Using historical data to get your settings in the ball park is an OK starting place but using current data for the final calculations will keep you system in the groove.
Instead of setting your program to function in the past why not set your stops dynamically with up-to-date data by calculating a stop offset using something like ATR. In my opinion it is dangerous to use optimized results of past market action to set trade parameters for future markets. Calculate them with current, live data. Barry --- In [email protected], "gonzagags" <[EMAIL PROTECTED]> wrote: > > Hello > I use very much the walk forward optimization, and I think that only > out-of-sample data are useful data to know how your system behaves. > So, I want to add to my system's code, both the two variables I > optimize, changing their value as the date change. > e.g, from 1-1-1999 to 3-31-1999, stoploss must be 5% and stopprofit, > 3%. And, from 4-1-1999 to 6-30-1999, the values must change, 4.5% and > 2%, for example.. and so on.. > I suppose using the switch command and some kind of variables I could > change my stoploss and stopprofit parameters when the date change, but > I haven't seen any similar example.. > Any idea? > Thanks >
