abstract: Using historical data from the third-party payment acquiring industry, I develop a statistical model to predict the probability of fraudulent transactions by the merchants. The model consists of two levels of analysis – the first focuses on fraud detection at the store level, and the second focuses on fraud detection at the merchant level by aggregating store level data to the merchant level for merchants with multiple stores. My purpose is to put the model into business operations, helping to identify fraudulent merchants at the time of transactions and thus mitigate the risk exposure of the payment acquiring businesses. The model developed in this study is distinct from existing fraud detection models in three important aspects. First, it predicts the probability of fraud at the merchant level, as opposed to at the transaction level or by the cardholders. Second, it is developed by applying machine learning algorithms and logistical regressions to all the transaction level and merchant level variables collected from real business operations, rather than relying on the experiences and analytical abilities of business experts as in the development of traditional expert systems. Third, instead of using a small sample, I develop and test the model using a huge sample that consists of over 600,000 merchants and 10 million transactions per month. I conclude this study with a discussion of the model’s possible applications in practice as well as its implications for future research. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2015

Thesis (MBA (Business Management))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: The mobile phone has not only changed the way the world works today, but also changed the direction the world is moving toward. The mobile phone changed the face of communication and enabled more people to reach more other people than ever before. The big four banks in South Africa represent 83.5 per cent of the total asset value of all banks in South Africa. Traditionally, banking customers (or potential customers) could be reached through the current 2 786 branches, 19 451 ATMs and potentially 4.59 million internet users. There are 47.9 million mobile phone subscribers in South Africa – increasing the number of potential contact points by order of magnitude. The possibilities for banks utilising the mobile phone are endless, however online banking and offering banking services through a mobile phone is increasingly more subject to fraud attacks. Online banking and credit card fraud is still on the increase. Today, SMS is used to alert customer of movements on their bank account. This keeps the customer informed and enables them to notify their banks and prevent subsequent fraudulent transactions. SMS can be sent from one mobile phone to another (P2P) or from a computer system to a mobile phone (A2P). In 2007, 2 trillion SMS's were sent worldwide and was responsible for 75 to 80 per cent of all mobile phone revenues. South Africa sent 34 billion A2P SMS in 2009 of which 29 per cent were sent as transactional SMS by the top three banks in South Africa. SMS-based transactional alerts are SMS sent each time a change occurs in a bank account, for example, when your credit card is used then you will get an SMS on your mobile phone. Each bank makes different functionalities available. Absa reported 2 million customers receiving SMS alerts in 2008. FNB reported 67 million transactional SMS per month in 2009. The core advantages for using transactional SMS are the cost, reliability and ubiquity. Research was conducted among five of the six largest banks. Data revealed that banks send between 16 million and 69.4 million SMS per month and have approximately between 4.5 and 5.1 million customers using this service. The impact was tested through personal interviews. The two common factors were fraud reduction and customer retention. The two key challenges the banks identified are: i) Capacity/throughput with the mobile network operators; and ii) Getting internal systems and processes defined and working together for the alerts. The advantages identified are competitive positioning, customer interaction, empowerment of people and revenue. Key findings of the research were: SMS-based transactional alerts offer strategic importance; Any system is as good as it is being utilised; Security controls are extremely complex; SMS capacity is a common challenge and big risk; Internal processes cause the most complexity; Return on investment is not adequately measured; Transactional alerts is a potential revenue stream; There is no interaction between the bank and the customer; SMS in South Africa create high dependencies; SMS-based transactional alerts are successful.