Financial liberalistion and breaks in stock market volatility

This paper proposes a new statistical procedure which aims at providing robust estimates of volatility around official liberalisation dates, by using data driven techniques to identify the number and timing of structural breaks in the variance dynamics of stock market returns. The paper illustrates the usefulness of the procedure by providing an empirical application that focuses on five East Asian emerging markets, all of which liberalised their financial markets in the late 1980s or early 1990s, namely (South) Korea, Malaysia, Philippines, Taiwan and Thailand. It is shown that (i) the detected breakdates in the volatility of stock market returns do not correspond to official liberalisation dates and (ii) the use of official liberalisation dates as breakdates is likely to result in inaccurate inference. By using data driven techniques to detect multiple structural changes a richer -and inevitably more accurate - pattern of volatility dynamics emerges in comparison to focussing on official liberalisation dates.