Rebalancing Discussion with CIO Don Pierce and Arun Muralidhar

In the new economic paradigm of higher interest rates and higher inflation, allocators are considering different ways to rebalance to meet their targets. In this discussion, Christine Giordano, editor of Markets Group's Institutional Allocator, interviews Donald Pierce, the Chief Investment Officer of the $14 billion San Bernardino County Employees Retirement Association in California, and Arun Muralidhar, who has a Ph.D. in managerial economics and finance from MIT and is co-founder of M-cube Investment Technologies LLC, (developer and licenser of  AlphaEngine® software) and the asset management company, AlphaEngine.


When Pierce joined the fund in 2001, he wanted to find a way to enhance the range-based rebalancing that had been in place. Since 2005, he and Muralidhar have been working together to improve SBCERA’s rebalancing methodology, and Pierce has gained about 100 basis points per year. Throughout this interview, the two will elaborate on how exactly they have done it and explore how their process responds to market volatility, from the Great Recession of 2008 to the high interest post-pandemic environment.



Christine Giordano: Donald, you've been at the fund since 2001 and you've initiated the software since 2005. Can we talk about what convinced you to put this in place and what it took for you to get there?


Donald Pierce: Thank you very much. When I first joined the system back in August 2001, the peak of the market that year was in March. So we didn't really start getting a sell off until later in 2002 and continued into 2003. The range-based rebalancing we had at the time didn't feel adequate for what we were dealing with. Unlike what was popular at the time, which was the idea of perfectly implementing your policy portfolio as the ideal, ranges were an acknowledgement of the frictions in implementing the perfect policy. But that rebalancing was not viewed as potentially additive, but as a risk mitigant. In general, they were right in the way they were doing it by just using range-based rebalancing and letting the market tell you when to move based on price movement. And so in 2003, I had proposed a feasibility study to review our rebalancing method and while the concept was strong, the proposed execution was lacking.

That is when around 2004, by happenstance, I met with Arun when he was showcasing AlphaEngine and I was floored because he was dealing with things I was only just scratching around at. I had a problem wherein the back of an Excel spreadsheet wasn't entirely convincing to the board. Instead, a model and software application that was pulling live data down to make better decisions was something we took back to the board in July 2005 when we went live. But in order to explore this, what we really wanted to do was make sure we're not transacting more than we would if we were just doing range-based rebalancing. So the board was engaged at least because we kept telling them that we have this concept and this idea, and I think that they were open to it to the degree that we didn't transact more than we would have otherwise. From there, we went live in in July of 2005. And then we brought on an overlay manager to help us as Russell Investments to help us in 2006. And it's been working that way ever since. The genesis was a two year incubation period, but with the accelerant of a software program that really dealt with the issues. I'm sure Arun can go over what prompted him to do it because at that point it was outside the viewpoints of asset allocation and asset management for pensions.


Giordano: Arun, can you tell us what led to this new idea?


Arun Muralidhar: Sometimes it's wonderful to make big mistakes. And I tell a story which I tell to all my students that in 1998, I worked for the World Bank's pension plan, and I put an asset allocation in place where the expected return on equities, for example, was 10% and that had been pulled from all the major banks and asset managers. Then I left to go to JP Morgan (in 1999). And when the tech bubble burst, we put very tight ranges on the assumption that it was actually good risk management.

To Don's point, why not have perfect policy implementation - when you get into a secular bear market, the tight ranges make you buy, then you lose again, then you buy again, and you lose, and they can actually be a risk-increasing activity. At the same time, I was in charge of the research on 20 billion of currencies at JP Morgan, which was being done on Excel spreadsheets. Enter the World Wide Web and it just struck me that finance is nothing more than taking data and converting it into decisions. Today, they call that Big Data AI, but all I wanted every morning when I woke up was to have a simple GPS that every asset owner could have had to apply on the one decision that affects 100% of the portfolio. But then I saw all the clients at JP Morgan getting smashed because their asset allocation took a beating in the tech bubble. Forgetting about the alpha, the question that came up was how we could help pension funds and endowments manage the single biggest risk in the portfolio, which is asset allocation, in a slightly more intelligent and informed manner? And so essentially, what we were trying to do with the software was allow intelligent folks like Don to take their ideas, put it in software, and then allow it (AlphaEngine) to process (regularly). That way, clients would have exactly the same process (on 100% of their portfolio) as a JP Morgan or a Blackrock would have in managing 2% of their portfolio. It was very idealistic because we really, genuinely thought we could transform the industry since nobody was providing the service at the level we were offering it at. And that allowed the CIOs who would be interested to do this in a much more efficient manner. 


Giordano: Looking into the data, and sorry if I'm kind of digging into your secret sauce, what exactly do you look for that makes the difference here?


Muralidhar: For us, data typically comes from vendors. So, you could have economic data coming from the Federal Reserve, the OECD, or the IMF, which is a lot of data that's being thrown at people. And then you have price data on things like securities, the price of oil, and the price of copper, along with credit spreads and volatilities. So, there's a lot of data that's sitting out in disparate places. Our idea was to centralize the data in one place, where somebody like Don wouldn't have to go hunting for it. And with a couple of clicks on our website, he could find it and make sure that it was cleaned up before he acted on it. That is a service we're providing with a team in Bangalore which still does that every day. They're gathering that data, scrubbing it, and making sure that it's clean. And then he can write whatever rules he wants to say that as the price of oil goes up, for example, he can reduce his stocks accordingly. Or if interest rates go up, to throw a lot of stocks into bonds. That's essentially all we were doing with data, with enabling technology to empower CIOs to make much better decisions.


Pierce: And so, if I could, the things that are our model methodology tend to rely more on valuation methods. We tend to have very yield focused analytics, so if the yield on bonds is higher than a certain amount of the dividend yield or earnings yield, or favor bonds over equity, we then use sentiment indicators like credit spreads, VIX spreads and things like that. And so are our data use tends to be valuation and sentiment driven as opposed to economic, but that doesn't mean that people can't do it, which is the real benefit of Arun's AlphaEngine. His thought process was that he could provide a menu of options for people. And whether or not you elect to to use it or not would be entirely up to you. But I think that was certainly our experience within the confines of a turnover limit. That's what we ended up with.


Muralidhar:  We went one step further, Christine, because we didn't want to have people dependent on us for data, as Don is finding out data vendors can be difficult to deal with. So we even allow clients to upload their own data into the system. Let's say your asset manager's willing to send you a certain data series that you want to make decisions with. The platform is meant to be data agnostic, so that if you wanted the data from us, we would collect it from you, but if you have the data from your own source, no point to paying for it twice. Because the data was just one part of the whole process, on how do we get you to better decisions, once the data is in the system and scrubbed and clean and ready for you.


Giordano: So many people are strapped, in which they cannot hire that extra analyst, access public pension funds, or don't have the permission to hire that full timer here and there. Does this actually take the place of an analyst? How does this work to kind of perhaps pay itself off?


Pierce: The value-add proposition here was entirely an improvement on the rebalancing methodology, as opposed to more of a cost savings for employees. We certainly made the argument that that range based rebalancing that we had in place, had a track record of reducing risk, but not really adding a lot of value. And then on any particular day, it might have added a particular amount of basis points to the plan, but if you were to fast forward six or seven months it might not have. As Arun made that particular point of saying, in a bear a secular bear market, taken in totality, range based rebalancing would be a risk reducer, but it didn't add value. One of the things that we had established is that in using market data to make better decisions, we hypothesized that we could use we could make 35 basis points on the total plan, and for every CIO out there listening, they're going to realize that is a staggering amount of added value. If you gave a money manager 350 basis points of outperformance, they would need to have a 10% allocation to get that sort of the same kind of buying value. And as Arun has shared and suggested, this is one of the areas that is just not fully engaged in. We spend a lot of time evaluating our asset allocation and managers. But we don't spend nearly as much time evaluating the shifts of the portfolio within the ranges that have been established. And so, for us, it was really a value-add argument, as opposed to just saving on some employee costs and things like that.


Muralidhar: I just came from the Netherlands where I met my first ever client, Roland van den Brink and Patrick Groenendijk from PME Pensioenfonds. Patrick was the equivalent of Don and Roland was the CIO. When I showed him the software, he told me, "Arun, don't fall in love with your software. For me, the most valuable thing, even if it doesn't make me a single euro, is you now told me I can delegate to my staff and the decisions will be transparent. This is exciting governance where you've pre-tested the idea before pulling the trigger, because that type of discipline and governance is worth a lot of money to me." I'll never forget that. In fact, it was one of my most enjoyable afternoons with him still; because he taught me how critical it was for a CIO to be able to trust a staff member, and how important transparency and this ability to say when he's pulling the trigger, there's a 50/50 chance that he'll get the direction correct. But if he's done all the hard work before, and the analysis is good, then he could defend that position very easily. And that was a very, very insightful comment.


Giordano: It comes from the top down as far as the thesis is plugged in. But it's malleable; you can shape it to your own thesis within the Investment Office. Are there certain theses that you've seen for this new economy that are impressive that you can share?


Pierce: We don't have an economic motive to change our model because I don't have a month-to-month track record that I have to necessarily sell. But I am incredibly encouraged by the robustness of the rules. The reason I say that is because it has served us in long stead and the most recent circumstances where we had a sudden shift from a very frothy market in 2021 with a recognition that the inflation was not transitory and that rates needed to rise to combat it. The value of using an overlay program to help you with rebalancing is that you're a lot closer to the market and I'm happy to go into that. As an information or knowledge management improvement, I cannot stress that just getting one step closer to the market makes you a much better investor, it really does. There are knock on effects to being a more engaged investor, on top of the 35 basis points of expected return, or in our case 100 basis points since inception of the program! We're going on 16 years, 17 years and what I'm proud of is a model that was developed out of research in the late 90s, early 2000s, has held its ground and has been very robust during our time period. Now, it has not worked in every period, but when it has, it pays off immensely.


Muralidhar: I would add that the central bank's intervening in markets has been a complete paradigm shift. It encouraged an enormous amount of risk-taking behavior, low interest rates, and basically bailing out the slightest stumble in markets. So, capturing sentiment became very critical and we still we run an investment business using exactly the same software license with very different objectives from what Don's doing. Using sentiment indicators becomes very critical because the market can shift on a dime. Because Don doesn't have to worry about managing assets for third parties, he has the ability to be more patient, which I'm very envious of. Because there are some rules that are so easy to look at, if you could just have the patience to wait it out. So for somebody who's faster moving, you have to worry about technical analysis and sentiment indicators. Somebody like Don, who can be a patient investor, can consider overweight equity positions for a number of years, and whether it is good or not.


Giordano: And of course, everyone is talking about inflation and a potential recession. With both of you actually having lived through the Great Recession financial crisis with this software in hand, how do you expect it? What would you have expected to do with it if we didn't hit a recession?


Pierce: Our model is based on relative value expectation. So in other words, we're not making forecasts as to whether or not stocks are going to go up eight or ten or twelve. What we're simply doing is saying stocks might be better than bonds, or in some cases, like now, we think bonds will be better than stocks. And that sort of differentiation of where you think those payoffs are is less a forecast and more an evaluation of the things that we've talked about, which is valuation and sentiment. This includes things like VIX levels, credit spreads, and things like that. And so we have a number of the rules that we've developed, based on the research from the 1990s and early 2000s. That gives us confidence that the portfolio on any particular month could be offside, like for the last six months, where it's been overweight bonds and underweight equity, which has not felt like a particularly smart thing. Watching the stock market rally in the face of interest rate hikes is not one of the most fun things to watch. But at the same time, we and the board have confidence that the program will work over time, and it does need that time. Earlier on, we probably had the same issues, whereas when we first started the program back in 2005, had we had a bigger misstep and we were running into issues, we probably might not have survived. But the longer you go, hopefully the more your clients will stay around.


Muralidhar: We had a really good 2008. We nailed it and I asked how we could make it even bigger.  We got lucky because we had David Deutsch at San Diego County who started as a software client (in 2006) just like Don. But the board was putting so much pressure on him saying you can't be this smart. He was making money, so he called us up one day and asked us to manage it because we knew how to do this and it was easier for him to hire us. He came to us with one (additional) request, which is don't just make the money, but make money when things go really bad because at that point, I've got no outlet. And so we ended up putting together these rules like Don was talking about that had a particular tendency to do well when the markets blew up. So we had a very good 2008, we had a very good 2018, which again was where people were struggling. And I hate to say this, but this smells and feels like 2008 on steroids. We've got a lot of allocation to illiquid assets and  Central Banks have huge balance sheets., Raising rates is only focused on inflation and nothing else. Silicon Valley Bank and Credit Suisse both folded. There's a lot of damage in commercial real estate. Not to sound depressing, but when this breaks, it's not going to be pretty because one of the things that happened was when it broke in 2008,  is the fact that people had so much allocated to illiquid back then - the Ivy League's particularly - the cash illiquidity forced them to sell more equities. And now every pension fund in the world, not just in the US, are so heavily allocated to illiquid assets, I think it's going to be catastrophic if you don't have a process in place to manage the risk of your illiquid assets. I think that will be the killer this time around.


Giordano: And we mentioned having rules in place that served in 2008. Can you share one or two of these rules?


Muralidhar: We have one rule, which is a very common one in the academic literature, called the Baltic Dry Index, which measures global economic activity. And when that slows, it's typically a bad sign for the economy, which is a bad sign for stocks. We also look at the ratio of price of oil or the price of copper, which industrial prices indicate. So, we see trends in those indicators within the economy. As Don says, we look at credit spreads and how different risky assets behave. And then investors start to shed risk that can move very quickly. So we want some slow indicators that take you into position and sometimes they're early. And you want some indicators that are fast, which are a little bit late. The goal is supposed to be never too early, nor too late. So in Don's case, six months is perfect. In the case of me, as an asset manager, six months is hell, so I've got to shrink that window down dramatically. Because as an asset manager, people won't be comfortable with the six months. That's why we've got slightly faster signals.


Giordano: You mentioned the Baltic Dry Index and the impact of these global indicators. What specifically does that mean for the economy?


Muralidhar: As the Baltic Dry Index is going down that means global economic activity is still declining. If the price of copper is going down that means the economic activity is declining. Typically, the price of oil is going down, that means oil demand is down. So they are very simple indicators and all of this is in the public domain. I don't think there's any idea and Don's model or in ours which doesn't have some economic or academic backing, too. Don, in your case, didn't your board require you to have academic backing?


Pierce: Yes, 100% of all our rules are backed by academic literature, because for some reason, they didn't want some 31 year old to make large decisions on the portfolio because he thought moving averages was a good idea. Most of the literature really focuses on the top and bottom deciles. So in effect, the top 10% of experience and the bottom 10% of experience have information and the rest of the 80% really doesn't have much information. That's our methodology. For all of our statistical friends out there who might say by definition you have the sample size because each day you get new samples, it's not as robust because the thing that might have been in the top 10% ten years ago is no longer there. That's why we don't use a 95% confidence interval or something like that because we don't have to prove beyond a reasonable doubt something works, we can simply use a preponderance of evidence. And so for us, it's the top 10% and bottom 10% of experience that is instructive. And then the rest of the 80% is largely not as actionable.


Giordano: What is the overlay process you use?


Pierce: When you use an overlay manager and using futures and derivatives to implement the rebalancing, you're now one step closer to the marketplace. I can safely say that as an improvement of your overall engagement with the market and your engagement with managers, is it is night and day. As a market participant, you are hearing and listening to what's going on in the marketplace much closer and getting more knowledge transfer. From my standpoint, it's been a very engaging part of the work and it's something we've embraced, frankly, because when we talk about allocating, we ask about the assets. You might have an awesome management team but fundamentally, what are the assets that they hold and how are they extracting the value, as opposed to being one step removed and simply relying on a more antiseptic view of the process. At the end of the day, it's the assets that are going to drive your performance. The management team is there to select them and manage the process but at the end of the day, it is the cash flows that you receive, or the sale of the asset that you've purchased, hopefully at a lower price, and you sell it at a higher price, or the income you've received from an asset and then finally paid back – that is what drives this. By getting that much closer to the market allows you to have that kind of level of conversation with the managers to understand what assets they are purchasing and how they're how they're trying to make money for you, as opposed to looking at your performance. It's just a night and day difference between being an investor in that way.


Muralidhar: May I just add, I got lucky that I started my career on a derivatives desk before I went to the pension fund. I then implemented a currency overlay then I went to another derivatives vendor. But because asset allocation was my passion, it was trivial for me to expect that futures. Forwards, and options and homers[AM2]  would be used by sophisticated institutions to manage their portfolios. And it's surprisingly underutilized. For example, even though people might use futures, there are many simple strategies that are just begging to be implemented by the use of options. Because of the lack of familiarity among boards and CIOs and even consultants about these strategies, I think the industry is underserved and it behooves some of these CIOs to get the familiarity that Don learned over time. Don, you did this on your own right? You didn't have somebody sit you down, you just basically learned by virtue of having to do it?


Pierce: Yes and the first thing that you will experience is that you engage in an overlay manager and they're using futures. They might come to you, or maybe the Street comes to you and effectively say, "We noticed that you have this allocation, maybe you would prefer a swap on it. So instead of now rolling your position every day, month, or quarter, you might put this on for a year. And we can give you a string of a fixed rate against that." So your roll yield might be volatile but I can lock in a rate right now on the S&P 500 and maybe you can get an S&P flat or perhaps minus five basis points. We use the Russell 2000 futures that way. There was an inherent short in the marketplace for a long time where you were able to accrue the Russell 2000 return and you would get a roll yield benefit of between 40 and 120 basis points. When we would look at small cap equity managers, it wasn't the small cap equity index that you had to beat. You had to beat the Russell 2000 index plus the roll yield that I would be getting, and we have zero of those. The first entree will be US swaps. So, you know, the decision that whether or not you want to continue to do roll yield is


I just want to say with all sincerity that Arun changed the trajectory of my career, to go from allocating to investing. And that's something I can never say thank you enough for, so thank you.


Giordano: Don, I'm sure other CIOs would love to learn exactly your philosophy, your movements, and how you're making it work?


Pierce: In talking about the progression from using futures and forwards to using swaps, there's a sort of put-call parity, so a future is equal to long a call and short a put. But once you introduced the concept that futures have an option equivalent, suddenly it becomes a very powerful tool to incorporate options because from time to time options also have interesting payoffs. The information growth and learning in adopting those tools just makes for a much more enriching career.


Giordano:  Where did it save you? Is there a case example in which you had if you hadn't, you would have suffered?


Pierce: One of the first large foot positions that we put on was selling puts during the European debt crisis. The insurance industry offers financial products that have embedded protections into them. So, if you buy a principally protected investment for some time period, three or five years or seven years, the insurance company goes out and buys that insurance. During 2010, the long dated implied volatility for the S&P 500, the Russell, 2000, the Nikkei 225, and Euro Stoxx 50 were all trading at staggering levels. We sold puts at a strike price where, if you included the premium that you received, you were buying the stock market below March 2009 levels. For us, that was a demarcation to say there's something broken that somebody's willing to pay insurance for something that would be worse than what then worse than what happened. That doesn't mean it couldn't have happened, but we were taking a calculated risk. As it turned out, that worked incredibly well. So well that that we got taken out of it within two years that we made close to 85% of the premium that we would have received. So, we closed that position out three years early and moved on. That was an upside participation. More recently, we purchased one year puts on the S&P 500 back in September of 2021 and also June of 2021. Having that kind of protection on into the teeth of 2022 was incredibly helpful. But you don't just jump into buying puts and calls and using some of these strategies without first getting the underlying knowledge from your derivatives experience, which starts with an overlay program.


Giordano: What should you be sure to have in place if you're going to go ahead to do this strategy?


Pierce: You could certainly implement an informed rebalancing on a physicals basis and that was what we did in the first year. I wouldn't recommend it, but to each their own.


Giordano: So, you wouldn't recommend an informed rebalance?


Pierce: I would recommend an informed rebalancing program, but a rebalancing method that uses futures is so much better. It's hard for me to imagine if we were simply stuck with physicals, but I suppose it's possible. And our estimate of 35 basis points on the forward estimate was implementation agnostic. But it's hard for me to imagine not having it because the tools are so valuable, and the learning has been so edifying and helpful as an investor that I can't, it's hard for me to extract that out, but I suppose it's possible.


Giordano: Last advice to other CIOs things that been working?


Pierce: When you have a model in place, it's not always going to be right and certainly the last six months has not been a lot of fun. The model has definitely preferred bonds over equity and that started increasing duration as rates went up. We started at the end of 2021 at a duration of less than a year; we were a quarter of the year, 0.25 that has now since moved up to about 3 or 3.5, so it's been steadily increasing duration. But watching the stock market rally in front of very aggressive rate hike sequences always challenges you.


Giordano: I think we all kind of feel your pain a little bit.


Pierce: Thank you. I appreciate the solidarity.


Giordano: I want to thank you so much for your time for your words of wisdom here and really drilling into something that's been working for you and helpful for other people to hear about.