[andypap]

hi

any suggestions on where to start for computing the correlations between the equity curve of a strategy and predefined popularions of variables,,,
or even
a group of equity curves of the same strategy among different data sets and sets of variables..

i am intrested in software or articles or web sites references or your own experience in such an endeavour if anyone is working on this field of finding the critical components of sucess or failure of a strategy..

the ideal process the way i can picture it is to start by setting manually begin and ending periods of success/ failure samples of the equity curve and try various sets of variables to see what is connected or random to these ''period" and IF something stands out try to writte a predictive code that will switch the strategy on and off

this is a point B to point A type of approach that i think is the step number 2(instead of parameter optimization) once a system is generating a satisfactory equity curve in a large data set with a big population of trades..

regards