Ejemplo n.º 1
0
        // McSimulation implementation
        protected override TimeGrid timeGrid()
        {
            Date          referenceDate = process_.riskFreeRate().link.referenceDate();
            DayCounter    voldc         = process_.blackVolatility().link.dayCounter();
            List <double> fixingTimes   = new  InitializedList <double>(arguments_.fixingDates.Count());

            for (int i = 0; i < arguments_.fixingDates.Count(); i++)
            {
                if (arguments_.fixingDates[i] >= referenceDate)
                {
                    double t = voldc.yearFraction(referenceDate,
                                                  arguments_.fixingDates[i]);
                    fixingTimes.Add(t);
                }
            }
            // handle here maxStepsPerYear
            return(new TimeGrid(fixingTimes.Last(), fixingTimes.Count()));
        }
        public override void calculate()
        {
            /* this engine cannot really check for the averageType==Geometric
             * since it can be used as control variate for the Arithmetic version
             *  QL_REQUIRE(arguments_.averageType == Average::Geometric,
             *             "not a geometric average option");
             */

            if (!(arguments_.exercise.type() == Exercise.Type.European))
            {
                throw new ApplicationException("not an European Option");
            }

            double runningLog;
            int    pastFixings;

            if (arguments_.averageType == Average.Type.Geometric)
            {
                if (!(arguments_.runningAccumulator > 0.0))
                {
                    throw new ApplicationException("positive running product required: "
                                                   + arguments_.runningAccumulator + " not allowed");
                }
                runningLog =
                    Math.Log(arguments_.runningAccumulator.GetValueOrDefault());
                pastFixings = arguments_.pastFixings.GetValueOrDefault();
            }
            else      // it is being used as control variate
            {
                runningLog  = 1.0;
                pastFixings = 0;
            }

            PlainVanillaPayoff payoff = (PlainVanillaPayoff)(arguments_.payoff);

            if (payoff == null)
            {
                throw new ApplicationException("non-plain payoff given");
            }

            Date          referenceDate = process_.riskFreeRate().link.referenceDate();
            DayCounter    rfdc          = process_.riskFreeRate().link.dayCounter();
            DayCounter    divdc         = process_.dividendYield().link.dayCounter();
            DayCounter    voldc         = process_.blackVolatility().link.dayCounter();
            List <double> fixingTimes   = new InitializedList <double>(arguments_.fixingDates.Count());
            int           i;

            for (i = 0; i < arguments_.fixingDates.Count(); i++)
            {
                if (arguments_.fixingDates[i] >= referenceDate)
                {
                    double t = voldc.yearFraction(referenceDate,
                                                  arguments_.fixingDates[i]);
                    fixingTimes.Add(t);
                }
            }

            int    remainingFixings = fixingTimes.Count();
            int    numberOfFixings  = pastFixings + remainingFixings;
            double N = numberOfFixings;

            double pastWeight   = pastFixings / N;
            double futureWeight = 1.0 - pastWeight;

            /*double timeSum = std::accumulate(fixingTimes.begin(),
             *                             fixingTimes.end(), 0.0);*/
            double timeSum = 0;

            fixingTimes.ForEach((ii, vv) => timeSum += fixingTimes[ii]);

            double vola = process_.blackVolatility().link.blackVol(
                arguments_.exercise.lastDate(),
                payoff.strike());
            double temp = 0.0;

            for (i = pastFixings + 1; i < numberOfFixings; i++)
            {
                temp += fixingTimes[i - pastFixings - 1] * (N - i);
            }
            double variance   = vola * vola / N / N * (timeSum + 2.0 * temp);
            double dsigG_dsig = Math.Sqrt((timeSum + 2.0 * temp)) / N;
            double sigG       = vola * dsigG_dsig;
            double dmuG_dsig  = -(vola * timeSum) / N;

            Date   exDate       = arguments_.exercise.lastDate();
            double dividendRate = process_.dividendYield().link.
                                  zeroRate(exDate, divdc, Compounding.Continuous, Frequency.NoFrequency).rate();
            double riskFreeRate = process_.riskFreeRate().link.
                                  zeroRate(exDate, rfdc, Compounding.Continuous, Frequency.NoFrequency).rate();
            double nu = riskFreeRate - dividendRate - 0.5 * vola * vola;

            double s = process_.stateVariable().link.value();

            if (!(s > 0.0))
            {
                throw new ApplicationException("positive underlying value required");
            }

            int    M   = (pastFixings == 0 ? 1 : pastFixings);
            double muG = pastWeight * runningLog / M +
                         futureWeight * Math.Log(s) + nu * timeSum / N;
            double forwardPrice = Math.Exp(muG + variance / 2.0);

            double riskFreeDiscount = process_.riskFreeRate().link.discount(
                arguments_.exercise.lastDate());

            BlackCalculator black = new BlackCalculator(payoff, forwardPrice, Math.Sqrt(variance),
                                                        riskFreeDiscount);

            results_.value = black.value();
            results_.delta = futureWeight * black.delta(forwardPrice) * forwardPrice / s;
            results_.gamma = forwardPrice * futureWeight / (s * s)
                             * (black.gamma(forwardPrice) * futureWeight * forwardPrice
                                - pastWeight * black.delta(forwardPrice));

            double Nx_1, nx_1;
            CumulativeNormalDistribution CND = new CumulativeNormalDistribution();
            NormalDistribution           ND  = new NormalDistribution();

            if (sigG > Const.QL_Epsilon)
            {
                double x_1 = (muG - Math.Log(payoff.strike()) + variance) / sigG;
                Nx_1 = CND.value(x_1);
                nx_1 = ND.value(x_1);
            }
            else
            {
                Nx_1 = (muG > Math.Log(payoff.strike()) ? 1.0 : 0.0);
                nx_1 = 0.0;
            }
            results_.vega = forwardPrice * riskFreeDiscount *
                            ((dmuG_dsig + sigG * dsigG_dsig) * Nx_1 + nx_1 * dsigG_dsig);

            if (payoff.optionType() == Option.Type.Put)
            {
                results_.vega -= riskFreeDiscount * forwardPrice *
                                 (dmuG_dsig + sigG * dsigG_dsig);
            }

            double tRho = rfdc.yearFraction(process_.riskFreeRate().link.referenceDate(),
                                            arguments_.exercise.lastDate());

            results_.rho = black.rho(tRho) * timeSum / (N * tRho)
                           - (tRho - timeSum / N) * results_.value;

            double tDiv = divdc.yearFraction(
                process_.dividendYield().link.referenceDate(),
                arguments_.exercise.lastDate());

            results_.dividendRho = black.dividendRho(tDiv) * timeSum / (N * tDiv);

            results_.strikeSensitivity = black.strikeSensitivity();

            results_.theta = Utils.blackScholesTheta(process_,
                                                     results_.value.GetValueOrDefault(),
                                                     results_.delta.GetValueOrDefault(),
                                                     results_.gamma.GetValueOrDefault());
        }