论文信息 - Bayesian prediction tests for structural stability

Bayesian prediction tests for structural stability

Abstract A Bayesian prediction test for structural stability is proposed which has several advantages over classical procedures. In particular, the small sample distribution of the test statistic is known under the null hypothesis, the same test statistic is appropriate for both regression and time series models, the test is valid even for nonstationary, multivariate AR and VAR models, and the test is robust to certain forms of nonnormality in the error distribution. A comparison of the Bayesian and classical tests indicates that the Bayesian test is equivalent to the predictive Chow test for regression models.

Jeffrey A. Mills | J. Mills

[1] Taku Yamamoto,et al. Asymptotic mean square prediction error for an autoregressive model with estimated coefficients , 1976 .

[2] Thomas F. Cooley,et al. Likelihood and other approaches to prediction in dynamic models , 1987 .

[3] C. Sims. Bayesian skepticism on unit root econometrics , 1988 .

[4] Small-Sample Properties of Predictions from the Regression Model with Autoregressive Errors , 1983 .

[5] H. Lütkepohl. Prediction tests for structural stability , 1988 .

[6] George E. P. Box,et al. Comparison of Forecast and Actuality , 1976 .

[7] Patrick A. Thompson,et al. Sampling the Future: A Bayesian Approach to Forecasting From Univariate , 1986 .

[8] G. Reinsel,et al. Asymptotic Properties of Prediction Errors for the Multivariate Autoregressive Model Using Estimated Parameters , 1980 .

[9] S. Chib,et al. Bayes prediction in regressions with elliptical errors , 1988 .

[10] Peter Schmidt,et al. THE ASYMPTOTIC DISTRIBUTION OF FORECASTS IN THE DYNAMIC SIMULATION OF AN ECONOMETRIC MODEL , 1974 .

[11] Richard T. Baillie,et al. Prediction from the Dynamic Simultaneous Equation Model with Vector Autoregressive Errors , 1981 .