AWT Quant is a next-generation quantitative research & financial modeling platform.
It combines the best of stochastic modeling, LLMs, portfolio analysis, and autonomous agents
to deliver a comprehensive quant toolkit for researchers, analysts, and traders.

💹 From SPDE Monte Carlo simulations to TimeGPT-style assistants, portfolio optimization, stress testing, and macroeconomic integration — AWT Quant covers it all.

• ✨ Features
• ⚡️ Installation
• 🚀 Quick Start
• 📊 Modules
• 🧠 LLM Forecasting
• 🧪 Use Cases
• 🛠 Development
• 📖 Documentation
• 📄 License
• 🙏 Acknowledgements

✔️ Stochastic Models: GBM, Heston, CIR, OU, MJD
✔️ Forecasting Engines: GARCH, SPDE, LLM-based, macro-informed
✔️ Portfolio Optimization: Constraints, objectives, efficient frontier visualization
✔️ Risk Analytics: Stress testing, VaR, CVaR, tearsheets
✔️ Multi-Factor Analysis: PCA, clustering, attribution, localized models
✔️ Macroeconomic Data: 800k+ time series from FRED, IMF, WorldBank, OECD
✔️ Agent Workflows: Autonomous research, trading, portfolio management

pip install awt-quant
# or with poetry
poetry add awt-quant

from awt_quant.forecast.stochastic.run_simulations import SPDEMCSimulator

sim = SPDEMCSimulator(
    symbol="AAPL",
    start_date="2022-01-01",
    end_date="2022-03-01",
    dt=1,
    num_paths=1000,
    eq="heston"
)

sim.download_data()
sim.set_parameters()
sim.simulate()
sim.plot_simulation()

from awt_quant.forecast.portfolio.portfolio_forecast import PortfolioForecaster
from awt_quant.portfolio.optimization.optimize import PortfolioOptimizer

forecaster = PortfolioForecaster(["AAPL", "MSFT", "TSLA", "AMZN", "GOOG"])
forecast_results = forecaster.forecast(horizon=30)

optimizer = PortfolioOptimizer(
    assets=["AAPL", "MSFT", "TSLA", "AMZN", "GOOG"],
    objective="sharpe",
    constraints={"max_volatility": 0.15, "max_per_asset": 0.25},
    forecast_data=forecast_results
)

weights = optimizer.optimize()
optimizer.plot_efficient_frontier()
optimizer.plot_allocation()

• Lag-Llama → Time series foundation model with macro context
• TimeGPT Assistant → Forecasts with narratives & natural language queries
• Macro Models → Granger causality, transfer entropy, PCMCI

• 🔮 Forecasting: Ensemble statistical + stochastic + LLM methods
• 💼 Portfolio Optimization: Forecast-informed allocation, constraints, Black-Litterman
• 🔍 Multi-Factor Analysis: Clustering, attribution, sensitivity analysis
• 📊 Risk Analytics: VaR, CVaR, drawdowns, stress tests
• 🤖 Autonomous Agents: Research reports, trading strategies, monitoring

git clone https://github.com/pr1m8/awt-quant.git
cd awt-quant
poetry install
pytest tests/

👉 awt-quant.readthedocs.io

Includes:

• API reference
• Interactive notebooks
• Tutorials & guides
• Case studies
• Benchmarks

MIT © 2025 William R. Astley / Pr1m8

• Lag-Llama
• AutoTS
• PyPortfolioOpt
• FRED API