Platform: Metatrader 4
Advisor version: 1.5, 1.6 (custom modification), v22 – the latest version of the robot
Currency pairs: EURUSD and GBPUSD
Working time: only at night
Robot type: trading from the borders of the night range
How Robotron trades
The exact algorithm of work is not described anywhere, but the settings and the nature of the deals made show that this is a night scalper. The profit for each trade is minimal.
Most likely, the author used either the oversold or overbought indicator, or Bollinger bands, or his own design. Our special course “Profit on the machine” includes a similar robot called 008, in which the trade is bouncing off the borders of Bollinger bands.
The visualization of trades confirms the assumption about the advisor’s algorithm.
Note that the EA puts stops and take-profits, but they are so large that they almost never work. For example, with the default settings, TP is 100 points for 4-digit quotes. SL and TP are needed here rather in case of unexpected movements at night – a kind of safety net.
Although signals with basic settings are searched for 2 hours, the trades themselves remain open longer. Most likely, the EA exits the market according to the indicators.
Expert advisor settings
You cannot influence the market analysis algorithm through the settings, only auxiliary parameters are available:
- MaxSpread, MaxSlippage – filters by spread and slippage in points;
- HourStartSummer, HourEndSummer – sets the time range for signal search;
- HourStartWinter, HourEndWinter – the same, but for winter time, consider whether your broker’s time is changed;
- Lots – trade volume;
- UseRisk – whether dynamic lot will be used. If true, the EA calculates the trade volume as a percentage of the deposit;
- MaxRisk – set the risk per trade for a dynamic lot;
- TradingStyle – Conservative, balanced and aggressive working styles are available. How exactly they differ is unknown;
- StopLoss, TakeProfit – loss and profit in points, respectively;
- MinProfit – the minimum profit, upon reaching which the EA will look for a moment to exit the market;
- DynamicCloseLoss, DynamicCloseLProfit – parameters for dynamic TP and SL;
- TradeFriday, TradeSunday – whether trading will be conducted on the night from Friday to Saturday and from Sunday to Monday.
I recommend that you be extremely careful with the temporary settings. The EA works normally only on a “quiet” market.
Robotron test results
In addition to the version of the Expert Advisor with the v22 index, we managed to find version 1.5 (basic algorithm) and a custom modification 1.6 on the network. All of them were tested on the EURUSD currency pair for the period from the beginning of 2018 to the beginning of November of the same year.
The initial deposit is $ 10,000, a dynamic lot equal to 3% of the capital was used. The work was carried out on the M5 timeframe. Signals were searched for from 00:00 to 02:00 Moscow time. The results were as follows:
- Robotron 1.5 – the deposit increased by 97.2% with a maximum drawdown of 15.32%. 94.66% of all concluded deals were closed with a profit. A little more than 10 months of trading, 637 transactions were concluded;
- Robotron 1.6 – the results are radically different from those of the basic algorithm. The deposit was not drained, but a loss was recorded, and the growth curve of the deposit leaves questions. The settings are the same, but the result is completely different;
- Robotron v22 – there were half the deals concluded than in Robotron 1.5, while the loss at the end of the test period was 11.1%. Winrate is worse at 66.56%.
The reasons for these differences in results are not entirely clear. Finally, we decided to draw conclusions about performance based on the results of trading on a real account.
Robotron results when working with real money
The initial deposit was $ 500, trading was carried out for a month with the recommended parameters. The v22 version of the EA did not execute any transactions during this period, so the results below are given for Robotron 1.5.
In total, 22 deals were concluded, most of the time the loss was approximately equal to the profit. The results of the deals fluctuated around $ 10. But at the beginning of November, there was a substantial loss of $ 50 on GBPUSD, which was the reason for the total loss for the month of trading.
The growth curve of the deposit shows that Robotron, with all the desire, cannot be attributed to reliable advisors. Large losses that occur periodically cancel out the small gains made the rest of the time.
In the course of our experiment, it was not possible to accelerate the start-up capital at times or at least get a solid profit. But we did not set this goal for ourselves. The task was to check how different the results in the tester and in real trading can be. As for me, this goal was achieved, while I’m not saying that all EAs are bad, but it is very difficult to find a worthwhile algorithm among the free ones.
As for Robotron, it cannot be optimized. Although the idea itself is worth considering, its implementation leaves much to be desired.
Perhaps you get to get the version with the v22 index in the title to work. Be sure to write about the results in the comments.