Sunday, December 25, 2011

Portfolio Optimization in R, a previous error.

I realized that I made a mistake in the calculation of the market weight portfolio from the previous post.  I hold constant the portfolio weights through time. These should be adjusted after each day as prices change.  The market portfolio requires no re-balancing.  What I have is, in essence, a re-balancing back to the original weights at the end of each day.

The equal weight portfolio is OK.  As we assume no transaction costs, it can be re-balanced daily.  In fact, the constant weight is the key assumption of this portfolio, so it SHOULD be re-balanced daily.

The optimal portfolio model also has the same issue.  We are implicitly re-balancing each day.  Our assumption in the optimization was holding for the entire period.

I will make the corrections and post results and updated code.  Being the holiday season, it might take a few days longer than normal.

Merry Christmas!

1 comment:

  1. using monthly return instead of daily return may make this more practical as we can rebalance monthly?