Abstract
The rapid growth of artificial intelligence has led investors to question whether the recent rally in AI-related stocks is a sign of a speculative bubble. This paper compares the current AI driven market to the dot-com boom of the late 1990s. Using a proprietary methodology, we identify AI exposure within the S&P 500 Index and construct AI and ex-AI portfolios. We apply a similar framework to the dot-com period, building TMT and ex-TMT portfolios. Then, we compare the performance, concentration, momentum dynamics, valuation, and financial characteristics of the S&P 500 Index, TMT, ex-TMT, AI, and ex-AI portfolios.
Even if there are a lot of similarities between the AI and dotcom booms, our analysis suggests that the current AI episode lacks the hallmark “explosive valuation dynamics“ typically associated with late-stage bubbles. However, this does not imply that the theme is low risk. The primary concern is concentration risk, explaining that a narrow group of AI related stocks has driven a disproportionate share of index level returns. This implies that standard equity allocations embed significant implicit exposure to the AI factor and to long-duration growth. Consequently, risk management should focus less on timing a potential “bubble pop” and more on mitigating drawdown amplification. Meanwhile, it remains crucial to assess the sustainability of earnings and capital expenditure dynamics while monitoring indicators such as market breadth, issuance activity and any re-acceleration in valuations, as these could signal a transition toward genuinely speculative behavior.
Authors
