Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Most traders fool themselves with backtests - GoshawkTrades on systematic trading
76 Plays1 day ago
In this episode, we’re joined by @GoshawkTrades, a systematic and algorithmic trader running 12+ live strategies across futures, equities, and crypto. We talk about how he got into trading, why he moved toward systematic strategies, what he got wrong early, and what his trading looks like today. We also cover where strategy ideas come from, how to test them properly, the most common backtesting mistakes, strategy decay, correlation across multiple algos, risk management, kill switches, quant vs discretionary trading, and how AI agents could change markets over the next few years.
00:00 Getting into trading and moving from discretionary to systematic
03:47 Shorting small cap gappers and finding his first profitable system
08:50 Moving from small caps into futures and crypto
13:44 Discretionary trading, FOMO, and scratching the itch responsibly
18:57 Where systematic strategy ideas come from
23:34 Building a portfolio across trend, mean reversion, and inefficiencies
31:41 Whether trend following still works in sideways markets
40:07 How he sizes and allocates between strategies
46:24 Dealing with drawdowns and knowing when to cut a strategy
53:33 The biggest mistakes newer algo traders make
01:05:50 Finding edge in new markets
Recommended
Transcript
Introduction and Guest Background
00:00:03
Speaker
you
00:00:12
Speaker
Welcome to a new episode of the InSilico Terminal podcast. Today with ah a face, not just like me talking into into random profile and picture, which is nice for for a change. And my guest today is, I'm probably going to pronounce the name wrong.
00:00:28
Speaker
go can Can you pronounce it actually? What's that's your name? ah Like channel name is GhostLogTrades, but my personal name is Manir.
Transition to Systematic Trading
00:00:37
Speaker
But yeah, very nice to be on this podcast. I've been watching it for a good while and you've many great episodes and you had a friend of mine on recently, which was Padma. So yeah um I'm excited to kind of see what we talk about.
00:00:47
Speaker
appreciate that. Can you maybe go go into a bit who you are and what you do for the viewers that don't know you already? sure. very high level primarily i run algorithmic strategies so primarily i run them in the crypto futures and equities markets um i also do um kind of like sub-sector stuff so maybe like small cap specific things in equities for example and i do slightly a bit of large caps as well but primarily everything nowadays is fully algorithmic so As of current date, so basically like May, I'm running about nine active strategies. I did actually have 12 before, but some of them had to close down, which we can kind of talk about in this podcast. But that's pretty much everything for how I got started. Mainly it was from the discretionary side of things. So I've been trading since...
00:01:29
Speaker
early slash mid 2018 so about seven years trading give or take um and originally i started fully just discretionary so i was you know just seeing if someone on youtube would tell me a setup i'll take the similar trade or i'd find a trade idea from youtube that sort of stuff yeah um and then i quickly kind of got more into the systematic route ah within about a year or two what what made you switch from uh discretionary to systematic Primarily because I suck at discretionary trading. I'm not very good at it. i Primarily, so on the discretionary side of things, I started with small caps. I actually went through a YouTuber, probably most people know as like HumbleTrader. She's a big kind YouTuber that kind of goes through discretionary trading. I bought into her kind of like group and was kind of learning from them. And one of the guys in that group called Luis, I think, he was trading small caps, just like fully only small caps. And I thought, hey, this is an interesting way to make money. Way more volatility, way more chance for me to become a millionaire. So, hey, let me like kind of mainly focus on that side of things. yeah That quickly led down to a road of just me being very unsuccessful. I never really got past maybe two weeks or three weeks of profitability.
Algorithm Development and Challenges
00:02:33
Speaker
And then from there, when I finally made the transition, which was in 2020, give or take, um it was mainly because I saw a couple of ah traders on Twitter that they were using kind of code and data for their own systematic ah process. And I know it sounds stupid because I think nowadays it's a bit more well-known, but back then I just really didn't think
00:02:51
Speaker
there was a data side of things to trading. I just thought it was all kind of like, you know, you had better instincts, you were better discretionary, and that's how you kind of traded. So for a long time, that was kind of what i was just stuck in. on For about a year and a half, I just thought I was really shit at trading. So it took a while.
00:03:06
Speaker
So what what kind of ah systematic things did you get in in early? And do do you have like any kind of ah coding background or are just... Yes, i um I've always been fascinated with making money online. So the first ways actually ever made money was creating micro servers back in like, God, I don't know, a very long time ago, it's like when I was 12 or something. And that's how i first made my first kind of like couple hundred dollars. And that's where I learned the value kind of skill of being able to code. And back then, you know, we didn't have AI or anything anything like that. So you just kind of learn from textbooks or online and things like that. um So that's where I started to mainly learn how to code. But I will say,
00:03:39
Speaker
um I did have to kind of bring my coding up to scratch since I started more trading because back then I was mainly focused on multiple languages. obviously there's a lot of it is like JavaScript and Java and stuff like that. But now I'm primarily just either Python, C++, plus plus learning a bit of Rust. I kind of suck at Rust, but apart from that is kind of the main focus.
00:03:58
Speaker
So what what kind of strategies did you run at the beginning? what What first caught your eye? At the beginning, systematic or discretionary? ah I guess discretionary, since you said you you sucked at it anyway, probably doesn't really matter. For systematic, the first one was a pretty much shawing the open on small cap tickers. When I first pulled some data on small gaps, I realized it was about like 70% of gappers closed right on the day, which is obviously a pretty big edge. You can pretty much just short that. And the main thing you had to control was your w risk management because some of these tickers would maybe move 10 20% in around the open.
00:04:32
Speaker
the open So as long as you didn't blow up during that time, you had a pretty high chance that it was going to close red and you could make money basically shorting it. um So once I kind of had that data, that was the best strategy by far and in my first kind of two years or so that was able to compound my my account quite a lot. I will say nowadays, it's become more crowded. Most people say it's stopped working, but I disagree. I still run some small cap strategies as well. Just constantly with trading, you always have to adapt. So some things are not going to work anymore as easily, and going have to maybe change some things here and there. But still, small caps is kind of like a great niche, and it probably attributes 10% to my overall portfolio, and that's kind of the the main part of the moment.
00:05:11
Speaker
So that's like the first thing that you ran that you run not you actually like made money with. And then how how did you continue? How did you come to run like 12 strategies consecutively?
00:05:22
Speaker
um So primarily in small caps, I don't know how much you kind of well know, or maybe the audience does. There's a couple of things that suck. I don't know. I don't really know anything. Oh, okay, okay. I'll break it down. So small caps primarily is where you're having a sub billion, sub 500 million market cap ticker. And it's normally gapping up on some very stupid news. So like, you know, they've cured cancer or they think they cured cancer or something like that. And a lot of times it's just bullshit. They're either replicating the same news they did maybe a couple of weeks ago, or they're just trying to pump up a stock so that they could, you know, dump some of those shares.
00:05:54
Speaker
So a lot of the time ah you could short it if it was maybe up 50, 20%, 100%. And then at by end of the day, you know, it'd be basically right on the day. um So that was the basic strategy for how you could kind of systemize it or why I moved away from it is there's a couple things that really suck. And one is halts. Unlike in crypto, you're in equity markets, right? If it moves too quickly, the whole market basically freezes or you know that particular ticker freezes. You can't get out. You can't enter. And that can be a really big risk problem to try and solve for because you can get very unlucky. Like maybe your stop was two cents above where it halted. You're now stuck in and it could you could move 10%. It can move 20%. To give kind of like a personal anecdote, I remember the worst trade or the most nervous I've ever
Diversification and Strategy Adaptation
00:06:35
Speaker
been. It wasn't even that much money, but I think i was trading maybe 10 grand at the time or something like that. I maybe had about... two or three grand on this one ticker it uh halted for news i wasn't able to get out of that ticker for i think four days so it's just four days of being stuck in a ticker that i had to pay for locates um and there was nothing i could do i didn't know when it was going reopen i didn't know if it was going reopen like 500 higher or if it was going to be you know dead stock so the market was halted for four days
00:06:59
Speaker
Yeah, for that particular ticker. Yeah. It's mainly when um I don't know all the complications from the halts, but primarily it's when if they've maybe done some fraudulent or there's been some discrepancy on the news article. So then the basically the NASDAQ or whoever holds that particular ticker and then goes, you know, we need to verify this. We need to check this to make sure there's nothing kind of illegal going on. But obviously during that time, if you're in a position, it is very scary to kind of like see what happened. So that was probably the worst experience. And ever since that kind of experience, I knew while small caps is great area to grow your small count and kind of like build up higher. Once you get, in my opinion, once you get to about like six figures or maybe above, um it becomes more of a a risk weighing system. Some people like I had ah the opportunity to talk to internal envy, which is a huge small cap trader. He's, you know consistently doing small caps and he does it with I I'm maybe misquoting here, but I think it's about 20 million or plus. So like you definitely can do it on bigger size, but you need that risk tolerance. Like I've seen at least from what he's posted on Twitter, he's gone into drawdowns of 10, 20%, you know, very quickly. And that that
00:08:01
Speaker
Personally, for me, I'm more like I'm fine being a bit slower and and kind of slowly growing the accounts. So I moved into futures quite soon after small caps. The main reason why is is futures, first, ah you don't really have common halts. There's lower volatility. And then two, you've still got amazing leverage for the sense that I can have very low intraday margin and still be able to run multiple strategies. I can have multiple positions open quite easily. And all those sort of let's say, more infrastructure things with small caps, they're a lot easier in futures, at least for running like trend-following strategies and things like that. So I moved into futures. And then ever from there, I realized, you know why am i only stopping at futures? I added in crypto, added in more large-cap strategies. And the whole reason for this is the main thing I kind of talk about on my channel is the...
00:08:46
Speaker
thing, at least as an algorithmic trader that you want to build towards is having uncorrelated bets or uncorrelated strategies. And the only way to really do that is you need multiple instruments, multiple different strategy types. So the more things you trade, normally the better. And that's kind of what I've worked towards.
00:09:01
Speaker
When you talk about futures, ah you mostly mean index futures, I guess, like NASDAQ and SPX? So before normally, but now I pretty much trade anything in futures. So like soybeans, pretty much anything like gold and obviously silver were the main ones. I know those those were big ah contributors to kind of P&L in the past couple of months. um And it's just because so much volatility. But primarily now I trade anything just because Especially when you're trading like NASDAQ and index. And then if you're trading like BTC and ETH as well at the same time, they're all kind of quite correlated. um So you do definitely need some things that, you are pretty uncorrelated for the most part. And especially even if they are slightly correlated, if the volatility is a lot less on them, that means you just have, you know, you can put some allocation more into a low volatility asset. So you don't have as much risk kind of tailing.
00:09:48
Speaker
Mm-hmm. It just kind of made me realize like how little I actually know about the other markets other than crypto. like Crypto, I don't know if it just seems that way from from being inside of it, but the the market mechanics are so simple in a way where it's like, this just a perp, it's just running indefinitely, you have like leverage on it, that's how it works. yeah And then I guess small caps are very similar to altcoins where it's just like, they also have insane volatility and they have news and whatever, but if the news is fake, they don't get halted, like the the market just keeps trading. so I knew about like halts and and stuff from... I actually kind of started trading... ah Well, my first experiences when GameStop happened. So like in... in When that? 2021. With like Wall Street bets and anything. And I remember reading about like market halts back then, but I thought it's only like a volatility thing or whatever. Like it just like if there's too much volatility, the market gets halted for like half an hour or like an hour or
Crypto Market Advantages
00:10:41
Speaker
something. I didn't know it could actually be poor. for this, that's pretty crazy. But it yeah it's interesting to hear about the yeah other kinds of markets as well.
00:10:49
Speaker
Yeah, I would say like one thing that hopefully, I think we're slowly kind of seeing is hopefully equity markets kind of transition more to a model similar to crypto in the sense that it's just continuous futures contracts. The most annoying thing really in trading futures is you have to move over to the new contract once your current contract acquires. Which um I know for like big funds and things like that, there's quite a lot of opportunities there because there's kind of like always a mispricing of some kind of kind, especially if you're doing like high frequency. But for most normal traders, it's just a hassle. It's just annoying, like having to switch over. And then additionally, it's not like necessarily the contract expires. And then on on that exact same day, you just move on to the new contract. It's normally that the older contract is are already still in play until there's enough volume on the new contract. So you kind of have to measure a volume on both contracts until the volume on the new contract is better or you know more than the old contract and you move over. So um there's just a lot of annoying things like that personally, at least from futures. And it's also why I loved kind of trading crypto in the sense where
00:11:48
Speaker
um I will still say there's quite a lot of nuances, at least in crypto, but for infrastructure side of things, you guys are all very developer-focused. So it's all very quite easy to integrate. ah The biggest thing I'll say is like your data is all basically free. Obviously, you can pay for high-quality data, but you can just get data from Binance or Hyperliquid or whatever, and you can easily plug in some small strategies or something you want to test, which is pretty awesome, at least from developer side of things.
00:12:11
Speaker
I feel like sometimes we forget, like because ah in crypto, especially when the market is sold, we all kind of love love to be bearish and like say, hey, this is all of kind of uses and and nothing is really like happening or whatever. But I think from a trading perspective, we sometimes forget that the experience is actually so much better than than like trading TradFi with all these these, like with the data being free, which you don't really have otherwise, and also with all the...
00:12:33
Speaker
like I can just go on Hyperliquid and trade the the gold perpetual
Discretionary vs. Systematic Trading
00:12:36
Speaker
and it's just like I don't have to worry about any futures, any contracts, what all all of that stuff. and like Even like options, most stocks and whatever I traded with like options a lot and they're way more complicated than perps. It's such an easy, like accessible experience and also like because there's no KYC and like everyone can like get to it. i think Maybe that's kind of like an underrated bull case that you don't really realize if you're like in crypto all the time because you're so used to it. But for normal people, it's like way more accessible and like easier to get into than TreadFi assets.
00:13:07
Speaker
Yeah, I would agree. And like I think Pedmo would be ashamed of me if I didn't plug kind of Hyperliquid slash him. Because like I remember when Hyperliquid first kind like ICO'd and it was a crazy day because we both went long basically the exact same time. i think he obviously had some points from before because he was using it earlier. But I primarily went long from that time. there was still one of my biggest singular trades with just longing Hyperliquid very early on. Obviously, I i didn't hold it as well as Pedma. I don't know if he spoke about it on the podcast that much, but I held it to, I think, like 36, 37, and then I kind of re-bought on some dips. But Hyperliquid was the first time, especially because no KYC, as you mentioned, and also the expansion of them, right? Like adding gold and even SPY with TradeYXE is insane. Like it's it's one of the, at least as a trade, it's one of the biggest developments, i would say, in crypto for quite a while, at least for for me personally.
00:13:57
Speaker
So you so you're still the do a couple of discretionary trades, I guess, because this is kind of like... I mean, I i would guess if there's like no no chart, no data to back up and you just buy something, it's not really a systematic or algorithmic decision. yeah Yeah. So um i I have a philosophy and I think Padman does as well that...
00:14:16
Speaker
we should always, you're always going to get an itch to place a discretionary trade. There's no way you you survive seven years in trading. There's no one trade you want to put on personally. So I always put like at least some of my allocation into discretionary bets. It's never going to be more than like two to maybe max like 5% if I'm really convicted. And that's just because if I don't do that, I'll probably put it on something worse. So if I if i know, hey, okay, I've got 5% allocation to some discretionary trades, I can look out for those, see if there's any good ones, especially with the whole like AI sector at the moment.
00:14:43
Speaker
My God, bro, seeing stuff gap up 20% every single day and you you miss it gets very annoying. So having at least like 5% say, I'm i'm just going to a bit of money on this particular trade and see what happens. Like it's fun to do. um I still would say my discretionary track record is is pretty shit. Probably the only best trades I've really made discretionary is Hyperliquid, Intel when Trump bought it,
00:15:04
Speaker
and oh bloom energy and that that's probably like the only ones that um have worked out very well for me but still because they're such a small size of the capital uh or the account sorry it it it's not like it's not it's not made me a millionaire like someone else you know trading intel or something like that um but is it's been at least a fun trade Is that kind of your take on the the whole ah discretionary versus versus quantitative debate as well that like pops up on on Twitter every couple of months where it's like all quant people are like, hey, never trade discretionary. This is like the stupidest thing you can do. And then there's like people that got rich from it that are like, hey, quants are stupid. they They just like think they're smart because they have all this data and whatever.
00:15:41
Speaker
i So i've I've had the benefit of, I run like a development agency. So we kind of like help traders with turning their strategies into algorithmic code basically. and Nowadays with AI that's become a lot easier, but still if you're doing something that's maybe more niche down, especially like DAS or other sort of platforms like that, we we kind of help out traders. And through that business, I've been able to work with a lot of amazing discretionary traders, or at least get to know them, see what they kind of do, and kind of help convert some of their process maybe into automation. um the thing I will say is I definitely won't ever be a person that says discretionary trading doesn't work. I've seen too many people do very, you know, incredibly well discretionary trading. I will just say it is more, at least how I think about it, it's it's more like a skill in in a sport, you know, some people are just naturally going to be way better at a certain sport, whether it be genetics, whether it be those sort of things. And they can perform very high level at that kind of side of things. And I wouldn't say
00:16:33
Speaker
Like discretionary trading is meant for every trader. I think the beautiful thing about trading is you can make money in pretty much like 100 different ways. You can do it systematically. You can do it in quant way. you You can do it through fucking meme coins if you wanted to, you know, like you can do it in any way you want. um But I will say discretionary trading from my experience and from a lot the traders I know, it's probably got the highest like hit rate for blowing up or for not doing very well long term. Normally the best people I've seen discretionary trading is that they realize they got maybe a bit lucky on a position. Like I know a lot of thing about discretionary trading, right, is you put in a huge concentrated bet into something you're very convicted into or you size up a lot more on that. Sometimes that pays off, you know, like if you if you went really long on Intel or something like that, you you could make millions just by it. But at some point you have to realize, one, there's not going that many good opportunities that are going to appear every single year. And then two, are you going to be able to catch all of those?
00:17:21
Speaker
Most likely not. So you kind of have to make sure that you slowly size into something that's maybe more long term. And probably the best trader I know that's is the that's kind of done this is Lucas. He's going to be in I think, the new Unknown Market Wizards book. um well the new I think he's called like New Generation Book or something. But he started in small caps, made a ton of money in small caps, mainly through 2020 and things like that. And then he more transitioned into large cap plays, lower volatility plays, and obviously has allocated his kind of like a portfolio now because it's in, I think, multi millions or maybe even 100 into a lot of different sectors. And I think that is kind of the most stable way to grow is that once you realize your account has grown enough to
00:17:59
Speaker
you don't need to be taking, you know, 100% conviction on a particular trade to kind of sustain your life or anything like that. I would say going out into different ways is normally the better method. I think that's a quite quite reasonable take. I think also what you said before with ah like acknowledging that there is always like this itch to be like do something or like be in a position, especially like when you are in markets all the time or when you watch markets all the time, like everyone will experience FOMO. Everyone saw all the, the ah the what's it called? memory, like all all the the memory stuff and whatever that like pumped now. And then you see all the Twitter posts and people are talking about if you bought this yesterday, then you you would have outperformed the S&P for five years or whatever. And it's kind of like, maybe I should do something. So it's like good that you that you have this method of like allocating a little bit and just like ah scratching that itch, but the also like just focusing mainly on on what
Portfolio Evolution and Experimentation
00:18:53
Speaker
you're good at. And I want to ask a bit more about like
00:18:56
Speaker
um where Where do your ideas come from and what kind of strategies? You mentioned you run across ah multiple different markets now. what What kind of strategies do you use ah if you can like talk about that?
00:19:08
Speaker
Yeah, sure. at least for where ideas come from. It's kind of changed ah over the kind of trading career, I guess. Initially, it was primarily all from other traders or books or like papers and things like that. Mainly just because i didn't really understand what made different strategies work or different strategy types, really. um There's a great quote.
00:19:28
Speaker
I'm going to forget who it was by, but primarily he was saying like everything around trading has been discovered. like most strategy types, all things that have been done have been done in X markets. And then really the biggest factor is transitioning that into a newer market.
00:19:41
Speaker
Or if that particular opportunity kind of appears, you know that, hey, it's this kind of strategy type. I can kind of employ that and and it's going to work well. But initially, primarily it was from books and people on Twitter. I'd still say Twitter is probably one of the best places for any trader. If you're not on Twitter, it just... i don't know. The space here is amazing. There's a lot of things, a lot of new ideas. Obviously, there's a lot of crap as well, but there's a lot of good things you can take inspiration from and kind of like learn from. For books, because I know people ask, like, what books? um I would always recommend Advanced Futures Strategies by Robert Carver. It's a great book. It is a bit more...
00:20:12
Speaker
like some parts are a bit more complicated or convoluted, but overall the ideas in that book are very sound and they've worked well in crypto, they've worked well in pretty much all markets. Cause primarily what he's discussing in that book is very simple kind of trend following strategies. I think there's like one or two mean reversion strategies, but most of the time, a lot of it's around trend and trending has worked in all markets. So for my own portfolio,
00:20:33
Speaker
It's one thing I've tried to fix, but I'm always going to be a bit more allocated than I want to be to trend. Most of my strategies going more trend focused. And that's just because most of the time, like 99% of the time, that's where most of your returns come from. It means you catch a trend in a particular market. It trends up for a long time.
00:20:50
Speaker
And you make a good amount of money on it. Nowadays, for where my ideas come from, most of the time it starts with me looking at my portfolio and seeing what current strategies I have. And then where am I lacking? Or where i have do I have too much correlation? um i've for the past couple of months. I was discussing kind of my YouTube that I was potentially going to remove one strategy and I actually ultimately did. And it was merely because that strategy just had very high correlation for about a quarter and a bit, about like two extra months or so. Whereas the years before that i'd been running it really never had very high correlation, which meant it was fine to put in the portfolio. But since that correlation changed and also kind of sustained for a good amount of time,
00:21:27
Speaker
meant no longer that I really needed that strategy. I had other strategies that were correlated to it. Might as well just remove it and have more the allocation on the other strategy, for example. So nowadays, I'm more looking the portfolio saying, OK, I have way too many trend strategies. I need to come some new ideas. And that's also a great thing about crypto in the sense that you can do things like funding rate arbitrages. You can do things that just more obscure, which are normally going to uncorrelated to your main portfolio. The only downside of those particular strategies is normally they're going to be short term or they're not going last very long. Or they're to be really annoying to like do like funding rate arbitrage is quite annoying infrastructure wise to of like actually pull off. But those are great ways to kind of add to the portfolio. And then really now, like I don't want to sound like it's all roses and like easy, but because I have around 10 strategies,
00:22:13
Speaker
I don't really need to be adding, you know, 20 more or 30 more strategies for it to actually work long term. Most of the time now it's going to be a lot of patience, making sure I don't blow up on anything, making sure risk is well allocated. And then, of course, like adapting if a strategy you know dies or if I need to add a new one, for example. um I think that's kind of like the most point for at least how I get ideas.
00:22:33
Speaker
So you're you're kind of in the sweet spot of like how many strategies you you want to run or you think is ah makes sense? Yeah. um So there's a great kind of graph on there um around like diminishing returns or like adding more strategies. For the most part, once you get, I think it's above like 12 or 15, your diminishing returns per strategy added going to be very like low. You're not going to really get much for adding that new strategy. um Most of the time the biggest gains for adding new strategies is going to be around from like two to five. So if you're going from one strategy to five strategies, that's a huge jump. You're way less correlated to any one strategy. And you also probably have new markets, you have new instruments. So your X-E curve is often going to be a lot smoother than if you're just running one strategy, for example. But when you get above like 10,
00:23:16
Speaker
it gets really hard to find things that are uncorrelated. I see some people and they say like they're running 50 strategies. Maybe they are. i'm not necessarily calling out bullshit. But a lot of the times, if you looked at their correlation matrix, they're going to be quite correlated. or they're to be doing very similar things. And like do you really need four strategies that are doing basically the same idea? you You could just probably put that into one strategy and then have it you know allocated to whatever different instruments. So um a lot of times, at least for me, that's what I try and kind of focus on.
00:23:45
Speaker
What makes your strategies different? Like how do you um differentiate between... because Because you said basically most of the stuff is like trend following. So and do do you have like multiple different trend following strategies that like work on maybe, I don't know different timeframes or like across different markets or how how can I imagine that?
00:24:03
Speaker
Yeah, good question. um Most of the time, it's it's going to be based on time frame. So that's going to one obviously key one. Obviously doing anything intraday to daily, while there's going to be sometimes slight correlation, I guess if you're doing on the same instruments, most of the time they're going to be a bit different. Though I know don't ever really go to like five minute or one minute time frames or things like that. For the most part, I've never really succeeded there. I also think there's just so much noise and edge. I mean, like a competition, sorry. You're never really to be able to to really beat in those particular areas. So most of the time, I'm kind of like four hour daily, weekly. um I'll sometimes do like monthly if I really wanted to on a particular thing. But most the time be between
Balancing Strategies and Market Inefficiencies
00:24:40
Speaker
those. And then for the main way I differentiate is just doing different instruments. Most of the time, not like, for example, like if you're trading soybeans and like BTC, they're not going to be very correlated. So you can kind of have slightly different trend following strategies are going to be running on those. Because ah one thing I often see is like a misconception is that your strategy has to work on multiple markets.
00:25:01
Speaker
I've ah personally rarely found that to be the case. Like doing a trend-flowing strategy on BTC, BTC has such different characteristics, such different volatility than, let's just go back like soybean example, right? That's so different um that in the case that you're trying to find one strategy that works on both, it's going to be very rare. So most of the time, it's going to be a separate trend-flowing strategy, maybe for, you know a group of commodities that more you trade similar-ish together, and then it'll be different for like, you know, crypto, for example. But then of course I do run mean reversion strategies as well.
00:25:29
Speaker
um I more kind of treat mean reversion strategies as kind of insurance. Most of the time you're you're goingnna it's going to be hard to find a mean reversion strategy that is doing 100, 200% returns over a couple of years. It's going to pretty difficult. um But they can be great things for padding drawdowns because most of the time um this is slight tangent, but it is kind of important. A lot of times people try and find and an ultimate indicator to tell you if the market is trending or if it's chopping. From my seven years of experience, obviously I'm not the smartest dude ever, but I've never found a single indicator that just says, hey, this is a chopping market, this is a trend marking. Most of the time it's to be a lagging indicator or it's going something that necessarily ah overcomplicates the the whole thing and then kind of ruins the over strategy anyway. So most of the time my simplest kind of metrics are just
00:26:12
Speaker
If it's above the 200 MA or above the 50 MA, then it's probably trending. If it's kind of in between those, then it's going chopping. And these kind of like very simple kind of regime testers just allow me to go, okay, don't take traceit like mean reversion trace during this time. It's not to be perfect. But what it can do, and kind of the point I'm coming back to is that if you're running trend and mean reversion at the same time, even if they're not detecting trend and chop perfectly, they could kind of balance each other out so that you're not... um getting just a huge drawdown if, you know, it's completely dropping and stuff like that. So that's kind of how i split the portfolio. So it's probably the top of my head, it's probably like 35% mean reversion, 40-ish percent trend, and then kind of remainder is going to obscure stuff. So like arbitrage or things that kind of come around so that i can kind of take advantage of because
00:26:59
Speaker
One thing I'd like to mention is that there's going to be inefficiencies that pop up every you once in while. They're not gonna be that common, though. oh Most of the time, a lot of these strategies, what I'd call them is like risk premiums. So I can collect some money kind of through those and they can run for a very long time. They've worked for years. They probably will work for years still. But when an inefficiency comes up like... um That's a good recent example. don't Like Polymarket was a good example. I didn't really take that much advantage of it. I know a couple of dudes that did really well. But for me, like one great one was like the Jelly Jelly trade ages ago with Hyperliquid. That was quite a clear and while may look discretionary, ah you could actually do that in a very systematic route just because there two different kind of counterparties kind of like having risk on each other.
00:27:40
Speaker
So it really easy trade you could kind of take exploiter exploit of and that was then able to add, you know, 5-10% to the portfolio returns, though it only lasted like three days. and So those sort of things, once they pop up, I always want to have some capital free so I can, you know, ah try and trade those or kind of exploit them.
00:27:57
Speaker
Apart from these like ah one-off instances, do you execute your other strategies all automatically or is there any like manual trading involved? ah No, the there all the risk premium stuff, I would never really say it's worth to do manual. I just have i have it automated, especially now with like AI. Even if you've got like a bit of coding knowledge, you can kind of probably do quite well for most of these kind of basic trend following strategies. um So I would always automate all of those. The rest of them,
00:28:23
Speaker
Even if it is something I could maybe do quicker manually, i would still prefer to do it automated, at least for me personally. I know a couple of traders that more trust their own manual executions. so They'll still do it manually, especially if you can kind of like get better speeds out to like actually exploiting it. But for me personally, um i I'm decent coding. And also nowadays, like ai is pretty fast with stuff. So I prefer to have it all automated just because I i don't really want to be stuck ah in a position trying to manage it. I'm very bad at it. I know that's not my strength. So I try and stay away from it as long as possible and just kind of let the computer do do do my thinking.
00:28:56
Speaker
do Do you have an example of a strategy that maybe used to be very lucrative for the past and like stopped working? ah Yeah, to be honest, it maybe he's actually still works. I think ah Scott Phillips actually mentions it as well. um I guess it feels like I'm kind of stealing his idea, but I'll mention it. I'll try and mention another idea as well. But one I did at the trade was around shorting all these meme coins as they were kind of like ICOing on like Binance and stuff like that. The main reason why was because I came from the small cap niche. It was basically the same thing as small caps. It was it was some dumbass, you know, stock ICOing or gapping up a huge amount. You could basically just short it and then wait a couple of days basically. and The only different thing really with those was that
00:29:37
Speaker
you um waited out kind of the first maybe day or first day or two, shorted it, waited on the backside, and then basically held it for like two weeks or something like that. And that worked very well for a good period of time. The only thing I'll say with that strategy is as with small caps and kind of like any mean reversion sort of strategy, you're going to get a huge whipsaw when that regime ah changes. So i I can't remember what day it was, but like BTC ran a bit and basically all meme coins ran a bit as well. like The whole market kind of like changed for like a week or so. And I got not ruined, but like I gave ah a lot of P&L back on those because I was still shorting those meme coins. And that's very similar to small caps where small caps will work beautifully for like two weeks. You'll think you're a god, you're you're making money every single day. And then for you know a week and a half, you just can't make a single
Market Volatility and Institutional Influence
00:30:21
Speaker
penny. So um it's going to be kind of a flip-flop between those. But that strategy...
00:30:25
Speaker
I bet still works nowadays. If like if BTC is in a downtrend, you could basically just short all alts and kind of like meme coin stuff or anything that's really junk and you can do very well. So that's one that that works. And then for my own personal idea,
00:30:40
Speaker
I don't trade that many like pure inefficiencies that have died. One okay one ah i one I'd put is there's an there's an indicator called internal bar strength. um You can actually use that indicator very well on most indexes slash commodities. doesn't work as well in crypto, but primarily you can use it on a daily chart and look for basically what the indicator shows is like how much range on that particular bar was like to the downside or upside, like in a simple manner. So you can see if the if the like a particular indice closed really far down on the bar and was like a really big red bar, most of the time in indices, that's going to revert. So you can basically buy those and then you'll you'll get basically free money on the next couple of days. But of course, there's some times where it doesn't revert, so you have to kind of manage that risk. But that can be a very simple strategy that works quite well. But personally for me, I haven't had that many strategies die um die on me. The main reason being is...
00:31:31
Speaker
ah most of my strategies are going kind of risk premiums. I'm not really inventing anything new. And then the ones that have died have lost in maybe like two days. I've never really found something that's worked for a month or like two months. And I've been just exploring every single day. There's smarter dudes out there that can definitely do that. But for me, I kind of stay in what I know I can kind of do and then just kind of build from there.
00:31:52
Speaker
Do you think um that's something that just came to my mind? Because the the market, like we've talked a lot about trend trend following strategies and with many people that are more on the the algorithmic side, they tend to do lot of trend following stuff.
00:32:07
Speaker
And the market that we've seen for the last, I don't know, 14 years or whatever, ah since the great financial crisis has ah subsided basically, has been up only forever. So like in ah in a huge uptrend since then. And what do you think would happen to those strategies? like Would they still be as effective if we went into stagflationary period like the 80s or like the the Japanese market where it's just like sideways for 40 years or something? Do you think you could still find the edge in them?
00:32:34
Speaker
It's good question. um This would be a complete guess, to be honest. Like, I haven't necessarily tested that far back on those. But for what I would imagine is in most trend following strategies, either way, you can do quite well, even if the market is trending down. If it was pure chop, like it really wasn't moving at all, um I would be quite surprised just because still nowadays, the biggest thing that's really changed over the years, and like this is the thing you normally have to model for, is that volatility has just increasingly risen. Like if you compare even like 2018, 2017 volatility, it is nothing compared to these days. Now obviously nowadays it's a bit because of Trump. He likes doing a lot of crazy shit. you You get a lot of volatility. But even during like Biden, there was way more volatility during Biden's period than if there was in you know like 2015 or 2012 something like that. So that I don't think is necessarily going to change mainly because as I don't know if you guys know, but like the PDT rule is getting kind of removed from stocks, which means you don't need 25K anymore to kind of like trade or like day trade. They based it on like some risk assessment, but basically it means you can trade a less amount of money and more people can be able to trade. Additionally, we're seeing things like Cauchy, Polymarket.
00:33:42
Speaker
Overall, the trend has just been that there's going more people gambling. And more people gambling means probably more volume, means more volatility. There's going to more kind of like fake outs on either side because most people are going to be you know trying to trade everything. um So I would be very surprised if we went into like a full stagnation market and just nothing trended or anything like that. um Additionally, let's say indices did go into kind of stagnation.
00:34:02
Speaker
Something is going trending. Like something somewhere is going to be trending. So the whole point of trend following markets or trend following strategies is you are ah across a wide range of things. Maybe it could be that you know Korea's market is fucking going insane and you can along that. Or it'd be you know different areas of different worlds, at different worlds, different like countries. And you can kind of go through those. So I would say as long as you're trend following across a wide range of assets and instruments, you're probably going to be fine. I couldn't really see any reason why.
00:34:30
Speaker
Yeah, that makes sense. i've just like is it actually I didn't know that volatility was actually rising throughout the last couple of years. I guess it kind of makes sense if you look at the market, but I feel like anytime I i listen or I see any people talking about options or like the options market makers or whatever, they're always like,
00:34:48
Speaker
Because i don't actually know how this saudi mechanic works because I'm too too stupid to like properly understand it. but But I think because you like sell options on all the all the stuff, it like dampens volatility. So like ah basically also what happened on on Bitcoin, or like as as assets grow, that like the volatility subsides and becomes less. like bitcoin became a lot bigger but it used to be like a 50 vol asset and now it's like a 30 vol asset and it's kind of like happening to all the indices like the the the vix is also like always trending down through of through a longer amount of time so how does that like work with you saying that the volatility is actually rising
00:35:23
Speaker
I will say I don't trade options. i'm I'm not, I'm saying like you not smart enough to really understand full options side of things, but I mean at least mean on like volatility and volume. So like, for example, if if you take a good really good example actually you of this is like small cap strategies and equities. Back in like 2012, no one was trading small,
00:35:40
Speaker
tickers slash small companies everyone was mainly into root chips there wasn't really much volume slash uh kind of like volatility happening in 2012. 2012 like if you got ticker that moved 20 that would be you know quite surprising it would be a big move it would be like a big thing nowadays nowadays you can have things run a thousand percent in a day and it's not common, but it's not even like that unforeseen, you know? um By that, I mean, is like the amount of volatility across the market, I think has definitely increased necessarily, maybe like on SPY or QQQ, I 100% on BCC, BCC has definitely become less volatile, but you're also maybe like, it depends what you're referring to, like comparing BCC when it was like 20,000 or something, it's very different to now that it's like an $80,000 asset. Also, the biggest thing that's changed there the,
00:36:26
Speaker
this is at least in my opinion, like now you have way more institutions market making for BTC and things like that. A lot more have stepped in to where that's always going to damper volatility because you just have ah better liquidity on both sides of the book. Whereas before, you know, when I was trading at maybe 10,000, 20,000, there's not going to be institutions necessarily market making for that or not necessarily as big. um So,
00:36:48
Speaker
That's just my opinion, at least for it. Like ah someone can pull up the the full research paper and kind of maps behind it. But at least from personal experience, nowadays volatility is is definitely increased, at least for me. i definitely like I definitely feel like you're right about that just from like a ah looking at the markets. and it kind of see like i'm I'm wondering a bit if that's a a phenomenon of us being such a big speculation thing. where it's like If you look at the past, like mostly when this happened was like during Tulip Mania or like I guess in the in the twenty s I don't really know, like when when the market like went really crazy and like everything went up at once. and like
00:37:23
Speaker
everything is going insane and then there's this crash and then everyone is is kind dead and like no one has any money left anymore so that the market doesn't really do anything for a while. um But it's interesting to think about it because it's like also an an accessibility thing, as as I said before. like Nowadays, I mean in 2012, who was trading? you know like Who really had access to to to use the markets? Pretty much like only people in finance or people are that like were accredited investors or whatever, you need to to like trade back then. but like Nowadays, everyone is trading everything. So you have like way more people, as you said, being there, and being in the market, being like there to create this this volatility. And I wonder little bit like what we're talking about, how that is going to play out over the the next couple of years. like
00:38:10
Speaker
Because but it probably won't end up with everyone becoming rich and we're just like going to see another Great Depression or whatever. for Probably not. But like, okay, let me actually put this maybe one one way better. Especially i like, I wish I could kind of just like do some research now on this and kind of check. But primarily, ah volatility on like singular tickers and things like that have definitely increased at least like, one easy example, right? A lot of these AI sectors now could this volatility be um like it like, it makes sense for the fundamentals. potentially but we've seen like dot com kind of moves in like a couple of days on some of these tickets right like i've i've never seen a ticket like intel bloom uh so many of the data center stocks move 20 on a day the next day move 20 the next day move another 10 it's absurd like there's just not that many times i can think over the past even seven years now obviously i haven't been training for incredibly a long time but um
00:39:04
Speaker
it' It's been quite a long time since we've seen that kind of consistent in in momentum and moves across like multiple different markets as well. It's kind of like alt season 2021 in crypto. Yes. With like yeah actually like billion dollar companies and not just like random random shit coins.
00:39:22
Speaker
Yeah, it's absurd. Like, especially the one thing I will say, like, this kind of goes back to the, I guess, discretionary and algorithmic side of the point. A key thing, if you're going be algorithmic, if there's something you kind of more want to do, you have to realize you're going be able to catch these kind of discretionary sort of trades. It's just a downside of it. I'm not going to necessarily to catch AI sector, you know, booming and catch all the best thesis plays. But I understand that, hey, okay, maybe I'm going to do better the next couple of years. And it's going to bit more stable because running a portfolio
Strategy Management and Daily Routine
00:39:50
Speaker
like this. And I think also that's a...
00:39:52
Speaker
big thing that draws people to discretionary trading is the idea that you can get really outsized returns. Whereas once you become more algorithmic and systematic, you realize, you know, if I if i can do 20% on the year, 30% on the year, like that's actually really good year. Where most discretionary traders are thinking, you know, I want 500% in a single year. want to flip my account into millions. um But most of the time, for some people that can work out, but a lot of times results in kind blowing up.
00:40:18
Speaker
how do How do you decide how to like allocate and size between your strategies? Yeah, primarily it's just a volatility sizing. So, you know, obviously if I'm trading like I guess crude oil now is like a lot more volatility, but back back in the day crude oil wasn't having as much volatility. So like those sort of things you would size a bit higher because that's more volatility. Whereas if I was sizing something for like QQQ or something like that, I'm going to be sizing a bit less because it's just way more volatility. So those particular strategies, that's normally how sizing within the instrument level and then also within the strategy level. So it'll be volatility sizing on each in and particular trade based on the asset. And then also overall in the portfolio per strategy is going to be sized based on that as well. um There's more complex methods, like methods to do it. But at least for me, for the amount of capital,
00:41:04
Speaker
to do the extra methods or the more complex methods were never really worth it. And I think this is a small side topic of that. A lot of times, especially when you're building things in algorithmic trading, you're going to see that there is an optimal method that you could work towards. But most of the time, that optimal method maybe gets you a couple of more basis points above the the previous method that's a lot simpler. And while that's definitely worth it, if you're managing you know hundreds of millions or you know billions and stuff like that, because each basis point is is worth so much, when you're doing 100k account or fifty k account or whatever it's it's not really that worth to spend all that extra time on infrastructure just to get a very small increment and also another side of things is sometimes these methods can be very annoying to just even maintain so you know maybe you build it once but then you have to maintain it because something happens or xxx kind of thing happens and that can be a real downside especially for
00:41:53
Speaker
each unit of time, at least how I like to think about it, when I wake up on the day, I want to see what's the most productive thing I can actually do that's going to move me forward, at least in my portfolio or something like that. And a lot of times it's really, really like fun to work on a new complex problem and like, you know, spend five hours trying to figure this out. But realistically, it's not going to add anything to the portfolio. So most of the time I try try and focus on, like okay, what's the main things I need to work on today and then kind of go through those.
00:42:18
Speaker
what What does that look like? Like what is your day, a typical day for you look like nowadays? My day recently has looked ah kind of crazy just because normally I'm based in Dubai. So I've been traveling around Asia for a bit with this whole around war stuff. But traditionally, if I was in Dubai, like waking up, um I'd normally wake up, have like very small kind of like just morning workout or just like ah move around. i used to be fat as fuck. So like I've slowly was losing weight. So in the morning, I'd normally do a workout or move around. And then by that time, I just get straight onto the desk and
00:42:48
Speaker
First thing I'd normally open is ah emails. Just because I run a couple of businesses now, um I'll normally have some people that I need to respond to, things like that. And then from there, after that, I'll be doing kind of my first coding project. So whether that be coding a new strategy, backtesting something, looking for a new idea, and that'll normally be most of the day, to be honest. um And then that will probably be about five to seven hours, depending on what I'm working on. And then in between there, they'll be, you know, having some food and shit like that. But for the most part, it's working on one or two singular projects during that day and then just spending time on that. Additionally, if I have to code something for someone or anything like that, I'll obviously allocate time between there of coding on that sort of things. But I will say for algorithmic trading,
00:43:29
Speaker
A lot of the work is upfront. Like once you have the strategies up and running, there's not that much stuff you need to do on a daily basis. yeah um I don't want to make it sound like you're you know on a laptop on a beach and you're just making money. It's definitely not like that. But it is, ah for the most part, there's not much maintenance um if you're running daily or weekly strategies. you're running says high frequency something like that, 100%, there's a lot involve of maintenance maintenance you have to kind of do on a daily basis. But for daily, weekly, you know, trend strategies, there's not much I really have to do on that part of things. Most the time, it's either testing a new idea, researching a new kind of like method or seeing if this maybe would improve something. So a lot of tinkering, a lot of being curious, I'd say is kind of like the main part.
00:44:10
Speaker
How much time do you like spend on these things versus like what the payoff is? Because I imagine it's quite difficult to find something new that's actually like worthwhile. And you're just like ah looking at these things. and and like How often do you find something that's actually like worth after the time?
00:44:28
Speaker
probably Like when i when I first started, it was if I probably tested like 12 ideas, maybe like one, like slightly worked or like was kind of close to working. um Nowadays, more the idea, like I have a way better filter nowadays just because ive I've tested so many strategies. Nowadays, I'm like probably one in five, one in six. But for for me to even find six is going to a lot more rare. Like a lot of the time it's going be me scrolling through stuff or reading about stuff and thinking that's going to work for me or that like that that doesn't really make sense. um So most of those times I'm not really testing it that much. But initially when I started, I'd be testing everything.
00:45:04
Speaker
And I'd recommend that to anyone because the only way you build that kind of filter of what slightly seems to work or what at least gives you... curiosity around just takes a lot of time. it takes It takes a ton of hours. So nowadays the payoff is is decent, but I will say, as I kind of mentioned earlier, you know, going from 10 strategies and above has been incredibly difficult. um I just haven't really found or been able to maintain the amount of new strategies I've kind of been adding every time as I used to in kind of previous years. So a lot of the time nowadays it's uh making small tweaks on current strategies to slightly improve them like if i can do a volatility measure a bit better or if i could maybe optimize an exit a bit better then i'll kind of tweak those but you have to be careful around those as well because you don't want to overfit or you don't want to ruin a you know a decently good strategy so a lot of the time it's maybe tweaking something and then letting it run for three months in comparison and seeing how it kind of does and seeing okay well was it really worth that tweak you know if the tweak maybe gave me
00:46:01
Speaker
2% extra, at least for my capital size, it's not really that worth it. So I would just leave it as it is because I know this one's probably more robust without that tweak anyway. But if it was 15%, 20% something crazy extra, then I'd be like, okay, that's definitely worth adding that addition or adding that additional filter. So a lot of it is is very slow testing like that. I'd kind of compare it to Well, it's all based on the science method, right? So you have a hypothesis, you you see if something potentially is going to work, you guess, hey, yes, I think this is maybe going work. Then you go through the testing phase, then then you kind of see how the results came out. So a lot of times just waiting and also then testing things.
00:46:35
Speaker
It sounds to me like you need a lot of ah patience in in in your work because it like takes quite a while, especially on the higher timeframe things to like see when things are like working, how
Psychological Challenges and Community Engagement
00:46:44
Speaker
they're working. And also, um how do you deal with, first of all, that and then also like um when you're in the drawdown period, which may also last maybe like a couple months or whatever,
00:46:55
Speaker
but what what do you do like How do you deal with that psychologically, where it's just like it feels like nothing is really working or you don't really have an impact on what you can do and you just have to like sit it out and and like wait? And when when would you say, like okay, I'm going to do something different now, I'm going to cut this, i'm I'm stopping this, I'm changing this or whatever?
00:47:12
Speaker
Great question. um I'm going to say, like, I wish I had the answer for this. um The only thing I've slowly discovered over the years is having having some hobbies or having, like, something you can really distract yourself with. For me, the sad thing is, like, a lot of the time, I like being on the computer. I like...
00:47:28
Speaker
you know, like maybe writing some code or or watching something or whatever. But the downside of that is you're very close to, you know, remote accessing and into your VPS and seeing how performance is or yeah looking at your accounts or something. So I've tried to find as much as possible going out and doing something for X amount of hours. and Like I know this kind of sounds weird, I guess, to to someone that's maybe trading manually. But nowadays, the worst time for me and and the time I want to be away from my computer is around Market Open. I do not want to be around my computer at Market Open. I ideally want to be, i don't know, ah rock climbing or out and in a city or something, just doing anything else or in the gym, for example, just because market open for me is always the time where I've consistently made bad decisions. It's because I'll see some really big P&L that I'm up or something and I'll be like, oh maybe maybe I trim some of this or maybe I maybe i close this position or something. I don't know And those are the times where I just never really want to be around. Sometimes I have to be right. Like if I'm testing a strategy or things like that, I have to be around market open. But then luckily,
00:48:24
Speaker
ah because I'm testing something, I'm kind of siloed in to that one particular strategy or one particular idea. So it kind of helps me distract a bit. um But to be honest, patience, the best word I've ever heard around patience is patience is simply focusing on something else while the other thing is running. So that's the mythology I kind of have to it. I just think, okay, I need to focus on one particular thing so I don't remember that this exists and in the background and kind of work on that. So most of the time that's that's hobbies, finding something else. It's also why, like I get a question sometimes, like why do I run a business or why do I do other things apart from trading? And it's because I think people have the idea that when you're trading, you're just, you're locked in 24 seven, you're making money every single day. You're just, I don't know, there's a million opportunities.
00:49:05
Speaker
Most of the time you're you're sitting on your hands or something's not happening or the market's kind of boring. So a lot of the times you have spare, you can work on businesses, you can see if there's other ways you can provide value. um And also,
00:49:17
Speaker
It's why I kind of did Twitter because um one, Twitter is a great kind of log of all your thoughts. I can kind of go back, see what I thought, see what maybe changed. Two, meeting people is amazing through Twitter, like meeting obviously yourself, meeting other people that I've met. um Like that is just an amazing ability and an amazing network, especially because trading is a very isolating sport for the most part. You're not really near anyone unless you're in an office or something. You're not really having anyone else to compare to. And then three, creates opportunities. You can have business opportunities. You can meet someone that can maybe help with your trading or you can help, et cetera. And then and that kind of leads, I'll kind of answer it now. One question from Twitter, is someone asked like, would I ever run a fund?
00:49:55
Speaker
I did get an offer for like um running some of their capital in their particular fund. It wasn't crazy amount. It was like a couple of seven figures. Obviously, in a fund level, that isn't that much money. like Obviously, personally, it's a lot of money, but like fund level isn't that much. Personally, for me, and this has been the experience, especially because I run a company called AskEdgar, which is like around, we process SEC filings through AI, and we get like structured data from unstructured data. The side effects of running more proper businesses when you're working with hedge funds is regulatory. I hate regulatory. there There's a reason I never went into lore. I don't like dealing with lawyers. Going through contracts, having all these sort of things that handle is just mainly annoying. So I think maybe in the future, if ah I really wanted to, maybe I'd run a fund. But nowadays, more and more, I more think of it as like, I'd love to grow my capital to multi multi seven figures.
00:50:45
Speaker
yeah manage it myself and kind of deal it from there. um i don't think it would really be an aspiration anymore to run a fund. Obviously, it's a great like status signal. it's ah It's a cool thing to do. But also working with funds, people don't like to say this, but a lot of running a fund is just basically being a sales team in in some way. Like a lot of the time you're fundraising. Like it's not that much of the time that you're taking your trades. A lot of the time you are going to someone saying, hey, could you put some money? Could you a like, hey, look, I was trying your performance. It's not that what people kind of think that you're jumping on a desk.
00:51:14
Speaker
placing million dollars into trades every single day is is not it's not really that bad.
00:51:20
Speaker
I think... um I don't know, I guess my perception kind of changed, but I remember like a couple of years ago at least, or I guess there's also like a couple, like many different kinds of funds and stuff, but... um
00:51:31
Speaker
i just I used to think that most funds are just kind of like a scam or like not really like a... I mean, actually most funds kind of are a scam, like if you if you if you think about it, because they're like most like don't ah outperform the market and then they take like 5% fee or whatever from from the capital that you put in. So it's kind of like, ah yeah, as you said, like very ah sales sales business more than than actually like a trading trading business. But of course, there are like many exceptions and many different ways to to do it.
00:51:58
Speaker
Yeah, I would say, um like, primarily... One thing people, I think, misunderstand about around funds is every person compares funds to SPY or, like, benchmarks.
00:52:10
Speaker
it's that's ah It's a very simplistic way to look at it. Most of the time, funds, especially from more working with, like, ah some family funds and things like that, most of the time they don't have the same objectives as beating SPY. Obviously, that's what we all want to do. But like most of the time it's that, hey, i like this family or this X amount of people have hundreds of millions. They they don't want 10%, 20% drawdown. They want like
00:52:35
Speaker
Exactly. Yeah. They may want maybe a 5% drawdown or as little drawdown as possible to make 5% or 6% or a bit more. Additionally, another thing is once you have that amount of money, especially like you know ah you know if you're if you're in hundreds of millions or billions, it becomes actually like an allocation problem. like You can't just dump all that money into one thing. like you have to kind of split it around. So, you know, putting some money into a fund that's maybe uncorrelated to your main portfolio. Because another thing most people forget is most people think, hey, ah this person is just taking all the money and putting in a fund. No, they have their own SPI allocation. They have their own whatever allocation as well. They they want something that's kind of separate. So um that's one thing. But I would agree. Yes, a lot of funds, though, is sales. It's a lot of fundraising.
00:53:18
Speaker
It's a lot of going to conferences, it's a lot of meeting people, going to dinners, um those sort of things, which at least for me, I'm still kind of a bit of a nerd at heart. I'm i'm not a huge, you know, I'm going to take you to dinner, wine and dine you. can't really be asked. So um most of the time I'm i'm fine managing my own money. And then if I get to work with funds, whether that be in in my capabilities, whether it be in coding or whatever process I'm doing, that's great. I like that. But in like actually running the whole thing, I don't think
Risk Management and Trading Mistakes
00:53:44
Speaker
so.
00:53:44
Speaker
Mm-hmm. After we've like ah been been through your your journey a bit now, what are like some mistakes that you made along the way that's ah like that you can like share for that are good lessons for other people and maybe where you had a big drawdown or something or just like did something really stupid?
00:54:01
Speaker
Sure. and i So biggest mistakes I see people making normally is around over-optimizing. So it's going to be people adding way too many indicators for their particular strategies. And it's going to be that they're trying to over-complicate most strategies. So most of the time, at least for my strategies,
00:54:17
Speaker
I would say 8 out of 10 times, they're going be less than 2 to 3 parameters. So there's really not going too much tweaking. There's not really going too much things that are going to be happening. And the reason why behind this is just the biggest enemy in algorithmic trading is overfitting.
00:54:32
Speaker
So if you fit way too many parameters onto historical data, it's going to look amazing. The strategy is going to look great. You're go to be really rich. But when yet you put it live, it's going to completely fail. So um Most of the time for me, I just think about it in like degrees of freedom. It's been kind of the easiest way that I can kind of think about it.
00:54:47
Speaker
Almost like a knob. So if I'm tuning anything or I have the opportunity to tune it, I want to try and eliminate as many as possible as those. Obviously, you still need to keep a a couple. Otherwise, you're not going to really have any signal or you're not going to really have anything that's useful. But normally around two to three, I think is around the sweet spot. Another thing I see people doing is somewhat related is that they never do out of sample testing. A lot of people just, you know, let's say download some data from BTC or EF or whatever, and they get five years of data, and then they just test over all those five years of data. They optimize a ton of it. Maybe they employ some machine learning because they want to be fancy. But ultimately what's going result in there is the same thing without fitting. So...
00:55:25
Speaker
what you want to do is normally have two, maybe three years of data that going to optimize over. going to test your strategies, test parameters, etc. And then you want two years of data that you're going to test on how did that actually perform on something that's out of sample. And primarily what out of sample is trying to replicate is once you actually turn a strategy on live, what's actually happening is new data is being fed into it that it's never seen before. And it's reacting to that new data. So that's really all we're trying to replicate with out of sample testing is just how is it going to ah react to new data it's never tested on before and how is it going to perform. Most of the time, your out of sample isn't going to perform as well as in sample.
00:56:00
Speaker
But if it's within margin, you know, within a ah kind of close, then you'll normally be fine and you can kind of like take it on to the next steps. and There is more complicated ways of doing this, but I'm kind of just simplifying it down. um Another one I'll say is a lot of people try and target returns, but it's way easier to target risk. So most of the time you want to be volatility targeting or something like that, controlling your risk as much as possible because that is ultimately the only thing in trading we can kind like can control. You can't really target, hey, I'm going to go for 50 or 100% returns. Most of the time you try and do that, you're not really actually going to achieve it or going mess up in in some other way. So try and control your risk as much as possible and then let the returns follow from the actual process or from the actual kind of different signals. um
00:56:42
Speaker
One of one, I guess i'll I'll end on for this one, is that having a obsession around just a single strategy or a single market or a single instrument. Really in trading, at least how I think about it, is trading is a business. And in any business, you want multiple revenue streams. You want to you know have different allocations to different things. And ultimately, you don't want to be super concentrated on just one particular product or something like that. Now, obviously, when you're starting out in business, you're going to have to do that. And it's similar in trading. When you're starting out, you're going to need one market you really focus on, you you build on it. But once you're kind of over that step, you really want to diversify as quickly as possible. You want different instruments. You want different things that are trading.
00:57:17
Speaker
And one thing I do see a lot is if someone's trading, for example, BTC, they'll have a strategy just for BTC. In most of my experience, I've never seen that to be successful for the main reason being is is you're limited to just one, BTC's opportunities. And then two, most of the time you're just um but underperforming the underlying. Like the time you could have just gone on BTC and you would have done better than trying to do all this kind of fancy stuff. So most of the better to have a group ah or a basket of instruments or tokens or whatever you're trading. um And normally there you're going to have more opportunities. There's going to more diversification and there's go be new opportunities that you can actually take advantage of. Like an easy one that I can reference recently was that a lot of our portfolio wasn't doing very well in crypto because of like overall just crypto being kind of choppy and not doing very well. But then we had a couple of amazing trends in like ZEC and a couple of those. And whether it's to the long side or to short side, as long as you catch a trend on either one of them, you can perform very well. So like that was then able to compensate a lot so that our returns were still positive and like way above BTCs over that kind of time period. so Those sort of things you never would be possible if you're just only focusing on ETH or stuff like that. And then to be kind of tactical with it, most the what we're looking for or as I'm looking for is going to be top dollar volume.
00:58:27
Speaker
um You can also just do top normal volume, but it's not as efficient really to to do it. But you can do dollar volume. You can do like most of volatility. You can also do by market cap if you want as well. But normally like top 50, top 20 by filtering it as as that basket will give you a lot more opportunities than just trading one singular like asset.
00:58:45
Speaker
Mm-hmm. I think that's ah that's a good response and also cut into the advice section a bit as well already. um yeah i just want to end on very briefly, if you still have the time, maybe to the Twitter questions because there are some of them left. Maybe you can just like, ah there's like the three three ah questions from this one guy. um Maybe you can just like pick your your favorite or whatever or like the ones you think makes the most sense to answer.
00:59:13
Speaker
Sure. I think I could probably touch on the first two. So, okay, ah how to know if you actually have an edge? um There's going to a very annoying answer to this question. Primarily, it's going live market performance for for the most part. Like, right if you're actually making money over a long period of time, you have an edge. um The...
00:59:30
Speaker
sign of things of like beforehand, like before taking it live, most of the time it's going to be, can you explain why you have an edge or why this would actually make money? um A lot of things I see people trade, they'll, you know, I don't know, like I'll see someone say like, you know, there's this chart pattern, it looks good, it's going to go up.
00:59:46
Speaker
I'm like, okay, cool. But like what's like, what's the underlying reason? Like, why would this strategy ever make money in the future? It's not just going be because there's this chart pattern, is it's going to work. and There's going something that's happening underneath it that's, you know, allowing this kind of like ah signal to happen. So most of the time it's going to be premium. It's most of the time going to be that this asset is more risky than going long in bonds or just having some bonds, which means this is more likely going to trend up over time or over a long enough time period because there's got to be a premium to it. Otherwise, why would did anyone touch that particular asset? and Two, there could be that there's more volume coming in, there's more things that are like opening up for that particular newer market, for example. Or maybe it's a newer market that's completely underpriced because like we've kind of seen with Korean stocks, for example. Korean stocks for a long time wasn't really accessible to most other people unless you kind of went through different brokers and stuff like that. Now it's like on interactive brokers because this whole AI stuff, like people are trading it more. And now a lot of these very underpriced stocks compared to like US stocks are now getting priced up higher. That's an underlying reason I can understand why I'm going to make money on it. Now, obviously, that doesn't mean that I now have a signal or a way to go long because of that underlying assumption. I just know that this potentially could make money. Now I need to create a strategy around that so that I can actually take a signal, I can take an exit, all those sort of things. So first thing is just understanding why am I making money
Applying Established Strategies to New Markets
01:01:01
Speaker
on this. Another great example I always like to give is just an arbitrage. arbitrage, you know, even someone very, like even a 10 year old could understand that I'm buying it cheaper somewhere and I'm selling it more expensive somewhere. I have a reason why i'm making money. So those sort of things are key things. Two, for the second question, constructing a portfolio with different strategy types.
01:01:19
Speaker
A lot of people overcomplicate it. i think get a lot of questions of people saying like, how do you run 10 strategies? Isn't it like you're I think it' it sounds more daunting than it really is. The main part of it is that you have a couple of different strategy types. So ideally, trend following, mean reversion. And the rest of it is really just different asset classes, different instruments. And most of the time, you're going to have to create variations of a particular strategy for those different types.
01:01:42
Speaker
types And that's going to be mostly your portfolio. Additionally, as I kind of mentioned, you can leave some allocation for shorter term inefficiencies. So if an arbitrage comes up or something like that, then you can allocate to it and kind of like trade those. um And then also, you know, I would always leave a bit of allocation for like curiosity. So for example, like you can try out Polymarket. There's a lot of cool new nuances there. Maybe you can see if you can make a couple of like hundred dollars or whatever from that and then maybe scale it. So those sort of things is normally going to be the way you build out a portfolio. But as I've learned,
01:02:12
Speaker
Getting above 10 strategies, that's where it's going very difficult. Most of the time, most things you add into that portfolio are going to be correlated or they're going to be very different. So if you're starting out, the two to five phase of strategies is probably the most fun. Like you're just adding tons of new strategies, getting way better returns because of it. But once you get past that, that's where it really gets become more of a fist fight. You're just going to have to keep on going to find something that's actually kind of useful. And then i think, what was his last question?
01:02:37
Speaker
ah The fun question I already ah answered. how How to find edge in your markets? Yeah. um This is... ah kind of answers ah the the answer in the in the question. Most of the time when you're in a newer market, you can just take philosophies from older markets and they're going to work even better in the newer market. Really great example this was crypto in the early days. A lot of the time you could just take in well-known strategies like arbitrage and things like that that were previously done in equities and other markets. and apply them in crypto. And because it's such a newer market, going to make a ton of, you know, more money because there's way more inefficiencies there. Now, there is a whole slight conversation which I'll touch on is that is it really an inefficiency or are you just getting paid way more premium because it's a newer market? There's so much more unknowns. For example, your broker may just disappear. Like like in crypto, you know, you you may have money on this broker, it may just disappear. And then, hey, it doesn't really matter if you made 100 extra turns if the broker disappears. So you've got a lot more of those sort of things that you're to plan for.
01:03:34
Speaker
And it's more called like operational risk, where a lot of new markets is operational risk that you're getting paid a fat premium on, but you're going to have to somehow counterparty for it or find some ways to lower that operational risk as much as possible, whether that be you know standard things that once you have a you know decent run, a a couple percent return or whatever, you withdraw as much money as possible and you just keep the base quite small so that you know that, okay, if the broker does disappear or something happens, then i at least have all these returns kind stacked in in different ways. For but like necessarily finding something new, as I kind of mentioned,
01:04:07
Speaker
you're probably not gonna find like a new holy grail strategy that's never appeared before. Most of the time it's gonna be something that's already been found for all there's a mythology to it. And that's why I would say while reading research papers is is a bit boring, now nowadays you do have AI to like kind of summarize them.
01:04:23
Speaker
A great thing about reading them is understanding what had previously worked and in previous markets way earlier on so that when you do find a newer market, like, you know, for example, Korean markets now or Polymarket or things like that, you can understand all the basic ideas and principles and then just translate them and then kind of build from there. So a lot of it is just replicating someone else's idea and and kind of slightly doing a bit better or just applying to a slightly different market.
Conclusion and Episode Wrap-Up
01:04:48
Speaker
That makes sense, yeah. I hope ah everyone enjoyed the the ah viewer question segment. New new thing in the in this podcast so far. But I think it was a cool edition.
01:04:59
Speaker
And I hope everyone also enjoyed this conversation. I definitely did. Thank you very much for coming on. My absolute pleasure. It's been very nice to meet you. And then also we had a bit of a conversation with my laptop cut out. But overall, I definitely, everyone checked previous episodes of this kind of podcast. there's a lot of great like gems. And I personally stolen some ideas from people in all the previous podcast. So I definitely recommend kind of checking those out.
01:05:22
Speaker
You heard that there's alpha in the podcast. There is alpha. There's alpha. You need to listen to every single one. There is good stuff.