Five More Questions: Factor Investing with Jim O’Shaughnessy

By Jack Forehand, CFA (@practicalquant)

This is Part II of my interview with Jim O’Shaughnessy. If you missed Part I, you can find it here. In part II, we talk about trend following, the future for simple value investing, and we take in in depth look at some of the arguments others have made in favor of the death of value investing.

Just a note before we begin. This interview was transcribed from a phone conversation so please forgive any grammatical errors.


Jack: In part I of our interview we talked about the multiple points of failure that investors face. The first being that they sell when the market is down and the second being that they sell when they underperform. I wanted to ask you about trend following in light of those points of failure. The goal of trend following is to help to limit the first point of failure by getting out of the market during major declines, but that often comes at the expense of the second because there are typically many false positives that get you out of the market and then it ends up going up. January of this year was a great example because the market turned very quickly and many trend systems missed ten plus percent of gains. I noticed that OSAM doesn’t offer a trend following model and I was wondering how you look at trend following in the context of the multiple points of failure that investors face.

Jim: I’ve always been enamored with trend following systems. I have studied as many of them as I could get my hands on. I mean, as many of the reasonable empirically realistic ones that I could find. We don’t offer a trend following system. And that is because for the most part, what we found is that the number of false positives that they give leads to exactly the type of behavior that you’re talking about, which is, you get a false positive like you got in December 2018. You then got a V shaped bottom. You get the market up 15%, and you’re up nothing. And then you get fired.

And by the way, that doesn’t keep me from still looking at different interpretations of trend following stuff. I would love to be able to find one that that didn’t have a lot of false positives and works very well. Because then I’d probably offer it to investors. But I think that for the most part, they’re so based on moving averages, on absolute ups and downs, there’s a whole different group of factors that get used in them. But a number of false negatives leads to that second error, which is a hard one to overcome, because if you’re going to be offering a trend following system, your goal is to you start offering it 2007, right? Because it keeps you out of the carnage. Now it puts you back in late. It doesn’t get you back in in February of 09, but it gets you back in by, I think, May or June of 09. So you still manage to avoid that massive loss and you get back in at a reasonable time. The problem is that the excess returns are really magnified by the massive bears. So you’ll see it got you out in 72 and then back in in 75. If you’ve got good data and you could go all the way back to the twenties it got you out not before the crash, unfortunately, but right after the crash in 29, and it didn’t get you back in until 33 and it didn’t get you out for the 45% decline until it was too late because it happened so fast in I think it was 1938. So what happens is that absent those really massive downturns; if you took them out and you look at just the results of all of the other signals, what you see is that you’re not adding any value right there.

Jack: So essentially they’re wrong more than they’re right by a pretty wide margin, but when they’re right, they’re right in a massive way. So the problem is sitting through all the times they’re wrong to get to when they’re right can be very difficult.

Jim: Yes, exactly. That’s a great way of putting it. And so kind of knowing what I know about the way investors behave, I just think that it’s going to be very difficult unless you have the absolutely good luck of offering one of those systems right before a serious downturn. And yet, by the way, the recent, history is a great example. So going into 2019, you got people out, the market went down nearly 20%, but then boom it went right back up. And so the investor might love you all the way until you’re still sitting in cash and the market’s up 15 plus percent, and you get fired. And so one of the things about trend following systems that I actually find more interesting is they are really good at highlighting the agency problem. The agency problem is: you might be right, but if you’re not right with enough consistency and your timing isn’t perfect, you get fired. And then that multiplies by; let’s say you’re dealing with an investment committee, and there’s five people on that committee. And let’s say that one of those people was a very smart young person who’s made a big study of all of this and started looking at, well, should I recommend a trend following manager to the committee? And then they look at their own research and they’re like, well, the odds are that I’m going to probably make that recommendation and then we’ll get a get a false positive and then that’s going to reflect badly on me, so I’m not going to do it. So, I guess what I would say is if you’re a do it yourselfer and you want to take a small percentage or some percentage of your portfolio and follow trend following just as insurance against a real bear. Well, you’re going to be wrong more often than you’re right. But let’s say you do it with a 20% of your portfolio. I mean, do the math, right. Look at how much it could have saved you and is that worth it at the end of the day and if it is, then do it, and if it’s not, then don’t.

I’m worried about the school of thought: when you look at particularly the United States stock market; and people say you shouldn’t just look at the United States. And they’re right. We shouldn’t. What about Japan? But there’s a different answer to that which is, well, yeah, the United States is kind of like the ideal setting for stock markets to work and work well. Could that change? Sure, sure it could. We could get bad laws. We could get the rule of law overturned. We could get a whole host of things that would negate all of the study of the past. So I am not dismissing that, but absent those kinds of radical changes, generally speaking, you’re much better off being bullish and being wrong and feeling lousy 30% of the time and being right 70% of the time. But to answer your question directly, yes, I do think the trend following it fascinating. I wish I could find one that I really felt good enough to put my own money in and recommend to the public. I haven’t done so yet, but that doesn’t mean we’re not still looking.

Jack: I wanted to ask you about the future for simple value investing strategies. Common investing wisdom states that simple investing strategies often beat complex ones. But as factor investing has grown, the approaches have become more complex. For example, it is now widely believed that just buying a basket of cheap stocks isn’t enough. These days, things like negative quality screens that take a much more detailed look at the fundamentals and try to eliminate companies that are cheap for good reason are widely utilized. As more and more computing power comes online and artificial intelligence becomes more widely used, there is an argument to be made that simple Ben Graham type value strategies may not work anymore because a greater percentage of companies that are cheap will be cheap for a reason. But on the flip side, as more and more computers try to one up one another and compete to find every edge possible, things could also come full circle and simple value could have its day again. Do you think the days of simple investing strategies working in areas like value are behind us?

Jim: I think that really depends on what we’ve been discussing earlier in our conversation, and that is human behavior. We ran it a long time ago, but still there was a lot of computing power available and we ran it. We ran one of Graham’s old school strategies that had I think like eight or nine different factors that you had to consider and boy it did really, really well. And there were a lot of computers when we were testing that as well. I think the challenge does come down to, are people willing to wait for things to work? And many times, they’re not. And therefore, you get a class of stocks that are sort of perennially undervalued and get swept out of the way by the more complicated filters that indeed organizations like my own do use.

So I don’t know that it’s that you need more complexity in the strategy, but I think that you need (a) the ability to stick with it, and (b) the ability to understand that every strategy, be it a growth strategy, be it a value strategy, be it a large cap or a small cap strategy, have cycles and they move in and out of favor. And then finally it gets down to simple things like definitions. We were talking about price to book earlier, right? There’s $1 trillion indexed to price to book and we think it’s awful. So there’s a massive difference of opinion. And so I do think that you’re gonna always, getting back to my theory of how the stock market works and why it works so well, you didn’t have heterogeneous opinion even on the various factors that we still call value. And some very, very smart people think we’re wrong in our take on price to book. And we’re happy to put the empirical results against the empirical results, but there’s enough fluidity to how you define a factor and how you use it to say that, if you’re talking about a Ben Graham portfolio that bought low price to book stocks, then probably, I wouldn’t want to do that. I wouldn’t want to do that not because it’s out of fashion or old fashioned, but rather because of the new learning that we’ve come to have about the factor itself.

Ultimately, if you’re a quant, you’re pretty much using the same data as everyone else.  Now we’ve been trying to change that and in what we call the, the correction or data cleaning that we do at OSAM. A lot of people will be very shocked to know that, say if you said, hey, what’s the PE of Google? You might find very different numbers at Reuters then you find at Bloomberg than that you find at Yahoo Finance than you find at any other information source. And that’s because people have different definitions for price to earnings. It’s because some use trailing 12-month. Some use forward 12-month, Some like  Value Line mix them up and use half and half. We found that there’s all sorts of joys if you will at the data level and we found it very helpful to recognize and scrub that data very, very well. Now that’s one of the most unglamorous things you could do as a quant. It’s not fun. It’s work horse stuff, not show horse stuff, but boy, it helps. And so I think that, yes, all the computer power that we all have now, we all have pretty much the same data other than people who are coming up with new and unusual datasets, and we have been for several years looking at creating a newer version of one ourselves. We do that more from my point of view, more for consummation than anything else. You look at a strategy first here. And then you say, does it work here? Yes it works here. OK, does it work in Canada? Yes, it works in Canada. Ok, so it works in North America. Does it work in Europe? Yes, it works in Europe. Does it work in Japan? Well, basically not. Ultimately, it comes down to the human investors’ willingness to sit with bad performance for x amount of time. So everyone and their brother is talking about how value is dead, dead as a doornail. And I don’t know, our value strategy has done pretty well over the last 10 or 15 years. Is that because we use different factors? Probably.  Using the term value or growth to be some monolithic thing I think is a mistake. And you know, big, well run companies that are reasonably valued and have high shareholder yield, they are good investments. A lot of people don’t agree with that. I mean look at all the debate people are having about buybacks and, and yet based on empirical evidence, we think it’s a winner and it’s been a winner. So I think you’ve got to look very closely at what the strategy is being covered under the model using the term value, dig down and say, okay, what are the factors you’re using? What’s the evidence for those factors? What’s the win rate versus the loss rate? So I guess it’s a little more complicated and nuanced than simply saying, you know, is Ben Graham dead? Well in theory the very simple Graham things are probably not going to work as well because they get picked off. Right? But that’s arbitrage. And talk about shooting fish in a barrel, back in Graham’s day people didn’t even think about looking at a balance sheet. You know, read how guys like Jesse Livermore and all those guys trade stocks, it was all trading humans for those guys. So all of the inefficiencies built into where you could buy 50 cents worth of earnings for 25 cents and sell it at 75 cents, that’s gone because machines can figure that out.

And you know, there’s a classic old arb where they used to arb between markets, right? Between New York and London. That worked for years and years. You’d buy a stock that trades on both exchanges and you’d notice that it was trading a dollar or pound higher or lower on one of the markets and you buy it where it was a pound or a dollar lower and sell it where it was a pound or a dollar higher. So all of that is gone. Any kind of informational advantage, that’s gone. And so it kind of comes back to what I’ve said for a long time, which is, you know, market’s change minute by minute. Human nature hasn’t changed at all really in a millennium. Arbitrage human nature. That’s kind of your last sustainable edge.

Jack: I wanted to talk to you about what some people have referred to as the death of value investing. I’m a big believer in value investing like you are. But one of the things that I understand about my own behavioral biases as a big believer in value is that I am likely to focus on reading people who agree with me. I’m going to read what you write. I’m going to read Wes Gray. I’m going to read Tobias Carlisle. I’m going to read the guys that are also big believers in value investing and my confirmation bias is going to kick in. And after that, I’m going to be even more of a believer in value investing than I was in the first place. And so what I did earlier this year is for our blog, I decided I was going to write an article taking the other side. I was going to say, if I wanted to say value investing is dead, here are the reasons I would come up with to say that it no longer works. Then if at the end of that didn’t convince myself, I would feel better about my conviction in value. And so what I want to do is give you some of the arguments I came up with against value and see if you could help me refute them.

The first comes from my recent interview with Ben Hunt. I know you are also a reader of Ben and the Epsilon Theory blog. He wrote a piece called The Three Body Problem where he talked about the fact that if you introduce a third body to two bodies moving in space, the movements of all the bodies becomes completely unpredictable. And so what he was arguing is that this whole quantitative easing that’s come into the picture has introduced a third body into market, and that makes future outcomes more random and makes us less able to rely on things like base rates to predict things that might occur in the future like mean reversion with value investing. Do you think quantitative easing could make it less likely that we can rely on the past as a guide for the future?

Jim: I read Ben and his colleagues. I think they’re all really smart guys and I read them religiously because I think it’s always really important to see a lot of sides to an argument. So the idea of quantitative easing being a third body, my response to that is, if you look at the history of monetary and fiscal policy in the United States, you have seen any number of times when the government did something that everybody at the time thought this is a game changer, it is going to be an end to X. And it proved (a) not to be an end to X and (b) often times proved to be something that the government did not continue to do.

So you know it’s pretty easy to say, well gosh, now that quantitative easing is here they are going to keep doing it forever. My response to that is why? Why would you say that? There are all sorts of reasons why that might not be a good idea. And you know, none of the reasons for why it’s a very bad idea to continue to do that have yet reared their ugly heads, but they may, and if they continue to do it, probably will rear their ugly heads. And that will call for a change in policy yet again. And so I think that taking a current policy and extrapolating it forever into the future is a bad idea. Because if you’re aware of history, you see that policy changes all the time.

And you know, the other side of the argument against me would be, well, Jim, yeah, but this is a really big change. Okay. Yeah, it is a really big change. Just like going off the gold standard was a really big change, right? And yet ultimately the outcome being random makes very little sense to me because as much as you want to believe in alchemy, it cannot change the basic rules of economics. And so because the Fed or because treasury is doing some particular activity, you’re not going to suddenly say, Jim, you know, there’s a food truck out here at Union Square. I love this guy. He’s doing a hundred thousand bucks of revenue every year, but he wants us to pay 10 million for it. I think we should do that. Because I’m going to say you’re crazy, Jack, that’s nuts. We’ll never make that money back. Right. So these fundamental rules of economics, you know, it’s why they call it the dismal science, right? Because its immutable. The underlying rules don’t change. We keep trying to believe that they do. It’s why we priced dotcoms to the moon; it’s eyeballs, it’s not sales or revenues. Or you know, housing prices never go down. That’s why they can take all these junky CDOs and give them a rating. Well, they did go down. And the number of eyeballs they do attract is meaningless if you don’t attract revenue that leads to profits. So you can dress this stuff up and be really fancy with it and say, well, this new alchemy is going to completely change everything, and my attitude is we’ll see. Good luck with that. You know, they’ve been trying to convert lead into gold since Isaac Newton, himself an alchemist, failed, and it’s going to continue to fail. Philip Dick has a great line, which is reality is that which once you stop believing in it, doesn’t go away. And so reality in economics doesn’t go away. Can it be exacerbated? Can different outcomes happen because of particular policies? Of course. Absolutely. But ultimately the rules of economics are not going to be rescinded. They’re not going to be overturned. And so I say, you know, I love the idea. It’s great and exciting writing. But I just don’t see how things either continue to stay a certain way, because they don’t, or much more importantly, you’re going to overrule the laws of economics and economic outcomes into random events.

Jack: Two other arguments I have heard against value are that too many people are following it and that they are sticking with it more than ever before. On the first one, there are more value-based factor funds than ever and firms like DFA have massive assets following value. Some have argued that all this money following value has reduced the value premium. The second point is one that the anonymous Twitter user Modest Proposal made on your son Patrick’s podcast. He argued that followers of value are more likely to stick with it now than they have been in the past. Since having the weaker hands fold is good for long-term value investors, that could also potentially be a problem. What do you think about those two arguments? 

Jim: So those are kind of contradictory in my mind. It would seem to me, and again, I could be wrong here, but it would seem to me that if there were a lot of people trying to buy a particular thing, that would drive its price up. And there could be some validity to that with all sorts of investment strategies. If you’ve got a lot of people who are clinging to an investment strategy through hell and high water, then that would be true. I just don’t see that being the case. Because again, I would love it if everyone would suddenly decide that they wanted to buy cheap stocks with high shareholder yield because I own them already. If everyone else is going to come and start buying them, I’ll say hallelujah.

I haven’t really seen that happening. Now we’ve done well but not to the point where I think that we are going to negate the premium altogether because again, back to heterogeneous, not homogeneous opinion basis.

The weak hands sticking to it? I don’t know, that’s an interesting idea. I just don’t know how you could measure that empirically. I guess you could look at some pure value-based fund, or kind of best-case quant funds where you know the rules they are using and kind of look at the cash flows into and out of that fund. My guess would be that that’s not proven to be true. That what we actually see is people moving their money into index funds, away from actively managed funds. And that’s something that you know, that’s definitely happening. Again, you’ve got the empirical evidence for that.

Now is that a good idea?  I don’t know. I used to say that if you couldn’t follow a simple rules-based way of investing, you are probably better off indexing because you’re going to get a cost advantage and at least you are going to get market exposure. But you know, if you have everyone piling into these index funds, one of the real problems with index funds is they don’t pay any attention to valuation. And what I said earlier, I think holds true. Valuations ultimately do matter and you can have as much fairy dust as you want, it’s not going to fix that problem. And the only way to fix that problem is for prices to adjust and usually than means downward. So I guess I would say not having right in front of me the cash flows to clear value investing or funds that you can see the rules that they’re using, I guess DFA would probably be a good example there maybe there’d be a couple others, I don’t see money pouring into values strategies. I do see money pouring into index strategies, but that’s a different argument.  

And then so as far as we can, I don’t know, I don’t see any evidence of that. What was the other one in addition to the weak hands?

Jack: It was that too many people are following value and too much money is pouring into it.

Jim: So I guess I will give you the same answer there. Does the data show that to be true? I haven’t looked at flow of funds for various funds by strategy in a long time. I used to do that, but I haven’t done it in years. But my guess would be that that would prove to be not true.

Jack: Yeah, you would think if that was true, you would see spreads narrowing, You would see people buying value stocks and driving up valuations and you’d see value getting more expensive relative to growth. And I think you’ve actually seen the opposite. I think in reality value is getting cheaper relative to grow. So that sort of refutes that argument.

Jim: Yeah. So again, I would say those are pretty easy questions to answer just by looking at flow of funds and see whether that’s true or not. My intuition, and I could be totally corrected here once we actually looked at the numbers, but my intuition would tell me that is not true. That money does not seem to be sloshing around in value or any of its strategies any more now than ever.

Jack: Another argument I have heard against value is this whole concept of alternative data. The concept that if you and I are using standard fundamentals to buy our stocks and let’s say you and I are both buying Walmart. But let’s say the hedge funds out there have satellite photos and credit card data and they know that the parking lots are empty and nobody’s buying anything. And so our standard fundamentals aren’t giving us the full picture. So I wonder if this alternative data to some degree can be a threat to standard fundamental value investing because more of the cheap stocks will be cheap for a reason. We just may not see the reasons in standard fundamental data.

Jim: That’s a really interesting point. And it’s something that I don’t have this definitive point of view on because I haven’t been able to run the tests that I would feel would be required to either support or refute that point of view. Intuitively it makes some sense, right? That if you see through satellite imaging that there are no cars there, that would be a useful I would think, but I don’t know. I know that in my history in the business, in my organization, we have tested lots of things that seemed really super smart and intuitive that ended up not working. And so, not having the ability to actually run the test, that type of data analysis, I would have to say I don’t know. I’d like to know. We are in fact through our OSAM research partners, where we bring people in as part of the research team, even though they don’t work for OSAM directly. We do have access to some very bright guys in machine learning and other places and so we’re going to be running some tests that might coordinate with that point of view. And then I guess that I’d say, then I’ll know a little bit better. But again, the sensitivity to initial conditions is very; I don’t know if you’ve read Manelbrot’s The Misbehavior of Markets, but you know, he makes a very strong case in there for the fact that chaotic mathematics does a much better job of describing markets, not just the stock market, but any auction based market, than traditional Bayesian analysis. And so I don’t like to punt, but I am going to have to punt because I don’t have evidence. I can’t argue in favor of something when I don’t have evidence. When I don’t have evidence, I don’t have an opinion.

Jack: That’s something more and more people should do to be honest because a lot of people tend to have opinions on things that they have no evidence on, especially in the stock market. That’s a very common thing. And you know, people need to learn to say I don’t know a lot more, I think.

The last argument against value that I wanted to ask you is that value is essentially both an implicit and explicit bet against technology. So it’s an explicit bet in that value strategies tend to underweight technology stocks and when they do buy technology stocks, they tend to buy the Western Digitals and the Micron Technologies of the world versus the Googles and the Facebooks. And then implicitly, you know, you’re buying a lot of stocks these days that people say Amazon for instance, will be taking down. You’re buying a lot of stocks that technology may have a negative impact on. So I’m wondering what you think about that, the theory that value is being impacted negatively in two different ways by technology.

Jim: It’s interesting. I do know that while we didn’t own it for a long period of time, we did in fact buy stocks like Apple in a value strategy because it had become, for a brief period of time at least, a value stock that was buying shares back. I think, you know, there may be merit to that. My only comment would be that at least between 1964 and 2009, if you look at the sector analysis that I did for the fourth edition of What Works on Wall Street, the one with the highest standard deviation of return and one of the lowest compound returns was information technology. Because, talk about having to wake up to do battle with death on the line every day, right? Technology is fascinating in that something that is standard right now can be obsolete overnight. I mean, think about Blackberry for example. Everybody was on their Blackberries until they were all on their iPhones.

And you can give all sorts of examples in technology where yeah, there was one leader and then their done, because of innovation that supersedes the other one’s moat. Tech folks are, in my opinion, and I think this is one of the reasons why information technology has had such a high standard deviation and actually one of the lower compound returns is because of the relentless improvement in technology, but unfortunately not by always the dominant player. And so it’s funny because I’ll say that to people and they’ll say, yeah, but now it’s different. Well, okay, but in 2009 a lot of these tech companies were around then too. And some of them thrived and some went bankrupt. And so I think that the very nature of the creative destruction that is prevalent in that industry is a huge benefit to those of us who are consumers of technology, but potentially could be lethal too. Again, if you’re the owner of Blackberry you think you are impervious to everything right up until the minute you’re not. I think that that’s something that if you want to write a book, it would probably be a pretty good book. Things that looked like they were going to go on forever and then innovation comes along and displaces them virtually overnight. I mean, Google was not the first search engine. We all know that. Yahoo, Ask Jeeves, the list is kind of endless. It’s kind of endless and they built a better mouse trap. Now does that mean that their mouse trap is going to be the best mouse trap always? Well, if they’re smart and they are innovating and they’re buying up any competitor that they see who might have an advantage then maybe. But maybe they miss one, and if they miss one and they come out with a vastly better search algorithm, then even the Googles of the world are at risk.

So I think that that it’s fascinating. I love technology. I’m an early adopter, I was one of the first people online. I was one of the first to buy an iPhone etc. But at least the numbers for a fairly long period of time, 1963 to 2009, show it to be a very high standard deviation, and because of that sort of creative destruction, not as stable as other industries. But again, I can’t give you a yes, I’m certain about that because I’m not. I think the Amazon example is actually the really interesting example because you know, they might actually have a competitive advantage that is so deep and that moat is so wide that it compounds on itself. It’s easy to spin a narrative where that happens, but you also have to think of a narrative where, you know, somebody else figures out a better algorithm or faster delivery or you know, you name it.

I love Amazon Prime. I’ve been a customer since they’ve had Amazon Prime, but I’ve got to tell you if some other competitor came along and offered me better terms and a better experience, see ya Amazon.

Thank you so much Jim. I really appreciate you giving me the time. I’ve learned a ton doing this. I really appreciate it. If investors want to find out more about you or OSAM, where are the best places to go?

Our website is osam.com. You can find me on twitter at @jposhaughnessy.

Photo: Copyright: 123rf.com / kentoh