Bayesian Estimation of a Pragmatic Model for Monetary Policy Analysis: The Case of Pakistan

We present Bayesian maximum likelihood estimation of quarterly projections model for Pakistan, as documented in Ahmad & Pasha (2015). Estimation results based on quarterly data from 2001Q3 to 2023Q1 show substantial differences in values of estimated versus calibrated parameters related to aggregate demand, aggregate supply,monetary policy rule and exogenous shock processes. Comparison of pseudo out-ofsample forecasting performance for key macro variables shows that estimated model provides more precise forecasts in case of headline inflation, real GDP growth, interest rate and exchange rate over 8-quarters forecast horizon. We use the estimated model for gap analysis and scenario analysis. Gap analysis, based on March 2023 data, shows that Pakistan is passing through a recession with overshot exchange rate. In scenario analysis, we incorporate implications of political instability, climate risks, commodity prices and global financial conditions for next three years’ forecasts of domestic variables under baseline and alternate scenarios. Results of scenario analysis, which may be utilized for macro stress testing exercise, show that simultaneous realization of assumed risk factors may lead to substantial deterioration of macroeconomic outlook. We explore different monetary and fiscal policy options to counter the recent crisis.Our results show that under current circumstances, using an expansionary monetary policy may lead to substantial rise in inflation and macroeconomic volatility without offering sustainable gains in GDP growth.