Bank of America flags tech bubble risk but tells clients not to sit out the rally

Bank of America has a message for investors nervous about the tech megacaps pushing the Nasdaq 100 to record highs: you can still join the party, as long as you book a ride home in advance.

The bank’s strategists argue that equity derivatives offer a way to capture the rally while sidestepping the damage if it turns out to be a bubble.

The bubble warning, by the numbers

BofA analysts, with work on the theme led by Michael Hartnett, have been drawing comparisons between today’s AI-fueled market and the 2000 dot-com bubble.

The bank’s Bubble Risk Indicator, or BRI, for the Nasdaq 100 and the broader tech sector has climbed to a range of 0.72 to 0.8. Higher readings mean the conditions that tend to precede corrections are building.

The analysts say the current setup resembles the market roughly six months before the March 2000 dot-com peak.

There is an important difference, though. Only 18 stocks in the S&P 500 scored above the 0.8 BRI threshold, and together they represent just 3.2% of the index weight. At the height of the dot-com boom, somewhere between 50 and 100 stocks cleared that bar.

Why volatility is flashing late-1990s signals

VIXEQ, a gauge tied to single-stock volatility, is up 46% year to date. The VIX, Wall Street’s better-known fear index for the S&P 500, has risen 13% over the same stretch.

BofA notes that this kind of divergence is reminiscent of the late 1990s, when single names lurched around even as headline indexes kept marching higher.

The analysts also flagged the effect of high interest rates. Tech and AI-linked assets have kept rising while other sectors lagged significantly.

The money behind the AI trade

AI-related capital expenditures by US hyperscalers are forecast at approximately $795 billion for 2026. The projection for 2027 is about $1.08 trillion.

BofA’s playbook: call spreads, not concentrated bets

Rather than buying the megacaps outright, BofA’s strategists suggest limited-risk derivatives, particularly call spreads.

A call spread works like a capped bet. An investor buys a call option, which pays off if a stock or index rises above a set price. At the same time, they sell another call at a higher price, which helps pay for the first one but caps the potential gain.

The result is a position with a known, limited cost. If the rally continues, the investor profits up to a ceiling. If the bubble pops, the most they lose is what they paid for the trade, not a large chunk of a stock portfolio.

According to the research, these trades could be structured on Nasdaq 100 exposure or on semiconductor sector ETFs, the corner of the market most directly tied to the AI spending boom.

What this means for investors

Capped upside means investors give up some gains if the AI rally turns into a full melt-up. Options also carry their own complexities, including timing risk: a position can expire worthless if the market stalls, even if the long-term thesis proves right.

With just 18 stocks above the 0.8 BRI threshold and only 3.2% of S&P 500 weight in that zone, the risk is narrower than in 2000. If that number begins expanding toward the 50 to 100 range seen at the dot-com peak, it would suggest the euphoria is spreading beyond the AI leaders.

A 46% jump in VIXEQ against a 13% move in the VIX shows stress building at the company level.

The projected climb from roughly $795 billion in 2026 to about $1.08 trillion in 2027 is a vote of confidence from the companies doing the spending. Any sign that hyperscalers are pulling back would hit the exact stocks driving the Nasdaq 100.