← All forecasts · October 19

FIX / Q3 2026

Comfort Systems USA

Reported: October 22, 2026

Published forecast

2026-10-20T08:25:59.720117+00:00

Opus 5.5 High in Shortcut

Shortcut

13.09

Shortcut normalized EPS

+1.9% vs Street

Off by $0.82 per share

Street

12.85

Street normalized EPS

Frozen October 18 snapshot

Off by $1.06 per share

Real

13.91

Published normalized diluted EPS

Shortcut win

Reported result · October 22 ↗ · Diluted EPS was $13.91; the public normalized actual is also $13.91, consistent with the original forecast convention. · Normalized EPS source ↗

USD per diluted common share. Percentages use unrounded forecasts.

Model download

Explore the assumptions and calculations behind the forecast.

Open in Shortcut ↗Download XLSX ↓

Shortcut’s reasoning.

The Mechanical and Electrical segment build produces normalized diluted EPS of $13.09 versus the Street estimate of $12.85. Under the model assumptions, normalized and GAAP diluted EPS coincide.

Electrical gross margin of 26.2% and a $6m earn-out expense are retained forecast judgments. Applying the exact historical averages instead would lower EPS by roughly $0.03; review preserved the original assumptions.

The guidance comparison uses an acquired-revenue upper bound as a cross-check. It does not drive the scored EPS. The workbook includes the segment detail, source references and retained sensitivities.

On the record.

Reporting basis
normalized diluted EPS; forecast equals GAAP diluted EPS
Forecast completed
2026-10-19T23:38:24.320309+00:00
Published
2026-10-20T08:25:59.720117+00:00
Model assumptions and review notes
  • Electrical gross margin of 26.2% and the $6m earn-out expense are judgments near historical averages, although the workbook labels describe the averages too literally. Applying the exact historical averages would lower EPS by approximately $0.03; the published assumptions are retained.
  • The guidance-implied revenue comparison uses a $150m acquired-revenue upper bound. It is an ancillary cross-check and does not drive EPS.
  • The read-only rendering engine disagrees with the cached overall check and wildcard guidance counts. Independent arithmetic reconciles the underlying check cells; these diagnostics do not drive EPS.
  • Some guidance labels clip or wrap awkwardly, and a few negative-number formats differ. These presentation limitations do not affect the verified EPS calculation.
  • Some webpages were retrieved without archived captures. The original consensus and forecast inputs remain unchanged.
Research sources