6 minute read 2 Dec 2019
Boxer training ring shadow

How data analytics is leading the fight against financial crime

By

EY Americas

Multidisciplinary professional services organization

6 minute read 2 Dec 2019

Show resources

With increasingly rigorous compliance requirements and growing levels of data, fighting financial crime has never been more difficult.

Combating global financial crime activity, from money laundering and market misconduct to sanctions, terrorist financing and bribery and corruption, costs an estimated US$1.3 trillion annually, according to a 2018 Refinitiv Survey¹. With more than US$26 billion in fines imposed by global regulators in the last decade for non-compliance with Anti-Money Laundering (AML), Know Your Customer (KYC) and Sanctions regulations², there is a material need for change.

Governments and regulators are putting financial services firms on the front line in the fight against financial crime, with increasingly rigorous compliance requirements. Trade institutions are finding it particularly challenging to meet these heightened expectations due to manual processes and legacy technologies that no longer keep pace with the huge volumes of data being produced and the complexity of the global banking environment.

Banks that innovate and adopt new technologies and techniques to address regulatory compliance demands will be industry leaders in the years to come.
Blair Delzoppo
EY Asia-Pacific FSO Data & Analytics Partner

From ‘pushing paper’ to learning machines

Traditionally financial institutions have relied heavily on manual, human intervention in the regulatory reporting process; humans literally putting pen to paper. This remains common practise today, particularly in the case management workflow. Several levels of case investigators physically review details and write disposition narratives before suspicious activity and other compliance obligations are reported to regulators.

However, with the enormous amounts of data flowing in and out of banking systems, it’s impossible for humans to keep pace with demand. Risk alert backlogs are more often than not growing faster than operations teams can handle. Advanced data and analytics techniques such as artificial intelligence, machine learning, natural language processing and cognitive automation can be used to accelerate or automate a significant portion of the labor-intensive work, reducing operational costs and leaving the people free to concentrate on preventative interventions.

As well as reducing operational workloads in case management, compliance teams are also leveraging advanced analytics in a range of preventative financial crime use cases including enriching the KYC process, enhancing sanctions screening performance, and monitoring transactional activity, helping to proactively identify risks and opportunities.

We have found that machine learning models not only accelerate the closure of a risk alert backlog, but in most cases have a higher degree of accuracy.
Yoon Chung
EY Asia-Pacific FSO Data & Analytics Director

Innovative solutions to age-old problems

Following are three examples of opportunities for banks to use advanced data & analytics techniques and technologies to help improve regulatory compliance, enhance customer experience and lower the cost of operational risk management.

  • 1. Transaction Monitoring (TM)

    In Anti-Money Laundering, the use of machine learning models can enrich transaction monitoring alerts and boost Suspicious Matter Report (SMR) conversion rates – predicting AML scenarios before they occur. Enrichment adds potentially significant details about the customers, accounts or beneficiaries associated with the alert such as:

    • Prior cases, SMRs or Transaction Threshold Reports (TTR)
    • Existing scoring processes that assess the risk of a transaction, series of transactions, customers or accounts
    • External information such as law enforcement inquiries, subpoenas or negative news

    The use of machine learning models has been shown to detect “true positive” results with greater accuracy than traditional methods, and even predict significant events before they occur.

    Machine learning and AI models are faster and more efficient than humans in AML transaction monitoring; the more data they ingest, the more accurate they become.
    Yoon Chung
    EY Asia-Pacific FSO Data & Analytics Director
  • 2. Know Your Customer (KYC)

    To be KYC compliant, organizations must collect, verify, manage and validate customer data to perform the required due diligence and allow for appropriate customer risk assessments or investigations. However, building a comprehensive ‘single view of customer’ across various internal and external source systems and digital interactions has been a consistent challenge faced by financial institutions for many years.

    KYC verification and data entry has traditionally been a manual and inefficient process – often coupled with critical errors, data gaps and quality issues. By augmenting human activity with machine learning techniques, it is possible to achieve a more holistic view of the customer, enhance the data used to conduct due diligence, and provide a more contextual basis for determining customer risk and detecting suspicious activity.

    Analytics can also enable customer segmentation and profiling for various business purposes, including compliance and marketing. For example, customer profiles could be used by compliance teams for customer risk assessments or investigations. Business or marketing teams could use the same data to create personalized banking offers based on customer preferences.

    By using data-driven insights, fighting financial crime can be faster, more accurate and cost less.
    Blair Delzoppo
    EY Asia-Pacific FSO Data & Analytics Partner
  • 3. Sanctions Screening

    The performance and effectiveness of screening engines are becoming under pressure due to rapidly changing and increasing regulatory demand, with risk detection capabilities of existing systems unable to keep pace. A typical symptom of poor screening efficiency is an ever-growing backlog of screening alerts and unsustainable levels of false positives, both factors having a direct impact on operational costs.

    At the core of an effective screening solution is an uplift of the completeness and accuracy of the data being ingested by the screening engine. Tuning the matching and filtering performance of the screening engine requires the data to be complete and of high quality, ultimately resulting in a boost in true positives detection rates and operational efficiency.

    In addition to ensuring the screening engine is operating at peak performance with accurate data, emerging AI and analytical methods can also be used to address operational efficiency issues related to case investigation.

    Machine learning techniques can be coupled with predictive calculations based on historical investigator decisions to substantially lower the number of alerts to be safely dispositioned, so investigators can focus on those with the greatest likelihood of being true positives.

    By helping to build processes that are more likely to result in complete and accurate data, and optimizing the engine to avoid false positives, the effort and cost involved can be reduced.

Intelligence-led and data-driven approach to fighting financial crime

It’s evident that financial service organizations are being challenged both internally and externally in keeping up with the onerous demands of mitigating financial crime risks.

To align operational effectiveness with these demands, organizations are having to seek innovative ways to address issues surrounding SMR conversion rates, KYC due diligence and screening alert management.

There is an increased appetite among banks to go beyond simply flagging suspicious activity for compliance purposes. The goal is to leverage data and technology to more cost-effectively identify potential criminal behavior and prevent criminal activity occurring in the first place.

Complete and accurate data is essential to resolving these issues and an uplift of data quality will have immediate effects on the performance of existing monitoring and screening engines.

Advanced analytics and cognitive techniques, such as AI, machine learning and automation, can help filter out false positives and improve inefficiencies in existing investigative processes. There are opportunities for data and analytics to not only drive efficiencies and operational cost reductions, but more importantly to identify intelligence-led and data-driven ways to tackle financial crime.

When I talk to clients, they believe that our combination of professional skills and advanced data and analytics products are what help them accelerate results.
Blair Delzoppo
EY Asia-Pacific FSO Data & Analytics Partner

Summary

Fighting financial crime has never been more difficult for banks, who are still relying on manual processes to identify potentially suspicious activity. Advanced analytics techniques such as machine learning and AI models can be used to automate risk detection and increase accuracy.

About this article

By

EY Americas

Multidisciplinary professional services organization