I'll research Caterpillar before writing the preview. Let me start by gathering recent filings, transcripts, and news.Interesting — CAT has become an AI/data center play. Let me read the latest earnings materials and get price data.I have Q1 2026 detail. Now let me get the year-ago Q2 2025 for comparison, recent stock price, and consensus estimates.Let me get consensus estimates and the latest news context.Let me fix the web search and get consensus estimates.The result is a JSON string. Let me parse it properly.I now have what I need. Let me compile a few reference figures before writing.Here is the earnings preview.
Caterpillar has been re-rated over the past year from a cyclical industrial into a de facto AI/data-center power play, and this print is the first real test of whether that narrative can survive both a stretched valuation and freshly emerging concerns about data-center demand durability. Wall Street, on average, expects the company to post quarterly EPS of $6.20 (+31.4% Y/Y) on revenue of $19.2B (+15.7% Y/Y) — with Zacks slightly higher at ~$6.25. The bar is high, and the stock is entering the print in a fragile technical position.
The bull case now lives almost entirely in the Power & Energy segment (turbines, reciprocating engines, oil & gas, power generation), which management has aggressively leaned into:
What to watch Q2: Order intake and backlog trajectory (does it build again off $63B?), the prime-vs-backup power mix, capacity ramp progress (management said it was tracking ahead of schedule), and any new gigawatt agreements. This is the number the market will trade on.
The AI-power thesis took its first real hit just days before the print. Baird analyst Mig Dobre downgraded CAT to Hold from Buy and cut his price target to $900 from $1,200, citing mounting regulatory risks to data-center expansion — including New York State's moratorium on data-center construction — which he argued raises costs, adds approval hurdles, limits site availability, and likely slows future investment. Caterpillar stock had become an AI darling due to its power-generation products supporting the data-center build-out, but cracks are appearing as investors question whether hyperscaler capital spending can sustain growth. Separately, analysts flagged Middle East geopolitical developments as a potential overhang.
The tension is clear: sentiment coming in is very bullish given the transition to an AI/data-center story, but the valuation is now considered quite rich and Power Generation remains a relatively small portion of overall revenue.
Construction Industries — the largest revenue segment and the swing factor on the "AI is only a small % of revenue" critique: - Q1 was a blowout: sales +38% to $7.16B, margin 21.4% (+160 bps) — but that was heavily flattered by a ~$1.5B dealer inventory build vs. a slight drawdown a year ago. Underlying sales to users grew 7% (fifth straight quarter of growth). - For Q2, management guided to minimal further change in CI dealer inventory and a more typical sequential build — so the Y/Y growth optics should normalize sharply. Watch sales-to-users (the cleaner demand signal), North America non-resi/infrastructure (IIJA + data-center construction), and price realization. - Note the year-ago comp is easy: Q2 2025 CI sales fell 7% to $6.19B on unfavorable price and a dealer inventory drawdown.
Resource Industries — the profit-margin wildcard: - Q1 margin cratered to 10.0% (–700 bps Y/Y) on tariffs and timing. Management explicitly guided RI margin lower Y/Y again in Q2, with price realization improving as the year progresses. - The offset: mining order intake was the strongest since 2012, driven by copper and gold, plus an aging fleet. Watch whether backlog strength starts translating into margin recovery.
Tariffs remain the single biggest earnings headwind and the reason margins look compressed despite strong volumes: - Q1 2026 tariff cost was ~$600M (better than the $800M feared). For Q2, management guided ~$700M (vs. ~$400M a year ago), split roughly 50% CI / 25% P&E / 25% RI. - Full-year 2026 tariff estimate was trimmed to $2.2–2.4B (from $2.6B), partly reflecting the Supreme Court IEEPA ruling and a shift to Section 232/122. - Guidance framework: adjusted operating margin near the bottom of the progressive target range including tariffs, but in the top half excluding them. By segment for Q2: P&E margin slightly higher Y/Y, CI higher Y/Y, RI lower Y/Y.
This is the crux of the setup. The stock has been on a historic run and then rolled over hard:
Given consensus of ~$6.20 EPS / ~$19.2B revenue and the recent 20%+ pullback, expectations have cooled somewhat — but the valuation still demands a clean beat and reassuring forward data-center commentary. The asymmetry looks skewed toward outsized volatility either direction.
Note: figures on prior quarters and guidance are drawn from Caterpillar's Q1 2026 and Q2 2025 earnings releases and the Q1 2026 call transcript; consensus, price-target and analyst-sentiment figures are from public news/analyst commentary dated within the last few weeks. Reminder that beginning in 2026 CAT reports under a revised segment structure (Power & Energy, Construction Industries, and Resource Industries — the latter now including Rail), so year-ago segment comparisons use recast historicals.