Over the years, my trading approach has evolved significantly, but the most profound transformation occurred when I began working on algorithmic trading projects. For the past four years, my primary focus has been running a research center dedicated to developing trading algorithms for both retail and institutional clients. During this period, we’ve built and tested over 3,000 algo solutions, each designed to perform across various markets.
Initially, I felt discouraged when comparing the performance of my manual trades to the results from the algorithms we developed. However, I soon realized that these algorithms could actually enhance my own trading strategy. By incorporating the insights and techniques from these systems, I began to see the greatest improvements in my 15-year trading journey.
Today, I’d like to share the most valuable lessons and strategies I’ve adopted after analyzing over 100,000 algo trades.
Every Trading System Can Be Automated, But Not All Will Be Profitable Without Oversight
No matter what others say, any trading system with a clear set of rules can potentially be automated with the right development team. The challenge, however, is that most systems are designed to perform well in specific market conditions. For instance, I like to classify systems based on volatility—some thrive in strong, fast-moving markets, while others perform better in ranging, sideways markets. From my experience, these systems require significant optimization to determine the best times for algos to operate and when they should pause to ensure optimal performance.
That’s why traders must remain involved to ensure algos are running only when market conditions are favorable. While it’s possible to fully automate this decision-making process, it takes time and rigorous testing. If you’re aiming for a system that can function fully autonomously, you need at least five years of backtesting data and a year of live testing to validate the results.
Algorithms Should Always Manage Your Trades
A trader’s role boils down to three key stages: research, trading, and risk management. If there’s one area where algorithms consistently outperform humans, it’s in managing risk and trade execution. In trading, a single minute can drastically change your position, which is why trade management is often the most challenging aspect for traders.
Every trading system has its own rules for managing risk during an active trade. The goal is to identify conditions that indicate a shift in the probability of success. For example, if you’re holding a long position, what market behavior should prompt you to start questioning the trade? Is it a series of lower lows and lower highs? A drop in an indicator’s value? Negative news? Once you understand these triggers, it becomes much easier to automate your risk management process.
Most of the strategies I use have multiple profit targets, and I prefer to take profits along the way. That’s why I have a risk algo integrated into my platform. It tracks the levels I manually input, and when those levels are hit, the algo knows exactly what to do. Additionally, I’ve built early exit strategies into my risk management algo, which continuously scans for potential exit signals based on my rules.
Of course, you can manage trades manually, but in my experience, most losing trades that should have been cut early occur when I’m either away from the screen or simply miss a signal because I’m tired. That’s why I can’t imagine trading without my risk management tools anymore.
Algos Must Be Continuously Optimized and Improved
A common misconception is that once you find an algorithm that works, your job is done. But this couldn’t be further from the truth. In our testing, we’ve found that less than 2% of algorithms continue to perform consistently after 12 months without additional optimization. The key to maintaining success in automated trading lies in continuously optimizing and testing different scenarios to match current market conditions.
For example, with just one algorithm we run in our company, we simultaneously test 30 to 50 different variations to determine which version performed most stably over the past 30 days. This allows us to ensure that the main algorithm is always running with the best possible optimization for the current environment.
In my early years of manual trading, I maintained a trading journal to track and improve my performance. While this was helpful, it limited the number of changes I could make at any given time. Nowadays, I apply a similar approach to what we do with algos. For every trade I make, I also execute the same trade on four different demo accounts, each with identical trade sizes but slightly different rules or confirmation criteria. At the end of each month, I review the performance of each account to see which approach yielded the best results. I then incorporate the best-performing strategy into my main account and continue to optimize from there.
This process has helped me maintain consistent results without drastic changes month after month.
Discipline is the Biggest Factor for Long-Term Success
Most traders have heard that discipline is essential in trading. I understood this before I started working with algos, having experienced the ups and downs of manual trading. But after testing over 3,000 different trading models in the past five years, I’ve gained a deeper appreciation for algorithmic trading and how it can reinforce discipline.
An algo works 24/5, executing trades based on the criteria you’ve programmed into it, and when you see an algo produce positive results month after month, it becomes clear that the biggest factor is discipline—following the same rules over and over.
View this post on Instagram
For all the trading strategies I use in my personal trading, I always tried to have a checklist to help me stay disciplined. However, there were times when I would relax and skip steps. To prevent myself from breaking my own rules, I now use an algo that monitors my trades. If the algo detects a mistake, it will close the trade and alert me to what I’ve done wrong. This forces me to double-check every setup before I place a pending order.
There may be other changes I’ve made after working with algos daily, but these three areas—automation, risk management, and discipline—have had the most significant impact on my trading.
Final Thoughts
I don’t believe you must use algorithms in your trading, but incorporating some form of automation can definitely improve your results. Consider the strategies you use, write down all the rules, and explore ways to automate your processes. You’ll see improvements within the first 30 days, guaranteed.
I hope you found this useful, and if you have any questions, feel free to ask in the comment section below.
P.S. If you’re interested in automated trading, check out my Trading Screener project.
0 Comments