Download MetaTrader 5
To add comments, please log in or register
Manage agents in your profile. It is easy!
Sergey Golubev
Moderator
53278
Sergey Golubev 2013.01.15 06:32  

I have the other example - with free EA for now (I can upload it with forward testing results). I forward tested it since February last year till end of December last year with annual forecasting grow as 87% for 5K initial deposit. But the dilemma is the follolwing: some people prefer to have 80% as a monthly/weekly forecasting grow :)
So, it depends on people's expectation and illusions. 
I remember some talking on Alpari forum many years ago when some trader told that he is trading on weekly timeframe (W1) and his system is consistently profitable weekly so ... everything is possible.

Jozef Bartek
243
Jozef Bartek 2013.01.26 13:04  

Hey newdigital,

I was deciding between MT4 and MT5 last September and as an experienced developer I was naturally inclined in choosing object oriented MQL5 over MQL4 (simplified C). However, I still went for MT4 because:

  • I don't like the position netting and FIFO in MT5  (I live in the UK)
  • none of the big brokers were planning Spread Betting on MT5 in a foreseeable future
  • MT4 code base is huge and there are many indicators/EAs around whose code I can use as an inspiration/learning material

Now after 5 months of coding in MQL4 I  don't regret my decision. Yes, I cried many times why MQL4 isn't syntactically at least on par with ANSI-C (yes, I can import DLLs, but ...) but on the other hand, the amount of already available code and knowledge on the Internet (and on this site of course :-)) is overwhelming and proved invaluable to me many times. So no plan/reason for me to move to MT5. I read somewhere an opinion that perhaps Metaquotes will realize their mistake, create MT6 with best features from MT4 and MT5 and we all will happily migrate. Let's see.

My other point to you newdigital after reading all posts in this thread is following: do yourself a favor and learn coding. Since you moderated a big forum for years, you know too well that you can come across very varied quality of posts on different forums and many times there are contradictory opinions about how stuff works. It happened to me few times and I had to sit down, write a test code and see for myself where the truth is. Specifically I'm talking about backtesting and MTF. It seems to me you simply rely on what you heard from others. They might have been right back then. But MT4 keeps evolving with each new revision and many things are being improved/bugs fixed. You are doing yourself a disservice if you keep completely rejecting backtesting results in MT4 as not trustworthy in general because if you could prove to yourself that backtesting results are actually useful and sufficiently accurate, the EAs you're interested in could be tested much faster to see if EA has any potential rather than wait for some results for few months on demo (which don't prove much IMO). BTW, my EA uses 5 different TFs (M1, M15, H1, H4, D1) and indicators for various bits of its logic. Main TF is H1 but for purpose of being able to test even visually the bits of logic which use other TFs, I  hard-coded TFs into EA and you guess what? If I run the same EA with the same setting on any TF, I'm getting identical backtest results so all TFs are modelled the same way regardless what TF you're testing the EA on.


J.

Sergey Golubev
Moderator
53278
Sergey Golubev 2013.03.23 17:21  

Hi fxjozef,

Thanks for your comment.

I am not rejecting backtesting results. But - according to my experience - no any backtesting results can be a proof for buyer that EA of seller is profitable - you know that there are a lot of sellers who are selling EAs based on backtesting results ONLY ... for my understanding - this is 'forex scam'. Why? Because we do not know what is inside the code ...  

As to MT5 so I like everything which is new one. Because I am NEWdigital (old by age by new by feeling) :)  

paco246y8
5
paco246y8 2014.06.12 07:52  
newdigital:

Hi fxjozef,

Thanks for your comment.

I am not rejecting backtesting results. But - according to my experience - no any backtesting results can be a proof for buyer that EA of seller is profitable - you know that there are a lot of sellers who are selling EAs based on backtesting results ONLY ... for my understanding - this is 'forex scam'. Why? Because we do not know what is inside the code ...  

As to MT5 so I like everything which is new one. Because I am NEWdigital (old by age by new by feeling) :)  


Hi newdigital,

I have recently migrated to MT5, and I can code everything, MT4, MT5, C++, java, or whatever is convenient to get something done. I read your complains at the beginning of this thread but I didn't read the whole thread, sorry for that.

People who find the system of their dreams abandon forums, because they have what they need for their lives. Great coders also go away. Why spend your time on screen when your EA does everything?

So for now, while the opportunity comes, I code free and share, just like anybody else. If you are still in need, just ask.

Happy trading.

Paco.

daniel
62
daniel 2014.12.03 21:18  
newdigital:

I have the other example - with free EA for now (I can upload it with forward testing results). I forward tested it since February last year till end of December last year with annual forecasting grow as 87% for 5K initial deposit. But the dilemma is the follolwing: some people prefer to have 80% as a monthly/weekly forecasting grow :)
So, it depends on people's expectation and illusions. 
I remember some talking on Alpari forum many years ago when some trader told that he is trading on weekly timeframe (W1) and his system is consistently profitable weekly so ... everything is possible.

 

Weeekly!!! thats amazing, its like the one you nee to you'r little childrens, for when they got out of school and ask you for a car for the ladies haha

 

by the way, i am stucked on an optimization for an Ea i created based on some indicator, takes 500 - 1000 hours to optimize, i was told the following:

" multi timeframe model and I test everything in open prices using M1 data. It's the fastest and most realible model i found till now." do you know where i can get this??? or someway how ... is there a program or so?? don't want to ask to my source because is the programmer i pay for eas to code ... :D this is finally the best result i have for gbpusd 4 months, looks like those oohhhh holy grail, but its not, i have just pusshed it to the limit..., normally makes not more than 1.7000 every 4 months (picture)

 

thanks, see you around 

/ /1...56789101112
To add comments, please log in or register