This document discusses building and forecasting ARIMA models on imports and exports of Pakistan from 1947-2013. It finds that an ARIMA(2,2,2) model fits imports and an ARIMA(1,2,2) fits exports. Time series analysis is important for forecasting and planning. The document outlines the ARIMA modeling process including identifying stationary patterns in the data using autocorrelation and partial autocorrelation functions to select appropriate ARIMA models.