Probability - page 2

New comment
 
 

I think the difficulty of the below example, at least in my opinion is any probability you come up with off of simply those numbers will not be as high of quality(reliable) as if you filtered it with something or another specific addition.

In your example from the open of the first to the close on the last you have a pretty decent range of .0179 or 179 pips.

Based simply on those numbers I think your best chance is to go off of variance or standard deviation. So you could say from the most recent close what is the probability that it hits a certain price based on the average standard deviation of the previous 4 bars. Again I still feel like just doing that would be somewhat hard to predict a trend, even though obviously if the trend is down(up) it is likely to continue until it doesn't.

range of 1st bar:181 pips

range of 2nd bar: 70 pips

range of 3rd bar: 177 pips

range of 4th bar: 120 pips

Average range is 137 pips based on that sample.

Variance

(181-137)^2=1936

(70-137)^2=4489

(177-137)^2=1600

(120-137)^2=289

ave variance 2078.5?

I could be going about this all wrong, when would you plan to open your position? At the open of the next bar? If you assume you opened your position at the open of each bar you would have a 25% chance of being right long, and a 75% chance of being right short(going off of open to the high, and open to the low). The reason I ask is depending on the trend you would have a much higher chance of being right(in this case you are in a downtrend with a much higher chance of shorting at the open). This is why I was saying it would be good to filter it. obviously until the 4th bar you don't know the probability, and you certainly don't know it at the first bar unless you have a better reference.

The trend will persist until it doesn't but again, I feel like to increase your chance of being right you have to consider the duration or distance of the trend move in comparison to average, because if say on average a trend lasts 200 pips, you may not like the risk of opening a trade in that direction, but if it lasts 500 pips you would likely want to stick with the trend.

Let me break out some books and see if I can come up with an actual good answer instead of stating the obvious. I'll let smarter people attempt to help you, haha.

Ben

 

Gidday Ben

You will have to excuse me my explaination will be a bit ass about face but never the less.

I plan to use the open of a new bar. if you look at the mm sheet you will see that there is a max open positions I think it said 13 so I could open only 13 positions at a time one per pair type of thing.

There is a max DD of 25% i use that as a close all so if the equity fall 25% below balance then close all.

So for all positions there is no SL I just use a TS of the 2% which is the size of the positions I opened and the main focus point the TP.

The TP is the 2% of the balance per pair divided by the pip value multipled by the ME. so if the SL is 44 pip and the ME is 2.33 then the TP will be 44x2.33 = 102 pip so if we have a higher probability on the short side then we are more like to hit that 102 pip and take profit. hense the main focus being the TP exit point.

My idea is to let a trade breath and what are the chances that all 13 positions will develop in to a loss.

Sorry about the messed up way this is comming to gether but at least it's comming to gether slowly.

Cheers

Beno

 

Great Start

Hello

I have been playing with this my self. I'm working on G/U for now. I will clean up my spreadsheet and post it here. I basically used Data analysis in Excel

to give me descriptive stats one year at the time. Here is example for the last year. For H1 bar

RANGE 2008

Mean 42.77020654

Standard Error 0.425736519

Median 33

Mode 19

Standard Deviation 33.38436744

Sample Variance 1114.515989

Kurtosis 9.383841696

Skewness 2.366877386

Range 376

Minimum 1

Maximum 377

Sum 262994

Count 6149

Confidence Level(95.0%) 0.834592528

and Frequency table

Bin Frequency Cumulative %

0 25 2175 35.37%

25 50 2243 71.85%

50 75 921 86.83%

75 100 429 93.80%

100 125 193 96.94%

125 150 96 98.50%

150 175 42 99.19%

175 200 20 99.51%

200 225 13 99.72%

225 250 9 99.87%

250 275 3 99.92%

275 300 2 99.95%

300 325 2 99.98%

325 350 0 99.98%

350 375 0 99.98%

375 400 1 100.00%

More 0 100.00%

Let me know if we are on the same page here or I'm way of the base.

 

Gidday dejnik

Welcome to the forum thanks for the input the more the merrier.

This is all one big experiment so all info is welcome.

I will look forward to looking at your sheet.

Cheers

Beno

 

Gidday

Here is some further reading you might want to have a look at.

Probability and Integration, I

I think it's similar to what we are trying to do TP = Target.

 

Beno,

How complicated or simple are you wanting this setup to be? Are you thinking of something as complicated as the Better EA that won the EA competition last year?

Sorry I am still trying to understand the trade setup, are you saying you are using(or wanting to use) a 2% trailing stop so once you determine the trade you just enter and stay in till you get stopped out with a 2% account loss or until you reach TP? Sorry still new to the terminology of all this, does ME=margin equity?

Just trying to get clarification so I can make sure I am trying to be as logical and efficient as possible with ideas.

thanks

Ben

 
whitemagicbmc:
Beno,

How complicated or simple are you wanting this setup to be? Are you thinking of something as complicated as the Better EA that won the EA competition last year?

Sorry I am still trying to understand the trade setup, are you saying you are using(or wanting to use) a 2% trailing stop so once you determine the trade you just enter and stay in till you get stopped out with a 2% account loss or until you reach TP? Sorry still new to the terminology of all this, does ME=margin equity?

Just trying to get clarification so I can make sure I am trying to be as logical and efficient as possible with ideas.

thanks

Ben

Yes I use a 2% trailing stop so the trailing stop will not start until there is 2% profit. or the TP is hit.

ME = Mathematical Expectancy which is the amount your account is expected to grow by formula is (1-AverageWin/AverageLoss)*(Accuracy)+1)

you could have an accuracy as low a 18% and still have a positive expectancy and still be making money not a lot but a small amount.

All this may look or seem complicated but if you look at the formula's on the mm sheet you will see how everything fits together in an intergrated fashion.

I hope that helps

Cheers

Beno

 

haha gotcha, yea I was referencing expectancy as well just didn't know the abbreviation of ME you were using. Thanks for clarifying, I will check out the MM file and see what you are talking about.

Thanks

Ben

 

Have a look!

News - Automated Trading Championship 2007

B.

123
New comment
Reason: