[R-SIG-Finance] Backing out implied parameters from HNGARCH option model

Tobias tobias_elbert at hotmail.com
Wed Jan 10 02:26:48 CET 2007


Dear all,

I am trying to back out the implied parameters using observed options prices
and the Heston-Nandi Garch model via NLS. 

For this purpose I am using Diethelm Wuertz's pricing algorithm in a
function called hnoption similar to this:

hnoption <- function(b)      #where b is the vector of parameters passed on
to the pricing algorithm
{
     ....
     hnoption <- value obtained via pricing algorithm
}

calccall <- function(b)
{
	for (jj in 1:num1)
	{
		counter <<- jj 
		qq[jj] <- hnoption(b)
		
	}
	
	calccall <- sum((c-qq)^2)
 }

# some initial guess of parameters
bstart <- c(.1,.1,1,-.5,.7)

p <- optim(bstart,calccall,hessian = TRUE)

The code works fine for a few iterations of optim(). However, at some point
it produces parameter values that cause an integral that's being evaluated
in hnoption to drift off to infinity. At least that's what I think happens
when R produces the error: error in integrate(): non-finite function value.
I have tried to constrain some parameters in optim() but keep receiving the
same error. 

I would really appreciate if someone could point me to possible solutions. 

Regards 

Tobias

-- 
View this message in context: http://www.nabble.com/Backing-out-implied-parameters-from-HNGARCH-option-model-tf2950123.html#a8250644
Sent from the Rmetrics mailing list archive at Nabble.com.



More information about the R-SIG-Finance mailing list