This thesis presents exploratory study on the short term forecasting methods applicable to the time series of exports of Korea.
Several forecasting methods, such as Census II, Winters', Box-Jenkins' univariate and transfer function models, are applied to the data and the performance of each forecasting technique is evaluated in terms of forecasting error.
An empirical but pragmatic forecasting method is developed taking into account the relations between the export amount and L/C and the procedure is shown to be conceptually simple and satisfactory.
The empirical results indicate that the methods which adopt the amount of L/C (letters of credit) arrivals as a leading indicator are substantially profitable.