Monte Carlo Engine
1,000-iteration Monte Carlo engine with a realistic Ornstein-Uhlenbeck operator model. Adjust parameters live — the probability distribution of outcomes updates immediately.
Efficiency fluctuates via an Ornstein-Uhlenbeck process (θ=0.35, σ=7.5%). Shocks persist ~3 weeks before normalising — capturing realistic human performance autocorrelation.
Position size = base_r × perf_mult(eff) × regime_penalty(streak) × 0.95. Efficiency tier directly gates the risk multiplier each week.
Slippage on positive weeks, edge decay per quarter, ATO taxation annually, 4-week capital injection gate, and configurable drawdown halt threshold.
Simulation Methodology
OU Operator: eff[t+1] = eff[t] + 0.35·(μ − eff[t]) + 0.075·Z, clamped to [40%, 100%]. θ=0.35 means shocks persist ~3 weeks before normalising.
Risk gating: applied_risk = base_r × perf_mult(eff) × regime_penalty(streak) × 0.95. Mirrors recommend_r_generator.py exactly.
Weekly compounding: Position size = equity_start × applied_risk (fixed for the week). equity_end = equity_start × (1 + applied_risk × captured_weekly_R).
Commission: Deducted when eff ≥ 88% and after commissionStartWeek. Formula: (base_r × 0.20 + max(yield, 0) × 0.05) × [1.5 if eff ≥ 95%].