Wells Fargo: AI spending is "spilling over" into the real economy, with industrial stocks being the biggest beneficiaries, but caution is needed regarding political resistance.

Wells Fargo strategist Ohsung Kwon recently pointed out that the AI spending by tech giants is "dripping over" into the broader real economy, with industrial stocks becoming the major beneficiaries. This assessment is not a vague optimistic expectationanalyzing macro data, corporate earnings reports, and policy dynamics, the spillover effects of AI investment are indeed happening, but they also face increasingly severe political resistance. The spillover effect is not just a slogan: data is validating it. Kwon's judgment is supported by solid data. According to Wells Fargo, manufacturing activity expanded at its fastest pace in over four years in July, with non-AI-related capital expenditure increasing by 10% year-on-year and the growth rate of commercial and industrial loans also accelerating significantly. The industrial sector of the S&P 500 index has risen a cumulative 20% this year, trailing only energy and information technology. A more macro perspective confirms this trend. According to ConstructConnect data, spending on the construction of data centers across the U.S. in the first five months of 2026 has reached $58.1 billion, more than four times that of the same period in 2025; the first quarter saw startup spending of $46.5 billion, a year-on-year increase of over 500%. The AI boom has completely transformed the U.S. economy, with capital expenditures from hyperscale cloud service providers expected to reach about $750 billion this year. IDC anticipates that global AI infrastructure spending will hit $497 billion in 2026, a year-on-year increase of about 56%. Who is benefitting? Orders pouring in from Caterpillar to Vertiv. The companies "selling shovels" are the first to reap the benefits. Caterpillar announced on Tuesday that driven by demand for power equipment and construction machinery due to data center construction, its quarterly revenue exceeded $20 billion for the first time, a 24% year-on-year increase, with sales in its construction industry segment jumping by 35%. The company then raised its guidance for annual revenue growth. Data center infrastructure providers are similarly benefitting. Vertiv has backlog orders surpassing $15 billion, with fourth-quarter orders up 252% year-on-year; Eaton's electrical division reported about a 50% year-on-year increase in data center revenue in the first quarter. Wells Fargo itself estimates that capital expenditures from hyperscale cloud service providers will reach $1.1 trillion by 2027, about 25% higher than market consensus. Kwon estimates that there are currently about 40 hyperscale data centers under construction in the U.S., with over 100 in planning, concentrated in Texas, Georgia, Virginia, and Pennsylvania. Concerns: Political resistance is becoming the biggest variable. However, Kwon explicitly warns: "The biggest risk to data center construction is political backlash, especially with the midterm elections approaching." This risk is rapidly becoming evident. A Gallup poll shows that 71% of Americans oppose the construction of AI data centers in their area; 77% are concerned that AI will drive up electricity prices. In July of this year, opponents launched 142 protest actions across 42 states. In just the first quarter, local groups blocked or delayed 75 projects, involving an investment of about $130 billion. There is even greater pressure from the policy level. In July, New York State signed the nation's first one-year moratorium on large AI data centers; gubernatorial candidates in at least 12 states have expressed support for pausing data center construction in the midterm elections. Tax incentives are also rapidly fadingfour states have already canceled or suspended tax incentives for data centers, and nine more are under consideration. Based on a 7% sales tax, the procurement cost of equipment for a single 1 GW data center will increase by about $3 billion. Although the Trump administration introduced a "Utility Ratepayer Protection Commitment," inviting tech giants to sign agreements not to pass on grid upgrade costs to residents, this commitment is not binding, and whether it can alleviate public backlash remains uncertain. Wells Fargo's judgment reveals a reality that is unfolding: the dividends of AI spending are spreading from semiconductors and cloud service providers, injecting growth momentum into traditional industrial sectors. However, the sustainability of this narrative increasingly depends not on technology or capital, but on politicshow the midterm elections reshape data center regulatory landscapes will be a key variable in determining how far the "trickle-down effect" can go. As Kwon stated, we might still be in a "very, very early stage"but early stages are often the ones with the greatest uncertainties.