Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 21
 5 Attachment(s)    

Thread: Ntk 07

  1. #11
    Junior Member
    Join Date
    Jan 2009
    Location
    Denver CO
    Posts
    4

    Default

    Very good point Roodstaart. The broker is Interbank FX. I use them for the optimizations, because this is the broker that I have both real and demo accounts with, so I hope that if I use the optimized values that they will give me relatively good results in real trading. Not that I think this is necessarily true!

    By the way, are you in the Netherlands?

    Take care

    Kelvin

  2. #12

    Default

    Quote Originally Posted by funyoo View Post
    Here is a backtesting done with a new EA : NTK 07.

    Time filter (chas) is GMT+2.

    EU M5. On one year.

    Total net profit : 820.43%
    RDD : 28.93%
    Looks a little different to my backtesting results. Any clues as to why that is?

    Attached Files Attached Files

  3. #13
    Junior Member
    Join Date
    Jan 2009
    Posts
    19

    Default

    Quote Originally Posted by kelvinrice View Post
    Thank you Funyoo for your prompt reply. Here are the optimization results for NTK07.

    I have 5 virtual servers with fast processors. Please let me know if there are other EA's that you would like me to optimize? If so, please let me have the .set files for the optimization as well as the EA and required indicators, if necessary.

    Thanks

    Kelvin

    I hope that the file uploaded. I did not see this on the draft of the e-mail??
    This EA looks promising, i will do some tests with Alpari to see how it performs.
    Last edited by paromi; 02-21-2009 at 09:24.

  4. #14
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by druss View Post
    Looks a little different to my backtesting results. Any clues as to why that is?
    Hi druss,

    You have to do your own optimization, each broker has a different tick data.

  5. #15

    Default

    Quote Originally Posted by funyoo View Post
    Hi druss,

    You have to do your own optimization, each broker has a different tick data.
    So if that is the case then would it be true that running backtest is a complete waste of time if each broker has different data?

    I have purposely chosen Alpari as the demo platform to run the backtest because I would of thought that Alpari Data if any data was reliable then their data would be the most reliable.

    What is the point of running backtests on unreliable data?

  6. #16
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by druss View Post
    So if that is the case then would it be true that running backtest is a complete waste of time if each broker has different data?

    I have purposely chosen Alpari as the demo platform to run the backtest because I would of thought that Alpari Data if any data was reliable then their data would be the most reliable.

    What is the point of running backtests on unreliable data?
    It is not a question of reliable or unreliable data, it's just that each broker has his own tick data (different spreads, GMT offset, etc...) generating different results.

    The backtesting result is only here to show us a potential. You know this famous rule : Past performance is no guarantee of future results.

    You have still then to do forward tests in demo.

  7. #17

    Default

    Quote Originally Posted by funyoo View Post
    It is not a question of reliable or unreliable data, it's just that each broker has his own tick data (different spreads, GMT offset, etc...) generating different results.

    The backtesting result is only here to show us a potential. You know this famous rule : Past performance is no guarantee of future results.

    You have still then to do forward tests in demo.
    Let me get this right so I'm on the same page.

    Assuming we use the same EA on two different brokers each have the same spreads we will still get different results on backtest because tick data maybe different?

    Should the open close high low for the 1min candles be roughly the same regardless of which broker it is? Will the open high low close for the 5min/10min etc be roughly the same?

  8. #18
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by druss View Post
    Let me get this right so I'm on the same page.

    Assuming we use the same EA on two different brokers each have the same spreads we will still get different results on backtest because tick data maybe different?
    That's right. Try to test a same EA on different brokers, you will see what happens.

    Quote Originally Posted by druss View Post
    Should the open close high low for the 1min candles be roughly the same regardless of which broker it is? Will the open high low close for the 5min/10min etc be roughly the same?
    Not necessarily. What you should do is to do optimization yourself on the broker you want to use for your forward tests (demo and real).

  9. #19

    Default

    What length of previous data do you feel running an optimisation over gives the most reliable results for a potentially good EA?

    Or does it not matter that much?

  10. #20
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by druss View Post
    What length of previous data do you feel running an optimisation over gives the most reliable results for a potentially good EA?

    Or does it not matter that much?
    It's up to you.

    Personally, I do it on one year. I can't invest more time on backtesting.

    For certain people, an optimization over one week and every week is enough.

    For other, it is at least on 20 years, etc...

Tags for this Thread

100, alpari, backtest, backtesting, breakout, broker, brokers, change, demo, eurusd, expert, experts, forex, high, how to, indicators, low, mql, mql4, mt4, profit, profit factor, profitable, real, test, thank you, time, trading, trailingstop, trend

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •