More documents from Cook, Steven
More documents from Quantitative Finance

Export to your Reference Manger

Please Copy & Paste



Bookmark and Share

The robustness of modified unit root tests in the presence of GARCH

[journal article]

Cook, Steven

fulltextDownloadDownload full text

(221 KByte)

Citation Suggestion

Please use the following Persistent Identifier (PID) to cite this document:

Further Details
Abstract The research of Kim and Schmidt (1993) is extended to examine the properties of modified Dickey-Fuller unit root tests in the presence of generalised autoregressive conditional heteroskedasticity (GARCH). Using Monte Carlo simulation, the properties of the tests are examined for a range of GARCH processes over alternative sample sizes. Oversizing is observed for all tests, with the extent of size distortion driven by the volatility, rather than the persistence, of the underlying GARCH process. While the original Dickey-Fuller test is found to exhibit greater size distortion than the modified tests, the modified tests are found to be substantially oversized when the GARCH process exhibits a high degree of volatility, even for large samples.
Classification Basic Research, General Concepts and History of Economics; Economic Statistics, Econometrics, Business Informatics
Method theory application
Free Keywords GARCH models; Time Series Economics; Numerical Simulation; Quantitative Finance
Document language English
Publication Year 2006
Page/Pages p. 359-363
Journal Quantitative Finance, 6 (2006) 4
Status Postprint; peer reviewed
Licence PEER Licence Agreement (applicable only to documents from PEER project)