TradeQuantiX and The Rogue Quant - Part 1
Opinions and Viewpoints of Two Different Systematic Traders
Introduction:
Hey everyone, this article will follow a unique format compared to others.
and , have collaborated to write a joint piece. In this article, we’ll share our personal perspectives on three distinct systematic trading topics. You’ll get a glimpse into how two different systematic traders view the world of systematic trading. Everyone has their own unique opinions on what’s considered optimal or a best practice, and you likely have your own views as well. The goal of this article is to offer two additional perspectives from two completely different systematic traders with diverse trading backgrounds. Now, let’s dive in!Background:
The Rogue Quant Background:
Hey, I'm Leo. I've been trading for over 25 years. I've lived through bull runs, crashes, and just about everything in between.
For the past decade, I’ve focused on one core question:
How do you build trading systems that actually hold up in the real world?
That’s what I share in The Rogue Quant newsletter. Every Saturday, I deliver one fully backtested strategy, complete with EasyLanguage and Python code. I break down the logic, explain the edge, and show exactly how to implement or adapt it to your own trading.
I have a master’s in financial engineering with a focus on algorithmic trading—but I don’t trade theory. I build systems that are tested, stress-tested, and then tested again. What I share isn’t just what looks good on paper. It’s what’s survived changing markets, volatility spikes, and the realities of execution. I’ll tell you what works—and just as importantly, where it might break.
Simple. Practical. Honest.
No magic formulas. No hype.
Just a battle-tested process for building strategies that can take a punch or two.
TradeQuantiX Background:
Hey, I'm TradeQuantiX. I run the newsletter "Systematic Trading with TradeQuantiX," where I explore a wide range of systematic trading concepts to share my knowledge with readers. My trading journey began when I graduated college as an engineer, just before the Covid lockdowns. I spent a year unemployed, waiting for the economy to recover so I could start working and earning an income. During that time, I discovered systematic trading and immediately fell in love with it. As an engineer, I relished the challenge it presented. For months, I lay in bed unable to sleep, my mind tirelessly obsessed with systematic trading.
In the beginning, I invested a significant amount of time and money into learning systematic trading, often learning the hard way by doing things incorrectly. Progress was slow, but after a few years, I successfully implemented my first profitable trading system, which I continue to trade today. Currently, systematic trading occupies my afternoons and weekends after my day job. I am a multi-country equities systematic trader, managing a portfolio of uncorrelated trading systems that focus on daily timeframes and higher. I am constantly expanding my portfolio over time, and my system research and development never stops. I look forward to discussing my perspectives on various systematic trading topics with everyone!
Where Should You Start If You're Just Beginning Your Systematic Trading Journey?
The Rogue Quant Response:
If you're just starting your systematic trading journey, the key is simplicity and clarity. Here's how I'd suggest beginning:
First, get comfortable with the idea of losing money.
No joke. Your initial trading system will probably be far from perfect, and that's okay. Mistakes are part of the game.
But there are foundational steps that every trader should master:
The basics: Start by learning core principles of systematic trading, such as what makes a strategy robust, proper backtesting techniques, and fundamentals of risk management. A great introductory book is "Systematic Trading" by Robert Carver.
Choose your tools: If coding isn't your strength, select reliable, user-friendly software like TradeStation or AmiBroker that allows you to test and implement strategies. If you're familiar with coding or want to explore it, learning Python will be extremely helpful.
Start simple: If you're already a successful discretionary trader, begin by automating your own rules. Or, start with straightforward concepts such as trend-following or mean-reversion strategies. Avoid overly complex indicators or methods initially. Your goal is clarity, not complexity.
Get data: First and foremost, data quality is crucial—ensure you're working with high-quality market data (avoid using free sources like Yahoo Finance). Always include realistic assumptions such as trading costs, slippage, and commissions.
Analyze key metrics: Develop a strong understanding of essential performance metrics, including Profit Factor, Sharpe Ratio, Maximum Drawdown, and Win Rate. Correct interpretation of these metrics will save you time and money.
Develop a process: Follow a solid methodology for developing trading systems. This structured approach will give you confidence when you transition your strategies to live trading.
Incubate your strategies: Before committing significant capital, test your strategies through an incubation period in a simulated (paper trading) environment to observe real-time performance and identify potential issues.
Keep learning: Books, courses and participation in trading communities can accelerate your learning curve.
But in short, develop patience because most quit before they get good. Your first goal shouldn't be creating the ultimate winning strategy but rather mastering the process of researching, backtesting, and analyzing trading strategies.
TradeQuantiX Response:
Many times, beginners want to trade using more advanced methodologies than they should. When I first started, I tried to create fully automated day trading strategies based on 5-minute candles. Needless to say, I didn’t make money, my automation frequently failed, and I grew frustrated. Looking back, I realize I began trading in one of the most challenging ways possible. On the 5-minute timeframe, I was competing with large institutions and hedge funds employing high-frequency trading (HFT) algorithms. I lacked the infrastructure, knowledge, and skills to compete with them and I still couldn’t compete with them today even if I wanted to!
If I could go back and advise my beginner self, here’s what I’d say: Focus on the simple stuff, follow the 80/20 rule, capture well-known market effects, trade in markets with less sophisticated players, and keep most of your money invested in the S&P 500 until you’ve built viable systems with proven results. Many beginners love to overcomplicate things, using way too many indicators and parameters, and overfitting backtests endlessly. Instead, concentrate on simple, proven concepts like trend following, momentum, arbitrage, or seasonality etc. Use as few parameters and indicators as possible, modeling the effect in the simplest way imaginable to avoid overfitting. Don’t spend weeks or months trying to perfect your trading systems—learn when to say “it’s good enough” and move on. You’re generally better off developing five decent, uncorrelated systems that each take a week to create than spending months perfecting a single great system.
Start by focusing on lower-liquidity stocks, non-US stock markets, or crypto, where the players are less sophisticated. Trust me, you can’t compete with multi-billion-dollar hedge funds, so target markets where they can’t easily participate. Finally, I’d tell myself not to sell all my investments to fund my first trading systems. Contrary to what online gurus might claim, your initial trading systems are unlikely to make you rich. If I could do it over, I’d allocate only a small portion of capital to my first systems to gain experience. That way, if things went wrong, I wouldn’t lose too much money. Again, this is all what I would tell myself, it’s not financial advice for you the reader 🙂.
If you’re a beginner eager to learn more about where and how to start your systematic trading journey, check out this other article I wrote on that exact topic:
What Requirements Do You Need To See In A Trading System To Allocate Capital To It?
The Rogue Quant Response:
One of the most overrated aspects when evaluating a trading system is the equity curve. Let me explain.
When you see an equity curve constantly hitting new highs with a steep upward slope, your first thought might be: “Wow, I want this strategy!”
But (and there’s always a but), when you dive deeper into the actual numbers and key metrics, reality sets in...
Things like:
Can I realistically trade a strategy that historically experienced a 40% drawdown?
Can I handle a system that might have 8, 9, or even 10 consecutive losing trades?
Am I comfortable trading a strategy with only a 30% or 40% win rate?
This deeper analysis is crucial because beautiful equity curves can hide these important facts. A strategy worth allocating capital to must align closely with your personal trading style and psychological comfort.
No matter how impressive an equity curve looks, every strategy eventually reveals its "ugly" side. And when that happens, if the strategy isn't aligned with your personal risk tolerance, you'll freeze. You'll start doubting your strategy—and doubt is often the first step toward losing money.
At that critical moment, traders often let intuition override a solid strategy - one that's already passed rigorous backtesting.
Thus, the most important requirement for me when allocating capital to a trading system is alignment with my personal trading style. I must clearly understand and accept the fluctuations and drawdowns that I can - and cannot - handle psychologically (and in my trading account).
TradeQuantiX Response:
In order for me to allocate hard-earned capital to a trading system, there are a few things I need to see. At a high level, those are the system is robust, the system has proper risk management, and the system adds value to my portfolio. Notice I didn’t mention anything about returns—that was intentional. To be honest, I don’t really care how much money a system makes or what its historical annual return is. I care far more about the three criteria I listed. I won’t dig too deeply into the robustness aspect because the next topic, which The Rogue Quant and I will answer, focuses on that. However, I will explore the other two aspects I mentioned.
First, if a system has poor risk management, I won’t trade it. Who cares if a system generates 100% returns per year if there’s a 90% chance it blows up (loses all your money) within the next five years? I’d much rather trade a more tame system with lower returns that has less than a 1% chance of blowing up, ideally a 0% chance. The other point I made is that the system must add value to my portfolio. This is critical, and I’ll illustrate it with two scenarios.
Scenario 1: You create a new trend following system on US stocks with a 20% return and a 25% drawdown. That sounds like a pretty exciting system! But then you look at my current portfolio and realize you already have five similar trend following systems on US stocks. Sure, this new system might add some value, but it’s unlikely to make a massive difference to the overall portfolio because all these systems will be highly correlated. It would be more I deal to add a completely different system than trend following, or add a different market than the US market, or a different timeframe etc. Just something different than what’s in the portfolio already.
Scenario 2: You create a new trend following system on US stocks, but this one has only a 12% return and a 20% drawdown—not nearly as exciting as the system in Scenario 1. Then you look at your current portfolio and see that you currently trade three mean reversion systems on the US market. Adding this trend following system could significantly improve the portfolio’s performance. Why? Because it introduces a non-correlated trading system that behaves differently and makes money during different market conditions. This is much more likely to have a massive positive impact on the overall portfolio than adding another mean reversion system.
I’d much rather add systems to my portfolio that bring value, as demonstrated in Scenario 2, rather than Scenario 1. Remember, you’re not trading an individual system—you’re trading a collection of systems that form an entire portfolio. The equity curve your account reflects is the collective result of all your trading systems working together, not just one individual system. That’s why I consider my portfolio as a whole and prioritize adding systems that enhance its overall performance. I focus on adding systems that have non-correlation compared with the systems I already trade, rather than obsessing over absolute returns or drawdowns of individual trading systems.
Okay, my answer to this question might feel a bit dry. I’m sure you’d rather hear what specific system stats I look for in backtests before trading a system. Like I mentioned earlier, the effect on the overall portfolio is nearly all that matters. In fact, I’m about to start trading a losing system simply because it adds value to my portfolio, like in Scenario 2. The system makes and loses money at completely different times than the rest of the systems I trade, smoothing the portfolio equity curve.
Alright fine, I’ll play along and make it fun. For trend-following and momentum systems on a daily timeframe or higher (which make up 80% of what I trade), I generally like to see annual returns of 12% or higher, ideally beating the S&P 500. I’d also prefer a maximum drawdown of no more than twice the annual returns. I want the system’s expectancy to be around 5% or higher after accounting for slippage and commissions. I look for stable returns over time—not a perfect 45-degree equity curve, flat periods and drawdowns are fine—but the performance shouldn’t appear to be decaying over time. Finally, I want out-of-sample test results to be at least 80% of the in-sample results. So, if the in-sample returns were 20% per year, the out-of-sample returns could drop to 16%, with the same logic applying to drawdowns and other stats. Essentially, this means 20% of the in-sample performance might be due to overfitting (there’s always some fitting involved), but 80% reflects the system’s true edge. Keep in mind, while it’s important to have general guidelines for system stats, nothing I do is set in stone or has rigid requirements. The overall impact on the portfolio is the true guideline I use.
How Do You Define Robustness In Your Trading Systems?
The Rogue Quant Response:
I like Bob Pardo's definition of robustness:
it's essentially the strategy's ability to perform as expected.
And "Perform as expected" means behaving similarly during live trading as it did in historical backtests.
In short, robustness is about achieving stability in performance. To evaluate this stability, you have numerous statistical tests at your disposal.
For me, validating a robust strategy involves various factors—from (the basics) choosing the appropriate size of in-sample and out-of-sample datasets, incorporating realistic trading costs into backtests, to applying rigorous (more advanced) methods like Monte Carlo simulations, variance testing, walk-forward analysis (WFA), adding noise and many more.
Having a robust system doesn't necessarily mean it works perfectly across vastly different instruments. However, it should maintain its edge when applied to instruments with similar characteristics - for example, trend-following assets or assets with a historical tendency for mean reversion.
Also, robustness doesn't require the system to trade both long and short positions. You can have robust strategies that operate exclusively long (or short). However, from my experience, strategies that trade both long and short tend to exhibit more consistent stability.
If this still seems somewhat vague, here's a simple exercise that might clarify robustness by examining its opposite:
So, what exactly defines a non-robust system?
Imagine your strategy uses a 50-period moving average. If changing that moving average slightly to 48 or 52 periods drastically reduces performance, it indicates a lack of robustness.
Similarly, if applying the strategy to two correlated indices like the S&P 500 and the Dow Jones produces vastly different results, that's another sign of fragility.
But the most critical validation occurs post-backtest. It’s during live trading, or what I call the "incubation period." Here, you test the strategy with 100% unseen data—real market conditions.
I usually keep my strategies in incubation for at least three months (often in a simulated account). During this period, I closely monitor executions, slippage, rollovers, and other nuances that only become evident when interacting with real-time market data.
If, after all these tests, the strategy performs as expected, you're on the right path to robustness.
Finally, it's important to note that even robust strategies can eventually lose their edge. That’s why you need to keep monitoring key metrics to identify when a robust strategy has stopped performing “as expected”.
In my article linked below, I discuss robustness in practice by analyzing the same strategy across 15 different markets.
TradeQuantiX Response:
Robustness is the most important aspect of a trading system in my view, with risk being a close second. The reason robustness matters so much to me is simple: without it, you don’t really have a trading system. I don’t care how impressive your backtest looks—without robustness, it’ll lose money in real live trading. Over the years, I’ve spent a lot of time studying robustness and learning how to develop trading strategies with it in mind. So, how do I define robustness? To me, robustness is what ensures a trading strategy continues to perform in real-time on live market data for many years into the future. No matter what challenging scenarios the markets throw at it, a robust system can handle them and recover from drawdowns.
Most people spend the majority of their time during the trading system development phase trying to craft a perfect backtest—optimizing parameters, adding multiple rules to smooth out uncomfortable drawdowns in historical performance. They’re trying to make their backtest results look robust because they assume a flawless backtest equals robustness. They tell themselves, “If the backtest survived all these scenarios, surely that will carry into the future!” Nope, that’s not how it works. That backtest is junk—not robust, just curve-fit garbage!
So, how do I assess robustness? There are many ways to test for it. Some of my favorites include parameter sensitivity tests, testing the system on similar markets, and out-of-sample testing, among others. There are plenty of robustness tests available to stress-test your trading system. While these tests are great, these days I prefer to build robustness into the system from the start of the development process. For example, if I’m creating a momentum system, I’ll ask myself what risks exist beyond what the backtest reveals. There’s the risk that my chosen trading universe underperforms in the future, the risk of picking poor parameters, the risk that my momentum ranking factors decay over time, or timing risk based on when the system rebalances its portfolio of stocks. These are all examples of risks that need to be addressed to call a trading system robust. I’ll incorporate methodologies into the system itself to mitigate these risks, making it less vulnerable to them in the future.
I like to think of these risks as “luck” in the outcomes of your trading. Sure, the system could keep performing well into the future if you get lucky—picking the perfect universe, parameters, and timing for your trades. But the reality is, you’ll most likely get unlucky. Markets shift, and future performance will fall short of your backtest. Instead, you should intentionally account for these risks to build in robustness, ensuring the best possible outcomes in an unpredictable future. This topic is pretty deep, so I’ve written a whole set of articles about it. They explore how to build robustness, control the influence of luck in your trading systems, and provide practical examples. You can check them out here for more info:
Luck By Design - Part 1
Welcome to the “Systematic Trading with TradeQuantiX” newsletter, your go-to resource for all things systematic trading. This publication aims to equip you with a complete toolkit to support your sys…
Conclusions:
Systematic trading isn’t about chasing some holy grail—it’s about building stuff that works and lasts for the long term. The Rogue Quant and TradeQuantiX both provided their perspectives about how they approach systematic trading concepts: Leo’s got decades of scars and a process forged in fire, while I stumbled in during Covid, obsessed over it, and figured it out the messy way. The point is, start basic, don’t overcomplicate it, and make sure your systems can take a hit. Focus on robustness, it matters more than a pretty backtest. Feel free to reach out to either of us with questions!