An Empirical Investigation of Economic Variables of Pakistan by Using Autoregressive Distributed LAG Model

Zafar, H., Arif, M., and Yaseen, M.
(2016) 14-th International Conference on Statistical Sciences, Karachi, Pakistan, 29, 203-214

In this research two main techniques of time series analysis, Co-integration and Autoregressive Distributed Lag (ARDL) model were discussed with the advantages of one technique over the other. This study used to investigate the relationship between five economic variables of Pakistan by using monthly data for the period 1975 to 2015. ARDL model was used to study the relationship between Exchange Rate on Consumer Price Index, Money Supply, Imports and Exports. This work inspects the Short Run (SR) dynamic and Long Run (LR) relationship with the aid of autoregressive distributed lag (ARDL) model. First apply Augmented Dicky-Fuller (ADF) test and Phillips-Peron (PP) Test for checking the stationary of variables. The second step was ARDL-Bound Test and it concluded that variables have long run relation with each other. After ARDL-Bound Test next step was selection of lag length for ARDL model. The diagnostic test inspected that model is good to fit. The ARDL model exposed significant and non-significant long run relation. The coefficient of ECM (-1) indicates how much of the disequilibrium in the short-run was fixed (eliminated) in the long-run.