“Testing of weak form of efficient market hypothesis: evidence from the Bahrain Bourse”

Efficient market hypothesis (EMH) states that financial markets are “informationally efficient”, implying that current prices fully reflect all available information. The present study aims at testing the weak form of market efficiency of the individual stocks listed on the Bahrain Bourse for the period 2011 to 2015. Weak form of EMH is tested using the Kolmogorov-Smirnov goodness of fit test, run test and autocorrelation test. The K-S test result concludes that in general the stock price movement does not follow random walk. The results of the runs test reveals that share prices of seven companies do not follow random walk. Autocorrelation tests reveal that share prices exhibit low to moderate correlation varying from negative to positive values. As the study shows mixed results, it is difficult to conclude the weak form of efficiency of Bahrain Bourse. Iqbal Thonse Hawaldar (Bahrain), Babitha Rohit (India), Prakash Pinto (India) BUSINESS PERSPECTIVES LLC “СPС “Business Perspectives” Hryhorii Skovoroda lane, 10, Sumy, 40022, Ukraine www.businessperspectives.org Testing of weak form of efficient market hypothesis: evidence from the Bahrain Bourse Received on: 2 of April, 2017 Accepted on: 11th of July, 2017


INTRODUCTION
Efficient market hypothesis (EMH) states that financial markets are "informationally efficient", implying that current prices fully reflect all available information. According to Fama (1970), there are three forms of EMH. They are: "weak", "semi-strong" and "strong". Weak form states that prices already reflect all past publicly available information. Semi-strong EMH states that price changes reflect past information, as well as reflect new public information. Strong EMH claims that prices instantly reflect even hidden or "insider" information.
The present study tests the weak form of EMH of the individual stocks listed on the Bahrain Bourse for the period 2011 to 2015. Bahrain Stock Exchange established in 1987, officially began operations in 1989, with 29 listed companies and a manual trading system. However, in 2010, Bahrain Bourse (BHB) was established as a shareholding company to replace Bahrain Stock Exchange. The market capitalization of the Bourse at the end of 2015 was BD 7,199, 907, 825 with 46 listed companies. The Bahrain All-Share Index lists shares under six sectoral indices: commercial banks, investment, insurance, services, industrial and hotels and tourism.

LITERATURE REVIEW
A study into the literature reveals large number of publications having examined the existence of the EMH theory in various developed and undeveloped markets with varying results (Ananzeh, 2014). Iqbal and Mallikarjunappa (2008, 2010, 2011) studied on Indian Stock Market and found that Indian Stock Market is not efficient in weak and semi-strong form.
Awan and Subayyal (2016) studied six stock exchanges in Gulf region (Oman, UAE, Kuwait, Saudi Arabia, Bahrain and Qatar) for the period 2011 to 2015. The results provide evidence that the stock prices at the Gulf markets do not follow the random walk model. Asiri (2008) measured the performance of the Bahrain Stock Exchange (BSE) by applying the Dickey Fuller unit root tests, the ARIMA model, and exponential smoothing techniques. Dahel and Laabas (1999) compared the stock markets of Bahrain, Kuwait, Saudi Arabia and Oman. Rao and Shankaraiah (2003) investigated the stock market efficiency and suggested strategies for Gulf Co-operation Council (GCC) stock markets. They concluded that these markets were neither developed nor informationally efficient and recommended that there must be better networking between these markets. Sharma (2005) tested the daily returns series of Gulf Co-operation Council (GCC) stock markets and reported that the returns follow a normal distribution.
Elango and Hussein (2007) examined market efficiency across the stock markets in the GCC countries using daily indices data from October 2001 to October 2006. Kolmogorov-Smirnov test has been applied for analysis and the study rejects the null hypothesis that the returns follow a normal distribution for all seven markets. Run test has been used for randomness. It is concluded that the market is not efficient in its weak form.
Very few studies (Solnik, 1973 Solnik (1973) investigated market efficiency for 234 securities in eight major European stock markets using daily prices. Ang and Pohlman (1978) examined 54 stocks of five stock exchanges of Japan, Singapore, Australia, Hong Kong and the Philippines and concluded that the markets are slightly efficient in the weakest form. Moustafa (2004) analyzed the weak form efficiency of United Arab Emirates (UAE) stock market by observing the behavior of stock prices. Three years data were considered from October 2, 2001 to September 1, 2003. Using nonparametric run test on 43 stocks of the Emirates market and the study concluded that stocks didn't follow the normal distribution.

METHODOLOGY
The present study uses daily prices of the companies listed in the Bahrain All Share Index. The period of study is from 2011 to 2015. The sample consists of 43 companies listed on the Bahrain Bourse for which data are available.
The stock returns are defined as follows: Log is the logarithmic price at time t and 1 pt Log − is the logarithmic at time The study analyzes and tests the weak form of efficiency for the individual stock movements.
Weak form of EMH is tested using the run test, Kolmogrov-Smirnov goodness of fit test and autocorrelation test.
Run test: run test is a nonparametric test. It is a statistical procedure that examines whether a string of data is occurring randomly given a specific distribution ( The standard normal Z-statistic used to conduct a run test is given by: . Kolmogrov-Smirnov goodness of fit test (K-S test): the K-S test is a nonparametric test and is used to determine how well a random sample of data fits a particular distribution (Mobareket et al., 2008; Salam, 2013).
Autocorrelation: Autocorrelation is defined as a mathematical representation of the degree of similarity between a given time series and a lagged version of itself over successive time intervals (Hamid et al., 2010). A higher value of autocorrelation indicates greater correlation in values. Fama (1970) proposes that by using a less restrictive condition of submartingale sequence, low serial correlation may persist in an efficient market.
where k P -auto-correlation coefficients at lagk , n -sample size.
Descriptive statistics of the returns of the listed companies on the Bahrain Stock Exchange is presented in Table 1. Delmon Poultry Co. (POLTRY) records the highest mean returns of 1.528%. The returns of companies that have thin trades are GLOBAL, INVCORP, TAKAFUL, UFC and UPI. Bahrain Telecommunication Co. (BATELCO) having the highest number of trading days has a mean return of -0.057% with standard deviation of 1.272%. The Bahrain All Share Index records a mean return of -0.013% and standard deviation of 0.455%. The results for GLOBAL, TAKAFUL and UPI are not arrived at due to lack of data.

CONCLUSION
The present study aims to test the weak form of efficiency of the individual stock listed on Bahrain Bourse for the period 2011 to 2015. Delmon Poultry Co. (POLTRY) records the highest mean returns of 1.528%, while the Bahrain All Share Index records a mean return of -0.013%. The K-S test result conScludes that in general, the individual stock returns do not follow random walk. Run test shows that the successive price changes are not random (Elango & Hussein, 2007).
Autocorrelation tests reveal that share prices exhibit low to moderate correlation varying from negative to positive values. The results of the tests do not support the randomness of the returns of the individual stocks. Therefore, it is difficult to conclude that successive returns of the stocks listed on Bahrain Bourse follow a random walk. These conclusions are similar to Solnik (1973), Rao and Shankaraiah (2003), Moustafa (2004), and Elango and Hussein (2007).