Credit cycle identification: A Markov-switching application

A Work Project, presented as part of the requirements for the Award of a Masters Degree in Finance from the NOVA – School of Business and Economics

This project aims to study credit dynamics and to identify phases of credit cycles at the country level. We applied a Markov-switching (MS) autoregressive framework and a MS with regime-invariant macroeconomic variables to a broad concept of credit, domestic credit. We used a sample of 10 developed countries. MS identification power is assessed using smooth probabilities of low growth states, collected as a by-product of models estimation, against historical databases of crisis events. Conclusions support that MS is accurate in identifying credit cycle phases, and that domestic credit is a good variable for such identification. Additionally, Credit Gap, excess growth over GDP and Broad Money contribute positively to the MS predictions.