Transcript
Speaker: Welcome to a new episode of the Ensilico Terminal podcast. My guest today is QuantXBT, doing his first podcast ever, I guess, and I'm very glad to have him on. um Have you been around since like the BitMEX times? That's where where your name came from?
Speaker: Yeah, so I'm actually a crypto native. I've been in the industry for around 12 years since 2014. ah Mostly um just as a participant, I wasn't ah trading back in the day. i just remember buying Bitcoin in 2014, doing some stuff with it, then losing interest and coming back ah interest in trading around 2016, 2017.
Speaker: And, you know, at the time, the biggest platform was BitMEX and also Kraken was very big and they used the XPT ticker. So I just went with it. And ah i don't know if you if you've been around that long, but ah it used to be like a signal of, I don't know, status or something to use the XPT. So I just, yeah, I just figured I'll go with that. And I don't know, it sticks. so and And the BTC one was not available, so I just went with this one. too
Speaker: i have not been around like, I mean, I guess I have been in crypto since 2017, but I didn't really like trade or use BitMEX or whatever, but I know about the whole XBD thing. So I guess it did it does signal your your OG status a little bit.
Speaker: um what What did you do back then in in crypto? Like, did you how did you get into trading? What sort of stuff did you do? Yeah, so um back when I just got into, I've always been like an early adopter of things. looked for my emails for early signs of crypto interest and my Coinbase account was like 2012. off or something like that. And obviously i I didn't do anything relevant with it, but I just um remember being sort of interested and um I have a software engineering background. So
Speaker: um It was like a natural interest, I guess, when there there i I graduated from college and I was looking for something to do. ah the An obvious thing was mining was like ah a hot thing at the moment. i can remember it was like 2014. Obviously, i um I was not experienced enough to go deeply into the into it. It would have been a great time, but... um
Speaker: I just felt like yeah it wasn't profitable enough and I didn't have the resources to venture and just like make the capex and not think about recouping my money within the next six months. So I just didn't do much with that knowledge and that um Yeah, involvement and I just ended up buying Bitcoin around 300 bucks. I did some, I bought something in the Silk Road. um I remember the funds got con confiscated by the FBI. Sometimes I wanted to to to access them again. So I guess all of that the compounded to me losing interest over crypto and
Speaker: You know, ah late 2016, 2017, it was impossible not to look at crypto. Like everyone was in group chats. Everyone was like chilling the next coin and XRP was big. Ethereum was like eight bucks or something. So there was people who were like... um very early into that and I guess I just got hooked from there. i remember when I saw Bitcoin hitting like a thousand bucks or something, I remembered when I had bought it for 300. So my instinct was I will short that thing, you know, I i didn't have any yeah
Speaker: economic thesis, I was not an experienced trader. I just decided like it was expensive because it was more expensive than what I saw it. Right. And you always heard that it was like a bubble and it was volatile.
Speaker: And needless to say, it was a terrible decision for me to go short Bitcoin. And um I just got very salty about Bitcoin. So I just went into the altcoin space and try to chase the icos in 2017 like ah everyone else right i didn't have any edge or anything special in the market i was just like a normie i was following big accounts and stuff and yeah i round tripped a few times in 2017 the cycles were shorter um i think we're pretty used now to having like
Speaker: six eight months of done markets and then uh two or three months of bold markets but back in the day it was like we had a bold market and a bear market in a period of like three three months and that was like a lot of time right we expected things to go up and down very quickly so i i got to experience that a few times in 2017 and eventually when you know, the the industry started like um dying my business partner and I was doing the the software thing and I just decided that it was a natural fit to start doing, try to do automated stuff right with trading.
Speaker: ah We were not successful at first try. The things that we tried to do were like trying to automate technical indicators and, you know, stuff like that that were very that felt complex enough that no other people we thought were like doing but that in hindsight were like stupid things that we were just trying right seeing what's it um and it wasn't until 2018 like me 2018 that we uh found a very simple arbitrage uh strategy it wasn't
Speaker: It wasn't fancy. we' were just trying to arm the price between BitMEX and Binance and it worked. So from there, I think um the the time that we spent was like 70% in our day-to-day job and then 30% on this venture. And it eventually like took all of our time and it made...
Speaker: the ah the obvious decision for us was to stop doing what we were doing before and just go full time into this automated trading journey. Would you say it was an easy back then to do stuff like this because it wasn't as competitive or was it already kind of like difficult to get into since it took you ah quite a while to like find figure out what what actually works?
Speaker: I think it was relatively easy. The only reason that it took us quite a big amount of time to figure out what sticked is because we literally had no trading background whatsoever. So I think had we had the experience of being like in a trading firm or even like knowing simple stuff about markets would have made it much easier for us to identify how mispriced was everything back then.
Speaker: I think we didn't have that experience and that's what made it difficult for us but in hindsight those were like the the the golden days of crypto you could do like very sloppy things and spreads were very wide and the dislocations were very consistent and you know right now we don't have that luxury anymore but i can remember things were were easier back then definitely yeah so you mainly just focused on arbitrage back then Yeah, most most of the things that we were doing were some kind of arbitrage, which ah in hindsight, they were more like cross-exchange market making type of edges, you know, but at the time, we ah it all felt like an arbitrage, right? Because we were trading like very correlative
Speaker: assets, they were not necessarily the same instruments, but they were on a very high degree of correlation and um they tended to move in the same direction. So yeah, we we called it an arbitrage, maybe more like a start art, but if that that's the main thing that we were doing back then.
Speaker: Did you already need like a lot of infrastructure to do that properly?
Speaker: Yeah, so I wouldn't say it's as competitive as it's right now that you need like super fast infrar and stuff like that. But it was definitely competitive because we tried a few different approaches to the market and the the very first things that we implemented, even on the right path of finding repeatable edges, were not successful because of infrar. So there was definitely some um challenge over there.
Speaker: But as I said, that was the technically easy part for us because we came from a software engineering background. So I guess that the trading side was the hard part because we didn't know what we were doing, but we were good at doing things in automated ways. I think that's also one of the things that made it very interesting for us. We were very technical back then. So if you gave us like a more technical challenge that could make us money, that, that was like a no brainer, right?
Speaker: Did you just find out about more profitable strategies by just like fucking around and seeing what works? Or did you like, because you didn't have a trading background and all, did you like gain knowledge from from Twitter, from looking at what other people do from group chats? Did you read books? Or like, how how did you teach yourself what to look for?
Speaker: So we tried everything. um i mean, like trying to follow as much people as we could and read everything. and I guess one of the the risks that I constantly talk about in my Twitter account is that it's very difficult for a beginner to identify what is good advice between ah from bad advice. You know, when when you're listening to to people who...
Speaker: talk something about a topic that you are not familiar at all, it it's very easy to think they're saying valuable things, right? And that's definitely one of the traps that we fell into. We were like following people that were just self-interest and um we're not really giving away. i don't judge and judge that part, but they were not giving away anything valuable.
Speaker: um But at the same time, i think we were gaining at that point enough market experience to gain some first principles of knowledge from the the training that we were doing so that whenever we were looking for more knowledge, it was like we were seeking for specific parts that were missing in our thinking, in our processes, rather than looking for systems that were already built and we expected to be profitable, if that makes any sense. Yeah.
Speaker: I think you you talked before about doing HFT in JavaScript, and as far as I understand, that's like a little bit unusual, I guess, because most people use like different kind of programming languages. I think JavaScript is really HFT thing usually. like how all did you come up with with doing that? What's the story behind that?
Speaker: that's That's funny because the only reason that my account has a decent number of followers is because of that tweet. i i remember i I had like 200 followers in in, I don't know, it was like 2023 or something like that. And I made this tweet about how HFT looked and that you don't, ah the equity curve of an HFT and how you didn't have to have the strongest infra if you wanted to make it in crypto. And I showed that I...
Speaker: yeah that our infra was in JavaScript and um it obviously isn't is not anymore that simple as it used to be. But yeah it highlighted the I think that the interest came from people who actually knew um that the bottleneck in Crypto HFT is not necessarily the programming language because there is a lot of jitter involved in the platforms. And one one thing that I constantly talk about is that edge can come from very different sources. And if you focus only on the compute side, you're missing very low hanging fruit in terms of like getting deals with the exchanges or faster connections and stuff like that. So we we've made a lot of. um
Speaker: um Benchmarking over the last few years regarding like programming languages and stuff and JavaScript, ah believe it or not, for for most of those things, it's it's pretty fast, at least on on the platform that we were using it with Node.js.
Speaker: ah We we tested against like obviously Python and stuff like that, but we were not never Python guys. So ah stuff like Rust and Java, my my colleague was like a very Java strong strong guy and they were very, very on the edge. Like the the the compute side only matters when things get very crowded in terms of like your um the the amount of information that you need to process. But if you are good at pre-filtering the amount of things that you need to to process, ah you'll find that the compute side is very, very minimal and you can have like one, two millisecond improvements in the compute side on like P99 events, but then you'll have a four millisecond improvement if you have like a collocation agreement with an exchange or you know, stuff like stuff like that, or or you find a better way to design your network setup. So I think the the reason that it worked for us is because we understood that from the very beginning. Like we knew that we could make a tiny impact by optimizing things that were more complex on the code side, but we could do a ah huge lever if we, we could move a huge lever, which is like,
Speaker: focused on the things that actually moved the needle on latency. I'm curious, how has this like landscape evolved? Because this was all more or less pre-AI, and now we're in quite a different situation. And I feel like I see like less and less talk about programming languages themselves. Is is that still like a relevant thing, or has I kind of like abstracted that away a bit?
Speaker: I think it has evolved a little bit in the sense that The actual programming language is not necessarily an edge anymore. Right. So it's very simple for anyone to spin up like a bot or something in the fastest language that you can think like you can do it in rust or whatever, but the essence of the things that, that are important to imp improve upon are, are, I think the hard part of the system and the the thing with AI, and I'm not saying this is going to stay the same for.
Speaker: a long period of time this might obviously change but you you can basically you can direct it to do the things in the path of the the knowledge that you already have in a way so if you're not sure or you're not familiar of the things that are important to optimize for a trading operation, you're just gonna end up with a system that like has a fast programming language but has bottlenecks on all of the places that were more relevant to optimize.
Speaker: Yeah, that makes sense. Do you like, um i'm how does your operation look like nowadays? Like, do you still run similar strategies as you did back then? Or what sort of stuff do you do now?
Speaker: So whenever, when we started our operation, we started like a very simple cross exchange ARV that worked for us. We we were able to like double our money in like a month. And we obviously thought that we were, that we found like the chicken that laid egg golden the golden eggs, right?
Speaker: And we found out very quickly that that was not not the case and that edge decays relatively quickly. Right. And I think this is true for many of the ah people that I've talked that i that trade in a similar way to us, like in a prop way.
Speaker: It's not that you're constantly evolving the idea, but the implementation, the venues, the trading pairs, all of those things are things that are like a more evolving process. In in the end, it's a market, it's still like participants needing liquidity and trading irrationally sometimes or trading very efficiently other times. So I think the core principles remain even to to this day, but the execution and the places where we do things and the way that we do it is is what has evolved. Not necessarily that the idea has been, um yeah, it has been changing a lot. It's just a constant improvement.
Speaker: Mm-hmm.
Speaker: and And do you just still do similar things like in a way of, ah I guess, HFT market making cross-exchange arbitrage or... I think it it has evolved ah more. yeah we're We're parting away from HFT mostly because the the edge has compressed a lot over the years.
Speaker: um The spreads have been compressing a lot and there's much more sophisticated players. So whenever, you know, there there's there's markets where you think that you can make an edge and...
Speaker: They have like widespread and stuff like that. there' There's mostly a reason for that. And they're very like toxic markets. And you know, that the operation itself has changed. We're trying to move more into positional trading rather than pure HFT. But ah in general, I think the 80% of our volume and P&L still comes from related strategies. So similar types of statards and market making mostly.
Speaker: Is it still on like the big new exchanges or have you also ventured more into smaller venues? So we're in the top venues. Obviously, I will say that we're trading like the large names. We're in Bybeet, we're in Binance, OKEx, all of those exchanges.
Speaker: But most of the edge... these days comes from having access to different market participants that are not necessarily in those venues. So I will say that ah we're in all the big ones, but we're also very heavy on the smaller ones.
Speaker: Yeah. So is is that also like on the on the DEX side? Are you exploring like new new DEXs or mostly just like centralized exchanges? It's mostly centralized exchanges. We used to do DEX stuff back in, um I think it was like 2020, 2021. But it was mostly like MEV types of strategies. um But the edge compressed really quickly on on that one. And we just felt like we were not, we we didn't have enough edge to keep um that,
Speaker: operation going. So we just shut it down. And it's still for us, centralized exchanges, the the the majority of our training, whenever we go into a DEX, it's mostly because we have a good reason to do it, or there's a very specific trade that we want to do, but it's not part of our day or day operation.
Speaker: Do you feel like it might change as like, more and more volume moves away from centralized exchanges to DEXs?
Speaker: I don't think a reason, I don't think there's an opposition for that to happen, but I also know that many of the reason why some of the platforms have grown is not because they're centralized or decentralized, but mostly because of user experience and access. I remember back in the day, there were there used to be decentralized exchanges, um even like in 2017, most people don't remember, but there used to be like these EtherDelta and other implementations, but the the user experience was very bad back then. So yeah I think it's not necessarily like a competition between ah DEXs or centralized exchanges. It's mostly ah like who's giving the better product, who's giving the the best, you know, liquidity and the fee structure and everything. So I'm
Speaker: Maybe things can change, right? There's there's more talk ah about self-custody of funds and stuff like that. So maybe that moves the needle a little bit, but I think Hyperliquid is like a great example of this. it it I think it would have been a centralized exchange or decentralized exchange. It wouldn't have mattered really. It's mostly that the user experience was great. The product and offering was great and, you know, ah that really has nothing to do whether it's centralized or decentralized.
Speaker: Can you explain to me a bit more about what mid-frequency trading means or what it looks like? Because I have like a vague idea of HFT, I guess, but the MFT is a bit more vague to me. like but how how does it what What sort of like time frame does that span? What sort of trades does that contain?
Speaker: I think this is ah an interesting question because um depending on who you ask, you're going to get a different answer, right? So yeah I don't think there's like a clear cut definition on what constitutes like HST or MFT or, you know, other types of trading. um The way that I, that,
Speaker: I defined it personally trades that last um anywhere from a few ah minutes to a few to a couple hours. um I know that that might be too much for some other HFTs that I've talked to. They consider MFT to be like,
Speaker: from a minute to 10 minutes and then anything from there is like ah long long holding or something like that. But that's the way that I define it. If if a position we hold like two, three, four, five hours that that I still consider to be MFT. So that's a way that we think about that.
Speaker: And how how does that differ from the HRT things that you've done so far? Like, ah is is it still arbitrage, has just like last a bit longer? Or how how can i imagine that?
Speaker: I think the the The distinction mostly is that it's more heavy on the modeling side rather than on the infra side. HFT tends to be like very simple in principle, like your infra needs to be good and you need need to have like a good estimation of what fair price means and at any given instant, right? And if there's like a venue or a pair or something that differs a lot from others, then you have a an opportunity there.
Speaker: MFT is a little bit different because on things might look very similar across platforms or instruments, but you're again, you are like trading another type of of edge, right? youre you're You're either like looking to get a carry trade or like a statistical reversion or something like that.
Speaker: So I think in principle, Even HFT, MFT it's, it's all the same. It's like mispriced is, is an instrument relative to other instruments that you're looking at. So it's just a different, uh, horizon, I guess. Uh, but yeah, the, I guess the strategies go more into the carry side or more statistical heavy side rather than just like pure infra or.
Speaker: Yeah. ah sir I think you you once tweeted that you had ah seven years without a losing month. Is that still accurate?
Speaker: It's still accurate. um So we've we've had down months, but unrelated to business operations. So for whatever reason, there there was like ah a dent on the P&L or something, but not necessarily because of the trading operation itself.
Speaker: I think when i I normally speak with people and and that are not so familiar with with this type of of trading, they they get very impressed. But that the The thing to be very aware of here is that this is not a purely predictive type of trading. So it's not because I'm any genius or something. It's just the nature of the business, right? The HFT strategies that tend to do a lot of turnover and a lot of trades. And um if you...
Speaker: are not profitable within a day or two, needless to say, like months, um you're doing something wrong because it's very hard for for a strategy to do like a thousand trades in a single day and and you ending up on the losing side, right? It's possible, obviously it happens, but there's there's something more fundamentally flawed in your system if you are running those types of strategies.
Speaker: The obvious... downside to this type of trading is that it's very capacity constrained. So the this the when I said that when we started, we we doubled our money in a single month, of of course, it was just like a thousand bucks, right? We we tried to replicate that with much more capital. And as you can imagine, it was like not not nowhere near what we were doing at the moment, but it's not unheard of HFT firms doing like ah double digit returns, like even triple digit returns, but it's, it's high, sharp, low capacity, you know, um and that's just the nature of the business. I don't think um an HFT operation is,
Speaker: So much like a portfolio return, but it's mostly like how much rent can you charge for the service that you're giving and your cap to the amount of service that you can provide. Right. So it doesn't really matter if I can do it with a thousand bucks or a million, it doesn't scale linearly. So. yeah That's that's why our trading is like we don't have down months, but at the same time, we don't have the variance of having like a month with a triple digit return as other trader may have, even though they had like small losses during the months, if that made sense.
Speaker: Yeah. um So I guess like like you like you said, if you like lose money, you're already doing something wrong. But um I imagine you still have to to worry a lot or like just make make sure everything in the operation itself is robust because we are in in crypto and we've had many ah terrible traumatic events that we all all went through where exchanges go down and all that sort of stuff. How do you make sure to manage those?
Speaker: i guess the right answer is you just can't avoid everything. I would be lying to you if I had a a great answer for this. It has affected like huge participants in the market, which tells you that it's rather agnostic of size of experience. It's just like...
Speaker: how deep in the risk curve you are and how reactive you are to news and stuff like that. Some things that I um personally ah implemented when whenever we had like all these rug pulls and stuff like that is whatever whenever there's like rumors of a platform going insolvent or stuff like that, we just like, it's a policy. We just close everything, withdraw, wait a few days and we come back if things get like better, but we're we're not in a wait and see position. um And I think that helped has helped us avoid at least a few um platforms that have gone under, including FTX.
Speaker: But at the same time, you can't avoid everything. Like I remember we started this formal operation in 2018 and Binance was hacked for like 7,000 BTCs or something. that That's not something that people remember a lot, but i remember clearly. And there was nothing that we could have done. And we that the things could have perfectly gone against us and we not recovering our money. And This also happened to us recently with ah two smaller exchanges and another recent one, we were like of the 0.05% of affected customers in Bybit that were not like in the big Bybit hack. we
Speaker: they they Everyone was already like calling victory on their funds being recovered and we were just like still not getting them. I think we were like in the last 200 or 300 clients three hundred times that they They made whole and we were all already thinking like, you know, we got fucked this time. They obviously prioritized like retail. It's better for them to make an agreement with, I don't know, 200, 300 firms rather than going solvent and stuff like that. And we obviously ended ended up like um recovering our funds, but we had seven-figure exposure there. So there was obviously like, you you cannot...
Speaker: you cannot avoid everything. And the, the, what one of the reasons that we were like, um, that, that we were with a a decent part of our operation in, in, in Bybit is that we felt that it was one of the safest exchanges, like even over Binance at the time, there was a lot of rumors going on Binance. So, um, yeah we, we obviously distribute funds, uh, to, to, uh,
Speaker: mitigate these risks as well. that's That's one of the reasons that we also operate in multiple platforms rather than just a handful. But at the same time, it's going to happen and you're there, it's part of the risk that you're taking, right? And the market's paying you for that as well.
Speaker: Do you still only focus on crypto itself? on Have you also ventured more into RWAs as that has become even on centralized exchanges a bit more of a where the volume goes?
Speaker: We have explored it, but it's still on the very research side of things. We have not dealt like any um big resources into that. We've also explored prediction markets a little bit more recently, okay but To be honest, we're because we are crypto natives and all of our focus has been on on crypto, I think that's one of the main reasons why we understand this market and we've been able to make it work the ah the way that we have.
Speaker: But I would be lying to you if I said that I had an edge on every market that I touch, right? So i it'd be great if I could just like trade any new market that has 5 billion in volume on Binance, right? that That'd be great. But ah the the truth is that The type of trading that I do and that we do doesn't work that way necessarily. so we need to go over a research process first and whenever we feel there is some validation there. um Yeah, we we go heavier into that.
Speaker: um That being said, that doesn't mean that we don't necessarily know or um are able to run strategies on those assets. But at the same time, the thing is that it's not that a strategy has to die for us to remove it. It's always about...
Speaker: opportunity cost. So it is not only that we, we feel like we can make money, but we need to make more money than we are already doing. Right. So it's, that's the, the core decision that it's driving, whether we're trading a place or not. So, uh, fortunately for us, that has not been the case for, um, trying to look for other places to, to, to do our thing, but we have obviously explored them. Um, and yeah, so, so far we're not doing any of that.
Speaker: Mm-hmm. How does your operation look like in in general in terms of like, um since you've been doing this for quite a while? How does your day to day process and stuff look like? Because I i imagine most of it is just a monitoring already already running strategies and then research and how is that developed with AI? Like, but what what does your day kind of look like?
Speaker: Sure, so when we first started, this was more like, it was a more involved operation. We had to do a lot of manual things and um it it felt more like an operation over the years. And this has been obviously an effort that has paid off very well. We have invested a lot in automation even before AI.
Speaker: So at the like, I don't know, four years ago, we stopped doing a lot of the monitoring and the day to day things that we were doing and mostly focused on on the research side.
Speaker: Needless to say, today we with AI, we've been able to leverage those tools a lot to have like finally a dashboard and not only like terminal commands and stuff like that, that we were doing all of our operation in in terminals and SSH and all that all of those those things. um But yeah, I guess...
Speaker: in In the day-to-day, today, it's more, um ah we have established ah projects that we are working on. Those those are mostly like research projects.
Speaker: And whenever there's like um a string that we feel like we have to pull, we just go deep into that. um One of the efforts that we've had recently with our operation is to have a more um robust engine to have or strategies replayed and um much more data in our in in-house so that we don't have to be collecting a lot of data from external sources and also be more confident of the timestamps and everything that we're analyzing. So that's mostly what has been taking our time, basically building robust internal tooling and information access.
Speaker: Have you found the the external data to be more inconsistent? Definitely. So not only is it and because inconsistent, it's also like very incomplete. And um we tend to find edge in more niche and obscure markets and events. And more often than not, you'll you'll find that the specific event that you're trying to replay or or that you're trying to come up with a hypothesis hypothesis or something,
Speaker: It's basically a missing piece of the data that the vendor is providing to you because the platform went down when they were like collecting it. And it it has happened to us a couple of times. So we just obviously have used data vendors in the past a few times and we still do whenever we want to fill gaps on on research that we're doing.
Speaker: But most of our efforts right now and and a big part of our budget goes to like having in-house data. And that's one of the things that we've leveraged leveraged AI to do quite well.
Speaker: Do you ever do any manual trading? um Only personally, and I'm very bad at it. ah So i I honestly know my strengths and my weaknesses and definitely like manual trading is not one of the those things. i I tend to go more on the risk curve with my personal trading.
Speaker: only because i I am very systematic in the professional side. So it's very boring for me to like try something manually that I can automate and like have as part of the business with more capital and everything. But I do from time to time with the manual trading like, yeah, I try it, but I wouldn't say I have an edge there, not even close.
Speaker: I see. do with Do you do manual trading? I do, yeah. I do, I guess, primarily menel manual trading and probably both of our listeners, since we're like a manual trading tool. But do you do feel, like, do you have a view on, I've discussed this a bunch, or like, this comes up on Twitter her every so often, of like,
Speaker: ah the quantitative versus the discretionary trader ah if like people should even do manual trading at all or it's just like the market is efficient and do you don't even chance to succeed you have like any views on that so my Like, first of all, the the the main opinion that I hold there is that if you are doing whatever you're doing and you're making money money consistent consistently, just keep doing it, right? Whether this is technical analysis or you're looking at the moon or reading the cards or whatever, right? It really doesn't matter what you do. If if it has proven to be working for you, like, by all means, go ahead and continue doing it.
Speaker: um That being said, I think um there's... there's there's a distinction between what we call like quantitative trading and systematic trading and also like manual discretionary trading i think people confuse discretionary trading with like pure like sentiment based or emotion based type of trading but my experience is that there's a lot of very good discretionary traders that have some sort of quantitative information, that doesn't necessarily mean that they have a system that puts the trade like places the trades for them or that it's very automated or that it's like, you know, ah trades a lot or whatever. If you have an an Excel spreadsheet and you have some sort of information and something stands out, I mean, even if it's some simple
Speaker: data-driven trading like that one, I wouldn't necessarily call it like purely discretionary, right? And the same way that I wouldn't call a a quantitative system purely quantitative or purely systematic if you do some type of discretion in the capacity that it that you put into it or if you change for whatever reason, like the type of market that you're trading, like there's some discretion into that.
Speaker: um But, you know, people get... quite mad, especially if they like feel attacked at the type of trading or the the style that they're trying to do. But I feel that the data is um does not lie.
Speaker: Majority of people will lose money in the in trading this way. And the people that are trading, as I said, like that appear to be purely discretionary, really have some either...
Speaker: data that they're using or even the intuition that they have is in some way data driven, even if they don't feel that way, right? You will see very good traders making calls that seem to be supported by a chart, but The reality is that the trade came from years of experience of seeing some certain market market condition and the chart is just like the the trigger that they use to place a trade.
Speaker: But really the trade didn't came from the chart, right? So if if the trades were... very easily like identifiable from chart patterns or whatever, you know you you you're 100% sure that the quants would have exploited that.
Speaker: And I guess that's the the the the main focus of the of of the conversation when whenever it arises online, which I know it happens a lot and it happened recently again.
Speaker: So I tried to keep myself like on a neutral tro position. Obviously I am on the quantitative side and I know that that thing works. I've tried the discretionary thing and hasn't worked for me.
Speaker: and That doesn't mean that it won't work for someone else, but yeah I don't know. That's the opinion that I hold. i don't know. What's your view on this on you being on the other side, I guess. I mean, it's almost a bit of a philosophical debate kind of because it's like, I guess the central question about it is like, can you quantize, is that the correct word? i don't know. Like, can can you categorize everything in a way where you can just like, where you could just automate it away?
Speaker: Because it it could kind of appear like that. But I think ah what what you mentioned about intuition, like there's just things in the market that like you internalize if you spend a lot of time interacting with the market. that are very difficult to to put it into a rules-based, easily writable down system where you can just say do it and automate it. There's like more... i don't know, because the human brain is also kind of a computer that the doesn't always work in ways that we completely understand, but it can definitely be useful.
Speaker: even if it's a waste that we don't understand. But also, as you said, most people lose money. So you can't really just be like, hey, I'm i'm using my intuition and just doing whatever and and not having any basis in reality at all and and just fuck around and make money with that. you I think you yeah you need to have like a balance or just, as you said, like find out what works for you. And if you're making money, then that's that's pretty much good enough. That's like telling you if you're doing something right or not, regardless of like what basis it's happening on.
Speaker: And I mean, for what it's worth, like there has been quant funds that have blown up as well. So that doesn't necessarily mean that just because you're doing quant, you're not going to lose money or you're going to be profitable all the time. That's far from the truth, right? It's, it's also the type of trading that you're doing that the risks that you're taking, even like there's these.
Speaker: If you're trying to have a system that it's 100% automated, like there's no intervention whatsoever for anything, you're very likely to lose money or at least stop making money. So you cannot say that a quant or a systematic type of trading, like algorithmic trading,
Speaker: does not require any sort of discretion. This might change, right? Or i might be too dumb to have a system that evolves itself all the time. But I found that even on the quant side, you still have to make some discretionary calls. There are some things that you've seen in the data that you are very easily comfortable trading. And there are some things that you should be able to make the judgmental of saying like, you know, this is definitely not something that the system has been trained for. And I'm not doing this out of like emotion. I'm, I'm taking the rational choice of, you know, turning off the system or i don't know, like even doubling down, this is like the best time to trade it. The system doesn't capture it in full. And I just need to go more deep into this opportunity. I guess those are tie, those are also like things that will come from experience and intuition. And I wouldn't call that like pure systematic. That's also some discretion there.
Speaker: How does your your system fare in different market conditions? Like, does it change a lot with with different regimes if there's more retail participation or where like if it's a bull market or a bear market, more activity, less activity?
Speaker: For the type of trading that we do, we benefit more from more market participation and more like general interest. Edge is one of those things that it's not equally distributed, but like the the fact that we make money every day doesn't mean that we make the same amount of money every day.
Speaker: That's ah like a different perspective. So there's obviously conditions that will benefit systems more like than others. For us, a high volatility, high retail participation is what we're after.
Speaker: um You know, volatility to create these, you know, Inherit um forced trading and that's what we're morally after. Like if we can find the type of traders and I've talked about this a lot. If you can find traders that that don't want to trade but need to trade, those are great people to trade against, right? So um yeah, and and I think that's more present with volatility crypto. We have these cycles that are very...
Speaker: um ah consistent that are basically like short periods of extreme volatility and then people get like scared and they lower leverage and then we get into like long long periods of low volatility and people start getting more confident using leverage and more confident using leverage and then some triggered like moves a lot of liquidations and we go in this cycle all over again. So it's one of those, um, persistent things just because it's the human nature, right? They, they we're after profits and you can only take what the market's giving you. You cannot force with a, like even the perfect model won't make more, won't make money in a market that it's not giving any. Right. So I think people try to force trading when there's none, uh, and we'll tend to,
Speaker: put very high leverage onto that. And that also creates this repeating cycle that, you know, if you're doing something that it's very profitable during volatile times, you'll you'll be able to do good on those times.
Speaker: Mm-hmm.
Speaker: What would you recommend ah for someone that wants to like start out now or like that is trying to go a similar route that that you have done and also maybe doesn't really have a trading quant style background?
Speaker: what What should they do or where where should they start?
Speaker: This is an interesting question because I'm not sure I'm the best person to give advice on on those types of things because i my my journey had no no no clear path. But um i would say that I guess try to come up with principles that will remain true regardless of the market that you're trading and try to avoid noise. And even if there's a a person online, including myself, that has like a decent following or something, don't follow their advice or don't
Speaker: Like, be curious about their their ideas, but don't make make them your own unless you have, like, tested them. i will say that for someone that has very little or no experience in in trading, the best... devest thing to do would be to start trading with a small account. I mostly discourage like test accounts or paper accounts because you don't have any like skin in the game and the the market dynamics are also very different because people are just like testing stuff. They're not putting their money what they what where they like their their ideas or their edge ease and the big participants are not there.
Speaker: So I will say like try a big market. put some small amount of capital and try to learn from what you're doing, what's not working. um Try to differentiate of the trades that you take and ah which ones are are good and how do they look at the times that you took the trade and the ones that are bad equally.
Speaker: um i know it's an oversimplification of ah of a very large process, but I wouldn't have much more ah advice than that, to be honest. ah I wouldn't recommend like buying a trading course or stuff like that. Just like try to identify things that are mispriced in the sense that you have some sort of sts of what the price of something is, right? If you see that the price of an asset is the same in 10 venues and it's not the same in one venue, you can either make the hypothesis that the one venue is like mispriced or you can make the hypothesis that the other 10 venues are mispriced. But there's something there that you need to explore, like the different things, um the outliers.
Speaker: I think that's where I would go and Yeah, you cannot have enough data for this. So get data, try to analyze what the market's doing, not necessarily only what you're doing. The useful thing about analyzing the things that you do is that you have the intention you know the intention of why you did it.
Speaker: But the market also gives a lot of information. if you You don't have to have like the intention of everyone. You just have to have the performance of the trades that they made. So yeah, I would start there. Just like try something, try to think for yourself, get ideas from other people. And when you get those ideas, don't make them your own, validate them first.
Speaker: Yeah, I think that's that's very sound advice. Do you have, um i' I'm curious to, what is your opinion on a crypto as a whole? I don't know if you have too too much of an opinion about the space that's going on, whatever, since you're like mostly on the on the systematic side, but as someone who has been in the space for quite a while, if you like have a Coinbase account from 2012 or whatever, what what do you think about where crypto is today and then where it's going in the future?
Speaker: I'm honestly like the only opinion that I hold about crypto in general is that I'm a very big Bitcoin bull. I think that would be reflected with the XPT part of my handle, right?
Speaker: um The thing is that I i thought... hard about this and I've i read books, I've heard, I've experienced for the first time and I think my the the way that I joined the industry was backwards to the rest of the people. Most people get like interested in trading and then they try to find a justification for um um the the value of the assets, right? For me, it was the opposite side. I first used Bitcoin, I bought it for a reason and I used it and um that
Speaker: That felt like a very interesting use case, right? And when you start seeing, like, I come from from a place where the currency got very um the ah depressed over the last few years compared to other currencies. And um there's a lot of capital controls and stuff like that. So I don't think it's very... um very clear for people who have never lived under those conditions to see the value in these types of things, because there's yeah like, everything is very stable and they just don't see the value very clearly more than speculation. But when you come from a ah place where there's actual value in like being able to move your funds around and not having your, your, your,
Speaker: I don't know, assets seized and the even like you talk about 50% devaluations in Bitcoin, you'll see a lot of currencies that have greater devaluations of that. So um in general, I'm um i'm a big bull of Bitcoin. I do have some diversification in in Ethereum just because I understand that there is the risk for m not risk, but the the ah the small chance of it overtaking like the market, even even though I know that people are not so ETH bulls recently as they used to be like two, three years ago.
Speaker: But that's the only opinion that I hold any other shit going, I think, ah unless it has like positive cash flows, I will say like BNB or hype or those tokens that have some sort of value accrue. I think they they will all tend to zero over time and they have.
Speaker: I've been enough time and you don't have to be OG to like... Steve is right. it's It's on the data, but there's also some sort of like survivorship bias happening all the time because you don't see the thing that is no longer here for a reason. Right. So you just like think that the things are that are here are are have been here all the time, but the the the market used to look very different back in 2016, 2017. Yeah.
Speaker: Yeah, i I don't have an opinion. I think um it's very easy to rationalize like dollar cost averaging into the wrong stuff and stuff like that and try to apply like investing fundamentals of the long term over the wrong assets. So the only thing that I that that i personally have a strong opinion on is on Bitcoin and if I don't have a strong opinion, but it's like my hedge, right?
Speaker: Mm-hmm. I think what you said is very interesting. If you look at the coin market cap top 10 or CoinGecko from every year, does look very different over the years. Even if you think, I don't know what's in there, like Solana, Airbnb hype and all of that.
Speaker: A couple years ago it was Cardano and then XRP. I guess XRP is still high, but like it it does change constantly and you can't really like rely on on the altcoins still being there. I guess nowadays it's a bit different because they have like actual businesses, cash flows and stuff attached to them but back in the day it was mostly just random stuff overvalued for no real reason.
Speaker: And I think ah to be honest that the boomers had a better view on these, like it like there there might be more, they may have missed the vote on Bitcoin, but they were kind of right on everything else. Like if the asset has no intrinsic value, if it has no like way to accrue any value or whatever it's just worthless and people were marrying their backs and saying the like the opposite of that and they've been proving right about that and and i will i i tend to be more and more into that camp as the time passes and we see yeah more and more tokens uh just exiting like even the huge protocols and layer twos and stuff they're they're
Speaker: dying slowly and I think they they're going to zero in my opinion most of them ah not to be salty for with anyone doing like meaningful stuff but it's a reality of it like um if if the token or as it is doesn't have a a good way doesn't have the network effect and doesn't have a good way to recoup some of the value that the protocols networks have generated, um it's very hard for it to survive. And there's also a lot of diustion of a dilution of capital. So, yeah, to to be honest, there might be a few and there will be a few that um will make like incredible returns.
Speaker: But trying to identify those in a market of like thousands of currencies, I think it's just like an almost impossible game, at least for me, there's people who will, who will do it spectacularly and yeah I'm not one of those. So my opinion is just like, I hold Bitcoin, I hold Ethereum as ah as a hedge and other than that, I just trade and hold other assets.
Speaker: Yeah. I think that's ah that's a reasonable view. Bullish Bitcoin, maybe a bit of Eiffel and short all the dogshit coins. Is there anything else that you want to to mention at the end?
Speaker: No, I mean, ah you said short all the altcoins and the only thing that I would say there is that you should be careful about shorting regardless of your opinion. Just be bearish. Yeah, and that's that's one thing, right? you You can, as we've seen time and time again, you can be very right on certain call and still blow up.
Speaker: That doesn't necessarily mean that the thesis that you were like... bearish on or bullish on doesn't reflect. It's just that timing matters when you're talking like trading, especially when we're talking about shorting. There is a cost of shorting. Even if you have like unlimited margin, there's also this opportunity cost of shorting. So of doing something, like ah any trade will take...
Speaker: capital and time from and effort from other things that you could be doing. So you could be right of writing shorting, like, i don't know, XYZ coin. And yeah, I'd go to zero in 10 years, but could have you done something any better with your time and effort and capital during those 10 years? Probably the answer is yes. So um and you can also get liquidated in the meantime and still the token went to zero in 10 years. Right. So yeah I don't advise like betting a a get against the industry, but at the same time, i think the altcoins, altcoins face and
Speaker: Many of the crypto assets in general, it's just they're all in a trend towards zero and it's just a matter of which ones get there first. That's a view I have.
Speaker: think that's a those are good good closing words. So thank you very much for coming on.
Speaker: appreciate the invite. Thank you so much for having me.
