Is the Stock Market Efficient? Evidence from Nonlinear Unit Root Tests for Nigeria

Abstract

This study re-examined the traditional research topic in finance—the efficient market hypothesis (EMH) within the context of a nonlinearity unit root test using daily data sourced on the Nigerian economy. This was premised on two key motivations. First, the study observed that most of the existing studies on EMH are based on an aggregate stock index and the presence of heterogeneity of listed firms on the floor of the exchange can make the results obtained from aggregate stock price-based tests misleading due to spuriousness. To overcome this, the study tested the validity of EMH at a sectorial level. The second motivation centers on the observed nonlinear property of the time series data used in the existing literature. The study first examines the unit root properties of the data by applying the Harvey, Leybourne, and Xiao (2008) methods. The results indicate rejections of the null hypothesis for all the series, an indication that stock market indices in Nigeria are nonlinear and asymmetric in nature. This suggests that results obtained from linear based models could be biased. In order to achieve more accurate results, the study applied the ESPAR model and observed that, overall, there is an abundance of evidence to show that the Nigerian stock exchange is mean reverting, thus investors are advised to adopt a contrarian investment strategy to maximize the opportunities in the market.