Trading Strategies

Trading Strategies Skill-Graph — CONCEPT:AU-AHE.assimilation.trading-ecosystem-spec

Reference catalog of strategy archetypes supported by the Emerald Exchange strategy lifecycle.

Strategy Types

Momentum

• Trend Following: EMA crossover, Bollinger breakout, ADX-filtered
• Relative Strength: Sector rotation, pair momentum, cross-sectional momentum
• Time-Series Momentum: TSMOM 1M/3M/12M look-back

Mean Reversion

• Statistical Arbitrage: Pairs trading via cointegration (Engle-Granger)
• Bollinger Mean Reversion: Z-score entry/exit on Bollinger bands
• Ornstein-Uhlenbeck: Mean-reversion speed estimation

Factor-Based

• Multi-Factor: Value + Momentum + Quality + Low-Vol composite
• Alpha Factors: IC/IR scoring via QLib factor pipeline
• Fama-French: 3-factor and 5-factor model decomposition

Machine Learning

• Regime Detection: HMM-based (Hidden Markov Model) state classification
• Signal Prediction: LightGBM/XGBoost on engineered features
• Deep Learning: Transformer-based price prediction (route to GPU via data-science-mcp)

Crypto-Native

• Funding Rate Arbitrage: Long spot / short perp when funding > threshold
• Cross-Exchange Arbitrage: Bid-ask spread exploitation across CCXT backends
• DeFi Yield: Liquidity provision optimization (requires on-chain integration)

Strategy Lifecycle — CONCEPT:AU-AHE.assimilation.trading-ecosystem-spec

draft → backtest → paper → live
  │        │         │       │
  │        │         │       └── Requires human approval (OS-5.1)
  │        │         └── 30-day minimum paper trading
  │        └── Sharpe > 1.5, MaxDD < 15%
  └── Hypothesis documented in KG

Integration Points

• Backtest Engine: Routes to data-science-mcp on GPU hardware
• Signal Generation: emerald_signals MCP tool
• Risk Validation: emerald_risk pre-trade checks
• Export: PineScript, MQL5, TDX via emerald_strategy(action='export')