Loading
Entity Resolution: The Key to Unlocking FinCEN Data
Entity Resolution provides a view of people, organizations, addresses, phones, and accounts to help identify money launderers and illicit financial networks.
As criminal organizations become more sophisticated and money laundering and terrorism financing continue to grow, reliance on government agencies to safeguard national security against these threats has become critically important. Gathering, analyzing, and disseminating financial intelligence is a huge job for agencies like the Financial Crimes Enforcement Network (FinCEN).
The challenge, of course, is that criminals use sophisticated tactics to conceal the origins of illicit funds, making it difficult for authorities to detect, investigate, and prosecute Analysts often face the nearly impossible task of sifting through mountains of data to identify potential leads or verify information related to known suspects.
Tapping the right resource
Government agencies like the IRS, FBI, DOJ, and those tasked with enforcing adherence to policies in the Bank Secrecy Act (BSA) frequently utilize the FinCEN Query (FCQ) to proactively identify new targets for their investigations often by searching for all reports filed within a specific timeframe or location. They also query the database to gather information on known suspects involved in active or pending criminal cases.
The sheer volume of queries, however, is staggering. IRS Criminal Investigation (CI) , for example, conducted nearly 2 million queries in FY23, with over 87% of all investigations involving a BSA filing related to the primary subject. The DOJ has also executed millions of queries to support their investigations across various offices, agencies, and bureaus.
Despite the availability of digital tools and analytical systems to help streamline investigations, save time and money, and expedite prosecutions, many Suspicious Activity Report (SAR) review teams and task forces still rely on traditional methods to drive investigations— printing documents, physically reading them, and manually highlighting key information. These teams often convene to discuss potential targets without having incorporated historical data and related content. Consequently, the quality of cases is diminished due to the limited scope and depth of the subjects being reviewed.
Improving data analysis with entity resolution
Ideally, BSA data should be presented in an efficient, interpretable, and manageable way. However, with so many different reporting institutions, it is inevitable there are inconsistencies in the way data is collected, recorded and reported. In addition, uncooperative adversaries trying to avoid detection will often misrepresent themselves and provide false or misleading information to throw investigators off their tracks.
A SAR report, for example, might list someone as “John Smith” while several Currency Transaction Reports (CTRs) might refer to the same person as “Jonathan Smith” or “Jon Smith.” If an analyst is trying to uncover related content, they’re at a distinct disadvantage because they cannot see the complete activity of their subject, or which results, or combination thereof, constitute the same entity.
Filling in these blind spots is a huge challenge for investigators. Fortunately, the rise of entity resolution solutions is allowing investigating teams to finally visualize entire networks to identify who the criminals are and who they’re interacting with.
Entity Resolution takes multiple, disparate data points — from external and internal sources — and resolves them into one distinct, unique entity. In the case of investigators looking into money laundering or terrorism financing, they can now work out whether multiple records are referencing the same real-world thing, such as a person (like the John Smith example above), organization, address, phone number, bank account, or device.
Overcoming the challenges in linking entities across BSA filings
With over 4 million SARs and 21 million CTRs filed every year, resolving disparate content from multiple data sources data to identify illicit transactions can be a major challenge.
The Fincen Query System empowers authorized users to download BSA data in various formats, including PDF and XML. All the specific details associated with parsing the XML content are published on the FinCEN website and provide the necessary structures, values, and layouts.
This data is rich in detail, encompassing names, addresses, identifiers, contact information, and financial data. By combining these elements using solutions like the Quantexa Decision Intelligence Platform, analysts can effectively identify individual entities within and across different BSA forms.
Figure 2 illustrates how distinct values from various sources are used to pinpoint specific entities. Applying various transformation, computation, and permutation techniques through platforms like Quantexa’s, analysts can effectively identify individuals even when data is presented differently.
For example, names are analyzed using aliases, Soundex codes, or translations. Dates of birth are matched based on year, transposed month/day, or in defined range. Addresses, phone numbers, and id-numbers are consistently standardized. Additionally, custom heuristics can be employed to handle specific cases or formats for special values.
This capability is crucial for investigators and analysts to connect the dots between financial activities to prioritize investigations and assess associated risks.
Figure 2. Entity Resolution Across BSA Forms
Bringing it all together
By identifying and linking entities across different BSA filings using Quantexa’s Platform, investigators can gain deeper insights into illicit activities and uncover complex networks. Figure 3 illustrates how entity resolution can be applied to connect various BSA sources.
In this approach, entities including individuals, addresses, phones, id-numbers, and accounts are represented as the integration of information from different BSA forms.
This consolidated network provides a more concise and comprehensive view of the data, facilitating the application of rules, scoring models, and the generation of intelligence reports.
Figure 3. Consolidated Entity Network
The level of detail within a network directly impacts the depth of insights gained from the analysis. A single entity can serve as a critical connection between disparate criminal networks when investigating money laundering or how a specific terrorist group is being financed. This same technology can also help identify a central figure in a human trafficking operation.
The implementation of Beneficial Ownership Information (BOI) requirements in 2025 will further enrich the data and enable the identification of even more intricate connections. With an anticipated 35 million filings, primarily from small businesses, the goal is to enhance entity resolution, data fidelity, and the overall understanding of the complex relationships within the data. Every detail, no matter how small, can contribute to a more comprehensive picture.
