Methodology

How Risk Horizon Works

Risk Horizon synthesises publicly available regulatory output into structured intelligence using AI. This page explains the methodology, scope, data sources, and important limitations of that process.

All intelligence is AI-generated from public sources. Not regulatory or investment advice. Verify before relying on any output.

Intelligence Generation

How signals are produced

Risk Horizon runs a daily automated scan that processes publicly available regulatory publications — speeches, enforcement actions, consultation papers, supervisory guidance, and thematic reviews — across 18 jurisdictions. A large language model (Claude by Anthropic) reads and interprets this material, extracting structured intelligence signals with identified impact direction, materiality assessment, time horizon, and required action.

Signals are then grouped into risk themes, linked to scenario packs, and mapped to regulatory alignment frameworks. The entire chain from source publication to structured pack is AI-generated. No human analyst has reviewed individual outputs unless otherwise stated.

Daily scan cadence

Sources are scanned every morning at 06:00 UTC and processed within the same cycle.

AI classification

Claude categorises each signal by risk type, jurisdiction, materiality, and time horizon.

Scenario synthesis

Related signals are combined into scenario packs with trigger conditions and control stress points.

Regulatory mapping

Signals and scenarios are mapped to regulatory frameworks with clause-level relevance narratives.

Field Definitions

What each field means

Confidence

AI-assessed certainty that the signal represents a genuine emerging risk. High: clear regulatory intent or enforcement action. Medium: indicative language or early-stage consultation. Low: background noise or speculative signals.

Materiality (1–10)

Estimated potential impact on a typical regulated financial institution. 1–3: monitoring only. 4–6: warrants management attention. 7–9: requires active remediation or escalation. 10: systemic or immediate board-level concern.

Impact Direction

Whether the risk represented by this signal is increasing (regulatory pressure or incident trajectory rising), stable (steady-state supervisory focus), or decreasing (resolved or de-prioritised by regulators).

Time Horizon

How quickly this signal is expected to crystallise into regulatory action or risk event. Immediate: 0–3 months. Near Term: 3–12 months. Medium Term: 1–3 years. Long Term: 3+ years.

Verification Status

Whether the signal has been cross-checked against a primary source. Verified means a source URL was retrieved and the underlying document was parseable. Unverified means the signal was inferred from context or secondary reporting.

Theme lifecycle classification

Emerging

A risk cluster first appearing in regulatory commentary or supervisory focus. Not yet the subject of formal guidance or enforcement. Requires monitoring.

Crystallising

The risk is actively being regulated. Formal consultations, thematic reviews, or enforcement actions are in train. Immediate attention required by risk functions.

Established

The risk has been formally regulated and embedded in supervisory frameworks. Compliance is table stakes; focus shifts to evidence and ongoing controls.

Declining

Regulatory focus on this risk is reducing — either because it has been resolved, superseded, or de-prioritised in favour of newer themes.

Jurisdictions

Covered regulatory bodies

Risk Horizon monitors publications from the following 18 regulatory bodies and standard-setting organisations. Coverage is limited to publicly available documents. Bilateral supervisory correspondence (e.g. individual firm SREP letters) is not accessible.

APRAAustralian Prudential Regulation AuthorityAustralia
PRAPrudential Regulation AuthorityUnited Kingdom
FCAFinancial Conduct AuthorityUnited Kingdom
ECBEuropean Central BankEuropean Union
EBAEuropean Banking AuthorityEuropean Union
MASMonetary Authority of SingaporeSingapore
OSFIOffice of the Superintendent of Financial InstitutionsCanada
OCCOffice of the Comptroller of the CurrencyUnited States
FRBFederal Reserve BoardUnited States
FDICFederal Deposit Insurance CorporationUnited States
SECSecurities and Exchange CommissionUnited States
CFTCCommodity Futures Trading CommissionUnited States
BISBank for International SettlementsGlobal
BCBSBasel Committee on Banking SupervisionGlobal
FATFFinancial Action Task ForceGlobal
IAISInternational Association of Insurance SupervisorsGlobal
IOSCOInternational Organization of Securities CommissionsGlobal
FSBFinancial Stability BoardGlobal

Risk Taxonomy

12 risk categories

Every signal and theme is classified into one of twelve risk categories drawn from the Basel framework and standard risk taxonomy used across major jurisdictions.

Operational Risk

Risks arising from failed processes, systems, people, or external events, including business continuity and third-party dependencies.

Technology & Digital

Risks from technology failures, digital transformation initiatives, model risk, and IT infrastructure concentration.

Third-Party Risk

Risks introduced by outsourced service providers, supply chain dependencies, and vendor concentration.

Geopolitical Risk

Risks from political instability, sanctions, trade policy shifts, and cross-border regulatory divergence.

Conduct Risk

Risks from behaviour that results in unfair outcomes for customers, market manipulation, or regulatory censure.

ESG & Climate

Physical and transition risks arising from climate change, nature-related financial risks, and sustainability obligations.

Credit Risk

Risks from counterparty default, concentration, and deterioration in asset quality across lending portfolios.

Market Risk

Risks from adverse movements in interest rates, foreign exchange, equity prices, and commodity prices.

Liquidity Risk

Risks from an inability to fund obligations as they fall due, including intraday and stress liquidity management.

Regulatory Risk

Risks from evolving regulatory requirements, enforcement actions, and supervisory expectations.

Strategic Risk

Risks from poor strategic decisions, misaligned business models, or failure to respond to structural industry change.

Cyber Risk

Risks from cyber attacks, data breaches, ransomware, and vulnerabilities in digital infrastructure.

Source Types

What we read

Regulatory SpeechSpeeches and public statements by regulators setting supervisory intent.
Enforcement ActionPublished penalties, orders, and sanctions against regulated entities.
Consultation PaperProposed rule changes and regulatory frameworks open for industry comment.
Thematic ReviewCross-firm deep dives into specific risk areas by regulators.
Supervisory GuidanceDear CEO letters, supervisory statements, and guidance publications.
Major IncidentSignificant operational, cyber, or market incidents with systemic implications.
Technology ShiftEmerging technology developments with material risk implications for financial services.
Geopolitical EventCross-border political or macroeconomic events with regulatory consequences.
Academic / IndustryResearch and reports from central banks, industry bodies, and academic institutions.
Legislative ChangeNew or amended legislation directly affecting regulated activities.

Limitations

What Risk Horizon is not

01

Risk Horizon does not provide legal, regulatory, or investment advice. All intelligence outputs should be reviewed by qualified professionals before being relied upon.

02

AI language models can misinterpret regulatory text, omit important context, or generate plausible-sounding but inaccurate summaries. Critical decisions should be validated against the primary source document.

03

Coverage is limited to publicly available materials. Firm-specific supervisory correspondence, confidential regulatory discussions, and non-public enforcement proceedings are not included.

04

Materiality and confidence scores are AI-generated estimates, not actuarial or quantitative assessments. They reflect relative significance within the platform, not absolute risk measurement.

05

The platform does not constitute a risk management system and should not replace an institution's own risk frameworks, internal governance, or regulatory engagement processes.