[R-SIG-Finance] Return.rebalancing contemporaneous calculation

[Brian G. Peterson]

On 10/08/2014 03:24 PM, Charles Duranceau wrote:
> Hi,
> I'm working with "PerformanceAnalytics" for calculating portfolio returns using the function Return.rebalancing. I don't understand why the results - portfolio returns - are lagged by 1 period when compared to Return.portfolio (assuming weight are time invariant).
>
> I read the definition that explain the rebalancing is taken place in the next period but since the input are returns (like in  Return.portfolio) they should be consolidated on same period.
> Is there a way not losing the first input?
>
> Here few lines to illustrate my question
>
> pfmf2003<-Return.rebalancing(mfr2003,wghts1)
> pfstat<-Return.portfolio(mfr2003,c(1,0))
> ** wghts1 = 1 for the period
>
> Results
>                          portfolio.Rebal       portfolio.returns.
> 2003-05-31      0.0087336245         0.203679369
> 2003-06-30      0.1093073593         0.008733624
> 2003-07-31     -0.0351219512         0.109307359
> 2003-08-31     -0.0404448938        -0.035121951
> 2003-09-30     -0.0063224447        -0.040444894

Return.portfolio and Return.rebalancing are the same function in current
CRAN version, and have been in the R-Forge version for quite a while.

Can you give us your sessionInfo, and a fully reproducible example?

Brian

-- 
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock