On the suboptimality of single-factor exercise strategies for Bermudan swaptions

This paper resolves the disagreement between Longstaff et al. [2001. Journal of Finance Economics 62, 39-66] and Andersen and Andreasen [2001. Journal of Financial Economics 62, 3-37] over the effectiveness of the common business practice of using best-fit single-factor term structure models to deduce exercise strategies of Bermudan swaptions. I examine the cost of using recalibrated single-factor models to determine the exercise strategy for Bermudan swaptions in a multifactor world. I show that single-factor exercise strategies applied in a multifactor world only give rise to economically insignificant losses. Furthermore, I find that the conditional model risk as defined in Longstaff et al. [2001. Journal of Finance Economics 62, 39-66] is statistically insignificant given the number of observations. Additional tests using the Primal-Dual algorithm of Andersen and Broadie [2004. Management Science 50(9)] indicate that losses found in Longstaff et al. [2001. Journal of Finance Economics 62, 39-66] cannot, as claimed, be ascribed to the number of factors.