Banking giant gives verdict on AI stocks bubble

Banking giant Citi has weighed in on the ongoing debate surrounding artificial intelligence (AI) valuations, concluding that despite rapid gains, AI stocks have not yet entered bubble territory.

The bank’s latest analysis found that overall sector valuations remain within reasonable bounds, even as certain pockets show signs of overheating. 

To this end, Citi’s internal valuation monitor flagged “only a couple of red flags” across broad AI exposure, suggesting that while the recent rally has been “abnormally strong,” it remains anchored to earnings fundamentals.

“AI does not look like a bubble yet based on our valuation monitor. <…> There are some growing pockets of valuation concern when we dive deeper within the theme, especially in more asset heavy sub-categories,” the bank said. 

Emerging AI sector risks 

However, the bank cautioned that risks are emerging within more asset-heavy segments of the AI ecosystem, particularly among U.S. industrial and infrastructure-linked companies that have seen outsized price moves. It advised investors to consider profit-taking in these areas while maintaining exposure to diversified AI plays.

At the same time, Citi continues to favor a “growth at a reasonable price” (GARP) strategy, focusing on companies with earnings expectations aligned to market forecasts and sustainable growth potential. 

The bank also distinguishes between asset-light and asset-heavy AI models, an increasingly important factor as more firms tie their future to the technology. 

It’s worth noting that concerns about an AI bubble have been growing across global markets. Analysts have compared the current boom to the late-1990s Dot-com era, pointing out that the top ten U.S. companies, many driven by AI, now make up nearly 40% of the S&P 500’s market capitalization, compared with about 25% at the dot-com peak.

The index’s price-to-book ratio has also climbed above its 2000 high, and investor behavior shows familiar signs of “bubble psychology,” with capital pouring into AI-linked stocks regardless of profitability. Some economists warn that valuations in parts of the AI sector may already exceed those seen during the dot-com mania.

At the same time, today’s environment differs in key ways. Unlike the internet boom, many AI leaders, particularly in semiconductors such as Nvidia (NASDAQ: NVDA), cloud, and enterprise software, are profitable, cash-generating businesses funding their expansion from internal resources rather than speculation. 

The technology is also being integrated into established industries, enhancing efficiency and productivity, rather than relying on untested business models with no clear path to revenue.

Featured image via Shutterstock