Posted by
Luigi Ballabio on
Feb 25, 2008; 4:34pm
URL: http://quantlib.414.s1.nabble.com/FlatForward-and-its-referenceDate-tp5539p5540.html
Hi Oliver,
On Wed, 2008-02-13 at 02:22 -0800, obeck wrote:
> I am new to QuantLib and I have got a very basic problem with the interest
> rate calculations performed by FlatForward (FlatForward (const Date
> &referenceDate, Rate forward, const DayCounter &dayCounter, Compounding
> compounding=Continuous, Frequency frequency=Annual)). The following question
> I could not find answered neither in the QuantLib docu, nor in the user
> groups:
> 1) In the above constructor 'forward' specifies a forward rate for the
> period starting at 'referenceDate' and ending at 'referenceDate' +
> '1/frequency'? Is this correct?
Correct (the forward rate is the annualized one, of course.)
> 2) What is the variable 'referenceDate' good for in the context of
> FlatForward anyway? If forward rates are equal (flat) for all time periods
> in the future, obviously the spot zero rates should be identical for all
> maturities, too. The other way round, if zero spot rates are identical for
> all maturities, forward rates will also be 'flat'. What information does
> 'referenceDate' provide in this context?
As you say, the rates will be identical for all dates. However, the
FlatForward curve (like all yield term structures) can also provide
discounts. The given reference date is the date at which the discount
equals 1. Depending on the conventions used on your desk for NPV
calculation, it can be today's date or the spot date (two business days
from today for Euribor rates.)
> 3) Given that it is correct that FlatForward describes a flat interest rate
> (or dividend yield) term structure, a change of the variable 'referenceDate'
> should not alter the net present value calculation of options. However the
> npv-calculation are affected by changes in the 'referenceDate' of
> FlatForward. Why is that the case? I have attached a simple code fragment
> mainly taken from the QuantLib examples to illustrate the effect of using
> different referenceDates in FlatForward on NPV calculations.
Same as (2). In the exp(-rt) you have in the Black-Scholes formula, r is
the same regardless of the reference date. However, t changes depending
on where you set it, hence the change of NPV.
Luigi
--
I am extraordinarily patient, provided I get my own way in the end.
-- Margaret Thatcher
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users