Chris Wood, a veteran equity strategist at Jefferies, has suggested that the results of the U.S. midterm elections provided major technology firms with a convenient justification to reduce their capital expenditures. Wood believes that the potential for a Democratic majority in Congress could impose constraints on spending, a development he views as positive for artificial intelligence investments.
In his weekly “Greed and Fear” publication, Wood explained that he prefers “disciplined deployment” of capital over indiscriminate spending. He noted that while AI capital expenditure is forecast to reach nearly $1 trillion this year and approximately $1.6 trillion in 2027, the returns on that investment are declining. The median return on invested capital (ROIC) dropped from 27.2% in 2024 and is expected to fall to 15.7% by 2028.
Wood and his colleague Desh Peramunetilleke identified a “dissonance” between these accelerating returns and declining investment efficiency. They argue that slowing the rapid proliferation of data centers may ultimately benefit the AI sector by preventing malinvestment. Major hyperscalers, including Amazon, Meta, Alphabet, and Microsoft, are now reevaluating their spending strategies in light of the recent election outcomes, which also threaten the tax incentives provided by the Republican One Big Beautiful Bill Act.
Wood is cleverly using the news cycle to cover up their own fear of overbuilding. Classic Wall Street spin.
One point trillion sounds insane, but if returns drop to fifteen percent, maybe they need to rethink the strategy.
Good for the grid, I guess. The strain on local power resources from all those servers was becoming unbearable.
I am skeptical that a congressional majority changes corporate capex. These companies operate on global cycles, not just US politics.
Why do we always call it ‘disciplined deployment’ when it is just cutting spending? It sounds nice but feels hollow.
The tax incentive threat is the real driver here, not the election result itself. Read the small print.
I am surprised Wood frames this as positive. Slower spending usually means slower AI innovation for everyone else relying on it.
If data center construction slows down, does that help the rest of the tech supply chain recover from layoffs?
The ROIC decline is actually scary. We are building a massive bubble that might not sustain itself financially soon.
Honestly, I think this is just convenient storytelling. Tech giants don’t need political excuses to cut costs when margins tighten.