The present article uses the dynamic factor model of Stock and Watson to construct a coincident index with a clear statistical foundation able to represent the level of activity of the processing industry of the state of Rio Grande do Sul. In addition to this linear model, we also employ a regime switching methodology in order to determine the asymmetry of the business cycle in the industry on a statewide basis, pointing out periods of economic growth and stagnation in this sector. This new indicator is compared with the industrial performance index developed by the Federation of the Industries of the State of Rio Grande do Sul (FIERGS). The results show that both linear and nonlinear models estimate components that are highly correlated, such as the weighted average index currently calculated by FIERGS.
Markov-switching; business cycle; coincident indicators; dynamic factor model
