|
|
Nikolaos Kourogenis |
|
''Polynomial Trends, Nonstationary Volatility and the Eicker-White Asymptotic Variance Estimator'' |
( 2015, Vol. 35 No.3 ) |
|
|
The problem of inference in autoregressions around polynomial trends, under nonstationary, possibly explosive, volatility is investigated. It is shown that the well-known t-statistics that incorporate the Eicker-White covariance matrix estimator are asymptotically standard normal. Simulation results show that the application of a residual-based recursive-design wild bootstrap reduces significantly the size distortions in small samples. |
|
|
Keywords: Autoregression, Polynomial trend, Nonstationary volatility, Eicker-White covariance matrix estimator, Wild bootstrap |
JEL: C2 - Single Equation Models; Single Variables: General C1 - Econometric and Statistical Methods: General |
|
Manuscript Received : Jun 01 2015 | | Manuscript Accepted : Jul 24 2015 |
|