CMStatistics 2017: Start Registration
View Submission - CMStatistics
Title: Dynamic semiparametric factor model with structural breaks Authors:  Weining Wang - City U of London (United Kingdom) [presenting]
Wei Biao Wu - University of Chicago (United States)
Likai Chen - Washington University in Saint Louis (United States)
Abstract: A dynamic semiparametric model with structural factor dynamics for structral breaks is considered. The observations are described by a few low dimensional factors with time invariable functions of covariates as loadings. This framework could incorporate both discontinuity (structural breaks)in space and in time. The discontinuity in space is to account for possible known spatial discrepancies, while the unknown structural break in time models the regime switching effects introduced by exogenous shocks over time. We provide identifiability condition for our factor processes with structural breaks. We develop an estimation procedure allowing to estimate and make inference on the break point in time. The procedure's precision is evaluated via a simulation study. Finally we show two empirical illustrations on modeling the dynamics of the minimum wage in China and the dynamics of trading volume from a limit order book dataset.