QuantLib › quantlib-users

Jump Points on Piecewise Yield Curve

Classic

List

Threaded

3 messages Options

Vlad F.

Jump Points on Piecewise Yield Curve

Hi,

I am trying to understand how Jump points work in Piecewise Yield Curve

My understanding is that a jump point should only effect the day that it’s applied on. For example, let say there is a RateHelper instrument going from Apr-12-2011 through May-25-2011 with a rate of 1.2% and Jump point with a date of May-5-2011 and value of 20 bps expressed as a discount factor.

What I was expecting to see if I were to extract the daily forward rate is for May-5-2011 to be affected by 20 bps.

What I see is that the jump point has affected the whole period roughly by 0.5 bps and March-15-2011 point was affected by 19.5 bps.

I am attaching a spreadsheet which contains a test cases that I put together in trying to understand it.

Jump_Test.zip -> https://docs.google.com/leaf?id=0BwNMjRmDi385MjAwYzNhMTgtNDMxNy00Yzc2LWEwODItMTZlMzZkMDEzN2Fk&hl=en&authkey=CObFm9EN

Any help would be appreciated.

Thanks,
Vlad

------------------------------------------------------------------------------
Enable your software for Intel(R) Active Management Technology to meet the
growing manageability and security demands of your customers. Businesses
are taking advantage of Intel(R) vPro (TM) technology - will your software
be a part of the solution? Download the Intel(R) Manageability Checker
today! http://p.sf.net/sfu/intel-dev2devmar
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Ferdinando M. Ametrano-3

Re: Jump Points on Piecewise Yield Curve

Hi Vlad

the jumps are modeled as exogenous multiplicative (discount) factor.

Any single jump does generate a spike for that single one day forward
rate, but if you consider a weekly forward rate there will be 7 rates
spiked up, if you consider a monthly forward rate there will be 30
rates spiked up, etc

Roughly speaking the (log) discount is the integral of the forward
rates: if you have a 1M spot rate of 2% and a 30bps jump in between
you won't get daily fwd rates of 2% with a single fwd rate at 2.30%.
Instead you'll get something like flat 1.99% with a spike at 2.29%,
keeping the average at 2%

Hope it helps

ciao -- Nando

On Thu, Mar 24, 2011 at 7:43 PM, Vlad F. <[hidden email]> wrote:

> Hi,
> I am trying to understand how Jump points work in Piecewise Yield Curve
> My understanding is that a jump point should only effect the day that it’s
> applied on. For example, let say there is a RateHelper instrument going
> from Apr-12-2011 through May-25-2011 with a rate of 1.2% and Jump point with
> a date of May-5-2011 and value of 20 bps expressed as a discount factor.
> What I was expecting to see if I were to extract the daily forward rate is
> for May-5-2011 to be affected by 20 bps.
> What I see is that the jump point has affected the whole period roughly by
> 0.5 bps and March-15-2011 point was affected by 19.5 bps.
> I am attaching a spreadsheet which contains a test cases that I put together
> in trying to understand it.
> Jump_Test.zip ->
> https://docs.google.com/leaf?id=0BwNMjRmDi385MjAwYzNhMTgtNDMxNy00Yzc2LWEwODItMTZlMzZkMDEzN2Fk&hl=en&authkey=CObFm9EN
>
> Any help would be appreciated.
> Thanks,
> Vlad
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>

Vlad F.

Re: Jump Points on Piecewise Yield Curve

Thank you for explanation Nando.

-Vlad

On Fri, Mar 25, 2011 at 7:35 AM, Ferdinando Ametrano <[hidden email]> wrote:

Hi Vlad

the jumps are modeled as exogenous multiplicative (discount) factor.

Any single jump does generate a spike for that single one day forward
rate, but if you consider a weekly forward rate there will be 7 rates
spiked up, if you consider a monthly forward rate there will be 30
rates spiked up, etc

Roughly speaking the (log) discount is the integral of the forward
rates: if you have a 1M spot rate of 2% and a 30bps jump in between
you won't get daily fwd rates of 2% with a single fwd rate at 2.30%.
Instead you'll get something like flat 1.99% with a spike at 2.29%,
keeping the average at 2%

Hope it helps

ciao -- Nando

On Thu, Mar 24, 2011 at 7:43 PM, Vlad F. <[hidden email]> wrote:
> Hi,
> I am trying to understand how Jump points work in Piecewise Yield Curve
> My understanding is that a jump point should only effect the day that it’s
> applied on. For example, let say there is a RateHelper instrument going
> from Apr-12-2011 through May-25-2011 with a rate of 1.2% and Jump point with
> a date of May-5-2011 and value of 20 bps expressed as a discount factor.
> What I was expecting to see if I were to extract the daily forward rate is
> for May-5-2011 to be affected by 20 bps.
> What I see is that the jump point has affected the whole period roughly by
> 0.5 bps and March-15-2011 point was affected by 19.5 bps.
> I am attaching a spreadsheet which contains a test cases that I put together
> in trying to understand it.
> Jump_Test.zip ->
> https://docs.google.com/leaf?id=0BwNMjRmDi385MjAwYzNhMTgtNDMxNy00Yzc2LWEwODItMTZlMzZkMDEzN2Fk&hl=en&authkey=CObFm9EN
>
> Any help would be appreciated.
> Thanks,
> Vlad
> ------------------------------------------------------------------------------
> Enable your software for Intel(R) Active Management Technology to meet the
> growing manageability and security demands of your customers. Businesses
> are taking advantage of Intel(R) vPro (TM) technology - will your software
> be a part of the solution? Download the Intel(R) Manageability Checker
> today! http://p.sf.net/sfu/intel-dev2devmar
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>