Wall Street veteran Ed Yardeni says financial, physical and political constraints could make the AI investment boom more durable, first reported by Fortune. That view comes amid mixed quarterly results, as investors question whether AI returns justify enormous capital expenditures and data centers face electricity shortages and limited grid capacity, while political backlash could slow the build-out.
Some on Wall Street see investors sorting AI winners from losers rather than abandoning the sector.
Alphabet's cash flow sank further into negative territory in the third quarter, and it fell about 12% Thursday, its worst day on record. Microsoft reported strong cloud growth and positive free cash flow but said capital spending would accelerate.
Yardeni identifies financial, physical and political constraints. Investors question whether AI returns justify enormous capital expenditures and whether capital supply can meet hyperscaler demand. If constraints worsen, he warned, stocks could fall further, with AI now a major driver of corporate earnings and GDP growth.
Yardeni says capital flowing less easily could lead to more efficient allocation and lower overinvestment risk. He expects hyperscalers to retain financing because of strong balance sheets and keep spending. Weaker-balance-sheet AI companies, including so-called neoclouds, have seen borrowing costs soar and may have to pull back.
Amazon, Alphabet, Meta, Microsoft and Oracle are projected to invest nearly $800 billion this year and $1 trillion annually from 2027 through 2030.
The investment forecast sits alongside differing financing prospects and practical limits on data-center growth. Data centers have run into electricity shortages and limited grid capacity, while political backlash could slow the build-out.
Oxford Economics economists say a Democratic Congress could step up AI scrutiny and shape debate over regulation, but federal legislation would unfold gradually and a nationwide pause on new data-center projects is unlikely. State and local governments will have greater influence over the build-out.
Oxford says reduced demand for chips and other tech inputs could give consumers some relief on tech-gadget prices. It also says a slower build-out could ease inflationary pressures through the end of the decade and relieve some upward pressure on long-term rates. If AI capex runs hotter for longer, this inflation component would prove stickier, keeping pressure on the Federal Reserve to maintain restrictive monetary policy for longer.