Swap Price QuantLib vs. Bloomberg
Locked
Swap Price QuantLib vs. Bloomberg
 
|
Re: Swap Price QuantLib vs. Bloomberg
|
|
Hi
the zip file seems not to work. Here is another version David -----Original Message----- From: dhoorens [mailto:[hidden email]] Sent: mercredi 24 mars 2010 13:04 To: [hidden email] Subject: [Quantlib-users] Swap Price QuantLib vs. Bloomberg Hi I'm trying to replicate a Swap price I can see on my Bloomberg page I 'm not using the Quantlib xl code, because some people in my company are not yet ready to use objects in Excel (pfff) => I have created my own interface with Excel Here is the code double vanillaSwapPrice(int todaysDate_, int tenorInYears, double fixRate, int forwardTenorInMonths, int fltFrequency_, xloper* yieldTermStructure, double nominal, char* type, int settlementDays, int fixFrequency_, char* calendar_, char* dayCounter_, char* fixDayCounter_, char* fltDayCounter_, double spread, char* results){ try{ QuantLib::Calendar calendar = toCalendar(calendar_, QuantLib::TARGET()); QuantLib::Calendar fixCalendar = calendar; QuantLib::Calendar fltCalendar = calendar; QuantLib::RelinkableHandle<QuantLib::YieldTermStructure> yTS; QuantLib::Frequency fixFrequency = toFrequency(fixFrequency_); QuantLib::Frequency fltFrequency = toFrequency(fltFrequency_, QuantLib::Semiannual); QuantLib::DayCounter dayCounter = toDayCounter(dayCounter_, QuantLib::Actual360()); QuantLib::DayCounter fixDayCounter = toDayCounter(fixDayCounter_, QuantLib::Thirty360(QuantLib::Thirty360::BondBasis)); QuantLib::DayCounter fltDayCounter = toDayCounter(fltDayCounter_, QuantLib::Actual360()); boost::shared_ptr<QuantLib::IborIndex> index = makeIborIndex(QuantLib::Period(fltFrequency),yTS) .withSettlementDays(settlementDays) .withCalendar(calendar) .withDayCounter(fltDayCounter); QuantLib::Date todaysDate = toDate(todaysDate_); QuantLib::Settings::instance().evaluationDate() = todaysDate; QuantLib::Date settlement = calendar.advance(todaysDate,settlementDays,QuantLib::Days); yTS.linkTo(toYieldTermStructure(settlement, calendar, dayCounter, yieldTermStructure)); QuantLib::VanillaSwap::Type typ; std::string type_ = std::string(type); if (type_ == "R") typ = QuantLib::VanillaSwap::Receiver; else if (type_ == "P") typ = QuantLib::VanillaSwap::Payer; else throw (std::string("unknwown type of Swap")); if (nominal ==0.0) nominal = 1.0; boost::shared_ptr<QuantLib::VanillaSwap> sw = makeVSwap(QuantLib::Period(tenorInYears,QuantLib::Years), index, fixRate, QuantLib::Period(forwardTenorInMonths,QuantLib::Months)) .withType(typ) .withNominal(nominal) .withFixedLegTenor(QuantLib::Period(fixFrequency)) .withFixedLegCalendar(fixCalendar) .withFixedLegDayCount(fixDayCounter) .withFloatingLegTenor(QuantLib::Period(fltFrequency)) .withFloatingLegSpread(QuantLib::Spread(spread)) .withFloatingLegCalendar(fltCalendar) .withFloatingLegDayCount(fltDayCounter); //Depending on what I want as results i can choose between return sw->NPV(); //return sw->fixedLegNPV(); //return sw->fixedLegBPS(); //return sw->floatingLegNPV(); //return sw->floatingLegBPS(); //return sw->fairRate(); } catch (...) { errMsg += "[" + functionCall->functionName() + "]" + "Unknown Error"; } //not the real return but adapted here to say somthg :-/ return -1; } The toXXX functions (i.e. toFrequency, toDayCounter, etc) are converter of the inputs into QuantLib types. The (optionnal) second input is the default value, if the first value is not right Here is the toYieldTermStructure function boost::shared_ptr<YieldTermStructure> toYieldTermStructure (Date settlement, Calendar calendar, DayCounter dayCounter, xloper* input, Compounding comp= QuantLib::Continuous, Frequency freq== QuantLib::Annual, bool endOfMonth= true){ std::vector<std::vector<double> >m = ObjectHandler::operToMatrix<double>(*input ,"Input"); unsigned int rows, cols, sizeInput; rows = m.size(); cols = m[0].size(); sizeInput = __max(rows,cols); boost::shared_ptr<YieldTermStructure> curve; if (sizeInput==1) curve = boost::shared_ptr<YieldTermStructure>( new FlatForward(settlement, m[0][0], dayCounter, comp, freq)); else{ std::vector<Date> vDate; std::vector<Real> vPrice; bool isDate = (m[0][0] >=367); //the first X-value of the curve is greater than 1/1/1901 => it's a date if (rows ==2 && cols > 2){ //horizontal data int j_init = 0; if (m[1][0] != 1.0){ vDate.push_back(settlement); vPrice.push_back(1.0); j_init++; } else { if ((isDate && toDate((long)m[0][0]) != settlement) || (!isDate && m[0][0] != 0.0)) throw (std::string("[toYieldTermStructure] Invalid first term value")); } for (unsigned int j = j_init; j< cols; j++){ if (isDate) vDate.push_back(toDate((long)m[0][j])); else //It is not a date => we assume the value is in Months and we add it to the settlement vDate.push_back(calendar.advance(settlement, (long)m[0][j], Months, QuantLib::Unadjusted, endOfMonth)); vPrice.push_back(m[1][j]); } } else if (cols ==2 && rows > 2){ //vertical data int i_init = 0; if (m[0][1] != 1.0){ vDate.push_back(settlement); vPrice.push_back(1.0); i_init ++; } else { if ((isDate && toDate((long)m[0][0]) != settlement) || (!isDate && m[0][0] != 0.0)) throw (std::string("[toYieldTermStructure] Invalid first term value")); } for (unsigned int i = i_init; i< rows; i++){ if (isDate) vDate.push_back(toDate((long)m[i][0])); else //It is not a date => we assume the value is in Months and we add it to the settlement vDate.push_back(calendar.advance(settlement, (long)m[i][0], Months, QuantLib::Unadjusted, endOfMonth)); vPrice.push_back(m[i][1]); } } else throw (std::string("[toYieldTermStructure] data has too much either cols or rows")); curve = boost::shared_ptr<YieldTermStructure>( new InterpolatedDiscountCurve<Cubic>( vDate, vPrice, dayCounter, calendar, Cubic(CubicInterpolation::Spline, false, CubicInterpolation::SecondDerivative, 0.0, CubicInterpolation::SecondDerivative, 0.0))); } return curve; } Finally in attach you can find the makeVSwap and makeIborIndex classes When I enter in Excel as inputs: int todaysDate_ = 40261 (ie 24/3/2010) int tenorInYears = 5 double fixRate = 2.21207% int forwardTenorInMonths = 0 int fltFrequency_ = 2 xloper* yieldTermStructure = the curve given by the the second gif file double nominal = 10000000 char* type = R int settlementDays = 0 int fixFrequency_ = 1 char* calendar_ = TARGET char* dayCounter_ =THIRTY360 char* fixDayCounter_ = THIRTY360 char* fltDayCounter_ = ACTUAL360 double spread = 0 it returns a value for NPV = -9252218 FixLegNPV = 1042594.49 FixLegBps = 4713.22 FltLegNPV = -1135118.66 FltLegBps = -4809.48 FairRate = 2.40837 Here are the questions Why do I have differences with Bloomberg. Is QuantLib coherent with Blmbg? Why the market value of each leg is so different (Blmg vs QuantLib)? (example QuantLIb FixLegNPV = 1042596.49 and Blmbg FixLegNPV = 9904395.99) Tks David http://old.nabble.com/file/p28014072/QL_Blmbg.zip QL_Blmbg.zip -- View this message in context: http://old.nabble.com/Swap-Price-QuantLib-vs.-Bloomberg-tp28014072p28014 072.html Sent from the quantlib-users mailing list archive at Nabble.com. ------------------------------------------------------------------------ ------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users -----Disclaimer----- This message may contain confidential information intended solely for the use of the named addressee. If you are not the intended recipient, you should not read, use, disclose or reproduce the content of this message. If you have received this message by mistake, please notify the sender immediately. Any views or opinions presented in this message are solely those of the author and do not necessarily represent those of AXA Belgium, AXA Bank Europe, AXA Tech Belgium GIE - ESV or any other entity of the AXA Group, unless otherwise stated by the sender and duly authorized by the said companies. --------------------- ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
QL_Blmbg.zip (71K)
Re: Swap Price QuantLib vs. Bloomberg
|
|
Re: Swap Price QuantLib vs. Bloomberg
|
|
Re: Swap Price QuantLib vs. Bloomberg
|
|
Re: Swap Price QuantLib vs. Bloomberg
|
|
Re: Swap Price QuantLib vs. Bloomberg
|
|
| Free forum by Nabble | Edit this page |