A comparison of four estimators of a first order autoregressive process
Loading...
Authors
Horn, Joseph A. Jr.
Subjects
Autoregression
Autocorrelation
Durbin-Watson
Prais-Winsten
Theil-Nagar
Beach-MacKinnon
Autocorrelation
Durbin-Watson
Prais-Winsten
Theil-Nagar
Beach-MacKinnon
Advisors
Boger, Dan C.
Date of Issue
1986-09
Date
September 1986
Publisher
Language
en_US
Abstract
Econometricans must choose between many methods for estimating Rho in a first order autoregressive process. This thesis examines the performance of four estimators in a Monte Carlo situation. The methods examined are Durbin-Watson, Beach-MacKinnon, Theil-Nagar and Prais-Winsten. The autocorrelation coefficient, Rho, was varied from .2 to .9 and each method provided estimates of Rho and beta for 1000 replications. The results presented here are similar to those found in previous comparisons. Specifically, Ordinary Least Squares was found to be an efficient estimator of beta when autocorrelation is present only to a slight degree. Of the four estimators examined, the performance of Theil-Nagar proved superior in estimating both Rho and beta for small values of the correlation coeficient. Beach-MacKinnon, on the hand, while containing a large bias in the estimation of Rho, is the more efficient estimator of beta for large values of Rho.
Type
Thesis
Description
Series/Report No
Department
Operations Research
Organization
Naval Postgraduate School (U.S.)
Identifiers
NPS Report Number
Sponsors
Funder
Format
48 p.
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.