Posted by
Michele Ravani-2 on
URL: http://quantlib.414.s1.nabble.com/RE-FdDividendAmericanOption-doesn-t-like-zero-t-ime-to-dividend-tp2179p2181.html
On Fri, 13 Sep 2002 09:51:38 +0100 Marco Marchioro <
[hidden email]> wrote:
MM> Hi,
MM> On a second thought you should be able to have 0 as a dividend date.
MM> I did not have a chance to try the C++ code you provided but the
MM> following,
MM> equivalent, python code prints 5.35988679227 as the option value
MM> and does not raise any exception.
MM>
MM> import QuantLib
MM> opt = QuantLib.FdDividendAmericanOption("Call",40.25,
MM> 35,0.0,0.05,0.1,0.25,
MM> [0.125],[0],20,20)
MM> print opt.value()
MM>
MM> anybody else has an opinion on this?
MM>
MM> Marco
MM>
MM>
MM>
MM> At 11:55 AM 9/11/02 -0500, you wrote:
MM> > Hi Vadim,
MM> > note that the vector dividendTimeA should contain
MM> > the ex-dividend
MM> > dates.
MM> > If the dividend will be paid today the ex-dividend date is tomorrow,
MM> > hence cannot be 0.0(should be a positive small number).
I would tend to stick to your original reply, unless one wants to get down
to counting the hours.
Assume for instance you are running an overnight risk analysis after
midnight using yesterday's closing prices, the dividend possibly wouldn't
reflect in the price yet, but I guess your risk numbers in the morning
should.
I know this is a bit extreme, but does it make sense?
Vadim, why do you want to have a 0 dividend time?
What is the business reason?
Ciao
--
Michele Ravani
[hidden email]
"Those who live hoping, die singing" My Gran