DeepBermuda is a neural network-enhanced implementation of the Longstaff-Schwartz Monte Carlo (LSM) method for pricing Bermudan options, with a focus on high-dimensional, over-the-counter (OTC) derivatives. This project extends traditional LSM by replacing polynomial regression with a feedforward neural network to more accurately estimate continuation values in complex multi-asset settings. It is specifically designed to handle basket-style Bermudan options with many underlyings, reflecting real-world OTC structures where early exercise occurs at discrete intervals. The goal is to explore the intersection of deep learning and advanced option pricing for flexible, high-dimensional derivatives.
To see findings and results check Conclusions.pdf
-
Clone the repository
git clone https://github.com/TradersAtUGA/FNN-Enhanced-LSM.git
cd your-repo-name -
Create a virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate -
Install dependencies
pip install -r requirements.txt -
Run the project
python main.py
The documentation for the params is in the config_docs.md file
You can customize the params by editing the config.yaml file
Due to the nature of American options—being exercisable at any time—it is extremely difficult and often unrealistic to accurately price a multi-asset American option basket.
If you attempt to test such a configuration, be aware that it may consume a significant amount of computational resources.