Lain
 Regular
   Posts: 77
Joined: 10/11/2012
User Profile |
Since Portfolios built with Port Wizard are dynamic I would assume if all the strategies AND parameters are the same then it shouldn't matter what the simulation end dates is since it always takes the selected strategies and picks the "top" ones from a prior period at the beginning of every month to trade for the next month.
Would this be a correct statement?
I've been doing a lot of tests with Port Wizard and I'm getting different end results depending on the end date I use.
Example.
I have 15 strategies selected. 100 period look back. Monthly. 1 Interval, 5 strategies with 5 trades per month. Evaluation Calmar
So from what I understand. At the beginning of every month. It looks back 100 periods and picks the top 5 strategies with the highest Calmar (and min 5 trades per month).
Correct?
When I run with a date of 1/1/2000 to 1/1/2011 (Portfolio 1)
And save it, the portfolio page shows it only uses 12 strategies, so I looks like 3 of them didn't make the cut for that period.
When I run with a date of 1/1/2000 to 1/1/2015 (Portfolio 2)
And save it, the portfolio page shows it only uses 13 strategies, so I looks like 2 of them didn't make the cut for that period.
The one specific strategy that didn't make it as a R24 strategy.
I took a look through Selection Output page and indeed R24 was NOT used in Portfolio 1.
I then took a look through the Selection Output page for Portfolio 2 and R24 was used..... BUT it was used continually all the way back to 2000.
So why was R24 used in Portfolio 2 and NOT Portfolio 1 when all the other settings were the same, just a different end date.
Thanks,
Lain
[Edited by Lain on 7/15/2015 3:04 PM]
|