ai-hedge-fund
Applicationvirattt/ai-hedge-fund
AI-powered hedge fund simulation with multiple investing agents.
Overview
An AI-powered hedge fund proof-of-concept that simulates trading decisions using agents modeled after famous investors. It includes CLI and web interfaces for backtesting and exploring multi-agent strategies, intended solely for educational purposes.
README Preview
# AI Hedge Fund\n\nThis is a proof of concept for an AI-powered hedge fund. The goal of this project is to explore the use of AI to make trading decisions. This project is for **educational** purposes only and is not intended for real trading or investment.\n\n> **🚧 The project is evolving.** We're rebuilding it into a persistent, always-on AI hedge fund — a *fund* as a first-class entity you can backtest, paper-trade, and (opt-in) run live, with the investor agents reimagined as pluggable, backtestable "alpha models." Read the **[Vision →](VISION.md)** and the **[Roadmap →](ROADMAP.md)**.\n\nThis system employs several agents working together:\n\n1. Aswath Damodaran Agent - The Dean of Valuation, focuses on story, numbers, and disciplined valuation\n2. Ben Graham Agent - The godfather of value investing, only buys hidden gems with a margin of safety\n3. Bill Ackman Agent - An activist investor, takes bold positions and pushes for change\n4. Cathie Wood Agent - The queen of growth investing, believes in the power of innovation and disruption\n5. Charlie Munger Agent - Warren Buffett's partner, only buys wonderful businesses at fair prices\n6. Michael Burry Agent - The Big Short contrarian who hunts for deep value\n7. Mohnish Pabrai Agent - The Dhandho investor, who looks for doubles at low risk\n8. Nassim Taleb Agent - The Black Swan risk analyst, focuses on tail risk, antifragility, and asymmetric payoffs\n9. Peter Lynch Agent - Practical investor who seeks "ten-baggers" in everyday businesses\n10. Phil Fisher Agent - Meticulous growth investor who uses deep "scuttlebutt" research \n11. Rakesh Jhunjhunwala Agent - The Big Bull of India\n12. Stanley Druckenmiller Agent - Macro legend who hunts for asymmetric opportunities with growth potential\n13. Warren Buffett Agent - The oracle of Omaha, seeks wonderful companies at a fair price\n14. Valuation Agent - Calculates the intrinsic value of a stock and generates trading signals\n15. Sentiment Agent - Analyzes marke
FAQ (2)
TroubleshootingHow do I prevent LLM training-data memory from biasing backtest results?
Use the opt-in blind mode introduced in PR #721. Run the CLI with --blind, or set HEDGE_FUND_BLIND=1 for the interactive/TUI workflow. In blind mode, FundamentalsSnapshot.render() removes the ticker and industry and replaces report_period/filing_date labels with relative labels such as t-0, t-1, and t-2. The default prompt and existing cached results remain unchanged. This reduces look-ahead recall but does not eliminate it: sector is still included, and distinctive financial figures may identify large companies. For the cleanest validation, evaluate backtest windows after the model's training cutoff; pre-cutoff results may contain memorized future outcomes.
TroubleshootingHow to fix a GitHub repository showing no license when the README says MIT?
Add a standard MIT license file named LICENSE at the repository root. It must contain the MIT text with the actual copyright holder and year, for example:
MIT License
Copyright (c) 2025 virattt
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.Commit and push the file. Existing license files inside dependency directories, such as app/frontend/LICENSE for React Flow, only cover that bundled dependency and do not grant a license for the project's own code.