Title：Investment Efficiency in China: A Semiparametric Stochastic Efficient Investment Model with Panel Data

　　　　 Schedule: Nov 23, Wed 1:30-2:30 PM

　　　　 Location: 腾讯会议 457 413 332

　　　　 Introduction：王钛宁，经济学博士，首都经济贸易大学国际经济管理学院副教授。研究方向为非参数计量经济学与随机边界模型。已有研究发表在Econometric Theory, Empirical Economics, The World Economy, Economics Letters  等国际学术期刊。更多信息详见 https://isem.cueb.edu.cn/szdw/xyjs/90091.htm

　　　　 Abstract：We propose a semiparametric model for analyzing investment efficiency. We assume that a firm may make an inefficient investment decision due to under-investment (driven by financial constraints) or over-investment (driven by agency costs). Thus, a firm is investment efficient when it does not under- or over-invest. Since these investment decisions are not observed by market participants, we assume that probabilities of the decisions are influenced by financial friction variables. Our model has several advantages over the conventional models. First, we allow a firm to under-, over-, or efficiently invest; and separate under-investment from over-investment. This improves over existing models, which either explicitly assume the absence of agency costs or implicitly assume an infeasible scenario where under-investment and over-investment appear simultaneously. Second, our functions for efficient investment are nonparametric, alleviating the risk of making misleading inferences on investment efficiency due to model misspecification. Third, our model disentangles fixed effects from investment efficiency, and circumvents incidental parameter problems through model transformation. We perform investment efficiency analysis  using a panel of Chinese listed firms during 2006-2020. We document significant nonlinear effects of Tobin’s q and sales on investment, which are largely masked by parametric models in the literature. We also find significant over-investment by the state-owned enterprises, and under-investment by the non-state-owned enterprises. These over/under-investment are, on average, found to be persistent which has increased over the past 15 years.