One very popular and widely used theory in the world of financial advice is "Modern Portfolio Theory." Since it was invented in the 1950s, it has become the cornerstone of financial planning. I was recently very privileged to interview the inventor of this theory, Dr. Harry Markowitz, winner of the 1990 Nobel Prize for Economics.
You can listen to our interview, which was first broadcast on The Goldstein on Gelt Show here. Below is a transcript of what we discussed.
Douglas Goldstein:Dr. Harry Markowitz is an economist, the recipient of the John von Neumann Theory Prize, as well as the Nobel Prize in economics. He's best known for his work in modern portfolio theory, which is the cornerstone of the work that I do as a financial advisor.
Can you give us a layman's definition of modern portfolio theory?
Harry Markowitz:I can't give you a definition, but I can give you a brief description. When somebody, when you or an institutional investor, or an advisor, applies portfolio theory, at the heart of this application is a computer program which we call a "portfolio optimizer" and like all computer programs, it has inputs and outputs. About the output: there's a couple of outputs. One is a risk return trade-off curve. It shows you what's the most return you can get on the average for a given level of risk or the least risk you can have to put up for a given level of return. There's a trade-off curve. You can go for low risk, for which you have low return, or you can go up the curve and get the higher return but for more risk, and then behind each of the other outputs of this program, behind each of these risk return combinations, is a portfolio combination of securities and the risk of return output. That's the output of the analysis. The outputs are just the logical consequences of the inputs. The inputs are when a user, like an institution, decides on what its investables are. Maybe it does the analysis at the asset class level or at the security level, but for each of the investable securities it has to supply an expected return. You have an average return in the long run, a supply of estimates, the volatility or uncertainty of the returns. It's got to supply estimates of the correlations between each security or some kind of a model from which these correlations can be derived, and then finally it states constraints that don't invest too much in this security, don't invest too much in this industry, and so on. Then what the mathematics does, what the computer does, is to figure out that curve, given your constraints and your estimates. It figures out what maximum return you can get for a given level of risk.
Douglas Goldstein:So in simpler terms, a lot of people will say, "Well, I've always heard that if I'm willing to take more risk, I'm going to make more money," but that's not a true statement, is it?
Harry Markowitz:No, there are ways of taking risks that will not be efficient, as we say. You can gamble. You can buy lottery tickets. By taking more risks, then you'll probably lose, like most people will. And also this analysis is a framework that helps you to think through the consequences of portfolios of your own estimates for risk of return on the individual securities and their correlations. So if you are overoptimistic on the inputs and do something really stupid like leverage, a high lay, then you get wiped down on a downturn, don't blame me.
Douglas Goldstein:One of the concepts, I think, that modern portfolio theory relates to is that people shouldn't just look at one asset, but should rather look at how each of the assets correlates and works together, which is why we always talk to people about designing a whole portfolio and not just looking at one thing at a time. Is that a good summary?
Harry Markowitz:I think it's all summarized in the title of my 1952 paper, my first paper on this subject. It was called Portfolio Selection, and that was the novel insight that as far as the analysis goes, you should think about the portfolio as a whole. I'd say the analysis was new because people knew about diversification before Shakespeare. In The Merchant of Venice, somebody says, "Antonio, you look sad, is your business going bad?" and he says "My goods are not too bottom trusted nor my fortune to just this one year" and so on. So he knew about diversification.
Douglas Goldstein:These days, the markets feel crazy, and I've been doing this for something like 25 years, working as a financial advisor. Every time we go through this kind of upheaval, I hear from clients, "Well maybe this time it's different, and look at what's going on with oil, or this problem or that problem." What is the best way for an advisor, like me, to help clients through these rough markets when we are applying your theories?
Harry Markowitz:I work for an outfit called "Guided Choice." It's my every Thursday place that I go to, and I wrote a little piece for them towards the end of 2008. I wrote a piece for them to distribute to their clients, saying, "The sky is falling. What do I do now?" and then the piece started like this. It said, "Well, now that you know what a standard deviation feels like, or a two-standard deviation move feels like, in real time, maybe instead of going so high on the frontier, we have seven portfolios where one is the most chicken and seven is the most aggressive." So instead of being at six and seven up at the high end, you should come down to four or five, but not with the intention that you are going to move back up when things are better, because the chief error that the small investor makes is they buy when the market has gone up. They assume it's going to go up further, and then they sell when the market has gone down and they think it's going to go down more, whereas the institutional investor, the financial advisor, does something which is called "rebalancing." You start off maybe with a 60-40 mix, but when the market goes up, it's now 70-30 and you sell off the 10%, so you get back to 60-40. So you're selling when it is high, and then if it goes down, it will be in 60-40 or 50-50, so you buy what's left. So if you take the complete cycle from where the market was in 2007, and then you go all the way down, and then you come back until it's up there again, some people have lost money, but some people have made money. So the small investor, who dumped at the bottom and who bought at the top and dumped at the bottom, lost money, and my folks made money.
Douglas Goldstein:I think that one of the reasons that people mess themselves up is in fact not because the rules that you are describing are wrong. It's just that people tend to mess themselves up. As you know, there's the whole theory of behavioral finance, researched by people like Danny Kahneman who won the Nobel Prize, and Terry Odean on your coast.
Harry Markowitz:I was going to say Terry Odean has some wonderful work out on the kinds of mistakes that individuals make. They tweak too much, he says, especially men more than women.
Douglas Goldstein:Is there something that people can do, when they realize that your systems are reasonable, in order to stop themselves from messing themselves up?
Harry Markowitz:Okay, if they don't have a financial advisor, this is the advice I give to waitresses who ask me. They said how do you balance between stocks and bonds? Bonds are complicated, so let's say stocks and a savings account. They should buy a well-diversified portfolio in an investment company. Vanguard is the usual, inexpensive and well-diversified. They offer a lot of other things, but in the portfolio they should balance between stocks and bonds. So if the stock market went down 38% or 40% - the big caps went down 38% in 2008 - if the stock market goes down 30%, 40%, or 50% maybe, due to some horrible event, it's going to come back up eventually. They're not going to chicken out, so if they have half bonds half stocks, my simple suggestion that I would say to a waitress is to put half of your money in stocks, half of your money in the savings account, or 60-40 if you are young, or if you are younger you could go all stocks. But make sure that you stick with the program, and as for rebalancing, it's a little complicated, and you should leave that to the institutional investors.
Douglas Goldstein:I was working on a book about how the strategies of chess can be applied to be investing. I wrote the book with world chess champion Susan Polgar, and in the book, what we realized was that all of the behavioral finance models that Kahneman and Odean talk about, all those problems seem to affect chess players as well. When I was talking to Susan about it, I would explain to her what we study in the world of finance, and she was nodding her head and saying, "Yeah, this is exactly the same type of problem or disease that infects the mind of a chess grandmaster." So I think that having a system like you are describing that you are just going to stick to is the most important thing, even if it's not the best system. Having any system is better than having nothing at all.
Harry Markowitz:I don't know about that, because other people may do momentum. They have a system that if it went up, it's going to go up, and if they follow that system and don't diversify, in chess you got to know a little something. You can get clocked, you can get fool's mate almost instantly, if you have a bad system. But it's true that what portfolio theory supplies is a framework and you've got to use it in such a way that you are going to stick with it, otherwise it's useless.
Douglas Goldstein:One of the tools that a lot of financial advisors use in order to develop the asset allocation model is a Monte Carlo simulation, which is basically testing many, many, many different possible market scenarios in a certain asset allocation model. Do you think this is a reasonable tool for people to bet their future on?
Harry Markowitz:Yes, but before I answer that question, let me go back to the chess situation. There's one major difference between chess and investing. With chess, it's a real book and with finance, it is true that every time it is different. Every time there's uncertainty, every time the world keeps changing like the '20s, the '30s, or with Hitler, with the things that are going on now, you're still trying to get controlled risk and get return on your portfolio as a whole. But you always have to do this forward looking - what things are more uncertain, what things do I have to diversify against?
Let me tell you about the Monte Carlo simulation. I said there is a risk-return frontier. There are various combinations of risk and return you can get. The analysis is a one-period analysis. In other words, we want to get a good probability distribution of return between now and let's say next year. And then next year we'll get a good probability distribution of return from then to the year after that, and then maybe if our beliefs don't change, we'll rebalance. But supposing I am 25 years old, and I expect that I'm saving up for my retirement at 65. I want a good probability distribution of what my wealth will be like at 65, and the state of the art is to do a Monte Carlo simulation, taking into account, even though the world is uncertain, that you need some kind of model of how prices evolve or how returns evolve, and you have to randomly draw over and over again. But you are also taking into account what wealth this person has now, the rate of their saving. Maybe it's a 401k, so what is the rate in which the employer will match, etcetera, etcetera. So thinking through the consequences of any particular point on the risk return trade-off curve is not a trivial thing, and Monte Carlo helps you pick the right point on the curve.
Douglas Goldstein: So the bottom line is that for someone who's trying to develop a long-term financial plan, a Monte Carlo simulation is a good tool to help get a sense of where he might be going within a certain let's say bandwidth of volatility?
Harry Markowitz:Yeah, obviously there's no pinpoint because the world is uncertain, and more of the impact of my prior remark is that there is more uncertainty than we actually put into our models because we have to make a model in that sense. But there's a lot of randomness in any reasonable model. We know that on the average, over the long run stocks have done well and we assume that on the average over the long run, they will do well, but there is a randomness in it and depending on whether you're going to retire five years from now or 50, 40 years from now and etcetera, etcetera, Monte Carlo is not the kind of thing which an individual is able to do for themselves. So this is one of the benefits of having an advisor who in turn has some kind of support from some large institutional investor that supplies frontiers and Monte Carlo and so on. One of the advantages, if you can, of having advisors is that they can help you pick out the right point on the frontier using in part Monte Carlo.
Douglas Goldstein:We certainly do that, and we do are very best, though as you point out, there will always be a certain element of randomness. How can people follow your work and follow the new ideas you are working on these days?
Harry Markowitz:I'm currently writing volume 3 of a four-volume book, and so volume 1 is out. It's called Risk Return Analysis: The Theory and Practice of Rational Investing, but this is really for the students and the quantum person and the scholar. There are popular books around that are non-mathematical. There's one by Peter Bernstein that explains all of these things, and it's called Capital Ideas that explains this in non-technical terms. But there are other books around. I don't have a list of them right now, on how to invest. Roger Gibson has a book on asset allocation that's now in the fifth edition, which is for a popular audience.
If you Google "Harry Markowitz Company," you'll get a page out of my website, which has a link back to the home page, and you'll see my prizes, and my publications, and my two very able secretaries at Harry Markowitz Company. All that is there.