Additional tests to challenge an EA | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

What kind of rules do you have for opening additional positions once in profit? 2 replies

Crosshairs with 2 additional Lines +/- X pips away 1 reply

How to draw an additional line? 3 replies

Additional Vegas 4h filter ideas that seem to work(+100% pips), need input please! :) 16 replies

Additional tests to challenge an EA

  • #1
  • First Post: May 22, 2013 8:49am May 22, 2013 8:49am
Good Day to all brilliant brains here at FF,

I have wrote an EA which gives very good results and which is already trading live since 1 month.

It's behavior is identical as the one observed during my tests.

On the tester, I tested it from 2008 to today and it survived, I tested it also starting the 1, 10 and 20 of each month from jan 2008 to today and it survived.

So my question is, which additional tests could I do, to test the solidity of that EA.

Any proposal or idea is welcome.

Thanks.
Simplicity is the Ultimate Sophistication.
  • #2
  • May 23, 2013 3:55am May 23, 2013 3:55am
You seem to use some kind of martingaling ... Right ?

If so, you should be very carefull. All kind of martingaling schemes have the same result on Live. Soon or later you will loose all the money in your trading account.
MT4/MT5 EAs/Indicators/Alerts coder
  • #3
  • May 23, 2013 4:08am May 23, 2013 4:08am
you can attached your test results to make things interesting.
the results give more insight into the EA and would allow members recommend additional tests based on the EA's behavior. better yet, you could give a brief introduction.
It is possible to make large sums trading, small losses, large wins
  • #4
  • May 23, 2013 4:46am May 23, 2013 4:46am
Quoting FerruFx
You seem to use some kind of martingaling ... Right ? If so, you should be very carefull. All kind of martingaling schemes have the same result on Live. Soon or later you will loose all the money in your trading account.
Thanks FerruFx, I know that martingale schemes are prone to failure but mine is some kind biased to remove some risk from it, I still have room to remove more risk from it but this also removes profits so I will leave it as is, acceptable risk for subtantial profits (in the long run).

Omadivi, I could publish the results, but don't ask me to publish the EA, it's a year+ work. and it's live since one month only. Still interested ?
Simplicity is the Ultimate Sophistication.
  • #5
  • May 23, 2013 4:49am May 23, 2013 4:49am
Quoting broketrader
... but mine is some kind biased to remove some risk from it...
Doesn't matter. No martingale (even "biased"!) survives in the long term. There's a lot to read about this on FF ...
MT4/MT5 EAs/Indicators/Alerts coder
  • #6
  • May 23, 2013 4:52am May 23, 2013 4:52am
If a specific martingale was found to survive in long term ... all casinos should be already ruined.
MT4/MT5 EAs/Indicators/Alerts coder
  • #7
  • May 23, 2013 4:53am May 23, 2013 4:53am
Quoting FerruFx
{quote} Doesn't matter. No martingale (even "biased"!) survives in the long term. There's a lot to read about this on FF ...
Doesn't matter, I wasn't expecting another response, I need to give it a try anyway... And doesn't matter if my account is gone in 3, 6 months or 2 years, it's the risk that I accept to take and as you probably know, in Forex, no risk = no profits.
Simplicity is the Ultimate Sophistication.
  • #8
  • May 23, 2013 4:58am May 23, 2013 4:58am
Quoting FerruFx
If a specific martingale was found to survive in long term ... all casinos should be already ruined.
If you have very deep pockets, you can win with a martingale in casinos this is exactly why they introduced betting limits !

http://en.wikipedia.org/wiki/Martingale_%28betting_system%29
Simplicity is the Ultimate Sophistication.
  • #9
  • May 23, 2013 5:04am May 23, 2013 5:04am
Quoting broketrader
And doesn't matter if my account is gone in 3, 6 months or 2 years, it's the risk that I accept to take
Yes exactly. My post was just a warning. There's so many newbies here (I don't know about you) and they must understand the risk of this kind of strategy before jump into it.
MT4/MT5 EAs/Indicators/Alerts coder
  • #10
  • May 23, 2013 5:05am May 23, 2013 5:05am
Quoting broketrader
{quote} If you have very deep pockets, you can win with a martingale in casinos this is exactly why they introduced betting limits ! http://en.wikipedia.org/wiki/Marting...ting_system%29
Limits apart ... a martingale survive ONLY with infinite funds.
MT4/MT5 EAs/Indicators/Alerts coder
  • #11
  • May 23, 2013 5:12am May 23, 2013 5:12am
Quoting FerruFx
{quote} Yes exactly. My post was just a warning. There's so many newbies here (I don't know about you) and they must understand the risk of this kind of strategy before jump into it.
Totally agree with you, and I DON'T ADVISE ANYBODY IN ADOPTING THIS KIND OF STRATEGY, I was just asking for ways to test and challenge an EA, I saw that you are a programmer, I'm also a programmer but not in the finance world. I have tested my EA and was just looking for other test ideas to see if mine is rock solid or not.

Yourself, when you write an EA, what kind of tests you do on it ? (only if you want to share off course)
Simplicity is the Ultimate Sophistication.
  • #12
  • May 23, 2013 5:16am May 23, 2013 5:16am
Quoting FerruFx
{quote} Limits apart ... a martingale survive ONLY with infinite funds.
Yes, but there are ways to partially around the limitation of not having infinite funds, one of them is starting small, the remaining of risk, is the risk you have to take, there is always a risk you need to take...
Simplicity is the Ultimate Sophistication.
  • #13
  • May 23, 2013 5:20am May 23, 2013 5:20am
Quoting broketrader
Yourself, when you write an EA, what kind of tests you do on it ? (only if you want to share off course)
If your test is to check its profitability, only real trading on a live account will give you the real picture.

Backtest for me is only to confirm that the EA is trading as per the defined rules ... nothing else.

So many examples showing great backtests ... then poor results on live. The opposite is also true ... sometimes!

Wish you all the best in this quest. Be safe ...
MT4/MT5 EAs/Indicators/Alerts coder
  • #14
  • May 23, 2013 5:21am May 23, 2013 5:21am
Quoting broketrader
{quote} Yes, but there are ways to partially around the limitation of not having infinite funds, one of them is starting small, the remaining of risk, is the risk you have to take, there is always a risk you need to take...
Yes but sadly the result is always the same.
MT4/MT5 EAs/Indicators/Alerts coder
  • #15
  • May 23, 2013 5:24am May 23, 2013 5:24am
Quoting FerruFx
{quote} If your test is to check its profitability, only real trading on a live account will give you the real picture. Backtest for me is only to confirm that the EA is trading as per the defined rules ... nothing else. So many examples showing great backtests ... then poor results on live. The opposite is also true ... sometimes! Wish you all the best in this quest. Be safe ...
Knowing the dangers of martingaling, my tests were first to check it's viability before it's profitability, when it got safe I went on a demo for three months and after seeing that it was behaving in demo like in tests, I went live one month ago.

Thank you. Good luck to you too.
Simplicity is the Ultimate Sophistication.
  • #16
  • May 23, 2013 5:28am May 23, 2013 5:28am
Quoting broketrader
{quote} Knowing the dangers of martingaling, my tests were first to check it's viability before it's profitability, when it got safe I went on a demo for three months and after seeing that it was behaving in demo like in tests, I went live one month ago. Thank you. Good luck to you too.
Great!
MT4/MT5 EAs/Indicators/Alerts coder
  • #17
  • May 23, 2013 6:57am May 23, 2013 6:57am
congrats. you are doing well. the key is controlling your draw down. i think your drawdowns are relatively large though, you should look into it.
It is possible to make large sums trading, small losses, large wins
  • #18
  • Last Post: May 23, 2013 7:39am May 23, 2013 7:39am
Quoting omadivi
congrats. you are doing well. the key is controlling your draw down. i think your drawdowns are relatively large though, you should look into it.
It's a bit early to say that I'm doing well, this baby is only 1 month old, so until it has not run for 1 or 2 years I will not say that I'm doing well. Maybe I will never reach that point...

The drawdowns are there by design, I mean it's the nature of martingale to have drawdowns so I cannot look into them or control them, I already did the maximum to reduce them during the development and test phase, now I can just watch the EA run and control that it does what it's expected to do.

Thank you for your comments.
Simplicity is the Ultimate Sophistication.
0 traders viewing now