Posted By: Mothership
A strategy's life-span... - 10/24/14 14:51
Hi!
Couple of weeks ago I somewhere read (can't really remember where I read or who said it) that strategies don't have infinite life - a strategy maybe works for 4-6 years or so.
A thought occurred to me consindering this. Often when I backtest my srategy the backtest window often consists of 10-12 years or so. Does this mean that the strategy "doesn't really work" on half (or so) of the backtest window? Won't the result be missleading then?
And, isn't the optimization-function supposed to take care of this "problem" since then optimization helps the strategy to adapt to new market conditions?
MS
Couple of weeks ago I somewhere read (can't really remember where I read or who said it) that strategies don't have infinite life - a strategy maybe works for 4-6 years or so.
A thought occurred to me consindering this. Often when I backtest my srategy the backtest window often consists of 10-12 years or so. Does this mean that the strategy "doesn't really work" on half (or so) of the backtest window? Won't the result be missleading then?
And, isn't the optimization-function supposed to take care of this "problem" since then optimization helps the strategy to adapt to new market conditions?
MS