interest-rate-models

Ornstein-Uhlenbeck mean-reverting stochastic process with exact transition distributions and Vasicek calibration

Ornstein-Uhlenbeck mean-reverting stochastic process with exact transition distributions and Vasicek calibration

This project aims to implement the Hull & While One Factor model and apply it to price Bermudan Swaptions.

The LIBOR Market Model (BGM) from first principles in pure Python/NumPy — measure-consistent forward-LIBOR drifts, Black-76 caplets, Rebonato swaptions, all pinned by algebraic identities. Adversarially verified before a line of code. The discrete-tenor cousin of hjm-lab.

Excel-based quantitative finance models: Brownian motion simulation, Vasicek short-rate Monte Carlo, and a full Basel SA-CCR counterparty exposure engine.

Stochastic short-rate term-structure models from first principles: Vasicek + CIR closed-form affine bond prices, the unifying affine ODE framework that reproduces both, Vasicek bond options, and calibration. Every formula identity-tested — the affine ODE IS the closed form. No rates library underneath.

Quantitative finance projects on yield curve modeling (Nelson-Siegel, Nelson-Siegel-Svensson), FRA valuation, and the Gaussian Heath-Jarrow-Morton (HJM) framework with caplet pricing. EPFL Interest Rate and Credit Risk Models course

The two-factor Gaussian (G2++) short-rate model from first principles: exact curve fit, ZCB options, caps/floors, and a European swaption priced three independent ways that agree (Brigo-Mercurio 1-D, 2-D quadrature, Monte Carlo). 67 identity tests, adversarially verified. No rates library underneath.

Heath-Jarrow-Morton forward-rate framework from scratch: the no-arbitrage drift condition, Gaussian-HJM bond prices & ZCB options, a Musiela forward-curve Monte Carlo, and reproductions of Ho-Lee / Hull-White (=Vasicek option vol) / G2++ from the vol structure. 35 identity tests, adversarially verified. No rates library.