|
PureBytes Links
Trading Reference Links
|
Thanks for the email
I agree with that ... irregular number of years of data and limited number
of years can create some nasty surprises. Some real "sigma events"<G>.
For that reason, I've been converting all of my spreads and crosses from 12
and 15 year data to 30 year data. It's a real eye opener. Since it's less
than 100 data points, normal distribution analysis is still suspect.
Re your event trade analysis, I use VBA to plot 20 to 50 trades on one
XLchart. Calculate the mean plus stdev and overlay, you'll soon pick up some
ideas for days-in-trade analysis. Maybe I should think about a "kurtosis"
indicator or rolling histogram <G>
Best regards
Walter
|