Assessing the effectiveness of fraud management strategies Employed by financial institutions: a case of equity bank Kenya limited

Abstract

Equity Bank Kenya Limited, a key player in the Kenyan banking industry, has implemented various strategies to detect, prevent, and respond to fraud. Despite these measures, the persistent recurrence of fraud cases necessitates a thorough evaluation of the bank's fraud management framework. This study sought to address this gap by examining the fraud detection, prevention, and response mechanisms employed by Equity Bank and exploring potential advancements to strengthen its fraud management strategies. The findings contribute valuable insights into fraud management practices within the Kenyan banking sector, offering lessons for other financial institutions in emerging markets facing similar challenges. The objectives were to evaluate the fraud detection strategies used by the bank, analyze its fraud prevention measures, assess the effectiveness of its fraud response mechanisms, and identify possible advancements to enhance its overall fraud management framework. Guided by Donald Cressey's 1953 Fraud Triangle Theory and Cohen & Felson's 1979 Routine Activity Theory, the study adopted a mixed-methods approach with a descriptive research design. Data was collected through structured questionnaires from a stratified sample of 333 respondents and semi-structured interviews with a purposive sample of 10 interviewees from key departments, including Internal Audit, Compliance, Risk Management, and IT Security. Quantitative data was analyzed using descriptive statistics, and qualitative responses through thematic analysis. The findings revealed that, while whistleblowing mechanisms exist, their effectiveness is undermined by employee reluctance due to fear of retaliation, with a section of the respondents expressing uncertainty about whistleblower protection. Although AI-driven fraud detection systems have contributed to a notable reduction in fraud-related losses, concerns persist regarding high false-positive rates, leading a number of respondents to perceive current AI and Machine Learning (ML) implementations as ineffective in overall fraud detection at Equity Bank. Regarding fraud prevention, measures like staff training and internal controls are present; however, a substantial number of employees consider the current fraud prevention training inadequate. While most respondents agreed that fraud containment is effective, key challenges include delays in fraud case resolution and unsatisfactory financial loss recovery rates, with many strongly disagreeing that recovery rates are adequate. To enhance fraud management, the study recommends strengthening whistleblower protection through clearer policies and enforcement at Equity Bank, refining AI-based fraud detection models to reduce false positives and improve accuracy within the bank's specific operational context, expanding and regularly updating training programs to address identified knowledge gaps among employees, and integrating blockchain technology for enhanced transaction security and traceability. Additionally, proactive leveraging of predictive analytics and machine learning models is proposed as crucial technological enhancements for Equity Bank's fraud management mechanism. These recommendations are particularly suited to Equity Bank’s expansive digital operations and can serve as a model for fraud management across Kenya’s banking sector and similar institutions in developing markets. The scope of the study was limited to the bank’s Kenyan operations, and findings may not fully generalize to smaller institutions or those with differing regulatory contexts. Additionally, restricted access to internal data and legacy technology systems posed constraints in evaluating all aspects of fraud management.