A study from the U.S. Government Accountability Office (GAO) entitled, “Countering Illicit Finance and Trade: U.S. Efforts to Combat Trade-Based Money Laundering” made an important conclusion:
“...financial institutions have limited visibility into the underlying documentation of the majority of trade transactions for which they process the payments, which makes it more difficult for them to identify suspicious activity.”
The innovation of cybercriminals coupled with market forces makes Trade-Based Money Laundering (TBML) an attractive proposition for fraudsters. As discussed in our first article in this two-part series, there are many moving parts to TBML, and the mechanisms of TBML often hide in plain sight. In this second article, the means of preventing TBML are explored with a view to the emerging world of TradeTech, artificial intelligence (AI), and blockchain.
The complex nature of TBML and the data disconnect
Criminals and fraudsters are experts at exploiting existing structures and human behavior.
The people behind TBML are no different, using existing trade networks and financial structures for their nefarious deeds: An example being the use of ‘open account transactions’, i.e., trades not financed by a bank. Even improvements in compliance by U.S. financial institutions such as the requirements under the Bank Secrecy Act (BSA) and related Anti-Money Laundering (AML) regulations, struggle to handle the extent of TBMLs reach.
The financial watchdog the FATF along with the Department of the Treasury said that “TBML is one of the most challenging forms of money laundering to investigate because of the complexities of trade transactions and the sheer volume of international trade.”
Getting to the heart of that expanded network and across country borders, is not easy, and requires due process and ‘TradeTech’ designed for the task.
When investigating suspicious trading activities, banks often have to wade through linked shell companies registered in several countries using multiple bank accounts. TBML events involving multiple companies and banks can easily go under the radar. The data involved is often purposely disconnected, highly unstructured, inconsistent and with poor visibility. There also exists the problem where useful data is purposely routed through multiple countries, some unwilling to share the data making it difficult to have a complete view of activities.
There is a need for a more seamless and agreed method of data sharing across the entire supply chain to help in TBML prevention. However, protectionism and anti-globalization have impacted the appetite to share trade data.
Although there continues to be a reluctance in data sharing, some movement in this area is evident. This is most keenly felt in the supply chain and transport sector. Post-pandemic, organizations such as the World Food Programme (WFP) and logistics companies are collaborating with public bodies on new initiatives to support humanitarian response activities.
In addition, companies can be less reluctant to share if they can benefit from sharing, hence the growing popularity of fraud consortiums where collaboration can be beneficial for individual contributor.
However, the opening up of data channels brings its own issues in the form of vulnerabilities in monitoring trade-based transactions. A fine balance must be met. The incorporation of technologies in trade (TradeTech) could have both good and bad consequences that need to be addressed to ensure TradeTech works for all companies regardless of their size, and all countries regardless of their level of development.
The issues with current AML monitoring models
Data sharing aside, TBML discovery is hard because the activities are often hidden in legitimate trade; the data obfuscated, within complex multi-layered systems, and across jurisdictions. The cost of combating fraud is massive, with banks spending around USD 181 billion in 2019 on anti-fraud measures. These costs are compounded by issues such as false positives: According to research from Microsoft, traditional Transaction Monitoring Systems (TMS) produce over 90% false positives during AML checks. These false positives cost time and money to investigate.
Traditional AML monitoring and verification checks are also hampered by poor visibility across multi-jurisdictional trading systems and disconnected data silos. Something has to give, and this looks set to be a smarter way to leverage data.
Also AML monitoring systems do not change with a change in the behavior of fraudsters. Fraudsters will always disguise their behavior, or repeat, evolve their actions over time. Any useful fraud solution must be able to evolve over-time to intercept and possibly predict fraud. Next generation fraud models will be able to forecast fraud.
The emerging ‘tech-net’ to catch the TBML thief
As the fraudsters behind TBML look forward to continued gains, smart TradeTech, designed to handle the complexities of TBML checks, can help to deliver the net to catch the TBML thief. The world of trade is at an interesting juncture as new technologies begin to mature. Businesses of all sizes, including the Micro, Small, and Medium-sized Enterprises (MSME) are using emerging and new technologies in their day-to-day digitization of business operations and processes. A variety of technologies that can feed into TradeTech for TBML prevention exists now, this includes:
Artificial Intelligence (AI)/Machine Learning (ML)
- Advanced analytics and machine learning, helping to reduce false positives and false negative. Also mMachine Learning is used in sanction screening to spot patterns and trends and contextualize trade documents.
- Trade documents are often in multiple and unstructured formats, making it easy to miss an anomalous transaction. Technologies including OCR (optical character recognition), natural language processing (NLP), and machine learning (ML) and emerging deep learning (DL) techniques can be used to decipher the multitude of document formats associated with trade transactions. AI-based tools augmented by human analysts can then interpret these documents, searching through the copious trade documents for red flags and sanctions subjects. The human element here suggest the AI solution with evolve how investigators do their job and not replace the human element.
- Supply chain visibility and compliance can be enhanced by AI-enabled TradeTech.
- Future AI solution will not only be able to predict who the fraudster is, but also when a fraudster will commit fraud. AI memory based time series model, can now forecast when a fraud will occur and its potential volume.
- Sanction lists are dynamic and continuously updated. This dynamism creates sub-optimal AML checks that are based on sanction lists, TBML techniques exacerbate this issue. For sanction lists to be effective and accurate they need to be optimized by automating updates. Solutions that provide this functionality use blockchain technology. Automated revisions of sanction lists reduce human error and fraud, and in doing so, reduce false positives.
The future is smart TradeTech, not TBML
A report from the FATF published in December 2020, on trends in TBML, came up with several recommendations including:
“Because trade-related transactions are often complex and multijurisdictional, innovative IT solutions, such as graph analytics and artificial intelligence (AI) and machine learning, can be particularly helpful in TBML-related analysis”
Being smart about data is important but how you then view and analyze that data is vital. This is no truer than in the case of Trade-Based Money Laundering. TBML is pushing the limits of compliance and technology. By applying intelligent analysis to the specter of TBML the world of trade and banking can work in unison to remove this heinous crime from our planet.
Deya Innab, Chief Strategy and Product Officer has contributed to the World Economic Forum’s recent report, “Mapping TradeTech: Trade in the Fourth
Industrial Revolution". The report was developed in cooperation with more than 50 global industry experts, policy-makers, academics, and civil society leaders. It provides a comprehensive overview of the impact of emerging technologies on the global trade system, including the top ten most transformative technologies.
Learn more about EastNets Trade Based Money Laundering Solution: SafeTrade.