The Exchanges

Q That is very, very kind of you. I can't believe I actually get to do it for a job, to be honest, but I want to start with you. Um, you're, you're a celebrity in venture circles, um, and I want to start with a question from Bill Gurley, which is like, what do you do, Michael, and how do you spend your time today?

A Yeah, exactly. Well, my title is Head of Consilient Research for Counterpoint Global, which is part of Morgan Stanley Investment Management. We're a long only investment management firm, also have a Ventures Investment Fund as well. But my job is really three parts. One is to work with our team on investment process. So things like, how do we think about markets? How do we think about valuation, competitive strategy, decision making? The second part of the job is really doing research, and that is thinking about topics and writing about them. And then finally is external things, things like conferences or, or podcasts or things like that. And the way I like to describe what I do is input and output. So try to find something that's interesting, spend some time to understand it, and then communicate it both internally and externally. So it's a, it's for me, it's a very fun and gratifying job. And, uh, allows me to, uh, to pursue a lot of interesting things.

AI assessment note: “my job is really three parts. One is to work with our team”

Q You say about input and output there, and that makes me think about kind of decision-making and process, but you've said some brilliant things before, especially on kind of luck and randomness, and you said randomness and luck are related, but there's a useful distinction between the two. What's the distinction between luck and randomness, Michael?

A It's a great question. I mean, let's start by defining luck, and by the way, this gets into philosophy very fast, but just to be trying to be practical. I'm going to say luck has three conditions. It happens to an individual organization, so to you or your company or your favorite sports team. Second is it can be good or bad. Don't mean to suggest that it's symmetrical, but there's a positive sign potentially and a negative sign. And then third, and this is important, it's reasonable to expect a different outcome could have occurred, right? So if we replay the tape of time, reasonable to expect a different outcome. So the way we think about randomness versus luck is randomness would be at the system level and luck would be more at the individual level. So, for instance, every year in my class, I do a little exercise to demonstrate this point. I flip a coin, and I ask the students to guess whether it's going to be a head or a tail, right? We do a bunch of them in a row. So what I know is given the size of the class that someone's very likely to get four or five in a row correct. The coin tosses themselves, assuming I actually, I actually do it with a simulator because I'm so bad at it, but the coin toss themselves, I'm going to deem to be random, but if someone, if you're the one that calls four or five in a row correctly, I would call that to be, I would call that luck. So one …

AI assessment note: “one would be system level, one would be more individual level”

Q And then I bring them all together. And then in the meeting, I can draw on that and say, oh, Michael, You actually said you really didn't like this for that reason. Talk to me about that, and you wouldn't have probably said that if it was in a collaborative doc where people are chiming in. Are there any other ways that you've seen it really work and bring out?

A No, that's, that's fabulous, Harry. That's fabulous. And the other one, I mean, the other technique is known as a premortem, which you're probably familiar with, which is we pretend we made the investment. It's now a year from now, or two years from now, and this investment's turned very sour. It's very bad. We're on, on, on displeased with it. And then each of us writes down today why basically the, the story from the financial times and wall street journal that's published two years from now, why this didn't turn out well. Right. So again, it's the same, the same mechanism that you just described, which is you're encouraging people to think on their own. Why this, you know, sort of what the downside is in these types of scenarios. And, and that again, surfaces potential problems. We have to think about upside as well, but surfaces potential problems, which are then we can identify and see, uh, A, are these risks worth, worth taking? And B, if these, some of these eventualities occur, how do we, how do we take care of them or, or deal with them in a way that allows the investment to continue to work out well?

AI assessment note: “the other technique is known as a premortem, which you're probably familiar with”

Q It's coming to the incredibly unfair moment when I ask you, I love also your background, and then my background, like books, books, books, nothing. Um, what's your favorite book, Michael?

A Um, can't, can't give you an answer to that. No, I'm just kidding. I, it, there are too many, but the one I would say probably is, I mentioned I'm a head of Consilient Research, uh, is a book called Consilience by E.O. Wilson. E.O. Wilson was a famous biologist. He died recently But consilience is about the unification of knowledge, and the argument that, that Wilson made that I, I'm very taken with this argument is that much of science is advanced through reductionism, so breaking things down into their components and understanding how they work, and many of the wonders we see around us are the consequence of that. Consilience is the unification of knowledge, and the argument he makes is many of the vexing problems we have in our world will require, uh, combining different disciplines together to try to solve them. So, so to me, consilience, and I've been very involved the last 25 years of the Santa Fe Institute, SFI is also very dedicated to this kind of a concept, so Consilience would be my book.

AI assessment note: “Consilience would be my book”

Q And then I bring them all together. And then in the meeting, I can draw on that and say, oh, Michael, You actually said you really didn't like this for that reason. Talk to me about that, and you wouldn't have probably said that if it was in a collaborative doc where people are chiming in. Are there any other ways that you've seen it really work and bring out?

A No, that's, that's fabulous, Harry. That's fabulous. And the other one, I mean, the other technique is known as a premortem, which you're probably familiar with, which is we pretend we made the investment. It's now a year from now, or two years from now, and this investment's turned very sour. It's very bad. We're on, on, on displeased with it. And then each of us writes down today why basically the, the story from the financial times and wall street journal that's published two years from now, why this didn't turn out well. Right. So again, it's the same, the same mechanism that you just described, which is you're encouraging people to think on their own. Why this, you know, sort of what the downside is in these types of scenarios. And, and that again, surfaces potential problems. We have to think about upside as well, but surfaces potential problems, which are then we can identify and see, uh, A, are these risks worth, worth taking? And B, if these, some of these eventualities occur, how do we, how do we take care of them or, or deal with them in a way that allows the investment to continue to work out well?

AI assessment note: “the other technique is known as a premortem”

Q It's coming to the incredibly unfair moment when I ask you, I love also your background, and then my background, like books, books, books, nothing. Um, what's your favorite book, Michael?

A Um, can't, can't give you an answer to that. No, I'm just kidding. I, it, there are too many, but the one I would say probably is, I mentioned I'm a head of Consilient Research, uh, is a book called Consilience by E.O. Wilson. E.O. Wilson was a famous biologist. He died recently But consilience is about the unification of knowledge, and the argument that, that Wilson made that I, I'm very taken with this argument is that much of science is advanced through reductionism, so breaking things down into their components and understanding how they work, and many of the wonders we see around us are the consequence of that. Consilience is the unification of knowledge, and the argument he makes is many of the vexing problems we have in our world will require, uh, combining different disciplines together to try to solve them. So, so to me, consilience, and I've been very involved the last 25 years of the Santa Fe Institute, SFI is also very dedicated to this kind of a concept, so Consilience would be my book.

AI assessment note: “is a book called Consilience by E.O. Wilson”

Q Can I ask, what do you think makes Bill Gurley so special and the incredible ambassador that he is?

A So I think, first of all, he's very bright, and he's very intellectually curious, um, and It is interesting. So I met him right when he got out of business school and joined. We were in the same research department and right away I could see, I started talking to him about certain frameworks and ideas, and he immediately not only understood them, but integrated them. So two big ideas. One is he, he understood, we were talking about DCF and everything's a DCF model. I think he's very early on understanding the importance of return on capital investment returns, cash flows, and he applied it right away. Very successfully, by the way, in investments. The second is he was very early on and understanding Brian Arthur's work on increasing returns, understanding the mechanisms of increasing returns and economics. We're taught the returns on the margin go down. Decreasing returns is a consequence of competition and maturation. But under certain circumstances, you get these increased returns. So he was very early on. So, um, and by the way, not not only a good investor, but also a very good communicator. So he writes very well. And, you know, you, you, anytime you see or give it, you know, see the talks that he gives, they're very thoughtful, very well prepared. So yeah, it's, it's a combination of all those things.

AI assessment note: “he's very bright, and he's very intellectually curious”

Q What mistake have you made in the investment world that in hindsight you wish you hadn't made?

A Yeah, I mean, those are easy because she just took things that turned out badly, but I'll mention one investment, which is I bought Sears Holdings. Um, Eddie Lampert is a C, was a CEO, and I knew Eddie a little bit at the time. I really liked the way he thought. I liked what he was trying to do with the business. It ended up being turned, turned out very badly. Um, I'm not sure it was, it was a mistake in the sense that it turned out badly. Um, but it's interesting because I liked his process, and I was saying, like, I'm gonna make a bet with the guy. Didn't turn out well, but that was, that would be one example.

AI assessment note: “I'll mention one investment, which is I bought Sears Holdings.”

Q Oh, a hundred percent. Um, I think it's the challenge that all spin out GPs face when they try and raise their own funds. Um, when we think about that, though, then, is investing in or building a company on a novel premise, is that a luck-based endeavor?

A I don't, I mean, I don't think so, of course. I, I, I think that you, what you strive to do is to develop a process that you think will allow you to have the most, uh, best chance of success over time. And one of the things I would just say, Harry, about building a process is the first thing to think about and, and think about deeply is actually what you're trying to do. Like, how do you think you're gonna achieve some sort of an edge versus other people doing it? And in effect, the process you develop is, is, should be something that contributes to that source of edge, you know? So for example, in public markets, you'd think about the two extremes would be, you know, Jim Simons at Renaissance technology, right? Tons of technology, very smart people trading very frequently. And on the other side of the continuum would be Warren Buffett, who reportedly sits in his office all day and reads and then makes these very few, but very consequential investment decisions. Both of those folks have been very successful over time, but their approaches and their processes are very different. So the key is that there is no one size fits all for process, I think. But rather that you want to make your process congruent with your perceived source of edge. And so that to me is sort of the key thing to bear in mind. It's, it's, it's often one thing people say, here's what I do all day. Usually whe…

AI assessment note: “I don't, I mean, I don't think so, of course.”

Q Can I ask you, what makes bad process, or where do you see most people make mistakes in that process?

A Yeah, I mean, it's usually two things. One is a deviation from what you say you're trying to do, right? So you just laid out kind of how you're approaching it, this combination of media and venture investing and so forth. You know, the question is, are you starting to do things that deviate from those things, or you bring in these three founders, your, your little investment committee, and, and they really have expertise and different points of view and you override them for whatever reason, right? So you just depart from process in a way that's, um, non-systematic. And by the way, it's not that hard to do because sometimes you may have a process and it's, you have a couple things that don't work out, right? Which is gonna be naturally happening. So you start to question what you're doing. So you deviate. That's, um, that's one problem. The other one is just letting biases creep in to what you're doing, right? So biases would be things like, and the, and these, you may be emboldened by success, by the way, but two big sources of bias, I think are one is overconfidence. And when you look at overconfidence, usually the, The, the, the sub-component of oral confidence of the problem is over precision, which is you start to think you know what the world is going to be rather than better than you do, and the other one is confirmation bias, which is sort of, you know, when you have a …

AI assessment note: “it's usually two things. One is a deviation... The other one is just letting biases creep in”

Q Speaking of the episodic nature of returns there, Doug Leone told me on, from Sequoia, Harry, just never sell your winners. And, uh, you know, I, I, I took that very much to heart, um, and I didn't sell some winners at the time. Um, that was a mistake. How do you think about, uh, Position exiting. And when's the right time to take the chips off the table?

A Yeah, I don't know if there's a good answer for that. I mean, part of it is at some point, if a company gets mature enough, you should be able to start to assess value on some, in some way, shape, or form. You mentioned early on having these models might be silly, and I don't disagree with that. But at some point, if there's some point of maturation, and certainly if the company's gone public or even It's gotten some degree of maturity. You can start to think about, uh, you know, values or ranges of values that you think are appropriate, and, and the answer is, yeah, when the company meets or exceeds that value you think is reasonable, that then you should scale, scale out of it. Now, I'm sympathetic to the argument is, you know, this is work by Hank Bessenbinder demonstrates that in public markets over long periods of time, most companies, public markets earn Returns less than the treasury bill rates. And it's like a handful, less than five percent of companies basically drive all the value for the stock market. So that's the argument for latching onto the winners and holding onto them forever. The premise is that you can find them and you know, you own them and you know which ones are gonna be the winners. Um, so that would be the argument in favor of doing that. So that's, that's finding the whatever, you know, apples when they're turning or Amazons or Microsoft when they're…

AI assessment note: “when the company meets or exceeds that value you think is reasonable, that then you should scale”

Q Do you look at VCs today? I'm just fascinated. Do you look at VCs and see like, thirty million price for like, you know, two people and a dog in a garage with nothing and go, what are these people doing?

A Well, I mean, yes and no. I mean, the one thing I would say that You know, there's a, there's a, I found to be a fascinating, not surprising, but fascinating academic paper published a couple of years ago where they looked at the distribution of returns for 30,000 venture capital investments. So these are individual investments. And then they looked at 15,000 buyout investments to look at the, the distribution of returns and, you know, in venture, you know, I'll just say what everybody already knows. The majority of investments lose money. Right. But a handful of investments make so much money that they more than common, you know, make up for everything else, right? So the overall returns for venture over time have been pretty good, mostly because it's a few companies pulling all the, all the freight for the whole thing. Now, that said, when you say, oh, thirty million dollars for two people in a garage, the key is even those investments that lost money were deemed to be positive expected value investments, right? They had option value. They had potential. We know in advance. So it's not, it's not how often you make money that matters. It's how much money you make when you're right. So on one level, it seems, you know, if it's, if, if the ideas are not good or the people are not good, that would be problematic. But on the other hand, if it's a positive expected value going into…

AI assessment note: “even those investments that lost money were deemed to be positive expected value investments”

Q people, you have students, you know, this is unprecedented market times, and you have a lot of young investors going, oh, am I any good at this? I thought I was, but this is a very new world, and this is uncertain times for someone who's never done it before, or seen it before. How do you advise younger investors looking at the markets today questioning whether they're any good?

A Yeah, I mean, well, it's sometimes hard to know if you are any good. The key is, are you sticking to your process and trying to think about things the right way? And in markets like this, there are a couple tendencies, uh, that can be, I mean, it depends what you're trying to do, but there are tendencies that can be problematic. One is to focus a lot on price swings. And, you know, price swings, uh, can, but both, both on the upside, by the way, and on the downside. So price swing, you know, you might think that upward price means that you're smart and downward. Okay. So, so try to focus less on that. And then the second thing is people tend to focus on macro events, right? So, oh, here's what the Fed's going to do. And here's what the inflation is going to do. And here's what interest rates are going to do and so forth. There is a large literature demonstrating that people are not good at forecasting those things. So it's not to say that you shouldn't be aware of them, or you should not think about how those things might affect your business or the businesses you're looking after. But by the same token, sitting around and spending a ton of time on macro in my, from my point of view is not is not great. So then the question becomes, what are we trying to do? And, you know, the key is we're trying to find we're trying to buy stakes in businesses. And that actually at the end of …

AI assessment note: “The key is, are you sticking to your process and trying to think about things”

Q My question to you is, you know, you've said before about everything being a DCF, and as a venture investor, I went, hmm, that's an interesting thought, and we forget that our job is tied to DCFs ultimately, so can you explain this idea of everything being a DCF specifically for a venture audience?

A Yeah, so Harry, I gotta say that the impetus to write that report was a venture capitalist actually sent me a note about value venture businesses. I was like, I don't really like it. So, uh, it got me to encourage me to write that. Um, listen, I think that the way I would think about this as a venture capitalist is not that I'm going to build some elaborate DCF model. Of course, that's very unrealistic given the wide range of potential outcomes. I think that the, for that, for venture in particular, I would step down to, you know, what I call the basic, well, many people call the basic unit of analysis, which is fundamentally how will this company make money and understand that. Now, obviously, sometimes we're going to build up to how the economic model is going to work, but it's really trying to get some clarity around ultimately what the economic model is going to be. So that economic model then becomes the foundation upon which the future cash flows are built. And when you think about the value of any financial asset is the present value of the cash flows. And as a venture investor, eventually we're going to have an exit. The number one form of exit, of course, is to sell to a strategic buyer. And when you do that, the problem becomes their problem. They have to figure out how to extract the cash flows from the business over time. Or if you go public, of course, then, uh, yo…

AI assessment note: “value of any financial asset is the present value of the cash flows”

Q mismanagement or wrong decision making, here I think so much of, you know, where we're at is because of, you know, a health pandemic in COVID. We have political conflict and, you know, a potential war that, you know, shakes so much of supply chains. What worries me is, The instability of the world, not just financial markets, and does that make sense? And it feels like it's a first.

A It, it does, but again, you know, go back to take off, go back to 1900 and look at the history from 1900 to today, and just ask yourself, oh, with pandemic, 1918, we had two world wars. I mean, we've had, we've seen all, we've seen versions of all these things over 125 years, right? On, on some level, it's always novel. It's obviously massively disruptive. Like you said, I mean, we've had more globalization in the last 30 years, and as a consequence, supply chain effects are greater. But on some level, we've seen versions of all these things in the past. And so again, it's, it always feels dramatic in the moment. It always, and it is. I'm not saying it's not. But, uh, I, I think we have seen things like this. And by the way, on volatility is a good, an interesting one. Go back and look at volatility. And, you know, you see that the volatility we've seen in the last 10 or 15 years, even including the financial crisis, really doesn't hold a candle to the volatility we saw in the 19 twenties and thirties in the United States. There was vastly more volatility in, in stock prices. So yeah, on the one hand, it's always novel. It's always scary. On the other hand, when you, when, when the sweep of history, this is not inconsistent with things we've seen before.

AI assessment note: “this is not inconsistent with things we've seen before.”

Q Can I ask, what do you think makes Bill Gurley so special and the incredible ambassador that he is?

A So I think, first of all, he's very bright, and he's very intellectually curious, um, and It is interesting. So I met him right when he got out of business school and joined. We were in the same research department and right away I could see, I started talking to him about certain frameworks and ideas, and he immediately not only understood them, but integrated them. So two big ideas. One is he, he understood, we were talking about DCF and everything's a DCF model. I think he's very early on understanding the importance of return on capital investment returns, cash flows, and he applied it right away. Very successfully, by the way, in investments. The second is he was very early on and understanding Brian Arthur's work on increasing returns, understanding the mechanisms of increasing returns and economics. We're taught the returns on the margin go down. Decreasing returns is a consequence of competition and maturation. But under certain circumstances, you get these increased returns. So he was very early on. So, um, and by the way, not not only a good investor, but also a very good communicator. So he writes very well. And, you know, you, you, anytime you see or give it, you know, see the talks that he gives, they're very thoughtful, very well prepared. So yeah, it's, it's a combination of all those things.

AI assessment note: “he's very bright, and he's very intellectually curious”

Q process or just improve it in some ways? Like I said, when I had two people on my ICs, It didn't work honestly, Michael. And then when I had three, it was like product market fit of investment decision making, but I could have kept going with two and it just meandering in the way that it was. How do you think about when to change a core investment process?

A Yeah. I mean, to me, I, I think it's, you know, all this, you can do a lot of pre-work and get sort of best practices and, you know, three is better than two. That's a best practice. And, and by the way, five is, is not five is better than four. For instance, these odd numbers tend to be good things in terms of, and there's good data. Pardon me. There's good data to back all that up, but to me, the, the issue, uh, so, so that get that hygiene down, that process down, uh, in terms of the best practices. The other part that's very difficult is literally how do we update our views of the world as the world changes? That becomes the big challenge, right? And that, you know, the fancy term is to be an appropriate Bayesian, which is I have a prior view of the world as new information tumbles in, I revise my view in the proper direction and the proper magnitude. That is the very difficult thing to me. So when do you, when does, when does the world have to change? What is your process of change to reflect that? Um, I'll give you a couple examples. One example, certainly in, uh, we do a lot of work on this in public markets, very relevant for private markets as well, which is we've had in the last few decades, a huge rise in intangible investment, right? So companies used to invest in factories and machines and inventory, old fashioned stuff. And the accounting for that was a certain wa…

AI assessment note: “to be an appropriate Bayesian, which is I have a prior view”

Q Can I ask, where do you find people most misunderstand this theory and this understanding? Where do your students misunderstand it? Where do you hear people just not get it?

A Yeah, I mean, I think that because it's about attribution, right? In other words, if things go well for me, I want to think that I'm skillful, that it's like volition, that I'm the one that's responsible for it. And if things go badly, I want to point the finger at somebody else. So I think a lot of it is just sort of taking responsibility for all these things. But even taking one step back here, I would just say that it's not always clear, you know, various activities have various combinations of skill and luck in determining the outcome. So you think about things like chess matches or running races, that's almost pure skill and luck plays a very, very fleeting role in those kinds of things. By contrast, we were talking about, you know, lotteries, but roulette wheels and certain, those things are all basically luck and very little skill. And most things are at some between those two extremes. And I think our minds are not particularly good at sorting out where that activity is on that luck skill continuum. And as a consequence, it's misattribution basically.

AI assessment note: “our minds are not particularly good at sorting out where that activity is”

Q Oh, a hundred percent. Um, I think it's the challenge that all spin out GPs face when they try and raise their own funds. Um, when we think about that, though, then, is investing in or building a company on a novel premise, is that a luck-based endeavor?

A I don't, I mean, I don't think so, of course. I, I, I think that you, what you strive to do is to develop a process that you think will allow you to have the most, uh, best chance of success over time. And one of the things I would just say, Harry, about building a process is the first thing to think about and, and think about deeply is actually what you're trying to do. Like, how do you think you're gonna achieve some sort of an edge versus other people doing it? And in effect, the process you develop is, is, should be something that contributes to that source of edge, you know? So for example, in public markets, you'd think about the two extremes would be, you know, Jim Simons at Renaissance technology, right? Tons of technology, very smart people trading very frequently. And on the other side of the continuum would be Warren Buffett, who reportedly sits in his office all day and reads and then makes these very few, but very consequential investment decisions. Both of those folks have been very successful over time, but their approaches and their processes are very different. So the key is that there is no one size fits all for process, I think. But rather that you want to make your process congruent with your perceived source of edge. And so that to me is sort of the key thing to bear in mind. It's, it's, it's often one thing people say, here's what I do all day. Usually whe…

AI assessment note: “I don't, I mean, I don't think so, of course.”

Q Can I ask you, what makes bad process, or where do you see most people make mistakes in that process?

A Yeah, I mean, it's usually two things. One is a deviation from what you say you're trying to do, right? So you just laid out kind of how you're approaching it, this combination of media and venture investing and so forth. You know, the question is, are you starting to do things that deviate from those things, or you bring in these three founders, your, your little investment committee, and, and they really have expertise and different points of view and you override them for whatever reason, right? So you just depart from process in a way that's, um, non-systematic. And by the way, it's not that hard to do because sometimes you may have a process and it's, you have a couple things that don't work out, right? Which is gonna be naturally happening. So you start to question what you're doing. So you deviate. That's, um, that's one problem. The other one is just letting biases creep in to what you're doing, right? So biases would be things like, and the, and these, you may be emboldened by success, by the way, but two big sources of bias, I think are one is overconfidence. And when you look at overconfidence, usually the, The, the, the sub-component of oral confidence of the problem is over precision, which is you start to think you know what the world is going to be rather than better than you do, and the other one is confirmation bias, which is sort of, you know, when you have a …

AI assessment note: “it's usually two things. One is a deviation... The other one is just letting biases”

Q because you develop a thesis in your mind, you meet a company that aligns to that thesis, and you go, yes, that is the one. And then you're going, well, it's probably a 50% chance that your thesis was wrong, but confirmation bias has made you commit to that deal. How do you think thesis-driven investors can not fall victim to confirmation bias and that situation I just laid out?

A Yeah, no, right. Exactly. I think this is a really big issue. And, um, the biggest thing I, the way I would think about this is to say, if you do have a thesis, and by the way, I'm not adverse to that, but if you have a thesis, you should be thinking about so-called signposts, which is to say, if this thesis is correct, here are the things that we should see happen as we travel down this road, right? And these signposts are things you should communicate in advance, not as we go, but in advance, And typically write them down and you should write them down specifics with probabilities, right? And so if this is, this is what we should have it. Every time you don't pass a signpost or the signpost is something very different than what you anticipated, that should be a stop point. And the stop point should say, okay, is there a reason for this for reasons? It could be, or should we kill, right? Should we quit what we're doing? And so you, in a sense, you've laid it out in advance. Um, you've laid out the potential problems. You laid out the potential opportunity. And you've, you know, so you're almost, you're pre-committed to changing your course of thought if, if, if your thesis isn't unfolding as anticipated, right? So the key, uh, Harry, is to do the work up front, essentially, um, and, and, uh, as a consequence, and think through the potential, uh, alternatives. And then that alm…

AI assessment note: “if you have a thesis, you should be thinking about so-called signposts”

Q My question to you is, you know, you've said before about everything being a DCF, and as a venture investor, I went, hmm, that's an interesting thought, and we forget that our job is tied to DCFs ultimately, so can you explain this idea of everything being a DCF specifically for a venture audience?

A Yeah, so Harry, I gotta say that the impetus to write that report was a venture capitalist actually sent me a note about value venture businesses. I was like, I don't really like it. So, uh, it got me to encourage me to write that. Um, listen, I think that the way I would think about this as a venture capitalist is not that I'm going to build some elaborate DCF model. Of course, that's very unrealistic given the wide range of potential outcomes. I think that the, for that, for venture in particular, I would step down to, you know, what I call the basic, well, many people call the basic unit of analysis, which is fundamentally how will this company make money and understand that. Now, obviously, sometimes we're going to build up to how the economic model is going to work, but it's really trying to get some clarity around ultimately what the economic model is going to be. So that economic model then becomes the foundation upon which the future cash flows are built. And when you think about the value of any financial asset is the present value of the cash flows. And as a venture investor, eventually we're going to have an exit. The number one form of exit, of course, is to sell to a strategic buyer. And when you do that, the problem becomes their problem. They have to figure out how to extract the cash flows from the business over time. Or if you go public, of course, then, uh, yo…

AI assessment note: “the way I would think about this as a venture capitalist is not that”

Q Do you look at VCs today? I'm just fascinated. Do you look at VCs and see like, thirty million price for like, you know, two people and a dog in a garage with nothing and go, what are these people doing?

A Well, I mean, yes and no. I mean, the one thing I would say that You know, there's a, there's a, I found to be a fascinating, not surprising, but fascinating academic paper published a couple of years ago where they looked at the distribution of returns for 30,000 venture capital investments. So these are individual investments. And then they looked at 15,000 buyout investments to look at the, the distribution of returns and, you know, in venture, you know, I'll just say what everybody already knows. The majority of investments lose money. Right. But a handful of investments make so much money that they more than common, you know, make up for everything else, right? So the overall returns for venture over time have been pretty good, mostly because it's a few companies pulling all the, all the freight for the whole thing. Now, that said, when you say, oh, thirty million dollars for two people in a garage, the key is even those investments that lost money were deemed to be positive expected value investments, right? They had option value. They had potential. We know in advance. So it's not, it's not how often you make money that matters. It's how much money you make when you're right. So on one level, it seems, you know, if it's, if, if the ideas are not good or the people are not good, that would be problematic. But on the other hand, if it's a positive expected value going into…

AI assessment note: “it's not how often you make money that matters. It's how much money”

Q Speaking of the episodic nature of returns there, Doug Leone told me on, from Sequoia, Harry, just never sell your winners. And, uh, you know, I, I, I took that very much to heart, um, and I didn't sell some winners at the time. Um, that was a mistake. How do you think about, uh, Position exiting. And when's the right time to take the chips off the table?

A Yeah, I don't know if there's a good answer for that. I mean, part of it is at some point, if a company gets mature enough, you should be able to start to assess value on some, in some way, shape, or form. You mentioned early on having these models might be silly, and I don't disagree with that. But at some point, if there's some point of maturation, and certainly if the company's gone public or even It's gotten some degree of maturity. You can start to think about, uh, you know, values or ranges of values that you think are appropriate, and, and the answer is, yeah, when the company meets or exceeds that value you think is reasonable, that then you should scale, scale out of it. Now, I'm sympathetic to the argument is, you know, this is work by Hank Bessenbinder demonstrates that in public markets over long periods of time, most companies, public markets earn Returns less than the treasury bill rates. And it's like a handful, less than five percent of companies basically drive all the value for the stock market. So that's the argument for latching onto the winners and holding onto them forever. The premise is that you can find them and you know, you own them and you know which ones are gonna be the winners. Um, so that would be the argument in favor of doing that. So that's, that's finding the whatever, you know, apples when they're turning or Amazons or Microsoft when they're…

AI assessment note: “when the company meets or exceeds that value you think is reasonable, that then you should scale”

Q people, you have students, you know, this is unprecedented market times, and you have a lot of young investors going, oh, am I any good at this? I thought I was, but this is a very new world, and this is uncertain times for someone who's never done it before, or seen it before. How do you advise younger investors looking at the markets today questioning whether they're any good?

A Yeah, I mean, well, it's sometimes hard to know if you are any good. The key is, are you sticking to your process and trying to think about things the right way? And in markets like this, there are a couple tendencies, uh, that can be, I mean, it depends what you're trying to do, but there are tendencies that can be problematic. One is to focus a lot on price swings. And, you know, price swings, uh, can, but both, both on the upside, by the way, and on the downside. So price swing, you know, you might think that upward price means that you're smart and downward. Okay. So, so try to focus less on that. And then the second thing is people tend to focus on macro events, right? So, oh, here's what the Fed's going to do. And here's what the inflation is going to do. And here's what interest rates are going to do and so forth. There is a large literature demonstrating that people are not good at forecasting those things. So it's not to say that you shouldn't be aware of them, or you should not think about how those things might affect your business or the businesses you're looking after. But by the same token, sitting around and spending a ton of time on macro in my, from my point of view is not is not great. So then the question becomes, what are we trying to do? And, you know, the key is we're trying to find we're trying to buy stakes in businesses. And that actually at the end of …

AI assessment note: “The key is, are you sticking to your process and trying to think about things”

Q mismanagement or wrong decision making, here I think so much of, you know, where we're at is because of, you know, a health pandemic in COVID. We have political conflict and, you know, a potential war that, you know, shakes so much of supply chains. What worries me is, The instability of the world, not just financial markets, and does that make sense? And it feels like it's a first.

A It, it does, but again, you know, go back to take off, go back to 1900 and look at the history from 1900 to today, and just ask yourself, oh, with pandemic, 1918, we had two world wars. I mean, we've had, we've seen all, we've seen versions of all these things over 125 years, right? On, on some level, it's always novel. It's obviously massively disruptive. Like you said, I mean, we've had more globalization in the last 30 years, and as a consequence, supply chain effects are greater. But on some level, we've seen versions of all these things in the past. And so again, it's, it always feels dramatic in the moment. It always, and it is. I'm not saying it's not. But, uh, I, I think we have seen things like this. And by the way, on volatility is a good, an interesting one. Go back and look at volatility. And, you know, you see that the volatility we've seen in the last 10 or 15 years, even including the financial crisis, really doesn't hold a candle to the volatility we saw in the 19 twenties and thirties in the United States. There was vastly more volatility in, in stock prices. So yeah, on the one hand, it's always novel. It's always scary. On the other hand, when you, when, when the sweep of history, this is not inconsistent with things we've seen before.

AI assessment note: “It, it does, but again, you know, go back to 1900 and look at”

Q capital, you obviously see increased competition and increased pricing, which obviously then reduces the size of your outcomes. If you come in at 50, not 10, that's a very significant return difference. And what worries me is we're going to see this denigration of venture returns, much more to PE like returns. How do you think about that? And do you think that's fair for me to worry about that?

A Absolutely fair. Anytime you see capital flowing into something, you should worry about returns. That's just, that's like a basic rule. And, you know, if you look at long-term public market equivalent, so, so this is how academics are measured is public market equivalent for venture capital. It's actually been pretty good over long periods of time, but it's extremely episodic. So almost all the returns are earned in very short windows. One was around the dot com. One was actually what we just went through in 20, 20, and 20, 21. And now we're back. We're probably back down to sort of PMEs closer to one, right? So for venture, it's very episodic. So buyouts, by the way, have consistently been a little bit better than, um, private equity have been a little bit better than venture. Um, they're, they're much less episodic or dramatic, and they tend to chug along a little bit over one PME. So, um, Yeah. I mean, I think that you always should worry about that. And the other interesting thing is just taking a close examination at how companies are exiting. Right. And so what's happened is, you know, a generation or two ago, a lot of companies went public for a lot of reasons. Now companies are not going public at the same rate. Uh, these are interesting questions as to how to think about exits as well. So the whole complex should be examined pretty carefully. And, um, and, and by the w…

AI assessment note: “Absolutely fair. Anytime you see capital flowing into something, you should worry about returns.”

Q Can I ask, where do you find people most misunderstand this theory and this understanding? Where do your students misunderstand it? Where do you hear people just not get it?

A Yeah, I mean, I think that because it's about attribution, right? In other words, if things go well for me, I want to think that I'm skillful, that it's like volition, that I'm the one that's responsible for it. And if things go badly, I want to point the finger at somebody else. So I think a lot of it is just sort of taking responsibility for all these things. But even taking one step back here, I would just say that it's not always clear, you know, various activities have various combinations of skill and luck in determining the outcome. So you think about things like chess matches or running races, that's almost pure skill and luck plays a very, very fleeting role in those kinds of things. By contrast, we were talking about, you know, lotteries, but roulette wheels and certain, those things are all basically luck and very little skill. And most things are at some between those two extremes. And I think our minds are not particularly good at sorting out where that activity is on that luck skill continuum. And as a consequence, it's misattribution basically.

AI assessment note: “our minds are not particularly good at sorting out where that activity is”

Q because you develop a thesis in your mind, you meet a company that aligns to that thesis, and you go, yes, that is the one. And then you're going, well, it's probably a 50% chance that your thesis was wrong, but confirmation bias has made you commit to that deal. How do you think thesis-driven investors can not fall victim to confirmation bias and that situation I just laid out?

A Yeah, no, right. Exactly. I think this is a really big issue. And, um, the biggest thing I, the way I would think about this is to say, if you do have a thesis, and by the way, I'm not adverse to that, but if you have a thesis, you should be thinking about so-called signposts, which is to say, if this thesis is correct, here are the things that we should see happen as we travel down this road, right? And these signposts are things you should communicate in advance, not as we go, but in advance, And typically write them down and you should write them down specifics with probabilities, right? And so if this is, this is what we should have it. Every time you don't pass a signpost or the signpost is something very different than what you anticipated, that should be a stop point. And the stop point should say, okay, is there a reason for this for reasons? It could be, or should we kill, right? Should we quit what we're doing? And so you, in a sense, you've laid it out in advance. Um, you've laid out the potential problems. You laid out the potential opportunity. And you've, you know, so you're almost, you're pre-committed to changing your course of thought if, if, if your thesis isn't unfolding as anticipated, right? So the key, uh, Harry, is to do the work up front, essentially, um, and, and, uh, as a consequence, and think through the potential, uh, alternatives. And then that alm…

AI assessment note: “you're pre-committed to changing your course of thought if, if, if your thesis isn't unfolding”

Q I love that as a distinction. You said before, as kind of a follow-up to that one, my natural kind of question was, why does putting yourself in a position to enjoy good luck, what everyone obviously wants, why does that put you in a position to lose?

A Yeah, I mean, the first thing is, I'll just say that there, there are like a bunch of aphorisms about luck, you know, luck is where preparation meets opportunity, or the harder I work, the luckier I get, and I don't really find any of those very appealing, candidly, if you, if you accept my definition of luck, right, um, because, and another way of thinking about it is that, you know, uh, skill is what's in your control, and luck goes with out of your control, so by definition, out of your control. Now, that said, Harry, I think that, you know, what you're picking up is on something like this. You say, oh, I, you know, My friend won the lottery yesterday, right? And we'd all agree, I think, that that was lucky. Now, you would not be in a position to win the lottery unless you buy a lottery ticket, right? So in a sense, you have to be willing to lose in order to potentially win. The fundamental question it all distills down to is, is it positive expected value or negative expected value, right? So, and, and investing by and large should be obviously positive expected value. Playing lottery by definition or gambling is a negative expected value. It's a consequence. People should be, you know, that they should do it for fun if they would like. And for other, there may be other psychic benefits of it, but just in pure monetary pros and cons, it's going to be a negative game.

AI assessment note: “you would not be in a position to win the lottery unless you buy a lottery ticket”