The investor who called Black Monday and made his fortune profiting from the 1987 crash is telling clients to keep buying the artificial intelligence rally — but to do it with one foot near the exit. In a Thursday morning appearance on CNBC’s Squawk Box, Tudor Investment founder Paul Tudor Jones argued that the AI-fueled bull market is roughly 50% to 60% of the way through its run, with another year or two left before peaking, and that he himself recently added to his AI exposure. Per the CNBC report on the interview, Jones framed the moment as analogous to 1999 — about a year before the dot-com peak in early 2000 — and warned that when the cycle eventually breaks, the drawdown could be “breathtaking.”

The reason this matters to working investors is not the headline call. It is the historical pattern matching that drove Jones to it. He explicitly compared the current AI moment to two prior transformative tech booms: Microsoft’s emergence in the early 1980s, and the commercialization of the internet in the mid-1990s. Both ushered in productivity miracles that lasted four to five and a half years. Both produced eye-watering equity returns before the cycle ended. And both, when they did end, ended hard.

What Jones Actually Said

Jones did not equivocate about how late or early the cycle is. “We’re kind of, I’d say, 50 or 60%,” he told the Squawk Box hosts. “If I had to pick a period, we’ve got another year or two to run.” That is a calibrated, actionable estimate from one of the most respected macro traders alive, and it puts the projected market peak somewhere between the spring of 2027 and the spring of 2028.

The analogy he chose for the current AI moment was Anthropic’s Claude. “I kind of think Claude, January of this year, would be the equivalent of when Microsoft came out in ‘81,” Jones said. The comparison is loaded with meaning. Microsoft’s emergence in the early 1980s coincided with the launch of the IBM PC and the start of a decade-long productivity revolution that delivered enormous equity returns to investors who held through the volatility. Jones is arguing that a similar productivity miracle is now beginning around enterprise-grade AI, and that we are roughly in the equivalent of 1985 or 1986 — past the early adopter phase, well into the broad-deployment phase, but still years from saturation.

He paired that bullish framing with a sobering warning about what happens when the cycle inevitably ends. “Just imagine the stock market went up another 40%,” Jones said. “The stock market GDP is gonna probably be good lord 300%, 350%. You just know that there’ll be some… breathtaking kind of corrections.” That arithmetic implies a peak market-cap-to-GDP ratio at roughly 3.5 times annual U.S. economic output, which would shatter every prior bubble high and set the stage for a drawdown that he says will be substantial.

Why the 1999 Analog Is Both Bullish and Terrifying

The 1999 comparison is the one investors should sit with longest. In 1999, the S&P 500 returned roughly 21% and the Nasdaq Composite returned 86%. Investors who got out at the start of 1999 missed the largest single-year tech gains in market history. Investors who stayed in until early 2000 captured those gains but then watched the Nasdaq lose 78% of its value over the next 30 months.

Jones’s argument is that we are in the equivalent of early-to-mid 1999. There is meaningful upside left for investors willing to hold AI exposure, but the upside is being consumed at the cost of an increasingly extreme valuation environment that will not unwind gently. His use of the word “breathtaking” to describe the eventual correction is a deliberate echo of the language used to describe the 2000-2002 dot-com unwind, and it is a warning that risk management discipline matters more in late-cycle bull markets than in any other environment.

The discipline he describes is straightforward and grounded in his own track record. Jones called the 1987 Black Monday crash because he was watching the parallels between 1987 price action and the 1929 setup and recognized the pattern in time to position for it. He is now watching 2026 against 1999 with the same pattern-matching lens. That kind of historical reasoning is not a substitute for fundamental analysis, but it is a powerful complement to it, and it is the kind of thinking that separates traders who survive multi-decade careers from those who blow up in single cycles.

The Productivity Miracle Argument

The reason Jones can be both this bullish and this cautious simultaneously is that he sees the underlying technology as genuinely transformative. He argues that AI represents a productivity miracle on the scale of personal computing in the 1980s and the internet in the 1990s. Each of those prior waves delivered four to five and a half years of sustained productivity gains and market upside before the speculative excess unwound, and the underlying productivity gains persisted long after the financial volatility passed.

That distinction is critical for investors. The dot-com bust did not disprove the productivity case for the internet. It just disproved the valuation case for many of the companies that were going to deliver it. Pets.com went to zero. Amazon, Microsoft, Apple, and Google went on to build the largest businesses in history. The productivity miracle was real. The financial structure built around it was a bubble. Both can be true simultaneously, and they typically are.

For AI in 2026, the same disaggregation will likely apply. Some of today’s hottest AI names will deliver against their valuations and grow into trillion-dollar franchises. Some will not, and their share prices will eventually reflect that. The challenge for investors is that no one knows in advance which is which, and the bull-market environment makes it nearly impossible to pick correctly because every name is rising. Jones’s solution to this problem is the one he has used for decades.

How Jones Is Actually Trading It

“I’m a macro trader, so I just buy baskets,” Jones said when describing his AI exposure. “What I would simply say is, it’s a crazy, crazy time. I love always to find historical precedents.” He did not name specific tickers or specific ETFs. The basket approach is a deliberate response to the single-name selection problem: if you do not know which AI company will dominate, own a broad swath of them and let the basket capture the productivity gains.

That approach has structural advantages and disadvantages. The advantage is diversification: a basket protects against the risk that any single name disappoints. The disadvantage is that a basket also dilutes upside from the eventual winners. Jones is signaling that, in his view, the macro trade — long broad AI exposure — is more important than the picking trade. For investors with less risk tolerance and less ability to time exits, basket exposure through AI-themed ETFs is a reasonable mirror of the Jones approach.

We have explored the specific names worth considering in our analysis of the best AI stocks to buy now in 2026, and the broader question of whether AI startup valuations are a bubble or a boom for investors trying to evaluate the picks-and-shovels case versus the platform case.

The Regulation Risk Jones Is Watching

Jones’s bullish call came with an important caveat that often gets buried in coverage: he is worried about the long-term risks of AI itself. He told Squawk Box that governments will ultimately need to step in with regulation, and he expressed concern about the possibility of AI becoming dangerous to humanity if left unchecked. That is a meaningful comment from someone who is simultaneously buying more AI exposure, and it should be read as a warning that regulatory risk is the single most likely catalyst for the eventual correction.

The pattern in technology history is that productivity revolutions arrive without rules, generate enormous wealth, eventually attract concentrated political attention, and then face a wave of regulation that compresses valuations even when the underlying business case remains intact. Telecom went through it in the 1930s. Television went through it in the 1940s. The internet went through it in the late 1990s and 2000s. AI is now in the pre-regulation phase, and Jones is signaling that this phase has a finite life.

For investors, the implication is that the timing of an AI peak may be less about valuation excess and more about political backlash. A serious AI safety incident, a major election turning on AI policy, or a coordinated regulatory crackdown by the EU, U.S., and Chinese governments could compress AI multiples sharply even before the productivity case is fully delivered. Risk managers should be tracking the political calendar at least as closely as the earnings calendar.

What Jones Did Not Say

It is also worth noting what Jones declined to comment on, because the absences are revealing. He did not address Federal Reserve policy, the question of whether incoming Fed Chair nominee Kevin Warsh can engineer rate cuts, the trajectory of inflation, or the Iran-driven oil shock that has dominated other corners of the market. He limited his analysis to the AI investment thesis specifically, which suggests he sees AI as a relatively rate-insensitive trade compared to broader equity exposure.

That stance is consistent with his macro framework: AI is, in his view, a once-in-a-generation productivity story whose returns are largely independent of the rate cycle. Higher rates may compress multiples but cannot stop the underlying business case. Lower rates would accelerate gains but are not necessary to make the trade work. The Fed cycle therefore becomes a second-order issue in the Jones framework.

For investors who do not share Jones’s confidence on rate insensitivity, our coverage of the Federal Reserve’s most recent rate decision and inflation outlook provides the macro context that complements his AI-specific analysis.

How to Use This Call

A Tudor Jones television appearance is a market-moving event. His call this morning is unlikely to change institutional positioning much because professional investors already have AI exposure baked into their books. The bigger effect will be on retail investors and individual portfolios, where fear of missing the rally and fear of getting caught in the crash both compete for the same investment dollars.

The framework Jones is offering is a useful resolution to that tension. Stay in the trade, but hold it in basket form rather than concentrated single-name positions. Recognize that another year or two of gains is plausible but the eventual correction will be severe. Build risk-management discipline now, while the trade is working, rather than trying to retrofit it during a drawdown when it is psychologically harder to act. And watch the regulatory backdrop as carefully as the earnings backdrop, because the political calendar may dictate the eventual peak more than the fundamentals do.

The dot-com analogy Jones is leaning on came with a clear lesson: the productivity revolution was real, the equity returns to long-term holders were real, and the people who got hurt were the ones who confused a bubble for a permanent state. AI in 2026 is, on Jones’s read, in exactly the position internet stocks were in 1999. That means the upside is real and so is the cliff. Both will arrive. Position accordingly.

Frequently Asked Questions

Who is Paul Tudor Jones and why does his market call matter?

Paul Tudor Jones is the billionaire hedge fund manager and founder of Tudor Investment Corporation, one of the most prominent macro hedge funds. He shot to fame after predicting and profiting from the 1987 Black Monday stock market crash, and his track record of pattern-matching across market cycles makes his macro calls market-moving events. He is also chairman of nonprofit Just Capital, which ranks public U.S. companies based on social and environmental metrics.

What does Jones mean when he says we're 50-60% through the AI bull market?

Jones is estimating that the AI-driven bull market still has roughly another year or two of upside left before peaking. He is comparing the current environment to 1999, about a year before the dot-com peak in early 2000, when the S&P 500 returned 21% and the Nasdaq returned 86% before the subsequent bust. His framework is based on prior productivity miracles (Microsoft in the 1980s, internet in the 1990s) lasting four to five and a half years.

What's the "breathtaking" correction Jones warned about?

Jones suggested that if the rally continues another 40% from current levels, total U.S. stock market capitalization could reach 300% to 350% of GDP — a level that would shatter every prior bubble high. He used the word “breathtaking” to describe the eventual drawdown when those valuations correct, a deliberate echo of the language used to describe the 2000-2002 dot-com unwind that wiped 78% off the Nasdaq Composite.

How is Jones positioning his own AI exposure?

Jones describes himself as a macro trader who buys baskets rather than picking individual stocks. He confirmed he recently added to his AI investments but did not specify timing or specific tickers. The basket approach is a deliberate response to the single-name selection problem in late-cycle bull markets: own broad exposure to the productivity miracle without trying to identify which individual companies will be the long-term winners.

What did Jones say about AI regulation and safety?

Jones warned that governments will ultimately need to step in with AI regulation and expressed concern about the possibility of AI becoming dangerous to humanity if left unchecked. He framed regulatory risk as a meaningful long-term concern even as he remains short-term bullish on the trade. The pattern from prior technology cycles suggests political backlash often compresses valuations even when the underlying business case remains intact.

How does this compare to other current market views?

Jones’s bullish AI call lands the same week Whirlpool warned of a “recession-level” decline in U.S. consumer durables tied to the Iran war, highlighting the sharp bifurcation in the U.S. economy: AI and megacap tech are powering index gains while consumer-facing categories are cracking. For more on that bifurcation, see our analysis of whether AI is replacing jobs and which careers are safe and our coverage of the 2026 economic outlook.