The Derivative - Trend following. The dynamic systems & strategies that make up a modern CTA, w/ EMCs John Krautsack

Episode Date: February 17, 2022

From his days as a clerk in the S&P pits to becoming Chairman and CEO of EMC Capital, John Krautsack is here to tell us about his journey in the futures industry — and nothing is off limits! From his wild west pit days to working for one of the original turtle traders, the late Liz Cheval, to what it's like implementing a new vision and everything in between. With their Classic program having been around for 40 years, John gives us a glimpse into EMC’s recipe for long-term success. We're talking about building robust systems for trend following (EMC is featured in our latest Trend Following Guide here), various strategies, automated research, AI and machine learning, plus more. Hold on to your seats; this episode will take you on an adventurous ride! Chapters: 00:00-01:28 = Intro 01:29-07:40 =Crazy Pit Days 07:41-17:29 =An original Turtle & Secrets to a 40 year run 17:30-46:16 =EMC Classic: Building a Robust set of Systems 46:17-56:14 =EMC Alpha, Alpha+ & Machine Learning 56:15-01:08:37 =Trend Following Future, Bond Trends & Trending Lumber 01:08:38- 01:12:12 =Two Truths & a Lie Before you go, check out these items mentioned in this episode: Blog post: Liz Cheval: From Turtle to Titan Podcast: Trend Following Turtle Tails (and Tales) with Jerry Parker Whitepaper: Newly Released Trend Following Guide About John Krautsack:  John directs all investment activity at EMC and started his career in the futures industry in 1985 as an assistant to a prominent S&P 500 trader at the Chicago Mercantile Exchange. From 1989 to 1995, he managed trading operations for De Angelis Trading/Crown Capital Management, JPD Enterprises, and ALH Capital. He joined EMC in 1995, overseeing trading and managing the portfolio until he assumed the role of Chairman in 2013. Don't forget to subscribe to The Derivative, and 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. with bonds selling off and commodities taking off. And we've got John Krautsack, chairman and CEO of EMC Capital, who's been trend following in one manner or another with clients harder than money since 1985. I was 11. Wow. So how do you stick around that long in the cutthroat hedge fund world? How do you stick with trend following

through all its fits and starts over the years? We're finding out. Send it. This episode is brought to you by RCM's new white paper on Send it. papers to download the newest paper where we get into what trend following is, why it tends to work in inflationary environments, why commodity exposure mostly sucks, unless you do it smartly via trend, and highlights on five trend managers we recommend. Many of them have been on the pot here, so you'll recognize the names. So go check it out today and let us know what you think. Now, back to the show. Okay, we are here with the main man at EMC Capital Advisors, John Krautzak. How are you, John?

Hey, good. Good to see you. Good to see you as well. Yeah. Hopefully, we're warming up here. And you're in the northern Chicago suburbs, right? That's right. Yeah. Hopefully we're warming up here and you're in the northern Chicago suburbs, right? That's right. Yep. So hopefully we're done with the winter. But I think it's supposed to snow like a foot on Thursday, right? Geez, I didn't hear that. Yeah. Yeah. So let's start off a little background.

So you started as a clerk in the S&P pit. Is that right? Yep, that's right. I had a good friend of mine whose brother-in-law worked as a friend of mine whose brother-in-law was a big S&P trader, a big local trader in the pit. And he asked if I wanted to go down to the mercantile exchange with them. And so I headed down there I headed down there with them and, um, you know, we went down and we got onto the floor. I've never been on the floor before in the, this is in the eighties. Um, and just total chaos in my eyes. Um, his brother-in-law comes rambling out of the pit. We go up to the Merc club for lunch. The guy's not even paying attention to us he's looking at his cards looking at the ticker um finally he uh he looks across the table and he tosses me his trading cards and he's like you know count my position he's like buy side on one side sell

side on the other so i go through his cards and I'm like, you know, you're long 80 S&P, big contracts. And he just, he looks at me and goes, you want to work for me? I wasn't even looking for a job. I just was going down there to meet him. So it was a really, a lucky start into the business. Yeah. For those watching on on youtube i'll show a little right this was basically the trading card just a simple two by two grid right and then i bought

five i sold ten i bought five um and then i was a clerk down in the bond pen and then our job was to go find the right to match up okay you did You did those 10. Make sure, get the other guy's little piece of paper. That's right. And because he was such a large trader, my role was to literally get into the pit. And I was probably one of the only clerks that was in the bottom of the pit for the whole day, just getting cards from him and giving him his position because he was trading. He just traded such huge size. whole day just getting cards from him and giving him his position because he was traded he just

traded such huge size he needed a count like continually all throughout the day and then i'd i'd move about in the pit and match the trades with guys who's you know who traded with them because he was just so crazy and tell the listeners if you can what they uh what that trading floor was like. I always say people would lose their minds if they saw the amount of paper on the floor, which we just talked about a little bit, but, um, how do you describe it to people who, who never got that experience? Yeah. I, you know, I, I kind of equate it to, you know, a, a sporting event where, you know, the bigger, the bigger the guy is, the stronger the guy is.

I mean, you'd have so many people packed into these pits that if one guy swayed, the whole row of guys would sway. And, yeah, and paper all over the place. The phone clerks would be shooting cards out to the brokers. You know, it was brokers. It was crazy. I know the first couple days working down there, I was so exhausted coming home on the train. It was just really intense. Yeah, I always said I didn't make it.

You'd get like dip spit spilled on you, fist fights every day. It was it was the Wild West to be sure. It really was. What you got any what was your craziest story from one of the on the pit days? Well, one of my one of my favorite stories is there's this one big local trader in the pit and he would just over trade and just get himself into, into trouble. So one day he shows up in the pit and you know, if you want to do big size a hundred lots or whatever, you got your hands over your head.

So he showed up in the pit and he had a shoestring and he tied his wrist to his belt loop. So he couldn't get his hands over his head to, to over trade. And I just thought that was so funny. I mean, just all sorts of crazy things. But he could still do a bunch of fives with the one hand. He just,

yeah. Yeah. Yeah. So there's so many stories down there. I mean, that was, that really was was a it's wild that, you know, business gets accomplished down there. It's just, you know, so much chaos. It doesn't seem up at 6 a.m. to go meet the other clerks and settle trades, out trades. Little did I know, like there are millions of dollars, right, of like, oh, we're missing a 20 lot in bonds. And it's from two basis points higher or something. Yeah, I remember right before Black Monday, like that morning.

My the guy I was working for, he was he was out a whole bunch of contracts and I was, and he wasn't even in town. He was, he was in Arizona and I was running around trying to square him up before that opening and that chaos. And, you know, giving myself a heart attack doing it. Right. I finally got ahold of him, and he left a couple of his cards in his jacket pocket. He didn't turn those cards in, so he was square, but I didn't know that. So you were running around for no reason. And so moving on, a little bit of the firm background so emc was founded by elizabeth chaval uh who's one of the original turtles right that's correct yeah and so we've

covered the turtle program on here before uh we'll put a link to our pod with jerry parker in there you could learn more about it uh so don't need to go too deep into that. But if you could, like give us the personal side of Liz, what she was like, how she viewed the markets, kind of what you learned from her as you were getting into the biz. Yeah, I think number one, she was, she was a wonderful person. She was, she was probably the best listener I've ever met in my life. I mean,

she, she can meet the janitor and the janitor could tell her what she's, what, what he's interested in. And she would go and like buy a book and give them a book about like that particular subject. She always wrote letters. She was just, her attention to detail was unbelievable. But I think probably one of the best things I learned from her is her discipline with, with the systematic style that the quantitative side of, uh, of how we do, how we research, how we, you know, don't change the portfolio just because things aren't working all those disciplines to our, to our style. Um, it is really what I learned. I think that's why we've been around for so long is because we've

been so disciplined with when we're back testing our models, making sure that we're not overfitting, over optimizing our strategy. Those disciplines, she constantly wanted us to replicate what our research was producing. And then what did she, what was her background before joining the turtle program? So she, she had a, I think a math degree. So, you know, she just, you know, she responded to that ad. And I think it was only, it was her and just one other person who who one other woman who who sounded chosen yeah yeah and then tragically she died what's it

been ten years or so well on a busy and been China right that's correct yeah she she was there to you know we started using our models and testing our models on Chinese commodity markets. The performance coming out of those markets was just fantastic. pitching a Chinese guy who owned an FCM, basically pitching how we started trading money for Morgan Stanley as an FCM and then how they grew that whole managed futures business out. So she was trying to form a partnership. And she went into the meeting and she had brain aneurysm before really the meeting even started so that was in 2013. 2013 all right yeah so let's move on to happier topics but her spirit lives on in the firm, right? It really does.

And so you guys have been running the classic program since 1985. Is that right? That's correct. Wow. 40 years. So before we get into the nitty gritty about your programs and all that, just let me ask sort of like ask a married couple that had been married for 40 years. What's the secret? What's CMC's secret for having been around so long, right? There's probably been 40,000 hedge fund firms that have launched and died in those 40 years. Yeah, I really think the secret is our discipline, our entire team's discipline to our researching, to sticking with our models. I think that's the most important thing because it's so easy when you're a quantitative firm to, if things aren't working, you start changing it to what, what, you know, the market's doing at this current time. And, you know, I think we see that currently that there's a lot of, you know, we were in a 40-year bull bond market, bull stock market.

And it's very easy. You know, there's a lot of firms who basically have like this long only element in their portfolios now. I think, you know, our discipline, there's some tough times not too long ago in the commodity markets, no big trends going on. It's very easy to want to kick those markets out of your portfolio. It's very easy to want to change your models for trading those strategies. And we just stay disciplined with the way we research, the way we run our optimizations. We stay disciplined with the diversification in our portfolio, whether that's with markets or with systems.

And I think I really believe that that is the reason why we've been around so long, because when it's time to for managed futures to perform we usually perform the uh and that's interesting usually people hear discipline right and i'm thinking like oh i only risk so much on a train i don't right i don't get crazy on the wrist side and that's why we've been around so long because we don't blow up so that's also true but it seems more like you're saying no discipline to stick to our knitting, um, within a framework of your evolving and whatnot, but right. It all too easy.

Like I did a, uh, we did a blog post on silver trend following silver. And I think it had 19 straight false breakouts over like 11 years, right? Like nothing. And then last year had this big move, or it might've been 2020, but it had this big move. It finally paid out. So that's the discipline you're talking about, right? Of like, in order to take that losing trade over and over and over, knowing at some point it's going to pay out. Absolutely. That obviously that was, you mentioned silver and oh man, we wanted to kick that market out of our portfolio.

But I also mean the discipline of it's all embedded in our optimization process, the risk you take per trade, the weightings of each market, your correlations between each market. Everything is built into that research optimization process. So sticking with the quantities that you're supposed to buy, the disciplines of putting on trades, buying at all-time new highs or selling it lows. That's a very hard thing to do. And what piece of advice would you give to some startup guys of eyeing that longevity, right? Of like,

it's so hard because if I'm a startup of I don't have the assets behind me, if I don't have the, perhaps your guys track record and, you know, capital buildup, it's going to be way more tempting to pivot and say, we got to stop trading this. We're going to go out of business. We're going to lose our clients. Right. So how do you kind of weigh those two things? If I'm a startup hedge fund wanting that longevity? Yeah. I mean, I think it's, that's what's so hard about it is that, you know, you got, you have to make sure that you're,

you have a good research process and you, you stick to that process even, even through hard times. And I agree with you. It's, it's a tough hurdle right now. And the cost of just running a business right now, the price, you know, the cost of exchange fees and getting live quotes and getting enormous amount of data. Like we paid so much money for data. So there's a big hurdle. There's a big hurdle for sure and that's been a refrain on this podcast for a long time of like hey what if you have the choice go out of business or change your model

most everyone's going to change their model right they don't want to go out of business right that's exactly right from an investor i appreciate the right i'd rather have you sticking in one lane so to speak right of like hey i know what I'm going to get with these guys. I'm allocating. This is the profile they're going to. But if you start changing all this stuff, then as an investor, I lose confidence in that not knowing what I'm going to get. That's exactly right. And we've seen that.

We've seen that with other managers all of a sudden turn into a completely different trader. And we owe it to our investors. all of a sudden turn into, you know, turn into a completely different trader. And, you know, we owe it to our investors. I mean, we have two large investors who've been in the classic program, one since 1989, and another since 1991. So they know what they're getting from us. Yeah, and they've probably, that's compounded rather nicely, right? It sure has. So let's dig into these strategies a little bit. The classic program that we mentioned, the core of that is trend following, right? That's correct. Yeah. We kind of bucketed in two areas. We do a technical trend following.

We have two core systems that are under that. And then we have two systems that are more statistical momentum. We're trying to capture the same big outlier moves, whether up or down. But we categorize them just a little bit different. Expand on that. So what's the difference between that momentum signal and the trend following signal? Like a breakout versus a relative value? Yeah, we don't do really breakout anymore. That whole style degraded quite a bit.

So what we're doing for the trend following slice of it is what we do is we have three core lookbacks. And we're trying to get confirmation on these lookbacks. So we'll look back maybe on the first parameter, anywhere from 100 to 200 days. And then there's a threshold that that market either has to be above or below in order to take a buy or sell. Then there's a medium term look back with another separate threshold. So we need confirmations on all three lookbacks in order to go to the next level. So the next level would be once we get confirmation that, let's say, the long-term lookback, medium-term, and short-term lookback are confirming that there's upside directional movement, then we have a volatility filter. Basically, that's looking at current vol and comparing to past vol. And if it's over a certain threshold, we don't take the trade at all. If it's under it, we're allowed to take the trade.

And then we have a little volatility move that must happen from the previous close has to move in the direction of that trend in order for us to put that trend on. So we have two systems that have that same core logic. But in research, the way we make those systems different from each other is we pair each of our systems with a unique super value. And that super value gives that system direction and how we want that system to contribute to the portfolio. So an example of a super value for our shorter term systems will pair a sharp ratio with a accelerated return numerator. And that will allow that system to be it creates a shorter term system. And it also is a more nimble system. So it's responding fast.

So when the super value is higher, you're more likely to get a trade there or you're going to trade it a little larger? Well, it basically is using that super value across all that market data to have to fit those parameters genetically have to be, we populate each year new parameters for this core logic. So it gives that system a roadmap to how we want it to perform. We want it to be really, you know, a real good risk adjusted return from that system versus those same core parameters on our longer term trend following system, we might pair that to a return and a sortino. And then that system becomes way more accepting of volatility and becomes a much longer term system. So each one of our systems has a unique mathematical super value. And it's, it tends to be weighted to the, closer to the most recent data, you know, so there might be a sharp over 10 years times a return numerator over two years. And so during that optimization

process, that's how these systems get new parameters. There's no wholesale changing in the parameters, and the parameters would be, okay, the first look back back maybe it was 150 days look back but but next year it might be 180 day look back and the parameter uh that indicates it's trending could could move as well so we're optimizing to get those parameters which make the systems quite different from each other and that's all machine learning, AI based? Or you run that every year, hit a button? Or is it ongoing? Is it like a rolling 12 month process? Oh, no, it's that the reoptimization is an annual process. And the genetic algorithms take place. So the you're basically populating the, the strongest gene in, in each parameter.

Yeah. Hit on that a little. Why do you call it genetic algorithms? Because what it is doing is it's learning over time, uh, through our genetic algorithm, it's learning what the, the, the best gene or parameter that is that will go to the next generation. So as we do a forward walk, every time we will, so a forward walk is basically we don't just optimize a system to the whole set of data from 1980, let's say until until present. We walk through that optimization. So, for example, we'll backtest from 1980 to 1985.

We'll come up with the core logic of a system. And in the sixth year, we trade that core logic. So we're trading out of sample. We're building an out-of-sample track record, basically, what we're forced to do in real life as a manager. Much easier if you could trade instant. Not possible without the time machine. Yeah. time machine yeah so so basically um each each time we hypothetically have to trade trade those new parameters on data that we aren't privy to just yet um that's that gets those parameters

are like get populated so those are the best parameters And every time we do a rolling optimization, the genetics of what comes to the top is the best parameters keeps on moving along. So that's across each of those timeframes, each of those systems. It'll change it on the longer term, on the shorter term. And then is that changing the portfolio construction as well, or that's static? Or you might add them over time, but it's not, the machine learning is not saying, and you should add carbon credits or something. Right. No, we would have to just include that data into our optimizations in order to, you know, add markets. I love it. And so what is, have you ever seen, what's that, when is that period are you allowed to say? And like,

does it work really well right after the period and start to degrade? So you have to switch it over. Are you asking when's the period where we add a new market in? No, when you re-optimize, basically. So if you do it annually, you said, or periodically. Yeah, so we have four systems in Classic. And so we have a set calendar date for each one of them. So we don't do all four of them at the same time.

So we might, you know, one each quarter is when we optimize them. And over the years, like five years ago, did it take three years to degrade? Now it takes one year. Does it degrade at all or you just find the next, the new best one? The part of re-optimizing is basically so we don't degrade. Degrade. Yeah. So it's just, it's creating parameters and really looking forward more than backwards.

Basically, we're trying to, we're trying to, you know, understand what's, what are the changes in the market? Do you ever do that? Look at the one that if you'd kept it just to drive yourself crazy or just take the new one and ignore what happened? No, we definitely do. We look backwards and see what moved what parameters moved and then try to hone in on like uh you know a range so for example the first look back is anywhere from 100 to 200 days we would we would like put a low look back and a high look back and have to optimize within that range but

now when we do genetic optimizations we can just let everything we don't have to put like ranges on on these optimizations we could just open up the parameters as wide as we want and just let let those parameters get populated you know from a natural selection process and how crazy could that be so could it jump from 50 days to 550 or something or there's still some yeah no they never we never get wholesale changes mostly because we're optimizing the systems to that specific unique um you know sharp ratio basically yeah um the super value is always geared towards so it always wants to it always wants to stay within its you know it parameters, shorter-term parameters. And that's kind of like the beauty of it. We know what kind of contribution each one of our systems is going to give us. I mean,

obviously over time, we can look back and go, geez, if we only traded the momentum system, the one momentum system, we would have been doing way better. But we want the diversification within the systems because they all go in and out of favor, just like markets go in and out of favor. And, you know, we just want to be around for that. And speaking of that over time and in and out of favor, like what is the model today look like in terms of what it was in 85, 95, 05, 15, right? So like, is it 50% the same, 5% the same? What does that evolution look like? Well, I would say it's more like 5% the same. These models are a lot more sophisticated models. Back in the day when Liz started the firm, it was one system and one risk management-like strategy that overlaid the portfolio with very little components to them. Now we have multiple systems.

Each system has multiple core logic to it. Probably one of the most important things we've done is our – so each system has its own risk management to it, where it gets in, where it gets out. But one of the best things that we added to the portfolio was this whole risk overlay. And it's built with components very similar to our system components. We take those components and we run it through an optimization every year. So the components would be, it would be like open trade, equity, a trailing P&L, a scale factor. And so what we see, what happens in Classic is we optimize those core parameters

to a utility function. And a utility function is really like a satisfaction. How much are you willing to make and how much are you willing to give back on that? So we'll see in big periods of time when EMC is making money and the trailing P&L builds up, once it gets past the scale factor, we start lightening up the portfolio across the entire portfolio. And we can get to a point where we're cut back in classic 50% because the trailing P and L has really kicked in. And this is something that our clients have really noticed over the past you know 10 years or so we we implemented this in 2007 and then we have continued to re-optimize it every year and add more you know more components to it but what happens is there's you know, more components to it. But what happens is there's, you know, you, you've been around this space for a long time. You see huge months that CTAs all do really well. And like the next month,

it's given all back. This component really helps us, you know, retain the, you know, p l um it lowers our volatility it lowers our drawdowns um and it helps us capture more of the more of the trade does it take you right does it like move you less positive skew like so i can see where you're coming from right that's the number one complaint about trend following great it but i made you know i rode crude from 30 up to 90 and then it sold back down and i didn't get out until it went back to 50. and people like you and me are like hey you still made 30 to 50 that's a great trade but the investor's like well i had to report that loss from 70 down to 50 and it was it was painful to report to my end investors or to my family office or whatever.

So your utility factor is solving for that? That's exactly right. It's like a transformation function. And so what it basically does is it's looking at, it's like ranking monthly returns using an arc tangent. So basically what it's doing is when you're looking back at, you know, your research and you're looking back at your monthly returns, it's going to reward a better negative return to an improved negative return more than it will reward an improvement in a positive return if that makes sense um so you really it's like a satisfaction thing so you know we would go to our clients and go hey you know would you rather us make 25% or 20% and only have a couple percent give back, a satisfaction, how much is enough?

Right. So did you do that with an eye towards the behavioral finance, right? That's a famous testing of like, would you rather make $100 or lose? I can't remember what the math is, right? The famous experiment, but people would rather avoid the loss than get the gain, even if it's economically in their favor to go for the larger gain. They want to avoid the loss. That's exactly right.

Yeah, I mean, that was what that concept was put in place for, the utility function. Which is weird, right? Because in your lab, in the scientist's lab, you'd be like, who cares what the investor thinks? Just I'm making the best product possible, right? And that kind of goes into like full Kelly betting and all that stuff, right? I'm like, well, the best product possible might be I fully run this thing at 80 vol and I have drawdowns of 80%, but over 30 years, I compound the highest. Great in the classroom, not so great for real world and real investors, right? That's right. That's right. But that also leads me to like, do you, so how much of that do

you give up? Do you give up some of these, right? You're not, you're in that scenario, you're probably never going to have the a hundred percent return or like the huge outlier move. Yeah. I mean, it was, you know, it was in place during 08, you know, and we had, we had a really big year in classic in 08, even though we scale that portfolio dramatically um you know so let's say you know if we if we made 50 in 08 we probably left 20 you know further in return on the table probably i mean it's yeah we're cutting that thing pretty, pretty dramatically. And we don't just do like all of a sudden a 50% cut. It's like, as that P&L builds, we're, we're doing a 5%, 10% cut. And, you know,

eventually when that P&L stops building, we stop cutting. But it's across the whole board. You know, we've had so many different, different strategies to reduce that give back. And, you know, one of them was just look at the market that's trending and making money, cut that, but that's a bad idea. It's a terribly bad idea. Yeah. So I was just going to ask next, right. So that's on the portfolio level. That's correct.

And so it's going to be, and does it look at each month? So if I, if my best three month return in the historically was 40% and now I'm at 38, it's going to start peeling back. It's it's actually different than that. It's if, if let's say we, we have a trailing P&L over X days that gets above, let's say, 9%. So once the trailing gets above 9% and it goes to 10%, we have a little scale factor that's built in. So what it really does is if we have an open equity, we don't trade in our classic strategy. We don't trade that open trade equity. So the scale factor is just using a multiplier against that open trade equity. So it makes the open trade equity bigger, which makes us take money off the table, basically.

Got it. Right, which is the classic from my days running a trend following, right? You would hit a big trade in corn. Now, right, say you started with a million-dollar portfolio. You make 200 grand in corn. Now, right, open trade equity, now you're at 1.2 million, and you get a trade in silver. Do I size it off the million or the 1.2 million, right? That2 million? Inevitably, you'd size it off the new

1.2 million. Now the corn P&L goes away and you lose on silver and it just increases volatility across the board. That's right. I was just going to say, from a standpoint of sizing trades in the portfolio, we do look at that market's volatility, short-term look at volatility to know how much we want to risk in that market. So each market has its own risk weighting, but then we take in consideration the volatility of that market, the current volatility of that market when we put a position on. So the more volatile the market is, the smaller the size and vice versa. If a market doesn't have a lot of volatility, something like Euro dollars, we're able to put on a bigger position. Right. Which is classic. Trend found 101, right? I'm doing 20 euro dollars, one palladium. And if there's the big outlier move, right? If it's for standard deviation move, I want to make the same amount in each of those markets. I don't want to unknowingly have like most of the portfolios

profits because this market's more volatile. And then you said something interesting before of if the volatility is too high, is that in each market or portfolio wide? If the volatility is too high, you're not going to take the trade. So that component's within the system. Got it. So we've tried to, it's funny, we've really tried to get rid of that parameter in all of our systems, but it screens out so many bad trades that we need it at some level. Each system has maybe a different threshold. So some of our longer-term systems are more accepting of expanded vol.

So we want to make sure that there's not a trend and just because of volatility, we're not a part of it. That's the diversification within all the systems as well. My experience is when volatility is expanding, 07, 08, 2020, 2021, 2022 year, that's when trend following does better. This is saying not just increasing vol, but if it's really high absolute level of vol, you're going to get whipsawed.

You're going to get stopped out too soon. That's right. Yeah. So on to the other programs. So we covered classic. Anything else to add on classic? I could just fill you in. The other types of systems that we have is the statistical momentum systems.

And so those are quite different than having confirmation on three levels, like the technical trend following. So the statistical momentums are basically we're looking at a shorter look back period of time. It's all time weighted to the present. And one of the systems is looking at a close to close basis. And so we do this analysis where we're looking at today's close versus yesterday's close versus the day before it closed. And we walk back and forth this what we call count against. So we're looking for closes in a certain direction and a lack of closes in the opposite direction. So almost like you would look at a manager where, where you have, you have a, a run-up,

and then you just want to make sure within that run-up, you don't have like these drawdowns, these big drawdowns. So a lack of counter closes will allow us to put that momentum trade on. And then our longest term system is more of a looking at a shorter term look at the market, but really looking at the drift and the magnitude of the drift. So that would be a market that, you know, if you looked at a chart, you just like, you know, you go, oh, you know, it's definitely going higher, but it kind of measures that on a short term. So the system, you know, is more accepting of volatility

and it allows this trade to get put on like awfully fast over the last five years that longer term system has been the momentum longer term system has been our most successful system in the portfolio but all four systems we equally weigh just for diversification and you know we we want to just make sure that we are uh you know we're not making changes just because of the results of our optimization which is weird right because you're like we're going to do optimization and this these this parameter set we think gives us the best chance, but then at the same time sort of ignoring optimization there on that level. So it's an interesting yin and yang there of like, okay,

I'm optimizing, but not over-optimizing. Right, right. That's critical in our research process. I mean, because it is really easy to say, oh, this system works great with this market, and this system works good with another market. So trade just those systems for those markets. That's just a real overfit strategy. What we do is we equally weigh all of our systems across all of our markets. So once

we run an optimization and come up with parameters for, let's say, our short-term trend following system. Those parameters that we come up with, we trade every market with those same parameters. So we're trying to build a robust set of systems that trades successfully in sugar as it would in euro dollars as it would in S&Ps. So we're not customizing the systems to each individual market or to each sector. Which comes back to the silver comment, right? So even to the point where you would add something that's lost 26 times in a row and no one in their right mind would ever trade on a standalone basis, right? Right, right. Which is crazy.

But that's the power. Coming back, it's interesting to me, right? Always the momentum factors usually talked about in stocks, right? Single name stocks, not really in futures markets where we tend to call it trend. So like for you, what do those correlations look like? Are those models 0.8, 0.9 correlated or they, you'll see much different trades? Yeah. They're, they're, they still have a high correlation.

Yeah. Because just like CTAs, trend following CTAs have high correlations because we're trying to chase the same outlier moves. And so that just moves the correlations because we're trying to chase the same outlier moves and so that just moves the correlations you know up pretty high because we want all these systems to be able to capture trend or you know momentum type moves so the correlations would probably be you know between our longest term system and our shorter term, that would probably be the lowest correlation. So the longer-term system goes out to about 120-some days, average holding period.

And the shorter one is more like a 15-day. So there's plenty of times where those systems are actually on the opposite side of the market. So one could be short and one could still be long. So that would probably be more like a 0.6 correlation with each other. I guess another way to ask it, you're capturing the momentum factor, even in the trend model, right? Is that the scientific explanation

for why these things can work over time yeah i would say yes so circling back so you guys have a few other programs so let's talk through those quickly. Sure. So the other programs that we have, the first one we call EMC Alpha, and we really custom built this program for a client of ours who wanted to start a 1940 Act Mutual Fund, which is a game changer in our business. You know, basically people want to invest in our programs. It was typically if we had a fund vehicle, then they could come in for a lower threshold. But if they just wanted to do a managed account in classic,

the minimum was $5 million managed account. So there's not a lot of guys who have a $5 million that represents a small portion of their total assets. Now this mutual fund structure is a completely different game changer. Number one,

you basically have a lower threshold in order to get into this alternative investment and you have, you know, hundreds of brokers selling this to their clients versus, you know, one sales guy or two sales guys, you know, pitching it. So we're the sub-advisor to this program called the alpha program and where it differs from classic is we have, we have 10 systems in the, in the portfolio. And then we have sub strategies for sub strategies. One of the sub strategies is just short-term interest rates. We call it on constraint rates.

We optimize a few systems to that core strategy. So that differs a lot from what we do in classic, whereas we're customizing systems to certain sub-strategies. Then we have a commodity-only sub-strategy that we optimize core systems to. And then we have a long-sh short global financial, which is basically currencies, fixed income, stock indices. And we optimize some more trend momentum systems to those sub strategies using different super values as well. And then the very final sub strategy is a long only rebalance sub strategy of stocks, bonds and all get flat and can get short. So right now in that strategy, we are short fixed income.

Well, there's a long only bucket, but then there's also the dynamic bucket that could go too. It might be long in the long only and short in the dynamic net correct yeah and how do you get that that's like to give you basically a positive carry that will keep you above water until the kind of alternative piece pays out or do you view it as the give you positive beta with protection via the alternative piece pays out or do you view it as the give you positive beta with protection via the alternative piece yeah i mean quite honestly the whole idea yeah the whole idea of this portfolio is we always you know when we just had classic we always said you know you want to combine this product with your other holdings. And that's usually long stocks, long bonds, long gold.

And so that's a hard sell because when an investor invests in the classic program, he doesn't look at it as a combined investment with his, you know, their current investments, their long investments, totally looks at it separate. So sometimes, you know, we go through periods where they're not correct. They're not great periods. And so line item risk, we call that. Like they're just sick of seeing that line item when if you bundled it, it's like, Oh yeah, exactly. So exactly so it's uh so we just decided that you know it's hard to sell people on just the standalone um so why don't we build like

a portfolio that's you know it's more balanced uh as as just a complete investment. So what's interesting is that mutual fund started out with $5 million in 2013. It's over $500 million now. And really where they're selling that product is basically to brokers' fixed income sleeves. They started out selling it to like these… Managed futures. Yeah, yeah, managed futures or even in someone's stock bucket. Now where the big attraction is happening is…

Bond replacement. … bond replacement income. So it's shooting more towards single digit returns with single digit volatility. Very different than what classic shoots for. And then what else? Other programs? And then we have another carve out within that EMC Alpha program that's in the mutual fund, we also trade ETFs in that portfolio. We have an EMC Alpha Plus strategy, which is a levered-up version from that Alpha strategy, and it's all futures no no etfs in that portfolio so it's it's very similar we built it very similar to alpha it's just a levered up version of that jumping around a little bit here we touched a little bit on the ai machine learning? Just give us a little bit of like,

what's your take on all that? Indispensable to you, should have done it sooner. It's just a fancy spreadsheet. What's your take on how you guys use the machine learning and how you value it? Yeah, I mean, we really value it a lot because it's an automated research process for us now. I mean, we used to sit at a conference table, everyone in the management group, and just, you know, spitball, just figure out what changed in the system. Is it good? Should we allow it to change by that much? This is a natural selection process

that, you know, coming to these conclusions, we're able to come to them a lot faster because of the genetic optimization that's happening. There's a lot of different interpretations of artificial intelligence. You know, there's, you know, artificial intelligence where people just throw a whole bunch of parameters into the code and let it try to figure out what's the best parameters. We don't do it that way. We basically have core logical parameters, and then we're directing these systems to get the best parameters within that core logic so it's quite different than some of the artificial intelligence we've had yeah i think i view it as like outsourced manpower woman power right like you're just instead of a room full of a thousand

people crunching through all this you have the machine do the work that's exactly right yeah um so we really believe in it from not only from a a system building idea but also from that overlay risk management uh that's really that's really helped us uh outperform a lot of our peers especially in good times that's easier to tell investors about too, instead of like, well, I don't know why we went long that, but the AI did it. Yeah. It seems to have worked over time. And the classic AI, right, is like going to optimize to what's working over this X period where you've kind of said like, no, we're going to have discipline. And even if something's not working in that period,

we believe in it philosophically and we're going to keep discipline. And even if something's not working in that period, we believe in it philosophically and we're going to keep, keep it in the portfolio. Right. Right. Yeah. We look at the whole portfolio, you know, when we run these optimizations,

we're not, we're not just singling out markets. We will, we will take some outlier performance in our back test and remove those markets. So they don't skew our results by too much like if you know the central bank you know devalued their currency or whatever and that had a huge move like way back when Mexico had their

big devaluation it really sp spiked our P&L. And so we kind of take some of that noise out just so it doesn't skew the results. Take out the largest loser too. I'm thinking of the Olympics here, the largest loser, largest winner, judge scores get thrown out. That's right.

Yeah, yeah. Just some of the ones that can really skew. If trend following starts to go really back in vogue here, right. It's had a great 21. It's doing great so far this year. If hundreds of millions of billions start flowing back into the space, do you have any reservations that it gets too big that that'll degrade the core signal there? What are your thoughts on the old trend following too big angle that was prevalent five, six years I personally, you know, it's, I think you can get skewed when, you know, if you look at the 40 years that the bond market has been, you know, has gone up and you customize your strategy to have a long bias or you customize it to, you know, have a long bias in stocks. I think that, that could definitely hurt, you know, you know,

if people have adjusted their strategies to to that past market environment, I think then it could terribly degrade. But from our standpoint, we, we build the same parameters that get you in a long trade are the same parameters that get you in a short trade and then shorter versions of those parameters get us out of those positions uh so shorter look backs uh and thresholds get us out of the position. So I, you know, I personally don't believe in, in like, you know, that the more money that comes in, you know, the less success we're going to have. We would have to probably weigh certain markets like lumber her and orange juice and some of those smaller contracts

we probably have to weigh them too low to actually participate in them but at this point right now they are light weights but they we can still trade those markets and have them add value gwen asked another way what would you have to see in order to be like, this has gotten crazy. Like these, as soon as there's a breakout, the thing spikes this much, we're not able to get in soon enough. Right. Like what kind of things would you have to see to be like, there's too much money chasing these trends,

which is a odd question in and of itself. Cause what timeframe is it on a 10 day timeframe, a hundred day, a 200 dayday, right? Yeah, I mean, we would definitely have to, you know, obviously we have to look at slippage, you know, type effects. We have to look at all our custom algos that we built for putting trades on. We'd have to look at our market weights. There'd be several things, you know,

if we start seeing, you know, moves that we can't get in in time and can't get out in time. We hope that through our optimization process that those parameters will all adjust to that current environment. Right. So, but my,

my counter argument would be get the more, the merrier that's going to drive more liquidity and more people into the market. And if your models are, are better than the next guys, there could be opportunity there. Opportunity. That's right.

Yeah. And so you mentioned everything being the same on the long side and the short side, the bond trade we've talked about, right? So bonds basically went straight up, rates down for 30, 40 years. Two things there. One, back in the day, everyone said managed futures returns were juiced because they could hold T-bills, but we'll ignore that for now. But just right. If they were a participant in that huge 30 year bond trend now, simplistically, we can say, oh, if that reverses, if rates go back to

15%, whatever, there's going to be a huge downtrend in bond prices rates up. But Roy Niederhofer, some others have pointed out like that won't look the same because of the cost of carry, because the curve will be different. So what are your thoughts on the rates up, bond prices down? Is that going to be a mirror effect? What's it going to look like for your programs and Managed Futures in particular? Yeah, I mean, we trade across the whole curve. So we're trading short-term interest rate products. We're trading medium-term. We're trading long-term products. We built these systems based on the data that we had, which was really skewed to the upside. For us in particular know i'm we don't really look at it can is it going to be a different type of trend we're just trying to capture directional price movement so i mean the the amount of p and l we've already picked up in the euro dollars um over the last you know couple years signals it's a it it's

it's a pretty good trend right now i mean almost better than uh the the amount we picked up on the upside so it you know it all depends i i think it depends on the volatility of the market. It doesn't, you know, Euro dollars, vol has just picked up big time in there. So our strategies could miss if they get screened out by vol, but we got in so early that all of our systems are already positioned in those markets. So, you know, they could be right that, you know, this is going to be a different type trend, but, you know, we basically are just building models that catch, you know,

try to catch outlier moves and however big that is, it is. I mean, we're hoping that our risk management overlay, you know, ends up capturing, you know, scaling back when we do capture these moves. And, you know, like right now, we have a pretty decent scale going on in the classic program. So if those euro dollars reverse, we're going to capture a big piece of that already which sounds to me like you're saying like yeah maybe that trend looks totally different you're not going to make as much as you made on the way up but unless you set your whole model up as like that's the main factor right that you wanted that your whole performance is based on it looking exactly the same right if you don't do that then it's not necessarily a problem. You're just going to get what you get and have a nice day and move on to the next one.

Right. I mean, look, I think probably most investors didn't think that interest rates could, you know, go negative and could have, you could have put into their models, you can't buy bonds when the rate is under a percent or whatever. And we've talked about that before. But what do you do with bonds at the zero bound? And why would you go long there? There's no more room to go but yeah yeah 2019 was it right where there was tons of made money made in uh boons and bubbles and all

that german stuff of going long at the zero bound in it right basically it went more negative and you could make money yeah and the only data that supported, you know, was prior to this all happening was Japan. You know, the euro yet the rates, the rates went negative. And the volatility just completely died out. So it was almost like a flat line movement on a chart. What we did as a manager, just not being in that type of environment before, is we just lowered our market risk in there because obviously with no ranges, we would be putting on massive positions based on the volatility of that market, which is a very dangerous thing to do. So we pretty much cut the leverage in that market.

That's definitely a discretionary move when it comes to our research. But we just felt that without any data supporting what could happen here, that we'd be best off protecting ourselves love it um and then let's talk to we mentioned lumber a little bit um we just had a pod a couple weeks ago on lumber talking about the supply chain issues the beetles because the climate change all this stuff so it's always amazes me right like no offense but you don't know anything about all that stuff exactly right but you might say oh no offense taken that i love being able to make money off something that i i know nothing about right so to me it's like a feather in the cap of trend

buying like you would need an army of analysts and all this exposure to be like i'm i researched all these lumber i bought into this lumber mill that's publicly traded or i bought this timber right and here it's just part of the portfolio and you get that exposure in a measured way um and it's like but you know it's a long call option on things like that happening right yeah that's right that's right. You're right. I mean, even if you think you've got all the fundamental reasons for what a market is going to do or what it should do, sometimes the markets completely go the opposite direction and you put your hands up like, why?

That's the greatest thing about the disciplines of being a trend follow systematic managers. You don't have to know why. You don't have to know why. And then something, another one, carbon credits. Are they part of the portfolio yet? What are your thoughts on that? To me, I think it's a weird thing to think about, right? Because by definition, they should go up. They've reset plus 15% a year or whatever. So it would seem to be a perfect thing for a trend following portfolio. portfolio yeah we do not have it in any of our portfolios right now and we don't we also don't have like bitcoin in our in our portfolio as well um there's a couple reasons for it from a standpoint of the mutual fund um the the um the the oversight oversight of that distribution of the mutual fund, they disallow it to be in our portfolio.

So that one program can't have that. And then we, you know, we've gone in front of, you know, several of our investors in the past to add the VIX or to some other markets. And they don't tell us we can't, but we just sort of look at some of the returns from some of those markets and we feel like it's a little bit of a dangerous move to get into, especially the VIX. Where you think you're going to get out and where you do get out of some of those markets is not the same. We're staying away from some of those markets. We still have a lot of diversification in our portfolio, you know, between commodities and financials, currencies. And the crypto always seems like a perfect fit for a trend power, right?

Like just take a little bit of risk and you're going to get some of these outlier moves and rinse and repaint, have the builder to go short. So I'm sure you guys will revisit that down the line. Absolutely. Yeah. As the futures become more ent will revisit that down the line. Absolutely. Yeah. As the future has become more entrenched too, and the liquidity gets better. Let's finish it up with two truths and a lie this year. Um, what do you got three, three things about you?

One of which is a bit of a stretch. Hmm. What do you got? Three things about you, one of which is a bit of a stretch. I'm a good golfer. That's a good question. I'm good at these podcasts. You're making it tough now and yeah three that you're not god and the sun beaming down from behind you there is um just the normal sun can you see that on your screen the sun's like oh yeah Good. No, I, you know, honestly, you know, I think I'm good at what I do here at EMC from a standpoint of keeping, you know, the philosophy of what we do from a research standpoint and, you know, managing the company. But I'm not the smartest guy in the world. I surround myself with some very brilliant people at the company here who've been with us for several decades, and I'm loyal to them. Liz was very loyal to me. I'm loyal to them that's liz was very loyal to to me um i'm loyal to them i you know i pay them what they're worth because um because i want them to stick around here yeah

teach them well enough to leave and pay them well enough that they won't want to i think that's how that's right that's how the line goes, right? That's right. Well, thanks, John. Any other last thoughts before we let you go? Tell them where they can find you and all that good stuff. Oh, sure. You can find us at www.emccta.com. emccta.com. Cool. And then we're doing a white paper on trend following which you guys will be highlighted a little bit in so check

that out and thanks for listening thanks John hey thank you 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. 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.