54 Futures Matrix & Kelly Sizing Guide
Rigorous historical physics: 10-year rolling win rates, Profit Factor (PF), Trader’s Equation, and Half-Kelly risk management.
1. 10-Year Rolling Forensics & Data Protocol
MarketSightAI systematically tracks 54 major Chinese continuous commodity and index futures contracts across CFFEX, SHFE, DCE, CZCE, and INE. Unlike conventional curve-fitting backtests, our engine enforces rigorous causality:
All daily settlement records are physically aligned to 15:00:00 CST with zero arbitrary clock offsets.
Bar N+1 cannot peek into Bar N+2. Indicator bands and regime gates recompute dynamically strictly on closed bars.
Orders strictly progress through PENDING ➔ TRIGGERED ➔ CLOSED states according to price touch mechanics.
2. Core Quantitative Metrics Dissected
Profit Factor represents the gross profits divided by gross losses across all triggered trades. A PF > 1.0 indicates profitable positive expectancy. Top empirical setups in MarketSightAI typically feature PF between 1.6 and 2.4.
Al Brooks Price Action emphasizes that win rate alone is meaningless without Reward-to-Risk ratio. Even at 40% win rate, a setup with 2.5:1 R:R yields remarkable long-term equity growth.
Measures the payoff asymmetry of executed setups. MarketSightAI computes this strictly using actual exit execution rather than theoretical targets.
3. Trader’s Equation & Half-Kelly Sizing Engine
The cornerstone of mathematical trading is the Trader’s Equation:
When EV is strictly positive, the Kelly Criterion calculates the theoretically optimal capital allocation to maximize geometric growth rate without risking total ruin:
Full Kelly assumes known infinite samples with constant distribution. In real futures markets, volatility clustering and fat-tail drawdowns happen. Using Half-Kelly retains ~75% of maximum growth rate while reducing peak portfolio volatility by over 50%.