Adam Parker on the AI Supercycle, NVIDIA and Where Investors Still Have an Edge

1. Meet Adam Parker

I'm Michael Monahan with Founder and Friends and today we have a great guest, Adam Parker from Triariat Research.

Where do you think Nvidia could go?

I don't see how their earnings aren't at least twice as high at the end of this cycle as where they are now. We said $10 trillion would be an easy thing to see at the end of the cycle.

Are these still deep cyclicals, right? Or have they built enough of a secular growth story that they're not deep cyclicals?

Look, this is a supercycle. Compute is going to grow way above GDP for a really long time. Physical AI is going to be massive. I think the bull case for America that is nowhere near in prices is that we're going to make that stuff in the U.S. We're going to make all the humanoid stuff. We're going to make all the silicon and software that goes in it. There's a real chance that health care is the best-performing sector in the stock market over the next 10 years.

What's the one tool you would ask people to use in the simplest fashion?

You've got to be flexible, and you can't be anchored to only one way of doing things. The older I get, the more mindful I am of how many places I just don't have any skill. If AI hasn't made you a little bit worried about your skill, you might be a psycho, right?

Hey, thanks for having me.

Great to have you today, Adam. I really appreciate you making the time.

Yeah, it's good to see you.

So, for our listeners, give us one line on what Trivariate does. We're going to get deep into the details over the next hour, but why don't you add not only what you do, but also the 25 or 30 years you've been doing investment research. What's the one tool you would ask people to use in the simplest fashion?

Trivariate—the idea behind the 3 variables is some quant, some fundamental, and some macro. Those 3 variables apply to U.S. equities. Our business is that we sell research, bespoke services, and events to big institutions. Honestly, in the last year, I feel like I've written 15 times about how there are things that I now believe that I used to believe the opposite of.

I think the thing is, you've got to be flexible, and you can't be anchored to only one way of doing things. I'd probably stress flexibility and adaptability more than at any time in my life.

I think that's good. Markets change, and you've got to be adaptable to those markets. But I think the one thing that's changed for me over time is that when I started out, I thought it would be easy to pick short-term events and hard to pick long-term events. What I've learned over very long periods of time is that time is probably the last piece of alpha. It's very hard to figure out what's going on in the short term, but if you step back, I think we can find things going on in the long term. Has any of your research shown that?

The penalty for missing estimates for U.S. equities right now is really harsh, and way harsher than the reward for beating them. I think there's this kind of double-edged-sword thing: I want to focus on the long term, where maybe I can have an edge, but I also just want to follow the dumb investment advice: don't own things that miss.

Right now, I think the unconditional probability of beating is around 72–73%, so most companies beat.

But I think you just have to be careful. When you and I were starting the business, it was pretty reasonable to say, “I know current conditions are terrible, but this business has suppressed margins and suppressed multiples. If we look at 18–24 months, it'll be improving, and I can buy it now.” Whereas I think now you would say, “I don't want to own something I know is going to miss,” because cheap stocks that miss have never been punished more.

I think it's regime-dependent, but you've got to see what the market's giving you and also take advantage of where you have skill. The older I get, the more mindful I am of how many places I just don't have any skill, and how many people I'm with all day don't even know they don't have skill.

Maybe the answer to your first question is to be aware of where you have skill. If AI hasn't made you a little bit worried about your skill, you might be a psycho, right? You've got to really focus on that more than ever. That hones you in on where alpha is available, where you might be able to have an edge, and where you can avoid things where you just don't.

2. Long Term vs Pod Shops

Your point on the short term is well taken. If there's a hundred analysts at every multistrat, and all they're trying to do is figure out the quarter on one stock, if you're running a diversified portfolio of founder stocks, why would you be better at calling the quarter than those hundred analysts whose whole job depends on getting that right, leveraged 6–12 times? You're not competing in a game where you have an edge. The edge is going to be more about a diversified, risk-managed book held over time, or whatever. That's how I'm thinking about it.

And that's what we feel like is the biggest change for me. I feel like 25 years ago, when I started, those levered pod shops would have been the smart money. I feel like the roles have reversed, and retail and long-only investors, because we have the ability to hold for much longer periods of time, can make more economic-based decisions.

Let's do Palantir as an example. We can buy and hold Palantir because we believe in Alex Karp's long-term vision. We don't have to get the quarter exactly right. We don't have to get valuation right. We believe that over very long periods of time, if a stock's compounding at 100%, margins are expanding, and the Rule of 40 is going 127, 144, 145, 155, good things are going to happen to the stock price. A guy sitting in a pod shop doesn't have that luxury.

I don't know if it's smart or not. I think it's just that they're solving different problems. There's no doubt that most multistrats have delivered double-digit returns independent of the regime, and they're doing it with some of the smartest investment minds in the world doing the risk management. They're just solving a different problem.

If you're going to have relatively low turnover and hold stuff, then you've got to lean into what your edge is, which is thinking 2, 3, or 4 years out. They're going to turn the book over 600–800% a year. They could be in and out of it. They're going to be market-neutral, sector-neutral, and factor-neutral. So, if they're long Palantir, they're going to be short something against it.

It's just solving a different game. I guess I've learned that everyone is playing a different game, and they've raised money under different rules. The arbitrageable thing is that everyone's playing a different game. As long as you know what game you're playing, I don't mind playing the game of thinking long term.

One thing I feel like I know, without any arrogance, is more than a lot of people: I know a lot of people who made a ton of money for themselves while they didn't make any money for their investors. I know a lot of those people because your incentive under that short-term game is, “I want to make $17 million for myself this year, and if I lose money next year, so what? I'm fired. I made $8.5 million a year for the 2 years, and I'll get to the next one and the next one.”

We've both seen résumés with people who have worked at 3 or 4 of the major multistrats. They gun it one year and get blown up the next. But their incentive is to gun it because, if you get it right, you can make massive wealth for yourself during that year. The big mothership doesn't care because they've got a hundred guys doing different strategies, then risk-managing it, and they can still generate double-digit returns.

3. Why Valuation Misleads

That's different from when you're running money for—or managing—an active ETF, or whatever, where you need to deliver above-S&P returns or at least have an upside-to-downside capture ratio over a cycle. That makes sense, right? It's a different game.

I think that's really good. One of the things I've thought about over the years is that I put public equities into 4 buckets: value, GARP, growth, and what I call public venture.

There are some companies that are public, but people really think about them as venture: SpaceX, Tesla, and even Palantir kind of live in that category. Then there are the ones you were talking about that were deep value. I've always thought they have different investors and trade on different metrics. One of the mistakes I see people making is that a value guy will try to put value metrics on a growth stock, and that's not going to work.

I'm not really sure. We've done a lot of work that shows valuation doesn't really work to pick stocks, especially 12-month forward.

Right. If you think about what investors have been wildly off on, Micron's a great example. 2 years ago, the sell-side analysts thought they were going to do—I have to check, but something like $6 billion in free cash flow over the next 8 quarters—and they now think it's around $300 billion.

It’s like the order of magnitude—you’ve got to get the E right. Yeah.

Right. And—

I think there’s also information in it now. If you look at stocks that are in the cheapest decile on price-to-earnings—about 8 times forward earnings or less in the S&P 500 universe—and they just got cheaper on price-to-earnings in the last quarter, they have about a 48% chance of beating estimates.

If they’re in the 8th-most-expensive decile—30, 31, 32, 33 times forward earnings—and they just got more expensive, they have a 75% chance of beating estimates. If you run a diversified portfolio, do you want to own something that has a 75% chance of beating estimates or a 48% chance?

So now there’s information in the multiple and in the change in the multiple that didn’t used to be there. I don’t want to ignore that.

I don’t think I want my portfolio to be cheaper than the broader market, and I certainly don’t want to own stuff that’s in the cheapest decile. If I show you, with your experience, all the mega-cap and mid-cap stocks that trade at, let’s say, 10 times price-to-forward earnings or cheaper on a 12-month-forward basis, there isn’t a single one where you say, “Wow, I didn’t realize United Airlines might be a good stock or a bad stock.”

We’ll see, but you’re not confused about why it trades at a low multiple.

So you can’t use valuation to pick securities in that way.

Right. Yeah, I agree with that.

And you’re absolutely right. The most negative person on U.S. equities is always a British economist or a British fixed-income economist. If you listen to that guy and his valuation framework for picking U.S. equities, you’re left behind.

Yeah, absolutely. And I think there’s a lot of credibility to you making that statement, just given our career arc. We’re going to do your career arc because I want to get all the way back to Micron and think about how we’re going to value these memory stocks.

4. Bernstein Semis Origins

You really had real deep-research value at the beginning. It’s interesting that you spent all these years in value and valuation and got to the conclusion that maybe that isn’t even a big driver of stock-price performance. Let’s go all the way back to when you and I met at Sanford Bernstein. We’ll fast-forward past the fact that you were the No. 1-ranked semiconductor analyst and went on to be the strategist. Let’s cover that part of your career. What were you building in your investing framework at that point?

You know, it’s funny. I launched coverage of semiconductors at Bernstein on October 1, 2002, so we’re coming up on 24 years after the launch. The title of the launch was “Share Gainers and Margin Expanders Are Multiple Expanders.”

For those listening, I spent 11 months in a room doing nothing but researching 7 securities. When I launched coverage, I was on 7 stocks in the semiconductor industry, and it took me 11 months to produce that document. Retrospectively, that’s a lot of time.

I’ve written a lot of things that aged like milk, but that one has actually aged okay. I think if you expand your gross margins, you’re generally going to get multiple expansion. I think that thesis still holds.

I think the market is a little bit less second-derivative than it used to be. People used to always say, “Everything’s about the rate of change.” This past year, I’ve been thinking less about that and more about how the absolute level of growth is so high that stocks can probably absorb multiple contractions and still be pretty good securities.

I think covering semiconductors was luck, but it turned out to be a great industry to cut my teeth on all those years ago. I didn’t know Nvidia was too small for me to cover by market cap and liquidity when I launched.

Yeah, it’s crazy. I remember 20 years ago I got introduced to Jensen at CES, and clearly I should have spent a little more time on that introduction.

Yeah, those companies had a competitor called ATI Technologies that got bought by AMD. They were just random-number generators.

If you asked a broad group of respected buy-siders 25 years ago about Nvidia—competent or incompetent, honest or dishonest—they probably would have put Jensen in the dishonest and incompetent quadrant. Maybe smart, but he was shilling some suboptimal GPU. The world has changed a lot.

5. Micron From Disaster to Titan

I think that’s another lesson. Micron—we’ve alluded to it a couple of times—but they had, and this sounds made up, exactly $365 million per year of free-cash-flow burn over a 5-year stretch. You can look it up.

I wrote a note around 2005 where, at the time, there were something like 200,000 people living in Boise, Idaho, where they’re headquartered, and Micron had 9,000 employees in Boise. If you got rid of old people who were retired and young people who didn’t work, it was a meaningful percentage of Boise citizens working at Micron.

I wrote a sarcastic, wise-ass note basically saying that if they had gone to the parking lot and just given the employees $1 million a day, it probably would have been a better use of capital than trying to produce DRAM chips in Idaho or whatever.

Those businesses were bad businesses, and through consolidation and this massive AI supercycle, you got yourself a trillion-dollar market cap 20 or 25 years later. Things change. I think that flexibility theme is worth emphasizing: don’t anchor yourself to one way of thinking. Incompetent, dishonest and terrible businesses can turn into the most important securities in the world.

And thinking about it, I was anchored on Micron. I was anchored by a good friend of ours in the business. You probably know who he was, but he was a banker at Goldman and is still on the Street somewhere else. Almost his whole career was refinancing Micron multiple times. He made his career being Micron’s go-to capital-markets banker.

You could have said 10 years ago that only through the idiocy of the buy side did Micron even make it.

Like the converts. They had some bad cycles.

Yeah, so he led all those converts for them.

But now it’s almost the other side of the spectrum. What the Street is telling you about the stock is that they’re going to lose money in a couple of years, because otherwise it would trade at a higher multiple.

I think the valuation comment is interesting because people make that statement about valuation independently of the balance sheet. Micron is an interesting one: it trades somewhere around 5 or 6 times peak earnings, maybe 10 or 11 times normalized earnings, to the extent we know what normalized means. But people don’t say there’s any difference if they have net debt or a third of their market cap in cash.

They’re going to generate something like $300 billion in free cash flow in the next 8 quarters on a round-number $1 trillion market cap. Shouldn’t that be valued somewhat differently than if they had net debt?

Your Goldman buddy would now have to figure out whether they can buy Constellation or Ciena with their inflated equity. If their equity is really that inflated, they should use it to buy something. If the stock gets killed, it gets killed. Who cares? They’ll be in a better position in 5 years.

The world has totally changed on that. But I do think one of the reasons valuation doesn’t work is that people can’t get the E right. Another reason is that they’re doing it almost independently of the balance sheet, and of course it shouldn’t be.

Yeah.

6. Are Memory Stocks Still Cyclical

Yeah. I want to go through your career progression, but I want to stay here on Micron. We’ll do the career later, so let’s—

I covered semiconductors, so it’s come full circle. It’s probably the most important sector in the market.

Do you have a differentiated viewpoint on whether these are still deep cyclicals, or have they built enough of a secular-growth story that they’re not deep cyclicals? I’m still anchored on “deep cyclical,” but I’m willing to be talked out of it.

Most semiconductors are not deep cyclicals. I’m speaking to the memory stocks specifically.

Yeah, memory—and the reason most of them aren’t is that when you and I learned the industry, a lot of the companies still manufactured the chips. The manufacturing causes some of the cyclicality because it requires you to invest capital to build the buildings—

—and buy the tools.

Very few companies manufacture anymore. Micron is one of them, along with Taiwan Semiconductor and Samsung, but there are very few. Nvidia doesn’t make chips; they get Taiwan Semiconductor to make them. So it’s a different world in terms of the capital intensity of the industry and the cyclicality.

But I think the answer to your question is that this is a supercycle. Compute is going to grow way above GDP for a really long time. Sure, it’s cyclical, and DRAM is more cyclical than other parts of the semiconductor chain, but we have a better grip on demand and supply over a 3-, 4- or 5-quarter view than we used to.

And we're short compute. We're short power for the compute. People can't get the chips they want, and demand is going to be massive. So, if semis were, say, 2% above GDP for 30 years, it might be 8%, 9%, or 10% above GDP for 10 or 15 years, and at a higher margin for the US companies that are in that space.

It's just a different paradigm than you learned 25 years ago, right? I think that's the part people don't get. I think the physical AI part isn't in anyone's—maybe it's a little bit in Tesla's stock price—but it's not in anyone's purview that we're going to have millions of robots doing things.

7. Nvidia to Ten Trillion

Those robots are going to have shedloads of silicon in them. I think there's a long runway. It's cyclical, but the periodicity and the amplitude are way different than history. Before you and I reconnected, I had read a piece of your research—or maybe Lauren had shared with me a meeting she had sat in with you—where you talked about Nvidia being, and I'll let you give the order multiple, much bigger than it is now.

Given that we love founders, let's drill in on Jensen, because we really think he's one of the ultimate founders. Where do you think Nvidia could go? You have a big forecast, and it has to do with these physical robots.

Yeah. Yeah. I mean, we said that a long time ago. I guess I'd start with this: We all misremember when things started, right?

Yeah.

Like, one of my children, who's 23, said to me a couple of years ago something about the Apple App Store and referenced that as when I was a child. I said to her, "Honey, there wasn't an Apple App Store until 2008. I was 39 years old when it launched. I wasn't in high school. There were no apps." She was like, "Holy cow."

So, I think we all—

—don't remember life when there was no App Store. But—

—for a lot of us of my age, I took a deli-counter ticket at the University of Michigan and waited in line to get access to a PC so I could type my papers when I was in college, right? The technology stuff—we misremember.

I had too many summer associates. People say, "Hey, can my kid, my nephew, my niece, my—"

—you know, and so I try to say yes into the pay-it-forward thesis.

Okay.

But since we've had a lot of these LLMs, it's hard to tell them to give them projects. The last 3 or 4 years, what I typically do the first couple of days is say, "Here's some year-ahead outlooks. Here's Goldman Sachs, Morgan Stanley, J.P. Morgan, and UBS." Each one of them has a year-ahead outlook for equities and for economics.

Sometime in the middle of the summer of 2023, I gave this kid those 8 documents. He read them over the first couple of days, went away, came back, and said, "You know what's interesting? None of these 8 documents have the letters AI in them."

In November 2022, when all these firms did their year-ahead outlooks for 2023—economics and strategy—the letters AI were not in any of those 8 documents. We all think we've been talking about AI our whole lives, but the truth is, very few people in the investing world knew what it was 3 and a half or 4 years ago, okay?

They weren't positioned for it. They didn't think it would be relevant to the economy or relevant to investing at Goldman Sachs, Morgan Stanley, J.P. Morgan, and UBS. Let's say the best 4—at least the 4—

—European and US firms. Okay. So, that's telling.

If you look at the misremembering of the App Store in 2008, and if this is a 10-year investment cycle and May 2023 was NVIDIA's first big upward sales revision, one way to think about it is that you're 3 years and 4 months into a 10-year investment cycle.

Part of me wants to say, why the hell would the stock have already peaked? Let's say we're usually 6 months anticipatory with semis, and it's 12 or 18 months anticipated here. We still have a long way to go in terms of the absolute growth rate.

I'm framing it a little bit that way in my head. NVIDIA has a couple of special situations associated with it. One that I don't think is part of everyday vernacular is that they have a software platform called CUDA. They lost money on it for 15 years, and so now it's this installed base where you can't really do any complex inference problems without CUDA.

They just have a massive lead that's going to be hard to replace.

And I think that's the 2 sides to the coin. What you, 20 years ago as a sell-side analyst, would have said was, "What's this guy doing in his leather jacket?"

He's got his vision. He sees something that no one else sees, has the courage to see it, and then, in his favorite word, the grit to go.

Yeah. And look, I get there's some criticism of the circular lending. I get that if you buy a hotel and then you sleep in it and count your own bed as occupancy, people could be critical of some of the circular lending and access-to-capital stuff.

At the end of the day, I don't see how their earnings aren't at least twice as high at the end of this cycle as where they are now. 15% a year for 5 years is a double. That's a very conservative view of their 5-year revenue.

I think the odds are that they'll accrue a lot of cash—there won't be a huge free-cash-flow-to-net-income disconnect over that whole cycle. I don't think it's crazy to say the stock could be 70% or 80% higher from here.

At the time, I had mentioned to Lauren that it was a double, but the stock's up maybe 20% or 25% since then. I think we said $10 trillion would be sort of an easy thing to see at the end of the cycle. You're at whatever, $5.5 trillion now, so let's call it 80% higher.

That embeds some continued multiple contraction. It's just that the growth rate is going to [laughter] be high.

Yeah.

8. AI Estimates Still Too Low

So, you can get to a double using very traditional spreadsheet math. If you were a Sand Hill Road guy who really gets charged up by, "How good could it be? If you were wrong to the upside, what could it look like if you're being too conservative?"

Well, it's an interesting framing. I like that question because, again, you and I were raised in a regime where we were used to sell-side analyst estimates being too high.

There was a central place to post forward estimates starting in 1978. If you look from 1978 forward, about 80% or 85% of the years, the analyst estimates in January for that year turned out to be too high. They would start with an average of 14% earnings growth, and the answer was something like 8%.

The market could still go up in those years. There were down revisions, as long as you thought the next year's earnings were up. But we were used to overly optimistic assessments.

Since 2023, since NVIDIA's upward revision, estimates have been consistently too low. While that's enough data to conclude that something's changed, most people don't ask the question the way you did: Could there be an upward bias to the estimates?

I think there will be. If I look out, I'd say the area where they're most low is probably 2028. I think a lot of people are just embedding this decline. At the same time, you're probably going to see more AI productivity than you've seen yet.

We do a lot of natural-language processing of transcripts and the like. We only think about 20% of US equities are even saying they're doing cost-related AI productivity today. There are going to be way more than that by 2028. So, I think there's a real chance that the 2028 estimates are too low.

The sell side has 18% next year's earnings growth, and if you tack on 10% for 2020, you get 455 in S&P earnings. I think there's a real chance it's more like 475 or 500. If you want to pay 19 or 20 times that, I don't see why you're not at 10,000 on the S&P.

I think that's in your statistical expectation. You assign a probability to earnings and a multiple. I think there's a higher probability of upside to earnings than what's in the consensus.

Yeah. I think so. And you know—

That's a good question, because I don't think people—I think people are assuming downside. They're assuming a fall off a cliff here because they're baking it all on access to capital to fund it all.

A smart investor said to me a couple of months ago, "Musk has a superpower, which is unlimited access to cheap capital," right? The core businesses generate a lot of free cash.

9. What Could Go Wrong

There are things to worry about. I don't want to whistle past the graveyard, but I think the estimate skew is probably to the upside.

Yeah.

No, I think on our platform—there are plenty of people worrying about what could go wrong, right? I feel like that's almost the tenor right now. I probably took calls from a half dozen or a dozen financial journalists this week. Every question was, "What could go wrong?"

If we think about a classic wall-of-worry framework, we're doing pretty good.

Yeah. I think "What could go wrong?" looks like a question that's hard to answer. There are a few of them, right?

If we're right that there are just so many use cases and the return is pretty high, there's a scenario 3, 5, or 6 years out where unemployment really does get to be a problem, where white-collar jobs in the $100,000-to-$500,000 range could have meaningfully—kind of 10% or 15%—unemployment.

We'll see. There is a thesis that there will be new jobs forming.

I think it'll be right around when physical AI takes off and you get a lot of jobs around humanoid production and all that.

What is it that's hard for you to say you're confident unemployment isn't a problem in 5 years? I would say that's one where I don't want to be on TV saying there's no problem, because that could age like milk. You know what I mean?

Yeah. Let's talk about one place. There was a big piece of news yesterday on white-collar job replacement: 3 of those major investment banks that you talked about have gone to the top 100 law firms and said, "You've got to restructure pricing," because the LLMs are taking over a big piece of Big Law. That was paralegals, first-year associates, and junior partners making word edits and billing that out, and the LLMs have taken that away.

So, it's there. We'll see what happens. But I want to go back to—you can comment on that if you want—but I want to go back to physical AI because I think that's part of the upside to chips that people aren't talking about. I want to give you both the credit and, for our listeners, the opportunity to learn from you, because you've been so early in outlining what physical AI can do with GPUs and other AI chips.

Yeah, on the lawyer thing—or the white-collar thing—I think there are a lot of industries that will be almost bimodal. If you're a few people and you have expertise in a niche, you'll do well, and if you're massive, you'll do well. But these sort of 1,000-attorney firms that are in the middle might get hollowed out because they're not big enough to pay the rainmakers at the top to get those huge private-equity firms and huge deals. These guys are like NBA players now, where they're paying $30 million a year for these stars. You can't really hire them and fund those unless you have a bigger mass of lawyers.

So, I think there'll be some consolidation and mergers, and maybe even private equity and maybe even public stocks at the law-firm level. I think there's a lot of change there. I will say—and this will probably annoy some of your listeners—but I still think lawyers have a lot more differentiated skill than investment bankers, on average.

Because if you really think about a lot of deals, I would say, look, I think some of the AI tools are pretty good at taking pro forma P&Ls. I could take 2 businesses and create a pro forma balance sheet, income statement, and cash flow using Claude pretty well. So the value-add of some whippersnapper from Wharton is kind of low if I can basically do that. Then all the real questions I have are: Is there an antitrust issue? What's the tax benefit of the SpinCo? What's the whatever? I still need to call the same 5 law firms and talk to the experts there.

I think the challenge for all these businesses will be: How am I going to get the 50-year-old MD in banking or the 50-year-old partner in the law firm? How am I going to get enough through the funnel? Because I don't need as many people, and their skills at the early level are not needed as much. So, we could probably apply it to a lot of businesses, but you maybe go down that rabbit hole.

10. Physical AI and Humanoid Care

In terms of the semiconductor thing, look, physical AI is going to be massive. I think people need to—it may sound odd to them—but imagine we're really short geriatricians. We don't have enough people to care for old people in this country. We have an aging population. So, would you let a humanoid take care of one of your parents in their 80s?

I think the answer is yes. I think if it has a lot of silicon in it, sensors, and it knows if your parent is taking their medication or if something's wrong with their blood pressure; if it's measuring their vitals, it can quickly contact an emergency vehicle, which will be driverless. There's just so many efficiencies.

I feel like the things I'm most confident in on a 5- or 10-year view are going to go way above GDP: healthcare services, compute, and power. An equity portfolio should have some exposure to those themes.

Okay.

Yeah. So you're going to have a humanoid in your lifetime for sure. Yeah, yeah, for sure.

Yeah. My view, what I've been telling friends—I have a friend who I surf with a lot, who also—he's retired and he loves to work on his classic cars.

And he's also really interested in AI, and I think his biggest single-stock position is probably Nvidia. So he pays attention closely. I said to him, "You and I and early adopters are going to buy a humanoid sometime in the next 3 years." And I said, "If you want to be an early adopter, you're going to have one helping you in your garage in 5 years, and everyone will probably have one in 10 years."

Yeah. I just think—yeah, maybe. I think you'll have a lot of them. It depends what the tasks are. If they're very measurable, concrete tasks, I think you'll rely on the humanoid to do that, like mow your lawn or different things like that. Those things are kind of easy.

I think the geriatrician is a perfect example. I think it's a perfect example, right? You just see old people, and they have an aide helping them. There just aren't enough of those aides. There aren't enough people willing to do it. If we have millions of people who are aging, I think you'd say, "Wait, I can have somebody whom my parent isn't going to accuse of stealing. I'm not going to worry if my parent says something inappropriate in front of this person, right?"

It'll measure their vitals. It'll have sensors and cameras. I can talk to them at any second, and they can talk to me at any second. It can text. It can know things. It's just the efficiency. There won't be any mistakes on drug delivery or whatever.

I just think you're headed to a world where it's going to be sturdy. It can hold your parent's arm. It can take them to the driverless vehicle. It can walk them into the doctor's office. It can learn. It can get the script. Think about all that.

11. Humanoids And New Jobs

I think the bull case for America that is nowhere near priced in is that we're going to make that stuff in the US. We're going to make all the humanoid stuff. We're going to make all the silicon and software that goes in it. There'll be some blue-collar jobs around all that that'll fill the void from the paper-pushing jobs that are going to get replaced by AI productivity.

There's a—I would assign a higher probability to that than the consensus. I don't want to get dreamy, but that's where I think people aren't thinking about the world yet. Maybe they are with Tesla's stock price and Musk talking about humanoids, but it's not an everyday vernacular.

No, I think it's fair. I think it takes more horsepower to think about what could go right than to be cynical about what could go wrong.

12. Debt Fears And Markets

Oh, you always sound smarter when you're bearish, right? I've been asked 50 times a year for the last 15 years, "Are you at all worried about US government debt?" If you're on TV and you say, "No, I'm not worried about it," at some point, you're a dick, right?

I think the question is not that. The question is, should I change my equity-portfolio positioning today, or my asset allocation, for the fact that that's about to matter right now? The answer almost every day for the last 15 years was no. The reason it's topical right now is, of course, people are worried about bond yields backing up. But today's an interesting day as we're filming this: the 10-year yield is up, but the tech stocks are also up. Sometimes the 10-year yield goes up and stocks go up because growth is good. I think right now the earnings upside is still pretty good. I get fired up about that question because I feel like it's just always wrong, and you get fired if you position your fund for that.

Yeah. No, I think it's a really fair answer. I get asked it a lot, too. I gently say, especially with newer people who ask me questions, "Look, that's a hypothetical. It's so deep into the hypothetical. First of all, it probably doesn't affect my investing time horizon, but secondarily, there's no data or evidence that supports it. I've got a limited amount of decisions I can make, and I have to feed some sort of data or evidence into that decision-making process."

You're talking about something where we don't have anything. We don't even have a beginning path to go down.

Yeah. Yeah. Yeah. I think that's right. Or it's just as simple as, if that's what the person asking you is worried about, they probably just shouldn't allocate to equities.

Right. If the question is, can I beat the equity market? That's a different question from whether you should be in equities altogether.

Right. Yeah.

13. AI Purpose And Joy

Yeah. I've got a philosophy question. I know you've got a PhD in statistics, but let's go back to philosophy, and then we'll go back to some of your career progression, because I think it's super interesting. I want to make sure we don't forget about the new newsletter that you're doing.

14. Wrap Up And Where To Subscribe

Yeah, great.

On philosophy, I was getting out of the subway at 57th, getting off the N/R, the other night, and the motorman is a guy my age. He just had ultimate joy on his face driving the train. He's pure joy.

He doesn’t care that it’s hot. He doesn’t care who the mayor is or isn’t. You could tell he had pure joy driving the train.

Right.

I wrote a tweet after this and basically said, “I’m an AI maximalist.” I’ve been using 2 to 4 LLMs every day for the last 18 months. I believe in robotics. One of the coolest experiences I had was getting picked up by one of the robotaxis the other day. It’s just a better experience than Waymo, right?

And so I’m all in on AI. My portfolio is all in on it. But I looked at this guy and I said, “We have purpose and prosperity for a lot of people operating these big machines.” Forget the white-collar workers, but how are we going to replace the purpose of these people? I’m thinking especially about the jobs people really don’t like. End-of-life care is very tricky, but how are we going to replace the purpose of these people? I think that’s the most important thing we have to think about. There may be no answer, but how would you frame that?

15. College ROI And Trades

Well, the return on the cost of education right now—the cost of education is way too high.

Yes.

It’s been many years that, for the median person, it hasn’t been a very good return. I think we all engage in things where the median outcome isn’t a very good idea, but the 90th percentile is so awesome—your business, my business, starting a restaurant, doing a Broadway show, or starting an ETF. The median one underperforms, doesn’t raise assets, and fails, but the 90th percentile is amazing.

For the median person, going to college is a dumb idea, but for the 80th- or 90th-percentile person, it’s a good idea. That distribution has shifted and stopped. It used to be pretty clear that the more education you had, the more money you made. I don’t think that distribution has grown at the rate that people thought.

There’s a lot of data out there that your average HVAC employee makes more than your average Harvard grad, because the skill they’ve learned translates to income at a pretty good level. I think those physical things will still be in demand.

Think about all the stuff that’s going to happen around humanoid robots. You’re going to need metals, copper, power, sensors, and software. You’re going to need to maintain them, and there are going to be huge factories producing them. It could be a sneaky-big reindustrialization of America, like an industrial revolution in the 30s, because I think you’re going to make a ton of these. I think we’re going to end up with several of these things.

Look, if I told you a couple of years ago that I’d have a sales agent, you would have been like, “What are you talking about?” But we have sales agents that are amazing. They do things that, when you and I started the business, salespeople would fax the research to somebody else.

Yes.

Right. So just think about what happens now. I can measure the mouse movement of every client on everything I produce. I can tell—I mean, the amount of data that you can analyze is incredible. I’m just very bullish on the efficiencies that will come from that, and I don’t think demand or supply will get in the way for 12 months or more unless there’s some weird policy.

16. Data Centers And Narratives

Twenty-three states now think that a data center is not something they want in their state. But Google—look up the facts. What do you think uses more water: a midsize data center or a golf course? The answer is, it’s close. The biggest data center in the world would be something like 3 golf courses in terms of the amount of water it consumes.

If you listen to your average politician—and this is both parties, so I’m being politically neutral when I say this—they act like you’re using the entire Pacific Ocean for a data center. The answer is 2 golf courses for the biggest ones.

Yeah.

Yeah. And I think people—

So people just have a misconception. I don’t know if it’s their Twitter feed or what it is that gets them into this sort of narrative.

I was asked this question yesterday on Bloomberg Surveillance, and my answer was that the narrative simply has to change. I think what happened was Sam Altman and Dario Amodei kind of scared everyone, and luckily Jensen Huang and Mark Zuckerberg are going back and cleaning it up a little bit. Mark has made some big investments in the community and in training workers. Jensen has talked about prosperity. Elon Musk talks about prosperity and abundance. Travis Kalanick talks about abundance.

What I mentioned on Surveillance was that people like Netflix. They like same-day Amazon delivery. They like Starlink bringing them internet anywhere. People kind of like those things. That’s what a data center is. Nobody had a problem with a data center then. So it’s about reeducating people that data centers bring them a lot of the things that make their daily lives easier.

17. Productive Years And Healthcare

Yeah. I mean, I was at a meeting maybe 10 years ago with Michael Bloomberg, when he was mayor. There was a lull in the questioning, so I asked him, “How do you know if you’re a good mayor? How are you going to measure whether you’re a good mayor?”

In my mind, he gave an incredible answer. He said that if the citizens had more productive years—if the average person lived longer and was more productive while they were alive at the end of his term than at the beginning of his term—then he had done a good job. There’s a distribution around that. It’s not just the median or the average.

The idea would be that if people were living until they were 75 and working and producing a certain amount of output up until they were 70 when he started, and that number became 80 and 75 when he finished, then he had done a good job. A lot goes into that: education, jobs, health care, safety, and so on.

I like that concept of productive years. The whole point of all the AI stuff is that we live longer and are more productive while we’re alive. In that mindset, I think people don’t quite understand what I’ve concluded isn’t in the price: compute, power, and health care services. I think the health care part in particular is something people don’t yet understand.

There’s a real chance that health care is the best-performing sector in the stock market over the next 10 years. The market is telling me there’s a 0% chance, and I think it’s 30% or 40% or whatever. I want to arbitrage that difference, because if you go to Nvidia’s website and click on the health care tab, Jensen knows a lot about health care.

Okay.

Yeah. I just feel like that’s an opportunity set that’s not being recognized. You’ve seen a little of it with Quest and Labcorp, but there are a lot of low-margin businesses with lots of employees that are going to get much more efficient as they predict their employee and customer behavior. I’m really interested in that area.

Yeah.

We’ll see.

Yeah. Makes sense.

18. Portfolio Strategy Not Calls

All right. So I wanted to go through your career, and we got derailed.

No, sorry.

No, we had a great—this was a fortuitous adventure talking about Micron, Nvidia, and some of these things.

I think we left off when you became the Bernstein strategist. Correct?

Right. And those are big shoes to fill, because the Bernstein strategist has been a big position over time. What did you learn in that role?

I think there’s a difference between being an analyst and thinking bottom-up about a stock and an industry, and being a portfolio manager, where you have a bunch of different stocks and have to create a thought process around how much alpha is available to me and how much alpha I can generate per unit of risk.

I think it became more about how I should think like a portfolio manager rather than an analyst. In my mind, that was the difference between being a strategist and being an analyst.

What we do in our business at Trivariate isn’t really what people think of as traditional strategy on the Street. They think of sector preferences or stock market calls. I made this comment earlier, but I know I don’t have any skill making a one-month stock market call. I also know that nobody else does, but they act like they do.

I feel like there’s power in knowing that isn’t my strength. When you have your own business, you can’t really build a massive business making stock market calls. When you work at one of the big firms, you feel like you have to do it because there are a bunch of financial advisers, and you make up some B.S. framework for your target that’s always wrong or whatever. I just don’t think that’s a good use of time.

When I think about portfolio strategy, and what I write a lot about, probably 60 of our clients on the institutional side send us their portfolios to do custom risk work. We do a lot of risk management. This week, I wrote a note about whether you should have a stop-loss program. If a stock goes down a certain amount, should you actually put a rule in place around that, or should you run more diversified than normal in this regime?

If you normally own 40 stocks, should it be 60? Could something make you go to 60 in the next 10 years? What is it? How do you think about adding to your losers versus trimming your winners, versus selling your losers and adding to your winners? It’s about how you trade.

So, it's more strategy and portfolio construction. Where's the alpha available if you're a bottom-up stock picker? Where should you spend your time, versus where should you own stuff? How many names do you own for alpha versus risk? To me, that's real portfolio strategy.

I think a lot about those issues. Today, as an example, I think you need to compute your beta for every stock. You need to have an AI beta and a non-AI beta because we have the most negative-beta stocks in the history of the stock market right now. If I have a 10% position in Exxon, it looks like negative beta to the S&P. I don't have a negative position in Exxon; it's just that the metric is meaningless, right?

Those are the kinds of things we're constantly writing and thinking about for our institutional clients. To me, that's real portfolio strategy, and there's real value added there, as opposed to, “I'm bullish on equities, and I think the stock market targets 8,000,” or whatever. Charlie Munger would say, “Beta is BS, earnings are BS, and that's BS strategy.”

19. Trivector Edge And Clients

I want to ask you what makes your firm so different, because it is really differentiated. But I think you just showed everyone.

Yeah, a little bit, hopefully.

Yeah, I didn't realize it was lucky.

I have a PhD in statistics. I covered semiconductors, and our business focuses on risk management. Semiconductors became the most important sector in the market, so I just got lucky that risk and semiconductors have been the 2 things that come up a lot.

We write a lot about where the alpha is available. If you're a bottom-up stock picker, where should you spend your time to differentiate from the index? We write a lot about management decision-making because human beings introduce volatility into stocks. We analyze a lot of buybacks versus accelerated share repurchases, spin-offs, M&A, M&A rumors, broken follow-ons, and IPOs—everything where humans introduce volatility.

I used to think quantitative stuff plus fundamental stuff combined was superior to either alone. I don't really think that anymore. I think some stocks are better for quant, some are better for fundamental, and some are better combined. If you're a fundamental stock picker, you should just focus on where the fundamentals are better and then have some sort of risk-managed group of stocks around that.

Our clients are all the biggest asset managers, hedge funds, sovereigns, private equity, crossover funds, large RIAs, law firms, boards, and management teams or companies that care about either management decision-making, available alpha, or risk management. We do some tech strategy, too, but it's not, “I think the S&P is going up 3%.” That's complete nonsense.

20. Career Speedrun And Newsletter

Yeah, I think that makes sense. I want to come to the fact that you're now making that available to non-institutional investors. Let's finish out your career, because I want to speedrun your career. I think that's what is actually going to help people, but I also think it'll be fun to learn about your career because there are some really neat embedded lessons.

So, let's speedrun it. You were a Bernstein analyst, then you went over to Morgan Stanley. You were their chief strate—

Yeah, chief equity strategist.

Chief strategist. You went to someone else's buy-side shop, started your own buy-side shop, and started a sell-side shop. Do a speedrun on each one of those, and then also what you learned switching from sell-side to buy-side and what you learned switching to running your own company.

Yeah, you nailed it. I worked 11.5 years at Bernstein, was semiconductor director of research for 1 year, and then became a strategist. Then I was at Morgan Stanley as the US equity strategist. It's a great privilege to work at both of those firms.

Honestly, at Morgan Stanley, I spoke at 44 conferences in 1 year and traveled for 32 weeks, so it was more of a figurehead kind of job. But you have a lot of access to the biggest hedge fund investors in the world, and you have very transaction-focused people. There are deals, fixed income, equities, economics, credit, and currencies.

I would describe it as Bernstein going from MIT football to Alabama football. Morgan Stanley is just so much bigger. It was a great learning experience, but I got the opportunity to work at an amazing buy-side firm called Eminence Capital, which recently closed, run by an amazing investor named Ricky Sandler.

My job there was to help the firm get more analytically rigorous around risk management, position sizing, and diagnosing trades, which was a natural segue into some of the risk-management work we do now. We started Trivariate's institutional business in May 2021, so we just had our 5th anniversary a few months ago.

About a year ago, as you alluded to, we started a business selling a newsletter to individuals and advisers. As 1 of my kids says, I'm sort of a D-list celebrity because I go on CNBC a lot or whatever. I don't know—maybe 2 or 3 times a week, somebody comes up to me and says something in an airport or on the street. I thought I should probably write something that addresses that group.

We write a newsletter with 3 or 4 short things a week. We do a video where I talk about thoughts from the meetings I have. We'll do a monthly webcast where people ask me questions, and we have some screens of ideas and that kind of stuff. It's $110 a month or $1,200 a year for any financial adviser or individual under Trivariate Research.

It's been a cool way to interact, and it's fun when somebody comes up to me and says, “Oh, I'm a subscriber,” and they take a picture or whatever. My kids think that's hysterical. I like it because we write things for that group such as how to dollar-cost-average new business or how to think about which ETF to choose. We actually have a Dave Portnoy pizza-style ETF score where we say, “This one isn't really very good, but this one is good, and here's why it's giving you the exposure.” Once yours is at a big number, we'll give you a letter grade and celebrate together.

I look forward to it. Did you get the pizza example from Lauren, or did Lauren get the pizza example from you? She describes our fundamental overlay, right? We mixed in some of our old Bernstein stuff. We run a fundamental factor model on top of the portfolio, and she describes it as checking the pizzas when they come out of the oven to make sure they don't have burnt crust.

Maybe we just both like pizza, because I like this Portnoy thing where he gives it an 8.4 or whatever.

I agree with him on the Rhode Island pizza, by the way. A good example is momentum. Momentum was an amazing factor for a while. There's an ETF called MTUM, and it underperformed the S&P while momentum was awesome. That was part of the reason we gave it somewhat of a poor grade.

It's not necessarily how it performed; it's whether it's giving the person the exposure they think it's giving them, right? That's how we score it.

21. Leveraged ETFs And Zero DTE

Yeah. I know we've got about maybe 10 more minutes. You've been super generous. I think I asked for a half hour, and I took you for a full hour, so I appreciate that.

Maybe another time we'll talk about ETFs, because that's another thing I've put into buckets. It's long-term compounders, right?

Right?

Okay. But they're not the best expression of compounding, and you're on the wrong side of the house edge. The long-term compounders are things like VOOs and SPYs. Those are okay, but they're not the best expression of compounding.

Then there are things that are just structurally inefficient: buffers, covered calls, because over very long periods of time, you're giving up a lot of compounding for some near-term certainty. Then the ones that are uninvestable are doubles and levers and things like that.

That last point is interesting. I wish I could get the number in front of me, but somebody will fact-check this. There's something like 700 leveraged or inverse ETFs today, up from 45 3 years ago.

Yeah. For those listening, it's the double-long Hynix, triple-long semiconductors, or triple-long Nvidia—those kinds of things. I think part of the reason you're seeing huge volatility is that that's 1 fact pattern: a massive, parabolic increase in leveraged and inverse ETFs.

The second fact pattern is that, depending on the exact day, around 65% of all options in US equities are zero-day, and about 70% of all options are retail. So, 7 times 65 is about 50% of all options: retail dudes doing zero-day options. Many of those are on the triple-levered ETFs.

You want to know why stocks are massively volatile on the prints? A lot of it is the zero-day options, so it's just a different world. Near-term volatility is certainly going to be higher.

Yeah, it's definitely an interesting space, and I think you're right.

More leveraged ETFs have closed this year than existed cumulatively up until 3 years ago.

It’s like 60 of them have already failed this year.

For people watching, you obviously spend a ton of time at risk. Do you think you have a simple explanation of why the daily time decay is so dangerous in them and why they don’t do what people expect?

So, I think the biggest misnomer is that people think if they buy 2x QQQs, for example, they’re going to get twice the Nasdaq 100. They don’t. I don’t know that I have a succinct way to explain that daily time decay to you. I think a lot of it is just the amount of capital that’s available at the time.

I think part of it is what we’re just talking about, where you can’t actually leverage in the time frame required, in the amount that’s required. So if you’re triple-long semis—I think it’s SOXL—it’s been, what, 220% or something. It’s not 3x over the time frame. Or, even worse, I think it has a negative asymmetry, meaning I think it’s closer to triple-levered on the downside but not on the upside.

Yeah, that’s how most of them work.

I think it’s harder to—I don’t have the exact numbers, but I want to say something like 80% of all S&P volume is at the open and at the close. So when you do get moves that are substantial intraday, it’s sometimes hard to true it up on the closing print.

Yeah.

Yeah, my personal opinion is that those are not great investment vehicles, but that probably was obvious when I was articulating it.

Yeah.

But you can trade them and make money. If you do that, then you can. If my dad is listening to this—which he probably isn’t, and probably won’t—he bought Amazon many years ago because he likes to read and he loved the Kindle. I’m not sure—he’s a smart guy, for sure—but I’m not sure I can untether from him the fact that Amazon is up because of the Kindle.

You know what I mean? I don’t think he cares. He’s 20x’d, and he’s like, “I like Amazon and I like to read,” and I’m like, “Yeah, this thing called AWS and whatever doesn’t matter,” right? So sometimes it’s just that simple: you buy SOXL, it went from 7 to 15, you doubled your money, and you’re like, “Listen to these academic jerks who worked at Bernstein tell me I shouldn’t buy triple-levered stuff.”

22. Running Your Own Firm

Sometimes it’s just that simple. People go back to what made the money, and I think that’s popular in the options market, the retail community, the zero-day market, and with crypto and all that stuff. When you’ve had a long run of risk-taking with easy financial conditions, everything seems easy. But if we get a more austere year or 2, some of that will get cleared out, I think.

Final question, and maybe this is because it’s something I experienced, and I want to ask other people if they experienced it. I feel like for every year I worked in someone else’s organization, it took me a period of time to change how I worked. I call it breaking bad habits, but maybe it’s just thinking differently, right?

Did you feel that you went through that when you left working for someone else? And, by the way, we worked at world-class organizations, but did you feel that to run your own business, you had to make some—let’s not make it a pejorative—pretty big changes in your framework of thinking to run your own business versus being really effective working for someone else?

You know, look, I worked at 3 firms: Bernstein, Morgan Stanley, and Eminence Capital. I would say they’re all incredible, well-respected firms. I think my firm doesn’t have a brand that’s well respected, right? When you think about a job, it’s the people you work with and interact with.

One of the best things is that I talk to super-smart people all day long and learn a lot. I think one of the things I’ve learned is that you probably always want to be able to talk to people in their 30s. Having kids in their 20s and talking to people in their 20s, they sometimes don’t have enough experience for me to totally respect where their head’s coming from, but people in their 30s are still excited about growth and they haven’t let technology pass them by. They’re sort of in that exciting part of their life, and you don’t want to be one of those people who only talks to people your age.

What’s awesome about my job is I talk to people of all ages, and I like that. Then there’s the content. I like producing and analyzing stuff on the equity market.

It’s funny: we did our monthly Zoom for institutional investors today, and I always do the Zoom on what I published on last month. The market had all kinds of stuff going on, and people want to talk about the Fed. I don’t think I used the word “Fed” one time in my research in the last month. We wrote about stop-losses, M&A rumors, and executive compensation.

We’re working right now on looking at the adjective-to-noun ratio and the -ing-to--ed ratio from earnings-call transcripts. Are they forward-looking or backward-looking? Just cool stuff you can analyze that I think isn’t just what everyone else is doing. I like the content.

Then it’s the autonomy, right? To do what you want when you want. At Morgan Stanley, it was amazing, but when I told you I traveled 32 weeks, it wasn’t always in my control.

I’ll never forget this: my kids were little, and they played soccer on this rooftop in New York, which is pathetic. They have these rooms that are this size, and your kids play soccer because there’s no space. There’s this dad, a nice guy, and I see him on Saturday morning. Then I leave, and I remember I went to Jakarta from New York. Then I went to London. Then I went to Miami. I landed at midnight, and the next Friday night I went to soccer again.

He asked me what I did this past week, and I almost felt like I had to lie to him, because if you tell a normal person you spent 60 hours on an airplane, they just think you’re crazy. Morgan Stanley was amazing, but I just didn’t have a lot of control. I would end up having 2 or 3 weeks a quarter like the one I described.

Now, when we do something, it’s all for me. The downside is, obviously, I’m the guy who washes the coffee mugs in the office. [Laughter] The person who started the business with me, who’s the head of sales, is one-third the president of the firm and the visionary, one-third salesman, and one-third my personal secretary. We’re all just—it’s a very humbling thing when you build your own business.

I think you just have to deal with the fact that you don’t have the resources that we had. I was thinking about this the other day, when you and I started at Bernstein. I don’t know if you ever took any of those European trips, but we used to have a 7-series BMW waiting for us with glass bottles of water to drive us half a mile between meetings. Now I’m connecting on 3 tubes and spilling water on myself.

All the small-business stuff is true, but we’re 5 and a half years in now. We’ve hit the upward part of the curve, and we feel like the investing and all that is behind us. But I do think in the first couple of years there are some low moments of just eating crow, or whatever the phrase is. You just have to soldier on.

I do think you also become aware, when you run your own business, of who has pricing power over you and who doesn’t. UnitedHealthcare sends me an email: “Your health insurance is up 9%.” There’s no negotiation. Some firms do, and sometimes you can push back and be like, “You know what? No, I don’t want that.”

So I do think you learn running your own business, and you get a feeling for who has pricing power and who doesn’t. I think there’s a lot of power in that, but there’s no 1 ideal thing. Everything has its pros and cons. You’ll see—you’re in that, so you’re living it. You know what I mean?

Yeah. [Snorts] This is the second business I’ve started, so—

Yeah, you know what I’m talking about.

Incredible.

Yeah. Great to see you and chat with you, and I wish you all the luck in the world raising assets for this Founders ETF. It sounds like a great idea.

Awesome. We really appreciate that and just want to remind people for TR vector. Yeah. www.trictorresearch.com. We're on X. We're on LinkedIn. If you happen to go to Triari because you don't remember Trictor for some reason, it'll take you over if you're an individual investor. It's 100 bucks a month or if you decide for the year and you just you get a bunch of content and again I think you can tell you're not going to get the same stuff you get from everywhere else. So we try to our view is like if you get it for free from the big firms then there's no value for me doing it. So we try to have give you some different color. We we'll take questions from people in a monthly zoom. I do weekly videos that are two three minutes explaining what people are asking me and then we write two or three short things a week. So it's it's pretty manageable and it's it's it's been cool. It's been growing and I'm excited about it.

Thanks a lot.

All right. Great. Thanks, Mike. Take care.