Time-Series Modeling and Decomposition
LINEAR AND NONLINEAR TIME SERIES MODELS
Self-Exciting Threshold AutoRegressive (SETAR) Model
Another type of nonlinear time series models are the Self-Exciting Threshold AutoRegressive (SETAR) models introduced in a seminal paper by Tong and Lim. They are typically applied as an extension of autoregressive models, in order to allow for higher degree of flexibility in model parameters through a regime switching behavior. Given a time series of data , the SETAR model is a tool for understanding and, perhaps, predicting future values in this series, assuming that the behavior of the series changes once the series enters a different regime. The switch from one regime to another depends on the past values of the series (hence the Self-Exciting portion of the name). The model consists of autoregressive (AR) parts, each for a different regime. The model is usually referred to as the SETAR(, ) model where is the number of regimes and is the order of the autoregressive part (since those can differ between regimes, the portion is sometimes dropped and models are denoted simply as SETAR(k).They allow for changes in the model parameters according to the value of weakly exogenous threshold variable , assumed to be past values of , e.g. , where is the delay parameter, triggering the changes. Defined in this way, SETAR model can be presented as follows:
where is a column vector of variables; are non-trivial thresholds dividing the domain of into different regimes. In each of the regimes, the AR(p) process is governed by a different set of variables: . In such setting, a change of the regime (because the past values of the series surpassed the threshold) causes a different set of coefficients: . to govern the process . The SETAR model is a special case of Tong's general threshold autoregressive models. The latter allows the threshold variable to be very flexible, such as an exogenous time series in the open-loop threshold autoregressive system, a Markov chain in the Markovchain driven threshold autoregressive model which is now also known as the Markov switching model.