Episode #121: “The Reality Is High-Risk Stocks Earn Low Returns”
Guest: Pim van Vliet. Pim is the Head of the Conservative Equities team, responsible for Robeco’s Low-volatility strategy ‘Conservative Equities’. He is also Portfolio Manager Quantitative Allocation strategies. He has published in the Journal of Banking and Finance, Management Science, the Journal of Portfolio Management and other academic journals. Pim is also a guest lecturer at several universities and advocates low-volatility investing at international seminars. He is the author of numerous academic research papers and a book on low-volatility investing.
Date Recorded: 9/05/18
To listen to Episode #121 on iTunes, click here
To listen to Episode #121 on Stitcher, click here
To listen to Episode #121 on Pocket Casts, click here
To listen to Episode #121 on Google Play, click here
To stream Episode #121, click here
Comments or suggestions? Email us Feedback@TheMebFaberShow.com or call us to leave a voicemail at 323 834 9159
Interested in sponsoring an episode? Email Jeff at email@example.com
Summary: In Episode 121, we welcome fellow quant, Pim van Vliet. If you’re a low-vol investor, or having been wanting to learn more about low-vol, this is the episode for you.
Meb dives straight in, opening with a quote from Pim: “The low-volatility effect is perhaps the largest anomaly in finance, challenging the basic trade-off between risk and return, as higher risk does not lead to higher returns. Still, it remains one of the least utilized factor premiums in financial markets.” He asks Pim to explain.
Pim tell us that low-volatility is the biggest anomaly of them all. People have trouble embracing the concept. We’ve been trained to believe that higher risk should be rewarded with higher returns, but Pim walks us through some counterarguments. He goes on to explain that CAPM (Capital Asset Pricing Model) is great in theory, yet bad at describing reality. He tells us that “the reality is high risk stocks earn low returns.”
Next, Meb brings up a paper Pim wrote called “The Volatility Effect” and asks Pim to walk us through it. Pim tells us one of the broad takeaways is that low-vol works cross borders (unlike some other factors). It’s not just effective in the U.S. – it’s also been proven out in Europe and Japan. In addition, this alpha seems to be getting stronger now rather than waning as have other factors when their visibility has increased.
Meb asks about Rob Arnott and factor-timing/factor valuations. Does factor valuation matter?
Pim agrees with Rob in that valuation does matter. If you only look at low-vol, you might end up buying “expensive defensive”. If so, then yes, your expected returns will be lower. That’s why Pim includes a value filter. He looks at “multi-factor defensive”. Pim mentions Cliff Asness and notes that he likes incorporating momentum into his approach as well.
The conversation bounces around a bit: where is Pim finding opportunities around the world now… additional details on how low-vol works across countries, sectors, and asset classes… and how low-vol complements a CAPE approach, pointing toward some effective defensive market strategies.
Next, Meb asks about potential biases. For instance, if you focus on low-vol, could that mean you’ll end up with a basket of, say, utility stocks and no tech? Pim tells us that, yes, if you focus purely on low-vol, you could get more sector and country effect. But he goes on to tell us how investors might mitigate that.
There’s plenty more in this fun, quant-driven episode – a discuss of the definition of risk (volatility versus permanent loss of capital)… factor fishing and data mining… how low-vol works from a portfolio perspective… Pim’s forecast of the future… and Pim’s most memorable trade. This is a great story, highlighting how an early loss delivered such a powerful learning lesson, that it probably ended up making Pim money in the long run.
Get all the details in Episode 121.
Links from the Episode:
- 0:50 – Welcome
- 1:25 – Pim’s thoughts on low volatility
- 4:26 – Defining CAPM
- 6:39 – “The Volatility Effect: Lower Risk Without Lower Return” – van Vliet & Blitz
- 6:54 – The New Finance – Haugen
- 6:55 – The Inefficient Stock Market – Haugen
- 11:06 – Why low volatility works
- 14:40 – Where Pim falls on the issue of factor valuation and stock timing
- 20:20 – Finding opportunity in low volatility strategies
- 22:30 – Does low volatility work everywhere
- 25:00 – Recent paper on Low Volatility and CAPE
- 27:59 – Best way to implement a low volatility strategy
- 28:20 – High Returns from Low Risk: A Remarkable Stock Market Paradox – van Vliet & Koning
- 30:44 – Will low volatility bias an investor toward specific sectors?
- 32:57 – Biggest benefit of low volatility
- 37:17 – Pim’s other areas of focus
- 40:44 – Pim’s thoughts on crypto
- 41:20 – Low volatility across asset allocations
- 43:56 – Looking ahead, and the trends that Pim sees as influential over the next 10 years
- 48:33 – What Pim’s students are thinking about
- 51:24 – Pim’s favorite surf destination
- 53:20 – Most memorable investment
- 57:08 – How people can follow Pim; @paradoxinvestor on Twitter, Pim’s paper, Robeco
Transcript of Episode 121:
Meb: 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.
Disclaimer: 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.
Meb: Welcome, podcast listeners. Today we have a great show for you, all you quant lovers. Our guest is the head of conservative equities and quant allocation at Robeco. You’ve read his writings in the journal “Banking and Finance Management Science” and the journal “Portfolio Management.” I think he’s written, like, 25 whitepapers on SSRN. He’s also a guest lecturer, and we’re really happy he’s joining us today. Welcome, Pim van Vliet.
Pim: Hi, Meb. Happy to be on your show to focus on quant stuff.
Meb: Yeah. Well, we can talk about anything. We can talk about surfing, we can talk about quant, we can talk about anything else. But let’s dive right in. All right, so we got a quote Jeff found that you’ve written that says, “The low-volatility effect is perhaps the largest anomaly in finance, challenging the basic trade-off between risk and return as higher risk does not lead to higher returns. Still, it remains one of the least utilized factor premiums in financial markets.”
So, why don’t you unpack this quote for us a bit, maybe explain low-vol, anomaly and your thoughts on low-vol in general as a good starting point?
Pim: That’s great. Let’s do it. So, low-volatility is, I think, the biggest anomaly of them all in the sense of that we really don’t understand why this is the case. And also, academic and practitioners having trouble embracing this concept.
So, when I started my studies, like any finance student, I learned about the positive trade-off between risk and return, and higher risk should lead to a higher return. But it’s empirically not the case.
So, Fama-French ’92 showed it, but also earlier on in the ’70s, this was documented. So, if you knew where to look, you were looking at it. The thing is, in academics, that’s the first point I wanna make, is that returns are usually measured as simple returns. So, that means that all academic studies using simple returns underestimate the size of this anomaly.
So, high-risk stocks look like they have a pretty low return, but if you add the compounding, then it’s really, really bad. So, if you read through the literature, like Jensen-Scholes in the ’70s, but also Fama-French ’92, then you should know that this is understating the size of the anomaly, namely that high-risk stocks underperform.
And if you compound it out, because high-risk stocks, by definition, go up and down a lot, but if you go down minus 50 and go up 50 again, then all investors know that you’ve lost money, but in academia, then, you assume that [inaudible 00:03:39]. So, any high-volatility strategy, the return of that is overestimated in academia.
So, that’s a very important point to make. And it’s not the mistake of the CAPM or the asset pricing model, but it’s the mistake or it’s the false assumption of the researchers using the CAPM to test it.
Even today, with all the multiple factor extensions, the size of the low-risk anomaly is understated because of this compounding. Because CAPM is a one period model, and this one period is often assumed to be one month, while in reality, there’s also, like, one year, or maybe 10 years.
And if you add to that, then you will see that the return goes down with volatility. So, if you short on volatility, or risk on beta or anything, risk metric, then you will underestimate the alpha.
Meb: Could you explain for our listeners who might not be familiar with CAPM exactly what that is? And it’s kinda crazy because you have so many of these ideas in academia that get taught that are, in some cases, not only not true, but 180 degrees not true. Like, it’s true, but the sign is backwards. Maybe explain a little bit what CAPM is and why it’s still accepted today if some of the tenets are not accurate?
Pim: CAPM is a great theory, and when I teach, I also use it. I also believe that it’s a good theory. It’s just very bad at describing reality.
In fact, a stock with a high risk should have a high return, and that’s written down in the CAPM. Only systematic risk should be priced. That’s an extension of the Markowitz model. Markowitz would state that you should diversify so you get rid of your idiosyncratic company risk, and the only true risk which is priced is market price. Stocks with have a high beta or a high sensitivity to market fluctuations should have a higher return.
And also, CAPM can be used in corporate finance settings where you make investment decisions in projects. Then you should allocate your capital to low-risk projects which have high returns. But it’s a normative framework, but as a theory of trying to predict, to describe markets, it’s really poor, because reality is that high-risk stocks earn low returns.
And it’s very funny in that when I found out on this, so, it was work by Robert Haugen, and I read it as an undergrad student, I was really shocked that low-risk stocks have high returns and high-risks have low returns. And by the way, he was, basically, really attacking this capital asset pricing model where he showed this model might be a great theoretical model but it’s really, really poor at predicting and describing markets.
So, that’s the CAPM. This is the theory. And I like it, but it’s just not…and I also base my own investments on it. Good to use it as a tool, but it’s a very bad description of reality.
Meb: You mentioned an early writer on this topic that doesn’t get enough credit or, sort of, headlines. He passed away a few years ago, but he wrote a handful of incredibly influential books. If I remember, “The New Finance,” “The Inefficient Stock Market,” and a couple ideas on…we’ll post these to the show notes, but fantastic reads on this topic that were really early in some of the ideas you’re talking about.
And so you’d written…maybe to unpack this a little more, you had written, about a decade ago… God, 2007, a decade ago, you wrote a paper called “The Volatility Effect.” And maybe walk us through some of, like, this concept of a simple, alternative approach to constructing portfolios and maybe, kind of, some of the original ideas of that paper, and takeaways.
Pim: Back in 2007, we wrote this paper, “The Volatility Effect,” where we basically laid out that low-volatility is a factor, and besides factors such as size, value, and momentum. In this paper, we take an international perspective. The first most-studied are US-centric. I guess it’s all about the market cap, but the other [inaudible 00:07:55] is international. But we’ve showed that this low-vol effect is also present in Europe and Japan, which makes it a very persistent anomaly.
Other factors sometimes don’t work in other regions. Like, momentum has difficulties in Japan. But that’s one thing we wrote down there. The second thing is that we found that this anomaly, or this alpha, it’s getting stronger over time. So, most alphas, once they’re documented, they become a bit smaller like we’ve seen with value nowadays, struggling for the past 10 years, much smaller than in the past 80 years. Like low-vol, it’s a factor which keeps getting stronger over time.
And thirdly, in our paper, we also gave explanations for why this might be the case. So, why could it be that low risk beats high-risk stocks? You asked about how do we do this. So, a simple way to test factors is to take, say, 2,000 large-cap stocks, so the 2,000 largest stocks in your universe. You show them based on their historical price movement.
So, you sort them from low-volatility to high-volatility. You can also sort them from low beta to high beta. And then you dynamically construct a portfolio where you systematically buy stocks which are the lowest and then, a quarter later, you do it again because some stocks might move from low risk to high risk, or then [inaudible 00:09:18] might move [inaudible 00:09:20].
So, if you do that, then you construct portfolios which have very stable risk properties, and then you’re just gonna test if this low-risk portfolio has, also, a low return, because that’s what the CAPM, as we’ve discussed, predicts. And the fact that it totally [inaudible 00:09:36] the world upside down, it’s the other way around, it’s negative.
If you add the compounding to it, so the longer your investment horizon, the bigger the alpha becomes, that’s also what we lay down in the article. And the interesting thing is that, of all the factors out there, the low-vol factors have been driven by practitioners.
So I’m a pracademic. So, I’m a practitioner with also academic ties, but also Haugen was not a main academic. So, he was not part of the opinion-makers on the big U.S. universities. He was, sort of, an outcast. And also, if you read his books, it’s very funny. He’s really critical of the Chicago School. He makes fun of them. But I also understand why he was not loved by his peers in academia.
But this low-vol factor is something different than the others. It’s a different factor, and nowadays it’s more accepted. Also, the work of Eng [SP] [inaudible 00:10:33] has contributed to this, and also some researchers like Frazzini and Petersen have also brought this anomaly more to the mainstream academia, and nowadays it’s more or less accepted.
The interesting thing is that the practitioners have, sort of, led it. And also, if you look at asset managers, I’m based in the Netherlands, a beautiful country, all listeners should come here, but it’s also driven by European managers while most of the innovation in finance is driven by Wall Street or the U.S.. Well, this is a little bit of a European phenomenon, and also led by practitioners.
Meb: What are some of the explanations? So, you think about low-vol. I mean, you would think intuitively that most people would love low-volatility stocks, you know, given the alternative, with the research showing that they outperform.
You’ve mentioned some possible explanations for the success of the strategy and/or why it’s inefficient, everything from leverage restrictions to inefficient industry practice, behavioural biases. What sort of specific reasons…you know, we always love to say, “Why does this work?” What are the major contributing factors in your mind?
Pim: Yeah, so, I spent four years of my life doing my Ph.D. on downside risk, but my thesis was…constitutes all these factors. So value, momentum, but also low risk, couldn’t be explained by downside risk, so couldn’t be [inaudible 00:11:58] because the CAPM also [inaudible 00:12:00] mean variance returns or quadratic utility.
I’ve worked on that model. Very complicated, technical. Long story, four years. Short story, not really. [inaudible 00:12:12] risk is not really expedited. You can partly explain some with the low [inaudible 00:12:19], some with momentum, and some with [inaudible 00:12:25].
I entered the industry. I came through Robeco, the firm I’m still working with. I met the head of research, except I found him working on multifactor models for the [inaudible 00:12:33] 2005. We did it multi-billion. We knew the work of Robert Haugen. We explained you should buy stocks with good [inaudible 00:12:42] and build factors [inaudible 00:12:45].
I was surprised to find out when I entered the industry that low risk/low-vol was not exploited. And I spoke with our head of research. I still remember it. And I said, “Hey, David, why aren’t we adding low risk as a factor? And so, low-vol?” And then his answer was very simple. He said, “Pim, you get equity [inaudible 00:13:04] returns might a bit better, but you get lots of tracking error.”
And I said, “Pardon me? What are you saying about errors? What kind of error?” And then he explained this concept of benchmark deviation, just it’s a more agnostic term. Tracking error. Tracking error is an industry concept. It’s very elegant, and I understand why [inaudible 00:13:24] use it, but it’s turning the concept of risk upside down.
low-vol stocks have huge tracking errors. If markets go up, they lag. If markets go down, they do better. That’s called tracking error. So, if you put it in a multifactor model where the aim is to consistently outperform a benchmark each year, after year, after year beating the index, then low-vol stocks, basically, it’s a weak factor.
And then I started to understand, like, “Hey, this is interesting.” There are limits of arbitrage, benchmarks, leverage constraints. And then it got me going. I’m a bit of a contrarian guy. I thought if it’s difficult to harvest for us, it’s probably difficult for others as well.
And yeah, we built strategy around it. We called it conservative equity. And that was 12 years ago. And yeah, ever since, I’ve been working on this fascinating topic, and I still, until today, when you meet clients, prospects, talk about the concept, and each time the benchmark comes back. So, that’s really an explanation I don’t expect.
Meb: I think you’ve done a little too good of a job starting to educate people because I feel like our world is becoming a little more knowledgeable about this concept, I would say, over the past cycle. And so, because of that, you have some of the industry titans like Azinus [SP] and Arnott starting to weigh in on, in the last few years, this concept of particular factors and valuations, and whether you can time factors.
So, in this quant deathmatch, where do you stand in, sort of, that spectrum of, yes, it matters, the factor valuations, and therefore you can also time them, or there’s times they work better than others? Or, on the flipside, maybe it’s nearly impossible to time them even though you can see, at times, when it may be expensive or not. Where are you on the spectrum?
Pim: Yeah, that’s a good one. So, I really like fights. So, catfights are great, but also quant fights or quant debates. I enjoy them every time they occur.
So, for example, ongoing…so, timing, back to timing. There’s a debate whether correlations matter. My standpoint is this. I think if you look at [inaudible 00:15:44] now, he stresses the fact that valuation matters. I agree with him on that.
If you look at low-volatility standalone, so you simply, only look at low-volatility and that’s it, then you might enter up buying expensive defensive. And that’s currently…for the last couple of years, that’s the case. And our research shows, and his research, that if you do that, then your expected return will be lower.
So, that’s why I include valuation, we include valuation, ever since 12 years ago, in our strategies. So, we do multi-factor defensive. So, we include valuation or net payout yield, which is also a valuation and quality metric. If you screen your low-vol stocks on that, you can still find cheapness. And you don’t want to go against value.
If you look at the generic low-vol strategy, so simply looking at the single-factor low-vol strategy, there is some finding in it. So, if it’s expensive, the returns are lower. If it’s cheap, returns are high. But if you do a consistently cheap low-vol strategy, then this goes away. Timing is much more difficult or, after transaction costs, is not possible anymore because you simply are always on the right side of value.
You also mentioned Cliff Asness. Yeah, great guy, also a pracademic. [inaudible 00:17:05], I respect him. He also emphasizes momentum. So, in his thesis, he we worked on that, and especially combining momentum with value. And I’m…also agree on that one. So, if you only do value, like if you see with value the last 10 years, it’s weak, but the trend is your friend. Also, he wrote on trend-following strategies. That, it’s a factor which worked for hundreds of years. It also captures beta. Academics also have a problem grasping with momentum.
So, yes, you should also always include momentum in your strategy. So, if you do a defensive strategy, make sure you do it multidimensional, multifactor.
So, I also posted another paper on SSRN, and it’s published in the “Journal of [inaudible 00:17:53]” of this summer, where we say if you build a strategy based on low-risk income and momentum, so income as shareholder [inaudible 00:18:04] that’s also how [inaudible 00:18:04] we refer to, if you do that, and we propose a simple formula, then basically, you get access to all the factor premiums out there, all of them. So that’s [inaudible 00:18:14] factors, [inaudible 00:18:15], and also the recently [inaudible 00:18:18] factors.
And the reason I wrote that paper is if you look in the whole academic literature for 50 years, basically, if you look at Robert Haugen’s work, you already know 90%. So, not much happened after that, I would say. It’s not that there’s some new insight. The problem is, how do you translate all these academic insights into a simple strategy?
And that’s what we’ve tried to do in this recent paper. It’s called “Quantitative Investing Made Easy: The Conservative Formula.” And there we say that the rule of three, so that’s value, momentum, yes, but also low risk, combining them gets you a very strong portfolio.
So, in the whole [inaudible 00:18:56] debate, I agree with the value guys, I agree with momentum guys, but I do see that low-vol or low risk is, sort of, not getting the attention it should get because it matters a lot, especially in the long run if you add compounding to it. And also, if you look at the academic evidence, it’s very persistent and strong, and stronger than some of the others.
And alos, the final thing about why you should look at low-vol is that your level of understanding, so why it’s working, that’s important for a factor. So you should know why it’s working. Otherwise, it can be arbitraged away if it’s irrational behaviour. That the guys buying the high-vol stocks are often very intelligent, highly-educated guys, CFAs, Ph.D., that just…who have another incentive.
And that makes it very interesting that this anomaly can be understood from a rational point of view, so not just the behaviour, but structural impediments.
Meb: You’re a man after my own heart, talking about some of these factors. So we’ve got low-vol…which, by the way, what’s a generic representation of low-vol, as you mentioned? Is it volatility over a certain time period? Is it volatility relative to a sector or market? What’s a generic, kind of, factor, expression of that?
Pim: Short answer? Three-year historical volatility, 156 weeks.
Pim: And you have something going.
Meb: Okay, perfect. So, in your book and in your paper, you mention some really simple multi-factor portfolios. You talk about vol and then pairing it with payout yield, or shareholder yield, which is basically dividends or net buybacks, which is one of my favourite factors. So that gives you, like you mentioned, a little value and quality, and it has pretty high correlation to free cash flow, priced free cash flow, and momentum.
And so, talk to me about, sorta, where that’s finding opportunity around the world right now. Is it something that you do globally and, you know, you’re finding the most opportunity in any regions or countries, or is it something that you’re finding opportunity everywhere, or there’s…nothing looks good? What would a screen like that look like today?
Pim: We are international investors, so I’m based in Rotterdam. If I step on my bike, then I can be in Belgium in an hour. So, we’re international investors by default. So, if I look at factors, I also look at them internationally.
The interesting thing is that value has problems a lot of the time, but especially in the U.S. If you look at emerging markets, value is pretty good, also recently. But we do see a lot of value in emerging markets. Also, right, we manage more than $20 billion in our strategies. A significant part of that, about 1/3, is in EM. So that’s where we see lots of added value. Also, there, the volatility is very high. These stocks are very volatile.
So, that’s where we see value. And in the U.S., you see that the U.S. market is quite expensive, and there you see that low-vol and value are fighting. And when factors are fighting, that’s often a tough period, and…but we still find relatively cheap U.S. low-vol stocks.
But as you know, that factor has been driving the U.S. for the past two years. Any multifactor strategy which gives way to value has no problems. That’s where momentum helps to keep up a bit and reduce the pain [inaudible 00:22:38].
Meb: You know, it’s interesting. I mean, we often say, for multifactor, our favourite combination is along the lines of my buddy Steve Sjuggerud says, “Cheap, hated, and in an uptrend.” So, kind of combining value and momentum.
You mentioned earlier an interesting point, which was maybe momentum doesn’t work, necessarily, in Japan, which is sort of odd. Is low-vol a factor that seemingly works everywhere in the markets you’ve tested, or is there some where it actually doesn’t work or it tends to not work as well? Is there any broad takeaways?
Pim: Yeah, that’s a good one. So, low-volatility works very good within all countries, all sectors. It’s getting a bit strong stronger, also, over time. We do see that in earlier samples, like in the ’40s and the ’50s, it’s a bit weaker. But still, it’s reducing risk-adjusted alpha.
What you do see is that it’s only not working across asset classes. So if you look at bonds versus equities, then you might say bonds are low risk and equities are high risk. But on that level, you don’t see it working. So, across markets, there’s not a low-vol effect. Also, across currencies, it’s more difficult.
So I’m working, currently, on a 200-year study of factor investing. So you will [inaudible 00:24:02] it when it comes out. Trends, of course, strong value, strong with low risk. We see it to be, across-markets, a bit more difficult, and we understand it because the reasons for why low-vol effect or low-vol works is because of relative utility benchmarking [inaudible 00:24:19] constraints.
And on the asset class level, that’s not really the case. You’re making your asset allocation decision not on anti or tracking error but just on your risk aversion. So, then, in the second step, then, when you move into the equity market, then you start to be, kind of, biased to high-risk.
So, that’s very fascinating. We also find it within credit markets. So we’ve done work on investment-grade [inaudible 00:24:42], also very strong there. And again, you see that within those two asset classes, it’s strong, and across, it’s a bit weaker. And you really see that it’s important within asset classes.
And yeah, Meb, maybe…I know you love momentum, and also CAPE you’ve been working on, but yeah, I would strongly suggest, also, to look at low-volatility as a factor because it makes the ride more smooth.
Meb: You just had a paper come out, like, a couple months ago that was kinda looking at low-vol and CAPE. And you know, a lot of people we chat with are probably rightfully concerned about the U.S. market here domestically, about it being expensive, and what should they do, and it’s a year…depending on who you ask, year-10 bull market, but certainly coming off some really long period of strong returns.
And you talk about in this paper how you examine low-vol strategy might be a good defence against a expensive market. Could you unpack a little bit about what that paper talks about and some of your conclusions?
Pim: Yes. So CAPE risk [inaudible 00:25:50] much talk. [inaudible 00:25:51] Nobel-prize winning. The debate basically goes around the fact, does CAPE predict returns? And the answer is yes, but in the longer term. There’s a lot of buts. There’s a lot of pros and cons to it.
But we live…in our study, we said we are not gonna use CAPE to predict returns. We believe that with high CAPE, expected return a little bit lower. However, we said, let’s turn it around. Let’s use CAPE and then predict risk and not look at returns [inaudible 00:26:26].
And what you then see happening is that it’s CAPE, it’s above 30 like today, then downside risk of equities is really bad. So, the [inaudible 00:26:36] portfolio of CAPE on risk is very high. And by putting these things together for CAPE, and not relating it to return because people might say, “Yeah, return is a bit lower, but who cares? Bonds are low.” But if you really show, and that’s what we do in the paper, that high CAPE gives you a high expected downside risk, then we see a real danger of high valuations. So, that’s why I think we can have a value strategy, because then you can use it.
What we do in the paper is that we then show, report how is low-vol doing in such a scenario. And then we see that low-vol stocks do what they should do, namely prevent risk and reduce risk, especially in a high CAPE scenario. But the point is, with low risk, is people…either, if people are afraid of the market, they often try to time it, and they often do it in the wrong moment.
So, if CAPE is high, you might conclude, “Let’s go out of the market.” But then you will miss out on long-term [inaudible 00:27:35]. If you do a low-vol strategy, then you can always be invested in the equity market and our prospective return without have to time the market, and then you can have it both ways. And you can have your return, but you will also [inaudible 00:27:51].
And that’s what we wrote down in our paper, and yeah, we see it resonates very well with our clients all around the world who can use it for their outlook. Becaues many are fearful of high valuations, especially in the U.S. But then, yeah, it’s too much to go out, and then a defensive study might be [inaudible 00:28:10].
Meb: Interesting. So, as an investor listening to this, obviously, they could just go allocate to some Robeco funds. But let’s say they had said, “Hey, I’m gonna do this on my own,” or, “I wanna implement…” Any general considerations to investors as to the best way to implement such a strategy or to think about it in the context of a portfolio?
Pim: Yeah, good one. So, I also wrote a book on low-risk investing, and I describe the story of my life about low-risk and how it can be done, and I especially focus on the implementation, as your question implies.
So, we say there can be three ways to do it. So, either it’s do it yourself, and then a great way to do it is screen your universe, if you have a broker account, and simply sort your stocks on beta, and then throw out all the high-beta stocks. And then you screen on a dividend or a shareholder yield, and then you throw out all the low-yielders. And then, finally, you screen on momentum, and then you kind of pick the ones [inaudible 00:29:13] do well.
Then you can end up with a portfolio of about 100 stocks, and then you can do it yourself, rebalance each quarter, and make sure you don’t trade too much. So, that’s for the guys who really have some money, who can diversify across 100 stocks, who like it. But they should be aware of the details because if you own 100 stocks, it needs some time to manage to your portfolio.
That’s step one. That’s one way to do it, but do it yourself. And in the book, we provide easy ways to do it. And also, if you read my paper which is free on SSRN, there are other ways to see how you can do this.
The second is to buy an ETF. There are trackers of low-vol indexes around, many of them also in the U.S. Then you buy, simply track a low-vol index. Then the problem is that you don’t include value/momentum, like, you don’t believe in that. So, you should also look at that.
And then you can…the third way we present is, it’s by active funds, active low-vol funds which have a multi-factor approach which do include value/momentum. And one of those [inaudible 00:30:23] I’m managing myself as fund manager, but that’s full disclosure.
But we wanna say, there are multiple ways to get exposure to this great premium. This low-volatility effect will benefit, yeah, your long-term wealth. And yeah, your show also says it’s about preserving your capital. I think that’s very important in investing and often overlooked. But risk…people often focus on return, but risk is really an important factor.
Meb: So, you mentioned…and by the way, your book, “High Returns from Low Risk,” we’ll post a link in the show notes. Great book, fun, easy read.
You know, and a lot of factors, intentionally or not, can potentially give you bias toward certain things, neither necessarily good nor bad. I mean, we often talk about, if you’re a global investor, whether you do top-down or bottom-up, you end up in certain countries just because certain countries have 3,000 stocks and some only have, probably, 100.
Are there any sectors that a low-vol approach traditionally will bias you to? So, I imagine a listener may be listening and say, “Well, am I just gonna end up with a bunch of utility stocks with 80% of my portfolio and never own any Amazon?” Is there any sort of tilt that this strategy ends up with? If so, is it good? Is it bad? Are there ways to correct it? All your thoughts on that.
Pim: Yeah. First of all, you do get tilts like with any factor, and you do get tilts to more defensive sectors. Currently, that’s telco or utilities. However, telco was not always defensive, so that’s time-variant. If you do a multi-factor approach, these tilts get less. So, if you do pure low-vol, nothing else, you get more pronounced sector effect and also contrary effect. And these can be mitigated with a multi-factor approach, and they can further be mitigated with sector constraints.
So you could say, “Hey, I want a maximum of 10% in one of the sectors, compared to an index.” Then you can manage this. However, it’s not…if you look at the simulations, these fields are not really bad, but they help your performance because it also works [inaudible 00:32:51] to time and to select checkers. However, we do get some quite big potential deviations, and those can be managed. That depends on your risk tolerance.
But we have very mild sector constraints and we have this multi-factor approach. Both of them help get a very broadly diversified portfolio.
Meb: I wonder what your response would be. You know, certain people out there, like Buffett, you know, would say, “Look…” You know, they don’t necessarily equate risk with volatility but, rather, the permanent loss of capital. And I imagine some critics listening would say, “Hey, okay, low-vol, sounds great. But you know, value sounded great going into ’08, and some of the value managers got taken out to the woodshed because they followed value all the way down and lost 80% drawdown during the bear market.”
Do you think that, in terms of drawdowns or permanent risk of loss, what’s the biggest savior or benefit of low-vol? Is it pairing it with value, or is it pairing it with momentum, or is there certain unique attributes that help protect low-vol in general to the, kind of, dual threat of large drawdowns and permanent loss of capital?
Pim: Generals tend to fight the last war. So, when the French were preparing for a Nazi Germany, they were basically repeating World War I in their simulations. And then the Germans came out with a Blitzkrieg and they did it differently, and then they crashed through the French.
What’s interesting, I often see the same. So, back in ’06, when we started our low-vol strategy, I met many, many people who said, “Yeah, Pim, low-vol is just value in disguise.” Because back then, value was low-risk, and low-risk was value. So they said, “Pim, I don’t…low-vol is nothing new. It’s old wine, new bottle. I have my value manager management style, and I’m fine.”
So, then, basically, they’re looking at ’01 and 2000 with the burst of the tech bubble, and they thought they were prepared for the next crisis. They were not. Nowadays, I mean, people I see…and back then, low-vol was cheap, so that was good. Back then, after the crisis, of course, people knew that value didn’t offer protection, low-vol did. So many people said, “Hey, that’s great. It’s different.”
Then low-vol became expensive, and then sometimes the same people said, “Hey, Pim, low-vol is expensive. I don’t like low-vol.” But then I said, “Hey, listen, at least you know it’s not value. So, you see? It’s a different factor.” So, at first they don’t like it because it’s value, and they don’t like it because it’s not value. And there’s always a reason not to like it. That’s interesting.
The second thing is that nowadays, many people who are preparing for the next crash might think that low-vol is offering them protection because back in ’08, it did. I’m pretty sure it will offer protection. However, I’m not 100% sure. So, that’s why I think, in a new crisis, it could be that value will offer protection again.
So, why would you bet on one factor? Why would you bet only on low-vol if you can also include value? So, that’s what I believe in, that to be prepared for the next crisis, it will not be ’08, it will not be ’00, but if you look at all the crises of the past 100, 200 years, a mix of momentum, value, and low risk is the best.
Because I’m totally agreed that past volatility is not the same as risk. Of course not. The thing is, risk is from relative loss and the amount of loss. That’s what you want to predict. And the thing is that historical volatility is pretty good at that. It’s pretty, pretty good. If you add some other stuff like credit spreads from the credit market, but also momentum helps a bit, and if you add correlations, then you can build a model that really predicts downside risk.
Of course, that’s, in the end, what matters. And volatility, for me, is just a way to predict it. And it’s a very efficient way, and it’s even better than using downside volatility. Many people think that downside volatility could be better to use, but the thing is that if you use downside volatility as a predictor of future downside risk, then you throw away lots of information because, also, what goes up often can go down quickly.
So, that’s my standpoint. To predict downside risk, don’t look at the past war, but look at all the wars and think about your tactics. And don’t bet on one factor, and use all of them, at least the best of them.
Meb: All right, well said. We’ve chatted the majority of this podcast on low-vol and multi-factors. You’ve written a boatload of papers, and publications, and ideas, and teaching. What else is on your brain these days? Anything in particular you’re working on, or any papers that you think are particularly interesting that don’t get as much interest?
You know, it’s always funny where we’ll write a paper or spend many days on a blog post or something, and we think it’s the most profound thing in the world, and the world is just crickets, and no one’s interested. So, is there anything out there that either you’re excited about or that you think deserves a little more discussion?
Pim: Yeah, a couple of things. First, I’m very happy with the p-hacking debates now going in academia.
Meb: Explain to the audience what that means.
Pim: So, the problem with academic research, not only in finance but also in medical research, is that researchers have an incentive to publish positive results. So, that means you do 20 tests, and then 1 out of 20 is significant. If you are a Ph.D. or you’re writing a paper, you do 20 tests, 1 of them is significant, and then you write a paper about it. But it’s not…it’s just fluke. It’s nothing. It’s noise. And you can always come up with a story quick fits the paper.
That’s called, in finance, factor fishing, and in social science, it’s called p-hacking. It’s data mining. It’s wrong, and because it gives us fake factors. And that’s a serious issue, and ever since I switched from academia to being a practitioner, I’ve found this to be the case. Then we found some great factors in academic journals, then we tested them on our own data, and then many we could throw away.
And then I was also shocked to see that the results which were not true were not true…falsified. Only knew it if you knew it, but it was not published. And nowadays, that’s switching in [inaudible 00:39:39] academia, and I’m happy with that.
So, it’s gonna be clean, and all the big editors of the big journals now at least admit that this is a problem and they wanna do something about it, which means several things. One is to accept more papers who address p-hacking, so, who try to falsify other results. The second is where you first pose an idea and then do the test. And that’s, with finance, a bit less possible, but especially outside finance, very relevant. But that your work will be published no matter what comes out.
I’m fascinated by this, and I’m also working on factor investing on a 200-year perspective. For example, I like history. I love it. I love to read about it because we can certainly learn so much about history. So, that’s a couple of things which come together, so that’s a project I’m on.
And I also am fascinated by taking factor investing outside equities into bonds and into allocation. And at my firm, I am also doing some very exciting stuff there and building solutions outside equities. And yeah, there’s huge demand for this, and it’s disruptive for hedge funds because you can simply do what hedge funds are doing, but then not with 2/20 but have much more attractive fees, based on solid academic research, and then outside equities.
Meb: I’m looking forward to you posting your next paper on low-vol investing applied to the cryptocurrency markets.
Pim: Yeah, that’s where I spend hardly any time on that. I’m a contrarian.
Pim: So I just let everybody tweet and talk about it and just leave it. For me, my favourite crypto is gold.
Meb: And I don’t think you can even claim low-vol even exists in crypto when…I mean, I was looking at some of them today were up and down 15% in a single day. I don’t even know if you can claim low-vol exists.
And so, you mentioned before, too, low-vol. Did…this works across assets as well, if you’re looking at, sort of, a cross-asset allocation sort of concept? I know you’ve written on some of those ideas as well.
Is that something…you know, how do you think about comparing assets? So, if you’ve got equities, and bonds, and commodities, and some are just naturally higher vol, and crypto which is extreme vol, how do you think about that and putting it, kinda, all together into a portfolio? Is it such that you need to adjust them and say, “Well, look, sovereign bonds are naturally lower vol”?
So, this kind of gets into the whole concept of, of course, risk parity and everything else. But what are your thoughts on putting the bowl of soup together, and putting together, kind of, an entire portfolio? Does low-vol have a place there too, or no?
Pim: Yeah, you can do it implicitly. So, if you create a [inaudible 00:42:25] asset portfolio, what you can do is put in low-vol stocks, low-vol bonds, and then you get a portfolio which has a lower vol, and then you can beautifully lever up. And that’s also what Fischer-Black [SP] recommended. So you can lever up your risky parts, and then…so you build a 60/40 portfolio with a 40/60 risk profile, for example. That’s how you can do it.
If you do it across markets, so, that’s where the evidence is mixed. That’s what I said. So, low-vol works within assets very strongly. So, within equities, within credits, especially when there’s leverage and benchmarks. Well, for example, crypto, there’s not so much leverage constraint because they’re all very volatile, and there’s also not a benchmark. So, my prior would be that I would expect there might be a low-vol effect that people go for the most risky crypto, but I expect not the strongest results there.
If you look across markets, also, it’s less strong. So, also within commodities, you see a weaker low-vol effect. So, it’s particularly strong within, and then you can use it, as you said, in a risk parity approach where, if you give equal weight to the different asset classes, then you often see that your risk/return profile goes up. Then if you go…then it ties into that.
And that’s risk parity is not a very high [inaudible 00:43:53]. That’s the problem with cross-market position. That’s just…for example, with risk parity, you often give more weight to bonds, and that’s sort of a binary thing. If [inaudible 00:44:03] go up or down, that will drive your return. I’m a bit cautious on the multi-asset side with applying low-vol, but you can really apply it strongly within asset classes.
Meb: So, we’ve kind of been chatting a lot about the past. Let’s shift our gaze to the future. You know, you mentioned, one, the pressure on fees, which I think is pretty universal but still pretty early days considering the average median fees around the world, but hedge funds being, kind of, the most egregious example at 2 and 20.
What are some other trends that you, kind of, look at our profession in the next 10, 20 years? You also mentioned the p-hacking and falsifying…I can’t tell you how many pitches I get that it takes about five seconds to look at and say, “Okay, well, that’s not even possible, this Sharpe ratio of three and no down years in the past 20 years,” yadda yadda.
What are some other trends you see developing, either well-known or not, or kinda any…this is hard for a quant-to-quant, but any forecasts you have about, kind of, the future of our world, and research, and everything else?
Pim: Yeah, first I think [inaudible 00:45:16] has a long way to go still, so it can grow further, and further, and further. So, in the U.S. is more progressed, but still there’s lots of room for growth. It can move beyond 50%.
So, that’s one. And why is that? Because it’s based on very solid reasoning. We trade too much and we pay too much for active management. So, that’s the truth, mathematically, and it’s gonna swipe away lots of our business. So, that’s disrupting.
Then there’s always room for fundamental stock pickers, but it will be less so than it was in the past, but there will always be room. Two reasons. First to set prices, price discovery, and second is that people are always overconfident and that they think they will have skill to select the right manager, and many people are overconfident that they can select the right stocks. But there will always be room for that, but it will be smaller.
And then, in the middle between these two trends, there is systematic investing, from [inaudible 00:46:17], and give it many ways. I think that’s where growth will be. And the problem there is, as you mentioned, there is also lots of…there are low barriers of entry. If you have a smart guy who [inaudible 00:46:31] with a degree in physics, and you have a [inaudible 00:46:35] who can just pitch your Sharpe [inaudible 00:46:39] strategy.
I think that’s where in the future there will be some shakeout, but brand will also be important. So, who is doing factor investing? Because there are so many factors out there, and there’s this p-hacking so you can pick the wrong ones, or you can do it in the wrong way. I think brands will come up where people will say, “These quant shops have a proven track record across cycle, across multiple cycles.” And that’s what I will see going.
And then another trend, so, this factor investing, quant investing, will also move outside equities. It’s now quite US-equities centric, but I expect it to move more to emerging markets. For example, we see much less data, and also to markets like [inaudible 00:47:25], bonds, investment-grade bonds, and international equities, small caps, international.
So, that’s if I look at the future. And I think that’s good for consumers and clients whose…if they pay a fee, they will get alpha for that, and that’s good.
And also, when you look at the factors, why they work. Suppose markets become a bit more efficient because of this movement. I also think it’s beneficial to society because capital will be allocated in a more, in a better way and in a cheaper way, and that’s for the real economy, in the end, good news. So, I’m pretty optimistic on that.
Meb: I like it. Yeah, I mean, we talk a lot about, you know, with this recent Fidelity news that they’re launching zero-fee index funds, this kind of barbell where the world’s gonna shift to market cap-weighted indexes for, essentially, free. And then people that deliver alpha, or at least are attempting to, you know, can still charge a higher fee, but probably not 2 and 20.
But it has to be something that just doesn’t look like a closet index. Particularly here in the U.S., we have so many of these mutual funds that charge a fortune, and you essentially get the S&P 500, which is…I think people are starting to become wise to a little more as the days go on, but I still don’t know that it’s a widely-held belief.
You have a lot of interaction with students. What’s on their brain these days? Is there anything in the general conversations you’re having that either makes you take a step back and get worried, or is it something that, you know, they all wanna start mining companies for crypto in Iceland? What’s, kinda, the general feedback there? Or does everyone just kinda hate Wall Street, and not interested, and wants to go work for Amazon and Google? What’s the general feedback there?
Pim: The interesting thing is that, like, banking is less attractive now than it was 10 years ago, for example, or when I studied. So, what I do notice is that investing has become more a profession than it was in the 10, 20 years ago. And so, like, for investing, if you Google it, there’s so much information. So, students are more knowledgeable.
I do notice that they’re not the wild, crazy cowboys. In fact, they’re quite conservative in a way. So, I do not get worried when I speak with students. I get, basically, quite enthusiastic. They’re quite down-to-earth, knowledgeable, very competitive. You do see that the international…lots of our students are international at our universities. Because I’m a Dutch native speaker, but our universities teach in English. Basically, most Dutch are bilingual.
But then, also, international students come to our universities and they put pressure. It’s a meritocracy here, and it’s really the best and the brightest ones winning. So they’re very focused on getting good results and either starting their own companies — that’s more in vogue now than it was 20 years ago, that’s the difference.
But, yeah, very knowledgeable, very intelligent. And so, the level of knowledge of our profession has gone up, quant finance, which I teach. So, quite interesting, and not that [inaudible 00:50:54] more encouraging out.
Meb: Yeah, I wonder if you’re gonna see that trend where investing and finance becomes almost more of, like, a boring profession where you have less and less of these major cowboy hedge funds of, like, the ’70s and ’80s, the Soroses and the Tigers, and you have, you know, the continued…I’d say just making it more professional, sort of buttoned-down, kind of boring. I think that’s probably a positive.
I love the trend towards young people being interested in entrepreneurship and starting companies, all those ideas. I think that’s…but these things go in cycles. You know, if you look at MBAs, any given year, what they’re going into, it’s a pretty oscillating trend over the years. Sometimes it’s Wall Street, sometimes it’s consulting, sometimes it’s Silicon Valley and everything else. So, interesting.
A couple more questions. We gotta wind it down. What’s been your favourite surf destination you’ve been to? Is it…you got a local spot you can’t tell anybody about in the Netherlands? Or where…what’s your favorite?
Pim: My favorite is the Canary Islands, I think, because there you can do both windsurfing, which I like, and the other parts. There’s Costa Calma where it’s shallow. You can stand. So if you didn’t do it for a year, it’s okay. And on the other side, then, you have some big waves. And not too big because I’m not that good. I’m just a fan. I love it, to be in the water and be washed away, to really feel the nature.
So, that’s something I can recommend. And South Africa, that’s such a beautiful country. Good surf there, but also [crosstalk 00:52:34]…
Meb: Too sharky.
Pim: Yes. Yeah, but no risk, no return, eh?
Meb: Yeah, but good wine and everything else there.
Pim: Yeah, I’ve been in there a couple of times. I love it because there’s also Dutch heritage there. People speak Afrikaans, which is, yeah, very close to the Dutch language. And also, they’re drinking wine there. And it’s a beautiful…
And it’s the world in one country. It’s the third, second, and first world just in one nation. And that’s fascinating because we’re all living in one world, only separated by borders. South Africa is basically a borderless country where you have the whole world in one country, and that creates challenges. And it’s fascinating to see that. But also, nature is very good, including the surf and sunshine.
Meb: I love it. I haven’t been to either spot. As listeners know, I’m a fairly terrible surfer, so it would be fun to get you and I on video and they can watch…the two-quant groms [SP] highlight reel of bloopers I imagine would be really funny.
Last question we always ask, the 2018 question, is looking back on your career, what’s been your most memorable investment or trade? Good, bad, horrific, anything in-between? Anything come to mind?
Pim: Yes, yes. It’s a trade I did when I was a teenager. I put, first, my saving amount in a mutual sort of bond fund. I did investing with my dad. You know, great guy. I also mentioned in the book. Good mentor. But I was a bit [inaudible 00:54:16] back then, and I wanted to do…to move into the risk and the stock market.
And I saw this great stock. It was an airplane manufacturer, Dutch Pride [SP], recently [inaudible 00:54:30]. It was in tough times, very volatile, bad momentum, bad value. But back then, as a teenager, I didn’t know about factor investing. I bought the stock hoping it would rise up, and it did not, and it was a painful ride down.
It cost me a lot. I lost most of my savings. 2/3 I put in, I lost. I did not put all my savings in, but back then, it was a significant part of my net worth. And why I like it is it taught me a lot. It was my college tuition fee for the real world of investing because ever since, at any time I step into the behavioural biases, I know that this was just one example.
And as a quant, I exploit human behaviour, but I’m also very well aware of my own mistakes and my own biases, and that makes you humble. And so, I do think this is a great example. I lost money, but I think in the end, I made money out of it because it made me wiser and more experienced.
Meb: Yeah, you know, it’s interesting. We talk a lot about this on the podcast, on how best to educate a young person about investing .And the challenge for so many is…the best lesson, almost, when you’re young is to lose money. And no one starts out with that as their goal, but certainly that’s…losses, you know, are the best teacher in the early days.
I think the worst thing that could probably happen is, you know, young people have many years of investing success only to build up a balance and then feel the pain when they have some more money. I don’t know what the right answer is, but I think coming up with a curriculum there would be pretty fun and interesting.
Pim: What I can add to that is, that’s also why volatility works, I think, low-vol. Because if you enter in high-vol stock or position, your learning curve is quicker. So, Eric Falkenstein, another pracademic who also wrote books on low-vol investing, writes about this element, that if you enter high-vol assets, your learning curve is steeper, so it’s also rational to do so.
Like for me, it was rational to enter this high-vol stock because it taught me a lot. It came at a price, but it also came at a return. And the return was not financial, but the return was I learned.
And that’s also interesting to note, that this learning behaviour, like you said, like you said that should be taught, maybe, to teenagers, then I would maybe also use more volatile assets to give them a more quicker way to learn. So, that’s interesting, where everything comes together.
Meb: I love it. All right, Pim, it’s been a blast. Where can people find your writings? Where do you they follow you? What’s the best place to keep up with all your goings-on?
Pim: Recently I joined Twitter, so I’m on there, @paradoxinvestor. My papers on SSRN. That’s where you, if you Google me, Pim van Vliet, SSRN, you will see my papers, set alerts for new ones. And also, on my company website, there is lots of stuff. Some of our whitepapers, like the one we discussed on CAPE, is also found there. So, and I’m active on LinkedIn. That’s for…the network I’m using a lot.
Meb: Robeco is one of my favorite institutions as far as putting out quality investment content, so absolutely check that out. We will post show note links to all these things, white papers, books, links to your website, Twitter, all that good stuff. Pim, thanks so much for taking the time today.
Pim: Thanks a lot. I enjoy it, and I just see myself on a good surf with you somewhere in Gran Canaria or South Africa. Look forward to that.
Meb: Perfect, I love it. Listeners, thanks for taking the time today to listen in. You can find show notes, everything else at mebfaber.com/podcast. Send us some questions for the feedback. Radio shows, all that good stuff. firstname.lastname@example.org.
If you love the show, hate it, leave us a review, and check out our new “Best Investment Writing Book, Volume II.” Leave a review there as well. Thanks for listening, friends, and good investing.