Testing the weak-form efficiency in Middle Eastern financial markets: empirical evidence from the Palestine Exchange and its main competitors

The Efficient Market Hypothesis can be defined by the simple statement that “security prices fully reflect all available information”. (Fama 1991, p. 1575) While there are different levels of market efficiency, the focus of this paper is the weak-form, in which information sets include only historical prices and “how well do past returns predict future returns?” (ibid.) This dissertation investigates the weak-form efficiency of the Palestine Exchange and its main competitors (Amman Stock Exchange and Dubai Financial Market) for the post 2010 period, in which these markets are characterized by illiquidity and thin trading. Main indexes daily closings for the markets under study during the period 2010-2013 were used to examine efficiency, by employing a number of the most common econometric techniques seen in recent empirical results covering Middle East North Africa (MENA) financial markets. Hence Augmented Dickey-Fuller test was conducted to examine the presence of unit root, then serial autocorrelations of returns at various lags were estimated, followed by employing a non-parametric runs test to investigate the randomness of returns. This research also takes into account the possible costs of infrequent trading, which may increase the tendency of rejecting market efficiency hypothesis even if it is true. Therefore to avoid the effect of thin trading, the econometrics techniques mentioned were applied again, using adjusted returns rather than observed daily indexes returns. Results obtained using the observed indexes returns reveal that the Palestine Exchange and Dubai Financial Market were found to be weak-form efficient during post 2010 period. On the other hand Amman Stock Exchange was found to be weak-form inefficient during the same period. However when adjusting returns for infrequent trading using the approach proposed by Miller et al (1994) – explained in chapter 4.2.4 – no conclusion could been drawn, as conflicting and contradicting results were obtained. This lead to the assumption that the proposed model cannot be used for markets under examination without further modifications, as it was developed for more sophisticated and liquid markets. Investigating the further attempts needed to develop a more suitable model for adjusting infrequent trading effect in the markets under study will be left for future research