Imagine you were building a hotel because every room is booked. Then you build another, and another. At some point, someone’s gotta ask if there’s enough demand to fill them all.
That’s exactly the question Wall Street’s been asking about AI stocks over the past several months.
Spending on chips and data centers helped propel stocks higher for years. For perspective, the S&P 500 was up 10.6% in 2026 through November 3, even as AI shares grew jittery. If performance holds, this year will mark the fourth consecutive year of double-digit gains for the benchmark.
While some worry that Democrats capturing both chambers of Congress in the midterm elections, as reported by AP News, adds scrutiny to a nervous market that could weigh on AI stocks, Bank of America now sees a twist investors might not have expected.
In a note shared with me, the bank’s analysts argue that slower AI spending, due in part to post-election regulatory roadblocks to data center builds, might actually prolong the boom, regardless of the midterm election adding to uncertainty. The question is: could the very thing investors fear most end up giving the AI-driven stock rally a second life?
TIMOTHY A. CLARY / Getty Images
Bank of America says slower AI spending could extend the boom
Bank of America is challenging a common theme on Wall Street that ever-higher spending is always better for chip stocks.
More Artificial Intelligence:
- AI’s biggest business opportunity may not look like AI
- AI racks are outgrowing old power: 2 chipmakers have a strategy
- BofA splits software into winners and losers as AI agents arrive
The bank’s argument turns on one condition. Demand for paid computing must remain strong even if the pace of new data center construction slows.
“Slower supply growth against resilient demand could lift utilization, support pricing and reduce overbuild risk,” the bank’s analysts wrote as reported by Seeking Alpha.
Naturally, idle servers produce little revenue. If too much computing capacity floods the market, providers may slash rental prices to attract customers, making the next expensive expansion harder to justify.
A more measured rollout could avoid that trap.
Importantly, demand hasn’t collapsed as investors fear. Based on the latest quarterly reports, Amazon Web Services revenue grew 39% year over year, Microsoft Azure climbed 48%, and Google Cloud surged 103%.
For Microsoft in particular, the improvements to its existing GPU fleet helped it serve more customers and earn more revenue per AI token. Better use of installed hardware, in other words, can be valuable without another spending surge.
The financial stakes are immense, though. BofA forecasts combined free-cash-flow deficits for eight hyperscalers and neoclouds to rise from $128 billion in 2026 to $335 billion in 2027, then narrow to $77 billion in 2028.
“The market therefore appears to be discounting capex risk before the demand data have cracked,” the analysts wrote.
BofA’s trade-off is potentially lower peak earnings but a longer profit cycle, which could support chip-stock valuations.
How AI spending became Wall Street’s biggest bet
You’d probably think AI is a story about software breakthroughs. On Wall Street, though, it has become effectively a story about who can spend the most money building the machines behind them.
Goldman Sachs estimates global AI investment will exceed an eye-popping $1 trillion in 2026. U.S. hyperscalers alone are expected to spend roughly $800 billion on capital projects, and even though not every dollar goes toward AI, the amount is staggering, to say the least.
Big Tech giants like Amazon have projected $220 billion in capital spending for 2026, while Alphabet expects $195 billion to $205 billion. Essentially, those budgets are a windfall for Nvidia’s chips, Broadcom’s networking equipment, memory suppliers, and semiconductor manufacturers.
Unsurprisingly, investors have followed the money.
State Street estimates the Magnificent Seven generated 62% of the S&P 500’s return in 2023 and 53% in 2024. AI Capex has effectively driven the market, and Goldman Sachs estimates AI investment accounts for nearly half of the index’s expected earnings-per-share growth in 2026.
The big assumption is that greater capacity today would produce bigger profits tomorrow. On the flip side, though, a larger data center budget also means larger bills for the companies writing the checks, without guaranteeing that customers will pay enough for AI to justify them.
Amazon illustrates the tension. Even as its cloud business accelerated, the company reported a $42.4 billion trailing 12-month free cash flow outflow, partly reflecting its aggressive infrastructure investments.
Wall Street has spent years rewarding the AI buildout. Now investors are asking whether the same spending could eventually undermine it.
What Democrats’ midterm victory means for AI stocks
The midterms offer a new reason for AI investors to question the spending boom.
Democrats won control of both chambers of Congress on November 3. The initial reaction, as of 10:30 am EST, is the Nasdaq and S&P 500 gaining 2.1% and 1.6%, respectively.
Yes, backlash against data centers has become a key issue in recent months, and state and local decisions on permits and power infrastructure could slow construction, particularly in major hubs such as Texas, as Democrats push for greater scrutiny of the industry’s expansion.
But Democratic control doesn’t guarantee sweeping AI legislation. President Donald Trump retains veto power, making gridlock a possibility.
Moreover, as I recently covered, Morgan Stanley analysts argued that chipmakers’ earnings depend heavily on a handful of major tech companies and that several manufacturers could be hit if hardware orders slow.
History offers encouragement, not a promise. Fidelity says the S&P 500 rose in 95% of the 12-month periods following midterms since 1938, a pattern not tied to either party winning.
For 2027, watch Big Tech’s spending guidance, cloud revenue growth, server utilization, free cash flow, and chip orders. The election may shape the rules. Whether AI investments earn their keep will determine much more.
Related: Morgan Stanley issues a fresh warning about AI stocks
