← Back to Noah Predict
FOR REINSURERS

Understanding aggregation risk in real time.

Reinsurance decisions are rarely about single events — they are about how risks interact across geographies, perils and portfolios.

Reinsurance desk NOAH PREDICT · Predictive political risk intelligence
The system, not the event

Aggregation emerges when signals align.

Reinsurance is about how risks interact. Not single events, but the way signals stack: geography, peril, actor, time. Noah is built on coupling — the measurement of two independent signal streams moving together — so it reads correlation natively rather than as an after-the-fact cross-tabulation.

When port closure and road closure couple at Jaccard 1.0, cargo movement has stopped. When K&R velocity and political-violence velocity couple in the same admin-1 region, the book behind it has already shifted.

What reinsurers get

Four lenses on the connected book.

Noah produces the same structured output as for direct insurers — peril sheet, sub-national scores, peer comparison — but with the coupling layer surfaced as a first-class block.

01 · Cross-peril couplings

Which perils are moving together

Every run lists the top coupled peril pairs with Jaccard similarity and a one-line meaning note. Data exfiltration ↔ State-sponsored intrusion coupled at 0.33; Insurgency ↔ Frontline clashes coupled at 0.33. Each coupling is evidence for why a peril score cannot be treated in isolation.

02 · Sub-national clustering

Geographic concentration, hard-data backed

Sub-national districts carry kidnap_freq_90d, ied_within_5km_un, ingo_convoy_incidents_90d counts — the inputs an aggregation team would want to see. When those counts spike in two adjacent districts, the book behind them stacks.

03 · Regime classification

What shape is this country in

Each country gets a regime read with probability: Fragile Calm, Pressure Release, Political Crisis, Stabilisation, Regime Instability. Aggregation risk is highest where multiple books sit in correlated regimes — Noah exposes that directly.

04 · Invalidation triggers

What would force the aggregate view to shift

Each country ships with a what_would_change_my_mind list — falsifiable, numeric, dated. Aggregation monitoring can subscribe to those triggers and re-rate when they fire.

Benchmarked against

Six external anchors, on every country in the book.

The same anchor layer the direct insurance product ships with. A reinsurer reviewing a Noah bundle can cross-check any country’s read against six independent reference points without leaving the report.

Advisory
FCDO

UK travel advice level + regions advised against.

Advisory
US State Dept

Level 1–4 + indicator codes (T, C, K, U, H, N, E, D, O).

Peace Index
GPI

IEP ranking 1–163 with overall score and three sub-scores.

Governance
World Bank WGI

Six-axis governance scores.

Sanctions
OFAC & UK HMT

Sovereign designations + named entity list.

Event data
ACLED

30d / 90d / 365d event counts, fatalities, top actors.

Across a portfolio Multi-country queries run the same engine per country and reconcile the outputs into a portfolio view. Coupling scores persist across countries — when insurgency velocity in Mali rises in the same window as food-insecurity velocity in Somalia, and both countries already sit in the book, that is a portfolio-level signal, not two unrelated country reads.
Why this matters

Aggregation is often recognised too late.

By the time patterns are clear in broker consensus or quarterly briefings, exposure has already stacked. Noah provides earlier visibility by measuring the system, not the events — and by surfacing couplings as part of the output rather than as a separate analysis.

It is not just about where risk exists.

It is about how it connects.

Elsewhere on Noah Predict

More from the system.

Free while in preview
Official systems show the current state of risk.
Noah shows where it is moving.

Google sign-in. Three free analyses a day. No card required.