TD Cowen warns that Microsoft might already have too many data centers

Microsoft’s boom in AI data center spending may have gotten out over its skis, according to a TD Cowen report published Friday that’s causing aftershocks across Wall Street and all the way over in Europe.

“Our recent channel checks indicate that Microsoft has terminated select leases with at least two private data center operators across multiple US markets to the tune of ‘a couple of hundred MW,’” wrote analysts led by Michael Elias, also saying that some of its real estate management tactics mirror measures that Meta took while curtailing its metaverse-related capex binge. Separately, they flagged that Microsoft may be reallocating some of its international spending budget to the US.

“Our initial reaction is that this is tied to Microsoft potentially being in an oversupply position,” they concluded.

European stocks with a link to the AI data center trade are taking one on the chin on Monday.

Chatter around this report, as well as the monthly options expiry, likely contributed to the magnitude of the losses across chip stocks on Friday, with Nvidia, Micron, Intel, and Microchip Technology all down 4% to 5%.

Of note: these are communication infrastructure analysts; Elias & Co. don’t cover Microsoft, but rather the likes of telecom infrastructure companies like American Tower, Crown Castle, and Equinix.

Last October, Microsoft CEO Satya Nadella said that supply constraints weighed on the company’s revenue generation from AI services “because data centers don’t get built overnight.” That message on the need to scale up was affirmed by the company’s massive capital outlay plans.

A Microsoft spokesperson reiterated Microsoft’s plans to spend over $80 billion on infrastructure this year, though they noted that the tech giant “may strategically pace or adjust” such spending.

The continued willingness of megacap tech companies to spend hundreds of billions on capex has been a massive boon for S&P 500 profit generation, and is probably one of the most critical factors to monitor in terms of judging whether the AI boom has gotten a little long in the tooth.