Watchlist
Organizations in various industries are facing pressure from regulators to better identify their customers, associates, and partners as they manage risk and profitability. Middesk's risk and compliance suite provides multi-sourced data and configurable regulatory and risk services to help identify entities associated with money laundering, terrorism financing, white-collar crime, and other unusual activities in support of your Know Your Customer (KYC) and Anti-Money Laundering (AML) programs.
The Anti-Money Laundering, Counter-Terrorism Financing Rules and Global Official List checks assist in identifying the risks associated with individuals due to specific or general involvement with Money Laundering (AML), Terrorism Financing (CTF), Politically Exposed Persons (PEPs), terrorists and wanted criminals.
How Middesk helps with Watchlist screenings
Middesk runs watchlist screenings against all businesses created in our system. Given that watchlists commonly lack unique identifiers, Middesk leverages its identity matching tools to determine if an entity matches any of the records available on all of the mentioned sanctions lists. Middesk will provide you with the status of each individual screening list result and will flag any potential matches for you to review in further detail.
In addition to the granular details about each watchlist, Middesk will also summarize the watchlist search as follows:
| status | label | sub_label |
|---|---|---|
| Success | No Watchlists hits were identified | No Hits |
| Failure | X Watchlists hit(s) have been identified | Hits |
Name Match Scoring
By default, Middesk uses a string comparison scoring algorithm to output a score (0-100) when comparing business and individual names to watchlist records. The scoring model follows a two-step process: first, it identifies potential matches using Levenshtein distance, then it scores them using Jaro-Winkler and Soundex algorithms. Jaro-Winkler prioritizes matches that start similarly, while Soundex ensures phonetic similarity. A composite score determines if a match exceeds a set threshold, flagging the entity as a potential match.
The watchlist score threshold must be between 80 and 100.
Model Reference
Note that the types of expected matches might vary depending on on the length of the string given how distance-based algorithms used in scoring are designed.
| Risk Tolerance | Expected Matches | Score Threshold |
|---|---|---|
| Exact Match | Exact matches Special characters, non-letters | 100 |
| High | Exact matches Special characters, non-letters, roman letters Single letter modified (added, removed or changed) Name initials added Tokens swapped | 95 - 99.99 |
| Medium | Exact matches Special characters, non-letters, roman letters Single letter modified (added, removed or changed) Name initials added Tokens swapped Extra or changed token Entity designator altered Stop words/non contextual words added One or two tokens removed | 91 - 94.99 |
| Low | Exact matches Name abbreviations Extra or changed single or double letter Extra or changed token Tokens Swapped Entity designator altered Stop words/non contextual words added Name salutations Surname replaced with initial Name replaced with initial 50% tokens removed | 87-90.99 |
Exact Match Detection
When comparing two names, our system first attempts to identify exact matches using the following process:
Both names are converted to lowercase and split into individual tokens (words)
We filter out:
- Common stop words
- Any tokens less than 3 characters in length
- any non alpha-numeric characters
A match is considered "exact" if either:
- All tokens from the input name are present in the sanction name (regardless of order)
- The sorted tokens from both names are identical when joined together
For example:
Input: "John Smith Company"
Sanction: "COMPANY SMITH JOHN"
Result: Exact match (tokens match regardless of order)
Note: This process helps identify matches even when word order differs or when one name contains a subset of the other name's tokens.
Example Records
| Threshold | List Value | Input Name | Score |
|---|---|---|---|
| 100 | JARRAYA, Mounir Ben Habib | @@JARRAYA, Mounir Ben Habib | 100 |
| 100 | THE JOINT STOCK COMPANY STATE RESEARCH CENTER OF THE RUSSIAN FEDERATION TROITSK INSTITUTE FOR INNOVATION AND FUSION RESEARCH | THE JOINT STOCK COMPANY STATE RESEARCH CENTER OF THE RUSSIAN FEDERATION TROITSK INSTI%TUTE FOR INNOVATION AND FUSION RESEARCH | 100 |
| 96 | WASHINGTON TRADING LTD | W4SHINGT0N TRADING LTD | 98 |
| 96 | MAAROUFI, Tarek Ben Habib Ben Al-Toumi | MAAROUFI, Tarek Ben Hablb Ben Al-Toumi | 96.71 |
| 94 | GIGANT KOMPLEKSNYE SISTEMY | GIGANT XOMPLEKSNYE SISTEMY | 95.72 |
| 94 | DROGANOV, Aleksey O. | 100 DROGANOV, Aleksey O. | 94.74 |
| 86 | ZAGARIA, Carmine | GARIA, Carmine | 86.90 |
In order to configure your watchlist settings, visit the settings page in the dashboard. The configuration selections will be applied globally across all businesses.
Other Settings
Middesk also offers several configuration options to reduce false positives, including:
Date of Birth Filter for Submitted People
When DOB is included with submitted people, the model will use DOB as an additional filtering parameter on top of name match scoring.
| Match YYYY/mm/dd (default) | Match on exact date provided |
| Match YYYY/mm | Match on any date within the month and year provided |
| Match YYYY | Match on any date within the year provided |
| Match within a range of +1/-1 year | Match on any date within prior and after the year provided |
Watchlist configuration
Select specific federal agencies and watchlist lists to be searched (see below).
Submitted and Found People names
Select Business only, Submitted people names and/or Found people names (found from Secretary of State Registrations) when searching for Watchlist records.
Lists in Business Verification - Verify
As a part of our Business Verification - Verify product, Middesk supports the following sanctions lists, including those on the Consolidated Screening List. If you would like additional lists added to the product, please reach out to [email protected].
| List | Organization |
|---|---|
| Entity List (EL) | Bureau of Industry and Security |
| Denied Persons List (DPL) | Bureau of Industry and Security |
| Military End User (MEU) | Bureau of Industry and Security |
| Unverified List (UVL) | Bureau of Industry and Security |
| ITAR Debarred (DTC) | State Department |
| Nonproliferation Sanctions (ISN) | State Department |
| Capta List (CAP) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Foreign Sanctions Evaders (FSE) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Non-SDN Menu-Based Sanctions (NS-MBS) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Non-SDN Iranian Sanctions (NS-ISA) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Non-SDN Chinese Military-Industrial Complex Companies List (NS-CMIC) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Non-SDN Palestinian Legislative Council List (PLC) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Specially Designated Nationals (SDN) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
| Sectoral Sanctions Identifications List (SSI) | Department of the Treasury Office of Foreign Assets Control (OFAC) |
Global Watchlists
Middesk also offers Global Watchlist screening as part of our Enhanced Screenings package to extend the coverage of our lists within the US and globally. To learn more, check out Global Watchlists for more information.
Watchlist List Refresh
Middesk's Watchlist lists are pulled from official government lists each day at 3:00 AM PT (the government website refreshes their data at 2:00 AM PT, so this gives their process time to complete, while still almost immediately pulling the results). A job then kicks off to ingest the lists into the database to be searched against ( ~3:30AM PT).
If the lists fail to be pulled or ingested into our database, there is automated alerting to escalate to the engineers and work to resolve the issue immediately.
API Reference
For more information about Watchlist or the Date of Birth Watchlist Filter, view our reference or reach out to [email protected]
Updated 3 months ago