Prediction: Wall Street Has the Post-Midterm Artificial Intelligence (AI) Trade Completely Wrong

Wall Street appears to have found a silver lining following the Democrats' sweeping victories in the 2026 midterm elections. With Democrats gaining control of both chambers of Congress, investors are considering whether more regulation could slow the construction of artificial intelligence (AI) data centers across the United States.

At first glance, that might actually benefit Microsoft (NASDAQ: MSFT), Amazon (NASDAQ: AMZN), Alphabet (NASDAQ: GOOGL) (NASDAQ: GOOG), and Meta Platforms (NASDAQ: META). After all, building AI infrastructure is expensive.

If Big Tech is forced to reduce capital expenditures (capex), their free cash flow could improve. Meanwhile, a shortage of compute capacity could allow existing data center operators to charge higher prices. While this all makes sense in theory, I think Wall Street may be overlooking an important possibility.

Image source: Getty Images.

Wall Street thinks AI regulation will benefit the hyperscalers

Over the last several years, the hyperscalers have committed hundreds of billions of dollars to expanding AI infrastructure. These investments include procuring graphics processing units (GPUs), building data centers, and securing electricity. However, a number of candidates used data centers as a controversial topic during their time on the campaign trail. For many people around the country, data centers have fueled rising utility costs.

Following the midterms, Wall Street is considering whether new legislation or tighter permitting requirements could make additional U.S. data center projects more difficult to pursue. The financial argument is pretty straightforward: If the hyperscalers slow down their infrastructure investments, capex would decline and free cash flow would rise. At the same time, continued demand for compute could create a supply demand imbalance. Companies with existing capacity would be in a position to charge higher utilization and command pricing power.

The combination of lower spending and higher revenue per unit of compute would be favorable for the hyperscalers. With that said, this assumes that AI developers respond to potential regulatory restrictions by slowing down the pace of their infrastructure roadmaps.

Follow the money

During a discussion with The Wall Street Journal on Nov. 4, Meta's president and vice chair, Dina Powell McCormick, acknowledged that the election results could complicate future data center development domestically. She says starting new domestic projects may be harder and raised a question that could reshape the AI investment narrative: "Do we need to build outside the United States?"

There's already evidence that Big Tech is willing to invest internationally. Consider some of their existing projects:

Data source: Company press releases, Wall Street Journal, Reuters.

These investments were announced before the midterms, so they shouldn't be interpreted as reactions to the election. Nevertheless, they demonstrate that hyperscalers already have international infrastructure strategies in place. If new restrictions make projects in certain states less attractive, I wouldn't be surprised to see Big Tech evaluate opportunities elsewhere and accelerate overseas development.

Why the post-midterm trade could backfire

Suppose new federal or state regulations delay data center construction in the U.S. Initially, the hyperscalers may postpone their projects, causing capex to decline and free cash flow to improve. In turn, investors may cheer these tighter cost controls and higher profits, leading to rising stock prices.

It's easy to forget that AI development is a global competition. If demand for capacity continues growing, then Microsoft, Amazon, Alphabet, and Meta would have little incentive to sit on the sidelines indefinitely. Instead, they may choose to redirect their infrastructure budgets toward international markets where electricity, land, and regulatory approvals are more readily available.

Of course, this wouldn't be the easiest thing to achieve. Overseas construction involves its own set of permitting requirements, energy constraints, and U.S. export controls on semiconductors. Furthermore, Democrats' congressional victories may not lead to new federal restrictions. President Trump retains veto authority, and the specifics around any proposed legislation are sparse at this point.

Still, I question whether the hyperscalers would willingly delay expansion plans for two years while waiting for the 2028 presidential election. That strategy introduces another layer of uncertainty without guaranteeing a more favorable regulatory environment afterward.

Most importantly, redirecting AI infrastructure investments overseas would not generate the sustained improvements to cash flow Wall Street is anticipating. The hyperscalers could easily continue pouring hundreds of billions of dollars annually on AI infrastructure, just in different countries.

Ultimately, I think investors need to distinguish between a slowdown in U.S. data center construction and a slowdown in global AI investment. Those are two different outcomes.

For now, Wall Street seems interested in the possibility that tighter regulation could improve Big Tech's financial performance in the near term. If hyperscalers respond by expanding internationally, the anticipated cash flow windfall could prove much smaller than expected.

Adam Spatacco has positions in Alphabet, Amazon, and Microsoft. The Motley Fool has positions in and recommends Alphabet, Amazon, Meta Platforms, and Microsoft. The Motley Fool has a disclosure policy.