The Score Risk System is a high-frequency, event-driven quantitative trading engine developed in Python. It is designed to navigate the high-volatility environments of Bitcoin (BTC) and Ethereum (ETH) by utilizing a proprietary multi-factor Risk Score to dictate position sizing and direction.
As of March 2026
| Metric | Value |
|---|---|
| Final Equity | $42,395.78 |
| Total Return | 323.96% |
| Max Drawdown (MDD) | 61.40% |
| Sharpe Ratio | 0.79 |
| Total Periods | 210,280 (15m Candles) |
The engine calculates a real-time risk metric (0-100) based on four primary sub-factors:
- Distance from ATH: Quantifies the "discount" or "overextension" of the asset.
- Trend Momentum: Measures price velocity relative to the 50-day EMA.
- Volatility (ATR): Adjusts entry spacing and sizing based on market chaos.
- ETH/BTC Divergence: Used as a leading indicator for market-wide regime shifts.
- Regime Filter: A 50-day Exponential Moving Average (EMA) acts as the primary switch between "Bull" (Long-biased) and "Bear" (Short-biased/Defensive) modes.
- Inventory Management: Utilizes a Bi-directional FIFO Queue to manage layered DCA (Dollar Cost Averaging) entries and exits.
- Position Sizing: Tiered entry logic (e.g., 15% / 8%) to ensure capital is deployed at optimal risk levels.
- Volatility Normalization: Pre-calculates 24-hour and 30-day ATR baselines to prevent "buying the cliff" during cascading liquidations.
- Dynamic Step Logic: Spaces entries based on ATR-multipliers (e.g., 1.5x ATR) to ensure mathematical distribution of risk.
- Circuit Breakers: Implements equity-based stops to protect the principal during "black swan" events.
score-risk-system/
├── data/ # local h5/csv cache for btc/eth history
├── src/
│ ├── engine.py # core event-driven simulation logic
│ ├── indicators.py # synthetic risk score & ema calculations
│ └── execution.py # fifo queue & order management
├── results/ # backtest charts and csv reports
└── main.py # entry point for backtest simulation