A threshold AR(1) model

1984/06/01 by Joseph D. Petruccelli, Samuel W. Woolford · 1 citation
Economics, Econometrics and Finance · Mathematics · Decision Sciences · #Financial Risk and Volatility Modeling #Statistical Methods and Inference #Advanced Statistical Process Monitoring

paper · doi:10.2307/3213639

Abstract

We consider the model where φ 1 , φ 2 are real coefficients, not necessarily equal, and the a t ,'s are a sequence of i.i.d. random variables with mean 0. Necessary and sufficient conditions on the φ 's are given for stationarity of the process. Least squares estimators of the φ 's are derived and, under mild regularity conditions, are shown to be consistent and asymptotically normal. An hypothesis test is given to differentiate between an AR(1) (the case φ 1 = φ 2 ) and this threshold model. The asymptotic behavior of the test statistic is derived. Small-sample behavior of the estimators and the hypothesis test are studied via simulated data.

Cited by

• Long-run relationship between the unemployment rate and the current account balance in the United States: An empirical analysis [europepmc]

Related

• On Parameter Estimation of Threshold Autoregressive Models
• Some Probabilistic and Statistical Properties of a Random Coefficient Autoregressive Model
• Asymptotic Inference for AR(1) Penal Data
• Testing for strict stationarity in a random coefficient autoregressive model
• Asymptotic inferences for an AR(1) model with a change point: stationary and nearly non-stationary cases