Predictive Analytics & the fight against Money Laundering
Soufiane OMRANA, FICA, CFE, CAMS, CFCS
Certified CO/MLRO - I turn compliance into a strategic advantage, fuelling growth with simplicity and impact!
This article was produced few weeks ago in the framework of the Digital Transformation Program (ESSEC Business School in France). It brings more perspective to the different articles shared related to the recent implementation of AI tools by the French regulator (ACPR) as well as the concept of "invisible finance" being more and more popularized in the industry.
Please do reach out to me if you are interested in sharing knowledge or have some specific projects to be discussed. Do not hesitate to share your views on my article and start the conversation!
Despite a stronger regulatory environment in almost all the jurisdictions around the world, the volume of money laundering and other financial crimes is growing world-wide. The techniques used to evade and detect money laundering and other financial crimes are more and more sophisticated. Collectively, with the help of local law enforcement authorities and supervisors, the banking industry (and specifically retail banking targeting individual clients with plain vanilla products such as saving or current accounts, credit cards and digital banking) invested heavily in improving its program against financial crime. This paper will try to demonstrate that Data Analytics and Machine Learning, when used properly, can significantly reduce the risks associated with financial crime. According to Napier (2020), “Predictive analytics is all about using machine learning techniques in conjunction with data, to predict what will happen in the future”.
Trust as a critical Key value driver in the banking industry
Indeed, and according to a recent report, financial institutions spent an estimated $214 billion on financial crime compliance. Other reports published in 2018 estimate that compliance costs annually $1.3 trillion. Indeed, with more than $26 billion in fines imposed by global regulators in the last decade for non-compliance with Anti-Money Laundering, Know Your Customer (KYC) and sanctions regulations, there is clearly a need for change not only in the leadership aspect of managing financial crime, but at the more operational level as well. The evolution in customer behaviour, combined with rapid innovation and never-ending regulatory requirements are-shaping the banking industry in an environment dominated by macro-economic shocks. Trust, innovation and macro-economic environment can be considered as key value drivers of the industry.
Being proactive versus reactive: opportunity versus threat in the compliance area
On the one hand, financial inclusion, sustainability and mobile banking represent growth opportunities. On the other hand, fintech players are disrupting the industry by offering innovative growth opportunities at lower cost thanks to their investment in technology and Artificial Intelligence in particular, and specifically in the Compliance area. Regulators have always been reactive in their requirements. However, there is now an opportunity for banks and financial services industry to be proactive in the compliance domain. Recent enhancements in Machine Learning (ML) are helping banks to improve their AML programs significantly, and specifically the transaction monitoring element of these programs. In addition, regulators around the world are establishing a relevant framework to support such initiatives conducted by the industry. It is the case in France with the recent initiatives offered by the local regulator (ACPR). According to a recent McKinsey study, 80% of the participants had begun the process of adopting ML solutions within their AML framework, while in the same time fines continue to grow (more than $26 billion in fines imposed by global regulators in the last decade). The rule and scenario-based approach used by the industry seem to be outdated. These rules and techniques (using basic statistical techniques) are driven by red flags, expert judgement and limited amount of data. As per the growing volume of money laundered around the globe, rules often fail to capture the latest trends in money laundering techniques and behaviour. There are great benefits in implementing a predictive analytics framework in the banking industry. However, there are some challenges and pre-requisites that need to be considered:
·??????Data Quality: the predictions made by Machine Learning rely mainly on the quality of the data put into the model. A broad range of data will be needed.
·??????Selection of appropriate model: there are various types of models that can be used for predictive analytics, and each has its benefits and externalities.
·??????Accuracy and Reliability: outcomes of the models need to be explained to a wide range of stakeholders. Therefore, models need to have the capacity to exploit the data and build relevant reports to take relevant decisions.
How are the current market leaders using Predictive Analytics and for what purpose?
Leading banks apply ML across the entire AML value chain (or broader within the entire Compliance value chain). However, Machine Learning and Data Analytics seem to be producing a better outcome when used specifically within the transaction monitoring system. Indeed, Machine Learning, combined with dedicated advanced algorithms (random forest or deep learning) can provide tremendous benefits. Machine learning models focus more on behaviour-indicative and sophisticated algorithms. In addition, they are more flexible and can be easily adjusted to new trends and continuously improved using unsupervised learning techniques. There are several best practices that leading financial institutions are implementing:
·??????Align stakeholders on vision and design, and more specifically on the strategy of the firm: Data Analytics is first a leadership problem. Stakeholders are clearly identified and engaged at an early phase to align on vision and decide on architectural design choices.
·??????Anticipate Change Management issues: successful transformation always involve change management.
·??????Enhance the existing Enterprise Risk Management framework
How are Predictive Analytics & Machine Learning disrupting the industry and changing the competitive dynamics?
According to a recent released by Mckinsey (2022), leading financial institutions improved suspicious activity identification by up to 40%.?The combination of both Machine Learning and Data Analytics is called “Predictive Analytics’. This approach can be used to detect money laundering at an early stage and relies on sophisticated algorithms. In a 2017 article published in the Journal of Financial Compliance, it is demonstrated that predictive analytics has been successfully used in a wide range of banking areas, specifically in applications for credit or debit cards, online and mobile banking and of course for AML purposes. Predictive Analytics is not only improving anomaly detection but is enhancing trend and behavioural analysis to understand associated predictions. All the deviations highlighted will warrant investigation reducing cost, increasing trust in the ecosystem and therefore providing a better client experience (less KYC & compliance burden), while providing transparency and securing the environment.
领英推荐
Conclusion:
In the fight against money laundering, banks have traditionally been reactive. Advanced analytics and cognitive techniques will not only drive efficiencies but will enable to tackle financial crime. Predictive Analytics might serve the strategy of the organization and can be considered as a distinctive capability that can facilitate reaching a competitive advantage.
References:
True cost of financial crime compliance study, LexisNexis, 2020.
How data analytics is leading the fight against financial crime, EY, 2021.
How can predictive analytics improve AML compliance, Napier, 2020.
Predictive Analytics in Fraud & AML, Journal of Financial Compliance, 2017.
Revealing the true cost of financial crime, Refinitiv, 2018.
The fight against money laundering: Machine Learning is a game changer, McKinsey, 2022.
ACPR initiatives related to Artificial Intelligence:
Founder & CEO SimpleAccounts.io at Data Innovation Technologies | Partner & Director of Strategic Planning & Relations at HiveWorx
9 个月Soufiane, Great insights! ?? Thanks for sharing!