Exploring the Turning Point of the AI Rally: Lessons from the Dot-com Bubble

Research evidence excerpt

Exploring the Turning Point of the AI Rally: Lessons from the Dot-com Bubble

IdeaMExhibit 1: Weight of NASDAQ technology stocks (high P/S) in aggregate hedge fund

portfolio versus weight in market portfolio

Peak of excess exposure

Source: Brunnermeier and Nagel (2004), Morgan Stanley Research

1.3 Rational investors cannot burst a bubble if they cannot synchronize

their selling

As discussed above, empirical evidence suggests that rational and sophisticated investors

such as hedge funds tend to participate in, rather than oppose, a growing bubble. In

'"Bubbles and Crashes", Abreu and Brunnermeier (2003) present a theoretical model

showing that the presence of rational arbitrageurs alone does not necessarily lead to the

immediate correction of a bubble.

The key is synchronization risk: arbitrageurs recognize overvaluation at different points in

time and are therefore unable to synchronize their selling sufficiently. Although all

investors expect the bubble to burst eventually, exiting too early means forgoing further

gains, while exiting too late risks being caught in the crash. As a result, it can be rational

for investors to remain invested in the bubble.

The study suggests that the direct trigger for a bubble's collapse is not adverse news

itself, but news events or sharp price declines trigger alignment in the timing of sales.

The bubble bursts when the cumulative selling pressure from arbitrageurs exceeds the

capacity of behavioral traders (boundedly rational investors who support overvalued

assets based on excessive optimism or trend-following behavior) to absorb that supply.

1.4 Inexperienced investors are more likely to extrapolate trends

Greenwood and Nagel's (2009) "Inexperienced Investors and Bubbles" examines the role