Today we have another awesome guest on the show: Michael Mauboussin. Michael is the head of Global Financial Strategies at Credit Suisse. He is one of my favorite authors and the author of three books, including, More Than You Know: Finding Financial Wisdom in Unconventional Places, which was named one of the 100 best business books of all time. Michael also serves as an adjunct professor of finance at Columbia Business School and is an expert in decision making, behavioral psychology, and all of those fields applied to the financial markets, especially. Michael, welcome to The Science of Success.
Michael: Thanks, Matt. Great to be with you today.
Matt: We are super excited to have you on here. So, to kind of kick things off and get started, tell us a little bit about... For listeners who might not be familiar with some of your books, tell us a little bit about your background, and how did you become so fascinated with the psychological aspects of human decision making, specifically within the context of investing, which you're obviously an expert at, but also, you know, even more broadly.
Michael: You know, Matt, I think part of it is you mention my association with Columbia Business School, and I started teaching there in the early 1990s and I was thinking a lot about what I was talking about with the students, effectively giving them tools to try to make them successful investors, and sort of had this growing feeling that what made for great investing had less to do with the tools--you know, accounting and financial statement analysis and valuation, although those things are obviously really important--and much more to do with decision making and temperament, especially under stressful situations. So, probably in the mid-1990s, I started to just open up my reading quite a bit. A lot more science, a lot more in the world of psychology, and sort of being exposed to this world as a lightning bolt of recognition that probably what makes for great... not just great investors, but really great in any field, is awareness of a lot of these psychological factors that improves the quality of decision. So, it sort of changed my whole tenor, recognizing that a lot of things we teach, for example, in business schools or actually any kind of school, are just the ante to the game, but the real success has to do with this whole other field of decision making. So, that was sort of my epiphany, was that recognition of where value comes from. The other thing I'll just mention is I was reading widely... You know, I was one of those guys who was... You know, I'd read something and I'd be like, oh, here's a connection to this, or here's a connection to that, and just sort of this recognition that we live in an extremely rich world, and that there are a lot of interesting connections between different things that may not be superficially obvious but that I think could really make... that could be some really fascinating connections, and I think really helpful connections to allow people to think about the world more effectively.
Matt: And that's essentially the concept of the idea of multi-disciplinary thinking, that Charlie Munger is a huge proponent of, and I know you're a huge proponent of, and something actually we touched on a little bit with one of our previous guests, Shane Parrish of Farnam Street. Can you explain a little bit more about, and maybe even provide some examples of, how different disciplines can impact each other or how maybe psychology can underpin finance, or something like that?
Michael: Yeah, absolutely. The way I like to think about this is that it's like a toolbox, the metaphor of a toolbox, right. You might have the best hammer or the best screwdriver of anybody, but what you really want to do when you're thinking about the world is to have the right tool to apply to the right problem. And so, I think the Munger approach... And I do. I give huge credit for my thinking to Charlie Munger, who I think is the most articulate. I'd also mention another book, which many of your listeners may be aware of, by E.O. Wilson called Consilience, and these ideas that many of the vexing problems in our worlds are at the intersection of disciplines and we need a sort of full toolbox to try to tackle them. So, to me, this is the way to think about the world. The other thing I'll just say is another quick comment, is that we've made huge strides in science over the last, let's say, 400, 500 years through reductionism, which is to say basically breaking things down into its fundamental components, and it's been extraordinary, and I think a lot of the things we take for granted in life, advancements, are the result of that amazing work. But I think increasingly, we're bumping into areas where we're dealing with systems that are complex, where reductionism really doesn't work, where, in a very real sense, the whole is greater than the sum of the parts. And that requires a very different way of thinking about the world. Now, if you think about academia in general, you get paid for specialization. You get paid for being narrow. But a lot of the problems in the world are kind of going the opposite direction, where it's important to think about things from different perspectives. So, one example I would give you, and I think is also a very powerful mental model in and of itself, and for me was another big eye-opening moment, is just thinking about markets as complex, adaptive systems. The stock market, right. So, if you say to an academic or a really traditional economist, "How should we think about how people behave?" they'll typically say, "Well, we've got these models of agents who are rational and they understand their different... They have information that comes in and they understand their preferences and they have utility functions, and then they make decisions on the basis of this. You know, we've known for a long time that empirically, that's not how the world works. So, if you try to extrapolate that into a model of markets, it just doesn't fit the facts all that well. Complex adaptive systems, by contrast, come at the world as thinking about the interaction of heterogeneous parts or agents, right, and you can think about other examples like ants in an ant colony, right? Absolutely fascinating, because the colony itself is almost an organism. It has a life cycle and is sometimes aggressive, sometimes passive, but every individual ant is really basically clueless. They're sort of bumbling little agents within this total. So, I think that's a much, much richer way... And by the way, your consciousness, for example, neurons in your brain, you can think about example after example, people that live in New York City are components of a complex system. And when we take that sort of set of tools and that way of thinking to the world of markets, it just opens up, again, new ways of thinking about things gives you good reason to understand why markets are generally hard to beat, but it also gives you some insight as to why markets go periodically haywire. So, to me, this whole mental models thing is just a really, really powerful way to think about the world. Now, let's talk about the pros and cons. The pros is, I think, that if you do understand big concepts from various disciplines, gives you a huge leg up in life. The con is it requires constant--basically--reading and thinking and learning. So, if you're going to get into this world, it ends up being sort of a commitment to perpetual learning. Now, that's not everybody's cup of tea, but if it is, I just think it's a really fun, exciting, and I think ultimately a great way to find success.
Matt: I love the idea that the traditional education or business school or whatever it might be is sort of the ante to get into the game, but if you really want to win, if you really want to compete at the highest level, you need to have a much richer and much deeper toolkit to really understand reality.
Michael: Yeah, and I really think that's the case. The other thing I'll just say is that's certainly true. I also think that there are gaps now in our education. Especially, for example, in high school and college students. I'll give you one example, and I don't mean... This is sort of a negative example, but I don't mean to be too negative. One of my sons went to a really terrific high school and they decided to develop a leadership center for the kids, which is great, right. So, they were working on things communication, cultural awareness, a lot of things you would say are really important. But what struck me as fascinating about it is there was actually no segment or module on decision making or on psychology. So, I went to the guy that ran the program and I said, "This is really interesting, because at the end of the day, our future leaders are really people that need to be equipped in understanding how to make decisions, understanding being [INAUDIBLE 00:10:39], or understanding the scientific method and what science tells us. These are actually very essential elements in the future, and we're just basically not teaching those things. So, that, to me, is another area that we should be spending... And by the way, I'm about to go back to one of my college reunions, and when I went to college, the kinds of things, the decision making courses--they're now much more common--didn't exist at all. So, if you're someone of my age and you're in your forties or fifties, chances are you didn't have any access to it in school. So, there's more of it now, but certainly not enough of it, in my opinion. So, yeah, I think you have to supplement a lot of what your curriculum has been in order to become a more well-rounded individual.
Matt: So, if you're somebody that's listening to this podcast, what are some easy steps or maybe some first steps they could take on the path towards starting to build this toolkit or starting to maybe understand human decision making more effectively, or make better decisions?
Michael: Yeah, Matt, and I think that you know my answer, which is probably to start, whether you can read or certainly listen to audiobooks or something, but there are a handful of books that'll probably get you off and running. One book that I always loved, and I'm sure you're fan of as well, is Bob Cialdini's book Influence: The Psychology of Persuasion. It's an easy book to read. It's got six big models about how you could influence people and their decision making, or you can also see or reflect how those things influence you and your decision making. So, that's a great starting point. Another great one, of course, is Danny Kahneman's book, Thinking, Fast and Slow. It's probably a little bit more of a challenge, but so rich in terms of its content. So, that would be another thing I would say, is people reading that and just really, I mean, the degree to which you're willing to wade into the, for example, the psychology literature is fantastic. So, that's one set of things. The second set of things is if you have an appetite to do so, it's really great to try to hang out with people who are different than you. And that might be if you're a finance person, hang out with artists or people who are into literature. You know, there was a very famous essay many years ago about the two cultures, sort of the literary culture and the scientific culture, and the argument was these cultures really didn't meld with one another, and I think those people who really tried to reach out, to understand different points of view, have diverse thoughts, I think that really just forces you into being actively open-minded about the world and, I think, really gives you a leg up in a lot of circumstances. So, I don't know if that's a gentle entry in, but probably the first thing I would say is to start to read some of these things and think about, be introspective about how they're influencing you or how your decision making processes work, and then just make an effort to reach out to people who are different. You know, is Brian Grazer the guy who wrote a book on creativity recently? Do you know that book?
Matt: I do not.
Michael: The Hollywood guy. So, the Hollywood guy.
Matt: We'll put it in the show notes.
Michael: [Laughs] Yeah, exactly. So, we'll track down the exact book, but I think it's just called Creativity. And he had this sort of extraordinary story, which I absolutely love, and he said he just made a point, is when he read an article... He's a pretty famous producer now, but he'd read an article about somebody, he would just say, "I want to meet that person," and he would call them up out of nowhere and say, "I'd love to have a cup of coffee with you. Can we make that happen?" And he'd reach out to people where it'd take six months, 12 months, 18 months to schedule something, but he was just reaching, going all over the place. One week he'd be talking to a lead athlete. Next week he'd be talking to an astronaut. Then he'd be talking to a Navy SEAL. Then he'd talk to a police commissioner. I mean, this incredible, fascinating array of people, and he just made it part of what he was about, and I think he argues that really helped stoke his own personal creativity and mindset.
Matt: That's fascinating. And that makes me think of two kind of quick notes for people who are listening. One is we actually did a whole... We did a six part series called Weapons of Influence where we basically... On the podcast, where we basically broke down each of the major pillars of influence and kind of dove deep into the research studies and the findings behind it. So, for people who want to kind of take that first step that Michael's recommending, that's a great way to get started. And the other thing, briefly, we also did a really cool episode recently on creativity, so, to kind of drill into some of this neuroscience behind that and how to spark your own creativity, for people who are listening.
Michael: Super cool. Super cool.
Matt: So, one of the things you touched on briefly was the idea of being numerate, and another way that I think Peter Bevelin called that in the book Seeking Wisdom is the physics and mathematics of misjudgment, and I know Munger did an amazing job in his speech about human misjudgment, kind of nailing all the different psychological factors. But two of the things I think that you've done an incredible job of really studying and explaining, Michael, are the concept of base rates and the concept of reversion to the mean, and I'd love to drill into talking about both of those, and I know there's a lot to unpack in each one of those, but in a way that we could kind of explain them to a layperson that's never heard of either of those concepts why they're important and what they are.
Michael: Yeah. So, great. Great question. The base rate, it really comes from the work of Kahneman and Tversky, so Danny Kahneman, Amos Tversky. They were examining how people... Well, actually, the ideas precede that by many decades, but they sort of codified this to some degree. And the idea is that there are two ways of making forecasts of the world, what they called the inside versus the outside view. So, the inside view--and Matt, this is how you and I typically operate, right. You know, if I give you a problem, you give me a problem, our classic way to solve it is to gather a bunch of information, right, combine it with our own inputs, and then project into the future, right. So, if you go to a college student and you say, "Hey, when will you be done with your term paper?" they sort of think about what their calendar looks like, how hard the paper is, and so forth, and they make some sort of projection. So, that's the natural way to think. The outside view, by contrast, we're calling the base rates, says, you know what? I'm going to think about my problem as an instance of a larger reference class. Basically, in plain words, I'm going to ask the question, what happened when other people were in this situation? Right, and it's a very unnatural way to think for two reasons. Number one is you have to leave aside your own information, this cherished information that you have, and second is you have to find and ultimately appeal to this base rate. So, for example, in our term paper example, instead of saying, "Hey, when will you finish your term paper?" and the student thinking about their own schedule and the difficulty of the paper, you basically ask a question of all the students who had a term paper due a certain day, when did they actually complete it? It's a very different question, and it turns out that what we see in the decision making literature is the introduction of base rates actually massively sharpens the quality of forecasts. So, we've applied it very specifically, for example, in the world of business to things like sales growth rates for companies. So, you might say, you know, hey, here's a company that has 10 billion dollars in sales. What's the sales growth rate going to be for the next three years or five years or ten years? So, you could model it. Again, bottom up. Sort of say, "Here's what they do. Here's how many new units they'll sell," and so forth. Or you can ask the question of companies of that same size over time, "What's the distribution of growth rate?" So, they're not mutually exclusive. Both of them go together, but that's the idea of base rates. And so, once you start to think about base rates, you start to see them, they're basically everywhere. But certainly realms like sports, realms like business, we have really good data on base rates and I think they can be really, really helpful. Reversion to the mean is another concept that is really important, and I think very, actually, quite tricky. So, reversion to the mean formally says that outcomes that are far from average will be followed by outcomes with an expected value closer to the average. So, the classic example of that is heights of people, right. Heights of fathers and sons, for example, specifically. So, what we know is that very tall fathers have tall sons, but the heights of the sons are closer to the average of all the sons. And likewise, short fathers have short sons, but again, the heights of the sons are closer to the average. So, there's sort of a squishing back toward the middle. So, that's an effect that happens, right, and it's just a statistical artifact. By the way, on the height thing, for instance, that sort of has to be true, if you think about it for a second, because otherwise there'd be people walking around who are 20 feet tall and two feet tall. That doesn't happen, right. So, here's an interesting way to think about the reversion to the mean, how powerful the force will be. So, if the correlation from one event to the next event is basically zero, then you should expect very, very rapid reversion to the mean. Let me give you one really concrete example from the markets. It turns out if you look at the standard [INAUDIBLE 00:19:52]500s. They're the most popular index in the U.S., and you look at the results from year to year. So, you take on X axis t=0, like what it did last year, and then on the Y axis, t plus one, what it does in the subsequent year, and you plot that going back to the 1920s. The correlation is basically zero. In other words, what happened last year tells you absolutely nothing about what's going to happen the subsequent year. So, as a result, the best estimate of what's going to happen next is some measure of the average, right. Reversion to the mean. And so, your best estimate for the market is basically the historical average. On the other extreme, if the correlation is perfect, very high, you expect no reversion to the mean at all. So, Matt, if you and I ran a sprint against Usain Bolt, he's going to win, right. And when we run again, he's going to win again. It's going to be perfectly correlated that he's going to win every single time, and there is no reversion to the mean. So, how we finished in prior races or how he finished in prior races doesn't really make a difference. He's going to win every single time. So, this idea of reversion to the mean, you can think about how correlated outcomes are over time. That also gives you an idea of how rapidly that idea of reversion to the mean takes effect. So, super powerful, super important, and often really overlooked. Even people who do this for a living--for example, sports executives--somehow get tripped up and don't fully take into account reversion to the mean.
Matt: And one of the things that I really struggled with, and I've read your chapters, and a bunch of Kahneman's stuff over and over again. I've read your chapters in The Success Equation five or six times, trying to really drill that concept into my head as the relationship between correlation and reversion to the mean. And also, you know, kind of going back to the simplest example is flipping a coin, and when people think about reversion to the mean, sometimes if a coin comes up heads four times in a row, people think, oh, I'm due a tails, right. But that's actually a completely incorrect way to think about and really understand how reversion to the mean actually functions.
Michael: Yeah, exactly, and I think that... Look, one of the reasons it's so challenging is because we have intuitions about how all this stuff works, but if we want to be slightly more formal, exactly what you said. So, when correlations are low, reversion to the mean is very, very powerful, and that's my stock market example. When correlation is very high, reversion to the mean is not a powerful force. In other words, what had happened before is, for the most part, a pretty good estimate of what's going to happen next. And yeah, no. By the way, that little heuristic, that's one of our tools in our toolbox. That's a mental model. It's an incredibly powerful mental model and, remarkably, very few people get it. The other thing, you know, Kahneman talks about this, but one of the other reasons that reversion to the mean is difficult is because our minds are wired to seek causality. If I give you an effect, some sort of an outcome, your mind is going to try to come up with a cause to explain it. And reversion to the mean is a concept that really has no cause and effect. And I'll give you an example that I always find to be fascinating. It turns out I mentioned before that the heights of fathers and sons, tall fathers have tall sons, but the heights are closer to the average of all the sons. But it turns out, and this is somewhat counter-intuitive, that if you plot the heights of the sons, it turns out very tall sons have tall fathers, but the heights of the fathers are closer to the average of all the fathers. And we know that sons don't cause father's, right. So, it gives you pause. You sort of say... So, in other words, the reversion to the mean has no arrow of time, and the notion of causality really doesn't apply. It's just it applies any time you have two series that are not perfectly correlated with one another. And by the way, the heights of fathers and sons, the correlation's almost exactly .5. So, in other words, if you're six inches above average, the best estimate of your son's height would be three inches above average, half the distance between your height and the height of everybody else. So interesting, right. So, I applaud you for going back to the concept. I did the same thing many, many times, going back to it, and there are some other people besides Kahneman who talked about it effectively. I just think it's a really hard concept to get your head wrapped around and it also is worthy of a lot of study.
Matt: I think the trickiest part is the very counter-intuitive notion that there's no cause and effect. That's what people think that it means, that there's some kind of cause that it's going to happen, cause something to happen, when in reality there's no arrow of time, there's no causality at all.
Michael: Yeah. So, I would say, Matt, to be a little bit more careful about it, it doesn't mean the causality isn't part of it. It just doesn't require causality, right.
Matt: Yeah, that's definitely a better way to say it.
Michael: So, the example I give that also... Well, I'll give you a quick story on this. I was presenting to... it was actually an academic conference, and it was on behavioral strategy. Super interesting. So, these are professors of strategy, corporate strategy, who have a behavioral bend. Super interesting topic. So, I was doing a presentation a little bit on luck and skill stuff, and I showed them a very classic, well-known picture where, if you take, say, 100--I'm just making this up--take 100 companies and you rank them in quintiles, so from top to bottom, and then you follow those cohorts, the highest returns on capital I'd say specifically to lowest returns on capital, and you follow those cohorts over time. What you'll see is the high return on capital returns go down and the low ones go up, which is exactly what reversion to the mean would indicate. So, I show that slide, and everyone's sort of, you know, amening and hi-fiving, and they all get that, right. But then I flip the data and I started with 2014 and I went backwards. So, I went from 2014 back to 2005. And again, what you do is you rank the companies on 2014 return on capital, again, highest to lowest, and then you follow those cohorts back in time. And what you find is the same picture.
Matt: That's wow.
Michael: So, it's clear for example that competition... So, you say, why would returns on capital go down over time? And the classic answer in economics is competition, right. So, if you're earning very high returns, maybe I'll come in and try to take part of your business away. That makes total sense. But clearly, competition can't work backward, right. So, it's the same idea that it's flummoxing, right, because competition is such a satisfactory answer as to why returns go down, but it doesn't really explain what we're after. Only partially explains what we're after. It's a really interesting point.
Matt: And I think that the mind invents the reasons why it's happening. Often it's just a statistical artifact.
Michael: Yeah. And that's the work... And that's another thing I would recommend. I find this to be almost infinitely fascinating, but the work by Michael Gazzaniga, who is famous for his work on split brain patients, so these are people that have suffered typically epilepsy and, to address these severe seizures, they sever the corpus callosum, the bundle of nerves between the two hemispheres of the brain. And what that opened up for Gazzaniga, Roger Sperry before him, was this opportunity to study modularity in the brain, and what Gazzaniga found was in the left hemisphere, where our language resides for most people, that there's a module they've now dubbed the interpreter, and the primary job of the interpreter is to find causes for every effect. So, it's a sort of cause and effect closing machine. And to your point, often in life, cause and effect are clear. You throw the rock at the window and it smashes. That's cause and effect, right. But the point is that if there's randomness, there's luck, going back to your coin tossing example, there's some sort of stochastic process, your mind is just going to make up a cause. It's fabricated, right, because it wants to close the cause and effect loop, and what Gazzaniga was able to show so brilliantly and so poignantly is that, with these experiments with these split brain patients, they could really isolate where this is happening and come up with these really fascinating results. And Gazzaniga wrote a book last year and he makes this point where... quite powerful, where he sort of makes this claim where he thinks that that module, that cause and effect connection, is the thing that distinguishes humans from other species most fundamentally, which is really interesting if it's true. So, I think that's a really important thing to keep in mind, too, is that our minds are constantly closing the cause and effect loops and it's not above any of us. We all do it and we just have to be very, very mindful of the stories that we're telling ourselves, because sometimes they're true and sometimes they're not.
Matt: And I don't know the specifics of those studies, but essentially, what they were doing, they had them open a door or something, right, and then the other hemisphere of the brain would invent a reason why they had done it or something, right?
Michael: Yeah, totally. Exactly. So, I mean, there are lots of different examples. They would show pictures or whatever it is, but one simple example, yeah, would be something just like that. They would flash some words to the left visual field, where it goes to the right hemisphere. Something that'll say, the patient sitting down will say "Stand up." So, the left visual hemisphere sees it. Right hemisphere connects. The patient stands up. So, it's interesting. Of course, the left hemisphere, the person knows that they're standing up. They have no access to that cue, but now the researcher will say, you know, "Patient, why are you standing up?" And the research is almost humorous. It's because these people would fabricate these sort of elaborate, crazy stories. You know, my left knee is sore and I want a stretch, or something like that, right. They would fabricate something that would sort of hold the whole thing together. But obviously, it was completely contrived. So, again, you get these chuckles as you see these things that these people are saying, but the more serious and fundamental point is that we're all doing it all the time and we're just not mindful of it. So, this is just shining a spotlight on something that we're all doing all the time. So, it's a really hard thing to do, but it's discipline to say, am I fabricating a narrative here or is this a luck-laden activity or a luck-laden field? Am I simply just capturing luck here and I'm making up a story to try to make for a cohesive world?
Matt: I think that's the critical point, is that just because... It's happening in the research, but the reality is it's happening every single day to everyone who's listened to this podcast, and both of us.
Michael: Precisely. Absolutely.
Matt: Well, I think that's a good segue into the idea of cognitive biases, and I know that's something you're very knowledgeable about. What are some of the most insidious or even some of the most common cognitive biases that you see people suffering from? And maybe specifically in the context of investing, or even broadly?
Michael: Yeah. So, there are really two things that I would mention in investing. There are many more. One of them, which is extremely difficult to sidestep, is confirmation bias. This is this idea that even if you struggle to make a decision--let's say buy an investment, buy a stock or what have you--even if you sort of struggle to come to that conclusion, once you've made a decision, we all have a natural tendency to seek information that confirms out point of view and to dismiss or disavow or discount disconfirming points of view. And one of the things we've learned, you know, certainly, and I think a lot of what we've been seeing in computer science the last 25 or 30 years has been strongly reinforcing, is this idea of updating information as new information comes in. So, it's a Bayes' theorem. So, you have a prior... you have a point of view of how the world works. New information comes in and, really, if you're doing your job properly, you should be updating your view, updating your prior, given this new information. And, unfortunately, the confirmation bias is this sort of huge brick wall that prevents new information from finding its way into your mind or finding its way into your decision making. So, that's the first one that's a really big one. The second one is probably overconfidence, and this is very trivial to demonstrate if you get a group of people. People tend to be very overconfident about topics that are a little bit away from their own bailiwick. So, if I give you questions that you know a lot about, you'll do fine, but things that are just a little bit on the margin from that, you'll tend to be overconfident. And the way that tends to manifest in an investing setting, for sure, is people tend to project ranges of outcomes that are too narrow. In other words, they think they understand the future better than they actually do, and they fail to consider possibilities, whether they're really good possibilities or really bad possibilities, and that's, I think, the more pernicious component of overconfidence. So, those are two that come to mind, but boy, you know, things like... We could go on and on. Loss aversion. So, we suffer losses more than we enjoy comparable-sized gains. That's a really big one that looms large in a lot of our decisions. So, there's a long list of them, but those two probably, confirmation bias and overconfidence, are probably the one-two that I would list first.
Matt: And what do you think are some ways that people can combat each of those?
Michael: So, confirmation bias is just really, the key is to be as open-minded as possible. Jonathan Baron at University of Pennsylvania's got this beautiful phrase. He called it actively open-minded, and this idea of really, truly trying to be as open as you can to new information or new input. And the second thing, I think it's very few people are going to be formal about doing something like Bayes' theorem, but understanding behind Bayes' theorem, which is, you have a point of view. New information comes in. Are you revising your view, both directionally the correct amount and the magnitude of the correct amount? So, those would be some ways to try to do that. Overconfidence, the key is to just... and we can go back to our discussion a few moments ago about Bayes rates, is just to continue to compel yourself to think about alternatives, right. I'll give you one example that's a very simple one. I joke with my students at Columbia Business School, often when there are stock recommendations, you know, you see someone on CNBC or something, or they recommend a stock for purchase, they'll often say, "Well, the upside is 30% and the downside is 10%." Something like that, so it sounds like three to one. Pretty good, right? But if you think about, just statistically for a moment, the standard deviation of the stock market, right, so how fat the bell shape is of the distribution of returns. It's about 20% standard deviation in the last 85 years or so. So, that's a diverse five portfolio, of course. So, the standard deviation of an individual stock is going to be higher than that. Let me just pick 30% just to make the numbers easy. So, the average stock, let's say roughly speaking, would be up about 10%, mean return, average, with a 30 standard deviation. So, just translate that into statistics. That would say that about 68% of the time, it's going to be between up 40%, right, 10% mean plus 30 standard deviation, to down 20%. So, 10% mean minus 30%. So, 40 to -20. So, I just joked about this 10 to 30 percent upside, 10% down. You know, just one standard deviation is wider than most analysts are willing to accept, and certainly going on two standard deviations, it's vastly wider. So, imposing this discipline on yourself to understand what the underlying distributions look like and to recognize, try to think about having ranges of the future that are wide enough. And then there are other techniques, which we could talk about, and I think you probably have covered some of these in some of your prior podcasts, but things like pre-mortems. So, these sort of structured ways to get people to think about different points of view are also some nice techniques to allow to do that.
Matt: You know, we actually use pre-mortems in our business, but it's not something that I've talked about at all on the podcast. I'd love for you to kind of extrapolate on that concept.
Michael: Sure. I mean, so most people know about post-mortems, right? So, in other words, the patient has died or something adverse has happened to the patient and we sit around as a medical community and say, given the facts that we had at the time and our technology, what could we or should we have done differently to get to a better outcome? And we're also very familiar with scenario forecasting. So, we sit here in the present. We peer into the future and say, "Here are the possibilities we should consider as we make a decision." A pre-mortem, as you've already gathered from the name, is a very different exercise. It actually effectively launches yourself into the future and you look back to the present. So, now it's June, for example, 2017 and we look back to today, June 2016. This was developed by a social psychologist named Gary Klein, and so, just to give props to him, he's the guy that developed this. And so, we can tie together two ideas here. So, here's the classic way to do this. You say, "Let's sit down. We'll meet in our conference room." I suspect this is what you guys do in your business. And you say, "We're going to think about making a particular decision." Let's say it's an investment decision or a business decision to expand or what have you. And what we're going to imagine, then, each of us, is that this decision turned out to be a fiasco. Total disaster. We're all embarrassed about it. But now it's June 2017, so it's a year from now. So, each of us is going to write a little narrative, write a little 200-word essay about why this decision turned south. And it's very important to do it independently, and it's very important to do it from the point of view of the future looking back to today, right. So, you might say, and then you combine the different inputs, and it turns out that that exercise tends to generate substantially more alternatives or scenarios than simply standing in the present looking to the future. And by the way, is that consistent, Matt, with your own experience in your own company?
Matt: Oh, yeah. Absolutely.
Michael: Yeah. And so, let's tie this back to the idea of the interpreter. You might say, "Well, hey, I'm looking at scenarios. I'm thinking about this already. Why is a pre-mortem adding value?" And the answer, I believe, is by launching yourself into the future, assuming that this particular outcome has occurred, what that does is it wakes up your interpreter, right. This little module in your brain, you've now given it a fact and you're saying, "Hey, interpreter, why did this go bad?" And the interpreter's like, "I'm up to this task," and starts generating particular causes for it, right? So, in a sense, your scenario planning, standing in the present, future, the thing isn't done. So, you're not really thinking about causes in a very rich sense. And the second, the pre-mortem, you're basically recruiting your interpreter, in a sense, to help you understand scenarios more richly. Isn't that cool? So, I think that's part of the psychological reason why pre-mortems, I think, can be more effective than simply scenarios. And, you know, my experience is very consistent with yours, that organizations that have adopted, embraced pre-mortems tend to report that they have much richer discussions, much more heated debates, and ultimately probably make better decisions as a consequence of going through the exercise.
Matt: Another related concept that we've used a number of times is something from the military called a Red Team. Have you ever heard of that?
Michael: Yep, absolutely. So, we wrote a piece about decision making, and we talked about different things. So, we talk about Red Team, Blue Team very specifically. And, you know, you may have mentioned this before, but red team typically is attacker, blue team is defender. I think today, one of the good... it's from military strategy, of course, but today, one great example, very relevant example is cybersecurity. So, you might say, "Hey, chief technology officer, are we protected from cyber-threats?" And he or she may say yes, but you might hire a hacker to be your red team, so to challenge yourself to see where your vulnerabilities lie. And so, red team... And, by the way, this was my prior job. If we had a particular investment that wasn't working out well or a thesis that didn't seem to be unfolding, we actually would do this, that you'd assign some people to go off and develop the counter case, the devil's advocate case. You'd have people defending the point of view of the firm and we just let people sit across from each other at a conference room, and everybody else would be judge and jury and we'd let them go at it, which was great. I'll tell you the one thing that I learned. A couple of things that I would just add onto that. One is that in Red Team, Blue Team, I think it's really important to distinguish between facts and opinion, and I think a lot of our discussions in general, by the way, we tend to not distinguish as carefully as we should or could between facts and opinion. So, this is a really interesting exercise I'd recommend all the listeners to do, if they have a few minutes, is to pull out an article. For example, something you either really agree with or something you really disagree with, right. So, something that's really polarizing for you. And then take two different color highlighters, say blue and yellow, and with one color, highlight what you would deem to be facts and then another color what you would deem to be opinion, and then simply step back from the document, and whether you agree with it or disagree with it, try to have a balanced assessment as to whether you're being persuaded or not persuaded by fact or opinion. That's super cool. The second thing I'll mention, which was a new thing for me, is that Adam Grant's a great professor at University of Pennsylvania, and he wrote a book called The Originals. I don't know if you guys talked about that. There's some stuff on creativity in there, as well.
Matt: Have not.
Michael: But Adam talked about Red Team, Blue Team, and he actually made a point that I didn't appreciate fully until I read it. And he said, "If you're assigning red team responsibility in your organization, what you want to find is someone who really doesn't believe in the thesis." You don't want to just say, "Hey, can you be the devil's advocate?" You want someone who actually doesn't believe in the thesis, someone who really is the devil's advocate, and he just says that enriches the dialogue greatly, versus having someone that's sort of an innocent bystander, grab them by the collar and say, "Go tell us why you're against this." So, that was another little wrinkle that I just learned about, which I think could add a little value in the process.
Matt: And another tool that I know you're a big advocate of our checklists. Can you talk a little bit about that, how important they are and how they can improve decision making?
Michael: Yeah, absolutely. You know, I was really inspired, and I think many others, originally, by Atul Gawande's article in The New Yorker, which ended up being a book, The Checklist Manifesto. But the protagonist of that original New Yorker article, and to a large degree, the book, is a guy named Peter Pronovost, who's a doctor at Johns Hopkins. And, actually, we had a conference a number of years ago where he invited Pronovost to come in. And the story's nothing less than astounding, where Pronovost basically... And by the way, he had lost his father to a medical error, so it was very real and very personal for him. Where Pronovost basically introduced a very simple five-step checklist for putting tubes in, intravenous tubes, and found that they could massively reduce infection rates, saved lots and lots of lies, and I think Gawande in the book argues that Pronovost may have saved more lives in the United States than any other person in the last ten years or so. So, this sort of informs us that... By the way, doctors, if you ask them what they need to do before putting a tube in, they know what to do. It's not like their lack of knowledge. It's really a lack of execution. And so, I think the point that Gawande makes in the book that I think is so powerful is that in every field where this has been studied, be it aviation, medicine, construction, a faithful... First of all, coming up with a good checklist and a faithful use of the checklist has led to better results, and this is without making the underlying users any smarter or any better trained. So, it's just hewing to the process more accurately, which is really fascinating. So, I think a lot about this in the context of investing. Now, investing is a little bit of art and a little bit of science, and I think where the checklists really do apply very effectively is in a lot of the process-oriented stuff. So, how to do certain types of calculations. Basically, it's sort of the fundamental components of investing analysis. Now, the art part comes into some other elements of interpretation, but I would just say if you have components of whatever job you do, and I think almost all of us do have components that are somewhat algorithmic, where consistency and accuracy are really, really helpful, you should be thinking about, if you're not doing it already, developing and applying checklists. Gawande's book is fantastic. Pronovost, by the way, himself, wrote a book about this topic, and maybe the last thing I'll say that came out of Pronovost's book, which I think is very important, is that he said one of the keys to checklists succeeding is actually gathering and analyzing data. In other words, being scientific about this, not sort of just a nice idea of having a checklist, and I think that was one of the keys to Pronovost's original wild success as a Johns Hopkins, was not just that they developed a proper checklist but they figured out ways to get the doctors to use it, and then they really kept track of it and gave the doctors feedback. And so, this idea of data collection and feedback is also a really, really key element to this whole thing.
Matt: Changing directions a little bit, I'd love to dig into some of the stuff you talked about in The Success Equation, kind of untangling luck from skill and the concept of the luck-skill continuum. One of the tools or mental models that you use to describe that phenomenon was the two jars model, which I found to be extremely helpful. I'd love for you to kind of explain that a little bit.
Michael: Sure. So, you know, and by the way, luck, skill, the whole topic of The Success Equation, it had been sort of lurking in the shadows for me for many, many years. I played sports in college and high school and a sports fan. Clearly a big deal in the world of investing, and also if you look at corporate performance, it's almost everywhere you look, this idea of luck was sort of there, but hard to pin down. And I read Fooled by Randomness by Nassim Taleb in 2001. That certainly got me thinking more about that, and I think Taleb does an incredibly effective job in that book of sort of underscoring the role of luck, but didn't really do much to help us quantify a lot of this. So, the cornerstone of the book, as you point out, is called the luck-skill continuum, and the way to think about this is that you just draw a line and on the far left you put activities that are pure luck, right. So, roulette wheels or lotteries, where really, there's no skill whatsoever. And on the far right, you might put pure skill activities. And things like maybe... a lot of things. Pure skill, but running races, or chess is probably over there. And then, just thinking about arraying activities between those two extremes. So, where does a basketball game fit on that? Where does bowling? Whatever it is, right. So, that in and of itself, the methodological approaches to trying to do that was really, really interesting. But, as I got into this, as you point out, I was trying to think about conceptualized the so-called two jar model. So, the idea is that your outcome for whatever activity is going to be the result of drawing a number from a jar filled with numbers for skill, and then drawing a number from a jar that's got luck. Right, so you're going to pull two numbers out, add them together, and that'll be your outcome. Now, if you're on the pure luck side of the continuum, for example, you'll have a luck distribution. You can envision it as a bell-shaped distribution, is fine. And your skill jar is filled with zeroes, right. So, only luck will make a different. If it's on the pure skill side, you know, you have a skill distribution and you're drawing zeroes from luck, so only skill matters, but almost everything in life is sort of these two rich distributions colliding with one another. And the question is, how much is each contributing? So, I just think that's... And by the way, one of the really nice things about the two jar model is it allows us to understand to some degree things like reversion to the mean, which we spoke about before. It allows us to appreciate the fact that great outliers--for example, streaks in sports of consecutive hits in a baseball game or consecutive shots made by a basketball player--are always, and almost by definition, going to combine great skill and great luck. Because, if you think about it for a second, that has to be true, right. Not all skillful players have the streaks in sports, but all the streaks are held by skillful players, right, because skill is the prerequisite and luck comes on top. So, to me, it's just a very, very vibrant way to think about a lot of things in life, and the key point of The Success Equation is not just thinking about these topics, but hopefully providing some people with some ways to think about the concrete, how they have to deal with the world differently concretely, as a consequence of understanding the role of luck.
Matt: And one of the things that I'm really fascinated with is the concept of deliberate practice, and you touch on that and how it relates to and applies more specifically in skill-dominated systems. But I'm curious, you know, how would you think about applying something like deliberate practice, or maybe the core lessons behind deliberate practice, to a field like investing or business or entrepreneurship?
Michael: Yeah. Super interesting. And so, deliberate... I don't know if you've... There's a brand new book by Anders Eriksson called Peak, on this...
Matt: I have not heard of it. I'll have to read that.
Michael: Okay, yeah. Check it out. So, Anders Eriksson just wrote a book called Peak, just as it sounds, which I just read a couple of weeks ago. So, that is his... you know, talking about deliberate practice, just to reiterate for all the listeners, deliberate practice is this idea of practicing that is at the cusp of your ability, so a little bit at or right beyond your ability, often where you have a teacher or coach, someone who can give you instruction, and you're getting quality feedback. So, you're proving at the cusp of your skill level. So, as he points out, a lot of us practice things, or we do things that's like we practice. We do things over and over, or even we practice but we don't really satisfy the requirements of deliberate practice. It's usually not beyond our or at the edge of our capability. We often don't have coaches. We often don't get the quality feedback. And, as Eriksson expresses it, deliberate practice is not a whole lot of fun, right. It's actually very tiring, because you're constantly pressing yourself. So, I wrote a piece about this actual topic of deliberate practice and 10,000 hours back in 2004. It came before Gladwell's book and so forth, and I've struggled since that moment of writing that piece about what deliberate practice means. What is this idea of working beyond our boundaries and getting feedback and so forth? So, I don't know that there's a perfectly good example of that, so maybe I can make two points. One is what I argued in The Success Equation, is skill improvement or skill development through deliberate practice is absolutely valid in fields where your output is an accurate reflection of your skill. So, what kinds of things would that be true? It would be, you know, music, if you're a musician. Athletics, it would be true. Chess playing, it would be true. So, there's certain fields where the output is an accurate indicator. There's very little luck that's filtering out the outcomes, right. So, that's where deliberate practice really is good. As you slide over to the luck side of the continuum, what happens is the connection between your skill and the outcome is colored greatly by luck. So, for the example I gave Matt that's a trivial one is, if you're a blackjack player and you enjoy playing blackjack and you go to Atlantic City, you may play properly with standard strategy and lose badly for a few hands, or you may play very foolishly and win for a few hands, right? So, this connection between your skill and the outcome are broken. And when that's the case, what I argue is you should focus almost exclusively on process. And process, it's got elements of deliberate practice, but process is going to have three components, as I would argue for it. One is an analytical component. That is both trying to find situations where you have an advantage and also how do you bet, given your advantage. I'm going to call the second component behavioral, and this covers a lot of what we've been talking about today, but are you aware of managing and mitigating the behavioral biases that we all fall prey to? And the third I'm going to call organizational, which is we all work for companies or parts of organizations or parts of teams. None of them are perfect. Agency cost can be a very big deal. What are we all doing collectively to minimize those organizational drags, right. So, to me, it becomes very process-oriented, and I think if you look at the elite performers, whether it is in sports betting or even sports team management or investing, you get a very common thread, that those folks are almost always and almost exclusively focused on process in the faith, the full faith that a good process leads to good outcomes over time.
Matt: I think that's great advice and that's something that I've struggled with a lot, is kind of how to reconcile that or how to deal with the challenge of getting whether it's accurate feedback or whatever else it might be in systems where there's a very fuzzy relationship between skill and outcome.
Matt: So, you've touched on this a little bit, but if you had to kind of distill it, what would one piece of homework be that you would give to the listeners of this episode?
Michael: Read. [Laughs] Read is probably the main thing, is to... And I actually say that I think working with people like you or following people like you is a great place to help curate some of this stuff, but I think it probably helps to have some thoughtful people. Shane Parrish, you mentioned, was fantastic.
Matt: He's great.
Michael: And Shane's another guy who can help you curate that stuff. But I think starting to just...making sure that you commit a substantial percent of your day to learning, continual learning, and, again, being diverse in what you're reading and thinking about; and forcing yourself, compelling yourself to have the stance of being actively open-minded, so making sure that you're considering different points of view, you're exposing yourself to different types of people. So, that maybe not. That's maybe a tall order, but, to me, that would be the first thing I would say. And, you know, I do find a lot of people struggle to find time--or at least they perceive they struggle to find time--to read, and the main thing I would just say is that life is about tradeoffs. So, the question is: Are there things that you're doing today in your moment to moment that you could trade off, that you could do less of, that would allow you to do more reading? Because I do think the return on investment is really, really... The return on time and the return on investment is really high for reading.
Matt: You know, there's a really funny study that Zig Ziglar talks about in some of his old speeches. And I think the study was in the '50s or '60s, but they basically looked at...they looked at a factory and they started with everybody from the factory workers up to the line managers, up to the office managers, up to the president, and they looked at how many hours a week they each spent watching TV. And there was sort of a relationship where, you know, it's like the factory worker spent 20 hours a week watching TV, all the way up to where the president spent half an hour a week watching TV or something. So, that's a great point, is that there's always a way to find time to read if you make it a priority.
Michael: That's right. Exactly. And I love that. And, again, it's maybe not everybody's cup of tea, but for people who are probably listening to this, it is going to be something that they'll find interesting and I would just jump in. And I would also encourage... Especially for young people it's a great thing to get going on. When you can work it into your habits when you're young, it's just a huge leg up through the years, for sure.
Matt: I mean, obviously you're a very active reader. Do you have any kind of methodology that you use to keep track of all of your kind of book notes or to keep...to sort of categorize everything that you've read and all the knowledge that you've accumulated?
Michael: [Laughs] So, Matt, I wish I had a good answer to this question. The answer is no, not so much. But I guess I...
Matt: I struggle with that, too. That's why I'm asking -- for myself, in this case.
Michael: [Laughs] But I benefit from a couple things, which are sort of offshoots of the way my career works. So, I have the fortune of being able to write a fair bit for my job and not just book stuff or just day-to-day stuff, and so that allows me to weave in a lot of the stuff that I read and implement it, and I think teaching and writing are two really powerful mechanisms to help consolidate thinking and consolidate ideas. So, that helps a lot. And, beyond that, it's just... Now, a lot of it is cumulative, right? So, it's just trying to make sure that whatever I'm reading clicks into place. I mentioned this Anders Ericsson book and, you know, I've been reading about... I think I have probably a half dozen books or more on expertise. Many of them were edited by Anders Ericsson. So, that was just adding onto something that I had a little bit of a foundation in. So, yeah, there's not much method to my madness, but I'm not sure that... Yeah, I'm not sure... I think just jumping in is probably the first and foremost thing to do.
Matt: Where can people find you and some of your works online?
Michael: So, probably the easiest thing to do is go to michaelmauboussin.com. So, that's a website that mostly highlights the books that you mentioned at the outset. The Success Equation, our skill lookbook, also has its own website, which is success-equation.com. Success-equation.com is also kind of fun because there are some interesting little simulations that you can play around with, including the two jar model you talked about. There's also some fun stuff on the Colonel Blotto game, which is a game theory model, and a little mind reader algorithm. So, there are some fun things to do there as well. And then it's harder... My professional writing is difficult to get access to through formal channels, but if you've got some fingers in Google, you can tend to find a lot of the stuff on there. So, I would just google it. [Laughs]
Matt: And I think valuewalk.com has a great list of a lot of your...a lot of your pieces.
Michael: Yeah. So, ValueWalk's a good example. Yeah, exactly. And Hurricane Capital's done a great job. So, a couple of these sites, those guys do a nice job of recapturing a lot of the stuff we do.
Matt: Well, Michael, thank you so much for being on The Science of Success. It's been great to have you and it's been an enlightening conversation.
Michael: Matt, it's been my pleasure the whole time, so thank you for having me.