Skip to content

Foundry

Le Global Risks Index

INT

Intelligence Hub

Pulse · Outlook · Horizon · analyst-reviewed briefings

Observatory Set up a watchlist

Synthèse

No brief yet

The daily brief is issued from analyst-reviewed signals once Pulse and Outlook are live. Each item names its sources and the authority that decides.

Signals

No signals yet

Pulse readings become signals for review when they cross a place’s own historical range. Nothing here is a warning; every handoff names the authority that decides.

on: pulse anomaly

Un signal à examiner

Waiting for approvalTriggered by Pulse nowcast5 jobs1 approval gate1 matrix job

Graphe
01Detect
02Vérifier
03Assess
04Transmettre

100%
Chronologie
Job0 d2 d4 d6 d8 d10 dTarget
Anomaly detectedPulse nowcast7 d
Sources verifiedData stewards1 d
Analyst reviewAnalysts2 d
Information for competent authoritiesSecretariat1 d
Logged and watchedPulse nowcastongoing

End to end on the longest path: about 12.3 d of proposed target service levels, before the jobs that run with no fixed end. Gates wait for the people named; bars with a fading end have no fixed length.

Étapes
  1. Anomaly detected

    Detect · Pulse nowcast · weekly
    Standardised anomaly vs own history
    Pulse nowcast
  2. Sources verified matrix · per stream

    Verify · Data stewards · 1 day
    Each stream checked · Licence admits use
    Licence registerProvenance (W3C PROV)
  3. Analyst review approval

    Assess · Analysts · 2 days
    Context, couplings, confidence
    Registre PostgreSQL (sécurité au niveau des lignes)
  4. Information for competent authorities

    Hand off · Secretariat · 1 day
    Shared for review, not a warning
    NXS-GRX (canonical data, provenance and risk) gateway
  5. Logged and watched

    Hand off · Pulse nowcast · ongoing
    Outcome tracked for verification
    PostgreSQL ledger (row-level security)Pulse nowcast
Artefacts
Artefact Tâche Qui Étape
Signal candidate Anomaly detected Pulse nowcast Detect
Source check Sources verified Gestionnaires de données Vérifier
Assessment Analyst review Analysts Assess
Handoff record Information for competent authorities Secrétariat Transmettre
Verification record Logged and watched Pulse nowcast Transmettre
TerminéEn coursEn attente d'approbationEn file d'attenteÉtape d'approbationTâche matricielleVous êtes ici
How a weekly anomaly becomes a signal for review by competent authorities, never a public warning. The canvas shows where a request stands at the marked job; it is illustrative, not the status of a real request. Durations are proposed target service levels. Select a job for its steps, the people who act, the technology it runs on and what it produces.

Horizon

No horizon cycle yet

Each cycle scans admitted sources, scores emerging risks for likelihood and impact with ranges, and links them to the domains they would couple into.

on: horizon cycle opens

Veille prospective et note

In progressTriggered by Intelligence team6 jobs1 approval gate1 matrix job

Graphe
01Scan
02Scan
03Assess
04Synthèse
05Synthèse

100%
Chronologie
Job0 d5 d10 d15 dTarget
Scan sourcesNXS-GRX Intelligence7 d
Cluster and deduplicateNXS-GRX Intelligence1 d
Likelihood and impact with uncertaintyAnalysts3 d
Systemic linksAnalysts2 d
Peer reviewReviewers5 d
Briefing issuedIntelligence team1 d

End to end on the longest path: about 17 d of proposed target service levels. Gates wait for the people named; bars with a fading end have no fixed length.

Étapes
  1. Scan sources matrix · sources

    Scan · NXS-GRX Intelligence · weekly
    Admitted sources and literature · Expert elicitation
    Horizon scanningLicence register
  2. Cluster and deduplicate

    Scan · NXS-GRX Intelligence · 1 day
    Themes and weak signals
    Horizon scanning
  3. Likelihood and impact with uncertainty

    Assess · Analysts · 3 days
    Structured elicitation · Ranges, not points
    Horizon scanningMonte Carlo engine
  4. Systemic links

    Assess · Analysts · 2 days
    Coupling to existing domains
    Réseau de dépendances
  5. Évaluation par les pairs approbation

    Brief · Reviewers · 5 days
    Challenge session · Disposition
    Registre PostgreSQL (sécurité au niveau des lignes)
  6. Briefing issued

    Brief · Intelligence team · 1 day
    Briefing with sources and limits
    Rapports Nexus
Artefacts
Artefact Tâche Qui Étape
Candidate list Scan sources NXS-GRX Intelligence Scan
Theme map Cluster and deduplicate NXS-GRX Intelligence Scan
Horizon matrix Likelihood and impact with uncertainty Analysts Assess
Link map Systemic links Analysts Assess
Review record Évaluation par les pairs Reviewers Synthèse
Horizon briefing Briefing issued Intelligence team Synthèse
TerminéEn coursEn attente d'approbationEn file d'attenteÉtape d'approbationTâche matricielleVous êtes ici
Emerging risks found, scored for likelihood and impact with uncertainty, and turned into a briefing after review. The canvas shows where a request stands at the marked job; it is illustrative, not the status of a real request. Durations are proposed target service levels. Select a job for its steps, the people who act, the technology it runs on and what it produces.

Outlooks

No outlooks yet

Outlooks give calibrated probabilities at 1, 3, 12 and 36 months for defined events, each with its verification scorecard.

Vérification

No scorecards yet

Every Outlook model version publishes its reliability diagram, Brier score, skill and ROC area on held-out cases before it is used.

on: model version submitted

Tester et valider un modèle

In progressTriggered by Model owner7 jobs1 approval gate11 checks1 deployment

Graphe
01Soumettre
02Tester
03Document
04Approve
05Approve

100%
Chronologie
Job0 d2 d4 d6 d8 d10 dTarget
Model version submittedYouinstant
BacktestVerification engine6 h
Out-of-sample scoresVerification engine2 h
BenchmarkVerification engine2 h
Model cardYou1 d
Approval gateMethod stewards10 d
Promote to productionNXS-GRX Platforminstant

End to end on the longest path: about 11.3 d of proposed target service levels. Gates wait for the people named; bars with a fading end have no fixed length.

Étapes
  1. Model version submitted

    Submit · You · instant
    Code, data and seed pinned · Intended use declared
    Model LabPostgreSQL ledger (row-level security)
  2. Backtest

    Test · Verification engine · 6 hours
    Rolling-origin backtest · Hindcasts over the archive
    Verification engineMonte Carlo engine
  3. Out-of-sample scores

    Test · Verification engine · 2 hours
    Brier and CRPS · Reliability and ROC area
    Verification engine
  4. Benchmark

    Test · Verification engine · 2 hours
    Against established indices and naive baselines
    Verification engineindicator standard
  5. Model card

    Document · You · 1 day
    Intended use and limits · Data, method and scores
    Replication packs
  6. Étape d'approbation approbation

    Approve · Method stewards · 10 days
    Disposition recorded
    Registre PostgreSQL (sécurité au niveau des lignes)
  7. Promote to production

    Approve · NXS-GRX Platform · instant
    Version pinned in production · Rollback kept
    Model LabPublic cut API
Artefacts
Artefact Tâche Qui Étape
Version record Model version submitted Vous Soumettre
Backtest report Backtest Verification engine Tester
Scorecard Out-of-sample scores Verification engine Tester
Benchmark table Benchmark Verification engine Tester
Model card Model card Vous Document
Approval record Étape d'approbation Method stewards Approve
Release tag Promote to production NXS-GRX Platform Approve
TerminéEn coursEn attente d'approbationQueuedApproval gateTest jobDeploymentYou are here
Every model version is backtested, scored out of sample, compared with established indices and documented in a model card before use. The canvas shows where a request stands at the marked job; it is illustrative, not the status of a real request. Durations are proposed target service levels. Select a job for its steps, the people who act, the technology it runs on and what it produces.

Workbench

on: watchlist saved

Analyst workbench: watchlist to handoff

Waiting for approvalTriggered by Monitoring team6 jobs1 approval gate1 matrix job

Graphe
01Set up
02Sense
03Examen
04Livrer

100%
Chronologie
Job0 d2 d4 d6 dTarget
Watchlist and thresholdsYouinstant
Pulse nowcastPulse nowcast7 d
Outlook updateOutlook models7 d
Analyst reviewDuty analyst6 h
Route the alertAlert routerinstant
Handoff to the authorityDuty analyst2 h

End to end on the longest path: about 7.3 d of proposed target service levels. Gates wait for the people named; bars with a fading end have no fixed length.

Étapes
  1. Watchlist and thresholds

    Set up · You · self-serve
    Entities and domains · Thresholds and channels
    Alert router
  2. Pulse nowcast matrix · entities × domains

    Sense · Pulse nowcast · weekly
    Anomaly against each place’s own history
    Pulse nowcastTested model kernels
  3. Outlook update

    Sense · Outlook models · weekly
    Calibrated probabilities refreshed
    Verification engineMonte Carlo engine
  4. Analyst review approval

    Review · Duty analyst · 6 hours
    Signal checked against sources · Context added
    Registre PostgreSQL (sécurité au niveau des lignes)
  5. Route the alert

    Deliver · Alert router · < 5 ms
    Email, webhook or dashboard · Never framed as a warning
    Alert routerWebhooks and streaming
  6. Handoff to the authority

    Deliver · Duty analyst · 2 hours
    Names the authority that decides
    Registre PostgreSQL (sécurité au niveau des lignes)
Artefacts
Artefact Tâche Qui Étape
Watchlist Watchlist and thresholds Vous Set up
Pulse readings Pulse nowcast Pulse nowcast Sense
Outlook readings Outlook update Outlook models Sense
Reviewed signal Analyst review Duty analyst Examen
Alert Route the alert Alert router Livrer
Hand-over note Handoff to the authority Duty analyst Livrer
TerminéEn coursEn attente d'approbationEn file d'attenteÉtape d'approbationTâche matricielleVous êtes ici
Watch countries, areas, sites or domains: Pulse nowcasts and outlook changes become signals, reviewed by an analyst before anything reaches you. The canvas shows where a request stands at the marked job; it is illustrative, not the status of a real request. Durations are proposed target service levels. Select a job for its steps, the people who act, the technology it runs on and what it produces.
Signals are information for review, not warnings · every handoff names the authority that decidesNot live yet

01Rules

How intelligence is issued

Signals, not warnings

A signal is information for review. Official warnings come from the authorities that issue them.

Reviewed before it reaches you

Every signal passes an analyst gate; the reviewer and the sources are recorded.

Handoff names the authority

Where action may be needed, the handoff names the authority that decides.

Verified in public

Every Outlook model version publishes reliability, Brier score, skill and ROC area.

02Parcours

Intelligence journeys

NXS-GRX · Govern

Un signal à examiner

How a weekly anomaly becomes a signal for review by competent authorities, never a public warning. 5 jobs · 1 gate.

Open the run NXS-GRX Intelligence · Monitor

Suivre avec des listes de veille et des alertes

Watch countries, areas, sites or domains: Pulse nowcasts and outlook changes become signals, reviewed by an analyst before anything reaches you. 6 jobs · 1 gate.

Open the run NXS-GRX Intelligence · Monitor

Veille prospective et note

Emerging risks found, scored for likelihood and impact with uncertainty, and turned into a briefing after review. 6 jobs · 1 gate.

Open the run NXS-GRX · Govern

Demander un ensemble de scénarios

Pathways to 2030, 2050 and 2100 and structured stress tests for planning and supervisory exercises. 5 jobs · 1 gate.

Open the run

03Données humanitaires

What the Humanitarian Data Exchange feed holds

194 countries synced on 2026-10-23 from 226 datasets whose licence passes the gate; 147 resources were left out for their licence. Each country’s page shows the values, with date, provider and licence. Source observations for review, not warnings and not index values.

Thème Pays With ADM1 figures With ADM2 figures Newest reference date
Population 109 98 59 2025-12-31
Personnes dans le besoin 22 6 16 2026-12-31
Taux de pauvreté 107 99 0 2024-12-31
Réfugiés accueillis 171 0 0 2025-12-31
Réfugiés à l'étranger 186 0 0 2025-12-31
Précipitations par rapport à la moyenne de long terme 112 109 44 2026-10-10
Humanitarian funding 107 0 0 2026-08-31
Prix alimentaires sur les marchés 85 0 0 2026-10-14

Bring intelligence to your team

Watchlists, briefings and outlooks are agreed with the Secretariat.

Contact

Statut

La méthodologie, la structure et les définitions de The Global Risks Index constituent un projet de travail préparatoire (2026), diffusé pour examen et commentaires et non pour mise en œuvre. Aucune valeur n'apparaît avant la publication d'une édition à partir d'une coupe de données probantes encadrée ; les valeurs comportent alors des intervalles, des classes, des sources et leur voie de correction. Les simulateurs appliquent les méthodes publiées à des nombres que vous choisissez et n'affichent jamais de valeur NXS-GRX.

Ce que NXS-GRX n'est pas

NXS-GRX est une information de recherche et d'analyse des risques. Ce n'est pas une notation de crédit ou ESG et il n'utilise aucune échelle de notation ; ce n'est pas un indice de référence pour des instruments financiers, ni une statistique officielle, ni une alerte d'urgence ou une classification officielle, ni un conseil. Il ne doit pas servir à évaluer des personnes physiques. Les classes de risque sont des bandes sur une échelle publiée, jamais des classements de pays ou de personnes.

Institutions et sources

The Nexus Consortium is stewarded by three independent not-for-profit institutions: The Global Centre for Risk and Innovation (GCRI), through GCRI US in Washington, DC and GCRI Canada in Ontario, with GCRI Canada as Lead Institution, Consortium Secretariat and lead contracting organisation; The Global Risks Forum (GRF); and The Global Risks Alliance (GRA). Technology, infrastructure and service providers work with the Consortium at arm’s length under written agreements, like every other external party in the public-private-planet model, and hold no role in its stewardship.

Chaque valeur porte l'attribution de ses sources, qui conservent leurs propres licences. Les noms externes ne sont cités qu'à des fins d'interopérabilité. Les textes rédigés par Nexus sont destinés à être publiés sous licence CC BY 4.0.

Avis en français

La méthodologie et les définitions du NXS-GRX sont un projet de travail préparatoire (2026). Le NXS-GRX est une information de recherche et d’analyse des risques : ni notation, ni indice de référence financier, ni statistique officielle, ni alerte, ni conseil. Les classes de risque sont des bandes sur une échelle publiée, non des classements.