Skip to content
All library documents

Finding an R Package for Multivariate VECH GARCH Estimation

Article Quant Q&A · Author: Nils
Read the original

Summary

The document addresses a software question about estimating a multivariate VECH GARCH(1,1) model in R. The question describes the model as representing residuals and a covariance matrix through a vectorized form of the matrix elements, and notes that the questioner had found packages for DCC and BEKK estimation but not VECH estimation.

The answer recommends the mgarch package from CRAN and points to an example associated with Professor Zivot. This provides a starting point for researchers seeking an implementation, rather than explaining the model’s equations, estimation procedure, assumptions, or interpretation. The document supplies no package version, code, comparison of alternative implementations, or evidence about numerical performance and maintenance. Users would need to consult the package documentation and example to determine whether its functionality matches their particular VECH specification and data requirements.

Key ideas

  • The question concerns multivariate VECH GARCH(1,1) estimation in R.
  • The response recommends the mgarch package as a possible implementation.
  • An example attributed to Professor Zivot is suggested as a practical reference.
  • The document does not explain model estimation or compare package capabilities and performance.

Tags

StatisticsVolatilityMachine learning
Full text
# Package for multivariate Garch Vech model for R?


# Package for multivariate Garch Vech model for R?












I`m new to programming and searching a package for R which inherents the estimation for a Vech Garch(1,1). This is a multivariate Garch model which forms the residuals and the covariance matrix from a NxN matrix to a N(N+1)/2 vector. I can only find DCC and BEKK estimation.

Does somebody know a package or source of code for this estimation?

## Answer by aureliano.bressan (score 1)

https://quant.stackexchange.com/a/27530

Try the mgarch package, it's available at CRAN. In this link you will find an example from Prof. Zivot.

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.