BlackSwan TRL-4 Simulation Baseline (first run)¶
Date: 2026-04-28 Configuration: 6 agents (3 Kelly, 3 Random), 100 steps, burn=0.5, drift=0.002, volatility=0.01
| Metric | Value |
|---|---|
| Agents alive | 0 |
| Survival rate | 0.0 |
| Kelly avg capital | 1123.66 |
| Random avg capital | 926.07 |
| Total trades | 236 |
All agents died due to burn rate and commissions. Next step: parameter sweep to find stable survival zone.
BlackSwan TRL-4 Simulation Sweep Results¶
Date: 2026-04-28 Sweep parameters: burn_rate ∈ {0, 0.1, 0.2, 0.5, 1.0}, failure_prob ∈ {0, 0.01, 0.05, 0.1}, 3 seeds each. Agents: 6 (3 Kelly, 3 Random), 200 steps.
Optimal Stability Zone¶
- failure_prob = 0.0, burn_rate ≤ 1.0:
- Survival rate: 100%
- Kelly avg capital: 2025–2369
- Random avg capital: 604–804
- Kelly advantage: ~1400–1565
Selected for TRL-4 Docker demo: burn_rate=0.1, failure_prob=0.0.
Docker Swarm Demo (Iteration 3)¶
Date: 2026-04-28 Configuration: 8 nodes, burn_rate=0.1, failure_prob=0.0, market drift=0.002, volatility=0.01 Infrastructure: Docker Compose, Redis pub/sub, Python 3.11-slim
- All 8 nodes started successfully and maintained capital above initial value.
- Node failure recovery works: after manually stopping a node, it automatically restarted within 5 seconds (restart_policy: on-failure).
- Communication via Redis EventBus confirmed.