The Information Content of Stock Prices in Ghana

Abstract

This paper investigates the weak form stock market efficiency in Ghana. We sought to establish whether investors in Ghana can form profitable trading strategies based on the information content of historical stock prices. We employed serial and cross-sectional correlation tests to ascertain the relationship between daily, monthly, quarterly and yearly stock returns. We find that stock returns are not normally distributed in Ghana and non-parametric and parametric tests applied to triangulate results did not generate any significant differences in the inferences drawn. We also find that for an overwhelming majority of listed stocks, daily stock returns are positively serially and cross sectionally correlated in a significant way irrespective of the method applied. Parametric test results for monthly returns indicate the dominance of insignificant positive serial correlation but the non-parametric test produced significant positive serial correlation. Cross sectional parametric and non-parametric tests show significant positive correlation in monthly returns. The differences between daily and monthly return correlation were insignificant. Quarterly returns were insignificantly positively serially correlated but the cross-sectional results provided evidence of significant positive correlation. Yearly returns were negatively correlated suggesting a price reversal phenomenon exists in Ghana but insignificant on the whole. Our test results suggest that the Ghana stock market is to a large extent weak-form inefficient.

Keywords: Stock Market Efficiency, Efficient Capital Markets, Ghana, Emerging Markets, Information Content, Stock Prices, Normality, Serial Correlation Tests, Auto-Correlation Test

JEL Classification: G12, G14, G15, N27