Probability-weighted branching outcomes of major geopolitical events. Click any scenario to see the full decision tree — winners, losers, and signals at every branch.
These probabilities are subject to quarterly calibration retrospectives against realized outcomes — first audit August 2026. See methodology.
Ongoing · Technology / Energy
Ongoing · Maritime / Trade
Ongoing · Military / Geopolitical
Ongoing · Credit / Real Estate
Recurring · Climate / Commodities
Watching · Monetary Policy
Watching · Energy / Geopolitical
Watching · Energy / Geopolitical
Watching · Military / Geopolitical
Watching · Sovereign / Credit
Watching · Cyber / Infrastructure
Watching · Sovereign / Trade
Low-probability, high-impact outcomes outside the modeled partition above. Surfaced for analytical hygiene — never ranked, never assigned a probability. Each card describes the trigger mechanism and the names that move if it actually happens.
Mechanism: Routing collapse and brief loss of intercontinental financial-messaging and cloud-control-plane connectivity, before traffic re-routes through degraded paths over days to weeks.
A near-simultaneous loss of multiple major transoceanic cables — Atlantic and Pacific — within a 48-hour window. Cause is plausible but not yet observed at this scale: state-actor sabotage, a fishing-vessel anchor cascade, or a Mediterranean / Red Sea chokepoint event that propagates. The economy assumes 95%+ of intercontinental data moves over these cables; the satellite uplink fleet cannot absorb that volume.
Methodology. Each scenario branches into conditional paths (Y) and sub-outcomes (Z). Percentages shown on paths are joint probabilities — the overall likelihood of that full chain occurring. These are analytical frames, not forecasts.
AI-generated hypothesis. Not investment advice. Always verify independently with a qualified financial advisor.
Contingency note — Watch the SubmarineCableMap incident feed and the Lloyd's List anchor-drop chronology. The hedge isn't a single ticker — it's a pair trade between satellite-comms exposure and hyperscaler cloud beta.
What would falsify this — A 12-month period with zero inter-continental cable incidents > 48h outage AND functional satellite backup absorbing rerouted commercial traffic would indicate the scenario premises no longer hold.
Anchored in: TeleGeography Submarine Cable Map + incident database · ITU submarine cable capacity utilisation reports · ENISA NIS2 critical infrastructure threat assessment · Internet Society routing resilience reports · Cloudflare Radar internet traffic anomaly data
Mechanism: Multi-quarter West-Coast supply-chain disruption (Boeing 737 line, Port of Seattle/Tacoma, Microsoft / Amazon HQ regions); flood of insurance/reinsurance claims; emergency rebuilding spend; durable shift of cloud workloads to non-PNW regions.
A magnitude 8.7+ event on the Cascadia fault, the long-overdue Pacific-Northwest analogue to the 1700 Vancouver Island earthquake. USGS one-in-300-year base rate places this materially above background; tsunami modelling suggests a 30-minute warning window for Portland/Seattle/Vancouver. Tech-corridor real estate, Boeing manufacturing, and West-Coast ports take simultaneous hits.
Contingency note — A long-tail position via reinsurance equity beta or via municipal-bond CDS on Oregon / Washington local issuers. The asymmetric play is at the reinsurance pricing turn, not the disaster spot.
What would falsify this — Publication of a revised USGS seismic hazard model placing the Cascadia CSZ recurrence interval above 1,000 years AND a structural decoupling of PNW tech/logistics from Boeing manufacturing would invalidate the core impact thesis.
Anchored in: USGS National Seismic Hazard Model (2023 update) · PNSN Pacific Northwest Seismic Network catalog · NOAA PTWC Pacific Tsunami Warning Center historical record · FEMA National Earthquake Loss Estimation (Cascadia scenario) · Swiss Re Cat bond / reinsurance exposure reports
Mechanism: EM hard-currency bond ETFs take an acute mark-to-zero on the affected issuer; spillover widens the EM credit-default-swap basket; a flight to USD safe assets; contagion math is the open question and the reason this is a black swan — neither a soft nor a hard restructuring fits the historical analogues cleanly.
A single large emerging-market economy — Turkey, Argentina, Egypt, or Pakistan — announces full unilateral repudiation of external sovereign and quasi-sovereign debt, framed as a sovereignty-restoration act, with diplomatic backing from a regional bloc (BRICS / Gulf states / others). Distinct from the existing partition because OpenWatch's current sovereign-default reference classes assume a managed restructuring, not a Soviet-style jubilee.
Contingency note — Watch the CDS-basket spread, the IMF Article IV calendar, and the BRICS-summit communiqué cycle. The hard-to-reverse part — once one country tests this and the diplomatic backing holds, the reference class for the next sovereign default has fundamentally changed.
What would falsify this — Successful completion of an IMF-coordinated restructuring by the largest stressed EM debtors (Turkey, Argentina, Egypt, Pakistan) within a 24-month window without unilateral repudiation would show the jubilee pathway is not the modal resolution.
Anchored in: IMF World Economic Outlook debt vulnerability screening · World Bank International Debt Statistics 2024 · JPMorgan EMBI spread and CDS data · Moody's / S&P EM sovereign rating history · Paris Club + BRICS debt-restructuring communiqués
Mechanism: Transformer destruction concentrates in 40–60° latitude grids (Northeast US, Canada, Northern Europe, Russia) with 12–24 month replacement lead times; satellite-comms backup demand surges as LEO mega-constellations degrade; precision-timing-dependent industries (high-frequency trading, GPS-dependent logistics, semiconductor lithography) face protracted disruption that no portfolio's standard risk model prices in.
A coronal mass ejection of 1859-Carrington-event magnitude impacts Earth's magnetosphere with little warning. NOAA Space Weather Prediction Center estimates roughly 12% per-decade probability — meaningfully above what OpenWatch's modeled scenario partition treats as background. Geomagnetically induced currents damage high-voltage transformers across mid- to high-latitudes; GPS becomes unreliable for days; LEO satellite fleets sustain degradation; aviation reroutes; financial-settlement and precision-timing systems run in degraded states for weeks.
Contingency note — Watch NOAA SWPC Kp-index alerts, especially during solar-maximum windows. The asymmetric hedge isn't 'buy satellites' — it's a pair trade between grid-hardening contractors and high-latitude utility credit. Reinsurance pricing turns lead the equity beta.
What would falsify this — Completion of NERC-mandated grid-hardening requirements for high-voltage transformers across North America and equivalent EU measures, combined with a sustained LEO constellation proving ≥48-hour recovery, would materially reduce the scenario impact thesis.
Anchored in: NOAA SWPC solar-cycle progression + historical X-flare catalog · NERC GMD Task Force transformer vulnerability assessment · FEMA / DoE Grid Resilience and Intelligence Project reports · NASA Goddard Space Weather Research Center CME arrival models · Lloyd's of London geomagnetic storm systemic risk report
Mechanism: API revenue at the frontier labs goes to near-zero for the pause window; hyperscaler GPU demand softens overnight; downstream AI-integrated SaaS loses its inference layer; the precision-timing for the resume window becomes the dominant volatility driver; the regulatory aftermath durably changes cost-of-compliance for every subsequent frontier deployment.
A frontier-model deployment produces sustained, visibly harmful output at consumer scale, and within a 48-hour window regulators in at least two of (US, EU, UK, China) impose a coordinated 30–90 day pause on commercial frontier-model serving. Distinct from the modeled AI-infrastructure scenario because the binding constraint is policy-driven, not compute-driven; modeled probability is implicitly zero in current OpenWatch reference classes.
Contingency note — Watch FedRAMP / EU AI Act enforcement bulletins, the NIST AI Safety Institute incident registry, and frontier-lab safety-team turnover patterns. The asymmetric play is compliance-tooling adjacencies, not shorting frontier-model exposure — equity beta there is too noisy. Signal-to-watch: a single coordinated regulatory action across at least two of US/EU/UK/China within a 48-hour window.
What would falsify this — A major visible AI harm event at consumer scale that fails to trigger coordinated enforcement action within 90 days in at least two of (US, EU, UK, China) would show that cross-jurisdictional regulatory coordination is slower than the scenario assumes.
Anchored in: UK DSIT / US AISI frontier-model safety evaluation reports · EU AI Act enforcement framework (high-risk AI provisions) · NIST AI Risk Management Framework 1.0 · Anthropic / OpenAI / DeepMind published safety incident disclosures · Stanford HAI AI Index 2024 (regulatory action tracker)
Tail contingency — deliberately no probability assigned. Outside the modeled scenario partition. Surfaced for analytical hygiene, not for ranking. AI-generated hypothesis. Not investment advice. Always verify independently with a qualified financial advisor.