Kelly Criterion Experiments
Optimal bet sizing visualized through Monte Carlo simulations of the Kelly criterion.
Implemented the Kelly criterion formula and ran Monte Carlo simulations to visualize how different bet-sizing strategies affect long-run wealth growth.
Key results:
- Plotted optimal fraction (f*) vs win probability and odds across parameter ranges
- Simulated 50,000-bet wealth trajectories for 0.1× to 4× Kelly multipliers
- Demonstrated that ~0.8× Kelly consistently dominates, while overbetting (>2× Kelly) leads to ruin despite a positive edge
- 10 independent experiments with fresh random seeds confirm robustness
Technologies: Python, NumPy, Matplotlib