Abstract

In recent years the study of regime shifts or structural breaks behaviour in time series has gained much attention. This is due to realization that many economic time series undergo episodes in which the behaviour of the series change quite dramatically as a result of financial crises or abrupt changes in the government policy. The paper use two difference approaches to capture the possibility of structural breaks in KLCI index of Bursa Malaysia between 1977 and 2008. The first approach is by using the Markov switching model where the movement between regimes or regime shifts are unrelated to the past observations of the process and enables probabilistic statements to be made about the likelihood of the series being in a particular regime in any time period. While the second approach is via the wavelets method where a time series is transform using a particular wavelet basis functions. The transform series is well localized in both the time (position) and the frequency (scale) domain. Hence, the study can detect precisely a sudden change in the data. Finally, the study compares the results from the two approaches and discusses the advantage and disadvantage of each approach. Several numerical results will be presented.

Key words: Structural breaks, financial time series, Markov switching model, wavelets transform.