Estimating GJR-GARCH Models in R with Available Packages

Summary

This exchange asks how to estimate a GJR-GARCH volatility model in R and whether suitable software packages exist. One reply points to the rgarch package and sketches a workflow: specify a GJR-GARCH variance model, choose mean and return-distribution settings, and fit the model to data. It notes that package organization and installation sources may have changed, and advises checking the current syntax and using rugarch after a split into univariate and multivariate packages.

A second reply lists bayesGARCH, gogarch, and ccgarch as packages available through CRAN. The document supplies package names and a brief example of model specification, but it does not compare implementations, explain parameter estimation, or discuss diagnostics and model assumptions. Its package availability and syntax guidance may be outdated, so the suggested libraries and interfaces should be verified before use.

Key ideas

• GJR-GARCH is a volatility model that can be specified and fitted in R.
• The reply identifies rgarch and rugarch as relevant package names and cautions that syntax may have changed.
• Other listed package options include bayesGARCH, gogarch, and ccgarch.
• The exchange does not compare package capabilities or explain estimation diagnostics.
• Package availability and usage details should be checked against current documentation.

Tags

# GJR-GARCH Model In R


# GJR-GARCH Model In R












Any idea how to estimate GJR-GARCH models in R? Is there any particular library like fGarch that supports such models?

## Answer by Karol J. Piczak (score 8)

https://quant.stackexchange.com/a/3481

You can have a look at `rgarch`. It's quite versatile. From what I remember, you have to get it explicitly from R-Forge, as it's not available from CRAN.

See the rgarch website for more details.

Last time I checked, usage was something like this:

```
spec.gjrGARCH = ugarchspec(variance.model=list(model="gjrGARCH", garchOrder=c(1,1)), mean.model=list(armaOrder=c(1,1), include.mean=TRUE), distribution.model="std")
gjrGARCH <- ugarchfit(data, spec=spec.gjrGARCH)
```

From what I see, it has been recently split into uni- and multivariate packages, so you would need to verify the syntax and install `rugarch`.

## Answer by ash (score 4)

https://quant.stackexchange.com/a/3478

CRAN has a few:

- bayesGARCH

- gogarch

- ccgarch

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.