[R-SIG-Finance] Extending the variance equation in GARCH models in the rugarch package (Realized GARCH)

[Riddell Red]

Dear all,

I am wondering if you are able to give me some direction. I am currently writing my thesis in Economics and Finance, and have been using the rugarch package for the GARCH modelling.

 

I am using the realized GARCH model, and comparing forecasts to a HAR model (that includes long memory of volatility). For a fair comparison, I want to extend the realized GARCH variance equation by including two parameters. These parameters are: the lagged 5 day average RV measure and the lagged 22 day average RV measure. This is the same as the HAR specification. The measurement equation remains the same, and thus the likelihood functions should be the same as far as I am aware.

 

The variance equation would be: log(σ2t) =ω+∑γ1log(RVt−1) +∑γ2log(RVt−2) +∑βlog(σ2t−1) + ∑θ1log(5davgRVt-1) + ∑θ2log(22davgRVt-1)

 

This is similar to a paper by Huang, Liu and Wang, named Modeling long memory volatility using realized measures of volatility: A realized HAR GARCH model. Huang is one of the authors of the original Realized GARCH model, so it lends credibility. I can send you the paper if you wish.

 

I am struggling to wrap my head around how to modify the realgarch R file.

 

Thank you for your time,

Regards,

Red.

 


	[[alternative HTML version deleted]]