Potential policy hurdles will slow construction of artificial intelligence data centers without stopping it, leaving opportunities to buy shares, Anna Rathbun, founder and chief executive of an investment advisory, said.
Rathbun said she had begun buying during the selloff and had kept her long-term investment approach unchanged after the midterm elections. She spoke on Bloomberg Technology, published Nov. 5.
Markets had priced in uninterrupted construction of AI infrastructure before the prospect of political obstacles changed expectations. Possible legislation, a moratorium and the loss of tax breaks could push back the point when companies receive returns on their investments.
“I think this is just a market readjusting itself,” she said, describing the disruption as a large interruption to the construction cycle. She did not give a numerical estimate for how far the broader market could fall.
Her distinction was between businesses paid during construction and those paid after a data center starts operating. Those groups face different consequences if projects take longer to complete.
Large cloud computing companies had borrowed money and were expected to spend it on construction and equipment. Suppliers such as Nvidia, other chipmakers and construction companies would receive payments as that spending continued, even if the pace slowed.
Businesses whose revenue depends on completed data centers would face a longer wait for cash they had expected sooner. That delay would change the timing of their returns and could lead investors to value the two groups differently.
The election result had not changed her view that construction would continue over the longer term. Some shares offered buying opportunities after the decline, in her assessment, although companies' ability to withstand delays would matter.
Heavy debt makes that wait harder to finance. Specialist AI cloud providers face both borrowing and dependence on data centers starting operations before they get paid, exposing them to both problems at once.
Political obstacles could also encourage companies to control spending more carefully. The earlier rush to borrow, build and secure capacity ahead of competitors had increased the risk of building more than customers would need.
A slower pace could reduce that risk by imposing greater financial discipline. That was a possible benefit of the disruption, rather than a change in her expectation that the infrastructure would eventually be built.
Her purchases were gradual, and most of her clients invest for the long term. She gave no names of shares she had bought, purchase amounts or timetable for further buying, and did not specify how long construction delays might last.
