Legendary macro trader Paul Tudor Jones sat down with Patrick O’Shaughnessy on Invest Like the Best for a sweeping conversation that spans 50 years of trading, the 1980 silver collapse, the 1987 crash, his evolving admiration for Warren Buffett, his alarming view of AI safety, and a daily routine that starts at 2:30 AM. This is one of the most candid and useful conversations a working trader, investor, or builder can listen to right now.

TLDW (Too Long, Didn’t Watch)

Paul Tudor Jones believes the United States is sitting on the most leveraged equity market in history at 252% of GDP, dwarfing 1929 and 2000. He sees a sovereign debt bubble, a coming wave of IPO supply that could reverse a decade of buyback driven gains, and a dollar yen trade setting up as the next big macro opportunity. He calls Bitcoin the best inflation hedge that exists thanks to its finite supply, but flags real cyber and quantum tail risks. He apologizes publicly to Warren Buffett for years of doubting him and calls him the OG of compound interest. He thinks AI is being deployed without any meaningful safety regulation, that watermarking AI content should be mandated by law, and that humanity is sleepwalking into a tail risk that could cost hundreds of millions of lives. And he closes with a simple life formula: God, family, friends, fun, and service, with a daily intentional act of kindness as the secret to a meaningful life.

Key Takeaways

• The US equity market is at 252% of GDP, the highest in history. For context, 1929 peaked at 65%, 1987 around 85 to 90%, and 2000 around 170%. A standard mean reversion to long term PEs would be a 30 to 35% decline, which on this base would shave 80 to 90% of GDP in market cap.
• We are in a sovereign debt bubble, not necessarily an equity bubble. But the country is over equitized, individual equity weightings are at all time highs, and private equity has more than doubled as a share of institutional portfolios since 2008.
• IPO supply is about to flip the buyback math. Buybacks have been retiring roughly 2% of market cap per year for a decade. Contemplated IPOs in the next year could equal 5 to 6% of market cap, reversing a structural tailwind.
• Hyperscaler capex will eat into tech cash flow, which is part of why tech has been dogging it and may continue to.
• The buy and hold S&P 500 advice is dangerous at current valuations. Historically, buying the S&P at a PE of 22 has produced negative 10 year returns. Valuation matters even on long horizons.
• Dollar yen is his current setup. The yen has been grossly undervalued for 24 months. Japan is the largest net international investment creditor, holding roughly $4.5 trillion mostly unhedged in dollars. The catalyst is a new Reagan or Thatcher style prime minister who Paul thinks will trigger a sharp yen rally.
• Bitcoin is the best inflation hedge in existence because it is finite and decentralized, more scarce than gold. The two real risks are kinetic conflict triggering cyber warfare and the eventual arrival of quantum computing.
• Every major crash he has lived through had the same DNA: leverage, usually derivative driven. 1987 was 100% portfolio insurance. 1998 was Long Term Capital and derivatives. 2000 was an IPO supply unlock cascade. Today combines all three risks with sovereign debt fragility on top.
• Trading is boxing, not chess. Most days you are jabbing and feeling out the market. A few times per cycle there is a real opening. Bitcoin in 2020 was a knockout. Two year rates in 2022 was a knockout. The job is to be ready when the opening appears.
• Great traders are 70% born, not made. Paul polled his top risk takers and the consensus was nature dominates nurture. The traits: type A, hyper curious, loves competition, loves games, intuitive grasp of probability.
• Liquidity is everything. His grandfather told him as a kid, “you are only worth what you can write a check for tomorrow.” He watched Bunker Hunt go from richest man on earth to virtually bankrupt in six weeks during the 1980 silver collapse. The lesson stuck.
• Warren Buffett apology. Paul publicly recants decades of skepticism, calling Buffett a flipping genius who understood compound interest at age nine and the OG of compounding.
• AI safety is a five alarm fire. Paul attended a small conference with modelers from the four biggest model labs. The consensus answer to how AI safety gets resolved was, paraphrasing, when 50 to 100 million people die in an accident. He thinks this is insane.
• Mandatory AI watermarking should be a campaign issue. He wants knowing violations made a felony after three offenses. He says deepfakes have already fooled serious people he knows twice this year.
• The build, break, iterate model is fine for most technology and catastrophic for AI because the break in this case can be civilization scale. The Atomic Energy Commission was created 18 months after the bomb. We are three years into deployed AI with effectively zero regulation.
• Daily routine for 50 years: wake at 6:15, work an hour, 45 minutes of hard cardio, screens for the open, meetings 10 to 12, lunch meeting, hour before close and hour after to plan the next day, walk with wife at 5, work, dinner, mindless TV, work 9:30 to 10:15, sleep, wake at 2:30 or 3 AM to watch the London open and do analytical work, then back to sleep until 6:15.
• Information overload is now the bottleneck. He works harder today than 40 years ago because the volume of inputs has exploded. The challenge is preserving what he calls exquisite execution: buying when there is blood on the ground and selling at maximum elation.
• Eli Tullis was his trading mentor. Tullis traded almost only cotton and was a master of executing at the maximum apogee of fear and greed. The biggest lesson came after a catastrophic loss when Tullis greeted his wife’s friends with a smile and total composure. When the going gets tough, the tough get going.
• Robin Hood Foundation was born from a wrong call. Paul was convinced 1987 would trigger a depression. It did not. But the conviction launched what became one of the most influential anti poverty organizations in America.
• Journalism 101 should be required at every college. Newspaper inverted pyramid writing taught him principal component analysis: lead with the most important fact, then the next, then the next. He says it is exactly how he ranks variables in a trade.
• If you do not use it, you lose it. A Palm Beach doctor told him “you retire, you die” and it changed how he thinks about working into his 90s.
• The principal components of a great life: God, family, friends, fun, service. Significance does not come from the trades. It comes from the people you loved and the people you served.
• Kill them with kindness. One intentional act of kindness per day, repeated, rewires you. “I should” becomes “I am.” It is the closing message of the entire conversation.

Detailed Summary

The Kindest Thing: A Three Year Old Lost in a Vegetable Market

Paul opens the conversation by insisting they reverse the usual order of the show and start with Patrick’s signature closing question: what is the kindest thing anyone has ever done for you. His earliest childhood memory is being separated from his mother around age two and a half at an outdoor produce market in Memphis in 1957. An elderly Black man took his hand, walked him up and down the aisles, and reunited him with his mother. When she tried to give him five dollars, a meaningful sum at the time, he refused, saying he knew she would do it for his child. That night Paul began adding the unnamed man to his prayer list. He repeated that prayer roughly four to five thousand times over the next twelve years.

Decades later, watching Harry Reasoner interview Eugene Lang on 60 Minutes, Paul saw the photo negative of his own story: an older man, this time helping kids of color in Harlem, promising to put them through college if they finished high school. Paul called Lang the next day and was redirected to Bedford Stuyvesant, the highest crime neighborhood in New York at the time. He adopted a class, ran after school programs, hired tutors, dealt with kids being murdered and teen pregnancy, and learned by failing what poverty actually requires to defeat. That work seeded the Robin Hood Foundation in 1987 and one of the first charter schools in New York, the Bedford Stuyvesant Charter School of Excellence, which became the number one ranked elementary school out of 543 in NYC within five years.

Aim High and Shoot Straight

Paul tells the story of his commencement address at what is now Rhodes College in Memphis. He polled the audience to see who remembered their own commencement speakers. Almost no one did. So he ended his speech by pulling out a bow, knocking an arrow, telling the graduates “whatever you do, aim high and shoot straight,” and shooting an apple off a table. Memorable.

Trading vs Investing: A 50 Year Career in the Trenches

Paul started in 1976 when inflation was raging and assets routinely doubled and halved in a single year. He cut his teeth on the floor of the cotton exchange and the COMEX, watching Bunker Hunt accumulate roughly 200 million ounces of silver at an average cost of $3.12 and ride it to roughly $50 an ounce, becoming worth $11 billion at the peak. When the COMEX restricted silver to liquidation only, the price collapsed from $50 to under $10 in eight weeks. Hunt was virtually bankrupt. The searing lesson: never trust permanence in any asset, and always preserve liquidity.

He contrasts his own life with Warren Buffett’s. Paul’s BBI Fund has run for 40 years with a negative 0.12 correlation to the S&P 500, meaning 100% of returns are alpha. He compares trading to playing right guard in the NFL for 50 years, fighting in the trenches every single day, while Buffett’s belief in America gave him a different kind of strength: the ability to ride out a 50% drawdown in 2008 to 2009 without flinching. After listening to the Acquired podcast on Berkshire Hathaway, Paul realized Buffett understood compound interest at age nine and sought out Benjamin Graham at 17. He calls himself an idiot for ever doubting him.

The AI Existential Risk Argument

Paul attended a small conference around 18 months ago with roughly 35 to 40 attendees, including one modeler from each of the four largest AI labs. When he asked them point blank how they expected AI safety to get resolved, the consensus answer was, paraphrasing, that meaningful action would only happen after a mass casualty event of 50 to 100 million people. He has been alarmed ever since.

His core critique is structural. The build, break, iterate cycle has been the engine of human invention since the beginning. The problem is that AI is the first technology where the tail event of a break could be civilizational. He compares the regulatory response unfavorably to the atomic bomb: the Atomic Energy Commission was stood up 18 months after Hiroshima. We are three years into widely deployed AI with no real regulation, no public referendum, and no convening with adversaries like China.

His specific policy ask is mandatory watermarking of AI generated content, with knowing violations made a felony after three offenses. He says deepfakes have already deceived people he trusts twice this year and that restoring trust in a basic shared reality is foundational to fixing American discourse. He also notes that a meaningful share of senior AI scientists openly envision a future of brain implanted humans with inalienable rights. He thinks most humans, given a vote, would reject that path. His point is that there has been no vote.

The Nature of Trading: Boxing, Not Chess

Trading, Paul says, is more like classic boxing than chess. You are jabbing, feeling out the opponent, looking for an opening. Most days you are gathering information and not doing much. A few times per cycle there is a real opening that you can land hard. He cites Bitcoin in 2020 and two year rates in 2022 as recent knockouts.

The genesis of every big move, he argues, is one of three things: the market got carried away, an imbalance went on too long, or a central bank or government did something they should not have. Right now he thinks dollar yen fits the pattern: the yen has been grossly undervalued for two years, Japan holds about $4.5 trillion in net international investment positions mostly unhedged in dollars, and the catalyst has arrived in a new prime minister he compares to Reagan, Thatcher, or Trump in his second term.

Bitcoin as the Best Inflation Hedge

Paul reiterates Bitcoin as superior to gold as an inflation hedge. Gold supply grows roughly two percent a year. Bitcoin’s supply is capped. Decentralization adds defensibility. The honest caveats: any kinetic global conflict will trigger cyber warfare, and electronic assets sit on the front line. Quantum computing, if and when it arrives, could enable hacks of any bank or any digital store of value. He is not predicting either tomorrow but he is unwilling to ignore them.

Are We in a Bubble? Look at the Numbers

The headline statistic is jaw dropping. Stock market capitalization to GDP is currently 252%. The 1929 peak was 65%. The 1987 peak was 85 to 90%. The 2000 peak was 170%. We have never been here before.

Bear markets since 1970 have mean reverted on roughly a ten year cadence. A reversion to a normalized PE from current levels would imply a 30 to 35% decline. On a 250% of GDP base, that is 80 to 90 points of GDP in evaporated wealth. Capital gains tax revenue would crater, the deficit would explode, and the bond market would suffer a self reinforcing negative feedback loop.

Add to this the IPO unlock schedule. Contemplated IPOs over the next year may equal 5 to 6% of market cap. For a decade, buybacks have removed roughly 2% per year. The math is about to flip. Hyperscaler capex commitments will further eat into the cash flow that funded the buybacks. Private equity has gone from 7% of institutional portfolios in 2007 to 16% today. Real estate and infrastructure allocations have grown. The system is dramatically more illiquid and more leveraged than it was in 2008.

Paul’s specific warning to anyone telling clients to just buy the S&P: at a starting PE of 22, history shows negative 10 year returns. Valuation always matters.

A Day in the Life of PTJ

The schedule is monastic. Up at 6:15. Work an hour. 45 minutes of hard cardio. At the screens for the open. Meetings from 10 to 12. Lunch meeting. Afternoon meeting. An hour before the close and an hour after to plan tomorrow and think about what is coming overnight in Tokyo and Hong Kong. Home around 5. An hour walking with his wife. Another hour of work. Dinner. Mindless TV. Work again from 9:30 to 10:15. Sleep. Wake at 2:30 or 3 AM to watch the London open for 30 to 45 minutes and do analytical work in the quiet. Back to sleep. Wake at 6:15. Repeat for 40 years.

He says he works harder now than ever before because of information overload. The opportunity cost of every distraction is exquisite execution: buying when there is blood on the ground, selling at peak euphoria.

Eli Tullis and Executing at Maximum Pain

Paul’s mentor Eli Tullis traded almost exclusively cotton. The defining moment came after Tullis was annihilated when a long awaited drought broke and cotton went limit down over a weekend. Paul watched in disbelief as Tullis welcomed his wife’s friends to a beautiful office for lunch with a smile, charm, and zero visible distress. The lesson, branded into Paul: when the going gets tough, the tough get going.

Are Traders Born or Made

Paul polled four or five of his best risk takers at a Christmas dinner. The unanimous answer: roughly 70% nature. The traits that recur: type A personality, hyper curiosity, love of competition, obsession with games, intuitive grasp of probability theory. Paul had a degree in probability theory without ever taking a math course on it. He played chess, backgammon, monopoly, gin rummy, gambled in college, and has never stopped playing bridge with friends.

Why Keep Trading?

Three reasons. First, his Palm Beach doctor told him retirement equals death. If you do not use it, you lose it. Second, his father lived to 100 and Paul wants to remain mentally sharp through his 90s. Third, and most importantly, he wants to make an absolute pot of money so he can give it away. The pursuit of nobility, as he calls it.

The Workless World

Paul used to despair about a future where AI does so much that humans no longer need to work. So much human significance comes from work. He has become more optimistic recently, watching how athletes find significance in sport and how he finds significance in bridge games with friends. Humans, he argues, are absurdly adaptable. We may find significance in something as small as a single intentional act of kindness per day.

Why Journalism 101 Should Be Required

Paul’s father ran a tiny trade finance legal paper in Memphis. Paul grew up writing for it and taking journalism classes. He argues that newspaper inverted pyramid writing should be mandatory in every college, more important than business school. Conclusion first. First sentence carries the most important fact. Who, what, where, when, why, how. Each subsequent paragraph drops one notch in importance. This is just principal component analysis applied to communication. It is also exactly how Paul ranks variables in a trade. At any given moment, ten things might matter, but only one is the catalytic variable today. The discipline of the inverted pyramid is the discipline of trading.

The Principal Components of a Great Life

Asked to apply the same framework to life, Paul answers without hesitation: God, family, friends, fun, service. He says he has actually thought about his own funeral with anticipation, partly because of the songs he has chosen. At the end, he says, no one thinks about the 1987 crash or Bitcoin. They think about who they loved, who loved them, what kind of relationships they had, and what they did to leave a legacy of betterment for others. Legacy, he insists, means deeds, not words.

Kill Them With Kindness

The closing message comes from his mother. Wake up some days you will be in a bad mood. Something on TV will make you angry. The temptation today is to demonize the other side. The antidote is intentional. One simple act of kindness per day, transmitted outward, repeated. Reps matter. “I should” becomes “I am.” Over time you become an organically kind person. Your outlook brightens. Multiply that across a country and the country changes.

Thoughts

The 252% market cap to GDP figure is the single most important number in the conversation. Most listeners will gloss over it. They should not. The structural argument Paul lays out is internally consistent and uncomfortably specific: an over equitized country, a sovereign debt bubble, an IPO supply wave that flips a decade of buyback math, hyperscaler capex eating cash flow, private equity more than doubled as a portfolio share since 2008, and far less liquidity than 2008 to absorb a shock. None of these are predictions of an imminent crash. They are descriptions of the kindling.

His Buffett apology is the kind of intellectual honesty that is rare in finance. Two operators with opposite styles can both be right for fifty years. Paul’s negative correlation to the S&P with 100% alpha and Buffett’s belief in America with patient compounding are not rival theories of investing. They are different jobs. Most retail investors are trying to do Buffett’s job with a trader’s emotional reflexes, which is why so few make it.

The AI section is the part of the interview that should make builders pause. Paul is not an AI doomer in the online sense. He is a 50 year career risk manager applying the standard framework: what is the size of the tail, what is the regulatory containment, who has the kill switch. His answer is that the tail is potentially civilization scale, the containment is effectively zero, and there is no kill switch. The historical precedent he reaches for is not science fiction but the Atomic Energy Commission stood up 18 months after Hiroshima. The contrast with our current trajectory is uncomfortable.

The watermarking proposal is unusually concrete for a trader and unusually politically tractable for an AI safety policy. It does not require slowing capability research. It does not require international coordination as a precondition. It restores the basic epistemic substrate of public discourse: knowing what is human and what is not. Whether you think AI risks are overblown or underrated, watermarking is a Pareto improvement.

For builders shipping software in the AI era, the meta lesson is that we are running the build, break, iterate playbook on a system whose break radius is no longer contained by the founders. That is a different kind of responsibility than the one most engineers have ever held. It does not have a clean answer yet. But the question is now visible.

The kindness frame at the start and end is not throat clearing. It is the actual operating system Paul has run on for 70 years. The four to five thousand prayer reps for an unnamed man who held his hand in a Memphis vegetable market produced a pattern interrupt 25 years later that founded one of the most effective anti poverty organizations in the country. Compound interest applies to acts as much as to dollars. That is the through line of the entire conversation, and it is the thing most listeners will forget by tomorrow morning. They should not.