Not Another Investment Podcast
Understand investing beyond the headlines with Edward Finley, sometime Professor of Finance at the University of Virginia and veteran Wall Street investor.
Not Another Investment Podcast
Mastering the Balancing Act: Decoding Absolute Return Strategies (S1 E15)
Embark on a journey with me, Edward Finley, to crack the code of absolute return strategies, the hedge funds aiming for consistent returns no matter the market mood swings. We're tossing aside the economic weather vane and zeroing in on how these strategies skillfully balance the scales between long and short stock positions and more complex trading strategies. By dissecting equity market neutral and managed futures strategies, you'll grasp the art of maintaining an even keel in the tumultuous seas of the stock market, where every wave of volatility is a test of precision and foresight.
As we chart the course for future episodes on arbitrage strategies, prepare for an education that's not just about surviving the market's storms, but thriving within them.
Episode Slides: https://1drv.ms/p/s!AqjfuX3WVgp8ukNZD4f14GAW6CNV
Thanks for listening! Please be sure to review the podcast or send your comments to me by email at info@not-another-investment-podcast.com. And tell your friends!
Hi, I'm Edward Finley, a Sumtime Professor at the University of Virginia and a Veteran Wall Street Investor, and you're listening to Not Another Investment Podcast. Here we explore topics and markets and investing that every educated person should understand to be a good citizen. Welcome to the podcast. I'm Edward Finley.
Speaker 1:Well, last time we introduced the idea that hedge funds aren't really an asset class. Instead, we learned that there are really strategies by which the nature of the trades creates a mix of common risks, like equity risk and interest rate risk, with some alternative risks, like non-normal risks, those that artificially depress volatility or have asymmetric payoffs, like options and futures, as well as leverage and illiquidity. By focusing on the mix of those risks in a strategy, we were able to identify three broad categories of hedge fund strategies, which is sort of like asset classes, because they share a common set of risks. Those three were dynamic market, absolute return and arbitrage. Last time, we also unpacked the largest category, which is dynamic market, consisting of equity, long short and macro. Today we're going to turn our attention to the smallest category, and that's the category we called absolute return. Just to give you some numbers there, this is the smallest category. It's about $400 billion, so really quite tiny, maybe a tenth of the size of, say, us equity markets. What about absolute return? What can we say in general about this category? Well, these strategies seek to identify anomalies in market pricing that allows them to earn a consistent positive return without regard to the economic or market conditions, and therefore they take very little exposure to equity or interest rate risk. I'll repeat that Absolute return strategies, like the name suggests, are those strategies that seek to identify anomalies in market pricing that allow them to earn consistent positive returns regardless of economic or market conditions. We're going to focus today on the two strategies that predominate in absolute return equity market neutral and managed futures.
Speaker 1:Let's start with equity market neutral. That's really very similar to equity long short that we talked about last time in terms of security selection. So first, the manager operates in the universe of equities, that's the universe of securities in which she's trading. The manager is either going to create straight longs and shorts, and you'll remember that means that she's selecting long positions in stocks that she thinks are going to go up fundamentally and she's selecting short positions in stocks that she thinks are going to go down. But, importantly, she's going to be doing this together or she might be engaged in pair trading, and, you'll remember, pair trading is long and short securities where the choice is not independent of each other. It's really relative to each other. The only difference here is that the manager here will attempt to completely neutralize any exposure to equity risk, and you'll recall that in directional long short, the manager is either doing straight longs and shorts or pair trades, but she's also making a decision about how much equity risk to own, which will change over time, ergo the name dynamic market risk. Here the manager is doing straight longs and shorts or pair trades and is instead seeking to strip out any equity risk, ergo the name equity market neutral.
Speaker 1:Because of the second quarter of 2023, equity market neutral strategies accounted for just about $50 billion in assets under management. That's quite small. That's about 1% of all hedge fund strategies and is about 20% of absolute return strategies. All right, well, let's take a look at what the trades look like in terms of the risk exposures. Well, unlike the equity long short that we looked at last time, where there's going to be some degree of market risk, here the manager is trying to strip out any market risk. That means concretely that during bull markets, her longs have to go up faster than her shorts are going down, and during bear markets her shorts have to go down faster than any of her longs are going up, and therefore the strategy's performance depends really heavily on the performance of both the longs and the shorts, which is very, very hard to do. And so this strategy, by virtue of that dynamic, is gonna have to exhibit some directionality, that is, some correlation with equity returns and some convexity, that is to say that the rate of exposure to equity risk will increase more quickly depending on market conditions, and so it's a bit of a misnomer to imagine that there's no equity risk here.
Speaker 1:The name is a little misleading. The attempt is to strip out equity market risk. Yes, it's true, but by virtue of that dynamic, it means that during bull markets and bear markets, the manager's strategy is going to have some pressure for the longs to do better than the shorts in the bull market period, for the shorts to do better than the longs, and that will give rise to some equity correlation. So although equity beta is gonna be close to zero, it doesn't mean there's no correlation to equity markets. Now recall that that will imply something about the kind of volatility of the securities that the manager is selecting. So beta is the covariance, the co-movement of the security in the market divided by the variance of the market.
Speaker 1:Well, we can rewrite that formula to state it somewhat differently. We can rewrite it to say that the beta equals the correlation of the asset, the security, and the market times a scale factor which is the ratio of the securities volatility to the markets volatility. So picture beta equals correlation times, the securities volatility divided by the markets volatility. Well, if you've got a strategy that's going to have some correlation to equity risk but it's going to show very low beta, that means that the low beta can either be because there's low correlation I've just said that I don't think that's very likely or it can be that the securities volatility is far lower than the markets volatility, making that scale factor, that fraction, very small.
Speaker 1:It's also likely to be the case that the manager is going to have to have net positive exposures to geographies and sectors, and so, while she might be market neutral in terms of equity market beta, she won't necessarily be beta neutral to, say, us equities, or beta neutral to industrials or consumer securities. And so these portfolios tend to be very, very highly diversified, hundreds and hundreds of positions. Again, a distinction from the earlier case. Okay. So if we understand the trades from that point of view, just to recap, we would expect that this kind of strategy probably has much more of the alternative risk exposures that are driving returns, not so much the common risk factors. We probably are going to expect that this strategy is going to show real evidence of non-linearity, because of the kinds of trades. We're going to expect there to be higher levels of illiquidity, we're going to expect there to be some degree of leverage and we will expect there to be correlation to equity risk, but just very low beta, which means that the volatility is going to be skewed in one direction.
Speaker 1:All right, so let's take a look at the HFRI equity market neutral index. Again. Our period is 1997 to 2023. First, what do the regression results tell us? Well, first, the regression results tell us that our common risk factors equity risk, tenure-treasury, investment-grade bonds and high-yield bonds explain only about 18% of the strategy's returns. So there are clearly drivers of return that are not captured by our common risk factors. In addition, the only one of our common risk factors that has any statistically significant correlation is equity risk. So that's not terribly surprising, but it's not economically significant, that is to say, for any 1% change in equity returns, the strategy's returns will change by about six basis points, and remember that means 6-hundredths of 1%. So very, very little equity exposure there. And so, true to its name, equity market neutral it would seem that linear exposure to traditional risks, including equity risk, is de minimis. There's very little, and that's consistent with what we built up from our bottoms-up analogy. How about the summary statistics? What does that tell us? Well, over that period, the average equity market neutral manager earned about 4% return, with 3.2% volatility, delivering a robust information ratio of 1.27. You might recall that equity markets in general have about a.6 information ratio and our model 60-40 portfolio had a similar sort of information ratio. So it's quite big 1.2% return for every percent of volatility. As expected, the correlation to US equities was 35%, so there's meaningful correlation.
Speaker 1:The strategy here is moving in the same direction as equity markets move, just not to the same degree, but, again, not terribly surprisingly. The beta is very low 6% beta. That would suggest to us that there's going to be some directional exposure to equity returns, but it's going to be pretty modest. There's seemingly no exposure to interest rate risks, with correlations of negative 6% and a beta of negative 2%. But what we do see is that while it's a symmetric distribution, there's no real skew. The tails are quite big. Kurtosis is 2.5, and that's telling us something about the non-normality of the strategy given its trades. And we also see that there's a relatively high auto correlation about 11% auto correlation, which is telling us that there's a pretty meaningful exposure here to illiquidity.
Speaker 1:The state-dependent returns help clarify the same picture. It shows a modest degree of directionality with equity returns and a fairly modestly positive directionality to interest rates. But primarily, what's there is convexity, so let's take it apart First, with respect to equity returns. It shows that modest directionality because we see that the worst average monthly returns occur during the worst monthly returns of equity markets and the best monthly returns occur in the best months of equity markets. And we can observe those returns as we move from quintile to quintile following the same direction as equity markets. But there's where the parallel breaks down, because we see that the percent to which the strategy participates in those equity market returns is inverse, that is to say it has very little participation in the lowest quintiles, the worst months of equity market returns, but it goes up pretty materially to about 15% to 20% during the remaining quintiles, and that confirms for us that while we have some modest exposure to equity returns, we notice that that exposure is increasing with equity risks increasing changes.
Speaker 1:We also see that the correlations within each quintile are reflecting a degree of roughly independence from equity returns. In the worst markets the correlations between their returns are rather high. So, despite very low participation, they're moving in the same direction. But by the time you get to the best monthly equity returns, the correlations between the strategy and equity returns are basically zero, so roughly, roughly independent.
Speaker 1:We can say something very similar around the interest rate exposure, because here the interest rate exposure shows some evidence of sort of very modest negative exposure to interest rates. So the short interest rates, which would be consistent when we think about how important it is for the short positions, which is really borrowing and it's sensitive to interest rates, how important the short positions are as well as the long positions. The participation in each of the quintiles, though, tells us a very, very different story. It tells us that, essentially, the participation is independent, that the strategy really shows very little connection with what interest rate returns are doing in any given period, and so when we think about market neutral strategies, I think what we come up with is. It's fairly consistent with its name it seeks to be equity market neutral. As measured by beta, however, we see that there is necessarily positive correlation to equity risk, and that's because of the nature of the trades, and we measure that in the correlations. We also see that that suggests that volatility is probably going to be understated, and it's going to be understated in part because they may be choosing lower volatility securities, but also in part because the nature of the trades might artificially dampen volatility, and that combination makes that fraction of the volatility of their portfolio of securities, divided by the volatility of the market, be artificially low relative to the amount of risk that the strategy is taking.
Speaker 1:All right, let's turn then to the second of the absolute return strategies, and really the quite larger of the two. Here we're going to be talking about managed futures. Managed futures is about $350 billion of the $400 billion in absolute return strategy, so it's quite a lot. Let's talk a little bit, though, before we jump right in. Talk a little bit about futures contracts and how they work, because that's really what this manager trades in.
Speaker 1:So a futures contract in operation is just a contract for the delivery of an asset at some point in the future. Very logical these contracts have standardized terms and therefore the contracts themselves trade as securities. You might remember from the episode in which we talked about how securities trade, this makes futures different from fallowards. So futures contracts it's a contract for the delivery of an asset in the future, with standardized terms and therefore the contract itself trades as a security. What are the terms of the contract? Well, the consistent terms are going to be the quantity of the asset, a price specified for the asset, a specified maturity or delivery date for the asset and a method of delivery.
Speaker 1:The long position on a futures contract is going to be the trader who commits to purchase the asset in the future and the short position is the trader who commits to deliver the asset in the future. So the long position trader commits to purchase the asset and the short the trader who commits to deliver the asset. But here's another important distinction the long holder of a futures contract is obliged to purchase the asset and the short holder is obliged to deliver the asset when the contract matures. And this is in stark contrast to, say, options, and you'll remember that options are like futures except, as the name suggests, no one's obliged to deliver on the contract. It's the holder of the contract who decides whether the other party is obliged to deliver Futures, not so Futures they are obliged to deliver. As a result, futures prices when futures trade and we're going to do much more on this later, when we talk about commodities, because that's the principle way in which investors invest in commodities.
Speaker 1:But for now we'll just say that futures prices when they're trading, are determined by a combination of four factors. The first is the spot price, that's the price at which the underlying asset is currently trading. Second, the risk-free rate, that's just the overnight treasury. Third, any cost of storing the asset. And then, last, the convenience yield. And the convenience yield is not something that we can observe directly. The convenience yield is just the implied expected return of holding an inventory of that asset. Now it's important to note that futures contracts can I just kept saying asset they can really be contracts in respect of any asset. Later, in another episode, we'll talk about commodities, which is a very big part of that world. But it's also possible to own futures on equities, on a particular equity, on an equity index. You can own futures on interest rates. You can own futures on underlying bonds. You can own futures on portfolios of securities. Futures can really be built and designed to deliver any kind of asset, and so when we talk about the nature of the asset's price, the nature of the futures price, what we really are referring to is how that trades relative to the underlying asset. Okay, so that's a quick and dirty of what futures contracts are.
Speaker 1:What do managed futures managers do? Well, here's their trading strategy. These managers are what's called systematic traders. So systematic traders mean that they're not making selections based on some fundamental view of the underlying asset in a futures contract. Instead, they're focusing on momentum, which is to say the nature of the underlying asset's price to continue to move in one direction once it begins to move that way, or convergence when an asset's price moves away from its long-term average, its tendency to revert to the mean. This is in contrast to discretionary futures strategies. Systematic futures strategies are really all about momentum and convergence, that is, to the degree to which the price is going to continue to follow a trend, or the extent to which it will reverse that trend and revert to the mean. Discretionary futures strategies might be used by macro managers, and that's because they're taking a very specific view of a directional nature, and so discretionary futures strategies do a very different thing.
Speaker 1:Here we're dealing only with systematic trading strategies. These managers then build models that seek to identify trends in more than 100 different futures markets not just commodities, but futures markets and equities, rates, currencies, etc. So what might we expect the risk exposures to look like for that type of trading strategy? Well, first, like all absolute return strategies, we would expect to see little or no beta to equity risk or interest rate risk. The managers here are focusing purely on either momentum in price or convergence in price, and they're not interested in their returns being helped or hurt by what the market is doing overall. So, like other absolute return strategies, we would expect to see little or no beta to equity risk or interest rate risk.
Speaker 1:These strategies are going to have pretty material tails, we might think, because they're highly levered. A futures contract requires only about 10% margin. So, to give you an example, if I want to own a futures contract with a nominal value of 100x, it means I only have to post about 5 to 10x in order to buy or sell that futures contract, and that means that it's a very levered strategy. That's sort of upwards to 9 or 10 times levered. In addition, these strategies are often going to be rather illiquid, particularly in times of distress. You can measure the liquidity of any particular managed futures manager by looking at what's called the margin to equity ratio. That's merely looking at how much margin the manager has on deposit divided by the manager's total assets under management.
Speaker 1:The kind of risk that we would expect to see here will also depend a little on whether the manager is a short-term or long-term trend follower. There's a distinction to be made here. In the short-term trend follower case, these managers specialize in the microstructure of markets, how markets clear minute to minute by individual trades. They enter a position immediately after they observe a trend reversal. They execute a large number of trades in a day, many, many, many trades in a day. They tend to be driven because they're so short-term in nature. These price movements tend to be driven primarily by behaviors and preferences. They're very, very hard to predict. In the case of short-term managers, when we think about the kind of risk exposures they'll have, we would expect to see that it will be very, very unstable because of how hard it is to predict these behaviors and preferences. Long-term trend followers, on the other hand, slowly increase their position as they observe a trend develop in the price of a futures contract and then slowly unwind their position when they begin to see a reversal of the trend.
Speaker 1:These managers execute far fewer trades. They might trade monthly or sometimes up to six months in duration. They tend to be driven, therefore, less by behaviors and preferences and more by secular trends, secular economic trends, which can be slightly more predictable. That means that, with respect to the long-term trend followers, we might expect to see some correlation to equity market risk. That's because those longer-term secular trends will feed through to their successes, in part because the trades that they're executing, and whether the price is in momentum or in convergence, are really more a feature of a secular trend, that is to say, more fundamental to economic exposure and less about behaviors and preferences. We would expect these managers to be very long volatility, that is to say, they are expecting to earn return for volatility. That's because in both cases, there's going to be a very large degree of unpredictability. They earn a return for that unpredictability. As a result, we would expect to see some equity correlation. We would expect to see not very much equity beta. We might see some tail risk. We might see some auto correlation, but it will depend a lot on the particular manager.
Speaker 1:Let's turn our attention, then, to the aggregate data. What do we really observe in the market? We'll start as we've done up until now We'll start with the regression results First. What we notice is that the regression results tell us that common risk factors equity risk, interest rate risk, investment-grade bond risk, high yield risk explain only about 12% of the strategy's returns. Very, very little Nearly 6% of the returns, and the strategy's average returns were only 6.6%. 6% of the returns are not explained by the common factor. There are clearly drivers of return not captured by common risk factors.
Speaker 1:We note that there are really two statistically significant correlations, and each of those is pretty meaningful economically. The first is we see that there is a statistically significant relationship to equity risk about 22%. Again, for any 1% change in equity market returns, the strategy would have a 22-basis point change in the same direction. We see a negative 22% correlation coefficient with risky credit, high yield particularly. That would tell us again if high yield returns go up by 1%, the strategy's returns would go down by 22 basis points. What's interesting about that is it suggests that really maybe shorting high yield was a source of return among all of the managed futures strategies in general, but it's also possible to imagine that really it's just an anomaly. It's just a way of showing us that the managed futures strategies have, on the one hand, some positive relationship with equity returns and, on the other hand, a pretty equally sized negative relationship with equity returns.
Speaker 1:It's not clear entirely what we're seeing. Investment grade credit nearly had the same negative relationship, but it wasn't statistically significant, leaving open the possibility that maybe there was some short exposure to interest rate risk. I mentioned already that the average returns were 6.6% and the volatility was almost 8%, which delivers a very strong information ratio of 0.83%. Not surprisingly, we see that the equity correlation is 23% and the beta is only 12%. That confirms that there's going to be some exposure to US equity market risk here. It's not entirely subsumed in very low beta. The beta is low, but there is some more meaningful economic risk exposure here in this kind of strategy. That's very likely just going to be the result of momentum and convergence in these trades. Because in general, equity markets exhibit momentum and convergence in general, it's not surprising that we would see some of that here. The correlation in beta to 10-year treasury is basically zero. We might conclude that, as we would expect, there's very little, seeming very little, exposure to interest rate risk.
Speaker 1:However, unusually, given our expectations because of the trades unusually, we see that the distribution here doesn't show a skew in one direction or another and really shows no tails that are different than normal tails. That's a bit of a headscratcher. However, if you look at an histogram of the monthly returns and an histogram is just a bar chart where you kind of create on the x-axis return intervals starting with, say like, negative 6% a month, and then at the other end of the extreme, positive 6% a month, and you count how many months the returns were in each of the, say, dozen intervals that you've selected, and it gives you a sort of a view of the distribution of monthly returns. Well, why would you do such a thing? Remember, when we talked about higher moment statistics like SKU and Cretosis, we said that you need many, many, many observations in order for the measurement to be valid, and typically in securities and in finance we don't have that many observations to be valid so often. If you get a very peculiar looking result from SKU and Cretosis that doesn't line up, it might be useful to take a look at the histogram to sort of see well, does it look like a normal distribution? And when we look at the histogram for the monthly returns of managed futures managers, we see that it looks like anything but normal. And, in particular, what we see is that there is a very large number of monthly returns that are either very modestly positive or negative, and that a large, or increasingly larger number of monthly returns are very significantly negative, that is to say, something like a third of the monthly returns are in the range of, say, negative 1% a month to negative 3% a month. In contrast, we can see that while there are lots of average monthly returns that are positive, they don't bear the same interval as on the negative side. And likewise we see that there are meaningful tails at the outskirts, at the sort of negative 6% end of the curve and the positive 6% end of the curve. And so it just might be that, because we have so few observations, the technical measure of skew and kurtosis is really masking what's really going on, and looking at the picture helps us see that, as we might expect for this strategy, we observe that, in fact, this is not a normal distribution, very much in line with what we expected.
Speaker 1:Turning our attention then to the state dependent returns. It kind of illustrates here the importance of nonlinear sources of return in the strategy. We see that kind of like equity long short. We see this kind of collared relationship with equity returns the worst average monthly returns occur at the same time as the worst average monthly returns for equities and conversely, the best monthly returns occur during the best equity monthly returns. What we see, however, in particular is very, very clear evidence of convexity. If we look at the participation of the strategy, we see that it's nearly zero in the worst returns. It increases through quintile two and quintile three and starts to decline for quintile four and quintile five, meaning that the nature of these trades, because their momentum or convergence trades, introduces a very large degree of convexity for which one can earn compensation, but it's not going to be something that we're going to be able to see in, say, a linear regression or by comparing its volatility. When we turn our attention to the 10-year treasury state dependent returns, we see much clearer evidence of no relationship between the strategy and interest rate returns. There's no relationship between the best monthly returns for one and the best monthly returns for the other. We see that the participation is negative for the first two quintiles, positive for the next two quintiles and zero for the top quintile, and so the correlation between the returns in each quintile is basically zero. So interest rate risk really doesn't play any role here at all.
Speaker 1:Okay, that's managed futures, and so I'll stop there. Next time we're going to take up the next and final category of hedge funds arbitrage strategies. Now, there are a lot of arbitrage strategies. In fact there are five convertible arbitrage, fixed income arbitrage, merger arbitrage, distressed and event-driven. That's a lot to do in one episode, so we're going to break it into two episodes, and next time we'll explore the bond arbitrage strategies, namely convertible ARB and fixed income ARB, and in the following episode we'll take up the equity strategies, namely merger ARB, distressed and event-driven. Until then, thanks again for listening and we'll look forward to seeing you next time.
.jpeg)