QuantLib › quantlib-users

globall update in piecewise yield curve?

Classic

List

Threaded

3 messages Options

Candy Chiu

globall update in piecewise yield curve?

Hi Luigi,

I just want to double check with you that the piecewise yield curve changes only local data in each step of the iteration? In other words, once it is done with the i-1'th instrument, all parameters indexed 0 to i-1 are fixed. I am not sure if this is true for the Cubic interpolation. One symptom that raised concern is when I apply a 20-year key rate shock, the 2y10y forward rate differs from the base case by 1bp. Since all market data from time 0 up to time 15 years didn't change, the shock should only impact discount factors after 15 years. The 2y10y forward rate depends only on discount factor up to 12 years, and should stay the same.

Your help is appreciated.

Candy

------------------------------------------------------------------------------
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. SALE $99.99 this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122912
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Luigi Ballabio

Re: globall update in piecewise yield curve?

Hi,
cubic interpolation is global; that is, even if the interpolation
moves just one node, the interpolation changes globally. If you want
shocks to be local, you'll need another interpolation.

Luigi

On Fri, Jan 11, 2013 at 4:00 PM, Candy Chiu <[hidden email]> wrote:

> Hi Luigi,
>
> I just want to double check with you that the piecewise yield curve changes
> only local data in each step of the iteration? In other words, once it is
> done with the i-1'th instrument, all parameters indexed 0 to i-1 are fixed.
> I am not sure if this is true for the Cubic interpolation. One symptom that
> raised concern is when I apply a 20-year key rate shock, the 2y10y forward
> rate differs from the base case by 1bp. Since all market data from time 0
> up to time 15 years didn't change, the shock should only impact discount
> factors after 15 years. The 2y10y forward rate depends only on discount
> factor up to 12 years, and should stay the same.
>
> Your help is appreciated.
>
> Candy
>
> ------------------------------------------------------------------------------
> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
> MVPs and experts. SALE $99.99 this month only -- learn more at:
> http://p.sf.net/sfu/learnmore_122912
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>

Luigi Ballabio

Re: globall update in piecewise yield curve?

Oops. I meant: "even if the bootstrap procedure changes just one node..."

On Sat, Jan 12, 2013 at 12:27 PM, Luigi Ballabio
<[hidden email]> wrote:

> Hi,
> cubic interpolation is global; that is, even if the interpolation
> moves just one node, the interpolation changes globally. If you want
> shocks to be local, you'll need another interpolation.
>
> Luigi
>
>
> On Fri, Jan 11, 2013 at 4:00 PM, Candy Chiu <[hidden email]> wrote:
>> Hi Luigi,
>>
>> I just want to double check with you that the piecewise yield curve changes
>> only local data in each step of the iteration? In other words, once it is
>> done with the i-1'th instrument, all parameters indexed 0 to i-1 are fixed.
>> I am not sure if this is true for the Cubic interpolation. One symptom that
>> raised concern is when I apply a 20-year key rate shock, the 2y10y forward
>> rate differs from the base case by 1bp. Since all market data from time 0
>> up to time 15 years didn't change, the shock should only impact discount
>> factors after 15 years. The 2y10y forward rate depends only on discount
>> factor up to 12 years, and should stay the same.
>>
>> Your help is appreciated.
>>
>> Candy
>>
>> ------------------------------------------------------------------------------
>> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
>> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
>> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
>> MVPs and experts. SALE $99.99 this month only -- learn more at:
>> http://p.sf.net/sfu/learnmore_122912
>> _______________________________________________
>> QuantLib-users mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>>