CODEWITHSUNDEEP
rtime-series
user2794659
user2794659

Reputation: 145

GARCH-M model estimation in R

is it possible to estimate a GARCH with volatility in the mean using R?

I read that it may be possible with rgarch package but I have some trouble installing it. Is there any other way?

The model is:

  r[t] = mu + c*s[t]^2 + a[t],

  a[t] = s[t]*e[t],

  s[t]^2 = alpha0 + alpha1 * a[t-1]^2 +  beta1 * s[t-1]^2,

Regards,

Juan.

Upvotes: 1

Views: 4986

Answers (2)

Robert
Robert

Reputation: 5152

In 2024, you can estimate this models with rugarch package:

returns <- data$rtn * 100

spec = ugarchspec(variance.model = list(model = "sGARCH", garchOrder = c(1, 1),   submodel = "GARCH"), 
      mean.model = list(armaOrder = c(0, 0), 
     include.mean = TRUE, archm = TRUE, archpow = 2))
#print(spec)

garchm <- ugarchfit(spec = spec, data = returns , solver = 'hybrid')
print(garchm)
coef(garchm)

See the help for more information.

Upvotes: 0

Paul Fletcher-Hill
Paul Fletcher-Hill

Reputation: 86

Ruey Tsay has published a garchM function. Save the code and load it into R using the source function:

source('/path/to/garchM.R')

The garchM function can be used as follows:

data <- read.table('/path/to/data.txt')
returns <- data$rtn * 100
garchM(returns)

Upvotes: 1

Related Questions