Statistical and Structural Approaches to Estimate Output Gap and Inflation Forecast: The Case of Pakistan

Abstract

The output gap and inflation forecast are important factors to analyze current state of the economy and stance of monetary policy. In this study we have measured output gap using different statistical and structural methods namely the Linear Time Trend (LTT) method, Quadratic Time Trend (QTT) method, Hodrick-Prescott (HP filter), Band Pass Baxter-King Filter (BP), Double Exponential Smoothing Method (DES), Structural Vector Autoregressive (SVAR) method. For the analysis we have used quarterly data over the period 1960 to 2014 for Pakistan. Moreover, the inflation is forecasted with univariate and multivariate models. The results suggest that Quadratic Time Trend (QTT) method and Structural Vector Autoregression (SVAR) captures the history of Pakistan economy well. Whereas, output gap estimated through Structural Vector Autoregression (SVAR) generate better inflation forecast compared to other methods.