Difference between a good and a bad trading robot
The automation of trading strategies has revolutionized financial markets. On MetaTrader 5 (MT5), traders can create or purchase Expert Advisors (EAs) capable of opening and managing positions without human intervention. But not all trading robots are created equal. A good robot can become a reliable ally in complying with the rules of a prop firm program such as FTMO, while a bad robot can ruin an account in just a few trades. This analysis explains how to distinguish between the two and how to evaluate an EA before entrusting it with your capital.
What is a trading robot?
A trading robot, or Expert Advisor, is software that automates the execution of predefined strategies. On MT5, an EA reads prices, applies indicators, and places orders according to coded rules. The goal is to eliminate emotion and enable rigorous algorithmic trading, adaptable to Forex or CFDs. Unlike a simple indicator, a robot manages entries, exits, position sizes, and sometimes fundamental filters.
Prop firms require strict money management and maximum consistency in strategy. To understand the basics, check out our article on automated money management for FTMO: essential rules.
Characteristics of a good trading robot
A robust and simple strategy
A good robot is based on a trading strategy that has been tested in different market contexts. Experts recommend avoiding overly complicated or over-optimized algorithms. According to the Nurp website, strategies that are too heavily calibrated to historical data (over-optimization) often fail in real-world conditions. An effective EA favors clear rules (Donchian, Bollinger Bands, moving averages, etc.) and adapts to multiple markets without constant adjustment. Simplicity facilitates maintenance and avoids hidden errors.
Integrated risk management
The best EAs incorporate risk limits. The PickMyTrade blog points out that one of the main causes of blown-up accounts is over-leveraging and poor position sizing. A good robot limits the risk per trade to 0.5–2% of capital and adjusts position sizes based on volatility. It also enforces automatic stop-loss and take-profit orders. In tests, you should see moderate drawdown and consistent gains. To learn how to configure these settings, read our guide on ideal risk settings for an FTMO robot on MT5.
Rigorous testing (backtesting and forward testing)
A good robot is not only profitable in a quick backtest. It is tested outside the sample, for example in walk-forward testing, and validated in forward testing on a demo account. The PickMyTrade blog warns that going live without backtesting or stress testing often leads to failure due to slippage, low liquidity, or over-optimization. A serious EA comes with a transparent history showing the capital curve, success rate, average trade duration, and max drawdown. For more on this, see our comparison of forward testing vs. backtesting: which is more reliable for FTMO?
Adaptability and consideration of external events
An effective robot does not just trigger orders based on technical analysis. It must also manage periods of high volatility. PickMyTrade notes that ignoring macroeconomic announcements can cause huge gaps and blown-up accounts overnight. A good EA includes an economic calendar and can suspend trades during major events, or at least increase stops to limit the impact.
Data quality and filtering
The reliability of input data is crucial. Nurp points out that incomplete or delayed data leads to poor decisions. A good robot uses reliable data feeds and checks price consistency. It incorporates filters to ignore erroneous signals and avoid counter-trend trades. If the strategy depends on the correlation between pairs, data quality is even more important; our article on Forex correlation and pairs to avoid in automated trading addresses this topic.
Simplicity of code and transparency
Finally, a good robot is easy to understand and modify. Overly complex algorithms are difficult to debug and generate errors. The author must document each parameter (indicator periods, volatility filters, etc.). This transparency allows traders to understand the origin of the signals and modify the robot according to the prop firm's requirements. A simple and transparent solution also encourages the use of proven configurations such as the Donchian + ATR strategy for breakouts, which we detailed in our article on the Donchian + ATR combination: high-precision configuration.
Characteristics of a bad trading robot
Over-optimization and the perfect curve
A "miracle" trading robot that displays a capital curve without drawdown may be too good to be true. Nurp warns that strategies that are overly tailored to historical data are often over-optimized and subsequently fail. They capture noise rather than real trends and are unable to adapt to new market conditions. A bad EA will not survive a change in volatility or regime (range vs. trend).
Excessive position size and lack of stop-loss
Bad robots bet on overly large lots or use martingale to recover losses. PickMyTrade reports thatexcessive leverage and the absence of stop-losses are the main causes of destroyed accounts. Without risk rules, a robot will continue to trade during flash crashes or illiquid markets. It can quickly exceed the drawdown limits imposed by FTMO. This is the mistake to avoid if you want to succeed in your FTMO challenge: our article on overly tight stop-losses and their impact explains how to size stops.
Lack of testing and blind trust
A bad robot is often sold without proof of performance over several years. PickMyTrade reminds users that going live without paper trading or forward testing is a trap. Users place their trust in the code without checking how it behaves during economic announcements, periods of extreme volatility, or widened spreads. This lack of rigor is incompatible with the discipline imposed by prop firms.
Ignorance of macroeconomic events
Some robots continue to trade during FOMC or ECB announcements. PickMyTrade indicates that ignoring macroeconomic news is a major mistake. "Bad" robots have no built-in calendar or settings to suspend trading, which exposes the account to massive slippage. To avoid this, we recommend using news filters or combining the EA with manual intervention during announcements. Our article on automatic FTMO validation: myth or reality? shows how to integrate an economic calendar into a robot.
Unnecessary complexity and opaque code
An EA riddled with conditions, loops, and esoteric calculations is often a bad sign. Overly complex algorithms are difficult to maintain and generate undetectable errors. Opaque code prevents traders from understanding what is happening and making corrections. A high-quality robot must be simple enough to be audited and modified.
Poor data quality and overlooked costs
A bad robot feeds on inaccurate price data, low resolution, or unsynchronized feeds. Nurp points out that incomplete or erroneous data leads to misleading signals. In addition, a poorly designed robot ignores transaction fees, commissions, and slippage, which significantly reduces actual profitability.
How to evaluate a robot before entrusting it to your FTMO account
- Strategy analysis: Verify that the robot follows understandable logic and is not based on arbitrary parameters. Ensure that it is based on a relevant signal (breakout, moving average, ATR, etc.) and does not use martingale. For an example of a complete plan, read our FTMO strategy in H1 with an MT5 EA.
- Serious backtesting: Perform backtesting over several years using high-quality data. Include fees, slippage, and variable spreads. Use out-of-sample testing to verify robustness. Read our article Forward testing vs. backtesting to understand the differences.
- Forward testing and demo account: Run the robot in real time on a demo account to observe its behavior in response to volatility and announcements. This test should last several weeks.
- Risk management: Verify that the robot allows you to set position sizes as a percentage of capital (1–2%), apply a stop-loss, and set a profit target. For more information, see our advice on ideal risk settings for an FTMO robot and automated money management for FTMO.
- Transparency and documentation: Demand clear documentation that describes the robot's logic, available parameters, and output conditions. A transparent developer will make it easier to adapt the robot to FTMO rules.
Conclusion
The difference between a good and a bad trading robot boils down to discipline and robustness. A high-quality EA combines simplicity, transparency, risk management, and rigorous testing. Conversely, over-optimized robots without stop-loss or testing are time bombs. Before committing to a prop firm like FTMO, take the time to study the strategy, test the robot thoroughly, and make sure it complies with your own risk management rules. Feel free to browse our other resources, such as How to succeed at FTMO with an MT5 robot: complete method and settings or Avoiding over-optimization of an Expert Advisor, to learn more about these topics.