Comparing Hedge Ratio Methodologies for Fixed-income Investments

Regression and duratio n are com peting hed ging models for reducing the risk of a debt position. This paper compa res these mode ls to determ ine if one method provides consistently superior hedging results. Both perfect forecast (in-sample) and historical (out-ofsample) hedge ratios a re emplo yed to hedge the long-term Bellwether bond and the two-year T-note. The regression procedure provides smaller d ollar errors for the B ellwether se ries, but neither method is consistently superior when two-year T-notes are hedged. Comparison against a no-hedge position and two naive hedge ratio methods shows the overal l superio rity of the regression and duration models. Previous claims that duration is superior w hen end-ofperiod prices a re known or that regre ssion and duration sho uld provide equivalent results are questionable.