Caterpillar's 5% Drop Puts Its Data-Center Power Premium to the Test

Caterpillar fell about 5% to $760.64 while Deere gained 2% and PACCAR slipped only 1%. With the industrial-sector ETF down roughly 1%, the divergence pointed to a company-specific valuation reset rather than a uniform retreat from heavy equipment.

The difference is Caterpillar's exposure to data-center power. Its Power and Energy segment sells engines, generator sets and turbines that can supply electricity on site. Second-quarter sales in that segment rose 17% to $8.24 billion, helping investors treat Caterpillar as an AI-infrastructure beneficiary as well as a machinery company.

Orders support the case

Caterpillar ended the second quarter with a $72 billion backlog, and management said some Power and Energy customers were booking as far out as 2030. Turbine lead times extended into 2029. The company also planned to restart a 10-megawatt gas-engine platform, with first shipments expected in the fourth quarter of 2026.

The official quarterly results add profit evidence. Power and Energy segment profit reached $2.027 billion, up 30% from $1.554 billion a year earlier. That means the demand story was not confined to orders; it was already contributing to earnings.

The share price had priced in a great deal of that strength. The source put Caterpillar at roughly 34 times trailing earnings and well below a 52-week high of $1,071.47 after the decline. A long backlog can support future revenue, but it cannot prevent the multiple from falling when investors demand a higher return.

Costs complicate the premium

Caterpillar expected about $600 million of third-quarter tariff costs in the source material. Official results also recorded $392 million of expected tariff recoveries in the second quarter. Investors should separate underlying price, volume and manufacturing costs from recoveries that can make quarter-to-quarter comparisons noisy.

Deere and PACCAR are useful controls because both are sensitive to financing conditions, but neither has the same data-center power narrative. Their better performance on the day suggests the market was questioning the extra premium attached to Caterpillar, not simply reacting to rates.

The next report should show whether Power and Energy growth, backlog conversion and tariff effects justify the valuation. Orders stretching years ahead are valuable, but delivery timing, customer concentration and project economics determine the cash return.

Caterpillar's selloff does not prove the data-center cycle is ending. It shows that even a strong order book can disappoint shareholders when the stock price assumes the best part of the cycle will last.