Sunbelt 4.0 was sold as margin progression. Six quarters of segment data show the opposite.

• General Tool decay. NA General Tool adjusted operating margin slid from 35.6% in FY24 to 32.7% in FY25 to 27.1% in Q3 FY26 — 850 bps in six quarters, in a segment that is roughly 60% of NA revenue.
• Specialty cannot outrun it. NA Specialty mix climbed from 33% to 36% at about 100 bps per year and earns ~74% dollar utilization versus General Tool ~48%, yet consolidated adjusted EBITDA margin still printed -259 bps year-on-year in Q3 FY26 to 41.0%.
• The arithmetic doesn't balance. If mix accretion outran segment decay, consolidated margin would expand. It is compressing for six straight quarters.

FY25 record FCF is 88% explained by one capex line management has already started to reload.

Net rental capex fell 39% to $2.18B in FY25 — below depreciation for the first time in five years — and accounts for 88% of the $1.55B FCF improvement. FY26 capex guide already raised to $2.2-2.3B, and the bad-debt provision rate was cut from 8% to 6% of gross receivables, releasing roughly $40M into the year. Treat FY25 FCF as cyclical at $1.2-1.3B, not the $1.79B headline.

Best balance sheet in the listed field. United Rentals is still the ceiling.

• Best-in-listed-field economics. 45.9% EBITDA margin, 12.7% ROIC, 1.6x net leverage with a $4.75B revolver to Nov 2029 — versus HRI at 34.3% margin / 6.4x leverage and WSC at 26.8% / 6.2x. S&P assigned BBB- stable in March 2026.
• Capex flex moat is real. FY24→FY25 capex cut from $685M to $456M (-33%), FCF jumped from $169M to $1.72B with revenue down less than 1%. HRI and CTOS printed near-zero net income in the same cycle.
• URI is denser. URI runs 1,768 branches versus SUNB 1,560 (+13%), $22.5B fleet versus $18.5B, 15% NA share versus 11% — and the gap has held flat for two years. The 20% NA share target requires consolidating the long tail of independents, not taking share from URI.

Multiple at the top of its 10-year band on a year reported earnings actually fell.

• Top-of-band valuation. 22.5x trailing P/E versus a 10-year average near 14x; 8.5x EV/EBITDA versus 6.5x. Forward 18x is only defensible if Sunbelt 4.0 margin progression delivers — which six quarters of segment data is contradicting.
• URI discount partly earned. SUNB 8.96x versus URI 10.07x EV/EBITDA. Bulls frame it as an index-inclusion artifact, but SUNB ROI slid from 19% (FY23) to 14% LTM faster than URI's, and SUNB prints into URI's already-decelerated rate (4.1% → 2.2% in FY25) three months later. Half the discount is fundamental, not technical.
• Hausfeld tail unpriced. March 2025 class action (N.D. Ill. 1:25-cv-03487) names Sunbelt in an alleged Rouse Services pricing-data cartel under Sherman §1. SUNB has not disclosed as material; sell-side carries no haircut. Motion-to-dismiss ruling expected late 2026 — sector-wide if it survives.

Q4 FY26 on June 23 settles three thesis variables in one print.

• Margin trajectory. Consolidated adjusted EBITDA margin ≥44% with FY27 capex guide ≤$2.5B refutes the bear case in a single release. Below 42% with FY27 capex above $2.7B confirms Sunbelt 4.0 margin progression is dead and triggers consensus EPS cuts.
• FCF run-rate. First US-GAAP full-year cash-flow statement. FY26 FCF above $1.8B with credible FY27 capex framing validates the run-rate consensus prices; below $1.5B confirms the $1.2-1.3B forensic mid-case and compresses the buyback math from a one-year sprint to a multi-year program.
• Specialty mix print. ≥+150 bps year-on-year for two consecutive quarters re-arms the structural margin lever. ≤+50 bps closes it. First read lands at Q4 — the second at Q1 FY27 in September.