Topic: Testing a Portfolio on an account?

Hi,

i like to know how you guys do that. I have a demo account and want to test my strategies. Is it better to have there different pairs on same timeframe or same pairs on different time frames?

Or only one pair all strategies with different timeframes.

Re: Testing a Portfolio on an account?

Hi,

i like to know how you guys do that. I have a demo account and want to test my strategies. Is it better to have there different pairs on same timeframe or same pairs on different time frames?

Or only one pair all strategies with different timeframes.

In the youtube visds posted for FXSBpro Popov uses multiple pairs on a single time frame. The choice is yours, IMO I think you should keep it simple. Trade what you can handle and process at the end of the day.

Re: Testing a Portfolio on an account?

Hello,

how can it be that backtests of some strategies are positive but then they completely fail on real accounts?

Best regards

Re: Testing a Portfolio on an account?

This is normally because of over-optimization of the strategy.

Monte Carlo and OOS can detect many of the potentially problematic strategies. However there is no guarantee. Every dollar you win is taken from another trader.

We also trade multiple strategies on different market in order to reduce the risk.

Re: Testing a Portfolio on an account?

Ok.
So if I do not optimize the strategy at all, it should perform better?

Re: Testing a Portfolio on an account?

I suggest......... very few indicators in a strategy, do optimize, do Monte Carlo, do Multi Markets thoroughly and carefully and do 'Walk Forward'.

Walk Forward will show you weaknesses quickly, it is well worth the time and effort and it is really easy in FSBPro. When doing it, examine the journal for each period to look for trouble spots.

I suggest using ony a couple indicators as that is where problems can be avoided. it may take more time to get a strategy but will save you from trouble when trading.

I also suggest that you stay away from the lower time frames and I know that few will agree with me hahaha

The data points are much clearer on the higher time frames... I am meaning 4 hour and daily.

Set high standards in your filtering, work with only the best..

I have been 'wahed and rinsed' by the market many times, learned the hard way.  lol

Re: Testing a Portfolio on an account?

Blaiserboy wrote:

I suggest......... very few indicators in a strategy, do optimize, do Monte Carlo, do Multi Markets thoroughly and carefully and do 'Walk Forward'.

Walk Forward will show you weaknesses quickly, it is well worth the time and effort and it is really easy in FSBPro. When doing it, examine the journal for each period to look for trouble spots.

I suggest using ony a couple indicators as that is where problems can be avoided. it may take more time to get a strategy but will save you from trouble when trading.

I also suggest that you stay away from the lower time frames and I know that few will agree with me hahaha

The data points are much clearer on the higher time frames... I am meaning 4 hour and daily.

Set high standards in your filtering, work with only the best..

I have been 'wahed and rinsed' by the market many times, learned the hard way.  lol

Washed and rinsed even with the EA Studio?

Re: Testing a Portfolio on an account?

Just starting to use EA Studio....

Re: Testing a Portfolio on an account?

Ok. Understand now.