It’s widely acknowledged that correspondent banks play a fundamental role in the global banking system. Without them, financial institutions would have no access to foreign markets without opening international branches.
Authorized in their regions to provide fee-based services to foreign banks, correspondent banks routinely assist in facilitating wire transfers, account deposits, settlements, currency exchanges and more. Respondent banks (the ones initiating the transactions) also benefit because correspondent banks open doors to capital markets. They also ease the burden of a respondent bank having to develop expertise in a variety of foreign regulations.
However, the very nature of correspondent banking is conducive to wide-spread money laundering and terrorist financing. Correspondent banks too often are at an AML disadvantage because they have to rely on a respondent bank’s ability to monitor transactions for risk. In correspondent banking, knowing your own customer isn’t enough. The correspondent bank must also know its customers’ customers to the greatest extent possible, and they must also be aware of a respondent bank’s other correspondent banking affiliations. Customer due diligence (CDD) is essential for correspondent banks to conduct their affairs in a compliant and safe manner.
The Financial Action Task Force (FATF) is well aware of the need for CDD in correspondent banking. In their recently updated publication, International Standards on Combating Money Laundering and the Financing of Terrorism & Proliferation, they devote two sections to this topic. A few of their recommendations above conventional due diligence include:
- Gather sufficient information about a respondent financial institution to learn the nature of their business, their reputation and any history of regulatory investigation or action.
- Clearly understand the respective responsibilities of each institution.
- Be satisfied that the respondent bank has conducted due diligence on the customers having direct access to accounts of the correspondent bank.
- Ascertain that the respondent bank does not permit its accounts to be used by shell banks.
- Assess the respondent institution’s AML/CFT controls.
While these are all valid recommendations, a host of challenges exist which can hamper effective correspondent banking CDD. One problem is that little information is available about the customer’s customers or pseudo-customers. Another challenge is that correspondent banks frequently must trust that the respondent bank has a robust, AML risk management program in place. Also, typical transaction monitoring systems (TMS) utilize rule-based detection, which often does not cover the complexity, scale and behavioral aspects of AML.
Fortunately, the evolution of more advanced technology has made correspondent banking due diligence easier, exceptionally accurate and readily accessible. New AML and fraud detection systems built on artificial intelligence (AI) platforms use a unique combination of powerful, enterprise-grade innovations to help correspondent institutions meet these once-elusive challenges. AI-driven AML provides superior awareness and management of transactions and relationships passing through a correspondent bank. Suspicious activities of correspondent customers are tracked by:
- Behavioral histories related to the volume of transactions, as well as to amounts of transactions over time;
- Expected volume of transactions and amounts;
- Transactional flow information;
- Graph machine learning;
- AI-trained change in behavior and outlier detection.
The FATF notes that, in an effort to avoid possible money laundering and terrorist financing, financial institutions are increasingly deciding to avoid business relationships with entire regions or customer classes. They warn that this “de-risking” practice can result in financial exclusion, less transparency and greater exposure to these financial crimes. As noted above, they recommend instead that correspondent banks employ due diligence.
And today, there is no greater due diligence than that provided by AI.
It’s widely acknowledged that correspondent banks play a fundamental role in the global banking system. Without them, financial institutions would have no access to foreign markets without opening international branches.
Authorized in their regions to provide fee-based services to foreign banks, correspondent banks routinely assist in facilitating wire transfers, account deposits, settlements, currency exchanges and more. Respondent banks (the ones initiating the transactions) also benefit because correspondent banks open doors to capital markets. They also ease the burden of a respondent bank having to develop expertise in a variety of foreign regulations.
However, the very nature of correspondent banking is conducive to wide-spread money laundering and terrorist financing. Correspondent banks too often are at an AML disadvantage because they have to rely on a respondent bank’s ability to monitor transactions for risk. In correspondent banking, knowing your own customer isn’t enough. The correspondent bank must also know its customers’ customers to the greatest extent possible, and they must also be aware of a respondent bank’s other correspondent banking affiliations. Customer due diligence (CDD) is essential for correspondent banks to conduct their affairs in a compliant and safe manner.
The Financial Action Task Force (FATF) is well aware of the need for CDD in correspondent banking. In their recently updated publication, International Standards on Combating Money Laundering and the Financing of Terrorism & Proliferation, they devote two sections to this topic. A few of their recommendations above conventional due diligence include:
- Gather sufficient information about a respondent financial institution to learn the nature of their business, their reputation and any history of regulatory investigation or action.
- Clearly understand the respective responsibilities of each institution.
- Be satisfied that the respondent bank has conducted due diligence on the customers having direct access to accounts of the correspondent bank.
- Ascertain that the respondent bank does not permit its accounts to be used by shell banks.
- Assess the respondent institution’s AML/CFT controls.
While these are all valid recommendations, a host of challenges exist which can hamper effective correspondent banking CDD. One problem is that little information is available about the customer’s customers or pseudo-customers. Another challenge is that correspondent banks frequently must trust that the respondent bank has a robust, AML risk management program in place. Also, typical transaction monitoring systems (TMS) utilize rule-based detection, which often does not cover the complexity, scale and behavioral aspects of AML.
Fortunately, the evolution of more advanced technology has made correspondent banking due diligence easier, exceptionally accurate and readily accessible. New AML and fraud detection systems built on artificial intelligence (AI) platforms use a unique combination of powerful, enterprise-grade innovations to help correspondent institutions meet these once-elusive challenges. AI-driven AML provides superior awareness and management of transactions and relationships passing through a correspondent bank. Suspicious activities of correspondent customers are tracked by:
- Behavioral histories related to the volume of transactions, as well as to amounts of transactions over time;
- Expected volume of transactions and amounts;
- Transactional flow information;
- Graph machine learning;
- AI-trained change in behavior and outlier detection.
The FATF notes that, in an effort to avoid possible money laundering and terrorist financing, financial institutions are increasingly deciding to avoid business relationships with entire regions or customer classes. They warn that this “de-risking” practice can result in financial exclusion, less transparency and greater exposure to these financial crimes. As noted above, they recommend instead that correspondent banks employ due diligence.
And today, there is no greater due diligence than that provided by AI.
