Comfort Systems USA, Inc. (NYSE:FIX) HOLD on all three horizons
43.63× trailing earnings against a warranted 16.9× keeps the Valuation Ceiling gate triggered, and a +9.9% base case is not compensation for a −39.1% tail.
Comfort Systems is the largest pure-play mechanical, electrical and plumbing contractor in the US. About three-quarters of its revenue is technology and industrial work.
What actually happened
There is no new fundamental information in this cycle, and it is worth saying so plainly rather than manufacturing a development. The second-quarter results were filed on the twenty-third of July, two days before the last report, so they were already in it. What moved was the tape. The stock fell a further twelve point three percent to close at fifteen nineteen ninety-six on the twenty-ninth of July, then recovered sixteen point seven percent close to close on steadily higher swing lows with momentum positive for ten consecutive sessions. Timing rose twenty-five points on that alone. The other four pillars moved by five points or less between them.
Higher swing lows from 1,491.24 on 29 Jul to 1,710.02 on 13 Aug
The quality was never the question
On the business itself the re-test came back one point better, at eighty-six. Return on invested capital is around forty-four percent against a cost of capital of eleven point one three. Revenue grew fifty point three percent year on year and diluted earnings per share ninety-one point nine. The balance sheet carries one point eight billion dollars of net cash and interest cover of a hundred and ninety-six times. And the backlog is fourteen point zero six billion, up seventy-three percent, of which ninety-seven percent is same-store rather than acquired — roughly fifteen months of signed work already booked. Nobody disputes the quality of this business. That is not what the signal turns on.
Revenue +50.3% YoY and diluted EPS +91.9%
The gate was re-run, and it still fires
Now the part that has not moved. The Valuation Ceiling gate was re-run and still triggers: forty-three point six three times clean trailing earnings against a warranted sixteen point nine, a ratio of two point five eight. The industrials guardrail sits at twenty-three times and is cleared on the trailing multiple, on this year's forward and on next year's alike, so the Expensive band does not depend on our discount rate. More importantly, the deep-expensive do-not-buy prohibition was re-adjudicated: both of its numeric legs are satisfied, and the only thing holding the prohibition off is the carve-out for exceptional, proven, durable growth. That judgement rests on contracted backlog rather than forecast, and it is the most load-bearing call in the report.
The price is 114% above the anchor's fair value of US$828
Why HOLD and a +16% target agree
The street's consensus target is two thousand and fifty-seven dollars, about sixteen percent above the price, and this report says hold. Those are not in conflict. Our base case is nineteen fifty at fifty-five percent, deliberately below consensus, because we do not think a forty-three times trailing contractor holds its multiple through a de-rating market. That is nine point nine percent of upside. The bear is ten eighty at twenty percent — minus thirty-nine point one. The number that matters is not the six and a half percent weighted value, it is the shape: about ten points of base-case upside against a thirty-nine point tail, on a business whose quality nobody disputes. No entry path is open.
Probability-weighted value US$1,888.50 — +6.5%
What could go wrong
The bear needs any two of three things. First, the cohort de-rating: this is a genuine member of the artificial-intelligence capex cohort, so it inherits a multiple compression from forty-three times toward twenty, which on reduced earnings is ten eighty. That is a forty to fifty percent move, not a wobble — though the macro report has the trigger currently receding as market breadth broadens, which is why it carries twenty percent rather than a prohibition. Second, a hyperscaler capex pause; the calibration for that is the late-July episode, minus twelve point three two percent across three sessions on sector-wide capex scrutiny, not on anything the company did. Third, competitive margin compression as rivals bid the same campuses. And a standing caution: thirty-five insider sales worth a hundred and fifty-two million dollars over twelve months, against zero purchases.
Risk vs Reward
Bull twenty-four hundred at twenty-five percent, plus thirty-five point three percent, and note it requires the multiple to hold rather than expand — hyperscaler capex confirmed above a trillion dollars for twenty twenty-seven and backlog through eighteen billion. Base nineteen fifty at fifty-five percent, plus nine point nine: earnings grow about twenty-three percent, the multiple gives back about ten, and the shareholder keeps the difference. Bear ten eighty at twenty percent, minus thirty-nine point one. Probability-weighted, that is eighteen eighty-eight fifty, or plus six and a half percent.
The verdict
Hold, on all three horizons, unchanged. This is a great business at the wrong price, and re-examining it this cycle did not change either half of that sentence — quality came back a point higher, the valuation gate came back still triggered, and the only pillar that genuinely moved was the tape repairing itself after a breakdown. What we are watching is a bookings number: the thesis-invalidation floor is sequential backlog at or below fourteen billion dollars, because the durability argument that keeps this out of a hard prohibition is built on contracted work rather than forecast. The next report is due the thirty-first of August.
It is a quantitative framework for educational purposes only, and it is not financial advice. Always do your own research.
Read the full report on donatien.ca →
