Create and Test Forex Strategies
You are not logged in. Please login or register.
Forex Software → Binary Options Tester → Trading EA Real Results?
I was wondering if anyone has tried any of the generated EAs to a real account? Sharing the results and seeing any discrepancies between the results of the backtested EA and the live EA would be much appreciated. Or, any forward tests would also suffice.
S
I designed BO Tester in order to create binary options strategies and experts for a trading company. They use the software and trade tens of experts on real accounts. As I know, they have a very good success rate and are completely happy with the applications. However, they are professional traders and can evaluate and choose strategies well.
I know they have plans of releasing real account statements and to sell BO experts and signals for several thousands dollars per copy.
They also request to stop selling the app 
because it provides opportunities for the traders to create their own experts.
There was slight discrepancy in the first version of the application, but after the change over to the native MT indicators, all noted discrepancies were eliminated.
We only observe now slippage and gaps in the broker's execution from time to time, but this is out of our control. However, the Monte Carlo tool is capable to simulate such distortions and this reduces the risks.
I also see that they trade on live accounts with experts that show bad statistics in BO Tester. They do that because they say they make money with this logic from years and rely on their experience.
There is also a human element, that introduces discrepancies. They have a group of operators that trade big volumes on different brokers manually using signals from BO Tester experts. However, sometimes the operators make mistakes.
However, they are professional traders and can evaluate and choose strategies well.
Is there a set of evaluation rules you'd recommend? We're not professional traders, at least I am not into trading math too much..this may be the case for many subscribers.
BTW I remember reading that when a strategy provides good results in many pairs is more rigid, is this valid?
I know they have plans of releasing real account statements and to sell BO experts and signals for several thousands dollars per copy.
I would do the same 
)
There was slight discrepancy in the first version of the application, but after the change over to the native MT indicators, all noted discrepancies were eliminated.
Can you identify the reasons the discrepancies issued please?
the Monte Carlo tool is capable to simulate such distortions and this reduces the risks.
Please share with us some of the settings that increase the chances of reducing risk?
I also see that they trade on live accounts with experts that show bad statistics in BO Tester. They do that because they say they make money with this logic from years and rely on their experience.
This makes sense. In binary the result is 1 or 0 so basically if you get many 0s i.e losses doing the opposite will give you many wins.
There is also a human element, that introduces discrepancies. They have a group of operators that trade big volumes on different brokers manually using signals from BO Tester experts. However, sometimes the operators make mistakes.
I mainly subscribed to create automated EAs. I don't like being stuck in front of my PC all day long and trade, despite the fact I have a good strategy.
I was wondering if anyone has tried any of the generated EAs to a real account? Sharing the results and seeing any discrepancies between the results of the backtested EA and the live EA would be much appreciated. Or, any forward tests would also suffice.
S
I am asking this cause I don't want to go live without having a good proof an EA works. Not being able to run forward test put so many limitations in the process. I tried fwd testing with MT4, it obviously didn't work due to the coding format used in the EAs. Do you know any possible ways of running a forward test on generated BOT EAs?
Do you know any possible ways of running a forward test on generated BOT EAs?
OOS in the Generator is the simplest and fastest way.
Do you know any possible ways of running a forward test on generated BOT EAs?
OOS in the Generator is the simplest and fastest way.
10%, 20% - What is best?
OOS in the Generator is the simplest and fastest way.
The following picture depicts a simple strategy run on EU, using 20% OOS.
There is only one indicator; BB -> Open below LB -> Apply to close -> Period 11 -> Deviation 3.76
Using 5M canldes, 15M expiry. Does this look like a nice strategy? I just want to get the idea of what a curve should look like and whether the OOS is applied correctly.
Note that under the generator I see a different strategy, the one that is placed at the very bottom of the collection tab. Despite the fact that I clicked on the strategy at the very top (the one in the picture) the Generator tab still shows the worse strategy (as seen in the 'generator picture'). Not sure why this happens?
'generator picture'
The Generator pushes strategies to the Collection.
When you click a strategy in the Collection, it goes to the Editor for review and export.
There is no way to send a strategy to the Generator. It doesn't accept strategies. It only generates new strategies.
All strategies you generate merge in the Collection.
The Collection may contain old strategies from previous generations.
It is possible to run the Generator with different conditions and the generated strategies to have lower profit than what the Collection contains.
The Generator shows the best strategy from the current run.
The Collection contains strategies from all runs (even from previous BO Tester session).
If you want to collect strategies from the last Generator run, delete the collected strategies before running the Generator.
All strategies you generate merge in the Collection.
Yes I have noticed. Are you thinking of allowing the testing of a selection of pairs combined, which will be displayed under one graph? Also, option to download the graph? Considering there is one graph per pair, per time frame, per expiry that would be very helpful in order to compare data from different strategies.
Also, I am trying to understand how the "The bar open below Lower Band after opening above it" works, as in the picture below it doesn't seem that it works the way it's stated.
The indicators are calculated at Close price and receive their final value at Bar Close.
MetaTrader forms (fixes) Close price and shows next bar Open price only when there is a tick from a new bar (shame for MetaTrader because it can take long time on slow markets).
At the moment MetaTrader gives the first tick (Bar Open) the expert determines the final values of the indicators and compare them with the new price. At that moment we have fixed Close price and Open price. However, the Close price and the corresponding indicator values are part of the previous bar. As a result of that, the Current price (Bar Open) is compared with the indicator value from the previous bar. So, when you look at the chart, you have to compare the bar Open price with the indicator value from the previous bar. This is the only safe way to do backtest. Unfortunately not many traders understand that.
This topic is discussed hundred of times in the forum and is described in the documentations. Please search for "Use Previous Bar Value". See also the FSB Pro True Charts.
The main purpose of our software is to provide correct backtest. We do that from 10 years.
Very interesting.. Thank you
Forex Software → Binary Options Tester → Trading EA Real Results?
Powered by PunBB, supported by Informer Technologies, Inc.