Topic: A few questions about practical use of the software.


I now have an FSB Pro license, and had a bit of time to play with the software. Please let me know whether you can answer the following questions, and anything else I should know about:

1) I noticed that the default currency spread is 20 points for all pairs. Do I calculate the average spread from my broker by hand, figuring in periods when spread goes up (Australian Session, etc.)?

2) Can I/should I back-test on data specifically from my broker (Oanda)? If so, how would I do this?

3) If discrepancies arise between the native FSB Pro back-test and the MT4 Strategy Tester, how is this typically addressed?

4) If discrepancies arise between FSB Pro/MT4 back-test profits and the profits of a live money account, how is this typically addressed?

5) What percent of seemingly good strategies fail when brought over to a live account? Why does this happen, beyond the occasional stochastic drawdown?

6) From my reading, it seems best to create a portfolio of EAs, and run that to exploit market pulses. Is that possible with FSB Pro, or do we use EA Studio?

I've already found a number of interesting strategies with the software, and I'll continue to tinker to see how far this can go.

Thank you.

Re: A few questions about practical use of the software.

1. Take the most reasonable amount. There's a custom indi called Spread Limiter or something similar, you can use it to prevent high spread trading as it inevitably will cost you the results. Even better is to use time limits and trade from Sidney open to NY close.

2. Certainly use your broker's data. Save it from your broker's terminal (Save As -> FSB's data folder). You can introduce yourself to other options as well ->

3. My opinion is that testing a strategy twice does not improve its abilities and future profitability. Secondly, some strats will never show similar results because of MT's bugs in its backtester. Usually they match but it is a fiddle to have correct settings and starting points and so on and so forth, it is a waste of time. Therefore the most vital thing a trader must do is to verify that his/her strategy behaves similarly in a live trading environment compared to a backtest. That is one must compare live trading signals to a following backtest on the same data set.

4. You will not have 100% exactness in profits due to spread variation, slippage, broker tricks etc. It is the trading signal matches which bear the importance.

5. The failure rate can be 10-90%, depends on many factors. There are a lot of books on this subject and web resources, a whole science.

6. For Pro visit ->

Re: A few questions about practical use of the software.

Thank you for your prompt response. Unfortunately, getting broker information from Oanda is a hassle (you have to pay every single time you need updated data) so I'll probably pay a couple of times, first, to see whether any real discrepancies exist on back and future data, then keep testing to make sure nothing substantial changes.

Also, I'm trying to figure out the real differences between EA Studio and FSB. I understand one is lighter, they generate EAs differently, and their use of indicators is different, but what is the core reason for EA Studio if FSB offers more options to play with and customize?

Re: A few questions about practical use of the software.

The use of indis is the same, just Studio uses native mql indis mostly. Secondly Studio is browser based, it is more simplified than Pro and therefore a lot faster. Thirdly its capabilities and the scope of use is somewhat larger, especially considering portfolio testing and trading. So if a user is into data mining approach, then Studio is a much better platform to do so.

Re: A few questions about practical use of the software.

Are there any downsides for being a browser-based solution by the way?

Here are my favorite fat burners for belly fat to read more about.

Re: A few questions about practical use of the software.

It depends what to consider as a downside smile One could argue that a stand-alone program would be better - no browser and internet connection dependability, easier setup of numerous instances, I can't tell for certain but maybe a slight performance advantage.