NIG process

[gigisplillo]

I've tried to price an american option based on the Normal Inverse Gaussian Process using montecarlo method. But the problem is that R (the sofware i'm using) does not recognize the code of the inverse normal distribution. Could someone help me? I'm putting my Rcode.

NIG:​

require(fImport)​

data <- yahooSeries("AMZN", from="2013-01-15",to="2013-04-15")​

S <- data[, "AMZN.Close"]​

X <- returns(S)​

​

S0 <- 142.2​

K <- 120​

T <- 61​

alpha2 <- 0.2​

beta <-0.5​

delta <- 1.2​

alpha <-3.2​

r <- 0.003​

d <- 0.002​

m <- 0.01​

​

Montecarlo:​

n<- 50000​

t <- rinvgauss(n,shape=delta*sqrt(alpha2^2-beta^2)),scale=delta)​

N <- rnorm(n,0,1)​

X <- beta*delta^2*t+N*delta*sqrt(t)​

omega <- delta*(sqrt(alpha2^2-(beta+1)^2-sqrt(alpha2^2-beta^2))​

S <- S0*exp(m*T+omega*T+X)​

payoff <- sapply(S, function (x)max(x-K,0))​

mean(payoff)*exp(-r*T)​

 

[diegosanaz]

Have you installed and loaded the correct package to use that function? I se there is a rinvgauss under the package "statmod", but it does not allow all the input you are writting in your code.

 

[gigisplillo]

No.i'll try thank you.

 

[gigisplillo]

The Montecarlo code is based on page 18 of the paper of Karsten Krog Sæbø entitled" Pricing Exotic Options with the Normal Inverse Gaussian Market Model using Numerical Path Integration" Master of Science in Physics and Mathematics Submission date: June 2009 Supervisor: Arvid Næss, MATH.

 

[gigisplillo]

Does anybody have a better R code to pricing options using the Normal Inverse Gamma process?