IMF economist: AI investment boom could trigger dot-com-style correction, though little more

The International Monetary Fund’s chief economist Pierre-Olivier Gourinchas warns that the current AI investment boom could lead to a dot-com-style market correction, though he believes it’s unlikely to trigger a systemic financial crisis. Unlike the 2008 housing bubble that relied heavily on debt financing, today’s AI investments are funded primarily by cash-rich tech companies, reducing the risk of broader economic contagion.

The big picture: Tech companies are pouring hundreds of billions into AI infrastructure without the leverage-based financing that made previous bubbles systemically dangerous.

• AI-related investment has increased by less than 0.4% of U.S. GDP since 2022, compared to the dot-com era’s 1.2% increase between 1995 and 2000.
• “This is not financed by debt, and that means that if there is a market correction, some shareholders, some equity holders, may lose out,” Gourinchas said. “But it doesn’t necessarily transmit to the broader financial system.”

Key similarities to the dot-com bubble: Both eras feature inflated valuations based on future promises rather than current productivity gains.

• Stock valuations and capital gains wealth have reached new heights, fueling consumption and adding to inflation pressures.
• The promise of transformative technology may not meet near-term market expectations, potentially triggering a valuation crash.
• Current productivity gains from AI investment have not yet materialized in the broader economy, similar to how internet stocks in the late 1990s weren’t based on actual revenues.

Why this matters: The AI boom is simultaneously supporting economic growth while contributing to persistent inflation without delivering expected productivity improvements.

• The IMF’s World Economic Outlook cites AI investment as a key factor propping up U.S. and global growth this year.
• However, increased investment and consumption are elevating demand and inflation pressures without associated productivity gains.
• Non-tech investment is actually falling, partly due to uncertainty over President Trump’s tariffs.

Inflation implications: The IMF has revised its U.S. inflation forecasts upward, now expecting slower declines than previously anticipated.

• U.S. consumer price inflation is forecast to decline to 2.7% in 2025 and only reach 2.4% in 2026.
• A year ago, the IMF had predicted inflation would return to the Federal Reserve’s 2% target this year.
• Other factors keeping inflation elevated include reduced immigration limiting labor supply and delayed tariff effects on consumer prices.

What about tariffs: Gourinchas confirmed that U.S. companies, not foreign exporters, are absorbing the cost of Trump’s tariffs.

• “So far, the evidence suggests that importers have absorbed it in margins, and they have not transmitted as much to the ultimate customers,” he explained.
• Import prices have not declined, contradicting Trump’s prediction that foreign countries would pay the price of protectionist policies.
• This assessment aligns with academic studies and business surveys showing companies on the U.S. side are bearing the tariff costs.

Potential risks: While direct financial system exposure appears limited, an AI correction could still create broader market instability.

• A correction might trigger shifts in sentiment and risk tolerance, leading to broader asset repricing that could stress non-bank financial institutions.
• However, Gourinchas emphasized there are no “enormous links from the debt channel” that would create systemic vulnerabilities like those seen in 2008.