NONPARAMETRIC ESTIMATION OF SIBUYA'S MEASURE OF LOCAL DEPENDENCE FOR TIME SERIES

Author(s):	Latif, Sumaia A. ; Morettin, Pedro A. Total Authors: 2
Document type:	Journal article
Source:	ADVANCES AND APPLICATIONS IN STATISTICS; v. 23, n. 1, p. 27-pg., 2011-07-01.
Abstract
The measures of local dependence are calculated on a function of the support of the variables, and may provide more information about the structure of dependence than the global coefficients which result in a single numerical value. In this paper, we study the function of Sibuya [11] in the context of stationary stochastic process both in the univariate and bivariate cases. We rewrite this function in terms of copula studying its properties. Two kernel smoothed estimators are proposed and their properties are derived. Monte Carlo experiments considering a stationary vector autoregressive process of order one and a Gaussian process are performed. Empirical illustrations are done using the series of daily returns of CAC 40 and DAX and of IBOVESPA (the Brazilian index). (AU)

FAPESP's process:	08/51097-6 - Time Series, Dependence Analysis and Applications
Grantee:	Pedro Alberto Morettin
Support Opportunities:	Research Projects - Thematic Grants