Measuring asymmetry

Abstract

A range of primary methods for dealing with price asymmetry, such as the approaches proposed by WOLFFRAM (1971) and HOUCK (1977), have been established in the literature, but consensus on which method should be preferred remains elusive. This note demonstrates that, theoretically, these two definitions are equivalent to a straightforward notion of asymmetry based on first differences. Using monthly data on gasoline prices and sales from the U.S., we illustrate, however, that, in practice, these approaches may yield divergent conclusions with respect to asymmetry.We argue that in such situations the asymmetry notion based on first differences should be preferred.