Applying Logistic Regression Models on Business Cycle Prediction

Abstract: The prediction of business cycles in the real economy is an important but challenging task. For this purpose we construct a logistic regression model that we use to predict turning points in the business cycles. With this approach we can determine probabilities of the economy being in a recession or an expansion in each time period. These probabilities are derived from a number of variables that should have an influence on the real economy. Our estimated model can predict business cycle turning points with a reasonable accuracy and could therefore be used as one potential tool for economic policy decision makers.