Price performance and Egyptian stock market efficiency : an intial public offerings perspective
The core of this thesis has involved an examination of the efficiency of the Egyptian stock market (ESM) with a specific focus on the price performance of the privatised initial public offerings (PIPOs). Recent structural changes in the Egyptian economy during the 1990s permit testing hypotheses about how these changes have affected the behaviour of ESM, in general, and PIPOs in particular. An analytical review of prior studies is provided in Chapter Two. Two documented anomalies of IPOs price performance, i.e. short-run underpricing and long-run overpricing, are revealed. Some researchers attribute these findings to the trading system of the developed capital markets. Our study refutes this explanation because we also find these anomalies in the ESM, although it is a market without an investment-banker (specialist) system. Accordingly, five empirical chapters are constructed to investigate the ESM. Before examining the price performance of PIPOs in the ESM, two chapters are assigned to examine the whole market at the domestic and international levels, as a preliminary exploration. From the domestic point of view, Chapter Four deals with questions of normality, volatility, randomness, and the efficiency of the ESM. Several basic tests were employed for testing normality. All indicated that none of the indices has a normally distributed return. Then, the Autoregressive Conditional Heteroscedastic model (ARCH) proposed by Robert Engle (1982) and the Generalized ARCH model (GARCR) of Bollerslev (1986) are employed to describe the process of stock returns. The findings show that the variance of returns is time-varying in the GARCH context. Also, the integratedness of the volatility of asset returns is analyzed using the IGARCH model. The results indicate that the volatility of stock returns is integrated. To test the stationarity of the ESM returns, unit root tests of Dickey and Fuller (1979) and the variance-ratio test of Lo and MacKinlay (1988) were implemented. The results support the notion that there is a relatively significant stationary component in past returns that can be used to predict future returns; therefore, returns do not follow pure random walks. Since the random walk hypothesis is not equivalent to market efficiency, we conduct the test of efficiency by using unit root and cointegration techniques, which are recently developed techniques in the time series literature. It is found that disaggregate stock price indices of the ESM are cointegrated which is interpreted as a violation of the concept of static efficiency introduced by MacDonald and Power (1993). Then, Chapter Five is assigned to test the internationalization of the ESM among eighteen emerging international stock markets. The Engle-Granger two-step methodology and the Johansen's multivariate cointegration tests were performed on these prices. The findings show that the eighteen emerging markets are cointegrated, indicating Granger-Causality in levels and these are suggesting of inefficiency. However, for the Middle Eastern and Mediterranean Rim markets groups, the results reveal an absence of any clear evidence of cointegration among them. Then, to measure the price performance of PIPOs, we use both the market-adjusted and risk-adjusted models. In the risk-adjusted model, both the general CAPM and the Returns Across Time and Securities (RATS) model were employed. Chapter Six illustrates that the Egyptian PIPOs are underpriced with average initial returns of 15.03 % and the observed distribution is heavily skewed and has a median of 13 %. Chapter Seven shows that insignificant positive excess market returns exist, on average, between the close in the first day of listing and the close in the fourth week of trading. It is suggested that these early positive excess market returns in the aftermarket may result from speculative bubbles which burst in subsequent trading in the aftermarket period giving rise to negative excess market returns. Also , the results indicate that the mean beta declines after-listing and varies around the market beta of unity. The mean beta in the Egyptian PIPOs market thus appear to behave nearly in a similar manner to the risk behaviour in other markets. Finally, Chapter Eight investigates the efficiency of the Egyptian PIPOs in the aftermarket. The results supported both the weak-form and semistrong-form of the Efficient Market Hypothesis of the PIPOs in the ESM.