South-East Asian Stock Markets Follow A Non-Random Walk

Venus, Khim-Sen Liew and Lim, Kian-Ping and Choong, Chee-Keong (2004) South-East Asian Stock Markets Follow A Non-Random Walk. [Working Paper]

PDF south.pdf Download (883kB)

Abstract

Are the returns of the major South-East Asian stock markets forecastable? This study resorts to time series models that estimate the current value of a variable solely based on its historical records. Technical analysis that explore, for instance, patterns such as head-and-shoulders are viewed as some kinds of non-linearity in the financial time-series. As such, linear and non-linear models are considered in this study, together with the benchmark random walk model for forecastability comparison. Stock market returns of five major South-East Asian countries over the sample period ranges from January 1990 to October 2000 are examined in this study. Two linear and four non-linear models are estimated to generate 1-day, 1-week, 1month, 3-month, 9-month and 1-year forecasts. Results obtained shows that the random walk model ranked last in terns of forecast accuracy performance in all cases. Hence, it can be concluded that these stock market returns follow a non-random walk and are forecastable by time series models.

Actions (For repository members only: login required)

View Item