TS Lombard's Beamish says Fed needs 5 rate hikes to curb AI boom

The Federal Reserve should raise interest rates as many as five times over the next year to contain inflation risks from the artificial intelligence boom, according to TS Lombard Chief Economist Freya Beamish.

"There are so many different types of shocks hitting the economy — that's why it's quite hard to tell how far behind the curve they are," Beamish said in a Bloomberg Television interview. "We probably are in an environment where this economy can sustain higher interest rates."

Beamish sees a "dual" economy in the US, with momentum in technology standing apart from the rest. Policymakers should respond by increasing borrowing costs as much as five times in the next year to curtail leverage building up in the tech sector, she said. The rest of the economy can probably withstand higher rates, she added.

The call comes as evidence mounts that AI investment is feeding through to consumer prices. Barclays estimates AI-related categories have added about 20 basis points to core PCE inflation and 25 basis points to headline PCE in recent months. Computer software and accessories prices have jumped 17% since December after being in deflation for much of their history, the bank said, driven by surging memory chip costs from data center demand.

The Fed itself acknowledged the dynamic in its June meeting minutes, with staff citing "AI-related price pressures" as a driver of core goods inflation alongside tariffs. The central bank said "ongoing strong demand for AI infrastructure would likely sustain upward pressure on prices for technology products and electricity."

For the Fed, the analysis suggests the inflation challenge has shifted. Even if oil prices resume their decline after the US-Iran ceasefire, Barclays said this alone is unlikely to deliver the disinflation policymakers had hoped for. Bond traders have already ramped up bets for a July rate increase ahead of Chair Kevin Warsh's congressional testimony and consumer-price data this week. The Fed's own survey showed interest rates expected to remain unchanged through early 2027, with market pricing implying one rate hike by mid-2027.

Cleveland Fed President Beth Hammack echoed similar concerns last week, saying surging demand for AI infrastructure could add to inflationary pressures and potentially require higher interest rates if price growth remains elevated. Hyperscalers "will pay almost any price" for critical data center equipment, she said, suggesting limited price sensitivity in the sector.

The last time the Fed faced this type of technology-driven inflation dynamic was during the dot-com era of the late 1990s, when then-Chair Alan Greenspan warned of "irrational exuberance" and the central bank raised rates from 4.75 percent to 6.5 percent between June 1999 and May 2000. The S&P 500 fell 10 percent over the subsequent six months as tighter policy eventually cooled the tech rally.

For investors, the stakes are high. Higher rates increase the cost of capital for AI infrastructure spending, potentially cooling a semiconductor rally that has lifted chip stocks 220 percent this year, according to Bull Theory. Productivity gains from AI are likely years away, the research platform said, keeping interest rates higher for longer — and higher rates are the single biggest risk to the AI valuations the market has been pricing in.

This article is for informational purposes only and does not constitute investment advice.