Robert Hagstrom on Pragmatic Truth, Multi-Disciplinary Investing, and the Concentrated Portfolio

Robert Hagstrom

I literally flailed about for years trying to figure out what my score in life was. My mother was a doctor — one of the first women out of Vanderbilt Medical School — and my father was a chemical engineer. They both seemed to have known what they wanted to do from day one; I couldn't figure it out. What I can say is that two people shaped me: Bill Miller and Charlie Munger. Those two influences converged almost coincidentally as I was writing The Warren Buffett Portfolio.

I met Bill as a stockbroker when I joined Legg Mason in 1984. He was director of research, and he is a true polymath. Our friendship was largely built around this whole curiosity about multiple disciplines. The blessing I had was that Charlie gave me the theory while Bill gave me the practicals. People have said: "Robert, you wrote your dissertation on Buffett. You did your practicals with Bill Miller." I cannot emphasise how important it was to do the practicals with Bill, because I could see how he used philosophy and biology and literature as we were actually making an investment. I could see the tangible payoff accruing in real time. When you get that first aha moment, you want more. And it just led me to keep reading, keep drilling into different disciplines.

Robert Hagstrom

It's by design. Being a concentrated, low-turnover portfolio manager opens up a tremendous amount of time. We own 20 stocks, plus or minus, with average holding periods of five to seven years. Seven of those positions I've held for over eleven years. I don't have to spend most of my waking hours worrying about markets and prices. We're market-aware and economically aware, but we're genuinely agnostic about where the market or the economy is headed. We feel like we own a collection of businesses and we just check in with those businesses to see how they're prospering.

The first thing I did, years and years ago, was turn off all financial news networks. I don't have a TV in my office. I watch Bloomberg for maybe 30 minutes first thing in the morning and then turn it off. Warren's secretary Debbie once told me that Warren sometimes has CNBC on, but never with the sound — he uses it as a ticker tape. I thought that was perfect. Beyond the portfolio architecture, it's also just natural curiosity: once you find a nugget and have an epiphany, you want another one. Bill had me reading the New York Review of Books, the London Literary Supplement, the London Review of Books — always scouting for ideas that might have relevance. It became a treasure hunt. Every day I think: what can I find today?

Robert Hagstrom

Mortimer Adler wrote a book called How to Read a Book, which I thought was presumptuous — until I realised I actually did not know how to read a book. The average book takes about 8 hours to read. But the key question is: does the book deserve my 8 hours? Adler sets up a multi-step process. First, what he calls inspectional reading: read the preface, check the bibliography for unfamiliar titles, skim the first and last chapter. You're trying to find out quickly whether the book is worth your time. Of 10 books you think you want to read, probably only two or three actually clear that bar.

For years I refused to write in my books — they were museum pieces. Adler changed that: make the book your own. Underline, highlight, question. Then, if it earns 8 hours, comes what he calls syntopical reading: find all comparable books on the same topic written by other people and bring in multiple viewpoints simultaneously. That cross-pollination is where the real insight lives.

I never get a shiver moment from TV or even the Wall Street Journal. I get those epiphanic moments from books. Bill put it perfectly: all reading is experiential. What moves you in a book becomes an internal experience, and that experience stays with you in a way that information from news never does.

Robert Hagstrom

A hundred per cent. My mother used to say: write it down, you won't forget it. And when you write a book, you own it. Whether the writing is informal or formal, three pages or a whole chapter, writing really does press the idea into me. It becomes part of me. It's not a flighty thought that comes and goes — it's deep inside you. Things I wrote in The Warren Buffett Portfolio twenty-five years ago I can still cite verbatim because the writing reinforced the learning.

And your second point — teaching amplifies it further. Once you share ideas with audiences, on a podcast, in a piece of writing, you verbalise them. You own them at a deeper level still. I tell young people looking for jobs: take one of your best research papers, clean it up with your English teacher's help, and when you drop your résumé on the interviewer's desk, drop that paper too. You'll separate yourself from the other nineteen candidates. More importantly, you'll demonstrate that you can think, reach a conclusion, and communicate in writing — which is exactly what employers fear AI and texting have destroyed.

Robert Hagstrom

Most people have opinions and no facts. If I'm going to say something or write something, you can be sure I'm going to footnote it. Nothing is more dreadful for a writer than to be called out on a claim you said was true and it wasn't. I learned that early when I wrote a profile of Sir John Templeton and got a small date wrong by a year. Someone wrote to correct me. I thought: if Templeton himself couldn't remember the exact year of something from his own life, I need to be obsessive about double-checking.

Bill was the same way. He'd send back my drafts: "Check that. That's misspelled. The university name has two Gs, not three." It's a constant paranoia — and that paranoia is what separates a well-argued position from a confident-sounding one that collapses under scrutiny. As for AI: it's better at surfacing information than it was. But it's still, to my mind, Google on steroids — a better retrieval engine. The genuine insight that changes how you think still comes from the kind of engaged reading we've been describing.

Robert Hagstrom

Bill identified two theories of truth. The correspondence theory holds that you have figured out how the world works and your thinking is anchored to that structure. For classic value investors, this crystallised into a dogma: value is low P/E, low price-to-book. That is how the world works. You are intellectually and emotionally attached to that as an absolute. The pragmatic theory — William James — holds that in a biological system, things are changing and evolving. Ideas are tools; the test of an idea is whether it works, whether it produces accurate predictions, whether it has cash value. The pragmatist investor asks: where is value actually working right now? What's the observation?

Bill used to say: value is always in the marketplace; it just migrates. It goes to different places. If you're pragmatic, you're not afraid to buy value in technology. You're not afraid to buy value in a financial. You just go wherever the opportunity set is. But the correspondence theorist is imprisoned in a sense of the absolute — not flexible in thinking. There are so many classic value investors who lost a decade or more of performance numbers because of stubbornness. Bill could look at those portfolios and say: I just made 20 times on Dell. That was a huge value investment. How come they didn't own Dell?

Robert Hagstrom

Bill would show us those attacks and say: do you see the biases these people have? They can't see it clearly. Their frame — value versus growth, P/E as the measure — makes them structurally unable to see the value that's sitting right in front of them. That's William James's blindness made operational. Howard Marks is the first to admit now that Bill was right. But the blindness at the time was genuine — it wasn't stupidity. These were brilliant people. The frame they were trapped in simply couldn't accommodate what Amazon was.

Robert Hagstrom

Let me back up to Dell first, because that's the foundation. Bill bought Dell Computer around 1993–94. What made it extraordinary wasn't buying it cheaply — it was holding it when it became the first company in history to achieve 100% return on invested capital. The model was negative working capital: you'd order a computer, call Dell, specify what you wanted, give your American Express number. That money hit Dell's account that night. They didn't pay suppliers for 30, 60, or 90 days. They grew the entire business on their customers' receivables. Dell went up 8,000% in the 1990s; Bill captured roughly 5,000% of that in the Value Trust.

Now we're in Las Vegas at a broker conference, and Bill meets with Jeff Bezos just before Amazon's IPO. Jeff described the business model. Bill asked: what are you doing and how does it work? And Jeff said: "It's Dell." With books, you can hold them for six months without paying for them, and sometimes take them back for free. Zero capital in the business. Bill went: I'm interested.

Then comes Wittgenstein. Bill did his PhD in philosophy — absent the dissertation — and one of his favourite philosophers is Ludwig Wittgenstein, the great Austrian philosopher of language. The insight: how you form a description ultimately determines your explanation. Language shapes descriptions; descriptions shape explanations; explanations drive action. Bill walked us through this: what is Wall Street's description of Amazon today? They said Barnes & Noble, because both sell books and Barnes & Noble is cheaper and has stores. Description: Barnes & Noble. Explanation: sell Amazon, buy Barnes & Noble. Or they said Walmart, because Amazon was adding non-book products. Description: Walmart. Explanation: avoid. Both descriptions were wrong, therefore both explanations were wrong. Once the description became Dell — negative working capital, capital-light model, unlimited scalability — the thesis was automatic. The investment followed directly. The description did all the work.

And I saw how many different ways you could describe the same thing, and how critical it was to identify which description was right. Because whatever the right description is, that tells you what's going to happen to the stock. That is not taught in CFA preparation. But Bill Miller taught me more about investing outside accounting, finance, and economics than I could ever have imagined possible.

Robert Hagstrom

He was doing the convertible bond analysis on Amazon and linear-extrapolating future distribution centre builds. Amazon had built six or seven large distribution centres and this analyst assumed they'd do that every year, indefinitely, until they went bankrupt. He never called management. Had he done so, Jeff would have said: we're done building centres for now. That's all we need. The analyst's thesis was superficially rigorous but built on a description that was catastrophically wrong. There were two ways to avoid the error: call management and ask how many more centres they need; or follow the cash all the way down the income statement, where there was plenty of it, even in loss-reporting quarters. A second-year finance student could have caught it, with a nose for curiosity.

Bill's habit was always: go find what's working, then figure out why it's working. William James's "cash value of ideas." Amazon revenues were growing, the business was getting bigger, it was working. So: let's figure out how it's working, and is it sustainable? That's it. No P/E ratio required.

Robert Hagstrom

In 1992, during the savings and loan crisis, Legg Mason made a very successful bet that the government would let financial firm equity survive. Banks in Colorado and Texas recovered, the S&P bounced, and we did well. In 2008, the instinct was to apply the same description: this is 1992; the government will let equity survive. That was the wrong description. The crucial difference in 2008 was the political environment. A government committee in 2008 was not going to be seen "lending to fat cats." The politics meant equity would be wiped out. The model from 1992 — government intervention preserves equity — was correct in 1992 and incorrect in 2008, not because the investing framework was wrong but because the description was wrong. We had not accounted for the change in the political variable.

The lesson I took: any time government is involved in a binary survival-or-destruction decision for a company, that outcome belongs in the "too hard" pile. Political committee dynamics are not reliably modellable.

Robert Hagstrom

I'd want to think carefully before equating Amazon and Bitcoin as investment theses, because the structure was different. As I understood it from Bill, his argument for Bitcoin was: I have no idea if this will work, but if it does, it could be very large. The maximum downside was 1% of net worth — not life-changing. The upside if correct was enormous. Great asymmetric odds. The hardest part, as with Amazon, wasn't the original bet. It was holding it at a hundred million dollars and watching it go to a billion.

I came at it differently. I don't own Bitcoin. But I thought the argument that it can't be valued because it doesn't generate cash was a poor argument — a Van Gogh painting doesn't generate cash, but supply-and-demand dynamics clearly drive value. What I found more compelling was the stored-value function. When we bought LVMH eleven years ago, one of the executives told us: in countries with fragile banking systems or hostile governments, people buy luxury goods as stored value — Louis Vuitton bags they can monetise when they flee. That clicked something for me. A woman at one of my client events came up afterwards and said her family in Argentina had been wiped out twice by the government — bank accounts to zero. They hold a lot of Bitcoin now. That made perfect sense to me: Bitcoin as portable, censorship-resistant stored value for people in genuinely unstable places.

Robert Hagstrom

Hendrik Bessembinder at Arizona State did a study covering the period from the Great Depression through to around 2020. He found that two-thirds of all listed stocks never beat Treasury bills over their lifetime. All net equity returns above T-bills accrue to roughly 4% of publicly traded companies. He updated his work for 1990 to 2020 and identified the 50 biggest contributors to total market capitalisation over that period — 35 US stocks, 15 international.

I took those 35 US stocks and looked at how an equally weighted portfolio of them performed over the most recent 10 years. The result: the market was up roughly 200%; this portfolio was up roughly 600%, excluding Nvidia, which was up 28,000%. So clearly home-run territory. But here's the kryptonite. On a monthly basis, these 17 outperforming stocks underperformed the market about 50% of the time. Quarterly, 60%. Annual, 63%. Over the period there were 102 separate instances of 20% or greater drawdowns. Average peak-to-trough drawdown: 41%. The question is: who holds that portfolio?

Robert Hagstrom

Cremers and Petajisto studied thousands of mutual funds and found that high active share combined with low turnover produces persistent excess returns. That's the mathematical justification for Buffett's portfolio style, validated independently using fund data. The corollary is that high active share alone isn't sufficient — low turnover is the other leg. The return is generated by holding, not by trading. And that corollary is precisely what prospect theory makes so nearly impossible for most investors.

Kahneman and Tversky established that people weight a unit of loss approximately twice as painfully as they weight an equivalent gain. The rational investor knows a 40% drawdown in a concentrated portfolio is a buying opportunity. The neurological reality is that almost no investor can experience it that way. This is not a failure of information or education. It is wired in. And the consequence is that the edge in concentrated investing is durable precisely because knowing about it does not fix it.

Robert Hagstrom

Rather than taking a maximum allocation from a client, I ask for 10% of what they intend to invest, for a mandatory three-year hold. Take less capital rather than more. JP Morgan had a saying: sell down until you can sleep. The direction is always towards the concentration level the client can actually hold through a 40% drawdown — because that drawdown will come. If you start a client at 100% allocation and they experience a 40% drop, they're gone. But if you start them at 10%, they experience it, they survive it, and three years later they're ready for more. You build conviction through experience, not through a prospectus.

Nick Sleep and Qais Zakaria are a beautiful example on the other side of that coin. The last time I checked, they still had nearly their entire net worth in three stocks: Amazon, Costco, and Berkshire. Their temperamental ability to hold through everything thrown at them over twenty-odd years is something genuinely rare.

Robert Hagstrom

If you go to any current AI and ask: what is the competitive advantage period for Nvidia? It can tell you what happened today and last week. It can monitor sales, margins, and return on capital. But it has no idea how to think about Nvidia within a financial ecology of competitors, how it intersects with AI infrastructure development, how long the GPU architecture advantage persists relative to potential alternatives. It can't model the competitive ecosystem over time. I spend 80 to 90% of my time on exactly that question — how long can this last? And there is currently nothing in AI that can answer it. That may be the remaining runway for human investors.

As for our own Nvidia position: we bought it in the autumn of 2022, knowing AI was arriving and that the GPU was the correct architecture from our reading of the Santa Fe Institute work. We didn't know ChatGPT would arrive in November of that year. When it did, Nvidia was 2–3% of the portfolio. It grew to 10–12% by market appreciation, at which point pension plan constraints required us to trim. But it earned its way there. It wasn't a speculation on an unproven idea — the earnings per share were growing faster than the share price through 2023–24. The multiple was actually coming down. Same with Microsoft, Meta, Alphabet, and ASML. These businesses have earned their valuations; we have simply never seen trillion-dollar businesses become four-trillion-dollar businesses before, and that scale disrupts people's intuitions.

Robert Hagstrom

My best clients have always been business owners, because we speak the same language. When I talk to a business owner, they understand cash, cost of capital, return on investment without any translation required. At my client seminars, I ask: how many of you are business owners? Maybe a dozen hands go up. How many of you own common stocks? Eighty-eight more hands go up. I then ask those 88 people: you own common stocks — what do you own? And they have no answer. They cannot connect "I own a share of Meta" to "I am a partial owner of a publishing business." They don't feel it.

Buffett could make that connection so instinctively because he owned private businesses and public companies simultaneously. He developed the habit of seeing a stock as a business by actually owning businesses, not just equities. The reason modern portfolio theory won in the 1970s wasn't because it was right — Graham was right, margin of safety was right — but because after the 1973–74 blow-up, there were no prominent practitioners to defend the business-ownership approach. The academics were waiting in the hallway with their broadly diversified, low-volatility portfolios, and nobody else had a ready answer. The result was an entire generation of investors who learned to manage a portfolio of non-correlated price series rather than a portfolio of businesses.

Robert Hagstrom

The most uplifting thing I take from Warren is simply this: he's approaching 95 and he never stopped. He kept playing the game he loved for thirty years after I wrote the book when he was 64. I'm just a little older than 64 now. That tells me: Robert, you're a puppy. There's no reason you can't stay in some form of this game for another twenty or thirty years if the health cooperates. I love solving puzzles. I love thinking about odds. I love the competition. I can't imagine anything more enjoyable to do over the next few decades. I'm a terrible golfer and I don't gamble. This is what I love to do, and by some extraordinary luck, it also worked out as a career.

And writing and investing reinforce each other. I'm a better businessman because I'm an investor. I'm a better writer because I'm an investor. I'm a better investor because I'm a writer. Both crafts sharpen the same underlying skill: reaching a well-founded conclusion and being able to defend it. When I look at Warren — and Charlie, who stayed active well into his nineties — I think: yes. I'll stick around.

Robert Hagstrom

That's the right way to think about it. I will never be as smart as Bill, Warren, or Charlie, and I'll never have whatever X factor gets them to their level. But by studying them, have I become better than I otherwise would have been? Have I become better than the index I'm competing against? The answer, over the last 11 years with Global Leaders, is yes — we've outperformed. That's not their level of excess return over 50 years. But the ride has been worthwhile. Not a replica of genius, but a genuine benefit from sustained study of it. And Charlie gave that formulation its clearest expression: you don't have to become an authority on every mental model. You just have to grasp the really big ideas, the stuff that does the moving and shaking. Once you have the big idea, you can move on. You don't have to write your PhD dissertation on Wittgenstein's Philosophical Investigations. But you do have to get a working knowledge of the philosophy of language — enough that when Bill walks you through the description problem on Amazon, you recognise immediately why it matters.

Robert Hagstrom

William, right back at you. I've learned so much from your writings and from this podcast. We need to keep the momentum going — there aren't many of us doing this kind of work and it matters more than ever. I look forward to continuing the conversation. Thanks so much for having me.