Critical Success Factors for Effective Fraud Risk Management by Banks

Today, fraud management has taken precedence as a core business issue for banks. Fraud Management is no more viewed as a mere compliance requirement because the scale and impact of fraud has grown significantly in today’s technology enabled world.
On one hand, technology advancement has provided banks with a chance to increase business by offering new services to customers and on the other, it has helped fraudsters become more sophisticated in their methods.

In this era of unparalleled competition, banks are exposed to various fraud related risks – Internal, external, regulatory and reputational. So it is imperative for banks to have a holistic view of fraud and have a robust fraud management strategy in place. Effective Fraud management provides enormous potential for the banks to get more business, reduce operational costs, and increase customer experience at same time.

This article discusses in detail critical the success factors for effective Fraud Management by banks.

i3 consulting over the years has advised multiple banks on fraud management. We have considerable experience in implementing fraud management solutions as well. Based on our thousands of hours of experience in fraud management for various banks in different geographies, we have identified three critical success factors that banks should adopt in order to tackle fraud and provide the best customer experience at same time.

1. Creating a sound fraud management strategy

Banks often tend to see technology as the only driver of Fraud Management Solution. However, the following aspects are very important:

• A robust enterprise-wide fraud management strategy that focuses on comprehensive evaluation of business processes and customer touch points
• A mechanism for regular evaluation of risks and threats to recalibrate the fraud management strategy considering latest risks and threats

Following are the key components of effective fraud management strategy:

2. Creating right enablers for sound fraud management system:

Right enablers for FMS spans across multiple dimensions. Banks can tailor fraud solution across these dimensions according to specific requirements.

Data Effectiveness: Providing quality data acts as a major hindrance to smooth implementation of FMS. It is imperative for the bank to establish right procedure for procuring quality data.

It is also important for the banks to view the entire bank-customer relationship through a 360-degree view of customer. Analyzing how a customer is interacting with various products and services across different channels will help in a deeper understanding of customer behavior. This 360-degree view of customer can be further used by different banking functions such as marketing, collections, recovery etc. It will also help the bank to move towards a unified data model with common data definitions across applications/systems.

Organizations need to view entire relationship between customer and bank

Selecting Right Business Rules: Banks should conduct a 360-degree evaluation of potential fraud risks before recommending first generation set of scenarios and rules

Integrated Fraud Management:
Banks should implement Fraud management systems across all channels or products, and there should be a robust mechanism to calculate fraud likelihood. Fraud score can have widespread implications for the bank. For instance, a new customer whose loan application has been rejected due to high application fraud score might have connections to existing bank accounts like common phone number, email ID or address etc. Some of these accounts might already be delinquent. By identifying these accounts, bank can treat them cautiously and optimize resources. Similarly, a customer with application fraud scores on the lower side but not below the rejection threshold can be monitored closely for few months for any fraudulent activity. Similarly, a customer with low transaction fraud score can be used to reject any additional lines of credit and collections department can closely monitor these customers for any signs of delinquency.

Integrated Fraud management solution will allow the bank to maintain customer profiles at various levels such as channel profile, account profile, customer profile etc. This can be used to monitor customer behavior at various levels and to reduce cross channel fraud.

3. Utilize advanced analytics and machine learning for effective fraud monitoring:

Most banks continue to face challenges that erode the effectiveness and efficiency of their Fraud programs beyond a certain point. Current fraud solutions are often rule-based solutions driven by an institution’s experience of known fraud events. These systems suffer from high rates of false positives, resulting in significant resources focused on investigating low-risk accounts and transactions. It is relatively easy for criminals to understand the linear rules applied by the institutions and then design approaches to circumvent them.
For effective monitoring and mitigation of fraud risk, banks should use advanced analytics for each fraud area such as:

• Using advanced unsupervised modelling techniques to identify homogenous customer micro-groups having similar transactional behavior. Then fine tune scenario thresholds for identified micro groups and use segment wise identified thresholds on top of existing rule engines to reduce false positives
• For new products and services where historical data is not available, self-learning models can be used for fraud detection. These models can quickly recognize anomalies at individual level or peer group level
• For areas where enough historical data is available for model development, supervised machine learning models will provide superior fraud detection. For instance, application fraud models can be trained to identify behavioral patterns of applicants who might turn out to be delinquent.

Opportunity cost of not having robust fraud management mechanism in place can be significant. Banks will have to invest significant amount of time and resources for developing such mechanisms. It is important for banks to focus on these critical success factors to get best results out of their fraud management program. i3c has advised various banking and non-banking clients in their fraud management initiatives.

We helped a leading bank in GCC in reducing fraud incidences by developing a predictive analytics based solution. Bank was struggling with sudden increase in fraud incidents across retail banking debit and credit card transactions resulting in fraudulent transactions of more than USD 20 million over span of 18 months. Client wanted a solution for near real time scoring of all card transactions that can help identify the riskiest transactions. We developed a statistical model by leveraging more than twelve months of card transactions across ATM, POS, and payment gateway, as well as customer behavior through other banking relationship. The resultant model’s predictive power was 78% (GINI) that was significantly better than banks exiting score-cards based solution (35%). Our solution has so far helped prevent more than 225 high value fraud transactions resulting in improved operating efficiency as well as enhanced brand reputation.

Opportunities for Fraud management (and thus the benefits) are not only limited to banking but span across various industries. For example, i3C worked as a thought partner to create and implement a data driven fraud detection strategy for leading pharmacy chain in KSA. Client was struggling with mounting losses due to mismanagement in different departments and individual store level. Client was looking for a framework which could help identify which suppliers and stores were most prone to fraud. We developed a comprehensive data-mart and reporting views across data from various departments including supplier database, inventory, sales, receivables etc. that allowed greater visibility into business performance. We identified 50+ scenarios by analyzing multiple factors through which internal fraud was happening at a supplier and store level. As a result, i3C was able to help the client reduce losses by more than 10 million SAR in a span of 6 months by recommending actions against 10 riskiest pharmacies as well as 25 riskiest vendors across the region.

To know more about our experience, feel free to contact us at  www.i3c.in

About the authors

Sarthak Agarwal
Sarthak works as an Associate with i3Consulting. He has specialized in Organizational performance improvement, Operations analytics and Forensics. 

Shravan Potnis
Shravan leads risk management practice at i3c. He is a risk management expert with an experience of more than 17 years in banking and risk consulting. 

Ashish Jain
Ashish Jain is Associate Director at i3 consulting. He has 12 years of rich experience in BFSI technology and has worked on multiple engagements in the areas of AML, Financial fraud, Compliance and others such as FATCA, Sanctions etc.