[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Equity Curve Filtering



PureBytes Links

Trading Reference Links

Be careful if you are attempting to find the best equity curve on a 
day-by-day, month-by-month basis.  If EC #8 performs better than the rest 
this month, then you are relying on a single month of data to make this 
conclusion. Reliance on a single month of data can be statistically 
problematic in terms of predicting the future performance of EC #8.

You can compare a basket of equity curves when you look at their entire 
data sets as a whole.  For example, if you are comparing 10 equity curves, 
they should each represent a significant number of trades going back 
several years.  Apply statistics to all of the data, rather than how your 
ECs are doing the past few months.


At 03:18 AM 3/15/2002 -0500, Mark Conway wrote:
>I've programmed some EL where there's an array of Equity Curve values. Now
>that I've got data where the EC mean can be calculated as well as the delta
>above or below the EC, it would be great to hear some ideas or get some
>pointers about how to interpret this data to implement some EC filtering.