The most interesting story in the equity market right now is not a chip company. It is a company that makes excavators. Caterpillar shares are up roughly 28 percent year to date as of late April 2026, the company reported a record 51.2 billion dollar order backlog on its most recent earnings call, and the trailing price to earnings ratio sits near 37. That is roughly double its ten year median. The market is repricing Caterpillar from a cyclical industrial into something closer to an AI infrastructure provider. The reasoning is straightforward. Data centers run on power and power runs on diesel generators, gas turbines, and grid build out, all of which require the kind of equipment Caterpillar sells.
Caterpillar is the headline name in a broader theme. The S&P 500 industrials sector has outperformed the S&P 500 by several hundred basis points this year. The drivers are three. First is the AI infrastructure capex cycle, where the four hyperscalers alone are expected to deploy more than 470 billion dollars in 2026 against roughly 350 billion in 2025. The dollars do not just buy chips. They buy concrete, steel, copper, transformers, generators, fans, racks, fiber, and the trucks that move all of it. Second is reshoring. A 2026 industry survey found that 74 percent of US manufacturers reported they had moved or were in the process of moving operations back to North America. That move is funded over years and produces steady demand for equipment, training, and logistics. Third is defense capex. The 2027 Pentagon budget request and the supplementals for Ukraine, Israel, and Taiwan have pushed contractors like Lockheed, Northrop, and General Dynamics into multi year visibility windows that few sectors can match.
Not every industrial name is winning. Deere is facing a difficult 2026. Management has guided to roughly 1.2 billion dollars in tariff costs across the year and the agriculture segment is being squeezed by retaliatory tariffs that have depressed export volumes. Construction and forestry backlogs are up about 50 percent, which is the bright spot, but the agriculture book is the larger business and remains soft. The stock has lagged Caterpillar by a meaningful margin year to date for that reason. The lesson is the standard one. Sector tailwinds matter but company specific exposures matter more.
CSX, Norfolk Southern, and Union Pacific are part of the same theme on the rail side. Reshoring and energy infrastructure both produce more freight ton miles, and the rails have spent the last two years investing in capacity. Volumes are growing modestly and pricing has held steady. Rails are not going to look like Caterpillar in the chart but they generate consistent free cash flow and tend to pay rising dividends. Investors who want sector exposure with less volatility have used the rails as a stabilizer.
For retail investors the question is whether the move is over or whether there is room. The honest answer is that the multiple expansion has run faster than the earnings expansion. Caterpillar at 37 times earnings is priced for execution. Any data center buildout slowdown or AI capex pause would compress that multiple quickly. The reshoring trend is more durable, but slower, and is unlikely to drive another 28 percent move in a single year. The industrial ETFs like XLI offer broader exposure with less single name risk and a roughly 1.6 percent yield. Active managers in the space have outperformed the index this year by overweighting power, defense, and aerospace and underweighting agriculture and trucking.
For first generation wealth builders the broader takeaway matters more than the trade. The AI revolution is not just a software story. It is a physical infrastructure story, and the companies building the physical layer are often more boring, more predictable, and more profitable than the companies building the model layer. Some of the steadiest wealth in this country has been built by people who owned shares in industrial companies for thirty years and reinvested the dividends without watching CNBC. That part of the playbook has not changed.
Risk to the thesis sits in three places. A cooling AI capex cycle would compress multiples first in industrials before it reached the chipmakers. A meaningful escalation in tariffs from current levels would damage Deere and other globally exposed names. And a recession in 2026 or 2027 would hit construction equipment demand harder than most other industrial verticals. Those risks are real. The trend is real too. Both can be true at the same time, which is most of what active investing looks like.
The next two weeks of earnings will say a great deal. Caterpillar reports April 30, Deere reports May 15, and the rails report through the first week of May. Watch the order books, the tariff disclosures, and the capex guidance from the hyperscalers. The signal will arrive in the language management uses, not just the headline numbers.