TIME SERIES MODELING AND VOLATILITY ANALYSIS OF EXPORTS OF PAKISTAN BY USING ARCH AND GARCH MODEL

Abstract

In this paper the identification and estimation of the mean and variance components of the monthly exports of Pakistan was done by using mean model ARIMA and variance model GARCH. Data, on monthly export of Pakistan, which are in million of US-dollar, was obtained from State Bank of Pakistan, covering the period January, 1972 to June, 2016.The residuals obtained under the best selected mean and variance is explained by the volatility structure. By applying the OLS methodology the parameters of ARIMA type simple specifications are normally estimated. The presence of ARCH effects or volatility has two disadvantages, the autocorrelation in error term and the presence of ARCH effect. This problem may be determined by employ the ARCH or GARCH variance models. In addition, the forecasting pattern for of exports of Pakistan might be helpful to suggest planners for effective future planning to attain sustainable economic growth.