EcoSta 2021: Start Registration
View Submission - EcoSta2021
A0161
Title: Estimation of a partially linear seemingly unrelated regressions model: Application to a translog cost system Authors:  Xin Geng - Nankai University (China)
Kai Sun - Shanghai University (China) [presenting]
Abstract: A partially linear seemingly unrelated regressions model is studied to estimate a translog cost system that consists of a partially linear translog cost function and input share equations. A simple and feasible estimation procedure is proposed. We show that both our parametric and nonparametric component estimators are consistent, asymptotically normal, and more efficient relative to the single-equation counterparts. We highlight that the relative efficiency gain of the nonparametric estimator for a particular equation, based on the Cholesky decomposition, improves with its position in the system and is maximized when this equation is placed at the end. A model specification test for parametric functional forms is proposed, and how to correct the between- and within-equation heteroscedasticity is also discussed. An Italian banking data set is used to estimate the translog cost system, yielding policy implications for risk management in banking.