Scholarlink

<p>This paper introduces a panel stochastic frontier model that accounts for unobserved factor structures to capture cross-sectional dependence and separate “decision-independent” from “decision-dependent” heterogeneity. By applying a transformation to remove time-varying heterogeneity, we estimate parameters using maximum likelihood while preserving the scaling property, thus enabling an accurate identification of inefficiency. Monte Carlo simulations confirm the method’s advantages over conventional within-transformation estimators, notably in mitigating bias and mean squared errors. An empirical application to US banks from 1994 to 2007 illustrates that overlooking unobserved common shocks can distort efficiency estimates before the financial crisis. Overall, our findings underscore the importance of distinguishing between time-varying and time-invariant heterogeneity, carrying significant implications for policy aimed at fostering stability and competition in financial markets.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Panel stochastic frontier model with multi-factor error structure: an application on bank efficiency

  • 期刊论文

  • 发表日期：

    2026年2月20日

查看全文
  • Chih-Chiang Hsu,
  • Chang-Ching Lin,
  • Shou-Yung Yin

摘要

This paper introduces a panel stochastic frontier model that accounts for unobserved factor structures to capture cross-sectional dependence and separate “decision-independent” from “decision-dependent” heterogeneity. By applying a transformation to remove time-varying heterogeneity, we estimate parameters using maximum likelihood while preserving the scaling property, thus enabling an accurate identification of inefficiency. Monte Carlo simulations confirm the method’s advantages over conventional within-transformation estimators, notably in mitigating bias and mean squared errors. An empirical application to US banks from 1994 to 2007 illustrates that overlooking unobserved common shocks can distort efficiency estimates before the financial crisis. Overall, our findings underscore the importance of distinguishing between time-varying and time-invariant heterogeneity, carrying significant implications for policy aimed at fostering stability and competition in financial markets.