Greetings !
Most of the systems on the MQL Market try to predict the future.
The only tool we have is history.
There is no future without history. But how to predict it?
Everyone in MQL strives to solve this question.
Most TOP systems stop working after some time.
1. Reason. Hackers.
As a rule and most often this happens due to constant hacking of these systems by hackers.
Hacking occurs using environment emulation via msimg32, oleacc etc.
All these DLLs allow you to emulate the operating system environment, activation is tied to these mechanisms.
Therefore, tracking all this from the MQL side is almost impossible.
After such a hack, the system falls into thousands of hands around the world and around the world, it is already put on stream.
Therefore, as a rule, the system shows itself perfectly at the beginning of the path, but then it changes.
Fortunately, we have learned to deal with this and all our systems are protected from such hacking.
We use an end-to-end encrypted security server and https, this is already working for us now.
2. Reason. Razor blade.
This is a OVER-optimization. We all want to have a perfect algorithm that never misses.
It's like walking on a razor's edge, one wrong step and you fall down.
Therefore, the developer falls into such a trap, the settings are so good that any change in the market breaks them.
It is like a fragile house of cards, it is big, tall and beautiful, but if the wind blows, it will collapse.
It is possible to create such settings, but will they predict the future.
And yes and no, it always depends on the degree of volatility of market situations.
Therefore, we always have inside protection against big news, such as FOMC etc.
Slip protection and constant adaptation to broker conditions, spreads, delays.
Now it already works for all our systems.
3. Reason. Not all brokers are the same.
All brokers have different quotes and different spreads.
This is due to different quote providers and therefore they differ.
Therefore, the simplest solution is to create a very simple strategy and make it work.
The more complex the algorithm, the higher the chances that it will not work in the future.
4. Reason. Margin of safety.
Also, don't forget about forward testing. Especially if it concerns Grid or Martingale.
The settings must have a margin of safety, this is also very important, you should not squeeze all the juice out of the algorithm.
Forward testing allows you to test the system up to a certain date.
This requires a fast algorithm and high optimization speed
For example, until December 2022, if you receive excellent settings and they continue to work until the current day, then you actually have several months to check the system and its settings.
The basic principle here is that if it worked before, it will work now.
5. Reason. DONT Use Martingale or DONT use only 1 pair.
Most systems on the market trade with only 1 currency pair, these are EURUSD/GBPUSD/USDJPY/GOLD
Such systems can be very expensive, although they only trade 1 pair.
At the same time, the systems are always closed, this is a black box, you will never get the parameters to adjust your algorithm and get away from the crowd.
Some systems do not even have a real signal, real trades and still use Martingale.
6. Reason Luck
All our systems are multi-currency. Moreover, they all trade simultaneously on several currency pairs.
For example GWay uses 5 currency instruments, DIVO uses 12 currency pairs for trading, Eve uses 10 pairs.
All these currency pairs have different settings, different optimization, different time frames.
At the same time, all systems are open, moreover, you can read the guide and create your own unique settings based on it.
Creating such a multi-currency system is always a lot of work and time resources.
It takes approximately 2-3 days to create settings for 1 instrument/currency pair.
It may take 1 month to create all settings for 12 pairs.
Therefore, some pairs may trade slightly worse than others.
Out of thousands of settings, we always choose the best in terms of profit factor, recovery factor and drawdown.
Then we carry out a forward check until today.
And even in this case, the luck factor plays no small role, because we can choose the settings a little worse or better.
Therefore, the luck factor also comes into play, especially when you have 12 currency pairs to trade, and not 1 and the one with the Martingale...
But on the other hand, we can always choose only the best currency pairs from a huge list just for your broker.
I hope these causes and consequences will help you choose the best way to check and analyze systems in MQL.
If you have read this to the end, perhaps you can find more in My Full Blog
