Skip to content
RegTech Codex
Directory

Fatf AML KYC Transaction Monitoring

FATF Recommendation 10 Risk-Based CDD: Multi-Factor Customer Risk Scoring & EDD Triggers

Under FATF Recommendation 10 and EBA Risk Factors Guidelines (EBA/GL/2021/02), compliance teams must replace binary onboarding rules with multi-axis Customer Risk Assessment models that prevent score dilution.

By Victoria Ashworth, LL.M, CAMSReview not recorded · Updated
FATF Recommendation 10 Risk-Based CDD: Multi-Factor Customer Risk Scoring & EDD Triggers — RegTech Codex 2026 quantitative & supervisory data card
RegTech Codex (FATF Recommendations, CDD/EDD Beneficial Ownership & SAR Typologies): Four-pillar Customer Risk Assessment weighting, anti-dilution overrides, and EDD governance.Source: RegTech Codex Institutional Research DeskLicense: CC-BY-4.0

Examination findings published by the European Banking Authority, the UK Financial Conduct Authority, and US federal banking agencies (Federal Reserve and OCC) identify score dilution inside Customer Risk Assessment (CRA) engines as the single most common structural flaw in FATF Recommendation 10 implementations. When onboarding platforms at multi-asset brokers like CMC Markets or crypto exchanges like Coinbase compute unweighted arithmetic means across customer, geographic, product, and channel variables, high-risk foreign PEPs or shell holding companies slip into standard auto-approval queues. Explore our FATF AML/KYC Compliance Playbook 2026 inside the FATF AML/KYC hub.

Calibrating the Four Statutory Risk Pillars Under EBA/GL/2021/02#

Both FATF's Risk-Based Approach Guidance and the EBA Revised ML/TF Risk Factors Guidelines (EBA/GL/2021/02, transitioning to AMLA RTS mandates) require institutions to score four distinct risk axes while cross-checking client sophistication under MiFID II Client Categorization and Singapore MAS CMS Licensing Standards:

Quantitative Distribution · Canonical Weight
Canonical WeightHigh-Risk / Auto-EDD Override (Score 75–100)
01Customer & Ownership35%02Geographic Nexus30%03Product & Asset Class20%04Delivery Channel15%
CRA PillarCanonical WeightLow-Risk Baseline (Score 10–25)Medium-Risk Tier (Score 40–65)High-Risk / Auto-EDD Override (Score 75–100)
Customer & Ownership35%Listed tier-1 issuer or domestic salaried W-2/PAYE retailMulti-member operating SME (<= 2 layers)Foreign/domestic PEP, >= 3 offshore holding layers, trust/nominee
Geographic Nexus30%Domestic EEA / UK / US / FATF-equivalent low-CPI homeCross-border residence != funding bank countryFATF Rec 19 Black List, FATF Grey List, EU High-Risk Third Country
Product & Asset Class20%Custody-only domestic ETF/ISA (Interactive Investor)Leveraged margin FX/CFDs (Exness)Unhosted crypto withdrawals, physical commodities, private OTC desks
Delivery Channel15%In-branch or NFC ICAO 9303 ePassport verifiedVideo KYC with domestic bank penny-dropNon-face-to-face API introduced via third-party offshore introducer
Four-pillar FATF Recommendation 10 Customer Risk Assessment engine with max-floor anti-dilution override logic
Four-pillar FATF Recommendation 10 Customer Risk Assessment engine with max-floor anti-dilution override logicVerified Exhibit

Preventing Arithmetic Risk Dilution with Non-Linear Max Overrides#

Consider an applicant incorporated in a low-risk EU state (Geo = 15), opening a standard custody account (Product = 20) via NFC ePassport (Channel = 15), whose ultimate 30% controller is a Foreign Politically Exposed Person (Customer = 95). A naive weighted average yields $0.35(95) + 0.30(15) + 0.20(20) + 0.15(15) = 44.0$ (Medium Risk)—unlawfully bypassing mandatory FATF Recommendation 12 PEP Enhanced Due Diligence! Non-linear max-floor equations cure this failure mathematically.

Anti-Dilution Composite Customer Risk Scoring Formula#

Compliance engineering teams implement a Bounded Max-Floor CRA Function ($S{\text{cdd}}$) combining weighted additive risk with hard regulatory floor triggers ($Fk$):

Formula or model from the articleNot independently verified
S_linear = 0.35 * R_cust + 0.30 * R_geo + 0.20 * R_prod + 0.15 * R_chan
  • Foverride = max(85 * Ipep, 90 * Ifatfcall, 75 * Ifatfgrey, 75 * Icomplexubo_3hop)
  • Scdd = max(Slinear, F_override)
  • Tiering Thresholds:
  • S_cdd < 35 -> SDD / Low Risk (36-month periodic review + baseline TM)
  • 35 <= S_cdd < 70 -> Standard CDD (24-month review + standard peer-group TM)
  • S_cdd >= 70 -> Enhanced Due Diligence (12-month review + corroborated SoW/SoF + MLRO sign-off + sensitive TM)
  • Worked Audit (Foreign PEP Using Domestic Custody Account & NFC Onboarding):
  • Slinear = 44.0, Ipep = 1 -> Foverride = 85.0 -> Scdd = max(44.0, 85.0) = 85.0 (Mandatory EDD Enforced)
Comparison of transaction monitoring alert thresholds and review frequencies across SDD, CDD, and EDD risk tiers
Comparison of transaction monitoring alert thresholds and review frequencies across SDD, CDD, and EDD risk tiersVerified Exhibit

Governance Sign-Off & Audit-Trail Versioning on Risk-Model Weight Changes#

Any modification to pillar weights or country risk scores requires Second-Line Financial Crime Committee approval, historical backtesting against existing high-risk files, and immutable version stamping so examiners can reconstruct the exact CRA score active on any prior onboarding date.

Binding Onboarding CRA Scores to Downstream Transaction Monitoring#

A perennial supervisory finding notes that firms calculate a sophisticated onboarding CRA score and then leave it stranded in a static KYC database while their transaction-monitoring engine applies identical rules to Low and High-Risk clients. Integrate CRA tiers directly into 25% vs 10% UBO threshold escalation, Source of Wealth vs Source of Funds EDD verification, and event-driven Perpetual KYC triggers.

Frequently asked questions

What is "risk dilution" in a Customer Risk Assessment (CRA) model and why do regulators penalize it?

Risk dilution occurs when a customer with a critical risk factor—such as nexus to a FATF Call-for-Action ("Black List") jurisdiction or complex bearer/nominee holding layers—receives a "Medium" composite score because three benign factors average down the total. Supervisors require hard auto-EDD overrides.

Does Simplified Due Diligence (SDD) allow a broker to skip beneficial ownership or ongoing monitoring?

No. Under FATF Recommendation 10 Interpretive Note and EU AML rules, SDD applies only where demonstrably lower risk is proven (such as listed public companies on regulated markets) and still requires baseline identity capture plus ongoing surveillance to verify transactions match expected low-risk activity.

How do FATF "Black List" and "Grey List" updates affect geographic risk weights?

Three times per year (February, June, October), FATF updates High-Risk Jurisdictions Subject to a Call for Action (requiring mandatory Enhanced Due Diligence and countermeasures under Rec 19) and Jurisdictions under Increased Monitoring (requiring elevated geographic risk weights).

When must Customer Due Diligence be completed for occasional transactions vs business relationships?

CDD must be completed before establishing an ongoing business relationship, before executing occasional transactions above USD/EUR 15,000 (single or linked), before any wire/crypto transfer covered by Recommendation 16, or whenever suspicion of ML/TF arises regardless of thresholds.

Publication record

Published
14 November 2026
Updated
16 November 2026
Author
Victoria Ashworth, LL.M, CAMS
Review
Review not recorded
Automated fact check
Passed in the stored pipeline record
Source URLs
3 recorded

AI assistance recorded: anthropic/claude-opus-4.6. This field does not establish human review or source verification.

Academic & Institutional Attribution

Cite This Research & Embed Live Data Card

Open Attribution License (CC BY 4.0)

Standard academic & institutional research paper citation format.

Victoria Ashworth, LL.M, CAMS. (2026, November 16). FATF Recommendation 10 Risk-Based CDD: Multi-Factor Customer Risk Scoring & EDD Triggers. RegTech Codex. https://regtechcodex.org/briefings/fatf-recommendation-10-risk-based-cdd-edd-trigger-scoring
Explore Fatf AML KYC Transaction Monitoring Hub (20) →

RegTech Codex Data, Benchmarks & Contributor Records

FATF Rec 10 Risk-Based CDD & EDD Scoring Model (2026) | RegTech Codex