A0888
Title: A new test for checking stationarity in variance for nonlinear time series with a trend
Authors: Lei Jin - Texas A and M University-Corpus Christi (United States)
Li Cai - Zhejiang Gongshang University (China) [presenting]
Abstract: The assumption of constant variance is fundamental in numerous statistical procedures for time series analysis. A novel procedure, such as GARCH models and Markov-switching models, is introduced to assess the variance stationarity of nonlinear time series. Unlike others, the proposed test relies on systematic sampling via Walsh transformations. It is developed under the process with a nonconstant mean function. Asymptotic pairwise independence is established among various Walsh transformation coefficients, and a max-type statistic is defined. The asymptotic null distributions of the test statistics are obtained. Furthermore, the consistency of the proposed methods is established under a sequence of local alternatives, contingent upon additional conditions on the mean functions. Through a comprehensive simulation study, we evaluate the finite sample performance of the procedure and conduct comparisons with existing methodologies. The findings offer insights into analyzing financial time series data.
