What Money Laundering Looks Like in Community Banks
Although they generally have smaller compliance teams and budgets, the nation’s community banks face the same arduous compliance challenges as their larger bank counterparts. Artificial intelligence and machine learning solutions can greatly assist community banks by detecting suspicious entities and transactions in an effort to improve regulatory compliance and AML investigation efficiency.
According to the United Nations Office on Drugs and Crime (UNODC), the amount of money laundered globally in one year is 2 to 5 percent of the global GDP, or upwards of $2 trillion in U.S. dollars. Financial institutions face significant risk from money laundering and other financial crimes, and community banks are no exception.
A crucial part of the U.S. financial system, community banks sustain local economies by supporting local businesses and enabling local commerce and employment. In fact, community banks make up more than 50 percent of small business loans and 82 percent of agricultural loans, and the Independent Community Bankers of America (ICBA) reported that one out of every five U.S. counties have no other physical banking offices except those operated by community banks.
The Risks Are Real for Community Banks
Unlike national and global banks that deal with sophisticated risk scenarios stemming from correspondent banking, securities trading and trade finance, community banks have a considerably smaller pool of transactions to monitor. However, they still share the same responsibility of mitigating their AML risk in order to comply with state and federal regulatory requirements. In contrast to community banks, larger international banks have compliance teams with staffs in the thousands, and they devote upwards of 20 percent of their operating budget to compliance. Community banks simply can’t afford the sheer quantity of personnel required to address AML and other compliance challenges as easily as their larger counterparts.
What’s more, white-collar criminals and money launderers are known to attempt to evade larger institutions’ compliance programs by operating through community banks. The risk is further compounded for community banks located in high-risk geographies, as well as those that transact business with risky customers in other regions. The current tendency of large banks to “de-risk” by avoiding or terminating customer accounts found to present a higher AML risk poses further risk; bad actors turn to community banks to carry out their illicit activities rejected elsewhere.
Community banks often take comfort in the greater familiarity they enjoy with their customers. However, some community banks put themselves into a vulnerable position by placing too much trust in their relationships with local customers that might be taking advantage. New, affordable and easily implemented data science technologies can help alleviate these risks even for smaller bank operations.
The Deficiencies of Today’s Anti-Money Laundering (AML) Processes
All AML practices leverage transaction monitoring systems (TMS) to identify suspicious transactions. Despite processing many fewer transactions, community banks are required to maintain effective TMS to seek out structuring transactions, funnel account activity and other common money laundering and terrorism financing red flags.
Unfortunately, in its search for suspicious activity, a TMS catches many normal transactions of legitimate clients. These “false positive” alerts trigger human investigations. The industry estimates that approximately 95 percent of the alerts generated by TMS are false positives. Nonetheless, regulatory statutes require all alerts to be adjudicated, which is a time-consuming and expensive process.
On the other hand, AML “false negatives,” which are instances of illicit transactions that are not flagged by a TMS, pose significant regulatory and compliance risk for institutions. It is estimated that 50 percent of financial crimes pass through TMS unnoticed.
Artificial Intelligence Radically Improves Compliance Efficacy
Unlike rules-based TMS that only consider if a transaction does or does not violate a static rule it has been programed to test for, artificial intelligence (AI) and machine learning can detect patterns of behavior, analyze the intent of those patterns and expose anomalous activities that a community bank would have simply missed. Artificial-intelligence-based solutions have the capability to cull, consolidate and analyze large volumes of data (i.e., all transactional data and KYC information across multiple lines of business) and compare that data to typical, licit patterns to identify and risk-score suspicious activity so that investigators can make faster, more accurate determinations.
The following AI and machine learning techniques can be employed to help community banks identify transactional anomalies worthy of further investigation:
- Collaborative filtering: capable of finding transactions with missing, matching and/or odd information
- Feature matching: utilized to identify transactions below a specific monetary threshold
- Fuzzy logic: used to find data matches with slight changes to names or addresses
- Cluster analysis: can detect abnormalities in transactions benefiting a single person or entity
- Time series analysis: detects transactions benefiting a person or entity over an extended period
- Focused keyword searches: ability to dynamically monitor, screen and filter transactions based on keywords from high-risk AML, CTF and financial crimes typologies
- Ability to learn from an AI-identified suspicious activity to enhance transaction monitoring and KYC platforms
One money laundering scenario that a community bank might encounter begins with a drug trafficker opening a business account at a local branch, stating that he recently started an export business. To avoid detection, he breaks up his ill-gotten gains and deposits one small portion into the community bank account each week. The account also receives transfers from two to three larger banks that he reports as employment-related reimbursements. The bank’s TMS does not trigger any alerts because the deposits aren’t deemed suspicious, and the drug trafficker continues to run his money laundering scheme. This failed AML effort is the result of the bank relying on the typical, manually-intensive customer due diligence investigation of the “business owner.” Alternatively, an AI-based AML solution would have linked the individual and his aliases to other shell companies and known criminal associates. In other words, artificial intelligence could have enhanced the community bank’s customer and entity resolution to solidify their risk management program.
While community banks have smaller compliance teams and lower budgets than their larger bank brethren, the risk stemming from money laundering and other financial crimes is just as high. Data science solutions, infused with AI and machine learning, are both available and affordable to today’s community banks. The proven capabilities of AI can help community bank operations by exposing suspicious entities and transactions, strengthening regulatory compliance exponentially and improving the AML investigation process.