Trade-based money laundering (TBML) is rapidly becoming a preferred method of money laundering for transnational drug trafficking and terrorist organizations. Simply defined, TBML is the process by which criminal organizations utilize legitimate international trade to mask their criminal proceeds. They do so by selling counterfeit goods, falsifying trade documents, over/under invoicing and falsifying financial statements.
TBML is effective for criminal groups because trade finance is document-intensive, requiring financial institutions (FIs) to expend significant effort to monitor and investigate the paper trail and other trade finance lines of business. That’s where the utilization of data science such as artificial intelligence (AI) and machine learning comes in.
International trade has remained constant since the 2008 financial crisis. The World Trade Organization and the World Bank Group both estimate that global trade reached approximately $16.5 trillion in 2015. Additionally, The World Bank Group estimates that global remittances will increase to more than $636 billion in 2017. Increased trade coupled with growing remittances will create challenges for an unprepared FI. Global trade needs remittances for down payments, payment of shipping costs and freight forwarders, purchase of goods/services, and payment on settlements. Each remittance holds clues that AI can exploit to better mitigate risk for FIs.
Some of the main red flags associated with TBML are:
- Trade does not match the normal business line for the shipper and receiver
- Inconsistent bills of lading for item(s) shipped
- Over/under valuation of merchandise
- Round tripping of the same merchandise
- Unusual land/sea shipping routes
- Double-invoicing for the same goods/service
- Third payments to vendors
AI can assist banks and covered institutions in unravelling the TBML web by analyzing NAICS (North American Industry Classification System) codes of related parties to ascertain if the transaction matches the lines of business such as the lines of business between a shipper of a manufacturing company and the product receiver. AI can analyze invoice numbers for patterns of re-use/falsification, compare values of merchandise to known correct values, and analyze vendor payments. Trade finance and TBML investigations are complex and time-consuming, but advanced data science solutions can automate much of the tedious and manual investigative process, allowing investigators to focus on the more complex aspects of the investigations.
Apart from the egregious nature of drug trafficking organizations, the terrorist organization Hezbollah is operating in our backyard from the tri-border area (TBA) in Paraguay, Brazil, and Argentina. Hezbollah cells have decades of experience in TBML and counterfeit goods sales to fund their terrorist activities. The Hezbollah TBML threat is increasing and FIs need to be better prepared to prevent, detect, and report TBML red flags.
An actual TBML case worked by the QuantaVerse team with the government that resulted in the prosecution of a foreign national is illustrated below to show how effective TBML is hiding and moving money for criminal organizations.
In conjunction with a trade finance transaction, a foreign national submitted invoices and bills of lading to a United States-based FI purporting the purchase and export of a crop duster aircraft (A/C). The documents indicated that the A/C was valued at approximately half a million dollars. The A/C was exported to a foreign country where new invoices showed it was valued at less than $100,000. When investigators traced the A/C, they found it was a previously crashed A/C and valued at only about $4,000.
Effective trade finance monitoring and investigation and TBML prevention and detection programs require FIs to have proactive training, good investigators, and a data solution to provide a holistic view of the transaction. Existing AI solutions available today provide that coverage and can easily assist FIs in mitigating their trade finance and TBML risks efficiently and effectively.