{
  "report_rows": [
    {
      "kpi": "Adjusted EPS (Q2'26, standalone Honeywell Technologies)",
      "prediction": "BEAT",
      "answer": "pred ~$1.85 vs. cons $1.80",
      "confidence": "MEDIUM"
    },
    {
      "kpi": "Revenue (Q2'26)",
      "prediction": "IN-LINE",
      "answer": "pred ~$5.02B vs. cons $5.00B",
      "confidence": "MEDIUM"
    },
    {
      "kpi": "Segment Profit Margin (Q2'26, company-wide automation segments)",
      "prediction": "BEAT",
      "answer": "pred ~20.6% vs. cons 20.3%",
      "confidence": "LOW"
    }
  ],
  "guide_rows": [
    {
      "kpi": "H2 2026 Adjusted EPS guidance",
      "prediction": "UNCHANGED",
      "answer": "guide ~$4.55 (mid of $4.40-$4.70) vs. cons ~$4.50 (H2 2026)",
      "confidence": "MEDIUM"
    },
    {
      "kpi": "FY2026 Adjusted EPS guidance",
      "prediction": "UNCHANGED",
      "answer": "guide ~$8.10 (mid of $7.90-$8.30) vs. cons ~$8.00 (FY2026)",
      "confidence": "MEDIUM"
    },
    {
      "kpi": "FY2026 Free Cash Flow conversion",
      "prediction": "BETTER",
      "answer": "guide ~80% vs. cons ~75% (FY2026)",
      "confidence": "LOW"
    },
    {
      "kpi": "Process Automation & Technology organic growth inflection (H2'26 commentary)",
      "prediction": "UNCHANGED",
      "answer": "guide ~high-single-digit H2 growth vs. cons ~mid-single-digit implied (H2 2026)",
      "confidence": "LOW"
    }
  ],
  "day1_residual_pct": -0.5,
  "day1_confidence": "LOW",
  "day5_residual_pct": -0.8,
  "day5_path": "STABILIZE",
  "day5_rationale": "Guidance was already raised just two weeks before the print (July 8), so most upside is pre-baked into consensus, capping the reaction to any in-line/modest beat. The quarter is mechanically messy (spin accounting, reverse split, Quantinuum deconsolidation, pending WWS/PSS divestitures), which will keep sell-side model rebuilds active for days without a clear directional catalyst. Absent a clean confirmation or disappointment on the PA&T Middle East ramp (the single biggest swing factor), estimate revisions should be modest and two-sided, pulling any initial move back toward flat as analysts normalize for one-time items rather than extending a trend.",
  "day5_confidence": "LOW"
}