Michael Burry just shorted Caterpillar’s 172% AI rally. One analyst says his bet won’t even matter

Michael Burry disclosed a Caterpillar short after the stock surged ~172% over 12 months and ~77% YTD, saying the shares are overvalued; Caterpillar initially fell nearly 7% after the disclosure and later dropped another ~4% to about $949 (lowest since mid-June). A Freedom Broker analyst pushed back, arguing the move is driven by a structural AI-infrastructure shift toward generator power for data centers, supported by Q1 results (sales +22% YoY to $17.4B, beating expectations). The analyst’s $910 price target implies a potential near-term pullback, contingent on whether hyperscalers sustain aggressive data-center spending and remain healthy on cash flow/debt.

Analysis

CAT is no longer being priced like a cyclical industrial; it is being capitalized as a hybrid AI infrastructure utility. That is the vulnerability: the market is paying for both normal machinery earnings and an embedded option on hyperscaler power buildouts, so any slowdown in capex can compress the multiple faster than EPS can fall. In other words, the short case is less about next quarter’s print and more about the stock’s fragile claim on a growth narrative that is still externally funded.

The cleaner beneficiaries of AI power spend are GEV and Vertiv-like proxies, because they sit closer to the bottleneck and have less legacy cyclicality hanging over them. CAT’s broader mix means it can look like a winner in a capital-expenditure boom, but it also gets hit by dealer inventory normalization and rental softness if the buildout pauses. A slowdown in hyperscaler spending would likely hit CAT first, then spill into NVDA sentiment as the market starts to question the durability of the capex flywheel.

The key contrarian point is that Burry may be early if grid constraints persist for 12-24 months: on-site generation is not a fad if interconnection queues stay clogged and data-center demand keeps outrunning utility upgrades. The near-term catalyst is the next hyperscaler capex and cash-flow commentary cycle over the next 1-3 months; the structural falsifier is continued backlog/order strength in power systems plus guidance that keeps the AI overlay intact. If CAT can reclaim and hold roughly the prior breakout zone around $1,000 while order growth stays firm, the short thesis loses urgency.

Key Decisions for Investors

• Sell CAT rallies into the $980-$1,000 zone; if you want defined risk, use a 1-3 month 950/900 put spread. Upside for the bear case is a retest of the low-$900s; stop if CAT reclaims $1,020 on rising power-systems orders.
• Pair trade: long GEV / short CAT on a relative-value basis for 1-3 months. The thesis is that GEV has cleaner AI-power leverage while CAT is paying an industrial-conglomerate multiple for a narrower AI option value.
• Use CAT as a hedge against overcrowded AI capex longs like NVDA on earnings weeks. If hyperscaler spending guidance softens, CAT should de-rate quickly without requiring a full market drawdown.
• Do not press a naked short if backlog and power-system orders keep accelerating; instead, wait for the next hyperscaler capex update as the real catalyst. The trade is most attractive only if consensus 2026 data-center spend starts to roll over.