TermStructure question.

[Siddharth Sharma-3]

Hi,

In the swapvaluation.cpp, I added a par rate
calculation after the creation of the
depoSwapTermStructure: i.e.

boost::shared_ptr<YieldTermStructure>
depoSwapTermStructure(new
             PiecewiseFlatForward(settlementDate,
depoSwapInstruments,                                
termStructureDayCounter, tolerance));

Rate tr = depoSwapTermStructure->parRate(10,
settlementDate);
-------------------------------------------------------

 the 10 year swap rate is 5.165% ( Rate
s10yQuote=0.05165;) but the value returned by
parRate is 5.16517962147765% The difference is too
small to be explained by a daycount
convention or fixingDays offset, but it exists
nonetheless and i can't seem to figure out
why. Could someone please explain why this discrepancy
exists, how to fix it and if not, how to explain it
away?
 I know it is in the 6th/7th decimal place, but the
numbers lose their credibility for some people if they
don't tie exactly to the numbers in the yield curve
inputs.
 
Anyone have any ideas?
 
thanks and regards
Sid


               
__________________________________________________________
How much free photo storage do you get? Store your friends 'n family snaps for FREE with Yahoo! Photos http://in.photos.yahoo.com

[Luigi Ballabio]

On 07/18/2005 12:17:39 PM, Siddharth Sharma wrote:
> In the swapvaluation.cpp, I added a par rate
> calculation after the creation of the
> depoSwapTermStructure
>
>  the 10 year swap rate is 5.165% ( Rate
> s10yQuote=0.05165;) but the value returned by
> parRate is 5.16517962147765%

Sid,
        the parRate method is just an approximation---it doesn't take  
into account all market conventions for swaps---and as a matter of fact  
I've a mind to remove it from the YieldTermStructure interface. The  
right way to evaluate the par rate is to instantiate a SimpleSwap and  
call its fairRate() method, as is done later in the example.

Later,
        Luigi

----------------------------------------

I have yet to see any problem, however complicated, which, when you
looked at it in the right way, did not become still more complicated.
-- Poul Anderson

[Ferdinando Ametrano-4]

Luigi Ballabio wrote:
>     the parRate method is just an approximation---it doesn't take  into
> account all market conventions for swaps---

the par rate is like a zero rate or a discount factor: all of them are
formulae/definitions that express the current (today) value of future
payment(s). As such none of them mind about any market convention at all.

> as a matter of fact  I've
> a mind to remove it from the YieldTermStructure interface.

I would not. It might be stressed in a better way that the par rate is
not a swap rate, but the par rate has its own right to exist.

> The  right
> way to evaluate the par rate is to instantiate a SimpleSwap and  call
> its fairRate() method, as is done later in the example.

Yeah!

ciao -- Nando

[Luigi Ballabio]

On 07/18/2005 03:13:28 PM, Ferdinando Ametrano wrote:
>>     the parRate method is just an approximation---it doesn't take  
>> into account all market conventions for swaps---
>
> the par rate is like a zero rate or a discount factor: all of them  
> are formulae/definitions that express the current (today) value of  
> future payment(s). As such none of them mind about any market  
> convention at all.

I beg to differ. There are conventions determining _what_ future  
payments should be considered when calculating the par rate. The fact  
that the method "doesn't mind" only means that you hard-coded them in  
the current implementation---for instance, fixed-rate payments are  
considered to have annual frequency, which is true here in Europe but,  
I'm told, false on the other side of the Atlantic.


>> as a matter of fact  I've a mind to remove it from the  
>> YieldTermStructure interface.
>
> I would not. It might be stressed in a better way that the par rate  
> is not a swap rate...

Even though the documentation in the code says otherwise

> ...but the par rate has its own right to exist.

Defined as what?

Later,
        Luigi


----------------------------------------

This gubblick contains many nonsklarkish English flutzpahs, but the  
overall pluggandisp can be glorked from context.
-- David Moser