Thesis (M.Com (Economics)) -- University of Limpopo, 2018 / This study investigates the impact of capital flight and investment on economic growth in South Africa using time series data from 1986 to 2016. It employs the Auto Regressive Distributed Lag (ARDL) bounds testing procedure and the Granger causality test as a method of analysis. The empirical findings reveal that the variables are cointegrated which is an indication of the existence of a long run relationship among them. It was further discovered that capital flight had a negative long run relationship with economic growth while investment showed a positive long run relationship with economic growth. The terms of trade and inflation which were added to the model as control variable were also found to have a significantly positive influence on economic growth. The Granger causality indicated a bidirectional relationship between inflation and economic growth, while the terms of trade is found to have a unidirectional relationship with economic growth and capital investment respectively. The results are in line with the neo-classical growth model and the accelerator theory of investment.

Seminal papers of Solow (1957) and Swan (1956) stimulated debate among economists on the role of technical change in productivity improvements and for that matter economic growth. The consensus is that technological change accounts for a significant proportion of gross national product (GNP) growth in industrialised economies. In the case of Australia, the aggregate productivity performance was poor in the 1970s and 1980s, but picked up very strongly by the 1990s, and was above the OECD average growth level for the first time in its productivity growth history. However, this high productivity growth rate could not be sustained and Australia started to experience a slowdown in productivity growth since 2000. This study empirically measures the performance of productivity in Australia’s economy for the period 1950-2005, using an econometric approach. Time-series data are used to develop econometric models that capture the dynamic interactions between GDP, fixed capital, labour units, human capital, foreign direct investment (FDI) and information and communication technology (ICT). The Johansen (1988) cointegration techniques are used to establish a long-run steady-state relation between or among economic time series. The econometric analysis pays careful attention to the time-series properties of the data by conducting unit root and conintegration tests for the variables in the system. / This study finds that Australia experienced productivity growth in the 1950s, a slow down in the mid 1960s, a very strong productivity growth in the mid 1990s and another slowdown from 2000 onwards. The study finds evidence that human capital, FDI and ICT are very strong determinants of long-run GDP and productivity growth in Australia. The study finds that the three, four and the five factor models are likely to give better measures of productivity performance in Australia as these models recognise human capital, FDI and ICT and include them as separate factors in the production function, This study finds evidence that the previous studies on the Australia’s productivity puzzle have made a very significant omission by not considering human capital, FDI and ICT as additional exogenous variables and by excluding them from the production function for productivity analysis.

A volatilidade é uma medida de incerteza quanto às variações dos preços de ativos. Este trabalho tem como objetivo analisar a volatilidade, através dos diversos modelos da família GARCH, de três índices de mercados financeiros: Dow Jones, IBovespa e S&P 500. Com este intuito, foram aqui utilizadas técnicas univariadas e multivariadas, bem como análises de Causalidade de Granger. Através das duas primeiras ferramentas, escolhemos o melhor modelo para cada um destes casos. Usando a terceira ferramenta, concluímos que o IBovespa é significativamente influenciado pela abertura do Dow Jones e do S&P500. Por outro lado, mostramos que a abertura do IBovespa não impacta, nem à 10% de significância, os índices Dow Jones e S&P 500. Também concluímos que a incorporação de um dos índices americanos ao modelo do IBovespa torna-o mais significativo, uma vez que o mercado acionário brasileiro é impactado pelos dois índices citados anteriormente. Desta forma, este trabalho mostra que os modelos GARCH multivariados aparentam ser mais eficazes na estimação da volatilidade de ativos financeiros do que os modelos GARCH univariados. / The volatility is a measure of the uncertainty of variations of asset prices. The main goal of this work is to analyze the volatility, by the use of several models of the GARCH family, of three financial market indexes: Dow Jones, IBovespa and S&P 500. With this purpose, we use univariate and multivariate techniques, as well as Granger Causality. Using these first two tools, we choose the best model for each one of these cases. Using the third tool, we conclude that the IBovespa is significatively influenced by the opening of the Dow Jones and the S&P 500 indexes. On the other hand, we show that the opening of the IBovespa does not impact, not even at 10% of significance, the Dow Jones and S&P 500 indexes. We also conclude that incorporation of one of these American indexes to the model involving IBovespa makes it more significant, once the Brazilian Stock Market is impacted by the two American indexes we mention before. This work shows that multivariate GARCH models seem to be more efficient in the volatility estimation of financial assets than univariate GARCH models.

A volatilidade é uma medida de incerteza quanto às variações dos preços de ativos. Este trabalho tem como objetivo analisar a volatilidade, através dos diversos modelos da família GARCH, de três índices de mercados financeiros: Dow Jones, IBovespa e S&P 500. Com este intuito, foram aqui utilizadas técnicas univariadas e multivariadas, bem como análises de Causalidade de Granger. Através das duas primeiras ferramentas, escolhemos o melhor modelo para cada um destes casos. Usando a terceira ferramenta, concluímos que o IBovespa é significativamente influenciado pela abertura do Dow Jones e do S&P500. Por outro lado, mostramos que a abertura do IBovespa não impacta, nem à 10% de significância, os índices Dow Jones e S&P 500. Também concluímos que a incorporação de um dos índices americanos ao modelo do IBovespa torna-o mais significativo, uma vez que o mercado acionário brasileiro é impactado pelos dois índices citados anteriormente. Desta forma, este trabalho mostra que os modelos GARCH multivariados aparentam ser mais eficazes na estimação da volatilidade de ativos financeiros do que os modelos GARCH univariados. / The volatility is a measure of the uncertainty of variations of asset prices. The main goal of this work is to analyze the volatility, by the use of several models of the GARCH family, of three financial market indexes: Dow Jones, IBovespa and S&P 500. With this purpose, we use univariate and multivariate techniques, as well as Granger Causality. Using these first two tools, we choose the best model for each one of these cases. Using the third tool, we conclude that the IBovespa is significatively influenced by the opening of the Dow Jones and the S&P 500 indexes. On the other hand, we show that the opening of the IBovespa does not impact, not even at 10% of significance, the Dow Jones and S&P 500 indexes. We also conclude that incorporation of one of these American indexes to the model involving IBovespa makes it more significant, once the Brazilian Stock Market is impacted by the two American indexes we mention before. This work shows that multivariate GARCH models seem to be more efficient in the volatility estimation of financial assets than univariate GARCH models.

A volatilidade é uma medida de incerteza quanto às variações dos preços de ativos. Este trabalho tem como objetivo analisar a volatilidade, através dos diversos modelos da família GARCH, de três índices de mercados financeiros: Dow Jones, IBovespa e S&P 500. Com este intuito, foram aqui utilizadas técnicas univariadas e multivariadas, bem como análises de Causalidade de Granger. Através das duas primeiras ferramentas, escolhemos o melhor modelo para cada um destes casos. Usando a terceira ferramenta, concluímos que o IBovespa é significativamente influenciado pela abertura do Dow Jones e do S&P500. Por outro lado, mostramos que a abertura do IBovespa não impacta, nem à 10% de significância, os índices Dow Jones e S&P 500. Também concluímos que a incorporação de um dos índices americanos ao modelo do IBovespa torna-o mais significativo, uma vez que o mercado acionário brasileiro é impactado pelos dois índices citados anteriormente. Desta forma, este trabalho mostra que os modelos GARCH multivariados aparentam ser mais eficazes na estimação da volatilidade de ativos financeiros do que os modelos GARCH univariados. / The volatility is a measure of the uncertainty of variations of asset prices. The main goal of this work is to analyze the volatility, by the use of several models of the GARCH family, of three financial market indexes: Dow Jones, IBovespa and S&P 500. With this purpose, we use univariate and multivariate techniques, as well as Granger Causality. Using these first two tools, we choose the best model for each one of these cases. Using the third tool, we conclude that the IBovespa is significatively influenced by the opening of the Dow Jones and the S&P 500 indexes. On the other hand, we show that the opening of the IBovespa does not impact, not even at 10% of significance, the Dow Jones and S&P 500 indexes. We also conclude that incorporation of one of these American indexes to the model involving IBovespa makes it more significant, once the Brazilian Stock Market is impacted by the two American indexes we mention before. This work shows that multivariate GARCH models seem to be more efficient in the volatility estimation of financial assets than univariate GARCH models.