Exclusive: Investors Are Watching AI Revenue Growth Against CapEx as Companies Face Pressure to Prove Returns, Wall Street Veteran Says: ‘The Bar Rises With the Size of the Spend'

Wall Street veteran Steve Neamtz says AI capex winners will be identified by measurable returns and revenue growth as spending rises.

Steve Neamtz, CEO of Yorkville America, said investors are increasingly examining whether the AI CapEx boom is producing measurable returns as the infrastructure buildout expands.

In an email interview with Benzinga, Neamtz said investors should compare rising capital expenditures with growth in cloud and AI revenue. He identified demand, financing, power availability and customer commitments as factors shaping the next phase of AI infrastructure investment.

Backlogs, contracted commitments and capacity utilization could help indicate whether spending is becoming productive. Neamtz said companies with named customers, defined project timelines and stronger evidence of progress each quarter may be viewed differently.

“The bar rises with the size of the spend, so the companies with a plan for returns will be separated from those with only a budget,” he said.

According to the Wall Street veteran, investors could penalize companies if increased spending fails to produce visible progress over several quarters or leads to greater dependence on debt.

Current consensus estimates indicate that six hyperscalers — Amazon.com Inc. (NASDAQ:AMZN), Alphabet Inc. (NASDAQ:GOOGL), Microsoft Corp., Meta Platforms Inc. (NASDAQ: META), Oracle Corp. (NYSE:ORCL), and Space Exploration Technologies Corp. (NASDAQ:SPCX) could collectively spend roughly $916 billion over the next 12 months, rising to nearly $1.2 trillion the following year, as per an August report by Apollo Global Management Chief Economist Torsten Slok.

In September, JPMorgan Chase CEO Jamie Dimon said AI-related spending across the hyperscaler ecosystem could reach $1 trillion in 2027, up from about $300 billion last year and $700 billion this year.

Washington Shift May Delay AI Projects

Neamtz said a shift in Washington after the midterms could slow federal permitting and alter energy policy, but state and local approvals and power availability will remain key constraints on AI infrastructure development.

He expects delays and project relocations rather than a reversal in demand. Projects tied to restricted locations, leveraged operators dependent on construction timelines and project-specific contractors face greater risks, while existing data center capacity and companies with secured power may be better positioned.

“We’d frame this by type of exposure, not by individual company,” he said.

Why AI Winners May Change in 2027

Neamtz described 2027 as a potential “shift in emphasis, not a reversal”, with investors asking which companies successfully converted infrastructure spending into earnings.

He also cited secured power, local opposition, financing costs, and contracted customers as considerations that may determine where projects are built and how quickly they generate returns. Neamtz added that short-term financing market fluctuations may affect the pace of growth and which companies benefit, but they won’t change the overall long-term trend.

“I believe patient funding and secured power may not be that exciting, but they are what advisors and investors ask about once the excitement fades.”

Image via Shutterstock