A dynamic factor model with regime switching is proposed as an empirical characterization of business cycles. The approach integrates the idea of comovements among macroeconomic variables and asymmetries of business cycle expansions and contractions. The first is captured with an unobservable dynamic factor and the second by allowing the factor to switch regimes. The model is estimated by maximizing its likelihood function and the empirical results indicate that the combination of these two features leads to a successful representation of the data relative to extant literature. This holds for within and out-of-sample and for both revised and real time data. 1 This paper was written with financial support from CNPq -Brazilian Council for Scientific and Technological Research. This material is based on my doctoral dissertation from the University of Pennsylvania. I am grateful to my advisor Frank Diebold for his invaluable advice during all stages of this research. I also thank James Hamilton for helpful suggestions. The author bears full responsibility for any errors.
This paper evaluates the ability of formal rules to establish U.S. business cycle turning point dates in real time. We consider two approaches, a nonparametric algorithm and a parametric Markov-switching dynamic-factor model. In order to accurately assess the real-time performance of these rules, we construct a new unrevised "real-time" data set of employment, industrial production, manufacturing and trade sales, and personal income. We then apply the rules to this data set to simulate the accuracy and timeliness with which they would have identified the NBER business cycle chronology had they been used in real time for the past 30 years. Both approaches accurately identified the NBER dated turning points in the sample in real time, with no instances of false positives. Further, both approaches, and especially the Markovswitching model, yielded significant improvement over the NBER in the speed with which business cycle troughs were identified. In addition to suggesting that business cycle dating rules are an informative tool to use alongside the traditional NBER analysis, these results provide formal evidence regarding the speed with which macroeconomic data reveals information about new business cycle phases.
We compare forecasts of recessions using four different specifications of the probit model: a time invariant conditionally independent version; a business cycle specific conditionally independent model; a time invariant probit with autocorrelated errors; and a business cycle specific probit with autocorrelated errors.The more sophisticated versions of the model take into account some of the potential underlying causes of the documented predictive instability of the yield curve. We find strong evidence in favor of the more sophisticated specification, which allows for multiple breakpoints across business cycles and autocorrelation. We also develop a new approach to the construction of real time forecasting of recession probabilities.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.