First Watch (FWRG): the model

The model tracks the 586 company-owned restaurants. The 79 franchised ones contribute a flat $12.5M a year (royalties and fees ran $6.0M in the first half of 2026), no growth, no inflation — the ~5 franchise openings a year are ignored. Existing stores keep their current $2.32M average volumes (prices rise with inflation; customer traffic changes by the traffic slider, default flat, compounding across the whole fleet — it affects revenue only). Store margin starts at the margin slider and drifts by the margin-change slider each year — a realistic pairing is ~0.3-0.4pts of margin lost per point of annual traffic decline, since fixed costs don't scale with guests. Openings are either a fixed count per year (the company's stated plan from 2027: ~50 company-owned; the ~55–60 figure often quoted is system-wide including franchisees) or — the default — a constant percentage of the store base (8%), so the opening pace itself compounds: 10% for 20 years takes 586 company stores to ~3,900, far past the stated ~2,200-location US market. New restaurants open 20% below the company's $2.8M target and ramp to it by year three (half-weighted in the opening year for mid-year openings). Store profit is First Watch's reported restaurant-level margin — rent and pre-opening costs already deducted (the reported reconciliation adds back G&A and depreciation but not pre-opening), so the build-cost slider is capex only. The calibration line under the headline numbers shows where that margin puts 2026 EBITDA against the company's own guidance. Corporate overhead starts at 10.0% of revenue — last quarter's actual cash overhead (reported G&A was 10.9%, of which ~0.9% was non-cash stock compensation; that part shows up instead as new shares issued each year worth ~1% of revenue — about $14M, the current run rate — at the prior year's share price, so a cheap stock dilutes more and nothing is counted twice). It declines by the points-per-year you set (default 0.4pt), never below a 4% floor (Chipotle runs ~5.5%, Texas Roadhouse ~3.9%). Renovation spending defaults to 1.8% of revenue — light while the fleet is young (Texas Roadhouse, older and open twice the hours, runs ~2.3%). Cash flow first pays debt down (or borrows up) to your debt target — default 1× EBITDA, since management's stated priority is self-funding and a stronger balance sheet, so buybacks only begin once debt is down — then buys back stock at whatever multiple you set, no cash hoarding. Interest charged is inflation plus the spread you set (default +4.5pts ≈ 6.5%, close to today's term loan at 6.59% and revolver at 7.24%), so debt costs the same in real terms regardless of the inflation slider. Inflation defaults to 2% — set it to 0 to see everything in today's dollars, with real returns. Recession years knock 5% off traffic, 2 points off margins, and halve openings. Store count is NOT capped — past ~2,000 company restaurants you're betting the whitespace is bigger than management's stated 2,200-location market. Taxes: 24% of profit after interest and depreciation (6.3% of revenue, the current rate — so new builds shield income the way they do in real life; the company's unused FICA tip credits, which would cut cash taxes further, are ignored). If the settings drive the company's value below its debt, the share price floors at zero — equity holders can't lose more than everything.

Things to try: move the multiple slider with buybacks on, then off — with buybacks on, the 10-year return cares less what multiple the market pays, because a cheap stock lets the company retire more of itself. The buyback is the gyroscope. With inflation at 0, everything is in 2026 dollars and the returns shown are real (on top of inflation). Deliberately ignored: franchise growth, new dayparts, any multiple re-rating, and anything going right that isn't already proven.