We examine the problem of predicting recessions from a machine learning perspective. We employ a number of machine learning algorithms to predict the likelihood of recession in a given month using historical data from a set of macroeconomic time series predictors. We argue that, due to the low frequency of historical recessions, this problem is better dealt with an imbalanced classification approach. We apply measures to compensate for class imbalance and use various performance metrics to evaluate and compare models. With these measures in place, ensemble machine learning models predict recessions with high accuracy and great reliability. In particular, a Random Forest model achieves a near perfect True Positive Rate within the historical training sample, generalizes extremely well to a test period containing 2008-2009 financial crisis, and shows elevated recession probabilities during the last few months of 2019, associated with the tightened macroeconomic environment and worsened by global pandemic.
The presentation will last approximately 40 minutes followed by 15-20 minutes for Q&A. Session will be recorded, and the recording link will be sent to those who have registered for the webinar.