-stock items and digital delivery throttles without filing a tech ticket, and a broader push behind Epic Hospitality. The company says the portal is already in more than 100 restaurants. This is not AI hype; it is blocking and tackling aimed at consistency across a 500-unit fleet that wants to be 1,000 by 2032.

It also fits the longer pattern. In prior calls, CAVA talked about new restaurant productivity holding near 100% and about using data to drive labor deployment and guest personalization. The new layer is execution: fewer manual workarounds, more real-time control. If it works, it should protect throughput as AUVs grow past $3 million and the company pushes into urban locations with longer development timelines.

The catch

Guidance is intact but not heroic: 76 to 78 net openings, full-year comps of 6% to 7%, restaurant margin of 23.7% to 24%, and adjusted EBITDA of $184 million to $189 million. For 2027, management is pointing to at least 16% unit growth and a low-to-mid-single-digit comp algorithm. The balance sheet is a fortress — zero debt, $427.9 million in cash and investments, a $100 million buyback authorized but unused — and the highest-return use of capital remains new restaurants at over 40% year-one cash-on-cash returns. “Our highest and best use of capital remains opening new restaurants, which have been generating over 40% Cash-on-cash returns in year 1.” — Unidentified Speaker · 2026-11-04

The risk is timing. If Q4 comps do not confirm the exit-rate improvement, CAVA will be spending pre-opening dollars and labor hours into a softer top line, and the market will read the margin pressure as structural rather than strategic. The company has been consistent on the alternative: do not buy traffic with discounts. “we're not going to get into that heavy discounting to combat any cyclical headwinds.” — Brett Schulman, Co-Founder and CEO · 2025-11-04 It has also consistently said lower-income cohorts are holding up better than the industry narrative would suggest: “we have not seen any shift in our consumer behavior... the lower end of the income strata is actually performed the strongest.” — Brett Schulman, Co-founder and Chief Executive Officer · 2026-05-19 That is still the bull case. The bear case is that a 3.1% comp, a 21% three-month drawdown, and a 59% peak-to-trough decline are telling investors the investment cycle is arriving before the demand recovery. This quarter does not settle that argument; it just makes the next two quarters unusually load-bearing.