项目名称: 非线性协整模型的有效估计、检验及其应用
项目编号: No.71201137
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 管理科学与工程
项目作者: 陈海强
作者单位: 厦门大学
项目金额: 22万元
中文摘要: 本课题研究非线性协整模型的估计方法、检验理论及其在经济建模和金融预测方面的应用。传统的线性协整模型假设协整关系为固定的线性形式,但很多经验证据表明上述假定与许多实际经济现象不符。因此,扩展线性协整理论,将非平稳和非线性结合起来,成为协整理论发展的主流方向。我们的研究通过允许协整关系为非线性甚至非参数模型,使得协整分析在模型设定方面更加灵活。我们的主要任务包括:①在考虑模型内生性和残差序列相关性等前提下,发展出一个有效估计方法来估计非线性协整模型;②设计一个非参数统计量来对非线性协整模型进行假设检验;③提出一个非参数单位根检验来区分非线性平稳序列与单位根序列,并进一步检验非线性协整关系。此外,本课题将考虑两个实证研究:①检验货币政策中性假说是否在中国成立,并考察货币政策对长期实质经济增长是否存在非线性影响;②通过检验市盈率等非平稳变量能否预测市场回报率来考察中国金融市场的有效性假说。
中文关键词: 时间序列;门限协整;非参数检验;有效估计;模型检验
英文摘要: We consider the estimation and inference for nonlinear cointegration models and their applications in economic modeling and financial forecasting. Traditional cointegration analysis assumes linear structure. However, this assumption is rejected by much empirical evidence. Hence,it is important to extend linear cointegration to nonlinear cointegration. By allowing for nonlinear terms or even nonparametric parts, our model enjoys more flexibility compared to traditional linear cointegration models. Our main tasks include: ①designing an efficient estimation method for nonlinear cointegration models with endogeneity and serially correlated error terms; ②proposing a nonparametric test statistic for nonlinear cointegration model specification; ③developing a nonparametric unit roots test to distinguish nonlinear stationary processes from unit root processes, and check the existence of the nonlinear cointegration. Moreover, two empirical applications will be considered: ① testing whether the monetary neutral hypothesis holds in China and whether the monetary supply generates a nonlinear impact to the long run real GDP growth; ②checking whether the efficient market hypothesis holds for Chinese financial markets by examining whether PE ratio, PD ratio and other highly persistent variables are useful in predicting market r
英文关键词: Time Series;Threshold Cointegration;Nonparametric Testing;Efficient Estimation;Model Specification Testing