Anti-money laundering (AML) compliance failures amassed $1.7 billion in fines in the first half of 2018 with over $1 billion of that resulting from actions taken by US prosecutors and regulators, according to the 2018 Mid-Year Anti-Money Laundering Review and Outlook by Debevoise & Plimpton.
Banks and financial institutions (FIs) need to ramp up their anti-money laundering transaction monitoring processes using artificial intelligence (AI), machine learning (ML), and big data to meet regulatory requirements now and in the future. As technology strategically predicts the consumers next need, such as a favorite apparel store aptly texting you a coupon for 15% off raincoats on a wet and dreary day, your AML transaction monitoring process can autonomously identify new threats and patterns of nefarious behavior before you finish your morning coffee (or beverage of choice).
Why move from a rule-led system to an intelligence-led one? A rule-led system fails to detect new potential suspicious activities, is unable to recognize complex patterns in large data volumes, and tends to produce a high number of false positives. Intelligence-led systems can learn right versus wrong detection quickly by interpreting regular behavior versus new behavior through pattern recognition from data such as transactions, behaviors, and time of instances. This helps identify anomalous activity in near real-time, thus allowing banks and financial institutions to identify and report illicit transactions faster as per the regulators expectation.
Taking on the task of moving to an intelligence-led AML transaction monitoring system can feel intimidating at first. However, it is incredibly beneficial to banks and financial institutions. Questions from who will build this technology to who can strategically set up a smart system without periods of extended downtime start running through your mind. Here are three areas of focus to help you get started.
1. Have the right strategy to reduce false alerts
A strategy conducive to blocking illicit transactions boils down to people and process. What is the lead time from data discrepancy to alert, and how long is it taking for an internal team to report the incident? This leaves a company with two choices. Either maintain the status quo by using existing technology and teams and continue to get the same results, or invest in compliance technology with AI and ML capabilities that detect financial crime in near real-time and reduce the number of false positives and negatives.
Choosing the right solution can make or break the strategy. There are many canned solutions on the market, but these come with rule-based automation and analytical limitations. Often, teams are still focusing on data collection and entry instead of investigative work and key insights for risk assessment metrics are often not available. These configurable constraints including poor scenario logic and threshold limitation contribute to compliance breaches. If inefficient or nonlogical scenarios are implemented, they can generate false alerts or create scenarios where the launderers can avoid actions that trigger government reporting, such as Suspicious Activity Reporting (SARs) for example. Working with a service provider to help define a strategy and build a solution that fits your unique business needs is key.
2. Maximize your data’s potential
Banks and financial institutions sit atop oceans of consumer data, and how this data is used is the critical factor in computing accurate results. Static segmentation, as an example, results from outdated or incorrect Know Your Customer (KYC) information recorded during account opening and can lead to false alarms. Inaccurate reporting of customer activity enables institutions to make misinformed decisions, which is why it is imperative to be able to quickly discern the important data from the rest.
The output is only as good as it is configured to be, which means companies need easy ways to aggregate and consolidate all of the data being collected and used. The way data is gathered, stored, and used should be an integral component of the overall AML transaction monitoring plan. This makes the experience less overwhelming and more enjoyable for all parties involved.
3. Standardize reporting for consistency
When there are unanimous standards for what constitutes risk and violation of compliance and what does not, customer identification and account monitoring can be defined and strategically built into a system to strengthen banks’ AML efforts.
With regulations come violations that result in massive amounts of company spending. Inconsistent reporting, such as too many or too little SAR and currency transaction reports completed by an institution, becomes a concern for regulators regarding compliance and the quality of the reporting. This exposes enterprises to regulatory sanctions and excessive costs. Well-defined standards and criteria for reporting developed in parallel to an AML transaction monitoring system can prevent many future headaches by setting the right standards early on in the process.
By letting your processes work (and learn) for you, work gets done quicker and more accurately. Many false positives we receive today stem from poor scenario or threshold logic, incorrect segmentation and profiling, and non-reconciliation of source system data. These can all be bypassed or eliminated by devising the right strategy paired with an intelligence-led technology solution.
The author, Sandeep Kumar Sahu leads the compliance practice for the banking and financial services business unit at Virtusa where he is responsible for practice development, solutioning, deployment of best practices and pre-sales.