REZI — Resideo Technologies, Inc.

TRACKER;2026-09-18;REZI;…;…;4/10;5/10;4.5/10;5/9;SPECULATIV;…%;60

REZI — Resideo Technologies, Inc. · deep-value reanalysis after the thesis broke

September 18, 2026 · price … · common market cap … · NYSE · Industrials / Building Products Reanalysis mechanically triggered: 3 of 7 falsifiers of the 2026-09-03 thesis triggered on 2026-09-17.

Executive summary (1 page: thesis, estimated value, verdict)

On 2026-09-17 the thesis checker reported three falsifiers broken out of seven: marja_operationala (operating margin), datorie_neta_ebitda (net debt/EBITDA), and marja_bruta (gross margin). This resumption’s task is to say whether the assumption that supported the verdict still holds. The answer, verified line by line in the 2026-08-12 10-Q and the FY2025 10-K, is that none of the three measures the business’s deterioration — all three measure a company that hasn’t existed since August 3, 2026 — but that the investigation they triggered found something else, real, that the 2026-09-03 analysis had missed: the owner-earnings base was ~8% too high, and this moves the verdict from “fair price, slightly below value” to “fair price, slightly above value.”

The autopsy of the three, with figures:

• marja_bruta = [0.3004; 0.2882] against the < 0.40 threshold. The two quarters are Q2 and Q1 2026, both consolidated with ADI, the distributor that made 64% of revenue at a 22.7% gross margin ((1,286 − 994)/1,286, 10-Q, ADI segment table). The remaining segment — Products and Solutions — had a gross margin of 43.60% in Q2 2026, versus 42.94% in Q2 2025. The falsifier is correctly calibrated for the post-spin entity and cleans itself up at the 2026-11-05 report.
• marja_operationala = [0.107; 0.0533] against < 0.15. The same contamination, plus one more, in the opposite direction: 0.107 isn’t the reported operating margin. Q2 2026 GAAP operating income is … on … of revenue, i.e. 6.61%; yfinance’s “EBIT” label adds the … of “Other income, net,” of which … is the one-off gain from terminating the Tax Matters Agreement with Honeywell. The falsifier was thus flattered by a one-off item and still broke.
• datorie_neta_ebitda = 4.6096 against > 4.5. The metric is named EBITDA but computes (total debt − cash) / trailing-4-quarter EBIT: (3,622 − 549) / 666 = 4.61. The gross debt of … includes the … of ADIG bonds held in escrow at 2026-07-04, which went to ADIG on 2026-08-03, and doesn’t reflect the … repayment made right after. The 4.5 threshold was set by the prior analysis with 2.93× on adjusted EBITDA in mind — a denominator mix-up.

So: transition noise, not structural deterioration. The positive evidence is the remaining segment’s series, taken from the 10-K’s and 10-Q’s CODM tables: P&S gross margin 38.62% (2023) → 40.95% (2024) → 42.08% (2025) → 42.68% (H1 2026), a 406-basis-point expansion over three years, achieved through tariffs, input inflation, and a weak U.S. residential market. Pricing power, which is the whole moat thesis, is intact.

What did deteriorate, and appears in none of the three falsifiers:

1. The remaining segment’s operating margin fell, from 21.14% (H1 2025) to 18.99% (H1 2026) — 215 basis points. Broken down: restructuring +…, litigation costs +…, R&D +… Without restructuring and litigation, the margin is 21.84% versus 21.22%, i.e. still rising. But segment restructuring ran … → 14 → … in 2023-2025 and … in the first half of 2026 alone, and adjusted EBITDA excludes it entirely.
2. Cash conversion broke in H1 2026: CFO of … on net profit of … Adjusted for the one-off … Tax Matters Agreement payment, underlying CFO is … versus 135 in H1 2025. Working capital absorbed … (receivables 149, inventory 45, current liabilities 50) on revenue growing 4.8%.

The owner-earnings base is thus rewritten from 250 to … FCFE (the bridge in the valuation chapter: adjusted EBITDA of … minus restructuring 45, SBC 32, interest 110, tax 95, capex 85, working capital 20). At …, 5% growth, and an 11% cost of equity — below the 12.0% CAPM implied by a beta of 1.566 — the five triangulated models give a range from … to …, with a median of …, and Monte Carlo over 20,000 scenarios gives a median of …, undervaluation probability …%, and only …% chance of a margin of safety above 30%. The 2026-09-03 analysis gave a … median and …% — and this despite the price meanwhile falling from … to …

Verdict: WATCH, no position. Not because the thesis broke — the underlying business is better than any consolidated figure from the last four quarters shows — but because at … there’s no margin of safety on assumptions I can defend, and the entire gap to the consensus target of … sits in a single parameter: the multiple. The analyst market pays 12.3× adjusted EBITDA at their target; I’m paying 7.9× today, and an 11% DCF refuses the re-rating. The first real checkpoint is 2026-11-05, the first report where ADI appears as a discontinued operation and where the post-spin balance sheet becomes fact, not estimate.