Backtesting, diversification & avoiding curve-fitting

Transcript

Andrew Swanscott 0:02: Okay. Welcome to the trading panel, our very first ever trading panel, I guess. So we're gonna see how this one goes. We've obviously got some people here. You can see it on the left. So we're gonna do a little bit of an introduction, and then we'll explain what this show is all about. This show is the idea of Jason's from Against All Odds Research Podcast. So we're gonna let Jason explain what this is and then we're gonna get to it.

So maybe I'll do a quick introduction of myself first. My name is Andrew Swanscott. I'm from the Better System Trader podcast. As you can probably tell from my accent, I'm Australian. I'm based in Melbourne, Australia. And my trading style is purely algorithmic. I don't use any discretion in the futures markets. I do breakout swing and mean reversion trading. Jason, do you want to go next? Yeah. So

Jason Kurz 0:54: I'm Jason. I run against a lots research. I have worked with hedge funds. I have a small fund still to this day. I've been trading markets for, over ten years at this moment now. So recently, we've been doing it, and Mayhem's also been involved in it too. You know, we've been on these spaces, and there's been going over and over again, and you're like talking to people who I don't know if they actually have money in the markets or not.

And so Andrew and I talked probably a month ago now, and we were like, hey, maybe it'd be a cool idea to do something a little different. And I was like, yeah, why don't we do something like this? Like, maybe we could get together, we could do a panel. Andrew mentioned using a service that could get us on multiple channels, and so we kinda the idea just started there.

And so the idea basically is Andrew and I will bring in buddies of ours, traders that we know, people we've interviewed. We have a lot of people that we've interviewed that are similar, but we also have a lot of people that aren't similar. So you guys can get a little bit from both of our audiences and meet different traders from all over, especially different traders, different services, different people. I mean, it's endless.

So I really think this is gonna be a cool idea. These are people who all have real money in the markets. There's nobody here just trying to sell you a service who or sell you a position that they're trying to talk about. These are real traders. So, anyways, getting to Mayhem. Why don't you introduce yourself next?

Panelist 2:19: Sure. Yeah. Happy to do that. Thanks for having me on. So I started trading well, I did my first trades when I was a teenager in the late nineties during the.com bubble, traded Amazon and Yahoo, borrowed some money from my dad against the earnings I had. I've been working since I was 12, had some money in the bank. But as a young person, obviously, I couldn't legally trade. So I did it through by proxy through him.

And those positions, a couple thousand dollars worth doubled the next day. And it was obviously I did obviously, I didn't understand something. And instead of going a whole hog and saying, let's just go super long the market. This is a money printer to infinity. Why would anyone work a job? I did the opposite. I cut everything out. I booked those gains, and I took a couple years off to figure out what was even going on.

Obviously, over the years ahead, the market crashed, the.com bubble deflated, and I started to get really interested in understanding more about how to trade. So in 2005, having earned enough money, becoming an adult by then, all that sort of stuff, I started trading my own money in my own account, trading stocks, then options then learning more about forex and futures and, navigating through the great financial crisis, actually surviving

and in some ways thriving. Obviously, I had some big losers, but I had some even bigger winners. And I started to get more interested in financial plumbing and where there were it's kind of a nexus between what was happening in the world of macro and where momentum was pointing price. And so that's kind of where my formative years happened, at least in terms of my swing trading. And that's the main kind of trading I'm into.

I like pair trades and swing trades driven by momentum primarily, but I also like to be, where there's where it's possible, try to find out why it's happening so I can understand some of the mechanics behind the scenes. So that's kind of the top down view of my trading style.

Mike 3:55: Nice. No, that's perfect. Mike's

Andrew Swanscott 3:58: here. Sorry, Jason. Mike's just dropped in. Good to see you again. Do you wanna we're just doing a quick little introduction. And since you just dropped in, do you wanna just give us a little intro about yourself?

Mike 4:11: I'm really famous. You guys don't know me. No. Hey. Mike. Been running a trend following font since 2011. And yeah. I mean, pretty simple business. Living in New York. Good buddies with Jerry Parker. Do some do some work for him as well. You know, we're we're good friends. We're we invest in each other's funds and do some do some work together. You know, latest thing was getting his ETF launched. So, yeah, just just simple guy. Isn't it?

That's all I got. We do a podcast too. Me and him do a podcast as well. Well, I do it and he shows up type of thing. But yeah.

Jason Kurz 5:13: Very nice. So we'll go to Kashyap, and then we'll go to Pavel. Kashyap, you're you're next. Yeah. Hey, guys. I'm Kashyap. I live in Dubai.

Panelist 5:21: So I got my start in finance writing investment newsletters. So I spend most of my life as a fundamental analyst. So that's my forte, and that also goes into my trading style. So thanks to my long association with Jason, I understand pretty much all the other styles of trading that are out there, but I tend to focus on the fundamentals.

And my specialty is crypto, which is kind of ironic because when you think fundamentals, think accounting and all that stuff. But, yeah, so that's me.

Andrew Swanscott 5:57: Nice. Well, we've got another crypto guy here as well. Pavel.

Pavel Kýček 6:00: Yeah. Yeah. Introduce yourself. Hello. Thank you for having me here. Well, I started trading almost eighteen years ago. I lost quite a lot of money on discretionary trading. Then I found a way how to make money in discretionary trading or I would say semi discretionary. It was the combination of systematic trading, like mini version kind of trading on indices with level two data.

This kind of like quick reversion move on basically mini SMP and in mini Nasdaq. This was how I used to be trading like eight years ago. Then I started systematically investing like these, I would say very simple ETF strategies. And when my daughter was born, I could see that discretionary trading was almost impossible, which was a little bit over seven years ago. And I went basically all in on systematic trading.

I used to be trading commodities, breakouts on commodities, and now I'm in stocks and in crypto because, yeah, in crypto, the potential is just huge. Lot of risks, of course, but the potential is also very, very different to any other traditional asset.

Andrew Swanscott 7:21: Yeah. Well, thanks for all those introductions. I just realized that we've got pretty much most of the continents of the world covered here. We've got The States, obviously.

Panelist 7:31: Pavel, you're in Europe, in Czech Republic, I think at Yeah, the exactly. Republic. Keshav's in Dubai. So I'm not sure. Mayhem, I think, is in The States as well. I am in The States. And I've actually got a drop off, but we'll still be able to cover all the continents of the world outside of Antarctica, right? Basically. So it was really cool hanging out with you all though, and I hope to do it again soon.

Jason Kurz 7:52: For dropping in, Mayhem. We'll see you next week or the week after.

Panelist 7:56: A great weekend, guys. Bye.

Andrew Swanscott 7:58: All right. Awesome. So so I just wanna encourage people as well because this is live. We all are sitting in our rooms, hotels, offices right now. So if you got any comments or questions, post them in the chat and we'll see what we can do. Now, Jason, this whole thing is your brainchild. It's your idea. How do you imagine we should start now that we've done the introductions? What's on your mind?

Jason Kurz 8:22: So something that would be interesting is just to see, especially Mike. I'll start with Mike because I do have a question for him, which is, Mike is a systematic trend follower. I'm also a systematic trend follower. We're looking at the NASDAQ and we're looking at small caps right now and everybody's arguing back and forth. I want to know how you're positioning in to handle that, Mike. You have two signals, both of them are saying to get long.

What do you do in that situation? Get long. Get long, yes. So this conversation has come up again and again and again. It's kind of funny because everybody's like, you gotta sell all your NASDAQ positions to get into the small cap positions, or you gotta sell all your small cap positions because the NASDAQ is booming and blah, blah. And it's like, trend following keeps it really simple. It's like, we don't know what's gonna happen.

The NASDAQ could go down today and small caps could take over, it's whatever, but we wouldn't be in both positions if we weren't already getting in those a long time ago because of the signals. So I just wanted to pass that to Mike, just considering I knew he was gonna answer it like that. And just for people to know, that's how professional traders like really handle things. It's like, you don't know the future.

I was just having a conversation with somebody the other day about this. Like, they were a new trader and they were asking me all these questions. Were great questions, but the one thing he was like really stuck on was like thinking that we all know so much. You like you get on Twitter and you see this person talking about all the money they're making or the trades they're making.

And he felt like we all knew exactly what was going to happen in the future. And really at the end of the day, it's like, nobody knows what's gonna happen. There's nobody who knows exactly what's gonna happen. What we do do well, is we put the probabilities on our side, whether it's systematic, fundamentals, however you trade, like you just try to put the probabilities on your side. So I think that's like the most important thing about trading.

So I guess to start off, like what is everybody's, what is everybody's strategy timeframe to just start there? Because I think that's really important. We should talk about that just so everybody knows. I know some people might even have multiple. We'll start with Pavel.

Pavel Kýček 10:34: Yeah, well, basically I'm running almost all my strategies on dailies with some regime filters on weeklies, but most most of my strategies are on daily time frames. But for example, on crypto, because we are just having not enough data, I would say, for all the testings and so on.

So that's why I'm making my out of sample or some kind of robustness testing on four, eight hours charts and so on just to get a little bit different point of view on our strategy. So this is how I'm thinking about it. But in fact, I don't care about it that much. I think I could go with a little bit lower time frames just to get some uncorrelated strategies even with the same strategies, which is quite interesting topic.

Would say that you can basically have almost the same strat, basically the same strategy. And if you run it on the small time frame like three x more or so, the correlation can get or the correlation can get to point 3.3, point 25, which is something which is very very interesting, I would say. That you you don't need to find anything else.

You just can reuse the same strategy on slightly lower time frame and get quite different quite different results. No.

Jason Kurz 12:04: I love that. And that's good. Yeah. Go ahead.

Pavel Kýček 12:07: Yeah. In general in general, I like daily time frames because the ratio like slippage slippage to average trade ratio is very good. Basically, you don't have to take care of fees in general or yes, you you know that fees are existing, but if you run robust strategy, fees are not an issue at all.

So this is also something why, for example, on crypto, I started almost exclusively with daily time frames because I didn't know how the slippage will be baked, what will be the real average trade and so on and so on.

So that's why I even didn't want to start with anything else even though I can see that crypto is much quicker market than stocks or even commodities and that's why I can see that there is added advantage, much bigger advantage compared to, for example, stocks go in with a little bit lower time frames like four hours or so than on one or daily time frames. So this is something what I'm thinking about.

I will definitely edit them, but no rush there.

Panelist 13:24: No. I'll go to next because I have the exact opposite of that approach, which is I focus on the fundamentals. Right? So what happens is, however, late I start my entry, it still takes about three weeks to, let's say, three months for the trade to start working. And that's the way it is because sometimes I'm very early to the news, and then the news starts seeping in after I have already taken my position, after I've been talking about it.

And sometimes it's like the market reprises to the news, just all of a sudden, right? Like you may have no action for, may have a long consolidation and then boom, it just spikes up and then just keeps running. So I've had this experience many times. So I just realized I should take a much longer time frame and just let my trades run. And that's the only way I can make a fundamental approach work.

Jason Kurz 14:29: So Mike, and you're looking at your timeframe. I know you run a breakout system, correct?

Mike 14:39: Yeah. Sure. So

Jason Kurz 14:41: you're looking at I mean, you you can feel free if you wanna keep it secret. You know, that's that's up to you too, but, nothing very secretive about it. Okay. You know? I was just I think it's all some Crypto Nick keeps coming up with a great question, which is, what do we use for back testing? You know?

Like, that's something that I think is a good starting point for a lot of traders because nobody, like, really thinks about that when they're trading. The first thing they think about is how to win, how to make money, but backtesting is incredibly important. So what platform do you use for backtesting like?

Mike 15:18: Oh, I built my own in Excel early days, and then then I just migrated over to a better language. I use Python. It's so easy to use. I won't say so easy, but it's it's more, I think, learnable than some of the older languages. But, yeah, there's some definitely well, at least for trend, I think there's there's some decent off the shelf, products to get you going.

But, but they might not they might not allow for a lot of creativity or your own rules, things like that. So sorry. This fucking alarm going off. Sorry. No. You're good. You guys might have heard it, but I heard it. But yeah. I mean, I think I think it's basic. I started yeah. Started Excel and went to just went to a better, more efficient, faster language.

And, I think you could probably you could you could code it up if you will, write it down on paper first. You can even before I even started building it in Excel, I would I would literally just just go through, go into a spreadsheet or write it down on paper and then and then go to the charts. I go to the charts and I write the day, write the price, write the, the position size and things like that.

Just to get the feel of trying to live through the ups and downs. You know, go slow with it. So you can actually see what's what's going on like, yeah, you can actually see it. It's like for a trend trader, you're kinda looking at, this huge rally off a bottom and then you get in. You're like, oh god. Now, you're like, like, when I when I was 22, 23, like, oh my god. I'm getting in now. What the hell? This doesn't make sense. Right?

And, just, let's see. Let's see if that works. And, go back as far as you want, test as many markets or stocks you want, just to make sure you're confident in it, and, for me, that's like fifty plus years with dozens and dozens of markets. You know? Because the increase the sample size as much as possible. So you can test this rule out as many times as possible.

You know, don't take one stock, where you have three different trades over the last ten years or something like that. That might not be, viable, when you go to another stock, or another market. So we wanna test it on as many different things as possible. And and then, yeah, I prefer that way to start. It's very thorough, very slow, and you get to write it down. Something happens in your brain when you write it down. You actually write it.

Don't you don't just get a, a report spit out at you after a thirty second test or something like that. And and all you do is look at the returns and the statistics of it and you get your pie in the sky, ideas of how wealthy you're gonna get and all this stuff. And you don't even really know how it works. You know? And you're not gonna have any confidence to do it when you get in the game with real money on the line.

So now those were those big those were big steps for me in my early days. Now I can I know what my rules are. I know what the testing. I know, I can I can I can look at a thirty second test after testing some things and be and be confident even even if I'm just looking at the the summary, statistics and things, because I know how they got there? I know where they're coming from.

I think I think if you're just starting out, I think I think it's it's good to go through those charts.

Pavel Kýček 19:48: Yeah. And

Mike 19:52: yeah. Just so you get the feel. Just get a little bit more of a daily p and l type of feel. You know? What types of situations you're gonna be buying and selling.

Pavel Kýček 20:05: Sir, good. If I can there were very good points in, I would say idea first style of testing and code first, I would say style of testing, because I think that one trader is very skilled. He can use both possibilities, especially if he has good robustness testings. But if you are a new trader, I don't think that you can basically start with a quote without knowing the markets, how they are behaving and so on.

And without, at least it was my path that because I started with discretionary trading, I had thousands of hours watching charts and then just from the quote you can see why something should be somehow working or some or why not, these these are quite good points I would say. Without Without it, without the start, even going over the charts and see what are the consequences if something happens then or how the trends are evolving and so on.

I think that it is very hard to say that something should be working or not.

Jason Kurz 21:14: Like that addition because it's How many times do you see people come into trading and then they talk about backtesting and they're like, Oh, the win rate is this. Oh, I can't do it because of this. I have to win more. But if you really understand the way markets work, like you said, if you're really writing it down and really going through it, because I started the same way now that I'm thinking about it. And I've made this point before,

but I never thought about it like Mike just brought it up which is like, I know if I would have started just with straight back testing the way I do it now and spitting out things in thirty seconds, a couple minutes, no way, no way would I be able to hang on to those

but understanding it and going through the back test and seeing every single pullback, the depth of the pullbacks, the drawdowns for each strategy, writing that all down really kind of makes you in tune with the market. So I think that's a really important point. I hope people also grasp because in general, it's like so many people love backtesting. Like they love the idea of it.

Then once it gets into, I'm actually running a backtest, how do I actually run this backtest? They go, Oh, well, it's a lot of work. This kind of sucks. The system is like, it has a 40% win rate. Whereas if you know any of the best traders in the world, they have 40%, 50% win rates.

The thing that I think makes you a really good trader is just understanding the possibility that you're gonna be wrong a lot and figuring out how to be wrong a lot and make money at the same time. So once you figure that out, you make your drawdowns smaller, you make your winners big, you have positive expectancy working your way in your backtesting and your trading, that's how you make money.

It's very hard to make money just putting money in the markets and hoping that it's gonna work and thinking that you're gonna be right all the time, because you're gonna just lose money all the time that way. You're gonna hang on to losers and you're gonna lose everything like that.

Pavel Kýček 23:11: Yeah.

Mike 23:12: Yeah. Think Yeah. Exactly. There's one more, like, maybe nitty gritty, detail about backtest is, okay. You're gonna probably people are gonna let's say, if there's stocks or cryptos or something like that or they're taking things from the present that are available now and bringing them back. Can't do that. You can't you have to have a selection methodology that selects instruments available at the time.

So if people that are wanting to get in futures trading as an example, they're gonna leave out a lot of markets that aren't available today, and they're gonna start their test in 1980 or '85, whatever. Doesn't matter. But there are markets available then that are not around anymore. So you would have been trading them then, so you need to include them.

You need to have this screener that scans all the markets available and then goes into your when to buy, when to sell, how much to buy and sell, all those things. But I think a lot of people leave that out. They just take, oh, well, these are great stocks. Okay. These are great markets. Okay. Let's go backwards and test our rules on them. Well, why would you be trading why would you be selecting those in 1985?

Why would you be selecting those in 1995? You know, or five years ago, ten, you probably wouldn't. So Yeah. You need to have the selection methodology as well. You can't be bringing your your current markets back. You know? So it's like when you're starting your test, you're bringing them back from the future. You know? Can't do that. It's a Yeah. This a is great point because it reminds me of all these billboard ads. Right? Like,

Panelist 25:12: I mean, in India, we have a big mutual fund industry. Are always advertising. And what this is, if you had invested in this particular fund in 2014, by now you'll have this much. Like, let's say you put thousand dollars into this particular fund in 2014. Now that thousand dollars, like $15,000 or whatever. And that just exactly gets to this point, right? If why would you have bought that fund in 2014? Yeah.

You wouldn't have like I mean, that's that those ads are completely out of context because no one would have done what they ask you to. It's the same thing with, like, all these articles on how do you put thousand dollars in NVIDIA at this particular stage, you'd be a millionaire. Like, yeah. Sure. But you would have bought pets.com also maybe, and you would have gone broke. Right? Like, but those examples don't make it to the advertisements.

Andrew Swanscott 26:06: I think another point there that Mike was talking about is this survivorship bias. And, people say, I'm going to trade all the stocks that are in the S and P 500 now without realizing that the S and P 500 changes. And, some companies that are in there now have had a massive run, but they weren't always in the S and P 500. So it really overstates their backtest results. Right? And then they get to live trading, and it's not the same.

And so I think that's something that a lot of, especially stock traders who do algorithmic or systematic testing forget to consider.

Pavel Kýček 26:48: Yeah, stock and especially in these days on crypto, I can see it everywhere. Know, the survivorship bias is huge because in stocks for example you have Norgate, in Norgate you buy, basically pay a few bucks and you got your survivorship bios free database.

But in crypto there are some, there are already some solutions but they are not known and yeah this is a huge issue and you can also see the bots that are sold everywhere like bot on Solana, on some crypto that was moving like crazy in last three years or so. It make like, I don't know, 100000% or so. And yeah, buy it, buy it because you will definitely make money with it to Justin Solana, of course. Know? This is very yeah. This it's everywhere.

Survivorship bias everywhere. For example, on Nvidia or top five Nasdaq Nasdaq stocks, Like this is really big issue, what I can see.

Andrew Swanscott 27:59: Yep. So I just want to jump back to a comment in the chat a little bit earlier, which we I'm gonna put it up on the screen here. Let's see. We've all been here crypto, CryptoNick. I've just discovered I'm a curve fitting champion. So we just talked a little bit about survivorship bias and the impact that can have on our testing. But what about what are some techniques that you use to kind of try and avoid curve fitting too much?

Who wants to start with that one?

Jason Kurz 28:35: I could go. Yeah. Curve fitting, God, that's my worst problem at the beginning, which was I continue to try to build everything into how I wanted it to move. Like, and not to mention, like, it's a thing I always go to, it's like a, it's adapted from Narcotics Anonymous or Alcoholics Anonymous, the first, the first thing is I'm powerless. And it's really like, I'm powerless over market direction.

It's the thing I always like tell myself constantly. I have no idea, I can't make the market move. It's like trying to make a wave move in the ocean, you can't do it. So, that's one thing that you have to learn to let go of over everything else. And once you get that part of it, then you start to stop trying to curve fit all your systems and strategies.

Like, because I was talking to a trader the other day and I completely forgot about this because I used to do this too. And it's mind blowing to actually think about it because I'm like, I would never even think to do that today. That's insane. Which is I used to, when I first started trading, started a weekly timeframe and you go, Oh, that looks like a good trend. And then you go to a daily and you're like, Okay, that's a good trend.

And let's say you buy there. And then as it's like going down against you, you're looking at like five minute and one minute charts going like, oh, it could turn around here. Oh, look at that green candle. And like, it's like, it's insane that you do that, but like new traders do that all the time. And so like really figuring out like are you going to trade? How do you make money in these markets? Like, we're all different.

Like even everybody up here, like even some of us are systematized, some of us are not, some of us are fundamental, really it's discipline. So how do you, how does your strategy work? Are you going to be a breakout trader trend follower type? Are you going to be a mean reversion equity mainly trader? Do you trade Forex?

Do you trade all markets like Michael and I know, I think Andrew does as well, but Michael and I, we trade all the macro markets in the futures. And then once you figure that out, then you go, okay, well, my strategy is blank. I'm a breakout trader. I'm a mean reversion trader. I do this. And so once you start to do that, you really figure out where you fit in. And once you figure out where you fit in, you stop like worrying about everything else.

Because really like what happens is you look for media out there or somebody to tell you how you should be trading And you pay for all these services and they're going to tell you exactly how you should trade. This is how you make money. This isn't how you make money. Do it this way. This way wins every time. You know, like you'll hear all these things and then you're changing your strategies around, you're doing all types of stuff.

So, one is figure figure out out who who you you are. Are. Two Two is figure figure out a strategy that you really like, and stick with that strategy. Like figure out what type of trader you really are. Most people have no idea for many years, and then they're in this circle of, Hey, I want to fit this in. Hey, I'm looking at news to try to make sure. I'm in this trade.

I know it's a good trade, but I'm also looking at news every day to get out of this good trade. Because I want to take these profits. I want to run home. I mean, especially Mike and I, we're, I know we're in the cocoa trade. Coco, there's been a million times I can tell you that I wanted to run off of this trade. A million times I wanted to run off of this trade, but, know, take the money and run basically.

That's just in my brain, but my system never has a sell signal. So what do I do? Well, I just listen to my system. If I would have got out early, I would have got out months ago, I would have got out, 60% ago, it doesn't make sense to hang on to it. Like, you can't think of how far a trend can go. So for me, my best edge is I build systems that keep me in trades, that don't take me out of trades too early. I can stay long.

This way, I'm not out thinking myself all the time. And then also I have complete conviction in my strategy these days. I always make the joke that like, God could come down from the heavens to me, and he could sit right in front of my face, and he could be like, Jason, I know that Coco's gonna crash 50% tomorrow, you've gotta get out of that trade. And I would look right back at God and say, Well, that sucks. My system's still long.

I don't know what to tell you. You know, like, that's the type of conviction I have in my own strategy now, but that comes with time. It's not something you're gonna get over on day one or day two, but over time, you'll build that conviction. But first it starts with figuring out what type of trader you wanna be.

Panelist 33:14: Hey,

Panelist 33:17: folks, I'm back.

Andrew Swanscott 33:20: Hey, welcome, Ahem. So we were just talking about curve fitting and what of what do you do to try and avoid overfitting your strategies? Do you want to contribute anything to that?

Panelist 33:32: Sure. Yeah. Could you ask I'm sorry. Could you tell me the question again? I just had a little hard time hearing you there. Yeah, sure. So what do you do to avoid curve fitting or overfitting? So like just overtrading the market?

Andrew Swanscott 33:47: Sorry, what was that?

Panelist 33:49: Am I coming through okay? I was talking about- Yeah, you're coming through fine. It's actually, so it's

Jason Kurz 33:54: Think of it this way, Mayhem, like fitting Going out and finding news to fit your narrative and your strategy. Yeah. Yeah. Okay. So just like kind of leaning into your bias. Yeah. You're good. You're a good one for this one.

Panelist 34:07: Yeah. So my whole thing is that I cannot do anything unless price agrees with it. Like, that's my baseline. If price doesn't agree with it, it doesn't matter what I think. I could be thinking we're going to go into a nuclear holocaust and the S and P is going to negative infinity. I can't short it if price doesn't agree. It just doesn't work. Now, there are exceptions to this rule occasionally.

One of those exceptions, I would say, is when expectations are crazy. And for example, let's talk about Fed funds futures going into this year, pricing in five to six cuts. Now that sounded absurd to me, and I started to look for opportunities there. And that was where the narrative started to drive my examination of momentum to comport to that idea.

So, I started to see momentum flip into your note futures, and I started to fade them as a result of that. But the original idea was expectations don't make sense here. How do I trade that? Where is the opportunity? Then when momentum started to agree, I started to see on multiple timelines opportunities to fade that, and I decided to express that trade in two year note futures.

It's been a pretty good trade, but I usually like to let price lead and then try to make a determination as to why that might be happening. But I can't trade just based on biases. Because if I did that, I would have already blown myself up, like, 12 times this year. You know? Like, there's so many things. Like, wow. I would love to shorten video here. It sure seems like a bubble. Yeah. But I can't trade on that because price doesn't agree.

It just doesn't, and the call flows will eat you alive. The the amount of intensive speculation on that stock has pushed skew to levels we haven't seen since June. With SMCI before today and even after today, that skew is incredibly blown out. And yesterday, it was the highest it had ever been towards calls. Am I gonna go faded on that? No. I'm gonna miss that 20% down day because price didn't agree.

But it's still like you can see those imbalances, and you can see the narrative it might be setting up for. But knowing how to trade that is a whole other matter altogether. So I just like to kinda consider myself a very small fish in a very big sea full of thirsty and hungry sharks and whales, and I'm just gonna swim in their wake. Right? And so that's what price is to me.

When momentum is showing me that, whether it's on a relative strength or relative weakness basis in a pair trade or whether it's in a single instrument, that's the trade setup I feel more comfortable with. And, Jason, you and I have been talking about this for almost four years now. It's also how we build our systems to trade the market.

We're we like you said, like, god could go and tap you on the shoulder and say, Coco's crashing the negative infinity, and you'll say, well, that doesn't matter. My system hasn't given me a sell signal. I feel the same way. Like, at the end of the day, every like, here's another contextual, like, bit of anecdotal observation. Near the lows of 2009, everyone in the world was saying, we're all gonna die. We're going to S and P, 300.

The Krausing market's going into a depression. It's a spiral. The economy will never stop going down. But when price told me that the market was reversing, I didn't care because price was saying, you know what? Like, the market's looking forward at something else, and it sees something.

And then you had FASB come in, and, of course, they marked the bottom by suspending mark to market, engaging mark to make believe, and it was up up and away from there. NQE supporting liquidity. So to me, to kinda put it all into, like, a quote, if price doesn't agree, I'm not in the trade.

Andrew Swanscott 37:24: No. So I guess we would oh, sorry. Jason, did you wanna

Jason Kurz 37:28: No, was just saying, I agree. Like it's always has to start with the price. Like you really have to figure out, because think about it this way. Mike brought this up on my podcast recently. We were talking about the NASDAQ trade and, we all have, all those trend followers have very similar signals. So we're all getting these long signals early twenty twenty three to go along the NASDAQ.

And it's like, oh, okay, we're going to go along the NASDAQ now. It's basically like, at that point, everything looked like recession was happening. And like, if you just went outside, like, you'd be like, I don't wanna buy the NASDAQ. At the same time, like we have enough faith in our systems at this point to be like, well, I could be completely wrong about that.

And if I'm completely wrong, then I'm gonna have to just take the trade, call it a day, like, that's the thing you just really have to do. Think figuring out how you filter yourself, whether it's price action, whether it's sentiment, fundamentals, however you do it, like, you really have to filter yourself and almost have bumpers on your trading like Bumper Bullet.

Panelist 38:31: It's true.

Pavel Kýček 38:34: I think is that curve fitting has a lot to do with novice traders that are trying to get rich on historical data. This is really something that not many think about that they just never can trade the history. You know, this is also something what Mike was talking about, but when I'm talking with new traders, they are always thinking about historical data as if the future will be almost the same

which it never will be, it will be always very different. So this is one thing, If they would have a understanding that the future will be very different than the past, then I think they wouldn't try to overfit that much. They would still try to overfit. I think because at the start of trading career, I think almost everyone is trying to get the best equity curve like 45% degrees, left corner to upper right corner and so on.

But then for example, how I gather it of pure fitting was going with many different strategies and creating broad portfolios. I know that most of you are trend followers. I like trend following and momentum trading too. But for me, for example, it makes perfect sense to add mini version strategies to it because I can use or reuse the same capital thanks to the uncorrelated strategies in one portfolio.

And once I really started seeing the simplicity, you can build the stability of the overall portfolio through more uncorrelated strategies in the portfolio. Then I really went through extreme, extreme, I would say even oversimplification. For example, I never optimize any parameter in trading strategy. My strategies are like one entry condition, one exit condition and one regime filter, and then some ranking

and these kind of like sub conditions, I would say but really I'm trying to keep everything as simple as possible on one strategy level and going a little bit more crazy on or a little bit more advanced on the portfolio level. This is for example, how I really got rid totally of the need of some kind of overfitting or so.

Panelist 41:06: So this works this works a bit different when you're looking at fundamentals because so, the way I would look at it is, for instance I mean, the best example I can think of is all the oil investors who used to follow in 2022. So the initial bull case for oil was pretty simple. Your demand is at 103,000,000 barrels per day or whatever, and your supply is at 102, so you're drawing down inventories, prices going up.

It's a pretty simple fundamental case. Right? Then you had the Nord Stream to pipeline getting hit. Then you had the disruptions in the natural gas market getting attention. And then you just go on and on. At one point, the oil investors were talking about German power prices, Norwegian power prices, and those charts were part of the bull case for why oil will be in a bull market forever.

Now when you see this kind of creep of the stats that you look at to justify your case, that's when you know that you're kind of getting into that overfitting stage because you are adding more and more news items which aren't exactly relevant, but they kind of help you make your case. They are parallel to your case. And another example that I vividly recollect because it just stuck with me over the years.

So I used to follow a mining newsletter where they would absolutely nail small cap stocks. So they would routinely hit 10 baggers. And the August 2007 issue of that newsletter was all about how there was a shortage of drilling rigs, and they should prioritize the projects where the rigs are secured and the assay labs are secured.

Now when you get into this level of when you see at this level of detail, it definitely means that something is amiss, like a top is near. So when I look at someone who follows the fundamentals and then all of a sudden they go into a different set of fundamentals to make their case, that's when I kind of see that there is a little bit of overfitting going there. Because, like, with fundamentals, you can't really backtest a strategy. Right?

You can't download historic data and backtest, but you can definitely look for information creep.

Mike 43:33: Yeah. I think for a systematic or, fixed portfolio type of approach, if you have two rules, if you remove the third of the data, how would it change if if the if the performance statistics don't don't change very much, say, 10%. Change more than 10%, you might got you might have something not so strong. And the same thing with an out of sample, batch of markets or stocks or something like that. Like, oh, okay. Yeah.

This works great on, Nvidia and Bitcoin and whatever. I don't know. Some other names that, have, gone into this bubble like pattern over the past several years or something. Oh, okay. Let me let me try it. Let me send me some of your rules, and I'll I'll test it on some other things. And, of course, not probably I can have anywhere near the same performance because these rules are for these markets this time. You know? That's not anything.

That's not that's not a strategy for the long run. It is great, if you can get lucky, which you won't, in the short run. But trade it long enough, you'll probably lose. So the out of sample is really, really important and change start and end date and, remove remove a big chunk of data, from your test. And, if if you have performance all around clustered in the same area, you probably got something okay.

But and it's safe for, like, the I have a vendetta against the buy and hold, equities crowd. You know, the Nikkei just got back to its all time hot. I think today or yesterday or something like that. You know, the buy and hold crowd will never will never use that as as part of their portfolio. They'll they'll never put the Nikkei in there because it's not a good one. Yeah. Well, we don't know.

Maybe the maybe The UK, the Nikkei was was a special case or not? US has gone into multiple twenty year period drawdowns from time to time, and it will again. And maybe this American exceptionalism, all these tailwinds behind the markets here, the stocks here, this and that. Like, maybe maybe the next fifty years isn't the same. You you still know. So, just so just the, so like a like a buying holders like, well, look at look at US equities.

That's their that's their argument for doing it. Look at yeah. We have you look it over here too. You're negating this. You're actively not looking at this. You know? That's a mistake. So I think you can kinda tell, like I think experience helps with this too is if you have an answer already in your mind that you want, you're you're probably in trouble. You'll probably find yourself in trouble.

You have to really, I think, stay honest with let's see if this works. And versus I want this to work, so let's make it work. You know? Figure it out somehow. Like? No. You'd be in trouble, I think. Sooner or later, you'd be in trouble.

Pavel Kýček 47:31: Maybe try to break it. You know? Try to break it to the code of the strategy, why it shouldn't be working, then Then push it to work. This is maybe a very good idea.

Jason Kurz 47:44: Yeah. Nick said it from a the angle of statistics, and that makes sense, when we're when we're back testing curve fitting. And the one thing I could add in on that, which we did kind of cover that. We did Michael certainly went over that a little bit too. But when we're looking at statistics and curve fitting, okay, here's a great example. I worked for a fund at one point. They give me a call and they go, I have figured out this strategy.

Now, if you, if you know a lot of the funds and, and if you hear a lot about what they're doing right now, they're all doing this basis point trades between cash bonds and selling bond futures. And so basically, I get this call that says, hey, I figured out this system. It wins every single time. We're always make money. I go, that's impossible. And to me, that sounds like you're ready to blow up. And he goes, no. No. It it wins every time.

It never loses money. So I go, send me what you're doing and I'll back test it. So I sent him the back test, and I'm like, dude, this system actually blows up every four or five years. What are you talking about? He goes, oh, you just gotta forecast recessions and then you're fine. It's like, what the hell are you talking? Like, nobody can do that. Like, there's nobody in the world who can do that as it pertains to the market.

Like, the market is the leading indicator. Like, most of the time, the market goes down, starts crashing, or bonds do something crazy. Like, you can't possibly do that, but this is something, this is where curve fitting really happens in, and it even happens on a professional level.

Like people think like they're watching professional traders and they're looking at these huge hedge funds and they're like, oh, these people are smarter than everybody. There's plenty of dumb money there too. There's plenty of people who do the same curve fitting type strategies because they have vendettas and they're trying to make their own money and they're trying to do their own thing too.

So once again, it's a, how do you just get past that one thing? Get past like, your biases. And I think it starts with identifying them. Like, what are your character defects as it pertains to trading? Are your defects, do you really like to trade too big? Do you believe that trading huge will help you as a trader? Do you believe that you can do you you do you believe your strategy is better than everybody else's and doesn't lose much?

You know, whatever it is, you just have to figure out your biases, your own defect. Because really it's like, I have my own biases, which are stupid. Mine used to be front running my systems. That was my worst bias that I had forever. So every single strategy I had, I'd be like, okay, well, let's say it's, we'll just make it something simple, like a one hundred day high.

It'd be like eighty days up and then you'd be like, it'd be getting closer to that point. I'd be like, oh, well, I guess I'll just buy it here. Why not? I think it's gonna go up. It's probably gonna go up. Then you buy it and then it crashes the other direction. And then if you realize if you waited for your actual system and your signal, you'd have made a lot more money. So that's one of mine. You just have to figure out like, what is it for you?

And it's not just gonna be statistical because looking at it just in a statistical basis will never get you to what you're actually doing wrong. It'll just get you to like going, like statistics spit it out at you. Figure out a way to do it completely unbiased.

Like if you're creating a system and you're doing it on a spreadsheet and you're writing the numbers a little bit funny because you want the system to sound better than it really is, that's a problem. But nowadays, like TrendSpider, very easy, great software. Everybody can use it. Everybody has, nobody has an excuse not to be able to backtest anymore. And it's dynamic too. You can switch the parameters.

You can do a lot of things that are really cool with that, with TrendSpider at the moment. So put something in TrendSpider. I think they even have seven days free. I'm not, I'm not a, well, paid by TrendSpider by any means, but just telling you it's a, it's a good one. I think they have seven days free still, and if you go on there, you get your seven days free, and you could just go from there, and okay, these are my systems.

And at least understand what your systems are doing without you having to write it down and just depend on yourself to be unbiased, because I think that's also important because you have to write it down and be unbiased but also your systems have to figure out a way to write them down and also code them in an unbiased way and not front run your systems which we all have been there. I think we've all been there. Yeah.

Andrew Swanscott 52:17: So we're we're coming up to the top of the hour. So we'll start wrapping this up soon. Did anyone have any comments I wanted to add to Jason's there before we move on. No? Alright. Jason, can you say that site again? Crypto Nick is was it trendspider?

Jason Kurz 52:37: Yes. Trendspider I dot think actually, mayhem, do you have a do you have subscription, like, discount through them still? I remember you used to. I don't know if he's on right now. He did have another call. But No. I do. I have a discount link that I should be able to provide for that. I'll get that to you after we wrap up. Okay, cool.

Yeah, I'll post it on my page and then I'm sure Bam will post it on his, but yeah, mean, way, I know they used to do seven days free or something, and you can actually hit them up and just say, Hey, how do you back test? And they will actually sit down with you for an hour and help you learn how to back test using their system, which is pretty cool. You're not gonna find many places that do that.

Pavel Kýček 53:21: Or for example, what I can recommend is RealTest. I don't know if you guys know it, especially those that are trading stocks probably do because I went with Excel spreadsheet and TradeStation and I'm non programmer, so I learned the hard way, I would say. So TradeStation, then Amibroker and then RealTest and then we have some proprietary backend that we built in last two years.

But I have to say that RealTest is something that non programmer can learn in one day. I would say this is really if then logics, very simple ones, and almost everyone can start using it very, very simply. So this is like my recommendation.

Panelist 54:10: I also think TradingView has a lot of great tools for backtesting. PineScript is a very versatile language. They've got a lot of data sources, so you can pretty much backtest anything you can imagine. And then if you really want to get into the nitty gritty and go granular, believe it or not, the Python programming language has so many resources for programmers.

If you're an adapter, you know someone that is, they've got all the libraries built out for finance already. So like pretty much anything you can imagine, including some of the back testing tools are already available there on GitHub. So you can download this stuff. You can run it on your computer locally. There's a lot of good documentation on it. Obviously, it takes a much deeper learning curve to get familiar with it.

But once you do, you can go to fully automated systems pretty quickly from your backtesting if you want to.

Pavel Kýček 55:01: The problem with TradingView in my opinion is that you can backtest portfolios. I think you can trade only single like assets, would say like one stock or one coin or so. And I don't think that there is portfolio backtesting, but not sure about it.

Panelist 55:22: Yeah. I don't think there's portfolio backtesting on TradingView or TrendSpider.

Jason Kurz 55:26: Yeah. Most of do not. You have to really do that. Pavel, that's on what you just said. You can back test portfolios

Pavel Kýček 55:35: on that. Yeah. And very, very simply, real test, really. Know no no sponsorship or so. No affiliation,

Jason Kurz 55:41: but What's the name of it again? Because I'm gonna I'm gonna look into that too. That's cool. A real test. Let me

Pavel Kýček 55:47: share share it here in the chat. Nice.

Mike 55:54: I know it. They they seem to have a good community too of people sharing ideas and stuff and help you out.

Pavel Kýček 56:02: Marston Marston built it and he was in the Unknown Market Wizards in the last one. And this is the creator of the software, he's a real trader. He's like super helpful because we are backtesting many quite advanced crypto strategies there. And if we need something almost one day maximum until he makes the update. So really very, very, very good software, very super quick, super quick. For example, to TradeStation, it was much, much quicker.

Very good one.

Andrew Swanscott 56:41: And Mayhem has just posted a code for Transpider get 30% off and a twenty one day free trial. Thanks for sharing that mayhem. That's in the Yeah. In the chat in YouTube. I don't think that will show up in Twitter, but check that one out. Alright. Well, we're just at our time now. It went so quick today. I can't believe it. So how about we all share how people can get in touch with us or follow us? So Pavel, do you want to go first?

Pavel Kýček 57:11: Yeah. Well, basically I'm active on Twitter. So my nickname is Pkycek. Maybe we should write it somewhere because Czech language is not the best one for learning, but pkycek basically. That's my Twitter account.

Andrew Swanscott 57:34: There you go. I'll put it in the chat here. Thank you. Yeah, Drey says make it two hours. I think yeah, we should make it longer next time. Absolutely. Jason. Yeah,

Jason Kurz 57:48: I mean, maybe next time we could start, maybe start a little bit earlier or something, whatever. I think it's been a cool format, like just having Mike just hop in and whatever, just send, we'll send it out to people and they can hop in as it's convenient to them and hop out if I think this was great. It was a great start as we got a lot of big stuff planned too. It's really cool to just start out here. Got some great people.

This was a lot of fun. Like I said, I just wanted to bring together real people in this industry to have real discussions about trading. Next time I'll have more lights. It got dark here as I was talking and I was like, I'm trying to grab lights and put it on me. In Ohio, the weather changes every five seconds. I'm looking outside, there's like a blizzard right now, so whatever. It was just 60 degrees a couple of days ago.

But yeah, no, this was great. This was a lot of fun. I can't wait to do the next one. And just really the most important thing to me is just bringing like good information and from good people and getting it out to good people.

And I think this is really that, like we can try to really take our group of people that Andrew and I have been talking to for all this time and bring in like people who are actually in these markets, trading it, using their own capital and really passionate about it, like Mayhem and I always talk about like, how much we love the markets and how much we talk about it and stuff.

And it's like, it's great to just be around these really passionate people who really love doing this to just have conversations about it, because that's what we're here for and this is how people learn. And I'm always happy to help people learn how to trade. And how you can get ahold of me, I completely forgot. I have a Substack Against All Odds Research, it's aaoresearchsubstack.com. I also have website aoresearch.com.

I'm on Twitter, I'm very active on there, jasonp138. And then my YouTube channel, where you could see interviews with all of these guys. I've had most of these guys on as well as, Victor Sprandio, Jerry Parker. You know, we just had on seasonality trader, Almanac trader just came on the other day, Jeffrey Hirsch, Anthony Crudelli, the list goes on and on.

I've had a ton of people on there, Tom Basso, but all these things that we're talking about, systematic trading, Andrew has tons of great information on that. I have tons of great information on that stuff. So please use all of our resources and that's why we put them out there because we want people to learn how to trade. I wanna help them on their journey because I felt like I didn't get enough a lot of help on my journey.

So this helps other people on their journey, and that makes me really happy.

Andrew Swanscott 1:00:43: Very well said. Kashyap?

Panelist 1:00:46: Yeah. You I'm mostly active on Twitter. It's kashyap two eight six. That's k a s h y a p, like in my name, followed by the numerals two eight six on Twitter. I also have my own website, which is kashyapseeram.com, which is my full name over here.

Andrew Swanscott 1:01:06: Awesome. Thank you. Mike?

Mike 1:01:10: Yeah. You can find me I'll I'll post my Twitter link right here in the in the chat. And your podcast. Sorry?

Jason Kurz 1:01:20: Your podcast.

Mike 1:01:24: What about Spotify. Right? The link to it?

Jason Kurz 1:01:27: Yeah.

Mike 1:01:28: Okay. Yeah. Podcast too. The talking trends podcast with me and me and Jerry Parker. And you can find my blog through my Twitter handle too and website to my my business too. So go to Twitter, you'll find everything right there. All the links up, but I'll I'll post the link to the top and trends on Spotify and Apple too. So people I know people listen to both.

Andrew Swanscott 1:02:06: I think what I'll do is I'll get everybody's links and I'll put them in the description of the stream here so that people watching can find them easily. Thanks Mike and Mayhem, for dropping back in Yeah. After you have to go. That's awesome.

Panelist 1:02:22: Yeah. Appreciate you having me back in. This has been a lot of fun. Look forward to the next one.

Jason Kurz 1:02:26: Yeah. Definitely shout out your channels Mayhem before you get off.

Andrew Swanscott 1:02:30: Yeah. A lot of great research,

Jason Kurz 1:02:32: a lot of great educational videos, everything.

Panelist 1:02:36: Thank you so much for the awesome introduction. And yeah, so feel free to follow me on Twitter, Mayhem4Markets, Mayhem4Markets. Same on YouTube, Mayhem4Markets. You can just search markets and mayhem. It should come up in the top results there. Then you could find my work for shorter term trading, more systematic trading at traderaid.com. We have educational resources. We share ideas. We have a podcast.

You can search it on any major podcast service under the name Traderaid. We've also just developed a plug in for Bookmap that visualizes S and P options, flow, and positioning on the right side of Bookmap in near time. You can also find my work at Macrovisor where it's longer term swing trading and investing. The theme being macro meets momentum equals opportunity.

And we also have a podcast there you can search on Apple, Spotify and elsewhere under the same name, Macrovisor.

Andrew Swanscott 1:03:28: Awesome. I have to say, Mayhem, you've got a voice. You sound like a radio announcer or something. Have you got radio experience in a past life or?

Panelist 1:03:36: You know, I just listened to enough that it kind of became an impression and then I started emulating it. But no, thank you.

Andrew Swanscott 1:03:45: Alright. And then so we've done everyone right. I'm the last one in the list. You gotta go, Andrew. Better system trader. Come and check us out on YouTube, Twitter, Facebook kinda, but, no one's there these days. But yeah. Actually, I've got a I do a live show every two weeks on it's currently on Mondays in the evenings, EST. But Pavel's gonna be my next guest, and it's on let me check. Was it the twin I think it's the twenty sixth. Yeah.

February 26 at, I think, 3PM eastern. We're doing it a bit early because he's based in Europe. So I'm gonna get up early and do that one. Come along. It's live just like this, and but there's gonna be a full interview style format. We're gonna be talking about algorithmic crypto trading. So I'm looking forward to that and trying to uncover some of the secrets of the little algo trading.

So and then this show, of course, is gonna be every Friday at 7PM Eastern. Is that right, Jason? Seven.

Jason Kurz 1:04:43: 7PM Eastern. So it is 4PM. 4PM Eastern. Oh, sorry. 4PM.

Andrew Swanscott 1:04:49: 4PM. Right. Okay. Pm Eastern. 4PM Eastern. And he's

Jason Kurz 1:04:54: in Australia. Andrew and I are talking all the time. We're like, I have no idea exactly how Australian time works sometimes. You guys just make it up It's morning. See It gets crazy. So it's morning there. We're all

Panelist 1:05:09: everybody's He's the future. He's in the future. Andrew is Andrew in might be able to meet.

Andrew Swanscott 1:05:15: That's all right. So 4PM Eastern on every Friday and the panel may change, may stay the same. We don't know. We're going to mix it up a little bit. So please come and join us. And thanks to everyone on the panel today. This was awesome. We really appreciate your time. Anyone want to have closing thoughts or

Jason Kurz 1:05:33: comments or should we just get out of here? We can get out of here. I just wanted to say thanks for everybody coming. A lot of my buddies on here. Glad to see Mike, Kashyap, Mayhem. Great to meet you, Pavel. Andrew, as always, this is a lot of fun, Glad like we could bring this to fruition and I hope everybody enjoyed and it'll only get better over time.

Panelist 1:05:55: Great to meet you all, Mike, Pavel and Andrew, really cool to catch up and great to see you, Jason, and Cash App again.

Pavel Kýček 1:06:05: Thank Thank you for having me here. Thanks

Andrew Swanscott 1:06:09: for joining us. Have a good one. Enjoy your weekend. Bye.

Mike 1:06:12: Cheers. Bye. Bye.