A0328
Title: Threshold models for high-dimensional nonlinear time series
Authors: Chi Tim Ng - Hang Seng University of Hong Kong (Hong Kong) [presenting]
Yan Wu - Hang Seng University of Hong Kong (Hong Kong)
Yuanbo Li - University of International Business and Economics (China)
Abstract: The threshold autoregressive (TAR) model is an important class of nonlinear time-series models and has attracted great attention in the literature. To apply the TAR model in high-dimensional settings, a threshold network autoregressive (TNAR) model is proposed to overcome the over-parameterisation difficulty by exploiting the network relations' available information. The sufficient conditions for the strict stationarity and the ergodicity of the TNAR model are established. A computationally efficient method is developed to estimate the multiple thresholds and the parameters. Grouped TNAR model is also proposed to reduce the number of parameters further. The asymptotic behaviour of the adaptive fused LASSO is explored, and the estimation consistency of both numbers of groups and group membership structure is established. The applicability of the proposed models is illustrated with the U.S. stock data.
