A Non-linear Analysis of Cointegration in South-East Asian Equity Markets

Sammanfattning: This paper investigates the presence of cointegration among the main stock markets in South-East Asia, namely those of Hong Kong, Singapore, Malaysia and Thailand. Part 1 of the thesis studies the relationship using Markov-switching models, while Part 2 uses regime-shifting models with one structural break. It takes into consideration two databases: one with indices in local currency, while the other contains USD-adjusted values. For the Markov-Switching model, in the USD-adjusted dataset, the markets are not cointegrated in the bear market regime, but they show evidence of cointegration in the bull market regime. For the local currency database, cointegration is present in both low-volatility and high-volatility regimes, but the strength of the relationship is stronger in the low-volatility regime. For the regime-shifting model, in the USD-adjusted database there is evidence of cointegration given a structural break on the 7th September 2007, which coincides with the outbreak of the Global Financial Crisis. The role of cointegration appears to be weaker after the crisis. For the local currency dataset, there is evidence in favor of cointegration given a structural break on the on the 9th January 2015, which coincides with the Chinese market bubble, and the strength of the relationship has strengthened thereafter.