Asset management giant: The market's expectations for AI productivity are overly optimistic.

Allianz, one of the world’s largest insurance and asset management groups, warned that markets are showing signs of irrational exuberance about expectations that artificial intelligence (AI) will boost productivity, and said AI’s actual impact on the real economy will be far more complex and uneven than what market pricing currently suggests.

On July 3, according to Bloomberg, Allianz chief economist Ludovic Subran said at an annual economic conference in France that day, “We don’t really know how widespread the adoption of AI will be and how it will affect the real economy, but the market is already very optimistic—especially about productivity—whereas reality will be a much more mixed picture. For me, this is where I see a certain degree of irrational exuberance.” He also expressed concern about the overall “market sentiment” surrounding AI investment.

Subran’s remarks echoed comments made earlier this week by International Monetary Fund (IMF) officials, and were consistent with a warning issued on Sunday by the Bank for International Settlements (BIS)—the BIS has listed AI as one of the four “stress points” threatening global economic prosperity.

AI impact will be a “mixed picture,” not the full dividend the market expects

Subran acknowledged that the changes heralded by AI will be revolutionary, calling it a “Renaissance-like moment,” and said AI will profoundly reshape the services-sector economy. However, he also emphasized that the technology is giving rise to some “odd phenomena” in corporate and investor behavior.

His core judgment is that AI’s impact on different economies will not be distributed evenly. Current market optimism is based on the assumption that productivity will rise broadly and quickly, but the reality will be a “mixed picture”—the degree to which different industries and different companies benefit will vary significantly, which is clearly at odds with the market’s current overall pricing logic.

Subran pointed the finger at the sharp surge in capital expenditures in the U.S. AI sector. Citing judgments similar to those of IMF officials, he specifically criticized some companies for being caught in a “debt expansion cycle”—large-scale capital spending increases debt, while the timing and scale of investment returns are highly uncertain.

Wall Street Insights reported that Tobias Adrian, Director of the Monetary and Capital Markets Department at the IMF, said that current valuations of AI-related stocks may not necessarily have already formed a bubble; what financial regulators should really be wary of is that global large technology companies are increasingly financing major investments in AI infrastructure—updating at a very rapid pace—through more and more mid- to long-term debt. This mismatch between the maturities of assets and liabilities could be a potential source of future financial stability risks.

Previously, on June 29, Wall Street Insights also wrote in an article that the BIS warned about three major threats: the bursting of an AI bubble, inflation, and sovereign debt. The report singled out that AI’s “circular financing” structure is opaque and carries risks of assets being pledged multiple times; when liquidity recedes, it could trigger a credit tsunami on the level of 2008. Coupled with rising risk of second-round effects from inflation and the fact that highly leveraged basis trades by hedge funds can easily trigger deleveraging sell-offs, vulnerabilities in the global financial system are intensifying.

Subran specifically mentioned a polarization in corporate behavior: companies such as Apple and Microsoft are “not taking many actions” in the AI space, while other firms are “overinvesting.” In his view, this divergence itself is a signal that the market structure is already imbalanced.

“If you issue bonds and use the proceeds to return value to shareholders, that doesn’t look like a good sign to me, ” Subran said. He also voiced specific concerns about the potential risks for data centers, including the risk of technological obsolescence faced by some data centers and the operational logic of monetizing capital expenditures.

Subran observed that the stock market and the bond market are pricing AI risks very differently. In the bond market, he believes investors have remained relatively rational—“When you look at the credit spreads of companies in that sector, especially the hyperscalers, they are more cautious than before,” he said. He noted that the bond market has not fallen into complacency: “the bond market still has plenty of ‘debt vigilantes’ (vigilantes).”

However, the situation in the stock market is completely different. “On the equity side, it seems the sky’s the limit—and of course that’s not the case,” Subran said plainly. In his view, this split between stocks and bonds is the most direct reflection of the irrational exuberance in the current AI investment boom—equity market pricing reflects the most optimistic scenario, while caution in the bond market indicates that real constraints still exist.