The correlation between the S&P 500 Industrial sector and the Philadelphia Semiconductor Index over a 45-day period has climbed to 0.75, nearing its highest level since last June. This industrial stock rally, driven by AI capital expenditure, is pushing a group of century-old manufacturing-based firms into a trajectory highly synchronized with semiconductor stocks. When NVIDIA releases its earnings report next week, both technology and industrial stocks will face the same pressure test.
Historically on Wall Street, industrial and technology stocks have followed different investment narratives. The former reflects the cyclical pulse of the global economy, while the latter drives the continuous expansion of productivity frontiers. However, the market since 2025 has been quietly rewriting this tradition. The indicator measuring the 45-day correlation between the S&P 500 Industrial sector and the Philadelphia Semiconductor Index has risen to 0.75, close to its highest level since last June. A reading of 1 indicates perfectly synchronized movements, and 0.75 signifies that industrial and semiconductor stocks are moving in step more closely than at any point in the past two years.
Recently, rising chipmaker stock prices have propelled the S&P 500 higher, while industrial companies such as power equipment supplier Vertiv Holdings LLC have also become among the index's top performers, with their trajectories aligning almost perfectly.
"If artificial intelligence is the sole engine driving the stock market and the economy, then any type of obstacle will ultimately have a greater impact on the entire industry," said Michael O'Rourke, Chief Market Strategist at JonesTrading Institutional Services. "Weakness in any key player will have a ripple effect across the entire sector."
Quantitative analysis by Neil Dutta, Head of Economic Research at Renaissance Macro Research, further reveals the breadth of this phenomenon. In a report to clients on May 7, he noted that there are currently 15 non-technology companies, with a combined market capitalization of $2 trillion, whose stock prices "fluctuate under the influence of AI capital expenditure." He warned, "If the AI cycle cools, the drag on consumption from its wealth effect will not be limited to the seven largest tech giants."
**From Diesel Engines to Data Centers: The AI-ization of Order Books**
The root of this high correlation is not difficult to understand. As hyperscale cloud service providers push 2026 capital expenditure to a scale exceeding $700 billion, the physical infrastructure required to build AI data centers—power management, cooling systems, backup generators—almost invariably points to industrial giants known for traditional manufacturing.
Dutta summarized this shift with a pointed observation: "These are not tech stocks. They trade like semiconductor stocks because their order books have become AI capital expenditure order books." He noted that Caterpillar is selling backup generators to data centers, while Vertiv provides cooling and power management equipment.
Data compiled by Renaissance Macro confirms this assessment: Industrial stocks, including Vertiv, Eaton, Caterpillar, and even engine manufacturer Cummins, all show correlations exceeding 60% with the VanEck Semiconductor ETF.
This transmission chain is not merely conceptual. Caterpillar's Q1 2026 earnings report showed total quarterly revenue reached $17.42 billion, a 22% year-over-year increase, significantly exceeding market expectations of $16.61 billion. Adjusted earnings per share were $5.54, far surpassing analysts' expectations of $4.62-$4.63. The biggest growth surprise came precisely from the Energy & Transportation segment—sales of large generator sets and turbines for data centers are rising sharply.
"Caterpillar is indeed benefiting from the AI construction wave," commented Dec Mullarkey, Managing Director at SLC Management.
Similar logic applies to GE Vernova Inc. and Vertiv. GE Vernova reported Q1 revenue of $9.3 billion and adjusted EBITDA of $896 million, both significantly exceeding market expectations. The company disclosed that its data center business alone generated more orders for electrification equipment in Q1 than in all of 2025, with backlog expected to reach $200 billion by the end of 2026, two years ahead of schedule.
Vertiv's Q1 net sales increased 30% year-over-year to $2.65 billion, and adjusted EPS jumped 83% to $1.17, well above the expected $1.00. The company has raised its full-year revenue guidance to approximately $13.75 billion, anticipating 71% growth in hyperscaler capital expenditure to about $650 billion in 2026, with Vertiv being a direct beneficiary as a key provider of power and cooling infrastructure.
Driven by the AI spending boom, the stock performance of these industrial companies has been staggering: Caterpillar, famous for its yellow excavators, has surged over 170% in the past 12 months; Vertiv is up over 260%; and GE Vernova has risen over 150%, all far exceeding the broader market's gains over the same period.
Emily Roland, Co-Chief Investment Strategist at Manulife John Hancock, provided a key footnote: Despite the industrial sector's valuation having climbed significantly, its "magnitude of earnings beats" ranks among the highest of all sectors. She noted that at the start of the earnings season, analysts expected an average earnings growth rate of 3% for the industry, while the actual growth rate reached 20%.
**When Industrial Services Become "Commoditized," High Correlation Means High Contagion**
Yet, it is precisely behind these impressive numbers that the contours of risk are becoming clear. Valuations for the industrial sector have climbed to historic highs. For the first time since 2021, the forward price-to-earnings ratio of industrial stocks in the S&P 500 is higher than that of technology companies.
Philip Straehl, Chief Investment Officer at Morningstar Wealth, wrote on May 7 that the U.S. stock market is "increasingly becoming a concentrated bet on artificial intelligence," as earnings growth has become deeply tied to AI infrastructure spending.
O'Rourke offered a more sober analytical framework: There is a fundamental mismatch between the valuation premium industrial stocks currently enjoy and the basic nature of their business models. "These elevated valuations further expose the group to a potentially difficult landing, as the services they provide are more commoditized than those in the technology sector."
What does this "commoditization" mean? Simply put, when the capital expenditure engines of hyperscale cloud providers are running at full speed, the power equipment and cooling systems needed for data centers seem to be in short supply. However, this pricing power does not stem from technological barriers or patent moats, but from a sudden surge in demand. Once the growth rate of AI capital expenditure slows, the commodity nature of these industrial services will quickly become apparent, putting pressure on profit margins far exceeding that on technology companies.
Historical precedent is not lacking. Straehl warned, "From railroads to fiber-optic networks, history has many examples where businesses built infrastructure based on optimistic forecasts of future adoption rates." He added that in these cases, the primary beneficiaries were consumers, not investors.
If the above represents structural risk, then a series of market signals since the beginning of 2026 provide a more timely warning. As early as early May this year, when news broke that OpenAI failed to meet internal revenue and new user targets, shares of Vertiv and GE Vernova fell 5.4% and 2.8%, respectively, that day. A performance slowdown at an AI software company can severely impact two stalwart American industrial giants—this is the most typical contagion path in a high-correlation market.
Wolfe Research issued a clearer warning on May 11: Investors may be overestimating the growth prospects of AI investment. The firm pointed out that power supply, raw material shortages, and regulatory bottlenecks could constrain the pace of large-scale AI infrastructure construction in the second half of the year. "If the capital expenditure growth rate slows, stocks highly dependent on AI investment—including semiconductors and industrials—will face significant impact."
This forms an unsettling echo with Dutta's judgment from two weeks prior: "Weakness in any key player will have a ripple effect across the entire sector."
From a macro perspective, while Morgan Stanley acknowledges that AI investment will contribute about 40% to earnings growth in 2026-2027, the bank also noted in a March report that market skepticism regarding AI construction financing pressure and investment returns has triggered a "strong rotation beyond large-cap tech stocks." Part of the capital in this rotation has flowed precisely into those industrial companies repriced as "AI concept stocks"—they enjoy the valuation premium brought by capital inflows but also bear the selling risk if the AI narrative reverses.
This judgment has received preliminary verification at the micro level. On April 26 this year, around the time AI correlation reached a cyclical high, Russell Investments' quarterly active manager survey showed equity fund managers were diversifying away from crowded AI leaders and maintaining valuation discipline on cyclical sectors. By the time Wolfe issued its warning in early May, market concerns about slowing AI infrastructure demand were no longer confined to the pure technology sector—the vulnerability of industrial stocks as "AI derivatives" is being factored into pricing considerations by a growing number of professional investors.
**NVIDIA Earnings: The "Non-Tech" Pressure Test for Industrial Stocks**
An upcoming key event will directly test the strength of this correlation iron curtain: NVIDIA is set to release its latest quarterly earnings report. This has never been just a test for tech stocks. Market analysts have explicitly linked the industrial sector to NVIDIA's earnings in research reports, and this week's market movements directly show the market is already pricing in expectations for the company's performance.
NVIDIA previously provided an exceptional Q1 revenue guidance range of $76.4 billion to $79.5 billion, far exceeding the consensus expectation of $72.2 billion. This is underpinned by the grand narrative of big tech companies' 2026 AI capital expenditure reaching over $700 billion.
However, information from the supply chain is not entirely positive. According to the latest data from TrendForce, due to geopolitical risks and supply chain calibration issues, the proportion of NVIDIA's next-generation Rubin series in 2026 high-end GPU shipments has been revised down from a previous estimate of 29% to 22%, while the Blackwell series will account for over 70%. This shift in shipment structure indicates that even within the AI industry, the pace of technological iteration is not entirely stable, and any disruption on the production side could be transmitted to the industrial sector through capital expenditure expectations.
Next week, when NVIDIA's numbers are in the spotlight, they will serve not only as a bellwether for the semiconductor industry but also as a yardstick for the fate of those stocks caught between traditional industry and the AI narrative, bound by high correlation.
Comments