"Pointing to the dispersion of returns among managers, it’s really difficult to desegregate skill from luck." (Tweet)
Top Traders is bringing you Top Traders Round Table, a series of conversations with industry leaders on the subject of Managed Futures. On this episode my guests are Ryan Abrams, Portfolio Manager at Wisconsin Alumni Research Association, Ela Karahasanoglu, VP at Workplace Safety and Insurance Board, and Carrie Lo, Director of Hedge Strategies at CalSTRS.
In This Episode, You'll Learn:
- How Carrie diligently prepared her new Risk Mitigation Strategies (RMS)
- How much CalSTRS will initially allocate to the RMS part for their $200bn+ portolio
- How Ryan uses risk mitigation strategies in his organization
- How Ela is slowly restructuring her alternative portfolio
- The risks of risk parity strategies
- Is trend-following really an opportunity cost question?
- How Carrie has worked with the CalSTRS board on implementing this strategy
- Why Ela has focused more on tail-risk strategies
- Carrie’s manager selection process
This episode was sponsored by CME Group:
Connect with our guests:
Learn more about Ryan Abrams and Wisconsin Alumni Research Association
Learn more about Ela Karahasanoglu and Workplace Safety and Insurance Board
Learn more about Carrie Lo and CalSTRS
"A potential Achilles heel, at least of a risk parity portfolio, is your exclusively long assets." - (Tweet)
The following is a full detailed transcript of this conversion. Click here to subscribe to our mailing list, and get full access to our library of downloadable eBook transcripts!
First of all, welcome and thank you so much for taking time out of your busy schedules to join me today for this conversation. Not just about managed futures, but also alternative investments in general.
Before we jump into today's topics, share with me a short version of your background and how you got to where you are today, and Carrie, why don't we start with you. Tell us about your investment journey and how that led you to CalSTRS.
Thanks, Niels, I'm happy to participate. I joined CalSTRS in 2009 when they had just formed the innovation and risk unit. Prior to that, by experience has been quite broad in the finance industry, on both the buy side and investment banking / corporate finance in New York, as well as on the buy side in equity research and convertible bond research at a mutual fund. My academic work has been in business as an undergrad at UC Berkley, and I also earned a Masters in Finance from the London Business School.
All those experiences have really culminated in my position at CalSTRS now, in the sense that I've worked at large organizations and small organizations on the buy side and sell side. Now, being an allocator, we've essentially created a new group within a large stable institution. The purpose of the Innovation and Risk Unit is to test strategies that are new to CalSTRS and can diversify the total plan - primarily it's heavy equity data. In the innovation unit, I spearheaded our first investments into hedge funds, namely global macro and trend following. Those have since become part of a formal asset class called Risk Mitigating Strategies, or RMS.
Exactly, that's great. I appreciate it, thank you so much. How about you Ela? I know that you've also been on the consulting side, advising institutions about alternative investments, as well as on the manager side. So I'm really interested in finding out what lead you to the investor side of our industry?
Oh thank you Niels, and thanks for this great podcast and allowing me to participate. So, I'm told that I actually have what people call the Trifecta - the asset management, the consulting, and now the institutional or the allocated role. Right now, I'm at Work Place Safety Insurance Board. I oversee out multi-asset strategies portfolio. We're roughly a 30 billion dollar plan based in Toronto. It's mostly an insurance plan, and I oversee roughly 20 odd percent of that portfolio, that includes hedge fund strategies.
I started out my career as a trader in Turkey. I grew up in Turkey, and I was a currency and bond trader. I went on to the States; I worked at Merrill Lynch, the credit derivatives as well as asset management group. I did my BA at George Town around the same time and then I moved to Europe. I worked for the European Union on several projects, and then I moved to Canada. I worked for a New York based quant global macro CTA, and that's where I got my first interaction with the macro and managed futures in a CTA world. I was the head of quant research and trading for that hedge fund for roughly six years.
I moved to New York in the meanwhile. I moved back to Canada and worked for Mercer, as you noted. I was the Senior Manager Research Consultant covering hedge funds as well as multi-asset and other side strategies participating in asset allocation. My focus was really on the quantitative strategies, although I did cover discretionary as well.
After that, I spent a year at CIBC Asset Management. I was the institutional portfolio manager for the currency and asset allocation team. So, that was going back to asset management. I moved to Work Place Safety Insurance Board roughly a year ago, and since then I'm overseeing the development of the portfolio as well as now I'm overseeing the investment team as well.
Fantastic. Thanks ever so much Ela.
What about you. Ryan? Please share some of your background and how you ended up in the alternative investment space.
Yeah sure, so I started my career right out of school in alternatives working as a manager research analyst focused on CTA and macro strategies. So, manager research and multi-manager portfolio construction have been 'home' for me - they're where I've spent my entire career.
I did that for four of five years before joining WARF in 2011. I am responsible for a 1.4 billion dollar portable alpha program here. We also run an internally managed risk parity strategy which is the foundation of our asset allocation. As part of that, I'm also responsible for a sleeve of risk premium strategies that are intended to be a compliment and a diversifier to the long-only asset betas that we hold in that portfolio. So that's a bit of my background.
Sure, and of course you left out the best bit, Ryan, because you're also the co-author of one of my favorite educational white papers that the CME Group offers. I'm sure you know which one I'm thinking of here.
Yeah, so the Lintner Revisited paper - I worked that with my boss, and Liz Flores at the CME - highlighting some of the benefits of managed futures and institutional portfolios, and a bit of an update and refresher on John Lintner's early 1980's paper describing the same.
Sure, absolutely. Now, our conversation today will focus on how the three of you use managed futures as part of your portfolio and the specific role that it plays. I think it's safe to say that managed futures, and trend following strategies, most often are used by institutional investors as protection during periods of stress in the financial markets to mitigate risk for the overall portfolio.
Of course, in your case Carrie, as you've mentioned, you have set up a specific part of your portfolio for risk mitigation strategies of about 9% according to what I could find. Which is of course, in your case, not a small amount of money given CalSTRS' size. Could you start out by talking us through how you define the crisis that you wish to mitigate, and what lead you to make this decision?
That's correct. CalSTRS manages about 206 billion dollars now for close to a million teachers. In 2015, we went through an asset allocation study where we modelled a new class called Risk Mitigating Strategies and modelled how it would improve our funding ratio over a 20-year time horizon.
Within Risk Mitigating Strategies, we included four investment sleeves: they are trend following, global macro, systematic risk premia, and long duration treasuries. The investment committee did decide to allocate 9% to this new class. We modelled quite conservative Sharpe Ratios for these four new sleeves to include in the asset allocation study. The Sharpe Ratio was approximately .3.
Even with that low Sharpe Ratio, it did show that by reducing drawdowns of the total plan (particularly when equities go through a downturn) and reducing the volatility of the plan over the long run, it would improve the funding ratio in 20 years despite this new asset class not expecting to achieve our 7.5% actuarial assumption.
We don't know what the next crisis will look like, which is why we wanted to include several tools in our toolkit. We did look at previous crises, as far back as we could with the tech bubble and the '08 financial crisis being the most significant ones. We evaluated how each of these strategies performed during those periods; how they were correlated to the plan, to equities, to bonds; and what kind of return profile they provided - whether it was primarily out-performance during downturns, and reducing their own tail risk within those strategies.
Sure, sure, absolutely. Ryan, if my memory serves me right, you've done something similar with trend following and with a not too dissimilar size of your portfolio. Can you go into some of the details about what lead you to your decision?
Sure, so our starting points may be a little bit different from CalSTRS from the perspective that ideally our portfolio is not so reliant on the performance of equities. We've balanced risk across multiple asset classes and multiple economic scenarios or outcomes by the allocations to levered bonds, levered tips, and commodity futures. For us, a potential Achilles heel (at least of a risk parity portfolio or approach) is exclusively long assets. With expected returns being low, or spreads on assets to cash being low, we're not as confident or comfortable that some of the assets in that portfolio, that are intended to provide protection or diversification in bad environments, bad outcomes, crises, will offer the same protection that they have historically - specifically, treasuries.
Hopefully, we're able to generate a positive performance crisis environment in our portable alpha portfolio. In looking at hedge funds there's this area between asset beta and alpha (I think a lot of people refer to this as risk premium) where you've got these strategies that have risk that you can explain.
There's often economic intuition behind why these things generate performance - ideas like value momentum, and then in most cases, you've got very long, rich, performance histories, or a lot of empirical evidence pointing to the notion that these things generate attractive returns over time. We've kind of lumped trend following into this bucket. It's basically momentum, long and short, employed on a cross asset basis. Looking at the properties of trend following, it has the potential to generate performance during any kind of dislocation.
I guess I'd kind of widen the definition from crisis to dislocation because it needn't be a bad environment, for equities, per se, when these strategies can generate nice strong performance. They can generate strong returns during an inflation surprise, or a stagflationary surprise as well - both of which would not only be bad for equity heavy portfolios but portfolios generally, especially on the inflation side, it would be... it's a catastrophic environment usually for bonds. So, with assets being universally expensive on a long-only basis, doing some more stuff that can benefit on the short side or from spreads between assets makes a lot of sense; then having a heavy risk allocation to something like trend following that I guess builds out the robustness of the portfolio to risks or environments where not a lot of things work, or potentially offer protection.
In a deflationary surprise, you've got bonds and in an inflationary surprise, you've got commodity futures. I think we're much more comfortable relying on, hopefully, more than one asset or more than one investment to generate diversifying performance in those kinds of environments.
Yeah, this is a healthy sized risk allocation in the policy portfolio, so we've got about a 15% risk allocation to alternative risk premia strategies, and trend following is about half of that. So you see these studies run by consultants where you'll do the optimizer and it will say that you should do something North of 20% in trend following. So, we're not quite there, but I think that 8% of risk is a lot more than many institutions hold. So hopefully we're able to get the nice diversification benefits from that.
Just one quick question, when did you start your portfolio?
Yeah, so I mean we've been invested in trend following strategies for more than a decade through the Portable Alpha Program. We moved that over to become part of the Policy Allocation in March of this year, so it's only been a few months.
Great, okay. And how are you using managed futures and trend following in your portfolio?
Sure, so maybe going back to your question as to how we define crisis and whether we use managed futures in relation to that. It's similar to what Ryan described. Our journey started roughly ten years ago - that roughly predates me. We had more of a traditionally structured portfolio with less alternatives. There was some but not much, and the idea was to reduce, to not just get the returns but also reduce the total fund risk. So, we started adding alternatives, and that's how we ended up with several different strategies that really did serve that purpose, one of them being the risk parity strategy.
So, like what Ryan described, in our portfolio, we do have a sleeve dedicated to risk parity strategies which are geared towards addressing not crisis, per se, but market distress - distress periods. I think Ryan described it as dislocation. I think that's referring to similar concepts there. Risk parity strategies, the way we've structured them, are quite dynamic in the sense that they're supposed to address different periods of growth and not growth, and that should be used to address part of that concern.
The other part is really in the hedge fund portfolio, so the risk parity is roughly 10% of our portfolio, and hedge funds are also roughly 10%. When the hedge fund portfolio was envisioned and created, it was to address also, similarly, that. I would agree with Ryan's view about the regular risk premia alpha, and right in the middle, we think of it as alternative risk premia.
So, the hedge fund portfolio is really a combination of the alpha and the alternative risk premia. That's how we put that portfolio together, and in there we view hedge funds, in general, as a way of mitigating, or creating if you will, a positive skew or buffering the equity risk premium.
In there, there is a portion that's allocated to macro strategies. When we say macro, it really does cover discretionary macro as well as the host of systematic strategies including managed futures and some of the alternative risk premia. So, the idea there is really convexity, and to get skew. The strategies have been structured because the size was relatively smaller.
We started out with Fund of Funds rather than direct managers, but over time we've added more direct mandates into the portfolio. What we're looking for there is that in periods where the equity risk premia is not driving the returns that we want to get a decent buffer and lowering risk from that portion of the portfolio. We want to keep a healthy level of that going forward.
Sure, sure. Speaking of level, in one of my recent Round Tables, which I encourage all our listeners to go back and listen to, I was joined by three of the world's leading consultants, and one of them said that when it comes to trend following strategies, most investors are unable to allocate as much to this strategy as they really should, which Ryan also alluded to. Do you agree with this? As investors who have embraced it and the benefits of trend following, do you feel still, deep down, that you should really have a much bigger allocation, but politically that would not be plausible to do? Carrie, if I can come to you first on this one again?
Sure, I think Ryan and Ela made great points in that this is an expected positive return strategy over the long run. It provides a different source of return because you can go long and short, multiple asset classes, multiple geographies, and it's a diversifier both in negative tails and, potentially positive tails as well. The capacity issue and allocation size issue, it is an important one that we grappled with in that trend following has several benefits but there are times when it doesn't perform, when there are not strong trends in the market, and those periods can actually be quite long, and multiple years as we've seen.
It's a lot about setting expectations upfront on the purpose of the strategy, how it's expected to behave and when. There are also sometimes capacity issues about the strategy itself. We hear about investors piling in, or withdrawing from the strategy in mass, and issues about potential market movements, and how quantitative strategies may be influencing markets. There is always a trade-off, and I think trend following, there's a general definition of it, but there are a lot of details in how you invest or what areas you invest in with trend following, whether it's more short term or more long term, and certainly on the short term side, there's less capacity and more transaction costs.
From our stand point, we began looking at trend following in 2011, and at that time there were a number of investor pressures about fees and performance, and so managers were responding with developing more medium and long-term products that were lower fee. That was the route we decided to go, to invest in because both of those things were attractive to us and certainly on the fee front we wanted to access this strategy at a lower cost.
Sure, sure, absolutely, and we may touch on that a little bit later as well. What about you Ela? Do you see, quickly, walk around wishing you had three times as much as you have in trend following that you currently have?
Indeed! So, I think Carrie touched on a very good point, one of the items that is very important to explain to an institutional committee or board is that while we would expect... the returns are expected on an absolute basis, the underperformance could be for extended periods of time. Though a lot of the studies will tell you that you need to have 15-20% in a portfolio that is pragmatic or practical, especially with an institutional portfolio where you're aiming to get a decent level of diversification with all the tools available. It's not exactly feasible to go for something of that size.
However, it needs to be meaningful. We've talked to our committee as we were putting the portfolio on, that in the hedge fund portfolio, it needs to be a size that would have an impact. Managed futures is part of it, but we do have other strategies that address a similar concern. So, since we're not only constricted by trend following, per se, to address what we're looking to address, we have roughly 3-4% of the portfolio in that specific type of strategies.
We do also have Alternative Risk Premia, I mentioned earlier. I think the important point there is the convexity of these strategies? How big will they get, and what is the gamma profile? As you know most of these strategies are considered long straddle, but I think the importance is are they long gamma? If they are, how much do they contribute to the performances as the markets go one way or the other? As Carrie mentioned, I think that's a great point, really on the either tail. In that sense if... it's really a portfolio construction question, so I think we are positioned well.
As institutional investors, we're very fee sensitive. Of course, fees don't make or break the deal, but on the managed futures side it has been systematically higher versus other hedge fund strategies. That has not made our case simpler, and therefore this has not been something that would be easy to go back and say, "Hey, we want to grow." That is not the only solution, there are longer-term, medium-term, relatively cheaper solutions out there, so we'd like to bundle the two, and that's how we get to where we want to get to.
Sure, I mean it makes perfect sense. Now Ryan, you, of course, did the analysis you call the Bible study on this, so you know that 25% is the right number. So, are you ever going to get there do you think?
I don't know, I don't know that there is a right number per se. It all goes back to the portfolio that any given investor is running. While it's tough to get boards to focus on what a single line item does to a portfolio, I think that should be the focus here, particularly because it is a positive expected return, highly diversifying asset.
I think with that one you can notionally fund it without putting up a lot of cash. If you were to invest in trend following through a managed account, this needn't be an opportunity cost question. You can take arguably more equity risk, or more of some other kind of risk in the portfolio just because trend following will likely bring down the overall volatility of your portfolio. So, all you're doing is moving on to a higher capital asset line and then increasing the efficiency of your portfolio.
I think that the focus should be, for boards and for investors serving boards, on the portfolio effect and what this can do for your portfolio rather than an opportunity cost question. You can't get into flexibility with leverage - that isn't an easy conversation with boards necessarily either, but to the extent that you can relax that constraint I think that you can get to the point where there are some interesting and powerful things that you can do.
Sure, absolutely, and speaking about getting these types of strategies approved by the board in the first place Carrie, was there anything unique to this mandating comparison with other investment mandates? I mean how difficult has it been for your institution to embrace these strategies in the first place?
It's been a long process. We began with macro and decided to start in a very focused way with the small allocation with a very specific objective. This was in 2009 that we presented Macro, so right after the financial crisis, and articles every day about the evils of hedge funds. In the end, we did the analysis, we presented the evidence, we showed that certain hedge fund strategies did have low correlation to equities or did have asymmetric return profiles, or took advantage of volatility and provided diversification to what we already had in the plan. In the end, we did receive unanimous approval to go forward.
As Ryan mentioned, managed accounts can solve several issues, one of them being the cost. We decided to invest directly in the funds in a managed account format to address the issues of transparency, control, risk taking, and gating, things like that. So, that's been a very important tenant of our program over time. It is an ongoing educational process. There continues to be a lot of scrutiny of hedge funds and again the fees come up, and we continue to just provide the quantitative evidence as well as the intuitive qualitative rationale for the program.
Ela, I would love to hear your experience as well, I know you may not have been there when it began on your side, but in terms of also dealing with investment boards and advisory boards in general about this strategy area.
This is interesting because I also as you mentioned, I did deal with this while I was at Mercer, and being on the consulting side of the business was interesting because people have different problems, and every portfolio is slightly different. The major driving factor was, or the way we approached it was, especially equity-centric portfolios where the main performance driver for the portfolio was traditionally 60/40 or equity risk premia was the main driver. We explain that there needs to be various drivers of returns to diversify away from one driver of return or two.
We showed the addition and how the value-add of such strategies could really make a difference. Especially for clear portfolios, you could simply go into a 60/40 portfolio and add managed futures or a portfolio of managed futures I should say, and you could make a big difference. So, it becomes very visible in that sense. In our case, so as I said, it predates me, but this was really a portfolio construction exercise. It was not just obviously hedge funds; there were several other things that were added.
Since then our committee and board, they've been very comfortable with progress. They understand the role the portfolio plays and the components, the pieces, do play in the portfolio, and the expectation that the next time frame or the stress periods that we've seen over the last four or five years. Late 2014, one, was a good example - same as early last year, or August of 2015. We've seen several these and it's been quite good to emphasize and validate the reasoning behind the portfolio.
What we're looking to do is education is endless. We want to move on to the next stage, and Ryan had made a good point that these can be powerful tools if you can move from a simple allocation. We also do have overlay structure, although this is not specifically for hedge funds. That's the way we would envision using these strategies because you can get quite impressive cash efficiency by just partial funding and more customizing of volatility. That could be an add-on and it could be quite a bigger impact on the portfolio, and that's where we would like to go and move on. We've been talking to our committee, and broadening the "toolkit" if you will.
Sure, great. You're absolutely right, education is never ending, and of course, as you should all read Ryan's study and listen to this podcast as part of that. Ryan, what about you? I mean you said you co-authored the study with your boss. So, I don't know whether that's your current boss, but did that help get the investment board onside so to speak?
So, unfortunately, a different boss, but fortunately a boss who already understood trend following and had been investing in it. It goes back to education, education, education, education. You must explain to the board what you're trying to do and why then review that and refresh their memory from time to time.
I don't know the composition of other boards, but these are often very busy, very important people, and you're only getting a few hours with them, in our case, every quarter. So, they've got a lot of other responsibilities that they're managing, including being on a board that you're serving. So, reviews, reminders from time to time of what you're trying to do, and why are always helpful from my perspective.
Sure, absolutely. So, once you have the approval of the investment boards and you've had to start finding managers that could give you the desired performance profile. What were you looking for, and how did you go about finding this Ryan? If I could stay with you, I know you might have gone in a different direction than Ela and Carrie.
No, so I think the direction is probably similar. Like Ela, we're focused on trying to get a convex return profile, and I think one of the best places to find that is in the medium to long term trend following duration. So, that's been our area of focus. We've talked about fees as well, but I think the reason why fees (for that segment of the trend following space) tend to be lower is correlations among managers or among strategies go up as you extend trend duration. Performance just looks increasingly similar, and then that also tends to be the area where you get the desired performance profile. That's been the focus for us as well.
Sure. What about you Ela? Is that how you went about it as well?
We've been looking to add a little bit more, and the portfolio structure has been not exactly in line with what I'd like to see in the long run, but the focus recently has been trying to extend maybe a little bit into tail risk strategies, and to complement the managed futures that we have in the portfolio. So, it's been a journey on our end, and I've been looking to expand our horizons in the sense that there are a lot of different things that we want to add into the way we deal with different time frames.
As Ryan mentioned, the fees have been a little bit difficult to deal with, and that has sort of created, and I guess I should say, encouraged us to look outside of what we are currently utilizing. Initially when constructing the entire subset of managed future and macro strategies, we did look at the whole host of (I like to call them) the beta strategies if you will, although they're not really the typical beta strategies... beta as well as alpha strategies into the mix, and that's how we took it forward.
Sure, sure, sure. Now in your case, Carrie, the adoption of this mandate has made you one of the largest investors in managed futures in the world. Did this make it easier or harder for you to find the right managers?
The manager selection process never seems easy, but we did want to start with a very broad universe, so there are nearly eight hundred trend following managers. Because of the size of our allocation, we did place more emphasis on the larger managers and those that were institutional quality, and that we wouldn't have too large of a holding ratio... or percent of their assets with our investment.
We did meet with nearly fifty trend following managers. We ultimately selected six, so we did try to take a very broad view. Ultimately when we decided on how we would construct the portfolio, we evaluated each manager on a standalone basis, but as others have alluded to how they interact or how they diversify each other was very important. So, along the lines of time horizon, we did want to include some short-term managers, so maybe down to a week or less - nothing high frequency, up to the longer term managers or having some that have some adaptive process across those time frames.
We also wanted diversification along the lines of the markets that they trade and the instruments that they trade. So, some will trade less than one hundred, others will trade north of three hundred, and we also wanted exposure to various geographies. Although the correlation of the managers can be very high, we found that sometimes there is enormous dispersion among the returns, because of how they construct their portfolio, how they manage the risk. So we also wanted to diversify along those types of approaches.
Yeah, I mean essentially you mentioned a dispersion of return. I think that dispersion between trend followers returns, are increasing in many respects. I think that in recent years it's not been that easy to really forecast how that space is reacting to certain periods. I'm sure you know more about maybe the reasons why that might be, but it's not quite as it was ten years ago, and twenty years ago regarding if you knew the performance of one manager in the trend space, you kind of knew the whole area, but I see differences as well. Which makes your job a little bit harder.
Ryan, I also wanted to ask you a little bit, because I think that from memory, that you've also gone more maybe in the route of the beta replication strategies for trend following? Am I right?
Yeah, that's correct. So, I think that there's a lot of differentiation among trend following strategies. Just kind of using that as a starting point, even something as simple as a very simple trading rule: buy when price is above a moving average and sell when it's below it, this somewhat incredibly generates positive performance over time and so there are all sorts of bells and whistles and risk controls that one can add around a set of very simple trading rules.
I think we're somewhere in the middle on the continuum. I think just because this is a widely known, well-understood strategy where there can be high correlation among strategies as well, suggesting that managers are doing something similar doesn't mean that there's not a lot of room to add value through implementation. So, I think that's something certainly to be aware of and cognizant of on the manager selection side.
Pointing to the dispersion of returns among managers, it's difficult to disaggregate skill from luck in that area. So, in a given year did manager A outperform manager B because they're better, or because of some kind of accident - because their asset allocation was slightly different, and that helped them, or their models were a little longer or shorter and that helped them? So, from my perspective, these are all difficult factors to try to understand and weigh, and I just think that you try to do the best that you can.