Great technology and great prices rarely arrive together. Artificial intelligence is the cleanest example.

Bain & Company estimates the industry needs $6 trillion in annual revenue by 2031 to justify its data center spending, and today’s products could supply as much as $1.8 trillion of that. The remaining $4.2 trillion has to come from businesses that barely exist yet, according to TechRadar.

Few people have poked at that gap as loudly as Michael Burry, who rose to fame by calling the housing crash before the global financial crisis. In September, he added to shorts on Micron, Nebius, Palantir, and a semiconductor ETF, arguing that the memory chip shortage will fade.

Now he has reorganized the whole bet.

Also read: Michael Burry’s blistering message on economic inequality

Burry consolidates into two index puts

On Oct. 11, Burry told his Substack readers he had rolled most of his options positions into two index shorts: puts on the Nasdaq-100 and puts on the VanEck Semiconductor ETF, known as SMH. He also said he sees a significant top within the next nine months and warned that the final stretch of the rally could get wild.

The contract details are specific. His Nasdaq-100 puts expire in April 2027, and his SMH puts in June 2027. Both sit roughly 20% out of the money. He called the Nasdaq-100 puts relatively cheap because volatility is low, and he expects both positions to gain significant value well before the index or the ETF actually falls 20%, according to Benzinga.

Breadth gives his argument some support. The Nasdaq-100 closed at a record 30,732 on Sept. 22, yet the share of its members above their 200-day moving average dropped from 77% to 56% since mid-August. Fewer stocks are carrying the rally.

He has not dropped every other bet. Burry still holds puts on Oracle, Palantir, MetLife, and Caterpillar as side bets.

Burry is still bearish. The consolidation makes that cleaner to see, but it does not tell you the size of the bet.

Astrid Stawiarz / Getty Images

The chip ETF is narrower than it looks

Burry calls SMH a concentrated bet on the hottest and most overextended chip stocks. Nvidia makes up 19.6% of the fund, Taiwan Semiconductor 9.3%, and Advanced Micro Devices 5.6%. Broadcom, Micron, Marvell, and Applied Materials each sit near 5%, according to Benzinga.

It’s easy to see why the fund caught his eye. SMH has climbed more than 67% this year. Still, a steep run on its own isn’t proof of a bubble.

Say the companies inside it doubled their profits over the same period. Prices up 67% against earnings up 100% would leave the fund trading at a lower multiple than when the year started. Burry has to show more than a chart that keeps going up.

His case leans on spending. Burry argues that Amazon, Meta, Alphabet, Microsoft and Oracle carry massive off-balance-sheet commitments that the market is not fully pricing in. Cloud growth stayed strong last quarter, and capital spending is expected to cross $1 trillion next year, so bulls and bears are staring at the same numbers.

He has also gone after the supplier itself. As reported by TheStreet, Burry compared Nvidia to Cisco, pointing to concentrated buyers, since hyperscalers account for about half of its data center revenue.

Why a correct call can still lose money

Here is how a put works. The buyer pays a premium up front for the right to sell at a fixed price before the contract expires. That premium is the most they can lose, and to make money, the drop has to be big enough to cover what they paid.

The price of the put keeps shifting as expiration gets closer and as volatility rises or falls.

Take a made-up example. An ETF trades at $100, and someone pays $5 for a put with an $80 strike. Ignoring fees, they only break even if the ETF sits at $75 when the contract expires.

Now say the fund slides 15% to $85. The put is worthless, and the full $5 is gone, even though the bet on direction was right. Burry’s puts are about 20% out of the money, so he has to clear the same kind of bar, as Benzinga reported.

More Wall Street:

• Nvidia and Micron are about to lord their successes all over Wall Street
• Nvidia, Palantir, and Alphabet send waves through Wall Street
• Wall Street elites will pocket windfall from record $46 billion profits

Then there is the timing. Burry said he expects a top sometime in the next nine months. Count forward from October, and that window stretches into July 2027, but both sets of his puts expire before then. If the peak shows up near the end of his window, he’d be right about the market and still lose on his contracts.

This is not new territory for him. Back in September, he closed out his Nvidia and Palantir puts before December 2026 rather than roll them forward, cutting loose ahead of time decay. The longer-dated Palantir puts stayed on the books, which tells you that the bearish conviction never left.

What investors should take from it

Even Burry’s shorts can still lag. As of late September, Micron traded about 2% above the roughly $1,051 where he first shorted it, leaving that bet slightly underwater, while 49 analysts rated Micron Strong Buy, according to TheStreet.

You don’t have to copy the trade to learn something from it. Look at how much of your portfolio is tied up in the same few AI winners. A chip fund and a Nasdaq index fund often own many of the same giants, so the overlap can be bigger than it looks.

Whether those prices hold depends on revenue catching up with all the spending, and by Bain’s math, $4.2 trillion in new yearly revenue still has to show up.

Burry is still bearish. The consolidation makes that cleaner to see, but it does not tell you the size of the bet.

Two positions can carry more or less risk than 10, depending on how they are sized. Getting the structure right is one thing. Getting the timing right is another problem entirely.

Own what has earned its place and take a harder look at anything that requires everything to go right.

Related: Michael Burry reveals his verdict on the ongoing AI bubble