Predicting recessions: analysing historical variables to forecast future economic downturn

Abstract

Recessions have always posed a significant risk to economic stability, policy planning and financial well-being. From decades to decades, economists have been studying a wide range of historical indicators, from GDP growth and inflation to yield curves and global financial markets, We intend to attempt to predict these downturns with greater accuracy. With the near advancement of machine learning and predictive analysis, the traditional ways have evolved into a data-centric and driven dynamic field of making educated predictions. This paper intends to investigate the various economic variables, deeply interlinked and often co-dependent, in contributing to predicting recession in the Indian context. The project's primary objective is to examine the performance of various economic factors and forecasting methods over time by analyzing two significant historical recessions: the Global Financial Crisis (2008) and the COVID-19-induced slump (2020). The objective of our critical analysis of these models' advantages and disadvantages is to draw attention to systemic biases, data issues, and methodological developments that affect recession prediction. The ultimate goal of our study is to support the ongoing development of accurate and trustworthy recession forecasting technologies. To help economists, policymakers, and financial institutions predict and react to future economic downturns, we hope to close the gap between classic economic analysis and contemporary data-driven methodologies.