[R-SIG-Finance] mcr, cr, and pcr at security level

[ezivot at u.washington.edu]

I think you are approaching the problem incorrectly. The easiest thing for you to do to get factor contributions at the portfolio is to run my function using the portfolio level inputs - portfolio betas and the portfolio variance. This will give you the right answer.

****************************************************************
*  Eric Zivot                  			               *
*  Robert Richards Chaired Professor of Economics              *
*  Department of Economics                                     *
*  Adjunct Professor of Finance                                *
*  Adjunct Professor of Statistics
*  Box 353330                  email:  ezivot at u.washington.edu *
*  University of Washington    phone:  206-543-6715            *
*  Seattle, WA 98195-3330                                      *                                                           *
*  www:  http://faculty.washington.edu/ezivot                  *
****************************************************************

On Thu, 27 Oct 2011, Arun Soni wrote:

> Dear community brethren,
>
> I have computed the factor risk decomposition using Dr Eric Zivot's script for factorRiskDecomposition that uses a cross-sectional fundamental factor model estimated from stock level to arrive at a factor model with 70+ factors.  Now that I have the factor level marginal contribution to risk, contribution to risk, and percentage contribution to risk, I would like to compute these measures at the stock level for each holding in the portfolio.  I was under the impression that since the cr is additive, I could multiply the matrix of weighted security exposures (betas) to each factor with the matrix of factor contributions to risk, which would yield me each stock's contribution to risk from all the factors.  I would then add each stock's contribution to residual risk to its contribution to factor risk to yield each stock's contribution to total risk, and the summing this measure across all the stocks in the portfolio should result in the portfolio risk but I am well short of t!
 hat figure.
>
> I would then like to similarly calculate each stock's marginal contribution to risk as well as its percentage contribution to risk.
>
> I would be extremely grateful if I could get some insight into my error.
>
>
> Regards
>
> Arun
>
>