Key Takeaways

Chapter 8: Sanctions Screening: Watchlists, False Positives, and Calibration

Core Concept

Sanctions screening is the process of checking customers, transactions, and counterparties against government-published lists of prohibited individuals and entities. The regulatory cost of a missed true match (multi-billion-dollar penalties, license revocation) is so high that institutions calibrate for maximum sensitivity — producing false positive rates routinely exceeding 99%.

Essential Points

1. Multiple Regimes, Multiple Lists - OFAC (US): SDN List, Consolidated Sanctions List, FSE List, SSI List - OFSI/HM Treasury (UK): UK Consolidated List — independent from EU post-Brexit - EU: Consolidated Financial Sanctions List - UN Security Council: Multilateral regimes (DPRK, Al-Qaida, Taliban) implemented through domestic law - Extraterritoriality is critical: OFAC sanctions apply to all USD-denominated transactions clearing through US banks, regardless of where the transaction originates — making US sanctions de facto global for dollar-clearing banks

2. The OFAC 50% Rule An entity owned 50% or more by a sanctioned person is itself subject to OFAC sanctions even if the entity is not named on any list. This creates a beneficial ownership screening requirement beyond list-name matching.

3. Name Matching Is the Core Technical Challenge - Exact matching: appropriate for document IDs; insufficient for names - Levenshtein/edit distance: good for typos; poor for transliterations - Phonetic (Soundex, Metaphone): good for English phonetics; limited for non-Latin names - ML-based matching: can learn language-specific similarity patterns with training data - All systems require name normalization before matching: Unicode handling, diacritic removal, transliteration, honorific stripping

4. False Positive Rates Are Systematically High - Common names from populations frequently represented on sanctions lists (e.g., Arabic, Persian, Russian names) generate disproportionate false positive rates - Watchlists contain multiple aliases and transliterations per entry — each is a false positive opportunity - Regulatory pressure toward maximum sensitivity calibration creates a structural false positive problem - This has equity implications: certain demographic groups experience higher rates of payment delay and screening friction

5. Supporting Data Is the Primary False Positive Management Tool - Date of birth: the single most powerful disambiguation field - Nationality and country of origin: moderate discriminating power - Document identifiers (passport, national ID): definitive when available - Physical address: useful but easily falsified - A name score of 0.90 + no DOB match + no nationality match = LOW priority - A name score of 0.90 + DOB match + nationality match = HIGH priority requiring urgent review

6. Real-Time vs. Batch Have Different Requirements - Payment screening must be integrated before transaction execution — milliseconds to seconds - SWIFT MT103/MT202 fields (originator, beneficiary, intermediary) must be screened - Customer screening uses periodic batch processing supplemented by designation-triggered re-screening - New OFAC designations require rapid customer-base re-screening "as soon as practicable"

7. True Match Consequences Are Immediate and Specific - US: Block or reject transaction; report to OFAC within 10 business days; maintain five-year records - UK: Freeze assets; report to OFSI within 10 days - EU: Freeze; report to national competent authority - Voluntary self-disclosure before OFAC discovery significantly mitigates penalty exposure

8. The Five OFAC Compliance Program Components 1. Management commitment 2. Risk assessment (institution-specific sanctions exposure) 3. Internal controls (policies, procedures, screening systems) 4. Testing and auditing (including threshold testing against known matches) 5. Training for relevant personnel

Key Distinctions

Connections to Other Chapters

• Chapter 6 (KYC): KYC data (name, DOB, nationality) is the input to sanctions screening. KYC data quality directly determines screening accuracy.
• Chapter 9 (Beneficial Ownership): The OFAC 50% Rule creates a beneficial ownership screening requirement — BO data from Chapter 9 feeds directly into sanctions exposure assessment.
• Chapter 11 (SAR/Case Management): Confirmed sanctions matches generate regulatory reporting obligations distinct from SAR filing — but managed through the same case management infrastructure.
• Chapter 23 (NLP for Regulatory Intelligence): NLP techniques are used to monitor regulatory publications for new sanctions designations — automating the detection of list changes that require re-screening.
• Chapter 29 (Algorithmic Fairness): The demographic disparities in sanctions screening false positive rates are a direct instance of the algorithmic fairness problem explored in Part 6.

Regulatory Reference Points

Next: Chapter 9 — Beneficial Ownership and Corporate Transparency →