Easydata-MD: A Monthly Dataset for Macroeconomic Research on Pakistan

Abstract

This paper introduces the monthly State Bank of Pakistan’s EasyData, for conducting empirical macroeconomic analysis and forecasting for Pakistan's economy. For this purpose. We perform a forecasting exercise using the conventional econometric models and the most recent machine-learning algorithms. We find that the machinelearning models outperform the benchmark and regression models based on observed factors. Furthermore, the dataset has a higher ability to predict the external variables, a possible outcome of Pakistan's economy and its persistent balance of payment problem. The focus of policy has been to address this issue.