Stewart: Welcome to another edition of the Insurance AUM Journal podcast. My name is Stewart Foley, and I’m standing with you at the corner of insurance and asset management with Michael Hunstad, the head of quantitative strategies at Northern Trust Asset Management. Welcome Mike.
Mike: Thanks Stewart. It’s great to be here
Stewart: You come out of a research background, right? You came out of quantitative research before this podcast you sent over this thing that you’ve put together called The Risk Report, and it is a very comprehensive study that you’ve done. Can you give us some background on that? Why it was done, the scope and so forth?
Mike: Sure. Absolutely. Yeah The Risk Report is sort of an amalgamation of a number of deep dives that we have done into client portfolios. So more than 200 portfolios, 64 institutional investors, more than $200 billion in assets, more than 1000 strategies were analyzed over a four year period. And the reason that we do these studies or these analysis is we get clients and prospects that come to us asking about the composition of their portfolio. And these studies are essentially a diagnostic check on the portfolio. And basically what it does is it looks at the portfolio as a point in time and decomposes the portfolios into the three primary sources of risk and potentially return. That’s beta or market risk, idiosyncratic risk and then more systematic risks like factors and sectors and countries and things of that nature. And we try to assess of all the risk you’re taking, how much of that risk are you getting paid to take, what’s working, what’s not working in the portfolio.
Stewart: It’s interesting. I think you know I teach finance. I’ve taught finance at a couple of different colleges and universities. And one of the things that is kind of the bedrock piece of finance is, modern portfolio theory or the idea that I can diversify away, uncompensated risks get a better risk return profile, so on and so forth. So the first point that you make, and I think many points out of this study stand some traditional views on their head, but this one is in particular, a strong statement to make. That institutions that you studied had two times as much uncompensated risk as compensated risk. How did you come to those conclusions?
Mike: Yeah, and I think the old rule of thumb that we all grew up with is that, more risk equals more return essentially. And that’s more or less embodied in the Capital Asset Pricing Model. But nowadays we know that just isn’t true. That there are types of risks that you do get paid to take, that are compensated. Risks like beta exposure, size, value, quality, momentum, low volatility et cetera. But there’re also a whole host of risks that you do not get paid to take over the long term, such as sector biases, region biases, country biases, things of that nature. And the innovation of this study is that we now have the ability to segment these portfolios because we have more data, more technology, into the relevant sources of compensated risk and uncompensated risk and gauge the relative quantity of those two risks in the portfolio.
Mike: So the finding and that’s one of the premiere findings of the study is that, institutional investors, as you said, took twice as much uncompensated risk as they did compensated risk. I think is not necessarily surprising in that we know anecdotally that a lot of these institutional portfolios struggle in terms of performance. But I think it reaffirms the intuition that something’s not quite right under the hood of a lot of these portfolios. So because we have now the technological ability, the data to do this kind of analysis I think, there’s actually a lot of intuition in that finding that there is a lot of uncompensated risks in client portfolios and that, we’ll get to this I’m sure later on, a lot of what we’re diversifying actually is the good stuff. It’s the compensated risk that we thought we’re keeping, but lo and behold, we’re kind of throwing the baby out with the bath water.
Stewart: It’s interesting. You kind of leading me to my next point or question I was going to talk about. So the traditional view is, I’ll get rid of my uncompensated risk through diversification, right? But your second finding shows that the portfolio, the underlying portfolio holdings can actually cancel each other out and effectively get rid of the good stuff that you were just talking about. How did you get to that conclusion?
Mike: Exactly. And again, it comes through the precision of the new data, new technology. But I think, it makes perfect sense. In the old method of diversification, let’s say that you have an equity portfolio and you have a number of managers in the line-up and you’re adding another one. And that manager isn’t perfectly correlated. Shouldn’t be perfectly correlated with another manager in the mix, therefore extensively, they provide some level of diversification. But what guarantees that, that diversification they provide is really the kind of diversification that you want? What’s the guarantee that it’s diversifying the uncompensated sectors, regions, countries, things like that versus the good that you do want to keep? And the answer is nothing. But what we have to do now is be much more precise about how we analyze these risks. Now we can look through the portfolio and say, “What kind of cancellation or what kind of diversification are we really getting? Are we diversifying away the bad stuff or we’re diversifying away the good stuff?” We do have the tools now to make that assessment.
Stewart: So I also have the opportunity to teach risk management and insurance in a prior life. And one of my gray haired old man quips is, the greatest risk you face is the one you don’t know you have, right? And you move into this and your third finding, which is that hidden portfolio risks cause unintended outcomes. I think that’s one of those things that intuitively we see over and over again, right? Where a company has a sound, at least based on traditional metrics, a sound risk management strategy and yet their portfolio may not perform the way they thought. What pieces of information were you looking at there?
Mike: Yeah. And this really gets to the level of understanding we have about the risks that we take. So embedded within most investment portfolios, there are risks that we probably do understand. For example, we may realize that we’re taking certain country bets or region bets or sector bets or style factor bets. But what we might not necessarily understand is a lot of these biases in the portfolio may have secondary and tertiary risks associated with them. I’ll give you an example. So if you’re a value investor today, and I know it’s tough to be a value investor today, but they’re still out there. If you’re not careful, it’s a good example though. So I’m going to go with it. If you’re not careful about how you build your portfolio, the bulk of your allocation is probably in the energy and financial sector because these are the sectors that have been classically more value-oriented over the last couple of years.
Mike: Okay. Well you probably understand then that you have these sector biases, but what you may not understand is that both of these sectors are strongly correlated to macroeconomic variables. So energy is oil prices, commodity prices, financials, interest rates, sensitivity. So I think one of the things that value investors, and we’ve seen a lot of value investors in this study, didn’t realize is that as interest rates fluctuate as oil prices fluctuate, which they have in 2020. And 2020 has been a great example of a year where there’s been a lot of unexpected due to hidden risks. As these macroeconomic variables change, all of a sudden there’s big, big deviations in my portfolios, or there’s a big price movements, big price swings in my portfolio as a result.
Mike: These are the kinds of things that we’re trying to understand. Is that as an asset owner, as an investor, do you really have the most precision, I’m going to use that term a lot today, most precision about exactly the kind of risk that you’re taking and how they influence the final output? And in many, many cases, as we did this study, we found that the answer was not necessarily. That there were a lot of pitfalls land mines, if you will, a lot of risks that we didn’t necessarily appreciate going in.
Stewart: And when you talk about conventional style investing and you talked about it a moment ago, right? Where I’m going to get a large cap growth manager, I’m going to get a small cap value manager, and to get this, I’m going to get that. And I’m going to put those together and I’m going to end up with a diversified portfolio and things will be good. Your fourth point, your fourth takeaway says conventional style investing lead to index like performance with higher fees, right? That’s not a good value prop for the investor. How did you get there?
Mike: Yeah sure. And I’m just going to rewind to 1993, when the style boxes first came out and there was great intention behind it. And the intention at the time was to say, keep in mind, Fama-French wrote their first paper on size and value the previous year. The early draft was released to previous year. So you have size, you have a value growth dimension. That’s exactly what the style box is constructed with-
Stewart: I, yeah. I want to say this real quick Mike. Your finance geekiness is an endearing thing for me. Cause I’m like, “I wish I could have pulled that date out myself.” That’s like, man, that’s a good stuff right there.
Mike: It’s around there-
Stewart: I’m sorry to interrupt. But it’s like, I’m such a frigging geek that I’m like, “Oh man, that’s cool talking about the dates. That’s awesome.” Sorry.
Mike: I thought you might like that.
Stewart: Absolutely. Absolutely.
Mike: Yeah. So, late 1992, Fama-French published, well before it was even published, they release the first version of their paper about size and value. Next year, Morningstar comes out with a style box, not a coincidence at all. The idea behind the style box in the original sentence was to say that, “Size and value are indeed dimensions of risk but we want to neutralize those risks in the portfolio.” That was the original intent of the style box. You want to neutralize large cap versus small cap? You want a neutralized growth versus value and let those managers shine with their security selection. And that was the intent because at the time, remember the Capital Asset Pricing Model suggests that it’s only idiosyncratic risk that earns a return. Everything else you can diversify away. I we’ll get real geeky there.
Mike: Anyway. So the style box, the prescription is to wash out your growth with value, wash out your large cap with small cap, but lo and behold, 1997, I’ll give you another paper. Mark Carhart, University of Chicago now shows that over a 30 year time horizon looking at 2,000 different plan sponsors that it was the style factor exposure was the sole contributor to excess return. So if a fund outperformed, it did so because of style factor exposure, not because of stock selection or security selection. So in that sense immediately, we had some problems with the style box implementation. Is that we’re washing out purposely exactly what the Carhartts study said was the compensated source of risks. So throwing the baby out with the bath water. We still find that half of the respondents to this particular study or participants in this study show evidence of style box orientation. But by large, this was a big source of problems in that for all the reasons that we talked about, you’re diversifying away the way the good risk and potentially keeping a lot of that bad risk.
Stewart: So I’m always cognizant of this whenever I hear you and I find understanding something and we’re on the same page, but I want to make sure that our audiences is as well. When you’re talking about exposure to factors, can you give some examples of factors versus some of the traditional metrics?
Mike: Absolutely. It’s a great question. So when you think about all the securities that are in an asset market, we can group those securities based on attributes. So some of those attributes may be their profitability, what’s their level of volatility, what’s their price-to-earnings price-to-book ratios. Some of those attributes have been shown to generate higher returns per unit of risk. And that’s a factor. So that’s what we define a factor. So factors like smaller size, higher value and in the content to Fama–French that would be a price-to-book metric, lower volatility, higher momentum, higher quality in the myriad ways that you can define quality and higher dividend yield. These six factors across the academic literature have been shown to produce higher returns per unit of risk. And that’s the key. A good factor should generate higher returns per unit of risk, which is contrary to the Capital Asset Pricing in modern portfolio theory. So these are the things that we consider compensated risks in our study. Everything else that adds to risk, but doesn’t add to return is considered an uncompensated risk.
Stewart: I just want you to know that Harry Mark can listen to this. And I want to make sure that he’s not absolutely out of his mind when we go to the next one, which is that you can over diversify and I’m assuming through an increase in transactions costs actually dilute your performance. That sounds to me based on my experience in managing money and that is something that’s often the case, but not frequently discussed. How did you come up with the over diversification idea? And where’s the right point? What’s the right amount if you will?
Mike: Yeah. And it’s a great question. So when we look at these called a portfolio factor analysis, when we do our studies, one of the things we look at is how many managers do you have in the mix? And in some cases it’s quite low, maybe one or two managers. In some cases it’s very, very high. We’ve had instances we’ve had hundreds of managers in the mix. And one of the things that is very, very apparent in the data is the more managers you have, the higher the amount of diversification that you have, the higher the tendency that you have essentially diversified away most of the active risk of the portfolio, but you are left with the fees. And remember that active risk is the source of active return in a portfolio. So if I have a hundred managers and they’re all diversifying against each other and at the end, my total portfolio has very, very little active risk, but the chance of earning a high active return is very, very low.
Mike: So you diversify away your active risk, but we do not diversify away the fees that you pay for these managers. And I would say one of the most common findings of this study is that a lot of these portfolios had so much diversification that they ended up essentially looking, feeling and behaving like the underlying benchmark minus the fees that were paid for that manager line-up. So I would say probably 70% of the cases in which we analyze this was essentially the outcome. Again, throwing the baby out with the bath water. We love the idea of diversification. We’ve been taught diversification, modern portfolio theory, but you can take it a little bit too far as well.
Stewart: Yeah. It’s interesting that is particularly the case. When you see times in the market, like we saw earlier this year, where transactions costs can go up, those transactions costs are not stable over time. And so you can exacerbate that problem over times when the diversification theoretically helps you, right? But actually can hurt you. All right. So the point number six here has to do with timing, right? Timing manager changes. And I think all of us have been taught if you’ve got a finance background, not to time, right? But there’s always the temptation to time, right?
Stewart: And it’s always like, “Oh, is this the right time to do this?” And it’s challenging. What did you come up with as far as what you saw people doing in terms of timing manager changes
Mike: Overall, we found a considerable evidence, the timing was in fact occurring in a lot of these portfolios. And so, let me just set the stage a little bit by what exactly I mean by that. Traditionally, when you hire a new manager and you hire that manager based off of a track record, usually a three to five year track record and the way that the process works. And it’s just a by-product of the way that people think, is that it’s very rare for a manager that’s underperforming to be on the short list in terms of the managers you’re evaluating or selecting. The challenge with that is if the source of the active return for those managers stems from uncompensated sources, meaning that you got lucky on a bet on a sector, you got lucky on a bet on a country or a region major three year track record that you’re being evaluated on.
Mike: The chance of that mean reverting is very, very high, right? So if it’s an uncompensated source of risk, it means it’s uncompensated over the long-term, very strong propensity toward mean reversion. So what we might expect to happen is that those active managers that outperformed in their recent track record may actually underperform going forward. And if you look at the academic literature on this topic, I would say there’s a, nearly a unanimous. So that’s very rare. And the academically.
Mike: Really a unanimous consensus on this, that, and I’m going to point out a paper here. It’s a phenomenal paper Journal Of Finance 2008, Goyal and Wahal. I think it’s called Selection Of Plan Sponsors or something like that. But they look at a very large number of planned sponsors over a many year period and confirm exactly what we’re talking about. That before a manager is hired, they tend to have a phenomenal three year track record only to find that once they’re hired, they tend to significantly underperform that track record. Also, very interesting is that once they’re fired following that period of under performance, they tend to go right back up to the top.
Mike: There’s a lot of cyclicality in manager performance, and we tend to hire at the peak and fire at the trough. Now that study was just updated this year. It’s not out yet published form, but it’s on SSRN. I highly encourage you to read it. It’s great stuff. Exactly the same findings. So precisely the same behavior going on. We find additional evidence of this in the risk report study as well, that there is a lot of evidence that we are timing manager’s selection, which tends to be a losing proposition from a kind of a portfolio wealth perspective.
Stewart: It’s interesting that right now, and I think we’ve been talking about this for a long time in the insurance asset management industry, but right now everybody’s talking about it, which is insurance companies have got to come up with a different paradigm in this lower for longer rate environment. They just can’t keep doing the same old thing, right? And everybody says it, GE insurance companies are slow to change and whatever else. And I think that’s true. And I think over time that may have served them very well. They make very long-term commitments and they don’t move quickly sometimes, but at the same time, I think it is really time for folks to have a good, hard look.
Stewart: And I was impressed when I read this piece that you did, that you really challenged a lot of traditional thinking with some hard evidence, some hard data, and you didn’t have an ax to grind when you started, right? Writing this research piece was a big commitment and I find it interesting that it is really challenging the traditional status quo approach on a lot of different fronts. I’m interested in knowing what you were thinking as you were going through this process.
Mike: This really is an outcome of much more than the last four years. We’ve been doing these kinds of portfolio factor analysis for at least eight years. And we kept seeing time and time and time again, the same issues pop up in these portfolios. So after a while, it just made sense for us to say, “Hey, we need to publish some of these findings to show that these things are pervasive problems.” And to your point, Stewart, right now they’re becoming more important than ever in that in a very real sense. We are at a difficult spot in terms of asset allocation. Especially for insurance companies, but really regardless of what kind of investor that you are very low return environment, very high expected volatility going forward, low levels of yield or income on a lot of these assets, very, very high tail risk, especially in the equity space.
Mike: How do we deal with this return gap, this rising volatility, a rising instance of volatility spikes, things of that nature? It really does require you to kind of think outside what I call the traditional box of traditional asset classes, a traditional approach to asset allocation. Because what we’ve shown in this study is that that approach leaves a lot to be desired in terms of uncompensated risk. And that same thing applies to a multi-asset scenario as well. So what I think this study is encouraging you to do is be precise in how you think about asset allocation. You can’t utilize your traditional approaches. I used to work for an insurance company as well, and I’m very well aware of kind of what goes on in that sausage making process. We have to be much more scientific, much more data oriented than that. If we’re going to achieve the kind of returns that we need to achieve going forward without stepping on these land mines of risks that we, that we don’t want.
Stewart: All right, last question. Only you have to be a finance geek dancer, this one, but it wouldn’t hurt, right? So I had students last fall who had multiple job offers and many of them more than one internship offer and just like liquidity in March, their opportunity set dried-up, right? No, their job offers got pulled. Their internship offers got pulled. It’s a tough situation for kids. They’re sitting at home. They’re doing classes on zoom and so forth. So here’s my question, Michael Hunstad on your graduation day, you walk across the stage, you shake the president’s hand and hand you your diploma. You walk off the stage, you’re getting ready to throw your mortarboard up in the air and you run into Michael Hunstad of today. What do you tell your 21 year old self in this environment?
Mike: Yeah, I would say the best piece of advice that I can give. It may sound a little bit bias, but it’s focused on the hard skills. You’ve acquired the skill set in college, but more and more as we go forward, if you’re in the investment arena, it’s less about intuition and more about hard data where the decision-making occurs. And that’s exactly the theme of this study as well, is that it’s about precision and data application. That I think is going to really the whole mentality, and I was there. I went to business school 20 years ago and it’s about fundamental analysis to stocks. And how do you feel about the management and all that kind of thing? And there will always be a place for that, but more and more, what we see is that it’s about risk measurement and can we quantify exactly what we’re doing in our portfolios?
Mike: So focusing on those hard skills is going to be much important today than it was 20 or 30 years ago. As that would be my first piece of advice. The second piece of advice, a little bit more soft. I give this to everybody. I always tell them, “In your job, treat every single meeting you have like a job interview because you never know where your next opportunity is going to come from.” So I’ll leave it at that. But I think those two things would probably be a pretty good base
Stewart: Advice. Good advice. Thanks for being on this is a very, very good piece of work that you’ve done you and your team have done. We’re going to make that available. I believe on the podcast when it gets posted as an interview article on our website. So thanks so much for being on.
Mike: Excellent. Thank you, Stewart.
Stewart: Thanks for being on Michael. Northern Trust Asset Management, head of quantitative strategies. If you like us, please follow us on all the major platforms. If you have ideas for podcasts, please email us at firstname.lastname@example.org. My name is Stewart Foley, and this is the Insurance AUM Journal podcast.