FRAUD OPERATIONS | PREVENT, MONITOR, DETECT, INVESTIGATE
EVERY $1 FRAUDULENT TRANSACTION REPRESENTS $3.25 IN TRUE COST FOR FINANCIAL SERVICES COMPANIES AND $3.44 FOR LENDING SERVICES COMPANIES
The banking industry is facing a multibillion-dollar problem as fraudsters are getting more resourceful, leading to increased costs for financial institutions. Fifteen years ago, more than half of all banking transactions took place in the branch. Today, that number is down to 10%, which has led to an increase in customers performing transactions though a digital platform . Due to the fast rise in digital channels, at least one in two transactions is likely to flow through a host of digital ecosystems, FinTechs and other third-party interfaces by 2022.
Various surveys and studies show that by 2022 organizations will be spending ~$9.3 billion annually on fraud detection and prevention tools. With all the digital advancements, fraud systems will become more complex, eventually slowing down service delivery, affecting not only the customer experience, but also the overall reputation of the organization, if these systems are not implemented correctly.
There are different kinds of bank frauds, with some of the most common types being check fraud, debit and credit card fraud, safe deposit fraud and ACH fraud. Modernizing and managing fraud through proper digitalization will enable institutions to stay ahead of financial crimes and reduce their fraud management costs.
WHAT THE INDUSTRY FACES
- Financial institutions continue to spend on fraud tools – big banks spend over $20M per year on fraud prevention
- Criminals are becoming more sophisticated and utilizing technology to commit financial crimes – costing institutions $525M in financial-crime, fraud and cybersecurity
- Too many false positives are increasing the cost of fraud investigation and impacting customer experience - $7 per transaction
- Constant pressure is on to build a well-structured fraud management model
- Choosing the right solution and keeping it up to date is a key to preventing, detecting and investigating a potentially fraudulent activity
TOP FINANCE INDUSTRY CHALLENGES
- Non-standardization, changing regulatory requirements across industries, regions and LOB impact client due diligence efforts
- Existing systems and algorithms are generic and ineffective
- Inability to deploy customizable models results in more false positives and false negatives
- Absence of aggregation of data across customers, products and geographies leads to time-consuming investigations
- Multiple hand-offs cause delays and increase regulatory risks for critical financial crime exit alerts
TYPES OF FRAUD FACED BY FINANCIAL INSTITUTIONS
- Money Laundering – AML fines in 2019 have breached $8 billion
- Credit Card Fraud – ~272K fraud cases in 2019 related to credit cards
- Mobile Fraud – Mobile related fraud has increased from 21% in 2017 to 50% in 2019
- Identity Fraud – 3.2 million cases in 2019 in the US
- Accounting Fraud – Related to business wherein the business “cooks the books” to appear more profitable – remember Enron?
- Loan Fraud – Mortgage, small business and personal loans are among the most common – Currently PPP loan fraud is a significant risk due to the changing regulations
- Wire Transfer Fraud – Real estate incidents are the number one contributor to wire fraud – 11,677 victims in 2019 with $221M in losses
ADDRESSING THESE CHALLENGES
Prevention, detection, investigation and monitoring are the four pillars of fraud management. Embedding operations management within a framework of detection and investigation, and utilizing advanced analytics to incorporate artificial intelligence, machine learning and other digital solutions for prevention and monitoring of fraud, will enable a financial institution to improve segmentation, reduce false positives, reduce complaints and provide effective incident handling.
Digitalization can help financial institutions further mitigate these frauds:
MONEY LAUNDERING – Banks can use smart transaction segmentation to spot money-laundering attempts. Lowering false positives will help banks spend fewer resources to catch fraudsters. Banks can use AI solutions to transform segmentation into a powerful anti-AML process. False positives can be reduced without compromising compliance with regulatory guidelines.
CREDIT CARD FRAUD – This is the most common fraud type wherein banks mostly rely on linear algorithms to differentiate between good and suspicious transactions. Banks today use advanced machine learning algorithms to further segment acceptable and fraudulent transactions. Cloud-based technologies can be utilized to run more sophisticated algorithms on suspicious transactions to potentially flag them as fraud.
MOBILE FRAUD – According to Guardian Analytics, 72% of mobile banking fraud is committed via mobile remote deposit capture. Banks can use mobile devices to their advantage to increase identity authentication, wherein a fraudster cannot access the customer data without using the customer’s phone.
IDENTITY FRAUD – Banking customers need more advanced solutions for protecting their identity and ensuring safe access to banking services. Deep learning algorithms are utilized to help financial institutions develop systems that can match DNA sequences and assist with identity recognition in many ways, which can range from physical and digital signatures to biometric recognition.
ACCOUNTING FRAUD – Utilizing process automation during financials close will help mitigate accounting mistakes made by humans. Organizations can create logical steps which can help close faster and with accuracy. No matter how capable the finance and accounting teams are, it is difficult to reduce risks without technology.
LOAN FRAUDS – Automated loan-processing systems integrated with automated KYC tools will enable financial institutions to identify potential fraudsters. Lenders are seeking solutions that will enable them to be on the digital journey with minimal risk to fraud.
WIRE TRANSFER FRAUD – Implementation of multi-factor authentication, along with biometric sign-on, can help mitigate the risk of wire transfer fraud. The use of voice recognition is one of the most recent innovations that has been introduced by banks to increase security, but still has a long way to go.
There will be a transformational shift in the next decade in areas of fraud detection and prevention, as financial institutions are looking for solutions that can yield benefits and ROI.
SHARED COMPLIANCE COSTS
Organizations, especially banks, are evaluating the concept of sharing cost of compliance and developing a common utility, but there are concerns over data management and scalability that are yet to be resolved.
MANAGED OPERATING MODEL
Stringent regulatory requirements are driving up compliance costs and, therefore, many organizations have implemented a cost-effective, dual-shore operating model.
TECHNOLOGY ENABLED COMPLIANCE HUBS
Organizations are streamlining functions through automation, using technology as a key enabler, establishing a centralized risk management view across divisions and customer segments, and creating global compliance hubs.
EMERGING TRENDS RELATED TO DIGITALIZING THE FRAUD MANAGEMENT PROCESS:
ARTIFICIAL INTELLIGENCE: AI will continue transforming customer profiling and fraud monitoring. AI-based models that will also increase operational efficiency will catch behaviors such as fraud or credit defaults in real-time, or even proactively.
BLOCK CHAIN: Distributed ledger technology, such as blockchain, will transform KYC. This technology will enhance user convenience and security when managing digital identities, while also enabling institutions to easily and reliably manage customer data. While some banks have formed consortiums to build and deploy prototypes of a blockchain-based KYC solution, it will require much global adoption and regulatory intervention before DLT-based technology gains momentum.
CLOUD TECHNOLOGY: Cloud-based solutions can be used for fraud detection, which will also help reduce costs. These solutions will need to be integrated with banking systems to enable internal bank users to have the visibility to make and review changes.
DATA COLLABORATION: Banks will continue to integrate some or all of their anti-fraud and (AML) systems into a single technology environment, in order to gain greater insights about clients by combining internal and external data.
BIOMETRICS: Some large banks have introduced account security features based around biometric data, including iris scanning, fingerprinting, voice recognition and facial identification, to varying degrees of consumer adoption.
DIGITALLY TRANSFORMING FRAUD OPERATIONS
When transforming fraud operations, financial organizations must take an approach that combines data analytics, domain expertise and operations management. A well-orchestrated digital transformation should cover the entire fraud life cycle by utilizing best-in-class processes, while integrating advanced analytics, RPA and artificial intelligence. The goal is to build and deploy a customized solution in a short timeframe, while ensuring no compromise on regulatory requirements. This needs to be combined with a robust and scalable operations delivery model for cost-effective operations and exception handling.
EXL’s digital solution for fraud management has enabled organizations to accomplish continuous fraud prevention, monitoring, detection and investigation via a perfect handshake between operations and analytics.
FRAUD ANALYTICS: PREVENTION AND MONITORING
- Customer and account segmentation will group customers and accounts for targeted monitoring. It will also help in risk differentiation among existing accounts.
- Initial score setting will set up new scenarios for initial monitoring and increased coverage, generating productive alerts.
- Implementing a prioritization engine to supplement alert generation will further reduce false positives based on alert scoring.
- Our tailored approach utilizes model validation concepts and applies them to the model framework.
- The monitoring framework tracks dashboards and provides continuous BTL monitoring
FRAUD OPERATIONS: DETECTION AND INVESTIGATION
- Fraud triage interprets incidents, assesses fraud severity and prioritizes incidents based on defined rules.
- Fraud investigation identifies anomalies, analyzes past trends and catalogs the fraud systems to help operations teams classify the fraud.
- Fraud identification enables data-based investigations, fraud classification and robust case management, which will empower the operations management team to achieve better resolutions.
- Fraud resolution is done in line with industry guidelines for better decision-making and an effective incident-handling process.
- Fraud closure is a preventive fraud control measures that involves a rigorous review mechanism and fraud risk management process.
- Fraud reporting must include fraud control performance dashboards for detailed operations reporting, to help stakeholders make informed decisions.
EXL DOMAIN EXPERIENCE
EXL’s experience in first- and third-party fraud management includes the utilization of analytics and operations management to help organizations enhance customer experience, increase revenue and reduce risks.
Account Takeover (ATO) remains a serious threat and is one of the fastest growing problems for the financial services industry. There have been annual losses of over $6.5 billion across all financial services, insurance, e-commerce and other industries. Advanced analytics solutions enable banks to predict ATOs and incorporate prevention measures to minimize ATOs.
Application Fraud in loan and credit card applications costs the financial sector millions every year and cause damage to the brand, while creating a negative profile for financial institutions. Loss of customers and poor customer experience also costs financial institutions. Early profiling across user interaction helps to avoid this fraud. Furthermore, preventive measures, such as biometric verification and multi-authentication can be utilized to reduce fraud.
Transaction Fraud is expected to soar to $32 billion by the end of 2020. Detecting transaction fraud is going to be of high priority for financial institutions. Machine learning- based solutions can proactively detect fraudulent transactions using classification algorithms, such as loan default prediction, spam detectors and recommender systems.
Payment Fraud has increased due to the shift to digital by mobile banking customers. Quick payment transactions have left banks and processors with less time to identify fund transactions. There has been an annual increase of 34% in average fraudulent transactions. Risk scoring models can help identify and reduce false positives, enabling banks to identify and monitor potential fraud transactions.
Synthetic Identify can exist due to unsatisfactory customer onboarding and due diligence. Integrating a robust digital onboarding solution and fake identity scoring models, along with the bank’s enhanced customer due diligence, helps to reduce fraud related to identity theft.
Straight Roller Fraud recognition is important to preventing nonpayment. $4 to $17 billion is classified as bad debt due to straight rollers. Early identification and intervention can be a swift way of improving operational efficiencies, coupled with collections operations. Deploying analytical models to identify potential straight rollers can also help banks detect these fraudsters.
EXL CAPABILITIES AND DIFFERENTIATORS
Our deep banking and regulatory expertise, coupled with our strengths in operations management, advanced analytics, robotics and AI solutions, has enabled us to deliver significant business outcomes across the financial fraud life cycle with enhanced customer experience..
EXL’s analytics-embedded digital fraud operations solution provides banks with an end-to-end suite of solutions to transform the fraud management process across various LOBs. This enables finance organizations to take a comprehensive, outcome-driven approach towards ensuring compliance, while accounting for changing regulations, enhanced customer experience, and faster identification and prevention of fraud.
Significant impact in 16 weeks in reducing false positive, increasing NPS scores and reducing complaints.
Vice President, BFS Operations