PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: value investing 2026

  • Mohnish Pabrai on How to Invest in 2026: The Ten Commandments of Investing, Charlie Munger Lessons, Cloning, Turkey Warehouses, Constellation Software, and Why Less Than 1% of Stock Pickers Beat the Market

    Mohnish Pabrai sat down with Shaan Puri to lay out exactly how he thinks about investing in 2026, walking through the ten commandments that have shaped a 27 year track record where every dollar invested in his oldest fund turned into roughly thirty. Watch the full conversation on YouTube here. Pabrai manages over a billion dollars, was close friends with Charlie Munger, has had lunch with Warren Buffett for 650,000 dollars, and has produced multiple 100 bagger investments in his career. This conversation is a complete operating manual for value investors, deep value hunters, and anyone trying to figure out how to compound capital in a market where the S&P trades at elevated valuations and AI capex is rewriting the rules.

    TLDW

    Pabrai argues that under one percent of stock pickers are actually good investors, that the game is a wealth transfer from the active to the inactive, and that temperament beats IQ every time. He walks through his core mental models: watching paint dry, the mistress versus the wife, introducing randomness into your life, cloning instead of inventing, taking a simple idea and taking it seriously, the too hard pile, no called strikes, the salmon spear, the inner scorecard, and don’t die at 25 and get buried at 75. He shares the full story of his 100 bagger Turkish warehouse company Reysas, his coal bets, his Constellation Software thesis around Mark Leonard, and why he is bearish on the S&P 500, bullish on pickaxe makers like TSMC and ASML conceptually but unwilling to pay current prices, and why GLP-1 drugs and Bitcoin both sit in his too hard pile. He retells Warren Buffett’s American Express salad oil crisis trade, the lesson Buffett delivered about Rick Guerin and leverage, the inner versus outer scorecard, and the Ed Thorp blackjack to Ken Griffin to Citadel chain. The closing punchline is that the most important investment any person can make is leading an aligned life, getting your music out, and discovering your calling before the wilderness years pile up.

    Key Takeaways

    • Well under one percent of Americans picking individual stocks are actually good at it. Index funds put you ahead of more than ninety percent of the crowd with zero effort.
    • The single biggest mistake smart investors make is impatience. Temperament, not IQ, decides outcomes.
    • Watching paint dry is the core skill. After making an investment, nothing may happen for three to five years. That is the nature of the game.
    • The mistress is always hotter than the wife. The stock you do not own looks more exciting than the one you do own because you do not know its flaws. The bar for swapping must be extremely high.
    • Raise your standards across the board: the investments you make, the people you hang out with, the relationships you keep. Buffett’s gravitational pull rule applies to both portfolios and friendships.
    • Introduce randomness into your life. Pabrai picking up a Peter Lynch book at Heathrow in 1994 led him to Buffett, Berkshire, Charlie Munger, bridge games, and his entire investing career.
    • Cloning works because almost no one will do it. Sam Walton copied everything. Walmart came from Kmart, Sam’s Club came from Sol Price’s Price Club, Burger King located across from McDonald’s instead of running their own site selection.
    • Elon Musk’s idiot index, calculating raw material cost on the London Metals Exchange and refusing to pay more than a small multiple over it, is the kind of simple framework no competitor will adopt even though it is publicly visible.
    • Take a simple idea and take it seriously. None of the other mental models work unless you commit fully to the first one.
    • The too hard pile is the most important box on a value investor’s desk. Buffett claims ninety eight percent of businesses belong there. Investing has no called strikes, so passing on ten thousand pitches before swinging is the right behavior.
    • The whale is swimming all the time, you only see it when it surfaces. Real investor activity is reading and studying, not trading.
    • Buffett at twelve gathered discarded racetrack tickets at Ax-Sar-Ben, found winners drunks had thrown away, and had his Aunt Alice cash them. He carried that pattern of finding anomalies into the Moody’s manuals in his twenties and into the Japan Company Handbook for two decades before pulling the trigger on the five Japanese trading companies.
    • The Japanese trading company trade was financed at half a percent in yen, the companies paid eight to nine percent dividends that later doubled, and Berkshire’s five billion has roughly doubled with almost no risk attached.
    • The American Express salad oil crisis taught Buffett to test the moat in the real world. He stood next to restaurant cash registers in Omaha, saw zero hesitation about accepting the card, and put forty percent of his fund into AMX.
    • The Buffett lunch lesson Pabrai still carries: a slightly above average investor who spends less than they earn and does not use leverage cannot help but get rich over a lifetime. Rick Guerin lost his Berkshire shares to margin calls in the 1973 to 1974 crash. Buffett bought them at forty dollars each, currently worth over seven hundred thousand.
    • Inner scorecard versus outer scorecard is the most fundamental life model. Buffett’s frame: would you rather be the greatest lover in the world and known as the worst, or the worst lover known as the greatest.
    • The Turkish stock market cycles through its float every seventeen days. Pure speculation. Indian quality companies trade at stratospheric valuations. The two together created a poker table Pabrai could sit at alone.
    • Reysas, the Istanbul warehouse operator, was bought at roughly three percent of liquidation value with a fifteen to sixteen million dollar market cap on eight hundred million in assets. It is now approaching a 100 bagger in dollars.
    • Pabrai’s thermonuclear event mental model: if ninety nine percent of humans were wiped out, someone would still produce Coke concentrate, because people will always trade fifteen minutes of labor for a Coke. Cement, paint, land, and steel are inflation indexed.
    • TAV Airports earned revenue in euros, paid costs in collapsing lira, traded at three to four times earnings on the Istanbul exchange. A natural monopoly hiding inside a panicked market.
    • The stock market is a church with a casino attached. Robinhood, prediction markets, zero day options, and two day options all increase the wealth transfer from the active to the inactive. Pabrai welcomes more casino activity because it helps his side.
    • On Polymarket, roughly one tenth of one percent of users capture sixty percent of profits. Two thousand traders made half a billion dollars in a year. The casual gambler funds the sharp.
    • On AI: invest in pickaxe makers. The alphabets and metas are playing a high capex game they have never played before. TSMC, ASML, and Micron are toll bridges. Pabrai is not making the bet because it sits between too hard, too expensive, and outside his circle of competence.
    • Constellation Software is Pabrai’s vertical SaaS bet because Mark Leonard built a mousetrap nobody else will clone. They acquire roughly two hundred small vertical market software companies a year, in delegated fashion, at five to six times cash flow that quickly becomes three to four times after revenue and license fee tweaks.
    • The market is wrong about AI killing software. Coding is one fifth of the pie. Adobe is not going out of business because someone can vibe code a Photoshop alternative. Incumbents reduce cost via automation while keeping cash flows intact.
    • Pabrai is bearish on the S&P 500. He agrees with Howard Marks that when the index trades at twenty three times earnings, the historical forward ten year return has bounced between minus two and plus two percent.
    • GLP-1 drugs sit in the too hard pile because industries with rapid change are the enemy of the investor. Ozempic to Mounjaro to upcoming tablets is too much turnover for valuations that already price in success.
    • Bitcoin sits in the too hard pile. Pabrai prefers gold and asks why a society that already has gold needs Bitcoin.
    • The four percent rule of compounding: roughly four percent of stocks have delivered the entire return of the US market over ninety years. Twelve investments built Berkshire across sixty years.
    • Investing rewards aging. Unlike basketball, the game gets easier with experience. Pattern recognition, expanded circle of competence, and the option to ride winners all compound.
    • Circle the wagons around your winners. Not selling Coke, not selling Apple, not firing Ajit Jain. The success of Berkshire was about not interfering with the four percent that worked.
    • Charlie Munger made an investment six days before he died at age ninety nine point nine. He invested like he was twenty five. Ben Franklin’s line: many people die at twenty five and are buried at seventy five.
    • Don’t save sex for old age. Don’t delay starting your real life until after the McKinsey rotation. Buffett’s frame transfers to careers too.
    • Ed Thorp wrote Beat the Dealer after the mob threatened him with a baseball bat for cleaning out single deck blackjack in Vegas. He then cracked options pricing before Black Scholes, ran Princeton Newport Partners, and became an early backer of Ken Griffin’s Citadel out of a Harvard dorm room.
    • Ken Griffin once told a Harvard recruit he wanted to quit at ten million dollars: please reject our offer, we do not want someone who dies at twenty five.
    • Lead an aligned life. Personality is largely baked by age five. The window to specialize is age eleven to twenty, which is exactly when the school system forces you to be a jack of all trades.
    • Get your music out. Every person has something specific they are meant to bring into the world. A misaligned life is the highest cost most people pay.

    Detailed Summary

    Why fewer than one percent of stock pickers are actually good

    Pabrai opens with a brutal estimate: well under one percent of the Americans who pick individual stocks are good at it. The game, he says, is a mechanism for transferring wealth from the active to the inactive. The good news is that index funds let anyone capture market returns with zero analytical work and end up ahead of more than ninety percent of active stock pickers. The implication is that anyone choosing to pick individual stocks is voluntarily entering a competition where the base rate of success is below ten percent, and the differentiator is almost never intelligence. It is temperament.

    That temperament shows up as patience. After making an investment, three to five years can go by with nothing happening. Sometimes the investment is a mistake and the patience converts into the discipline to reverse course. But on the whole, the less activity an investor takes, the better the outcomes. The first commandment is to enjoy watching paint dry.

    The mistress is always hotter than the wife

    The investments you already own are the wife. You see every flaw because you live with them every day. The investments you do not own are the mistress. Glamorous, unknown, exciting precisely because the temperament and the warts have not been revealed. Guy Spier, Pabrai’s longtime friend, deliberately stays reluctant to act on his portfolio. The point is not that you never act. The point is that the bar for action needs to be extremely high, and you have to learn to be comfortable passing on everything below that bar. Pabrai extends this directly to life: raise your standards about the people you spend time with, the projects you take on, and the investments you select.

    Introduce randomness into your life

    Charlie Munger told Pabrai over and over to introduce randomness into his life. The example Pabrai uses is his own origin story. He was an engineer running an IT company in 1994, sitting in Heathrow with his wife, looking for something to read on a flight. He picked up Peter Lynch’s One Up On Wall Street, finished it, picked up Beating the Street, finished that, encountered Buffett through a mention in Lynch, found the first two Buffett biographies fresh off the press, dove into the Berkshire and partnership letters, and within three years started attending the Omaha annual meeting. Every flight to Omaha on a Friday in May has both seatmates pre filtered for above average humans, all going for the same reason. The randomness exploded outward into Charlie Munger, the bridge games, Charlie’s friends, and decades of compounding social and intellectual capital.

    Shaan tells the parallel story of his own randomness bet. He flew to FarmCon in Kansas City for no particular reason, met newsletter operator Kevin Van Trump, cloned the model, launched The Milk Road for crypto, built the largest crypto newsletter in the world inside a year, and sold it for millions with one employee. Two mental models stacked: introduce randomness, then clone.

    Cloning is the cheat code no one will use

    Sam Walton freely admitted he had no original ideas. He walked into more competitor retail stores than any human in history. He looked at Sol Price’s Price Club, said no brainer, and opened Sam’s Club. He took his managers into a competitor and when they complained the store was a mess, he pointed at the one good candle display and said you can learn from anyone. He bought donuts at five thirty in the morning for distribution center drivers because they had ground level intel on every store. Walmart’s market cap dwarfs that of every competitor combined, and every system came from somewhere else.

    Tesla, SpaceX, and the Boring Company exist because Elon Musk applies the idiot index. He asks what raw materials go into a part, looks up the price on the London Metals Exchange, and refuses to pay a multiple over it without a fight. None of his competitors think this way even though he has written about it publicly and Walter Isaacson devoted a book to it. SpaceX intentionally blows up rockets to learn. Blue Origin tries hard not to blow up rockets. SpaceX is miles ahead. Burger King famously assigned two guys to track McDonald’s site selection and just put a store across the street. The reason cloning works so well is that almost no one is willing to do it.

    Take a simple idea and take it seriously

    This is the bedrock model that makes every other model work. Without total commitment to one simple idea, the rest of the mental models stay theoretical. Pabrai went to Turkey on a hunch in 2018 because the market screened cheap. He discovered that Turkish public companies cycle through their float every seventeen days, meaning every shareholder turns over more than twenty times a year. Compared with Berkshire, whose float may take a decade to rotate, Turkey is a hyperactive day trader’s casino. India, by contrast, has roughly one hundred to one hundred fifty quality companies, all picked over and priced at stratospheric multiples. The Turkish market gave Pabrai a poker table he could sit at almost alone. He chose to be an inch wide and a mile deep.

    Circle of competence and the too hard pile

    Pabrai’s eighth commandment is thou shalt not use Excel, and his ninth is that if you cannot explain an investment thesis to a ten year old in about four sentences, it is a pass. Investing is journalism more than spreadsheet work. Buffett stood at restaurant cash registers in Omaha during the salad oil crisis to test whether AMX’s brand had cracked. He walked into Snow White with his briefcase to study Disney. Peter Lynch told amateurs to make a list of every brand they consume because that is the most authentic intel they have. Buffett keeps a too hard box on his desk and claims ninety eight percent of businesses go in it. Investing has no called strikes, which means an investor can let ten thousand pitches go by and only swing at the fattest center cut pitch.

    The whale is swimming all the time. You only see it when it surfaces. Buffett at age twelve gathered discarded racetrack tickets at Ax-Sar-Ben to find ones drunks had thrown away, then had his aunt Alice cash them because he was underage. In his twenties he flipped through Moody’s manuals page by page looking to be hit in the head with a two by four. Western Insurance at fifteen dollars made twenty five dollars a share and had forty dollars of cash on the balance sheet. He has been flipping through the Japan Company Handbook for at least twenty years before pulling the trigger on the five Japanese trading companies, financing the entire five billion in yen at half a percent against eight to nine percent dividend yields that have since doubled. Berkshire’s five billion is now ten billion paying eight hundred million a year, essentially risk free.

    The 650,000 dollar lunch and what Buffett actually said

    Pabrai paid 650,000 dollars to have lunch with Warren Buffett in 2007 because his net worth had hit eighty four million dollars almost entirely from intellectual property he had taken from Buffett for free. He wanted to look him in the eye and thank him. Buffett’s stance on the lunches was that whoever paid should feel like they got a bargain, so he came prepared having studied biographies of every guest. Pabrai asked an innocent update question about Rick Guerin, the third partner of Buffett and Munger in the sixties and early seventies who then disappeared. Buffett’s answer became the lesson of the lunch. Charlie and I knew we were going to be rich, but we were not in a hurry. Rick was in a hurry. He was always levered. The 1973 to 1974 crash, the slowest motion crash in modern history, gave him margin calls. Buffett bought Rick’s Berkshire shares for forty dollars each. They are over seven hundred thousand now. The lesson: if you are even a slightly above average investor and you spend less than you earn and you do not use leverage, you cannot help but get rich over a lifetime.

    The other lunch lesson Pabrai still cites is the inner scorecard. Buffett’s framing: would you rather be the greatest lover in the world and known as the worst, or the worst lover known as the greatest. The answer determines whether you can resist external stimuli and stay centered. The way Pabrai practices it is by remembering that even Gandhi has critics. If Gandhi is fair game, so is anyone else with a public footprint.

    The Turkey trade and the thermonuclear event mental model

    The headline Turkey investment is Reysas, an Istanbul warehouse operator Pabrai started buying when the market cap was fifteen to sixteen million dollars on eight hundred million dollars of assets, roughly three percent of liquidation value. He told the broker to take out every ask up to the ten percent daily price limit. Templeton Fund called offering five percent of the company for a million dollars and Pabrai said why are you even calling, just take it. Templeton was exiting Turkey because of currency instability and inflation, both of which Pabrai considered irrelevant for the specific kinds of assets he was buying.

    The mental model that unlocked Turkey was the thermonuclear event scenario he discussed with Charlie Munger. If ninety nine percent of humans were wiped out, someone would still produce Coke concentrate and rebuild a bottling plant. The remaining seventy million people will trade fifteen minutes of labor for a Coke regardless of currency or exchange rate. A warehouse is land, paint, cement, and steel. All four are inflation indexed. Whatever happens to the lira, those assets do not care. When the lira collapsed ninety percent against the dollar in seven years, Reysas went up roughly ninety times in dollars and effectively to infinity in lira. He applied the same logic to TAV Airports, which collected revenue in euros while paying costs in collapsing lira. A natural monopoly trading at three to four times earnings on a panicked exchange.

    The casino, prediction markets, and Polymarket

    Buffett’s line at the most recent Berkshire meeting is that the stock market is a church with a casino attached, and the casino is getting crowded. Robinhood, two day options, leverage, and prediction markets like Polymarket all funnel casual gamblers into a transfer game where the sharps already know the prices. Pabrai notes that on Polymarket, roughly one tenth of one percent of users capture sixty percent of profits, and two thousand traders made half a billion dollars in a year. The horse track keeps twenty one percent of every dollar, Vegas keeps two to four percent on a great blackjack game, and yet some players still make a living off horse racing by spotting odds that make no sense. The casino activity is bad for society and great for any investor patient enough to wait for the obvious mispricing.

    AI, pickaxe makers, and the too hard pile

    On AI, Pabrai says invest in pickaxe makers. The alphabets and metas are playing a high capex game they have never played before, which is a recipe for surprise. The capex must pass through TSMC, ASML, and probably Micron. But all of those toll bridges are either too expensive, outside Pabrai’s circle of competence, or in his too hard pile. He is not making the bet. There is no scenario where he sells his Turkish warehouses to buy TSMC. The mistress, in this case, looks uglier than the wife and there are no bonus points for clever valuation work.

    Constellation Software, Mark Leonard, and vertical SaaS

    Where the market gets AI wrong is the assumption that AI coding kills software. Coding is at most one fifth of a software business. The market assumes Adobe is dead because someone can vibe code Photoshop. Pabrai disagrees. Incumbents will reduce costs through automation, keep cash flows intact, and possibly cut prices without losing margin. He invested in Constellation Software specifically because Mark Leonard has built a mousetrap nobody else will clone. Constellation’s M&A team touches seventy to one hundred thousand private vertical market software companies twice a year by phone and twice by email. They acquired roughly two hundred companies last year alone, never using bankers. They pay five to six times cash flow, then bump revenue and license fees about twenty percent, and the effective purchase price drops to three or four times within a year or two. The model is delegated, with deal authority pushed out to teams that do not need headquarters approval up to a threshold. They buy and hold, which scares away private equity that wants to flip. The universe of vertical SaaS targets is too small for private equity to bother with and big enough to keep Constellation compounding for decades. Mark Leonard is the kind of unicorn operator who does not appear twice in a generation.

    S&P bearish, GLP-1 too hard, Bitcoin too hard

    Pabrai is bearish on the S&P 500 because at roughly twenty three times earnings, historical forward ten year returns have ranged between minus two and plus two percent. Howard Marks’s analysis matches his own. GLP-1 drugs like Ozempic and Mounjaro are generating roughly seventy nine billion in revenue annually, more than the entire AI economy, but Pabrai puts them in the too hard pile because industries with rapid change are the enemy of the investor. Ozempic to Mounjaro to upcoming oral tablets is too many turns of the wheel. Bitcoin sits in the same too hard pile. Pabrai prefers gold and asks why a society that already has gold needs Bitcoin, which is widely used by scammers and ransomware operators.

    The four percent rule and circling the wagons

    Over the past ninety years, roughly four percent of US stocks have delivered the entire market return. The other ninety six percent have treaded water. Buffett himself has made three to four hundred investments and only twelve of them built Berkshire Hathaway. Index funds work because they are too dumb to sell Nvidia and too dumb to sell TSMC. Active investors and portfolio managers second guess winners and trim them. The most important investing discipline is circling the wagons around winners: not selling Coke, not selling Apple, not firing Ajit Jain. Capitalism is brutal and most businesses go to zero eventually. The thin slice of enduring moats, like FICO, McDonald’s brand, prime Istanbul warehouses, airport monopolies, and Coke bottlers, are what compound for decades. Pabrai’s bets on coal, airports, warehouses, and Constellation do not all need to work. If half work, the portfolio is a home run. If forty percent work, still a home run. Investing is a forgiving game.

    Ed Thorp, Ken Griffin, and the chain of investing genius

    Ed Thorp used MIT’s mainframe in the early sixties to crack single deck blackjack with basic strategy and card counting. He cleaned out mob run Vegas casinos until they showed him a baseball bat. To get back at them he wrote Beat the Dealer, which sold millions of copies and forced the industry to introduce multi deck shoes and rule changes. He then cracked options pricing before Black-Scholes and skipped the Nobel Prize to make money on it through Princeton Newport Partners, compounding at twenty five to thirty percent a year with no down years. He met a young Ken Griffin running Citadel out of a Harvard dorm and not only handed over algorithms but became an early backer. Pabrai’s first meeting with Thorp happened in a racquetball locker room while Pabrai was completely naked, copy of The Wall Street Journal next to him. Thorp introduced himself, Pabrai’s excitement overcame his sense of decorum, and they have been friends ever since. The Ken Griffin lore extends to his recruiting filter: a Harvard recruit who said he would quit at ten million dollars was told to please reject the offer, because Citadel does not want people who die at twenty five.

    Don’t die at twenty five and get buried at seventy five

    Ben Franklin’s line that many people die at twenty five and get buried at seventy five becomes Pabrai’s closing frame. Charlie Munger made an investment six days before he died at age ninety nine point nine. He invested like he was twenty five. The whole point of life is to keep growing, keep learning, keep finding the alignment between who you are inside and how you show up in the world. Personality is largely baked by age five, and after twelve the most a parent can really influence is the peer group. The window to specialize runs from age eleven to twenty, which is precisely when most educational systems force kids to be jacks of all trades. Bill Gates slipped out of his house to code through the night and accumulated ten to twenty thousand hours by his early twenties. Buffett picked stocks at eleven. Michelangelo sculpted at ten.

    The most important thing Pabrai wants viewers to take from the conversation is that an aligned life is more important than a great investment record. Get your music out. Find what energizes you. If you do not know your calling, work with a thoughtful industrial psychologist like Jack Keene or pay attention to which activities and people genuinely energize you and which drain you. Pabrai himself wandered the wilderness until his mid thirties when he finally understood his own calling. Buffett’s frame applies: do not save sex for old age and do not save your real work for after the McKinsey rotation.

    Thoughts

    The most useful thing Pabrai does in this conversation is collapse the gap between life philosophy and portfolio construction. Most investing content treats temperament as a soft skill on the side of the spreadsheet. Pabrai puts it where it belongs, at the center. The reason the four percent rule matters is not statistical, it is psychological. Almost everyone can identify the few enduring compounders. Almost nobody can sit on them for forty years without selling, trimming, switching to a hotter mistress, or breaking discipline on a leverage call. The actual edge in public markets is not analytical, it is the willingness to be inactive in the face of constant pressure to act.

    The Turkey trade is the most instructive case study in the whole conversation because it is genuinely replicable. Not the specific market, but the architecture. Pabrai stacked four simple mental models on a single trade: take a simple idea seriously, identify a market where the float churns so fast that price has no relationship to value, isolate assets whose intrinsic worth is currency independent, and run the thermonuclear event sanity check on the underlying demand. The result was a 100 bagger held in roughly the worst macro environment of his investing career. The lesson is not to go to Istanbul. It is that real edge tends to come from combining three or four boring frameworks at the same time, in a place where nobody else is bothering to combine them.

    The Constellation Software section deserves more attention than it gets in most investor decks. Pabrai is making a clean bet that vertical SaaS is misread by the market because of generic AI fear. He is probably right. Coding is a labor input to software, not the moat. Switching costs, regulatory tangles, integration depth, and decades of accumulated workflow customization are what keep customers paying. Mark Leonard has industrialized the act of acquiring those moats two hundred times a year. If the DNA holds after Mark eventually steps back, the math is hard to beat. The asymmetric risk is leadership transition, not technological disruption.

    The AI commentary is more interesting for what Pabrai refuses to do than for what he says. He acknowledges the pickaxe makers thesis, names the toll bridges, and then explicitly declines to make the bet because the valuations are too high and the path forward is genuinely uncertain. That is the discipline of the too hard pile in action. Plenty of investors right now are putting money to work in TSMC and ASML at multiples that bake in success scenarios, telling themselves they have done the homework. Pabrai’s position is that even when you are largely right about a trend, paying any price for it is a mistake. The structural humility there is the actual lesson.

    The aligned life closing argument hits hardest because it reframes the entire conversation. The ten commandments of investing are a subset of a broader operating system: figure out who you are by age twenty if you can, raise your standards on the people you spend time with, do not borrow against tomorrow, do not chase the mistress, and do not save your real ambitions for old age. Investing is just the highest leverage application of those rules. The viewers who walk away with one usable change probably should not be picking new stocks. They should be auditing whether they are leading the life that fits.

    Watch the full conversation with Mohnish Pabrai and Shaan Puri on YouTube here.

  • Howard Marks on Why Most Investors Lose, the AI Bubble, India, and the Hunt for the $10 Bill Nobody Picked Up

    TLDW

    Howard Marks, co-founder of Oaktree Capital and the author of the memos every serious investor reads first, sat down with Nikhil Kamath for a wide-ranging conversation on his 50+ year career, the philosophy of Mujo (the inevitability of change), why he chose bonds over stocks, the difference between drifting down the river and seeing it, where we sit in the current cycle, AI as both threat and opportunity, why active management lost to indexation, and why the only way to outperform in a world full of smart, motivated, computer-literate competitors is “superior insight.” His core message: investing is a puzzle that cannot be solved by formula, and the only edge that lasts is being more right than the other person, more often, with the discipline to stay calm when everyone else is panicking or partying.

    Key Takeaways

    • Mujo is the operating system. Marks took Japanese literature at Wharton and walked away with one idea that shaped his whole career: change is inevitable, unpredictable, and uncontrollable. You cannot predict the future, but you can prepare for it.
    • Cycles are excesses and corrections, not ups and downs. The S&P 500 has averaged about 10% per year for 100 years, but it is almost never between 8% and 12% in any given year. The norm is not the average. Greed and fear push the pendulum past equilibrium every time.
    • The recovery is two years older. When asked where we are in the cycle, Marks notes the bull market continued from April 2024 through January 2026, so by definition we are deeper into the cycle, with a recovery distorted by the unique man-made COVID recession.
    • Drifting versus seeing the river. Marks describes the first 35 years of his career (roughly age 14 to 49) as drifting. Starting Oaktree in 1995 was the first truly intentional decision he made. Entrepreneurship forced proactivity on him.
    • Why bonds over equities. The contractual, predictable nature of debt suited his conservative temperament (his parents were adults during the Depression). He was not voluntarily moved to bonds in 1978; a boss reassigned him just in time for the birth of the high-yield bond market.
    • Distressed debt is the bigger story. Bruce Karsh joined in 1987 and has run roughly $70 billion in distressed debt since 1988, with profits well over 90% of the total profit and loss.
    • Excess return is getting paid more than the risk warrants. If the market thinks a borrower has a 5% default probability and you correctly conclude it is 2%, you collect interest priced for 5% risk while taking 2% risk. That gap is the alpha.
    • Oaktree’s default rate is about a third of the market. Over 40 years, roughly 3.6% to 3.7% of high-yield bonds default each year. Oaktree’s rate is roughly one-third of that, achieved through process discipline, institutional memory, and analysts who stay analysts for life.
    • If you are starting a career today, understand AI. Marks says the investor who will make the most money over the next 10 years is the one who best understands AI and its capabilities, whether they bet for or against it.
    • AI is excellent at pattern matching, but cannot create new patterns. Can AI pick the Amazon out of five business plans? The Steve Jobs out of five CEOs? Marks bets no. Most humans cannot either, which means there is still a role for exceptional people.
    • Indexation won because active management lost. Passive did not become dominant because it is brilliant. It dominated because most active managers failed and charged high fees for the privilege.
    • Bad times create openings for active managers, but most cannot take them. Panic drives prices down, but the same panic prevents most investors from buying. Wally Deemer: when the time comes to buy, you will not want to.
    • The job is simple but not easy. Find the best managers, the best companies, the best ideas. Charlie Munger told Marks: anyone who thinks it is easy is stupid.
    • Where is the $10 bill nobody picked up? Marks thinks it is around AI, but only for those with insight above the average. If you are average and you crowd into AI, you get average results in a bull case and worse in a bear case.
    • Quantitative information about the present cannot produce alpha. Andrew Marks (howards son) pointed this out to his father during the COVID lockdown. Everyone has the same data. Outperformance has to come from somewhere else.
    • Buffett’s edge was reading Moody’s Manuals when nobody else would. The pre-internet research process favored those willing to do tedious work alone. The format of the edge changes; the fact that edge requires doing what others will not, does not.
    • You cannot coach height. Marks can tell you that second-level thinking, contrarian insight, and the ability to evolve at 80 are essential. He cannot tell you how to acquire any of them.
    • India: Marks declines to opine. He has deployed roughly $4 billion in India but refuses to claim expertise on the Indian stock market or recommend a sector.
    • History rhymes. Marks credits Mark Twain. The lessons that repeat are lessons of human nature, which changes incredibly slowly.
    • Investing is a puzzle, not dentistry. Quoting Taleb, Marks observes that engineers and dentists succeed by repeating the right answer. Investors face a problem with no certain solution. If you need to be right every time, do not become an investor.

    Detailed Summary

    From Queens to Wharton: The Accidental Investor

    Howard Marks grew up in Queens, New York, in a middle-class family. Neither of his parents went to college, but his father was an intelligent accountant. Marks discovered accounting in high school, fell in love with its orderliness, and chose Wharton because he was told it was the best undergraduate business school in America. Wharton required a literature class in a foreign country and a non-business minor. For reasons he no longer remembers, Marks chose Japanese studies, then took Japanese civilization and Japanese art. He calls it the most important academic decision of his life because of one concept he encountered: Mujo.

    Mujo, Independence of Events, and Why You Cannot Predict

    Mujo, the turning of the wheel of the law, teaches that change is inevitable, unpredictable, and uncontrollable, and that humans must accommodate it rather than try to control it. Marks pairs this with his deep belief in the independence of events: ten heads in a row do not change the odds on flip eleven. Roughly 20 years ago he wrote a memo titled “You Can’t Predict. You Can Prepare.” A portfolio cannot be optimized for both extreme upside and extreme downside, but it can be built to perform respectably across many possible futures, if you suboptimize for the middle of the probability distribution.

    Why Cycles Exist

    If GDP averages 2% growth, why is it never simply 2%? Marks’s answer is excesses and corrections. Optimism leads producers to overbuild and consumers to overspend, growth runs above trend, then satiation and oversupply pull it back below trend. The S&P 500 averages 10% per year over a century, but the return in any given year is almost never between 8% and 12%. The norm is not the average because human beings are not average; they are alternately greedy and fearful.

    Where Are We Now?

    Two years ago Marks told the Norwegian Sovereign Wealth Fund’s Nicolai Tangen that we were near the middle of the cycle. Two years later, the bull market in stocks continued through January 2026, so by simple math the recovery is older. The COVID recession was a man-made anomaly: one quarter of negative growth followed by the best quarter in history, triggered by a deliberate global shutdown rather than by accumulated excess. That distorts every traditional cycle metric.

    Drifting Versus Seeing the River

    One of the most personal moments in the conversation is Marks’s confession that he drifted for the first 35 years of his career. He did not pick his career, his first job, or his transition from equities to bonds in any deliberate way. Other people pushed him; he said yes. The first proactive decision of his life was co-founding Oaktree in 1995 at age 49, and even that came largely because his wife and his partner Bruce Karsh pushed him into it. Once he had to lead, he had to be intentional. Leadership cannot be passive.

    The Bond Decision

    Marks did not choose bonds; bonds chose him. In May 1978 his boss at Citibank moved him to the bond department to start a convertible fund. Three months later another phone call asked him to figure out something called high-yield bonds being run by a guy in California named Milken. Marks said yes both times. He arrived at the front of the line for high-yield in 1978 and has been there for 48 years.

    The conservative temperament fit. Marks’s parents were adults during the Depression, so he grew up hearing “don’t put all your eggs in one basket” and “save for a rainy day.” Bonds offered contractual, predictable returns. The phrase “junk bonds” was a bias that made the asset class cheaply available to anyone willing to do the analytical work.

    Distressed Debt and Excess Return

    When Bruce Karsh joined in 1987, Oaktree launched what Marks believes was the first distressed debt fund from a mainstream institution. Karsh has managed about $70 billion since 1988 with well over 90% of the total being profit. The core skill is predicting default probability better than the market. If consensus prices a borrower at a 5% default risk and you correctly assess 2%, the interest you receive is overpaid relative to actual risk. Marks calls this “excess return” and credits Mike Milken with the foundational insight: lend to borrowers others will not, demand interest beyond what compensates you, and the math works.

    Over 40 years, roughly 3.6% to 3.7% of high-yield bonds default annually on average. Oaktree’s default rate has been roughly one-third of that. Marks credits institutional culture (analysts who stay analysts for life), psychological stability in volatile periods, and a process that forces every analyst to ask the same eight questions of every company every time. In equity research, you can buy a stock for great management without examining the product, or for a great product without examining the management. In Oaktree’s bond process, you cover every base every time.

    Beginning a Career Today: The AI Question

    Asked what he would do today, Marks says the front of the line is AI. The investor who will succeed most over the next decade is the one who best understands AI, whether they bet for or against it. He notes that he was shocked by his own experience using Claude, but adds that he has not fired a single person and does not intend to.

    His view: AI excels at extracting patterns from history and applying them with discipline and without psychological wobble. But investing also requires creating new patterns. Can AI sit with five business plans and identify the future Amazon? Can it sit with five CEOs and pick Steve Jobs? Marks bets not. Then he adds the killer line: most humans cannot either. Which means the role for exceptional humans survives, but the bar gets higher.

    Why Indexation Won

    When Marks went to graduate school at the University of Chicago in 1968, his professor pointed out that most mutual funds underperformed the S&P after fees. Index funds did not exist yet; Jack Bogle launched the first one in 1974. Today, most equity mutual fund capital is passive. Marks’s controversial take: indexation did not win because it is great. It won because active management was so bad and so expensive. Even at equal fees, if active decisions are inferior, passive wins.

    Bad times create openings for active managers because panic drives prices down, but the same panic prevents most people from buying. Marks quotes the old trader Wally Deemer: when the time comes to buy, you will not want to. The advantage of an AI nudge that says “this is one of those moments, get your ass in gear and buy something” might genuinely add value, because it removes the emotion.

    Second-Level Thinking and Why You Cannot Coach It

    Marks’s first book, The Most Important Thing, has 21 chapters, each titled “The Most Important Thing Is…” Each one is different because so many things matter. The chapter on second-level thinking came to him spontaneously while writing a sample chapter for Columbia University Press. The argument is simple: if you think like everyone else, you act like everyone else, and you get the same results. To outperform, you must deviate from the herd and be more right than the herd. Different is not enough. Different and better is the bar.

    Can AI become a contrarian thinker? You can prompt Claude to give you only non-consensus answers, but the catch is that consensus is often close to right because the people building consensus are intelligent, educated, computer-literate, and motivated. Forcing non-consensus often forces wrong. The real edge is being non-consensus AND correct, which is a much narrower target.

    The $10 Bill That Nobody Has Picked Up

    Marks references the joke about the efficient market hypothesis: there is no $10 bill on the sidewalk because if there were, somebody would have already picked it up. He then concedes that the bill is probably around AI today, but only for those whose insight rises above the average. If you are average and you crowd into AI, you go along with the tide if it works and get crushed if it does not. Quoting Garrison Keillor’s Lake Wobegon, “where all the children are above average,” Marks notes that the math does not allow it. Most investors will not be above average, and acknowledging that is the first step toward becoming one of the few who are.

    Learning From Andrew, Buffett, and Onion-Skin Manuals

    Marks lived with his son Andrew during COVID and wrote a memo about it called “Something of Value” in January 2021. Andrew’s most important contribution was a near-revelation: readily available quantitative information about the present cannot be the source of investment alpha because everyone has it. Buffett’s edge in the 1950s was reading Moody’s Manuals (giant books printed on onion-skin paper with tiny type and zero narrative) when nobody else would. The medium changes; the principle that edge requires doing what others will not, does not.

    India

    Kamath asks Marks directly about India. Marks has deployed roughly $4 billion there but politely declines to claim any expertise on the Indian stock market or recommend a sector. He cautions Kamath about taking advice from people who do not know what they are talking about, and includes himself in that category on the question of India. The honesty is striking and is itself an investment lesson.

    History Rhymes, and Final Advice

    Marks reads Andrew Ross Sorkin’s 1929 and references it in an upcoming memo on private credit. He likes Mark Twain’s reputed line that history does not repeat but it rhymes, and Napoleon’s line that history is written by the winners of tomorrow. The lessons that rhyme are lessons of human nature, which evolves incredibly slowly. Fight or flight from the watering hole still drives behavior in financial markets.

    His final advice: investing is a puzzle, not engineering. A civil engineer calculates steel and concrete, builds the bridge, and the bridge stands. Every time. A dentist fills the cavity correctly and it stays filled. Every time. If you need that kind of reliability in your work, become a dentist. Investing is the act of positioning capital for a future that cannot be predicted accurately. You will be wrong sometimes. If something in your makeup cannot tolerate being wrong sometimes, do not become an investor. The puzzle has no final solution, which is exactly what makes it endlessly interesting.

    Thoughts

    The most useful thing Marks does in this conversation is admit, repeatedly and without ego, what he does not know. He does not know whether AI models differ in real intelligence. He does not know which sector in India to bet on. He does not know how to teach second-level thinking. He drifted for 35 years and only began making intentional decisions at 49. This honesty is the inverse of every guru selling certainty, and it is the actual content of the lesson he is trying to convey: epistemic humility is the precondition for superior insight, because you cannot acquire what you already think you have.

    The deepest insight in the conversation might be the one Andrew Marks (Howard’s son) gave his father during COVID: readily available quantitative information about the present cannot produce alpha because everyone has it. This is devastating in the AI era. If everyone is asking the same large language model the same question, the answers converge, and convergence is consensus, and consensus does not pay. The arms race for proprietary data, novel framings, and unconventional questions is the only thing that can break the convergence.

    Marks’s framing of cycles as excesses and corrections rather than ups and downs is genuinely useful. It reframes volatility from something to fear into something to expect, and reframes the question from “where are we going?” to “how far past trend have we already gone?” The 8 to 12 percent observation about the S&P (that the average return is almost never the actual return) is the kind of fact that should be taught in every introductory finance class but is almost never mentioned.

    The most contrarian claim in the conversation is the one about indexation: that it won because active was bad, not because passive is great. This is a useful inversion. Most defenders of passive investing argue from efficient market theory; Marks argues from the empirical failure of active managers. The implication is that if you can find the small population of active managers who genuinely outperform, the indexation argument falls apart for that subset. Most cannot. The hardest job in investing is the meta-job of identifying the few who can.

    The exchange about AI as a contrarian engine is one of the most clarifying short discussions of AI’s investment limits I have read. Different from consensus is easy. Different and better is the actual goal. Forcing different gets you wrong more often than right because consensus, built by smart, motivated, educated competitors, is usually close to correct. This is why “use AI to find non-consensus ideas” is a worse strategy than it sounds.

    Finally, the Buffett-Moody’s-Manual story is the most quietly profound moment in the interview. The edge in 1955 was the willingness to read tiny type on onion-skin paper alone in an office in Omaha when no one else would. The edge in 2026 is whatever the modern equivalent of that is, and the only honest answer is: nobody knows yet, which is precisely why finding it is worth so much money.