A0632
Title: Adaptive group fused Lasso for panel threshold model with cross-sectional dependence
Authors: Lulu Wang - University of Southampton (United Kingdom) [presenting]
Zudi Lu - University of Southampton (United Kingdom)
Abstract: Panel threshold regression has been one of the most popular methods in nonlinear panel time series analysis. The most common method to determine the number of threshold parameters is using a bootstrap procedure to approximate the sampling distribution under the assumption of cross-sectional independence, which may work poor in dealing with the strong dependence on most climate and finance data. We consider estimation of panel threshold model where both regressors and residuals are allowed to be cross-sectional dependent via adaptive group fused Lasso. We show that with probability approaching one, the proposed method can correctly determine the unknown number of threshold parameters and estimate regression parameters consistently. We establish the asymptotic theories of the Lasso estimators of regression coefficients. Simulation studies demonstrated that the proposed estimation method works well in finite samples under cross-sectional dependent conditions. We finally apply our Lasso estimation method to study the effect of precipitation on the stocks in FTSE 100.
