A collection of educational implementations for derivative pricing and risk modelling. The repository focuses on transparent numerical examples rather than a unified production pricing library.
- European option pricing with finite-difference methods
- Cox-Ross-Rubinstein binomial and trinomial trees
- Equity-option valuation using lattice methods
- Normal-inverse-Gaussian and other Lévy-process examples
- Credit-default-swap valuation under constant intensity
- Swaps, swaptions and related fixed-income exercises
git clone https://github.com/AIM-IT4/QuantitativeDerivativeModels.git
cd QuantitativeDerivativeModels
python -m pip install -r requirements.txtSome examples are self-contained notebooks and may require only the packages imported by that notebook. Where no requirements.txt is present for a specific example, create an isolated environment and install the displayed imports explicitly.
Start Jupyter from the repository root:
jupyter notebookThen open the notebook corresponding to the model you want to inspect. Each implementation should be treated as an educational reference and independently validated before use in trading, valuation, risk or regulatory workflows.
The repository contains multiple standalone models with different assumptions and levels of completeness. It is not a single calibrated pricing engine, and it does not provide a stable package-level API. Numerical results depend on the model assumptions, discretisation choices, market inputs and convergence settings used in each notebook.
Issues and pull requests are welcome, particularly for:
- reproducible numerical tests
- convergence and stability checks
- corrections to model assumptions or formulas
- clearer documentation of inputs and outputs
- comparisons against analytical benchmarks or established libraries
Please describe the financial instrument, model assumptions, expected result and validation method in each contribution.
This project is licensed under the MIT License. See LICENSE.md for details.
For technical questions, open a GitHub issue so the discussion remains searchable and useful to other readers.