On the validity of conventional statistical tests given evidence of nonsynchronous trading and nonlinear dynamics in returns generating process: a further note

Abstract

Given the growing empirical evidence that returns predictability follows an evolutionary path, it calls into question not only the usefulness of conventional statistical tests of market efficiency as highlighted by Saadi et al. (2006), but also the adequacy of the efficient markets hypothesis to explain observed market dynamics.

Acknowledgement

The first author would like to thank Universiti Malaysia Sabah for giving him a scholarship to pursue his PhD study in Monash University.

Notes

¹ Instead, the more stringent condition of market efficiency that rules out the economic significance of the observed predictability has received less attention, and this has always been used by proponents of EMH to downplay any serious challenge posed by evidence of returns predictability (see, for example, Malkiel, 2004).

² The third criticism is more specific, focusing on the inappropriateness of unit root tests for testing EMH since they are not predictability tests.

³ Due to space constraint, references are kept to minimum. However, they are available upon request from the authors.