In this episode I speak with Roxton McNeal, Head of Multi Asset Investment Strategy & Allocation at the UPS Investment Trust.
Before landing at UPS, Roxton’s career took him through the world of CTAs, developing hedge models for bonny light oil, and working in asset/liability management at General Motors. Each of these roles likely deserves its own podcast, but I do my best to pull a nugget of wisdom from each experience.
Where we spend the bulk of the conversation is in Roxton’s current role at at the UPS Investment Trust. We touch on many of he hot-button issues among institutional allocators, including the role of glide paths, private investing, tactical asset allocation, and tail risk hedging. I think what makes this conversation particularly interesting is how the constraints and realities of liability-driven investing shapes Roxton’s views in these areas.
Please enjoy my conversation with Roxton McNeal.
Transcript
Corey Hoffstein 00:00
321 Let’s do it. Hello and welcome everyone. I’m Corey Hoffstein. And this is flirting with models the podcast that pulls back the curtain to discover the human factor behind the quantitative strategy.
Narrator 00:18
Corey Hoffstein Is the co founder and chief investment officer of new found research due to industry regulations he will not discuss any of new found researches funds on this podcast all opinions expressed by podcast participants are solely their own opinion and do not reflect the opinion of new found research. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of newfound research may maintain positions and securities discussed in this podcast for more information is it think newfound.com.
Corey Hoffstein 00:49
This season is sponsored by simplify ETFs simplify seeks to help you modernize your portfolio with its innovative set of options based strategies. Full disclosure. Prior to simplify sponsoring the season, we had incorporated some of simplifies ETFs into our ETF model mandates here at New Found. If you’re interested in reading a brief case study about why and how. Visit simplified.us/flirting with models and stick around after the episode for an ongoing conversation about markets and convexity with the convexity Maven himself simplifies own Harley Bassman. In this episode I speak with Roxton McNeil, head of multi asset investment strategy and allocation at the UPS Investment Trust. Before landing at UPS. Roxanne’s career took him through the world of CTAs, developing hedge models for Bonny light oil, and working in asset liability management at General Motors. Each of these roles likely deserves its own podcast, but I do my best to pull a nugget of wisdom from each experience. Where we spend the bulk of the conversation is in Roxanne’s current role at the UPS Investment Trust. We touch on many hot button issues among institutional allocators, including the role of glide paths, private investing, tactical asset allocation, and tail risk hedging. I think what makes this conversation particularly interesting is how the constraints and realities of liability driven investing shapes Roxanne’s views in these areas, please enjoy my conversation with Roxton McNeil Roxanne MacNeil. Welcome to the program. excited to have you here. We have so much to unpack in this episode. I mean, I think in chatting with you originally on it, when we had our first conversation, I will admit I was overwhelmed by your background, we started going into all the different pieces of your career. And I said, There’s no way we can make a podcast out of this. But we’re going to try, we’re going to do our best and I have faith that you’re going to be able to break it down for us. So excited to have you here. Thank you for joining.
Roxton McNeal 02:51
Well, Cory, thank you for having me. It’s certainly an honor and a privilege to be participating on your podcast as well. So thank you for having me. And hopefully, we’ll have a good show here.
Corey Hoffstein 03:00
Privilege is all mine. So let’s start with a little bit of your background. Because as I mean, again, it is a incredibly diverse and successful career. And again, I think we could probably spend an episode on each chapter of it. So before we dive in, and we are going to be sort of hitting different episodes of your career, can you give us a broad overview of what the career arc has looked like?
Roxton McNeal 03:22
Sure, absolutely. Coming out of college. My first job opportunity was on the agency trading desk of a actually it was a Canadian bank that had just gotten its section 20 power. So they’re trading us fixed income. So I got an opportunity to trade US agency debt on their desk as a trader. And at that point, I don’t know if you recall, I’m a little older than you, Cory. But the FICO market was going on this look just went under the savings and loans crisis that was the insurance arm of the savings and loan, and there was no funding for him. So essentially, the US government had to create these bonds is a stopgap measure. They’re called FICO bonds. But the banks didn’t want to fund it up and the government didn’t want to fund it up. So there was this battle going on. So the spread on these was through the roof. Although it was backed by what I thought was the US government they were it was still it was issued by them. So why wouldn’t they have the full faith and credit of the US government? So I just thought the spreads were ridiculous. And part of it was probably naivety now that I look back, but I said these things gotta come in. So I was taking as many bikers as I could get on the desk, I was stripping them to create zeros out of them, and then selling them to my sales traders to sell to their clients. Like we’re just pushing those and we had a really, really good run of it. I think it made a lot of sense. They are trading three 400 pips above the comparable Treasury out I think they went up to 2019 was the last maturity. And then I would take the principal and I’d have the hedge trade on the back end of it. And that’s pretty much all I did for two years was just that worked out really well. The government finally came in and said yeah, we’ll fund it up and the spread compressed. We made money on the back end, as well as those and right when that was winding up I got a call from a headhunter who was looking to place someone at a hedge fund. And I’ve heard a little bit about hedge funds, but not a lot. It’s not something that they really taught us in school back in the late 80s, early 90s. So, you know, I entertained it and eventually took a position at a commodity trading advisor, where I was responsible for creating a macro trading strategy. Reason I took it as I had a lot of leeway in how I wanted to create it. There was things changing in the hedge fund industry, especially the CTA industry at that time, and the, you know, late 70s 80s, maybe even early 90s. It was the Wild West. I mean, you’d have, you know, like John Henry, he had his hedge fund that could make 50% return in one year, but it could lose 60%, the following year, and when I came in, investors are really, they think that they can handle that kind of volatility until that volatility hits them. So when I created my strategies, which I know, we’ll get into more detail later, it was really to kind of minimize that equity curve. I didn’t want to incur those bigger draw downs. And what could I do to create a strategy that people could stomach a little better than the kind of the Wild West days that everyone was trying to stay away from? So I was in that space at a couple firms for about 15, almost 18 years. And, as always, you know, built out some pretty good programs developed as I had the experience in them. And at that point, you know, the last firm ran it’s, it’s called net risk, right? Where you have a firm that set up across different functions, you may have an equity group, a CTA, group, and etc. and hedge fund gets paid on a watermark. So there’s something called net risk. So even if you’re performing very well, if the firm doesn’t make money, you don’t get paid. So those issues were going on. And I was kind of a little, I guess, dismayed. At the industry. At that point, I said, What else am I going to do. And at that point, I started at a young family as well. And maybe it was looking for, you know, I don’t know what was really going on in my head, but maybe looking for something a little dip in terms of quality of life, and had an opportunity with a colleague that I went to school with who was actually got involved with a group that was trading physical oil out of Nigeria, it was Bonny light oil, and they needed someone to come in who had commodity experience. And that’s where my CTA background came in, to hedge a physical allocation that they’re getting from Nigeria, as they were shipping it to the refineries in the Caribbean, where they refined the crude oil, and they were getting this allocation for the government, but they wanted to hedge the price. As you know, it takes up to three months in these you they call them ulcc or big tankers to get from Nigeria to the Caribbean. And obviously, prices fluctuate. So my job was to head that go into the Caribbean, that lasted about two or three years, the allocation was gotten from the current president of Nigeria at the time, and he did not get reelected. So that allocation was going to dry up. So again, before to see what I wanted to do. I was at that point reconsidered, going back into the CTA, the hedge fund world, but got an interesting call from someone who was they were looking to beef up their quantitative desk at a big pension. And it was the General Motors, asset management to the pension that defined benefit plans for General Motors. So I went and had a couple interviews there. And it really sparked my interest in how I could apply, you know, my past experiences to call it the pension dilemma at the time. So I worked there for about eight years, which was really interesting, you know, we went through a lot, obviously 2008 happened. At that time, GM, who was the sponsor of the pension went in chapter 11. So there’s a lot of things they’re coming out of that a lot of LDI, a lot of the big de risking the largest de risking of a pension in history was done during my tenure there as well. So I got to learn a lot of things about a lot of different things. And then most recently, I had an opportunity to kind of jump over to a UPS which had similar characteristics, I’d say to GM when I was there in oh eight and they wanted someone to come in and kind of guide that ship as they are going from an asset only perspective to more of a asset liability management or as they say ALM perspective in kind of drive that initiative at UPS, and I’ve been there running the strategy and allocation group at UPS for about five years now.
Corey Hoffstein 09:04
Yeah, so as I said, Dear listeners, we’ve got a lot to unpack here. So let’s get going rocks. And I want to get started with your experience in the CTA industry. And in a prior conversation. You described your job to me as quote, creating convexity in the Delta space. Can you explain to me what you mean by that? And what you think distinguishes someone who’s effective at it versus someone who isn’t? Yeah,
Roxton McNeal 09:30
absolutely. I mean, this was my thesis coming in before. Like I said, I had a lot of leeway to start my own program and what that looked like. And the first thing I do is I always try to set a very high level framework before I do anything that kind of level sets everything. So I have always been even in bath I always like to understand what the high level was before you start digging into the details. Things just make a lot more sense. And in doing a lot of the research and the analysis, I realized that straddles and trench systems are really two sides of the same coin. According to derivatives theory. You can replicate a trend following strategy with straddles, right? If you look at the Delta, even the standard Black Scholes is going to give you a delta for a straddle, which ultimately ends up being like the error function. But locally around the error function, you can linearly, it’s in linear, you know, approximations very good around that the money. So I said, Let’s do that if I can go in and intraday, I’m actually looking every hour or every two hours, I get a good locality approximation. So my linear approximation to the error function should be good. And I can just replicate a trend following strategy doing that. And the reason I chose intraday is because my holding period I wanted very short, because what I had for mentioned, I didn’t like the big drawdown, and there’s a lot of risk that happens overnight. And if I could control it kind of intraday, the way I was looking at it, then I could earn a substantial, you know, I could still get a return, but really lower the risk. And that’s pretty much what I meant by that, you know, if you look at CTAs, they’re charged with creating convexity using Delta products. And because of my personality, the way that I traded, maybe my risk level my risk aversion, that’s what came out of the strategy for me.
Corey Hoffstein 11:06
There’s a lot of people who have tried to run CTA products with varying degrees of success. Where do you think your particular edge came from? Yeah,
Roxton McNeal 11:16
that’s a good question. And that took a while to come by. At first, when you obviously, you’re charged with a program like that. All my effort was on the signals, I wanted to be right. And when do I go long? And when do I go short. And there was a lot of effort made on that. And as you know, the market is only a very small part of that is mean reverting and the rest is all noise, right? And is a huge signal to noise ratio, or noise to signal ratio in the markets. So after realizing that my signals weren’t going to get me there, I really concentrated more on the money management in the position management aspect of it. And that’s what really drives convexity anyway, because if I just kept the same level of position on I just have a delta security, you’re not going to create convexity unless you know when to add on and when to take off positions. That’s what creates the convexity. So I really started developing systems that allowed me to have better position and management program overlaid on top of the signals. And in the end was, the strategy was, you know, only in signal related effectiveness, it was only right 40% of the time, 45% of the time, but my gains or losses were 10 to one. So my net profit ratios were very, very high. And that’s, again, I attribute that to, you know, part of my risk aversion. What I was looking at maybe my objective function was to minimize equity drawdown. So I want to take that risk, but it all kind of fit together in a program that, that I was comfortable with trading. And the end resulted in, you know, some pretty good performance.
Corey Hoffstein 12:42
And I’m just curious, were you carrying risk overnight? Or was everything closed by end of day?
Roxton McNeal 12:47
Yes. So I would close out some of the positions over a day, but it really depended on the intraday versus the daily variance. So I was looking at early chunks. And if I had a trend that was developing, then I would hold the position overnight. If it was mean reversion, I would close it out. And I wouldn’t close out all of it either. I mean, to be very honest with you, most of the return that you make in these trades are actually overnight. So you can trade all day as you want. But I think just having that risk parameter to know where things are going. It did actually tend to carry overnight, and I’ll do it again, with liquid markets from around the world. I think a very successful strategy could be made that actually traded 24/7 and actually skipped from market to market. But that’s probably another conversation.
Corey Hoffstein 13:33
Well, speaking of Market to Market, let’s jump forward a little bit and talk about hedging Bonny light, which, candidly, was not even something I was aware of, I think until we first spoke, and so I’m curious first order approximation, why can’t you use just something like a very liquid proxy like WTI? To hedge out Bonnie light? What makes Bonnie light so unique?
Roxton McNeal 13:56
Yeah, it’s a really good question. So I didn’t know either, you know, who maybe you know, got the opportunity to work with this firm. Bonny light is a very light sweet crude with a very low sulfur content. So in that case, it’s it, you know, a lot of refineries like it made, it’s very easy to crack it, which means create the different products from it. It has a very low carbon density as well. And it makes it very easy to do. And you can get a lot more product and better products and a body. So it’s always going to be at a premium. And that was kind of the baseline, it’s always going to be at a premium over kind of the two benchmarks that people would probably heard of, which is the David Brent benchmark, which is typically how oil is priced in Europe, and then the West Texas Intermediate, which is pretty much the Louisiana, Texas oil and the benchmark for that. The other problem is, you know, a lot of people don’t understand but they think David Brent is some kind of you know, it comes from the Brent sea, and that’s its benchmark and there’s a price. Well, people didn’t know very shortly after dated Brent became this benchmark. Their supply dried up At the North Sea. And now it’s actually a mix of four or five different petroleum ZZ that are coming from that area and then are flowing down. So it’s not one price that actually dictates what David Brent is. It’s four of them, and it’s very variable. So you take the fact that your benchmark is really not investable, say that, a couple of the fact that body isn’t a different grade, and the supply and demand dynamics are very different on a body and demand kind of high grade oil than dated. And I just felt like the correlations that I ran, really were showing, not as high as they expected, you expect like a 9394 correlation, there actually, in some cases, depending obviously, on your back tests in the high 70s, low 80s. But more important, the variability around those prices was drastic. So that was the main impetus for me to say, Okay, well, in order for me to really create a good benchmark on this, I gotta underlying, figure out what the underlying factors that drive oil prices are in general, and then be able to pull out those factors and apply different weightings on the body as opposed to dated Brent, that can give me a better forecast on exactly what that exposure is that I wanted to hedge.
Corey Hoffstein 16:13
Can you expand a bit on what that process looked like? And what some of those key factors ended up being?
Roxton McNeal 16:19
Sure, yeah. So essentially, I hopefully this doesn’t get too technical. But really, I think it’s the process that really created the hedging. So first of all, you realize that nothing’s normally distributed. So what I did was okay, in order to create a distribution profile of Bonny light, I actually used a filtered historical simulation, where I was actually scaling the volatilities of the time series, and then re scaling them up to come up with a really good distribution. And then once I got that distribution, I applied a PCA or principal component analysis to it, because I wanted to be able to separate those pieces because my thought was, if I could separate the loadings, then they’re just additive, because just by the way, that PCA is constructed. And then if I could weight those exposures differently, I could actually use a maximum likelihood to figure out what weightings would give me the distribution of that filtered historical simulation. So that was kind of the process. There’s a stepwise fit in there as well, which is a critical component, because PCA itself is just a weighted average of factors. It’s very unintuitive. So I needed to apply some kind of intuition and addition to being able to trade stuff that was tradable, right? I mean, it’s great that if it loaded on GDP, but how do you trade GDP, right? I mean, you had to make sure that it was tradable. So applying that stepwise fit, and then finding the factors, and I did a bunch of tests to because if it’s PCA, then all the loading of that factor should be in one loading, it shouldn’t go across the loadings. They’re orthogonal, right. So I did a lot of tests on that it seemed to work out, there’s some kind of overlay, but I kind of forced the system not to load on any factor that was in PCA one, it couldn’t load on it, and PCA two, or PCA three, which is also pretty critical. And then once I got those loadings, I just optimized, I said, What’s it look like? Now, I optimized to minimize the variability between the prices I was getting from the Platts bulletin, which is the only place I could get Barney pricing. And I said every day just just run the optimization. But if it’s working for me, don’t do anything. So those factors are loading. And there’s going to be a 50% chance that I’m going to do better or worse, because there’s definitely basis, but I just let them run. If I was losing money on the hedge, I would restructure it right away reoptimize using genetic algorithm, and then it would reset it. So that way, I was always capturing some kind of drift component. But if the drift was bad, I was cutting it real quick. That’s how I set up the program. I guess the more interesting part is the conclusions that you came in, right? What did it load on? It loaded? It was kind of interesting. I had a proxy for the OPEC demand and supply constraints that they put on. And it’s funny it it had so little loading on anything that OPEC was stating it was going on in the world. But it did load a lot on US production, US demand US growth, the s&p 500. And those were kind of intuitive. I thought that those that kind of made sense, if that were the demand was coming from in fact, I would think it would load very differently. Now. In fact, I think it would probably load on China a lot. Now just so happened at that time, it was loading a lot on the US economy, the dollar index, I thought was kind of neat. The dollar index drove a lot of it that was kind of in the middle of the road, but where it was really loading was on copper, steel and gold. It loaded really heavy on those where the correlations were up in the 90s on that with it an r squared, like through the roof, so that throw that primary loadings are, and it also loaded, which was a very hard factor to implement on local catastrophes, and that’s what really created the fat tails. And if you couldn’t get that into your model, it was a very bad thing. So that was the other trick to this knowing. And the reason I want to filter historical simulation is because Was it was able to load on that scale volatility of those returns in the past that created the fat tails that I couldn’t get the normal distribution. So that was pretty much it was super interesting, learned a lot. Fortunately for me, I didn’t have to do a lot of hedging, because I think when I started, oil was trading at $60 a barrel. And when I finally stopped is $160 a barrel. So that was a fortunate enough to apply some of the trend trading skills that I learned prior to not have too much of a hedge on during that time period.
Corey Hoffstein 20:29
So I’m curious, you went from this incredibly mathematically driven trading background with the CTAs, you’re working on a very mathematically driven hedging product. But I know you had a front row seat to really a lot of the meetings that were driving political risk, that were a large factor to what was happening with this commodity. And so I’m curious as you got to see sort of the behind the scenes machinations, how did it change your view about trading commodities,
Roxton McNeal 21:00
that was super interesting for me, because coming from school and being in the trading world, I was always under the impression that everything was driven by you know, Adam Smith, and the invisible hand of capitalism, and its free market. And that’s how things move it’s supply and demand, but really getting detailed look inside, you know, just this one commodity, though the oil commodity, you realize that there’s other factors that are just, you know, call it idiosyncratic to the general market conditions. And what I mean by that is, there’s a lot of political influences that determine the price of commodities. And whether that’s politically that’s been influenced by other economies, or if that means there’s political unrest in one oil producing company and how that fit, you know, affects the oil coming out of your country. That is all interrelated. And it’s really, it’s not a market conditions, it’s something that comes in, you can’t control for it, you can’t control for the risk of it. But it has a huge influence and a lot higher influence than I could have imagined. I know you and I talked about kind of the personalities that I met in this. And I don’t know if this is the right place to mention those names. But needless to say, it was some pretty high up US government officials that had a lot of influence in what was going on in Nigeria at that time. And, you know, the ramifications spilt through even trying to get the former president where we’re getting the allocation from reelected, although the constitution of Nigeria at the time said it’s a two term, but we’re willing to go out there as a nation and say, you can run a third term, as long as it’s democratic, I just thought it was super interesting stuff. And to be privy to those conversations, certainly is influenced the way that I look at the world and shaped the way that I invest, even today. So
Corey Hoffstein 22:46
well, let’s keep jumping forward through the career. Now, onto your time at General Motors asset management. And you’ve held a few roles there related to asset and liability management and liability driven investing. This is sort of an interesting niche of asset management that I’m not, I don’t want to assume all of the listeners necessarily know about. So I was wondering if you could give a quick background as to maybe how liability driven investing differs from traditional strategic asset allocation.
Roxton McNeal 23:15
So at a very high level, if you think of just a regular asset management firm, or anyone who’s just running your own individual portfolio, you’re looking at the asset side of your balance sheet, you say, Okay, well, they did my equity, my fixed income, all your different factor exposures, beta exposures, and really what you’re doing there is your objective function is something like you maximize return with minimize risk or minimize risk with some kind of constrained on an expected return, right, something along those lines, but it’s all acid focused, it looks at what kind of return you want to get, when you start looking from an asset liability management perspective, all of a sudden, you bring your liability, the other side of your balance sheet into perspective. So as a pension, what we’re really doing is we’re managing a bunch of assets, to make sure that we can pay the beneficiaries benefit payments at the end of their life. And the way that that’s typically modeled that liability is a big zero coupon bond strip of zero coupon bonds, you have these benefit payments every year, out 80 years, which are based on actuarial assumptions. And you have that whole big short like we call it, it’s almost like a short bond exposure, because it think of it as you have to make these payments. These coupon payments, as it were, they’re actually called benefit payments, but look at them like that, that are becoming due. And your assets are used to generate the return to be able to cover that liability. So once you start looking at it from that perspective, the only return that really matters is the return that you can make above your liability growth, which is dictated by its discount, right? So again, it’s a bond. It is mandated in the United States that these cash flows are discounted by a double a corporate bond. So not only is it a fixed income instrument, it’s a corporate instrument. So there’s a rate component and There’s a spread component. So because if you want to manage that effectively, then the discount rate is your bogey in ALM it’s not your return, it’s your relative return compared to your discount rate return. And then your risk isn’t the risk of your assets and your typical covariance matrix or your you know, roll it all up, figure out what your aggregate volatility is, it is your volatility or your tracking error between your assets and that same liability. So there’s your difference, you have a return that is relative to your discount rate. And your volatility or your risk is the tracking error that you bring against your liability. And then there’s a whole host of different terminologies they use for ALM like immunization defeasance, and, and all these terms, but really all you think about there is, the less risk you want. If you want to mimic what your liabilities are, then you have to have more assets holding what your liability looks like, which means you go more corporate bonds, more fixed income. So the more fixed income you have, that you’re long on your asset side, you are mitigating the riskier liability. And depending on how big your fixed income allocation is, in respect to your total portfolio, has different definitions attached to how effectively you’re hedging that liability.
Corey Hoffstein 26:13
So your first role when you were at General Motors was to set up a set of models and reports to sort of better collate the risk and positions across the different asset classes and make that something that’s legible and understandable for the board. Curious as to how you thought about trying to find a balance between having a report that’s detailed enough to convey sort of the necessary information, but still simple enough to be useful for a board that maybe doesn’t have all the nuanced understanding that you do.
Roxton McNeal 26:45
That’s honestly, credit is still something that I’m trying to learn and deal with. coming from my background, very quantitative background, I always thought if you left information out that you weren’t telling the whole truth, you know, it’s very pure mathematical in terms of the way that you present things. But you’re right, it’s sometimes it’s like, what good is it if you use the audience within the first minute, they’re just lost. So I really had to adapt to a lot of different techniques in order to be able to explain it at a high level. And I think what I ended up reverting to at the very end, is just a series of very high level factor exposures that they could understand. And I would even relate those exposures to something even more so are easier to understand, like the s&p 500, right. So when I was running risk models, and I showed them what the risk was, I’d say, you know, for a given move in the s&p, our expectation is that this portfolio should go up or down by XML, and then relate the different exposures that made up the asset class, and explain the risks based on that exposure. And the loading of that exposure within the asset class, that seems to actually work pretty good. There’s still a learning curve, because he was just talking about factors and everything else, they get lost. But I think after two or three times have the same message, and it’s a lot of it’s about communication, they get to start thinking the way you think, and actually really understand it, and then it starts generating a lot of really good questions. So I think it worked out pretty good. Keeping it at that high level.
Corey Hoffstein 28:07
One of the things I often hear when dealing with allocators who have to manage committees is that there’s this big risk of groupthink, particularly when you’re talking about more opaque strategies that people might be scared to admit they don’t understand something. And so they’re just more willing to go with the direction of the group. I’m curious as to how you’ve dealt with that in the past.
Roxton McNeal 28:28
Yeah, and that happens a lot. It always seems like on these boards, there’s one person that seems to have a little more experienced than the others on a particular subject, and kind of leads the group to making the decision, the only way you could really counter that, again, you got to be very polite, right? Because they may not be at your level of understanding, so you can’t be condescending, or any of those things. But the beauty of a management committee or an oversight committee is it’s typically made up of a bunch of different functional groups. And even though the guy who has a little bit of experience who really wants to lead the charge on the decision, he’s typically in one group. So if you can more readily relate the topic that you’re talking about across the different functions, so they understand how it fits into their pocket. I think you get away from that a lot more. And an example of that is very simply, there’s always this loggerheads going on between the balance sheet and the income statement that a sponsor who has a big pension, the balance sheet is always looking for immunise this thing, I don’t want any volatility on my balance sheet, and then the income statement guy because the accounting rules in the US are like, I don’t care what you do, you just get me a 10% return because that flows through my EPS. And if I get those numbers, we’re gonna look good. So when you’re presenting something, if you can do it from both their perspective so you understand where they’re coming from, I think it helps a lot and it entitles the person that really knows their stuff within their crowd. as functional group to really relate and speak to you about it, it’s not always easy to do. And it’s times it’s not possible to do depending on the subject matter. But I always go into these meetings trying to be able to relate it across so that groupthink doesn’t dominate as much as it maybe it could in the past,
Corey Hoffstein 30:17
when you’re managing LDI portfolio, how do you think about trying to quantify the benefit of adding a new strategy or asset class to the lineup?
Roxton McNeal 30:27
I’ll revert back to, you know, what I was talking to you about in terms of ALM in general, when you start looking at an ALM plan, your return bogey is the discount rate that is being earned by the liability because that’s what its growth rate is, and your volatility is the tracking error between your asset return and your liability return? If that’s the case, then myself and a colleague at UPS have actually written a paper on this. It’s in a white paper. Now, it hasn’t been published yet. But we’ve had a lot of interest, where we said, Well, if that is what the real bogeys are for the ALM study, then isn’t that a Sharpe ratio, but from an ALM perspective. So if you look at the numerator, you’re looking at a Okay, well, you have your return on your assets minus your return on your discount rate, all over the tracking error between your assets and liability. That is your efficiency. That’s how good you’re doing running an ALM portfolio. And if that’s the case, then just like the Sharpe ratio, you can use it to decide what whether you want to bring another asset class or another beta exposure into the mix, just like the Sharpe ratio, what’s the correlation between its return and the return of the combined portfolio? And does it fit? Is it a creative to the ALM Sharpe ratio, and if it is, then it’s a viable strategy to be included?
Corey Hoffstein 31:47
last jump forward in time here, you are now the head of multi asset investment strategy and allocation at the UPS Investment Trust. I think this is going to be a fun part of the conversation because we’re going to talk about some contentious topics. I didn’t realize there were so many contentious topics in LDI. So this will be fun, but maybe it’s a bit of a level set. Before we go into all these contentious topics. Can you give a high level overview of the UPS Investment Trust and sort of its organizational structure, long term goals, liabilities, constraints, that sort of thing?
Roxton McNeal 32:18
Perfect. Yes, absolutely. So UPS pension is actually a series, it’s three pensions that are grouped together under a trust, that’s actually the UPS group trust, it’s three defined benefit plans that were responsible for managing the totally, um, is about 50 billion of assets now. And it’s one management plan, and two union plans. So those are the three make it up and about 50% management plan right now in terms of AUM and then 2525. For the two union plans. Just recently, they’ve completely bifurcated in terms of what their goals and objectives are. So the management plan was announced frozen a few years ago. And as of 2023, it’s going to be completely frozen. So that’s really in an ALM state, that’s where we’re really loading up on the fixed income, because we don’t need to generate returns for service costs, it’s closed. So all the entrants that are currently there, no new entrants can come in, and there’s no accretion of service costs in that plan. That’s what it means to be frozen. So if that’s the case, it looks like a static fixed income portfolio that’s not growing, we can immunize and hedge yet very easily. The other two hand, plants, however, are growing at a very fast rate. And for the work that we’ve done is like at a faster rate than the industry norm, they’re very long duration as well, which makes them risky. And they’re growing very fast. So they’re, it’s almost like an ALM aware type strategy that we’re implementing, where we still need the return seeking assets to cover the service costs and some of the funded gap. But we also want to be wary that any kind of risk that we take in the asset class portfolio against the discount rate is uncompensated, and we’re trying to explain that to the board. It’s just an extra volatility that you don’t need. So if I can hedge out all that discount rate risk, which I can with overlays, then yeah, you still have risk in the portfolio, but it’s compensated, I’m getting equity risk premium for whatever that premium is. So at a very high level is kind of where we’re at in different states. So we applied different strategies to them, which is the responsibility of my group is to set those allocations based on the different risk profiles of those plans.
Corey Hoffstein 34:25
All right, now, let’s have some fun, because as I said, there are I guess, some explosive debates around some of these points of managing LDI. I think this will be a fun part of the conversation, because I expect there are going to be some listeners that are perhaps going to be seeing read and other listeners that are like this doesn’t seem like a big deal at all. So like, for example, as it turns out, the idea of a glide path is a very explosive topic in the world of LDI investing. So I’m curious, how do you think glide paths should be implemented? Yeah, so
Roxton McNeal 34:55
I can see why people think glide paths don’t work. And it’s because their approach to them, I believe, are just too naive. Why would you want to lock in a glide path? If it’s going to take you 20 years to get to the next level? Risk is not bad. It feels like when people start implementing a glide path, they think risk is bad, you need risk, especially if you’re underfunded. Let me kind of level set here. First, though, the ALM strategy that you employ is dependent on a spectrum of things. So there’s this whole spectrum of what you should do depending on a host of issues. Namely, are you over funded? Are you underfunded? Are you at funded? That’s a big one. If you’re underfunded, you have a completely different strategic allocation than you do if you’re at funded or over funded. The second one is what’s your plan doing? Is it frozen? Is it closed? Is it open? Completely different strategic objectives for all of those? So when I talk about what I’m going to be talking about here very generically, is the biggest risk our plans, I think, if you’re funded, and closed, it’s very easy. Just shut it down. Why take that risk? So that one’s an easy one. I think the more interesting one is, you’re underfunded, and you’re still open. That’s a quagmire. That’s a problem. So how do you handle that? So what I’m speaking about is that, so you need risk. There’s no way about it. Some of these plans are 70% funded, and you know, great mathematically, they can’t earn their way out of that, you know, especially there’s a level of no return in math, where it’s like, you’re 60% funded, put a lottery ticket on because that’s how you’re gonna get out unless your sponsor gives you money or anything else. I’m talking about just pure return wise here, you’re not gonna get out of that. So I can see there’s a controversy there. But at the same time, the way I look at it is any kind of glide path. If you look at the assets that you have of a pension you have, essentially you can break it up between two big blocks, return seeking assets and fixed income and those fixed income are your ALM assets. If you were to create an outperformance option, which is call it equities minus fixed income, and have that equity be your equity benchmarks, the s&p 500 and your fixed income be the TLT or some 30 year bond right about your hydration. What you’re doing your funded status is going to be based on that outperformance, right? It’s the outperformance of equity to fixed income and obviously, you know, equity outperforms and it becomes negative right? And the way I look at it is really more relative so called equities divided by fixed income. If I were to create that profile and buy a put option on that outperformance option, that is exactly a pensions risk profile for a funded satisfied bet. Exactly. And if you’re short a put or call overwrite really you can look at as a call right or shorter put, you can create a Delta profile on it. And that delta profile is your glide path. As you start moving up the glide path and you’re overfunded, obviously, your lawn that put your start cutting it down, you’re all fixed income and you cut everything off. If you start going down on the other side of the putt, what would you do it, you got to risk, your delta is telling you to add more RSA, as you go up, your delta is telling you to cut RSA and add fixed income. And if you look at it like that, I think it makes all the sense in the world because you’re just delta hedging your funded status volatility. And that’s how we look at it. That’s something that we developed. Obviously, it’s tricky, because it’s a five year horizon you’re looking at, you have to put thresholds, you can’t be so super reactive. But at the same time, it tells you what to do. And I think it would it, it tells you what to do effectively in the same time.
Corey Hoffstein 38:36
So let’s talk a bit about those risk seeking assets and the way they can be put to work. Again, another hot button issue is sort of a time horizon liquidity match or mismatch. There’s a lot of debate as to whether folks in the pension space who have far dated liabilities should be taking on more liquidity risk and doing more in the private investment space. And there’s other people who push back on that, arguing that it has really important implications for what’s going on with the liquid assets. And I’m curious as to what your take on the situation is.
Roxton McNeal 39:07
Yeah, another great question. But again, I’ll frame it in terms of an open, underfunded plan. Again, if you’re funded and you’re on snuffed, then there’s no reason that you have to take that illiquidity risk. But you brought up some really good points. I mean, I think private equity does some good things, I think it matches, especially if it’s an open plan, it typically has a longer duration. So the private equity will match that maturity profile of fairly well, hopefully, you’re picking up some liquidity premium. I don’t necessarily believe especially in buyout space Friday, you’re getting any kind of liquidity premium at this point. But that’s besides the point. Even if that’s not the case, I still think there’s an added benefit to private equity. And that is the dampening of the volatility that it brings a new asset class. So if I can earn a risk premium that’s commensurate with equity, like you aforementioned, that earn me a commensurate equity risk premium, but I’m only reporting third of the volatility when something happens, that’s a beautiful thing for me, because that helps with my liquidity management, it helps with everything and helps it flows through the whole system. Because if I’m not a forced seller to get cash to pay for rebalancing, because my public equity markets are marked every day, well, that frees up a lot of opportunities for me in dislocations where I can use that cash somewhere else, and really generate some good compounded annual growth rates.
Corey Hoffstein 40:29
So just to clarify, it’s not that you actually believe the private equity is less volatile. It’s the reporting part. That’s key there.
Roxton McNeal 40:37
That’s exactly right. I do not believe it’s any less volatile. In fact, we, for a lot of our risk, we take a lot of time trying to dissect what the actual volatility of privates is, right? So we say, what is it it is a six month moving average? Well, how do we dissect that six month moving average and do it daily, and we do some work on that. But yeah, it’s the reporting, it’s the accounting part that we don’t have to worry about. Liquidity is very real. And when you have to rebalance, or equity markets are going down, or you have swaps on that are requiring a cash payment. That is a real problem for pensions. I mean, I’ve talked to a lot of women in the first quarter that were, you know, they couldn’t do the right thing because of the constraint on liquidity. And if you have a lot in private equity, there’s no hit on that they’re not even reporting, you don’t even know what they have, you know, for another two quarters at least, the only downside of that the risk that you’re taking is, if it’s an exacerbated move, and lasts for a long time period where you start getting hit on these marks. Yeah, you can run into some problems. But at that time you bought time, you’re not in the throes of it where it’s putting a lot of constraints on your balance sheet. So absolutely. It’s the quote unquote, fake volatility that we like that the private equity brings.
Corey Hoffstein 41:44
I’m happy to start a public equity hedge fund, by the way, that just reports on delay, if that would be useful for you.
Roxton McNeal 41:51
That’s awesome. That would be great. That would be great. I think there’s actually a, I can’t remember his name. There’s a Harvard professor who’s doing that. And he it’s called PE or something like that. And he’s trying to he’s trying to do just that.
Corey Hoffstein 42:04
So I’m gonna spoil this next question. I know your stance, you have in the past adopted tactical tilts, you went longer, you increased exposure to longer duration fixed income, at times, you’ve increased risk exposure coming out of 2017, because it’s some momentum signals, curious as to what your processes for trying to identify when the right time to take tactical bets is, and then how you go about implementing and managing those positions once they’re put on.
Roxton McNeal 42:35
So it’s kind of surprisingly simple. We play for a tactical, our fundamental advantage is that we have a long timeframe. And we have a lot of real money. And that really drives the way that we put these tactical tilts on. And the way we do it is we look for dislocations in the market, we look for Miss pricings in the market. And that’s when we will enter the trade now whether they’re Miss pricings or not, I don’t know if that’s the right word. Let’s put it discounts to where they were trading a while back. Right. So we’re armed with the knowledge. And I mean, everyone’s armed with the knowledge that if you look historically over the last 100 years of data, the s&p 500 that a 10% drawdown in the s&p is going to happen at least once a year. I know it doesn’t feel like that because we you know, we had such a big bull run before the first quarter that it felt like every year was, you know, a 10% drawdown was not unheard of. But on average, it’s about once a year. So what we’ll typically do is if there’s a 10% drawdown, we will typically try to tilt into that asset class that had the drawdown, we can hold it for 20 years. And if we believe that it’s not going to recover from that, then why are we even in that asset class to begin with? So that’s kind of our feeling there. We look for correlation dislocations to a good example of that is the credit to rate dislocation that happened during the first quarter, credit spreads blew out and rates came all the way back in. I mean, for a pension, that was probably the easiest rotation you could ever do. I went in and I bought a bunch of corporates, and I sold a bunch of rates and then just waited for that squeezed to come in, again, a big dislocation. And that’s one example a 10% location. And then if it hits 20%, we’ll add more risk on it to get even a bigger dislocation. We’re also armed with the fact that about every five years you get a 20% drawdown in the s&p and conditional on a 20% drawdown, it would typically the expected drawdown is actually 33%, which just so happens to be right where it hit in the first quarter of 20. Right. It was a 33% drawdown. So historically, we’re armed with this stuff. And we know that if these are we’re talking about one year and five year drawdowns, and if our investment horizon is 20 years, then why wouldn’t we take advantage of these opportunities? And that’s when we tell and that’s when we manage them. Typically, we wouldn’t go in delta one securities because there’s a lot of risk there because we’re definitely not market timers. I have no idea if tenza bottom or 20 is the bottom or 30 threes of bottom, but typically during sell off, there’s a lot of dislocation and the convexity markets that you can take advantage of an example is the first quarter of 2020 Yeah, I mean whiskies were the way to go. I mean, the SKU levels were so high, we just rolled whiskies on all the way down. We know the risk was one of them on them. And we just kept managing that risk. So we also look for maybe appearing, put spreads looked really well, that during the sell off again, because of the the SKU side. And then at high levels of all, I mean, vol is very highly mean reverting, we were putting some trades on that were actually had a ball trigger. I mean, we had a vol trigger of 80. It was a pretty good trade, right? It knocked out of all of 80. It worked out for us. So we look for those dislocations we try to be smart about putting the trades on. So we know what our losses, our max losses, we don’t want to put in delta one where we’re naked, because we’re certainly not market timers. But that’s pretty much you know, how we put on our technical trades.
Corey Hoffstein 45:47
As a larger institution. I know you also have the ability to manage certain strategies in house, I’m curious as to what your decision making processes around when to build out a capability in house versus when to hire an external manager?
Roxton McNeal 46:04
Yes. So we are fortunate that we do have some capabilities in house to develop strategies, I’d say, depending on the asset class. Now I’m mostly responsible, like I said, for the strategic and tactical allocation, as I told you, I it’s mostly in derivatives. So for us, our hurdle is can we develop it in house, and it’s cheaper than two and 20? What is the cost of this because that’s part of the carry, too. So if I’m paying two and 20, plus whatever the cost of these strategies are the Theta decay, that is my total costs. So if I can do that in house for cheaper than that, or substantially cheaper than that, then it makes a lot of sense. And that’s how we look at it. Now the equity side of the house is very different to manage an active equity portfolio, even a passive equity portfolio would require a ton of people, a ton of execution traders, and it’s just not worth it. Right. And if we can go get it passively for three pips, that’s what we’ll do. So it’s really a cost function type analysis, I think it also based on expertise and trust we have and the ability of the people that are managing it. So that’s how we look at it, you know, we’re lazy, most of the time, we just can’t get over that too. And 20 hurdle carry is just too much as you know, hedging and is a decaying asset. It’s a lot of times not cheap. So if we can develop it in house and do it for bips as opposed to two and 20. It certainly helps the conversations when we talk about the performance of our program.
Corey Hoffstein 47:27
Well, you teed me up right there, because I wanted to save the best for last, maybe the most explosive topic when it comes to LDI plans is tail hedging. And so I think you sort of leaned into it, maybe gave away the answer. But I am curious as to where you come down in this debate as to whether LDI plan should be adopting tail hedges?
Roxton McNeal 47:46
Yeah. Again, this is something I am fairly passionate about, I think tail hedging, if it’s done, right, adds a ton of value. And I’m not talking about your typical, you go in there, let’s buy an add of money that put you know what I mean, which is outrageously expensive from an implied ball perspective. And it’s just stated, again, your portfolio, I’m talking about being smart about what you’re wanting out of a tail hedging program. And for us, it’s when things get bad, we want to make return, I don’t care if things are going good, it doesn’t matter to us. So really, I’m not looking to source Delta, I’m looking to source Vega, and gamma. And because delta is the most expensive part of the Greek profile that you get, if I can sell that, or at least not by that, and by the gamma or the Vega that I need, that’s going to lower my cost. And it’s going to give me the convexity that I need when bad things happen. As I said, when I start anything, I start with a framework, right? What does this thing look like? And we really have five pillars to the framework. We know every hedge has them. Its reliability, which is the you know, the correlation to the trade. It’s the cost, obviously, how much does it cost. And the way we define cost is the sum of all the negative returns that we made from a trade historically, we have benefit, which is your between your all positive returns, and the distribution between zero and 98 percentile. convexity is all the returns positive greater than 98 percentile, and then decay. And that’s a very important factor for us too, because decay happens very fast. And if we can’t monetize these trades, then the hedge wasn’t worth anything. For us. There is a lot of debate going into who should be the buyers of tail hedging. And I’m going to cut up flip 180 degrees here, because the way that we’ve been talking about they call it tail hedging, and I don’t think that’s the way we look at it. We don’t look at it as a hedge. I think it’s a secondary benefit. I think as the markets are going down and I’m getting marked to market on these tail programs that I have a mitigating the risk as I’m making money as the s&p is going down. So in that way, it’s a hedge. But really what we’re doing is we want to be able to make money so we can put it back into the risk seeking assets. So when the market is drying up, and people are running, I want to supply that risk back into the market. Because that’s where my fundamental advantage comes in. I have a lot of money, and I have a long time frame. So when you get those big 20% dislocations, I want to be putting money back into the market. So do I think that we should be buying tail hedging programs for the sake of tail hedging programs? No. In fact, I think the world’s kind of shifted the wrong way. I think the banks who are selling these programs should be buying the tail risks themselves. They’re the ones that have to report quarterly earnings. They’re the ones that have all these constraints and pressures on him for daily p&l. I don’t have any of that. So actually, I should be selling these products to them. And if you take that mentality now, I also think, obviously, they want to have some on right, it just prudent, but I want to be buying this stuff they’re buying. So if they are going in there and buying tail hedging, I want to buy that I want to share in that with them. So from that perspective, we kind of flip everything around, we kind of say no, we shouldn’t be buying tail hedging. But if we can source it really cheap, where we can put it back to work to earn a higher CAGR. That’s what the program does. And that’s kind of kind of, I guess, not your typical hedging program. I guess, if you you know, look at it from that perspective,
Corey Hoffstein 51:11
from an actual operational perspective. Can you talk about maybe what the implementation looks like the different tail hedging strategies you have in play and how you think about how they sort of interact with one another?
Roxton McNeal 51:25
Absolutely. So again, we look at the framework where like, which strategies are going to be bringing in a good carry component? Obviously, when you’re looking at any kind of hedging strategy or convexity strategy, you want to make money, you want low bleed or low, no carry? And you want a lot of convexity, right, that’s the unicorn, it doesn’t exist. But we do think that if you can replicate what the Greek profile looks like, of what you want, when a crisis happens, and work backwards, so you run your model, where you say, if I had these strategies in place here, what does the Greek profile look like in a big sell off? The crisis Greeks, that gives you your input that you need to design what you’re looking for. So that’s kind of the tack that we play. And it’s like, are you adding convexity? Are they diversifying, one may add a lot more benefit than convexity. And we kind of balanced those out across the Greeks to come up with a really good diversified program. We don’t wanna be loaded up on gamma, you know, to capture the realized if it’s only implied moving. So we always want a really good combination of the different profiles that we’re looking at, for diversity reasons,
Corey Hoffstein 52:28
are you typically keeping sort of the same set of strategies over time, or you cycling through strategies depending on what’s being offered in the market or what the conditions are?
Roxton McNeal 52:38
Yeah, so we definitely cycled through strategies, there are some that we keep that have performed very well, they have low bleed, some we monetized during the first quarter, and we don’t have money anymore, their profile just doesn’t look the same. A lot of it has to do with spot levels of all and also skew levels of all most of our strategies to keep them cheap, have a funding leg and a hedging leg. Again, hedging used loosely, but the only way to get them cheap is to have some kind of a funding leg. And as long as that funding leg is the right Ray risk for your firm, it’s easy to do. And to give an example, that just in great space, which I think people will get this really easy. It’s like I could care less about rates at 350. So I will sell pairs there all day long. I don’t care if they’re rich or cheap, because I know if I get there, I’m fully funded. And something happened good for me. So I’ll sell those all day long, maybe to buy in at the money receiver option. So that would be an example of me funding up this kind of expensive at the money receiver by selling a bunch of convexity, you know, pay or convexity on the upside. And sometimes it’s worth it. I mean, I’ve had trades where it was right before the Trump election where the payers and the receivers, the receivers at zero, had such a skew. It’s such a high smile on him that I was selling both the fund up and at the money and actually they were paying me to buy a receiver to hedge out some of the liability. So that’s one example in equity space, you know, you can sell a meeting sell a put spread to buy VIX, so you’re it’s really a follow vol trade. You’re kind of equalizing the Vegas, you’re looking for the convexity of the VIX to kick in and dislocations that happens. And a recent one we’re looking at is, you know, funding up the difference between weekly and intraday volatility, to capture that premium going short, the one week variance and going along an hourly variant strategy, and then using the proceeds of that to fund up a four week variant swap. So your funding like is this weekly to daily dislocation that you get mean reverting markets, and then obviously, if it reverts or becomes trending, that will kind of collapse and go to zero, but I have my four week on the other end, that’s my hedging, like, so that’s a definition of a really good carry strategy, but it’s defensive at the same time, our group will never go naked. I will never go short. The one week that just will never happen. We saw what you know, a lot of the Canadians got beat up very badly on that. So, but we do think it’s a good source of premium, but I’ll hedge it with something else. So
Corey Hoffstein 54:58
last question for you. vaccine rollouts are underway. More and more people are getting vaccinated, fingers crossed mid summer into summer world is getting back to normal. What are you most looking forward to?
Roxton McNeal 55:11
Yeah, for me, so I’ll kind of break it down between work and kind of pleasure. Family definitely can’t wait to get back to traveling. I missed that tremendously. I think there’s a lot of value add when you’re sitting across the table from people that are, you know, the structures and the dealer, quant structures on the desk. So you can really hammer him with questions and really go over things hammer him out on the whiteboard, for example, just that interaction, I miss a lot, and I can’t wait to get to traveling again, won’t miss going back into the office, that commute is gone from, you know, an hour and a half in the morning to about five seconds, because my office is just on the other side of my bedroom here. And I feel like I’m a lot more efficient. I think there’s a lot of social stuff going on at work or around lunch and whatnot, I just feel like so it’s more effective here. My quality of life is better, I’m able to spend time with the family. I teach my kids across teams and the whole nine yards. So it’s so much better. And also, I guess, as a corollary to that is, you know, if I don’t go back to the office, then I may have some time to try to delete, relocate for a little while, like to a Caribbean island, and just, you know work from there. So I’m going to try to take a page out of your book. Alright, that sounds
Corey Hoffstein 56:21
nice. That sounds nice. Well, hopefully on the other side of this, I can come visit and sit down and pitch you my delayed s&p 500 hedge fund. Yes, we’re here to anytime. Well, I’d love to hear all rocks. And I can’t thank you enough. It’s been a phenomenal conversation. Thank you for joining me.
Roxton McNeal 56:38
Oh, great. Thank you so much. It’s been a it’s all my pleasure. And you know, thank you so much for having me.
Corey Hoffstein 56:48
If you’re enjoying the season, please consider heading over to your favorite podcast platform and leaving us a rating or review and sharing us with friends or on social media. It helps new people find us and helps us grow. Finally, if you’d like to learn more about newfound research, our investment mandates mutual funds or associated ETFs. Please visit think newfound.com. And now welcome back to my ongoing conversation with Harley Bassman Harley did steal a turn of phrase from you, when there’s a blow up. Leverage is often seen at the scene of the crime. And so for a lot of investors, leverage has become a pretty unwelcome word when they start reading an investment prospectus? How do you think about the leverage provided by options versus other forms of leverage?
Harley Bassman 57:37
I have to call you out. Again, you’re a professional here, there’s more than one type of leverage. You have financial leverage, you have economic leverage. Let’s talk about the first financial leverage, you can go and buy a million dollars of the two year treasury and you’re gonna make point in three quarters up and down your freight scoop up and down by 1%. But you could also do is you could take the million dollars, put that into overnight one week money and then buy the two year futures contract. Unless you think they’re gonna move rates dramatically, like hundreds of basis points overnight, that one week or one month, cash instrument ain’t moving at all. So your economic risk is unchanged if the two year Treasury risk, but now you have $2 million of line items on your books. You have the one month t Bill let’s say and the futures. Or how about this? What if you go and take your one month cash and put that into a T bill and then go and buy the first eight euro dollars? Well, now that’s basically more or less mathematically equal to a two year swap, which is more or less equal to two your future which is more or less equal to a two year Treasury plus or minus mere pennies. You have the exact same risk as being one your million dollars of two year treasuries. Yet you have 9 million of open line items on your books on your balance sheet. Are you levered nine to one? I don’t think so. Yet, if someone asks you if you live a nine to one, you kind of got to say yes. Now, in contrast, you can go and buy an option or some other context instrument where you own something. It’s a one for $1 invested, but it can perform like two to one or three to one in one direction or the other. I call that economic leverage. Is that good or bad? That depends. But it’s two different things entirely. And most times people tend to quote financial leverage because that’s what shows up on the books and records of firms or people’s whatever their personal balance sheet. I think it’s the most important to recognize as far as convexity per se. Negative convexity is not Dead by itself at a price. So as I’m fond of saying, no bad bonds, just bad prices, there is a price where I will sell an option. Just because I’m the convexity neighbor does not mean I’m always long optionality. But for price, I’ll sell it. There’s a price where I will buy a mortgage security over a Treasury or a future when it gets to be a wide enough spread. So the embedded optionality that I’m selling is very expensive for the risk profile. So that’s fine over there. What tends to happen is that people tend to sell options, sell convexity has a reach for yield. When the curve is flat, and spreads are tight. If people are desperately trying to go and hunt for yield, they sell optionality, and they then sell it too cheap. And then when bad things happen to good markets, and you’re in a risk profile, where you could lose five versus making one, if you’re not getting paid enough money for that, clearly it could be a problem. And that’s what tends to happen in almost all the major blow ups, which I wrote about recently. And 8794 98 And most recently, liquid financial crisis, Reuters crisis is the market is structurally short convexity and it becomes unmanageable. And the cost of managing that risk tends to get larger than the liquidity of the market. And that does not end well.