Episode #51: Mark Kritzman, Windham Capital Management, “We Have To Question The Assumptions That Underpin Our Models… Nothing Is Simple”
Guest: Mark Kritzman. Mark is a Founding Partner and Chief Executive Officer of Windham Capital Management, LLC, the Chairman of Windham’s investment committee, and a driving force behind Windham Labs. He is responsible for managing research activities and investment advisory services. He is also a Founding Partner of State Street Associates, and he teaches a graduate course at the Massachusetts Institute of Technology. He has written numerous articles for academic and professional journals and is the author of six books including Puzzles of Finance and The Portable Financial Analyst. Mark has won multiple awards including the Graham and Dodd Scroll, the Bernstein-Fabozzi/Jacobs- Levy Award, and the Roger F. Murray Q-Group Prize.e.
Date Recorded: 5/8/17
To listen to Episode #51 on iTunes, click here
To listen to Episode #51 on Stitcher, click here
To listen to Episode #51 on Pocket Casts, click here
To listen to Episode #51 on Google Play, click here
To stream Episode #51, click here
Comments or suggestions? Email us Feedback@TheMebFaberShow.com or call us to leave a voicemail at 323 834 9159
Summary: In Episode 51, we welcome Mark Kritzman. Per usual, we start with Mark’s background. He tells us a bit about his 40-year career in investing, leading to Windham, where he focuses on asset allocation and risk premia strategies.
But it’s not long before the guys jump in, starting with Mark’s 7th book, A Practitioner’s Guide to Asset Allocation, which will be coming out soon. Mark describes the process of asset allocation, starting with the basics, then taking us a layer deeper, discussing asset allocation as a way to balance the goal of increasing wealth while minimizing drawdowns. In essence, you need to identify the asset classes you want, evaluate their expected, long-term returns, then estimate the volatility of each and – just as importantly – estimate the correlation between the asset classes. With all this, you then find the particular allocations that give you the highest return for the same level of risk – the efficient frontier.
Next, the conversation takes a turn toward investing fallacies, including the idea that asset allocation drives more than 90% of performance. Mark tells us there are some flaws with this idea, then explains in detail. Another fallacy discussed is that of time-diversification – the assumption that investing over the long-term is safer than investing over shorter periods. Again, Mark provides details that call into question this belief.
The guys then get into investing in illiquid assets, and how to appropriately structure them in an asset allocation. It can be hard to maintain a balanced portfolio consisting of illiquid assets. Mark’s approach is to treat liquidity as a shadow investment. In essence, you attach a shadow asset as well as a shadow liability to the appropriate parts of the portfolio. You’ll want to listen to this part of the episode for all the details.
This dovetails into hedge funds, since hedge fund investing can also be illiquid. Meb asks how Mark thinks about hedge fund investing, and given limited information, is an investor’s only recourse to be able to pick the best managers? And if one doesn’t have that ability, should he/she just stick with investing in the S&P?
Mark has a great answer about how most of the historical premium of private equity over public equity can be attributed to the sector exposures of private equity funds. So investors can build a portfolio of public sector ETFs in a way that can approximate much of the hedge fund sector allocation. You’re probably going to be surprised at just how much of the premium of private equity over public equity doing this which would have delivered to an investor.
As usual, there’s plenty more in this episode: the role of fees and taxes… the concept of “turbulence”… the absorption ratio, and how we can use it to evaluate risk… and lastly, what Mark’s most useful idea is for listeners.
What is it? Find out in Episode 51.
Links from the Episode:
- Mark’s new book, A Practitioner’s Guide to Asset Allocation
- Windham Capital Management
12:07 – Brinson/Hood/Beebower paper
20:24 – The Forever Fund – Meb Faber
38:35 – Mahalanobis distance
Transcript of Episode 51:
Welcome Message: Welcome to The Meb Faber Show where the focus is on helping you grow and preserve your wealth. Join us as we discuss the craft of investing and uncover new and profitable ideas all to help you grow wealthier and wiser. Better investing starts here.
Dsiclaimer: Meb Faber is the co-founder and Chief Investment Officer at Cambria Investment Management. Due to industry regulations, he will not discuss any of Cambria’s funds on this podcast. All opinions expressed by podcast participants are solely their own opinions and do not reflect the opinion of Cambria investment management or its affiliates. For more information, visit cambriainvestments.com.
Sponsor: Today’s podcast is sponsored by Global Financial Data. We’ve been using data series from GFD for almost 10 years, ever since I wrote my first white paper. The data has been vital in our research in areas such as CAP [SP] ratio calculations and historical simulations. For almost 20 years now, global financial data has been aggregating and transcribing data from original sources with many sources no other data provider has published before. Please have a look at their website at globalfinancialdata.com for more info and to set up a trial account. If you mention that I sent you, they’re offering a 20% discount on all new business subscriptions. Again, that’s globalfinancialdata.com.
Meb: Ladies and gentlemen welcome to the podcast. Today we have the first podcast post-procreation, so I am recovering from a week of sleepless nights and no more podcasting from a hospital waiting room, so back at home, and really excited. Today we have a special guest with us, he’s the founding partner and CEO of Windham Capital, Mark Kritzman, welcome to the show.
Mark: Thank you very much, happy to be here.
Meb: So today is going to be a fun chat. Mark is chatting with us from Boston. And if you’re not familiar with Mark, we mentioned him a number of times in the blog over the years, I mean going back probably almost 10 years now. And Mark is interesting because he’s, on one hand, has a leg in the academic side of the world, he’s written six books, over 70 articles, won a bunch of industry awards, but he’s also a practitioner. So unlike a lot of academics that just kinda have their head in the clouds and you know, talking about ideas and publications that may not have any application to reality, Mark has been doing this for a long time. So Mark, why don’t we get started. Give us a little background information to help listeners get some context, to kind of where you are, what you’re doing, and kind of how you got there.
Mark: Sure, thank you. So as you mentioned, I’m the CEO of Windham Capital Management and I’m also a finance professor at the Sloan School, MIT, and at Windham, we focus on capital asset allocation as well as risk premium strategies. So I’ve been in the business almost 40 years, and I think I’ve been pretty lucky to have been born at a time…I was entering the business when interest rates were very high, had a huge tailwind at my back, and also, I got to know the originators of modern finance personally, back when they were introducing these theories and looking for people in industry to pay attention to them, so many then become good friends. I feel pretty fortunate to have arrived in the field of finance when I did.
Meb: Awesome, well, so look Mark, by the way, has his seventh book coming out this month, I highly recommend you to pre-order it. It’s called “The Practitioner’s Guide to Asset Allocation.” It’s a pretty cool book because first of all, it’s great, has a forward by Harry Markowitz. So anytime you get that forward, it’s a good sign. But also, that the book takes investors through really the process of asset allocation. So I’m not gonna tell you whether you need to buy the S&P [SP] tonight or tomorrow, but really how to build a portfolio and how to think about it from a practitioner standpoint. So I figured we’d talk a little bit about the book. And by the way, Mark has got a lot of other great books, I’ve read a few of them in the past weeks and in years even, but the “Puzzles of Finance,” the practice…what’re some other ones? “The Portable Financial Analyst,” all by Wiley, so we can commiserate about that.
But let’s get started, so in the book you start out at the very basics. And it gets as wonky as you want by the way if you get to the appendices etc., but let’s start with the basics. You start out, say what’s an asset class? And I figure I’ll let you take it from there as a good starting point.
Mark: So there are several characteristics that define an asset class. First of all, the securities within an asset class have to be relatively homogeneous. The asset classes have to be sufficiently distinct from one another. Asset classes have to offer the opportunity to improve the quality of a portfolio without depending on investor skill, and asset classes also should be sufficiently large to absorb a meaningful fraction of your portfolio. So I think those key characteristics define an asset class. What’s not an asset class, for example, would be hedge funds, right, because hedge funds are not internally homogeneous, they tend to invest in all different kinds of securities. And they also…it’s not the case that you should expect to add value to your portfolio by picking the average hedge fund. In order for a hedge fund to benefit your portfolio, you would have to have some skill in identifying superior hedge funds.
Meb: You know, it’s funny we have…and I can’t remember who to attribute this quote to, but I’ve heard it mentioned so many times. I have a friend who says, “Hedge funds are a compensation scheme masquerading as an asset class.” And I can laugh and say that because we’ve managed plenty of private funds in the past, but I in a world of today, particularly, with seeing so many mutual funds and ETF’s now that have, you know, look like hedge funds, as well as hedge funds that almost just look like closet indexers, the world has become much more blurred. It was a term that really meant something, probably in the ’70s, but today has kind of lost a lot of meaning. So some asset classes, you know, we typically start from the Big Four. Stocks, bonds, commodities, and currencies. And you mentioned some, you know like look, art, it’s not an asset class. And so you from the starting point of asset classes, what’s basically the fundamentals of asset allocation? How does someone get started? Once you’ve identified the universe, what’s next?
Mark: The way I think about asset allocation is to consider what it is investors really want. They want two things, they want to grow their wealth, and they want to avoid large draw-downs along away. And asset allocation is something that balances those two conflicting goals, right. They’re conflicting because the more you design a portfolio to increase wealth, the more you expose it to draw-downs. So an asset allocation policy is intended to provide that particular balance. Now, the typical approach for doing asset allocation is based on Harry Markowitz’s seminal article in 1952 called, “Portfolio Selection.” He showed that you want to diversify in an efficient way, and you need to take into account not only the risk of the different assets that you’re using but also how they covary together, how they either move together or diversify each other. So what you need to do is identify the asset classes that you’re interested in, estimate their expected return. I would not advise investors to rely too much on short-term historical averages. I think investors should look at either very long term returns or better yet, long-term risk premiums and add those risk premiums to the current risk [inaudible 00:08:30] rate, or come up with some fundamental approach for estimating expected returns.
However, I do think it’s pretty reasonable and pretty common to rely on historical measures of risk. So in addition to your expected return, we have to estimate the volatility of each asset class, and the correlation between the asset classes. And then what we want to do, what Harry Markowitz taught us to do, is to find the particular allocations that will give us the highest expected returns for a given level of risk. So we do that and we can trace out a frontier of portfolios that satisfy the criterion that is that for each level of risk…and by risk, we typically measured a standard deviation or volatility. So we have a…if you can imagine a graph where the vertical axis is expected return and the horizontal axis the standard deviation, looking at going from the lower left to the upper right and trace out a concave curve that’s always upward sloping, and the portfolios that lie on that curve are the ones that offer the highest expected return for each level of risk.
So the first step is to identify the menu, that is the asset classes. The second step is to construct the efficient frontier. And the third step is to pick the particular portfolio on the efficient frontier that best matches your preferences, your aversion to risk for example. There’s a theoretical argument for doing that which is simply to find a point of tangency between your utility function and efficient frontier, but nobody does that in practice. That’s just a nice theoretical concept.
So what we do in practice is we take each portfolio, it’s expected return, and standard deviation and we convert it to a probability of experiencing a particular loss or achieving a particular gain. Most people don’t really know the difference between…you know, if I said, “Here’s a portfolio that has an expected return of 5% and a standard deviation of 4%. Here’s one that has an expected return of 8% and standard deviation of 20%,” that’s not intuitive. But if I were to say, “This portfolio has a 5% chance of losing money on average over the next two years. This portfolio has a 10% chance of losing money,” then I can tell you what’s the likelihood that they’ll generate a certain amount of wealth, then that’s much more intuitive. So that’s what we do, is we just simply map the return and risk of each portfolio onto statements about the likelihood of achieving certain goals or failing to achieve certain goals. Portfolio that makes the most sense to us that way.
Meb: So perfect, we can see in the podcast there, simple advice, everyone understands that. But this is really where people usually go off the rails, and it seems like a pretty simple task to kind of run this through you know, and optimize or say, “Here’s a particular portfolio.” But most in particular individual masters, but also pros tend to now, kind of get into trouble here. And so, you know, I mean, I can’t tell you how many times I’ve been to a presentation and seen the famous study, or when they start talking about asset allocation and they say, “You know, 90% of the investment returns are due to asset allocation.” The famous Brinson, Beebower study. We talk a little bit about that and then we’ll kind of start to get into some more nuances and specifics on the asset allocation side.
Mark: Yeah, so that’s why in the book we include certain fallacies, we have several chapters that deal with the fallacies and several chapters that deal with challenges. The first fallacy is the notion that you just mentioned that asset allocation determines more than 90% of investment performance. And that’s based on this article that Gary Brinson, and Randy Hood, and Gil Beebower wrote back in 1986. What they did is they examined large pension funds, and then they defined the asset allocation policy as the average asset mix of those funds or you know, the particular fund invested in passive indexes. And then they defined market timing component of performance as the return attributable to deviating from that average asset mix. And then the residual left over, they attributed to security selection. So what they did is, they regressed the total return on these pension funds, on these two respective components of returns. And the regression analysis showed that about 94% of total return variation through time was associated with the asset mix policy. So they, therefore, concluded that asset allocation explains more than 90% of performance.
This is a fundamentally flawed article, the flaw is as follows. There’s no normal or default asset mix year, right. So it’s not saying if you know, the typical asset next might be 60, 40, or it can be whatever the average is of all of these funds. And then you know, the importance of asset allocation would be how much does deviating from that average change performance. But they don’t have any average, they assume that the average asset allocation is not investing at all, not even in riskless assets. So what they end up showing you is that when you incur risk or when you invest you incur a risk, but everybody sort of knows that. So what we do in the book, is we create this hypothetical world that has the stock market with only two stocks, and a bond market with only two bonds, and we can try to [inaudible 00:14:55] that stock A has the same returns as bond A and that stock B has the same returns as bond B, so that both asset classes end up having the same return. In this hypothetical world, the only possible way that you can affect performance is choosing between the two stocks and choosing between the two bonds. So it’s a world in which asset allocation has no relevance whatsoever, it can’t possibly affect performance because every single asset mix will have the same return.
We then applied the Brinson, Hood and Beebower methodology, we did the regression. And what their analysis revealed is that asset allocation explained 100% of performance. So you know, we created this hypothetical world in which asset allocation simply does not matter, we applied their methodology and it showed that all of the performance is due to asset allocation. It’s just you know, an extreme toy example to reveal just how silly their analysis was.
Meb: It’s also a good example of people just kind of reading the headlines, you see that a lot, just in media, in general, today. People read the headlines of an article, maybe if they are feeling particularly enterprising, they may read the abstract, but if you read a lot of papers today and spend some time thinking about what people come up with, some weird conclusions. All right, so you think about allocation, you get your portfolio. One of the really interesting chapters I thought was your take on what you call, “Time diversification.” And the concept where you say, “Look, it’s widely assumed that investing over long periods is less risky than investing over short periods because people think the likelihood of loss is lower over long horizons. You see so many charts talking about that. ” Maybe talk about how you think that that may not necessarily be the case.
Mark: That’s another fallacy, the fallacy of time diversification, and this has really provoked a lot of debate over the years. And the origin is from a former colleague of mine Paul Samuelson, in MIT. He engaged with one of his colleagues and offered his colleague a bet whereby his colleague would be favored to win the bet. And his colleague said, “You know, Paul, I realize that the odds are I’ll win, but should I lose I can’t afford to lose that much money, so I don’t wanna do this one bet. However, if you allow me to do the bet 100 times, I’ll take you up on that.” So that prompted Samuelson to write this very famous paper called, “The Fallacy of Large Numbers.” What he showed is that yeah, if you do the bet a hundred times, you’re less likely to lose on average, but the amount you potentially can lose is a hundred times as great. So it turns out that the diminishing likelihood of loss as you add more and more years to your horizon is exactly offset by the increasing amount that you can lose, so the magnitude offsets the probability. This is Samuelson’s argument against time diversification and it holds for a wide range of descriptions of how people you know, the kind of risk preferences that people have.
Now the reason that we are led to believe that time diversifies risk is because the probability of a particular loss does decline as we add more years to our investment horizon, as of the end of the horizon. But it turns out if investors care about what happens along the way, so let’s say you’re not so much concerned about the probability that you’re going to lose 10% or more at the end of 20 years, you care about being down 10% at any point during that 20-year horizon. That probability always goes up with time and never goes down. And then another very simple way of seeing that time does not diversify risk is that if you want to buy a put option to protect your portfolio, the price of that put option goes up with the time to expiration. So if it weren’t the case that risk increases [inaudible 00:19:25] over time, you wouldn’t be expected or required to pay more and more money for longer dated options.
Meb: You know I think this is such an important topic because you’ve touched on a couple of really interesting things. One, and these couple of sayings that we use that go along with what you’re talking about, part of which is, “Your largest drawdown as always in your future,” you know. And mathematically, that kind of has to be the case, it can’t get smaller. And the challenge for most people when they build portfolios, is like you mentioned, it’s the daily, weekly, monthly, quarterly, even yearly timeframe challenging of mucking around with them. You know, the endowment style challenge of sitting through a 50% drawdown in all of the angst that it causes people, and investors, and trustees, and everything else. So a lot of people are rational on paper, totally irrational in the real world, and that’s a big challenge. We think a lot about…we wrote an article recently called, “The Forever Fund.” Where we talked about, “Hey if investors were really serious about these long-time horizons, you could develop a structure that locks them up for 10 or 20 or 50 years and tell them to put their money where their mouth is.” But historically the path is the hardest part for people. They may understand it but it’s the most challenging.
Mark: You know, in the book we cover this a little bit, we talk about within horizon risk and in the second a book is where I got interested in this topic. I was doing some consulting years ago for a foundation and they asked me to help them determine the asset mix that would have no more than say 5% chance of losing money on average over a 20-year horizon. And I thought for a minute and I said, “So what you’re telling me is you’re going to meet with your investment committee only once and it’s gonna be the end of the last day of the 20th year?” And they said, “No, we meet quarterly.” So I said, “Well, would you be concerned if your portfolio was down 20% next year or the year after at any point during that horizon?” And they were, “Yes, of course.” So that led us to you know, figure out the method of determining within the horizon risk and also within horizon…you know, within horizon probability of loss, as well as within horizon value at risk. And it turns out the probability of loss was…I think would surprise, actually shock, a lot of people is that the likelihood of say, a 10% loss within a 5-year horizon is about 10 times as great the likelihood of a 5% loss at the end of the 5-year horizon. You know, it would be more likely than not at some point within that horizon, and only about 5% as of the end.
So that’s just something that I think investors should keep in mind if they’re a concerned about what happens along the way. But with most investors whether…even institutional investors, whether they’re willing to admit it or not, they do care about what happens along the way.
Meb: It’s a good business idea for you and I, we’re gonna start a endowment consultancy and say, “Here, we’ll give you an allocation, you can then disband your committee and in the rules, you’re only allowed to meet 20 years from now.” That would definitely be two-thirds of probably all the endowments out there if not more. It’s a great idea. So let’s get a little bit more into the weeds now. So we talked about the basics you know, coming up with the assets, putting them into a portfolio mix. Let’s talk a little bit about a couple of specifics and thoughts. And so you know, you guys, I know, have a lot of software and do a lot of work internally where some of the assumptions, as well as understanding of asset classes, and everything else really can make a big difference. And you know, one, you talked a little bit about real estate for example, and how the optimal amount of allocating to real estate whether you adjust for it or not adjust for it. Maybe you wanna talk about that a little bit because I think it can have kind of a big takeaway on how people actually implement a portfolio you know and because the assumptions have kind of big outcomes on what you end up investing in.
Mark: Now that’s a good point. So one of the challenges in putting together a portfolio is to determine what fraction of the portfolio should be allocated to illiquid assets. Real estate being a very good example, private equity is another example, some hedge funds with lock-ups etc. And this really you know, gained a lot of attention during the financial crisis because a lot of funds could not meet their cashflow commitments. The problem is that liquidity is measured in different units than return and risk. Most people when they try to figure out how to treat illiquid assets in a portfolio, do it in a very puristic way, they might simply just you know, assign liquidity scores and say, “I want my portfolio to have you know, a score greater than or no less than some threshold.” It’s pretty arbitrary.
The other issue is that most people think that you need liquidity only to meet cash demands. It turns out that you need liquidity for lots of reasons. You need liquidity to rebalance a portfolio, you need liquidity to take advantage of new investment opportunities, you need liquidity to terminate managers who are no longer performing up to expectations. So there are lots of uses of liquidity.
So the approach that we came up with is to treat liquidity as a shadow allocation. The way to do that is to figure out how a particular investor deploys liquidity. So a typical institutional fund, for example, will use liquidity to rebalance the portfolio. You put in place what you deem to be the optimal asset mix, prices change, and the actual asset mix is now different from what you believed to be optimal. What do you do? You want to rebalance it, so you come up with some rebalancing schedule, but to the extent, a large fraction of the portfolio is illiquid but you can’t fully restore the optimal weight, and you know, that lowers the quality of the portfolio, so we need some way of measuring that.
Another use of liquidity is to engage in tactical asset allocation. You may feel that you have some skill in tilting your portfolio to more defensive assets during certain regimes and to more growth oriented assets during other regimes. Well, again, if part of that portfolio is locked up in illiquid assets, you can’t take advantage of that skill. And then, of course, lots of funds have, you know, cash demands, pension funds have to pay out benefits. Balance funds have to contribute to the operating budget of the school, foundations have to pay out a certain amount of money. And then if you invest in private equity, you might have capital calls when the investments are scheduled. So there are lots of ways you use liquidity.
So the way to deal with this is to estimate either by simulation or some other process, the return and risk associated with these different uses of liquidity and then to attach a shadow asset to the liquid asset classes within your portfolio that you can move around to pay some of these liquidity needs. And then to attach a shadow liability to that part of the portfolio that can’t be moved. So the way I think about it is you attach a shadow asset you know, to play off and to engage in activities that improve the quality of the portfolio. And you attach a shadow liability to the illiquid part of the portfolio when you are playing defense, that is, you’re trying to preserve the quality of the portfolio to prevent it from suffering deterioration in its quality.
So if you do that then you literally introduce the shadow assets and liabilities, and you set their weight equal to the sum of the liquid assets, the shadow assets weight is set to the sum of the weight of the liquid assets, the shadow liabilities weight is set to the weight of the illiquid assets, and you re-optimize, and what you find is that your allocation, your optimal allocation to illiquid assets will be lower and sometimes significantly lower than what you would otherwise think is optimal in the absence of taking liquidity into account.
Meb: And we’ve also seen a lot of research and comments where, particularly for individual investors or advisors you know, private equity is an asset class, notoriously hard to do the research as well as to you know, you have to assume you’re picking quartile managers, otherwise you should just be in the S&P, I mean, is it worth…is there a public markets equivalent? Do you think that it even makes sense to consider private equity in asset class for kind of individuals or advisors, is it a good substitute or should you just ignore it altogether?
Mark: That’s a really good question. One of the earlier studies, actually by one of my colleagues in MIT, Antoinette Schoar, and a fellow at Harvard, Josh Lerner, showed that you did have to identify better performing private equity managers to do better than the public market. But more recent research shows that private equity on average…so if you just stick to private, the average private equity fund, on average, has added about 500 basis points a year, to the public market, on a risk equivalent basis. But it turns out…and this is not in the book, but based on some research that we’ve done here at Wyndham, and together with some of my colleagues at State Street, it turns out that most of the premium, this historical premium of private equity over public equity, can be attributed to the sector exposures for private equity funds.
So you can build a portfolio of public sector ETFs in a way that the sector exposures match the exposures of private equity funds, and that portfolio historically would have delivered to you 75% of the premium of private equity over public equity. So that’s something that individual investors can do quite easily. I think it’s also a very useful portfolio for institutional investors, anybody who invests in private equity. You know, most people probably realize that you commit a certain amount of capital but that capital isn’t called right away and you have to park it somewhere. Most people probably just park it in the stock market in the S&P 500. It might make more sense to invest it in a portfolio of public sector ETFs or securities that are structured to match the sector exposures of private equity funds.
Meb: And do you recall off-hand, I mean, is it like tack and health care or consumer discretionary? What’s the main allocations? Do you remember?
Mark: Well, they change through time. I’m not sure off the top of my head, and I should say, you know, you said we might get into the weeds here that I’m not sure I should, you can stop me if you think I am. When I say the public sector exposures, I’m not talking about the sectors that private equity funds think they’re in. Not the accounting exposures, but rather the economic exposures. And what I mean by that is if you take private equity performance, and you regress it on the returns of public market sectors, it’s those regression coefficients that determine the sector exposure. So it’s just to what extent do private equity funds covary with public sector performance? And I…off the top of my head, I don’t know what the current…
Meb: Well, that’s all right, we’re gonna post some show notes, and so we’ll do links to this paper, as well as some other private equity papers that kind of tackle this topic. And it’s an interesting area to me because you know, for the average investor, for the average advisor, one of the things that they spend most of the time thinking about certainly is asset allocation, but the implementation, you know, the things that are equally, if not more important…and you’ve talked about this and man, we’ve posted links to your papers on the blog years ago, is also the impact of management fees and you know, hedge funds are often two and twenty, for private equity funds similar, and then also taxes. And so a lot of hedge funds and private equity are run without any regards to taxes, whatsoever, so that can often have…the vehicle can have a very large impact on what you would consider to be included in a portfolio. So for both taxable as well as tax exempt.
Mark: Yeah, I think if you’re tax paying investor, the two most important things you can do is avoid taxes. So that’s basically lower turnover, and also try to postpone gains and realize losses, short term losses. And then the other thing is to diversify efficiently, you know, that’s way more important than trying to pick the manager who’s gonna outperform.
Meb: Yeah, there’s a great…and we’ll post again to the show notes there’s some great charts you have from some papers where it breaks out you know, index fund, versus mutual fund, versus hedge fund, and shows just how much alpha a really active and expensive mutual fund as well as a typical hedge fund with a 200% turnover in two and twenty. I mean, the hedge fund has to, on a gross basis, perform like nine percentage points higher just to get back down to a normal return for an index fund. So it’s a very high bar and a lot of people don’t think about it because they think of the sexiness of hedge funds and active management, but really, once you incorporate taxes, it’s a much higher bar.
I wanted to shift into one more topic because, man, this hour is flying by. There’s an area…so now let’s talk a little bit about the real world, and we’ve talked a little bit about draw-downs and the challenges of sticking to a basic portfolio once you’ve allocated to it. And over the years, you’ve talked a lot about you know, investors care about what happens in the meantime, and they care a lot more about losses, and you have a concept of what you call “turbulence.” And it has to be the first formula that I’ve ever seen inspired by human skulls, that has to be a first in finance. Maybe talk to us a little bit about turbulence and what that means for a portfolio?
Mark: A while ago, back in the late ’90s, I was trying to come up with a way of developing more reliable inputs to the portfolio construction process so we have to input standard deviations and correlations which in combination give us the covariances. And it occurred to me that it might be that you know, you observe returns every day, most of the returns we observe just reflect the fact that prices are noisy. Yet every now and then there is you know, some significant event that legitimately causes prices to change. And wouldn’t it be interesting if we could somehow distinguish returns that are attributable to noise from returns that are event driven?
And I wanted to do this in a statistical way so you know, we could be more efficient about it. From my co-authors and I at the time, developed this formula that measured how statistically unusual a set of returns is for a given period. So if we looked at the returns of the major asset classes yesterday…well, not yesterday since it was Sunday but last week. We want to be able to say, “Was that an unusual week or not?” And unusual in a statistical sense. So what it captures…so what “unusualness” captures is extreme price moves. So if one or more of the asset classes had a very large positive or negative return, that would qualify that period as unusual. But also, unusual interaction, so if two asset classes that are highly correlated, if they experienced divergent performance that would be statistically unusual. Two assets that were uncorrelated and their performance converged, that would be unusual.
So we came up with this formula that measured statistical unusualness of a set of returns taking into account, not just the size of the returns but also how the returns interact with one another. And we called that, “A Measure of Turbulence.” So it really captures instability in the markets. Now a friend of mine…so this was in the late 1990s, a friend of mine was presenting in our paper and you know, he drew up some new empirical results that we developed based on that paper, at a conference here in Boston. And a gentleman in the audience raised his hand and he said to my friend, “Well, that formula looks like the Mahalanobis distance.” So my colleague didn’t really know what the Mahalanobis distance was, but he was a pretty good marketing-type person. He says, “Oh yeah, you’re exactly right.” And then he returns to my office and he said, “Mark, I was just giving this talk and this guy said that our turbulence formula is the Mahalanobis distance.” And I said, “What?” And he said, “Mahalanobis.” So I said, “How do you spell this?” He goes, “I don’t know.”
So we went to Google and we started typing in how we thought it would be spelled, and sure enough, the Mahalanobis distance shows up. So we researched it and we found that there is a statistician in India who in 1927 wrote a paper where he introduced this, his name is Mahalanobis obviously. So what, he was doing he was working with an archaeologist and they had a collection of skulls, of human skulls from a particular you know, graveyard somewhere in India. And then they had another collection of skulls that they got from a distant battlefield, also in India. And these, the people who were in the graveyard and those who died on the battlefield were of different races within India. So he figured out that if he measured you know, the height and width of the skulls, the distance between the eyes, and everywhere you can possibly take measurements of a skull, and then he devised this formula. Did it in such a way that if you just gave somebody the number, the distance, he could tell you whether that skull came from the graveyard or from the battlefield. So it turns out that that’s the exact same math that we used to measure statistical unusualness in market returns, he was using it to measure, you know, to distinguish, “Is this skull unusual from this other skull?” Right, and the graveyard skull would be unusual if the benchmark is the battlefield and vice versa.
Anyway, I thought that was a really interesting connection, and if I have time, I’ll tell you one more connection, but let me go on about turbulence. So we came up with this way of measuring turbulence. What we found is it has two really important empirical properties. One is that returns to risk are much lower when markets are turbulent than when they’re calm. So if you look at the return of hedge funds, for example, and if you measure turbulence this way, if you look at say the 10% most turbulent periods, the hedge fund returns are gonna be much lower than they are the rest of the time. The same thing is true with stock returns, security trade and the currency market, small versus large or growth versus value. Anything where you’re increasing risk, it does worse when conditions are turbulent rather than when they’re calm. So that’s a big deal, and significantly worse like 10%, 15%, 20% worse, annualized.
The other thing is that turbulence is persistent, it’s not unlike the turbulence that we encounter when we fly. It may begin without warning, but we know if we’re in the airplane that it’s going to take time for the plane to pass through the weather system or to find a different altitude where it’s calmer. The same is true in the financial markets, turbulence may start without warning, it could be due to a political event, or some economic shock, but once it begins it typically takes several weeks for investors to digest and respond to what’s going on. So the combination of this big difference in returns along with persistence means that investors have an opportunity to tactically adjust their portfolio when markets become turbulent. By doing so they could…you know, we think that people can add value if they do it in an intelligent way.
Meb: I assume 2008, 2009 was kind of the granddaddy at least in our lifetimes, of turbulent periods. Is that right or is there some other periods that are pretty up there?
Mark: No, that would definitely be the largest, the most turbulent. Others would be you know, the ’98 Russian default on LTCM. You know, certainly, 1987. Some of the…in recent times, the events in China. So you know, there are a lot of periods and if you grasp the turbulence you’ll see it lines up with all of these events. But you’ll also see spikes in turbulence where there’s not really any news out there, so it’s capturing things that are sort of under the surface or are not on the radar screen of what investors typically have access to.
Meb: And so two comments there, one so my assumption is that right now it tends to be pretty not that turbulent, and you can comment on that. And two is, is there a way for people or listeners to follow along or track this? Do you guys publish any turbulence numbers or is there any…? Or should they just go build an Excel sheet on their own? What’s your thoughts on both?
Mark: You can use this in a couple of ways. One is to build portfolios that are more resilient to turbulence, and the way you would…so without trying to dynamically adjust your portfolio, if you wanna build a portfolio that is more, in my judgment, more efficiently diversified, you shouldn’t worry so much about correlations that prevail during calm markets, you should worry more about correlations that prevailed during turbulent periods because it’s during turbulent period that you want to diversify against losses. You don’t wanna diversify against gains, right?
Meb: And so what is that from a practical standpoint? Does that mean investors should have more in government bonds? Are there other asset classes? I mean, is it like managed futures or some sort of hedging vehicles? What’s practical?
Mark: It might be yes, certainly more in bonds and certainly, perhaps more in commodities. It depends on you know, all of the assets in your portfolio. So what you should do is you should estimate your correlations and volatilities based on the subsample throughout history that were turbulent, and then you know, use that as inputs to your optimization process. And the other thing is you should do if you want to stress-test a portfolio, you should stress-test it based on you know, simulations that are from turbulent subsamples rather than the full sample.
So that’s something that everybody can do, you know at Wyndham, you know, we’ve developed a suite of software applications and separating turbulent periods from calm periods is one of the features within that software, so it’s pretty easy to do.
Meb: And I imagine with…right now it seems like a pretty smooth sailing period. Is that showing up in the numbers as well? I feel like it’s got to be, the VIX, I think is closing today below 10?
Mark: Right, yes, it’s been very calm, lately. I would say that turbulence is a much better measure than VIX for measuring instability because it takes into account interactions whereas VIX does not. But one of the other measures that we rely upon is a measure of risk concentration. And this is whereby we do a factor analysis to determine the factors that are driving variability of returns, and then we compute what fraction of the total variability of returns is explained or absorbed by two most important factors. And we call that the absorption ratio. And it’s not widely used throughout the world, but what I think about it is, if the absorption ratio is high, in other words, if a few factors explain a large fraction of the very variability of returns, that tells us the risk is very unified, very concentrated. And when that’s the state of the world, shocks travel more broadly and more quickly. And the most draw-downs, historically most large draw-downs were preceded by spikes in the absorption ratio. If the absorption ratio is low, what that means is that these same key factors explain only a small fraction of the variability of returns. And what that means is that risk is distributed broadly across many different sources. And when that’s the state of the world, it’s relatively resilient. So that measure also currently is somewhat low, but it’s been trending upward for quite a while now.
Meb: It’s always interesting to me too to think about it because often the media and kind of geopolitical news flow doesn’t necessarily line up with markets, you know, there can be times when the news flows seems crazy and markets just kind of shrug it off and vice versa. There’s times where the news flow seems kinda chill and markets are already…like and I kind of thinking of ’07 as a good example here where the real estate had already cracked and you’re starting to see all these, you know, fissures going on and markets were starting to react but the news flow hadn’t really caught up. And maybe that goes a part of the theory that markets are a leading indicator, but it’s interesting to kind of have this perspective that there are some additional metrics that may be useful and starting to look at least not necessarily predictive, but warning signs on when things are starting to get a little yellow warning light.
So Mark, we only have you for a few more minutes so I got to kind of wind up quickly here. We should have blocked off two, three hours because there’s so much to talk about. So for the listeners here that listened in, we’ve kind of thrown a lot at them, a lot of theory, a lotta kind of ideas. What’s kind of your main takeaways? If you’re to give advice to an advisor or an individual on asset allocation at this point, with everything they’ve kind of heard today, what are your kind of final summary ideas from the book and just your career as a money manager, that you think would be most useful to our listeners?
Mark: So my advice is to approach investing as scientifically as possible. Which is not to say you want to just apply formulas without any judgment, but it’s to, you know, it’s to use the math that we have, to use the data that we have in an intelligent way, and to combine it with good judgment. And I think investors you know, should you know, either on their own, they should approach investing in a very structured, disciplined, and mathematical way, or engage with people who can provide those kinds of services.
I think we need to take into account that the world…the models that we have, the asset allocation models that we have are…they are just that, they are models, they’re simplifications of the real world. And we need to embrace the complexity of the real world, it’s not often convenient to do that but I think it’s very important to do that. So we have to question the assumptions that underpin our models. A lot of these assumptions are necessary to get tractable solutions, but they’re not…typically are not always reflective of reality.
So I guess, my main message is nothing is simple, use the science as best as you can, but don’t get over reliant on models and simplifying assumptions. Put in the effort to engage with the, you know, to embrace the complexity of the real world.
Meb: I like it. So we get to wind down here, sadly. Mark, where can people find more information about you, follow your writing, follow along with everything we talked about today? What is the best places to find more info?
Mark: windhamlabs.com I guess would be the easiest way. Or people can just you know, Google me and send me an email if they’re interested.
Meb: Oh, you just invited a ocean of emails, I apologize ahead of time. Look, Mark, it’s been fun. So we’ll add show notes, links to Windham Labs with everything else. Mark, thanks so much for taking the time out to chat today.
Mark: Thank you, I really appreciate it. It’s been a pleasure.
Meb: Listeners, check out Mark’s new book, “A Practitioners Guide to Asset Allocation”, out this month, as well as all of his other writings which are a lot of fun. Thanks for taking the time to listen today, we always welcome feedback. Please send us questions to the mailbag at email@example.com. As a reminder, you can always find the show notes and other episodes at mebfaber.com/podcast. Please subscribe to the show on iTunes. If you’re enjoying the podcast, leave us a review. Thanks for listening friends and good investing.
Sponsor: Today’s podcast is sponsored by the Idea Farm. Do you want the same investing address as pros, the Idea Farm gives small investors the same market research usually reserved for only the world’s largest institutions, funds, and money managers. These are reports from some of the most respected research shops in investing. Many of them cost thousands and are only available to institutions or investment professionals but now they’re yours with the Idea Farm subscription. Are you ready for an investing edge? Visit theideafarm.com to learn more.