Dating Business Cycle Turning Points for the French Economy: a MS-DFM approach

2014.

Doz C., Petronevich A.

The purpose of this paper is to develop a tool that would permit to identify the turning points of the business cycle of the French economy in a more timely and transparent manner than the existing institutions, such as OECD, do. We use the basic two-regime Markov Switch- ing Dynamic Factor Model and estimate it in one step and two steps. Both methods proved to work well on the 1993-2014 monthly data sample. We determine the major economic series that perform well under information-constrained one-step method and the best specication for the information-rich two-step model. We also nd that inclusion of certain series leads to false signals of recession as dened by OECD. We determine the core economic series for the one-step method that do not produce these signals, and therefore provide an instrument that is able to reproduce OECD dating with a less lag. The chronology produced by the two-step method necessarily describes a more comprehensive notion of economic activity, where the false signals are justied as correct alarms of a possibly beginning recession, which fortunately did not happen

Keywords: business cycles, Markov-switching, nowcasting, Dynamic Factor Models