By Hazem Mulhim
As published in the European Financial Review
The pressure on financial institutions to police money laundering has never been more intense. Regulations are getting stricter and fines are increasing, according to a study by National Economic Research Inc. When banks are accused of negligence or wrongdoing, the financial markets are unforgiving. Shares of Danske Bank — where nine executives face preliminary charges over €200 billion in suspicious transactions — have dropped by more than 50% since the allegations surfaced.
Making matters worse, the challenge of rooting out nefarious transactions is becoming more complex. Money is moving faster than ever. Criminals are becoming savvier, and technology has introduced new ways to hide illegal earnings and integrate them into the legitimate economy. An estimated 2% to 5% of global GDP is laundered each year, a staggering $800 billion to $2 trillion, according to a United Nations estimate. In other words, money laundering is not only occurring, it is prevalent within many financial institutions, unbeknownst to their risk and compliance officers. This renders them vulnerable to the stigma, expense and destabilization of a scandal.
Anti-money laundering lapses rank among the top non-financial risks to a bank’s credit rating, citing several downgrades of higher-rated developed market issuers. In May, a Fitch Ratings report said anti-money laundering lapses rank among the top non-financial risks to a bank’s credit rating, citing several downgrades of higher-rated developed-market issuers.
Many banks are getting the message. They’re using artificial intelligence and other fast-advancing technologies to identify money laundering that may be in their midst. Emerging technologies are being fine-tuned to revolutionize compliance with anti-money laundering (AML) directives. At a minimum, bankers and their IT and risk officers need to keep an eye on them because they are aimed at overcoming one of banks’ biggest disadvantages in detecting nefarious transactions: the immense amount of manual work that must be done.
Yet banks must also scrutinize these technologies and try them out before writing big checks. That will help them reduce the ever-rising risk of not meeting increasing regulatory obligations. It will make them familiar with tools that enforcement officials may soon consider to be required best practices. Moreover, by harnessing techniques such as automation and machine learning, the new tools can improve compliance efficiency and offer relief from the high cost of compliance.
Some of the new technologies improve upon existing tools, as we described in a recent report. Most financial institutions already use software-based screening to check clients and transactions against watch lists. But in the case of trade-based finance, legacy tools have proven only partially effective. They tend to conduct a single check, often when a transaction is initiated. But in the weeks or months between when a letter of credit is issued and the seller collects, transaction details may be modified, watch lists may be updated, or regulations can change.
A shipment may be rerouted via a blacklisted port or vessel. New, advanced AML-compliance screening software scans SWIFT messages and other trade documents not just once but continuously until the transaction has been completed. This reduces the chance that a financial institution will complete a transaction that has been blacklisted by regulators during the transaction lifecycle, or that has been rendered illegal via an amendment to the documents.
Other technologies, such as link analysis, streamline and improve upon tasks that previously demanded expensive and time-consuming investigations. Link analysis exploits the fact that criminals tend to work in networks. It uncovers hidden relationships among clients and accounts, and visually displays them so investigators can immediately understand them. Using analytics, such technology can identify links several steps removed, including ones with individuals outside the institution.
These nefarious connections can be difficult, if not impossible, for investigators to find manually. Upon identifying a network, the system ranks its risk profile. That helps a bank focus on those most likely to be involved in money laundering. Link analysis is particularly useful in complying with “beneficial ownership” provisions, such as FinCEN’s Beneficial Ownership Rule, and the EU’s Statutory Instrument No. 560 of 2016 Anti-Money Laundering: Beneficial Ownership of Corporate Entities.
AI-based tools can now scan bank data to reliably prioritise such anomalies according to which ones are the most likely to be suspicious. This reduces false positives and improves investigators’ ability to find real criminals. Artificial intelligence is also boosting the speed, efficiency and accuracy of anti-money-laundering activities. Investigators have traditionally been burdened by the need to manually examine each anomalous transaction – such as large deposits at bonus time. AI-based tools can now scan bank data to reliably prioritize such anomalies according to which ones are the most likely to be suspicious. This reduces false positives and improves investigators’ ability to find real criminals.
Remarkably, other AI applications are capable of finding criminal patterns previously unknown to bankers or enforcement officials. Using a technique called “supervised learning,” investigators help such applications become experts in identifying criminal activity. The applications’ machine-learning capabilities then scan the institution’s data to find new typologies common to these criminals. The machine applies this new knowledge across the bank’s data to identify additional suspects. Importantly, regulators have indicated they will not penalize institutions which report criminal activity that previously went undetected using previous methods.
Money laundering poses an increasing risk for every financial institution. Deploying these advanced tools can go a long way toward identifying suspicious activity, improve investigator efficiency, and significantly reduce the chances that a bank faces a money laundering crisis.