FORECASTING STOCK MARKET VOLATILITY:
Tanweh, Chumbon (2008)
2008
Kuvaus
Opinnäytetyö kokotekstinä PDF-muodossa.
Tiivistelmä
The purpose of this thesis is to investigate and evaluate the GARCH, Threshold GARCH (GJR), Exponential GARCH (EGARCH), GARCH based on student t distribution and GARCH based on generalized error distribution models on equity volatility forecast at index level on the OMX Helsinki all share price index (OMXH) and OMX Helsinki Cap price index (OMXHCAP) in the Helsinki stock exchange.
The overall return series of both markets are consistent with empirical results of other financial markets that is why the symmetric and asymmetric GARCH class models are explored to predict the daily volatility of the two indices. Dynamic and Static in-the sample and out- of the sample forecast was introduced. Three statistics error were used, the root mean square error (RMSE), the mean absolute error (MAE) and the mean absolute percentage error (MAPE) were used, with the Theil inequality coefficient, and with the implementation of three different proportions that is the bias, the variance and the covariance proportions the main conclusions were arrived at.
EGARCH came out to be the best model in forecasting both markets because of its capability of capturing asymmetric in both indices.The evaluation results for in-sample and out-sample forecast were consistent in both markets. The GARCH was only able to forecast to some extend the static in-the sample for OMXH index; the TARCH forecast out-of the sample for OMXHCAP index partly. The GARCH (t) and the GARCH (GJR) came out to be the worse forecasting models in this thesis.
The overall return series of both markets are consistent with empirical results of other financial markets that is why the symmetric and asymmetric GARCH class models are explored to predict the daily volatility of the two indices. Dynamic and Static in-the sample and out- of the sample forecast was introduced. Three statistics error were used, the root mean square error (RMSE), the mean absolute error (MAE) and the mean absolute percentage error (MAPE) were used, with the Theil inequality coefficient, and with the implementation of three different proportions that is the bias, the variance and the covariance proportions the main conclusions were arrived at.
EGARCH came out to be the best model in forecasting both markets because of its capability of capturing asymmetric in both indices.The evaluation results for in-sample and out-sample forecast were consistent in both markets. The GARCH was only able to forecast to some extend the static in-the sample for OMXH index; the TARCH forecast out-of the sample for OMXHCAP index partly. The GARCH (t) and the GARCH (GJR) came out to be the worse forecasting models in this thesis.