Data analysis - the catalyst for a win-win result against financial crime

‘The Hedgehog and the Fox – the fox knows many things; the hedgehog knows one big thing’. 

‘A Whole New Mind’, written by Daniel H. Pink, analyses the left and right brain hemispheres, with their main characteristics and benefits for humans. One of the key differences between the two hemispheres is the fact that the left one analyses the details, participates in the analysis of information, while the right hemisphere synthesises the big picture; it is specialised in synthesis, as it is particularly good at putting isolated elements together to perceive things as a whole. ‘Analyses and syntheses are perhaps the two most fundamental ways of interpreting information’, the author says. To help the readers understand better the concept, he refers to an ancient Greek adage ‘The Hedgehog and the Fox – the fox knows many things; the hedgehog knows one big thing’.  Thus, ‘the left side is the fox; the right side is the hedgehog’.

To perform analyses and syntheses of things, humans have access to lots of data/information about the world. Let’s take this metaphor now in our sphere – that of fighting financial crime by leveraging data and tech.

Fraud detection and AML/KYC require understanding connections and identifying anomalies in links among people, transactions, payment methods, locations, devices, times, and more. This process implies working with huge and ever-changing datasets, in real-time. But there are many situations when data is not available or is not of the right quality; it needs to be integrated into the banks/FIs tools, other times different tools/data are used for different cases, or different teams are working on different topics.

And even when businesses do have access to the data that could reveal illicit activity, they are unable to link the data and the relationships together (due to legacy monitoring systems, different data sets, different teams working on different tools, etc.). This old tech is sometimes expensive to tune, validate, and maintain, it involves manual processes and is generally incapable of analysing the massive volume of customer, entity/institution, and transaction data stored across various locations, formats, and protocols.

Advanced data and analytics techniques such as artificial intelligence, machine learning, natural language processing, cognitive automation, and others can be used to order and analyse large volumes of data while accelerating or automating human labour. Then, clear sources of data can be joined to develop a holistic data strategy, available for every function – KYC, AML, fraud prevention, to fight and prevent financial crime. The goal is to leverage data and technology to identify potential criminal behaviour more cost-effectively and prevent criminal activity from occurring in the first place.

What is currently happening?

The current KYC, AML, and reporting processes are extremely inefficient and time-consuming – these are (in many cases) paper-based processes, leading to errors and risks. Some companies are still using spreadsheets for these processes, thus introducing their own inherent issues.

No focus on customer experience – most of the times customers are unhappy with the difficulty and duration of onboarding processes from the start. After an account has been created, clients might get frustrated because they must repeatedly resubmit all the same information. Depending on the client’s risk profile, this can occur every year. If a client has business with different financial companies, they will be asked for the same information from each of the companies. But this is about to change, as the Millennials and younger generations are finally refusing to put up with substandard options—something Signicat believes is the end of ‘learned helplessness’.

Silos are particularly problematic – disjointed arrangement of teams, systems, policies and procedures, multiple suppliers, and deep silos inhibit data sharing. Silos create a barrier between the ability to connect client risk with transaction risk and might cause poor visibility and unavailability of analyst decision making. This not only creates costly compliance inefficiencies, but the more serious issue of risk is overlooked, and threats are unidentified.

All in all, many banks are focusing on reducing liabilities and efficiency costs, but losses in customer experience, revenue, reputation, and even regulatory compliance are mounting.

Still, if these challenges can be overcome, there are clear benefits.

Tech solutions for KYC and AML

Traditionally, KYC verification and data entry have been manual processes, however, 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.

As the KYC process is mono-dimensional, static, and backwards-looking, Marius-Cristian Frunza, Founder of Schwarzthal Tech, suggests the use of a dynamic KYC process. This is an efficient screen tool, made by a myriad of connected parties’ links that enables FI access to the complex picture of a client. Therefore, a paradigm shift is necessary to improve the concept of KYC, with a dynamic, multi-dimensional, and forward-looking concept aiming to explore the network behind a customer. To reach this goal, financial institutions could tap into tools such as AI and NLP, to collect, visualise, and qualify the information about the network and the underlying risk of a client.

Network analytics can improve the effectiveness of AML programs and detection methods by finding the hidden links between entities. Furthermore, it can reveal the relationships and interconnected transactions that characterise money-laundering activities. Customer risk-rating today is sometimes inaccurate, producing scores that miss high-risk customers and misclassify legions of low-risk customers as high risk. This tech enables financial institutions to create far more effective customer-risk rating models based on better data analytics. These models assess risk factors such as the customer’s occupation, salary, and the banking products used.

According to McKinsey, large banks have adopted an approach that integrates aspects of two other important AML tools: transaction monitoring and customer screening. The approach identifies high-risk customers more effectively than the method used by most financial institutions today, in some cases reducing the number of incorrectly labelled high-risk customers by between 25% and 50%.

Another highly praised tech used successfully to analyse, detect, and visualise complex data patterns, patterns that indicate the potential for fraud, is graph database and analytics. Gartner analysts have highlighted Graph Database & Analytics as a ‘top 10 trend for data and analytic’, with an estimated annual growth of 100% through 2022. This tech is a set of analytic techniques that allow businesses to ‘drill down’ into complex interrelationships among organisations, people, and transactions.

For instance, a bank could use advanced graph analytics to improve its fraud avoidance initiatives, specifically fraud detection for debit and credit cards. It can add graph analytics to its machine learning system to find data connections between ‘known fraud’ credit card applications and new applications. As a result, the financial institution can identify more questionable patterns, expose fraud rings, and terminate fraudulent cards faster.

As important as data is, businesses can’t properly evaluate the data unless it’s in an easy-to-digest format. That is why the TigerGraph team stresses the importance of data visualisation; visualisations are what allow developers to present ‘meaningless’ data in a way that is efficient, effective, and impactful.

Clearly, no one can deny the important role data plays in distinguishing between good guys from bad guys, however one solution/technology does not solve the challenge. Everything needs to be combined, orchestrated, and shared.

This editorial was first published in our Financial Crime and Fraud Report 2021 - How to Fight Fraud and Master KYC, Onboarding & Digital ID, which provides a comprehensive overview of the major trends driving growth in fraud prevention, identity management, digital onboarding and KYC, transaction monitoring, financial crime compliance, regtech, and more.

About Mirela Ciobanu

Mirela Ciobanu is a Senior Editor at The Paypers and has been actively involved in drafting industry reports, carrying out interviews, and writing about innovation in payments and fintech. She is passionate about finding the latest news on AI, crypto, blockchain, DeFi and she is an active advocate of the need to keep our online data/presence protected. Mirela has a bachelor’s degree in English language and holds a master’s degree in Marketing. She can be reached at mirelac@thepaypers.com or via LinkedIn.