A class of continuous-time garch-type processes (Q4238142)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1275708
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | A class of continuous-time garch-type processes |
scientific article; zbMATH DE number 1275708 |
Statements
A class of continuous-time garch-type processes (English)
0 references
29 September 1999
0 references
GARCH and Lévy processes
0 references
past dependence
0 references
stationary increments
0 references
0.836768627166748
0 references
0.8351190686225891
0 references
0.8050918579101562
0 references
0.798231840133667
0 references
As is known in discrete time the increments in log prices of a financial asset are often successfully modelled as a generalized autoregressive conditional heteroskedastic (GARCH) process. Successfully because these models were able to explain for example some phenomena in financial index's behaviour. The purpose of this paper is to delimit a class of continuous-time processes resembling the discrete-time GARCH processes in several ways. The problem of defining continuous-time GARCH models has been considered by several authors: Nelson, Kallsen and Taqqu, Hobson and Rogers. This approach is related to the ideas of Kind, Lipzer and Runggaldier in the sense that continuous-time GARCH-type processes presented in this paper have a past dependent structure defined in terms of the quadratic variation. But more generally dependence of the previsible compensator \(\langle\cdot\rangle\) of the quadratic variation \([\cdot]\) and of the present state is here allowed as well. Besides, the Wiener process may be replaced by an almost arbitrary (Lévy) process with stationary independent increments. In Section 2 the author recalls some basic facts about discrete-time GARCH models. Then he defines his class of continuous-time GARCH-type processes and argues that these processes are natural counterparts to the discrete-time GARCH processes. They are defined as solutions of stochastic differential equations and he also proves that these equations have a unique solution subject to common regularity conditions. In the next section the talk is about the subclass consisting of processes with past dependence defined in terms of \(\langle\cdot\rangle\) and \([\cdot]\) only. Consequently, the present state does not influence on the future evolution of the increments. This subclass in fact has a structure resembling the discrete time setting even better. Finally in Section 4 the author relates his processes to complete markets.
0 references