Semiparametric Factor Models with Time-varying Covariates with Applications in Asset Pricing

Chen Qihui received his Ph.D in Economics from the University of California, San Diego in 2017. Prior to that, he earned Bachelor’s degrees from Xiamen University and Master’s degree from both Xiamen University and Singapore Management University. He joined the School of Management and Economics of CUHK-Shenzhen in July 2017. His research interests include econometric theory and applied econometrics.

This paper provides a trustworthy and simple method for estimation and inference on semiparametric factor models with time-varying covariates. Specifically, we establish asymptotic properties of the estimators for the intercept function, the factor loading functions, the unobserved factors, and the number of factors. In particular, we establish a strong approximation for the distributions of the estimators of the intercept function and the factor loading functions. We also develop a bootstrap inference for testing their significance and linearity. The results do not require large T or smoothness of covariates over time, and therefore have a wide application in asset pricing.