Derivatives pricing, priced correctly
A pricing and risk engine, exposed over REST and gRPC. Vanilla through autocallables, thirteen model families, full Greeks, portfolio VaR and XVA — each pricer implementing a published method, and named against it.
Instruments
Vanilla, barrier, Asian, lookback, binary, spread, variance swap, autocallable, convertible bond, IR swap, cap/floor, CDS, inflation.
Models
Black-Scholes, Bachelier, Heston, SABR, Merton, Bates, Dupire, Hull-White, CIR, Vasicek, LMM, G2++, Jarrow-Yildirim.
Pricers
Analytical, Monte Carlo including Longstaff-Schwartz, finite difference, binomial and trinomial trees, and FFT.
Risk
Full Greeks, portfolio VaR, stress testing, hedge optimisation, and XVA across CVA, DVA, FVA and KVA.
Contract DSL
Composable algebraic contracts with a serialisable spec layer, so a custom payoff is data rather than a code change.
Methods you can check, and a status that means something
Every pricer implements a published method and is named against its source — Black & Scholes (1973), Garman & Kohlhagen (1983), Heston (1993), Dupire (1994) and the rest. The capability matrix carries the citation for each one, so a model choice can be checked against the literature rather than taken on description.
A capability marked validated is compared against an independent implementation of the same method on every build. One marked experimental is not, and the matrix says so against that capability, with the approximation responsible written beside it.
The matrix is behind sign-in. It states, for each of the ninety-three capabilities, the method, its validation status, the market data it needs and its stated limitations.