Testing for the Existence of a Generalized Wiener Process- the Case of Stock Prices

In this article, we present two nonparametric trispectrum based tests for testing the hypothesis that an observed time series was generated by what we call a generalized Wiener process (GWP). Assuming the existence of a Weiner process for asset rates of return is critical to the Black-Scholes model and its extension by Merton (BSM). The Hinich trispectrum-based test of linearity and the trispectrum extension of the Hinich-Rothman bispectrum test for time reversibility are used to test the validity of BSM. We apply the tests to a selection of high frequency NYSE and Australian (ASX) stocks.