Backtesting a Scalper

[gazuz]

Hello everyone,

this is also a bit directed to scoprion, im trying to change the CCI I've ben working on into a scalper where there CCI opens the trade, and then the rest plays out with a trailing stop or take profit (not sure right now). BUT I don't know how I am suppose to test it since there is no tick data. The rpice, for all I know, could be going up and down the candel like crazy but I can't tell that from the candle. I'd like to know how it is possible to backtest with tickdata like scorpion did.

Thanks

[scorpion]

Yes, candle-based data tell you so little about the ticks movement, so that's why backtesting with tick data is highly essential. As you're not a programmer, you have limited choice on tick backtesting and improving your own systems.

For instance, you may backtest by using backtesting suite such as WealthLab at www.wealth-lab.com, or you may ask someone to assist you; Fabrefactum.com will backtest your system at a cost. I'm thinking about helping you out to backtest the systems, but i'm afraid that i'm now a bit busy in finishing fxfisherman (just a little bit to go, we'll make it).

I have a few years worth of EURUSD ticks on fxfisherman.com. PM me if you wanna download it.

[scorpion]

If you wanna play big in forex, you should learn to backtest your own strategy, coz the knowledge will save you time, effort and capital. I'm sure that man smart like you can learn to do that in no time. Check out WealthLab (not sure if it can backtest ticks data).

Good luck to you,

[gazuz]

Thanks scorpion,

I will wait until the fxfisherman bot will be done and then my strategy will hopefully be an add-on when chosing which scalper to use. This is all yours, I don't even know if it even trades right with this script. Just to let you know use CCI on a 15 minute chart with settings at ~60 I believe that is the one that I used when I found out. Also no stop loss, no take profit, just a trailing stop. BUT the whole point of the backtesting is to find out which trailing stop would work better, finding which one would make more money without decreasing too dramatically the number of wins to losses

Regards,
Gazuz

[Ravique]

Originally Posted by scorpion

(just a little bit to go, we'll make it).

That is inspiring. =)

And thank you a lot for the Welath-Lab link, looks very useful.

[scorpion]

Originally Posted by gazuz

Just to let you know use CCI on a 15 minute chart with settings at ~60 I believe that is the one that I used when I found out. Also no stop loss, no take profit, just a trailing stop. BUT the whole point of the backtesting is to find out which trailing stop would work better, finding which one would make more money without decreasing too dramatically the number of wins to losses

Also backtest the system without trailing stop. Just want to see which one is better. Sometime, trailing stop closes the trades too early, for the fact that ticks' movement is non-linear, which go up and down frequently in each candle.

Again, check Wealth-Lab out. You'll need to backtest a lot more in the future as I've experient. At first, I backtested the systems just for documentation only, and then I did about 300 more backtestings and change the code just to tweak my systems to the right settings. I guess you're gonna need it somedays.

[scorpion]

Ravique, you're welcome.

[gazuz]

I went on wealth-lab quickly and found it by stocks... Am I missing something or where is it hidden?

[Ravique]

I took a quick look , and you can upload data there from different sources, i saw that you can upload ASCII files , which you can export from metatrader in csv format.

[gazuz]

Originally Posted by Ravique

I took a quick look , and you can upload data there from different sources, i saw that you can upload ASCII files , which you can export from metatrader in csv format.

Oh alright, and scorpion, I PM'ed you