$CAT · Caterpillar Inc.

Thesis

CAT prints Tue BMO into one of the most-stretched setups in the industrial tape — stock +43% YTD and +88% over the trailing year at a forward P/E ~33x on a cyclical, because the market has repriced it as an AI-power derivative. NY Executive Order 62 (Jul 14) paused >50MW data-center permits and Baird downgraded to Neutral + cut PT $300 in response. Options are pricing ±6.5% (2x typical realized), positioning is one-sided long, and a clean beat is fully priced in. Base case: modest EPS beat to ~$6.35, guide reaffirmation without discrete Power Gen upside amplification, and any state-level permit-contagion silence from management prints -3 to -5% on the DE 5/15 sector-multiple-cut template.

What would flip it

Power Gen backlog surges to >$70B and management raises FY guide with explicit 'NY EO 62 contained' language — the DOWN call inverts to +3 to +5%.

The market's narrative

Caterpillar enters Q2 as one of the most-stretched industrials in the tape — stock +43% YTD, +88% over the trailing year, trading at a forward P/E of ~33x on an industrial cyclical because the market has repriced it as an AI-power derivative (Power Generation +41% YoY Q1, $63B backlog, $725M large-engine capex expansion, 2.1GW PROPWR data-center supply deal). NY Executive Order 62 (Jul 14) paused >50MW data center permits, Baird downgraded to Neutral and cut PT $300 — the multiple is now entirely dependent on whether the NY moratorium spreads and on tariff pass-through in Resource Industries.

Where the Street may be wrong

• Forward P/E ~33x on an industrial cyclical is 2x the 5yr average — this multiple only holds if Power Gen keeps compounding at 40%+ YoY. Q2 print needs to CONFIRM the Power Gen trajectory AND expand backlog visibility beyond $63B to hold the multiple. Simple clean beat = flat, not up.
• NY EO 62 is a state-level catalyst that COULD spread — the market will read any Q2 commentary on state-level data-center permitting risk as either 'contained to NY' (multiple holds) or 'contagion risk' (multiple compresses). Silence from mgmt on the call = tape assumes contagion.
• Options implied ±6.5% is roughly 2x the 4Q realized reaction — market pricing for a fat-tail event, most likely DOWN given the position of the stock. Historical CAT reactions at these multiples fired negative 4 out of the last 5 times a Power-Gen-narrative name printed at parabolic-highs.
• Tariff impact trim from $2.6B to $2.2-2.4B for 2026 was already-known — that's in the tape. The forward color that MATTERS is FY27 tariff planning, and any 'we're preparing for a wider tariff regime' comment is a -3 to -5% signal that Resource Industries margins get squeezed.
• Baird downgrade + PT cut $300 8 days pre-print is a HUGE analyst tell — Baird is a serious industrials shop and their move was NY-EO-62 specific. Rating mix 15/11/2 still Buy-tilted but the trajectory is deteriorating heading in.

Peer read: DE Q2 (May) beat but CUT FY guide on ag weakness — construction/mining diverging from ag but the sector-multiple-cut precedent is in the tape. PCAR trucks orders soft on freight cycle. ETN/GE Vernova riding same AI-power thesis at similar-stretched multiples. Closest cautionary template: DE 5/15 (-6% on beat + FY guide cut), applied to CAT because both are cyclical-industrial names that repriced on secular narratives.

Reasoning

• Base-case print: EPS $6.35 (+1.6% vs $6.25 street), rev $19.50B (+1.0% vs $19.31B). Beat driven by Power Gen +40%+ continuing, Construction Industries steady, Resource Industries tariff-cushioned. Clean beat is priced in.
• Forward P/E ~33x on an industrial cyclical + stock +88% 1yr + Baird downgrade 8 days pre-print + NY EO 62 fresh overhang = the exact hype_high configuration that gets punished on clean beats. This is the AVGO conglomerate-segment-gating parallel applied to an AI-power-derivative multiple.
• Options implied ±6.5% (2x typical) = market pricing a fat-tail event. In the absence of a discrete UP-amplifier (NY EO 62 reversed, new $1B+ data-center supply deal, tariff FULLY absorbed), the fat-tail skew is DOWN because positioning is one-sided long.
• Guide raise on Power Gen backlog (>$65B new backlog) = the ONLY thing that saves this from selling off. Reaffirm-only = -3 to -5% base case. Any language about state-level data-center permit risk contagion = -6 to -8%.
• 3-day exit at -5% (vs -3.5% 1d) reflects the DE 5/15 template where sector-rotation-out-of-AI-power-industrials extended across 3-4 sessions as ETF flow rebalanced.