Reputation: 45
I am a beginner in R and have been working with the ImpliedVol function in the “FMStable” package to calculate the implied volatilities of European-type call options. What’s been bothering me is that the function seems to give wrong outputs. I conducted an experiment to see if it’s the case, and here’s what I did:
#using certain values to do the test
> ImpliedVol(spot=8,strike=10,expiry = 1,price = 2,intRate = 0.05,tol=1.e-9)
[1] 0.8643101
#use the B-S formula to obtain call option prices
> test=function(x){
+ s=8
+ X=10
+ T=1
+ r=0.05
+ d1=(log(s/X)+(r+x^2/2)*T)/(x*T^(1/2))
+ d2=d1-x*T^(1/2)
+ c=s*pnorm(d1)-X*exp(-r*T)*pnorm(d2)
+ return(c)
+ }
#testing list containing possible volatilities
> sig=seq(from=0.5,to=1,by=0.00001)
#test results
> calltest=test(sig)
> plot(sig,calltest)
> abline(h=2)
> abline(z=0.8643101)#the result from the ImpliedVol function
> abline(v=0.7919,col='blue')
As you can see, the true Implied Volatility should be approximately 0.7919 instead of 0.8643101.
I looked at the codes wrapped in the ImpliedVol function and noticed that the author used the uniroot function to obtain the root in f(ImpVol)=0. I don’t see what is wrong here and hope that someone can help me out of it.
(This question is raised because I couldn't get in touch with the author of the R package "FMStable" after many failed attempts.)
Upvotes: 0
Views: 64
Reputation: 13128
I don't do finance, but I see that the ImpliedVol
function has a carryCost
argument which has a default of 0. Setting carryCost
equal to intRate
gives you what you call the "true Implied Volatility."
ImpliedVol(spot=8,strike=10,expiry = 1,price = 2,intRate = 0.05, carryCost = 0.05,
tol=1.e-9)
# [1] 0.7918341
See e.g. the formula here: http://help.cqg.com/cqgic/default.htm#!Documents/blackscholesgeneralizedextendedmodel.htm
Upvotes: 1