The Potential for an AI Market Bubble

Could artificial intelligence be on the brink of a bubble bursting? And what would that mean for economies in the U.S. and globally?

As the U.S. stock market hits unprecedented highs, largely driven by impressive profits from AI companies, this concern is gaining traction. The current market valuation resembles the peaks seen during the tech boom of the late 1990s.

During a recent meeting for the International Monetary Fund (IMF), its chief economist raised alarms, suggesting that the current surge in AI investment echoes the dot-com bubble of the 1990s. Former IMF head Kristalina Georgieva indicated that financial instability indicators are on the rise, advising investors to brace themselves for potential challenges ahead.

In contrast, analysts at Goldman Sachs maintain that “expected investment returns are sustainable.” Their analysis hinges on the belief that U.S. productivity could increase by 1.5 to 1.9 percent annually in the early 2030s. If they are right, companies investing in data centers and expanding software production could see sufficient profits to warrant current stock valuations.

While there is a consensus among experts that AI could substantially reshape the U.S. economy over time, some dispute whether the anticipated productivity enhancements are realistic.

There’s also uncertainty regarding how a downturn in AI stocks would affect the broader economy, raising questions about whether the fallout would primarily hit stock investors or have wider effects.

For instance, during the dot-com crash, falling stock prices didn’t lead to a significant contraction in bank credit. The Federal Reserve responded by easing monetary policy, resulting in a relatively mild recession with real GDP only decreasing by about 0.3%.

In contrast, the housing market collapse in the mid-2000s caused deeper turmoil across the financial system, significantly reducing overall credit availability and leading to a much sharper recession, with real GDP tumbling by 4.2% from 2007 to 2009.

Up until recently, many experts believed that the AI boom largely stemmed from well-capitalized companies. If this holds true, the risk of a decline in AI stocks negatively influencing the economy might be minimized.

However, concerns about a potential bubble persist. One worry is that a few dominant AI companies may be essentially funding each other through circular trading, which can create a misleading image of growth and inflate valuations.

Another issue is that companies are increasingly relying on bond offerings to finance expensive infrastructure projects. While infrastructure investment has bolstered the economy, rising interest rates could bring additional risks.

The IMF has also pointed out the lack of oversight for non-bank financial entities, such as pension funds and hedge funds, which have rapidly expanded to control about half of global financial assets.

So, what could this all mean for the economy, both in the U.S. and beyond?

Gita Gopinath, a former IMF chief economist now at Harvard, suggests that a market correction similar to the dot-com crash could erase about $20 trillion from American household wealth—nearly 70% of U.S. GDP. He estimates that foreign investors might also incur losses exceeding $15 trillion, roughly 20% of global GDP.

Gopinath believes that today’s structural vulnerabilities and macroeconomic conditions pose greater risks than those present in the early 2000s, especially with factors like higher tariffs, potential Fed independence loss, and declining confidence in the U.S. dollar.

I’m not entirely convinced that a stock market decline now would resemble that of the early 2000s, when the S&P 500 and Nasdaq Composite dropped 50% and 75%, respectively. The reasons were multifaceted, including the impact of the September 11 attacks and the Enron scandal, which led to widespread skepticism about accounting practices.

Still, I share Gopinath’s concerns regarding the current economic landscape. It’s also worth noting that margin debt has ballooned, hitting an all-time high of $1.1 trillion—a 34% increase from the previous year. A market downturn could amplify losses for leveraged investors.

This suggests that we might see a stock market decline of 20-30%, potentially leading the U.S. economy into a mild recession.