Minimum distance estimation for ARMA and GARCH processes

機構典藏 > College of Business and Management > Graduate Institute & Department of Banking and Finance > Thesis > Item 987654321/108157

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/108157

Title:	Minimum distance estimation for ARMA and GARCH processes
Authors:	Chung, Huimin
Contributors:	淡江大學財務金融學系
Date:	1997
Issue Date:	2016-10-27 09:49:28 (UTC+8)
Abstract:	This dissertation considers the estimation of the parameters of ARMA and GARCH processes by the Minimum Distance Estimator (MOE). The estimator is obtained by minimizing a quadratic distance function between the sample autocorrelation and theoretical autocorrelation functions and has the advantage of imposing very little in terms of distributional assumptions on the innovation process. The asymptotic properties of the MOE for ARMA processes are discussed. The exact asymptotic efficiency of using a block of sample autocorrelations is derived for some low order MA and ARMA models. The MOE is surprisingly efficient for some parts of the parameter space. We also investigate the properties of the MOE when used to estimate linear GARCH models using autocorrelations of the squared process. Monte Carlo results show that the MDE performs better than Quasi-MLE with certain conditional densities which exhibit extreme departures from conditional Gaussianity. An application of the MOE to estimate a GARCH(l, 1) model from high frequency exchange rate data is provided.
Relation:	108 pages
Appears in Collections:	[Graduate Institute & Department of Banking and Finance] Thesis

Files in This Item:

File	Description	Size	Format
index.html	館藏資訊	0Kb	HTML	141	View/Open
目次.pdf		403Kb	Adobe PDF	119	View/Open