文章摘要
马薇,曹晓舟.组合方法视角下复杂协整关系的非参数识别[J].数量经济技术经济研究,2022,(5):170-189
组合方法视角下复杂协整关系的非参数识别
The Nonparametric Identification of Complex Cointegration Relationship in the Perspective of Combined Method
  
DOI:
中文关键词: 协整关系  非参数方法  组合方法
英文关键词: Cointegration Relationships  Nonparametric Methods  Combined Method
基金项目:
作者单位
马薇 天津财经大学统计学院 
曹晓舟 天津财经大学统计学院 
中文摘要:
      研究目标:构建对复杂协整关系进行非参数识别时有效、可操作的组合方法。研究方法:综合考虑核估计方法、窗宽选择以及协整检验三类关键因素,提出待分析的多种组合方法,通过蒙特卡洛模拟和中俄两国购买力平价理论在有限样本下对各组合方法的性质进行对比分析。研究发现:“局部线性核估计(LL)+交错验证窗宽选择(CV)+方差比检验”组合方法在识别真实协整关系时扭曲水平(Size)低,ADF检验在识别虚假协整关系时检验功效(Power)高。研究创新:扩展了E-G协整检验步骤,以非参数组合方法的视角提取最优残差序列并对复杂协整关系进行有效验证。研究价值:利用组合方法为复杂协整关系的建模探索出一条可行之路并丰富其应用场景,也为中俄两国进一步推进经贸合作提供了理论依据。
英文摘要:
      Research Objectives: Construct an effective and operational combined method for nonparametric identification of complex cointegration relationships. Research Methods: Considering the three key factors of kernel estimation method, bandwidth selection and cointegration test, multiple combined methods need to be analyzed are proposed. The properties of these methods are verified by Monte Carlo simulation and the Purchasing Power Parity theory between China and Russia analysis. Research Findings: The combined method of “Local linear estimation + Cross Validation method+ Variance Ratio test” has good size properties in recognition of the real cointegration relationship, and the ADF test has good power properties in identification of the spurious cointegration relationship. Research Innovations: The E-G cointegration test is extended, the optimal residual sequence is extracted in the perspective of the combined method, and finally the cointegration relationship is effectively verified. Research Value: This paper explores a feasible way for modeling complex cointegration relationships in the perspective of combined method which enrich the application scenarios, and provides a theoretical basis for further promoting economic and trade cooperation between China and Russia.
查看全文