- Solve real problems with our hands-on interface
- Progress from basic puts and calls to advanced strategies
Interactive Options Course
Finance Courses and Lessons
Topics
Language
Finance Courses and Lessons
Topics
Language
Posted July 28, 2026 at 10:53 am
What does it really take to become a successful trader? IBKR interns put veteran systematic trader Adrian Reid to the test with questions about AI, risk management, backtesting, and building a lasting edge, revealing the lessons every aspiring trader should learn.
The following is a summary of a live audio recording and may contain errors in spelling or grammar. Although IBKR has edited for clarity no material changes have been made.
Hello, and welcome to IBKR Podcasts. My name is Chris McLaughlin. I am a computer science student at Purdue University, and I’m currently interning here at IBKR for the Enterprise Architecture Department. I’m super excited to be hosting today’s special podcast episode featuring IBKR summer interns.
Today, we’re very excited to welcome Adrian Reid to the podcast. As background, Adrian is the founder of Enlightened Stock Trading. He is a trader, educator, and author who specializes in systematic trading and evidence-based investing. He has spent years developing and teaching systematic approaches to trading with a focus on building repeatable trading systems, portfolio construction, risk management, and removing emotional decision-making from the investment process. Through his educational programs, books, and coaching, Adrian has helped thousands of investors and traders improve their approach by relying on data, process, and disciplined execution rather than prediction or intuition. Today, we’ll be discussing Adrian’s journey into systematic trading, how traders can build and test robust systems, the psychology behind successful execution, and how emerging technologies like AI are shaping the future of trading. Adrian, welcome to the podcast. How are you?
I’m great. Thanks so much for having me. I’m looking forward to the panel interview. I’ve not done one of these for a long time and this is gonna be fun, so
Of course. Yeah, we’re super excited to have you. So just to kick things off, I’m gonna go ahead and hand it off to Jay to ask the first question.Â
Hi, Adrian. As Chris said, we really appreciate and have been really looking forward to having you on the podcast. Just a little bit about me. I’m a Rutgers student, rising senior studying computer science. I’m a software dev intern at IB. And what I wanted to ask was, Adrian, you’ve been a mentor and educator to thousands upon thousands of traders. I wanted to ask on your sort of come up, did you have any sort of figure like that?
Yeah, it’s a good question. I think mentoring is really important. It makes such a difference, pardon me, it makes such a difference to how fast you can master something. And you can probably see here, you know, there’s a lot of trading books that I’ve kind of used and benefited from over the years. When I was starting out, mentoring was not really that available. So it was really a case of kinda from a distance observing people who I respected and could learn from, and trying to, you know, get as close as I could to them without necessarily being in their physical presence. Particularly being in Australia, a lot of the major kinda trading figures at the time, and this is, you know, more than twenty-five years ago now, a lot of the major trading figures were in the US, and so I was a long way from them.
But probably one that stands out that I really was influenced by a lot early in my journey is Dr. Van Tharp. Unfortunately, Van’s no longer with us, but I benefit a lot from his books and we had a couple of courses that I did online and live over the years that really shaped a lot about my trading journey, especially in the areas of risk management, actually long-term survival. So I kinda credit Van, I guess, with my survival in the markets, at least in the early stages, and that’s, you know, I think that’s really important, ’cause one of the most important things in trading is you’ve gotta survive. You know, if you blow up your account, you’re done, and you have to start again, and that just wastes years and years of compounding time.
So I think, you know, in terms of mentoring in the trading space, at least, that would be one. There’s several other authors who are major influences in my life, my trading life, but Van would be the biggest one because I think he’s who was responsible for my survival long term.
Thank you, Adrian. And I had a follow-up. How would you sort of illustrate the sort of impact– I know you didn’t necessarily have a direct mentor, but, you know, how would you accredit those sorts of figures, those sorts of resources to your success?
When I was starting, no… I didn’t know any traders, so that’s the first point. So I had to learn from someone, and I didn’t even know any aspiring traders at the beginning, and so I just started reading. And I went through this divergence convergence kind of process for my learning, and the divergence started with just getting whatever books were available locally on trading and investing and trying to kind of learn broadly to figure out what it was all about and what I really needed to learn. And eventually, I discovered systematic trading, and I discovered systematic trading through the Market Wizards series from Jack Schwager, and that was a pivotal point because as soon as I discovered systematic trading, I realized that was my thing. That was gonna be my thing because I was an…
I’m an analytical person. I studied engineering. I’m quite quantitative. I like numbers and analysis and math and quantitative trading really just leapt off the page at me. And so after that, I basically spent time on Amazon finding every book I could on trading, technical trading, quantif- quantified trading, technical analysis, all of those things, and basically was ordering boxes and boxes of books to help me on my journey. Again, a lot of this stuff wasn’t available in Australia at the time, so I was having, you know, crates of books delivered from Amazon, reading them, throwing a whole bunch of them away ’cause they were garbage, and really just going through the ones that had value and, you know, dog-earing the pages, writing notes in the margins and so on.
So I think that divergent process to identify my niche in the trading world and then converging on that and finding the authors and the teachers in that space, and going deep on learning with their content is what’s really responsible for helping me get there or get through the initial stages to some level of consistency and some level of profitability.
Does that kind of answer the question?
Oh, for sure. And I just think like, you know, coming from a starting point where you didn’t really have a lot of people, a lot of resources, and to accomplish what you’ve accomplished is so impressive. I had one more follow-up to that. So you mentioned, you know, you’re really interested in numbers, quantitative data. Did you always sort of have that interest, or was there like a moment or a stretch of time where you realized, like, this is something that I’m really passionate and interested about?
Look, the analytical side of life is probably always where I’ve been most comfortable. I mean, in school, wouldn’t say I hated, but pretty much I hated English and I hated any of the social subjects. I was no good at languages. I was no good at music and art, like any of the creative stuff. It just wasn’t my jam.
But in math class, in chemistry, in physics, I felt comfortable and at home. So it’s– I think it’s some– an area of life that has just always been more natural. I won’t say it comes completely naturally. I mean, obviously I had to learn and whatever, but they’re the sorts of subjects and topics that I like. So going from school to university, the choice of something analytical was again, obvious. Engineering was the choice at the time for people who were kind of good at math and liked physical things and was fascinated by how the world worked. And then engineering led to problem-solving, and problem-solving led to kind of the markets in some way.
And trading is really a problem-solving kind of journey, figuring out what works, what doesn’t work, why, and putting rules around it, and creating trading programs that work. So I don’t know. In hindsight, it looks kinda natural, but yeah, I guess I like the analytical side of life.
Yeah, I can totally relate. I’m much more of a math guy too. So I’m gonna pass it back to Chris.
Thanks, Jay. I kind of actually have a question that builds off of that a little bit. You mentioned really appreciating the analytical side of things, and I am sure there’s a lot of other people out there, maybe even some people listening that are kind of in that same boat. But systematic trading can seem like a really intimidating world to break into. From your experience, what would you say are some of like the non-negotiable skills and foundations needed before a new trader can create their first system?
Adrian Reid
Yeah, it’s a good question. It– and it does look intimidating from the outside, and I think it looks intimidating because a lot of people overcomplicate it. And, you know, when you think about quanti-quantitative trading, often you kind of get these images of hedge funds with PhDs in math and complicated coding and analysis that you look at on the page, and you actually don’t understand what it means. So I wanna be upfront and say that’s only one aspect of systematic or algorithmic trading. Like, it doesn’t have to be that complicated. I won’t– I’m not a math genius. I’m not a programming genius. You know, and you don’t need to be. So what are the non-negotiated– non-negotiable skills? I think the first one is fascination with the markets, and that’s because it’s not easy.
It’s not easy conceptually. Like, there’s some concepts you’ve got to grasp and some skills you’ve got to develop, and you have to learn to do some things which are not taught in schools. You know, they’re not natural. And it’s also not easy emotionally because you’re gonna be challenged when your account is going up and down, when you have a sudden loss that you don’t expect, when you’re sitting there f- in fear worrying about pressing the button to enter or exit a trade if you’re doing the right thing.
Your emotions are going wild. And to cope with those things, you need a reason, and just making money is not really a good enough reason. You’ve got to love the game, and I think most people who really win in whatever discipline in life love the game they’re playing. And I think you’ve just gotta love the markets. You’ve gotta be interested in, like, why did that happen and what is going on and who’s driving that and how does that work? And if you have that fascination, then it helps give you the fuel to get through some of those challenges, both the emotional and the technical. So fascination with the markets is one.
The second one, also a non-technical skill, is curiosity. And that’s because most things about the markets are not intuitive. You know, we come in, we study whatever we study, and we have these ideas about how the world works. And when– and that comes from our training in school, it comes from our family interactions, it comes from our training at university or whatever. And then we come to the markets, and all of a sudden there’s this emotional, irrational beast called the markets, and it does things that you don’t expect. And it reacts to good news badly, and it reacts to bad news well, and it causes things that don’t happen elsewhere in life. And so we’ve got to get curious about that and dig into why and dig into what is actually happening.
Because if we bring all of our preconceptions from life into the markets, we’re gonna be dead wrong, and we’re gonna lose money hand over fist. So we’ve got to be open to the fact that we might be wrong. We’ve got to be open to the fact that the way we think money and wor- the world and, you know, economics works might be wrong. And an announcement that looks at face value to be hugely positive might push the share price down because it’s not as positive as what people were expecting or because people were expecting something different or because people didn’t like what they heard. And so it’s not always rational, and it’s not always intuitively obvious.
In fact, frequently, it’s not intuitively obvious. So if we’re– if we can get curious and be open to learning about how the markets actually move rather than how we think they should be moving, then we can really start to make some inroads. So the two things are fascination with the markets and curiosity, completely not technical things.
Then we get to the technical stuff. I mean, they– you only really need basic math, and you only really need basic programming, particularly now in the world of AI. I mean, I can dream up an idea by looking at some charts, throw it to my AI kind of environment and have it fully define the idea, fully code the idea, run it through my fifteen-step process to test and evaluate and fine-tune and improve it, and then it spits out a written report to me in plain English. You know, I don’t even need the kind of real technical stuff anymore, as long as I understand what’s going on under the hood. So those technical barriers, I think they’re reducing a lot, and it comes down to some of the things that I talked about earlier. Plus, probably a little bit of creativity. You need the math, yeah, you need the programming, but increasingly, they’re not really barriers anymore. It’s more about those other things that keep you in the game.
Yeah, I think that’s really interesting. I think the emotional side of trading is something that not a lot of people consider when they’re looking at it from the outside. But I know I’m personally starting my trading journey here pretty early, and it’s something that I’m learning to deal with pretty quickly, is kind of coping with things not going as you expect them to, and maybe the markets start– just aren’t as easy as like numbers on a page. So I think that information and that advice is very valuable. Moving forward, I’ll go ahead and hand the next question on to Thu.
Thanks, Chris. Just want to say it’s been really great to hear about your experiences and insights. Adrian, my name is Tu. I’m a rising senior at Harvard studying econ and linguistics, and I’m an intern on the marketing project management team this summer. I just kind of want to expand upon your point of curiosity and the emotional aspect of trading and talk a little bit about your evidence-based investing approach as well as kind of what that means in practice, specifically of like how a trader gather, tests, and actually trusts evidence before committing real money, especially in times of like high market or economic volatility.
Yeah. Okay. Good one. Good question. So evidence-based trading is– Basically what I like to do is never place a trade that I haven’t tested and validated with some sort of data. And of course, you don’t know in the moment how that trade’s gonna pan out, but the conditions in which you’re placing a trade have probably happened many, many times over the last several decades.
And so what I wanna do is look for data that I can use to model a strategy. And in its simplest form, and frankly, in the markets, often just keeping it simple is the best thing to do. So in its simplest form, that looks like getting the historical open high, low close volume prices for any instrument that you’re trading, and then applying your trading rules to those. So think pure technical analysis, you know, a price breaks above a two hundred day high, it’s probably going up. This is all hypothesis by the way. So you imagine you look at a chart and you say, “Okay, the chart’s going up. If I bought here, and if I sold when that happened, that should make money because look at this one great example on the chart.”
One great example means nothing because the markets are so noisy. But if you look at that and say, “All right, well, I got a hypothesis that if I did that over and over again on many stocks over many years, I would make money,” then you take those rules, you apply them to that data, and you test it historically, and you see how it would’ve performed. And so, you know, this is the process of back testing, and that’s, I think one of the most important skills for a quantified– quant- quantitative trader is to really be able to take your hypotheses and test them on data that exists in the markets. And that data might be price data, interest rate data, it could be economic data, it could even be social cues or social kind of chatter about, you know, how frequently a stock is mentioned or something like that.
But the easiest data and the best data to start with is pure price data because price allows you to develop strategies that work. Not all price-driven strategies work, but pure price strategies can work. And so that’s simple data. It’s easy to get. It’s easy to model because there’s not too many variables that are uncontrollable. You know, if you’re just creating a strategy on the price history of a stock, you get that price history every single day. It comes in at a certain time. It’s available when you need to make the decision. So basically you take that data, you overlay your rules on it, you see how it performed, and then you te- validate or invalidate your hypothesis.
There’s of course a whole process around this. So when you’re back testing a system, you know, we’re creating the hypothesis rules. We’re running some tests to validate that there is actually an edge there. Then we refine those rules. We probably optimize, vary the parameters to make sure that we’ve got some sort of stability in the rules. You know, if the two hundred-day moving average adds value, and the hundred and fifty-day moving average adds value, and the three hundred-day moving average adds value, then the moving average probably adds value. But if the two hundred-day moving average adds value and the hundred and ninety-day moving average doesn’t, and the two hundred and ten-day moving average doesn’t, then it’s probably just a fluke in the data.
So a big part of our job is figuring out what is real and actually adds value and what is just a fluke in the data so that we don’t fall victim to data mining and overfitting. So we need a process to go through that, but essentially, we’re taking a hypothesis, back testing it on real data that would have been available at the time we were making the trade, and seeing if we can design a system that worked in the past, then validating it on unseen data, and then going live and testing it with real money, and then scaling it up.
Does that explain sort of clearly enough what you’re–
Yeah, for sure. And I think it’s interesting ’cause given how uncertainty or un- yeah, c- it can seem, trading can seem as a young investor, it’s reassuring to hear that there are some concrete methods and good practices to follow in order to take some of that risk and uncertainty away.
Yeah, the uncertainty is actually really important here because one of the things that holds people back is the belief that there should be an answer and one right way to do it. So what’s the best rule? What’s the best moving average? What– You know, how wide should my stop loss be? Where sh- what should my risk-return ratio be?
Like, people ask questions like there’s one answer, but there’s not one answer because the data is noisy. But that’s okay. We just need to think beyond the trade that’s in front of us. We need to think about hundreds of or thousands of trades in the future and make sure that our method over many trades over– with a lot of noise and a lot of variability will give us an edge. And what’s interesting is often the edge is not what people expect. You know, when we come out of school, we expect that being right gives us the edge, right? I mean, no one wants to get forty percent in an exam, okay? But you can make tons of money being right forty percent of the time in the markets. And this is a massive mindset shift that most people can’t cope with because if you’re right forty percent of the time, that means you’re wrong on sixty percent of your trades. But if when you’re right, you win really big, and when you’re wrong, you lose really small, the math works out that you make a ton of money if you can place enough trades. So we don’t actually need to be right. We just need to have an edge, and having an edge means we need to often accept being wrong, but make sure that we don’t lose much when we’re wrong and we win a lot when we’re right. So that uncertainty is a real mindset shift, and the variability of the data is a real mindset shift, and we’ve gotta elevate our thinking from the trade that’s in front of us to the strategy that’s driving the trades, and then elevate from the strategy that’s driving the trades one level further to the portfolio of strategies.
Because even a strategy applied over and over again has a degree of uncertainty, and if we apply many strategies that are diversified, non-correlated to each other, then we start to do a lot better, and we start to get more consistent returns. We’re sort of abstracting away from the uncertainty or the variability, and the portfolio gets smoother and smoother and smoother.
Yeah, that’s great. I think your explanation on the nuance of, you know, having an edge and kind of taking a look at the bigger picture is really, really insightful. And then so with that, I’ll turn it over to Drew for the next question.
Thank you. Sort of building on that, I wanted to ask how often should you be changing your parameters and/or copying the automation rules of potentially more successful portfolios? And as part of that, how do you avoid going into, like, the kinds of subjectivity and second-guessing that automated trading is supposed to avoid when you’re setting up these different types of parameters?
Yeah, there’s a few bits to that question. So the first part I heard was changing the parameters in your system. The second part was copying other people’s rules and, you know, automation approaches, and then the subjectivity. So let me cover those in turn because I think they’re all really important.
The first one is about your own systems. How often should you change your own system to, you know, adapt to the market, to tune it up, to make sure that it’s relevant? And the answer is generally much less often than you would think because what you want to think about is why you’re driven to make a change to the strategy. And usually, if you’ve got a strategy and you’ve proven in the past that it worked, and you’re applying it, you’re generally driven to change that strategy because you just had a losing trade or you’re in a drawdown. It’s like, oh, I’m trying to avoid losing trades and I’m trying to avoid drawdown, so let me just optimize this or fine-tune it, change it a little bit or a lot so that it works now.
But it’s a trap because re-optimizing and fine-tuning or changing the parameters to make the system work now, now as in based on the trade you just took, that’s still the past, and the next trade is still unknown. So the parameters you’re about to change it to may be no better than the ones that you had. It’s back to the previous point about the uncertainty and the variability of the markets. We have to accept that there’s losses and there’s more of them than we would like, and they’re often more frequent and bigger than we would like. We’ve got to try and keep the losses small to survive. But the key is to develop sufficient confidence in the strategy so that you can trade through a drawdown without being tempted to tinker with it.
Because if you change the strategy and then you have a loss, you change the strategy again, you have a loss, you change the strategy again, you end up in this spiral where you just keep going into bigger and bigger and bigger drawdowns, and you don’t recover because you’re changing the strategy the whole time. But if you keep the strategy constant, then you have a drawdown. Usually, if it’s a good strategy, you come out of the drawdown. But if you change the strategy, now you’ve got a new strategy, and there’s nothing to stop you having another drawdown, and then change it again, another drawdown. So it’s actually a bit of a death spiral.
If you keep changing the strategy, you keep going into drawdown, and your account basically just gets whittled away. The key is, is it a good strategy? And that’s something that, you know, that’s a harder question to answer. You need that testing process that takes you through, you know, multiple stages to evaluate the edge and to check that it’s stable, check that it’s robust, check that all the rules are significant, check that it works on seen and unseen data on related markets.
There’s a whole bunch of tests you can do to build that confidence so that when you go into a drawdown, you’re not tempted to change the strategy. So not very often is the answer to the first question. But you need to monitor it often because we want to monitor our strategies and make sure that they’re still behaving the way they’re supposed to. And drawdown is something that every strategy has. There’s supposed to be drawdown. You cannot avoid it, and if you try and avoid it, you can’t succeed in trading. There’s always drawdowns. There’s always losing trades. So we need to monitor our strategies to make sure they’re not broken. And things can change, so strategies can break.
What can change that causes a strategy to break is all sorts of things. It could be market behavior, it could be market rules. I mean, in the past, there’s a whole bunch of things that have caused different strategies to break. Changes in commission levels, chan-like, changes in the market. So when the markets moved to decimalization from fractions in the share price, that changed things.
When high-frequency trading started, that changed things. It changed the way the markets moved, which killed some strategies. So you need to monitor the strategy on a regular basis, weekly to monthly, depending on the duration or the timeframe of the strategy, and check that what it’s doing is what it’s…
is within the bounds of normal based on its historical behavior. So we don’t change our strategy very often, but we monitor it often, and if it looks like the strategy is diverging from how it has behaved in the past and how it should behave based on our testing, then our job is to intervene and ask why. So back to that idea of curiosity. You know, what happened here? Why has this changed? Is it behavioral? Is it, you know, did I make a mistake? Is it more fragile than I thought? Dig into why.
The next part of your question… So does that cover the first part? You know, how… Yeah, good. So the next part of the question was, how often should we be copying other people? And I think, look, we all learn from other people. Look at the books over my shoulder. I’ve got tons of ideas from other people. I read websites, articles, journals about trading. I listen to podcasts. I get ideas. So I think that’s really important. But what we don’t want to do is jump from strategy to strategy to strategy.
What I’m trying to do is build a portfolio of strategies that complement each other nicely. So if I read something and find a strategy or an idea from someone else that is a good complement to my portfolio, then I’ll test it and add it in if it adds value to the portfolio, and that’s a testing process because not every strategy adds to what you’ve already got. But if it’s low correlation, if it adds diversity, if it improves your risk-adjusted performance, if it improves your performance in an extreme market dislocation, then adding someone else’s strategy is really powerful. Make sense? So I think we all gotta be humble enough to learn from other people, but we’ve also gotta be skeptical enough to test what other people say and build confidence in it ourself. So just because, you know, some guru or some big name talked about this approach doesn’t mean it’s gonna work for us. We have to still test it and validate it and still build the confidence in it ’cause it’s our money on the line. And when it’s our money on the line, it’s our emotions that are gonna muck it up.
So we have to have absolute confidence in the strategy, so we’ve really gotta go through that testing process to get all our questions and concerns out so that we can then follow it as if it was ours.
Now, I think there was a third part to the question, but I forgot what it was.
Yeah, it was just when you’re making those calls, how do you avoid being subjective as much as possible?
Oh yeah, being subjective, the emotional side of it. ‘Cause quanti- quantitative trading, systematic trading, algo trading, whatever you wanna call it, they’re all basically the same, you know, similar sort of things. It’s designed to take the emotion out. And so in the day-to-day when you’re following the buy and sell rules, yeah, it does.
But the trader still has the ability to press the override button or to press the eject button, or the stop button, or to jump in and reoptimize the rules like we were talking about earlier. So that comes down to confidence, and I think the… There’s a couple of tools that you can use, not technical tools, but processes. The first one is journaling everything that you see and do in the markets and learning from how you’re feeling. So this happened and made me feel this way, and then taking that and turning it into testing. So one of the great techniques that I think is massively underused, mostly because no one talks about it, I talk about it a fair bit, but is backtesting your emotions. And so when you’ve got a strategy and something happens and it causes an emotional response which makes you wanna react, I step back from that and say, “All right, what happened, and what was my emotional response?” And I try and turn that emotional trigger into a rule that I can test. Oh, that stock was way more volatile than I thought. Let me put in a rule to filter out volatile stocks so that I can avoid trades like that in the future and see if it improves my strategy. And when you test these triggers of your emotional reactions, you generally see that your emotional reactions aren’t helpful. So, oh, this stock is really gappy from one day to the next. Let me put in a rule that eliminates stocks that gap overnight and see if it improves the strategy. So I’m not doing it to tinker with the strategy, I’m doing it to prove or disprove that my reaction to the markets was real or meaningful. And most of the time, your emotional reactions, you disprove them by doing that because they’re just emotions. They don’t actually help in the markets. And when you disprove the emotional reaction by analysis, you go, “Oh, okay, I can just ignore that in the future,” and it calms the whole thing down. So the more you test your ideas and your hypotheses and your emotional reactions to the market, the more you can just have confidence to let your strategies run.
So I’m doing a lot of testing, but I’m not doing a lot of changing of what’s live in the market because the testing is all about making sure that I’m maintaining confidence. So I test fast, but I change very slow.
That was great. Thank you so much. You’ve really opened up my eyes to like what the world of trading could…
Okay, cool.
And I’m a lot more interested. So with that, I’ll pass it on to Will.
Thanks, Drew. And thank you, Adrian. It’s been really interesting to kinda hear how you apply your thinking to different levels of uncertainty. And I actually had a question regarding the actual testing process, if that’s all right.
You mentioned having a lot of confidence in your systems and kind of the mistake of looking at limited sets of data that might appear to work for a strategy that end up being kind of uncorrelated. So what are some examples of potentially dangerous mistakes that you see people make when they’re back testing systems that kinda give them that false sense of confidence?
There’s so many mistakes, I can’t tell you, honestly. Some of the biggest mistakes are really driven by trying to get certainty and trying to build an outstanding strategy. You know, trying to build the one strategy that will make you all of the money that you want and generate all of the profits to achieve your goals. Because when you’re trying to develop a great strategy, you miss so many good strategies. And the trouble with trying to develop a great strategy is that all you can develop on is past data.
And so what looks great in the past, at best, is probably only going to look good in the future. So trying to develop the best, you know, an outstanding strategy that has a super smooth equity curve, ultra high Sharpe ratio, very low loss, you know, rate of losses, high win rate, big wins, small losses, you can only really do that by overfitting the data. And overfitting the data means your rules are fine-tuned to exactly what happened in the past. And so when you apply them to the future, they don’t work because they were really precisely tuned to the exact circumstances in the past. And the future is not going to be exactly the same as the past. You know, they say the future’s, the past performance doesn’t predict, past doesn’t predict the future, but it kind of rhymes.
And I think that’s a good way to think about it. It’s going to look and feel somewhat similar, but the exact movements from day to day, from week to week are not going to be the same as they were in the past. So the biggest mistake is really trying to develop an amazing strategy and adding too much complexity, too many rules, doing too much optimization because you end up overfitting and you end up with a strategy that looks great in the past, but doesn’t perform in the future. The goal of a systematic trader should be to develop a strategy that survives through multiple market regimes and survives on unseen data because the future is unseen. So we need some simple, blunt, robust rules that keep us safe and allow us to capture some profit and allow us to get out before we give all that profit up.
And so good strategies are typically far simpler than most people think when they’re starting out. And so if you find yourself adding rules and filters and exotic kind of conditions and pulling in lots of different data sources to get consensus, you’re probably overfitting and you’re probably not going to end up with a strategy that works in real-time trading. So keep it simple is the best advice I can give. The fewer rules, the better. And the more trades in your sample set that allow you to profit from those simple rules, the better.
Got you. Thank you so much. I had another question about overfitting. I know a lot of AI and machine learning are increasingly used to kind of scan through data for different patterns and edges that humans might not be able to see themselves. And so do you kind of see that AI as like a genuine edge, or does that risk of overfitting cause a lot of potential issues?
Both. Absolutely both, depending on how you use it. It’s kind of like… You know, if you’ve got a hammer then that’s a really great tool for some things, and it’s not a really great tool for other things. You know, a hammer’s pretty good at hammering in nails and removing nails. It’s not very good at performing surgery or tightening a bolt or something like that. AI is really great at some things and really not very useful at other things. So what I use AI for is research to find ideas. So, you know, scour the web, scour social media, pull in what certain people are saying in terms of people I respect or other people that develop systems to give me ideas to develop strategies to trade with.
So it’s really great at kind of scanning what’s available, looking at papers, and pulling out ideas and rules from finance journals and those sorts of things. It’s great at that. If you give it– If you ask any AI tool for a profitable trading system, you’re going to get garbage. Yeah, because the AI tools are trained on data sets that are public, and if an AI tool is scraping social media and all sorts of random websites and pulling together the consensus kind of answer on what is a profitable trading system, what’s the probability that that’s actually right? I mean, if you read every website and every social media post about trading and took kind of the average most common answer, do you think it would be profitable? Of course not, because most people aren’t profitable in the market. Most people lose money. So asking an LLM to give you a trading system doesn’t help.
However, you can give it your process and have it run your process for you to speed you up. So as we speak, Claude Code is doing analysis for me the way I would do the analysis. So it’s coding up the system rules, and it’s running the tests for me the way I would run them, including all the steps in my process because I’ve programmed that. I’ve created the skills to do each step of my system development process, and then it spits out a report at the end to tell me whether the idea was any good. And not only that, it looks at the analysis that it did and the conclusions that it found, and it brainstorms new ideas and puts them in at the front.
So that process, I would do that myself, and I have done that for over twenty years now. It just happens faster. So it speeds it up. So it speeds up the discovery. That can speed up the discovery of luck, you know, flukes in the data and so on. So you don’t want to trust that blindly. You know, I have a incubation process which any strategy will go through and I’ll of course, when I get a strategy out the end after Claude Code has followed my whole process, I’ll really critically evaluate how that was developed.
Just the same as if a student gave me a strategy, I would very cri-critically evaluate how that was developed and make sure that the testing was sound, the conclusions were sound, that the right ideas were added in, and I might go back and retest certain things. But it’s a really great tool to accelerate us as quantitative traders.
You know, we don’t need to write our own code anymore. I don’t write code anymore. You know, I just kind of give it the i– the rules in plain English, and it says, “Hey, you missed this and you missed that,” and then it codes it up, and then it can run the testing process for me. So I think it’s a really powerful tool. In and of itself it’s not an edge. I would say it’s an accelerator. It’s an accelerator for idea generation, and it’s accelerator for coding and testing. But I very much doubt that if you plug an AI into an account and said, “Make me money,” that it would work yet. But, you know, I could be proven wrong.
I just know that if you have back-tested strategies that have survived multiple market regimes over multiple decades, and the strategies are very robust and stable, then it’s possible to make great returns in the market, and the AI can help me discover more edges like that. So that’s how I’m using it.
Gotcha. Thank you so much. That makes total sense.
I’ll pass it off now to Devika. I think she had a couple of questions.
Hi, Adrian. My name is Devika. I’m a finance and fintech student at Georgia Tech, and I was just wondering, I know you’ve traded through some pretty brutal market environments like two thousand and eight, the twenty twenty COVID crash, and rising rate environments. So could you just describe emotionally what it feels like to sit in a drawdown when your system is telling you to stay the course?
Yeah, it’s interesting, and I think sitting through drawdown is probably the hardest thing traders have to do because no one likes to see their money erode. I mean, when we go to work, we like to get paid, and when you go to work in the markets, often you don’t get paid. Often you have to pay for the privilege in the form of drawdown and losses.
So look, emotionally it’s tough, but what we need to do is develop the resilience to sit through it, develop the confidence to sit through it. And so you do that a couple of ways, and we’ve talked about testing a lot, so I won’t go back into that. You really need to understand how your strategy has performed through environments like that, so that when you go– when you come across environments like that, you know what to expect.
And so a good technique is to look at the equity curve, the back tested results of your strategy in detail. Not just, oh, over thirty years it made, you know, huge amounts of money. That’s fantastic, and it had a ten percent drawdown. Great, I can cope with ten percent drawdown. You know, looking at the superficial surface stats is not enough. You need to go down to the day-by-day level and the trade-by-trade level and see what actually happened under the hood. Because when you’re at the hard right edge of the chart, and you’re placing the trade and you’re watching your account, that’s when the emotion comes up. So you test at the macro level, not macro as in macroeconomics, but macro as in zoom out twenty or thirty years of data, right? You test at that level and go, “Wow, this is great. You know, I’m gonna be rich. Woo-hoo.”
But then you zoom in and you trade at the micro level, which is, “Holy shit, I had another loss, another loss, another loss. Oh, I’m in drawdown. Oh, I’ve been in drawdown for three months. Oh, is that normal?” And so that’s all the emotional stuff.
So you’ve got to test at that macro level, but then zoom in and look at what to expect at the micro level, the day-by-day, trade-by-trade, and see what’s normal. Because when you’ve seen in the data what’s normal, when it happens, it doesn’t freak you out, and then you’ve got the confidence to see it through. So I think that’s probably the biggest tip. Go from that big w- long term fra- time frame testing, develop the system so it works, but then really zoom in and inspect and interrogate what could happen. You know, what’s the best trade? What’s the worst trade? How fast can these things move against me? How fast can they move in my favor?
How many days in a row? How many losses in a row? You know, how long is the drawdown? H-how did that drawdown look day by day? Do that so that you know what to expect, you can sit through it.
Got it. Thank you so much. That was very insightful and…
Two great questions. Two very important stuff.
I’ll pass it over to Konark for the next question.
Hi, Adrian. My name is Konark. I’m a master’s student studying quant finance and interning at IBKR in the electronic trading compliance team. I completely agree with you. Everyone should have a system, rules, understand risk management in trading, especially with AI help. It has become very easy. So my question is about how AI has changed systematic trading. We all can see AI and LLMs are rapidly entering retail strategy development. As per you, over the next few years, do these tools make your students better system developers or just fast curve fitters? Like what guardrail should one follow while using AI?
Yeah. I think it’s a good question, and I think similar to what I said before, it has the potential to do both. It has the potential to make you a better systematic trader, but also to be a faster curve fitter, and we really need to avoid that. And I think the biggest mistake that people will make is completely deferring to the AI. You know, “Test this strategy for me.” Okay? If you tell AI to do something, it’s gonna have a good crack at doing it, right? It’s not gonna say, “No, tell me exactly how.” It’s gonna look at its information and come up with an approach and have a crack at it, and that may or may not be sensible. So I think using it superficially is dangerous and will remain dangerous.
But if you give it the process and you tell it exactly how to do the analysis, then it has the potential to speed up the good analysis that you would do yourself. If you give it a bad process, then it has the potential to speed up the curve fitting, the overfitting, and give you bad outcomes. So the fact that AI exists doesn’t mean we don’t need to understand the fundamental principles of solid systematic trading and correct back testing and all of that. We need to learn that so that we can properly drive the tool. You know, if we’re gonna be a carpenter, we need to learn how to use a hammer. We need to learn how a house goes together. We can’t just blindly trust the robot to build a house. We need to kind of be able to supervise, I guess, is a similar analogy.
We can’t just blindly let AI develop our strategies. We need to supervise it. And it’s our money, so we need to be responsible. And I think, you know, responsibility is, you know, back to, all the way back to the beginning, we talked about mentoring and Van Tharp, who was really big for me in my early days, was personal responsibility is critical. How did I make this happen? What did I do that caused this to happen? And being fully responsible for everything that happens in the markets in our account. And if we abdicate responsibility to AI, then we’re not maintaining that control and accountability ourself, and it leads too easily to blame.
It leads too easily to being removed from the process and bad results. So I think to sum it up, actually learn and understand how to be a quana- a quantitative trader, and then use the AI as a tool with the processes that you’ve learned to accelerate you rather than deferring. Lots of people are gonna come to the markets without any education and assume AI knows and defer to AI, and I think that’s a mistake. Learn to be a trader first, use the tool, and then it’s gonna be very powerful. I mean, my testing now, the speed at which I can evaluate a hypothesis is so much faster than it was just two years ago. But I’m still using the same process. It’s just all happening under the hood now.
You know, my back tester doesn’t even pop up on my screen anymore. It all happens in a silent window that’s hidden. But the same tests are going on that I would’ve done. So I think we need to use the tool correctly. Does that help? I mean, there’s so much more to this. We could talk about this for hours, but I think that’s a good start. What do you think?
Yeah, that answers well. Like I also have same thought process on these things to have the guardrails and everything intact, and human can’t be replaced by AI on these things.
Yeah, we are augmented, right? I think we’re augmented by these things, and the guardrails are the process that we follow, the testing process, the bounds, the rules about what is safe in the markets and what’s not, and the rules about how to make optimization decisions. You’ve got to kind of educate the tool about how you would make the decision so it can make the same decision you would so you can accelerate.
Thank you so much, Adrian. I will pass it over to Drew to close us out.
Yeah. Adrian, thank you so much for joining us and sharing your insights and experience. We’ve really enjoyed hearing about your journey, your approach to systematic trading, and your perspective on the entire process. We appreciate you taking the time to share your knowledge with us and our audience.
Thanks so much. Been super fun. There’s been some really great questions here. I hope it’s been valuable for you guys, but also for the listeners. I think– Look, in summary of all of– for all of this, systematic trading is such a blast. It’s so interesting and fascinating to drill into what drives the markets and how to extract an edge, and also to learn about s- you know, the psychology of how to actually, as a human, step back and let the strategies do their work despite the uncertainty, despite the fear. So if anyone’s interested in pursuing this path, I would say it’s been so instrumental in my life and it’s unlocked so many great kind of opportunities, and I’ve learned so much about the markets and myself by doing this. So keep asking these sorts of questions and look, if anyone wants to pursue systematic trading, I’m more than happy to answer questions and so on in the future, so feel free to reach out.
Awesome. Thank you so much. And to our audience, if you enjoyed today’s episode, please subscribe wherever you download your podcasts from. And we’ll be back soon.
Thanks everyone.
References to third-party products or services by guests are the guest’s own opinion and experience and do not constitute an endorsement by IBKR.
The analysis in this material is provided for information only and is not and should not be construed as an offer to sell or the solicitation of an offer to buy any security. To the extent that this material discusses general market activity, industry or sector trends or other broad-based economic or political conditions, it should not be construed as research or investment advice. To the extent that it includes references to specific securities, commodities, currencies, or other instruments, those references do not constitute a recommendation by IBKR to buy, sell or hold such investments. This material does not and is not intended to take into account the particular financial conditions, investment objectives or requirements of individual customers. Before acting on this material, you should consider whether it is suitable for your particular circumstances and, as necessary, seek professional advice.
The views and opinions expressed herein are those of the author and do not necessarily reflect the views of Interactive Brokers, its affiliates, or its employees.
Join The Conversation
For specific platform feedback and suggestions, please submit it directly to our team using these instructions.
If you have an account-specific question or concern, please reach out to Client Services.
We encourage you to look through our FAQs before posting. Your question may already be covered!