项目名称: 半参数空间自回归模型的理论研究及应用
项目编号: No.71501119
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 管理科学
项目作者: 金飞
作者单位: 上海财经大学
项目金额: 18万元
中文摘要: 本项目首次提出半参数空间自回归(SP‐SAR)模型以刻画空间相关程度的非线性性,并提出半参数空间面板数据(SP‐SPD)模型以控制个体的异质性。这些模型相比空间计量经济学中的模型更为符合实际,填补了文献中的空白,有着广阔的应用前景。我们将从理论上严格证明SP-SAR 和SP-SPD 模型中参数估计量的相合性(consistency)和渐近分布,以及检验空间相关程度非线性性的检验统计量的渐近分布。这些分析和证明难度较大,有利于推动空间计量理论的发展。我们将会把跟我们模型相关的matlab 程序封装成软件包,以供其它理论与实证经济学研究学者(包括业界从业人员)使用,从而推动这些模型在学术与实际商业领域中的应用。研究中得出的理论结果将应用于研究中国经济增长问题和金融产品收益率。通过这些实证研究来验证空间相关程度是否确实是非线性的,从而为政策制定者、企业和个人提供科学的建议和指导。
中文关键词: 空间自回归;半参数;非线性;序列估计量;异质性
英文摘要: This project proposes semiparametic spatial autoregressive (SP-SAR) models to capture the nonlinearity in spatial dependence and also proposes semiparametic spatial panel data (SP-SPD) models to further control the individual heterogeneity for the first time. Compared with the existing models in the spatial econometric literature, our models are new and more realistic, which are expected to be applied widely. We shall prove rigorously in theory that the series estimators for the SP-SAR and SP-SPD models are consistent and asymptotically normal, and the test statistics that test for the nonlinearity of spatial dependence are asymptotically normal. These analyses and proofs are theoretically challenging and contribute to the developments of spatial econometric theory. We shall compile the matlab codes related to our models into software packages, so that they can be easily used by other theoretical and empirical economists including people in industry, promoting the applications of our models in academia and industry. The theoretical results will be applied to study China economic growth and the yield rates of financial products. In these empirical studies, we can investigate whether or not the spatial dependence is nonlinear, so that scientific suggestions and guidance can be provided to policy makers, companies and individuals.
英文关键词: spatial autoregressive;semiparametric ;nonlinear;series estimator;heterogeneity