In the 2nd Quarter 2004 release of SPV we have re-estimated the four standard models using an updated sample of mortgage bonds. In order to illustrate this effect on model parameters and model performance we have also re-estimated the M1 model using the old sample.
Furthermore we have changed the valuation model used to back test the model performance when it comes to valuation and total return calculations. However, we will continue to the traditional valuation model as well.
This quarterly theme concerns the performance of a hedging strategy based on factor durations where the hedge portfolio is a set of par swaps. We compare realized monthly returns on the mortgages and their hedge portfolios. We find that we are able to remove a large part of the interest rate risk. However, we still find significant residual variance in the hedged returns, in particular for the mortgages that are burned out, as they are more or less uncorrelated with interest rate factors used in the analysis.