Monte Carlo Simulation
Jump-diffusion Monte Carlo projection with scenario weighting, GARCH volatility, and Merton jump compensator
Simulation Paths
Forecast Horizon
Price Dynamics
Jump-diffusion GBM with Merton 1976 compensator. Student-t innovations from GARCH-estimated degrees of freedom for fat tails.
Volatility
GJR-GARCH(1,1) with skewed Student-t innovations. Ornstein-Uhlenbeck dynamics with leverage effect (rho = -0.7).
Scenarios
7 scenarios weighted by ML crash probability, VIX level, yield curve slope, and recession indicators. Weights re-normalize dynamically.
Mean Reversion
Calibrated from drawdown recovery data. Asymmetric: stronger pull from below fair value (0.08) than from above (0.04).
Block Bootstrap
21-day overlapping blocks preserve volatility clustering and serial correlation in historical residuals.
Validation
Realism checks: 2-15% annual return, 10-30% vol, 30-90% crash frequency, kurtosis > 3 (fat tails).