Modelling the persistence of conditional variances (Q3756387)
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 4005417
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Modelling the persistence of conditional variances |
scientific article; zbMATH DE number 4005417 |
Statements
Modelling the persistence of conditional variances (English)
0 references
1986
0 references
autoregressive conditional heteroscedasticity
0 references
nonlinear conditional heteroscedasticity
0 references
exchange rate determination
0 references
ARCH-type models
0 references
modelling of risk and uncertainty
0 references
GARCH
0 references
modelling conditional variances
0 references
asset pricing theory
0 references
integrated in variance
0 references
generalizing the conditional density
0 references
Student-\(t\) distribution with unknown degrees of freedom
0 references
conditional kurtosis
0 references
time aggregated models
0 references
0 references
0 references
0.8532047271728516
0 references
0.8399104475975037
0 references
0.8269030451774597
0 references
0.8269030451774597
0 references
0.8233121037483215
0 references
The authors survey the recent stand in ARCH-type modelling and they also present new developments of this area of research. The increasing interest in ARCH-type models and its extensions stems from microeconomic theory in which risk-averse agents have to make decisions based upon the distribution of a random variable some time in the future. This requires the adequate modelling of risk and uncertainty. The authors start with a simple version of the ARCH and generalized ARCH (GARCH) formulations for modelling conditional variances. The properties of the models are discussed and by use of asset pricing theory an example is given. The next part of the paper defines a new class of models which are defined as integrated in variance. These include as a special case the variance analogue of a unit root in the mean. A further extension consists of generalizing the conditional density from a normal to a Student-\(t\) distribution with unknown degrees of freedom. Procedures for estimating these unknown degrees of freedom are given. With these estimates implications about the conditional kurtosis of these models and time aggregated models can be drawn. A final generalization allows the conditional heteroscedasticity to be a nonlinear function of the squared innovations. The paper is discussed by F. X. Diebold, John Geweke, David F. Hendry, S. G. Pantula and St. E. Zin.
0 references