Optimal Autoregressive Predictions
This paper proposes a new, optimal estimator of the AR(1) coefficient that minimixes the prediction mean-squared-error. This estimator can be used to generate an optimal predictor. The new estimator¡®s asymptotic distributions are derived for the cases of stationarity and a near unit root. The optimal estimator is also derived for the AR(p) model (p>=2) and its asymptotic distributions are reported. Simulation results confirm advantages of using the optimal estimator for prediction.
|Date of creation:||Feb 2016|
|Contact details of provider:|| Postal: 1 Sinsu-dong, Mapo-gu, Seoul 121-742|
Web page: http://econdept.sogang.ac.kr/laboratory/information.do
More information through EDIRC
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Phillips, Peter C B, 1988. "Regression Theory for Near-Integrated Time Series," Econometrica, Econometric Society, vol. 56(5), pages 1021-1043, September.
- Andrew Atkeson & Lee E. Ohanian, 2001. "Are Phillips curves useful for forecasting inflation?," Quarterly Review, Federal Reserve Bank of Minneapolis, issue Win, pages 2-11.
- Choi, In, 1993. "Asymptotic Normality of the Least-Squares Estimates for Higher Order Autoregressive Integrated Processes with Some Applications," Econometric Theory, Cambridge University Press, vol. 9(02), pages 263-282, April.
When requesting a correction, please mention this item's handle: RePEc:sgo:wpaper:1607. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Jung Hur)
If references are entirely missing, you can add them using this form.