|
/ | Forum |
  |
rortiz77
2009.12.22 18:21
Does it make more sense to run an optimization for 2-3 years and use those settings and hope that it captures 99% of all scenarios? or Run a 1-2 month optimization every month? and use those settings for a month...and then run the optimization again? Thanks, -Ron |
 |
Mechanical Trading System "Chuvashov's Fork" This article draws your attention to the brief review of the method and program code of the mechanical trading system based on the technique proposed by Stanislav Chuvashov. The market analysis considered in the article has something in common with Thomas DeMark's approach to drawing trend lines for the last closest time interval, fractals being the reference points in the construction of trend lines. |
|
Jellybean
2009.12.22 21:40
rortiz77:
Does it make more sense to run an optimization for 2-3 years and use those settings and hope that it captures 99% of all scenarios? or Run a 1-2 month optimization every month? and use those settings for a month...and then run the optimization again? Thanks, -Ron Hi Ron It all depends on the strategy (especially it's timeframe). I suggest you try optimising for a 2 year period starting, say, 3 years ago and see how those parameters perform in the year just passed. Then do the same for, say, 3 months starting 6 months ago and test those parameters on the last 3 months. I've tried optimising strategies for various timeframes and found little consistency. I think it is more important to be able to identify when the strategy isn't performing as expected so you can stop it until conditions change to suit it again. Cheers Jellybean |