The AI Trade Is Collapsing, What Comes Next?

November 2026

The post-election selloff in the AI trade has centered on companies tied to the data center buildout, but the selling has spread broadly across markets.

The catalyst was the midterm election result, which signaled a sharp rejection of the administration’s policies. Candidates tapped into public concern about AI, including supporting moratoria on data center construction. That helped trigger a broad selloff in technology stocks.

So where do we go from here?

The euphoria phase—when nearly everything tied to the central investment theme rises—is over. The next phase is beginning: a period of sorting out which companies have the strongest businesses, balance sheets, and competitive advantages.

What is likely to happen?

The pace of data center construction may slow, and fewer projects may be built. But the buildout will continue. Some communities will welcome data centers for the jobs, potential energy-cost benefits, and other economic gains they can bring. Others may negotiate for additional benefits. Some projects may move offshore, where economies are eager for investment, though that raises national security concerns. And the main competitor, China, is unlikely to slow its efforts.

A slower buildout would mean less incremental demand for the components and resources that go into data centers: semiconductors, servers, and energy. That could put downward pressure on computing prices. At the same time, existing data centers could become more valuable—and more expensive to operate.

Who stands to win or lose?

Established hyperscalers, with strong balance sheets and access to capital, may emerge stronger. Newer cloud providers could be more vulnerable. Frontier model companies such as OpenAI and Anthropic could benefit from lower computing costs, which would improve their economics. Software companies that embed AI agents and inference into products may also gain, especially when those tools deliver measurable returns for businesses. Consumers could benefit from productivity gains as they adopt AI agents.

What are some second order impacts?

Lower semiconductor costs, particularly memory, will help vendors moderate prices and could stimulate demand. Some semiconductor capacity could free up benefitting players that have had difficulty in getting parts as well as fabrication commitments. To the degree the AI buildout contributes to inflation, this could moderate as well. Concerns about access to capital may be alleviated as weaker players will be shut out. Here again the strong get stronger.

For now, investors are reassessing valuations considering these possibilities. That adjustment should happen very quickly. Valuations have already fallen substantially; the correction may not be over. But corporate and consumer adoption of AI is accelerating, which gives us confidence the investment opportunities on the other side will be very exciting.

Michael P. DeSantis
Managing Member

Bruce M. Lupatkin
Managing Member