EcoSta 2018: Registration
View Submission - EcoSta2018
Title: Uncertainty and business cycle asymmetry: An application of a serially-correlated two-tier SF model Authors:  Hung-Jen Wang - National Taiwan University (Taiwan) [presenting]
Yu-Fan Huang - Capital University of Economics and Business (China)
Sui Luo - Capital University of Economics and Business (China)
Abstract: A two-tier SF model is proposed where the inefficiency components are serially correlated. We show that by extending the frontier function to a stochastic trend, the model can be interpreted as an unobserved component model (UCM) of business cycle (BC) studies. A Bayesian estimation method is proposed where the result of a scaled mixture normal distribution is adopted to help the estimation. Compared to existing UCM in the BC literature, the new SF model is more intuitive and provides greater flexibility in modeling BC asymmetry. Using this model we are able to test three types of BC asymmetries: intrinsic, dynamic, and no asymmetries. To our knowledge, this is the first paper to discuss the different types of BC asymmetry in the literature. The model also allows us to study the correlation between the asymmetry and the macro/policy variables. In particular, we are interested in how economic uncertainties affect the asymmetry and the BC in general. Using measures of macro uncertainty and financial uncertainty as examples, we find that macro uncertainty is strongly linked to output deviation in the US and the effect is asymmetric. The impact of financial uncertainty on business cycles is less prominent.