Re: Building Yield Curve of Zero Rates (Theo)

Posted by tallent_e on Jan 09, 2013; 10:00am
URL: http://quantlib.414.s1.nabble.com/Re-Building-Yield-Curve-of-Zero-Rates-Theo-tp13887.html

Hi Theo.

Notice that the result you get is close to your benchmark.

First, check you use the right curve library: http://quantlib.org/reference/class_quant_lib_1_1_yield_term_structure.html as suggested by another person.

Then, watch your evaluation date. You don't show your complete snippet but this might be tricky (and firm-specific):

	// http://quantcorner.wordpress.com/2011/01/31/us-treasury-notes-and-quantlib/
    // Evaluation date
	Integer fixingDays = 1;
	Natural settlementDays = 1;
	Date todaysDate = calendar.advance(settlementDate, -fixingDays, Days);
	Settings::instance().evaluationDate() = todaysDate;

Last but not least, pay attention to the business day convention: adjusted/unajusted, modified, modified-following ?

I hope this helps a bit,

Best,
Édouard

--
http://quantcorner.wordpress.com
twitter: @tallent_e
linkedin: fr.linkedin.com/pub/édouard-tallent/8/145/7a/


4. ??: FW: Building Yield Curve of Zero Rates (Cheng Li)

---

Hi All,

I have Sample of USD Zero Yield Curve Data as follows: 

Observation Date
03/08/2012

Maturity Days
0,3,4,11,35,67,96,137,227,320,411,501,591,734,1099,1466,1830,2195,2560,2
926,3293,3657,4387,5484,7311,9135,10961,

Yield
0,1.57151762841849E-03,1.5841928756397E-03,1.80193133074194E-03,2.374223
75119134E-03,3.30406684491067E-03,4.35679286119092E-03,3.56334288917869E
-03,3.69745160830227E-03,3.77125648757358E-03,3.84515157443031E-03,3.928
32387415253E-03,4.03993077763243E-03,4.35625780531457E-03,5.080685648854
12E-03,6.4706614338575E-03,8.38591062422842E-03,0.010332793648027,1.2171
0204339663E-02,1.38041853050158E-02,1.52910208160213E-02,1.6727919706485
3E-02,1.89718781527944E-02,2.12626395949329E-02,2.31753447176866E-02,2.4
1372894297357E-02,2.47383920090547E-02

InterpolationType
Linear

Extrapolation Type
near

DayCount Convention
ACT365FIXED

Compunding Frequency
CONTINUOUS,

So I build a vector<QuantLib::Date> yieldMaturities, vector<Rate> yields
from the information above.
In the case of vector<QuantLib::Date> yieldMaturities, I just add the
Maturity Days to observation date.

I now build the yieldCurve as

Handle<YieldTermStructure> yieldCurve(
boost::shared_ptr<YieldTermStructure>(
new ForwardCurve(yieldMaturities,yields, 
dayCountConv)));

Then pass to 

boost::shared_ptr<BlackProcess> blackProcess(
new BlackProcess(underlyingH, yieldCurve,
flatVolTS));

as I am cant determine what the zero rate will be for the exercise date
of the option.

Now given this option is on a commodity forward, I want to discount from
payment date.

Hence have 

DiscountFactor expiryDiscount =
blackProcess->riskFreeRate()->discount(europeanExercise->lastDate());
DiscountFactor paymentDiscount =
blackProcess->riskFreeRate()->discount(paymentDate);

Real forwardValue = europeanOption.NPV()/expiryDiscount;
Real presentValue = forwardValue * paymentDiscount;

My PV is slightly off the PV from another system.

The market data is the same in both systems, except for the difference
in discount factors.

The paymentDiscount factor in Quantlib is 0.9990450902703203, whilst
that from the other system is 0.9987391782033116, hence diff in PV.

I am wondering whether my yieldCurve of Zeros has been constructed
properly. 

Also its seems YieldTermStructure takes in market quotes and boostraps
zero rates from it. So Given I already have Zero rates, have I
contstructed the Zero Yield Curve propely? Looks like I have not done
something right.
Regards
Theo
Regards

Theo
------------------------------------------------------------------------------
Master Java SE, Java EE, Eclipse, Spring, Hibernate, JavaScript, jQuery
and much more. Keep your Java skills current with LearnJavaNow -
200+ hours of step-by-step video tutorials by Java experts.
SALE $49.99 this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122612 
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users