The application of time series models to the Korean stock market

Abstract

Autoregressive Moving Average(ARMA) models are stationary time series models which contain broad class of parsimonious time series processes found useful in describing various time series. Autoregressive Integrated Moving Average(ARIMA) models are widely used nonstationary time series models which improve nostationarity in the mean of ARMA model. Generalized Autoregressive Conditional Heteroscedasticity(GARCH) models are good for the regression analysis of nonconstant error variance. In this thesis, we use ARIMA model and ARIMA-GARCH model for regressing and forecasting financial time series, such as KOSPI 200 index.