Revisiting the Economic Crisis after a Decade: Statistical and Machine Learning Perspectives

Abstract

This paper examines the 2008-09 crisis experienced by the world economy as one of the largest of its kind in four decades. Based on ample research published in the decade since, a simple statistical correlation coefficient is applied to selected variables, and a principal component analysis is conducted for significant indicators that an economic crisis sends as signals of its imminence. Recovery from the crisis is also analyzed not only for the reinstatement of original relationships but also for new signs that falling into a crisis and recovering from it may take quite different paths.