edo2017ardo
14 posts
msg #137164
- Ignore edo2017ardo |
7/30/2017 3:55:00 PM
Does a filter that gets a minimum of 1% weekly profit with a 90% accuracy exist?
Or... is it asking too much or for something too good to be true?
|
four
5,087 posts
msg #137165
- Ignore four
modified |
7/30/2017 4:15:39 PM
The 10% that don't
versus
The 90% that do
--
http://www.wisdomtrading.com/risk-drawdown-ruin-calculator/
https://joehentges.net/outliers-trading/
https://en.wikipedia.org/wiki/Taleb_distribution
http://www.investopedia.com/articles/trading/09/risk-management.asp
|
Kevin_in_GA
4,599 posts
msg #137167
- Ignore Kevin_in_GA |
7/30/2017 4:33:29 PM
More important than accuracy is expectancy -
(winning% * average gain (in %)) - (losing%* average loss (in %))
if a filter doesn't have a positive expectancy, do not trade it.
|
edo2017ardo
14 posts
msg #137174
- Ignore edo2017ardo
modified |
7/30/2017 6:26:31 PM
Yes, "expectancy" is definitely a more correct concept and makes perfect sense!
Because it doesn't matter "how many" times your system wins or loses, if we do not consider at the same time "how much" it is won or lost...
Thank you Kevin
|