Possible Enhancements

Class AmericanCondition
unify the intrinsicValues/Payoff thing

Class AmericanExercise
check that everywhere the American condition is applied from earliestDate and not earlier

Class AmericanPayoffAtExpiry
calculate greeks

Class AmericanPayoffAtHit
calculate greeks

Class AnalyticContinuousGeometricAveragePriceAsianEngine
handle seasoned options

Class AnalyticDigitalAmericanEngine
add more greeks (as of now only delta and rho available)

Class AnalyticDiscreteGeometricAveragePriceAsianEngine
implement correct theta, rho, and dividend-rho calculation

Class BicubicSpline
revise end conditions

Class BinomialVanillaEngine
Greeks are not overly accurate. They could be improved by building a tree so that it has three points at the current time. The value would be fetched from the middle one, while the two side points would be used for estimating partial derivatives.

Class BivariateCumulativeNormalDistributionDr78
check accuracy of this algorithm and compare with: 1) Drezner, Z, (1978), Computation of the bivariate normal integral, Mathematics of Computation 32, pp. 277-279. 2) Drezner, Z. and Wesolowsky, G. O. (1990) `On the Computation of the Bivariate Normal Integral', Journal of Statistical Computation and Simulation 35, pp. 101-107. 3) Drezner, Z (1992) Computation of the Multivariate Normal Integral, ACM Transactions on Mathematics Software 18, pp. 450-460. 4) Drezner, Z (1994) Computation of the Trivariate Normal Integral, Mathematics of Computation 62, pp. 289-294. 5) Genz, A. (1992) `Numerical Computation of the Multivariate Normal Probabilities', J. Comput. Graph. Stat. 1, pp. 141-150.

Class BlackCallableFixedRateBondEngine
set additionalResults (e.g. vega, fairStrike, etc.)

Class BlackVarianceCurve
check time extrapolation

Class BlackVarianceSurface
check time extrapolation

Member BoundaryCondition::Side
Generalize for n-dimensional conditions

Class CallableBond
models/shortrate/calibrationHelpers

Class CallableBond
OAS/OAD

Class CallableBond
floating rate callable bonds ?

Class CashFlows
add tests

Class CDO
Investigate and fix cases $ E_{i+1} < E_i. $

Class Cdor
check settlement days, end-of-month adjustment, and day-count convention.

Class CliquetOption
  • add local/global caps/floors
  • add accrued coupon and last fixing

Class ContinuousAveragingAsianOption
add running average

Class CubicInterpolation
implement missing schemes (FourthOrder and ModifiedParabolic) and missing boundary conditions (Periodic and Lagrange).

Class DirichletBC
generalize to time-dependent conditions.

Class ExplicitEuler
add Richardson extrapolation

Class FDDividendEngineBase
The dividend class really needs to be made more sophisticated to distinguish between fixed dividends and fractional dividends

Class FDDividendEngineShiftScale
Review literature to see whether this is described

Class FittedBondDiscountCurve
refactor the bond helper class so that it is pure virtual and returns a generic bond or its cash flows. Derived classes would include helpers for fixed-rate and zero-coupon bonds. In this way, both bonds and bills can be used to fit a discount curve using the exact same machinery. At present, only fixed-coupon bonds are supported. An even better way to move forward might be to get rate helpers to return cashflows, in which case this class could be used to fit any set of cash flows, not just bonds.

Class FittedBondDiscountCurve
add more fitting diagnostics: smoothness, standard deviation, student-t test, etc. Generic smoothness method may be useful for smoothing splines fitting. See Fisher, M., D. Nychka and D. Zervos: "Fitting the term structure of interest rates with smoothing splines." Board of Governors of the Federal Reserve System, Federal Resere Board Working Paper, 95-1.

Class FittedBondDiscountCurve
add extrapolation routines

Class FittedBondDiscountCurve::FittingMethod
derive the special-case class LinearFittingMethods from FittingMethod. A linear fitting to a set of basis functions $ b_i(t) $ is any fitting of the form

\[ d(t) = \sum_{i=0} c_i b_i(t) \]

i.e., linear in the unknown coefficients $ c_i $. Such a fitting can be reduced to a linear algebra problem $ Ax = b $, and for large numbers of bonds, would typically be much faster computationally than the generic non-linear fitting method.

Class FixedRateBondForward
Add preconditions and tests

Class FixedRateBondForward
Create switch- if coupon goes to seller is toggled on, don't consider income in the $ P_{DirtyFwd}(t) $ calculation.

Class FixedRateBondForward
Verify this works when the underlying is paper (in which case ignore all AI.)

Class Forward
Add preconditions and tests

Member GeneralizedBlackScholesProcess::drift (Time t, Real x) const
revise extrapolation

Member GeneralizedBlackScholesProcess::diffusion (Time t, Real x) const
revise extrapolation

Class GenericRiskStatistics
add historical annualized volatility

Class IntegralEngine
define tolerance for calculate()

Class InterestRateIndex
add methods returning InterestRate

Class InverseCumulativeStudent
Find/implement an efficient algorithm for evaluating the cumulative Student t-distribution, replacing the Newton iteration

Class Jibar
check settlement days and day-count convention.

Class MCVarianceSwapEngine
define tolerance of numerical integral and incorporate it in errorEstimate

Class MixedScheme
  • derive variable theta schemes
  • introduce multi time-level schemes.

Class MultiCubicSpline
  • allow extrapolation as for the other interpolations
  • investigate if and how to implement Hyman filters and different boundary conditions

Class NeumannBC
generalize to time-dependent conditions.

Class OneFactorCopula
Improve on simple Euler integration

Class OneFactorGaussianStudentCopula
Improve performance/accuracy of the calculation of inverse cumulative Y. Tabulate and store it for selected correlations?

Class OneFactorStudentCopula
Improve performance/accuracy of the calculation of inverse cumulative Y. Tabulate and store it for selected correlations?

Class OneFactorStudentGaussianCopula
Improve performance/accuracy of the calculation of inverse cumulative Y. Tabulate and store it for selected correlations?

Class RandomizedLDS
implement the other randomization algorithms

Member SampledCurve::valueAtCenter () const
replace or complement with a more general function valueAt(spot)

Member SampledCurve::firstDerivativeAtCenter () const
replace or complement with a more general function firstDerivativeAt(spot)

Member SampledCurve::secondDerivativeAtCenter () const
replace or complement with a more general function secondDerivativeAt(spot)

Class Solver1D
  • clean up the interface so that it is clear whether the accuracy is specified for $ x $ or $ f(x) $.
  • add target value (now the target value is 0.0)

Class SwapRateHelper
use input SwapIndex to create the swap

Class Swaption
add greeks and explicit exercise lag

Class SyntheticCDO
Investigate and fix cases $ E_{i+1} < E_i. $

Class Tibor
check settlement days and end-of-month adjustment.

Class TimeGrid
what was the rationale for limiting the grid to positive times? Investigate and see whether we can use it for negative ones as well.

Class TRLibor
check end-of-month adjustment.

Class UnitedKingdom
add LIFFE

Class YoYCapFloorTermPriceSurface
deal with index interpolation.

Class Zibor
check settlement days, end-of-month adjustment, and day-count convention.

File syntheticcdoengines.hpp
Add further engines for analytical expected tranche loss cases - large homogeneous pool with Normal Inverse Gaussian, Gamma copula