Please refer to the article at:

https://arxiv.org/pdf/0804.0162.pdf

Hi,

I whote and article about improving the GARCH model utilizing high and low data:

https://www.researchgate.net/publication/299997143_High-low_range_in_GARCH_models_of_stock_return_volatility

I did all the calculations in Eviews. Relevant commands in Eview are the following:

genr logr = dlog(price)

genr park = (log(high/low))^2/(4*log(2))

equation eq1.arch(1,1) logr ....this is a basic GARCH(1,1) model with zero mean and normally distributed errors

equation eq2.arch(0,1) logr @ park(-1)   ....this is the model suggested by me, again with zero mean return and normally distributed errors.

Article High-low range in GARCH models of stock return volatility

Hi,

Faisal Nawaz Please refer to the article at:

https://www.efmaefm.org/0EFMAMEETINGS/EFMA%20ANNUAL%20MEETINGS/2012-Barcelona/papers/EFMA2012_0173_fullpaper.pdf