Real-world key correlations: Assessing the bond-equity correlation over a one year horizon
Document ID: 2008-1230 (previously 2008/85)
Published on: 1st November 2008
Author: Steffen Sorensen & Martin Skrk
Due to the strong flight-to-quality effects in financial markets the short-term (conditional, ‘point in time’) correlation between bond and equity returns can differ from the long-term (unconditional, ‘through the cycle’) correlation. Although the unconditional target set by Barrie&Hibbert does include information on historical flight-to-quality events, care needs to be taken if we want to assess a suitable target for a one year projection. As we have shown elsewhere, we need appropriate multivariate statistical estimation techniques to assess the bond-equity correlation. Such estimators can be complex to use for out of sample one year forecasting and it is therefore important to consider whether a more simple approach can be used.
This note discusses an alternative method that can be used to assess the one year conditional (‘point in time’) correlation between bond and equity returns
