A0950
Title: Using Kendall Tau to assess the excess co-movement in time series
Authors: Ying Zhang - Acadia University (Canada) [presenting]
Will Sutherland - Acadia University (Canada)
Abstract: Kendall Tau is first introduced as an accordance measure between two stationary time series, and then two methods are discussed to estimate its variance and construct its confidence interval. Motivated by an economic phenomenon, the prices of commodities have a persistent tendency to move together; the proposal is to use the Kendall Tau confidence interval method to assess the excess co-movement between two-time series. To assess the excess dependence of two series, one needs to work with the residuals after modeling the trend and other effects, which provides the ideal conditions for our proposed confidence interval construction. The use of this method to investigate the excess co-movement of seven commodities from 1990-2020 is demonstrated.
