MacroQuant is an advanced analytical framework designed to decode the complex relationships between U.S. Public Debt dynamics, Global Liquidity cycles, and the price action of major financial assets (Bitcoin, Equities, and Bonds).
Using a blend of Econometrics and Machine Learning, this project aims to quantify how the "Invisible Hand" of central bank liquidity dictates market regimes.
- Central Bank Balances: Tracking the Fed, ECB, and BoJ to calculate global net liquidity.
- TGA & Reverse Repo: Analyzing the U.S. Treasury General Account and RRP facility to predict short-term market stress.
- Fiscal Dominance: Modeling the impact of U.S. Public Debt expansion on long-term Treasury yields and inflation expectations.
- Correlations: Quantifying how Bitcoin and Gold act as "Liquidity Barometers" compared to traditional Equities.
- Feature Engineering: Extracting alpha from macro-indicators (CPI, PMI, Yield Curves).
- Algorithms: Utilizing LSTM (Long Short-Term Memory) networks and XGBoost for non-linear price prediction.
| Field | Technologies |
|---|---|
| Data Science | Python, Pandas, NumPy, Scipy |
| Machine Learning | Scikit-learn, TensorFlow/PyTorch (LSTMs), XGBoost |
| Data Sourcing | FRED API, Yahoo Finance, Glassnode (On-chain) |
| Visualization | Plotly (Interactive Charts), Matplotlib, Seaborn |
data/: Multi-source datasets (Macro, On-chain, and TradFi).notebooks/: Exploratory Data Analysis (EDA) and Correlation Heatmaps.models/: Serialized predictive models and backtesting results.scripts/: Modular pipelines for automated feature scaling and training.
- Phase 1: Multi-asset data ingestion pipeline (BTC, SPX, GOLD, DXY).
- Phase 2: Correlation matrix analysis under different liquidity regimes.
- Phase 3: (In Progress) Integration of U.S. Debt Cycle & Net Liquidity indicators.
- Phase 4: Deployment of the Price Prediction Engine and Backtesting framework.