A1939
Title: The sequential estimation of quantile factor models
Authors: Long Yu - Shanghai University of Finance and Economics (China)
Manyao Shi - Shanghai University of Finance and Economics (China) [presenting]
Abstract: Quantile factor models are widely used to capture heterogeneous factor structures across different distributional quantiles. However, the original iterative quantile regression (IQR) algorithm estimates all factors simultaneously, which does not guarantee nestedness across different factor numbers. A sequential estimation framework for quantile factor analysis (SQFA) is proposed. Algorithmic convergence of the proposed procedure is established, showing that the objective function is monotonically non-increasing. The estimated factors and loadings converge in probability at the same rate as the simultaneous estimation approach. A rank-minimization criterion is embedded into the procedure to consistently determine the number of factors at each quantile. Monte Carlo simulations and an empirical application demonstrate that SQFA performs well in finite samples, offering both computational efficiency and reliable statistical performance.
