Are Gulf stock markets efficient? Evidence from new multiple variance ratio tests

Abstract

The purpose of this article is to test for the Random Walk Hypothesis (RWH) for seven stock markets in Gulf Cooperation Council (GCC) countries, and to determine the effect of the correction for thin trading. Three new multiple variance ratio tests are applied to both observed returns and returns corrected for thin trading. It is found overall that the RWH does not hold for the GCC stock markets at both daily and weekly frequencies. This evidence is particularly strong when daily returns are used, where the RWH is soundly rejected for both observed and corrected returns.

Keywords:

• joint test
• random walk
• variance ratio tests
• wild bootstrap

JEL Classification::

• G14
• G15
• GCC

Acknowledgements

Professor Jasim Al-Ajmi would like to thank Deanship of Scientific Research at the University of Bahrain for the research grant number 427/2002.

Notes

¹ http://www.mscibarra.com/products/indices/gcc/.

² Chen and Deo (2006) also proposed a joint intersection test where one-sided alternative hypothesis is imposed. In this article, we consider their joint test with the two-sided alternative to be comparable with the other tests considered in this article.

³ This cutoff point coincides the time when several major US banks failed or were subject to government takeover. This is widely taken as the point where the subprime mortgage crisis hits the global financial market (http://en.wikipedia.org/wiki/Subprime_mortgage_crisis#Financial_market_impacts.2C_2008).

⁴ For further details about the FTSE classification see http://www.ftse.com/Indices/Country_Classification/Downloads/FTSE_Country_Classification_Sept_08_update.pdf.