Invest Like the Best with Patrick O'Shaughnessy - Jesse Livermore – The Search for the Truth with the Anonymous Master - [Invest Like the Best, EP.136]

Episode Date: June 25, 2019

This week I have a very special guest years in the making. Like another favorite episode, with anonymous guest Modest Proposal, this conversation is with one of the stars of the financial twitter universe who writes anonymously and goes by the pseudonym Jesse Livermore. I met Jesse 6 years ago after reading his unbelievably unique investing research, which tackled all the big and interesting issues in markets. He now also works with me as a research partner at OSAM, where’s he’s used our data to continue to his search for truth in markets. Despite being one of the brightest minds I’ve encountered he is also as humble and unassuming as they come. I’m at least a slightly better person because of trying to emulate how he conducts himself. I get to have many conversations with him that go from 0-100 fast, and I’m thrilled to be able to share one of those with you. For more episodes go to InvestorFieldGuide.com/podcast. Sign up for the book club, where you’ll get a full investor curriculum and then 3-4 suggestions every month at InvestorFieldGuide.com/bookclub. Follow Patrick on Twitter at @patrick_oshag   Show Notes 1:33 - (First Question) – Jesse’s origin story for investing 4:37 – Exploring his ways of problem solving starting with intuitive             7:53 – David Epstein Podcast Episode 11:46 – Looking at the analytical way of problem solving 15:42 – Statistical inference 24:45 – Should we opt for simplicity in the investment process 25:26 – Does his own investing include all three, intuition, analysis, and statistics 26:09 – The evolution of his research, process, and thinking on various investment factors. 31:38 – Thoughts on inflation and its impact on market valuation 40:05 – The Earnings Mirage 46:25 – Free Cash flow and valuations 50:51 – What should investors take away from this research 53:01 – Thoughts on trend as an interesting market signal 59:00 – The problems with trend 1:00:34 – Post on “The Single Greatest Predictor of Future Stock Market Returns” 1:11:15 – His work into understanding factors 1:15:36 – Looking at momentum 1:18:16 – His curiosity into the current market cycle 1:20:04 – Lessons learned from his time in the military, an effective way to create an environment where people can safely disagree with their co-workers 1:30:10 – The concept of progress in meaningful work 1:33:08 – Kindest thing anyone has done for him   Learn More For more episodes go to InvestorFieldGuide.com/podcast.  Sign up for the book club, where you’ll get a full investor curriculum and then 3-4 suggestions every month at InvestorFieldGuide.com/bookclub Follow Patrick on twitter at @patrick_oshag

Hello and welcome, everyone. I'm Patrick O'Shaughnessy, and this is Invest Like the Best. This show is an open-ended exploration of markets, ideas, methods, stories, and of strategies that will help you better invest both your time and your money. You can learn more and stay up to date at investorfieldguide.com. Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shaunacy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaughnessy Asset Management may maintain positions and the securities discussed in this podcast. This week I have a very special guest years in the making. Like another favorite episode with anonymous guest modest proposal, this conversation is with one of the stars of the financial Twitter universe who writes anonymously and goes by the pseudonym Jesse Livermore. I met Jesse six years ago after reading his unbelievably unique investing research. which tackled all the big and interesting issues in markets.

He now also works with me as a research partner at OSAM, where he's used our data to continue his search for truth in markets. Despite being one of the brightest minds I've encountered, he's also as humble and unassuming as they come. I'm at least a slightly better person because of trying to emulate how he conducts himself. I get to have many conversations with him that go from zero to 100 fast, and I'm thrilled to be able to share one of those with you. So, Jesse, this is a long time coming.

people are going to be excited for this one. My favorite of these conversations are often with people that are anonymous on Twitter. And you've got an interesting reason for being anonymous. We won't go into it, obviously. I'll start by saying you're not in the field of investing professionally. And yet many consider you to me, I certainly do, maybe one of the most thoughtful investing thinkers out there. So tell us your sort of investing origin story. How did you get interested in this topic, given that it's not what you do professionally and what has fostered that interest and that love? It all started for me. in probably the late 1990s when I was in college. I dabbled in some of what was going on in the tech bubble, lost some money, made some money. And that started my interest in investing as a hobby. I ended up when I got out of college around 9-11, joining the military. And I spent five years in the military. When I got out of the military, I went to grad school. I was trying to become a physicist. And that was around 2007 to 2009 timeframe. And that really sucked me in in terms of all the things that were happening. And it really became interesting to me, more interesting than what I was doing. And then I started trading on my own very aggressively during that period.

I had a lot of time during the day, a lot of time, a lot of free time. Because in grad school, you can make your own schedule if you get your work done. And then I had some successes, especially when the market turned in 2009. I started getting really, really interested, really enjoying the activity of it, the way that it brings all different disciplines together in a way that physics really doesn't. And so I end up taking a job in STEM just to make more money to be able to trade. That's kind of where I've been since. It all happened very slowly, the blog and everything, because it started out. I was just on Twitter, having debates with people and having fun conversations.

Over time, I started to learn from Twitter, like, learn a lot from really, really smart people. I know that's probably been your experience as well. And I think it was Josh Brown, actually, who had basically told me, he gave me the advice that, you know, there's a lot of really good content on my feed. I should probably make a blog out of it. And I started doing that and started making posts and just grew from there. And then this is around 2012, 2012, 2013 timeframe. And then, you know, I started really getting into it and enjoying it. I've always enjoyed writing. We've talked a lot about how deep you'll go. I know last night at dinner we were talking about you going through this crazy dense accounting textbook lately. And so you'll kind of topic by topic, build up expertise or toolkit in a sub-discipline, let's say, underneath investing. I'm curious how you first came to it. So when you were first trading or investing, what were the instruments that interested you? Was it equities? Was it bonds? Was it options? What was your first step into this whole thing? It was small-cap equities. It was macro in 2007, but in 2009, when I really started to develop kind of a real passion for spending time on it, it was small-cap equities. And just at that time,

I didn't have a really solid experience base to know that the kind of gains that were being generated at that time we're not normal. I was thinking that it was my talent and my abilities. And that obviously is really, it's conducive to building more interest and building more excitement. And in terms of just the thinking style, I think I have a background obviously in physics, but also in philosophy. And that's helped, in terms of breaking problems up and trying to think deeply about things. And that's helped. I mean, you can probably speak to that as well because you're also a philosophy major. But I did a one of my undergraduates was in philosophy. That's actually how we connected. Yeah, exactly. Originally, which was like a mutual love of an obscure philosopher named Schopenh, which is kind of a

wild way to make the connection. But as our origin story, which is kind of fascinating, again, we connected on Twitter. We actually wrote a research paper together on profit margins. Scott, I don't know, a long time ago now. And that was sort of the beginning of this interesting relationship. What I learned during that exercise, the first paper we wrote together, maybe your superpower as I'll call you an analyst, is that you get to the base level of everything. And maybe that's, maybe that's because of the physics, your training in physics or I don't know what, just the way your mind works, which is that more than anyone I know, you leave no stone untowelior. turned in a research project. And I think that's a fascinating way of thinking. And so I'd like to spend

the first part of our conversation. We'll obviously spend a ton of time on interesting investing topics later on. But I'd actually like to form a sort of groundwork of thinking about the way that you apply statistical, analytical, other intuitive thinking toolkits to problems, interesting problems that you're trying to solve. So maybe you could lay that frame. You haven't written about this publicly, but we've talked a lot about it offline, this sort of analytical, statistical, intuitive ways of thinking, what they are, the strength and weaknesses of each, and what kinds of problems those three methods are useful for solving. I think one of the problems that we have, and this is also from reading the blogosphere as well,

is that we have a lot of different tools that we use to kind of make inferences, which I think is what investment is really ultimately about. It's about making good inferences. Forecasting. We have different tools that we use that are kind of innate and built in, and we don't necessarily know how they work or when we're using them that we're using them. So I've been trying in my own thinking to try to take my own thinking and, you know, and kind of put it into different bins in terms of what the basis for the inference is, right?

So, simplistically, I can think of three things, three bins. Intuition, analysis, and statistical inference. So intuition would just be automatic inference. Your mind picks up on patterns in the environment based on, you know, hundreds of millions of years of evolution to be able to do that. And you kind of just have a sense in your gut of what's going to happen, what's true, what's not true, that kind of thing. It's a very rapid form of intuition, kind of like the system one form of inference. And the advantage to intuition is really the cost.

It's very cheap. You don't have to spend a lot of resources to use it. It just kind of presents itself and you have it. Like anything, inference is something that has to be thought of in terms of a cost-benefit analysis. How much validity or truth do you get for the amount that you have to put in? And if it takes a whole lot to put in and you don't get a lot of truth out, that's a bad proposition. If it doesn't take a lot and you do get a lot of truth, then it's a great proposition. And so I think intuition is something that I do use.

All traders and all investors use it to a sort. certain extent. It's kind of the, it's how you start. It's how you kind of, we wouldn't even notice all the ways in which intuition is acting on our thought process. But, so that's the advantage would be the low cost. And the disadvantage, though, and this is something that Daniel Kahneman talks about a lot is, and Gary Klein, is that you need to have, to be able to develop valid intuitions about things. There needs to be, you need to have experience and it needs to be experience in a high validity environment. It's an environment that presents stable, robust patterns consistently. If there's a lot of noise, if there's a lot of fluctuation, a lot of

of meaningless movements. You get into a situation where you end up picking up on false things. Your intuition gets confused and it's hard to develop a sound intuition. You interviewed David Epstein, talked about wicked and kind. That would be a similar type of situation. Like a wicked environment is one where it's hard to develop intuitions, but a kind environment is one where it's easy. And the problem is that investing finance is a really, really messy, really complicated system to analyze, the financial system. And so intuition can be somewhat, it's kind of hamstrung in terms of its ability to be used because it's easy to develop intuitions about what the market's going to do and it's really easy to be wrong about them because

there's so much noise in markets. I thought about ways that we can research to understand what other people have proposed for ways to improve our intuitions. And I think one of the things that I found that was interesting in my thought process on this or my research on this was that there was a study that was in 2011 on cognitive reflection. And what it found was that people that are high in cognitive reflection are better at using their intuitions and they're better at forecasting as a technique, as a tool, and it's just a general kind of pattern that's seen in the data. People that have the ability to reflect make better inferences and make better forecasts. And what's interesting is that another study, similar to this one, found that traders, of all people.

Traders have some of the highest levels of cognitive reflection of all people. And that's kind of consistent, I think, with my experience, and I'm probably with yours as well, that people that are involved in investing that have real money on the line, they're forced to think about their thinking and think about ways in which they can be wrong in ways that a lot of other people don't think about. So I think that's a real plus that I've kind of, for my own thinking, to try to force myself when I have intuition to think about those intuitions and think about the possibility that they might be wrong, to think about where they might have come from. Why do I have this feeling? Is it just because of repetition of something? Is that thing that was repeated to make my intuition be this way? Is that actually going to continue the way I think it is? And just to think that way. Another interesting point on the intuition side is that this study that I was talking about in 2011, it found that people that make very, very quick,

decisions in a forecasting competition that are like very rapid and they don't reflect at all, they do very poorly because they don't have a lot of cognitive reflection. They're not taking that second step in the thought process to maybe check their intuitive processes. But at the same time, people that take very long on the opposite extreme, on the opposite end of the pendulum, that take really long time to deliberate on things, they also have poor performance. And one of the things that I found fascinating, the author of the study had hypothesized that deliberation as a practice, like the process like system two thinking, where you deliberate, where you analyze, where you try to lay out the possibilities rationally, that process is good for fleshing out information. It's good for getting things

on the table that might have been missed, but it's not good at making the actual choice. Like in other words, pulling the trigger. The pulling the trigger process of a decision or of a judgment, it leans a lot on the intuitive, rapid, kind of felt what's the right move here, you know? And people that don't use it that way, end up getting stuck. You can't make a decision. You can't figure it out because rationality itself or analysis is not going to tell you once you've got everything laid out what to do necessarily. It's not going to necessarily make it, it's not going to pull the trigger for you. So that's something that I've actually been using myself in my own thinking as well is this idea of intuition needs to be used in a certain way and you want to have it be, you don't want to have it

be too fast, but you don't want to have it to be too slow to where you overthink everything and you disturb that whole process and you use tools to try to resolve a decision that can't really resolve it. You have to just go with your gut sometimes. So summarize that section, it's an incredibly valuable tool because of its low cost and its speed. But everyone says, you know, they're like magic black boxes, like things that just produce a good outcome. But you can't just, you can't just say, go with the gut. You have to, you have to foster. You have to cultivate it. Yep. To speak to that point is that it's also not transparent. And that's what makes it hard in a conversation and a dialogue for it to be productively used. Like, if you and I are having a discussion about an investment and I try

tell you, hey, you know, it's my gut. It's my gut. You know, well, you can't do anything with that, right? You have no way to use that or to measure that or assess it. So that's a big disadvantage as well. And that's why I think that, you know, there's limits to the extent that it can be used, but it does have a role in the process of trading and investing, in my opinion. So the second two categories are obviously critical for the present and future of investing. So I would say the trend has been statistical dominance in the last, you know, however many years, the rise of quantitative strategies, let's call it. But the analytical is also the original tool for doing investment analysis. So to talk about both of these categories, pick whichever one you want to start.

We'll start with the analysis as a tool. I guess we could define that as just you're using, you're representing a system or a problem, you're applying principles to it. In that application, you're coming to a conclusion. And the advantage to that type of thinking, that's kind of like, that's like theoretical thinking, the type of thinking that a theoretical physicist might engage in. The advantage to that is also cost, because you can, it's not as expensive as data collection, right? If you can do something in theory space, you don't necessarily have to collect data in the way that you would if you were doing something in empirical slash statistical space. But, you know, it's also the other advantage to it is that it can deal with regime change. So like, we might

enter a situation in investing or in trading where there isn't really a history yet. But we may have the tools in terms of principles that exist that we could apply to this case to be able to draw a conclusion. And if we're hamstrung and we're tied up and we need to have data, that's going to limit us. But in that analysis has the potential to be able to get us to an answer without having had the experiment already have played out at least a few times. So that's an advantage that can deal with unique cases, and it can also deal with situations where variables that matter have changed. And that's kind of what physics is, right?

So physics, you come up with some sort of model, and then you can apply that model to any case you want, and it's supposed to be true and it's supposed to work. And then also, finally, it's transparent. Unlike intuition, analysis, I'm fleshing out how I'm getting where I'm going, which makes it much more useful in dialogue and a joint discussion. Disadvantage, obviously, this is a huge disadvantage, and I think it's the one that really, everyone,

a lot of people appreciate this, just intuitive. but it's the fact that it's so easy to be wrong. It's so easy when you represent a system, especially a system that's very complex like a market, a financial market, it's so easy to just miss one thing. And all you have to do is miss one thing. And if you miss that one thing that matters, it can change everything. The fact that it's that easy to miss things, combined with the fact that it's very easy to tell good stories, it's very easy to use other tools like persuasion and rhetoric in the process of an analysis makes analysis dangerous. And I think, you know, I don't have any gripes on it. I think people realize that. People realize that, you know,

talk is cheap. Theory is kind of, yeah, sure, let's see if it works. It's very easy to tell stories. It's very easy to narrate. But oftentimes, theory just ends up being wrong. In terms of ways to improve theory, I think, and this is something that engineers use, this comes from the STEM world, is that if you're going to use analysis, you want to be cognizant of the fact that it's easy to make mistakes and to miss things in the representation and the application of principles. It's easy for things to be wrong, just one thing, and that's all it takes. In that process, you therefore need to build in margins for error. That's kind of the key thing, is to build in margins for error so that if you are wrong, it's still okay.

Your end conclusion will still probably be right. You still have room for there to be small errors or hopefully large ones. To speak more on that, it's also important to think openly and honestly about the analysis and to stress test it and to try to find the weaknesses in it. Try to find the weak link seriously because a lot of times, you know, if we come into a situation and we want to come to a certain conclusion, analysis might take us there very easily. But we got to step back and say, okay, what could we be missing here? What could be wrong here? What are the things here that we are not thinking about right now that are going to drive this to be incorrect?

And we need to get our minds on that now and think about that in a very critical way. And then finally, obviously, you don't want to just use one tool. So a great way to validate analysis is to actually test it. So if you test it, the beauty of testing is because testing, let's say you have a whole bunch of different links in an analytical chain. If any one of those links is wrong, you're at risk of being wrong in the overall conclusion. But testing can test all of them at once because it just does the experiment. True or not, yeah. True or not.

If it works, and then if it doesn't work, you'll know that something in that chain was somewhere mistaken. So to move on to statistical inference, I mean, we'll just transition right there. I mean, obviously, its advantage is that it's able to test things in a reliable way to get a lot of coverage at a low cost.

It's able to, in situations where there's a high probability of making an analytic mistake, it can tell you that you've made one. And it can also, because it's rooted in the actual result, it's closer to reality than these other approaches, which are, you know, they're in their own worlds. But empirical statistical approaches are literally, they're touching reality, they're touching the result, they're kind of responding to what they're seeing. So what I've thought about, and this is something that I spent the most of my time in the last, when I was thinking about this question, is

what's the basis for a statistical inference? What is the actual justification for extrapolating out, for example, a pattern? So, you know, a good way to maybe explain it would be just with an example. Like, let's suppose that right now I pull out a coin and I flip. it in front of you and in front of the audience 10 times. And every single time it comes back heads in that flip. So that's like a one in a thousand chance. And the question posed to you or to the audience would be, okay, well, I'm going to flip it again. What's the likelihood that it will turn out on the 11th flip to be heads? Or which way would you bet? That would be the question. And it would be a tough call there. I think most of us, you know, in this situation, we might

actually bet heads. Why not? Right? I mean, it's possible that this is an unfair coin. It's possible that the coin is biased. The very fact that I brought it up, suggest that maybe I have some tricks up my sleeve. We pick up on all that. But ultimately, I think the basis for extrapolating that out, and that's what we do in the statistical analysis, it's a Bayesian base. It uses Bay's rule. So it's kind of like, in your mind, you're thinking about, let's just isolate the possibility space. It could either be the case that the coin is a fair normal coin, which is how it looks to you. And if that's the case, then what just happened is that I just got really, really lucky. I had a one in a thousand event just happened,

let's say roughly. That's one possibility. The other possibility is that I've got some trick off my sleeve, and this is not a fair coin, and this was all pre-stage somehow, and it's biased. So what we're doing in statistical analysis is we're looking at those two possibilities, and we're saying, look, it's so unlikely that this would have come about in the case of an unfair coin. There has to be some bias, some signal, some causal feature that is making this pattern occur. It could not have, it's so unlikely. So you're kind of picking, I don't want to anger the mathematicians here and, you know,

a misframe-based theorem, but you're kind of picking, you know, you're picking your poison. here. And it's clear that the worst poison here in terms of what's likely to be true is that this happened randomly. If it was maybe 10, it would be hard. But let's say it's a 10,000. And I do it 10,000 times. And it comes back heads every time. You know that there's just no way that could have happened on a fair coin. It's just that's so ridiculously impossible. There's not enough atoms in the universe to kind of convey how unlikely that would be. And so you would have to say that you did something. You cheated in some way or to translate that into it into investing. There has to be some causal signal that's going to make this recur. And I think the key thing. So everyone, we all

get that intuitively at least. We kind of understand that that's the basis, but the problem is is that there's ways to undermine that basis that we're not necessarily perceptive of. And I know that I wasn't, through a lot of my career in terms of doing research, is that you can subvert the Bayesian basis for statistical inference in various ways. Let me give you an example. So in this case, I said that I flipped the coin 10 times. And that's pretty clear. You know that the trial size here was 10. But it's possible for me to misstate what the true sample size is because there could be correlations between the flips. That's a bad example to use because it's impossible. for their to be, yeah, they're obviously independent. But you could have some other pattern where they're

not independent. And I could have a huge string of things that happened that all happened because of one thing that happened. And that one thing would be the true sample size. It would be one. But you're thinking it's 10,000. We've been doing this in investing. That's a great example. Exactly. I mean, and it happens all the time. It's just, and so it's really important, I think, in statistical analysis to have a correct assessment of what the sample size. In fact, I even like to, I like to call it the trial size. Instead of calling it the sample size, it's the trial size. How many unique challenges to the hypothesis or to the theory have you applied in the experiment? Ways for it to come and come back at you as being, and give you a result of being wrong or whatnot.

Just to think about that, another example of ways that you can subvert the process is that this all relies on the idea that you're going to call your shot. In this example where I'm flipping these coins, if I pull the coin out like right now in front of you and I say, I'm going to flip this thing 10 times and it's going to come back heads every time. And I say that before I do it. And that happens. Okay, that's going to count for a lot because I had to call my shot. I didn't have the luxury of any kind of gamesmanship in this. I actually put it on the line in front of you before the actual result. And I called it.

But it's possible, we can't do that in investing, right? It's hard, like, a lot of the times we have to go backwards. We have to look at past history and use the past history to make inferences. We can't just go off of the patterns that are emerging that someone has called immediately. It takes too long to run experiments that way. we might have a theory about how markets are going to work over the next 10 years. It would take 10 years to do the experiment to determine if I'm right or wrong. And that's too long.

And so we have to go back in the past and look at 10-year periods and say, okay, what happened in history? And the problem is that when you do that, and I'm not saying that it invalidates the logic, but there's a subversion of the basis there because there you can do certain things. Like, for example, you can engage in what's called multiple comparisons. So like I can, an egregious example of multiple comparisons would be, and this would be actual outright fraud, But if I were to just flip a coin 10 times, I get a random expected result for the coin being fair. And then I don't like that result. So I throw that out.

And then I flip it again 10 times. And then I keep doing that until I get some statistically significant outcome. And then I take that outcome and I say, look, here it is, this proof, you know. And that's kind of odd. Everybody gets that that's, I'm not even, I'm limiting the data that I'm presenting. And I'm hiding the true universe of data that I generated. Obviously, that's fraud. But there's ways that you can actually end up doing that and not knowing it.

So let me give another ridiculous example. Let's suppose I have a theory that the color of a person's shirt affects the likelihood of flipping heads, which is just obviously wrong, right? But let's just assume that that's my theory and we're going to test it out. So I've got like a purple shirt on right now and I say, okay, I'm going to flip the coin 10 times to test my theory. That purple is a causal color for coin flips. And it comes back random, comes back 50, 50, whatever. There's no signal.

So I say, okay, well, maybe it's not purple. Maybe it's orange. So I try orange. And I change the hypothesis, do it again. Now I try orange. And again, it comes back random as expected. for a fair coin, maybe half and half, 50-50. So I say, well, let's try, you know, green.

And then eventually, you can see where this is going. Eventually, by just pure random chance, I'm going to end up randomly stumbling upon to a pattern that is clearly the result of a random process. And I'm going to think, well, gee, I just established, through a statistical justification, I just established that there's a causal relationship between a green shirt and 10 straight head flips.

And the point is that I'm doing that in a way that's much more subtle. I'm not excluding data. I genuinely believe that the color of the shirt matters, and I'm basically segregating off different colors, and I'm testing all of them. And in the process, I'm stumbling upon one thing that had this pattern. It's obviously spurious, but I'm going to conclude that it's causal. And so that happens in investing because we come up with a theory.

Maybe it works if you trade on Thursday afternoons and you sell on, you know, or you buy on Thursday afternoon, you sell on Monday morning or whatever. You know, you can come up with an infinite space of things that you can test. If you keep testing, you're going to eventually stumble upon the patterns just because patterns emerge randomly. And that's such a problem. So what do you do about that? Well, you have to try to find ways to constrain the process so that if that's happening, you're going to catch it.

And one way to do that is like, you know, as the machine learning people will tell you, is to try to use the holdout method, to test stuff out a sample, to like to call your shot, to figure out what the theory is and then test it, not use the testing process to develop a theory or to arrive at whatever the claim is that you're trying to arrive at. The subversion aspect is also a problem with respect to the concept of search. So we have computers now. If it was 50 years ago, it might be harder to play this game. But the fact that we have computers means that we can do searches for patterns much more efficiently and effectively and quickly,

which makes it much easier to unknowingly and unwittingly engage in multiple comparisons as a phenomenon and end up subverting the basis. So I think, you know, what does that mean in terms of how to invest? Well, it means that you got to call your shot and you have to find ways to incorporate the idea of a called shot into the investing and the testing process. So it means that when we arrive at a theory or we arrive at a conclusion or some belief, based on evidence that we see in the data, we tested out a sample. We tested in different markets. We tested in different environments. To be fair, that creates its own complications because sometimes when you go into a different

environment, you're actually changing variables that matter. So it really helps if you have a theory. It really helps if you have a way to know that the thing that you're tweaking doesn't matter and shouldn't matter. If you don't know that, then you never know whether or not the failure in the out of sample is due to the fact that differences between the in-sample and the out-of-sampal were such that they were capable of causing a change in the pattern. So if you can tweak things, if you can tweak variables that you know don't matter, that's always great. If you can change, like, for example, on a factor test,

if you can change the day of the week or the month that you're doing the rebalancing, if you can change something you know should not matter. And if it still survives, that's going to increase the probability that's true. But if it fails, then you know you have a problem, and you know you're probably playing with noise. Do you think that one of the primary conclusions of kind of summarizing everything you've said so far is to opt for simplicity in the investment process? I want to caveat that there's advantages to simplicity at all levels.

So a simple analysis is better than a complex one because there's fewer ways to be wrong. There's fewer things that you have to check. And then in statistical inference, simplicity, I mean, I don't know if I can tie that as clearly to statistical inference. I think that complexity creates room for gamesmanship. It increases the space available

for you to subvert the various basis for doing things, right? I mean, as a general rule, complexity is dangerous because it makes it easier to get around the difficulty of producing a true conclusion in testing. When you comes to your own investing, so, you know, like you said, call your shot, would you say that once you've arrived at a decision, it's usually a combination of all of these tools in the toolkit? Always. It's always a combination. Now, there have been cases where I can't, because I do discretionary for fun. I don't do it because I think it necessarily works. I do it because I just enjoy it. It's a hobby to just pick my own stuff and play around with it and take positions. And sometimes there just isn't a statistical, there isn't enough good data to be able to do a solid statistical analysis. And all I can go off of is just you read the financial statements. You kind of do some deep diving and digging.

So that becomes mostly analytic. But obviously, even in a situation like that, you can kind of use the empirical feedback of the investment itself as a way to check yourself. Like, for example, you can take the position. And if it starts to go against you, that is probably, that's information that maybe you were wrong and you need to check things again. So I do use all of them when I can, although it can be very hard to use statistical inference in certain areas of the market where you don't have a case history that's sufficient or you don't have good enough data. It's a critical thing is you need to have good data. And before, you know, I had access to some of the stuff that I have working with you guys, I didn't have anything but the publicly available stuff. And there it's even harder to try to draw conclusions that are strong and valid.

So I want to set aside a couple of the other thinking toolkit items that we had just because this is a great spot to move into some of the interesting investment. topics that you've explored. I would say you've kind of tackled most of the biggest questions that people have, things like active versus passive, trend, valuation, et cetera. And so I'd love to start at the highest level, which is sort of at the market level. A lot of this work has been in equities, but we can go into other asset classes if you like as well, and get your take on your own research and the process, the evolution of your thinking on things like valuation, like trend, like the impact of macroeconomic variables on these things like valuation. You pick an interesting starting point you mentioned before maybe talking about this idea of markets as banks.

Yeah. And that might be interesting framing. In the mid-2000s, I kind of bought into a lot of the valuation theory, Cape ratio. I was really an officinado of the Cape ratio. And that was beneficial because it helped me to be out of the market when the market ended up having a big mess-up. And I attributed that mess-up to the valuation. And one of the ways in which my thinking has evolved subsequently is to have realized that actually that's not the reason why the market. did what it did in 2007, 2008. The valuation in 2007, it wasn't necessarily cheap, but it wasn't

egregiously high. It wasn't even that high. I mean, we'll say more about the fact that there's reasons to think that maybe valuations are going to be permanently higher. It's dangerous to say that. But I attributed the rightness of my positioning to the rightness of that belief, and they weren't necessarily related. And it took, it took several years to kind of undo that effect, right? I had to go through 2010, 2011, 2012, and realize that we're going right back up to the highs here. This is not, you know, we haven't popped a bubble. In fact, we're going to make a new one. We're going to make new highs and we're going to surprise everyone. So anyways, asset markets as banks, I think that it's almost immoral to suggest that history is not going to repeat to say

that things are different this time. It's taken to be a blasphemy, or at least it was 10 years ago or seven years ago. But I think that if you think about the progress of an economy, the evolution of an economy of a financial marketplace, I think it actually makes sense for things. For things, things to become more expensive. And let me give the, the asset market says bank's concept is just a way of kind of explaining that. Reason why, like, if it's the year 1870 and you're putting your money in a bank, the money is not in the, I mean, obviously they're taking the money. I don't want to anger the MMT crowd here, but in terms of the theory of money, but let's just simply say that they're taking the money and they're lending it out at some rate. So the money's not there. And somebody

else has got it. If I were to tell you, why are you lending your money to you're crazy? You're nuts. There's nothing there. It's like the money's, that money could be lost. I mean, why are you doing that? Why are you accepting such a low interest rate for the risk that you're taking on? Your answer would be, well, I know that people aren't going to panic. I know that everyone else is going to be there to keep this system intact. And in the year 1870, that might not have been a valid argument because of the way the banking system was structured at the time. You didn't have the Federal Reserve. You didn't have some of these systems to try to keep things stable. But I think valuation is a function not only of the cash flows that you demand or the returns that you

demand. It's also a function of the liquidity that you have. And the liquidity is preserved. through structures, through confidence, through structures that create confidence, that create networks of confidence where everyone believes that everyone else is going to be there to take out their positions if they want to sell. And I think that over time, as an economy evolves and we learn things,

we should be able to build those structures more effectively so that you can be more confident that we're not going to have calamities. And if that's the case, then you should be willing to accept a lower return. And I think that if you look at the evidence, if you look at the way the Fed responds to the market these days, much more so than maybe 70 years ago,

if there's a problem, the Fed steps in. And that's scandalous to talk about. A lot of people don't like that. But I think that is a reason for people to be more confident in the system because we understand how it works. We understand the things that we've done in the past that were wrong, and we're not going to do them again. At least the risk of us doing them again isn't as large.

And to me, that's a reason to believe that maybe the equity risk premium, the premium that you get over bonds or cash, shouldn't be as large as it was in those prior periods. To speak more to that point, I think one of the key things that actually is cyclically, to talk about views about the actual economy, I think that our cycles are going to be slower. We're going to have lower growth. And a lot of people might think that's going to be bad. But I think that actually is good for equities because what that does is it reduces the boom bus cycle. And in doing so, it kind of reduces the liquidity risk. It reduces the risks of panics because you don't have the same highs and lows. I don't want to say it reduces volatility, but in that direction type of thing. I think another thing on that same point is just, I think that inflation is going to be low forever. I'll come out and say it. I think that inflation is going to be low forever. I think that interest rates are going to be low forever. I won't even stop and say like, oh, we're going to do 50 years of a reset. Never. No, forever.

That's my view. And I think if that's the case, it gives the Fed a lot more room to be supportive. The one thing that would stop the Fed from being able to support markets to be able to keep conditions right for people to be bullish would be if you had a legitimate inflation problem. Then they wouldn't have their hands would be tied. But I don't think that's ever going to happen. Can you say as much more about that as you can? Inflation specifically, this is something you and I haven't discussed in a lot of detail. So what's the basis of that belief? Well, I don't understand inflation. I don't think anyone does. I think it would be foolish to pretend like you do. I think part of my conclusion there is just having believed that inflation was going to come back for 10 years. I'm at the

point now where I'm ready to give up. But also I think that there's something demographic, there's something in terms of the structure of a mature economy, in terms of the way that we become anchored to prices. I have one example to use, and that's Japan. And I think that's the best example that we have, and I think that there's no reason for us to think that things are that much different. First, it was Japan. Now it's Japan in Europe. Why not the U.S. as well? Demographically, as we become older, maybe we invest less. Maybe we, you know, maybe you don't have the same boom, bus cycles. Maybe you have more timid behavior by consumer. I don't know what it is. I don't understand. But I think that from what I can see, just I think it's valid at this point to extrapolate out the

trend that we've had for a pretty long time because there's been no reason, it's given us no reason in the data to think that inflation is going to ever come back. You've written quite a bit about, and we've talked a lot about recently the effect of inflation on market valuation historically. Maybe you can talk about that. You mentioned CAPE ratio as well. So why that concept of market valuation was appealing in the first place, maybe what flaws you uncovered in it when you really did a deep dive and how inflation specifically impacts that historical data set and how we might want to be careful about it.

Sure. Yeah. I mean, that's a good one. The CAPE ratio, it's not as much of a problem now, but it was a problem, let's say, like in 2011, 2012, is there's been some accounting changes in terms of the way that. that depreciation is accounted for or asset writedowns are accounted for under FAS rules. And I don't think people realize the significance of those changes back then, but they were really affecting the CAPE ratio significantly. You had lots of write downs in 2001, 2003 time frame,

and also in 2009 to 2010 timeframe that were depressing the CAPE ratio. And, I mean, as a general rule, I'll say this. We'll probably get to this more later. But I think that there's pretty good evidence that accounting as a practice in the past was much less rigorous, especially around the idea of depreciation was much less rigorous than it is today. And that matters to valuation questions, particularly earnings-based questions on valuation, because what you're doing when you're engaging in the valuation process is you're trying to compare the current market to prior markets. You know what those markets delivered in terms of returns. And so you want to compare the current valuation to those valuations to see what

kinds of returns you're going to get now. And I think it's just a, it's a very difficult comparison to make because of, or at least it was in 2011, 2012, because the CAPE ratio had been depressed by those write downs. There's also the fact that kind of dividend payout ratios have changed. This isn't a huge effect, but it's also a small effect. So those are two things that I've read about about those a long time ago. Everyone knows about that stuff. That's kind of well-known. Jeremy Siegel has also written about that very extensively and makes a really good case for that being an important difference. But I think one of the most important expenses in calculating earnings is depreciation. It's the cost of maintaining the current value of assets in the presence of

wear and tear, decay, but not just those two things. Also, obsolescence, which we'll talk about probably later, it's an important concept in all that. But the problem with the way that depreciation is calculated in U.S. generally accepted accounting principles, and also pretty much around the world is that it's calculated against the historical cost of assets. So it's like, let me, let me distinguish between historical costs and replacement cost, right? So if I bought a house in Newport Beach, in the year 1955, and I paid $10,000 for it. On a historical cost framework, I would carry that cost on my personal balance sheet. I would carry that asset at the cost that I paid for it, $10,000, net of depreciation, but let's assume that we're not talking about, we're not depreciating it yet.

It's land, let's say, so it doesn't depreciate. I would carry it at $10,000. But the true value of it today is going to be in the millions. And so a replacement cost, A historical cost framework would carry it at the cost that was paid for it. A replacement cost framework would carry it at the cost of replacing it right now, which is kind of like the market value. You can use that as a proxy for the replacement cost. It's a pretty good proxy. A lot of people don't realize this, but in U.S. accounting, assets are not carried at their current values.

They're carried at the values that the companies initially paid for them. And so because assets are carried at those values, depreciation is calculated against those historical values. It's not calculated against the current value of an asset. It's calculated against the purchase price however many years ago. So let me give an example. If I build a factory and it's the year 1960, and I build a factory and I pay, let's say I pay $100 million for it, and I depreciate it over, let's say, a useful life of, let's say, 10 years, or 20 years, whatever,

I'm going to basically deduct the cost of that capital expenditure each year in equal bunches if I use a straight line method. And I'm going to deduct that cost from earnings as an expense. I'm not going to deduct it all at once in one year. I'm going to deduct it over the useful life of the factory. The problem is that that expenditure, or that kind of that non-cash cost is meant to capture the cost of maintaining the physical condition of the factory and also maintaining its competitiveness over time, right? So I may have to make capital outlays into it to renew some of its equipment, to improve things that are breaking, to keep it on pace with everything else in the economy. And the problem is in the accounting framework that's used in U.S. Gap, I'm going to use the historical cost. So in 1960, I'm going to be using the price that I paid, $100 million.

But inflation is going to be making that cost, that the true value of that asset is going to be rising. And the true cost of maintaining it or of preserving it in the face of obsolescence, wear, and tear, et cetera, that's going to be increasing. And it's not going to be reflected in the depreciation charges that I'm taking because the depreciation charge is going to be locked in reference to the historical cost. And so in periods of really, really high inflation, you get a situation where the inflation causes the historical. asset values, the values at which the assets on the company's balance sheet are being carried, it causes those values to become very understated in current price terms. And therefore, the depreciation becomes very understated. And therefore, the earnings become overstated. And so, like, you think about the year 1981. You look at the Cape ratio. The Cape ratio is like, you know, six times,

seven times, right? Well, part of that's obviously because of where interest rates were, part of where the economy was, and whatever else. But I think a significant part of it is because the, The market recognized that earnings back then were overstated relative to the norm. They were not what they appeared to be because the depreciation expenses were being understated because you had had a long period of inflation that was causing the historical costs of assets on the balance sheets of companies to become understated over time. So I think that's an important detail in all of this that gets missed. I know that a lot of people realize that there's a relationship between inflation and valuations,

P.E. ratios. And there's various explanations that are flogations. around for that. Like maybe, maybe high inflation leads to high interest rates, which then causes PE ratios to fall. That's plausible. Fine. I think the more important variable there would be the real interest rate because just like inflation causes high interest rates, it also causes high nominal growth. It causes high sales growth, non-inflation adjusted growth, which should, therefore, that should be buoyant to valuations. But that's plausible, right? You can use that as an explanation. Maybe inflation also causes uncertainty in terms of what the, what policy is going to be,

and maybe the markets don't like that. That's plausible. But I think, a really important factor in the relationship between inflation and PE multiples that is very rarely discussed is this idea of the way that inflation in a historical cost framework causes asset values and therefore depreciation to be understated and therefore causing earnings to be overstated. I think the market gets that. The market is very, very wise. The market knows when earnings aren't what they appear to be. And the market therefore downrates them in terms of the multiple that it applies. And you see that. If you look in the, you know, if you look in the the periods of history where P.E. multiples have been the lowest. They've always been periods

of high inflation, either that or periods of mass crisis, like the Depression or 09, right? You look at the 70s, early 80s, very low P.E multiples. You look at the late 1940s after the World War II experience where there was a lot of inflation, very low P.E. multiples. You look at the late, the early 1920s after World War I period. That's actually a dual period of inflation and deflation, but there's a spot in there where the trailing history was very, very high with inflation. And there you had a very low multiple. So the market seems to get that. And I think that's important when we talk about valuations today. It's really important to remember that, hey, we have a very low inflation rate. To use an average that incorporates periods that had really high inflation and apply that average to now as

if that's that averages the norm, I think that's a mistake. And I think you can show that via this argument. So I think this is a perfect time to talk about how you came to investigate some of this stuff in such detail more recently, which is the most recent piece that you've written. We don't have, I guess, a final title. Yeah, we'll call it the earnings mirage for now. I think that's what we settled on. But discuss why you started this investigation. And then there's a lot of things that you've mentioned there that I want to pull more on, things like the importance of free cash flow, things like accounting standards, almost the philosophy of depreciation, which I know you've started to think about beyond just the financial context, the accounting context. So describe how

I'm always interested in when you've gotten deep into something where it began. So what was the original question? Well, the original question is, so I've actually, had this thing where as a personal tool, like if I'm bearish, part of me will be bearish on the market, but part of me will know that I shouldn't be and will know that I need to be getting in. And so there have been a couple times in this cycle where I had been kind of maybe like in cash at the right time and wanting to get back in but not having the ability to do so because it's hard to make switches and it's hard to get yourself to acquiesce and chase something that's risen a little bit. And what I use this tool where I was like, okay, instead of thinking, because my mind is

anchor to whatever the right price level is, right? Like maybe let's say, let's say that it's 2013 and I had been out and the S&P was at 1350 and then it goes up to 1,500. It gets much harder to invest at that point. It just feels bad. If you feel like you're doing something immoral and you're dangerous and you're giving into your temptations and so you come up with excuses. At that time, I thought to myself, okay, well, I actually had been positioned correctly for this one. I'm using it as an example. I thought to myself, you know, the company in the meantime, I shouldn't think about the level of 1350 or some level as kind of an absolute value point because the point is the companies are making money over time. So if 1300 two years ago, or let's say in this case,

it would be like, let's say, 1900 and four years ago as a price is not the same as 1900 today. So I shouldn't be holding out for a price like that because the company has been accumulating earnings. So that's where this all started is I thought about this idea of retained earnings and trying to use that as a way to value the market, as a way to create something like the Cape, but with retained earnings. So what I did there was I basically said, okay, I built this little technique for trying to calculate the book value of the S&P 500 just using retained earnings, just using that. And the accumulation of retained earnings over time and kind of using that as an anchor point for where the market should be. So if the market should have been at, let's say,

1900 three years ago, and if the market has retained $180 since then, I'm talking about the S&P, if it has retained $180, then it should be at $2080 today. That kind of logic, right? So I built a kind of evaluation system based on that. And then I thought, you know, you had actually brought up this point. Why don't you look at return on equity as a place, you know, because you can actually use the same, the book value that you calculate via this method of accumulating retained earnings over time.

You can also use that as for other purposes, like maybe measuring returns on equity. And when we did that, we found that the return of equity was incredibly low. if you calculate the book value in this way and you compare the earnings to the book value over time, you end up getting an average of like 4% for the return on equity. To just step back and clarify

what this is all saying, it's like every year the corporate sector is earning some amount of money and they're taking that money and they're even doing one or two things with it. They're either paying it to you in cash or they are somehow reinvesting into the business. And the proposition is that

if it comes to you, then it's real. You got it, it was your dividend and it's cash in your pocket, right? So you're not you're happy. If they're taking the money and they're putting it back in the business, we've got to keep track of that. That's supposed to be creating value and we want to know that it is. So what we're doing in this whole model is we take the amount of earnings that are being retained, just that amount, and we add that up over time. And we compare that to the amount of earnings that are subsequently produced.

And when you do that, you find that the amount of earnings that are subsequently produced is very, very low relative to the amount that is actually being invested through the retained earnings process. To be clear, each chunk of retained earnings is itself adjusted for inflation. Yes. Well, in this process, we adjust it for inflation because that's how you can get around the historical cost problem, right? Is you have to adjust it for inflation. And you adjust the earnings for inflation as well. So you adjust all the terms for inflation. If you do that, you find that the actual returns that are being generated on this investment process that the corporate sector is allegedly engaging in is extremely low, much lower than expected. And so the question is why that, why is that? And I think that the best explanation for that is that the earnings are overstated because of inflation. Maybe because of other reasons as well, because depreciation is just hard to kind of capture.

But we know right now that earnings are systematically overstated in an inflationary environment. We always have inflation. So we should therefore expect there to always be some amount of overstatement. And here's why. If earnings are overstated, then the amount of money that the corporate sector is actually reinvesting, that's less than they're telling you. So the actual return, if you compare it to that smaller number, which is the true amount of earnings that they're generating and retaining, if you compare earnings to that number, it's going to be much higher, the ROE. But if you use this inflated number that they're giving you using the accounting method, it's going to end up being, it's going to make the return look very low. That kind of has been an eye-opening experience.

And I want to be clear that this theory around overstated earnings is not something that's new. It's been known for a long time. And there's actually an author named Andrew Smithers, a British economist, who's a real strong proponent of. of this idea that earnings are overstated. If you go back to the 1940s or in the 1970s, there is actual discussion of the problem that inflation is creating in terms of earnings overstatements. So like it's funny that like in the late 1940s, I think this is the first experience that the U.S. had with this. DuPont, Sears Roebuck, and U.S. Steel. They all tried to use a different accounting method, a replacement cost accounting method, to get their

depreciation expense to rise and to get their earnings to fall. which is so counterintuitive today. And they were doing that because this is post-SEC Act, right? So the unions all had access. The unions all had access to their earnings reports. And they knew that their earnings were overstated because they had this huge inflationary environment. They wanted to get their earnings down to reality so that they could have a better negotiation with the unions. And the SEC said, no, that's not allowed because the SEC has always been very, very averse to anything but a historical cost approach.

I want to clarify some terms here. So the method that you developed, again, basically integrating in each year's retained earnings, everything adjusted for inflation for, we'll call it like an improved book value. So this method is called integrated equity. That's the name you've given it in the paper. So then now what you've got is a really interesting data set against which to measure other stuff. Return on integrated equity as one example. Price to integrated equity as a second example. So let's talk about that valuation piece.

So what I think is fascinating is, and here's where I want to. ask about free cash flow. So when people read this paper, I think what you'll find is just this incredible dive into the history of U.S. profitability and a lot of questions you should be asking of data that you're used to taking for granted. So when you're used to hearing an ROE of 10 or 12 percent or a price to book of, you know, now two or whatever it is, I think you'll look at those very, what you would think of as obvious and simple statistics in a very different light after reading this paper. So let's talk about how the market is, in fact, quite smart. And what the levels of something like what we call pie, priced integrated equity, as a really clean measure

of valuation, might tell us about how markets discount this impact of inflation and why free cash flow for analysts out there and investors today is so important. We talked about inflation as a driver of understated depreciation, but it's important to realize that that's not the only way that depreciation can be misstated. There's no way to empirically measure in depreciation. There's no way to like, it's very difficult to like make that empirical and statistical. So you have to kind of use accounting models. You have to use the straight line method. You have to take the asset and just depreciate it over some useful life. And the question then becomes, what's the useful life? Well, you have to just use thumb rules. And that's just very unreliable. And I think a lot of times people

think that maybe depreciation is overstated because they hear these very low useful lives for ask. Like maybe I say the useful life of this factory is 10 years or 15 years. And they say, well, that's too low. This thing's going to last for much longer. But I think the important point there to realize is that depreciation is not just about accounting for the cost of maintaining the physical condition of an asset. It's also maintaining the competitiveness of the asset in the market. Its ability to generate profits relative to all the other things and other ways of producing that output. So let me give an example. It's the year 1995 and I go out and I buy $100 million with a Kodak cameras. If I look at the depreciation of those Kodak cameras as being purely

a physical process that I have to account for, like the physical decay, like the camera breaking or not working anymore or kind of over time just losing its quality. I might have a useful life of 30 years or 35 years. Who knows? But we all know that the economy progresses and we develop a new way of doing camera photos. The depreciation could happen like in two or three years to zero. It could be economically non-viable. And so depreciation has to account not only for the physical decay of assets, but also the competitive decay of assets as technology improves. The very process of thinking about all this, it makes you very, very nervous about earnings as a measure because you realize that like this huge chunk of earnings depreciation, it's such a guess.

It's such a difficult thing to put your mind around and to get accurate on, to accurately quantify. So I think, you know, one thing that we tested was, okay, well, if you use free cash flow instead of earnings, do you get a better, let's say, value signal? And the answer is absolutely you do. In a lot of different contexts, free cash flow works much better. And in that, with free cash flow, the beauty is you're letting the actual capital expenditures themselves be the depreciation. You're using the actual flows themselves. You're not applying some theoretical thumb rule and hoping that it hits the mark. You know, you're just, you're using the actual flows. And what's really fascinating as well is that if you, to come back to valuation,

if you look at the free cash flow in the 1970s and you compare that to the free cash flow today, we had said that maybe inflation was causing depreciation to be understated and, it causing earnings to be overstated, that is perfectly borne out in the free cash flow data, because if you look in the periods of high inflation, the free cash flow plunges. Because the point is, is the true cost of maintaining the assets, right, is not captured in the earnings accounting, and therefore the earnings are being overstated relative to the true distributable cash flows that the companies are generating. And so that's another data point.

It's in the piece. You can look at the charts kind of proves that really there is a problem here. There is a problem that has to be addressed in the investment process. You and I were going back and forth on this kind of done so endlessly, right? it's been this well that never seems to end, the butt we never find the bottom. You've mentioned to me a few times that this is sort of the deepest dive you've ever done on any one topic, probably six months, I guess now into this. Having done such an insane deep dive, what are in your mind the primary takeaways from this bit of research that you think are most useful or interesting for

investors? Well, obviously the depreciation point, the free cash flow point, that free cash flow is an important check on earnings, whether you're discretionary investor or a quant investor, you need to be using free cash flow in your models because it's one of the only ways of accurately capturing or accurately preparing or mitigating the possibility of misstated earnings. Additionally, I think in terms of valuation, the ultimate conclusion here is not that different from the CAPE. So I think if you use the integrated equity method to value the market, in other words, if you sum up the retained earnings and you compare the price to that sum over time, you get a signal that's very, very close to the Cape. And just as an interesting point, not to put too much weight on this, which I know we want to be

careful of, but Pi, the ratio, is more statistically predictive of forward returns than any other measure of valuation that we've ever looked at. So again, take that with a grain of salt, do with it what you will. But just to put like a proof point around why this integrated equity concept is so interesting, it itself is the most price versus that is more predictive than anything else that people talk about, including Cape, including earnings, including Tobin's all the things you could throw at it, it's way more accurate. It is. It's a worthwhile point.

Yeah, and I mean, it gets closer, I think. It's probably a little bit closer to truth. But the truth ends up being the same truth that we've been getting from these other measures, which is that, look, markets are overvalued relative to the past, right? And there's really no way around that. I mean, that's something that's confirmed by all angles. My take on all that is that, yes, and that's the way things should be. The market should be more expensive than it was in the past, and it's going to stay that way,

whether you like it or not. And so I would offer that as another way of modifying expectations. I think we should expect lower future returns. I think we should expect valuations to stay high. I think we should expect opportunities to buy cheap markets to be constricted going forward. I think that's a great conclusion to this thinking on market valuation in general. The historical problems with it. The current opportunity set may be prospects for the future.

I'd love to turn now to another huge investing category, which is trend. What I've always found very interesting about really, I guess, any price-based signal, whether that's cross-sectional momentum or market trend or whatever, is that you sort of don't have to know anything else. It's very pure and it's very adaptable. And so I'd love to hear your thoughts, having done a ton of exploring on using trend signals. We'll hold momentum off to when we talk about factors, but trend at the market level and trend alongside other maybe interesting indicators of underlying economic trends as a useful tool for investors. So maybe if the conclusion for valuation is, what it's telling us is it's a blunt instrument for market timing. It certainly has basically

no relationship with forward returns less than maybe 10 years. So it's a strategic tool, not a tactical one. And maybe it's predicting lower than average returns, but we should accept that. I want to use trend and talk about trend in sort of the same way. So start by summarizing kind of how you feel today about that as a useful signal, maybe in contrast to valuation and what you think the most interesting aspects of trend research are. The way that I got introduced to trend was obviously like a lot of meb Faber, obviously his paper. And I also read a lot of Corey Hofstein and Adam Butler, those guys. You know, I would call myself maybe an agnostic on trend. So let me let me segue into this whole idea of macroeconomic variables in the context of trend with just by explaining why I'm worried

about trend or why I think it would be useful to try to find a way to improve it. Going back to the whole point about how I think they're going to be longer cycles that are, that have less of the boom bust and lower inflation, the Fed being able to step in more often. I'm actually worried. I don't want to present this as like a theory that everyone has to believe. But I myself, I'm a little bit worried that that can affect trend as a way of investing. And here's the reason why. So if you look at the data, trend, the way it works is a trade-off.

You trade-off, you take a lot of little whipsaws, a lot of little nicky cuts, you know what I mean? You take a lot of whipsaws, and in exchange for that, you end up escaping from really bad drawdowns. And it's in those drawdowns where they're really, really deep and drawn out, where the moving average or the momentum line is able to drop over time to where when you get back in, you end up getting back into the lower price. That's how it ends up being profitable. When there's a whipsaw, usually that doesn't happen. And so you end up selling low and buying high, right? You end up buying higher than you sold at. So if you look at the data and you kind of, I had that little program that kind of spelled out all, like take every single switch in various

momentum and trend strategies and just look at what happened. And each year and each month, what were the switches? What was the true trial size? When I do that, it spawned concern that if we have in the future, in the next 50 years, if we're in a situation where we have more whipsaws and fewer big drawdowns than we had in the past because the Fed is intervening, because the cycles are smoother, because the economy has gotten, maybe things don't get as bad because people don't make the same kinds of mistakes that they made in 1929 in terms of policy, 1931, and it's something.

If that's the case, then trend may actually end up disappointing. That's the concern. I'm not saying that I know that. That's the fear. And we see that in this cycle, right? Trend has not done well because of the fact that it keeps getting whipsawed. We have drops and we think this is it. Here we go.

And then in comes the Fed or income something else. Or maybe it wasn't as bad as we thought it was. I'm worried about that. So the whole idea, you know, I kind of put together this little strategy for myself. I called it growth trend timing, where I try to like segment off the periods where the trend strategy actually does the switching. In other words, I'm trying to get the strategy to not switch during times where there's a low probability of the switch being successful. If you just want to simplify it down, you can simplify it down to recessions. If you look in the past,

the recessions, the big recessions, 2009, 1924, 1929 through 32, those are the periods where trend really, really, really generated big deltas on the market. And those deltas are what have helped it to generate positive alpha or your better sharp ratio over time. even with all of the whipsaws that it incurred in between. So this strategy was just a way of trying to segregate off the periods where you have a good risk reward in making the switch, periods where there's a high probability of recession. Now, I have to be clear, and I would be fully transparent. The analysis that backs growth trend timing, it's thin.

The statistical evidence for it, it's limited, and it's more limited than trend because it's very hard to test it out a sample. You know, you don't have a big sample size in terms of the true sample size. You only have a few recessions. So I want to caveat all of that and say that this is not some panacea. but that's the motivation. The motivation is to try to find a way to make trend robust for environments where the ratio of whipsaws to big drawdowns may be different than it was in the past. And just the very fact of doing that kind of deviates from the way that systematic investors go.

You're not supposed to be treating each case on its own. But again, I'm not a pure systematic investor. You know, I like to do discretionary. What are those kinds of things? So in the growth in that title, what does that mean? What are sort of some of the underlying economic ideas? that you found interesting or compelling both, you know, analytically and in the data. Retail sales, real retail sales is a very strong indicator. If you compare it to the prior year, industrial production, you can use earnings itself. Housing starts. There's a lot of, there's several ones. And I want to be clear as well that, can we talk about statistical method? I take that very seriously. This is probably overfitted. There's a significant risk of overfitting here

because it's very easy to go back and find these variables and to build the model to where it gives you the right outcome. If you look just pulling the data, if it's, we have to use something, right? If we're going to predict recessions, those are the best variables to use. Real retail sales is a great one. And then industrial production, earnings, housing starts. There's a couple other ones that you can use, but those are the best. So another way of thinking about... Unemployment rate.

I'm sorry, I forgot what I think. Unemployment rate. The trend in the unemployment rate I found was a really, really good indicator of when recession start and end. So an interesting way to think about the prospective results for trend is that it relies on large drawdowns. large drawdowns can be caused for some underlying recession economic reason or because of some valuation blow off. But without those, I think it's a pretty clean case that you don't want to be a trend investor.

So maybe just a point, again, we don't know what's going to happen in the future. But for prospective trend investors out there, just realize that the causal source of your return is some large equity drawdown. And by the way, it takes a little bit of time. So even 2008, which we acutely remember for September and October, you know, it started in, November 2007. So it did take some time for that sort of pattern to form, but the 1987-type bizarreo correction is not going to be saved by Trent. Yeah, trend doesn't protect you from 87. It doesn't protect you from 2011 type stuff. You know, it gets you out when you want to stay in because there's just not enough time for the average or the momentum line to come down and force

you back in at a lower price. So you end up getting back at a higher price and taking a lot, taking a relative loss. I want to be clear there that I'm agnostic. I'm not atheist on trend. I'm just agnostic. I just, I'm worried. about the possibility that it might not work. And so all of that work was motivated on trying to deal with that. And what I found was simply that trend works better inside recessions.

And so if there's some way that I can narrow the applicability of it to recessions and kind of not use it outside of recessions, would it work to create a better signal? And it does. Acknowledging the fact that there's probably a lot of overfitting going on, that there's a lot of risk of overfitting

in the fact that I get to build all this stuff in hindsight. Let's see how it works in reality now and see if it, I mean, you know. To close the section on kind of market-level stuff, You wrote a post that I don't know if it was your most red post ever, but certainly got a lot of attention, which maybe because of the title, which is the single greatest predictor of equity returns. So I'm just curious for you to describe what this was, how you came up with the idea and what you think about it now, whether you think, whether your, what your level of conviction is in this idea? Yeah, I mean, so, yeah, that, that is. Am I right that that got the most? What has gotten the most?

Oh, by far that one. I mean, I still get emails about that today. Like weekly emails, people ask me questions about it. The basic insight that if you treat markets like a physical system, everybody has to hold every asset that exists in the market. And I thought at the time, I was thinking, okay, well, if there are changes in the relative supplies of assets, those changes are going to translate directly into the average portfolio allocation by definition, because somebody has to hold everything. So the average investor, like every investor as a group, has to have the same asset allocation as the actual supply ratio. of all the various assets. That prompted me and thinking,

okay, well, is there any relationship between that allocation and future returns? In other words, in periods where the allocation is very low, do you end up seeing

strong future returns coming from that position? And the answer is yes, you do, and the signal is a little better than the CAPE ratio. But I have to caveat that too and just say,

that whole process of trying to correlate returns to current valuations, it's just the sample size on that is not as large as you think because you're taking cycles. A 10-year period can be 10 years. It can be 120 months. It can be, again, if you have a calculator, you do the math. I don't have the calculator.

It could be like some number of days, some number of minutes, some number of seconds. And the point is that however you chop it up, you can get a different sample size of N. But that's not giving you a different trial. And in the same way, when you do correlations of acid, valuation and future returns, you can say, okay, maybe in January of 1963, the CAPE ratio was some value. And then I say, okay, the return from that year of January, 1963, to January, 1973, that 10-year period, maybe I had some return.

And maybe that checks with my model. So we've got one data point. We're good. Now I go to February, February of 1963 to February 1973. Then I go March, March of 1963 to March of 1973. The point is those are not different samples. Those are not independent of each other because they're overlapping. happening. And so if you actually look at the true sample size of those correlations in my graph

and in the graph that others have provided, they're not like these huge numbers. They're like four or five, four or five independent segments of data that are not connected to the other segments. And a sample of four or five is not enough, especially when you have the luxury of hindsight to build your own model, a sample of four or five is not enough to start drawing strong conclusions. So, you know, I think that there's a significant level of overfitting in both my chart and the charts that others provide there. And I want to be fully transparent about that. But I do think the concept of asset supply, of thinking about supply as a driver of market outcomes is a valid concept. And I think it's interesting because it's so difficult to parse that out in an analysis because there's so many other really important variables that are mixed in with it that you can't ever figure out what role is supply playing.

I do think, for example, that the fact that the corporate sector right now is buying back a lot of shares, is not issuing shares, they're not increasing the supply naturally of equity. and you have large issuance of government debt, money supply growing, credit to GDP is much higher than in the past. I think that means that, therefore, unless the allocation preference changes, the market valuation of equities to GDP has to also change. The market prices of securities has to go up because that's the only way the supply can keep up with the supply of everything else in the market. I think that's a valid way of thinking.

And like, you know, if I had to try to explain the point intuitively around supply, I would just say it this way. Let's suppose that you sold a stock at 1030 this morning and you sold Apple. you sold it to somebody else. You're not the only person in the world. There's other people that own Apple. And let's say some percentage of them, by chance, is going to decide today to sell their Apple shares.

Well, there has to be a buyer for all those Apple shares. So somebody has to be on the other side of that where they wake up this morning and they say, you know what, I want to buy Apple today. And you all have to meet together in the market. And you'll meet through the market maker, right? Obviously, you'll put your order in it all happen. The point is that if I doubled right now the supply of Apple stock,

right, if I literally right now just airdropped a bunch of Apple shares into the market, you would think, that the number of people that are stumbling into the market at 1030 to sell their shares is going to rise because you've doubled the pool of people that are eligible for doing that. If the ratio or the probability stays the same, the number of people who will show up will end up, they will also, they'll end up causing the flow in one direction to increase, right? And so, like, I think that actually has an effect on markets, and it's an effect that we can't ever parse out because there's also news flow.

There's also all 1,000 other things that are happening in the market, and it's very difficult. One area where I think that there's clear evidence, of a supply impact in markets is in the preferred stock space. This is fascinating. So the interesting thing about equity is common equity, is that common equity, there's a growth process that takes place in common equity,

that the earnings grow over time, the dividends grow. So there's a basis for the price to be going up constantly, and that tends to increase the supply. Now, the corporate sector can mess with the increase because if they stop issuing shares, they're going to affect the rate of increase. And if the other stuff in the economy is increasing, the price is going to have to increase even more to keep up

to keep the supplies balanced to meet the investor preferences for how they want to allocate, what percentage of their portfolios they want to put their assets into. But the beauty of preferred stocks, or actually the interesting feature of preferred stocks, is they don't have that feature, right? They're just a promise to pay a certain amount of money. And most preferred stocks, first of all, they're not used as often today as they were in the past. They used to be much more popular as a way of generating financing. Today they're just used by banks and other entities that have reasons in terms of regulatory reasons for using them,

in terms of how the regulatory structure is set up. But there are certain preferred stocks that actually have been around for decades and that are not callable or that have never been called, that were issued during periods where the yields on preferred stocks were very low. So these are very, very attractive securities to the companies because they're very, very low yielding. And what's interesting is that back in the 1940s and 50s or the 1930s, right, when these things were issued, and they still trade today.

when these things were issued, they were issued in the supply that was appropriate for that period of time, right? When the money supply was, let's say, like, 10% or 5% of what it is today. And so, like, let me just give a couple examples. So, Union Electric is a utility in the Midwest. They issued a preferred stock on April 30th, 1946. It was redeemable any time at 10% above the par value, and the par value was $100. The yield, and this is, prepare yourself for something crazy here, the yield that it was issued at, the yield is 3.85%. That's the yield. And it's trading slightly above that right now. But that's a pretty low yield.

For a security that's callable that has no legal recourse if the dividend doesn't get paid. I mean, you can make a claim on the common that they can't pay their dividends until you get paid. But there's no bankruptcy proceeding. There's no way to force them to sell assets to pay you. You don't have any recourse like that. So that security is yielding right now in today's market. It's yielding 3.85% or thereabouts. And it's redeemable. So you don't even have upside. If we have rates go negative and we have this Japanese scenario, they might be able to call it and take away some of your upside. The thing that's fascinating about prefers is that you can compare the preferred to the same company. You can make comparisons inside the same company to their debt.

For example, the same exact company, Union Electric, has a 40-year first lien mortgage bond that's yielding 3.55%. So this is like top of the food chain in terms of liquidation. The only difference that you're getting in those two securities is 30 basis points. and one of them you have full legal recourse. You are first in line. Is that worth 30 basis points of difference of Delta? There are absolutely not. Now, there are some tax differences that you can factor in,

but even then I don't think it can explain why that preferred is priced at such a low, a ridiculously low yield. Give you one more example. People will have heard of like Alcoa, the company Alcoa. Part of it is now called Arconic, which I didn't even know until recently,

but they did some kind of weird spin-off or whatever. But the preferred stock that was issued by Alcoa is still in the market. And this preferred stock, This is crazy. So this preferred stock is trading at a yield of 4.25%. And it was issued on January the 20th of 1947. If you compare that, let's compare that to the actual senior debt of the same company, Alcoa. The preferred yields 4.25%. The senior debt yields 5.76%. So you might ask you what in the world is going on. Why would anyone literally give up 150 basis points or whatever to a

invest in a security that's like several rungs lower on the totem pole, you know, in terms of the food chain for the credit, right? And the answer is, I think part of the answer is the supply question, right? Because the point is these securities were issued way back then when the amount that a company would raise was very low because the economy was much smaller, right? And these companies were much smaller. If you look at the Union Electric preferred, there's only $7.5 million of that preferred in the market right now. It's a very, very, very, very thin, scarce security. It's almost like a baseball card. It's a vintage issue. There's not a lot of it. So every day, you need to have people that are waking up and wanting to sell it. There's not a lot of people to do that. Now, the flip

side of this is that the other side of the equation is the people that wake up and they want to buy it, right? And that's influenced by like the visibility and by like the attractiveness of the price and whatnot. And that's how everything equilibuates too, because if there aren't enough buyers for the sellers, right, then the price will change and, you know, maybe it makes the news or it shows up on a screen and then that will create buyers. So there's still that equilibrium process. But still, there's no way to make sense of this configuration. And it's because I think that's one of the best examples of the hidden effect of supply affecting a market. Because here, it's the same exact security. So there's no way you can rationalize this in terms of like efficient markets theory.

Even if you take into a consideration some of the tax benefits with the Alcoa case, it doesn't justify them. It's like you're literally taking a lower yield in exchange for like significant more credit risk. There's nothing else that you're getting. There's no upside and nothing. It's longer duration. so you're taking more inflation interest rate risk, makes no sense. So that's one area where that makes me confident in two things. One, it makes me confident that supply matters, even though I can't prove it, even though there's no way to ever parse things out perfectly. And two, it makes me confident that markets are inefficient, that there are inefficiencies that develop that are real, that make no sense. And that, you know, if you want to be strong on the EMH theory,

I think that's wrong. I'd love to talk about factors now. Probably the most interesting thing that we've done in this what we call our research partners program, of which you are the flagship member, is finding people who have really interesting takes on markets and investing, who are good with data, obviously, and are just looking at our process for the first time with no extra grind. And a true people talking about cognitive diversity, this program has made us realize the true value of cognitive diversity, because you just get people with a completely different set of interests and skills looking at the same problem. So one of the first things we did was to basically look into value and momentum and ask sort of an elemental question, which is just like,

what is going on here? Why do these factors actually work? What can we map it back to in simplest possible terms to better understand why they work and why they might continue to work into the future? So what the primary levers are? So maybe you could describe sort of attacking this problem and how you did it. So obviously having access to the data that, you know, OSAM has, it opens up a lot of doors in terms of what you can look at. And I thought, you know, we talked about looking at, this idea of what is happening. There's different ways that the value factor could work. You could have it work because maybe value stocks pay higher dividend yields and those dividends accrue over time. That's one possibility. They're cheaper so they should have higher dividend yields. We talked about

there's another possibility that maybe the multiple is expanding. And I've always wanted to know what it was. I wanted someone to get deeper into it because it just saying it works isn't enough. I just wanted to know what was going on under the hood. So what we did was we kind of went under the and looked at the earnings. And what was really fascinating at first to us, and you can remember this, was finding out that these companies, their earnings drop.

Their earnings go down during the holding period. You might buy a stock at a 10 times PE ratio and think that you're not getting a 10 times PE ratio for ad infinitum. You're getting like an E that's dropping. That makes sense if the market is kind of getting in front of what it thinks is happening, right? So if you look at a stock, like, let's say, Best Buy,

or whatever the retail stock that's getting hammered today is, I don't know what it is. I haven't kept track of that. There's a reason why those things are priced cheap. It's not because the market wants to give away value. There's a reason. It's because the market believes that the future earnings are not going to be, the current earnings are not going to be reflective of the future earnings.

So the question would be, what would make value justified as a, like, what would make it justified to buy a low P.E. stock? And the true answer would be if the future earnings over the long term were actually, if the company was, even though its earnings are going to decline, if the earnings in the going, looking out way into the future, if they were going to recover. And actually, in the final analysis, when you hold this thing for the long term, you end up getting more earnings for the dollar you spend on the price.

That's what would make it a justified strategy where, like, literally it has an investment basis for outperforming. If the future earnings that you're going to get from this company, even though it's going to drop in the initial phase, would still be better because of the fact that you bought it is cheap. And so we tested it out, we tested out value on longer horizons. And we looked at how the earnings fluctuated over those longer horizons. And we saw that there was a recovery. The companies would initially have a decline like the market expected, but over time, the earnings would come back. And they would come back to a sufficient level to actually make them more attractive at that initial price than the growth stocks, where you were getting growth the whole time, but the pricing was kind of expensive.

And so you weren't getting as much in the end for the amount that you spent. And so that was a key insight for us. We realized, okay, what's happening here is that the market is probably going through a period where it's getting very bearish on a, on a company, and it's over-discounting the company. But then over time, the earnings are turning out to be not as bad as had been thought, and therefore the market is having to adjust and rewrite. And we saw that in the form of the multiple expansion that happens inside the holding period. You can see, literally, if you look at it, that the multiple is expanding. Now, part of the reason why the multiple is expanding, obviously, is because the earnings are dropping. If the earnings

drop and everything else is held constant, the multiple is going to go up because the P is constant, the E is falling. But we found that actually, no, there's more multiple expansion than just, just that. There's actually an additional multiple expansion that's very significant. And I think our theory that, and I think a working hypothesis there, is that the market is overreacting to bad news, overreacting to a gloomy outlook. It's very difficult to continue to hold stocks when you don't know what's going to come, what's going to happen. But then over time, those companies end up recovering. And they end up recovering their earnings to a level that is sufficient to make them attractive value at the time that you buy them. And in that process, you end up generating alpha.

Let's talk about momentum now as a weird and interesting complement to value. So again, back to like the simplest possible attribution of returns being fundamental growth, multiple expansion, return of capital, right? It doesn't get much more simple than that. And you can even count most people reinvest the capital. So it's really just fundamental growth and multiple expansion. So if we just think about returns in those two lenses and you've described how you get value, now describe momentum.

So it's a completely different story, but one that's really interesting in its own ways. Yeah, I mean, so momentum, you find that there's like super normal. earnings growth that's happening in the period that after you buy the company. In the momentum, like the market is in the process of trying to price in a very, very strong growth outlook for the company. And what's weird is that, and I mean, this is open to debate, whether this is the right interpretation, but what's weird about momentum is that, so if you buy at the beginning of the holding period, if you buy the momentum stock when it's a momentum stock, when it becomes a momentum stock, and you sell it in a year, you do great. But the person that you sell to doesn't do

great. The person that you sell to ends up like underperforming big time. So that leads me to believe personally that what's happening is kind of like the opposite of what's happening in value where in value, the market is overreacting in the beginning of the holding period to maybe a weakening outlook and is basically pricing that in and overpricing that in and end up making a mistake that it has to recover from. I think the opposite is happening in momentum or at least the average case of momentum because this is all this is like there's lots of variance in this. But I think that what's happening is that the market is seeing really, really strong fundamentals. it's extrapolating them out. And in that process, it's basically the prices is catching up with the

fundamentals. And the market is trying to basically price all that in. And it kind of overdoes it. It kind of goes too far at the end of the holding period. It takes it, it takes the process too far to where whoever you now sell to, the guy who holds your bag, he ends up taking a beating as things, you know, mean revert. But I have to be, I have to be clear that I don't, I mean, that's tentative. That's a tentative theory, tentative hypothesis. I don't want to offer that as like truth. I mean. Right, right. Like always, momentum is confounding, incredibly strong. wrong statistical pattern that's hard to explain. Right. And, you know, another point on that is that like value to a fundamental investor value would probably be more comfortable because you actually can,

if you buy a value stock, at least historically, I don't know about the future, but historically, if you bought value stocks and you hold them for the long term like Warren Buffett style, you actually do get some alpha from that. And so it's a process where the market is writing itself. You're making money off of the market correcting a mistake. But in momentum, at least the interpretation that I just presented, which is not something that anyone has to necessarily take as a rule. It's open to disagreement, but the interpretation that I just described is one where you're making a profit off of the mistake. The mistake is happening at the end, not the beginning, right? So you're actually selling into a mistake.

That might make you a little more uncomfortable maybe with the process because you're not really buying a fundamental value stock. You're buying a stock that maybe is either fairly priced or a little bit overpriced, and you're selling it for a price that's even more overpriced at the end. But the point is that, look, you know, the market, all of these things are just vibrations in the market, right, that are created by human behavior and whatever else drives outcomes. And there's nothing wrong with profiting from vibrations that are going in the wrong direction. If it works, it works. And momentum clearly works or has worked in the past. And obviously, one implication for future research is playing with these levers. So in a value stock understanding the prospects for future growth, value works because those companies recover. If they don't recover, value will not work. Exactly.

So parsing value stocks by some sort of handicapping, you know, moneyball style of future fundamental outcomes, sales and earnings. Not that that's, easy. It's incredibly hard. But improving value signals seems to rest almost entirely on that sort of work. And you've written about that as well. So, you know, just a fascinating treatment of the source of returns in factors. In closing on investing and markets in general, if you could have any question answered. Oh, that's a great question. Man. In markets, what would it be? Oh, my gosh. This is kind of cheating, but I want to know how this cycle ends. I want to know whether or not. I want to know whether or not, I want to know what the average valuation of the market is going to be

over the next 30 years from today forward. That's what I want to know. I want to know, I want to know what's going to cause us to have the next recession. I mean, that's kind of, obviously, of course I would ask that, but not just a profit, not just to make profits off of that, but to just, I find that to be really fascinating. I've been most of my most intense periods of investment, obviously I was, I played around the market in prior periods, but this, this period has been my most involved period, and it hasn't yet ended. And it's like, it doesn't show any signs of ending. And I'm very fascinating and understand how it's going to. Given your unique personality and set of interest, that closes some non-investing questions. So the first of which you mentioned at the beginning

sometimes spent in the military. And I would love just to hear now with some perspective in hindsight, the benefit of hindsight, what lessons you took from your time in the military that might be interesting to people. It was interesting because at that phase of my life, I hadn't really developed a lot of insights about how people relate. And in the military, I was on submarines, and that forces you into an environment where it's very, very tightly packed. You develop relational skills with people, simply because you have to live next to them constantly for months at a time, away from your family, away from things that you enjoy. And I think, well, I think one of the biggest takeaways that I developed there, the military has a big emphasis on this idea of praise in public, criticized in private,

and lots of egos, and we have lots of disagreements about stuff, about anything. I mean, people, you're on a submarine, right? You're inside this tight space and you're there for months at a time. And I think that the biggest insight that I gained from that is just the idea that that everybody's status matters to them a lot. And that in all of our communications as people, we need to be respectful of that and be sensitive to that empathetically. And so, you know, I had a situation once where we were sitting in a roundtable and we had squadron and we had big wigs that were on our ship that were inspecting us. And my weapons officer, one of my supervisors was basically like we were in a disagreement about something and we were trying to figure out who was at

fault and it just clicked to me. Like, I need to think not just about, in this situation, I need to think about egos right now because this is not going to diffuse itself in a way that's going to be optimal for me if I don't respect that fact. I can be right, but there's sometimes a cost to being right at the expense of someone else being wrong. And that's actually influenced my thinking really, really heavily because it goes, it kind of goes contrary to what you hear, which is like we say, okay, well, we're a truth-based organization. If you, if you are right, if you have some concern, raise it, bring it up, express it. And if it leads to a, disagreement, then, you know, have the, have the fight, you know what I mean? And the problem that I've

always had with that is that there's a cost to that, right? There's a cost to the relationship that gets damaged. And that's the big thing that I learned in the military. There's a cost to that. And I was able in the military to utilize that. I picked up on that real quick. And I was able to utilize that to kind of build good relationships with the people that I worked with, especially my supervisors, because, again, military status is very, very important in the military. Like, you don't rank, you know, that type of stuff is very, very sensitive. And when that stuff gets challenged or you cross those lines, no good, right? Not good. So anyways, how does that affect my current thinking? To tie into this idea of truth-based organization, I think that it's not enough to just say we're like an open thinking,

like free thinking, honest, truth-based organization. I think you need to like build incentives that respect those risks that people have to take and that kind of work around the costs of voicing a dissenting opinion. So let me give an example, which you call the pass-through method, right? I think you get a great name to it. So let's suppose that, well, I could use an example from like an area like regulation or prosecution, but let's use an example from like, let's say, like an investment case, right? Let's say that there's some guy, he's like the golden boy of the firm, and he's got this thesis that he wants to invest in Xon mobile. So let's say the chief investment officer or the portfolio manager is going to have the big group of like peers get together and he's going to present

on his thesis and he's going to explain why he thinks we should invest in Xon mobile. So let's suppose that I see some big problem. I'm not, I'm like this, I think this is crap. Let's suppose that that's the honest truth. I think this is just a crap idea, we're going to lose money, okay, this is just a terrible investment, bad risk reward, he's making mistakes in his presentation. I got a couple options. I can just, if I, if I come out in front of everybody and I raise my questions, I can try to like, I can spend time trying to like massage my challenge to him in a way that hides it and conceals my, that I'm attacking him, or I'm attacking his thesis. But that takes a lot of resources and a lot of effort, right? It's like a,

it's a huge drain to have to do that when you're trying to raise an objection. And even then, it's not clear that you're going to be able to raise the objection in a way that's not going to hurt people's feelings. That's the first concern is I don't want to make John my enemy because I might need him later on or I might just want to be in a good relationship with him because I work with him. So I might have to tone down and that's not good because that suppresses true disagreement and true dissent. It suppresses an honest take. I might massage things or make things softer than they should be. Maybe I should hit hard. Also, I'm worried about what the, if it's the Navy, it's the captain, what the captain thinks. If it's, you know, a hedge fund, maybe it's the owner of the hedge fund

or the manager of the mutual fund or whatever. If I'm going to be thinking, well, what does he want to hear? Because that's going to, how do I endear myself to him, right? Those are all incentives that we cannot get out of. We have to be honest and acknowledge. Those are big incentives. And we're all in this for ourselves, ultimately, to do well and to build status for ourselves and to build success

and to have people like us and have everything work out great. The process of dissent can sometimes threaten that, and that's why you have to hold back sometimes. So what I think needs to be done in those kinds of situations is, I think the man. or the person that's running the situation, needs to be aware of that and be honest about that and realize that you can't just say to everybody,

hey, say what you think and have, you know, disagree and because there's costs. And he has to understand those costs. And he has to structure the way he gets information in a way that provides workarounds to that. So let me give you the way that I think that it should be done is like to use the example of John and his ExxonMobil thesis. I think that the best way to do it would be

instead of having everybody in front of everybody else have a big fight and, you know, who gets to look the best in front of the boss. I think the best way to do it would be for the boss to say, okay, John is going to give a presentation. He's going to present on his thesis. Now, everybody in this room that's watching, you have a job. Your job is to listen carefully to his thesis.

Take it in. Think about it. I want you to then come up with reasons why, or I want you to identify the biggest weaknesses in his thesis for yourself. Here's what I want you to do with those things that you identify. don't want you to voice them in here because that's going to create a, you know, it's going to create a big brawl. I don't want to do that. I want you to take your insights, your thoughts, your findings, your, what you think is the weakest aspects of his case, right? And maybe the strengths, too. You can keep it

both ways. And then I want you to send them to me, the boss. So what does that do? First of all, it eliminates the fight. It eliminates all the politics that might happen if there's disagreement and challenges. What it also does is it preserves the incentive of the employees to actually voice the opinion of dissent because the reason why they would voice it is because they want to look good in front of the boss. They want to add value. They want to be seen as, hey, I found this mistake in your work. So there still is in that

context because everyone is sending in the feedback to the boss. The boss is maintaining and preserving the incentive for the actual employees to actually put some effort into this process. If I'm going to write something to the boss and he tells me he wants to hear the best arguments against this thesis,

I'm going to make an investment in that because he's going to see it. And he's he's going to judge me in the future on the kind of work that I do. So you're preserving that you're aggregating that you're getting all the information. You're providing, I don't want to use this word, but it's a good word for this context. You're providing a safe space to be a dissenter because you're saying you have to dissent. You have to give me the best that you've got to challenge this idea. And so you're preserving that aspect of the incentive process too. You're making it very comfortable. And then in diverse environments, you have different personality types. A lot of times

people that are more timid, they end up getting drowned out by like really strong, strongly voiced. People that have strong personality types and strong presences, they can sometimes be interruptions and you can have all kinds of things that happen there that make people not like where they work. This approach is also bypassing that. So in my own work, in my job, I use that whenever I can. I call it from your naming, I call it the pass-through method, which is that I basically tell everyone, tell all the employees, you have to come up with the best challenges to this idea. You send them to me.

When I get them, I'm going to take them. them. I'm going to pick the ones that are the best. I may contact you to get more information. I'm going to pick the ones that are best. I'm going to then take that to John and pose them to him. And then we can iterate. We can keep going. We can redo it again and again. But I'm basically taking this process and putting it inside of something that preserves all the incentives in the right places to make the true, honest, legitimate, strong dissent happen. And I don't think that we do that enough in corporate America. I think that a lot of times we just go with what's simple, which is just have everyone kind of just, if you have a disagreement, just voice the disagreement. But like there are huge costs to doing that. And you can't get away from, you can't eliminate those costs just through saying something, like saying, well, I'm a, as the CEO, I want you all to be disagreeing with each other. I want you all to voice your views. That's not how it works. I mean,

there are costs whether or not he says that. And by saying that, he doesn't eliminate the costs. But if he sets up a program or a system where he can take in that feedback and have it be honest, and then he can take it and he can dispense it to the right people in the right context. and then he can do all the work around managing emotions. Like in other words, instead of having each person have to learn workplace poker, right, have to learn how to play that game, how to voice a disagreement in a very, very respectful and, you know, non-threatening way, all that work can be thrown out.

He can do it one person. He can do it from a position where he doesn't have as much risk. I absolutely love that. And I realize two things as you're describing that idea. Sometimes I love when there's episodes of the podcast where I realize what the title's going to be during the conversation. So I just realized that the title of this one is going to be the search for truth. Because if you think about physics, philosophy, everything you talked about at the beginning

and just then, the unifying theme and kind of what you seem to be doing is just searching for truth in all sorts of different ways. And the second is a new concept that maybe we can, you know, another time flesh out more, which was brought to my attention by a new friend named Jake Weinreb. And Jake told me about this idea in companies, I guess, is the context that he introduced me to it of the line, that if you're an above-the-line organization, you're committed to that search for truth and learning. And if you're below, you're more focused on ego and being right. And if you decide consciously to be the former, magic starts to happen. And if you're the

latter, it's an extremely frustrating exercise. And what you just described is basically a way of engendering that in a business, which is fascinating. Given the idea of this search for truth, I want to close, I'll do my standard closing question, but one more prior to that, which is at dinner we were talking about this, this really elegant idea of progress, what you call progress in meaningful work as sort of a true north, if you will, for your own activity. Can you describe that concept for people? Yeah, sure. I mean, and these are not, I don't have any expertise in this. This is just personal philosophy. I've tried to look at my own experiences and try to see where, like, when do I get really, really, really enjoyable experiences in life? And they always come, that phrase is not my phrase, by the way. They always come for me in situations where there's progress in meaningful work. So like, what does progress mean? Progress means that, like, I'm engaging in solving a problem and it's actually, there's actual traction. I'm making progress. I'm getting somewhere. My skill, are adequate to meet the challenges, and I'm seeing myself move forward. I'm like getting towards the truth, let's say, or getting towards something really interesting or something good. The meaningful is that, like, I find that, like, for me to be interested in something,

it has to matter in my worldview, right? It has to have some significance. And we use, like, an example might be, like, if I was in seventh grade, you know, meaningful might be being good at playing basketball or playing soccer. But today it wouldn't be, it wouldn't be, right? But, you know, now meaningful is, like, can be used to make money. that's one thing that I find to be very, I just get into it. Like, if it's some way for us to make money, I just, I find that to be, it just gets me going.

But like, for whoever it is or whatever it is, it has to be something that, not only do you need to have the traction and the progress happening, you also need it to be something that is in an area that matters that actually has, that makes a difference. And I think that a lot of jobs are jobs that just don't matter. And we don't have that feeling. In fact, I'm not going to say anything about that, but a lot of jobs are like that. There's no sense of like being good at this is something that makes me proud of myself. Being good at what I'm doing. You need to have that, right? And then obviously the work aspect is it has to have some sort of out. It has to involve you in terms of your engagement. It has to be yours. It has to be something that you own.

I mean, you know, those are, again, I want to caveat that. That's just a very, very crude personal. I think it's simple, obviously, but I think it's just a great framework. And reflecting on periods that you view as positive in your own history and what they had in common, like that's a pretty powerful way of orienting oneself. And I think, especially importantly, there's a lot of people that don't do that or haven't had that experience for, variety of reasons that are both sad and tragic and everything else. But if you can push yourself to be more in that kind of loop, it's a positive thing. Yeah, I think to say to a little more on that, I think that, you know, the challenge is just trying to find out how do you set yourself up for that

kind of experience. Like, how do you, you know, what are the kinds of activities that you should be involved in to get more progress and meaningful work type experiences to happen? I talked about Daniel Gross. He's a guy who's having a lot of influence on my thinking as well around the idea of flow and how to create flow and how to like tackle a problem that becomes overwhelming by like focusing on making the problem easier in terms of traction right because that's idea of progress sometimes you hit a roadblock I think that in that situation for me what works a lot is to kind of make the problem like try to tackle the simplest problem or the easiest problem that's in front of me so that I can get some traction and if I can get some traction from that that'll create some

momentum and some like some fuel to keep going if I just hit a roadblock and I try to solve everything at once, it just becomes overwhelming. But if I can just stay, okay, what do I know? What do I not know? Where can I make a step here? A simple one step to get closer to truth. Let's just find a way to do that. And oftentimes, just by doing that, I'm able to then get fuel and get momentum and then it starts to build. And then I start getting in flow. And there we go again. So you know my closing question, which is to ask you what the kindest thing that anyone's ever done for you is. Yeah. So, I mean, obviously, I'd be a full amount to come prepared for this one. So obviously, I do want a caveat first. You know, this is not an allowed answer, but I have to have to, I have

to put my parents in there. They've just amazing parents. My mom and dad have basically been helpful at every stage of my life in terms of providing me whatever I needed, whether it was supporting me through grad school, through whatever else. I mean, just unbelievably good parents. Obviously, you're not allowed to use that answer because obviously that's everybody's answer. But this is interesting. I thought about this. It's tough to answer because most of the time, I don't want to be cynical, but most of the time when people do things for you, they're doing them, you know, not because they, I don't say because they want to say because they want something, but because they, they're self-interested. Yeah, there's a,

a self-interest component to it and it's hard to really sit to pretend like that's not there, right? It may be because they just like you and they like your friendship, but there's something that they like that they're getting for them. So I'll give you, I thought about this and the example that I think is the best example. This is not really, this isn't like out of a storybook situation, but it's actually a pretty kind thing. So I, like several years ago, I was basically looking to get a surgery for an injury and I was really stressed out about it and I didn't know if it was the right move to make. And so what I did was I, I was on, I was on, this website called PubMed.com, which has like all the journal articles. So I went on there and I

basically looked at, I was researching the injury to see, okay, what's the best way to get surgery for this? And so I found that this guy's name kept coming on. I'm not going to say who it was because he would get in trouble if I kept going. But I found the name of this guy who was like the top chief medical expert on this in the entire country. MD, you know, like just the top pedib you can imagine. And he had had a massive research. So I'm like, you know what's the risk? I found his email address. I sent an email to him. I sent him all of my information, right?

X-ray, stuff like that. And I just kind of tried to frame it in a way that hopefully maybe he might just say something about what he thinks. And this guy, I'm talking like, he comes back to me at night. Like apparently he was in bed or something like that. He's on his iPad. It's sent from my iPad, right? He sends me back like this three-page write-up of like just, hey, I mean, like, here's what you should do. Here's what I would try first.

Here's some advice. Here's what I found in this study. study. I'm like, whoa, man, you don't even know who I am. And you're taking on the risk, because it's not fully kosher to be doing that, you know, to be talking to patients that you've never seen before and giving medical advice like that. He gave me everything. And he actually helped me make a really good decision in that area by just basically providing me with free medical service, free medical care for nothing. And in a way that actually was not only a net positive to him, possibly a net negative, because he's actually now taking a small amount of risk that,

hey, what if I don't like what happens and then I sue him, right? I mean, that's something that I would be thinking about. If I was a doctor, I'm like, hell, no, I'm not going to work for free and just give you information out of the goodness of his heart. And it's just, it was really like, wow, you are an awesome person. And I would say further that I think a lot of the best people in our society really are the people that work in medicine. If you think about surgeons, you think about, you know, doctors, they just, I don't know

what it is, but my experiences with them have been very positive. Well, really cool and unique answer and just an awesome conversation all around as I knew it would be just like all the many conversations we've had over the year. so thank you for all the time. I'm glad to do it on the record. All right, man. It was a blast. Hey, everyone.

Patrick here again. To find more episodes of Invest like the best, go to investorfieldguide.com forward slash podcast. If you're a book lover, you can also sign up for my book club at investorfieldguide.com forward slash book club. After you sign up, we'll receive a full investor curriculum right away

and then three to four suggestions of new books every month. You can also follow me on Twitter at Patrick underscore or OSHAG. If you enjoy the show, please leave a quick review for us on iTunes, which will help more people discover invest like the best. Thanks so much for listening.