The Derivative - Trend, Prop, and Being Allergic to Optimization with Bill Gebhardt of 10Dynamics

Episode Date: August 29, 2024

In todays episode, Jeff Malec sits down with Bill Gebhardt, the founder of 10 Dynamics, to discuss his interesting career journey and the development of his unique systematic trading approach. Bill shares his background, from working on the floor of the CBOE in the early 90s to earning a PhD in finance and transitioning into the energy trading space. He provides insights into the evolution of the commodity markets, including the rise and fall of Enron, and the challenges of maintaining an edge in fundamental trading. The conversation then delves into Bill's transition to systematic trading, sparked by his exposure to a successful systematic team at his previous firm, Trailstone. This led him to develop the 10 Dynamics model, which is built on 10 core signals that closely mirror Bill's own decision-making process as a trader. Bill explains his "allergic to optimization" philosophy and the benefits of using multiple time frames to generate positions. He also discusses the importance of risk management, operational efficiency, and adapting to changing market conditions. Throughout the episode, Bill shares his unique perspective on the markets, the role of human behavior, and the future of systematic trading. This insightful discussion offers valuable lessons for both aspiring and experienced traders looking to navigate the complex world of alternative investments. Buckle up, because this conversation is about to "send it" into the world of alternative investments. Chapters: 00:00-01:40= Intro 01:41-4:18= Puddle jumping to across the pond / ski adventures 4:19-14:46= Quants, Derivatives, Energies & Enron: Starting 10Dynamics 14:47-31:55= Prop trader behaviors, philosophy 101, random walk paths & systematic approaches 31:56-43:39=Time frames – why they matter, deploying risk & flat investments 43:40-53:47= Models vs traders, Quant trading, fundamental filtering & never cherry picking 53:48-01:05:09=Mispricing, Alpha decay & Delta hedging From the episode: Whitepaper 10Dynamics: “The Freezer” – Automated Risk control systems for systematic trading 10Dynamics.com HI HO Silver! Blog post Follow along with Bill on X (Twitter) @BillGebhardt1 and on LinkedIn. Don't forget to subscribe to ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠The Derivative⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, follow us on Twitter at ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠@rcmAlts⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ and our host Jeff at ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠@AttainCap2⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, or ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠LinkedIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ , and ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Facebook⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, and ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠sign-up for our blog digest⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Disclaimer: This podcast is provided for informational purposes only and should not be relied upon as legal, business, or tax advice. All opinions expressed by podcast participants are solely their own opinions and do not necessarily reflect the opinions of RCM Alternatives, their affiliates, or companies featured. Due to industry regulations, participants on this podcast are instructed not to make specific trade recommendations, nor reference past or potential profits. And listeners are reminded that managed futures, commodity trading, and other alternative investments are complex and carry a risk of substantial losses. As such, they are not suitable for all investors. For more information, visit ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠www.rcmalternatives.com/disclaimer⁠

Welcome to The Derivative by RCM Alternatives, where we dive into what makes alternative investments go, analyze the strategies of unique hedge fund managers, and chat with interesting guests from across the investment world. Hello there. Well, the kids are back at school. It's 98 degrees in Chicago, and I'm not wearing sunglasses. If you didn't get that, please go watch the Booze Brothers. But anyway, let's get to the pod.

We're sitting down today with Bill Gebhardt of 10 Dynamics, so named because of the 10 core trading signals Bill systematized to build out his multi-timeframe model. We go from Bill's days in the Chicago option pits, to his energy trading days, to his work as an allocator of sorts, hiring traders into a Deutsche Bank prop trading arm. In between all that, we touch on quantum mental trading, prop versus multi-strat, discretionary versus systematic, and one of my favorite lines of late, being allergic to optimization. Send it. of unique managers to meet your unique risk and reward parameters. All right, everyone. We're here with Bill Gebhardt of 10 Dynamics.

How are you, Bill? I'm great. How are you? I'm great. Beautiful, sunny, 60-degree day here in Chicago in August, which is super rare. I was seeing it pop up at coldest on record, I think. I was going to say, I lived in Chicago for a bit, and that sounds awfully cold for August. I remember sweating in bars in August.

That's the way you're doing it. If you're going to be sweating, do it in a bar. Yeah, exactly. How's London this time of year? Lovely? Cold? Yeah, it's, London's pretty nice.

Um, we're, we're cooler than you guys. We're probably, you know, 69, 70. Well, actually not you right now.

You're, you're older than we are, but, but typically we don't get really hot like, like you do in Chicago, but it's, how long have you been in London? Uh,

23 years now. Wow. It's been years now. Wow. So it's been a while. Yeah. All right. And before that? Well, originally from Colorado, but I moved around a lot for jobs. I was in Chicago, as I said, for a while.

I worked on the CBOE for a bit way back in the early 90s. So I got my floor training. Not floor training. I was a clerk. But I got the experience of being on the floor, which is, I think, one of the coolest examples of complete chaos resulting in order that you'll ever see. And it's a shame that kind of the open outcry has gone away with technology. But yeah, that was interesting. But yeah, I moved around a bit and ended up at one point down in Houston working for an energy company. and they um they did a joint venture uh that brought me to London so I've been here since then I was coming I came on a two-year expat deal and

just never left never left uh and my listeners know I'm not gonna let you say Colorado without talking skiing for a minute so where were you from in Colorado uh I grew up in Fort Collins I went to school in Boulder at CU awesome yeah so you. So you're a Dion fan? You liking what's happening with primetime? Yeah, I think, I think it's pretty interesting. I mean, they've, you know, I mean, my, my, this is how old I am. I was in the McCartney era, right? So my, my first year at CU, I think they went one and whatever the number of games was there one in 10. And then McCartney came in that next year. And by the time I left, they won the national championship. So it was a, yeah, it was a pretty good run, but yeah, it's been kind of up and down since then so good love it uh what were

your favorite ski areas uh well my brother lives in steamboat which which i like quite a bit um and uh i mean as a being in in boulder was just really easy to get up to sort of winter park and mary jane and i was back in the day i was obsessed with modal skiing so mary jane was my my go-to uh that's a fun one i love it so not here to talk about skiing unfortunately but uh 10 dynamics you guys are doing some cool stuff in trend 10 different signals 15 different timeframes. I want to dig into that. You've got a bit of an interesting background too. Give us a little bit about how you got to 10 Dynamics, how you founded the firm and the financial background you have.

It's a little bit of a long story since I've been around for a while. I worked for a bit. As I mentioned, I was on the floor of the CBOE for a while but um yeah i i worked for a bit as i mentioned i was on uh on the floor with the cvoe for a while and then i went off and um i worked for chase back when chase was separate uh worked in their mortgage bank in florida and did some finance stuff and ended up getting a phd in in finance from cornell so i went back to school for for a few years and didn't kill yourself you didn't jump off the cornell bridge no i did not i did not but um it's yeah they they work you pretty hard there i'd say it's it's um it's a great school though i love cornell's it was really really good um i finished

my phd in 2000 and then uh one of my actual former professors from colorado uh worked at coke industries and convinced me to get into the energy trading business. And so, yes, I moved to Houston and worked. This is kind of even far back in time. This was the Enron days. And my first job was on the weather derivatives desk. So that was a very quanty sort of derivatives product at the time. And that's where I got my first exposure to energy trading.

Was Enron at the time like just the elephant in the room? Was it hard to actually get trades off and do things with them being such a big player? They were a great source of liquidity. I mean, I think they were obviously super successful in what they did, but they were really important for the liquidity in the market everywhere they went. So, you know, when I joined within a year, I ended up coming to London because Coke sold their energy trading business and put it into a JV with Entergy. And so Entergy Coke was this joint venture and Entergy had an office in London. And so when I came over, there were already, I think, 21 U.S. energy companies active in Europe at the time. And that was that was led by Enron. Right. They came in. They sort of structured the market.

They got liquidity. I mean, there were existing markets here, but they did a lot to to bring liquidity and change contracts and, you know, bring a help the market development develop. And that's why all the American energy companies were here. And then Enron went under, and within, I think it was two years, there were only two left, and it was us, Entergy Coke, and Sempra. So out of the 21, there were only a few suppliers, and that was just because both of us had higher credit ratings. So everyone else was kind of triple B+, and they couldn't survive the implosion. I derailed you, but Enron it has lots of questions right yeah yeah exactly um and then from from uh from there we got bought by merrill lynch and um uh worked at worked at

merrill and in the energy business and was head of at the time i was head of energy coke's uh fundamental uh trading or fundamental analytics group. So we had a very close, one of the things that we did that kind of Enron did too, was we had a very close pairing with a fundamental analyst with prop traders. And that was, it was a really good model at the time. And at the time, that's really where I started my trading career was developing fundamentals and trading off the back of that. So kind of bottom up fundamental modeling. And, you know, in 2000, it was a really great time to be doing that because there was a lot of information that wasn't available at the time.

So you could actually find information to beat into models that other people didn't have. And a lot of the models were pretty new. So you could do a better job of forecasting demand and supply for just on the pure modeling side. So, you know, we had an incredible run and I think some of those inputs are like storage numbers that people didn't know and things like, I mean, we had, you know, in Europe, you had things like hydro production and, you know, the governments would produce the hydro numbers, but they but at the time, there wasn't this sense of everything should be open.

So even though they would put it on the web, they wouldn't make it easy to find it. And they would only allow you to download it once every so many, like 24 hours. They had all these restrictions on getting the data. And so you could find data that was just because the market hadn't developed enough, that people weren't using or hadn't really thought of, particularly in the power market, the more granular you get, the more accurate you get. And so initially you just say, well, let's look at temperatures and try to forecast demand. And then, then you got to figure out, okay, well, yeah, there's a lot of different kinds of power production, but you've got at the

time there was hydro production, run a river, which is basically driven by precipitation. And so there were lots and lots of things that you could add to your, to your model to, to make it more accurate. And, um, uh, and that was, that was the super successful, uh, approach. And it was an approach that, um, led to eventually I ended up at, um, uh, at Deutsche Bank in 2007 and, and was ended up heading the European, uh, trading energy trading group there. And so trading power and gas and emissions credits came along at the time. And it was really that bottom-up fundamental prop trading approach. And one of the things that was clear though, is that more and more people were doing the same things and getting access to the same data. And so you were losing your edge, if you will, or you're losing a bit of diversity of opinions. You had lots of people coming to the same result.

And the result that was, it got harder. So, you know, I think if you look at performance or my career for fundamental trading, it's been in a steady decline in the commodity space. I think it's still, you know, some people are still making money there, but it's gotten way harder than it was. And you have to invest a lot more to get there to do it than we did at the time. So isn't that always the story? Nobody ever comes on the podcast and it's like, it's way easier now than it was back then. It's always the case of like, oh, it was all this info was just right there. You could grab it, make a lot of money.

Yeah, yeah, for sure. It depends, too, on what you're, you know, if you're trying to, particularly if you're going after mispricings, right? If you're going after mispricings, then, you know, regardless of whether you think markets are truly efficient or not, they're efficient enough that if enough people are doing the same thing, it's not going to work forever, right? So it's harder to find those places where you get mis pricings and um and so yeah so you know from there um you know i ran that ran the desk for for i guess about six years and um at deutsche for six years yeah deutsche and started

in 07 didn't their stock go down 90 probably right after you started it did it did i had one of my best years ever in 2008 as a desk we had we we used to joke we made more money than deutsche bank because the bank as a whole lost money like everybody did in 2008 and we had we had a great year so do you get netted out right like a lot of these people leave because their bonus is like oh sorry you did great but i'll tell you you know people make a big deal about being uncorrelated or particularly about being a hedge like let's say you run a hedge business that sounds great in theory, but the reality is when the hedges pay off, nobody has any money or nobody has any capital. So bonuses

don't go along with being the great performer when you're a hedge. And it really is. I've seen it happen in my career a couple of times where you have an outperformance when everybody else is not performing. And unless you have some contractual guarantee with somebody who's got deep pockets, you're probably not going to have a great year yourself. Right. So it's, it's tough. It's tough being in that, in that spot. So that's super interesting, right? It's like misaligned incentives because it'll create this thing within the banks where everyone wants to be on the correlated trade. So they get paid. Completely. You're way better. The people who were on riding the tide, they made way more money than the people who were on riding the tide they made way more money than the people who were there trying to you know be the be the ones who you know performed in bad times or or

and i'm sure there's lots of people with similar stories depending on what they were doing where they they were they had their one great year and yeah it didn't work out yeah a lot of the traders that become good ctas good hedge fund managers are right that was the trigger that got them out of wherever they were and said, yeah, this is crap. I made all this money. I didn't get paid. I might as well go on myself. Exactly. Exactly. Yeah. It was pretty interesting. Interesting time too, to see, you know,

just all the things that can go wrong. You know, I think when you're in the business for a while, you start to see, you know, how, how easy it is to be overconfident. You know, one of the things, we can talk about it too, like it was sort of the foundation of one of my philosophical principles is I am allergic to optimization. That's my, that's my saying allergic to optimization in all of its forms. So I really, we don't really optimize our system at all.

That's kind of one of the core things that we don't, we don't have any parameters that we optimize. And I think that's, that comes from seeing how easy it is. Even really smart people can fool themselves through optimization, thinking that they know a lot more than they do. And then in the case of 2008, correlations all go to one when supposedly they don't. Or you have things that never dislocate and suddenly they're 400 basis points apart when they're usually five you know things like that and um you know if you optimize too much and and uh you can really get yourself

in trouble and we saw that you know a couple times my career um you know we're sort of yeah try as hard as we can not to do that i call that uh statistical correlation and fundamental correlation right like yeah cool all this is statistically non-correlated. It looks great in the back test, but what are they actually doing? And is what they're doing going to become correlated in a crisis? You got to have both pieces, right? And I think that's one of the big difference, one of the big things you learn being in commodities that you don't get from financial markets.

Financial markets people that I meet, they are way more confident in correlations and covariance matrices and things like that. And, and I get why, because they're, they're, they are more stable. We're in commodities. You know, if you start doing bottom up fundamental and modeling, you have switching points. You have things where the correlation can be close to a hundred percent, and then you switch a substitution point and suddenly the correlation is negative. So you can't, that's why you don't see people in commodities doing the same types of things you see in like equity space and whatever with really relying heavily on correlations because they're highly volatile. So you survived the first downdraft in Deutsche. But then the second downdraft, you said? Well, I think it was, you know, you had the crisis, right?

And then everyone started to figure out what was going to happen. And just like I think the same thing happened, you know, around the Great Depression. It took like four or five years before the SEC and the CFTC and all those rules got put into place to change everything. So you just had Dodd-Frank and some of the other changes in Europe, you had limits on bonuses that are based on salaries. And so you had a lot of changes. And then the big thing was accounting. You had changes for how derivatives are accounted for in banks, which really increased the balance sheet charge that you got for trading. And energy was still a very prop business. Even inside a bank, we didn't have lots of customer flow. The one, again, big difference for commodities inside a bank is typically you're dealing with people who know more

about the commodity than you do. If you're trading oil, you're trading against VTOL and other oil traders who are as smart as you are, where if you're selling a financial product to a corporate that's hedging their, you know, price risk for some input, they probably don't know as much about the market as you do so that you're unlikely to be taken advantage of where, you know, in the energy space, if you're offering solutions to clients, a lot of times you could be on the wrong end of that. So, so it was really a prop business and as you know, prop trading and banks sort of went away for all the reasons, you know, everybody's aware of. And so then I left in 2012 and started with some partners. So the management team at the commodity business at Deutsche, our head of commodities, started Trailstone.

And myself and the head of oil and the head of metals, we all left Deutsche to join Trailstone. And so started a kind of new energy merchant in 2012. And I was responsible for the European business. And it was the same basic approach, bottom-up fundamental trading. But we had an idea, at least in Europe, our strategy was around assets because assets were super cheap, power assets at the time. So we wanted to, looking to buy power assets and then build a fundamental kind of trading shop around that. That means like owning the actual power plant?

Yeah. Owning the actual power plant. Yeah. That's a really good model when you can do it at the right time of the cycle. And that's one of the, again, I talk a lot about commodities energy, because I think it's such a great market, but one of the reasons why there's such good markets to trade is because of the huge investment cycle. So you go from oversupply to undersupply repeatedly. You can look back over the last, as long as you want to look and you just see these cycles of excess supply and then price tightness. And that is driven by the investment cycle is so long. If you want to build a new power plant because prices are too high, it takes you a while to do it. So, so you get these, these nice cycles and energy in particular that, that I think are, are kind

of what leads to, you know, as we'll talk about a bit, kind of the, the, the money that you can make in terms of trend following same, same sort of cycle that you're catching really. And then was that tough? You're competing with like Glencore and those kinds of groups of like big, huge. Yeah, we were a bit more niche because we were just doing power and gas at the time. So, and we did oil as well. We owned an oil refinery in the US for a bit. And so, yeah, we were competing

with kind of the big names in that space. Had our own, you know, one of the other things they did, we had a very renewable energy focus, which, which they still have at Trailstone. That's a big part of their, their business. Um, they've just been acquired, um, recently, but, um, so we were, we were, um, you know, pretty, I think we've successful over, over the time building that business. We didn't end up buying any assets though.

And, and that eventually led to kind of changing the business at Trailstone. And that led to me leaving and, and starting, starting 10 Dynamics. Trailstone, you guys were launching a few different managers and things of that nature as well. Well, Trailstone was, you know, we had a big prop trading team. And one of the things that over my career, I've seen many, many attempts at systematic trading and quantitative trading, usually driven by quants. So it's, you know, tends to be more the technical, and I don't mean technical analysis, I mean, more technical in terms of mathematical type of approach to that business. And it was never successful. And over, you know, many years and many different attempts and for various reasons.

And, and, and then we ended up hiring a team of systematic traders that had been successful over, over quite a long period of time. And they had a great track record and convinced us they could do it. And, and so we, we hired those guys into Trailstone and they were successful. But we realized- Kind of a pod shop before it was a thing? Yeah, I mean, we didn't really want to spin them out because we didn't know we'd need to.

We didn't realize how broadly applicable their strategy was, I think at the time and how much more capacity they could manage than we could ever fund. Because we were an energy trading company. We didn't want to be a systematic hedge fund, right? So we realized, oh, these guys, we could probably be better for them if, if, you know, we help incubate them and

spend them out. So we spend them out into their, their own business. But, you know, it was really influential on me because, you know, for the first time seeing what a good systematic team does and kind of how they approach it. And, and, and what I learned from that, you know, was, was that some of the things that I, because I had a PhD in finance over my career, both trading and managing traders, I built up a lot of tools that I was using for my own trading to help mostly around timing for fundamental type trading. Cause the, the problem you have with, with fundamental trading is in this is the sort of the bane of fundamental traders it's what we're into early which is i aka we got stopped out right so um and and that is a

thing because you you because at the end of the day most fundamental trading is counter trend so if a market's low you think it's cheap it goes lower you think it's even cheaper so you tend to like it the more it goes against you and so know, timing becomes, if you're not paying attention to timing signals, you end up, you know, getting stopped out or having big drawdowns or whatever. So I was using timing signals to help alongside the fundamentals. But then what I saw with our systematic group was that, you know, that the competition on the fundamental side had gotten so tight that actually the systematic systems were outperforming. On an average year, we're outperforming most people on the fundamental side. You'd have a great year of fundamental trading where you have some shift in fundamentals,

and you'd have an amazing year where the systematics wouldn't be able to keep up necessarily. But by and large, it was as good or potentially better in a lot of different circumstances than fundamentals. And so I've gotten to the point with these tools that I was using at the time that I thought all I needed was the tools. I didn't really need the fundamental inputs anymore. But I needed to make them more systematic because they were, you know, it's tough to do everything by hand. You can only watch one or two markets at a time. And, and to me, you know, the idea of, of being able to be diversified across a lot

of different markets is, is really valuable. And that was kind of the seed in my mind, I guess, to, well, maybe I should take the tools that I've gotten and see if I can put them into a system and apply them to a broader range of markets. So, and that was, and then that became 10 dynamics and your main 10 dynamics. Yeah. And so we're 10 dynamics is because, yeah, it was because we have, there's, there's 10 basic signals that I've, I've developed over time. And, and, and these are kind of old and tried and true. I mean, you know, the first one of these signals actually wrote in 1992. So it's been around for, for a while. And I have, my approach has been maybe which ones work, keep the ones that work and assume they're going to keep working and throw out the rest, right? Without any real connection to

the market itself. And because I had spent my career looking at the market and working with traders and that kind of stuff, I wanted a system that traded like I would trade. So as I was developing the signals, if I would get a signal that I would say to buy, even if it made money, I wouldn't want that signal because it's a trade I wouldn't take. So I wanted to strip out every signal that the system would produce that I would not believe in and not take. Hand on heart. So I'd like to try to do the, you know, put your hand over the chart and say, would I take this or not? And try to ignore what it was doing. Which is sort of the opposite of classic trend following, which is like, I have no idea why this market's breaking out. I don't care.

I'm just going to get in line. Exactly. Exactly. So what I wanted was a system that modeled my decision-making as closely as possible. So that's, that's really what we, we, we built was something that, that, and, and now I would say, you know, I used to think, you know, as I was developing the tools over the years, that there were times where the tools would say you know i used to think you know as i was developing the tools over the years that there were times where the tools would be kind of like ah you know maybe you should buy here and i'd be nah that doesn't that doesn't look right occasionally my intuition would be right but probably i would say it's been got at least six years since i felt like that i feel like now the system is always right if i'm trying to out guess the system it's a mistake you know it's a mistake. I feel like the system

is a better version of me somehow. It's taking all the trades I would take without any of the emotion or conflicting signals or other things that can influence you. Which is the point of systematic in the first place, right? Take all that away. Yeah. So that's really what we try to say, look, what we're trying to do is we're trying to model a prop trader is what we're effectively trying to do. Now, it's true. We're using prices to do that. And we're using, you know, things that, that aren't necessarily what, you know, like a fundamental trader would use, but it's that same concept.

So, you know, even how we do risk management, you can talk a bit of all that stuff, but there's, there's things that we do specifically that are modeled on how I think a good prop trading desk works and how you manage risk and traders and that sort of thing. So you wouldn't, do you consider yourself a trend model or no? Yes. Cause we follow the trend for sure. So we're definitely trend following. But we don't know, we can talk a little bit about the signals. Basically as a prop trader would follow a trend, not as a systematic would follow a trend. I guess in just the, yeah, I guess the thought process around it, exactly. So I used to have this way of looking at a market and saying, oh, or even looking back and saying, well,

here's a reasonable place where you should have gotten in. What signals would you get to get in there? You're not going to pick the bottom. You're not going to pick the top. You're going to pick some reasonable place to get in. Well, what would make you get in there? You know? And so kind of thinking through how the signal should work as a whole to try to identify those, those areas. And, and I started with a couple of philosophical things. One is my PhD. I was doing behavioral finance at the time. So I was sort of a big believer and, and one of the early, you know, this was, this was in the sort of, you know, early nineties. Right. And that was just at the beginning of the behavioral finance movement. So you were still a heretic and in the eyes of

anybody from Chicago, if you start talking about behavioral finance. Um, but, uh, yeah, I, I, so I did some research in that and, and, and really started to believe that markets at the end of the day are just driven by human psychology. And there's kind of, you know, underlying factors that push it one way or the other. And that, you know, I think, you know, unlike efficient markets where it's like the market moves exactly where it's supposed to go and then something happens and it moves exactly where it's supposed to go. And I think it's the opposite.

I think it's like, you know, we use waves as this analogy all the time. I think, you know, I think I always view it like a pond. And when a pebble drops in a pond, that's like news. It's news hitting the market is like a pebble dropping in the pond. And then there's this wave that propagates, you know, and I think that's the way markets really work is this sort of propagating psychology that repeats itself. And so that's kind of had that in mind when we were you know when i was developing the the signals over the years and the other thing that i i got experience with in in the 90s

um that was when sort of mental brought and and fractal mathematics was kind of a hot topic right and um and i i get interested in that and i also had in my in my PhD, I had a class from a kind of famous professor who's a big market pricing guy. And he did this thing that all, I think all efficient market guys do. They put up a chart of unlabeled chart of a stock or whatever, quote unquote stock. And they say, you know, can you guys guess what stock this is and go around the room and everybody has their guess. So it's whatever, Microsoft, whatever.

Oh no, this is a random walk. See how you guys were all fooled. A random walk looks exactly like stock prices. But then what I did is I went and I generated a hundred random walk paths. And what I realized was that was cherry picked. If you generate a hundred random walks, they don't look anything like stock prices. 90% of them look nothing like stock prices, right? So it was okay. Yeah. Statistically, it's great for kind of mathematical finance from a statistical point of view, but it didn't really look to me the way markets behaved. And then, you know, taking kind of the fractal idea and what that kind of stuff Mental Bar was doing, you generate price pass from that and it looks a lot like what markets do. And you're suddenly like, oh, wait a minute, this has to be somehow closer. And so I spent a lot of time trying to figure that out. I never figured anything out worth really worth using, but because the math didn't play out, but it worked in my mind

in terms of thinking about how the market is structured. Right. So the path was worth the, right. Yeah, yeah, exactly. Exactly. So, so kind of those two things like keeping call, it's not a, it's not a mathematical fractal, but keeping this fractal behavior kind of in mind, along with market psychology and kind of putting those two things together to come up with, with ideas for identifying when a market is behaving in a particular way. And in layman's terms, what you mean by fractals, it's going to jump, there's going to be jump points. Well, there are a couple of things for me that mean like fractals. One is that it doesn't matter what timeframe you're looking at, right? You should see the same type of behavior on a weekly chart as you do on a 30 minute chart, really. And there are limits to that. Like if you go into the really short term, then I think all bets are off there, it's, it's kind of a different, different arena, but, but in general,

you know, you should be agnostic on timeframe. And I used to have this with, with traders that I work with, you know, somebody come up and go, Oh, look, look at this 60 minute chart. If you would have bought gold here at this price, look at what the 60 minute chart did. It's like, Oh wow. It's amazing. And then they come, you know, two weeks later and they'd say, oh, look at this daily chart. Look at this daily chart. If you use this daily chart on this market, you'd have done an amazing. I'm like, well, when do you use the 60 minute? When do you use the daily? How do you choose what time frame you're picking? And what I came to the realization is you can't. There is no way. I don't think there's any way. So maybe they're all right. So maybe all think there's any way so maybe they're all right so maybe all the

time frames are right maybe they're all telling you something different in a different um level of abstraction within the the market right so so that was kind of the the basis for why why we use all the time frames at the same another way to think of that is we can't know which one is right so we might as well participate in all of them so we're sure to get the one that will be right yeah and i think it came back from you know in my in my finance career you know you have this have you heard the the the fundamental law of active management and that kind of stuff which basically basically you just want to if you have an edge you want to bet as many times as you can like make it as broad as possible and that kind of idea made sense to me that if i if i'm a system that works on every timeframe, why not

bet on all the timeframes at the same time? And if it works in one market, why doesn't it work in all the markets? Why shouldn't I be betting in all the markets at the same time and trying to, you know, really make as many different, um, trades as you can because you think you have an edge, right? And that that's that was the i guess underlying basis for for 10 dynamics and how we got to got to where we eventually got to i'll real quick the far end of that logic right is that you'll just earn the t-bill right that you've removed all the risk and you just get the risk free right if you did that across the world in every single market every single time frame are you just getting the risk-free rate?

Well, I don't know. It depends on if you can lever that or not, I guess. But what you get is you get an ever-increasing Sharpe ratio. And that's what we see in our portfolio. The more markets we add, the more assets we trade, the more the Sharpe ratio goes up. So it is kind of that, you know, I guess implementation of that idea. Talk to me a second about these timeframes. So we're talking what's the shortest, what's the longest? We started a 30 minute timeframe and we go all the way out to weekly basically. But then we divide the market

onto lots of different time frames. So we don't stick to necessarily what people would consider the normal time frames or whatever. We just want to make sure we have enough time frames that are different enough. If you divide it too much, then you end up with signals that are exactly the same across a lot of different timeframes. So you don't get any diversification there. But we've kind of arrived at like between 12 to 15 timeframes, depending on how many hours the market trades. Because we do trade down to intraday. So it could be some like 1.75 days timeframe, right?

It's like custom. Yeah, yeah, exactly. And even, you know know you can even get different looks we haven't really done this with some of the things i think you know everybody just uses the clothes for their daily chart well why not use 12 o'clock or why don't you use three o'clock you know get in before everybody else here you know there are some different behaviors around the clothes but you know there's different ways of if you think about you know the the difference between kind of precision and accuracy i'd rather have

three different charts that are all slightly different giving me the same trade right but if one if if one out of the three is giving me something different then that makes me think is there something weird about just that very special pattern that you know so so again to me it's like having lots of different signals give you more robustness. Is it like a voting machine? So you have seven say buy, six say sell, I'm long one? Yep, exactly. And it's netting out or you place it all 13 trades?

No, no, no. We just net it all out. So we boil everything down to, I mean, our process is we have a risk. We basically start at like what volatility are we targeting for a given account or fund or whatever. We start with annual volatility. We back that into a daily volatility target. And then we scale the portfolio to try to achieve that target. And that amounts to allocating a risk budget to each asset. So you might say, okay,

to hit a 10% vol, depends on how big the account is, but let's say you'd say, okay, well, we want to have at most 100,000 risk when we use VAR. So 100,000 VAR on say corn futures or whatever, then we can back that into a number of contracts based on the volatility. And that tells us our maximum contracts. And then what our system does does it just scales between those based on all these different things moving around so um so we're always scaling in and out um so we have our 10 signals that we apply to each time frame independently is it ever flat or it's always scaling at some it can be flat sure sure but it'll but it'll pass through that So we tend to scale in increments of five to 10%. So we will take all 10 signals and apply them to a given timeframe.

That'll give us a position on that timeframe. And then we take the positions on all the different timeframes and add them up. And that'll give us kind of between minus 100% and plus 100% type of number. And that gives us our scaling for the position. So the model is just constantly sort of scaling in and out based on how the confident, I guess it would be, we call it a confidence measure on the trend. How better than a confidence game. Confidence measure.

Yeah, exactly. Training might be the same thing. We don't know. How often is it 100%, right? That everyone's voting in the same direction across all the timeframes? Yeah, I mean, it happens in really strong trends. Obviously, you know, Cocoa has been the darling of the early part of the year and Orange Juice. So there you were, you know, you were running at 100% for weeks at a time.

But overall, I'd say it's pretty evenly, it's not quite, it's definitely not a normal distribution. So it's something like the tails are probably 100% long or 100% short, maybe 8% to 10% of the time, something like that, and then kind of up from there. So it's, and, you know, it can be flat for significant periods of time, too. It just depends on how the model is looking at the different trends and the different time frames and how it's all netting out. And then how many markets? So it's not just energy. Energy is your root. No, no. So we do all the futures and all the liquid futures, I guess, in the U S we do European energy futures.

We do some financial futures in Europe. We do, you know, treasury bonds and notes and we do currency futures equity indices. So pretty much the whole, what I guess would be the traditional sort of broad CTA kind of space. We also do, but we also do on the equity side, we do large cap single name equities. So we'll do, you know, kind of the big, the big companies in the S&P 500.

We also do ETFs. NVIDIA? Yeah, do NVIDIA. And the system, what's interesting about the system, and this is probably true of all trend following in the equity space, it's kind of what you'd expect, right? That the more behavioral something is,

the better it works. So the big like tech companies and things that are hot, so your high volatility, your large caps that are really moving, system works great. When you get down the small cap space, you're basically talking about beta and earnings announcements, right? So that's like jumps and sort of beta, know is not what you're trying to capture so so the system doesn't really work for for smaller cap stocks because the the dynamics are different um yeah see i know jerry parker chess speaks big on single name equities but the rest of the trend world has been hesitant to go there for whatever reasons. Yeah. Yeah, it works well.

The hardest part is talking to investors about it because it's not normal. And they already have that exposure, they think. Yeah, they think that there's a lot of things that go into it. A, if you're talking to the big allocators, they have teams. They have a kind of CTA type team or systematic team or quant team. Then they have a market neutral equity team, and then they have a long only team. And then when you say, well, we got a model that trades equities, it can be anywhere from 80% net long to 80% net short. So you've got beta timing in there, which equity guys typically don't like, right? Their head explodes. Yeah, exactly. So it's a tough, that's been a tougher education hurdle, I think,

than, you know, on the future side, people are much more used to kind of what we're doing. So it's... So talk about the allocators for a second. Where are they putting you in that trend bucket? Do you guys correlate with the trend indices and whatnot? We do. We're probably, I think we're about between 40% to 45% correlated with trend, something

like that. So there's definitely trend exposure there. I think it depends on what we do, what markets you include, because if you want to get rid of trend, you can take out some of the markets. I think one of the things that adds to trend a bit is we trade rates. It's a rates exposure that gives you some pretty significant weight on the trend at the end of the day, but also equity indices as well too. Do you allow investors to do that? Like, I just want the commodities? Yeah. So that's one of the things that because we have, we didn't really talk about it,

but we just mentioned how we don't optimize. We use the same model for everything. So that means that we're happy to customize to any portfolio. If somebody wants to do a managed account, if they have enough capital to do it, they can customize to whatever exposure they want. And that's what our current managed accounts do. So we'll have people say, we just want these markets, whatever it is, ags and energy, or we just want, we'll take all the futures or we don't, we can even do just equities if we want to do that. So we're happy and kind of set up to be able to manage the different benchmarks we call them.

And then on the risk side, you will dig into that for a minute, but do you view it, that's probably where your prop trader routes help the most, right? In my experience, prop traders hate losing even for 10 minutes, right? Versus systematic guys will be like, no, it's in a 30% drawdown. That was in the back test. It's all fine. So, right. Is that part of what you built into the model of like? Well, yeah. I mean, for sure. Some of it. I mean, one of the things which we kind of talked about is the system scales in and out. So one thing that we don't do is we don't try to make sure that the portfolio is hitting a specific vol target every day, right? Because sometimes it could be in an environment where there's not a lot of trends. And so one of the things that we learned from... Yeah, don't put on risk just for the sake of putting it on. Yeah, exactly. And you could see it, you know, with, it's one of the, the kind of psychological challenges for being a non-systematic trader is when do you know something and when do you think you know something?

And the reality is if you're going to be a fundamental guy, you know, generally big fundamental changes in a market might happen once or twice a year if you're lucky. And, and if you're a good fundamental trader, you make a lot of money and you tend to outsize your bets when you're confident like that. But then what happens? Well, then what happens is nothing. You think you have an idea and then you get bored. And there's very few traders who can make one trade a year and then sit on their hands

for the rest of the year. And usually what happens is they make a lot of money and then sort of bleed for the next six or nine months until the next big idea comes along. And so we built the system specifically so it would scale like when the market, when the model doesn't have a lot of confidence within an asset, it will be close to flat. And when there's a lot of assets that there's not a lot of signal, you'll be close to flat. So, you know, for a 10% annual vol, for us, in our VAR model terms, it's like a 1.5% daily VAR. That's kind of our target VAR. But our VAR actually ranges, we could be at 50 bps.

So we could be at less than a third of the target VAR for weeks or even a month. But we can also go as high as 2 and a half to 3% VAR. So our VAR range is actually 10 X in terms of, you know, we're low deployment versus high deployment. So, and that definitely adds value. So I I'm, I'm amazed that people can scale up their systems all the time and have it be successful as if they allowed it to sort of reflect some sort of confidence in the signal. And, you know, we can do it in our, if we want to, we can force our model to be fully deployed. And what we see is it deteriorates the returns pretty significantly.

So that being fully deployed when you have confidence and not deployed when you don't, I think is a key lesson from the prop side that works in our system for sure. And it's funny, how do the investors feel about that? I've seen plenty of times where an investor quits an investment, a CTA, because it's flat. And I'll be like, well, its peers are all down 10% to 20%. It's like, yeah, but I just want it to be doing something. I want it to be.

So I think it's not just the trader, it's the investor too who hates seeing just nothingness. For sure. And I know for a fact a couple trader, it's the investor too, who hates seeing just nothingness. For sure. And I know for a fact, a couple of the big multi-strats, if you go work for a multi-strat, they force you. Like you need to be, they want you to deploy risk all the time. And if you don't, it's not the shop for you, right? So there are people like that.

And not every shoe fits every foot, right? So that's not our approach. And usually people like that, they're looking for, I don't know, looking for something different than we are, right? Our whole thing is we want to deliver the best risk-adjusted returns that we can, right? And that's without benchmark

and without where other people have correlations that they're worried about or they want different things out of their their investments so talk to that for a minute since you've kind of been on all sides of this and have some insight into the like do you think you can doesn't sound like you're trying but do you think people can recreate a pod shop multi-strat right with systematic or quantum mental is the new term, right? If you can create that same kind of look and feel and return profile with models instead of actual traders.

Well, so our experience at, at trail stone was that, um, our systematic guys were slightly negatively correlated with our prop guys, which in a way makes sense because our systematic guys were trend following. And as I mentioned, you know, fundamentals by nature are counter trend. So you've got guys doing counter trend trading alongside a trend followers. What does that mean? That means in market extremes,

you tend to be flat as an organization, and then you tend to make good money in the middle, usually. So, and it's super efficient from a, from a firm level return on risk. So could you replicate, so, you know, and I, again, my, my experience with systematic is purely on the, you know, we are trend, trend following. I don't, and this is a, this is a, I think it comes down to your philosophical makeup. I, I really have a hard time trading countertrend. I can't do it. But people who are quants, they tend to love countertrend. It's cheap. Look at the model says it's cheap. Let's- Yeah. If you like it at 40, you'll love it at 20. Yeah. And that's why you see a lot of quants doing mean reversion

strategies and pairs and all that kind of stuff. That's, that's really, you know, that, that kind of mean reverting idea. And that's the problem that I really have with mean reverting strategies. They tend to be negatively tailed, right? So you have negative tails and distribution. And again, that goes so against my, this goes to the prop pair thing, not wanting to lose, right? The idea of having a tail event. So our, our returns are positively skewed, right? And, and that means that at the end of the day, we probably lose on more trades than we went on, but our winners are far outsized from our losers. So that's how we, you know, that's by design, right?

Yeah. But it sounded when you were talking about building the model, you sounded more like a quantamental, right? Of like, I'm taking everything I know from being a fundamental trader and putting it into the system, which is what this new quantum mental is trying to do right of like hey now that everything's digitized we can take all these inputs that used to just be the trader and a bunch of screens and using them put it into a model and trade off it well i think

the the quantum mental combination works really good for mean reverting systematic with fundamentals. Because if you think about, it took me a long while to sort of figure this out, but let's say you've got a system where you lose on nine trades, but you make a ton on one. So you're going to, you have this sort of classic trend. Now you're going to use a fundamental of classic trend yeah classic trend now you're gonna use a fundamental filter to say well i want to i only want to trade when the fundamentals are in my favor because that's going to eliminate all my losers the problem is if you eliminate that one winner by accident you turn a profitable system into not making anything right? So it's quite risky to combine a filter with a trend following strategy

where you have a few winners because you can easily wipe out the winners. And if you do that, you kill your returns. Right. The other way around, if you have a system that wins on nine trades and only loses once on the 10th, and it's a big loss.

Yeah. If you're trying to filter that out. Negative skew, positive. Yeah. If you can filter out the negative skew, that's valuable right so so that kind of combination of of mean reverting and and fundamentals works pretty good because you're going to filter out those big losers and if you can do that that's a good combination right where i think it's it's a lot tougher if you're if you're in the trend following space to to make that combo yeah which is most

trend followers like hey we're taking all the losers but they're small right they just they know that it's an option profile we're buying all these straddles buying buying buying and then boom it's going to pay off eventually exactly exactly um so what's next more more markets more stocks yeah we just well we actually interestingly just today, I just finished, we had an investor who was interested in the Chinese domestic futures. And one of the things I really love about our model is how we look at backtesting. For us, backtesting isn't anything to do with looking back with the markets. It's give us a new market.

We've never seen it. Let's see how we do. So yeah, so we took... Your perfect out of sample markets. It's give us a new market. We've never seen it. Let's see how we do. So, yeah. So we took. Your perfect out of sample test. It really is. Right. So, you know, so I just took out of the box, you know, the same model we use for anything.

We loaded up the sort of 30 plus futures contracts, domestic Chinese futures contracts and ran it. And we get basically the identical sharp ratio that we get in the other global futures with less correlation to the trend factor. So it's like, oh, well, this looks all right. So we did the same thing a few years ago on the crypto side. Somebody was interested in crypto and we said, well, let's have a look and crypto works fine. So the model seems to be pretty robust as long as it's a normal market. And one of the things we don't do is we don't cherry pick the portfolio. So, I mean, Coco was a good example. And I don't know if you've messed around with any trend following

stuff, but I can remember doing some trend following testing in the 90s. And the one market that stood out as really difficult was Coco. Coco was really not very trendy. And I think it has not been trendy for 30 years. So if you would have not put it in your portfolio, you know, you wouldn't have caught this trend. Which we've seen. I've seen that in a lot of portfolios. Exactly. So, you know, as long as a market is a normal market, just because it's never had a big move doesn't mean it won't. So we tend not to throw things out of the portfolio as long as they don't, you know, behave in a way like, like are very jumpy. So the model doesn't do good with jumps right and no trend following i think can really do good with jumps because it's so like power

doesn't work and no no power does so this is this is one of the other funny things that we've you know given my background everyone who's a power fundamental analyst like oh trend following can't possibly work in power and it's like well it actually works just fine doesn't seem to have a problem what doesn't work though is like we do some commodity spreads. So, so things like Brent TI that sort of move over time or gold platinum or things like that, that are, that almost behave like a normal market. Those are great. What doesn't work is spreads that are tightly linked that then tend to have dislocations. So like certain regional spreads will be very tight. Like what are the, there's the wheat, wheat spread in Chicago. What is it? The Minneapolis versus Chicago. Yeah, tight. Like what are the, there's the wheat, wheat spread in Chicago. What is it?

The Minneapolis versus Chicago. Yeah, exactly. So you wouldn't, you wouldn't trade that with trend following, right. It would, it would,

it's kind of a jumpy sort of shot. Cause it's all, it's a nice straight line until it isn't. Yeah, exactly. Exactly. So we exclude anything that's, that's too kind of like that.

That's that doesn't move, doesn't move or, or, and we also exclude anything that's really too illiquid to like that that's that doesn't move doesn't move or or and we also exclude anything that's really too illiquid to to follow but um yeah so we tend to be we don't want to i don't like the idea like investors will say well what you know if we only wanted to trade 20 futures what would you recommend i'm like yeah you you tell me which ones you want to be in because i can't pick them i don't know which i don't know what what's going to be coco next year yeah we did a blog post we'll put in the show notes i think silver it was maybe five years ago but we it hadn't made money on a trend trade in 15 years

or something yeah it was like 30 losers and then had this huge winner it's like well that's that's what you have this one exactly to your point if if you try and silver doesn't work, we're leaving that out. Yeah, exactly. And volatility is a bit like that. So we trade the VIX just from the long side, because again, we don't want the negative tails in there. And so we trade VIX and some of the volatility related ETFs and we do make money with them, but it's not over time. It's not a lot of money, but it really pays in certain times when, you know, trend in particular is getting smashed. Right. So it's like the hedge benefit. If you can, if you can have a market where you're not steadily losing money, even though this is, it's not trending, that's a great market, right? It just tells you when it does trend,

you're going to, you're going to make a lot of money in it. So, um, so yeah, we tend not to, and then you try to cherry pick that way to that point do you have this crisis period type profile you believe you will in a 08 type long drawdown a 22 type drawdown yeah i mean one of the best years for the for the model was during covid um it was a fantastic fantastic year for the model um where everything was was trending dynamically and volatility was paying off and all that so 2008 would be would be good as well because you had to remember back then you had huge commodity trends right oil went to whatever it was yeah yeah i get i've gotten in that debate on twitter

and elsewhere like you'll people will think about a tail hedge and be like okay i'm willing for it to pay 200 basis points a year whatever for this hedge. But then they look at a trend follower and it's only, say, it averages 3% a year positive averages. I'm like, I don't like that. It only, I'm like, but it's providing that profile with a positive carry. If it can hedge with a positive carry, that's the whole rail, right? What are you doing? Yeah, I know.

That's why I get it. I know they don't get it. Like our, you know, we don't sell it. We don't, I haven't talked to anybody, but just it. Like, you know, we don't sell it. We haven't talked to anybody. But just our volatility portfolio, it'd be a great thing to add. Like I said, it doesn't cost you any money. If things, you know, go derail again, it'll be there for you. And why not do that in the VIX futures versus the ETFs?

We do both. So for people who just want futures, they'll just get, obviously, the futures exposure. We do the ETFs? We do both. So for people who just want futures, they'll just get the, obviously the futures exposure. We do the ETFs as well. The other thing that's interesting, I don't know if other people do it, but the triple levered sector ETFs like ERY, which is the energy triple, that actually acts exactly like the VIX.

It decays through time because the trend is, because of the costs of being triple levered, plus the fact that it's generally upright, but then it spikes. And so our system works just as well in triple levered sector products as it does in the VIX. And they tend to spike it. Well, the nice thing about energy, it has different spikes, right? So you get different chances to capture spikes.

So it's kind of an interesting thing that we found. Anything else we should know? Well, the one thing that I think is interesting, and we mentioned a little bit about if you're trying to capture mispricing, I don't think that what we're doing is capturing mispricing. I don't think that what we're doing is capturing mispricing. If I look at our, like, if you can look at systematics as two types of camps, you have the alpha decay people. And if you're an alpha decay person, if you believe it's constantly decaying, then you're constantly re-optimizing to try to catch whatever's changing to catch that alpha decay, right? And that's your system. So that is a mis, to me, that's a mispricing basis. But for us, we're using the same system. And if you look back to the 90s with the same

system across the same markets, you don't see any alpha decay at all. What you see is clear cyclical behavior in either how you look at, whether you look at Sharpe ratios or alpha, we tend to look at Sharpe ratios. So you have years where the Sharpe ratio is really good. And then you have a pattern of years where the Sharpe ratio is not great. And you think about what, we've kind of dug into it, what's driving that?

Well, from a trend following basis, right after 2008, what happened? Well, the economy kind of contracted across the globe. What did that mean? That meant you had oversupply in almost every commodity market, which meant you had downtrending or chopping around messes, right? So trend following during those years on a global basis wasn't working very well because of what was happening from a macro point of view.

So there's this tie back to what systematic is catching. It's catching these big investment cycles in energy, let's say. It's also sensitive to these big macro changes. There's times where trend following works well and times where it doesn't. That's not mispricing, right? And people want to argue that it's a factor. But if I look at what people call trend, the return on the trend factor or momentum factor, it's not great, right?

The Sharpe ratio is not great on that. And the good trend followers are delivering is much higher. So I don't know what it is. Is it smart trend? Is it smart? I don't know what it is. I just know that it's not, it doesn't look statistically like mispricing and we're not chasing alpha decay. So I don't know what it is out there, but there's something to what we're doing. That's, that's kind of core to the way markets work. You rang my bell of what I was going to ask, which ties into what you just said. Do you think it's because of the multiple timeframes gives you better monetization, right? Like are you getting X in cocoa at a better place, for example,

or are you doing things like that because of the multiple timeframes and the multiple- For sure. The multiple timeframes definitely helps, for sure. I think what really happens in our model is the long-term timeframes generate the overall positioning and they generate the overall return. And the short-term timeframes are basically like hedging. They're like smart stops, basically. So they're stopping you out of longer-term positions

when the short- term is turning against you. And so they work kind of collectively together really well. And the weeks long. So you're using basically weekly bars, as we would have called it back in the 90s. Weekly bars. So that trend may hold for 18 months or something like that. Yeah, we look at two different kind of measures of our holding period.

If you look at when we adjust our position from, say, 10% long to 20% long, that typically lasts about two weeks. So that's a relatively short holding period. But if you look at how long we're net long or net short, that's much longer. That's closer to two and a half, three months on average. So, you know, if you look at, you look at, you know, I was using the example of Apple, you know, Apple stock, it was basically an uptrend for how, I don't know how many years was that? It was an uptrend for a long time. So the system was never net short Apple for, you know, eight years, but it was also flat a few times. So it got to flat and then it got to a hundred percent long and flat. So you're, you kind of have this long position that you're always just sort of Delta hedging it based on what the

short-term is doing. And there seems to be some nice synergy there. That's another cool way to look at it. Delta hedging these long-term positions. Exactly. Exactly. Awesome. Well, I think we're going to name the podcast Allergic to Optimization. I like that line. Cool. All right. Yeah. Well, thanks for being with us. Where can they go find out more? Website? Yeah, we've got our 10dynamics.com. We've got an inquiries tab. You can reach there. We also have a lot of content on there now. We just put out a white paper on how we handle risk management and operational risk from a systematic fund which is really important one of the things we've kind of learned by doing i would say we hadn't really and i think a lot of people don't don't think about it there might be some interesting stuff there for people how do you view that operational risk in terms of getting

your signals off you mean well just all you can think about if you're trading systematically, you know, and you're really doing it from where you're basically the system's doing everything from selecting the trades to automatic execution and everything else. If you have a problem, there's a lot of ways that that can go wrong, right? So how do you, what's the best way, what's the best framework to make sure that you're managing that risk, but also that you're not just like stopping the system because one little thing went wrong and then you shut everything down and, you know, how do you keep from doing a thousand trades in a day and, you know,

all this sort of stuff. And I think we came up with a pretty, I'm pretty proud of it, I would say in terms of what we do. And the system really runs, really runs well now. And we have a whole framework for how we how we manage that so i think did you read the uh jim simon's book i can't remember what's called the uh i think i have it right over here the man who solved the market it's about renaissance and jim simon oh yeah somewhere in there early on they were losing money with this model and he couldn't figure it out and they had basically the sign flipped i

think for, for like multiply it by the value of the E-mini, for example, or something. And they were like, Oh, and switch that. And it just started going. Yeah. So yeah, that's, that kind of operational risk that's more model risk, but. Yeah. Yeah. There's all, all that plays into it. You've got, you've got basically your pure risk management in terms of, you know, any limits or metrics that you have. Like we have, we have limits and metrics that we manage.

But that goes alongside operational risk, things like you have a database and then you need the data from the database to feed into the model. Well, what happens if the database gives a missing value for something, then, you know, what are you going to do? Are you going to shut down the whole fund? Are you going to, you know, just isolate it and, you know, can you solve it automatically and all those kinds of things are

things that we we sort of dealt with over the years so yeah if na do what yeah yeah exactly exactly yeah yeah and we're lucky we're um we're lucky we are where we are today in terms of technology because i think it's way what we've done we couldn't have done with the time and people that we have. It would have taken so much more 10 years ago than it did today. And just even having Python to do everything in this process, this this operational and market risk management process that we have. Python made it conceptually much easier to deal with, I think, than had we been, you know. And did you use some of did your team use some AI tools

to even write that Python? Yeah, that's kind of new. Most of what we built we haven't. We do use it a little bit every once in a while, but it's interesting. It's very good at writing boilerplate. Yeah. Like, you know, give me a function that does this.

Okay. But most of what we do, it gets really confused really easily. Not boilerplate. Yeah, it's not boilerplate yeah i i keep i use it to write some market commentary stuff for some groups and it can't get past and i even have it in the script don't make this error but right if the smp is down 1.5 and the fund is down 1.8 yeah it'll say that the fund did better than the s&p oh really yeah i'm like no it was more negative more negative it's not better yeah and you even have it in the script don't make this

more negative mistake in an example it keeps doing it yeah exactly so yeah it's not quite there right so you don't want that in your trading model to be like whoops no no and we've we i guess we're as just philosophically we're as far away from ai as you can get given that we don't do any optimization so we're i don't know if we're dinosaurs and don't know it yet or what but we're fighting the fight yeah i i have some personal worries about that that as more people coming into the space have grown up so to speak on the ai they're just gonna unknowingly optimize by using the ai yeah well and it goes even further to you know what i mentioned about how the traditional approach to at least

that i've seen with technical trading was come up with a thousand different indicators with a thousand parameters and see what what with no connection to the market or what the market's doing or how the market trades or what's a pattern that makes sense you know and this kind of stuff so and then the next iteration of that though was is okay. All of these don't work. I'm going to trade the ones that are working now and then rotate, right? Like you keep moving the ones that are working to the top and moving the others to the bottom and the others can come back in. Yeah. You have a thousand AI traders and then you have one AI trading manager deciding which one of the AI traders to bet on, you know, this is where all this is going to go.

I don't know. And then you need the AI to write the investor letter every month because no one will know what happened. Yeah, exactly. Exactly. All right, Bill. Thanks so much. Good talking to you.

Thank you. Yeah, appreciate it. We'll look you up when we're in London. You going to go to the Bears game? Actually, that's not a bad idea yeah i haven't seen a bears bears game since the 90s but yeah they'll be there at tottenham stadium i believe against jacksonville yeah the problem they always have here is the um the football as they call them the football pitches here they're not used to such heavy

guys and so the turf is like every time they have time to seem to wembley the guys are like slipping and sliding all over the place because they're just tearing the turf up so hopefully they they sort that for the game but yeah they're big yeah uh awesome good talking to you all right you too jeff see you take care yep thanks bye okay that's it for the pod. Thanks to Bill, thanks to RCM, and thanks to Jeff Berger for producing. We'll likely be off next week. Have yourself a good Labor Day weekend. But back soon after that. Peace.

You've been listening to The Derivative. Links from this episode will be in the episode description of this channel. Follow us on Twitter at RCM Alts and visit our website to read our blog or subscribe to our newsletter at RCMAlts.com. If you liked our show, introduce a friend and show them how to subscribe. And be sure to leave comments. We'd love to hear from you. This podcast is provided for informational purposes only and should not be relied upon as legal, business, investment, or tax advice. Thank you.