|
IDENTIFICATION OF STATE SPACE SYSTEMS WITH CONDITIONALLY HETEROSKEDASTIC INNOVATIONS
Dietmar Bauer*
* Institute for Econometrics, Operations Research and System
Theory, TU Wien, Argentinierstr. 8, A-1040 Vienna, Austria
In this paper consistency of estimates of linear dynamic systems obtained by using subspace algorithms under quite general assumptions on the innovations are derived. The assumptions include i.a. GARCH type of errors as well as E-GARCH. Also the consistent estimation of the model for the conditional variance is discussed. A small simulation study shows the potential of subspace algorithms in the context of GARCH modelling in comparison with the method implemented in MATLAB.
Keywords: subspace methods, GARCH models, finance, asymptotic properties
Session slot T-Mo-M02: State Space Identification and Subspace Methods/Area code 3a : Modelling, Identification and Signal Processing
 | 