STEVEN is a high-performance quantitative framework designed for automated futures trading and macro-economic hedging. It leverages recursive Bayesian estimation to predict volatility spikes in leveraged derivatives markets.
STEVEN operates on three primary layers to ensure capital preservation during tail-risk events:
- Macro Overlay: Aggregates real-time fiscal data and yield curve inversions.
- Liquidity Scraper: Monitors order book depth and liquidation clusters in futures markets.
- Execution Layer: Deploys capital using a dynamic Kelly Criterion for optimal position sizing.
| Component | Architecture | Purpose |
|---|---|---|
| Engine | Rust (Low-Latency) | Sub-millisecond order execution. |
| Analysis | Python / Pandas | Statistical arbitrage and mean reversion. |
| Risk | C++ | Monte Carlo simulations for VaR (Value at Risk). |
| Storage | TimescaleDB | High-velocity time-series economic data. |
To initialize the STEVEN environment and begin the data ingestion cycle:
# Clone the private repository
git clone [https://github.com/GarrettBullish/STEVEN.git](https://github.com/GarrettBullish/STEVEN.git)
# Initialize the Macro-Environment variables
export LEVERAGE_LIMIT=10x
export RISK_TOLERANCE=AGGRESSIVE
# Boot the engine
./steven --mode automated --asset BTC-PERP --strategy mean-reversion