This project is an end-to-end automated quant trading pipeline.
It doesn’t just backtest one strategy — it creates, allocates, and evaluates entire portfolios of strategies using AI.
-
Strategy Generation 🧬
- A custom neural generator (FGEG) invents trading strategies.
- Instead of a human tweaking indicators, the system learns profitable patterns directly.
-
Capital Allocation 📊
- A reinforcement learning agent (PPO) decides how much capital each strategy gets.
- Ensures diversification, risk control, and max returns.
-
Evaluation ✅
- Final portfolio is stress-tested on unseen market data.
- Reports total PnL, annualized returns, and drawdowns.
git clone https://github.com/yourusername/quant-strategy-gen.git
cd quant-strategy-gen
python main.py