The South East Asia Capital Markets: 1995-2015

Abstract

[eng] The South-East Asia region is an increasingly economically important region due to its large population and development potential. The region has however experienced phases of substantial economic turmoil such as for instance the South-East Asia financial crisis of the 90’s. This dissertation analyses the short-term and long-term impacts of the financial crisis from a stock market point of view. This dissertation adds to the existing literature by focusing on the equity market rather than on the foreign exchange market which is the area covered in most of the existing literature. The results shows that the South-East Asia crisis was a rather complex event with quantitatively distinct phases. Granger causality and adjusted volatility analysis were also carried out. In both cases controlling for preexisting relations. The analysis shows that the South-East Asia financial crisis was a rather complex event, perhaps more complex than it is normally assumed, with dynamic interactions among the equity markets of the countries/jurisdictions analyzed. It will be shown that there was no country/jurisdiction that consistently drove the performance of the other countries/jurisdictions in the region. It will be also shown that the importance of some equity markets in the region shifted with for instance the equity market of Thailand becoming less regionally important and other countries, such as South Korea, becoming more important compared to the pre-crisis period. It was also analyzed the impact of the legal system in the performance of the equity markets in the region. Most of the analyzed countries/jurisdictions analyzed, with the noticeable exception of Thailand, were colonized and the colonizing country tended to impose their own legal system. Three groups of major legal systems were analyzed including the English, French and German legal systems. Typically in the existing literature there is a four group usually called the Scandinavian system. However, this group was not included because Scandinavian countries did not colonize South-East Asia. The results suggest that the type of legal system has a statistically significant impact on equity performance. The results also suggest, but with less statistical robustness, that the English system appears to have an advantage, from an equity market performance point of view, compared to the French and German. The analysis was carried out using classical econometric models, controlling for several drivers of the stock market performance, as well as using a more systematic approach for model factor selection, using a Lasso algorithm. The Lasso regression automatically choses which drivers to use (from a pool of drivers) for a model. In this way the driver selection is more objective. Finally, it was proposed an approach to try to detect Black Swan events such as financial crisis. The algorithm automatically selects the parameters of the forecasting algorithm used. For example, the length of the training data and the number of neurons in a neural network but can be extended to other forecasting techniques. This automated approach presents two advantages. First, it avoids the risk of biased model selection. After a financial crisis has happened it is tempting to find a quantitative model that (a posteriori) is able to detect the crisis, such as for instance changing the length of the training dataset until the model fits the data. Second, it is also allows for comparison among techniques that might require different parameter selections, such as the above mentioned length of the training data.