Fraud Prevention – 3 Data Strategies for Financial Services
Fraud awareness in the Financial Services industry is more important than ever. According to the September 2020 benchmarking report conducted by the Association of Certified Fraud Examiners (ACFE) in response to the coronavirus, 77% of survey respondents, representing a range of industries, have observed an increase in the overall level of fraud as of August, compared with 68% in May. The report reveals 68% of respondents have observed an increase in payment fraud schemes (vs. 60% in May) and 85% anticipate a further increase over the next year.
Given the increase of financial fraud this year and the upcoming holiday shopping season, which historically also leads to an increase, I am taking this opportunity to highlight 3 specific data and analytics strategies that can help in the fight against fraud across the Financial Services industry.
1- Break down the Silos
The specialized nature of transaction processing platforms (account opening, acquiring, ATMs, wire platforms) results in disparate data sources and data management processes. The unique approach each platform has towards a specific crime type creates an environment that is complex and costly to retrofit to counter new crime patterns. This results in duplicate efforts and divides the business, risk, and crime teams, thereby limiting collaboration opportunities. Synthetic identity fraud – where criminals combine real and fake information to create a new identity – is an example of a fast-growing area of financial crime where disparate, siloed systems make identifying this type of fraud more difficult.
Eliminating functional data silos is critical to establishing a horizontal and holistic approach to financial crime prevention. A shared, scalable data store that spans the enterprise enables a holistic approach. A converged data approach enables more comprehensive analysis while reducing duplication of data storage. It can be used by third-party platforms, analysts, data scientists and the lines of business. It transcends silos by supporting shared analytics and collaboration across teams and enables the unification of data and security across fraud prevention, anti-money laundering, and cybersecurity.
An example of a financial institution that successfully implemented an enterprise platform approach to address disparate data challenges is Santander. By combining workloads to a single data platform, the company improved the customer experience and found new proactive control alerts that protect customers from poor outcomes due to financial crimes.
2- Leverage Real-time Data and Machine Learning
Traditional deterministic rules-based fraud solutions flag known or old/known types of illicit activity. These rules are difficult to revise on the fly in order to be effective on newer, sophisticated attacks. As digital interactions increase and new payment models emerge, so too will new varieties of crime. We need fast and flexible solutions to quickly adjust to newly identified fraud schemes. It’s time to fight back with an approach that is equally advanced and agile to address advanced fraud schemes.
Effective fraud prevention requires sophisticated analytical approaches driven by real-time data and Machine Learning. Machine Learning techniques used in simulation models can prepare a financial institution for potential fraud and significantly improve existing financial crime detection systems. Artificial intelligence can be applied based on new types of analysis such as network and graph databases or explorative analytics. This helps increase the effectiveness and efficiency of financial crime investigations and can help to reduce false positives by being more responsive to customer behavior than outdated deterministic business rules. This enables financial institutions to redirect manual efforts to more likely and higher-value suspicious activities.
Bank BRI (PT Bank Rakyat Indonesia (Persero) Tbk), is another example of a financial institution using advanced capabilities in fraud prevention. Bank BRI developed a machine learning model for fraud detection by creating a behavioral scoring model based on customer savings, loan transactions, deposits, payroll, and other financial data. BRI automated the processing of data from multiple customer touchpoints such as ATMs, electronic data capture, and internet banking channels and have very successfully reduced the fraud rate by 40% to record low levels.
3- Get a little Help from your Friends
A third and very critical pillar of fraud prevention is collaboration. This takes several shapes including industry consortiums, public-private partnerships, and solution providers. Participation in consortiums and partnerships offer a wealth of information to help better establish a financial crime prevention strategy.
Consortiums and partnerships are typically organized by a specific focus – industry verticals, data sharing opportunities, crime typology. They focus on prevention approaches across the spectrum of best practices, shared data networks, privacy considerations, machine learning, and AI technologies. For example, a new consortium was introduced a year ago, Payments Risk & Fraud Consortium, focused on the risks associated with the evolving payments ecosystem. The desire to cooperate and best enable technology and innovation, and by extension, better fight financial crime can be seen in examples such as the US regulatory organizations joining the Global Financial Innovation Network GFIN.
Collaborations such as these are at work across the globe. An example of an initiative taking a leading role in such cooperation is the UK’s Joint Money Laundering Intelligence TaskForce (jmlit). It’s a collaboration across the Financial Conduct Authority, over 40 financial institutions, Cifas, a not-for-profit fraud prevention membership organization, and law enforcement agencies. Since 2015, they have been successfully fighting money laundering and wider economic threats.
Solution partners are of course critical to fraud prevention. There are many providers across the space with specific areas of expertise. Here I would like to highlight two of our innovative partners that have implemented cutting edge financial crime solutions leveraging the Cloudera platform.
Simudyne performs fraud simulation utilizing Agent-Based Model (ABM) generated synthetic transaction data. They enable simulation of potential fraud scenarios in a cost-effective, GDPR compliant virtual environment to significantly improve your financial crime detection systems. Simudyne identifies future fraud typologies from millions of simulations that can be used to dynamically train new machine learning algorithms for enhanced fraud identification. Learn more about Simudyne here.
Quantexa connects the dots within your data, using dynamic entity resolution and advanced network analytics to create context around your customers. This enables you to see the bigger picture and automatically assess potential criminal behavior.
Watch this webinar to see a Quantexa demo.
Fraud prevention will always be a challenge. Using state-of-the-art tools and collaboratively working together, we have the best chance to stay ahead of the fraudsters. Visit the Cloudera Fraud Prevention Resource Kit for more information.
The post Fraud Prevention – 3 Data Strategies for Financial Services appeared first on Cloudera Blog.