The Simulation View

An agent-based scenario simulation runs alongside the Probability Desk’s governed ensemble as a dedicated leg. It builds an actor-level model of each tail scenario — sovereign-wealth allocators, pension funds, insurers, analysts, governments — and lets them interact to surface second-order dynamics a single analyst can miss. It is governance-gated: until it earns a calibrated track record it carries zero weight on every published probability. What it does today is tell the desk where to look.

The simulation leg is capped at 5pp of ensemble weight and stays at 0pp until it clears calibration. This page updates as new scenario simulations report. Status as of 2026-06-02.

The ensemble — where the simulation sits 7 legs

Every published probability is a weighted blend of these legs. The simulation leg (highlighted) is wired in but held at 0pp; the other legs carry the number today.

Simulation pipeline 1 reported · 5 in build

What the simulation has flagged · Chokepoint concentration as a standing factor (Hormuz + Taiwan + Malacca + Panama)

• Insurance/freight cost-push as a distinct inflation channelDesk follow-up: size historical insurance/freight cost-push vs oil in past chokepoint episodes; add to inflation stress test
• Friend-shoring eroding the globalization dividend in CMAsDesk follow-up: quantify globalization-dividend assumption in our return model; sensitivity to friend-shoring
• Reflexive actors who trade the volatility and amplify itDesk follow-up: measure order-book reflexivity in prior strait incidents; second-order feedback in scenario model

These are structural blind spots the simulation surfaced and the desk is now investigating. They are research prompts, not published probabilities.

Explore

Editorial scenario analysis only. Not investment, actuarial, legal, geopolitical, or financial advice. The simulation leg carries no weight on any published probability while calibration is pending.