The chief economist argues that AI-driven capital spending is stoking inflation faster than it can deliver productivity gains, and the Fed needs to act before leverage spirals
Here’s a take you don’t hear often on Wall Street: the AI boom is a problem the Fed needs to solve with higher interest rates. Not eventually. Now.
Freya Beamish, chief economist at GlobalData TS Lombard, is making the case that the massive wave of AI-related spending is generating inflationary demand well before any offsetting productivity gains can materialize. Her argument, laid out in client notes this week, is blunt: tighter monetary policy would be “less ugly” than letting leverage pile up in an overheated market.
The inflation-before-productivity problem #
The core of Beamish’s thesis is a timing mismatch. Companies are pouring money into data centers, power infrastructure, and semiconductors at a breathtaking pace. TS Lombard projects that US AI and data-center infrastructure spending will hit roughly 2% of GDP in 2026.
The problem, as Beamish sees it, is that all this capital expenditure is creating demand right now, pushing up prices for energy, construction, chips, and skilled labor. The productivity benefits that AI promises, the kind that would actually help tame inflation by making the economy more efficient, are still somewhere down the road.
Why Beamish thinks the Fed is falling behind #
Beamish’s concern isn’t just about current inflation. It’s about what happens if the Federal Reserve stays accommodative while this spending wave continues.
When AI investments create wealth effects, think rising stock prices and executive bonuses, people save less and spend more. That additional consumer demand layers on top of the already-inflationary capital spending. The result is an economy running hotter than the Fed’s models might suggest.
If the Fed doesn’t respond by tightening policy, Beamish argues, two things happen. First, borrowing and leverage accumulate across the financial system as cheap money chases AI-adjacent opportunities. Second, long-end Treasury yields start rising on their own as the bond market prices in inflation the Fed isn’t fighting. TS Lombard’s research has drawn explicit comparisons to the 1990s dot-com era, noting similar bubble dynamics. The firm’s older analysis has flagged the potential for tech “melt-ups,” periods where prices rise far beyond fundamentals before the inevitable correction. Beamish’s position is that proactive rate hikes would prevent the kind of leverage buildup that made the dot-com bust so painful.
The equity market implications #
TS Lombard’s analysis isn’t purely bearish. The firm has acknowledged that under a soft-landing scenario, the AI boom could continue to support equity markets. If the Fed manages rates carefully and the economy avoids recession, tech stocks could keep climbing on the back of genuine revenue growth from AI adoption.
Beamish has been tracking AI’s impact on tech sector performance since ChatGPT launched in late 2022, giving her analysis a multi-year foundation rather than a reactionary hot take. She joined TS Lombard that same year, and the firm’s research has consistently emphasized the gap between AI hype cycles and actual economic impact.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our