A Robust Entropy-Based Test of Asymmetry for Discrete and Continuous Processes

We consider a metric entropy capable of detecting deviations from symmetry that is suitable for both discrete and continuous processes. A test statistic is constructed from an integrated normed difference between nonparametric estimates of two density functions. The null distribution (symmetry) is obtained by resampling from an artificially lengthened series constructed from a rotation of the original series about its mean (median, mode). Simulations demonstrate that the test has correct size and good power in the direction of interesting alternatives, while applications to updated Nelson & Plosser (1982) data demonstrate its potential power gains relative to existing tests. ∗The authors would like to thank Cees Diks, Estela Bee Dagum, and anonymous referees for their valuable comments and suggestions. Racine would like to gratefully acknowledge support from Natural Sciences and Engineering Research Council of Canada (NSERC:www.nserc.ca), the Social Sciences and Humanities Research Council of Canada (SSHRC:www.sshrc.ca), and the Shared Hierarchical Academic Research Computing Network (SHARCNET:www.sharcnet.ca).