STOCK MARKETS IN THE EAST ASIAN ECONOMIES : FURTHER EVIDENCE OF SEASONAL PATTERNS

Abstract

There has been a growing number of studies exploring market anomalies where stock returns tend to deviate from the normal paradigm of market efficiency. Evidence of significant and persistent seasonal variations in mean returns were discovered in major stock markets. Of particular interest are the calender anomalies, where systematic variations in stock prices in relation to the calender year were observed. This challenges the foundation of the efficient market hypothesis. This paper investigates the existence of market anomalies in the stock markets of Japan, Hong Kong, Singapore, Malaysia and Taiwan. It offers further evidence of seasonal patterns to establish whether they are analogous to those found in other established markets. The seasonal patterns analysed are the day-of-the-week effect, turn-of-the-month effect, monthly effect, turn-of the-quarter effect, January effect and the holiday effect. An overview of the five East Asian countries is presented. The significance of these markets in Asia is demonstrated by its market capitalisation and total value traded over the period of the analysis. Some key characteristics and the development of these stock markets are provided. Prior research on seasonal patterns are discussed in this paper with particular emphasis on possible explanations for these effects. Although empirical evidence strongly suggests that seasonal variations in mean returns exist for most major markets, the causes are not clearly defined. Nevertheless, some suggested explanations for these observed seasonalities are highlighted. An outline of the data and methodology is presented. The limitations of the data for Japan and Taiwan and the implication on empirical results are discussed. Given the biased data set for these two countries, there is a tendency to overstate Monday mean returns. A test using unbiased data was conducted for Japan. A shift in the distribution of the mean returns across the day of the week was observed. The relationship between changes that occur in stock prices from one trading day to the next is examined to study the weekend effect. Two hypotheses were tested to determine whether the stock returns generating process is defined by trading time or calender time. Results indicate that both hypotheses do not explain the stock returns process for all markets except Taiwan. Subsequently, mean returns for the holiday effect across the day of the week were examined for the five indices to determine whether the weekend effect is the outcome of some 'closed market' effect. Based on the empirical findings, the day-of-the-week effect, January effect and holiday effect are significant for the majority of the markets. Other seasonal effects namely the turn-of-the-month effect, monthly effect and the turn-of-the-quarter effect are generally not significant for the markets analysed. These findings suggest that the market, based on their respective index, may not be efficient. Some possible suggestions for these observed seasonal patterns are discussed in the concluding chapter.