Episode Transcript
[00:00:00] Speaker A: Generating an uncorrelated alpha stream and bolstering the overall portfolio's resilience by using this program as a tail risk mitigation tool that shields the portfolio during periods of extreme market stress. They're tired of constant bleeding from the option premiums. You know, from my experience running this strategy, I kind of believe that the more efficient and cheaper hedging strategy is this beta adjusting through an overlay portfolio instead of paying insurance that might expire worthless.
[00:00:40] Speaker B: Welcome to the Institutional Edge, a weekly podcast in partnership with Pensions Investments. I'm your host, Angelo Calvello. In each 30 minute episode, I interview asset owners, the investment professionals deploying capital who share insights on carefully curated topics. Occasionally we feature brilliant minds from outside side of our industry, driving the conversation forward. No fluff, no vendor pitches, no disguise marketing. Our goal is to challenge conventional thinking, elevate the conversation and help you make smarter investment decisions. But always with a little edginess along the way.
Welcome to the Institutional Edge. I'm Angelo Calvello, your host. Well, here's a problem every asset owner knows, but rarely talks about openly. You have a long term allocation target and your portfolio almost never actually sits at that target. Markets move, valuations shift, suddenly you're out of position.
Global tactical asset allocation is a strategy that lets you manage that gap systematically and at scale. GTAA's intellectual foundation goes back to Brinson et al. That is their 1986 study showing that 90% of a portfolio's return variation comes from asset allocation decisions, not stock selection. The practical tools came a few years later when Rob Arnott and Roger Clark at First Quadrant started using stock index, futures and other derivatives to execute these tactical shifts across global markets. That was the blueprint. What's interesting is that today many asset owners are building this capability in house rather than outsourcing it. And my guest today has done exactly that and has done it well over the years. My guest is Joy Xu. She's a senior Investment Director at Verizon Investment Management Corporation, where among her many other responsibilities, Joy has led a tactical overlay portfolio that generated over 1.2 billion in profit during her tenure. Tenure. I've asked Joy to share her experience building and managing this program so others could learn and benefit from it. Joy, welcome to the show.
[00:02:51] Speaker A: Thank you, Angelo, for having me.
[00:02:54] Speaker B: My pleasure. I'm going to start with a few warm up questions like I do with every guest. You ready?
[00:03:00] Speaker A: Yeah.
[00:03:00] Speaker B: Okay. City or nature?
[00:03:03] Speaker A: Nature.
[00:03:04] Speaker B: Early bird or night owl?
[00:03:06] Speaker A: Night owl.
[00:03:07] Speaker B: See, I told you these would be easy.
Books or movies.
[00:03:11] Speaker A: Books.
[00:03:12] Speaker B: Me too. And finally, do you like to stick to a plan or are you always open to a pivot?
[00:03:18] Speaker A: Open to a pivot.
[00:03:20] Speaker B: There you go. Now that we're warmed up, Joe, we could jump in and I really want to spend a little time at the beginning learning about the genesis of Verizon's GTAA program. I mean, if you could take a few minutes and tell us, where did the strategy come from, what was the genesis and the problem it was designed to solve. So, you know, give us some context before we get into the strategy itself.
[00:03:46] Speaker A: This strategy was born out of a practical challenge around ongoing cash raising, rebalancing and liquidity management.
Rather than mechanically reverting to strategic asset allocation or saa, we began asking for whether we could position the portfolio more effectively given prevailing market conditions.
So this led to the development of a GTAA based dynamic overlay program, a disciplined, macro aware and relative value driven approach designed to complement the strategic asset allocation or saa.
As we know that SAA is traditionally built on the long term capital market assumptions that assume the markets are in equilibrium. Our core philosophy is that while SAA serves as a long term anchor, markets frequently deviate from equilibrium, often in meaningful and persistent ways. These dislocations create opportunities to add value and manage risk through thoughtful tactical positioning. In the high side, our program has delivered meaningful, largely uncorrelated alpha relative to traditional assets over the past decade. More importantly, it has enhanced the stability and robustness of the overall portfolio, particularly during periods of market stress. During those environments, the overlay has at times generated positive returns which can be used to fund our pension benefit payments, reducing the need to sell temporarily dislocated assets at distressed level. As a result, the program not only contribute to return generation, but also play a critical role in preserving core asset exposures. This enables a full participation in subsequent market recoveries.
[00:06:04] Speaker B: So Dwight, GTAA has been around since I think Brinson et al. Gave us the paper in 1986. And then in my world it was really Rob Arnott and Roger Clark at first quadrant back in 19.
Shortly thereafter in the late 80s because you had derivatives now of listed derivatives.
The key for me was the transition from a discretionary approach to a systematic approach.
So can you tell me a little bit about your approach, your investment process and how you make these tactical decisions? Discretionary, systematic. Just let me know how you do it.
[00:06:44] Speaker A: Sure, yeah, those are a lot of questions. Let's tackle each of them.
First. Investment Process the GDA based overlay program utilize a systematic data driven framework that combines quantitative model driven process with qualitative oversight. In addition, by implementing tactical tilts through liquid derivatives, we ensure high capital efficiency and the ability to adjust global market exposures in a timely manner. Now let's dive in the four theme systematic factor models. The models were originally developed in collaboration with our team and external asset managers. The models generate overweight or underweight trade signals based on four distinct valuation and fundamental. We monitor historical valuation and relative values against cash and other assets in the portfolio. Second, technical factors focus on market momentum and volatility regimes. Third, sentiment factors measures market participant sentiments alongside real time economic activities. Finally, liquidity and risk analyze term structures and liquidity premium.
In addition, we have a continuous feedback loop that provide timely evaluation of factor performance and their evolving correlations. These allow us for constant model refinement.
[00:08:30] Speaker B: So, just a question you mentioned at the beginning, the systematic four theme model was developed through internal research. But also it sounded like discussions or ideas from some external sources.
Would these have been, for example, managers that you may have a strategic partnership with or some commercial relationship with?
[00:08:53] Speaker A: Yes.
[00:08:55] Speaker B: So you drew ideas from outside the organization and you made them your own, right? You put them in the context of solving this dynamic asset allocation problem. Okay, so we've got a systematic four theme model.
You mentioned sentiment. I'm going to go and ask are you using any alternative data in this process at this point in time? Like something pulled from social media would be perhaps a source of sentiment.
[00:09:23] Speaker A: We constantly looking at those as a data source. However, not implementing using those data sources because they are very expensive for us to deploy those data. But constantly watch out for the data from that source as well.
[00:09:44] Speaker B: And I think you mentioned your first theme. Looking at my notes, you talked about valuation in kind of a fundamental way, am I right? And I think you talked about, I mean, at least this is what I understood. There's a relative value issue here.
So can you give me an example like how you might think about S and P versus ifa? I mean there must be a scoring system you have or something that allows you to decide. But I'm guessing here, Joy, I have no idea. So go ahead.
[00:10:12] Speaker A: Yeah, sure. Relative value is a very important theme throughout our quantitative model. So. So if you look at the US versus IFA and look at their PE ratios and you can notice that US equity market has always been trading at a premium versus IFA market that because of a lot of reasons and we don't delve into that now. And what we do is that we kind of looking at the historical prediction premium of US vs IFA and then look at it and standardize it using statistics. Basically we take a Z score and it's telling us, yes, we know it always has a premium of US equity, but during the past history where we are in how much premium we have currently compared in the historical norm.
[00:11:12] Speaker B: And can I ask a little bit about how you measure sentiment? Since we're not talking about social media tweets, you're doing something else. I'm going to guess it's some kind of consumer sentiment. Maybe, but without sharing too much of the secret sauce, I mean, how do you, I mean what are the inputs and how do you use those inputs to make a decision?
[00:11:34] Speaker A: Yeah, normally we normalize those data in a statistical way.
So we look at the business sentiment, the capex spending and we look at the consumer spending, the consumer sentiment, and also a lot of economic activities.
[00:11:55] Speaker B: So it sounds like you're pulling from a lot of data sources. You've got four themes. Everything is systematic. I mean, this is not a discretionary program.
It's clearly quantitative in its nature.
I'm assuming there are risk targets that are part of this model, am I right?
[00:12:15] Speaker A: Yeah, yeah. So you are talking about risk management.
The risk management is through two primary constraints to ensure a prudent exposure. First, we have a static, over or underweight limit. This defender maximum allowable tactical tilt for each asset class.
Second, we have a dynamic risk limit. This dictates the relative risk of the GTA overlay program as a percentage of the overall portfolio risk. So together these measures ensure we maintain an appropriate risk level across different market regimes.
[00:12:59] Speaker B: Okay, and what's the cadence with which you run the model?
[00:13:04] Speaker A: The frequency we keep a bi weekly rebalancing and treating cadence prevents overreaction to short term market noise and allow us for gradual, disciplined portfolio fine tuning. However, our process remain agile enough to respond to black swan events, we can reposition the portfolio very quickly if needed.
[00:13:31] Speaker B: Let's talk about that Joy, about the repositioning because it takes us into something that was close to my heart and that is implementation.
So you've got your model a biweekly cadence.
You could do what you want, but it's. On a biweekly cadence you've got your output, now you gotta implement these decisions. Maybe this takes us into something you and I had talked about, the flexibility that's offered as well. But let's start at a high level. How do you implement these decisions? I mean, what are the asset classes that you're balancing and how do you implement them?
[00:14:07] Speaker A: Yeah, the asset Classes are the main asset classes in our portfolio. They are US Large cap, small cap, ifa, emerging market and high yield, emerging market, debt, credit and Treasuries.
[00:14:23] Speaker B: So how do you get exposure to these? Because you need flexibility if you're going to be biweekly.
[00:14:28] Speaker A: We get an exposure from using derivatives. Before we talk about the implementation, I would like to also mention that we have a qualitative assessment in place as well. These act as a critical risk assessment tool.
By looking through a macro lens and capital market cycles, we determine the appropriate magnitude of risk taking. These assessments allow us to scale overall risk up or down without altering the directional and the relative trades dictated by the model and then the implementation processes that.
We have all the risk management parameters established in the policy statements and we have an EC executive committee and their main role is to approve our strategic and tactical asset allocations accordingly. Under this process, I propose tactical positions based on the quantitative models and qualitative changes judgment. The EC then assess the market condition and treat the sizes before granting a final approval.
And operationally, we manage our derivatives by monitoring collateral levels daily. And as you mentioned, Angelo, how to implement it? We opted for derivative implementation because they offer greater capital efficiency and lower trading costs. This approach also allows for timely execution without causing any disruptions for our external asset managers. Finally, you know, we decided to outsource our execution because we have a smaller internal team. The outsource seems to be a simple solution and ensure everything runs smoothly.
[00:16:37] Speaker B: 30 episodes in the Proof of Concept has been established. Guests like Don Pierce, Venus Phillips, Andy Green, John Grable, Paul Greff, Brad Brindle, Mark Speed, Leo Svoda, Joe Eppers, Renee Diresta and Greg Brown. Hey, they don't come on a podcast that hasn't earned their time.
So we're ready to take the next step. We're now accepting sponsors for the institutional edge. And before you ask, here's how it works. Sponsors support the show, but don't get editorial control over the content. And no, you will not be hearing me reading ads on energy drinks and tax preparation services. This is still the same show. If sponsorship interests you, contact Julie Parton at Pensions and Investments. She can tell you about all the benefits. Her contact details are in the show notes. Let me go back just a bit. We talked about a qualitative assessment.
Is that done by you Call it the ec. Is that like an executive committee? Is that what it stands for? Ec?
[00:17:43] Speaker A: Yes.
[00:17:44] Speaker B: Is that something that's done at the committee level? Is that something that's done just by you alone? I mean, how is that qualitative layer? I mean, what is that qualitative layer?
[00:17:55] Speaker A: Well, that's a good question, Angelo. The qualitative level is also pretty structured. So we look at three different areas. One is macroeconomic conditions. As we know that business cycle comes with four phases.
We have a number of macro indicators can show where we are in the market cycle. And then we also have a framework of looking at a capital market cycle using capital market indicators. As we know that a lot of times the economic in economic cycle are not coincide with the capital market cycle. So we want to take both into consideration and this help us to assess the general risk level of where we are comfortable of taking. And finally we also look at the central banks and the monetary policy and fiscal policy and see if there is any impact from that area as well. So distill this all together, we have a discussion at the EC level and then we decide the risk level that we are comfortable of taking. And then with the overall risk level decided, we can implement the individual trade using the model. The quantitative model signals.
[00:19:30] Speaker B: I want to understand just a bit more here, Joy.
You've built this GTAA model, you took some ideas from some of your strategic partners, you've done it internally, you've built this model.
When did you implement it? I didn't ask you that question at the beginning, that's my fault. When did you implement this model?
[00:19:52] Speaker A: We implemented using derivatives since 2014.
And before 2014 we also implemented by using physical assets. Basically we either give cash to physical asset managers or taking cash using redemptions out of their strategies. And we had a good experience before and post 2014, we want to do it more on a timely manner. That's why we opted to use derivatives to transact more frequent and if we wanted to and with little market frictions and it's easier to implement it on a timely manner.
[00:20:41] Speaker B: So 2014 and beyond you've been using derivatives, you still have the same underlying systematic model, four themes with a qualitative, I'll say overlay, but it doesn't sound that qualitative to me where you're just saying, yeah, I don't know, maybe today we're going to see what happens. I mean it sounds like it's very structured, I think was the term you use. So if you got this a quantitative model, a qualitative overlay that's very structured you're implementing with derivatives.
The decision making process is how would I say? It has a qualitative level to it, it seems to me, because unless I'm missing Something it's not like the model just runs on its own and every two weeks it makes a change.
You're sitting there at the helm of the model, looking at outputs. You work with a few, I assume, a handful of your colleagues and you decide maybe kind of like a governance structure, executive committee, this is what we're going to do. Now, is that fair? I mean, is that how it all kind of funnels down then to that decision?
[00:21:50] Speaker A: Yes, you're correct, Angela.
[00:21:53] Speaker B: Very rarely, Joy, am I ever that correct. But I get it.
So are there instances where you've decided to override the model?
[00:22:03] Speaker A: Yeah, that's a very good question, Angelo. Yes, we do, occasionally. I'll give you an example that following a sudden geopolitical shift on February 28th with Epic Fury, our team overrode a risk on model signals generated right before the military action. After two emergency discussions in the first week of March, we elected to bypass the model recommendations and instead trimmed the previously risk compositions back to strategic weights to protect the portfolio during uncertainty. Later on, we reposition the portfolio to slightly risk off.
[00:22:46] Speaker B: Pretty good timing.
[00:22:49] Speaker A: That is an example that shows that we can use this tool to reposition our portfolio in a timely manner and also in a low cost way of doing so.
[00:23:04] Speaker B: You've given me one example of an override and a flexibility.
But let me step back. That's one example. How do you measure success?
You've been doing this now for almost 12 years with derivatives, and I know longer before that. I'm assuming this program would have been dead if it didn't work. No disrespect, am I right? If it wasn't working, but it seems like it's working. So how do you measure success?
[00:23:32] Speaker A: Yeah. Well, we measure success through two primary lenses, generating an uncorrelated alpha stream and bolstering the overall portfolio's resilience. By using this program as a tail risk mitigation tool that shields the portfolio during periods of extreme market stress. Well, since it's inception in 2014, the overlay program has been highly effective and delivers strong risk adjusted returns generating over 1.2 billion in profit with a Sharpe ratio of 1.4. The annual risk of the program is below 50 basis points.
But I will highlight that the program's value is more most pronounced during market downturns. For instance, at the height of the 2020 COVID 19 pandemic, the program generated almost 400 million in cash.
This provide critical stabilization and liquidity when the broader market were most volatile.
[00:24:46] Speaker B: So it's working. That's the Good news both in terms of risk return, but also in terms of helping you fund the liabilities at the end of the day, correct?
[00:24:58] Speaker A: Yes.
[00:24:59] Speaker B: Okay. What haven't I asked you about the program that I should ask you about your GTAA program here. Just summarizing. We talked about why it was started and the benefits it sought. We talked about your model quantitative with a qualitative structure. We talked about the decision making process, the implementation process using derivatives. And now the proof is in the pudding because you're showing me it's actually contributed significantly to the success of the management of the portfolio both in terms of risk and return and liquidity. Right?
[00:25:36] Speaker A: Yes.
[00:25:37] Speaker B: What am I missing, Joy? I mean, you're the one that built this, you're the one that runs this program and I know you didn't do it alone. You've got a group of people, but you're the only one on the camera today. So what am I missing?
[00:25:49] Speaker A: I don't think you're missing anything.
[00:25:51] Speaker B: Yes.
[00:25:55] Speaker A: What I want to emphasize is that, you know, I don't. I haven't seen a lot of people using these tools. And from my experience of running this strategy, I would say that it can be utilized widely among asset owners and it can be utilizing as a self funding tail risk hedging tool.
Because we've all seen that the traditional tail risk hedging is notoriously expensive to carry. Many funds end up abandoning their protection right before a crash because they're tired of constant bleeding from the auction premiums.
You know, from my experience running this strategy, I kind of believe that the more efficient and the cheaper hedging strategy is this beta adjusting through an overlay portfolio. Instead of paying insurance that might expire worthless, we can use this strategy to dynamically tilt our exposure and change our beta. This allow us to mitigate downside risk without heavy carry cost, making the whole portfolio far more resilient, especially during market downturns. And another benefit of this is that it solves the correlation convergence problem.
As we know that in the crisis, the safe and uncorrelated assets in the traditional portfolio always become highly correlated.
[00:27:38] Speaker B: Yeah.
[00:27:39] Speaker A: Yeah. The benefit of diversification disappears exactly when you need it most. And we saw this happen in a number of episodes such as in the 2008 and 2020 crashes. The liquidity dries up and everything sells off together. I think a well structured GTA overlay program provide an independent and uncorrelated source of alpha. By having a low correlation with traditional benchmarks, often zero or even negative correlation during stress, the program Ensures that when the rest of the portfolio is under pressure, the overlay is functioning as a true diversifier. And furthermore, this generates a dry powder effect.
Cash gains provide the liquidity you would need to rebalance into undervalued assets or meet spending obligations.
[00:28:42] Speaker B: Is this what you mean by self funding? I'm sorry to interrupt, but at the top of my mind, is this what you mean by it's a self funding mechanism?
[00:28:49] Speaker A: Yes.
[00:28:49] Speaker B: There you go. Okay. Sorry to interrupt, but yes, that's what I wanted to get to. That's pretty cool.
[00:28:54] Speaker A: Yeah. Thank you.
[00:28:56] Speaker B: You're welcome.
[00:28:58] Speaker A: So all of this you can do without being forced to sell distressed assets at the bottom of the market.
[00:29:06] Speaker B: I should ask you about tail risk hedging and I see how it's an alternative because you're not just buying puts at the end of the day and watching the premium get sucked out of them. I get that now. We've gone through how you got here. It's an interesting story and it's even more interesting because it worked and it's still working as recently as the end of the month. And I know that's only one data point, but you're able to show 11, 12 years of, of. Of auditable performance.
So we've got some of your peers sitting there thinking, boy, I'd really like to think about this for myself because this is an internally managed strategy. I know you outsourced the execution, but you know, the real value added comes from the quantitative approach, the qualitative structure, the decision making, the governance. Yes, implementation is important, but your concern is just getting the decision right and outsourcing a decision.
So this is something your peers could certainly run.
But I assume, what would you tell them? What are the biggest challenges that you've seen in trying to get this up and going and running it over time?
[00:30:19] Speaker A: I think the way we structured is that we want the process to be a systematic and a structured way so that the rebalancing can be more disciplined. And my recommendation is that starting from the basic, building individual factor models and testing out to see if it is work over time and also to see if it is work if you have a delay in execution and make sure the robustness of the model itself and then building up multiple models and building multiple models with low correlation so that they can offer diversification benefits over time. The second recommendation is to have a prudent and sound process.
You set up enough collateral to withstand the market movements and has a sound investment process. Make sure that the investments are well thought and the risk are well under managed as well as the implementation was taking place in a timely manner.
[00:31:39] Speaker B: Joy, it's interesting you mentioned that and you mentioned it earlier and I forgot to come back to this. And that is there's an operational piece that's important to this.
So if you're a CIO or a senior investment professional and you're thinking about introducing the strategy, you're right. You need to have good process, good implementation. But you gotta have the ops piece too, because there are moving pieces and you do have cash flows. You could have margin calls in your futures or collateral calls in your swaps. Right? So you got to get the ops right to make this thing work. Because you could have the best ideas and your managers or your external manager could implement well.
But boy, there's a lot of pain collect going on underneath, I would think.
[00:32:24] Speaker A: Right, Angelo? That's a good question. That's why I think risk management is really important.
That's the reason we have two primary risk levels to manage this program.
First, you have a static asset class level risk limit and then you have a dynamic risk level that maintain the proportion of this portfolio risk as a percentage of the overall portfolio risks.
[00:32:52] Speaker B: Got it? Okay. The only thing that's still in the back of my mind, Joy, and I'm a believer in this, believe me, whether you demonstrate it to me empirically, I actually had a chance to work with Roger Clark and Harinda Silva back in the 90s and I saw how these guys were building and deploying models. I'm a believer. But the one thing I always heard was you think you're timing the market with this. Have you ever heard that where people say you're trying to time the market? What's your response to that? Because that may be what a CIO wishing to implement something like this might hear from his or her board. What's your response?
[00:33:29] Speaker A: Well, people are always using timing the market this term to include everything like this. However, if you look at the asset managers, they are doing the timing as well because.
Because they're looking at the relative value all the time. No matter is multi sector portfolio. They're looking at different sectors. The valuation one versus the other, the other or the valuations of one security versus the other. So we are doing this all the time. And I don't know why this market timing get so much bad name on it.
[00:34:11] Speaker B: Okay, well done, Joy. This has been wonderful. I do have one question I ask all my guests and that is what's the worst pitch you ever heard? You've been in the business a long time. I'm sure people have pitched you strategies and joy. I'll say. As an aside, were you prepared to answer this? Because if not, we don't have to do it.
[00:34:34] Speaker A: The worst pitch is one that I heard during the market when the oil price was $150.
People came to me and said, the oil will shoot over $200. Make some investment on the energy side. And I didn't buy that.
[00:34:54] Speaker B: There you go.
I mean, you may have that pitch again soon given oil prices. So you better go back and dust off your notebook.
[00:35:03] Speaker A: We'll see.
[00:35:04] Speaker B: Well, Troy, this is great. I appreciate you sharing the process and the performance and the advice. So thank you very much.
[00:35:13] Speaker A: Thank you, Angelo for giving me this opportunity. Really appreciate it.
[00:35:18] Speaker B: Thanks for listening. Be sure to visit P and I's website for outstanding content and to hear previous episodes of the show. You can also find us on p and I's YouTube channel. Links are in the show Notes. If you have any questions or comments on the episode, or have suggestions for future topics and guests, we'd love to hear from you. My contact information is also in the show notes, and if you haven't already done so, we'd really appreciate an honest review on itunes. These reviews help us make sure we're delivering the content you need to be successful. To hear more insightful interviews with allocators, be sure to subscribe to the show on the podcast app of your choice. Finally, a special thanks to the Northrup family for providing us with music from the Super Trio. We'll see you next time. Namaste.
[00:36:11] Speaker A: The information presented in this podcast is for educational and informational purposes only. The host, guest and their affiliated organizations are not providing investment, legal, tax or financial advice. All opinions expressed by the host and guest are solely their own and should not be construed as investment recommendations or advice. Investment strategies discussed may not be suitable for all investors as individual circumstances vary.
Sam.
