Brett Nelson – A Pioneer on The VIX Path
In this episode, we talk with Brett Nelson, CIO and founder of Certeza Asset Management. Certeza Asset Management focuses on investments in volatility using proprietary algorithms to capture perceived mis-pricings in the VIX term structure.
We talk about the start of VIX trading from VRO to modern VIX products. His background from winning math competitions, isolating Vega on S&P options, to being on the vanguard of VIX trading. Brett has a unique entrepreneurial journey and has found his way outside the traditional paths to being a portfolio manager and volatility trader.
I hope you enjoy Brett’s insights as much as I did…
Listening options:
- Listen on Stitcher
- Listen on iTunes
- Listen on Spotify
- Listen on TuneIn
- Android users can grab our RSS feed here.
Have comments about the show, or ideas for things you’d like Taylor and Jason to discuss in future episodes? We’d love to hear from you at info@mutinyfund.com.
Transcript for Episode 21:
Taylor Pearson:
Hello, and welcome, I’m Taylor Pearson, and this is the Mutiny Podcast. This podcast is an open-ended exploration of topics relating to growing and preserving your wealth, including investing, markets, decision-making under opacity, risk, volatility and complexity. This podcast is provided for informational purposes only, and should not be relied upon as legal business investment or tax advice. All opinions expressed by podcast participants are solely their own opinions and do not necessarily reflect the opinions of RCM alternatives, Mutiny Fund, their affiliates or companies featured. Due to industry regulations, participants on this podcast are instructed to not make specific trade recommendations nor reference past or potential profits, and listeners are reminded that managed futures, commodity trading, forex trading and other alternative investments are complex and carry a risk of substantial losses.
Taylor Pearson:
As such, they’re not suitable for all investors and you should not rely on any of the information as a substitute for the exercise of your own skill and judgment in making such a decision on the appropriateness of such investments. Visit www.rcmam.com/disclaimer, for more information.
Jason Buck:
This is Jason Buck with Mutiny Fund and I’m sitting down with Brett Nelson of Certeza Asset Management. Certeza is an algorithmic trader that takes advantage of mispricing along the VIX curve term structure, to create absolute return value trades along the VIX curve. Now, what’s interesting to me about Brett is, he’s traded the VIX longer than anybody I know. Quite frankly, nobody could trade longer because he’s traded the VIX since the inception. But, we’ll come back to that piece. I want to start off, maybe, us talking about his childhood, the middle school and high school years and going around to different math competitions and trying to hide the fact that he was a math nerd. So, if you could tell us a little bit about that, Brett, that’d be great place to get us started.
Brett Nelson:
Sure. It’s a little funny, people that grow up in the current age, they maybe don’t realize this, but being smart, and being into nerdy things wasn’t cool, back in the 90s, and 80s. So, being an 80s kid, and growing up in that era, it definitely wasn’t cool to be a math enthusiast or a mathematician, so I had to be a closet math nerd. I joined some math programs when I was in about sixth grade, because one of my teachers came to me at the time and said, hey, I really want you to separate yourself from the rest of the students here and pursue this thing that you’re obviously quite good at.
Brett Nelson:
So, I pursued that at a low level at first and sixth and seventh grade and then started getting into regional, state and national competitions. To put it in perspective, you’re talking about doing what you would consider as intermediate level college mathematics in middle school. Then, obviously, once you hit more of high school years, you’re doing advanced mathematics at that point, and trying to compete. So, that’s really what spawned my interests and in my genesis in the financial markets.
Jason Buck:
So, yeah, having that mathematical background. Then, you first start learning about the financial markets, what really caught your attention from a mathematical perspective, what was really that impetus? What was that trigger? What was that aha moment where you’re like, this is where I can use applied math and use that practice in real life?
Brett Nelson:
Well, first off, I always liked the concept of making money. I was entrepreneurial in nature anyway. So, I was trying to start little businesses when I was 12, 13 years old. And, not like your typical lemonade stand type stuff, I was doing more wholesaling, and things like that, finding wholesale distribution stuff and trying to resell it at market prices and things like that. That was when I was quite young. So, that gives you a little bit of an insight into the type of person that I am.
Jason Buck:
Natural [inaudible 00:04:14]. That’s how you do it. [crosstalk 00:04:16]. Exactly, for the search. But, just like you said, you’re a math nerd when it wasn’t cool to be a math nerd. We’re both old enough to know we’re entrepreneurs when it wasn’t cool to be an entrepreneur. Now, it’s really cool to be an entrepreneur.
Brett Nelson:
Yeah.
Jason Buck:
So, sorry to interrupt you there, so going from wholesaling, on the entrepreneurial side, taking you into markets?
Brett Nelson:
Yeah. So, with a focus on just financial interest in general, then you have to realize that now we’re getting up into the mid 90s. In the mid 90s, the internet was becoming a thing pretty heavily and particularly, online trading platforms had become a thing by the mid to late 90s. And, you have really broad access cheaply to direct access trading platforms, interactive brokers, for example. Obviously, being as young as I was, I was undercapitalized, so I found myself in options because what do you do if you don’t have capital, but you want to do more sophisticated financial arbitrage type mechanisms.
Brett Nelson:
I landed in options and I started to, obviously, flirt with the different types of positionings within options and specifically, whether you want to actually manifest a directional opinion with options or whether you’re looking deeper into the mathematics, and it’s looking deeper into the mathematics, is where I got lost in it and finding the alpha that actually exists within that market.
Jason Buck:
Yeah, we always like to think about options as, it provides you with a finer paintbrush. There’s a lot of things you can do with options that you can’t do with other instruments. So, what was the fascination for it? It sounds like you weren’t interested in the directionality, but maybe trading the Greeks or thinking about different ways that pair off the Greeks against each other. How did you [inaudible 00:05:57]?
Brett Nelson:
Absolutely. Yeah. So, I made my fair share of directional trades. Then, once I got far enough into the math, when you dig deeper, you start realizing that what you thought was a directional trade and options actually wasn’t. When you fine tune in, you do some attribution and everything, you actually start realizing that most of the alpha that you’re picking up is due to vega, right? I started to really focus in on the vega component. A lot of people will say, well, I’m gamma Scalping and things like that. No, and then actuality, you’re not gamma scalping, you’re vegas scalping, even if you think you’re gamma scalping. Because, vega affects gamma.
Brett Nelson:
So, the alpha that’s inherent in the vega component, was fascinating to me. But then, you have to jump into the idea that, in fact, forecasting variants and forecasting vol, is one of the hardest things to do mathematically. It’s hard enough to predict direction or to forecast direction accurately, but to forecast the volatility or the variance of the direction, is much more difficult. So, that’s really what led me down this path in saying, I can utilize an inclination toward mathematics to actually create much more sophisticated positioning than anyone else that I was around, was doing at the time. Obviously, being in Utah, I wasn’t surrounded by Wall Street types. So, I spent just an absurd amount of time on quantum forums, mathematic forums, trading forums, and trying to figure out who was doing what I’m doing. And, was I by chance, pursuing a path that was incorrect, or was I actually just doing something that was uncommon, but not incorrect, right.
Brett Nelson:
And, by the early 2000s, I had found a small group of individuals that were more vol-focused like myself, and we were forced into trading options to extract vega. But, that wasn’t really ideal, and everyone that was doing it, and those who are honest, still do it, actually recognize that in many cases, it’s not ideal. Because, if you’re trying to be purely focused on volatility, options might not be the best way to do that, right. But, if you’re willing to take on some directional exposure and synthetic exposure and things like that, then maybe options are for you. But, for pure volatility, there didn’t exist a product at the time.
Brett Nelson:
So, we had this core group of individuals that spend a lot of time just isolated from the rest of the public and saying, how do we build these algos, and how do you forecast volatility correctly? Right. Unfortunately, you can’t buy a book on that. So, it was really us doing what we could do with the mathematics that we knew. Admittedly, amongst that group, I was more mathematical in nature and others were more finance-based. Some that were in prop groups in Chicago. But, as luck would have it, 2004 came around and the CBOE came out and said, “Hey, this VIX index that you’ve all been using is now going to have futures.” And, that was the defining moment.
Jason Buck:
Yeah, I was building a VIX arbitrage or intermarket spread models on the VIX in 2010, 2011, and I thought I was early and the liquidity is low, I can’t imagine 2004, right. So, part of that is, you don’t have the liquidity but also in a nascent market like that, there’s a lot of opportunities to get advantage of if you can model better. When they launched it 2004, it was a different VIX index. What year did they switch to now using the full spread that we use now?
Brett Nelson:
Well, it’s not just the different VIX index, it was, the liquidity really had to be there for VIX to be a usable instrument, right? It really didn’t do us a whole lot of good mathematically to say, you and I are the only two trading it and I’m trading against you for bragging rights. For example, we might have an April contract and a September contract to, maybe, a December contract. That’s not entirely helpful, right? In actuality, to really model variants appropriately, you really need a term structure.
Brett Nelson:
The term structure really didn’t evolve consistently until later in 2006. So, sometime around October of 2006, right? Really, what led to that was, the options came online for VIX a little while after the futures did, and the options actually gained in popularity faster than I thought they would. We dabbled in the options as well, and I was doing spread trades and options and things like that. But really, the beneficial effect was that the liquidity of both and the depth of both books actually increased dramatically. So, you actually ended up by October of 2006, you actually had what we’d consider a usable term structure that could be modeled.
Jason Buck:
Originally, they used a much more at the money strikes, and then eventually, though, they switched to out of the money strikes.
Brett Nelson:
Yeah, in 2004, yeah. So, the VRO and-
Jason Buck:
Right, VRO, yeah.
Brett Nelson:
Yeah, they were focused very much at the money before and now it’s a much broader selection. In either case, we still have the uniqueness of VIX that most people don’t realize. I like to call it the structural alpha of the VIX markets. The reason I said there’s structural alpha’s, because whether they were using the more narrow at the money strikes or a much broader swath of further out of the money strikes that they do now, it really doesn’t change the fact that you can’t replicate the synthetic index. And, because you can’t replicate the synthetics, there’s no physical to ARB against. So now, you’ve got a structural condition, mathematically, where the prices of the futures are allowed to wander much further than they would in a normal physical commodity market.
Jason Buck:
Great. Just to go back for a second, how would you describe VIX to the layman, instead of, maybe, 30 day, 40 variants? Maybe, tying it to the options on the S&P or the SPX, how would you describe the VIX index before we get into the term structure?
Brett Nelson:
Yeah, really simply, VIX is actually best viewed as even though it represents an annual number… so, it tries to suggest that it’s supposed to be telling you how much fluctuation up and down in an annual term, the S&P 500 should see. But, that’s not entirely accurate or helpful. So for example, if VIX is at 20, right now, right, it’s near 20, as we speak, it’s not accurate to say that it’s predicting or forecasting a plus or minus 20% fluctuation within the next year. That’s not helpful at all. And, that’s not what it is.
Brett Nelson:
What it actually is closer to, most accurately is, what type of daily fluctuations do we expect in the market right now? Okay? And, once you annualize out that, you’re getting more close to what VIX is representing. But, basically, what you have to do with the VIX to get a little bit more technical, is to say, what is this number that it’s stating, right? Now, bring it back into what the representation of that is, in daily terms. So, is it saying there’s going to be a half a percent daily fluctuation typically? Or, is it more like 3%? So, if you go back to, for example, March of 2020, and you’re seeing VIX at 80, VIX at 85, things like that, it’s not really saying plus or minus 80% fluctuation in the market, what it’s saying is, plus or minus 7% daily fluctuations or 5% daily fluctuations, which is wild in and of itself.
Brett Nelson:
But you remember, if you go back to that period, that’s exactly what we were seeing. So, the representation by VIX actually wasn’t far off of the daily fluctuations we were actually seeing in the market. So, that’s really what it is if you calculate out the number a little bit, it’s telling you how much per day would you expect the market to fluctuate.
Jason Buck:
Perfect. Let’s go back, so, 2004, they launched the VIX products on CBOE, you start trading those products, when do you decide, I can make a business out of this and I can go off my own, I can start a fund around this idea.
Brett Nelson:
That wasn’t till 2010, actually. The landscape changed completely again, in 2006. Or sorry, 2008. Because, 2006 said, I can model this. 2008 was the financial crisis. And, suddenly VIX became the darling of the markets because all the institutions and everyone else out there was trying to basically say, this VIX instrument is anti-correlated to the markets and perhaps a better hedge than the traditional products like puts and things like that, that we’re using. Right? So, institutions started piling into VIX, and it became incredibly popular. If you remember, the volume in VIX was tripling every year, sometimes quadrupling every year.
Brett Nelson:
During that time, it was still pretty thin, by market standards, and there was a lot of alpha coming off of it, and we were all making a lot of money if we were familiar with it. But, there was an inclination to say, this isn’t an institutional product, this can’t be scaled, this can’t be offered publicly. And, we were seeing the amount of inefficiency we were seeing, was obviously not going to continue, right? So, it wasn’t really realistic to think that we wouldn’t be able to have a public offering off of that, right?
Brett Nelson:
But, by mid 2010, I was finishing my algorithms, and the liquidity and volume continued to scale up and I said, hey, maybe this actually will be something that can be offered. So, I finished the algos and went into a dedicated prop account for the next 12 months to say, how will this turn out in real time with no other focus, no other input and no continual altering of the strategy, right. It went fantastically well, for 12 months ago, and I started to form Certeza Asset Management. By January of 2012, we launched to manage accounts, and that’s essentially the beginnings of the company.
Jason Buck:
As you said, you had that entrepreneurial nature, or you were always an entrepreneur-minded person, but you were trading the VIX for a long time. But, it takes a lot to build that confidence to make that leap, though, to go from your prop account, where you’re your clients, so you understand the deep mechanics of what you’re doing, so you can handle draw downs, etc. How nerve wracking was it to actually go to managing outside money with the SME structure and then eventually a fund structure?
Brett Nelson:
Well, it wasn’t that bad actually, because I eased my way into it. I managed for a couple more friends and family. And, only with a handful of millions. Like, a few million dollars. That allows you to at least, not jump in with both feet, necessarily and really make sure that… It’s not that the model breaks in a lot of those cases, it’s just that someone isn’t capable mentally of managing outside money, and especially larger amounts. A person to manage a portfolio, especially large portfolios, really needs to be able to disconnect from the dollar values, right? You really can’t consider what it was like to trade your own account or watch the dollars fluctuating. And then suddenly, daily swings are in the millions. You have to be able to be the right kind of person that can handle that, right, and-
Jason Buck:
It’s fairly sociopathic in a way, right? You have to think about gambling or whatever, that’s actually the proper way to have a fiduciary responsibility, though, you have to think in percentage terms, is essentially-
Brett Nelson:
Yeah, incentives terms. Then, the harder thing is actually to not just think in percentage terms, but to not consider what any fluctuation might mean to the revenue that you create for yourself. Right? You cannot make decisions based on, oh, the fluctuation today just took money out of my pocket in terms of fees or something like that. You have to turn that off completely and say, the model is the model. If there’s no reason to think that the model is broken, then don’t think that the model is broken, and don’t make decisions based on what your financial situation might be.
Jason Buck:
Exactly. I know you’ve modeled back to the ’87 crash, but as you know, it’s extremely difficult to model VIX and especially, as I always say, I don’t really trust a lot of back tests beyond 2010, just because, like you said, it was such a nascent market. We didn’t have that liquidity before the ETP started becoming online and becoming real traded. And, because that vol surface is so undulating and the real money players are changing, how do you think about that? It’s one of the hardest things. Everybody’s, can you back test this up, big strategy back to 1900? You’re like, no, you can’t. They’re like, why? I can do with stocks and bonds or whatever.
Brett Nelson:
Yeah.
Jason Buck:
You got to speak to that of, how it’s so hard to model or back test with any sort of VIX strategy.
Brett Nelson:
Yeah, really, a lot of people say, we’ve taken our VIX strategy back to 2005 or something like that. Like, okay, first of all, let’s be real, it wasn’t until October 2016 that here was even consistent pricing within VIX, okay. So, at earliest, it will start there. Second thing is, any of us that were trading it prior to about early 2010, is about the period, we know how inefficient it was. Okay. There were inefficiencies weren’t just mathematical inefficiencies or some type of legitimate arbitrage opportunity. It was obviously a lot of players entering the market that did not know what they were doing and were giving away money, right.
Brett Nelson:
So, I would say to anyone that’s looking at historical returns or back tests that are back into the ’08, ’09 period, a big grain of salt there. We actually refuse to take our spec that far, because I know what the market was like. Then, if you want to go back and further and you say, yeah, I want to model this clear back into the 90s or into the 80s with the old calculation on VIX. And, you could say, yeah, I can tell you that, in the crash of ’87, VIX, quote unquote, the VIX index, printed over 125, right, something like that. But, you’re saying, I don’t know how terribly useful that is because I still don’t have a representation of how I would have been able to monetize it or capture that value. What do I do when I still don’t have an indication of what correlations there might have been in the first or second month futures? So, really, it’s nice to see how far the index values can go, but you have to be super conservative on what you would actually think that you can monetize out of that?
Jason Buck:
Yeah, there’s a vast difference between the actual index values and actual term structure of the actual contracts. You could trade in the liquidity and the spreads and everything on that. So, we left off, let’s go back again a little bit. So, you launched Certeza. What was the original thesis and has this thesis continued to this day, and how did you look at what was your primary strategy for trading, VIX term structure?
Brett Nelson:
Yeah, the thesis actually hasn’t changed, really. The reason for that is that the origins of my VIX trading gave me a pretty broad perspective on what was the appropriate way to utilize VIX. So, for perspective, that other small group that I was working alongside of, their primary mechanism they were going towards was, what you would call the risk premium trade, right, the short vol trade. Even back then, it was very juicy and incredibly consistent. I found myself on the other side of their trade a lot. That’s why I split off and started doing my own thing. It was weird because I was on the other side of their short vol trade a lot, but not because I wanted to be long vol all the time. But, I could see the problematic nature of the short vol trade and how difficult it was to control in certain events.
Brett Nelson:
Then, on the other hand, I wasn’t necessarily trying to express long volatility bias all the time, although all of us would admit that if there was a way that we could just be long volatility all the time and make us money as consistently as the short vol guys, we would love to do that. But, that’s just not realistic. So, I really had to jump in and say, there’s a thesis on this market that says, it’s not about the direction of vol. It’s not about being long or short, there actually is structural alpha in this market. It is unique, there is no other instrument like it, and the characteristics that make it unique are a dream for a mathematician. So, I can exploit that and I can jump into it and say, my thesis on it being an inefficient product and how to exploit the inefficiency hasn’t changed. Although, I will say, over the years, we’ve gotten much better at it.
Jason Buck:
So, without giving too much away, how do you think about the inefficiency of it? Is it part of the term structure and having that roll yield, creates unique opportunities, especially if you’re, maybe, picking off some of that left tail and then having a bias towards the right tail and explosive environments? What is the unique structure of the VIX markets that allows for those opportunities?
Brett Nelson:
Yeah. Well, first, let me tell you a couple characteristics that VIX has that nothing else has, right? It’s non-trending, right? It’s very difficult to deal with something that trends. For example, the S&P 500 trends higher over time, it’s undeniable, right? Inflation pushes it higher, productivity increases, pushing higher. So, you’re looking at and saying, as a mathematician, I have to then play some nice little mathematical tricks to eliminate the trend before I can actually find any useful data in it. Okay?
Brett Nelson:
VIX doesn’t have that because VIX doesn’t just simply drift higher over time, it actually drifts horizontally over time with spikes up and spikes down. But, it always seeks back to that, roughly, that mean value, right, which is exactly what it’s designed to do. So, that makes it really nice to work with to begin with. Then, the other thing is that, when you talk about the explosive events, the right side, there’s no left tail index. There’s, what we call, a really fat right tail, which means VIX can’t really go below nine for any reasonable period of time, mathematically, but it sure can go clear up to 100 in a hurry, right?
Brett Nelson:
Normally, when something is between 10 and 100, you would say, okay, where’s the mean? Is the mean around 50 or is the mean around 45 or something? No, in this case, the mean is really… Or, I would say, the moat for a low vol environment is about 13, which is very close to about left, that bottom, that floor. So, you’re saying, oh, so now there’s a virtue of us saying we don’t have to protect against certain types of moves, if we position ourselves correctly, because there is no move in that direction, right? So, you can be single directional in your protection, right.
Brett Nelson:
Then, once you’ve identified that you say, the uniqueness of what I talked about before, you say every other market that has futures, whether you’re talking cattle or grains or anything, financial futures, there is always a physical, right? That you can buy the physical and if the futures prices get too far away from what that fair value is, [inaudible 00:25:38] will come in and push them back very, very quickly. So, they maintain this really tight range. VIX is unique in that it’s impossible to create that physical. So, you can’t buy the VIX index. You could buy the VIX index, it wouldn’t be particularly helpful all the time. So, you’re actually dealing with futures that can vary. I like to say they wander around, what you might consider could be the fair value. You notice how uncertain that is, that statement. It’s like, they wander in a wide range around what maybe probably might be the fair value, because no one can actually say what the fair value of the future is.
Brett Nelson:
That’s what I loved about it, because you say, suddenly, you’re in a situation where you realize, hey, a lot of these big institutions out there, they have massive long equity portfolios, billions of dollars that they have to put to work. And, when they put that to work, they do so with long stock, right? They have to add value to their clients, so what they do is they say, oh, we’re going to buffer the downside. How do they do that? Well, they come out and they start buying puts. Put options. And, when they buy those put options, they typically do it, in my experience, in a relatively, what you’d call, an unsophisticated manner.
Brett Nelson:
In other words, they’ll do a put by once every quarter, let’s say, and when that quarter is over, they get rid of those puts, and they put new puts on. They have a budget, something like 2% a year, that they’re willing to spend on put buy, okay. That type of indiscriminate put buying leads to alpha within VIX. The reason for that is because they don’t care if VIX is at 17 or VIX is at 16 and a quarter, they don’t care. They just need their puts, right? But, someone who comes in like us, we say, we absolutely do care about the difference between 16 and a quarter and 17. Right. And, we know whether there’s been some herd mentality or just some institutional pressure in one direction or another that has pushed a certain VIX contract off of… You can’t say that they’ve pushed it away from fair value, because you don’t really know where the fair value is. But, you can say, they’ve pushed it seemingly to the extremes of what could be considered the real fair value. So, it becomes highly unlikely that the current fair value is the accurate fair value.
Jason Buck:
I want to touch on something you said just a little bit previously, too. It’s like, VIX doesn’t trend but it mean reverts or spikes or clusters. So, as we know, from our Mandelbrot, it’s clustering where it’s mean reverting and you don’t really know, either way. As you alluded to, a lot of times, people say, VIX has a bimodal distribution. You have a low vol environment or high vol. I almost say it’s like a trimodal distribution, because then you have the spikes as well. Obviously, they usually mean revert pretty quickly, but you have a low vol environment, a medium vol environment and a high vol environment, that’s like a trimodal distribution.
Jason Buck:
That offers different opportunities. And, when we’re talking about the actual VIX index or the spot VIX, that’s one way to talk about, like you’re saying, it mean reverts, and it trends sideways. But, that’s the actual index. Then, you overlay, like you were alluding to, the term structure of those VIX futures. So, that can give you opportunities where maybe you’re playing that left tail where you have that barrier of lower volatility. So, you have that term structure that’s rolling down where you can maybe bring in some income or play that left tail side. Then, you could be putting on trades at other parts in the curve. They’re like that long volatility or looking for that big spikes, or vice versa. You could do from month to back month and or switch those ratios around. You guys use a very dynamic way of looking at putting on that term structure calendar trade. Can you talk a little bit about how you look at that, or whether, depending on which volatility environment you’re in, or what kind of indicators or workflows do you look at, depending on how you’ll assess that trading parameter?
Brett Nelson:
Sure. The way that we do it, is we take out a really universal approach to it. Right? By that, I mean, we come in and we say, what data do we have at our fingertips, right? Let’s say 50, 60 years worth of S&P 500 data. We’ve got new methodology VIX data going back to 1990 with all the VIX term structure data, going, at least, usable back to October 2006. So, we bring all that in, plus a couple other things and we say, we’re going to create an exhaustive data set. And, we’re going to run what you would consider an incredibly complex and sophisticated Monte Carlo simulation.
Brett Nelson:
So, for those who aren’t familiar, you’re basically saying, we start at time zero, at now. And we say, what possible moves could this future make in the next day, right? The thing about a lot of Monte Carlo simulations, as they step from one day to the next day to the next day, they’re doing it in a pretty, what you would call, static fashion, a static probability fashion, which doesn’t resemble volatility at all. Volatility is very dynamic. So, what you actually have to do, is you actually have to say, I’ve stepped forward a day. Let’s say, I’ve run 10,000 simulations for one day, right? So now, I have 10,000 new data points that are possibilities of what could have happened over a one day period of time. I can’t simply step to the next day. I actually have to step backwards now and say, let’s re-categorize what now looks like for all 10,000 of those points, and then step 10,000 points forward for each one of those 10,000 points.
Brett Nelson:
So, what we call it is, more of like a regime fit, a micro regime fit, right? You say, you have to realize that the universal set of data for VIX is not always applicable all at once. The reason for that is, you talk about volatility clustering as a prime example. It does you very little good to think that you’re going to get an outcome tomorrow, that is exclusive to a low vol environment when we’re currently in a highball regime. Right. So, don’t count on those outcomes and don’t weight them heavily in your probabilistic outcomes because they’re not going to happen, right. So, you have to actually regime fit very precisely or accurately.
Brett Nelson:
And then once you’ve regime fit, you have to bring that into every step of every 10,000 simulations that you do, right. So, step forward one step, 10,000 data points, and to each of the 10,000 data points, you have to re-regime fit, and then step forward again, and then re-regime fit, and it goes to millions and millions and millions. You create an exhaustive look of what could possibly happen all the way along the curve at every point along that futures curve. Right. Then, you say, where’s my exposure? Right. And, you figure out what the most probabilistic outcomes are. And, you say, we categorize everything in terms of what we call expectancy ratios, right?
Brett Nelson:
So basically, what is my expected profit or my expected loss on this specific trade for a given unit of risk, right? And, we do that for all possible combinations. So, we’re looking at saying, somewhere in there, there’s going to be an identified probabilistic inefficiency. And, once we found it, it gets even harder because then you say, okay, we found something that actually is representing an opportunity, now there’s 10s of 1000s of different ways that we could trade that. We could buy one contract per million, buy two contracts per million. Buy one now and wait for the price to improve and buy another one at a slightly improved price. There’s any number of ways you could trade it, which we call the methods, right?
Brett Nelson:
So then, you have to have an algorithm that comes in and optimizes the method and says, first, identify the inefficiency, then optimize the method around that one trade. Right. [crosstalk 00:33:26]-
Jason Buck:
Sorry to interrupt you, as part of the methods, are you taking into account the steepness of the curve, whether you’re in steep backwardation or shallow and vice versa for contango, if you’re in steep or shallow? Is that overlaying on your Monte Carlo simulations to then figure out that best trade and addressing the methods through the steepness of the curve as well?
Brett Nelson:
Yeah, so the slope of the curve is one variable. It’s one input in this regime fit, right? The curve might be inverted with spot VIX at 20, right, with the index at 20. The curve might be inverted. Or, there’s another time where, for example, like more recently, where spot VIX is at 25 and the curve is not inverted, right? So, you can’t look at it and say, oh, just because VIX, is at 25, we’re going to do this all the time. You have to say, okay, what does the curve actually look like with VIX at 25? And, how many circumstances do we have that are like this, right?
Jason Buck:
So, going back to the methods, you run through the methods, give us some examples of what trades you’re putting on or what kind of probabilistic trades it’s giving you? If you’re putting on different trades, they may be battling or offsetting each other? Or, are you thinking about the ratio of those and what’s the optimal ratio? How much are you hedging it out? All those little nuances, actually, the trading construction.
Brett Nelson:
You bring up one interesting point is that, there are possibilities you might have two trades that are pinging on your system there and saying both of these are good trades, and they might actually be contrary to each other. Right? Then, there’s always this question of, should they both be employed because both have positive expectancy. And yet, they do offset each other slightly. What you realize in VIX is that, yes, in fact, they both can be employed. Because, the VIX curve doesn’t necessarily behave the same way. It’s not necessarily creating a wash, there’s a time difference. There’s a time period difference that you’re looking at. So, both actually can be exploited at the same time. So, you might actually be long and short vol at the same time, right, at different points along the curve.
Brett Nelson:
For us, I will say, in general, we don’t really love the front month contract. Not as a rule, but as a general preference. The reason for that is, when you start categorizing things in terms of expectancy, which some people… Let me get out of the technical term expectancy and get into what someone would call, more like a risk multiple, right? Like, how many dollars are you expecting to gain per unit of risk? What happens is, there’s a lot of inefficiency in the front month future at the very front end of the curve. And, there’s a lot of opportunity to make absolute return there, that’s for certain.
Brett Nelson:
The problem with it is, once you’ve run these exhaustive data sets in these simulations, you realize that there’s a disproportionate increase in the risk in that zone. So, let’s say, in the second or third month, or the middle of the curve, as we say, I can only make $2 per contracts trading, right. I’m like, well, that’s not anything close to the $10 that I can make in the front month futures, right? Then you say, I’ll actually my risk in the middle of the curve is actually only $1 and a half. Whereas, my risk in the front month is $25, right? So, you’re, yeah, it’s a disproportionate increase in risk there. And, we’re not going to accept lower risk adjusted returns, simply because we can get higher absolute returns.
Brett Nelson:
I won’t say that we completely avoid the front end of the curve, but we have a preference to stay away from it. And, you’ll see us adjusting our positions out once we start crossing over into the front month.
Jason Buck:
Yeah, it’s not as sexy, but you get better risk metrics per unit of risk, if you’re in the second, or third, fourth month. So, if you’re playing in the midterm of that term structure, are you putting on calendar spreads directly? Are you putting on triangle spreads? What kind of trades are you putting on?
Brett Nelson:
Think mostly in terms of calendar spreads and butterflies. What most people would call a butterfly. It depends on the regime at the time, but you’ll see us in a lot of calendar spreads. They might be long, they might be short. We have a particular affinity to what you would call a ratio spread or a back spread, right. So, for example, there’s some really unique things you can do with VIX in that you can take… Everyone knows that the short vol trade is quite consistent and highly probable, right, but that it has blow up risk. Once you nail down the characteristics of how vol operates… In other words, for example, if I were to tell you that this curve that we’re talking about it, if everyone envisions that nice, beautiful curve where it’s a little steeper on the front end and flattens out to the back end, and then that with a volatility spike that can invert and go up, so the front end is inverted, quite steeply, and the back end is flatter, but lower.
Brett Nelson:
If I were to tell you something, as a little insight, it doesn’t make a smooth transition between those shapes. So for example, what might happen is that you’ll have this nice beautiful, what we call a, contango curve, right? This upward sloping and flattening curve. And, if I told you, that actually tends to shift first, and then it pivots and flattens out. So, basically, more shifting with not really a whole lot of change between the individual months, shifting, then it pivots. In other words, not gaining a whole lot, but the front month is coming up to flatter stance, then goes inverted. Then, once it’s inverted steeply enough, it goes into shift mode again, and just starts shifting completely higher.
Brett Nelson:
Now, if I say that to someone who actually deals with VIX, they say, oh, yeah, that’s actually yeah, that is how it behaves. If we go back into March of 2020, for example, or back into 2008, you’ll notice that, yeah, it inverts quite rapidly. Then, once it’s inverted far enough, the entire curve starts to shift up. And, you say, oh, well, that means that with different spreads, I can exploit this tendency for the back month of the spread, to actually continue higher while the spread doesn’t continue to widen any further. Or, the widening is minimal relative to the increase.
Brett Nelson:
So, this is where different types of ratios might become fun to play, right? Because, if you know where to expect those behaviors, you say, I can create something that, at its core is actually a short vol trade. It gets all the benefit of that consistent gathering of risk premium, as we say, and the high probability of profit, and all of those virtues, and, yet, it’s a little bit heavier on the back leg, on the long vol leg, right. So, if you know how to do it in the right proportions, you’re saying, the moment that I’m going to start to get in trouble on this calendar spread, that back leg is going to kick in and give me a safety net. Not only that, but if it keeps going, that back end is going to become the return driver, right?
Brett Nelson:
So, you end up with what we call a smile return curve, right? Where, if 90% of the time, VIX stays in his low mode, it’s calm mode, right, and you’re in that ratio spread, you’ll basically just gain the risk premium, the short vol trade. Then, there will be a little bit of a dip where you’ll go, negative P&L on that. You’ll go negative expectancy in the middle of that. Then, on the back end, it just goes parabolic because that long vol kicks in, you become crisis alpha at that point.
Jason Buck:
Yeah, and just to use a toy model as an intuition, pump, the idea would be of a ratio back spread like that, you could be short, the third month, and then you could put on two or three contracts of long, the fifth month, and whatever your ratio is that your model creates… That’s why I’m just using a toy model. That gives the audience an intuition of what that trade looks like. But even better, like you’re saying, as that curve moves from contango, or backwardation goes flat, and he kind of, rises and lifts in both directions, it gives you time also to adjust that trade, too, I assume as well. You can start taking off your closer to position, your short vol position. And really, like you said, ride the wave of the multiple contracts of the back month. Is that the way you look at it?
Brett Nelson:
Yeah, that’s the way you look at it. So, there’s any number of ways you can adjust it, right? Because, on the one hand, we don’t pick bottoms and tops. So, it’s almost unheard of, for us to say, we’re going to take a full-sized position from day one, right? We almost always have a partial position with the algorithm saying, that there is a reasonable probability that the price is improved, and we’re going to be able to add at a slightly improved price. So, it’s like $1 cost averaging for those who have done that in stocks before.
Brett Nelson:
But, once you’ve averaged into the right price, you can always add more exposure or take off exposure in unit, what you would call unit chunks, right? Or, you can actually change the actual back spread ratio to your point of saying, oh, well, I was too short third month, and then I was five long fifth month, right? That could go to a two to four ratio, or a two to six ratio, depending on whether or not… Without divulging too much, I can tell you that you can get into situations where the beauty of a trade like that is, for example, let’s say that you have to short third months and five long fifth months, right, and you’re saying the market has moved substantially, but this trade hasn’t particularly moved because it hedged itself off really well. But, if you do a quick attribution, right, and say, why has the market moved substantially, but the trade P&L is actually relatively flat?
Brett Nelson:
What that means, is that a portion of that trade has moved substantially, and then there’s a portion that’s left. Then, you have to run your analysis and say, is it time to take off the portion that has benefited substantially. Then, you’ve got, not necessarily, quote unquote, free, but you’ve got basically an easy entrance into a reversion trade from that point, right. Which actually happens quite frequently, where you say, the short volatility component of this trade is played out, right? You take it off, and you’re left with a long volatility trade, which happens to be just the time that you should be long vol.
Jason Buck:
Which, serendipitously, perfect time for long vol positioning. So, part of that, too, is, I assume that you’re constantly adjusting the volatility of those different points in the curve for your ratios, and you’re looking at those on a daily basis, like has anything changed? Do I need to change these ratios around? In a way, going back to the very beginning of conversation when you said, people aren’t necessarily gamma scalping, they’re vegas scalping. That’s a way to almost make it scalp the position as it moves around on you on a daily basis.
Brett Nelson:
Absolutely. That’s exactly what it is. Yes, we’re vegas scalping the position constantly. Sometimes we’re not necessarily the vegas scalping it, we’re just adding to sizes, pure money management, right. Pure money management techniques are employed all the time. But then, there’s the vegas scalping component in addition to that.
Jason Buck:
Then, right around that time, it might be the time to put back on a mean reversion trade. But, how do you guys think about… A lot of times, quants are just looking at mean reverting pairs trades, and they say, oh, this is in the 90th percentile, so it’s a high probability it’s going to mean revert. Then, that spread just absolutely blows out on them. We see a lot of that, 2020. So, how do you manage-
Brett Nelson:
2020 was crazy.
Jason Buck:
Yeah. How do you manage a lot of those concepts or positions?
Brett Nelson:
Interesting one. So, you mentioned try modality, right? I’m just going to throw out a really rough number here so nobody hold me to it. But, let’s just say, for easy reference, 99% of the time, VIX is in, what you would call, a diversionary state, right? The trades that you’re making are, by definition, convergence trades, right? What that means for anybody not quite as familiar is, that you’ve got a price that has wandered out to its extreme ends of its normal range, and you’re expecting it to converge back toward a more normal price. Right?
Brett Nelson:
So, 99% of the time, you’re saying that’s how VIX is, and it doesn’t really matter whether we’re on a low vol regime or a high vol regime. So, that’s what you would consider the bimodal portion, right? Low VIX regime has one mode or one peak, where the structure wants to revert back or converge back to that mode. Then, in a high vol regime, like you’d see in a 2008, 2009, and more recently, it’ll want to converse with that point. Then, you have that third mode, right, that you refer to. Yes, it’s a third mode and it’s also described as the divergent regime, right?
Brett Nelson:
Because, the most tricky thing about VIX is that occasionally it stops being mean reverting and it goes divergent. And, it does it for very short periods, but when it does it, it’s incredibly damaging to those who don’t count on it, right? So, you’re looking at it and you’re saying, I can’t just categorize this instrument as being just a perfectly 100% mean reverting instrument. Because, it’s absolutely certain that VIX will come back down to a lower level once it hits 80 or 90. It absolutely will. The question is, could you handle the ride, right? And, money management will get you most of the way there, but you have to be able to say, we recognize the divergent characteristic of VIX in certain environments. Ideally, we can exploit them. If we can’t exploit them, let’s make sure that we’re not vulnerable to them, to a devastating extent.
Jason Buck:
Get out of the way of that VIX moonshot. So, it’s the reason why we really love the VIX arbitrage trade, because, essentially, you could take advantage of both convergent and divergent trades, and there’s very few trades that have that. Most mean reverting pairs trades are going to be convergent all day long. They don’t have a divergent element to them. And, only, really, the big says that divergent element in a way which adds to the uniqueness that you’ve been highlighting. I’m curious that, we had a really unique… you have probably seen it, even, in the entirety of your career, in 2020, you have the Mark Swain ’20 sell-off and so you have this echo of volatility coming into Q4 of 2020. Then, we had this kink in the VIX curve due to election volatility or perceived election volatility-
Brett Nelson:
It was absurd.
Jason Buck:
Yeah, it made really difficult trades because you had high contango and then high backwardation around the event. If you haven’t had that in your data set, how do you model that?
Brett Nelson:
Well, you don’t model it. I’m not going to pretend that we had a fantastic time through that period. That was a very strange period, to see that kink propagate itself for months through the curve. Algos have to learn, there’s no way that we can build something back in July of 2010. In my case, build it up to that and to say it’s going to sit here and operate with no improvement on, at the time, was four years worth of data, right? I’ve got four years worth of data, and it’s a perfect system. No.
Brett Nelson:
So now, we’re getting up to where we have a larger data set. That was actually a really nice addition to the data set, because we’re saying, you can’t even categorize in standard deviation terms what that kink looked like. It was representing prices, relative prices, that, there’s no way to say whether it was a four sigma, five sigma, six sigma, ten sigma. You don’t even know because you can’t actually categorize it in normal sigma terms, right, it was that absurd. Most people didn’t even know what’s happening unless they were plugged into the volatility world. Because, in the equity world, there was what you would consider, a pretty ridiculous rally going on. Looking back on it, you start hearing reports about the SoftBank trade and all of these feedback loops that were happening, that were pushing the market higher, simultaneously pushing volatility higher, which is a whole different story.
Brett Nelson:
But, if it can happen, you need to be able to handle it. And, for us, what it comes down to is, more strict money management techniques rather than more precise modeling. Because, we’ve always taken the approach that not everything that can happen has happened. And, if something can happen, it will happen. That’s the other side of that. So, you look at it and you say, what is considered a devastating loss, a devastating drawdown for your particular strategy. And, some more aggressive ones might be 60%, 70%. Some more conservative ones might be 5 to 10%, I don’t know. But, we figure that out and that’s built into our model. Then, we say, we run our data set to pretty, really, strict risk limits. We say, this would be considered an optimal sizing for this trade, and then we back off even further from that.
Brett Nelson:
So, the point for us is, for a trade like that to blow up our strategy, it would have to go multiple times outside of its previous limits. In the case like you referenced with August of 2020, we were seeing prices that were roughly double their previous limits. So, that was enough to sting a little bit if we were on the wrong side of some of those trades, but it wasn’t enough to blow up. Right. That’s really what you have to consider, especially a young market like VIX is, you want to make sure that whoever’s running the strategy takes a somewhat conservative approach and doesn’t say, look, I can make 100% a year in VIX and, then, suddenly the account is zero.
Jason Buck:
As we both know, we’ve seen a lot of those over the years, how many people are making triple digit annual returns on the VIX, always. But, part of that is, I wonder, you’re somebody that’s highly quantitative, but also with the background as an entrepreneur, so that gives you some interesting insights. I know you follow Rentec a lot. So, in these Q4 2020s, how much of it is entirely quantitative and algorithmic versus the human element of overlaying, my algos may not have all the data they need, there may be something wonky here. And, as a human, I need to step in a little bit and reduce my risk metrics or my risk models, or my position sizes. The way we look at it is, that’s really the best model, is man plus machine. We wish everybody was highly quantitative, but there are times when you may not have enough datasets in your look back where you need the human element. How do you deal with that?
Brett Nelson:
Yeah. I can tell you that, to a small degree, whether you want to call it human stepping in or just a slightly greater degree of caution, that we use, both, in March and August of last year. And, in both circumstances, what you’ve got is that, we can run that simulation. That incredibly exhausted simulation that drew a picture of, it should be able to tell you, within the story dataset, how absurd the levels are that you’re seeing at the time. And, what it’s going to do, inevitably, every time, is tell you that this is a can’t lose trade. It’s virtually certain with $1 risk and 1000s of dollars of profit potential right, and that kind of thing, because it’s not within the dataset.
Brett Nelson:
Now, hopefully, someone’s done something more like what we’ve done, and they’ve gone into their model, and they’ve fattened the tails preemptively, right? So they’ve gone and said, we recognize the historic data set as not complete yet. One key consideration there is, if you generate the scatterplot of prices that have happened in the past, you’re looking at that, there’s little points where there’s these outlier events. And, I’m saying, well, with a huge, enormous data set, they might in fact be outlier events, with a smaller younger data set, the data set just hasn’t filled in yet, is the more likely conclusion. So, you’re saying that, probably, those are going to happen at a greater frequency than what I think they’re going to happen at, they’re not outliers. And in fact, they’re probably not going to be considered the outlier once the full data set is there, they’re going to be considered the three sigma event. And, there is a four or five sigma event that goes further, right?
Brett Nelson:
So, you’re looking at that and you’re saying, hopefully, someone will recognize that preemptively, and went in and fattened the tails out of their distributions and said that, outliers are going to be more common, and to a greater magnitude than what the data set is suggesting. Then, once you’ve done that preemptively, you can at least sit back a little bit and say, okay, we’ve already been quite conservative. But, that doesn’t mean, to your point of, marriage of man and machine, that doesn’t mean that you can then just sit back and say, press go, and I’ll let the black box run, right? You still need to come in and say, okay, there are times… Allow me to reference March 2020. I don’t know if others have used this term to describe it because I haven’t seen anyone use it, but we’ve talked about the commoditization of volatility in about a week and a half period in 2020.
Brett Nelson:
What I mean by that is, there should be a not predictable, but a relatively consistent relationship between volatility and the underlying market. Right? There were noticeable times in March of 2020, where volatility actually commoditized and the futures were being driven by supply and demand independently of one another. Okay. So, what you’re saying is that. normally, month to month to month along the VIX futures curve, what is it like, a 97, 98% correlation, right? It’s an incredibly high correlation, not when they commoditize. Each individual contract becomes a commodity in and of itself and then, suddenly, you start seeing these spreads behaving in not unpredictable, but definitely unusual circumstances, right?
Brett Nelson:
And, on the one hand, it’s an incredible opportunity. But, make sure you still have dry powder, is really the lesson there. If you’ve trusted a historic data set completely up to that point, it’s highly likely that not only will you not have dry powder, but you’ll be past your limits by then already, right? With each one of those occurrences, August 2015 was one, February 2018 was one, March 2020 was one, you look at these and you reassess and you say, were we conservative enough? And, if we were more conservative, what’s the impact to the overall model, right? Certain trades might not actually be worth it once the dataset starts filling in. So, the model, like ours, for example, has seven different types of primary trades that it makes, and there’s nuances within those.
Brett Nelson:
But, if you lump them into categories, you might say seven different types of trades. It might be time to look at one and say, we’re not eliminating it from possibility, but with the new data set that has extreme prices in it, when I rerun the model, or when I rerun the simulation, there’s so much more risk inherent in this trade now, that it might not actually ever come up as being an appropriate trade anymore, right? And, that’s just the reality of it. Because, you look at the VIX ETFs, for example, and trying to short VIX ETFs, and, for many years, it was basically, say, something like a two to one or a three to one. So, you’d expect one unit of gain and a possible three units of loss, right, on a spike event. Then, suddenly, in March of 2020, it showed its true colors, and it was actually 10 or 11, to one loss to gain, right.
Brett Nelson:
So, you look at that and you say, okay, the data set hadn’t filled out yet, right? Now, the question is, after you’ve got that data set, that short vol trade doesn’t even make sense if you actually have a possibility of losing 11 times the potential gain on it. Because, now you’re saying, I really still want to generate at least 10% a year, but if it can go against me 10 times, that’s 100% loss. I don’t know how to come back from that.
Jason Buck:
There’s so many things you touched on there that I thought were fantastic. One, no, I haven’t heard anybody say, they commoditize the VIX in those markets. That’s a great way of putting it because, like you said, people will say things like, the correlations breakdown, or the VIX is broken and the correlation’s broken. It’s not broken, it became a supply and demand market that’s decoupled from the S&P. It happens in these really stressed unusual times. I think that’s a great way of looking at it. That’s usually, yeah, when people get smoked, because they’ve been counting on this negative correlation, and that’s what they built their models on.
Jason Buck:
Then, like, as you said, they always have this Minsky moment, right? The volatility’s gotten so low coming into 2020, that, that spike is going to be so violent, it’s going to be a 10 to one loss that you weren’t expecting and in your model base. So, part of that, though, I’m wondering about is that, like you said, sometimes the base can be commoditized, and it’s due to supply and demand. What we’ve also seen in 2020, is that we went from more of, people have said, fundamentals to flows. So, it’s really about these real money players in those flows, which leads back to probably your supply and demand dynamics of actually the VIX curve or just options pricing in general.
Jason Buck:
So, we work with a lot of managers that come from a market making perspective or come from investment banks, and they’re trying to model those flows. It’s more watching how those flows are affecting pricing or pending pricing or whether it’s getting the dealer’s off side and they have the gamma hedge and you get a gamma squeeze, all of those little factors to flows. But, I’m wondering, when you guys are modeling that VIX term structure, you’re actually seeing the flows in real time as well, aren’t you?
Brett Nelson:
Yes. Yeah. So, one of the things for us is, we don’t like to jump in and say that we’re going to try and base a decision off of flow, right. It’s a difficult market to do that. And, for those who have the specialty in it, I’d be interested in hearing them break down how they would actually do it. Because, for us, we say, the flow dictates the outcome, but we’re not flow traders. So, we’re more interested in the now, and what the opportunity is in the now, right, rather than trying to say what the momentum or something might be based on, different characteristics of order flow and things like that.
Brett Nelson:
So, you will look at us and say that, most of the time, we’re actually looking for the relatively frequent or absolutely frequent occurrence of inefficient pricing. And then, leaning more on the law of large numbers to say that, I don’t want to be so heavily concentrated that I can be wrong once and it would ruin me. But, I want to be consistent. We’re going to set our bars really high, right. So, for example, people often mix up probability profit with expectancy and they try to use them interchangeably, but they’re not the same thing, right? You can have something that’s incredibly consistent, 99% probability of profit, and, yet, it’s negative expectancy, because the one time that you lose out of 100, it just wipes you out, right?
Brett Nelson:
On the other hand, you can have something that only wins once out of 100 times, but when it does, it’s so huge that it’s a positive expectancy trade, you think about the crisis alpha side. And, for us, we actually want the best experience for the client. And, the best experience for the client, in our opinion, is not waiting around for years for something to pay off huge. It’s not picking up dimes and then getting run over by the steamroller, right? The best experience is one that has reasonable probability of profits, something in the 75 to 80% range, right?
Brett Nelson:
So, that, more often than not, the trade is going to win, right? Then, looking at it and saying, we’re on the other side of saying, once we’ve got the 75 to 80% probability profit as a minimum, we’re looking at the expectancy exclusively there. And, we’re saying, okay, now, what’s the return per unit risk, right? And, saying, let’s make sure that this is an identified convergence trade, that’s our core, this convergence trade, and we’re going to count on that convergence, knowing that there’s a divergent element sometimes, right? And, produce this nice, hopefully smooth, consistent return, but with significant amounts of alpha.
Jason Buck:
That’s great. I think, part of the undercurrent of the last second part of this conversation we’ve been having, is the idea that, especially, in volatility markets, as a volatility trader, we’re trading these [inaudible 01:02:52] markets. And unfortunately, we have an end of one experiment that we’re all running. So, at the end of the day, survivorship is the greatest accolade you can have. You guys have been surviving just fine. You’ve been doing this since 2004. So, that’s the key, right, is, at the end of the day, is just surviving to trade the next day.
Brett Nelson:
Yeah. That’s really the thing. Let me give you an example of that. The biggest trading volatility out there is still the short vol trade, right? So, it’s undeniable that the largest portion of those who are trading volatility are doing so in a manner that they’re mostly selling volatility, and gaining that volatility risk premium, right. And yet, if you were to jump onto some of the different indexing sites, like Eurekahedge, for example, or something like that, and you pull up their little compiled indices of volatility funds that they’ve put together. That’s a really convenient way to look at, okay, what exists in the market, actually?
Brett Nelson:
And, you would say, well, because the short vol trade is so huge, and so many people do it, obviously, if I pull up one of those indexes, it’s going to be 40 short vol funds, and then a few crisis, alpha funds or a few long vol funds, right? In fact, that’s not the case. You pull up those indices, and you find out, well, actually, there’s only like four or five constituents in the short vol fund, or in the short vol index. Then, there’s like 16 in the relative value, right? Then, there’s maybe six or eight in the long vol side. That speaks to the survivorship problem, right? Because, it wasn’t that there weren’t a lot of people doing the short vol trade, it’s just, they’re all gone now. And, they’re gone for a reason, right?
Brett Nelson:
And, it’s not that there weren’t a lot of people doing the long vol trade because that has usefulness, but they, also, most of them, weren’t able to survive either. Because, there’s something to be said for not sitting around for three years without a gain or four years without a gain. The entrepreneurial in me says, you still have to generate a profit and generate revenue each year, right? So, it speaks volumes that the relative value trade is the largest of those constituent indexes. You look at them and say, okay, what is it to the relative value trade? Well, in fact, I would argue that the relative value trade is where the alpha is generated, primarily.
Jason Buck:
Yeah, it’s a great way of putting it is like, the short vol traders blow up, the long vol traders bleed long enough that they get [inaudible 01:05:27]. So, the relative value, where you want to be sitting as an entrepreneur. So, I’m going to ask my last question. As always, I’m fascinated by little details. At the top of your pitch deck, you have the quote by Oscar Wilde, “To expect the unexpected shows a thoroughly modern intellect.” So, that must mean a lot for you to put it on your deck. So, why did you put it in there?
Brett Nelson:
It goes back to that, not everything that can happen, has happened, right? I’ll veer off a little bit. So, we’ve all heard the quote, there’s the known unknowns and the unknown unknowns, right, all that kind of thing. That’s really common. It’s, maybe, a little cliched at this point. But, I’m going to throw one at you and say, there’s the known unknowables, okay? That’s the part that’s missing from that quote. The market actually does two things all the time. One is that it doesn’t expect something that hasn’t happened before. But, those things are going to happen all the time, right?
Brett Nelson:
So, first of all, expect the unexpected. That’s rule number one. Then, on the other part, it’s the known unknowables of the market thinks… It’s the hubris, right? What I mean by known unknowables is, when you study VIX long enough, you start realizing, let’s reference back to that kink in the curve around with the election season last year, and you say, that’s what I would call a known unknowable. Because, the reality is, that all of those people who think they know that there’s going to be some type of volatility there, don’t know that there’s going to be volatility there. It’s like saying, I know that this upcoming season on August 27th, there’s going to be a hurricane that makes landfall in Houston, right? You’re saying, there’s a chance that, that can happen, but if I’m running an insurance company that’s insuring homes in the area, I’m certainly not going to build that into my premiums and say that I know that this event is going to happen.
Brett Nelson:
But, the volatility markets do that all the time. They execute trades based on a known unknowable. And, it provides huge opportunity for those of us who jump in and say, this thing that you’re suggesting that you know is going to happen, is an unknowable. You can’t possibly know it, because no one could possibly know it to the certainty that you’re implying with those prices. So, I’m going to take the other side of that trade. And, you might be right. But, most probably, if we ran this simulation 1000 times, we’ll come out on the winning side of that.
Brett Nelson:
So, on the one hand, you say, exploit the known unknowables. And, on the other hand, anticipate the unexpected. Both of them are a question of hubris, right? Don’t ever get so confident and arrogant to think that you either know the future is going to happen, or you know that something is not going to happen.
Jason Buck:
Yeah, I probably use epistemic humility too much, which makes both of us no fun at dinner parties. We’re always questioning everybody’s knowledge base. But, I thank you for coming on. I really enjoy that you shared your story with us. I love what you guys do at Certeza. And, we’ll provide links and everything in all the show notes. But, thanks again for coming on, Brett and sharing with our audience, and we appreciate it.
Brett Nelson:
Yeah, thanks for having me, I enjoyed it.
Taylor Pearson:
Thanks for listening. If you enjoyed today’s show, we’d appreciate it if you would share this show with friends and leave us a review on iTunes, as it helps more listeners find the show and join our amazing community. To those of you who already shared or left a review, thank you very sincerely. It does mean a lot to us. If you’d like more information about Mutiny Fund, you can go to mutinyfund.com. For any thoughts on how we can improve this show or questions about anything we’ve talked about here on the podcast today, drop us a message via email. I’m taylor@mutinyfund.com, and Jason is jason@mutinyfund.com. Or, you can reach us on Twitter. I’m @TaylorPearsonMe, and Jason is @JasonMutiny. To hear about new episodes or get our monthly newsletter with reading recommendations, sign up at mutinyfund.com/newsletter.
