Well.
I will answer on all the question even if you all did not ask.

Q: What is Delta?
A: It is MA channel. Attach any MA (EMA or any) indicator on the chart. And set "Levels" to +7 and -7 (or what ever). So you will see the MA channel with Delta=7.

Q: What the entry conditions was programmed in this EA concerning Delta?
A: When SignalEMA is crossing ther upper border of the channel (MainEMA+Delta) - buy. If SignalEMA is crossing lower border of the channel (MainEMA-Delta) sell.

Q: Why we need this Delta?
A: No idea. I wanted to have some filter. Anyway we may set Delta to zero (0) and EA will work without this Delta. Probable we don't need any Delta. It was easy to code so I coded it.

Q: There are 2 versions of EA. What is the difference? Which EA to use?
A: The first version is more corresponding with the main theory. It is ready to test. The version #2 is still under development.

Q: Why we have different settings for version #1 compare with the theory? It should be 8, 21 and 55 EMA but it is not in pe-set files ...
A: I optimized the settings and backtested it back to 2001 with 90% modelling quality to find the profitable settings. Profitable in long way run. The settings (pre-set files) and EA you may find in this thread somewhere - look at my previous posts. So it may be profitable if we are using this version #1 during the 1 or more year.

Q: Why we have version #2?
A: many people said that version #1 is not trading often (it is true). And many people do not want to wait 1 year for example to have good profit (future possible profit!). Besides as we see from backtesting results of version #1 so it is profitable for some years only. It is not profitable every year. So it is necessary to use this version during the 2 or 3 years. Big distance from now . That is why the other members suggested to develop the second version.

Q: What was changed in EA to be a version #2?
A: It was some suggestion about re-entry. For example, we have uptend concerning MainEMA. EA took the order (buy). But suddenly the order was closed by SignalEMA-CloseEMA crossing. But the main trend may be continuing without us! So it was a suggestion to enter once again. I asked Igorad to help me with version #2 but seems it was many bugs in coding and I corrected it in version #2.1. But this version 2 (or 2.1) still need to develop more. The version #2 is still under develpment.

Q: Which version is more risky to use?
A: Version 2 of course. This version is still under development but it is much more risky than version #1. Why? Look at the accrete's post above. It was the explanation: EA may take some good pips according to the first entry condition and close the order. But downtrend is continuing and the EA may open new sell according the second condition. To get more pips. But as i said the downtrend was continuing but it was started long time ago and it may be reversal soon. So if we have good continuing trend so the second condition is helping us a lot. But if not? If trend just started in the morning and finishing in the evening and we re-entered (the other sell) in the end of the trend (reversal)? Yes it will be losses.

Q: What is the main problem with this kind of EAs?
A: This EA is EMA-crossing kind of EA. There are two programming approach concerning EAs based on EMA crossing:
- "Russian" approach: no one EMA crossing EA will be profitable for the long way run because market is changing and it is necessary to change the settings of the EAs together with the market. It is very risky EAs. MA indicator is the most bad/unreliable indicator. This approach came from some old american books/magazines.
- "American" approach: EAs based on EMA crossing are very profitable. MA indicator is the most good/reliable indicator. We need just optimize the settings of EAs once per week/month for the next month. This approach came from some old american books/magazines as well.
- new approach: EAs based on MA crossing should adjust the settings by themselves without any manual inervention/backtesting. It should be fully automatically (self-adjusted EMA crossing EAs). We don't need to optimize the settings of this EA every wek/month to find the settings for the next week/month. EA should do it automaticall and it is possible to code. Codersguru started some thread about it (don't remember the name of this thread now).

Q: so what to do now with the second version?
A: develop it more. Need some ideas.