Finance Simulator

This project runs a Monte Carlo simulation using Geometric Brownian Motion to generate potential future stock prices based on historical data. The simulation is visualized via a web interface, allowing a user to see possible price movements for a given stock.

Features

• Monte Carlo Simulation – Generates price paths using historical volatility and log returns.
• Web Interface – Displays historical stock data and simulation results using Svelte and Chart.js.
• Thumper – A lightweight OCaml web server built specifically for this project.
• Spice – A simple (and opinionated) logging utility.
• Real-time Streaming – Simulation results are streamed to the client as they are generated.

---

Running the project

Requirements

• OCaml and Dune OR

• Docker

• A JavaScript package manager (for example, pnpm, bun, npm)

Run

• First, the Javascript files need to be built

; pushd web
; <bundler> install
; <bundler> run build
; popd

• If you have OCaml (and Dune) installed, you can run with Dune

; dune exec finance_sim -- serve

• If you have Docker installed, you can run with Docker

; docker compose up

Screenshots

Next steps and limitations

• Graphing performance - Currently, simulated data is added to the graph slowly. Optimizing rendering performance is a priority.
• Stock selection - Right now the simulation only supports IBM stock data ending in early February 2025. Expanding stock selection is a future goal.
• UI enhancements - Improving the interface for a cleaner, more intuitive experience.