Stan Druckenmiller on AI, Markets and the Next Ten Years | TomorrowX Summit.
Stan DruckenmillerIliana Bouzali
- Druckenmiller has shifted from January’s bullish portfolio to capital preservation as the macro signals reverse. Then, Fed funds of 3.5% against 6% nominal growth, with rate cuts expected, signaled stimulus; now the Fed is hiking, credit spreads are widening and he is short copper. His equity book is roughly flat, with small dollar positions: “I’m not playing.”
- If equities have another big advance, Druckenmiller thinks AI use cases could lead it. Infrastructure “may be a great tactical trade” after positioning washed out, but he doubts it will provide the same boost it did over the last four years. His personal belief is that labs will stop releasing models, sell tasks instead of tokens and create businesses such as cybersecurity: “Who’s going to compete with these guys in cybersecurity?”
- His risk management follows actual P&L because correlations can fail precisely when investors need them. Unexpected behavior prompts cuts; he uses neither VaR nor stress tests. Rejecting Bouzali’s description of a cold streak, he says: “I’m not cold. I’m on an out streak.” Returns have come in spurts, with 5–8% years punctuated by three- or four-year runs; this pause could last two days or six months. His record of no down calendar years still included big drawdowns.
- His quieter structural worries are a weakening US talent pipeline and a financial-market crisis that forces fiscal reform. Imported talent has compensated for deteriorating K–12 education, but immigration policy threatens that workaround; he is optimistic policy will change. Reserve-currency status and zero rates delayed the deficit reckoning, yet he thinks a market crisis will produce reform sometime in the next ten years: “Slowly, and then suddenly, all at once.”
- His decision process starts with a meaningful bet on roughly 15% of the information a traditional investor would want. He then spends two or three weeks researching whether to add or exit—“investing, then investigating”—while emphasizing that concentration requires flexibility. Buying back the Nasdaq on the day of its 2000 high illustrates how discipline fails: he knew better, but “I just couldn’t help myself.”
- He puts 20% odds on having fusion or knowing it is coming within ten years, and at least 70% on data centers in space. He also believes cancer will become “a managed disease” in “10 years, not 20 years,” citing genetic sequencing, immunotherapy and AI. The AI bear thesis he respects most is politics and financing: “both concerning.”
1. January’s bullish setup has reversed
- Bouzali recalls long equities, short Treasuries, long copper and a healthcare thesis. Druckenmiller’s answer: “Basically, almost none.” Fed funds were 3.5% against 6% nominal growth, with further rate cuts expected. Strong breadth, tight spreads, rising commodities and enthusiasm among AI insiders around Claude Code made recession within 6–12 months seem almost impossible.
- A war triggered panic selling. When they quickly bought back, AI infrastructure’s potential liftoff looked clearer, and the portfolio had been cleared of holdings he thought would not rise much. Now the Fed is hiking, copper is a short and spreads are widening. Breadth peaked August 14 despite October index highs; defensive leadership has him alert to possible economic disappointment in the second half of 2027.
- He likes to go short, but would not dare short innovation given what is happening in the sector; much of it is also oversold and washed out. His equity positions are roughly flat across longs and shorts, and he is preserving capital: “I can’t remember a period when I had less clarity” about the next 12 months. He says the portfolio shift followed a reversal in the conditions that had supported his January view, including monetary policy and commodities.
2. AI labs could become application businesses
- Infrastructure “may be a great tactical trade” after positioning washed out, but he doubts it will provide the same boost it did over the last four years. He is looking for use cases.
- Druckenmiller’s personal belief is that labs will stop releasing models and move from tokens to tasks, creating businesses themselves. Cybersecurity is his example: “They created the problems,” and could deploy swarms of defensive agents. Healthcare offers further use cases.
- Asked for the AI bear thesis he respects most, he answers: “Politics, the financing—both concerning.”
3. Education, immigration and deficits worry him
- Deteriorating K–12 education makes imported talent more important. At a tech conference, he guessed that eight of the top ten tech companies are run by immigrants; he worries that immigration policy threatens this flow. He remains optimistic policy will change.
- In 2011–12, he warned college students about entitlements and demographics. He says he did not anticipate $5 trillion in COVID spending, followed by further spending supported by zero rates. Reserve-currency status has delayed the reckoning; he thinks a financial-market crisis will force fiscal reform sometime within ten years, while remaining optimistic that reform will happen.
4. Concentration requires flexibility and discipline
- “The P&L is my true signal”: unexpected behavior prompts cuts because correlations can fail in chaotic markets. He uses neither VaR nor stress tests; he says correlations failed for some investors in software and AI infrastructure earlier in the year. Rejecting Bouzali’s “cold streak” description, he says, “I’m not cold. I’m on an out streak.” Duquesne has made money in spurts: 5–8% years followed by three- or four-year runs, with sideways periods in between. The break could last two days or six months.
- Trading commodities, bonds, stocks and interest rates means he does not need to be heavily invested in any one category. He has sometimes withdrawn from equities to play other macro instruments; right now, with little clarity, the aim is to avoid losing money. He says his record of no down years was “the luck of the calendar”: he had big drawdowns, but was never down from December 31 to December 31.
- Concentration demands an open mind: “Put all your eggs in one basket and then watch the basket very carefully.” He prizes passion, curiosity, extreme competitiveness and discipline; high IQ can breed attachment to a thesis.
- His “investing, then investigating” approach starts with a bet large enough to hurt if wrong, using roughly 15% of the information a traditional investor would want. He then researches for two or three weeks before adding or exiting.
- After selling the Nasdaq in January 2000, FOMO drove him back on the day of the high; he thinks he missed the top by an hour. Asked what he learned: “Nothing. I knew it. I just couldn’t help myself.”
5. Ten-year optimism spans fusion, space and cancer
- Druckenmiller assigns 20% odds to having fusion—or clarity that it is coming—and “at least a 70%” chance of data centers in space.
- After 30 years on Sloan Kettering’s board, he cites genetic sequencing, mutations, immunotherapy and AI in predicting cancer becomes “a managed disease” in “10 years, not 20 years.”
Full transcript
I can’t remember a period when I had less clarity on where we’re going to be in 12 months.
Who’s going to compete with these guys in cybersecurity?
1. Looking back at January's portfolio
You have the biggest thing going on in the history of American business.
I think we’re going to beat cancer. I think it’s going to be a managed disease, and I mean in 10 years, not 20 years.
It’s an auto-reply: “Don’t talk to me. I’m not playing.”
Hi, Stan.
Hello, Iliana.
Gavin, thanks for giving us an excuse to do this again. We did it about nine months ago.
Before we start, I know we’re time-pressed, but I have to tell a little story about Iliana and Gavin. Iliana runs the global derivatives business at Morgan Stanley. Back then, we were called “macro legends”—more appropriately, in my opinion, “macro has-beens.” We had this Zoom call about a month after ChatGPT, and all the macro legends were opining on interest rates and other matters. This guy I’d never seen before, Gavin, comes on the Zoom and says, “You guys all have your head up your you-know-what, and you have the biggest thing going on in the history of American business, and you’re all here talking about nickels and dimes.”
Gavin, I want to thank you, because that was right when it all started. You predicted the whole thing, and you predicted its severity. Iliana, I want to thank you, because I would not have met Gavin if it weren’t for you. So thank you both.
You’re welcome. I went back to our January conversation, similar to this one, with a smaller audience. You were bullish on the economy, long equities, short Treasuries as a hedge, and long copper. You also believed healthcare was going to be a breakout winner. How much of that portfolio do you own today?
Basically, almost none. When you interviewed me in January, it was a unique time. Sometimes in my career I’ve had clarity, and sometimes I have no clarity.
When I have clarity, I like to play big. When I don’t have clarity, I’ll sometimes almost not play. At the time, if you’ll recall, the economy was very strong. More importantly, looking forward, Fed funds were at 3.5%, and nominal growth was 6%. Monetary policy was ridiculously loose. If you looked at the so-called Fed dot plot, more interest rate cuts were supposedly coming later in the year. Commodities, including copper, were breaking out to new highs. The advance-decline line and other market technicals I look at were all making new highs. Credit spreads were very, very tight. So it seemed almost impossible to me that we could have a recession within 6 to 12 months, with monetary policy overly stimulative.
Another thing was happening: Claude Code, which had come out the previous fall, pretty late, had just gotten everyone—or let’s say, the insiders—excited about an acceleration in AI. So we were quite bullish on capital spending. This all made us quite bullish on equities and, as you said, the economy.
2. Why everything has flipped
Unfortunately, about a month later, something called a war started. I panicked. We sold a bunch of stuff, but it actually ended up being lucky because when we quickly bought it back, it was more apparent than ever that AI infrastructure was going to have a complete liftoff. It cleaned out all the stuff in the portfolio that wasn’t going to rise big.
Now let’s look at today. The economy is still strong, but everything else I just talked about has completely flipped on its head. The Fed is not cutting; they’re hiking. We’ve had regime change at the Fed. There’s a gentleman I know who sat next to me for 15 years. He is not happy that we had free money for 15 years while the deficit and everything else went through the roof. He’s even less happy that inflation has been above target for five years.
So we have monetary tightening going on. Commodities, which were breaking out in January, look like they’ve topped and are going down. In fact, we’re now short copper. The advance-decline line, which was making a new high when we were interviewed, made its high back on August 14, despite the fact that the averages made their high in October. Credit spreads are widening. I think you’re getting the joke here.
3. I'm not playing
Everything is the exact opposite of the way it was then. We are not long equities. We’re not short equities. I like to go short, I admit it. But for all the reasons I think most people came to this conference, I wouldn’t dare go short the innovation and everything going on here. Also, a lot of that stuff is oversold and washed out. I think positions are down tactically.
My choice here is to do what I’ve sort of always done when I don’t have clarity. I’ve been doing this for 50 years. I can’t remember a period when I had less clarity on where we’re going to be in 12 months, unlike last January. You made fun of me on one of these Zoom calls the other day and said, “Stan has a button now. When I call, it just says, ‘I’m not playing.’”
It’s an auto-reply: “Don’t talk to me. I’m not playing.”
That’s where we are. We still have some small positions in the dollar. Our equity positions are basically flat, but there’s a mismatch. There are some shorts and longs in there, but we’re basically staying away.
They mentioned some kind words about my record in the introduction. It’s a record that has been—
You mean you’ve never lost money? You’ve never had a down year?
That’s correct, but it was the luck of the calendar. I had big drawdowns; I was just never down from December 31 to December 31.
My record is really partly due to a couple of things. Number one, because we trade four asset classes—commodities, bonds, stocks, and interest rates—we don’t feel the need to ever be heavily in one or the other.
You invented that style, to be clear. You’re not a mandate investor.
I probably did invent it, but I invented it out of necessity because I didn’t know what I was doing back then. Anyway, by pure coincidence, bonds and currencies tend to make their biggest moves when equities are in bear markets. So it made me more willing to put equities aside and not have to be invested in that asset category. You had other alternatives. It kind of made its own discipline.
We’ve had periods where we’ve withdrawn from the market, and then periods where we’re playing in other macro instruments besides equities. Right now, because I don’t have a lot of clarity on anything, we’re pretty much in one of those periods where I’d say we’re in capital preservation. We’re trying not to lose money, as opposed to make money.
We were backstage, and we were having a laugh because you were telling me how someone has analyzed your style over 50 years. Can you share with the audience what we’ve learned from that?
You must not be listening, because I think it’s what I just said. Duquesne has made its money in spurts over the years. We go marginally up, like 5% to 8% years, and then all of a sudden there’ll be a three- or four-year run that goes up. Then there’s another sideways period. We’ve avoided the downs, and I feel we’ve made money mainly because I was lucky enough to have a lot of young people at my firm. The Gavin call didn’t hurt, either.
4. Hot streaks, cold streaks and risk management
We just had a great three- or four-year run. A lot of the stuff I just mentioned underlines that there’s a lot of uncertainty around, so I’m just taking a break, which could last two days or six months. I don’t know.
So you’re in one of these cold streaks. How do you know when you’re back?
I’m not cold. I’m on an out streak. I think you’re mentioning the fact that, if you look at my risk management, it’s a little unusual. We don’t use VaR. We don’t use stress tests. We don’t use any of that stuff. I remember when I was running the Quantum Fund for Soros, banks wouldn’t lend us money because we didn’t do VaR. So I gave them their stupid VaR.
What I’ve always done, because we’re a matrix investor across the four categories I just mentioned, is look at my P&L. If it’s not behaving the way it should, if it’s not correlating properly, I’ll start cutting back. The P&L is my true signal. Why do I think that’s better than VaR? Because when you get into chaos, when you get into bear periods, none of these correlations work. I think some very talented people found that out the hard way with software and AI infrastructure earlier in the year.
Shorting software, and then—
5. What he's watching: AI use cases
Crazy stuff just starts happening, and your P&L will pick it up. We do that. You mentioned one of the most important parts of my job: I believe in streaks. People in Texas know sports. In baseball, way beyond random, you get streaks: hitting streaks, winning streaks. It’s the same in my business. Sometimes I’m seeing the ball; sometimes I’m not.
A big part of my job is to know whether I’m cold or hot. Believe me, I’ve had some cold periods. I play big when I’m hot, and when I’m cold, I try to make little bunts. If my average starts going up, I figure I might be seeing it better, and I’ll play harder.
I had a multiple-choice question choosing between themes or stocks that you want to buy here, but it doesn’t seem like you really want to buy anything. Let’s scratch that. I’m going to ask you a similar question. You’re not ready to buy things in size, but what are you watching that you’re intrigued by that you might—
The equity market—
In the equity market and broadly?
Well, given the way commodities are acting, and the fact that I didn’t mention that a year ago cyclical stocks were outperforming defensives.
Yeah. And now all the defensives are leading. I’m really alert to the fact that the second half of 2027 could be economically disappointing, so I’m thinking about that in terms of interest rates.
In equities, we’re looking for AI use cases, as opposed to infrastructure. Infrastructure may be a great tactical trade because those positions have really been washed out, but I don’t think it’s going to provide the same boost it did for us over the last 4 years. I think use cases could be interesting. Ironically, maybe the most interesting use case is an AI lab, and that’s because I don’t personally believe they’re going to be in the token business much longer.
6. What's happening quietly
My own belief is that they’re not going to be releasing models. If you have the number one intelligence in the world, you can start selling tasks instead of tokens, and you can create your own businesses. It doesn’t take a genius. Cybersecurity: who’s going to compete with these guys in cybersecurity? They created the problems, so now they’re the ones who could have swarms of agents defending us from them. Healthcare—there are all kinds of use cases. So, if there’s going to be another big push in the equity market, I think that’s where we’ll want to go.
Selling tasks, no tokens. You heard it here first. Okay, a slightly different question. Let’s leave AI aside for 5 minutes. Social media is very good at picking up things that are here and now, but many important things happen quietly, over longer periods of time, and under the surface. What might be some important things happening quietly and gradually in the world now that we should be paying attention to?
One thing that’s happening quietly, and that I think is a disgrace, is the deterioration of K–12 education in this country. There’s a partisan angle to it, which is that we’ve worked around that in the United States by importing all this amazing talent. I mean, we’re at a tech conference. It’s hard to imagine the top 10 tech companies—I don’t know them off the top of my head—but I bet you 8 of them are run by an immigrant.
Now I’m worried not only that K–12 is deteriorating, but that no one is connecting it to the fact that our immigration policy is threatening to stop the flow of that talent into the United States. I’m also optimistic that policy will change, but it’s something I’ve been thinking about a lot, independent of our business.
The other thing that’s happening slowly reminds me of the old saying: How do you go bankrupt? Slowly, and then suddenly, all at once. For those of you who don’t know, I spent 2 years running around colleges in 2011 and 2012, trying to convince the kids that my generation was ripping them off with entitlements and that we had to do something about the deficit. All you had to do was look at the demographic tables, and sometime between 2020 and 2030, you knew what was going to hit the fan.
Of course, it hasn’t hit the fan. The reserve currency has continued to let us get away with it. Even back in 2011 and 2012, I didn’t anticipate $5 trillion would be spent on COVID. Fine, but then they just kept spending on top of it, bailed out by zero interest rates.
7. How smart people get in their own way
I’m worried that at some point we’re going to be looking at what’s happening in France and the United States. I’m also optimistic that if it happens, we’ll do the reform. But as a macro investor, that would be a very different crisis than now. I do think sometime in the next 10 years we’re going to get fiscal reform, engendered by a crisis in the financial markets.
So, last time, about 9 months ago, you said some pretty cool things—
I like that she’s quoting that interview, because I’ve been wrong a lot in my life, too. That was good timing.
Oh, yeah. I’m cherry-picking for sure. You said, “My edge is not IQ, it’s pulling the trigger.” You said, “Contrarianism is overrated. People sacrifice a lot of P&L just to be different.” And you talked about how smart people get in their own way because they don’t know when to stop doing research. So why don’t we add one more thing to the list? How do smart people get in their own way?
I’d say this business requires intellectual humility. Intellectual humility requires an open mind. I’m all for conviction. I’ve plagiarized Mark Twain and said, “Put all your eggs in one basket and then watch the basket very carefully.”
Yep.
8. What separates great investors
I’m good with making big, concentrated bets, but sometimes people with very high IQs forget the second part: watch the basket very carefully. They’re so in love with the thesis they made that they’re not open-minded. You can have a great thesis, it can be a great bet, and things just happen, and they don’t seem to adjust as quickly as other people do. I don’t know why, but it tends to breed inflexibility. Not always, but a lot of times.
And adjacent to that, you’ve met and mentored some incredible investors. What separates the very good ones from the great ones?
They’ve got to be unbelievably passionate about our business. They’ve got to be intellectually curious. They have to be competitive to the point of a sickness.
You’ve got to be sick.
So driven to win and so competitive, not wanting to lose, that they—
It’s not about work-life balance.
9. How Stan makes decisions
No work-life balance. Every time you buy something, there’s a seller on the other side. If you don’t have an incredible work ethic, you’re going to lose in our business. And the fourth, and maybe most important, is that they’re all disciplined. Their disciplines may be very different, maybe different from mine, but whatever their discipline is, they stick to it. I’d say you put those 4 things together, and that kind of answers your question, in my opinion.
So, if our audience followed you around for a month—which would be weird—what would surprise them most about the way you make decisions?
10. Predictions for the next ten years
What I call at our firm “investing, then investigating.” I’ll hear an idea. I’m a technician, too, so I’ll see a chart I like. I like the story; it makes sense to me. Or the person telling it to me is in their wheelhouse, and I can see their passion. I’ll make not a giant bet, but a big enough bet to really hurt if I’m wrong, with about 15% of the information that a traditional investor would want.
Then I’ll give it to our research department and delve into it for 2 or 3 weeks, and decide whether to add to it or get rid of it. So there’s a lot of betting going on, a lot of intuition with things that I probably have no business buying, but it’s worked for me over time.
We’ve had some pretty cool conversations, bouncing through history, psychology, and markets, and a lot of wild hypotheticals. I’d like to bring our audience into one of these conversations. Give me 1 or 2 predictions about the next 10 years that sound not that likely, but you’d quietly put a 20% chance on.
I think there’s a 20% chance we’ll either have fusion or it’ll be clear we’re going to have fusion. Number 2: I’m in Austin, so I’d say there’s at least a 70% chance, if not higher, that we’re going to have data centers in space.
11. Fill in the blanks
Number 3: I’ve been on the board of Sloan Kettering for 30 years. The first 10 years, we went nowhere. But then we got genetic sequencing. We figured out it was a personal disease: mutations, immunotherapy, and now AI. I think we’re going to beat cancer. I think it’s going to be a managed disease, and I mean in 10 years, not 20 years.
I love it. So, we’re going to end before they kick us out with a fill-in-the-blanks.
These horrible red lights. Go ahead.
I know. I think somebody’s going to physically move us offstage, but I want to do this one. So, we’ll do fill in the blanks. In the first 15 minutes, I meet someone and I can tell whether—
They’re a phony.
The personal trait that has made me the most money is—
Open-mindedness.
And work-life balance.
The personal trait that has lost me the most money is—
Well, I screwed up my discipline a couple of times. I got emotional and did something I knew I shouldn’t be doing.
Recently?
It was 26 years ago, but I remember it like it was yesterday. I had sold the Nasdaq in January 2000, and I bought it back on the day of the high. I think I missed the top by an hour, because I had the worst FOMO. Somebody said, “What’d you learn from it?” And I said, “Nothing. I knew it. I just couldn’t help myself.”
You still have PTSD.
I know.
The AI bear thesis that I respect the most is—
Politics, the financing—both concerning.
Well, it’s a pleasure sharing the stage with you, Stan. Thank you for the excuse to do it again. More than anything, Stan, thank you for your friendship.
Thank you, Iliana. Appreciate it. Thank you, everybody.