i only trust on forward testing
you should compair that result with forward test to ensure that
how much you can trust on it
Hello,
Last few days I've written and did A LOT of optimization for my first EA which is based on a very simple strategy (haven't seen it before but it could be used already). Backtesting seems to give good results, see attached photos.
My question is how much can you trust these results? Any comments on them? I am new to Forex and although I try to read and learn as much as I can there is still a lot of things I don't understand
Strategy was tested on
EUR/USD
Daily
2006.01.01 - 2006.08.25
Initial deposit: 1000 USD
Regards,
"c4n"it should be 90%, not 45.
it should be 90%, not 45.
Thanks, that's exactly the input I've been looking for. Will Google a bit around and see if I can get the accuracy up.
Thanks, that's exactly the input I've been looking for. Will Google a bit around and see if I can get the accuracy up.
Hi, better forward test for next week, that's 100% accurate. If you really wants to use it on your real account, my suggestion is try it on real time until you get enough confident in it.
Just don't let yourself get too excited. You're one step forward, just keep in line.
Rgrds
No harm in backtesting and optimizing.
But do it the right way.
For example test/optimize only on data from 2005.
Don't change any setting and test, NOT optimize, on 2006 data.
If your results still are great you are on the right track.
Thanks for the input.
I don't want to rush anything and I certainly don't consider this a "get rich quick" scheme so I will go patiently step by step. For now my goal is to try and develop a strategy that will be profitable, even if only by a small percent.
I will keep the EA active at least for a month to see how it goes. Now it's been up for about 12 hours and has 4 wins and 1 loss, total of +30 pips. I still have to optimize the stop loss level but so far so good
May I ask what is the reason to optimize for 2005 and not 2006?
Thanks
No special reason. Optimize on 2001 and test on 2003 is fine too.
The danger of optimizing is that the settings only work on the data you used.
If you optimize on year X and test on year Y you have proof that the settings 'always' work and not just on the backted data.
(Over) optimizing, also called curve fitting, can give you +100% on the data you optimized on and -100% on all other data.
Signs of over optimization are often that there are a lot of loosers and the profit comes from one big killer profit.
No special reason. Optimize on 2001 and test on 2003 is fine too.
The danger of optimizing is that the settings only work on the data you used.
If you optimize on year X and test on year Y you have proof that the settings 'always' work and not just on the backted data.
(Over) optimizing, also called curve fitting, can give you +100% on the data you optimized on and -100% on all other data.
Signs of over optimization are often that there are a lot of loosers and the profit comes from one big killer profit.I think optimizing for 2005 and 2006 maybe better, at least you test in on latest market behaviour, it will give you a bit confident but nothing is compared to live test. Good luck then.
No special reason. Optimize on 2001 and test on 2003 is fine too.
The danger of optimizing is that the settings only work on the data you used.
If you optimize on year X and test on year Y you have proof that the settings 'always' work and not just on the backted data.
(Over) optimizing, also called curve fitting, can give you +100% on the data you optimized on and -100% on all other data.
Signs of over optimization are often that there are a lot of loosers and the profit comes from one big killer profit.Yes you are right.
Yep, that makes sense, thanks for the heads up regarding over-optimizing.
The strategy I am trying to develop involves taking small profits and (hopefully) a nice percentage of wins versus losses. If you check the screenshot you will see it had 85% profitable trades in 2006. If I run it on 2005 data I get about 75% wins. The forward testing I am doing has 80% wins at the moment, but like said I will let it run at least for a month, then see if I can tweak it up, test again and then decide if I will spend a little on a real test.
I still need to get the MQ up, so far I have it up to little less than 70%.
Oh, and I was wondering today for about 2 hours why my MetaTrader isn't working (prices don't change) only to figure out that it's weekend and forex is closed on weekends, he he
![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hello,
Last few days I've written and did A LOT of optimization for my first EA which is based on a very simple strategy (haven't seen it before but it could be used already). Backtesting seems to give good results, see attached photos.
My question is how much can you trust these results? Any comments on them? I am new to Forex and although I try to read and learn as much as I can there is still a lot of things I don't understand![](https://c.mql5.com/forextsd/smiles/wink.png)
Strategy was tested on
EUR/USD
Daily
2006.01.01 - 2006.08.25
Initial deposit: 1000 USD
Regards,
"c4n"