Abstract

The forward premium puzzle needs to be reformulated since extant studies address the negative slopes associated with the long dollar swings of the 1980s. By contrast the insignificant coefficients from recent data spans can be explained by an unbalanced regression problem caused by asymmetries in spot returns. These stem from market frictions such as transaction costs and are associated with overshooting of spot rates. Monte Carlo experiments show that asymmetries and overshooting effects produce widely dispersed and statistically insignificant slope coefficients whose small sample mean is close to zero.

Citations