IFC ADVISORY SERVICES | ACCESS TO FINANCE 94737 IFC Mobile Money Study 2011 SRI LANKA In Partnership with the Republic of Korea IFC ADVISORY SERVICES | ACCESS TO FINANCE IFC Mobile Money Study 2011 SRI LANKA In Partnership with the Republic of Korea International Finance Corporation 2011. All rights reserved. 2121 Pennsylvania Avenue, N.W. Washington, DC 20433 Internet: www.ifc.org The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. IFC encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly, and when the reproduction is for educational and non-commercial purposes, without a fee, subject to such attributions and notices as we may reasonably require. IFC does not guarantee the accuracy, reliability or completeness of the content included in this work, or for the conclusions or judgments described herein, and accepts no responsibility or liability for any omissions or errors (including, without limitation, typographical errors and technical errors) in the content whatsoever or for reliance thereon. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the govern- ments they represent. The contents of this work are intended for general informational purposes only and are not intended to con- stitute legal, securities, or investment advice, an opinion regarding the appropriateness of any investment, or a solicitation of any type. IFC or its affiliates may have an investment in, provide other advice or services to, or otherwise have a financial interest in, certain of the companies and parties (including named herein. All other queries on rights and licenses, including subsidiary rights, should be addressed to IFC’s Corporate Relations Department, 2121 Pennsylvania Avenue, N.W., Washington, D.C. 20433. International Finance Corporation is an international organization established by Articles of Agreement among its member countries, and a member of the World Bank Group. All names, logos and trademarks are the property of IFC and you may not use any of such materials for any purpose without the express written consent of IFC. Additionally, “International Finance Corporation” and “IFC” are registered trademarks of IFC and are protected under international law. Contents Foreword..................................................... ix 6. Conclusion..............................................32 Acknowledgments...................................... xi Appendixes Abbreviations............................................ xiii A. Fact Sheet and Demand Estimates................... 33 Summary.......................................................1 B. Persons Interviewed......................................... 37 References..................................................38 1. Introduction..............................................4 2. Demand Perspective.................................5 Box Bill Payments (Utilities).......................................... 6 3.1 Developments in Mobile Money Person-to-Person Transfers...................................... 6 Regulations.................................................13 Formal versus Informal Transfers and Loans........... 7 Figures Government-to-Person Payments...........................7 Public Transport.....................................................7 2.1 Potential Monthly Transactions in Key Other.....................................................................8 Mobile Money Market Segments in Sri Lanka......................................................5 3. Parameters of the Mobile Money 2.2 Percentage of Overseas Sri Lankans Ecosystem....................................................11 Reporting Various Means of Remittance Enabling Regulation.............................................11 Transfer........................................................ 6 Existing Access to Financial Services..................... 14 2.3 Percentage of Loans Issued by Formal and Existing Mobile Access and Market Situation....... 16 Informal Institutions.................................... 7 2.4 Degree of Commuter Preference for 4. User Survey Findings..............................19 Mobile Phone–Based Ticket Sale and Survey Design......................................................19 Purchase System........................................... 8 Respondent Identification ...................................19 3.1 Sri Lanka’s Mobile Money Market in the Socioeconomic Profile of Respondents.................20 Porteous Regulatory Environment Model.... 11 Profile of Mobile Money Use................................22 3.2 Poverty Rates and Bank Account General Use of Money.......................................... 23 Penetration in Selected South Asian Marketing and Advertising...................................25 Countries....................................................14 Attitudes and Perceptions.....................................25 3.3 Ratio of Average Bank Account Size to Conclusion ..........................................................26 Income in Selected South Asian Countries.. 15 3.4 Credit and Debit Cards: Number of Cards 5. Business Models......................................28 and Value of Transactions...........................17 Existing Business Model: Dialog .........................28 3.5 Subscriber Market Shares of Sri Lanka’s Other Business Models......................................... 28 Mobile Operators....................................... 17 Challenges............................................................30 3.6 Sri Lanka’s 2G and 3G Coverage................. 17 Potential Business Opportunities.......................... 30 3.7 Mobile Subscriber Growth and Mobile Risks ...................................................................30 Penetration................................................. 18 vii  viii  IFC Mobile Money Study 2011: Sri Lanka 4.1 Types of Bank Accounts Held.....................20 4.16 Perceived Mobile Money Benefits............... 26 4.2 Mobile Phone Access to GPRS/Mobile 4.17 User Reasons for Mobile Money Usage....... 26 Internet......................................................20 4.3 Socioeconomic Characteristics of Mobile Tables Money Users and Nonusers .......................21 S.1 Mobile Money Opportunities in Sri Lanka....3 4.4 Mobile Banking Service Provider................22 1.1 Distribution of Poorest 40 Percent of Sri 4.5 Mobile Money Services Used......................22 Lanka’s Population by Province (%)..............4 4.6 Users’ Reported Level of Knowledge of 1.2 GDP per Capita in Various Southeast Mobile Banking Services.............................22 Asian Countries (US$).................................. 4 4.7 Typical Bill Payment Channels...................23 2.1 Potential Mobile Money Market Segments.....5 4.8 Knowledge of Bank Services Available 2.2 Formal Private Remittance Transfers through Branch..........................................23 from the Middle East, 2000–07 ................... 6 4.9 Respondents Reporting High Level of 2.3 Volume of Financial Services Transacted at Knowledge of Various Financial Services Post Offices...................................................9 and Channels..............................................24 3.1 Parameters Affecting the Success of 4.10 Cash Withdrawal Sources Used Most Mobile Money Services............................... 12 Frequently..................................................24 3.2 Access to Formal Financial Sector as a 4.11 Typical Cash Withdrawal Amounts............ 24 Percentage of Country Population.............. 14 4.12 Fund Transfer Destinations......................... 25 3.3 Number of ATMs at Top Four Banks by 4.13 Preferred Source of Information on Province...................................................... 16 Mobile Banking Services.............................25 4.1 Survey Locations ........................................19 4.14 Mobile Money Services in Which Users 5.1 Dialog Business Model...............................29 Express High Interest..................................26 A.1 Fact Sheet................................................... 33 4.15 User Perceptions of Mobile Banking Fees.... 26 A.2 Demand Estimates......................................35 Foreword F inancial inclusion—access to a range of Mobile technology is a channel that, once in financial services and products for every- place, allows for the delivery of other low-cost one needing them, in a fair, transparent, financial services bringing banking to unbanked and cost-effective manner—is a goal of IFC and underserved people. Mobile money—the (International Finance Corporation) and a prior- transfer of funds using cell phones—is an innova- ity of the Group of 20 development agenda. tive method for both individuals and small busi- nesses to transfer money. Mobile money is becom- IFC has committed to achieving greater financial ing common in developed countries for small, inclusion by 2013 by providing more diversified frequent payments such as mass transit fees. In financial services and by deepening outreach to some developing countries, it offers an opportu- microclients and small and medium enterprises. nity for unbanked people to pay bills and transfer IFC also helped support and shape the G20 global funds without using cash. Some businesses use it financial inclusion agenda that calls for the pro- throughout their supply chain. motion of a range of financial services beyond credit—including payments, savings, remittances, Why has the development of mobile money sys- and insurance. tems been so successful in some countries, yet seem blocked in others? What can be done to More than 2.7 billion people in developing coun- encourage its development globally? tries do not have access to basic formal financial services, such as savings and checking accounts. This report looks at the technology required and Many governments have made savings accounts the business models used by mobile network oper- widely available, but to make payments and trans- ators, banks, and others in four developing coun- fer funds, the poor must often depend on costly tries—Brazil, Nigeria, Sri Lanka, and Thailand. It and unreliable informal financial services. Low compares these countries with Kenya and Japan, levels of financial inclusion also represent an which have successfully developed mobile money obstacle to economic development. operations, and with the United States. Developing innovative methods of retail payments Perhaps more importantly, it offers a framework is essential to increasing financial inclusion. New for a quick market study of a country to determine technologies and new business models are open- whether or what type of mobile money services ing new methods of retail payments, as well as bill might be developed commercially. It offers models payments and transfers of funds among people of user perception and demand surveys, then and businesses. develops a set of parameters—such as regulatory ix  x  IFC Mobile Money Study 2011: Sri Lanka environments, current access to financial services, telecommunications equipment and handset and the requirements of potential mobile money manufacturers, and others that could be involved service providers to run viable businesses—that in the development of mobile money businesses. can spur or block mobile money development. By using these survey techniques and examining I would like to express sincere thanks to the gov- the relevant parameters, a government or develop- ernment of the Republic of Korea for its support ment agency can assess a country’s potential for a of this study through the Korean Trust Fund. successful mobile money business. We hope this report will contribute to mobile money business development globally. It is Peer Stein intended for regulators, mobile network opera- Global Business Line Leader tors, commercial banks, microfinance institutions, IFC Advisory Services, Access to Finance Acknowledgments T his study was commissioned to increase Dialog in Sri Lanka, and TrueMoney in Thailand. understanding of mobile money (m-money) Other organizations, companies, and individuals and help address key issues in scaling up in each country gave generously of their time and further development of m-money ecosys- knowledge, including the Central Bank of Brazil, tems globally. the Central Bank of Nigeria, the Central Bank of Sri Lanka, and the Bank of Thailand. Appen- First and foremost, we are grateful to the govern- dix B of each country report lists the many people ment of the Republic of Korea for its leadership interviewed during the study; their participation in the area of information and communications is greatly appreciated. technology for development, and for funding this study to promote the m-money agenda for the The following IFC and World Bank colleagues public benefit. in the respective countries provided local insights and liaison with the above-mentioned partner- Intelecon Research and Consultancy Ltd of Van- ing institutions, and helped the team conduct couver was contracted by IFC (International meetings and field surveys: Alexandre Darze and Finance Corporation) to conduct the IFC Mobile Terence Gallagher (Brazil), Theophilus Adewale Money Study 2011, including in-country field- Onadeko (Nigeria), Asela Tikiri Bandara Dis- work. Andrew Dymond, Steve Esselaar, and sanayake (Sri Lanka), and Frederico Gil Sander Sonja Oestmann authored the reports, assisted and Ratchada Anantavrasilpa (Thailand). by the rest of the Intelecon team. The team also included Jenny Hoffmann from RiskFrontier Several individuals within IFC, infoDev, the Consulting (United Kingdom) and local research World Bank, and the Consultative Group to Assist partners in each country: Antonio Bothelo of the Poor helped create this report, providing ser- Diálogo Regional sobre la Sociedad de la Infor- vices including Trust Fund administration, proj- mación (Brazil), Ike Moweto of Research ICT ect management, project design, expert advice, Africa! (Nigeria), Harsha de Silva of LIRNEasia peer review, administration of in-country surveys, (Sri Lanka), and Deunden Nikomborirak of Thai- coordination, printing, and public relations. land Development Research Institute (Thailand). We are grateful for the insightful inputs and peer We are also extremely grateful to our partner- reviews by Hemant Baijal, Deepak Bhatia, Mar- ing m-money operators for their cooperation: Oi garete Biallas, Massimo Cirasino, Andi Dervishi, Paggo in Brazil (a new company, Paggo Soluçoes, Janine Firpo, Soren Heitmann, Eriko Ishikawa, has since been formed), eTranzact in Nigeria, Nikunj Jinsi, Samuel Kamau Nganga, Tim Kelly, xi  xii  IFC Mobile Money Study 2011: Sri Lanka William Kerr-Smith, Yong Hyun Kwon, Samia D’Costa, Philippe Dongier, Gilles Galludec, Mat- Melhem, Harish Natarajan, John Irungu Ngahu, thew Gamser, Dianne Garama, Idawati Harson- Mark Pickens, Christine Zhen-Wei Qiang, gko, Oleh Khalayim, Sujata Lamba, Henna Lee, Wiebke Schloemer, Josef Skoldeberg, Hourn Thy, Kent E. Lupberger, Trang Nguyen, Marcia Roa, Michael Trucano, and Shinya Yoshino. Colin Shepherd, Peer Stein, Stephanie Von Frie- deburg, and Ann-Marie Webster. Mary Paden edited the text to make it very user- friendly. Nita Congress gave it a wonderful design. The project could not have been completed with- out the administrative and managerial support Arata Onoguchi, Leila Search, and Piya Baptista of Greta Bull, Catherine H. Burtonboy, Valerie IFC Mobile Money Study 2011 Project Team Abbreviations 2G second generation 3G third generation ATM automated teller machine B2B business to business e-money electronic money e-payment electronic payment e-wallet electronic wallet G2P government to person GDP gross domestic product GKCCC Golden Key Credit Card Company GPRS general packet radio service ICTA Information and Communication Technology Agency IFC International Finance Corporation KYC know-your-customer LSM living standard measure m-banking mobile banking m-money mobile money m-payment mobile payment MNO mobile network operator NDB Bank National Development Bank NFC near-field communication P2P person to person POS point of sale SIM subscriber identity module SMS short message service STK SIM Toolkit USSD unstructured supplementary services data The average exchange rate for the year 2010 of 113.06 Sri Lankan rupee/1 U.S. dollar is used throughout. xiii  Summary O f the four countries included in the IFC study in early 2011, new regulations released by (International Finance Corporation) the Central Bank of Sri Lanka allow electronic Mobile Money Study, Sri Lanka has money (e-money) and e-money accounts by the greatest potential for mobile money licensed service providers, provided they main- (m-money) expansion. However, existing initia- tain a custodial account at a licensed commer- tives have not been able to take advantage of the cial bank that agrees to perform the duties and m-money opportunities for three main reasons: responsibilities outlined in the legislation (see box 3.1). ƒƒ Public access to financial services is already good. Sri Lanka has high bank account pene- ƒƒ Mobile phone coverage is growing, but the tration, but low electronic payment access, such dominant player is losing ground. Mobile as debit and credit cards. The banking sector access is rapidly increasing outside of Colombo. is dominated, in terms of the mass market, by The sector is experiencing increasing competi- government-owned banks. The sector is fairly tion and pricing pressure. Dialog, the incum- inefficient, which partly explains the low auto- bent operator, is growing more slowly than its mated teller machine (ATM) and point-of-sale competitors. The major focus for investments (POS) device roll-out. As in other parts of the by the mobile network operators (MNOs) is world, the roll-out of ATMs and POS devices rolling out the 3G network and not m-money. will be a crucial future focus of Sri Lankan Despite these uncertainties, there are several banks. The country’s high bank account pen- potential opportunities for m-money in Sri Lanka. etration means that m-money is unlikely to be Table S.1 provides details on each opportunity. adopted as aggressively as in Kenya, for exam- ple, where financial sector penetration was very ƒƒ Government-to-person (G2P) payments. low. Instead, m-money will have to be an addi- A large government social welfare program tional, but more targeted, service. (Samurdhi) provides small payments to 1.6  million households. The current system ƒƒ The status of government regulation is uncer- is highly inefficient and could be improved tain. In the absence of m-money regulations, through an m-money system. the status of m-money has not been clear. For example, it has not been clear whether a bank ƒƒ Payroll (informal sector). Sri Lanka has a account is a precondition for m-money or if large informal sector. The post office maintains an m-money account can be opened without a substantial network of accounts; mainte- a bank account. After the conclusion of this nance of these accounts is mostly manual, with 1  2  IFC Mobile Money Study 2011: Sri Lanka no Internet access. Using the post office’s net- these opportunities comes with a set of key ele- work of branches to provide efficient payroll or ments that must be implemented for Dialog person-to-person (P2P) transfer services repre- to take advantage of them. These elements are sents a major opportunity. detailed in section 5. ƒƒ Business-to-business (B2B) payments. Some Table S.1 shows several market opportunities for retailers, such as Cargills, have already identi- m-money in Sri Lanka. The first is a mass market fied an opportunity in the agricultural supply strategy aimed at providing a safe and convenient chain for m-money services and are piloting an way to pay bills. At the moment, the middle and initiative with small farmers. upper segments of the market are the targets of ƒƒ International remittances. The Middle East commercial banks and their mobile banking region has the most Sri Lankan expatriates. (m-banking) products. But at the lower end of the In 2007, 58 percent of all international remit- market, there is an opportunity to provide a faster tances were from the Middle East, a value of and cheaper mechanism to pay bills. As an indica- US$1.4  billion or about 4.3  percent of gross tion of the size of this market, the Sri Lankan post domestic product (GDP). office currently handles over 6.5 million electricity bill payments per year. In addition, there is a major opportunity in public transport, where m-money could provide more Another mass market opportunity is in the public efficient services quickly. However, this requires transport sector where there are over 10  million near-field communications (NFC) technology, a daily commuters. Leakage from the current tick- substantial investment in card readers, and NFC- eting system is massive—between 15  percent enabled mobile phones or cards. and 25  percent.1 The challenge is to develop a common NFC phone standard (rather than a par- Based on the analysis of the m-money ecosystem ticular brand of phone, which then has to pen- parameters in conjunction with the potential mar- etrate the market from scratch), something that kets for m-money, a set of recommendations was has not yet been achieved anywhere in the world. generated. Market opportunities were found in Alternatively, marketing an m-money solution bill payments, P2P transfers, and public transport. that includes a debit card (like the existing eZ Pay system) that is also a smart card and that can be Dialog, Sri Lanka’s largest telecommunications used on the public transport system could capture provider with a 50 percent market share, launched a sizable market share. Such a smart card solu- its m-money venture, eZ Pay, in 2007 in collabo- tion breaks out of a closed loop—in the sense that ration with National Development Bank (NDB most smart cards used for public transport can Bank). At present, only a small number of transac- only be used for public transport—and represents tions go through the system, mainly because there a significant opportunity. is no clear value proposition to the consumer and minimal marketing. This report details the oppor- tunities available in the Sri Lankan market, some of which Dialog could realize. However, each of 1  See de Silva 2010a.  Summary 3  Table S.1  Mobile Money Opportunities in Sri Lanka Potential Potential transactions/ market Assessment Description Challenges and obstacles month Bill payments ƒƒ Low ATM and POS penetration ƒƒ Building an agent network to areas not 6,440,168  (utilities) ƒƒ High bank account penetration covered by banking infrastructure ƒƒ Many Sri Lankans queue to pay bills ƒƒ Large retailer, Cargills, offers utility payments at stores P2P transfers ƒƒ Rural population ƒƒ Mobile operators are short of money Unknown  ƒƒ High use of informal channels to invest ƒƒ Expensive current offerings ƒƒ No agent network ƒƒ Competition from financial sector G2P ƒƒ Large-scale government social welfare ƒƒ Amounts being paid are very small 1,600,000 payments program (Samurdhi) uses an inefficient,  paper-based system, which could be improved and made cheaper for the government using m-money Payroll ƒƒ Relatively large informal sector ƒƒ Market the service to people who 4,708,418 (informal sector)  currently use money orders from the post office Public transport  ƒƒ Large-scale opportunity with clear value proposition to replace existing system ƒƒ Needs NFC to succeed, requiring investment 264,000,000 B2B payments ƒƒ Growing opportunity for some large ƒƒ Most businesses have little knowledge —  retailers about m-money ƒƒ Cargills is offering utility payments at point of sale International ƒƒ The Middle East region has the most Sri ƒƒ Most workers deposit pay directly into — remittances Lankan expatriates: in 2007, 58% of all family bank accounts at home  international remittances were from the ƒƒ Well-established informal black market Middle East, a value of US$1.4 billion system (hawala) in place ƒƒ Up to 85% of international workers carry or send money home as cash Credit and ƒƒ User survey respondents showed an ƒƒ Collaboration with government will be — microfinance interest in m-money microloans, with a challenge approximately 30% suggesting the ƒƒ Credit market is dominated by need Samurdhi Bank Societies, a very  inefficient system: nearly 65% of microcredit—including loans for consumption, income subsistence, and microenterprise start-up capital—is provided through these government banks Source: IFC Mobile Money Study 2011. Note:  = significant and unrealized opportunity for m-money: many of the preconditions for m-money exist, such as demand, supportive regulation, and an identifiable group of customers;  = potential opportunity but there are substantial challenges; — = not available. 1 Introduction S ri Lanka’s population is still largely rural— Its GDP per capita places Sri Lanka near the nearly 85 percent lives outside of cities. There average of comparable Southeast Asian countries will probably be rural-to-urban migration (table  1.2). Malaysia is clearly an outlier with a in the future, which represents a potential considerably higher GDP per capita, but Sri Lan- opportunity to m-money providers. People work- ka’s GDP is higher than that of the Philippines, ing in cities often wish to repatriate their savings to where m-money has taken off dramatically. Pov- their rural families conveniently and at a low cost. erty is less of a problem in Sri Lanka relative to countries like Bangladesh or Cambodia, where Income is fairly evenly spread across Sri Lanka’s GDP per capita is much lower. provinces, with the exception of the Western Prov- ince where Colombo, the largest city, is situated. The key point is that Sri Lanka is at a different This province is richer than the others. Table 1.1 stage in its economic development and is unlikely shows the percentage of the poorest 40 percent of to have the same socioeconomic conditions that the country’s population living in each province. made m-money in Kenya accelerate so rapidly. Table 1.1  Distribution of Poorest 40 Percent of Table 1.2  GDP per Capita in Various Southeast Sri Lanka’s Population by Province (%) Asian Countries (US$) Province Poorest 40% Country GDP per capita Central 13 Bangladesh 574 Eastern 15 Cambodia 775 North Central 14 India 1,031 North Western 14 Philippines 1,746 Sabaragamuwa 14 Sri Lanka 2,041 Southern 15 Indonesia 2,329 Uva 14 Thailand 3,940 Western 9 Malaysia 6,897 Source: Department of Census and Statistics n.d. Source: IMF 2009. 4  Demand Perspective 2 T he potential market segments for m-money described in table  2.1 were investigated. Table 2.1  Potential Mobile Money Market Segments Where appropriate and possible, additional potential applications were also examined. Market Figure 2.1 shows estimates of total monthly vol- segment Description umes (not value) of transactions in five of the Bill In developing economies, it is common to pay potential market segments. Because m-money payments bills by queuing outside the utility company. must compete with both traditional payment (utilities) Although this may be a niche market, the value methods and other e-money options, it is unlikely proposition is to provide a convenient, safe, and fast mechanism to pay bills. to be able to capture all of this potential. P2P The success of Kenya’s M-PESA indicates that Depending on data availability, the size of each of transfers there is a large unmet demand in transferring the demand markets was estimated to establish the money between people. relative size of the m-money opportunity. G2P Governments make regular payments to at least payments 170 million poor people worldwide.a The value proposition is to provide a more cost-effective and time-saving service to citizens. Payroll This segment might overlap with the P2P (informal market, but is a more specific opportunity for an Figure 2.1  Potential Monthly Transactions in sector) m-money application allowing small businesses Key Mobile Money Market Segments in Sri in the informal sector to pay their staff. Lanka Public The success of NFC technology in Japan indicates Millions transport that there is potentially a massive market, 300 particularly for NFC-enabled phones. 264,000,000 B2B B2B payments in rural areas beyond the reach of 200 payments banks are difficult and handled mainly by cash or check. M-money could provide mobile payment capabilities at each stage along the value chain. 100 Retail Cash is less secure than e-money. Consumers may 6,440,168 Unknown 1,600,000 4,708,418 payments find paying with an NFC-enabled card or phone 0 Bill P2P G2P Payroll Public more secure and more convenient than using cash. payments transfers payments (informal transport (utilities) sector) Source: IFC Mobile Money Study 2011. Source: IFC Mobile Money Study 2011. a. Pickens, Porteous, and Rotman 2009. 5  6  IFC Mobile Money Study 2011: Sri Lanka Bill Payments (Utilities) Table 2.2  Formal Private Remittance Transfers In Sri Lanka, electronic bill payments are a bur- from the Middle East, 2000–07 geoning industry. Existing bill payment mecha- Value As % of all nisms are largely aimed at higher-income groups remittances to (LSM 6 and above).1 For example, many com- Year Million SL Rs US$ Sri Lanka mercial banks offer m-banking services (including 2000 55,252 489 63.0 some bill payments), but none of the commercial banks interviewed plan to expand this offering 2001 62,680 554 60.8 beyond their customer base. However, Cargills, 2002 75,579 668 61.4 a convenience store network, currently takes pay- 2003 77,579 686 56.8 ment for utility bills for a small fee at its conve- nience stores: for paying a water bill, it charges 2004 87,871 777 55.5 0.2 percent of the value of the bill; for all other bill 2005 111,179 983 56.2 payments, it charges SL Rs 15 (US$0.13). 2006 128,282 1,135 57.1 2007 160,502 1,420 58.0 Person-to-Person Transfers Source: Central Bank of Sri Lanka annual reports. Given that more than 75 percent of Sri Lankans have bank accounts, it is not surprising that a sig- nificant percentage use traditional bank accounts to transfer money. The Middle East represents the Figure 2.2  Percentage of Overseas Sri Lankans Reporting Various Means of Remittance region with the most Sri Lankan expatriates. In Transfer 2007, 58 percent of all international remittances were from the Middle East, a value of US$1.4 bil- Deposit into recipient bank accounts lion or 4.3 percent of GDP. This volume of money Take money with me when I go home transfer represents a major potential opportunity Send cash with a friend or relative for m-money and was identified as a potential for Money transfer service (e.g., Western Union) m-money by survey respondents. Purchase goods & send through others Postal money order/postal mail Table 2.2 shows that official remittances coming Hawala into Sri Lanka (not counting the monies being Other (including Internet) transferred using the informal hawala system) Mobile payment systems grew threefold from 2000 to 2007. 0 20 40 60 80 100 Percentage of Figure  2.2 shows that of the Sri Lankans living respondents overseas, nearly 50 percent still take money home Source: LIRNEasia 2009. physically, and 40 percent use friends and relatives Note: Multiple selections were allowed. to deliver money. Domestic workers tend to trans- fer money to family using bank tellers (63  per- cent) or ATMs (14 percent), as indicated in user In a recent survey by LIRNEasia, a Sri Lankan survey responses. Multiple methods are used to think tank, demand for domestic remittances transfer money. was estimated for both Thailand and Sri Lanka. Given Thailand’s higher financial services penetra- tion (bank accounts, ATMs, etc.), it is not sur- 1  Living standard measures (LSMs) are segmentation prising that there is a substantial contrast between tools used in consumer marketing as a wealth proxy, calculated on ownership of household goods/assets and the amount of Sri Lankan domestic migrants who the degree of urbanization. The LSM categories range physically take money home—85 percent—com- from 1 (very poor and rural) to 10 (wealthy and urban). pared with only 22 percent in Thailand. 2. Demand Perspectives  7  Formal versus Informal To modernize the system (i.e., increase efficiency), Transfers and Loans the eSamurdhi Project has been launched by the Information and Communication Technology Getting accurate statistics on formal versus infor- Agency (ICTA) of Sri Lanka. mal money flows is difficult. The Central Bank of Sri Lanka conducts a consumer finance and Information and Communication socioeconomic survey about every five years. The Technology Agency 2003–04 survey2 shows that informal loans have been declining since the mid-1980s compared ICTA is the single high-level coordinator involved with loans from the formal sector (figure  2.3). in information and communication technology Even so, the informal sector is large, with about policy and e-government in Sri Lanka. It is wholly 39 percent of loan origination. If the movement owned by the government of Sri Lanka and is the of money outside the country using the hawala implementing organization of the e-Sri Lanka Ini- system is added, it is even larger. tiative, the country’s overarching electronic devel- opment project that harnesses information and communication technologies to achieve socio- Figure 2.3  Percentage of Loans Issued by Formal and Informal Institutions economic development. Major donors include the World Bank. Percentage of loans 100 ICTA’s focus is on the interoperability among different government services and organizations. 80 It has created a gateway or portal for electronic 60 Informal information and electronic interactions with gov- ernment, generally referred to as the Lanka Gate 40 initiative. 20 Formal Lanka Gate is a messaging platform and portal 0 that provides a short code for all government- 1978–79 1981–82 1986–87 1996–97 2003–04 based information services, such as railway time- Source: Central Bank of Sri Lanka, Consumer Finance and Socioeconomic tables. The vision is to integrate all the banks and Surveys. government departments, as well as include a mobile payment platform linking to the national Government-to-Person central switch. Payments ICTA is working on several projects, including piloting a credit card payment system as part of Samurdhi the online vehicle license registration process. Samurdhi is Sri Lanka’s main poverty alleviation program; it distributes grants and microloans. It is a Public Transport government program introduced in 1994 and estab- lished by the Samurdhi Authority of Sri Lanka Act Sri Lanka’s massive public transport system offers of 1995. Samurdhi distributes monthly payments an opportunity for a faster, more reliable payment of SL  Rs  400–1,000 (about US$3.50–US$8.80), system. Because there is little rail infrastructure, to 1.6 million families. For its loans, Samurdhi uses buses are the primary means of transportation. a passbook system; repayments are made at Samur- According to LIRNEasia (2010), 10 million com- dhi offices or collected on the doorstep. muters travel daily on 18,000 buses. The current ticketing system is mainly manual No survey had been released for 2008–09 as of this 2  and operated by both private and public compa- writing. nies. Public (government-owned) companies lose 8  IFC Mobile Money Study 2011: Sri Lanka approximately 15 percent of the fare in transit, and appointments. Using eChanneling, a mobile sub- private companies lose approximately 25 percent in scriber (all three major mobile providers offer this transit. A portion of the leakage is due to fraud, but service) phones the eChanneling call center and a considerable amount is due to the length of the books an appointment with the relevant doctor. manual payment process, whereby conductors are Doctors are allied to a particular hospital. The not able to complete all the transactions necessary hospital and doctor’s consulting fee, an eChannel- in the time available. A large percentage of people ing fee, and a small transaction fee for the MNO surveyed indicated that transport was one of their are deducted from the subscriber’s airtime balance common expenses. A fairly high percentage of com- (prepaid or postpaid). The MNOs charge a pre- muters (37 percent) would prefer a mobile phone– mium rate for the phone call, but nearly the entire based ticketing system (figure 2.4). The success of fee goes to eChanneling and the hospital and/or eChanneling, in which people can make and pay doctor. eChanneling charges a transaction fee of for doctor appointments over their mobile phone, SL Rs 120 (about US$1) for a SL Rs 1,000 (about means that consumers already have some comfort US$9) medical bill (the average size of an eChan- with using mobile phones to make purchases. neling bill). In our user survey, many m-money users knew Figure 2.4  Degree of Commuter Preference for about or used eChanneling, but only a few non- Mobile Phone–Based Ticket Sale and Purchase users did. eChanneling’s primary target market is System Sri Lanka’s 500,000 upper-class families, and its secondary market is the middle class. The program averages 3,000 transactions per day on its mobile Prefer Do not prefer at all platform, and 5,000 more transactions occur via very much 37% 14% the Internet. Do not Microfinance and Microinsurance prefer much 9% Microfinance Prefer somewhat The microfinance sector in Sri Lanka is dominated 13% Indifferent by Samurdhi Bank Societies. Nearly 65 percent of 27% microcredit, including loans for consumption, income subsistence, and microenterprise start- Source: de Silva 2010b. up capital, is through the government’s Samur- dhi Program. Of these loans, about 20  percent are less than SL Rs 20,000 (US$177) and almost To combat the fare leakage, two solutions are being 65 percent are less than SL Rs 50,000 (US$442) considered: first, a smart card using an embed- (SAMN n.d.). In the user survey, about 30  per- ded chip, and second, a mobile phone equipped cent of respondents expressed a need for m-money with NFC technology so payment can be made by services to provide microloans. swiping over a receiver. Microinsurance Other The microinsurance industry reaches 1.46  per- cent of the population and is offered by a lim- Business-to-Business or Business- ited number of institutions. The target market to-Employee Transfers for insurance products is the middle and upper eChanneling is a unique product in Sri Lanka; it classes, and nearly all companies have avoided tar- is a system that semi-automates booking doctor geting the lower classes (ADB 2010). 2. Demand Perspectives  9  Retail Sector: Cargills Post Office Cargills Food City (part of the Cargills Ceylon The Sri Lankan post office has approximately 4,500 Group) has 140 outlets across 19 districts, con- outlets, of which 160 are computerized; it is cur- sisting of Cargills Food City supermarkets, Car- rently undergoing a major computerization drive. gills Express convenience stores, and Cargills “Big The remaining post offices operate manually. Post City” super stores. offices exist in all districts and provinces in Sri Lanka. Cargills’s vision is for each store to be a “conve- Besides delivering the mail, post offices maintain nience center” for all of a family’s needs. As such, accounts for government pensions for 811,194 Cargills offers payment services for several utility former employees, and accept 6.5  million elec- bills through its Food City network. For a water tric bill payments from nearly one-third of all Sri bill, it charges 0.2 percent of the value of the bill. Lankans (table 2.3). For all other bill payments, it charges SL  Rs  15 (US$0.13) Table 2.3  Volume of Financial Services To supply its Food City stores with fresh pro- Transacted at Post Offices duce, Cargills has set up a supply network of Government about 10,000 farmers. To ensure quality, Cargills pension accounts Electricity bills pays the farmers approximately 20 percent more Province maintained paid than they would receive at the market. These pay- Central 95,034 1,085,527 ments are deposited directly into bank accounts— although in remote areas, cash is used. Eastern 7,554 289,805 North Central 33,484 344,178 To make up for the 20  percent premium, Car- gills has implemented a supply chain management North Western 101,949 1,962,122 system (e.g., using crates and not bags), which has Northern 0 288,591 cut down on waste by 40 percent. Cargills employs Sabaragamuwa 59,916 814,481 field officers who confirm the crops, times of deliv- ery, quantity, and quality by visiting farms at var- Southern 140,469 683,610 ious times of the year. By managing the farmers Uva 24,713 366,017 directly, Cargills reduces the need for intermedi- Western 348,075 667,565 aries, and thus further reduces the cost of supply. Total 811,194 6,501,896 Cargills has also started supplying farmers with Source: Sri Lanka Post Office. fertilizer and seeds as part of its supply manage- ment program at a wholesale (or even subsidized) cost; it also gives its supplying farmers short-term The government transfers funds into post office loans for capital equipment. accounts, where retirees can claim their funds with the proper identification. Post office employ- Cargills is discussing an m-money pilot project ees may even deliver the cash payments to retirees with Dialog to increase efficiency and convenience in remote areas as part of their rounds. for its entire network of farmers (eliminating the need for a special trip to a bank) and as an added Electricity bill payments data are placed on a com- convenience to farmers who live far from bank pact disc and uploaded daily from central branches branches. Of the respondents surveyed, 67  per- because most post offices do not have Internet or cent said they were familiar with closed-pay- intranet connectivity. ment systems such as prepaid cards and a similar mobile-based system has good chances for success- The main challenge in expansion is that the post ful application. office does not have a dedicated managerial role 10  IFC Mobile Money Study 2011: Sri Lanka to handle bill payments or expand the prod- office network, including HSBC Bank and the ucts it offers at its branches. New products and National Savings Bank, but these are mainly as refinements to products take a long time to work stand-alone operations, and infrastructure (such through the bureaucracy. Some banks, both gov- as computers) is provided by the bank. ernment and privately owned, are using the post Parameters of the Mobile Money Ecosystem 3 A set of parameters that affect the roll-out of an m-money system have been identi- Figure 3.1  Sri Lanka’s Mobile Money Market in the Porteous Regulatory Environment Model fied through a review of existing litera- ture and refined during the field visits. High Table 3.1 on the next page provides an overview 2 1 of the parameters selected. The following sections Low certainty; High certainty; provide an analysis of relevant parameters in Sri high openness high openness Lanka. OPENNESS Enabling Regulation 4 3 Summary Assessment Low certainty; High certainty; low openness low openness There is some confusion about the regulatory Sri Lanka environment for m-money in Sri Lanka. For example, the chief executive officers of the two Low main mobile operators believed that only a bank- Low CERTAINTY High led model of m-money was allowed by the Central Source: IFC Mobile Money Study 2011, based on Porteous 2006. Bank of Sri Lanka; in fact, the regulator indicated that a telecommunications company–led model is allowed. The Central Bank recognizes the need for greater certainty, and specific m-money regula- Two Schemes Approved by the tions are pending (see box 3.1 for update). Never- Central Bank of Sri Lanka theless, they are accepting applications for licenses The Central Bank’s initial preference was that an to operate m-money businesses. The lack of for- “e-wallet” be linked to a bank account. After dis- mality, however, contributes to a lack of certainty. cussions with Dialog revealed that people did not Since the openness to new entrants on the part necessarily want to open a bank account, a second of the Central Bank has not translated into actual scheme was proposed. The second scheme issues operational businesses, apart from Dialog, it is dif- an e-wallet that must be mirrored by funds in a ficult to claim that there is a high level of open- custodian account, which is held by the mobile ness. Consequently, Sri Lanka is assessed between operator and can bear interest for the mobile oper- Positions 3 and 4 in figure 3.1. ator. Under this scheme, the Central Bank must 11  12  IFC Mobile Money Study 2011: Sri Lanka Table 3.1  Parameters Affecting the Success of Mobile Money Services Category Parameters Socioeconomic Population GDP/capita Geographic area Remittance flow context Poverty GDP by region Urbanization; rural Gini coefficienta population Regulation Clear and risk-based Know-your-customer Agent regulation ID system regulatory framework regulation Interoperability Pricing restrictions on M-money license Bank outsourcing requirements accounts requirements Mandatory services banks Regulations on new Level of expensive Obstacles to international must offer branches requirements remittances Existing access Reach of networks/agents Penetration/use of cards Penetration/use of prepaid Internet banking usage to financial Informal financial access Nonbank provision of cards Unbanked population services financial services Cash-electronic transaction Competitiveness of banking industry ratio (use of cash) Existing mobile Population penetration/ Geographical coverage Level of competition 3G penetration/usage market situation coverage Level of fragmentation of Churnb industry Potential Bill payments Public transport P2P transfers G2P payments demand B2B transfers Credit and microcredit International remittances Savings Retail payments Retail sector Retailers with national Level of fragmentation Postal network Other distribution networks coverage Payment system POS terminal penetration Mass payment acceptance Card penetration National switchc Dominant payment Third-party payment methods in the economy processors Pricing Distortion through Banking services pricing intervention/regulation User perceptions Trust in mobile operators Willingness to pay for Cultural factors versus banks m-money service Sources: IFC Mobile Money Study 2011; CGAP. a. The Gini coefficient is a measure of the inequality of a distribution, with a value of 0 expressing total equality and a value of 1 maximal inequality. b. “Churn” in the telecommunications industry means customers move from one network operator to another. c. “National switch” here means an online interbank fund transfer system. approve the mobile company as an agent, and a Multiple Types of Deposit-Taking bank must be able to provide an audit trail for the Institutions custodian account. There are several types of financial institutions in Regulations in draft form were sent out for com- Sri Lanka, authorized by disparate acts that govern ment with the deadline of October 18, 2010 (see the sector, including the Samurdhi Authority of box 3.1). Sri Lanka Act of 1995, the Agrarian Services Act 3. Parameters of the Mobile Money Ecosystem  13  Box 3.1  Developments in Mobile Money Regulations In March 2011, the Central Bank of Sri Lanka approved two guidelines—one regarding bank-led m-payment services, and the other custodian account–based m-payment services—enabling payment schemes under the leadership of either a bank or a telecommunications company. Under the bank-led scheme, licensed commercial banks can offer three types of m-payment services to their account holders: ƒƒ Basic—provides account information, but not fund transfers ƒƒ Standard—basic services plus fund transfers including bill payments, fund transfers, and third-party transfers ƒƒ Extended—basic and standard services, plus the ability to deposit and withdraw though agents of respective banks Banks that are not licensed commercial banks can provide only the basic type of m-money services. Under the guidelines for the telecommunications company–led scheme, a telecommunications company can provide e-money services— specifically, an e-wallet that must be mirrored by funds in a custodian account, held by and earning interest for a mobile operator at a licensed commercial bank. The Central Bank of Sri Lanka must approve the mobile company as a licensed service provider (under the guidelines for service providers of payment cards), and the custodian bank must agree to perform a set of duties for account operations, including audits, reporting, and formulating know-your-customer and customer-due-diligence procedures. Licensed service providers (mobile operators) can appoint merchants to perform authorized functions such as cash-in/cash-out. M-payment services must only be in Sri Lankan rupees and can only be used for domestic transactions. Sources: Central Bank of Sri Lanka 2011a, 2011b. of 1979, the Banking Act of 1988, and the Bank- require the following (Central Bank of Sri Lanka ing Amendment Act of 2006. 2009a): The many different types of financial institutions ƒƒ Customer’s name from an original document make consistent regulation difficult. For example, issued by an official authority an institution accepting deposits under the Samur- ƒƒ Customer’s permanent mailing address and dhi Act is not regulated by the Central Bank. Con- supporting evidence confirmed through corre- sequently, the various acts are being harmonized. spondence For future license applications, and to ensure uni- formity of regulation, the Central Bank of Sri ƒƒ Authenticity and integrity of the identity docu- Lanka has set up the Inter-Regulatory Council so ments confirmed by the bank that any new deposit-taking institutions are autho- ƒƒ Independent verification of introducer’s rized and regulated by the Central Bank. address. Know-Your-Customer Regulation Agents A national identification card exists in Sri Lanka. Sri Lanka has been one of the most liberalized All Sri Lankans are required to apply for it on markets in South Asia in terms of agents’ abil- their 16th birthday and to carry it with them at all ity to work on behalf of banks. Even so, Dialog times. Sri Lanka has not adopted an incremental had to overcome a number of bureaucratic hur- know-your-customer (KYC) process. dles. To become an agent, it needed approval from M-money providers are required to comply with the board of the Central Bank of Sri Lanka. This the same KYC requirements as banks. The latest requirement has subsequently been relaxed for regulations from the Central Bank for KYC agents that have been audited by a bank to ensure 14  IFC Mobile Money Study 2011: Sri Lanka compliance with KYC anti-money-laundering one reason being the government banks’ mandate regulations. With the exception of bill payments to increase financial inclusion. through retail stores, agents have not taken advan- Sri Lanka has a substantially smaller econ- tage of the low barriers to entry. These financial omy than nearly all its neighbors. Whereas the services can be performed by agents in Sri Lanka regional median in South Asia is 318 accounts per (CGAP 2009): 1,000 adults, Sri Lanka scores 1,650 accounts per ƒƒ Private operators can provide financial services 1,000 adults (CGAP 2009)—more than 100 per- at post offices. cent penetration (figure  3.2). However, since many people have more than one account, this ƒƒ Banks can formally contract with companies as translates into an actual bank account penetration banking agents. of 59 percent (table 3.2). ƒƒ Agents can receive and forward applications to open accounts. Another measure of broad access to bank accounts is account size (figure  3.3). A relatively large ƒƒ Agents can open accounts on behalf of banks. ƒƒ Agents can receive payments for taxes, utilities, and the like. Figure 3.2  Poverty Rates and Bank Account ƒƒ Agents can accept funds for deposit to client Penetration in Selected South Asian Countries accounts. ƒƒ Agents can pay withdrawals from client — Number of deposit Share of population accounts/100 adults living on <$2/day accounts. 180 100 ƒƒ Agents can receive and forward loan requests. 160 140 80 ƒƒ Agents can collect loan payments on behalf of 120 banks. 60 100 Note that agents cannot evaluate credit and 80 40 approve loan requests on behalf of banks. 60 40 20 20 Conclusion 0 0 The m-money environment in Sri Lanka is uncer- Pakistan Bangladesh Philippines India Sri Lanka tain. On the one hand, the Central Bank is open to Source: CGAP 2009. m-money applications and prepared to modify its requirements based on feedback from licensees. On the other hand, the Central Bank has not released formal m-money guidelines, particularly regarding agents and their role in the m-money ecosystem. Table 3.2  Access to Formal Financial Sector as The uncertainty is likely to have an impact on the a Percentage of Country Population commitment of m-money providers. Country % of population Bangladesh 32 Existing Access to Financial India 48 Services Pakistan 60 Formal versus Informal Access Sri Lanka 59 Compared with other South Asian countries, Sri Source: World Bank 2009. Lanka scores well on its financial inclusiveness, 3. Parameters of the Mobile Money Ecosystem  15  constitute 48.2 percent of the assets of the bank- Figure 3.3  Ratio of Average Bank Account Size ing system. The 13 state banks licensed as com- to Income in Selected South Asian Countries mercial banks contributed 68.2  percent of the total growth in banking sector assets. Two of the Pakistan state commercial banks have a 41.2 percent share Bangladesh of the assets in the licensed commercial banks South Asia sector. State-owned licensed specialized banks represent nearly 84.5 percent of the assets of the India licensed specialized bank sector (Central Bank of Sri Lanka Sri Lanka 2009c). 0 0.5 1.0 1.5 2.0 2.5 3.0 Because of the dominance of government-owned Ratio banks, commercial banks target niche markets, Source: CGAP 2009. usually the middle- to upper-income segments. Commercial banks want to expand their range of services to their existing customer base rather deposit balance is indicative of the general level than expand their base by competing with the of income and the probability that banks are less government-owned banks in the lower-income likely to serve a broad market. In Sri Lanka, the market. ratio of account size to income is lower than com- parable South Asian countries. Banks in Sri Lanka In December 2008, the Central Bank of Sri Lanka serve a broader market than in other countries in placed a large private bank, Seylan Bank, under the region. state custody because its subsidiary, Golden Key Credit Card Company (GKCCC), went bank- rupt. GKCCC was a pyramid scheme. As a result Market Concentration of the GKCCC debacle, the Central Bank has The retail banking sector in Sri Lanka is domi- been cautious about nonbanks participating in nated by three government-owned banks: the the financial system and the risks of innovations Bank of Ceylon, the National Savings Bank, and such as m-money. Its focus has been mainly on the People’s Bank. All have a mandate to extend ensuring that the existing financial services sector their branch network to rural areas. They are is healthy. guaranteed by the government1 and perceived as being risk-free providers of savings accounts. The Banks government banks have almost exclusive access to large-scale retail savings as a cheap source Most of the banks interviewed perceived m-money as m-banking. All banks see m-money/m-bank- of funds allowing wide margins. This tends to ing as a niche market that is an additional chan- encourage inefficiency as well as allowing the nel for services for existing customers. The user banks to cross-subsidize an uneconomic branch survey indicated a general lack of awareness about infrastructure. m-money as well as of the distinctions between Sri Lanka has three categories of banks: licensed m-money and m-banking. commercial banks, licensed specialized banks, Nearly all the major banks have some kind of and regional development banks. The 21 state- m-banking bill payment scheme under con- owned banks that fall under these three licenses sideration. For example, the National Savings Bank hoped to begin a pilot program offering 1  The People’s Bank received four capital tranches from bill payments and introducing insurance pre- the government of Sri Lanka between 2004 and 2008 to meet its capital adequacy requirements of 10.5 per- mium collection using the mobile application cent. on the unstructured supplementary services data 16  IFC Mobile Money Study 2011: Sri Lanka (USSD) channel. They planned to use a third Point-of-Sale Devices party to manage relationships with the operators The situation with POS device roll-out is simi- to provide the channel. The Bank of Ceylon and lar. There are 24,977 POS devices in Sri Lanka, Nations Trust Bank also indicate they are looking representing a penetration of 117.2 per 100,000 at m-banking. inhabitants. Most POS devices are located around Colombo. This POS penetration is dramatically Automated Teller Machines lower than the average for developing coun- Sri Lankans have higher access to bank accounts tries—170 per 100,000 inhabitants; 2,088 per than their peers in neighboring countries. How- 100,000 is the average for developed countries. ever, their access to electronic accounts is lower, with only 1,876 ATMs across the country, most Credit and Debit Cards of them concentrated in the Colombo area. The The credit card market in Sri Lanka is nascent. In top four banks have 42 percent of all ATMs and the fourth quarter of 2009, there were more than 50 percent of these are based in the Western prov- 4.7  million transactions, representing a value of ince (table  3.3). User survey findings show that US$154 million and an average transaction value ATMs are highly popular for cash withdrawals, of US$36 (figure 3.4a). In contrast, the number but less used for bill payments and fund transfers. debit card transactions was just under 1  million Further, existing users of m-money services in the for a total value of US$5.7 million and an average survey also had a higher percentage of ATM use transaction value of US$4.5 (figure 3.4b). than nonusers. Credit cards are clearly focused on the upper end of the market, where people are able to make large- value transactions. Debit cards are targeted at the middle-income market. Even with the high bank account penetration, electronic payment penetra- tion is low with low debit card use, low credit card Table 3.3  Number of ATMs at Top Four Banks use, low POS penetration, and low ATM penetra- by Province tion. Nat’l Bank The user survey shows a lack of awareness and People’s % of all Bank of Hatton Ceylon Seylan financial literacy with respect to POS, debit, and ATMs Bank Bank Province credit cards. Central 34 10 27 14 11 The lack of electronic payment mechanisms rep- Eastern 23 4 14 3 6 resents a significant opportunity for m-money. Rolling out a cost-effective, secure, and accessi- North Central 18 4 8 7 5 ble payment mechanism would offer substantial North Western 26 5 14 7 7 advantages to the average Sri Lankan. Northern 9 5 10 5 4 Sabaragamuwa 17 7 15 8 6 Existing Mobile Access and Southern 36 4 19 12 9 Market Situation Uva 15 2 7 6 4 The mobile phone sector in Sri Lanka is quite competitive, with five mobile operators. Dialog, Western 101 58 159 72 50 Etisalat, and Mobitel are the main operators. Total 279 99 273 134 100 Airtel (owned by Bharti India) is increasing its Source: Commercial banks’ Web sites. subscriber base rapidly (though its market share currently remains low) (figure 3.5). 3. Parameters of the Mobile Money Ecosystem  17  Figure 3.4  Credit and Debit Cards: Number of Cards and Value of Transactions a. Credit cards b. Debit cards Value of transactions — Number of cards Value of transactions — Number(millions) of cards (thousand US$) (thousands) (thousand US$) 940 164 35 920 5.7 162 30 900 160 25 5.5 880 20 158 5.3 860 156 15 840 5.1 820 154 10 152 5 4.9 800 780 150 0 4.7 4th quarter 4th quarter 4th quarter 4th quarter 2008 2009 2008 2009 Source: Central Bank of Sri Lanka 2009. Figure 3.5  Subscriber Market Shares of Sri Figure 3.6  Sri Lanka’s 2G and 3G Coverage Lanka’s Mobile Operators Dialog Hutchison 44% Telecom 5% Bharti Airtel 12% Mobitel 23% Etisalal 16% Source: Dialog 2009. Network 3G services have been aggressively rolled out in Sri Lanka. Dialog has the largest 3G net- work and significant investments are being made by competing operators, such as Etisalat. Figure  3.6 shows 2G and 3G coverage in Sri Lanka. As the map indicates, 3G is concentrated in the western part of the country, specifically around Colombo. Outside of 3G, Sri Lanka continues to experience Source: Dialog 2009. high mobile subscriber growth. The compound 18  IFC Mobile Money Study 2011: Sri Lanka annual growth rate of mobile subscribers between 2004 and 2009 was 45 percent (figure 3.7). How- Figure 3.7  Mobile Subscriber Growth and Mobile Penetration ever, Dialog’s compound annual growth rate for the same period was only 32  percent (Dialog 2009), indicating the level of competition the Total mobile — Mobile penetration rate subscribers (millions) largest company is facing. Although the level of 16 80 market concentration is still high, as the number 14 70 12 60 of subscribers increases and new entrants such as 10 50 Etisalat make inroads, it should decline. As com- 8 40 petition continues, mobile operators face tight 6 30 4 20 margins as they compete for subscribers. 2 10 0 0 2004 2005 2006 2007 2008 2009 Source: Dialog 2009. User Survey Findings 4 T his chapter summarizes the results of the Respondent Identification survey of a small sample of Sri Lankan citi- zens in urban and semi-urban areas on the In-country activities involved close collaboration use and potential of m-money in general with NDB Bank to identify users and agents of and specifically Dialog’s eZ Pay product. No agent the system, as well as appropriate locations for survey was done in Sri Lanka because the number the study. Efforts were made to select respondents of subscribers using eZ Pay is small and few out- randomly, but also to include people from semi- lets see any customers at all. urban areas and from diverse socioeconomic back- grounds and circumstances. Survey respondents The survey is not intended to be a statistically sig- were polled in the locations shown in table 4.1. nificant sample of m-money users and nonusers. Its purpose is to provide an overview of people’s Urban locations are areas in close proximity to attitudes, preferences, issues, and recommenda- services and economic activities and whose popu- tions regarding m-money services, including trust lation is generally economically well-off and edu- in mobile operators versus banks, willingness to cated, but which can include significant numbers pay, and other user factors. of poor inhabitants. Survey Design Table 4.1  Survey Locations The survey methodology involved the identifica- Urban Semi-urban/rural tion of localities and stakeholders to engage in the Ampara Anuradhapura suburbs survey, as well as criteria for design and sampling. Target populations included users of the eZ Pay Anuradhapura Badulla m-money service.1 Nonusers were not current Colombo Kurunegala suburbs users of eZ Pay or any other m-money system, but used both mobile phones and financial products. Galle Matara Gampaha Puttalama Kalutara Ratnapura Kandy Kurunegala 1  eZ Pay is an m-money service with an active (non- bank branch) agent network that supports transactions Source: IFC Mobile Money Study 2011. including payment, sign-up, cash-in/cash-out, etc. 19  20  IFC Mobile Money Study 2011: Sri Lanka Semi-urban/rural locations include smaller towns and cities outside metropolitan areas or Figure 4.1  Types of Bank Accounts Held suburbs, characterized by fewer services and Percentage of respondents whose economic emphasis may extend into rural 100 and agricultural areas outside the urban core. 80 A survey of 100 eZ Pay users and 100 nonusers were Users Nonusers conducted via face-to-face and phone interviews. 60 Approximately 10 percent additional people were 40 surveyed in each group to account for any survey errors and to ensure quality. 20 0 Socioeconomic Profile of Savings Current Micro- Welfare None of Other account account finance payment these Respondents account The surveys were conducted in all central suburbs/ Source: IFC Mobile Money Study 2011. districts in Colombo. Peri-urban areas included Note: Multiple selections were allowed. were Nugegoda, Maharagama, Dehiwala, and Mount Lavinia, which are 10–15 kilometers from central Colombo. Figure 4.2  Mobile Phone Access to GPRS/ Mobile Internet The main m-money service in Sri Lanka, eZ Pay, is provided by one of the major mobile operators, Percentage of respondents Dialog, in partnership with NDB Bank. There 100 Users Nonusers are currently only about 3,300 users. Transactions 80 started out relatively strongly at the beginning of 60 2009, but nearly ground to a halt by February 40 2010. No new marketing has been done since the launch. Staff at Dialog outlets were aware of the 20 product, but there was no marketing material vis- 0 ible to the customer. Yes No Source: IFC Mobile Money Study 2011. Figure  4.1 shows that more than 95  percent of both users and nonusers have at least a savings account because of the strong presence of the three Sri Lanka’s m-money users are predominantly government banks, which are subsidized and have young male white-collar workers: a clear mandate to provide bank accounts to all. However, current accounts, which are more trans- ƒƒ 79 percent are male action oriented and typically provided by com- ƒƒ More than 40 percent are 25 years or younger mercial banks, are mainly held by m-money users, (and more than 70  percent are 35 years or at approximately 25 percent. younger) Figure  4.2 shows that 88  percent of m-money ƒƒ 46 percent are college graduates users have mobile phones that are general packet ƒƒ Almost 70 percent are single radio service (GPRS)/Internet access capable, compared with 39 percent of nonusers. ƒƒ More than 50  percent are in a junior admin- istrative/managerial position, 16  percent are Figure 4.3 provides the sociodemographic details students, and 15 percent are self-employed/in of users and nonusers in Sri Lanka. business 4. User Survey Findings  21  Figure 4.3  Socioeconomic Characteristics of Mobile Money Users and Nonusers a. Gender b. Marital status Married Male Users Nonusers Female Single 0 20 40 60 80 100 0 20 40 60 80 100 Percentage of respondents Percentage of respondents c. Age d. Occupation Agriculture 15–25 Administrative/ managerial 26–35 Laborer/industrial/trade Clerk 36–45 Professional Public service 46–55 Self-employed/ business 56–60 Housewife/retired Student No answer Unemployed/retired 0 20 40 60 80 100 0 20 40 60 80 100 Percentage of respondents Percentage of respondents e. Highest level of education completed f. Average monthly income Up to grade 9 US$43–US$217 Up to US$218–US$261 O level Passed US$262–US$304 O level Up to US$305–US$348 adv. level Passed US$349–US$435 adv. level Graduate/ >US$436 professional 0 20 40 60 80 100 0 20 40 60 80 100 Percentage of respondents Percentage of respondents Source: IFC Mobile Money Study 2011. 22  IFC Mobile Money Study 2011: Sri Lanka ƒƒ More than 40  percent are in the top two income groups, while more than 60 percent of Figure 4.5  Mobile Money Services Used nonusers are in the lowest income group. eChanneling Bill payment Profile of Mobile Money Use Balance inquiry Fifty-four  percent of m-money user respondents use the eZ Pay service, offered through the mobile Airtime recharge operator as a separate service to any existing bank Purchasing account (often with government banks). Fund transfer As figure  4.4 shows, there is some variation in Other opinion over who provides the m-money ser- vice. Only 40  percent of users were aware that 0 5 10 15 20 25 30 Percentage of respondents the service is provided in partnership with a bank, though a bank account with NDB is not required. Source: IFC Mobile Money Study 2011. Figure 4.4  Mobile Banking Service Provider Figure 4.6  Users’ Reported Level of Knowledge of Mobile Banking Services Bank Bank and 21% mobile operator Low 37% 12% Another company 1% Medium 37% Don’t know 1% Lowest 7% Mobile operator 40% Highest 29% Source: IFC Mobile Money Study 2011. High 25% Figure 4.5 shows that the main application used Source: IFC Mobile Money Study 2011. by respondents is eChanneling, the service that allows callers to set up and pay for doctor visits (see discussion in chapter 2).The second most- used m-money service is bill payment, used by Payment Methods more than 50  percent of respondents, illustrat- The most typical method of bill payment for all ing an important demand, as many Sri Lankans respondents was cash direct to the company or a have to stand in line to pay bills. Balance inqui- bank teller. Sri Lanka is clearly still heavily depen- ries, airtime top-up, and fund transfers were also dent on cash for bill payment and use of electronic popular. payment mechanisms is low. Overall, m-money users rated their knowledge Figure 4.7 shows that users were more apt to pay of m-banking services as quite high, with almost their bills to companies directly (54 percent) com- 45  percent stating their knowledge is high or pared with 39  percent of nonusers. Conversely, highest, and over 35 percent stating it is medium 30 percent of nonusers made greater use of bank (figure 4.6). tellers versus only about 12 percent of users. About 4. User Survey Findings  23  significant percentage of the population is simply Figure 4.7  Typical Bill Payment Channels not aware of the services, but could become inter- ested in using m-money services. Cash direct to company Bank teller General Use of Money ATM Company pay outlet Users Financial Literacy Third-party outlet Nonusers Virtually all nonuser and user respondents had a M-banking bank account (97–99 percent), which they iden- tified as savings accounts. Twenty-five percent of Check users reported they also had current accounts, Credit card compared with only 6 percent of nonusers. Most Prepaid card respondents had a medium to high knowledge of bank services available, but more users (30  per- Not person paying bills cent) reported highest knowledge versus 10  per- 0 20 40 60 Percentage of respondents cent of nonusers (figure 4.8). Source: IFC Mobile Money Study 2011. Figure 4.8  Knowledge of Bank Services Available through Branch 10 percent of users paid bills via alternative chan- Percentage of respondents nels such as m-banking or third-party pay outlets. 40 Users Nonusers 30 Reasons for Not Using Mobile Services 20 The main reasons for not using mobile services were a lack of necessity, a lack of awareness about 10 the service, and unavailability of the service. Inter- est in new types of mobile services such as e-health 0 Lowest Low Medium High Highest (e.g., scheduling and paying for medical services) and greater promotion/awareness building of Source: IFC Mobile Money Study 2011. m-money services are seen as drivers for increased use of m-money services. Respondents indicated high knowledge and abil- “No need for services” was cited by 53  percent ity to use mobile phones, ATMs, and debit and of respondents, while 24  percent cited nonuse prepaid cards. Sixty-four percent of nonusers indi- because they were not aware of the service or the cated high knowledge of prepaid cards compared service was unavailable to them. Unavailability of with only 40 percent of users. Conversely, approx- service and security issues were cited by 5 percent imately 60  percent of nonusers indicated low of respondents. knowledge of POS transactions and debit card usage, and 40 percent of Internet banking. Currently available m-money services often mirror services offered by traditional means such Although overall knowledge levels were moder- as banks. Nonetheless, there seems to be an inter- ate to high, a significant portion (20–30 percent) est in new types of Internet- and mobile-enabled of respondents, especially nonusers, cited moder- financial services such as eChanneling. A key bar- ate to low understanding of some m-money ser- rier to adopting m-money services may be that a vices (beyond basic transactions), such as POS 24  IFC Mobile Money Study 2011: Sri Lanka devices, credit/debit cards, and Internet bank- ing (figure 4.9). The lack of penetration of POS Figure 4.10  Cash Withdrawal Sources Used Most Frequently devices and debit cards may explain the low level of knowledge about these services. Percentage of respondents 100 Users Nonusers Figure 4.9  Respondents Reporting High Level 80 of Knowledge of Various Financial Services and Channels 60 40 Mobile phone 20 ATM Internet banking 0 Bank teller ATM Users POS Source: IFC Mobile Money Study 2011. Nonusers Credit card Debit card Typical Withdrawal Amounts Prepaid card 0 20 40 60 80 Sixty-three  percent of users typically withdrew Percentage of respondents more than US$22 at a time. Only around 38 per- Source: IFC Mobile Money Study 2011. cent of nonusers made this large a withdrawal, with amounts spread evenly (10–15  percent) around the smaller denominations (figure  4.11). Again, Mobile Money Interface the target market for m-money is the middle- to upper-income brackets. With respect to users’ ease of conducting m-money transactions via the handset, 80–90  percent of users found the service interface and transactions Figure 4.11  Typical Cash Withdrawal Amounts of payments, transfers, and bill inquiries to be easy. Percentage of respondents 70 60 Primary Place of Cash Withdrawal 50 Users Nonusers The most common places where cash was with- 40 drawn were ATMs and bank tellers. Nonusers 30 used both places almost equally, with a slight pref- 20 erence for ATM transactions. Users overwhelm- 10 0 ingly (90  percent) withdrew cash from an ATM US$4– US$9– US$14– US$19– >US$22 (figure 4.10). US$8 US$13 US$18 US$22 Source: IFC Mobile Money Study 2011. A majority (56  percent) of nonusers withdrew funds less than several times a month, usually monthly or less frequently, compared with 89 per- cent of users who withdrew cash several times a Time to Nearest Bank Branch month to several times a week. The low use of bank The majority of respondents (50–60 percent) said tellers by users, along with their higher financial it took them between 5 and 20 minutes to get literacy, is indicative of higher incomes and educa- to a bank branch, while 35  percent said it took tion levels. Clearly, m-money is not being targeted between 20 and 30 minutes. Twenty-one percent to the lower end of the market. of nonusers said it took more than 40 minutes. 4. User Survey Findings  25  How Transfers Are Conducted Figure 4.13  Preferred Source of Information For nonusers, both ATMs and bank tellers were on Mobile Banking Services chosen almost equally as the main place where Percentage of respondents they conducted transfers. For users, 63  percent preferred to conduct transfers with bank tellers; 60 Users Nonusers only 14  percent used ATMs. This behavior con- trasts with users’ reported cash withdrawal behav- 40 ior, with 90 percent of withdrawals accomplished via ATMs. Users did conduct more Internet-based 20 transactions (13 percent). The majority of transfers by both users and nonusers were to family: 50 per- 0 cent and 45 percent, respectively (figure 4.12). Mass media An agent Directly from SMS from Friends/ bank operator family Source: IFC Mobile Money Study 2011. Figure 4.12  Fund Transfer Destinations Note: Multiple selections were allowed. Percentage of respondents 60 Users Nonusers mechanisms such as e-mail (67 percent compared with 10 percent of nonusers). Nonusers preferred 40 mass media such as TV advertisements. 20 Attitudes and Perceptions Overall, 90  percent of respondents indicated 0 medium to high trust in m-banking, which was Family Friends Employer/ Government Other similar to trust levels for banks in general. Of non- employee users, 25–35  percent had high levels of trust in Source: IFC Mobile Money Study 2011. m-banking technologies, third-party agents, and security from fraud. Users were more conservative, with high trust in these categories only 15–20 per- Marketing and Advertising cent of the time. Among users, knowledge of m-money services and eZ Pay is largely facilitated though short message In terms of specific m-money services that were service (SMS) marketing and mass media, includ- perceived as high value, salary deposits, micro- ing print media, TV, and billboards. Among non- loans, and international remittances were chosen users, mass media (51  percent) and family and (figure  4.14). Nearly 30  percent of users stated friend recommendations (47  percent) are key that international remittances were a high-value (figure 4.13). Respondents had heard of available item to be offered. This supports the point that m-money services that include eZ Pay and various an international remittance service would find a other m-banking applications of the major banks significant market. such as Hatton, Sampath, HSBC, and Standard Chartered. Value for Money When given a choice of how they would prefer to About 68  percent of respondents thought receive information on m-money services, respon- m-banking services were affordable to very cheap dents indicated a mix of SMS text messages, (figure 4.15). The same percentage of respondents e-mail, TV advertisements, and telephone. How- felt the services were cheaper than traditional bank- ever, users were more comfortable with electronic ing services. Those who thought the services were 26  IFC Mobile Money Study 2011: Sri Lanka Figure 4.14  Mobile Money Services in Which Figure 4.16  Perceived Mobile Money Benefits Users Express High Interest Percentage of respondents rating bene t “high” Percentage of respondents 100 30 Users Nonusers 80 60 20 40 20 10 0 Cost Time 24-hour Physical Immediacy of savings savings access security fund transfer 0 Salary deposit Microloans International Source: IFC Mobile Money Study 2011. remittances Source: IFC Mobile Money Study 2011. Respondents said the main reasons they used m-money services were, in order of preference, ability to conduct safe transactions with notifica- Figure 4.15  User Perceptions of Mobile Banking Fees tion, wider acceptance of m-money, and the abil- ity to cash out at more locations (figure 4.17). Percentage of respondents 40 Figure 4.17  User Reasons for Mobile Money 30 Usage 20 Percentage of respondents 100 10 80 0 60 Very Somewhat Affordable Somewhat Very 40 cheap cheap expensive expensive 20 Source: IFC Mobile Money Study 2011. 0 Safe transaction Wide acceptance More cash-out expensive indicated that lack of widespread use, with notification of m-money locations technical hurdles, and availability of more conve- Source: IFC Mobile Money Study 2011. nient methods to pay, increased the relative cost of the service. Respondents’ income had an impact on the outcome of this question. Users usually had a higher income than nonusers, and therefore their Conclusion perception of what is affordable may be different eChanneling is one of the most popular m-money from that of the lower-income nonusers. applications in Sri Lanka. It has a clear value proposition, and the fact that nearly 70  percent Benefits of Mobile Money of users in the survey have used it emphasizes its success. In contrast, most Sri Lankans continue to The convenience factors of m-money trump the pay their bills at the bank teller or directly to the cost factors. Respondents gave high ratings to company using cash. There is clearly substantial time savings, 24-hour access, and physical secu- demand for bill payment services. rity (figure 4.16). Cost savings was chosen less fre- quently, with only 20 percent of users selecting it One of the main challenges for m-money is lack of as a main benefit. awareness. Dialog launched eZ Pay in 2007, but the 4. User Survey Findings  27  lack of recent marketing may have kept the service In terms of marketing m-money services, users are from growing: more than 20  percent of nonusers more comfortable with electronic means of com- were not aware that there was an m-money service. munication, but nonusers prefer mass media. This Because of the low POS device and ATM penetra- is important information for future marketing ini- tion, 60 percent of nonusers have low knowledge tiatives to nonusers. of these devices. As the banks promote debit cards, knowledge will increase. M-money providers and Finally, the survey results show a relatively high banks can be expected to compete and to promote demand for a follow-up remittance survey, based their services to potential customers in the future. on the findings of the key informant interviews. 5 Business Models Existing Business Model: The two other major mobile operators in Sri Dialog Lanka are Etisalat and Mobitel, neither of which have any m-money products. Etisalat is the most Table  5.1 provides a brief summary of the aggressive MNO at the moment and is appar- Dialog business model for eZ Pay, a partnership ently adding the most subscribers to its network. between Dialog and NDB Bank. eZ Pay trans- Market share statistics for 2010 are not available, actions started strong at the beginning of 2009, but estimates are that Etisalat now has 25 percent but nearly ground to a halt by February 2010. No of the market, compared with Dialog’s 50 percent. new marketing has been done on eZ Pay since its Mobitel has a 15 percent share.1 initial launch. Staff at Dialog outlets are aware of the product, but there is no marketing material Etisalat is investigating the introduction of an visible to the customer. Unless a customer signed m-money product focused on the international on to eZ Pay at launch, there is little chance that remittance markets, particularly from countries in any new customers would be aware of it. Few cus- which Etisalat operates and where large numbers tomers know that cash-out is possible at Dialog of Sri Lankan expatriates can be found, such as stores. Saudi Arabia and the Gulf. eZ Pay does not offer a compelling business model. The roll-out of m-money requires significant It has not targeted a segment of the Sri Lankan upfront costs, particularly regarding the agent population with a clear product. The eChanneling network. Dialog’s current financial position in business makes it clear that Sri Lankans are pre- a competitive mobile market makes this invest- ment somewhat unlikely. Its earnings before inter- pared to experiment even if the service is expen- est, tax, depreciation, and amortization are at the sive. lower end of the scale compared with the other mobile operators. Other Business Models As a result of increased competition in the mobile sector, Dialog is under financial pressure. In the last quarter of 2009, its mobile operations lost SL Rs 9.3 billion (about US$82 million), a sub- 1  Data for 2010 are roughly in-line with 2009 market stantial loss; in comparison, in the fourth quar- share data, where Dialog had a subscriber identity ter of 2008, Dialog lost SL Rs 272 million (about module (SIM) market share of 48 percent (see Dialog US$2.4 million). 2009). 28  5. Business Models  29  Table 5.1  Dialog Business Model Element Description Business ƒƒ Dialog—reduce distribution costs and bring new revenue streams objective ƒƒ NDB Bank—opportunity to obtain new accounts and more distribution ƒƒ Bank had three-year exclusivity agreement with Dialog, which has now expired ƒƒ Dialog wants other operators to join Strategy ƒƒ Focus on maintaining variable costs and variable revenues (other than marketing costs), although there are high acquisition costs because of SIM swap ƒƒ The high acquisition cost constraint has been overcome by the USSD registration option, which is available now Target ƒƒ Potentially all Dialog, NDB Bank, Seylan Bank, and other business partners coming on board via the mobile commerce market network ƒƒ Ideally need 100,000 customers to break even ƒƒ ƒƒ There are currently 3,300 customers, though transactions are at a lesser rate ƒƒ SL Rs 85 (U takes SL R Marketing ƒƒ Minimal marketing effort ƒƒ All other tra strategy on whethe Revenue ƒƒ The transaction fee is shared between the software provider (mChek), the acquiring bank, the issuing bank (NDB Bank), and traditional streams the issuing and acquiring operators (Dialog) ƒƒ SL Rs 25 (US$0.22) transaction fee is paid by the customer for electricity and water bills ƒƒ All other transactions cost customers SL Rs 10–25 (US$0.09–US$0.22) (depending on whether a promotion is being offered) compared with SL Rs 25 (US$0.22) bank charges ƒƒ SMS costs (one per transaction) are SL Re 1 (US$0.01) Costs ƒƒ Platform is provided by mChek, which takes a cut of every transaction ƒƒ Marketing costs are shared between bank and operator ƒƒ New SIM is required to operate eZ Pay in the case of the SIM Toolkit (STK) application; customers also have the option of registering via USSD ƒƒ Cash handling not yet seen as a big issue (mainly because there is virtually no cash handling) Transac- ƒƒ Utility bill payments tions ƒƒ Cash-in and cash-out (tiny) Merchants ƒƒ All Dialog service centers and approximately 3,000 non-Dialog retail merchants nationwide that have been acquired onto the eZ Pay network ƒƒ Dialog outlets can do KYC on behalf of NDB Bank but sent daily to bank for final verification and approval; embossed cards are created, and the file is sent to mChek for activation within four working days ƒƒ Agents currently not incentivized to open accounts, but agents get paid for selling SIMs (this is linked to the lack of marketing) Users ƒƒ Very limited use Pipeline ƒƒ Merchant-initiated transactions—bill payments ƒƒ Bringing in other banks and operators (e.g., Seylan Bank) ƒƒ Seylan Bank USSD and NDB Bank USSD application already available ƒƒ Merchant acquisition—Ceylinco VIP added to the network of merchants Model/ ƒƒ Model: Mobile operator–centric partners ƒƒ Partners: Acquiring bank—NDB Bank; issuing bank—Seylan Bank Source: IFC Mobile Money Study 2011. 30  IFC Mobile Money Study 2011: Sri Lanka Challenges ƒƒ Mass market—based on low cost and increased efficiency Dialog has several challenges in its current model of m-money. At the moment, transactions using eZ ƒƒ Segmented market—providing a particular ser- Pay are minimal. The initial launch of Dialog was vice to a specific target market. aimed at the lower end of the income spectrum. The lack of mass market debit and credit cards However, customers using the service are from the means that there is a potential opportunity to use middle- to upper-income segments. Most impor- m-money for goods purchases. Another potential tant, eZ Pay’s overall value proposition is unclear. mass market strategy is in the P2P or remittances Is it intended to act as an alternative access chan- market, where current products are either aimed nel for NDB Bank, or is it intended to add value at the middle- to upper-income markets or are for existing Dialog customers? If the latter, then inefficient (for example, the post office’s money the lack of a marketing strategy ensures that few order system). customers will perceive its value. A strategy that focuses on a particular segment The key issue Dialog must resolve is the value of the market is also possible. There is a massive proposition of its services. What problem does public transport market with an identified inef- eZ Pay address that makes it more attractive than ficiency. The formalization of this market could existing products (such as cash)? Which markets is potentially benefit both consumers and bus com- eZ Pay targeting? Once these questions have been panies. In the absence of an NFC-enabled phone, resolved, fundamental business model challenges linking a smart card to alternative mechanisms of such as the marketing strategy can be addressed. payment and allowing the smart card to be topped The next section provides some insight into the up using an m-money account represents a poten- value propositions available to Dialog. tial business opportunity. Another potential strategy is the payment of bills Potential Business such as water and electricity. Bill payment is often Opportunities done at the post office, with more than 6.5 mil- The Sri Lankan financial sector is unique. It has a lion utility payments made there each year. There very high bank account penetration rate of 59 per- are two ways to address this market: either as a cent. Even assuming that many people have mul- partner with or as a competitor to the post office. tiple accounts, the penetration rate is still high. In As a partner, the post office has the advantage of the user survey, 98 percent of respondents had a providing an existing network of agents with the bank account. Yet penetration of debit and credit ability to pay out cash and an extensive branch cards is very low, and ATM and POS device roll- network. The risk associated with this strategy is out is very small. Most of the bank accounts are that Dialog would have to deal with an inefficient with government institutions, but private sector state-owned organization which is only start- commercial banks are making inroads with the ing the process of computerization. The alterna- middle and upper classes. Bank accounts are pri- tive strategy of competing against the post office marily used to store value—that is, they are used means that Dialog does not have to worry about as savings accounts rather than transactional the inefficiencies of the post office, but it must accounts. Even with its high bank account pen- find an agent network. etration, the country has a large informal financial economy, especially for P2P transfers and interna- Risks tional remittances. Each strategy has a risk profile. Public transport Therefore, potential business opportunities can be as a strategy for m-money requires a very quick divided into two groups: payment process. NFC provides this function, but 5. Business Models  31  has yet to be successfully integrated into the wide areas as well as transit. Breaking out of a closed range of mobile phones that are available. Thus, loop (i.e., where smart cards can only be used for for NFC to successful, virtually all subscribers public transport purchases and no other payments) would have to replace their phones. As the expe- has been difficult in other parts of the world. Con- rience of replacing subscriber identity module veniently, the eZ Pay service is already allied to (SIM) cards for eZ Pay has shown, this would be a debit card via its partnership with NDB Bank. is a major obstacle. Being used for transit would help eZ Pay develop If a smart card option is chosen for public trans- a clear value proposition that would attract exist- port payment, it should be functional in other ing and new customers. 6 Conclusion T he Sri Lankan market has several m-money penetration of the formal financial sector means opportunities. Existing m-money imple- that the concurrent operation of an m-money mentations, such as eZ Pay, do not have a system alongside the current financial system will clear target market, and roll-out has basi- not work. To a degree, Dialog has recognized this cally been suspended. There is some regulatory problem by partnering with NDB Bank. How- uncertainty because no mobile payments regula- ever, a key problem remains: Dialog and NDB tion has been issued. Despite this, the market is Bank must determine where their eZ Pay product relatively open, and the Central Bank of Sri Lanka fits in its overall product basket. In other words, is agreeable to m-money ventures. As Kenya has eZ Pay should not be seen as an experiment, but as shown, the lack of a formal regulatory frame- part of the existing product portfolio. Currently, work—at least initially—is not an obstacle. this is not the case. Dialog and NDB Bank must clearly segment their customer base to determine While there are several m-money opportunities, precisely which customers will be targeted. As a Dialog is not well placed to take advantage of general offering, as is currently the case, eZ Pay them because it has insufficient capital to invest gets lost among the wide range of other financial in m-money initiatives, lacks a clear value propo- products. sition for its m-money service, and has a limited marketing strategy. Once customers have been appropriately seg- mented, a formal marketing strategy can be The issue of insufficient capital is based on a very adopted. Currently, there is no clear marketing competitive mobile market and Dialog’s alloca- strategy because Dialog and NDB Bank have not tion of resources to other business areas such as determined which customers would find eZ Pay mobile broadband. Dialog is currently the incum- appealing. bent mobile operator, but its market share is declining as more aggressive competitors such as Dialog is facing pressure in its core mobile busi- Etisalat enter the mobile market. The adoption of ness. In other parts of the world, m-money has an m-money strategy using any of the proposed been successful in capturing an increased propor- opportunities would reduce the challenges that tion of customers’ wallets, reducing churn, and Dialog faces, as well as shore up its market share. building loyalty. Thus, if Dialog can clarify its strategy, this would be a favorable time for inves- An m-money strategy requires a clear value propo- tors to provide funding for Dialog to take advan- sition. This report has argued that the high level of tage of these m-money opportunities. 32  Appendix A Fact Sheet and Demand Estimates Table A.1  Fact Sheet Country Profile Population 21.3 million a Sector shares of GDP Geographic area 65,610 sq. kma Agriculture GDP US$41.32 billionb 13% GDP per capita US$2,041b Gini index 41.1c Industry Services 29% 58% Rural population 84.9% (17.1 million) d Rural poor 7.9% (1.35 million)d Population below poverty line 22.7% (4.8 million) (calculated; 2008)d Financial Profile Number of banks 36e Deposit market share of nancial institutions Total branches 5,703e Total correspondent banking — agents All other nancial Number of bank accounts 7.0 million depositorsf institutions 33% Banking penetration 59% (2008)g Number of POS devices 24,977h; 117.2 per 100k pop Top six banks 67% Number of ATMs 1,876h; 8.8 per 100k pop Number of payment cards 5.95 million (3rd quarter 2009)i 33  34  IFC Mobile Money Study 2011: Sri Lanka Financial Profile (continued) Credit card growth 605,930 (3rd quarter 2005); Number of payment cards 852,312 (3rd quarter 2009);j (millions) 8.9% average yearly growth 2005–09 5.1 Debit card growth 2.2 million (3rd quarter 2005); 5.1 million (3rd quarter 2009);j 23.4% average yearly growth 0.852 2005–09 Remittance flow—inbound US$2,947 millionj (Canada, Credit Debit Saudi Arabia, United Kingdom) Remittance flow—outbound US$385 millionj (India) Use of e-payment channels (transactions per capita) Number of microfinance 11 (2009)k 1.2 institutions ATM Number of microfinance 2.6 million depositors (2008)k institution accounts Credit card Tele/mobile/Internet Debit card 0 Herfindahl-Hirschman Index — 2007 2008 2009 (HHI) Mobile Profile Mobile operators 5 Mobile operator market share Mobile coverage 90% (2007)l Dialog Number of mobile subscribers 14.1 millionm 44% Mobile penetration 69.0%n Hutchinson 5% Internet penetration 1.2%n Mobitel 23% Broadband penetration — Bharti Airtel Tigo 12% Herfindahl-Hirschman Index 2851 (Etisalat) (HHI) 16% Note: All data are for 2009 unless otherwise stated. — = not available. a. CIA 2010. b. International Monetary Fund, World Economic Outlook Database, April 2010; http://www.imf.org/external/pubs/ft/weo/2010/01/ weodata/weorept.aspx?sy=2008&ey=2010&scsm=1&ssd=1&sort=country&ds=.&br=1&pr1.x=33&pr1.y=9&c=524&s=NGDPD%2CN GDPDPC&grp=0&a=. c. United Nations Development Programme, Human Development Report Statistics 2009, http://hdrstats.undp.org/en/indicators/161.html. d. IFAD 2010. e. Central Bank of Sri Lanka, http://www.cbsl.gov.lk/htm/english/05_fss/f_1.html#2. f. Central Bank of Sri Lanka 2010, appendix. g. Honohan 2008. h. Central Bank of Sri Lanka 2010, chapter 8. Appendix A. Fact Sheet and Demand Estimates  35  i. Central Bank of Sri Lanka 2009d. j. Central Bank of Sri Lanka 2005. k. MIX Market, Microfinance in Sri Lanka: Country Profile, http://www.mixmarket.org/mfi/country/Sri%20Lanka. l. International Telecommunication Union, World Telecom ICT Indicators, http://www.itu.int/ITU-D/icteye/Indicators/Indicators.aspx. m. Telecommunications Regulatory Commission of Sri Lanka, http://www.trc.gov.lk/information/statistics.html. n. Central Bank of Sri Lanka, http://www.cbsl.gov.lk/pics_n_docs/10_pub/_docs/efr/annual_report/ar2009e/PDF/12_chapter_08.pdf. Table A.2  Demand Estimates Socioeconomic data Population (millions) 21.3a GDP per capita (US$) 2,041b Gini index 41.1c Financial data Bank accounts (million) 7d Banking penetration (percent) 59.0e Number of POS devices 24,977f POS devices (per million inhabitants) 1,173g Number of ATMs 1,876f ATMs (per million inhabitants) 88h Payment cards (million) 5.95i Payment cards (per million inhabitants) 279,343 Mobile data Mobile operators 5 Mobile penetration (percent) 69.0 Number of mobile subscribers (million) 14.1j Potential demand E-payments (per month) 1,521,750k G2P (transactions per month) 1,600,000l Payroll, informal sector (transactions per month) 4,708,418m P2P (transactions per month) Unknown Public transport (trips per month) 264,000,000n Unbanked (persons) 4,885,396o Utility (payments per month) 6,440,168p a. CIA 2010. 36  IFC Mobile Money Study 2011: Sri Lanka b. International Monetary Fund, World Economic Outlook Database, April 2010; http://www.imf.org/external/pubs/ft/weo/2010/01/ weodata/weorept.aspx?sy=2008&ey=2010&scsm=1&ssd=1&sort=country&ds=.&br=1&pr1.x=33&pr1.y=9&c=524&s=NGDPD%2CN GDPDPC&grp=0&a=. c. United Nations Development Programme, Human Development Report Statistics 2009, http://hdrstats.undp.org/en/indicators/161.html. d. Central Bank of Sri Lanka 2010, appendix. e. Honohan 2008. f. Divide number of POS devices by population (million). g. Central Bank of Sri Lanka 2010, chapter 8. h. Divide number of ATMs by population (million). f. Central Bank of Sri Lanka 2009d. i. Telecommunications Regulatory Commission of Sri Lanka, http://www.trc.gov.lk/information/statistics.html. j. E-payments per year (18,261,000), divided by 12 months. k. Central Bank of Sri Lanka 2008. l. Samurdhi payments to households. m. Informal sector = 61.9 percent (Sri Lanka Labor Force Survey, http://www.statistics.gov.lk/samplesurvey/LFS_Annual%20 Report_2009.pdf). n. Public transport data (de Silva 2010b): Public transport (per month) = 264,000,000 Bus trips (state-owned) per day = 2,000,000 Bus trips (state-owned) per month = 44,000,000 Bus trips (privately owned) per day = 10,000,000 Bus trips (privately owned) per month = 220,000,000 o. Unbanked data: Number of bank accounts = 7,030,204. Assuming 59% penetration, total number of potential accounts = 1,915,600. Unbanked = 4,885,396 (Central Bank of Sri Lanka). Calculation of total number of potential bank accounts (11,915,600), minus the number of actual bank accounts referenced above. p. Utilities consists of Postpaid subscribers: 2,000,000 (estimate that 16 percent of subscribers are postpaid; see de Silva 2010b) Satellite television (DTV): 160,000 (Lanka Business Online 2010) Electricity: 4,280,168 (Ceylon Electricity Board). Appendix B Persons Interviewed Asanga Karunaratna, Chief Executive Officer, S.D.N. Perera, Senior Deputy General Manager, eChanneling National Savings Bank Dedunie de Silva, Senior Legal Officer, Group Legal Saliya Rajakaruna, Chief Executive Officer, Nations Strategy and Operations, Dialog Trust Bank Druvi Sirisena, M-Commerce Manager, Group Sanjaya Karunasena, Head of Technology and Chief Commercial, Dialog Software Architect, ICTA, Sri Lanka Dumindra Ratnayaka, Chief Executive Officer, Sathika Wickramasinghe, Nielsen Sri Lanka Etisalat Sri Lanka Shabbir Raheem, Assistant Manager, Card Fariq Cader, Senior General Manager, Group Finance Operations, NDB Bank Operations and Revenue Assurance, Dialog Shaheen Cader, MD, Nielsen Sri Lanka Gehan Dias, Assistant Vice President, Cards and Shakila Wijewardena, MD, SEEDS Alternate Distributions Channels, NDB Bank Sheahan Arasaratnam, General Manager, Retail Janaki Mampitiya, Assistant Governor, Central Banking Products and Head of Marketing, Bank of Sri Lanka Standard Chartered K.B. Dissanayake, Postmaster General, Sri Lanka Sunil Jayantha Nawaratne, Director, Human Post Office Capital and Youth Development, Cargills Kanchana Abeywickrama, Operations Analyst, IFC- (Ceylon) PLC SEDF Colombo Supun Weerasinghe, Chief Executive Officer, R.P. Bandula Thilakasiri, Director General, Dialog Samurdhi Authority of Sri Lanka Sushanya Samarakkody, Coordinator Rangana De Silva, Project Manager, ICTA, Sri Lanka M-Commerce, Dialog Reshan Dewapura, Chief Operating Officer, ICTA, Trevine Fernandopulle, Chief Risk Officer, Bank of Sri Lanka Ceylon Rohan Peiris, Head of Information Technology, Wasantha Deshapriya, Program Director, ICTA, Sri Bank of Ceylon Lanka 37  References ADB (Asian Development Bank). 2010. —. 2010. Annual Report 2009. http://www. “Developing the Micro-Insurance Sector in cbsl.gov.lk/pics_n_docs/10_pub/_docs/efr/ Sri Lanka.” http://www.adb.org/Documents/ annual_report/ar2009e/ar09_content_2009_e. Brochures/Social-Protection-Project-Briefs/sri- htm; accessed June 18, 2011. microinsurance.pdf; accessed June 15, 2011. —. 2011a. “Mobile Payments Guidelines No. Central Bank of Sri Lanka. 2005. “Payments 1 of 2011 for the Bank-led Mobile Payment Bulletin 3Q2005.” http://www.cbsl.gov. Services.” http://www.cbsl.gov.lk/pics_n_ lk/pics_n_docs/10_pub/_docs/periodicals/ docs/09_lr/_docs/directions/psd/13_mobile_ payment_2011_1.pdf; accessed June 18, 2011. payment_bulletin/Payments_Bulletin_3Q2005. pdf; accessed June 15, 2011. —. 2011b. “Mobile Payment Guidelines No. 2 of 2011 for Custodian Account Based Mobile —. 2008. “Payments Bulletin, 3Q2008.” Payment Services.” http://www.cbsl.gov.lk/pics_n_ http://www.cbsl.gov.lk/pics_n_docs/10_pub/_ docs/09_lr/_docs/directions/psd/14_mobile_ docs/periodicals/payment_bulletin/Payments_ payment_2011_2.pdf; accessed June 18, 2011. Bulletin_3Q08.pdf; accessed June 15, 2011. CGAP (Consultative Group to Assist the —. 2009a. “Directions, Circulars and Poor). 2009. Financial Access: South Asia: A Guidelines Issued to Licensed Commercial Regional Snapshot. http://www.cgap.org/gm/ Banks.” http://www.cbsl.gov.lk/pics_n_ document-1.9.40542/2009%20Funder%20 docs/09_lr/_docs/directions/bsd/LCB_upto_ Survey%20Regional%20SA.pdf; accessed June November_2009.pdf; accessed June 15, 2011. 15, 2011. —. 2009b. “Financial Sector Development CIA (Central Intelligence Agency). 2010. CIA and System Stability.” Central Bank of Sri Lanka Factbook 2010. https://www.cia.gov/library/ Annual Report, http://www.cbsl.gov.lk/pics_n_ publications/the-world-factbook/geos/ce.html; docs/10_pub/_docs/efr/annual_report/ar2009e/ accessed June 14, 2011. PDF/12_chapter_08.pdf; accessed June 15, de Silva, Harsha. 2010a. M-Payments in Public 2011. Transport: The Case of Bus Tickets in Sri Lanka. LIRNEasia. http://lirneasia.net/wp-content/ —. 2009c. “Financial Sector Stability Review.” uploads/2010/04/Harsha-de-Silva.pdf; accessed http://www.cbsl.gov.lk/pics_n_docs/10_pub/_ June 15, 2011. docs/efr/financial_system_stability_review/ fssr_2009.pdf; accessed June 15, 2011. —. 2010b. The Need for an Efficient Bus and Train Ticket Sale-and-Purchase System in Sri —. 2009d. “Payments Bulletin 3Q2009.” Lanka: Possibilities for a Mobile 2.0 Solution. http://www.cbsl.gov.lk/pics_n_docs/10_pub/_ LIRNEasia. http://lirneasia.net/wp-content/ docs/periodicals/payment_bulletin/Payments_ uploads/2010/06/Full-Paper-English.pdf; Bulletin_3Q09.pdf; accessed June 15, 2011. accessed June 14, 2011. 38  References  39  Department of Census and Statistics, Sri Lanka. —. 2010. “Possibilities for Bus Tickets to n.d. “Household Income and Expenditure Go Mobile in Sri Lanka.” Policy Brief. http:// Survey 2006/07.” http://www.statistics.gov.lk/ lirneasia.net/wp-content/uploads/2010/04/ HIES/HIES2006_07Website/Publications/ mBusPass.pdf; accessed June 15, 2011. HIES200607Final%20ReportWeb%20.pdf; accessed June 15, 2011. Pickens, Mark, David Porteous, and Sarah Rotman. 2009. “Banking the Poor via G2P Payments.” Dialog. 2009. Annual Report. http://www.dialog. lk/about/investors/financials/annual-reports/; Focus Note No. 58. Washington, DC: accessed June 14, 2011. Consultative Group to Assist the Poor. http:// www.cgap.org/gm/document-1.9.41174/FN58. Honohan, Patrick. 2008. “Cross-Country Variation in pdf; accessed May 23, 2011. Household Access to Financial Services.” Journal of Banking and Finance 32: 2493–2500. http://ideas. Porteous, David. 2006. “The Enabling repec.org/a/eee/jbfina/v32y2008i11p2493-2500. Environment for Mobile Banking in Africa.” html; accessed June 15, 2011. http://www.bankablefrontier.com/assets/pdfs/ IFAD (International Fund for Agricultural ee.mobil.banking.report.v3.1.pdf; accessed May Development). 2010. “Country Statistics, Sri 22, 2011. Lanka.” Rural Poverty Portal. http://www. ruralpovertyportal.org/web/guest/country/statistics/ SAMN (South Asian Microfinance Network). n.d. tags/sri%20lanka; accessed June 18, 2011. Microfinance Sector in Sri Lanka, http://www. samn.eu/index.php?q=microfinance-srilanka; Lanka Business Online. 2010. “Alternative Media: accessed May 19, 2011. Lanka’s DTV to Target More Subscribers, Higher Revenues.” March 27. http:// World Bank. 2009. “Bringing Finance to Pakistan’s www.lankabusinessonline.com/fullstory. Poor: A Study on Access to Finance for the php?nid=825351420; accessed June 15, 2011. Underserved and Small Enterprises.” http:// LIRNEasia. 2009. “Teleuse at the Bottom of the siteresources.worldbank.org/INTFR/Resources/ Pyramid.” http://lirneasia.net/projects/2008-2010/ Paper-NenovaNiangandAhmad.pdf; accessed bop-teleuse-3/; accessed June 15, 2011. June 19, 2011. Photo credits: IFC Photo Library (top); True Money (bottom) 2121 Pennsylvania Avenue, NW Washington, DC 20433 USA www.ifc.org November 2011