Posted by
Luigi Ballabio on
URL: http://quantlib.414.s1.nabble.com/Day-counting-offset-tp3882p3883.html
On 07/06/2005 10:29:11 PM, Guowen Han wrote:
>
> The error message can be removed (suggested by Michelotti Enrico) by
> adding the following statement.
> myTermStructure->enableExtrapolation();
>
> However the offset is still there, and I got more time points than
> expected. Here is my code, any suggestions are appreciated
>
> std::vector<Date> dates;
> dates.push_back(refDate + Period(0, Months));
> ...
> dates.push_back(refDate + Period(360, Months));
It's probably a date-adjustment thing. You're passing as the last date
on the curve the reference date plus 30 years, unadjusted. But that
date might be a holiday, so the swaption maturity (calculated inside
the swaption helper) is set one or two days later and falls off the end
of the term structure.
You can try initializing the curve dates as
std::vector<Date> dates;
Calendar calendar = ...; (the one relevant for your calculation)
dates.push_back(refDate);
dates.push_back(calendar.advance(refDate, Period(1,Months)));
...
dates.push_back(calendar.advance(refDate, Period(360,Months)));
This might push the end of the curve a day or two further, which is the
lag you need.
Later,
Luigi
----------------------------------------
The shortest way to do many things is to do only one thing at once.
-- Samuel Smiles