This study seeks evidence on whether the return series on Bangladesh's Dhaka Stock Exchange (DSE) is independent and followsthe random walk model. The study focuses on assessing if the DSE deviates from idealised efficiency. The sample primarily includes all the listed companies on the DSE daily price index over the period 1988 to 2000. The results of both non-parametric (Kolmogrov-Smirnov: normality test and run test) and parametric test (Auto-correlation test, Autoregressive model, ARIMA model) provide evidence that the security returns do not follow the random walk model and the significant auto-correlation coefficient at different lags reject the null hypothesis of weak-form efficiency. The results are consistent with observations in different sub-samples without outlier and for individual securities. This anomaly with the efficient market hypothesis supports the thought that the market does not respond to new information instantaneously. This may be due to a delay in dissemination to new price sensitive information or biases (under or over reaction) in the response of market participants to such information. It may also be for the momentum effect related to herding in particular 'positive feed back trading' or 'trend following' the trading strategy by the average investors. JEL Classification: G12, G14, G34

This paper investigates the underlying factors that determine share returns on the Dhaka Stock Exchange. The empirical analysis does not support the critical condition of the Capital Asset Pricing Model of a positive relationship between share return and beta. However, it shows that variables such as size, price to book, volume of shares traded, earnings yield and cash flow yield have a significant influence on share returns. The degree and direction of relationship among the variables are similar to other emerging markets, but are not always consistent with developed markets perhaps due to lack of homogeneous expectations regarding risk return characteristics and different market microstructure.

Market efficiency is an area of enormous interest in financial literature. Numerous researchers conducted empirical studies in testing weak-form market efficiency in several stock markets and employed various techniques but the empirical evidence is controversial. Triangulation econometric approach is employed to assess the predictability of daily return series of Botswana Stock Exchange (BSE) and to test the null hypothesis of random walk model. The empirical results reject the null hypothesis of random walk model for the daily return series of BSE for the period of 1989–2005 and evidenced serial autocorrelation of return series, which clearly indicate predictability and volatility of security prices of Botswana market. However, the empirical evidence of both non-parametric (Kolmogrov–Smirnov: normality test and run test) and parametric test (Auto-correlation test, Auto-regressive model, ARIMA model) reject the hypothesis of random walk model and indeed violate the notion of weak-form market efficiency.