Dan Loeb has spent three decades turning a $3 million fund into Third Point, a roughly $25 billion collection of hedge fund, credit, insurance, and venture businesses. In this Invest Like the Best conversation with Patrick O’Shaughnessy, Loeb walks through how he reinvented his strategy from deep value and event-driven trades into quality and thematic investing, why he now believes every serious investor has to be a technology investor, how he reads the AI cycle and the semiconductor melt-up, where activism and corporate governance still pay, and the single mistake that taught him the most. It is a rare, unhurried look at how a famously sharp-elbowed activist actually thinks about markets, businesses, and people.

TLDW

Loeb covers an enormous amount of ground: his daily process for staying ahead of the information firehose, Jensen Huang’s AI stack as a mental model, and why Nvidia, Anthropic, and Elon Musk’s companies are the three most consequential firms he tracks. He traces Third Point’s roots in credit and event-driven investing at Jefferies, the influence of Joel Greenblatt’s “You Can Be a Stock Market Genius,” and his later pivot to quality investing shaped by “The Outsiders” and Lawrence Cunningham’s “Quality Investing.” He argues the AI rally is not a dot-com-style valuation bubble because the leaders generate enormous cash, explains why human judgment and structural market quirks still create alpha, and makes the case that AI will never fully run a capital system. He digs into corporate governance and his father’s influence, the Sotheby’s and Sony activism campaigns, the hard reality of activism in Japan, and what investing in Danaher’s operating system taught him. He names FTX as his hardest lesson, breaks down Third Point’s evolution into a 60-percent-credit platform spanning CLOs, structured credit, reinsurance and annuities, describes how he is pushing his analysts to use AI and Claude daily, and closes on kindness and the friend who let him sleep on a couch before he made it.

Thoughts

The most striking thing about Loeb is that he treats his own strategy as a thing to be disrupted rather than defended. He built his reputation on Greenblatt-style special situations, spin-offs, demutualizations, and post-reorg equities bought cheap because of forced selling and sandbagged guidance. Most investors who win that way spend the rest of their careers protecting the formula. Loeb instead watched the people who stayed rigid about deep value and low multiples underperform or disappear, and deliberately retrained himself and his team around business quality and thematic conviction. The willingness to abandon a winning identity is the actual edge here, more than any single trade. It is the rare investor who can say his current strategy would not fit cleanly on a PowerPoint deck and treat that as a feature.

His AI framing deserves attention because it is unfashionably calm. The bear case on AI is usually about valuation, and Loeb dismantles it on the leaders’ own numbers: these are companies investing off their balance sheets, generating enormous cash, trading at multiples that do not resemble 1999. He was short the dot-com bubble, so he is not a permabull cheering from the sidelines. His real point is subtler, that the danger is expectations, not valuations. The semiconductor index ran up 40 percent on genuinely strong fundamentals, but Micron and Nvidia both put up monster quarters and saw their stocks fall because expectations had simply outrun even great results. That gap between fundamentals and price is where he thinks the human investor still earns a living, precisely because quant strategies, CTAs, and risk-managed pods are forced to sell into weakness rather than buy it.

The governance material is the most quietly radical part of the conversation. Loeb defends shareholder primacy against the Business Roundtable’s softer stakeholder language, but his argument is not the cartoon version where shareholder value means strip-mining a company. It is that boards have one job, accountability for capital allocation and management, and that vague multi-stakeholder mandates become an excuse for directors to avoid the hard work. His read on bad governance is almost always relational: directors who let loyalty to an underperforming CEO override their duty, or who sit on boards for status and income. The Sotheby’s story is the clean illustration, a centuries-old, high-status business run unprofitably because nobody treated it like a business. Loeb’s pattern is to find the gap between claimed status and actual performance and to raise the social cost of coasting.

What is genuinely new in Loeb’s posture is how he talks about AI inside his own firm. He is not pitching it as a moat or a headcount-reduction story. He frames Claude and AI tools as a way to make each person a more autonomous self-improver, something that gives back whatever you put into it, with some analysts running agents overnight and burning tokens while he personally uses it more for queries. Coming from a 30-year fundamental investor, the absence of defensiveness is the signal. He pairs it with Brad Gerstner’s nod to “Essentialism”: the firehose is now infinite, so the scarce skill is deciding what is actually relevant. That is a more honest answer to the AI question than either doom or hype.

Finally, the FTX confession is worth sitting with because of how he frames it. He does not retreat into cynicism about venture or crypto. He notes that Sam Bankman-Fried, fraud aside, had a real nose for value, with stakes in Anthropic, Cursor, and Solana that would have made him a top venture investor of the era. The lesson Loeb extracts is procedural, not philosophical: their due diligence now includes checking bank balances, the most basic verification that would have surfaced the problem. It is a useful reminder that even sophisticated capital can skip boring fundamentals when a company is growing fast and the cap table looks good. The discipline is not in having a grand theory of fraud, it is in never skipping the unglamorous checks.

Key Takeaways

• Loeb’s macro focus right now collapses to two variables: where oil goes, dictated by war and geopolitics, and what AI does on the spending and infrastructure front and its impact on society and the economy.
• He argues you can no longer punt on technology and focus on industrials or consumer; tech is a big, growing, compounding part of the economy that affects everything else, so every investor has to become a tech investor.
• He uses Jensen Huang’s AI stack as a mental model: power and energy at the bottom, then chips and infrastructure, up through large language models, software, and applications.
• The three most consequential companies he tracks are Nvidia, Anthropic, and Elon Musk’s companies collectively.
• Third Point’s roots are in credit and event-driven investing, shaped by his time at Jefferies watching investors like David Tepper before he founded Appaloosa, Eric Mindich at Goldman, and firms like Angelo Gordon and Farallon.
• Joel Greenblatt’s “You Can Be a Stock Market Genius” was his foundational framework: spin-offs, demutualizations, privatizations, and post-reorg equities where a new, illiquid security gets dumped by holders who will not do the work.
• Spin-off managers often sandbag guidance because their incentive packages get set at the time of the spin-off, creating a predictable gap between conservative numbers and real value.
• From 1995 to roughly 2013-2015, event-driven special situations were Third Point’s bread and butter; those opportunities still exist, but the real edge now is overlaying them with a business-quality lens.
• The pivot to quality and thematic investing was influenced most by “The Outsiders” (capital allocation plus great operations) and Lawrence Cunningham’s “Quality Investing” (high-moat, high-return-on-capital businesses to own for years).
• AI disruption made last year one of the worst for many apparently high-quality companies, as businesses that looked durable rapidly became less so.
• Loeb sees the AI rally as fundamentally different from the dot-com bubble: the leaders invest off their balance sheets, generate enormous cash, and do not carry the valuation excess of 1999.
• The danger in semis is expectations, not valuation: Nvidia and Micron posted spectacular quarters yet saw stocks fall because expectations had outrun even great numbers.
• Structural forces still create alpha for fundamental investors: quants, CTAs, and multi-strategy pods have risk metrics that force selling on the way down, the opposite of what is rational for long-term holders.
• He believes AI will not fully run a capital system; private equity, restructurings, creditor committees, and high-touch negotiation will always need humans.
• His interest in governance came from his father, a securities lawyer and corporate governance expert who sat on the boards of Mattel and Williams-Sonoma and pushed ethical sourcing ahead of his time.
• Loeb defends shareholder primacy, citing Milton Friedman and Warren Buffett, and criticizes the Business Roundtable’s move away from shareholder value as a distraction from the board’s real duty.
• Bad governance usually comes from directors letting loyalty to a weak CEO override fiduciary duty, lacking the knowledge to do the job, or serving for status and income.
• Writing is a core activism lever: great writing is clear thinking, and social pressure through writing and PR is one of the most effective ways to move a board, alongside financial and legal levers.
• The Sotheby’s campaign targeted a high-status, centuries-old business run unprofitably; Third Point bought 9.9 percent, eventually brought in Tad Smith from MSG, who cleaned up operations and technology before the company sold.
• Third Point increasingly prefers to back great companies with excellent management and cheer them on rather than hunt for mismanaged businesses, because bad management tends to cluster into a morass.
• Third Point is a collection of businesses; the flagship hedge fund grew from $3 million to about $9 billion and is roughly 30 percent credit, with the broader firm closer to 60 percent credit.
• The firm spans a roughly $7 billion CLO business, structured and corporate credit, an insurance company, asbestos liabilities, a small private credit unit, and a venture capital arm.
• The unifying thread is valuing enterprises across early, mid, and mature stages and investing in whichever fulcrum security offers the best risk-reward, from equity to senior debt.
• Loeb cites buying Twitter’s financing debt near 96-97 cents at a 12 percent yield when most credit investors were scared, and a difficult xAI debt financing, as examples of cross-discipline conviction.
• He is the portfolio manager only of the hedge fund; the credit, CLO, structured credit, and high-yield businesses have their own PMs and investment committees he does not sit on.
• The Sony campaign saw Third Point own up to 7 percent and push to separate the conglomerate; management resisted for years before spinning out the semiconductor and financial services businesses.
• He learned that activism in Japan is hard, but the government often wants reform; he co-wrote a paper with Larry Lindsey and Niall Ferguson urging corporate governance and return on invested capital as a fourth arrow of Abenomics, picked up as a Wall Street Journal editorial.
• Investing in Danaher was his most instructive experience, teaching him how the Danaher Business System drives continuous improvement (Kaizen) and how the company celebrates rather than shames underperformance because problems are fixable.
• FTX was his hardest lesson; it looked great and was verifiable on the blockchain, but was not what it appeared, and now Third Point’s diligence includes checking bank balances.
• He notes that, fraud aside, Sam Bankman-Fried had a strong nose for value with stakes in Anthropic, Cursor, and Solana.
• Recent mistakes also include shorts where Third Point thought certain info-services businesses would resist AI disruption; he still expects a shakeout with some phoenixes rising from the ashes.
• He is pushing his whole team to use AI daily, hiring native computer scientists and system integrators, and describes Claude as a tool that makes you autonomous and gives back whatever you put into it.
• Third Point’s distinctive edge is optimism about AI creating net jobs and the ability to default into credit investing during stressed times, as it did with investment-grade credit in 2020.
• Credit is hard to copy because it runs on relationships, not electronic trading; that is why Third Point built into CLOs and eyes the roughly $6 trillion structured credit market rather than treating it as tourism.
• The great analyst has changed: 20 years ago it was someone who could model fast and crack a complex restructuring (Loeb made a career-defining bet on Drexel Burnham claims); today it is a Gavin Baker type who deeply understands an industry, like the analyst who flew to Texas and realized Casey’s General Stores was really a pizza chain.
• Outside the US, Loeb is more bullish on Korea, Taiwan, and Japan as hunting grounds, finds Europe tough on regulation (though he owns Rolls-Royce and ASML), and finds the Middle East the most vibrant region.
• What worries him most is not the business but running out of time for family, surfing, and reading; what excites him is incorporating everything relevant about the world and forming relationships with people building interesting things.
• His closing reflection is on kindness as a top-tier value, and the friend, Carter, who let him sleep on a couch and seeded his early fund, echoing a Palmer Luckey line that money cannot buy friends who believed in you when you had nothing.

Detailed Summary

Staying ahead of the firehose and reading the macro

Loeb opens by admitting he does not have a perfectly organized system for processing the modern flood of information. He checks the news for what is relevant to the economy and to Third Point’s positions, tries not to obsess over minute-to-minute moves, and leans more tactical than strategic. When people ask him about macro, he says the usual government-reported metrics (growth, unemployment, inflation, rates, currencies, gold, crypto) are trumped right now by two things: where oil goes, which depends on war and geopolitics, and what AI does on the spending and infrastructure side and its impact on society and the economy. To understand technology, he leans on Jensen Huang’s framing of the AI stack and talks to smart people regularly, and he watches three companies above all: Nvidia, Anthropic, and Elon Musk’s companies as a group.

From event-driven roots to quality investing

Third Point’s DNA comes from Loeb’s time as a credit investor at Jefferies, where he watched some of the best distressed, event-driven, and risk-arbitrage investors operate, from David Tepper to Eric Mindich to firms like Angelo Gordon and Farallon. His first lens was event-driven: spin-offs, demutualizations, privatizations, and post-reorg equities, where a newly created and illiquid security gets dumped by holders who will not do the work, and management sandbags guidance because incentive packages are set at the spin date. He barely thought about moats or returns on capital; he just wanted to buy something genuinely cheap with those characteristics. That was the firm’s bread and butter from 1995 until roughly 2013-2015. Those opportunities still exist, but Loeb describes deliberately evolving toward business quality and thematic investing, influenced by “The Outsiders” on capital allocation and Lawrence Cunningham’s “Quality Investing” on durable, high-return businesses. He organized the team around industry experts rather than generalists. The twist: AI disruption recently turned many apparently high-quality companies into much lower-quality ones, fast.

The AI cycle, bubbles, and the human edge

Loeb resists the bubble narrative. He was short the dot-com bubble and remembers the valuation excess; today’s AI leaders, by contrast, invest off their balance sheets and generate enormous cash, so unless you believe the capex yields no return, the earnings and multiples do not look like 1999. The real driver of volatility, he argues, is expectations: the semiconductor index ran up 40 percent on strong fundamentals, but Nvidia and Micron both delivered blowout quarters and still saw their stocks fall because expectations had run too high. That dynamic is exactly where a fundamental investor earns a living, because quants, CTAs, and risk-managed pods are structurally forced to sell into weakness. He also doubts AI will ever fully run a capital system, since private equity, restructurings, creditor committees, and high-touch credit always need humans. He cites “Reminiscences of a Stock Operator” and Ecclesiastes: there is nothing new under the sun, and human nature, with its bubbles, panics, and extremes, does not change.

Governance, his father, and the duty of boards

Loeb traces his governance interest to his father, a securities lawyer and corporate-governance expert who served on the boards of Mattel and Williams-Sonoma and championed ethical sourcing before it was common. He calls the American board system beautiful: directors are answerable to shareholders and accountable for strategy and key financial decisions. Governance breaks down when directors lose sight of their fiduciary duty, lack the knowledge or talent diversity to do the job, or prioritize things other than shareholders. He invokes Milton Friedman and Warren Buffett to argue that caring about communities, employees, and conduct is not inconsistent with shareholder value but part of it, and criticizes the Business Roundtable for muddying the board’s core duty. The most common failure he sees is directors letting loyalty to an underperforming CEO override their duty. Most of the time Third Point redirects existing boards without even taking a seat; the extreme proxy fights are the exception.

Activism, writing, Sotheby’s, and Sony

Great writing, Loeb says, is clear thinking and organizing your thoughts to get a desired outcome, and it is one of activism’s most effective levers alongside financial and legal pressure. Social pressure through writing and PR can move a board on its own. He sees a pattern in his campaigns: targets that hold themselves out as high status but are not living up to it. Sotheby’s is the clean example, a centuries-old, high-status business run unprofitably, where Third Point bought 9.9 percent, gave the existing CEO a year, then helped install Tad Smith from MSG, who modernized operations and technology before the company was sold. Sony was a two-act campaign in which Third Point owned up to 7 percent and pushed to break up the conglomerate; he recounts sharing the thesis with Andrew Ross Sorkin at the New York Times under embargo, the panic it caused, and how management resisted for years before spinning out the semiconductor and financial services units. The lesson: activism in Japan is genuinely hard, even though the government wanted reform. He co-authored a paper with Larry Lindsey and Niall Ferguson arguing corporate governance and return on invested capital should be a fourth arrow of Abenomics, which ran as a Wall Street Journal editorial.

The Danaher operating system

Loeb calls Danaher his most instructive investment. He and his partner persuaded the company to compress its five-day Danaher Business System training into a single day, and he came away with a deep appreciation for how a real operating system drives continuous improvement. The standout lesson was cultural: Danaher holds people individually accountable, but when it finds someone underperforming it celebrates rather than shames, because the problems are addressable and fixable, and it does this relentlessly across operations and working capital. He also points to the diaspora of Danaher executives, including Larry Culp and the leadership at Ingersoll Rand, as evidence of the system’s depth. The investment worked for about four years before COVID-era order surges and inventory swings turned tailwinds into headwinds; Third Point sold and has recently bought back in modestly.

The structure of Third Point and the fulcrum security

Third Point is not one fund but a collection of businesses. The flagship hedge fund grew from $3 million to about $9 billion and is roughly 30 percent credit, generically around 110 percent long and 30-40 percent short on the equity side. Across the firm the credit weight is closer to 60 percent, spanning a roughly $7 billion CLO business, several billion in structured and corporate credit, an insurance company, a couple billion in asbestos liabilities, a small new private credit unit, and a venture arm. The unifying thread is valuing enterprises at any stage and investing in whichever fulcrum security (the one with the best risk-reward) makes sense. Loeb illustrates with Credit Suisse’s takeover by UBS, where the holdco paper proved the fulcrum, and with buying Twitter’s resold financing debt near 96-97 cents at a 12 percent yield when other credit investors were scared, plus a difficult xAI debt financing that few credit people wanted. He pushes back on the idea that he sits atop everything: he is the PM only of the hedge fund, while the other businesses have their own PMs and committees he is not on.

Insurance, the FTX lesson, and recent mistakes

Loeb started a Bermuda reinsurance company in 2010, backed by himself, Kelso, and Pinebrook, on a barbell thesis of investing the float in Third Point and treasuries to defer taxes and lever capital. The reinsurance side soured, and about three years ago he concluded they had the right idea but the wrong vehicle, that plain-vanilla annuities (which can only invest in credit) would have fit better. Third Point merged the reinsurer into its UK closed-end fund, Third Point Offshore Investors, reincorporated from Guernsey to Cayman, and repurposed it into an insurance company managing private credit, structured credit, whole-loan mortgages, real estate lending, and investment-grade debt. His hardest lesson was FTX: it looked great, was verifiable on the blockchain, and had a strong cap table, but was not what it seemed; diligence now includes checking bank balances. He notes Sam Bankman-Fried, fraud aside, had a great nose for value (Anthropic, Cursor, Solana). Other recent mistakes were shorts where Third Point bet certain info-services businesses would resist AI disruption; he still expects a shakeout with some survivors rising from the ashes.

AI inside the firm, the analyst of the future, and kindness

Loeb is pushing his entire team to use AI daily, hiring native computer scientists and system integrators, and describes Claude as a tool that makes you an autonomous self-improver and gives back whatever you put into it, with some analysts running agents overnight while he uses it more for queries. He pairs this with Brad Gerstner’s recommendation of “Essentialism”: you cannot do it all, so you must decide what is most relevant. The great analyst has changed: 20 years ago it was someone who could model fast and crack a complex restructuring, as Loeb did with the Drexel Burnham bankruptcy claims early in his career; today it is a Gavin Baker type who deeply understands an industry and its technology, like the analyst who flew to Texas and realized Casey’s General Stores was really a pizza chain in disguise. On the rest of the world, he is more bullish on Korea, Taiwan, and Japan, finds Europe tough on regulation (while owning Rolls-Royce and ASML), and finds the Middle East the most vibrant region. He closes on what worries and excites him (time with family, surfing, and reading versus the joy of incorporating everything relevant about the world), and on kindness, crediting his friend Carter, who let him sleep on a couch and seeded his early fund, and echoing Palmer Luckey’s line that money cannot buy friends who believed in you when you had nothing.

Notable Quotes

“I think you have to be a tech person today. It’s a big and growing and compounding part of the economy. It affects everything else.”

Dan Loeb, on why no serious investor can punt on technology anymore

“Hold on to your seats because things are only going to accelerate from here.”

Dan Loeb, recounting a 2013 Davos warning about technological change he now applies to AI

“Maybe that’s where the human element comes in, to understand and to be able to make those tough trading decisions when fundamentals are going one way and stock prices are going the other way, and to be able to take the pain of losses in the short run.”

Dan Loeb, on where a human investor still has an edge over machines

“It’s very different from the dot-com bubble, which we were short going into. You don’t have the valuation bubble now on those companies that you had back in those days.”

Dan Loeb, on why he does not see the AI rally as a 1999-style bubble

“When they found someone that was underperforming, it was celebrated instead of shamed, because look at all these things you’re doing wrong, we can fix those. And they did.”

Dan Loeb, on the accountability culture he learned from the Danaher Business System

“I would have to say our investment in FTX. It looked great. The company was growing fast. We could verify it all on the blockchain.”

Dan Loeb, naming his hardest investment lesson

“Be kind to people you have no idea how it will ever benefit you. And sometimes it will and sometimes it won’t.”

Dan Loeb, on elevating kindness in your hierarchy of values

“The one thing money doesn’t buy you is friends that believed in you when you had nothing.”

Dan Loeb, quoting Gavin Baker quoting Palmer Luckey, on the friend who seeded his early fund

Watch the full conversation between Dan Loeb and Patrick O’Shaughnessy here.

Related Reading

• Third Point LLC — the official site of Dan Loeb’s investment firm, the primary source for the strategies discussed.
• You Can Be a Stock Market Genius (Wikipedia) — Joel Greenblatt’s classic on special situations that became Loeb’s foundational framework.
• The Outsiders by William Thorndike (Wikipedia) — the capital-allocation study that helped shape Loeb’s pivot to quality investing.
• Essentialism by Greg McKeown (Wikipedia) — the book Loeb cites for deciding what is actually relevant in an age of infinite information.
• Invest Like the Best by Patrick O’Shaughnessy — the podcast where this conversation took place, with a full archive of investor interviews.