Hi, I wish to ask about how to combine a series of log return time series to derive the portfolio variance - I understand that log returns is additive across time but not across components - hence for portfolio mean, the procedure is to change them to arithmetic form before doing the weights and then take the log of it. May I ask what is the procedure to do so for variance then? I haven't seen any code examples transforming it back to percentage returns so I thought perhaps I am missing something.
Portfolio Variance of Log Returns
- Thread starter myndelsg
- Start date
