This project implements a modified Monte Carlo simulation algorithm for pricing American options. Unlike the standard recursive Monte Carlo approach, this method integrates a Gated Recurrent Unit (GRU) neural network to model and enhance the stochastic flow of stock price trajectories.
The idea is inspired by and built upon the research paper: “Pricing American Options using Machine Learning Algorithms” by Prudence Djagba and Callixte Ndizihiwe
This repository contains -
- Code for training the GRU
- Code for the Quantconnect Implementation
Below is a backtest done from January 2023 to December 2024 that showcases the code working
