Show simple item record

dc.contributor.authorMaviti, Michael K
dc.date.accessioned2018-10-17T12:10:18Z
dc.date.available2018-10-17T12:10:18Z
dc.date.issued2018
dc.identifier.urihttp://hdl.handle.net/11295/104086
dc.description.abstractMobile Money Transfer (MMT) services provided by Mobile Network Operators (MNO) enable funds transfers made on mobile phones of end-users using digital equivalent of cash (float) without involving the bank(Jürgen Repp∗, Roland Riek 2014Fraunhofer Institute SIT, Darmstadt, Germany). Mobile money fraud is therefore the illegal implant in the electronic financial chain where the perpetrator applies deceit in order to extort money electronically from the holders. While MNOsin Kenya have done substantial work to advance fraud detection and prevention mechanism, incidents of successful losses due conning and defrauding are still high and affect the mobile money agents on a daily basis. M-Pesa from Safaricom Ltd is the dominant MMT platform in Kenya with a market niche of over 80%.With advancement in technology and especially digital systems capable behavioral analysis the fraudsters have equally developed ways which operate above the systems thus the resulting loses to mobile money agents are huge. The focus of this research is to evaluate the probability that an attempt by fraudster to penetrate the Mobile Money chain players will be successful and that the agent will lose money thus affecting the business operation. This penetration is not ICT driven rather a social engineering of the human interface within the chain by false pretence in various forms. Having identified eminent failure of the controls which are in place we seek to offset the risk to insurance and determine the premiums which the agents would be needed to pay to get covered. This research focused on risk profiling of M-Pesa agents in Murang’a and Nyeri in order to evaluate the probability that an attempt by authors of fraud will successful and lead to loss of money. The research undertakes a survey on agents operating in the business region of Murang’a and Nyeri. The region is further is further subdivided in to sub-region as per Safaricom regional business model. Risk profiling data was collected data through a questionnaire examining various parts of the MMT chain. While the Mpesa agent faces other risks in business, fraud is one of the major implicit risk facing MPESA agents. The general objective of this research is to determine actuarially fair premium without loadings which an insurance carrier would charge to cover MPESA agents against the risk of loss of money through fraud and conning. Some of the key findings were as follows;  Out of 490 respondents 379 were female while 111 were male. This translates to 77% and 23 % respectively. This bias may be explained by the fact that most agents prefer to employ ladies to run their shops.  The age between 18-40 years accounted for the 88%. This implies that the day to vi day business of M-pesa is being run by the digital age  Over 65% of the M-Pesa shops which participated in the survey have existed for at least over 1 year and above.  Out of the 490 M-Pesa shops 53% have had an attempted fraud.  Out of the 128 successful attempts, 80% were female while 20% were male. Murang’a, Sabasaba and Upper Nyeri contributed the biggest number of victims.  Midpoints of the loss cohorts were used to come up with the frequency tables above. The expected responses were a total of 128 but 132 responded. This implies 6 agents who may not have been defrauded gave false responses. Amounts lost between Kshs 5,000 – 60,000.00 contributed to 75% of the losses and thus the likely claims to the insurance company.  Training and learning from similar experiences plays a key role in fraud prevention. Though the data has an error of 2 points 36% of the unsuccessful cases were due to training and previous experiences.  69% of the participants on whom fraud was successful did not respond to this question. However instructions issued to agents via a phone call are the most commonly applied methods fraudsters.  Out of 128 cases of fraud only 23 cases (33%) were reported to the police. 67 % of the participants went silent to the question for various reasons.  Out of the 128 cases only 8% of the cases recorded some action by the police but have never been concluded.  Out the 490 participants only 23% have a form of insurance cover.  46% of the participants expressed the desire for insurance to cover against fraud. We adopted the Shapiro Wilks W Test of Normality and establish that the variables under consideration were not normally distributed. The research established that fraud within M-Pesa agents is an insurable risk since the risk profile for each individual can be uniquely determined and hence used to determine the premiums payable to cover against fraud.en_US
dc.language.isoenen_US
dc.publisherUniversity of Nairobien_US
dc.titleDetermination of premiums payable by Mpesa agents to cover against frauden_US
dc.typeThesisen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record