The Professionals Predicting AI Stocks Disagree by 80%

One strategist expects the S&P 500 at 5,000 by the end of 2027. Another bank expects 9,000. Same index, same deadline, forecasts that differ by 80%. Here is who is saying what about AI stocks, with names attached — and why the spread between them is more useful than any single number in it.

Why we are publishing this at all

We do not make market calls. So reporting other people's looks, at first glance, like a loophole.

It isn't, and the distinction is worth stating plainly because it is the same one you should apply to everything you read. Repeating an attributed forecast is journalism. Endorsing one is advice. We are doing the first and not the second: every figure below carries the name of the person or firm who said it, so you can weigh the source, and we have deliberately included forecasts that contradict each other.

If you only ever see the bullish case or only the bearish one, you are reading marketing. The honest version shows you both and admits nobody knows.

The bears

Joachim Klement, Panmure Liberum

Klement, head of market strategy at Panmure Liberum, expects the S&P 500 to fall to 5,000 points by the end of 2027 — roughly a 36% decline from where it stood when he said it — and the European Stoxx 600 to drop more than 30% to 430.

In his words: "My core conviction is that the AI bubble will either burst in 2027 or in 2028, so sometime in the next two years."

His reasoning: hyperscalers' free cash flows are "largely depleted", borrowing costs are rising fast and could become prohibitive, and Bloomberg Intelligence estimates hyperscaler data-centre capital spending will more than double in 2026 to $713 billion. He argues investors fixate on tech earnings and "excuse every macro, credit or whatever headwind" that appears.

Two things you should know before weighting this. First, Klement is the most bearish of eight strategists tracked by Bloomberg — he is the tail of the distribution, not the middle. Second, and more instructive: in mid-September he expected the S&P 500 to reach 8,300 by the end of 2026. Weeks later he was forecasting 5,000. The same professional moved his own number by a enormous margin in roughly a month.

Michael Burry

Burry, known for betting against subprime mortgages before 2008, has moved from short positions into put options — a more leveraged bearish bet. He said: "As such, I want more leverage in my short positions. Better timelines make leverage more palatable."

His disclosed positions include puts on Micron with strikes around $500 and on Nebius at double-digit strikes, both expiring June 2027, plus puts on other AI-linked stocks and index funds. Position sizes were not disclosed.

His argument is structural: securities tied to a handful of big technology companies that are simultaneously each other's customers, vendors and financiers do not deserve the low-risk pricing the market assigns them. On Micron specifically, he argues profits depend on memory demand outstripping supply, and that faster demand loss or rising Chinese output could bring the pricing peak sooner and lower.

The caveats matter. The implied "more than 50% drop" is an inference from strike prices, not something Burry stated. And his subprime thesis, though eventually correct, took years to pay — during which it looked wrong.

Rohit Sipahimalani, Temasek

The chief investment officer of Singapore's state investor has named a reversal in the AI trade as a key risk to global markets.

The bulls

The bearish case gets the headlines because alarm travels further than calm. The consensus sits somewhere else entirely.

• Jefferies projects the S&P 500 could reach 9,000 by the end of 2027.
• Citigroup strategists say solid 2027 earnings could support further global equity gains despite higher rates and geopolitical risk.
• Several brokerages expect the S&P 500 to finish 2026 at or above 8,000.
• Of the eight strategists Bloomberg tracks, the seven other than Klement expect an average gain of about 14%.

Put them side by side

Klement and Jefferies are forecasting the same index on the same date. One says 5,000, the other 9,000. The higher forecast is 80% above the lower one.

These are not amateurs. They are full-time professionals at established firms, with research teams, data and incentives to be right. And their honest views of the same market differ by nearly double.

What the spread actually tells you

Nobody knows. That is not cynicism, it is what an 80% disagreement among informed professionals means. If the future were legible, the range would be narrow.

A forecast is a snapshot of an opinion, not a schedule. Klement's own number moved from 8,300 to 5,000 in about a month. Forecasts are revised constantly, and the revision rarely gets the coverage the original did.

Confidence and accuracy are unrelated. The most quotable lines in this article are the most extreme ones. Extremity gets attention; it does not get accuracy.

Watch for the inference layer. The widely repeated claim that Burry expects Micron to halve is not something he said — it was derived from his option strike prices by journalists. By the time a number reaches a headline it has often passed through several hands, each adding certainty it did not start with.

Even a correct call can be untradeable. "The AI bubble bursts in 2027 or 2028" is a two-year window. Someone positioned for it in early 2026 could be badly wrong for a long time before being right, and may not survive the wait.

Something both sides agree on

Buried in the disagreement is one shared fact worth more than either forecast: hyperscaler data-centre capital spending is set to more than double in 2026, to roughly $713 billion, per Bloomberg Intelligence.

Bulls and bears do not dispute the number. They dispute what it means. To the bulls it is evidence of conviction and future capacity. To the bears it is a bill that has to be justified by revenue that has not arrived yet.

That is usually how market arguments work. The facts are shared; the interpretations are not. Finding the fact underneath a forecast is more useful than collecting more forecasts.

What we are not telling you

We are not telling you who is right. We do not know, we have no edge over the people quoted above, and our editorial policy commits us to not publishing price targets or buy and sell calls at all.

What we would say is this: if a single forecast ever feels compelling enough to act on, go and find the best-argued forecast that says the opposite. If you cannot find one, you have not looked. If you can, you now understand the decision you are actually making — which is choosing between two futures, not reading one off a chart.

Sources

1. Invezz, "AI bubble could burst in 2027, sending S&P 500 down to 5,000, strategist warns," 8 Oct. 2026 — Klement's targets and quote, the Bloomberg strategist survey, Jefferies, Citigroup and Temasek — invezz.com
2. The Motley Fool, "Michael Burry Thinks the AI Bubble Will Burst Sooner Than Later," 4 Oct. 2026 — Burry's put positions, strikes, expiries, reasoning and quote — fool.com
3. Yahoo Finance, "Michael Burry Turns More Bearish on AI With New 2027 Put Bets" — finance.yahoo.com
4. Wikipedia, "AI bubble" — background on the debate and its participants — wikipedia.org