Functional cointegration: definition and nonparametric estimation
Functional cointegration: definition and nonparametric estimation
复制标题
DOI：
10.1515/snde-2013-0083
复制
发表时间：
2014-12
影响因子：
0.8
通讯作者：
A. Banerjee;Jean-Yves Pitarakis
中科院分区：
文献类型：
--
作者：
A. Banerjee;Jean-Yves Pitarakis
摘要：我们正式定义了一个泛函协整的概念，通过一个泛函系数将两个时间序列的动力学联系起来。这是通过使用可和性概念来替代不再适合非线性动力学的I(1)性来实现的。我们随后介绍了一种估计未知泛函系数的非参数方法。该方法基于分段局部最小二乘原理，计算简单，易于实现。我们建立了它的一致性，并评价了它在有限样本中的性能。随后，我们通过一个应用程序来说明它的实用性，该应用程序通过一个情绪指标来探索股价与股息之间的联系。
Abstract We formally define a concept of functional cointegration linking the dynamics of two time series via a functional coefficient. This is achieved through the use of a concept of summability as an alternative to I(1)’ness which is no longer suitable under nonlinear dynamics. We subsequently introduce a nonparametric approach for estimating the unknown functional coefficients. Our method is based on a piecewise local least squares principle and is computationally simple to implement. We establish its consistency properties and evaluate its performance in finite samples. We subsequently illustrate its usefulness through an application that explores linkages between stock prices and dividends via a sentiment indicator.