Has the AI bull market already completed 60%? The legendary investor who accurately predicted "Black Monday" has rung the bell: it could continue for another year or two.

One of the most prominent macro investors in the world, Paul Tudor Jones, founder and chief investment officer of Tudor Investment Corporation, dropped a bombshell in a media interview. This legendary hedge fund manager, who gained fame for successfully predicting and profiting from the 1987 "Black Monday" crash, has provided a timeline for the current AI-driven bull market in US stocks with a mix of optimism and warning: there is still about one to two years of upside potential in this round of AI-driven bull market, and he himself has recently increased his holdings in AI-related stocks through a "basket" approach. However, if the market rises by about 40% more, the valuation expansion will create conditions for a "choking" correction. Nevertheless, the context of this statement is far from calm. Just a day before Jones made his remarks, the S&P 500 and Nasdaq indices had just fallen from their historical highs, and the Philadelphia Semiconductor Index plummeted by 2.7% in a single day. In this sensitive moment of "historical high points with corrections," Jones's combination of "end of the bull market" and "warning of correction" statements is bound to create ripples in Wall Street. Historical Analogy: The Time Axis from 1995 to 1999 Jones's thinking is not based on speculation, but on a rigorous historical benchmark framework. He has accurately positioned the current AI revolution in relation to two technological productivity miracles that changed the world: the rise of Microsoft Corporation's personal computer software in the early 1980s and the acceleration of the commercialization of the Internet in the mid-1990s with the release of Windows 95. "I believe that the appearance of the Claude model in January of this year is equivalent to the moment when Microsoft Corporation emerged in 1981," Jones said, referring to the Claude large model released by Anthropic at the beginning of this year as a milestone in the current AI revolution. "Both of these periods were the starting points of productivity miracles, each lasting about four to five years. We are now at around 50% to 60% completion." If we accurately project the time axis back to 1995, the year when the commercialization of the Internet truly accelerated, then Joness statement of a "one to two year" upward window is actually based on a core judgment: it is not the eve of a bubble burst, but the midway point of a productivity miracle. This positioning has great market significance: if the progress is only halfway complete, it means there is still a lot of room for AI profitability to be realized, and the reasonableness of continued capital expenditure expansion has not been disproven. However, what is most notable is not the analogy to 1995, but Jones's positioning of the current market sentiment in 1999, rather than the bubble peak of 2000. This distinction is crucial: 1999 was the period of the most frenzied valuation expansion in the internet bubble, with about a year until the peak in early 2000. Anchoring the present there not only confirms further upside potential, but also predicts the inevitable end of the bubble. 300% to 350%: Ringing the Alarm Bell Jones's most chilling warning focuses on a precise number: if the US stock market were to rise by about 40% from its current level, the total market capitalization to GDP ratio could climb to 300% to 350%. "At that point, you know in your heart that there will inevitably be some kind of... suffocating steep correction," Jones bluntly stated. This is not an empty threat. According to the latest data, the current total market capitalization to GDP ratio for US stocks is about 252%, far higher than the peak of the internet bubble in 2000 at 180%. Within Jones's framework, a further 40% increase would actually describe a "late-stage bubble" classic scenario: the end of a bull market is often accompanied by the most severe valuation expansion and the heaviest backlash. This judgment forms a set of precise mathematical coordinates: if the stock market were to rise by a further 40%, the valuation/GDP ratio would leap to the 300% to 350% range - no period in history has achieved a soft landing after reaching this level. Jones did not provide a specific time frame, but offered a conditional trigger: once the valuation ratio reaches this threshold, the correction is no longer a question of "if" but of "how severe" and "how fast." Current Market Environment: AI Sector Still Strong, But Differentiation Signals Arise Jones's remarks came at a crucial moment for the AI sector, which had just experienced a significant surge and shown signs of differentiation. On the day before Jones spoke (May 6th), both the S&P 500 and Nasdaq indices reached historical highs, with the Philadelphia Semiconductor Index skyrocketing by 3.5%, pushing its cumulative increase for 2026 to about 60%. AMD (AMD.US) surged 17% on better-than-expected financial results, NVIDIA Corporation (NVDA.US) rose nearly 5%, and the AI frenzy was in full swing. Year-to-date, the S&P 500 index had a cumulative increase of about 7.2%, the Nasdaq rose by 11%, and AI infrastructure-related assets outperformed the broader market by a wide margin. However, on the day Jones spoke (May 7th), this strong trend underwent a subtle change. The major US indices all closed lower: the Dow Jones Industrial Average fell by 0.63%, the S&P 500 by 0.38%, and the Nasdaq by 0.13%. The Philadelphia Semiconductor Index plummeted by 2.7%, while Intel Corporation (INTC.US) and AMD (AMD.US) both dropped by about 3%, and Arm (ARM.US) saw a significant decline. The Russell 2000 index recorded a drop of 1.6%. In contrast, some AI stocks continued to rise against the trend: NVIDIA Corporation and Microsoft Corporation (MSFT.US) saw gains of about 2%, Datadog (DDOG.US) surged by 31% due to an upward revision of full-year forecasts, and CrowdStrike (CRWD.US) and Palo Alto Networks (PANW.US) rose by 8% and 7% respectively. This high degree of differentiation within the sector - with some AI stocks still rising strongly while others are being heavily sold off - corresponds to Jones's historical positioning of the "end of the bull market": the late stage of a bubble does not necessarily result in a wholesale collapse, but rather a cruel selection between winners and losers. As Jones warned, when valuations approach extreme levels, the market's margin for error will sharply narrow. Another Voice: Warnings from Bubble Theorists Jones is clearly not the only one contemplating the ceiling of this current AI bull market. Investment legends from different camps have given their own judgments, forming a reference system for Jones's arguments. One of the most direct comparisons comes from another legendary investor, Jeremy Grantham, co-founder of GMO. Grantham recently issued a clear warning on a podcast: "The current US stock market is as dangerous as the 2000 internet bubble period, and the AI craze cannot reverse the market's overvaluation risk." Known for accurately identifying historical bubbles throughout his nearly 60-year investment career, he points out that "half of the assets in the current market are severely overvalued," and although AI has temporarily delayed a bear market, "the bubble will burst." The fundamental difference between Jones and Grantham is not whether there is a bubble - both actually acknowledge that current valuations are at historical highs - but rather in their timing judgments of the bubble stage. Grantham is more inclined to believe that the bubble could burst at any time, while Jones more accurately judges that the bubble is still in the "1999" stage rather than the "2000" stage, meaning there is still one last and most powerful wave of valuation expansion to be completed. These two legendary investors have reached a subtle consensus and disagreement: they both agree that a correction is inevitable, the difference lies in how many laps are left on the countdown timer. The Shrewd Game of Macro Hedge: Why Increase Holdings While Issuing Warnings? The seemingly contradictory aspect of Jones's remarks - publicly warning of a significant correction while revealing that he recently increased his holdings in AI-related stocks - actually reflects the holistic thinking of a top macro trader. "I am a macro trader, so I buy a basket of stocks, not specific individual stocks," Jones explained, and refused to disclose which AI-related positions he had increased or the specific timing of the trades. This "basket" approach embodies his overall bet on the AI trend and also provides him with maximum flexibility in terms of the specific exit path. This is not speculation, but a disciplined time-risk arbitrage anchored in history: he believes that the market will continue to be structurally supported by the AI productivity revolution in the next 12 to 24 months, with room for action before reaching the finish line. However, he also realizes that every party has an end, and the severity of the end will be proportional to the fervor of the beginning. The key is whether one has the ability to find the exit before the music stops. Jones's investment record shows that he belongs to the minority who believe they can do so. In addition, Jones's judgment on Federal Reserve policy also provides macro-level support for increasing holdings. He stated directly that under the incoming new Fed Chairman Kevin Warsh, a rate cut is "absolutely impossible," and even the possibility of reconsidering rate hikes when inflation pressure recurs is not ruled out. This judgment means that the market should not expect a "dovish rescue" from monetary policy to support AI's high valuations - the rise can only come from the profitability of AI itself. This essentially sets a disciplined constraint for AI investments: only companies that truly create value will receive returns, while targets that rely solely on speculative valuation will face elimination. Regulatory Dilemma: The Shadow Looming Over the AI Banquet Beyond trading strategy, Jones's remarks have another dimension that is often overlooked but of great significance - his concerns about the long-term risks of AI and his calls for government regulation. "If AI technology is allowed to develop unchecked, it could become dangerous," Jones warned, stating that governments around the world will eventually need to regulate AI. This concern is timely. On December 11, 2025, US President Trump signed an executive order on AI, signaling a federal attempt to coordinate AI regulation. On March 20, 2026, the White House released the "National AI Legislative Framework," proposing that Congress establish a unified federal AI policy to replace the current patchwork of state regulations. The framework's seven legislative priorities cover areas such as state law priority, child safety, energy impact, intellectual property protection, and others. However, the federal-state game of dominance over AI regulation has become intense. States like California, Colorado, Utah, and Texas have been actively pushing forward their own AI legislative agendas in 2026: Colorado's comprehensive AI bill will come into effect on June 30, 2026, California and Texas have passed the "AI Transparency Act" and the "Responsible AI Governance Act" respectively. Meanwhile, states like Florida, Virginia, Utah, and Washington are also actively advancing their own AI-related legislation. This game of "federal want to unify, states want to regulate" has real implications for the AI industry: on the one hand, a comprehensive federal regulatory framework, if passed, could provide a clearer compliance path for AI development; on the other hand, the divergent state legislations increase the compliance costs and uncertainty faced by companies. Jones listing regulation as a potential tail risk is not sensational - it represents a variable that the market generally overlooks when linearly extrapolating future profit expectations. Conclusion: Structural Opportunities and Tail Risks Under the Countdown Jones's remarks have stirred the real power of the market, not in the conclusions themselves, but in providing a precise timeline in years for this AI bull market: 50% to 60% completion, one to two years remaining. This clear positioning - whether ultimately proven as prophetic insight or precise hindsight error - will have a profound impact on risk exposure adjustments for institutional investors. Looking at it from a structural opportunity perspective: Jones's judgment of "50% to 60% completion, one to two years window" implies that the ongoing construction of AI infrastructure, data center investments, and chip supply chain expansions still have strong momentum in the medium to short term, all of which will continue to drive physical commodities like energy, industrial metals, and semiconductor materials. S&P 500 companies are expected to achieve their strongest profit growth in over four years in 2026, with over 80% of companies beating earnings estimates for the first quarter, and the logic supporting this growth has not disappeared. Viewing it from a tail risk perspective: Jones's warning threshold of a "further 40% increase reaching a valuation/GDP ratio of 300% to 350%" provides a clear "green light-yellow light-red light" risk map for risk managers. If valuations continue to rapidly inflate, approaching this historical limit, any trigger factor - lower-than-expected profits, worsening geopolitical conflicts, a resurgence of rate hike expectations due to rising inflation, or a sudden tightening of AI regulatory policies - could become the catalyst for a significant correction. Jones listing AI regulation as a core concern is not only shocking but also represents a risk dimension that the market generally overlooks when extrapolating future profit expectations.