Time-Series Model With Periodic Stochastic Regime Switching

This paper provides a historical chronology of economic activity in 16th- and 17th-century France that is based on wheat price series in Paris and Toulouse. A stochastic regime-switching model enables us to benchmark eras and summarize the salient features of a development difficult to appraise in all its complexity. A new class of Markov regime-switching time-series models is introduced to allow for nontrivial interdependencies between different types of cycles that make the economy grow at an unsteady rate. With a predominantly agricultural cycle, we uncover a strongly periodic Markov switching scheme for recorded wheat prices from the grain markets of Paris and Toulouse. Besides the periodic nature of the Markov chain, we also study whether a common factor determined the state of the economy in Paris and Toulouse or whether each series moved independently.