Topic: Monte Carlo

I use this tool every day, many times. It is a major feature of FSBPro and I really appreciate it.

I do not have a math background so I have done some reading to learn a bit about what this tool does.

Here are some resources for people who may be interested:



http://www.palisade.com/risk/monte_carlo_simulation.asp

http://statweb.stanford.edu/~owen/mc/

https://en.wikipedia.org/wiki/Monte_Carlo_method

http://www.investopedia.com/terms/m/montecarlosimulation.asp?layout=orig&adtest=5noninfinite

http://www.goldsim.com/Web/Introduction/Probabilistic/MonteCarlo/

http://www.columbia.edu/~mh2078/MCS04/MCS_overview.pdf

http://blog.minitab.com/blog/understanding-statistics/monte-carlo-is-not-as-difficult-as-you-think

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2924739/

Enjoy!

Re: Monte Carlo

Hello everyone,
elaboration of strategies it gives much importance to the analysis Monte Carlo.
But always I get to a zero point.

https://s32.postimg.org/75ymiv9z5/image.png

In the first photograph, there is a strategy that would not be bad, for different values.
Then with Monte Carlo, everything vanishes because of Randomly Close Position.

https://s31.postimg.org/b9c50o9uv/image.png
https://s32.postimg.org/8bnckmu41/2_2.png

The questions are:
• what methods used to try to improve the Monte Carlo?
• there is a way to force the slots of FBP only seek solutions for closing, without redoing the opening strategy?
I tried with the blocking of the opening slots, but I do not see improvements, it seems that FBP can not concentare only improve a part of the trade (closing) but the whole. These tests I've done with hundreds of different strategies, with different indicators opening.

Re: Monte Carlo

Myself..... because FSBPro generates so many strategies so quickly...... I disregard the poor results in Monte Carlo and move on to the next one.

Re: Monte Carlo

Freesby wrote:

Hello everyone,
elaboration of strategies it gives much importance to the analysis Monte Carlo.
But always I get to a zero point.

https://s32.postimg.org/75ymiv9z5/image.png

In the first photograph, there is a strategy that would not be bad, for different values.
Then with Monte Carlo, everything vanishes because of Randomly Close Position.

https://s31.postimg.org/b9c50o9uv/image.png
https://s32.postimg.org/8bnckmu41/2_2.png

The questions are:
• what methods used to try to improve the Monte Carlo?
• there is a way to force the slots of FBP only seek solutions for closing, without redoing the opening strategy?
I tried with the blocking of the opening slots, but I do not see improvements, it seems that FBP can not concentare only improve a part of the trade (closing) but the whole. These tests I've done with hundreds of different strategies, with different indicators opening.

Popov mentioned in a video overview of the monte carlo too l that the randomly close option was something that likely wouldn't happen (and if your broker is randomly closing positions on you you might want to find a better broker smile )  If that is the only setting that seems to be of issue and all the others pass I wouldn't discount the strategy unless you viewed that as a real world scenario.

Re: Monte Carlo

thank you

6 (edited by amenobh 2016-08-01 20:59:14)

Re: Monte Carlo

What GizmoTN said is completely right. I would add that this close test may be broken . Its not possible 2 % chance of randomly close a position completely break the strategy. In my tests I have not used this option anymore.

If you wanna improve your testings skills, i recommend this videos from popov http://forexsb.com/wiki/videos/expert_advisor_series

Re: Monte Carlo

Blaiserboy wrote:

Myself..... because FSBPro generates so many strategies so quickly...... I disregard the poor results in Monte Carlo and move on to the next one.


Thank you for this tip Good Sir..I am following in your footsteps and I believe very soon I will generate a robust and profitable EA with Monte Carlo even on any time frame..Thank you once again..You opened my eyes.