Posted by
MonkeyMan-3 on
URL: http://quantlib.414.s1.nabble.com/Missing-USDLibor6M-Actual-360-fixing-tp7050p7054.html
Luigi Ballabio <luigi.ballabio <at> gmail.com> writes:
> It's a calendar mismatch. USD libor is fixed in London and uses a
> calendar which joins US and UK holidays.
Makes sense. So I changed my date calculation to this:
Calendar calendar = UnitedStates();
Date todaysDate(29, April, 2005);
Integer fixingDays = 2;
JointCalendar usdgbpCalendar = JointCalendar(UnitedStates(), UnitedKingdom());
Date settlementDate = usdgbpCalendar.advance(todaysDate, fixingDays, Days);
settlementDate = usdgbpCalendar.adjust(settlementDate);
Then I also changed my rate helpers for deposits and swaps to have the joined
calendar:
boost::shared_ptr<RateHelper> d1w(new DepositRateHelper(
Handle<Quote>(d1wRate),
1*Weeks, fixingDays,
usdgbpCalendar, Following,
true, depositDayCounter));
and
boost::shared_ptr<RateHelper> s2y(new SwapRateHelper(
Handle<Quote>(s2yRate), 2*Years,
usdgbpCalendar, swFixedLegFrequency,
swFixedLegConvention, swFixedLegDayCounter,
swFloatingLegIndex));
And my code doesn't blow up. One last question, though. When I'm building a new
swap to price do I need to change both the fixed and floating calendars or only
the floating calendar. Like this:
Schedule fwdFixedSchedule(fwdStart, fwdMaturity,
Period(fixedLegFrequency),
calendar, fixedLegConvention,
fixedLegConvention,
DateGeneration::Forward, false);
Schedule fwdFloatSchedule(fwdStart, fwdMaturity,
Period(floatingLegFrequency),
usdgbpCalendar, floatingLegConvention,
floatingLegConvention,
DateGeneration::Forward, false);
It seems that since the standard is for the fixed leg to be unadjusted, I would
just use the joined calendar for the floating leg, but I may be confusing
conventions.
Thanks for your help. I have been flailing around looking at day count
conventions and in all the wrong places.
Monkey
------------------------------------------------------------------------------
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users