Report No. 32418-MX Mexico Broadening Access to Financial Services Among The Urban Population: Mexico City's Unbanked (In Two Volumes) Volume I: Main Report September 30, 2005 Colombia and Mexico Country Management Unit Finance, Private Sector and Infrastructure Department Latin America and the Caribbean Region Document of the World Bank ABBREVIATIONSANDACRONYMS ATM Automated Teller Machine CCO Check Cashing Outlets CDFI Community Development Financial Institution CDVC Community Development Venture Capital CFNI Chartered non-bank intermediaries CRA Community Reinvestment Act CNBV Comisidn Nacional Bancaria y de Valores CONDUSEF Comisidn Nacional para la Proteccidn y Defensa del Usuario de Servicios Financieros DANE Departamento Administrativo Nacional de Estadistica EFT Electronic FundTransfer ESW Economic Sector Work GDP Gross Domestic Product IDA Individual Development Accounts IDB InterAmerican Development Bank IFS InstitutionalFinancial Self-sufficiency IMF InternationalMonetary Fund INEiGI Instituto Nacional de Estadistica, Geografia e Informacidn IPAB Institute for Bank Savings LMI Low and Moderate Income MESBIC Minority Enterprise Small Business Investment Company MFI Microfinancial Institutions NBFI Non-Bank Financial Institutions NGO Non-governmental organizations OED Operations EvaluationDepartment ROSCAS Rotating Savings and Credit associations SCIS Credit Unions and Savings and LoanAssociations SME Small and MediumEnterprise SOFOLES Sociedad Financiera de Objeto Limitado (Mortgage associations) iv MEXICO BROADENINGACCESSTOFINANCIALSERVICESAMONGTHE URBAN POPULATION:MEXICO CITY'SUNBANKED Contents Page Introduction to the 2005 Publication................................................................. vi11 ... Executive Summary................................................................................................................. i x Chapter 1. Access to Financial Services Matters !............................................................... 1 1 Accounts, But Only a Minority Has Them................................................................................ Chapter 2. Introduction............................................................................................................................. Demand for Financial Services inMexico City: The Majority Want Bank 6 1. Who Uses Banks InMexico And Who Doesn't?.......................................................... 7 2 11 Why Are the Unbanked, Unbanked?........................................................................... How do the unbanked get along without financial services? ...................................... 3. . 25 Don't They ReachMost Mexicans ? ...................................................................................... Chapter 3. 4. Conclusion - There i s Significant Room for the Financial Service Market to Grow ..22 What Financial Services Are Available inMexico City, at What Cost and Why 28 28 2. 1. Mexico's Financial Sector - Shallow And Still In Shock? ......................................... 38 New Initiatives for IncreasingAccess ......................................................................... Products and Costs - A Focus On The HighEndCliente ........................................... 44 4. 3. 49 5. Conclusions.................................................................................................................. Structural Obstacles to Broadening Access ................................................................. 54 Experiences............................................................................................................................ Chapter 4. Recommended Approaches to IncreasingAccess and Reflections On Other 56 1. Product Based Approach ............................................................................................. 57 2. Price-Based Approach ................................................................................................. 59 Neighborhood Based Approach................................................................................... 60 4. 3. 62 Information and Education Approach ......................................................................... Technology Based Approach....................................................................................... 5. 62 6. Next Steps.................................................................................................................... 63 Bibliography: de la Torre. Leipziger. Gasha. "Behind Credit Fluctuationsin LatinAmerica:Old and New Suspects" -World Bank. 2003 and Adolfo Barajas and RobertoSteiner (2002). "Why Don'tThey Lend?Credit Stagnation inLatinAmerica". IMFStaffPapers. special issue. vol. 49. V Listof Figures Figure 2-1 Use of Financial Services is Very Low in Mexico; Compared to Other Countries .............................. 6 Figure 2-2 Use of banking products i s low. more so for more sophisticated products........................................... 7 Figure 2-3 Most bank account holders belong to the highest income brackets ..................................................... 8 Figure 2-4 Use o f formal financial services increases with education .................................................................. 9 Figure 2-5 Most people. including the majority of the unbanked. are paidby check........................................... 13 Figure 2-6 Most people would rather be paid in cash -" (Preferred methods of being paid) ............................... 13 Figure 2-7 The most commonly report savings instrument i s the informal tanda. ....... 16 Figure 2-8 Commercial stores make up the primary source of credit in .............................................................. .................................. 18 Figure 2-9 What do people borrow for? It depends on whether or not they have bank accounts........................ 20 Figure 2-10 Unbankedpersons don't attempt to open ............................... ..................................................... 22 Figure 2-12 The unbanked want loans for other things than business investments............................................. Figure 2-11The unbanked want savings accounts............................................................................................... 24 Figure 3-1 The Decline in deposits and lending by commercial banks has not affected GDP growth ...............25 34 Figure 3-2 Are non-bank financing institutions replacing commercial banks...................................................... 36 Figure 3-3 Banks select locations according to income Rather than population.................................................. 37 Figure 3-4 Financial services offered inMexico City by he biggest 5 banking institution...specialization in product types and lack of competition ................................................................................................................. Figure 3-5 Non-bank financial institutions concentrate on rural areas, outside of Mexico City .......................... 39 45 Figure 3-7 Different regulatory agenciesrequire the same reports from banks ................................................... Figure 3-6 Each bank sends 260 reports per year to 5 different regulatory agencies in Mexico.......................... 50 50 Figure 4-1 Profitability of CRA Lending ............................................................................................................. 61 Figure 4-2 Profitability of CRA lending ............................................................................................................. 62 List of Tables 9 Table 2-2 The unbanked are disproportionately independent workers and economically inactive. ..................... Table 2-1 Use of formal financial instruments increases with income.................................................................. 10 Table 2-3 Non-bank credit terms are expensive............................................................................ Table 2-4 Why don't you have a bank account................................................................................................... 23 Table 3-1 Mexico's financial market is shallow compared to peer countries (1999/2000) ................................. Table 3-2 Mexico's two largest commercial banks control 50% of the market.................................................. 28 30 Table 3-3 Non-Bank institutions differ as to their degree of formality ................................................................ Table 3-4 The spread between passive and active interest rates is high in Mexico...and growing.................... 32 36 Table 3-5 Location, Location, Location - % of respondents living within !hhour.............................................. 38 Table 3-6 Even basic banking products are costly and inaccessible to much of Mexico City's population ........41 Table 3-7 Bankingproducts are easier to access in the U.S. than in Mexico....................................................... 43 Table 3-8 ATM and electronic banking services inMexico are beyond the reach of most Mexicans - but a US Credit Card is not ................................................................................................................................................. 44 Table 3-9 Costs of bank regulation are higher in Mexico .................................................................................... 49 Listof Boxes Box 2-1 New data to look at access to financial services...................................................................................... 7 Box 2-2 Informal Savings Institutions ................................................................................................................ 15 Box 2-3 What it costs a Mexican Family to be Unbanked................................................................................... 21 Box 3-1 Regulatory and legal framework for banks ....... ................................................................................ 31 Box 3-2 Globalization could have an effect on domestic lending........................................................................ 34 Box 3-3 High end" products are growing............................................................................................................. 35 Box 3-4 Strategic thinking of banks and the absence of low-endproducts.......................................................... 42 Box 3-5 Attitudes can cost money and loose clients in the low income market ................................................ 52 Box 4-1 The Individual Development Account Program in the United States uses bank accounts to encourage 58 Box 4-2 Municipal Smart Cards Facilitate Electronic Payments for Public Services .......................................... asset building for the poor.................................................................................................................................... 58 Box 4-3 Mexican Banks introduce innovate ways to facilitate commercial credit .............................................. 59 Box 4-4 Induced competiton can lead to lower prices for banking services ........................................................ 60 vi Box 4-5 Attitudes can cost money and loose clients in the low income market ................................................. 61 62 Box 4-7 Public disclosure and information as a tool to increase access to financial services............................ Box 4-6 Electronic Transfer payments can lower cost and lead .......................................................................... 63 Illustrations Illustration2-1 The line to pay telephone bills in Mexico City........................................................................... 12 Illustration2-2 Facade of the "Caja de Ahorro Economia Solidaria. A.C." ....................................................... 17 vii INTRODUCTIONTO THE 2005 PUBLICATION The present report was written in 2003,based on research carried out inMexico City in2002. While no subsequent evidence has emerged to suggest that the data presented here on the use of financial institutions and the unbanked has changed in the interim, information on the banking sector may be somewhat out of date. The report relied on data from 1995-2000. Since that time certain changes have been detected which could have interesting implications for access to financial services. A reduction in the difference between short term and long term interest rates has made investments in government papers less competitive and as a result domestic lending as a percentage of GDP i s expected to pick up significantly. At the same time, charges for banking services appear to have reached a maximum, creating strong incentives for Banks to broaden their clientele base in order to increase revenues. A new player, Banco Azteca, has entered the field with a strategy aggressively directed toward offering services to the previously unbanked. Banco Azteca's success could create a new business model for banks in Mexico. At the same time, there i s a noted tendency for companies to shift from salary payments in check to payment through direct deposits in banks. While there i s no survey or statistical evidence available, this phenomenon could lead to increased use of banking services, although bankers observe that without some financial education, workers tend use their new accounts only to draw down cash. Finally, the presence of new technologies, such as pre-pay cards, ATM deposit facilities, and point of sale card readers could lead to a rapid spread of financial services among the unbanked. The Mexican government has already begun a policy to encourage card readers at "points of sale" ("teminalizacion") and there are several efforts focused on remittances, to lower costs and to develop banking through money transfers. The reader should bear in mind that the financial services represent a rapidly developing field and that it has gone beyond the scope of this paper to include the latest innovations. Its fast evolution notwithstanding, Mexico's financial sector still eludes the majority of the country's citizens, as this report argues, with consequences for national economic growth, for the financial sector and for the "unbanked" themselves. ... VI11 EXECUTIVE SUMMARY 1. The present report looks at issues of access to financial services in urban Mexico, drawing on existing documentation as well as on new research (surveys, interviews and focus groups) carried out in Mexico City during 2002. It is one of several initiatives in the region to explore the issue of financial exclusion: companion efforts include work on financial exclusion in urban Brazil ("Brazil: Access to Financial Services), in Bogotfi, Colombia (ongoing) and in rural Mexico (Mexico's Financial Markets and the Rural Economy). These initiatives differ in their methodology and emphasis, although all discuss the unmet demand for financial services, the characteristics of the "unbanked," and, to varying degrees, what can be done about financial exclusion. 2. The motivation behind this interest in financial exclusion i s two fold. First, it i s well documented that being unbanked has costs (Dunham, 2001; Caskey, 1997; and Stegman 1999; Carr and Scheutz, 2001; Skricki and Fernindez, 2001; Rhine et al., 2000; Green andLichter 1996.) It makes it more expensive to engage ina number of transactions (paying and being paid) and more difficult to save while maintaining the value of an asset As such, greater access to financial services i s desirable from an efficiency point of view. Second, financial exclusion i s an issue that primarily affects the poor. Access to financial services can help with poverty alleviation, particularly to the extent that it encourage asset buildings and help cope with shocks and overcome liquidity constraints. The literature on development banking highlights low-income families' need for a wide range of financial services both for everyday life and for asset building purposes (Dichter, 1997; Caskey, 2001; Robinson, 2001; Rutherford, 2001). As a result, the U S and most developed countries have policies to encourage greater access to financial services for the poorer segments of the population. In developing countries, this i s also the case, although this i s typically done through micro-credit approaches (rather than through the mainstream banking sector), and often with a more rural orientation. 3. Financial exclusion i s of particular concern in Mexico, where banks are still reeling from the twin Peso and banking crisis of 1994 and 1995. With bank credit to the private sector amounting to only about a quarter of pre-crisis levels, most firms and households rely on non-bank institutions and informal providers (including suppliers credits) to satisfy their demands for financial services. The persistent low levels of access to financial services by both low and middle income households has therefore become an important policy issue inMexico. 4. As to the present paper's focus on urban Mexico, it is motivated by the fact that Mexico i s now three quarters urban, with cities accounting for about 85% of the country's wealth and half the country's poor. Thus, access to financial services in urban Mexico i s of interest both from a social perspective and from a growth or efficiency one. Data was unfortunately not available for urban Mexico as a whole, so the report mostly relies on the new data gathered in Mexico City through a survey of 1,500 households. The survey was applied by INEGI as part of its regular labor and i x employment surveys to a sample statistically representative of Mexico City in terms o f location, age, sex and income level of the person interviewed, and a series of focus group meetings with unbanked persons. Note that since access i s likely be higher in Mexico City than inthe other cities in the country, the report probably underestimates the problem of financial exclusion in urban Mexicol. Nevertheless, the report does provide a good complement to the parallel work on the rural economy mentioned above. In both reports, it i s found that the shallowness of Mexico's financial sector means that financial exclusion i s an issue for the poor but also for the middle class and entrepreneurs. 5. The report i s organized as follows. Chapter 1 sets the stage by discussing the importance of access to financial services and what other countries are doing to promote such access. Chapter 2 presents an analysis of actual and potential demand for financial services, and offers a profile of the unbanked residents of Mexico City. Chapter 3 then looks at the supply o f financial services by the formal financial sector and possible explanations for Mexico's extremely low rate of financial access. Chapter 4 concludes and discusses possible means of improving financial participation. The findings presented inthese chapters are summarized below. I n Mexico City the Majority WantFinancial Services- But Only a Minority are "Banked" 6. Three quarters of Mexico City's adult population has no bank account In the context of this work, financial services or dealings with formal financial refer to the wide range of products and services institutions. This i s almost twice the provided by financial institutions, including percentage prevailing in Brazil, a savings and checking accounts, credit and debit cards and other payment services, short term country with similar income and working capital and consumer loans, long term income distribution characteristics, investment capital for fixed asset purchases, and six times the percentage in the mortgage financing, sight and time deposits, US. While Mexico City's unbanked remittances, transfer services, and insurance population is composed of a products. The term unbanked refers in general to individuals without access to formal disproportionate number of financial services, although in the context of minimally educated and the statistical work conducted here, we defined economically insecure persons, it it more narrowly to apply to people without also includes a large fraction of the deposit accounts in a regulated financial middle class, more than half of all institution. managers and three quarters o f salaried workers, suggesting that fina cia1 exclusion is not a poverty issue pe se. Seen another way, bank account holders are concentrated in the upper 15% of the income distribution, but about half o f individuals in the top decile do not have bank accounts. At the same time, some 16% o f banked persons earn below median income, and hold 8% of the checking accounts and 7% o f the credit cards. This shows that even relatively poor households can be interested inholding bank accounts. IResults from a marketing survey carried out in a total of 45,000 households in Guanajuato, Monterrey and Mexico City show a markedly lower incidence of banked persons than the Mexico City survey. X 7. Yet many unbanked exhibit "bankable" behavior. The unbanked have an ability to accumulate assets since they account for almost three-quarters of Mexico City's homeowners and most of them borrowed - from friends, family or informal institutions -- in order to acquire these homes. And a significant number of them (28%) have financial savings -held outside formal institutions. This, combined with the comparison between urban Mexico and urban Brazil, does suggest that a significant proportion of the unbanked could, and should, move into the formal banking system. 8. Being unbanked i s costly. The study shows that being unbanked makes it difficult to accumulate financial savings in a way that maintains the purchasing power of those savings (the majority lose to inflation (the majority lose to inflation and many have additional costs.) Survey results suggest that the unbanked account for almost half of the financial savers in Mexico City and that close to a third o f the unbanked hold savings in cash. But it will not be easy for the unbanked savers to accumulate financial savings since their funds lose value with inflation and they may lose more without security measures. Being unbanked also means spending more to make payments - say for services such as utilities. Once a month unbanked persons must make the trip down to the water company, the telephone company, electric company (over 20 blocks away from home or work for 70% o f the surveyed population) to face longlines and pay by cash. 9. One unexpected finding of the survey i s that nearly four-fifths o f the unbanked are paid by check. This suggests that up to 80% of the unbanked may have additional costs in travel time to the bank which wrote the check or in commissions from a second bank of check-cashing agency. Finally, being unbanked makes it harder and more expensive to borrow. Commercial "lay-away" plans at stores or "popular" savings institutions like the "Caja's" charge from 6% to 30% per month. All told, an unbanked two-earner household may lose up to 15% of potential income in real and opportunity costs, a high price for families who tend to be among the poorer two thirds of the population to begin with. (See Box 2.3) Besides the financial costs being unbanked may carry additional costs for the poor in terms of self-esteem, or "disempowerment" to apply a currently popular concept. Focus group discussions with unbanked persons revealed that they frequently do not consider the possibility of usingbanks because they assume they would be rejected. As one person commented, "we just aren't the kind of people banks would want..." The sum of the costs, financial and psychological, of being unbanked i s be difficult to estimate precisely, but clearly adds to the burdensof the poor. 10. The unbanked surveyed were near unanimous in affirming their interest in having access at least to an interest bearing savings account or to formal loans, for a variety of purposes. Consumer loans, loans for health, education, and home improvement figured high in their lists. Families in the upper 25% o f income distribution mentioned car loans. Flexible lending, or consumer loans tend to be available more frequently directly from the provider than from commercial banks or from micro- finance institutions. xi 11. Poverty i s by no means the only reason for remaining outside the formal financial sector. Only 42% of the unbanked invoke a lack of lack of income as the main reason for staying unbanked. Others find formal sector financial products for reasons such as distrust, inconvenient hours or location (30%)2 A significant proportion (15%) of the unbanked report having held bank accounts in the past, but most closed it on their own volitions citing lack of income, high cost or lack of usefulness. 12. Both the banked and the unbanked groups reported using alternatives to formal financial services. The most frequently used source of credit, among both groups are commercial stores, followed by "other" (mostly informal, community based sources). A noticeable difference between lower and higher income groups, are that the lower- income groups mentioned friends as a third source of loans, while the upper income groups mentioned banks. For savings, the unbanked mostly rely on informal systems such as tandas (rotating savings pools) or simply keep cash at home. None of the Government's low income banking programs (postal savings program or BANSEFI) were mentioned b y people surveyed, possibly because these are more active in rural areas. 13. In sum, while the report does not attempt to empirically determine whether Mexico's high rate of financial exclusion is due to demand side or supply side issues, it does suggest that a lack of income or other demand side issues i s only part of the explanation. Other problems appear to center on the lack of appropriate products, on high prices, and possibly insufficient outreach. These issues are discussed in the context of a review o f the supply of financial services. Mexico's banking sector is characterizedby its shallowness and high costs 14. In comparison with peer countries, Mexico's financial sector is shallow and underdeveloped with financial assets representing only about 68% of GDP (a ratio that i s about half of Brazil's and compares poorly to peer countries or other emerging markets.) Inaddition, domestic financial depthhasbeen following anegative trend since the mid-nineties with the domestic credit to GDP ratio, has been falling since 1995, dropping to 26% by 2003. This places Mexico alongside Gabon, Latvia and Rwanda, ranking 105 out of 138 countries in terms of credit to GDP ratio (CONDUSEF, 2003.) Increased reliance on foreign borrowings and liability contracted abroad has allowed the Mexican economy to continue growing, but access to internationalborrowing i s mostly limitedto larger firms (World Bank, 2003.) 15. The commercial banking sector continues to be highly concentrated despite an increase in the number of banks and their assets have grown in Mexico since the Tequila crisis of the 1990s. In the 1980s there were a total of seven commercial banks, two of which held over 50% of the total market assets. By 2003 there were a total of sixteen commercial banks, two of which hold 51% of total market assets. This category includes those who simply tink they don't need banking services, who don't trust banks, who find the banks too far, with inconvenient hours or bad service or who do not know how to go about opening a bank account xii Only ten banks account for 100% of the country's savings accounts, 99% of checking accounts and 99% of term deposits. Many banks do not offer retail services. 16. The banking sector i s increasingly focused on the high-end market as evidenced b y the fact that bank revenues from investments and services i s growing while `traditional' deposit and lending i s in decline (De Sousa Shields, 2002.) The commercial bank's traditional portfolio - business and housing-related lending - has shrunk annually since 1997, both inabsolute terms and as apercentage of bank assets. According to the Mexican bank consumer protection agency, CONDUSEF, Mexican banks derived 40% of total income from service charges in 2002, up from a minimal percentage in the 1990s. As new technologies have been introduced, the number and variety of options for banking - instant credit, debit, transfers, account consultations, telephone banking, ATM banking etc. - have blossomed, as have investment services and options. 17. The reason behind the continued drop in traditional banking activities i s unclear. The decline in deposit and lending activity i s surprising considering the fact that the spread between active and passive interest rates i s high and growing. The Association of Mexican Bankers (ABM) argues that despite recent improvements in the legal framework, their recourses are still limited in case of default so that lending remains a risky business. Faced with such disincentives on the "traditional" side, it appears that the banking sector has found a profitable alternative between charging service fees and investing in government papers. Representatives of regulatory agencies point out that the commercial banks have little motive to expand their clientele or to returnto traditional banking services. Alternatives to traditional commercial bankingproducts are beginning to emerge, but with mixed success 18. Although most of the major commercial banks have developed products aimed at attracting lower-end or unbanked depositors, their impact remains minimal. Analysis of the products' terms shows them to still have high costs of entry and/ or maintenance, costs that are estimated to be beyond the reach of households below the richest 25% of the income distribution. The terms of commercial banks' low-end savings account programs are such that the average savings account inMexico (which holds US$64) would lose value over a year. More generally, Mexico's banks charge rates significantly higher than those found in the US, Canada and Spain.3 While checking accounts are unaffordable to all but the wealthiest 25% o f Mexico City residents, a fairly standard checking account in the U S (e.g. WachoviaFirst Union Bank in Washington D.C.) with no monthly fees, no minimumbalance requirements, and unlimited use of the bank's ATMs and check writing, i s accessible to anyone who can provide identification and an address. 19. The Government has taken some steps towards developing the social interest banking sector - cooperatives, credit unions, and savings and loan associations (SCIs) --but so Reforma May 28,2002 ... XI11 far this has had limited reach. This follows a campaign pledge of President Fox's to promote micro-finance and access to banking as part of a broad agenda to combat poverty. Thus far, efforts have included creating a regulatory agencyhecond tier bank, BANSEFI, as the primary vehicle for social interest banking. BANSEFI offers technical assistance and fulfills general regulatory functions to assist the development of the social interest banking sector. It also has absorbed the former postal savings offices, known as PAHNAL, of which they were 60 in Mexico City. However, BANSEFI and the social banking institutions are mainly rural. Their reach i s limited with 140,000 persons for each branch institution, particularly in urban areas. None of the 1500persons surveyed in Mexico City under this initiative mentioned either SCIs or the PAHNALoffices. While social banking sector products are affordable, they are limited to savings - with very low interest rates - for the former PAHNAL offices and to certain types of credit (mainly micro-business and housing) for the SCIs. 20. Other institutions are beginning to emerge to fill in the lending function the commercial banks used to fulfill. The most promising initiative however i s Banco Azteca, which began operations in January 2003, growing out of Elektra, a commercial outlet specializing in selling domestic appliances and furnishings on credit. Banco Azteca, whose offices are located inside Elektra stores, has profited from Elektra's previous knowledge of the sector and from its network of 850 stores nationwide. It offers unsecured personal loans up to MX$5500,charging 6 - 10%per month, compared to the average commercial bank active interest rate of 24% per annum. Banco Azteca also offers savings instruments with very low (US$5.00) minimumdeposits requirements and that pay annual interest rates ranging from (1%) for savings and up to 8% for time deposits. Banco Azteca's growth has been spectacular, opening 1,500 new accounts per day after its first six months. As of 2003, after one year, the bank counted some 1.75 million accounts nationwide and the peso equivalent of over a billion dollars in savings4, a strong indicator of a pent-up demand for financial services. Banco Azteca has already paid off its initial investment costs for equipment and physical plant. In addition, CONDUSEF management speculates that lenders - from pawnshops and neighborhood "prestarnistas" to Walmarts, Home Depots and used car lots -- are increasing their role. Lending by non-bank institutions has remained steady and grown slightly as commercial bank credit has plummeted. Finally, the department store SANBORN's has been offering limited savings programs. 21. Some new experiences suggest that low-end banking can be profitable on a wide scale. The most promising recent initiative Banco Azteca, which began operations in January 2003, grew out of Elektra, a commercial outlet specializing in selling domestic appliances and furnishings on credit. Banco Azteca, whose offices are located inside Elektra stores, has profited from Elektra's previous knowledge of the sector and from its network o f 850 stores nationwide. It offers unsecured personal loans up to MX$5500,charging charges 6 - 10% per month, compared to the average commercial bank active interest rate of 24% per annum. Banco Azteca also offers savings instruments with very low (US$5.00) minimum deposits requirements and Telepone interview with Luis Nino, Financial VP of Banco Azteca, February, 2004. xiv that pay annual interest rates ranging from (1%) for savings and up to 8% for time deposits. Banco Azteca's growth has been spectacular reaching 5,000 per day after its first six months. The bank now counts some 1.75 million accounts nationwide, a strong indicator o f a pent-up demand for financial services and, perhaps, that low-end banking could be a good business proposition. Banco Azteca has already paid off its initial investment costs for equipment and physical plant. CONDUSEF management speculates that other lenders - from pawnshops and neighborhood "prestamistas" to Walmarts, Home Depots and used car lots -- are increasing their role in the low-end market, undetected as commercial bank credit has plummeted. Commercial outlets, both small and large scale, have long offered savings and lending services, albeit directed toward purchase of their own products (lay-away plans, clubes and time payment systems.) 22. A large demand for financial services still remains unmet.It could be met if the social banking system expands, if the commercial banks reach down-scale, if non-bank agents, like commercial outlets, move into banking, or if a combination of these occur. While the profitability of "wholesale" banking (small scale operations, large scale clientele) has yet to be tested, experiences in other countries have demonstrated that business can become more profitable as the depositors and service users increase in number. However, the bank, like any other business, must have a good understanding of and customer relations with its clients to succeed. 23. The Association of Mexican Bankers (ABM) points to regulatory obstacles to banking in general and to low-end banking entry in particular. Having to deal with seven different regulatory agencies i s in itself time-consuming, costly and frequently wasteful for the banks. Regulatory costs run higher for "low-end" banks, since each transaction i s charged separately, regardless o f size. A bank which specializes in many small transactions ends up paying more than banks which focus on fewer high price transactions. Banco Azteca has mentioned that regulatory complications delayed its opening, and that costs of regulatory compliance limit its scope of action. It also cites other examples of how the regulatory framework i s not currently designed to deal with "low-end banking" and create additional costs for programs which serve the poor. 24. Besides regulatory issues, the ABM points to weaknesses inthe judiciary which make lending more costly. The time and litigation required to collect collateral on defaulted loans, particularly after the financial crisis of 1995, has made commercial banks argue for stronger laws and enforcement practices. On the other hand, this does not appear to be an issue for Banco Azteca. Instead, dealing with its low-income clientele, it consciously spends more resources and time on credit ratings, interviews and collection than do the highend commercial banks. 25. The strong sector inroads toward highend banking, however, suggest that it will take more than regulatory andjudicial reforms to stimulate banks to move to increase their clientele base or to stimulate non-bank institutions (such as Elektra or SCIs) to offer appropriate financial services on a wide scale. Additional interventions may be needed to create incentives or rewards for financial institutions, which promote xv programs accessible products and services. Among other areas for incentive building, representatives of the banking sector in Mexico have mentioned technology and education. The first involves taking full advantage of electronic banking. The second concerns the preconceptions and misunderstandings both the unbanked and the financial institutions have about one another. Examples from other countries have included rewards for increased low-end services (increasing multipliers for savings from certain client groups, or reducing risk factors for loans to others) or tying permission to open additional branches in high income areas to evidence of work in low income areas. Recommendationsfor Government and Reflections on Other Experiences 26. The research has demonstrated that the vast majority of the population o f Mexico City (and of major urban areas) does not have access to financial services, and, as a result, pays higher costs and enjoys reduced opportunities of asset accumulation. Research also shows demand exists, capacity to save, borrow and pay for financial services at considerably higher costs than presently offered. For its part, the commercial banking sector, the social banking sector and the commercial outlets appear to be positioned to increase their coverage to the unbanked. 27. On key approach involves information gathering, public disclosure, and financial education to measure the use o f formal financial services, to track unsatisfied demand, to educate the unbanked in advantages offered by financial programs and to promote dissemination of successful banking practices which increase access to the unbanked. Much of the success of the community banking movement in the United States has been attributed to the public disclosure of banking records and to public awards programs which demonstrated that the poor can be a very good business. The Mexican government has used information sharing to great advantage under the F O V I program. By reviewing and publishing comparative regulatory costs for housing in different states and cities, the Government enabled the private sector to know where its investments in home construction would have greater impact, and where it should pressure for deregulation. Information sharing and public disclosure could be used in a similar way to promote interest in offering financial services to .unbanked and low income families. The regulatory agencies should develop ways to assure the sharing o f information about financial services, both among clients and providers. Mexico already counts on an able agency (CONDUSEF) for analyzing and transmitting existing data. CONDUSEF's role could be bolsteredto include following and disseminating information about the unbanked and informal institutions, the costs and profitability o f different product lines for different income levels. It could also survey unbanked and low-income groups both individuals and SMEs, to determine which products do result in asset accumulation and inincreased income levels. 28. Education and financial literacy are also part of information dissemination. The suspicions, misconceptions the unbanked and the banked hold about one another form serious obstacles to increasing access to financial services. Programs to allay such fears and to educate the unbanked in financial basics have been undertaken by NGOs and community service agencies, frequently with government support, in the United xvi States. The Mexican government could push the access agenda forward considerably by promoting such efforts in mutual education between the unbanked and the financing institutions. 29. A second approach could be to promote the creation and marketing of specific products attractive to the unbanked and/or of importance for overall economic development. The demand-side analysis pointed to the interest held by the unbanked in savings programs with interest, in consumer and other unsecured loans, and in service payments, areas where neither the commercial nor the social interest banks have focused. O f course there could well be more areas where the unbanked would like to have access to financial service. However, it i s worth pointing out that making at least savings and service payments more accessible could be of interest to governments as well, and would not affect overall system risk. 30. A third approach entailsfinding ways to reduce the costs of basicproducts so as to make them affordable to more individuals. Better than price controls, which tend to reduce the supply of services, are increased competition, minimum standards and public disclosure which can also help reduce prices. Certain states in the U S require banks to offer "lifeline" or basic minimum services without stipulating prices. In others, consumer defense agencies publish comparative service charges as a way to encourage lower prices. Given the many ramifications of financial product costs (defined risks, required reserves, etc.) there i s a wide latitude for the Government to encourage financial institutions to lower prices for particular clients. 31. Fourth, it may be helpful to look at means of bringing financial and The US experiencewith geographic targeting banking services to specific unserved and unbanked Legislation to promote access to financial services in the U S began on a geographic basis, communities both to increase access mainly to break the common practice of and to support specific community "redlining" certain city districts -- refusing to development. The absence o f banks consider requests (for loans or other services) from low-income and minority from persons or businesses within an area neighborhoods may contribute to the characterized by the presence of low-income andfor minority groups. This was viewed, and circle of poverty, or it may be a demonstrated to be, a form of racial and income reaction to it. While the physical discrimination. The passage of the Community absence of banks in certain Reinvestment Act (CRA) in 1977 and other delegations i s not the main measures such as the creation of the explanation, it may help explain the Neighborhood Reinvestment Corporation aimed at ending this practice and at stimulating banks to high proportion of unbanked people. get involved in neighborhood development by At the same time, depressed lending for property improvement and neighborhoods and the goal of community proiects. community improvements have long been a concern to both the national and municipal governments. 32. Last, it i s important to look at ways of using new technologies to reduce costs and increase the banking presence or access in underserved communities. New technologies used in banking and financial services should open opportunities for xvii lowering costs and for reaching marginal areas. The example of Guatemala City's proposeduse of electronic transfers to facilitate payments to public service companies has already been mentioned. Combinations of change machines, payment machines and ATMs could overcome the problems mentioned earlier by bankers in reaching low income persons with ATM services. Banks tend to locate in high income neighborhoods, leaving low income areas underserved. `Mobile banking' services such as ATMs (or ATMs in vans) provides a cost effective way to expand service and reach areas where branches may otherwise not locate. xviii CHAPTER 1. ACCESS TO FINANCIALSERVICESMATTERS ! "To be unbankedis to be under an economic disadvantage." Joseph I.Lieberman, US. Senate, in support of legislation to supportIndividual DevelopmentAccounts and Electronic Funds Transfer. Introduction 1. The present report looks at issues of access to financial services in urban Mexico, drawing on existing documentation as well as on new research (surveys, interviews and focus groups) carried out in Mexico City during 2002. It is one of several initiatives in the region to explore the issue of financial exclusion: companion efforts include work on financial exclusion in urban Brazil ("Brazil: Access to Financial Services,"), in Bogoti, Colombia (ongoing) and in rural Mexico (Mexico's Financial Markets and the Rural Economy.) These initiatives differ in their methodology and emphasis, although all discuss the unmet demand for financial services, the characteristics of the "unbanked," and, to varying degrees, what can be done about financial exclusion. 2. The motivation behindthis interest in financial exclusion i s two fold. First, it i s well documented that being unbanked has costs (Dunham, 2001; Caskey, 1997; and Stegman 1999; Carr and Scheutz, 2001; Skricki and Fernindez, 2001; Rhine et al., 2000; Green and Lichter 1996.) It makes it more expensive to engage in a number of transactions (paying and being paid) and more difficult to save while maintaining the value of an asset As such, higher access to financial services is desirable from an efficiency point of view. Second, financial exclusion i s an issue that primarily affects the poor. Access to financial services can help with poverty alleviation, particularly to the extent that it encourage asset buildings and help cope with shocks and overcome liquidity constraints. The literature on development banking highlights low-income families' need for a wide range of financial services both for everyday life and for asset building purposes (Dichter, 1997; Caskey, 2001; Robinson, 2001; Rutherford, 2001). As a result, the U S and most developed countries have policies to encourage greater access to financial services for the poorer segments of the population. In developing countries, this i s also the case, although this i s typically done through micro-credit approaches (rather than through the mainstream banking sector), and often with a more rural orientation. 3. Financial exclusion i s of particular concern in Mexico, where banks are still reeling from the twin Peso and banking crisis of 1994 and 1995. With bank credit to the private sector amounting to only about a quarter of pre-crisis levels, most firms and households rely on non-bank institutions and informal providers (including suppliers, credits) to satisfy their demands for financial services. The persistent low levels of access to financial services by both low and middle income households has therefore become an important policy issue inMexico. 1 4. As to the present paper's focus on urban Mexico, it i s motivated by the fact that Mexico i s now three quarters urban, with cities accounting for about 85% of the country's wealth and half the country's poor. Thus, access to financial services in urban Mexico i s of interest both from a social perspective and from a growth or efficiency one. Data was unfortunately not available for urban Mexico as a whole, so the report mostly relies on the new data gathered in Mexico City. Note that since access i s likely be higher in Mexico City than in the other cities in the country, the report probably underestimates the problem of financial exclusion in urban Mexico. Nevertheless, the report does provide a good complement to the parallel work on the rural economy mentioned above. In both reports, it is found that the shallowness of Mexico's financial sector means that financial exclusion i s an issue for the poor but also for the middle class and entrepreneurs. 5. Note that this report does not attempt to distinguish systematically whether the unbanked stay away from formal sector institutions by choice, or whether they are spurned or prevented entry. We therefore use the term financial exclusion broadly, considering it synonymous with a lack of access to financial services. As to financial services refer to the wide range of products and services provided by financial institutions, including savings and checking accounts, credit and debit cards and other payment services, short term working capital and consumer loans, long term investment capital for fixed asset purchases, mortgage financing, sight and time deposits, remittances, transfer services, and insurance products. The term unbanked refers in general to individuals without access to formal financial services, although in the context of the statistical work conducted here, we defined it more narrowly to apply to people without deposit accounts in a regulatedfinancial institution. 6. Instead, this report focuses on two key questions: what are the benefits the unbanked could gain from using formal financial sector institutions; and what discourages them from doing so. The objective i s to help to identify measures to increase access to financial services in Mexico. As such the report i s organized as follows. The present chapter briefly reviews the literature on the importance of access to financial services and the experience of other countries in promoting such access. Chapter 2 presents an analysis of actual and potential demand for financial services, and offers a profile of the unbanked residents'of Mexico City. Chapter 3 then looks at the supply of financial services by the formal financial sector and possible explanations for Mexico's extremely low rate o f financial access. Chapter 4 concludes and discusses possible means of improving financial participation. Increasedaccess can lead to greater macro-economicgrowth, equity?and stability 7. In 1873 Walter Bagehot noted the important role of the financial sector in England's economic growth when the financial markets were able to mobilize savings to finance the implementation of new technologies in England. In more recent times, the modernist development theory implied a link between financial services and economic development (Rostow, 1960, 1962) with the argument that countries achieving an aggregate savings level above 5% o f GDP will reach a level of reinvestment leading to economic "take-off '. Subsequent empirical studies have 2 demonstrated not only that the development of financial sectors accompanies economic development (Goldsmith, 1969, King and Levine, 1993), but also that financial development predicts economic development. More recent work has convincingly demonstrated that deeper financial markets can help reduce aggregate volatility (World Bank, 2000). Certainly a formal banking system with widespread coverage i s a consistent characteristic of developed economies, while widespread financial exclusion i s associatedwith lesser-developed countries. 8. Links between access to financial services and growth are several. Rostow's argument that countries achieving an aggregate savings level above 5% of GDP will reach a level of reinvestment leading to economic "take-off" implies a certain link between financial intermediation and growth, insofar as it i s access to financial services which convert aggregate savings to aggregate investment. Savings kept at home (under the mattress or in a tin can) cannot contribute to macro-economic growth. B y the same token, unbanked entrepreneurs who cannot obtain credit also represent an opportunity cost to a national economy. Hernando de Soto has argued that access to capital and to financial services i s the key to economic growth both in advanced economies and in the developing world (de Soto, 2000). Other literature linksbetter income distributionto macro-economic growth. 9. Access to financial services can help promote equity, and can, in this context, be linked to better economic equilibrium and to faster economic growth (Stegman, 1999.) Recent research has centered on the impact o f access to credit on micro- economic development, suggesting that improving the access of micro-enterprises to financial services could have an important positive impact on a country's income distribution (Westley, 2002.) Surveys of the literature on financial intermediationand poverty reduction conclude that development of the financial sector contributes to economic growth and thereby to poverty alleviation (Holden and Prokopenko, 2001). It has been widely documented that access to banking services makes savings much easier for a majority of persons studied. (Stegman, 1999; Carr and Scheutz, 2001; Skricki andFernfindez, 2001; Rhine et al., 2000; Green and Lichter, 1996). 10. A number of recent studies have looked at the positive results stemming from access to a broad range of financial services and payment system services (Christen, 1999; Wenner and Campos, 1998). Reports from the development banking community and studies of international NGOs (Dichter, 1997), strongly suggest that lower income families need a wide range of complementary financial services both for everyday life and for asset buildingpurposes (Caskey, 2001;Robinson, 2001; Rutherford, 2001). 11. Access to saving programs can allow households to better build a nest-egg that i s available for emergency needs and/or facilitate entry into higher-risk and higher return lines o f business (Westley, 2002.) The difficulties of managing to save without access to banking services are widely documented. (Stegman 1999, Carr and Scheutz 2001, Skricki and Fernfindez 2001, Rhine et.al 2000, Green and Lichter 1996) Lacking access to formal savings mechanisms, low income households save in traditional ways, but do not get the full benefits o f their savings. Keeping extra animals, jewelry, participation in savings groups, purchases on an installment plan, 3 loans to extended family or simply cash under the mattress means putting one's savings at risk of loss or theft. 12. Literature on the access to financial services points to the high costs the unbanked pay for not having those services. This i s well documented inthe United States, where the unbanked rely on a variety of unregulated service providers, dubbed "fringe banking" (Dunham, 2001; Caskey, 1997; Stegman, 1999). It includes "cash-and-carry" methods, check-cashing outlets (CCOs), "pay-day lenders" and pawnshops. Such institutions charge much higher rates than mainstream financial institutions - partly because they face higher risks and costs, but partly because they operate in dispersed neighborhoods where limited competition and lack of regulation. The research in Mexico turned up similar costs associated with life outside the formal financial sector. (See Chapter 2.) It is Feasible to Broaden Access to Financial Services 13. Close to half of the wealthiest 20% of the income distribution inMexico does not use formal financial markets. At the same time, individuals earningbelow median income of 2 minimum wages make up 16% of all bank account holders in Mexico, belying the notion that the poor cannot pay for financial services. Altogether this suggests a large untapped market for the formal financial institutions. 14. This is consistent with research in Latin America and the US that suggest that a "wholesale" approach to banking- many small deposits and small loans making up in volume for a few large deposits - can be sustainable and profitable5. Lendingto low income groups i s already a sustainable and profitable business in Latin America. It has been estimated that 29% of all microfinance services inLatin America are already provided by commercial banks (Christen, 2000). The experience of Latin American banks and associations specializing in consumer and business loans for low income individuals and households, suggests that the lower income groups, with lower education levels tend to be least likely to default or be late-payers on loans - particularly if they are female headed households. Access to financial services can also be seen as an issue of civil rights and economic welfare. 15. Many developed countries have pursued policies to assure access to financial services at all income levels. Inthe USA, the responsibility of government to "bring its broad- based experience in capital markets and financial services to bear on the inner city" was underlined in 1997 by the then Secretary of the Treasury, Robert Rubin. Referring to the initiative that gave Treasury the responsibility of bringingunbanked welfare recipients into the mainstream banking system, Rubin declared. "If this works.. . it will not only give them a more efficient way to cash checks and access to other financial services, but it may also encourage people to save, to plan financially and, therefore, to improve their economic life over time." Notably the experiences of the Caja Social in Colombia, Megabanco in Colombia and Banco Sol in Bolivia 4 16. Mexico's government has for many years supported access to financial services in ruralareas. More can and needs to bedone there, but an urgent needexists also urban areas. Other compelling reasons for considering access to financial services a matter for government involvement, include the fact that the concentration of financial services among the wealthiest implies that a small group of high income earners are benefiting disproportionately from government expenditures in bank supervision insurance and bailouts and from government bond payments. 5 CHAPTER2. DEMAND FOR FINANCIAL SERVICES IN MEXICO CITY: THE MAJORITY WANT BANK ACCOUNTS, BUT ONLY A MINORITY HASTHEM 1. Less than 25% of Mexico City's adult population useformal sectorfinancial services. This i s half the proportion prevailing in Brazil, a country with similar income and income distribution characteristics, and a quarter the proportion in the US (figure 2.1). This chapter looks at possible explanations for this low coverage, using recent surveys of both households and firms, combined with qualitative information derived from focus group interviews (Box 2.1) The first section examines who uses formal financial services in Mexico and who doesn't. The second section looks at how the unbanked manage without financial services - how they pay and get paid, how they send and receive money, how they save and borrow. It demonstratesthat the costs of being unbanked are high. The third section examines why the unbanked remain unbanked, what types of financial service could be attractive to the unbanked and concludes with a discussion of whether there might be substantial unsatisfied demand for financial services inMexico. Figure 2-1 Use of Financial Services is Very Low inMexico; Compared to Other Countries 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% I Mexico D.F Brazil USA Source: Mexico- INEGI survey [Sample Size: 1198 Respondents]; Brazil-World Bank 2002; USA-Consumer Finance survey 1998 6 Box 2-1 New data to look at accessto financial services Few of the usual data sources (census, household surveys, etc.) collect information on the access to financial services. One survey recently commissioned by the Media Planning Group, (a consulting firm specializing in marketing), collected information on users of commercial banking services. This survey, carried out by BIMSA, covered a wide sample (45,000 households) in three cities (Mexico City, Guadalajara, and Monterrey). However, since its purpose was to collect information on people with access to banking services it is biased towards higher income groups and may not be fully representative of the population as a whole. The BIMSA survey was therefore complemented by another one, commissioned for the purpose of the present study with the Universidud Nacional Autdnornu de Mexico (UNAM). This second survey was conducted as part of the regular employment surveys carried out by the Mexican Institute for Statistics, Geography and Information (INEGI) and applied to a representative sample (interms of geographic distribution and socio-economic levels) of 1,500 households. Inaddition, a survey of small and medium sized businesses in Mexico City was commissioned with an independent survey institute in collaboration with CONDUSEF, the Consumer Council for Users of Financial Services in Mexico. Finally, focus groups and interviews were conducted as part of the basic research for this report with groups of low-income unbanked individuals to glean their perceptions of financial services, costs and needs in their own words 1. WHOUSESBANKS MEXICO WHODOESN'T? IN AND 2. Financial services, whether provided by banks, by cooperatives or by micro-finance institutions, reach a very limited percentage of Mexico's urban population. The INEGI survey found that less than a quarter of Mexico City's households use any type of financial sector institutions. The BIMSA survey confirms this, reporting that less than 15% of the adult population of Mexico's largest cities has any type of bank account. The implication i s that three quarters or more of Mexico's urban households are without any type of formal financial service. As shown in Figure 2.2, the more sophisticated the banking products become, the fewer people access them. Figure 2-2 Use of banking productsis low, more so for more sophisticated products Source: BIMSA Survey [Sample Size: 45,000 Households] 7 Who are the unbanked 3. The unbanked are notjust thepoor. They include a large fraction of the middle class - people whose income puts them inthe 6thto 8thdecile of income distribution. However, users of financial services are primarily the rich and the upper middle class-more than half (56%) of bank account holders belong to the top two deciles of income and less than a fifth (16%) belong to the group earning less than the median income of two minimumwages according to BLMSA (figure 2.3.) The INEGI survey reports similar findings: the bottom half of income earners hold only 20% of Mexico City's savings accounts, 8% of its checking accounts and 7% of it's the credit cards and time deposit cards. 4. Seen another way the data shows that even relatively poor households can be interested in holding bank accounts. It is also of note that about half (48%) of individuals in the top decile manage without bank accounts, suggesting that use of financial services i s not universal even among the rich. These findings, combined with the high proportion of unbanked, particularly as compared with other Latin American and OECD countries, raises the question of whether the unbanked are better off without formal sector financial services, or whether there i s an unsatisfied potential demand. Figure 2-3 Most bank account holders belong to the highest income brackets 8th decile 11% 7th decile I 36% Source: BIMSA survey [Sample Size: 45,000 Households] 5. Unsurprisingly, the unbanked have lower income and educational attainment levels than those with bank accounts. Figure 2.4 shows that the education o f those using formal banking products tends to move away from the survey distribution and towards more education -- this i s particularly true for checking account users. In comparison, the users of informal savings products such as `cash at home' and `tandas' have education distributions closely matched to that of the survey population. Having said that, the education distribution of savings accounts users i s fairly close to that of that of informal `tanda' and `at home' users. This suggests that 8 educational attainment may not be an impediment for accessing formal banking servicesfor those that have used informal financial services. 6. At the same time, the data reviewed above suggests that the unbanked are far from being a homogeneous group. While they tend to be poorer and less educated as a group than the banked, they also include individuals from high income groups. Around half of the top decile of income earners are unbanked, and around ten percent of those with elementary school education have checking accounts. Figure 2-4 Use of formal financial servicesincreaseswith education 'El 100% 2o8 90% 80% 5 25 E 70% 60% h .E 50% =2g 40% 30% 20% u 10% 0% niterate Unfinished Unfinished Unfinished elementary junior high senior high school school school Level of Education Source: INEGISurvey [Sample Size: 1198 Respondents] 7. A disproportionately higher percentage of the unbanked work in the informal sector - - 50% of the unbanked versus 44% of the overall sample. This makes sense given the 8. Lower income and education level of the unbanked. It is also in keeping with formal bank requirements which emphasize steady employment references from employers. Debit card accounts and "AFORE" accounts (the state required pension accounts) are among the most widespread formal instruments. The median income of AFORE users reaches the 75th percentile of income distribution (Table 2.1). Both debit card accounts and AFORE accounts are associated with formal sector employment. Table 2-1 Use of formal financial instruments increases with income Type of Financial Instrument Income group of median user (decile) Informal banking instruments: Tandas* 5 th Savings at home 6th Coop or "Caja Popular" 6" Overall .". __ Unbanked Sh - I "I " " _ I -I " _ ". " Formal Banking instruments: Savings, debit card or "AFORE" account 71h Checking account 91h Investment account 9" Overall Banked 8Ih * tandas are informal rotating savings club. Source: - INEGIsurvey [Sample Size: 976 Respondents] 9 9. The unbanked group tend to have lower income security than their banked counterparts, as evidenced by the comparatively higher incidence of independent workers,6 workers without pay and the economically inactive and by the lower percentage of managers and salaried workers among the unbanked group(Tab1e 2.2). That informal, or irregular workers should be unbanked could also be a function of the banks' preference for clients with steady employment, reflected in requirements for references and proof of work (see Chapter 3). At the same time, a full 56% of managers and three quarters of salaried workers are unbanked, suggesting that informality i s not a primaryreason for being unbanked. Mexico's banking sector may be overlooking a large, profitable and bankable market. Table 2-2 The unbanked are disproportionately independent workers and economically inactive, but they also include a majority of managers and salaried workers Professional characteristics of: Banked Unbanked % of group who are unbanked Manager 2.8% 0.9% 56% Independent worker 13.0% 21.7% 87% Piece worker 7.7% 5.6% 74% Salaried worker 71.5% 49.1% 73% Family worker without pay 1.4% 100% Non-family worker without pay 0.1% 100% Economically inactive 4.9% 21.0% 94% Other 0.2% 100% 100% 100% Source: INEGIsurvey [Sample Size: 1198 Respondents] 10. The unbanked have a high home ownership rate - 63% - thereby their capacity for accumulating assets. Although 10% of the unbanked populations surveyed are still paying for their homes, their overall ownership rate i s not much lower than that exhibited by the banked group (70% owners with 12% still paying for the house) despite their differences injob security and income level^.^ 11. The unbanked are more likely to consider their economic situation as having deteriorated in the last 5 years. When asked: "How does your economic situation now compare with five years ago?" the unbanked replied "better" only in 23% of cases and "worse" in 34%. The proportions were reversed for the banked: 36% replied "better" but only 26% "worse." Although this could be a reflection of the fact that the banked families tend also to be higher income earners, no pattern emerged relating economic situation*. On the other hand, savers, whether higher or lower 'Generallyhomeownership a euphemism for informal worker Data on across Latin American cities reveal similarly high home ownership rates, which may result from the easy access to informal settlements and self-built homes. * Personsearning over 10MWs - Le. over 90" percentile - did consider their situation had improved by a factor of 12 to one. Other groups, however, showed no particular pattern: From 70-90th the majority considered 10 income earners, banked or unbanked, tend to reply that their situation has improved over the past five years. This i s consistent with research from the U S that shows that savings tend to provide individuals with security and improve their well being.' To the extent that access to banking services allows for increased savings, this would be solid backing for the claim that financial services provide a form of empowerment and sense of participation inthe economy." 12. The literature on financial exclusion frequently equates banking with a public service. While certain public services (water, health, sanitation) are considered key to human survival, financial services are grouped with those considered key to economic upward mobility and community development (along with energy and communications.) In the focus groups the unbanked described their feelings towards the formal banking sector in terms of exclusion from the mainstream. Insofar as they can see themselves cut off from a basic public service they also view access to financial services as a form of empowerment. 2. HOW DO THE UNBANKEDGET ALONG WITHOUT FINANCIAL SERVICES? 13. To answer question raised earlier about whether the unbanked are better off without formal financial services, the survey looked also at how unbanked individuals manage their finances in Mexico City and whether they bear additional costs compared to people with bank accounts. The survey asked specifically: how the unbanked make payments on goods and services; how they receive and transfer money; and how they save and borrow. The research findings confirm that being unbanked imposes severe financial and transaction costs, and that the low penetration of banking services in Mexico i s far from being socially or economically optimal, at least from the point o f view of users - and the unbanked non-users of banking services. How do the Unbanked Payfor Goodsand Services? 14. Making payments for public services represents a high costfor the unbanked in terms of time, travel and servicefees. In Mexico City, where payment by mail i s virtually unheard of, bank payment services are a booming business - among bank account holders. The unbanked rely on payments made in person which imply substantial time lost in transit and waiting in line. For some 70% o f those surveyed, the public service company i s more than 20 blocks from both home and place of work. Close to half reported spending over two hours intravel time to make service payments. Inthe focus groups, 98% reported that banks in their neighborhoods accept payments for public service and public service bills (water, energy, solid waste removal, telephone, and municipal services-including property and vehicle taxes). Typically, however, banks charge the equivalent of US3.50 to non-account holders and US$ 1.50 to account holders for each payment. All of those interviewed preferred to pay in themselves better off; from 50thto 70th,the same; from 40thto 50th,worse; from 30thto 40th,better, from 20" to 30thworse and from 0 to 20thbetter. See Stegman 98, Barr 2001for examples as well as U.S. Federal Reserve and OCC publications. loSee an interesting interpretation in "Savings Incentives for the Poor", The American Prospect Magazine, May 2003 11 person, although the time spent traveling to and from the payment offices runs two hours on average, at an average travel cost of US$l.OO without counting time spent waiting in line. None o f those interviewed considered their time as an added cost. Instead they viewed their trips to the public service companies as overall savings. Illustration 2-1 The line to pay telephone billsinMexico City Waiting in line to pay public service bills "Z'd Rather Wait in ine than Pay the Bank's charge", Says One lady Waitingto Pay .->r TelephoneBill How do the UnbankedReceiveand Transfer Money? 15. One unexpected finding from the survey was that a full 70% of the population surveyed, banked or unbanked, is paid by check (Figure 2.5). The majority (88%) of those paid by check are unbanked. Where and how this group cash their paychecks, and how much time and cashing fees mean in terms of lost income i s a topic for a further survey. Banks have no standard policy on check cashing fees. Commissions run from 0% - 10%. Money-changers can charge up to 15 %. Being unbanked can entail a de-facto pay cut for those paid b y check. In contrast, only a third o f the banked are paid by check. Presumably they do not have to pay a surcharge for cashing or depositing their paychecks as long as they use their own banks. The majority o f banked individuals (55%) reported being paid through direct deposit in their bank account. 11 In view of what seems an obvious opportunity for banks to gain new customers, a leading official of Banco Azteca remarked that the institutionalization of payment by direct deposit - by government in particular - would go a long way to stimulate expanding access to financial services 12 Figure 2-5 Most people, including the majority of the unbanked, are paid by check (How people are paid inMexico City) 70% 60% 50% 40% 30% 20% 10% 0% Check Deposit Cash Other Source: INEGI Survey [Sample Size: 977 Respondents]. Note that a small percentage of the unbanked (6% of those surveyed) are also paid by deposit in a debivsavings account - these are usually established by employers inthe employee's name. Given their limiteduse ofthe account, this group do not consider themselves account- holders. 16. The vast majority of the unbanked would prefer to be paid cash, suggesting that they view checks and other bank payments as costly or inconvenient (Figure 2.6). The bankedpopulationprefers non-cash forms of payment by a factor of about two to one. Evidently the population already banked sees greater advantages in receiving payments through banks than do the unbanked. A small, but non negligible, share (9%) of the unbanked named direct deposit as a preferred method for getting paid. The most striking result i s the lack of interest or lack of confidence which the unbanked express for Mexico's banks and for modern payment systems. Figure 2-6 Most people would rather be paid incash -"(Preferred methods of being paid) Cash Deposl InaSavingsAcct Check Depost ina Checking Accl Source: INEGI Survey [Sample Size: 976 Respondents] 13 17. The survey could not shed much light on the issue of remittances - international or domestic -- possibly because they are much less important in Mexico City than inthe rest of the country. Only 2% of household's surveys had received remittances from abroad and 1% from elsewhere in Mexico so responses on costs of such transactions could not be considered representative. The surveys did show the unbanked population to be greater users of remittances than the banked by almost two to one. Additional information was gathered through interviews with banks and remittance services such as Western Union.12 I 18. The unbanked pay more to send and to receive remittances. Western Union - admittedly one of the more expensive money senders-- charges 12% to the sender13 and another 7% to the receiver to send US$100 from New York to Mexico City, and to change dollars to Mexican pesos. The process implies an additional step (andcost) for the unbanked, who cannot have remittances deposited in personal accounts, but must claim them in person. Whoever sends such a remittance has to advise the receiver, either by phone, mail or personal message to give the order number, since the remittance company does not offer this ~ervice'~.An email alert i s the cheapest method, for those with internet access. The rest usually must add the cost of an international phone call. Finally, the unbanked must count the cost of travel to and from the bank, of waiting in line and claiming their remittance, before receiving the cash. In contrast, for people with bank accounts the flat fee for transferring money from a US bank to a Mexican bank can be as low as $9 payable by the sender and costing the recipient nothing.15 In some cases the bank announces the transfer by email to its account holder. The art of remittances i s evolving but continues to favor bank-account holders: Banamex-Citibank has just announced a new service for Mexican migrants to the US that includes a magnetic card with a passkey number for families to use for electronic deposits and withdrawals. The card i s to cost $3.00 and presumably will be re-usable. How do the Unbanked Save? 19. Only 28% of the unbanked hold some form of financial savings (which they keep outside of formal institutions.) Nevertheless, the unbanked still account for close to half (47%) of all financial savers. A rough calculation based on the answers to the questions "how much did you save in the past twelve months?" concludes that unbanked in the Distrito Federal have saved roughly 60% in informal mechanisms what the banked have accumulated in savings and checking accounts (US$500 `*Remittances ~ ~~ between banked persons can take the form of inter-account transfers or wires, while the unbanked rely on specialized companies (mainly Western Union and Wells Fargo) and bank wire. While an informal network for remittances certainly exists, from hand-carried cash to dollars sent through the post, the formal systems do not include any micro finance or "popular" banking entities as yet. l3BANAMEX charges a flat US$60 for sending any amount up to $1,000 - a costly rate for small amounts. Further study should advise how other institutions scale remittance charges and whether they are reduced or increased as the amount sent increases. l4This is curious since Western Union, a major provider of remittances, traditionally specialized in delivering personal messages. l5Banks used for this example were the Bank Fund Staff Federal Credit Union and Banco Santander Serfin. 14 million). Interestinglyenough, the unbanked are not the only ones who use informal saving mechanisms. A full 40% of the bankedalso rely on them. Box 2-2 Informal SavingsInstitutions Tundus are rotating savings pools that operate as both loans and savings instruments. A tanda might work as follows: a group of 20 people agree to contribute MX$200a week to a common pool for 20 weeks. Each week the proceeds are given to one member of the pool (who doesn't contribute that week) so that those at the beginning o f the pool effectively receive a loan which they pay off equal monthly installments until the end of the cycle. Those who receive money at the end of the cycle have effectively been saving with each pool contribution until the final withdrawal. From anecdotal evidence it appears tandas are usually made up of 5 to 20 "partners" each contributing US$10 to US$20 on a weekly or bi weekly basis. The prevalence of tandas in low income communities provides stark evidence that the poor have both the capacity and the willingness to use financial services, even with small amounts. Some quotations from tanda participants follow: ``My tanda is sacred. I cannotfail to make a payment because it hurts the others"; "Iparticipate only if I know I can fulfill my obligation. I'd rather not eat than fail to make a payment"; "My savings are small but drop by drop they make a puddle. In the bank they want us to save large quantities and they askfor a lot of papers. Iprefer my tanda." Clubes (included under "other" in figure 2.7) are managed by commercial stores and follow the same scheme as the tanda in which a group of individuals make regular and equal payments, in this case to the store and in this case the "pot" is not cash, but a specific article for sale. The articles may vary from person to person, but the amounts paid are usually the same. Clubes select the weekly or bi- weekly winner in different ways, but most involve a social meeting and a game of chance (door prize, bingo, etc.) Cujus de uhorros are similarly formed in groups of 15-50 members associatedby a common place of work or through a church. Unlike the tanda, however, a caja usually requires a full year's commitment after which members can withdraw their accumulated savings with interest, depending on the Caja's earnings throughout the year. Cajas also make loans to members and to non-members (when recommended by members). Since the Cajas are not regulated there i s no way to discover the range of interest rates they use, active or passive. However, a sampling of Cajas in the DF turned up deposits ranging from US$10 - $20, active interest rates from 5%-6% per month for members and 10% for non-members, and, in the most profitable cases, passive interest paid annually at rates up to 30%-40%. Besides the Cajas examined here, an unknown number o f Cooperutivus exist. The Cooperatives tend to offer interest on savings at a somewhat lower rate than the "Cajas". Systematic information is not available. The percent of the population surveyed which listed Cooperatives, Credit Unions or Micro Finance Institutions was too small to be considered significant on a city-wide scale. The faGade of a Cooperative "Caja" appears in illustration 3-2 below, with the services offered: Member's and Children's' Savings Plans Personal and Business Loans Fixed Term Investments Checking Accounts Financing for Funerals Consumer Loans 15 20. The Tundu, described in Box 2.2. i s the most widespread form of savings reported in Mexico City. Tundus appear under different names in most Latin American countries and also in West Africa under the name tontines and they turned up in 20% (both banked and unbanked individuals) of the responses to the question "how do you save?" on the INEGI survey. (See Figure 2.7.) Cajas de Ahorro and Clubes constitute somewhat more developed and more institutionalized forms of the tunda. These informal savings and loans systems rely on personal trust and confidence in other people, either co-savers, managers of the "cuju" or the informal borrower. The Tandus and Clubes pay no interest on deposits and there is, consequently, a certain opportunity cost involved for the member. The Cajus and informal loans can pay interest, indeed the first sometimes pay more than a commercial bank, but according to the focus groups they vary. A good Caja or a good borrower i s hard to find and a good Cuju i s not easy to join. People take their Tundus and cajas extremely seriously and their comments suggest that they frequently equate savings groups with extended family or friends. "The tunda. It's part of my life. I save. I gab with my colleagues and I enjoy myselfu lot." Figure 2-7 The most commonly report savingsinstrument is the informal tanda, followed closely by cash at home 25% Tanda Debit Card Saving At Home Savings Acct Caja de Ahorm Checking Other Investment Directed Saving Caja Solidaria Account Account Type of FinancialService Source: INEGISurvey [SampleSize: 740 Respondents.]Sample includesbothbankedandunbanked(only those who save). Answers are not mutually exclusive. Cajasolidaria are inthe process of beingregulatedand as such it could be debated whether these are formal or informal. 16 Illustration 2-2 Facade of the "Caja de Ahorro Econom'a Solidaria, A.C." Iztapalapa, MCxico D.F. 21. Some forms of consumption can also be thought of as savings in the case of the unbanked. Two which the survey did not deal with are the purchase of lottery tickets and the use of lay-away plans at stores. Lottery tickets can be thought o f as consumption expenditures, since they represent a form of gambling, and gambling i s most often equated with entertainment. At the same time, if the lottery ticket i s considered an investment in an unlikely and delayed gratification, then its purchase - usingthe difference between income and consumption - can be thought of as a form of saving for those who have no other options. Lottery tickets are purchased widely by both banked and unbanked, but the motives, entertainment versus investment opportunity, could be very different. A study o f who buys lottery tickets and how much purchasers spend as a part of their income could provide interesting information inthis respect. 22. Lay-away plans (buying on installments), like the "Clubes" can be seen as a form of directed savings used by stores and retail outlets. However, lay-away plans usually convert from savings to a loan at the point when a down payment or agreed upon amount has been paid to the store. The customer then takes the article in question and continues to pay the remainder as a loan. The customer usually pays a final amount in excess of the article's list price. A 30% mark up on list price i s standard when a lay-away plan i s used. Advance payments can cover anywhere from 25 - 50% of the marked-up cost. A monthly 15% interest charge on outstanding debt i s also standard for the credit portion of the plan. So rather than earn interest on savings, customers pay an additional amount for the deposit service. Layaway programs exist because they can count on a client group with a low level o f financial literacy, or simply with very few alternatives. This study did not look at the percentage of banked or unbanked who use these programs although, in the authors' opinion at least half of the economically active urban population has used lay-away plans irregularly, and close to a quarter rely on them regularly. 17 23. Of the various informal savings mechanisms used by the unbanked, only the ``cajas" can offer positive returns to savers -- very good returns in fact. However, no systematic information is available to allow a calculation of their real earnings or of their rates of failures. The unbanked who save by other means, (the tanda, home savings, and the informal "no-interest" loan) fare considerably worse. Providing their capital i s not lost or stolen, it loses its purchasing power with inflation, representing an opportunity cost. As to lay-away plans, they actually have negative interest rates. Insum, it is very difficult for the unbanked to accumulate financial savings in a way that maintains the purchasing power o f those savings. How do the Unbanked Borrow Money 24. Only about 10% of those surveyed had requested loans from formal sector institutions in the past 36 months. Half of this group belonged to the top 25% of the income distribution. It would seem a good guess that the rich request more loans from formal sector financial institutions than the poor because they enjoy better chances o f being accepted. Perhaps for that reason, borrowing - at least through the formal banking sector - seems more common among the rich. (17% of the upper 25% o f income distribution took out loans inthe past three years versus 8% from the lower 75%.) Figure 2-8 Commercial stores make up the primary source of credit in Mexico City-Secondary sourcesvary with income levels Commercial Shop Others Friend Bank Credit Card Credit Union State Government AFORE SOFOL Savings and Loan Money Lenders Caja NGOs Source: INEGI Survey [Sample Size: 1301 18 25. The unbanked do borrow however, as do persons earning within the lower 75% of income distribution. The most common source of credit by far proved to be commercial stores - providing 38% of all loans reported regardless of income level. The second most frequent source of loans, listed as "other" on the survey, refers to a variety of informal sources and appears to be evenly distributed among income groups. In third and fourth place, however, (banks and friends) there is a notable divergence between income groups. The higher income group favors banks while the 75% lower income group relies on friends. Referring to other sources, the lower income groups appear to be almost exclusive users of Credit Unions, Savings and Loans (both informal organizations) and Money Lenders. The high income group appears to be sole users of NGOs, Cajas and the principal users of AFOREs and SOFOLes. Table 2-3 Non-bank credit terms are expensive US$15 Bi-weekly Stores offering `lay- "clothes 30% Weekly US$10to One to three away 'plans *home furnishings US$50 months "sports goods Chain Stores Consumer durables 15% Weekly US$lSO to Six months consumers loans Electro-domestic us$looo items, Building materials "Cajas" 6% to 10% Twice US$lOOto Three to six Popular Savings Open or three US$500 months Associations times a Ye= Source: Martha Ferrer 26. Borrowing costs morefor the unbanked. As Table 2.3 shows, non-bank credit terms are tough; interest on loans outside the formal banking sector runs from four to twenty times higher than the average for commercial bank lending rates in 2002 (24.2% per annum - see Table 3.4). Even bank-issued credit cards which charge 0.66 to 1.66% per month (8 to 20% per annum), are less expensive than the non-bank systems, with the exception of the Cajas. Cajas can offer relatively attractive lending rates, but tend to have rather low loan ceilings and are hardto find. 27. What did people do with their borrowed funds? Responses varied considerably depending on whether or not people have bank accounts (see Figure 2.10). Among the 75% in lower income groups, the majority (60% or 45% of the total) borrowed to purchase or to renovate a home, as did a similar majority (65%) among the unbanked, while the principal purpose o f credit among the top 25% as among the banked 19 population, was consumer goods, which also accounted for 23% of the total. Considering that financing is readily available from commercial stores and i s also promoted by many public agencies for housing, borrowing for housing and for consumer items could well be supply-driven to a certain degree. Figure 2-9 What do people borrow for? It depends on whether or not they have bank accounts 70% 60% 50% 40% 30% 20% 10% 0% Buy or Purchase Investment Health Automobile Others Renovate Goods House Purposeof Credit 28. In sum, the unbanked populations manage to save and to borrow without formal banks and finance institutions, but they pay a higher cost than their compatriots with bank accounts. Box 2.3 describes the costs of financial services for a low income family inMexico City. The estimates conclude that apoor unbankedfamily could pay as much as two full months of salary or 15% of household income to cover its financial costs. 20 Box 2-3 What it costs a MexicanFamily to be Unbanked Rogelio is an informal vendor of candiesand other snack foods indowntown Mexico City. He and his wife (who works in a factory) makeroughly the same income: 2 minimum wages or MX$80per day which tranSCItes to around MX$2000a month each. This putsthem near the 40th percentile of Mexico City's' wage earners. Rogelio's wife has to make the family's paymentsas Rogelio can't leave his stall unattended for extended periods to go to make payments: Time and Travel Costs. Cashing checks- Rogelio's wife cashes her paycheck twice a month at abank costing her 4 hours and MX$lOin travel cost. PavingBills- Rogelio's wife also pays electricity, water and telephonebills during the work week (offices closed on weekends): 7.5 hours of waiting inline and 4.5 hours of travel once a month for all 3 offices for a total of 12hours lost plus MX$10 intravel cost. Yearly Total: for lost time i s MX$1920 Yearly Total: for travel cost i s MX$240 An Informal Loan-Typicalloanfor Rogelio i s a month and a half of salary or MX$30002.At most he uses one loan every year for a consumergood. MX$3000paid in6 months has a rateof 30% for the period in aMexicanbank inMay 20033.The market stall vendor rate i s 50% per month. If Rogelio takes 2 monthsto pay the loanhis cost i s MX$900 at the bank and MX$3000through a vendor. Yearly Total: An opportunity cost of informality of MX$2100. Unrealised Profits - Rogelio's Average Daily Sales are MX$200 and his Average Cost: MX $120. Cost with unrealised 10% discount for buying inbulk with more funds: MX$108. BulkBuying Profit minus his actualprofits amount to MX$92 -MX$80 =MX$12 per day or MX$312per month (26 days as he works Saturdays). Yearly Total: An unrealized Profit of MX$3744. Loss of Intereston Savings.Assume Rogelio saves one month of his salary per year under the mattress.A bank's minimum interest rate in 2003 i s 1%per year. IfRogelio had savings in Banco Azteca's time deposit account with a one monthterm then the interest rate would be 5% per year in 2003. Saving ina tanda for 3 separate terms (each time he saves MX$2000) will give him enough for the opening balance(MX$5000 assuminghe is able to put away this large sum of money). His lost interest earnings would be: Yearly Total: Keepingjust the minimum balance, yearly interest would pay MX$250 that he would not get ina tanda. Adding this up implies aYearly Cost of: MX$7254for Being Unbanked. This i s equivalent to nearly two full month's income for Rogelio's family or around 15% of their total annual household income. 1. Income distribution in Mexico City i s higher than in the rest of the country. Thus on a national level Rogelio's family would be higher on the income distribution scale. Therefore the costs of being unbanked for a family of medianincome on anational scale are likely to be a greater as apercentage of total family income. 2. As apoint of comparison, Bital and Bancomerbanks offer credit limits at 3 months of income for their lower income clients (e.g. Anticipo de Nomina Bital). 3. See Banamex.com 21 3. WHY ARETHEUNBANIUCD,UNBANKED? 29. Close to three-quarters (72%)of the unbanked have never even attempted to open a bank account, mostly because they felt they did not have enough money or because they objected to the very high minimum balance (Figure 2.10.) Many others simply don't find banks attractive -either because of the low interest rates they pay (10%) or because of high fees (6%) - or because they are difficult to get to (too far, inconvenient hours.) A non negligible fraction (7%) of the population either feels it doesn't need banks, or that banks are not trust worthy. These factors appear to be similarly represented across income groups, although the richer 25% tends to be more sensitive to cost factors (fees and low interest rates) and the poorer 75% to a lack of resources and to intangibles such as a lack of trust in the bank and bad service. Surprisingly a lack of knowledge on how to open a bank account i s involved equally frequently across income groups. Figure2-10 Unbankedpersonsdon't attempt to open bank accountsbecauseof highcosts Doesnt have enougn money 5: Very high minimumbalance Low interests rates Doesn't need it Didn'ttrust in the bank Highfees The bank is too far tElwhole sample OAbove 3 M.W. Didn't know how to open 0Upto 3 M.W. Other Bank hours inconvenient Bad service Missing basic documentation 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% i Source: INEGI Survey [Sample Size: 4961. Answers are not mutually exclusive. 22 Table 2-4 Why don't you have a bank account 30. About 15% of the unbanked have held accounts in the past. Of those who had closed down accounts, 25% Never attempted 72% explained that they could not maintain a minimumbalance It was rejected 3% (high maintenance cost), 23% said they didn't have you cancelled it enough money (a mixture of those who have little to save 13% and those who find cost requirements prohibitive), 22% Bank c ~ ~ e l l ite d 2% did not need the account, while 10% blamed low interest Other 10% rates, and 8% mentioned high fees. This suggests that more than half dropped out of the formal system because it Source: Inegi-Survey did not meet their needs. Interestingly, only a fraction of the unbanked reported having been rejected by the banking system either when they applied (3%) or having their account closed by the bank (2%.) 31. Physical distance does not appear as a majorfactor. Only 5% of the unbanked listed location as the main reason for not having a bank account. This suggests that physical inconvenience i s not a driving force in limiting usage of banks. Location could, however be a factor for potential outreach, as discussed in Chapters 3 and 4. Location in the neighborhood could be correlated with use of Cajas de Ahorro and Cajas Solidarias, but this i s inferred from a very small measurement - only 9% of the survey which use informal savings organizations. The data do certainly suggest that difficult locations can discourage the unbanked from establishing relationships with the formal financial sector.16 32. Discussions in focus groups on the perceptions of formal banks showed that the question why they have never approached a formal bank appears moot to many of the unbanked, given that banks appear to have no interest in them. While it has not been studied systematically, the banks in Mexico City do not seek clients actively as compared to U S cities, where banks advertise on the radio, at the cinemas and in newspapers. In contrast, the sector seems extremely introverted in Mexico" In addition, many of the less well-to-do unbanked react to the formal and forbidding aspect o f Banks, which they associate with the wealthy, the well-dressed and well- heeled. '' 33. The focus groups gave their reasons for not using banks: high commissions, impossible requirements, non-negotiable terms and conditions; low interest on deposits, insecurity and bad treatment. The last point came up repeatedly. A comment made in one focus group may well speak for many lower-income unbanked. When the discussion leader asked one group member why he hadn't gone to a bank when he l6No one mentioned it in a focus group, but it is likely that the absence of physical presence of banks in low income neighborhoods contributes to the unbanked's feeling of exclusion. Officials from Banco Azteca -see Chapter 4 for details - credit its presencein low income neighborhoodsand "barrios populares" for in large part for its success. l7This claim has been disputedby several who report that Mexicanbanks are doing aggressiveradio campaigns for clients l8Note that in the US, making banks less forbidding by making the locale less luxurious, and by avoiding ties and suits among tellers, is a key part of communitybankingstrategies. 23 needed a loan, laughter filled the room and one voice spoke up: "Don't you see how we look, Compaiiera? We just aren't the kind of people the banks would want...." Many group members made clear that their negative reactions towards banks were a function of the fact that they were anticipating bank rejection before the fact. Why should they consider or wish for what they knew lay beyond their reach. "I have no proof o f my income," said a street vendor. A plumber corroborated, "What can Iuse for a guarantee or a reference?" What kind offinancial services would the unbanked want? 34. When asked what kind of deposit account they would like to open, the majority of the unbanked persons mentioned savings accounts. This preference i s confirmed by the fact that among the very small group of unbanked who had attempted to open bank accounts in the past, most tried to open savings accounts. At the same time, a steady percentage of unbanked persons (almost 20% in total) named checking accounts, investment, debit and deposit certificate accounts as their first choice, indicating that certain unbanked persons have knowledge of different deposit models and forms of investment. Figure 2-11The unbanked want savings accounts What Sort of a Bank Account Would You Like to open ? Source: INEGISurvey [Sample Size: 44 Respondents] 35. As to the kind of financial service most desired by the unbanked, deposits and loans are about as popular, named by 41% and 43% of the unbanked respectively while payment services are o f most interest to 15%. Within the deposit category, savings with interest were by far the most popular choice (26% of the total unbanked) followed by deposits with a debit card and checking account^.'^ Those who named lending services as their first choice, opted between consumer loans, mortgages, loans for health, loans for education and for social insurance. While the last picked up fewer than 2% of the votes for loan products among both banked and unbanked, the l9Notably, 41% of those already banked mentioned savings with interest as their first choice, suggesting that the minimumbalance required for interest payments i s beyond many account holders. 24 first four shared group preferences almost evenly. Payment services referred to payment for utilities and government services, credit cards and remittances. The credit card was most in demand here (led, again by the banked population who may well see a credit card as a more possibility desire than do the unbanked), followed by public payment service. Altogether the survey suggests that there may be substantial pent-up demand for credit: eight times as many people indicated that they would like to obtain credit financing than acknowledged taking a loan in the previous three years. Figure 2-12 The unbanked want loans for other things than business investments 25% 20% 15% 10% 5% w,. cmsumer loan mortgageloan loanfar health loanfor education wial insurance Source: INEGISurvey [Sample Size: 339 Respondents] 4. CONCLUSION THEREIS SIGNIFICANT ROOM - FOR THE FINANCIAL SERVICE MARKETTO GROW 36. Three quarters of Mexico City's adult population has no bank account, nor any dealing with formal sector financial institutions. While this population is composed of a disproportionate number of poor, minimally educated and economically insecure, it also includes a large fraction of the middle class, more than half of all managers and three quarters of salaried workers. So are the unbanked "bankable"? They certainly have an ability to accumulate assets since they account for almost three- quarters of Mexico City's homeowners and most o f them borrowed - from friends, family or informal institutions -- in order to acquire these homes. And a significant number of them have financial savings. The evidence suggests that a significant proportion of the unbanked could indeed move into the formal banking system. Brazil, with a similar level of income and an even worse income distribution than Mexico, has half of its proportion of unbanked. 25 37. Being unbanked i s costly. The study shows that being unbanked makes it difficult to accumulate financial savings in a way that maintains the purchasing power of those savings (the majority lose to inflation and many have additional costs.) Survey results suggest that the unbanked account for almost half of the financial savers in Mexico City and that close to a third of the unbanked hold savings in cash. But it will not be easy for the unbanked savers to accumulate financial savings since their funds lose value with inflation and they may lose more without security measures. Being unbanked also means spending more to make payments - say for services such as utilities. Once a month unbanked persons must make the trip down to the water company, the telephone company, electric company (over 20 blocks away from home or work for 70% of the surveyed population) to face long lines and pay by cash. 38. The unbanked finding o f the survey i s that nearly four-fifths of the unbanked are paid by check. This suggests that up to 80% of the unbanked may have additional costs in travel time to the bank which wrote the check or in commissions from a second bank of check-cashing agency. Finally, being unbanked makes it harder and more expensive to borrow. Commercial "lay-away" plans at stores or "popular" savings institutions like the "Caja's" charge from 6% to 30% per month and informal. All told, an unbanked two-earner household may lose up to 15% of potential income in real and opportunity costs, a high price for families who tend to be among the poorer two thirds of the population to begin with. (See Box 2.3) Besides the financial costs being unbanked may carry additional costs for the poor in terms of self-esteem, or "disempowerment" to apply a currently popular concept. Focus group discussions with unbanked persons revealed that they frequently do not consider the possibility o f usingbanks because they assume they would be rejected. As one person commented, "we just aren't the kind of people banks would want..." The sum of the costs, financial and psychological, of being unbanked i s be difficult to estimate precisely, but clearly adds to the burdensof the poor. 39. Despite their limited experience, the unbanked surveyed were near unanimous in affirming their interest in having access at least to a savings account which paid interest or to formal loans, for a variety of purposes. Consumer loans, loans for health, education, and home improvement figured high in their lists. Families in the upper 25% of income distribution mentioned car loans. Flexible lending, or consumer loans tend to be available more frequently directly from stores than from commercial banks or from micro-finance institutions. 40. Both the banked and the unbanked groups reported using alternatives to formal financial services. The most frequently used source o f credit, among both groups are commercial stores, followed by "other" (mostly informal, community based sources). The lower income groups mentioned friends as a third source o f loans, while the upper income groups mentioned banks. Only persons from the top 25% of income distribution mentioned using NGOs or savings cajas. For savings, informal systems "tandus" and savings at home are the most frequent systems for both banked and unbanked. No one mentioned the postal savings program or BANSEFI.In general, it appears that people prefer community-based systems, or agents they know well, 26 judging from the lead that commercial stores and informal systems have on formal sector banking institutions, including micro-finance programs. 41. Problems appear to center on the lack of appropriate products and on their high prices. Considering the additional costs the unbanked pay for financial services, the lack of information also appears to be a factor in limiting access, from the client's side. However, being unbanked cannot certainly be considered a pure demand side issue. Chapter 3 turns to the supply side, to look at the extent to which the limited penetration of formal financial services stems from the banks' own strategies. 27 CHAPTER 3. WHAT FINANCIALSERVICES ARE AVAILABLE IN MEXICO CITY, AT WHAT COST AND WHY DON'T THEY REACH MOST MEXICANS? 1. This chapter discusses how supply side factors may explain the limitedpenetrationof formal financial services in Mexico City. The first section describes the financial sector in Mexico, noting that it i s much shallower than in other comparable countries, with banks still reeling from the Tequila crisis of the mid-nineties. Their conservative outlook is reflected in their products, which, as discussed in the second section, are geared towards a high-end clientele. The third section examines recent initiatives which show how attempts are being made to do business with a lower end clientele. The last section concludes with a discussion of the legal and regulatory frameworks, their effects and how it might limit expansion towards a larger clientele. 1. MEXICO'S FINANCIAL SECTOR SHALLOW ANDSTILL INSHOCK? - 2. Despite boasting Latin America's second largest financial sector, Mexico's financial depth is shallow. Financial assets represent about 68% of GDP, a ratio that i s about half o f Brazil's and compares poorly to peer countries or other emerging markets (Table 3.1). The credit to GDP ratio has been falling since 1995 to reach 26% by 2003.20 This places Mexico alongside Gabon, Latvia and Rwanda, ranking 105 out of 138 countries interms of credit to GDP ratio.21 Table 3-1 Mexico's financial market is shallow compared to peer countries (1999/2000) Total Financial Assets*Private Credit / GDP Total Financial GDPper capita 1999 / GDP (%) (%) Assets* (US$m) Atlas Method ( US$) Brazil 139.7 50.9 780,739 4,350 Argentina 81 24.2 231,166 7,550 Mtxico 68.3 32.3 386,366 4,440 India 92.8 36.4 437,293 440 China 199.4 128.9 2,153,7 17 780 Malaysia 370.9 144.8 331,302 3,390 Korea 267 140.3 1,091,499 8,490 Simplified definition base on aggregate of assets described in this table. Source: World Development Indicators; IFS; IFC -Emerging markets Database *'CONDUSEF web page 21 From CONDUSEF - Considering Mexico's economic growth of the past decade, some regulators speculate that the role of informal and non-bank lendershas grown to a significant, albeit unmeasured dimension. 28 The Formal Banking Sector 3. Modernization of the commercial banking sector over the past decade has opened more opportunities for high end customers, thanfor low and middle income groups. Following the country's opening up to international trade and particularly to international investments in 1998, most of the commercial banking sector has been bought out by transnational financial groups.22 BANORTE, owned by the Monterrey Group i s the only remaining commercial bank with a majority o f Mexican nationals as stockholders. New banks with foreign owners appeared in Mexico City, increasing the total number of banks, but not the total number of clients. A period of mergers and acquisitions has reinstated the traditional pattern of concentration in Mexico's banking sector. In the 1980's there were a total o f seven commercial banks, two o f which held over 50% o f the total market assets. B y 2003, there were a total of thirty- two commercial banks, two of which hold 51% of total assets. Today, the two largest banks account for half of the banking sector's assets (table 3.2). The majority of banks do not offer retail services. Only ten banks account for 100% of the country's savings accounts, 99% of checking accounts and 99% of term deposits. With one branch banking office for every 12,000 persons, Mexico City has one of the lowest bank coverage rates in the hemisphere. 4. Significant regulatory reform has taken place in Mexico since the Tequila Crisis o f the mid-90s, leading to sector modernization as well as to more prudent lending practices (See Box 3.1). However, the reforms appear to have mostly encouraged the development of "high end" products (mutual funds, derivative trading and, recently, purchase of secondary mortgages.) None of the regulatory reforms aimed to help banks downscale their services. 5. Following the Tequila Crisis, the regulatory system was strengthened and broadened with the establishment of new institutional responsibilities. The Banco de Mkxico i s responsible for regulating the fees that commercial banks can charge but i s not the only regulator. In addition, the banking department of the Ministry of Finance and Public Credit (SHCP), the National Commission for Banking Securities (CNBV) also set banking policy and regulations. The Institute for Bank Savings (IPAB) offers deposit insurance, but supervises bank reserves and exposure. Finally, banks must report their earnings, assets and liabilities, products and prices to the CNBV's consumer protection branch, CONDUSEF. 6. Sector reform also23sought to include formal, semi-formal and non-bank financial institutions under government ~upewision.~~ A new legal and regulatory framework for non-bank financial institutions was enacted in 2001. The "Law on Savings an3 Credit Institutions" (Ley de Ahorro y Crkdito Popular, or LACP) defined three types 22 Examples are Banamex, bought by Citicorp; Bancomer by the Spanish Banco Bilbao Vizcaya Argentaria; Serfin and Santander Mexican0 by Banco Santander Central Hispano; and Banco Internacional by the Hong Kong Shangai Bank. 23This Section is based on "Mexico's" Financial Markets and the Rural Economy", Sanchez, 2003. 24This does not include tandas and cajas solidarias or cajas de ahorro, which are informal institutions discussed in chapter 2. 29 of institutions: the Sociedad Financiera de Objetivo Limitado (SOFOL), Credit Unions and savings and loans associations (SLAs). The last two make up the so- called "popular" sector of savings and credit institutions (SCIs). SOFOLs are strictly credit institutions. When first opened they specialized in mortgages and loans for housing construction, but they have since expanded to include automotive and other consumer credit. There are fewer SOFOLes than other non-banking institutions but they handle more assets. Table 3-2 Mexico's two largest commercial banks control 50% of the market Market share (%)* Assets in millions of Profit reported for U S dollars 2003 BBVA BANCOMER 28.83 43,422 19.7 BANAMEWCITIBANK 22.66 38,284 35% SERFIN/SANTANDER 8.24 14,289 15% VITAL 9.97 14,123 5% SANTANDER MEXICAN0 8.30 11.936 6.7% BANORTE 6.74 10,396 3% SCOTIABANK INVERLAT 5.99 8,929 6% INBURSA 1.82 5,175 (0.05%) J.P. MORGANtCHASE) 0.44 1,759 (0.25%) BANCO AZTECA 0.31 200 (0.03%) TOTAL 93.31 154,158 Source: UNAMStudy from: CNBV website July, 2002 "Market share estimated as the ratio of bank deposits to total deposits. 7. In contrast to SOFOLs, SCIs and credit unions, which are regulated by CNBV's department for "non-banking institutions", Cajas Solidarias and Cooperativas are non-bank financial institutions that are legally recognized but still unregulated. Cajas Solidarias operate mostly in rural regions. \ 30 Box 3-1Regulatory and legal framework for banks Economic reforms enacted in the early 90s such as the reprivatization of the commercial banks, the dismantling of interest rate controls, the updating of financial sector legislation, the elimination of capital controls, and the easing of reserve requirements motivated significant growth in the Mexican financial system, particularly among commercial banks. However, the lack of a strong regulatory framework and poor supervision led to a deterioration of commercial banks portfolio quality and eventually to decapitalization (OECD, 2000). The system's weaknesses were dramatically exposed during the 1994-95 peso crisis and further exacerbated by the regional economic crisis in Asia and South America, leading ultimately to a credit liquidity crisis whose consequences are still being felt. A new round of sector reforms immediately following the Tequila crisis focused on sector modernization. Legislation in 1996 lead to the creation of a derivatives market to help banks hedge currency risks. Congress also passed a securities market law to establish better corporate governance of publicly traded companies and to strengthen minority shareholders rights and share issue transparency. Both laws were intended to deepen financial markets, increase capital liquidity, promote economic stability, and, most important to the financial institutions, to increase access to capital. InDecember of 1998,Congresspassedanother reformpackage to recapitalize the banking sector and to encourage more prudent lending practices. Key provisions of the reforms included the removal of foreign ownership caps on commercial banks and the creation of the Institute for Bank Savings (IPAB) (responsible for selling off banks' non-performing assets in the wake of the peso crisis). In September 1999,the CNBV, responsible for oversight of the financial industry, instituted stricter capitalization and provisioning rules on all commercial bank operations. Regulatory advances included the passage o f dual bankruptcy and secured lending laws in May 2000, which removed a pro-debtor bias in the judicial process. Enacted in December 2000, the full impact of this law has yet to be felt. In addition, the creation of a credit bureau in 1997 has helped commercial banks to better manage credit risk, although there is a debate currently as to the appropriateness of having one single credit bureau and its ability to serve the needs of a lower end clientele. Further regulatory advances facilitating seizure of collateral and property have been made law in 2003, to the satisfaction of the Association of Mexican Banks. A recent decision allows banks to purchase and to sell debt. I t could lead to new partnerships between "popular" savings and loan institutions and commercial banks, both of which are now permittedto trade credits. It could also open the way to the formation of a secondary mortgage market. Source: World Bank, 2002 (This section i s taken from the Mexico FSAP. For a more rigorous review of regulatory reform in Mexico's financial sector, see Sanchez, 2003) 8. The regulated non-bank financial institutions (SOFOLs, credit unions and SCIs) claim some MX$22billion, of outstanding credit, or 8.6% of the total credit issued by the financial sector.25 The SOFOLs' credit line for automobile loans has grown at rates averaging over 95% annually since 1995, suggesting a strong latent demand for consumer credit. This i s confirmed by reports of 40% annual growth in the consumer credit lines of credit unions during the same period. Since they are unregulated, no data i s collected on the size or penetration o f Cajas solidarias and cooperatives. 25World Bank-NAFIN 2002 31 Table 3-3 Non-Bank institutions differ as to their degree of formality Source: CondusefWebsite 2003 and own estimates 9. Together with the SOFOLs, the Credit Unions and SCIs are regulated by the department of CNBV for "non-banking institutions" according to the LACP. This law was designed to strengthen the SCI sector by reinforcing its supervision and incorporating all types of SCIs into the same regulatory framework. The law adopts a functional approach to regulation and supervision, focusing on the functions and services provided by the different institutions rather than on the charters under which they operate. The LACP envisions the organization of SCIs into a 3-tiered system, coupled with BANSEFI's branch network, with an auxiliary supervisory approach and a private deposit protection mechanism. The SCIs will be grouped into federations and federations into confederations. Ten SCIs are needed to create a federation and five federations are required to form a confederation. This tiered structure will facilitate the implementation of an auxiliary supervision scheme and the creation of a private insurance fund. Although the CNBV maintains its supervisory authority, the LACP allows using federations to enforce the legal framework through a Supervision Committee authorized by the CNBV, acting as auxiliary supervisors on behalf of the CNBV. The LACP establishes a deposit protection mechanism (Fondo de proteccidn y saneamiento). Excluding an initial contribution, the insurance fund i s not government-backed and will be funded by annual fees paid by licensed SCIs. 10. These reforms have led to the creation o f the National Financial Services Bank (BANSEFI), a new development bank, in charge of promoting implementing an ambitious program to develop the SCI sector. The Organic Law for BANSEFI was approved in April 2001. This law establishes BANSEFI as a national development credit institution in charge of promoting the SCI system. BANSEF126was created on the basis of the previous government-owned Trust for National Savings (PANHAL). The law gives BANSEFI three tasks: (i) promote savings from among low-income to populations, investing those savings in a conservative investment portfolio (mainly government bonds), (ii) to become the "Central Bank" of the popular banking system, and (iii) to coordinate the Federal Government actions oriented to promote the orderly development of the SCI system. BANSEFI plans to do the treasury and liquidity 26 The World Bank is supporting the strengthening of the SCIs through a Savings and Credit Sector Strengthening and RuralMicrofinanceCapacity Building Projectof US$64.6 million, approved in July 2,2002. 32 funds management services for the SCIs, develop new products, manages remittances transfers (from workers abroad), develop databases, and carry out technical studies and investigations oriented to promote the development of the system. BANSEFI focuses preferentially on first-tier savings mobilization, operating in areas where SCIs do not have branches and product development. BANSEFI also coordinates training and advisory services for the Federations and SCIs in their process of adjusting to the requirements of the LACP, including their possible transformation, merger or liquidation. These reforms have led to the creation of the National Financial Services Bank (BANSEFI), a new development bank, in charge of promoting implementing an ambitious program to develop the SCI sector. The Organic Law for BANSEFI was approved in April 2001. This law establishes BANSEFI as a national development credit institution in charge of promoting the SCI system27. BANSEFIwas created on the basis of the previous government-owned Trust for National Savings (PANHAL). The law gives BANSEFI three tasks: (i) promote savings from among low-income to populations, investing those savings in a conservative investment portfolio (mainly government bonds), (ii) to become the "Central Bank" of the popular banking system, and (iii) to coordinate the Federal Government actions oriented to promote the orderly development of the SCI system. BANSEFI plans to do the treasury and liquidity funds management services for the SCIs, develop new products, manages remittances transfers (from workers abroad), develop databases, and carry out technical studies and investigations oriented to promote the development of the system. BANSEFI focuses preferentially on first-tier savings mobilization, operating in areas where SCIs do not have branches and product development. BANSEFI also coordinates training and advisory services for the Federations and SCIs in their process of adjusting to the requirements of the LACP, including their possible transformation, merger or liquidation. BANSEFI's annual budget for 2003 was 1,066 million pesos, around $125 million dollars*. As of August it was showing a cost overrun, having executed 1,084.4 million pesos or 101.7%of the total yearly budget.28 11. Inaddition, SCIs receive training and technical assistance from BANSEFI, created by the 2001 law out o f the former postal savings program (PAHNAL). Charged with incorporating all "Cujus" and Cooperutivus with over 250 members as Credit Unions or Savings and Loan Associations, BANSEFI now counts a total of 350 SCI and Credit Union offices in its directory29.BANSEFIhas also absorbed some 550 former postal savings offices throughout the country.30 These are due to be converted into independent non-bank financial institutions by 2005. At that point BANSEFI will become a second floor reserve bank for the regulated non-bank financial institutions. This "social banking sector" is described in greater detail in Section Three, "New Initiatives for IncreasingAccess". 27The World Bank is supporting the strengtheningof the SCIs through a Savings and Credit Sector Strengtheningand RuralMicrofinanceCapacity BuildingProjectof US$64.6 million, approved inJuly 2,2002. *'The comparableagency for channeling finance to agro-businesses,FOR A is assigned7,163 millions of pesos or around US$ 700,000. As of August 2003, FIRA had executed 30% of its budget (information form CONDUSEF. 29The 350 offices represent branchoffices of some 10nationalagencies 30formerly part of the NationalProgramfor Savings- PAHNAL 33 Box 3-2 Globalization could have an effect on domesticlending A Mexican Government official tells of inviting the newly arrived executive director of a foreign owned bank out to lunch. When he proposes a restaurant 10 minutes drive away from the bank buildings, the executive director offered a counter-invitation at the Bank's executive dining room, explaining that his kidnap insurance provided coverage only at his home, his office and along the route between them. With this restriction the executive director can't drive through Coyoacan, San Angel or other well-off neighborhoods in Mexico City. The Government official remarked that such limits could have an impact on the bank's lending patterns. Sector Trends 12. The assets of Mexico's commercial banks grew by almost 50% over the past five years while both deposit and lending activities have decreased (Figure 3.1.)31 The commercial bank's traditional portfolio - business and housing-related lending - has shrunk annually since 1997, both in absolute terms and as apercentage of bank assets. Banks derived 34% of total income from service charges in 2003, up from 31% in 2002 and from a minimal percentage during the 1 9 9 0 ' ~ . ~Another important ~ percentage of bank income i s currently drawn from earnings on investments, Government papers above all. The possibility that government paper i s "crowding out'' other investment and lending activities has been discussed in connection with Latin American banks in general. (De la Torre, Leipziger and Gasha, 2002; Steiner and Barrajas 2002). That financial development in Mexico i s below that of other developing countries would not be so worrisome if it were not for the fact that the trend over the last five years has been negative. Domestic credit to GDP in the country declined from 30.2% in 1997 to 20.9% in 2002. Similarly, over this period, M2to GDPfell from 28.3% to 21.7% (Sanchez 2003). Figure 3-1 The Decline indeposits and lending by commercial banks has not affected GDP growth Q) Q) 40% 360 n E 30% 340 m v) r E 3 320 5 ie 8 0 20% 10% 300 E n 0% 280 Q 1997 1998 1999 2000 2001 Year Source: "Domestic credit" and "GDP" from WDUGDFdata. "Savings Captured" from CONDUSEF data. 31De Sousa Shields, 2002 32Banamex show income from commission growing at 20% annually- Reynaldo Re, "Valores y Dinero" in El Economista, Oct 22, 2003 34 13. The simultaneous growth of the commercial banks' assets and decline of their traditional deposit and loan activities suggests that the banking sector i s shifting to new product lines. Over the past ten years the number of services offered by commercial banks has grown considerably. As new technologies have been introduced, the number and variety of options for banking - instant credit, debit, transfers, account consultations, telephone banking, ATM banking etc. - have blossomed, as have investment services and options. The new emphasis on service fees has also drawn a new client profile, the investor as opposed to a depositor or borrower (See Box 3.3). Box 3-3 Highend" products are growing Following on the reforms of the 1990's, all of the 14.Bankers interviewed explained the major commercialbanks have introducedmutual funds new emphasis on fees for services as andother high net worth client productsand services to capture the upper-income market. The average a reflection of their efforts to set minimum opening balance in mutual funds, for prices which reflect real costs. example, is MN$20,000 (US$2,000), well beyond the Representatives o f the Mexican reach of low-income earners. The mutual funds grew 84% from 2000 to 2001, representing one segment of Bankers Association explain that the savings sector to be on the rise. Industry experts paying interest on small savings and predict that the trend will likely continue as the on checking accounts amounts to a segment continuesto benefitfrom new regulations, low returns on traditional savings instruments, and cross-subsidy. In the interests of increased promotion. Mutual fund growth rates for transparency, bankers explain, the 2002 are projectedto range from 30% to 50% . new regime of charges for withdrawals, deposits, bank statements, etc was adopted. Analysts from the Bank of Mexico, however, question the reality of the Source: De Sousa Shields, 2002 presumed costs, pointing out that the marginal costs of electronic transactions are close to zero, and that service charges appear to reflect what the market will bear given the limited competition inMexico associated with the increased sector concentration. 15. The continued drop in deposit and lending activity i s surprising considering the fact that the spread between active and passive interest rates i s high and growing (See Table 3.4). The Association of Mexican Bankers argues that despite recent improvements in the legal framework their recourses are still limited in case of default so that lending remains a risky business33. On the other hand, representatives of the regulatory agencies suggest that between service fees and government papers (CETEs - defined as risk free investments), commercial banks have found a highly profitable alternative business and have little incentive to expand their clientele or to return to traditional banking services. 33 See discussion in Box 3.1 and in Section 4 35 Table 3-4 The spread between passiveand active interest rates is high in Mexico,.. and growing Year Average Passive Average Active Average Spread Interest Rate Interest Rate 2001 6.5% 18.6% 12.2% 2002 6.4% 24.2% 17.8% Source: CONDUSEF 16. A related question concerns whether other institutions may have stepped in to fill in the lending function the commercial banks used to fulfill. Borrowing from abroad has partially at least compensated for the drop in domestic lending, but this i s only affecting larger firms. It i s also possible that entities other than the banks have begun to respond to demand.34 CONDUSEF management speculates that lenders - from pawnshops and neighborhood "prestamistas" to Wal-marts, Home Depots and used car lots -- are making up the difference. Figure 3.2 shows that lending by non-bank institutions has remained steady and grown slightly as commercial bank credit has plummeted. In any case, the decline in traditional deposit and lending activities in commercial banks, and the upsurge in non-bank institutions reflects findings described in Chapter 2 that commercial stores, friends and informal institutions are the primary sources for credit. Figure 3-2 Are non-bank financing institutions replacing commercial banks as primary lenders ? 70 Non-Bank Finance Institution 94 95 96 97 98 99 `00 `01 `02 Year Source: CONDUSEF GeographicDistribution Banks 17. Banks, not unexpectedly, locate disproportionately in higher income areas. InMexico City, the delegations with the largest informal settlements (such as Iztapalapa, Contreras, Madero, Carranza, Milpa Alta and TlBhuac) are also those with the least 34Villareal and Carreno, 2001 and 2002 and in various interviews 36 number of banking institutions per person (Figure 3.3.) Consultation with the Association of Mexican Banks confirmed that banks analyze economic activity in the areas where they are considering opening new branch offices, and are drawn, in particular, by the presence of transnational corporation headquarters and possibly a "high end" market. Thus it is no surprise to find a high concentration of banks in the downtown business districts, and very few banks in low-income residential areas. Iztapalapa, a poor neighborhood of Mexico City, boasts the largest wholesale market area (Central de Abastos) in the Federal District but has only 6 bank branches inside the entire area and around 16,000 residents per branch. At the other extreme, the high income neighborhood of Bosques de las Lomas has 30 branch offices [Ruiz, 20031. 18. About 92% of the persons with bank accounts claim to have a bank branch or ATM located within half an hour of their home while only 72% of the unbanked do. This findingis also reinforced bythe survey analyzed in Chapter 2. The relative scarcity of banks in low-income neighborhoods reduces convenience factors and increases the costs of doing business with banks.35 Figure 3-3 Banks select locations according to income Rather than population Y 40,000 2C 35,000 30,000 ;25,000 Q 20,000 w- 15,000 0 5 10,000 f 5,000 zS 0 0 10,000 20,000 30,000 40,000 Median income of the neighborhood formal" neighborhoods Neighborhoodswith more informalsettlements @ Source: UNAM survey 35Unless banks are conveniently located, or accessible through electronic and other methods, there i s little advantage to paying bills, cashing checks or even saving at a bank. 37 Table 3-5 Location, Location, Location- % of respondentsliving within Y2hour of financial service Respondentswith % of All Type of Financial % of Bank Account % of UnbankedWith Service Holders With Service in Service in Neighborhood Neighborhood Service in Neighborhood ATM 46.8% 36.8% 39.1% Bank 44.7% 34.7% 37.1% Money Exchange 18.7% 13.5% 14.7% Credit Union 2.5% 1.5% 1.8% Caja de Ahorro 6.7% 5.0% 5.4% Money Lender 4.2% 5.3% 5.0% Source: INEGISurvey 19. In sum, commercial banks in Mexico have been modernizing rapidly over the past decade, enjoying growing profits and attracting international investments. At the same time, they have become increasingly concentrated in terms of number and clientele. Having reduced traditional bank operations based on deposits and lending, they have begun to offer specialized investment services for a higher end clientele while investing on their own, particularly in government papers. As to non-bank financial institutions, a growing number has moved from informal status to become regulated or at least legally recognized, with the support of the Government's BANSEFI, but most are only active in rural areas. Although they are growing, the non-bank financial institutions have failed to fully compensate for commercial banks reduced lending, which seems to stem from a "crowding out" given alternative sources of income. Nevertheless the general decline in credit appears not to have affected GDP growth, leading analysts to speculate that commercial stores and informal lenders have been stepping in. 2. PRODUCTSAND COSTS A FOCUSONTHEHIGHENDCLIENTE - 20. The Commercial banking sector is focusing increasingly on electronic banking servicesfor an upscale clientele. Less than half o f the commercial banks in Mexico City offer deposit account services and nine hold all of the savings deposits. But the majority of commercial banks offer ATM banking (19), Internet banking (17) and telephone banking (16) and the number i s growing. Figure 3.4 below shows the major services offered by the leading ten banks, which, as the data show, hold 98% of the accounts in Mexico City and over 95% of the branches and personnel. The remaining 16 banks work with less than 1% of the total accounts36. 36Butquite possibly with a higher percentage of the total assets. 38 Figure 3-4 Financial servicesoffered in mexico City by he biggest5 banking institution ...specialization inproduct typesand lack of competition 80% 1 0Number of checking accounts HNumber of savings accounts 0Number of term deposits Number of credit cards BITAL BBVA Banamex Santander Serfin (Citibank) Bank Source: CONDUSEF, July 2001 21. Judging from the number of banks which survive in Mexico with no retail -services (deposit or checking accounts, credit cards or payment services), high end investment services are profitable, although the highest profits, indeed the only two figure profits, are reported by the banks with large retail portfolios. (See Table 3.2.) As Bankers do not calculate the management costs or profitability for particular products so it i s impossible to compare profitability for high and low end profits. Proxy indicators for efficient management, the ratio of accounts to personnel, of accounts to bank branch, of personnel to bank branches and of personnel to assets show no particular advantage for one type of service or another. Checkingand SavingsAccounts 22. Most of the "low-end" products are out of reach of low-income families. Table 3.6 shows that checking accounts, term deposits and credit cards vastly outnumber savings accounts in Mexico City. Savings accounts, containing on average the equivalent of US$64 are the lowest amounts. Checking accounts are the most widely used and hold, on average, the equivalent of US$3,000 while term deposits average US$28,000. These amounts suggest that the savings accounts are more likely used by low income families. (Recall that median personal income for Mexico City hovers around $lOO/month.) Most checking account and term deposit holders could be expected to earn above the upper 10 - 15`h percentiles by virtue of the account sizes. Of the ten banks in the market only Banco Serfin appears to specialize in savings deposit services, with 45% o f its own accounts held in savings, making for 72% of all savings accounts in Mexico City. The two largest banks (BBVA Bancomer and Bana~nex~~)and one of the smaller banks (Imbursa) hold more term deposit and credit 37Which hold 60% of all credit cards in the city. 39 card than other kinds of accounts. The rest of the banks, with the possible exception o f Scotia bank, which reports that 15% o f its accounts are for savings, do most of their business with checking accounts, followed by term deposits. 23. All together, the five largest banks hold 95% of savings accounts in Mexico City. Among their array of savings and deposit plans, each offers a "popular savings" program, purportedly aimed down market. Table 3.7 shows the costs of these plans, including minimum initial deposit, minimum balance and monthly fees. Serfin, with 72% o f all savings accounts, charges one peso to open an account. But all banks require proof of identity, of address, and of employment as well as a myriad of forms to fill out which can intimidate lower income persons and close the door on informal workers. 24. The average account loses value under the special plans for low-income or small savers. Savings accounts, most frequently mentioned by unbanked and low-income persons as the financial service most desired, offer limited real gains to the user. In this sense they resemble the informal savings plans described in Chapter 2. Small accounts do not accrue interest and the cost o f maintenance can eat up savings. Table 3.7 shows what would become of an account of MX$642 (US$64 - the average savings account in Mexico City) after a year. One bank would have refused to open the account, one would close down the account for failure to maintain a minimum balance, in a third case the account would lose about 15% of its value to maintenance fees. Only in one case, BITAL, would the savings maintain their nominal worth, perhaps even winning a lottery - but losing to inflation. (Tanda Bital pays no interest, but runs a lottery to award prizes to selected savers.) This could explain why persons who can afford to, put their savings into term deposits, mutual funds and other instruments 25. Charges on checking accounts vary depending on the product and on the bank. All banks require a minimuminitial deposit except for Serfin, which requires a symbolic one peso to open the account, as in its savings program. Beyond that, customer obligations include keeping a minimum balance, paying an annual charge for administration, a charge for each check written and penalty fees for bounced checks. 26. Most people couldn't afford to open a savings or checking account at most banks. To determine affordability, the review assumed, perhaps optimistically, that a person could set aside one third of monthly income to make an initial deposit, could pay up to one percent of monthly income for maintenance fees and could maintain half o f monthly income as a minimum balance. Reviewing the products of the five leading commercial banks in Mexico, savings appear to be more "affordable" than checking accounts. But for those on the lower 50% o f the income distribution curve, only one bank, however, (BBVA Bancomer) would pay interest, assuming that savers maintain a month's income in their account. For all those in the lowest quartile o f income distribution, savings in a bank account would represent a net cost rather than a benefit, as it would for most savers in the second quartile. The survey results reported in Chapter 2 showed that the majority of unbankedpersons claimed too highcosts as the reason why they did not use banks. This supply analysis confirms their claim. 40 Table 3-6 Evenbasic bankingproducts are costly and inaccessibleto much of Mexico City's population Product Costs Type of Chargefincome BBVA and Affordability Group Banamex Banca Serfin* Bancomer Bital Savings Product Name Invermatico MaxiCuenta Libreton Tanda Bital Minimuminitial deposit $1,000 $1 $750 $500 Monthly fee $7.19 $9.58 $14.85 -0- Costs of Basic Interest paid on amount $2,000 $3,000 $750 No interest* Savings Products Over. MinimumBalance $1,000 $1 $750 -0- Balance after a year for Cancelled 527 ($115 lost) Could not open $642 an average balance over the year of MX$642** Bottomquartile No Yes No No [MX$1,547] SecondQuartile No Yes Yes Yes Affordability O f [MX$2,380] SavingsAccounts Third Quartile Yes Yes Yes Yes [MX$3,703] 4th Quartile [$11,619] Yes Yes Yes Yes costsO Initialdeposit $2,000 $1 $1,000 $2,000 f Basic Monthly fee $14.39 $12.46 $16.77 11.5 Checking Products MinimumBalance $2,000 $5,000 $1,000 $2,000 Charge per check $28.75 $28.75 $28.75 $28.75 Bottomquartile No No No No AffordabiliQ SecondQuartile [MX$1,547] Of No interest No No No No Checking [MX$2,3SO] Accounts Third Quartile [MX$3,703] No No Yes No 4th Quartile [$11,619] Yes Yes Yes Yes Bottomquartile [MX$1,547] No No No No Affordability of SecondQuartile Interest Checking [MX$2,3801 No No No No Accounts Third Quartile [MX$3,703] No No No No 4th Quartile [$11,619] No No Yes Yes Source: own estimates form data of INEGI and CONDUSEF (Amounts in 2002 MX$.Based on 12/2002data) "2% of savings accounts inMexico City are with Banca Serfin. See Table 3.6 ** Average Savings Account Balance in Mexico City 41 Box 3-4 Strategic thinking of banks and the absenceof low-end products "The formalfinancial sector has beenpoorly advisedfrom many quarters, including people who: e Advise thatformal institutions cannot provide microfinance profitably because of the high transaction costs the institution would have to bear. Think that low-income people are uneducated and backward, and so unable to participate in theformal financial sector. e Assume that low-income people cannot afford commercial loans and so require government or donor-funded credit subsidies (thereby insuring that demand remains unmet)." MargueriteRobinson. "The Microfinance Revolution" As part of the research on the unbanked, the consulting firm Enterprising Solutions interviewed executives from several commercial banks in Mexico to explore current and potential strategies. Interviews found that the majority of banks are reluctant to move resources into the low income markets because of persistent concerns about the ability of small and microbusinesses and low-income individuals to accumulate savings and to repay loans. Strategically they do not see a profitable business due to the operational costs of serving such clients, and the lack of know-how to develop appropriate products, improve client selection, streamline client processing, and control transaction costs and risks. They are content to remain serving a small high income clientele while finding ways to cut costs (e.g. fewer staff and more electronic banking), raise services charges and develop new and innovative products for the existing market To "downscale" a commercial bank needs a clear understanding of the target market. Certain banks which have carried out market research in the sector have determined that low-income clients are interest rate sensitive to the degree that they prefer small fixed payments over variable interest rates. They also have a unique credit risk profile: very limited disposable income, change jobs often, typically with a non-income generating interval, and possess a small asset base from which to respond to family emergencies. Thus, a commercial bank interested in the low-end clientele must invest more time and resources to maintain payment vigilance - reflected in higher interest rates. Low- income client relationships also require communications skills. Banks which have begun popular savings programs train account managers to use colloquial language when speaking with low-income clients and to explain products and services using simple words, and diagrams. The recently launched Banca Azteca (linked to the Elektra store chain) has already staked its future explicitly on the low income market, and early results show important market gains and demonstrate that the strategy of building access to banking services for low income groups with a consumer credit history can be viable. The bank is particularly suited to expanding financial services to lower income clients (particularly in the urban market) because of its labor intensive interaction with clients and its widespread outlets boasting a benign, unsophisticated decor. Other bankers have taken note. As one stated rather offhandedly- "I'm worried about Banca Azteca, though I'm not sure why." Each bank that targets the low income segment has had to compare the perceived and actual risks of reaching out to the urban unbanked markets against the potential for profits and the compatibility with its existing strategies, product lines, infrastructure investments, and overall approach to growth. Most banks lack knowledge of the sector, do not see a favorable comparison between risks/costs and profitability and pursue incompatible strategies. The rapid growth of Banca Azteca shows that there is an important, potentially profitable niche for financial institutions in the low income market. 27. According to the same assumptions, checking accounts are also unaffordable to all but the wealthiest 25% of Mexico City residents. To put the Mexicanbank costs into perspective Table 3.8 applies the same analysis to the costs of a typical U S checking account (e.g. from WachoviaLFirst Union Bank in Washington D.C.) With no monthly fees, no minimum balance requirements, and unlimited use of the bank's ATMs and check writing, the account would be accessible to all Mexicans who can provide identification and an address. 42 28. Mexico's banks charge rates significantly higher than thosefound in the US, Canada and Spain3*. Table 3.8 compares rates charged by banks in the US and in Mexico City. In certain cases the entry price for U S services in the States - although they have a larger income per person - i s below the Mexican price of entry in the case o f Banks checking accounts. A Mexican could afford a checking account in the States, when he couldn't afford one in Mexico. Table 3-7 Banking products are easier to accessinthe US.than inMexico (inMX$of 2002) Mexico Checking USA* Free Mexico passbook US passbook non interest Checking Account Minimum opening $1,923.15 $793.81 $786.34 balance and $0 monthly charges $12.27 $8.18 $18.82 Income NO YES NO NO percentile 25 th: $1,547 Percentile 50th: NO YES YES YES $2,380 Percentile 7Sh: NO YES YES YES $3,703 Percentile looth: YES YES YES YES $11,619 Source: Own estimates from data of INEGI, CONDUSEF, Hannan 2001 and *www.wachovia.com. Digital Banking 29. Digital banking, including ATM, telephone and electronic banking from home, has been surging in Mexico. In theory electronic banking reduces the costs of financial transactions. Payment of salaries and wages into accounts handled electronically, first by "debit cards" and then by "credit cards" could bring large numbers of unbanked persons into the system. Such proposals have been made, and acted upon in the USA, but electronic banking in Mexico focuses on a more exclusive client group, with computer and telephone for internet access. Using the same analysis applied to checking and savings accounts, Table 3.9 shows that only middle classes (75`h percentile) and high income groups (looth percentile) can afford digital banking services. 38 See Reforma May 28,2002 43 Table 3-8 ATM and electronic bankingservicesinMexico are beyond the reach of most Mexicans but a - USCredit Cardis not Institution Banamex Banca USA BBVA- Bital Setfin Check Card Bancomer Income percentile 25 th: No No Yes No No $1,547 Income percentile 50th: $2,380 No No Yes No No Income percentile 75`h:$3,703 No Yes Yes Yes Yes Income percentile I00`~: $11,619 Yes Yes Yes Yes Yes Source: Own estimates from data of INEGI.Condusef and *www.wachovia.com. 30. The US example (but based on Mexican income levels) shows that electronic banking can be affordable to very low income groups. Wachovia's check card can be used by persons below the 25thpercentile of income distribution, while the Mexican bank products reach no farther down than the 75thpercentile of income distribution. This, added to the fact that certain banks can and do offer free ATM services and set very low entry fees for basic savings accounts implies that low-cost low end banking i s not necessarily a losing proposition. Note however that banks in the United States can be motivated by other important reasons such as public policy mandates (e.g. CRA), to set fees below their costs. 31. In the case of the Mexican market, this review shows that banking products are predominantly high priced and directed at an upper income clientele. The few products geared to the small saver (or potentially for the unbanked) are not easily affordable and even the basic savings accounts tend to lose their purchasing power due to minimal interest rates. Banking strategy and know-how appears to favor the high-end clientele, and the sector seems devoid of competitive forces driving prices down and/or seeking to open new markets. The situation appears ripe for change, and may be changing. The next section looks at some recent initiatives from the public and private sector in the area of broadening access to financial services. 3. NEWINITIATIVES FOR INCREASINGACCESS 32. A key campaign pledge of President Vicente Fox's government was to promote micro-finance and access to banking as part of a broad agenda to combat poverty. Thus far the government has concentrated on developing the "social interest banking sector" as a means of providing low-income households and individuals access to financial services. Government emphasizes the micro-finance sector, made up o f small, private institutions offering credit and the not for profit sector (cooperatives and credit unions). The Government offers technical assistance and general regulatory functions through BANSEFI. Government regulators differentiate between high end and low end commercial banks. However, the private-for-profit sector has recently made advance in developing new financial products for the unbanked and lower income groups. The commercial banks, already mentioned in the first section o f this chapter, make up one group. Commercial stores, discussed in Chapter 2 in terms of their prominent role as credit providers, are another. 44 Figure 3-5 Non-bank financial institutions concentrate on rural areas, outside of Mexico City G uStatej u a t C lients 56 Total a n a 616.53 Jalisc 354.49 3129% % Q u e r 6 t a r sN 340.57 1 8 % .L.P 86.497 5% s . L , p , N.L u e v o 79.356 4% Yucata 61.760 3% M i c h o a c a 53,653 3% N a y a r i 39.423 2% C o l i m 35.652 2% O r D.F. J a l 0 tros 707.37 35,422 21%% Total 1.910.744 110 0 % I S o u r c e : C N B V , D i r e c c i o n G e n e r a l de S u e r v i s i o n de E n t i d a d e s d e A h o r r o Y C r e d i t 0 P o p u l a r The Public sector and the social interest banks 33. The Government has designated BANSEFI as its vehicle for developing the social interest banking institutions and for increasing access to financial services. BANSEFI came into being in 2001 as part of the Ley de Ahorro y Crkdito Popular (2001) which transformed the 50 year old postal savings system (PAHNAL) into a state development bank to support cooperatives, credit unions and other social interest banking organizations. BANSEFI's primary mandate i s to provide the technical assistance necessary to regularize all informal financial organizations which meet minimum standards39through a process and requirements described by BANSEFI officials as costly. However, by providing technical support BANSEFI expects to reduce costs for these organizations4'. 34. The social banking sector is strongest in rural areas and has limited impact in urban areas.. Reporting the regularization of 500,000 accounts through BANSEFI's program to date, its directors hope to increase membership in the social interest banking organizations from the current estimate of 3 million to 4.5 million by the end of 2005. BANSEFIhas also taken over the management of over 550 former national postal savings offices (PAHNAL) and, as of 2003, has approximately 1.5 million clients. However, most of BANSEFI's offices are rural - there are only 60 branches in Metropolitan Mexico City. Only 2% o f the social interest banks have offices in Mexico City and their presence in the most urbanized states of Nuevo Leon, Estado de Mexico and Jalisco i s also minimum. Of BANSEFI's affiliates, only (Caja Popular Mexicana) maintains a national presence with some 400 affiliates (Fig. 3.5). 39 A minimum of 250 associates, and capitalization requirements at an estimated 9% of assets and 40% provisioning of non-performing portfolio. 40 Ibid. Supervision and Federation membership costs are quoted at between 1.35% and 4.35% of assets for large institutions. These could rise significantly in the case of Cajas Solidarias a significant number of which are small. 45 35. The social banking sector offers low cost products, but its organizations tend to be limited to one service. BANSEFI offers two savings products through the former PAHNAL offices. Both are affordable by persons within the first decile of income distribution, according to the same analysis of bank products carried out earlier in this chapter and both would pay interest to the average savings account (MX$640). Cuentahorro, i s a basic savings account that can be opened with 50 pesos. Deposits must be for 30 pesos or more. Cuentahorro pays variable interest (currently froml.35% to 2.7%) depending on the amounts deposited. Withdrawals are unrestricted but a 50 pesos minimum i s required to keep the account open. The second product, Tandahorro, works like a tanda. The client chooses a term and monthly deposit amount and participates in a lottery. Tandahorro also pays interest ranging from 2.2% to 2.85%. BANSEFIpays the difference between the interest and inflation rates (maintaining the purchase power of savings) for all accounts with over MX$1800 pesos. BANSEFI does not charge for account maintenance and has no credit products. Its earnings derive from the difference in returns on investments (mostly in government paper) and its interest payments to clients. Savings in BANSEFI have grown 14% for the period 1995 to 2001 in real terms representing a total number of deposit accounts close to 1 million41. According to BANSEFI officials, these products will be phased out by 2005 when the BANSEFI clients will be turned over to other social interest banking organizations and BANSEFI will become a regulator. Commercial Banks Open Special Servicesfor the Low-End Market and the Unbanked 36. Commercial banks are moving cautiously into the middle class market. As mentioned in an earlier section, all five of Mexico's major commercial banks offer lower end savings products. However, as the earlier analysis demonstrated, the terms of the products focus favor persons above the 75`h percentile o f income distribution - i.e. the upper income but unbanked population. Two of these "popular" savings programs are described in the following paragraphs. 37. BITAL Bank has created a product called Tanda Bital which mimics the informal Tanda structure and has few restrictions to entry. For a minimumopening balance of approximately US$50 (500 pesos) Tanda Bital offers a regular raffle with cash and consumer goods as prizes instead o f providing an interest rate. The other entry requirements are proof of residence, o f age (over 18) and o f citizenship (official identification). If a Tanda Bital account balance falls below $50 it i s charged a $5 monthly commission. As an incentive to save, Tanda Bital asks clients to set monthly deposit goals (e.g. $10 a month). Each month the clients meet their deposit goals they qualify for the raffle. 38. Conversations with BITAL managers reveal that the product has been successful but that it appears to have reached its peak growth even with BITAL's high accessibility and widespread bank branches. The managers say that their potential clientele distrusts banks, are wary of complicated products and are fearful that having a bank 41See www.bansefi.gob.mx 46 account might mean having to pay taxes which they generally avoid by remaining in the informal sector. Some of these factors echo the findings described in Chapter 2, and the analyses o f benefits inparagraphs 90-96 of this chapter. 39. Bancomer's Libreton targets low-income formal sector workers. The minimum opening balance i s MX$750 but the Libreton requires two references in addition to Tanda Bital's entry requirements. As with Tanda Bital the Libreton offers a regular raffle to accounts with over MX$750. However, every month the account balance falls below MX$750, the account i s charged MX$75. The Libreton pays interest on accounts with over MX$3000 deposited. Bancomer managers report the Libreton to be one of their most profitable products. Nevertheless, as the previous review and analysis of low end products offered by the commercial banks shows, with entry minimums of MX$500 and $750, neither Tanda Bital or the Bancomer Libreton are reaching persons below the median income level. A Commercial Sales OutletMoves into Banking 40. A commercial outlet which catered to lower and middle income groups uses its knowledge to design banking products to attract the urban unbanked. Banco Azteca, Mexico's sole commercial bank entirely focused on the low income market grew out of Elektra, a commercial outlet specializing in domestic appliances and furnishings - mostly electronic - and in selling on credit. Aiming for the households with incomes between 2 and 5 minimum wages, or between the 35th and 65thpercentile of income distribution, Banco Azteca has profited from Elektra's previous knowledge o f the sector and from its network of 850 stores nationwide. Banco Azteca's offices are located inside Elektra stores. The bank keeps the same long hours as the store and provides credit to Elektra customers, benefiting from Elektra's credit information system. Banco Azteca's urban presence i s strong (150 branches in Metropolitan Mexico City compared to BANSEFI's 60.) 41. Initially Banco Azteca made loans only for commercial goods but it now offers personal loans (unsecured) up to MX$5500. The average loan size i s MX$2500 and payments are collected weekly. Elektra requires official identification, proof of income and proof of residence from its clients. (The residential information i s an indicator.of capacity to pay, not a form o f guarantee.). Credit officers (riding Elektra- provided motorcycles) visit each potential client at home where they assess the value of the client's property and possessions. Informal workers present receipts for goods purchased for business to demonstrate their incomes: e.g. carpentry tools, gasoline receipts, etc. Banco Azteca officials estimate that one third of their clients are drawn from the informal sector. Once a credit i s approved (usually within a couple of days) the credit officers follow the repayment and make weekly visits to the person's home whenever there i s a late payment. One official puts it this way: "after three months o f delayed payments we've visited the client 10 times. A commercial bank makes one visit only after three months of late payment". Thus, operation costs are high but these are offset by higher interest rates. Banco Azteca charges 6 - 10% per month, compared to the average commercial bank active interest rate o f 24% per annum. Credit officers earn bonus commissions when credits pay off, but must pay for 47 defaults out of their wages. Bank officials credit their practice o f close tracking o f clients for their high repayment rate - 97% and for a low late payment rate (80% on time). 42. Banco Azteca opened its operations in 2001 with one savings product. "Guardadito" which pays a minimal annual interest rate (1%) but requires a minimum deposit and balance of only MX$50 (US$5.00). "Guardadito" accounts average a balance of MX$1300. There i s no maintenance charge for accounts over MX$50 and accounts over a certain amount qualify for monthly lottery prizes. Four months after launching the "Guardadito" Banco Azteca introduced a new savings instrument for clients who could maintain a minimum of MX$5000 (US$ 500.00): time deposits paying annual interest o f 5% for 1month, 6% for two months, 7% for 90 days and 8% for up to six months. To keep costs down the bank does not send out statements. Clients can request balance information at the office. 43. Banco Azteca's initial growth has been spectacular. In its first 6 months, Banco Azteca opened an average of 1,500 accounts each day and after that, with the introduction of the time deposit, new account openings increased to 5,000 per day. The bank now counts some 1.75 million accounts nationwide, a strong indicator of a pent-up demand for financial services. Banco Azteca's deposits total over MX$240 million (around 80% of the total social interest banking sector). Banco Azteca estimates that there are 16 million unbanked families in Mexico today. It plans to increase its menu of services, but bank officials comment that the current regulatory framework complicates matters. "The regulations are designed for commercial banks working with high income groups. They don't understand our kind of operation. We could have opened six months earlier if it weren't for the regulations," asserts Financial Vice President Luis Nifio de Rivera. While he welcomes competition, he also thinks that other commercial outlets will face regulatory complications if they attempt to enter the market. 44. What other commercial ventures could follow Elektra? S a n b ~ r n ~has~ been ~ s rumored to be considering opening a deposit program in addition to the ATM services available in all stores. Wal-Mart has begun to offer financial services to customers in the United States and in Canada. In Mexico it offers remittance services between outlets both in Mexico and internationally. Another possibility could be alliances formed between commercial banks and NGO or SCIs. At least one major bank in Mexico i s discussing the possibility of issuing debit cards to all clients of one of the largest union of micro-finance organizations. In this way the MFI clients would be able to save in a commercial bank, and the MFIwould be able to expand its lending by refinancingthrough the commercial bank. 45. As earlier analyses have demonstrated, even the "low-end" products of the commercial banking sector favor a clientele drawn from the upper quarter of income distribution, or the richest 25% of the population. Banco Azteca, developed from a commercial store, Elektra, has been moving effectively into the niche of urban ~ 42a major department store chain in Mexico 48 unbanked with a program aimed at the lower-middle class (3rd to 7th decile). The social interest banking sector products are affordable by persons in the first decile, but are mainly available in rural areas. 46. Given the very small scale o f the social interest banks and their limited growth, observers have commented that the social interest banking organizations, like many non-profit organizations, lack expansionary motives. On the other hand, the commercial banks which have significant drive to grow, lack knowledge or interest in the low-income market. In focusing exclusively on social interest banking organizations to expand financial services to the unbanked and to low-income families, government may be overlooking other possibilities. Commercial ventures or hybrid arrangements like Banco Azteca combine the advantages of commercial banks with aggressive business strategies with those of the social interest banks, with a knowledge of the lower-income market. 4. STRUCTURALOBSTACLESTO BROADENING ACCESS 47. Certain banking regulations can represent high costs for a bank focused on small savers and lenders. B y the same token deficiencies in judicial and local law enforcement which discourage commercial banks from lending also affect banks aiming for low income clients. There are technological and perceptional issues which also hinder efforts to bring financial services to the unbanked. However, the main problems for would be low-end bankers seem to be the general lack of regulations designed to deal with low-end banking. This, in turn, stems from a lack of information about target markets and about low-end banking. CNBV has been working to overcome this gap and to refine requirements to take into account the special conditions of the low-end market. Banking Regulations are Costly and Complex in Mexico Table 3-9 Costs of bank regulation are higher inMexico Cost Indicators as of 2000 117 1,510 .076 0.55 Canada 101.4 1,592 0.77 0.66 Hong Kong 169.5 1,035 2.1 9 1.49 M i x i c o 257.8 2,372 0.2 5 0.24 Source:ITAM Law Department"Efficiency of FinancialRegulationandReform of Supervision Authorities", August 2002. 49 48. Bankers in Mexico claim that having to deal with seven different regulatory agencies becomes time-consuming, costly and frequently wasteful since supervision processes are often duplicated. The seven banking regulators are: The Central Bank of Mexico, Banxico; the National Commission on Banks and Stocks, CNBV;The Institute for Bank Savings, IPAB; The Ministry of Finance and Public Credit, Banking Department, SHCP; National Commission of the Defense of Users of Financial Services, CONDUSEF; National Commission for Insurance and Finance, CNSF; The National Commission for Retirement Savings, CONSAR. Banks pay an annual quota for supervision and they devote significant resources to complying with audit requirements. Table 3.10 compares Mexico's expenditure on regulation to that of other countries with similar GDPs, while Figure 3.7 shows the possible duplications of effort within the regulatory system43. Figure 3-6 Each bank sends 260 reports per year to 5 different regulatory agencies inMexico Sources: Banco Azteca, 2003 Figure 3-7 Different regulatory agencies require the same reports from banks InformacicinFinandera ` P D 43Both Table 3.10 and Figure 3.7 are furnished by the Association of Mexican Banks and do not necessarily reflect the conclusions of the Government of Mexico or of the World Bank 50 49. Regulatory costs are high, but smaller and retail banks may end up paying more Banks pay fixed costs to the regulatory agencies to CECOBAN and to the Association of Bankers which total MX$17 million annually per bank.44 Minimum capital requirements to open a bank stand at MX$189.945million. These costs could weigh more heavily on a small operation or one which works with low margins on a "retail" approach. Notwithstanding, the non-fixed costs can represent a higher burden for the low-end bank. Each information request to the credit bureau (required by law for loan approval) costs MX$18 and each consultation takes five business days. For a low-end bank which makes a great many small loans, the required requests could make up a significant cost, particularly in comparison with the traditional commercial bank which specializes in a few very large loans per year46. 50. The Banks are not comfortable with regulatory measures which could favor low-end banking (stimulating domestic deposit taking and capping service charges.) Commercial banks have protested the liquidity and reserve requirements for deposits in foreign currency through the ABM. At the same time, the ABM has protested (and won, it seems) against attempts by regulatory authorities to cap service charges. It could be argued that these requirements encourage banks to increase capture of local savings and to move back into more traditional banking services of deposit and lending for a wider market. However, as the banks point out, such requirements could also threaten their stability. Local Government, Judicial Enforcement, and Government Papers as Obstaclesto Low- End Credit 51. According to the ABM, local governments and the judiciary may support the borrower in case of defaults on loans, despite legal entitlements and recent amendments to the existing legal code allow banks to sign an agreement with clients to guarantee repossession of collateral 3 months after default without resorting a court order. Bankers continue to consider judicial reform, along with corruption and inefficiency in the courts to be key obstacles to their expanding credit operations, particularly to low-income groups. They are also discouraged b y high security costs they associate with branches in low-income neighborhoods. 52. Infact, there is ample evidence that the default rate for low-income clients taking out small loans, i s While no systematic studies have been done, there i s some evidence that security costs of branch banks in low income neighborhoods are no higher than in better off neighborhoods4*. The information gap, in this case, has transited into higher costs and greater reticence for the bankers. The ABM also 44As calculated by Banco Azteca. 45According to Banco Azteca 46CNVB has agreed to waiver the requirement to consult the national credit bureau but only for six months, and the waiver cannot be extended. 47Marguerite Robinson, Op. Cit. 48Interviews with the Caja Social de Ahorro, MEGABANCO and with Banco Popular in Colombia affirm that security risks are no higher in low-income neighborhoods than in better-off neighborhoods. Experience in the U.S.has beenmixed 51 points to weaknesses in the judiciary which make lending more costly. The time and litigation required to collect collateral on defaulted loans, particularly after the Tequila effect, has made commercial banks argue for stronger laws and enforcement practices49. Box 3-5 Attitudes can cost money and looseclients inthe low income market Account managers at a major Mexican bank were alarmed that at one of its smaller branches away from any major city, the low income lending portfolio had just run an astronomical default rate. Further investigation revealed that the majority of small scale borrowers weren't bankrupt and were quite able and willing to pay. Perplexed, the big city managers investigated and discovered why at this particular branch, low income clients had stopped paying their debt. A new branch manager had arrived and decided that low income clients, who usually showed up in work clothes, would need to form a separate queue and wait until higher income clients had been attended to by bank staff. Bank officials report that the bank manager had also used a disparaging tone in addressing low income clients. Feeling humiliated, the clients had stopped going to the bank branch. Having made all their payments in cash and not possessing any checks, they had simply no way to continue making payments. The branch manager was soon relocated and the bank had to make a concerted effort to convince their clients to resume banking at their branch. Technological Obstacles 53. Electronic banking could reduce transaction costs and the costs o f branch banking, making it easier to expand services5'. The ABM refers to conversations among its members concerning problems with existing technologies. In the first place bankers cite the high costs of basic service installation and maintenance. Second, they point out that ATM machines are programmed to deliver cash in multiples of MS$50, while many low-income persons would need smaller denominations or change. Finally, they comment that lower income clients frequently need more transaction time as they tend to be unfamiliar with ATM machines or uncomfortable with them. Pre-conceptions as an Obstacle 54. Tanda Bital officials believe that their business has suffered because of their target clientele's preconceived anti-bank biases. Interviews of unbanked persons described in Chapter 2 suggest that many distrust bankers and anticipate disparaging treatment. Anecdotes from banks suggest that the unbanked may have good reason to be leery. Bankers are not usedto receiving poorly dressed and indigenous people (see Box 3.5) and banks seeking their business need to institute special strategies and awareness 49Banco Azteca has not argued for judicial reform. Instead, dealing with its low-income clientele, it consciously spends more resources and time on credit ratings, interviews and collection than do the high end commercial banks. Seen from this vantage point, it could be argued that the ABM's proposal would have the government (judiciary) doing work the bank itself would otherwise do. Before the Tequila Crisis lending represented a higher proportion of GDP, default rates were lower, and weaker laws were in place than today. 50 Notably, Banco Azteca relies exclusively on personal transactions although it intends to open electronic banking facilities in the near future. 52 training. Lack of information combined with lack of sensitivity51 again generates serious obstacles. Developing Regulationsfor "Low-End" Banking 55. The regulatory framework has no model for a "low-end" commercial bank. The Mexican regulatory framework conceives of the social interest banks as working with low-income groups, but has no organization like BANSEFI to assist commercial banks which might wish to focus on the same clientele. When Banco Azteca opened the first multi-service bank for a specifically low-end clientele it faced a two year process to meet requirements imposed by the regulators. According to a Banco Azteca official52,the Mexican banking regulations have been conceived and written with commercial banks working at the high-end of income distribution in mind. 56. Traditional regulations can be, and have been, altered to accommodate low-end conditions. Regulations require that proof of income or of net worth be obtained before authorizing credit. Since proof of income and of net worth i s only applicable to formal sector employees, the CNBV has accepted Banco Azteca' s socio-economic studies as an alternative. In a similar vein, CNBV has accepted the redefinition of default or late payment based on monthly delays, rather than on delays by payment. (Since Banco Azteca's clients make weekly payments, they would fall in default in less than a month if standard definitions were applied.) 57. Other issues are still to be resolved. Under present regulation, all Mexican banks must use (and finance) the country's only credit bureau, owned by the country's commercial banks, even though this bureau covers a "high-end" clientele which i s useless for a bank working with a "low-end" clientele. CNBV has agreed to allow Banco Azteca to rely only on its own data-base for a non extendable six month period. Similarly, CNBV has granted a six month waiver of its requirement that each credit be approved personally by a member of the Bank's credit board allowing instead for electronic signatures. With its nationwide network of offices and hundreds of credits per week, Banco Azteca i s attempting to institute an "electronic" credit board. 58. As CNBV has worked with Banco Azteca in an ad hoc manner to adjust regulations to fit a commercial venture into the low-end market, it has become clear that a gap exists in Mexico's bank regulatory framework. So far CNBV and Banco Azteca have managed to devise adequate stop gap measures. However, Banco Azteca's products are limited. Further adjustment will be needed as Banco Azteca's menu grows. If, as hoped, more institutions move into multi-service banking for the poor, through the graduation of SCI's or through the down-scaling o f commercial banks or through invasion from more commercial or other ventures, it i s clear that additional regulation will be needed. ''Experience with "popular" banking in the US reveals that information often needs to be disseminated with sensitivity training. People often don't believe even when told and cultural biases must be overcome. 52Most of this text i s based on interviews with Luis Nino, Financial Vice President of Banco Azteca 53 5. CONCLUSIONS 59. Broadening access to financial services could be important for the health of the overall financial sector as well as for the Mexican economy. The closing out of a large segment of the population from the formal banking sector tends to produce a lopsided or poorly distributed pattern of growth, and of itself contributes to a redistribution of income upwards. At the moment three groups of formal sector institutions are poisedto enter the market of the unbanked: the commercial banks, the social banking sector, and commercial outlets. Most of the major commercial banks have opened modest programs aimed at attracting low income or unbanked individuals. However, these various schemes represent costs of entry and/ or of maintenance beyond the reach of persons below the 75th percentile of income distribution. 60. The Government i s supporting the development of the social interest banking sector made up of the credit unions and savings and loan associations (SCIs) which have been regularized (there are twelve in Mexico City) and o f former PAHNAL, postal savings offices (60 in Mexico City), now part of BANSEFI, which i s also the public sector regulatory agency for the social interest sector. BANSEFI and the social banking institutions are mainly rural, with 140,000 persons for each institution in Mexico, compared to an already high 7,000 persons per commercial bank outlet. While social banking sector products are affordable, they are limited to savings for the BANSEFIs and to certain types of credit (mainly micro-business and housing) for the SCIs. 61. Inthe face of declining credit and deposit-taking, both in absolute terms and interms of GDP ratios, we suspect that new, non-bank financial institutions (NFBIs) have been filling the gap with a growing market of banking services. The recent entrance of commercial establishments (Sanborn's, ELEKTRA) into banking for a lower- income public and the initiatives of commercial banks to link with SCIs appear to be making important market gains in broadening the base of Mexico's financial service. Banco Azteca, the leading example, offers deposits with interest and a line of consumer and personal credit. Its widespread outlets, well-positioned for physical access and with "popular" or un-intimidating atmosphere give it distinct advantages for meeting the needs of a low income clientele. In addition, it counts on Elektra's ample client data base with credit histories. Active interest rates are high, however, and as Banco Azteca enjoys a unique position at this moment virtually without competitors, the interest rates will likely remain high. 62. Banco Azteca's experience makes clear that the regulatory and legal framework does not yet have means to deal with the particular case o f a for-profit multi-service bank offering services to the low-income market. (Government has relied on BANSEFI, as regulator o f the SCIs to cover financial institutions working with low end products.) Existing bank regulations make for duplication o f credit checks for banks working with low-income persons and informal sector workers. Regulatory fees cost more for institutions doing many small transactions than for institutions doing fewer large 54 ones. The current regulatory framework should be reviewed and revised to take into account the needs of multi-service banks working with low-income markets. 63. In addition to revisiting the regulatory frameworks, Mexico's banking regulators (SHCP, BANIXCO, CNBV, BANSEFI) should devise incentives to encourage the development of financial services for low-income populations in urban areas. Insofar as BANSEFI i s focused on the creation and regulation of specific institutions, new incentives should encourage the development o f new products, the provision of services to a wider public at lower prices and the application of new technologies to expand access to financial services at appropriate locations and neighborhoods. Supermarkets, gas stations, churches or other agencies may emerge as appropriate institutions for either carrying out or housing financial transactions. 64. Finally, the regulatory agencies should develop ways to assure the sharing o f information about financial services, both among clients and providers. Mexico already counts on an able agency (CONDUSEF) for analyzing and transmitting existing data. CONDUSEF's role could be bolstered to include following the unbanked and informal institutions, the costs and profitability of different product lines for different income levels. It could also survey unbanked and low-income population groups to determine which products do result in asset accumulation and in increased income levels. 65. The next chapter also discusses specific ideas to provide incentives for banks to set up operations in low income neighborhoods, for banks to begin new operations, for new agencies to provide financial services for low-income populations and the unbanked and for banks (or other institutions) to develop and to offer products for the unbanked and low-income populations. 55 CHAPTER4. RECOMMENDEDAPPROACHES TO INCREASING ACCESS AND REFLECTIONSON OTHER EXPERIENCES 1. The research has demonstrated that the vast majority of the population of Mexico City (and of major urban areas) does not have access to financial services, and, as a result, pays higher costs and enjoys reduced opportunities of asset accumulation. Research also shows demand exists, capacity to save, borrow and pay for financial services at considerably higher costs than presently offered. For its part, the commercial banking sector, the social banking sector and the commercial outlets appear to be positioned to increase their coverage to the unbanked. 2. Why the percentage of unbanked remains so high in Mexico, remains a question. More research i s clearly needed to better understand the systemic risks to the sector, to quantify the costs of increased services and to assess the presence of competitionkoncentration in the sector. It has not fallen within the scope of this paper to analyze the financial sector in Mexico in this sense. At the same time, this paper has identified several structural obstacles to broadening access to financial services (see Chapter 4) and the research undertaken leads us clearly to recommends a dialogue to discuss further the structural obstacles in order to identify programs, incentives, and regulatory changes which would allow the Mexican Government to support increased access to financial services. The research has indicated several key areas for developing a dialogue: rn An "informational approach,"to promote the creation and marketing of specific products identified as important to the unbanked and/or to overall economic development. rn A "product" approach,to find ways to reduce costs of basic products to become affordable to most unbanked individual^.^^ rn A "price"approach to find ways to reduce costs of basic products to become affordable to most unbanked individuals. rn A "location or neighborhood"approach to bringfinancial and banking services to specific unserved and unbanked communities both to increase access and to support specific community development. rn A "technologicalapproach" to find ways to apply new technologies both to reduce costs and to increase the banking presence or access inunderserved communities. 53 Although encouraging competition and market forces to drive down prices is the most appealing interpretation of the price approach for the author, the objective here is to review and discuss different experiences. 56 An "information and education approach" both to measure the use of formal financial services and to track unsatisfied demand as well as to promote both public disclosure of successful banking practices, of ways to increase access to the unbanked, and to educate the unbanked in advantages offered by financial programs. 3. Ineach of these cases, government should review existing regulations which apply to the commercial banks, social banking institutions and commercial outlets to determine which, if any, apply and how their application could have a more positive impact. It i s likely that programs and policies to encourage increased use of financial services will also overlap with community development programs and housing programs aimed at low income families. To identify appropriate incentive programs, the national regulatory agencies should open dialogues with local and state governments. Together they should review policies and programs used elsewhere to help design programs appropriate to the Mexican context. The following sections review the approaches listed above and outline some experiences used in other countries which could prove useful to the Mexican dialogue. 1. PRODUCT BASEDAPPROACH 4. The demand review pointed the interest held by the unbanked in savings programs with interest, in service payments and in consumer and other unsecured loans products which are not widely available in financial institutions. 5. Increasing savings capacity, both for short term liquidity management and for long term asset building among low income families and individuals i s considered desirable by many government^.^^ Guaranteeing the existence of adequate savings programs has been the objective of the PAHNALBANSEFIprogram, of the Cajas de Ahurrus in Spain and of the Individual Development Account program in the USA (IDAs). In the first cases, the government legislates and supervises the operational framework for savings institutions. In the second, government acts as a partner, investing savings deposits and paying returns to small savers. In the third case, government provides an incentive program for banks to open special savings accounts to help the unbanked learn to save and to manage assets55.These cases are mentioned only as examples to suggest that the products sought by unbanked persons may coincide with public sector interest. 54 Total savings in Mexico are high and they are also mainly concentrated among wealthier and better-educated families, meaning that the poor are less prepared for economic shocks. In fact, a considerable portion of low income families actually register negative savings since only a minority of households belong to the category of better-educated. Policies aimed at promoting savings should therefore consider the different sectors of the population, and should consider education. (Attanasio and Szekely 1998) 55 It i simportant to note that this i s only one way banks are involved in IDAs. In some programs, banks establish the accounts and NGOs and/or governments provide the match to the deposit, not the withdrawal. In other programs, banks establish the accounts and also provide all or a portion of the match. Government funds have been used for closing cost assistance, second mortgages or additional down payments when IDAs are used to assist in home buying. 57 Box 4-1The Individual Development Account Program inthe United States uses bank accountsto encourage asset buildingfor the poor The IDA program stimulates commercial banks to open deposit accounts with maximum interest rates for qualified poor opening their first account hence unbanked. (As noted in Chapter 3, the social banking sector in Mexico offers extremely low interest payments, while the commercial banks offer virtually none at all to families earning below the 75Ih percentile of income distribution because minimumbalances plus monthly maintenance charges are unaffordable for low income levels.) Under IDA, national or state governments might subsidize savings accounts by matching deposits after a certain amount has been saved for a defined time period. In fact, the participating banks make the matched payments and deduct them from corporate taxes. Evaluations of the IDA program, which started in 1998, are just becoming available. They tend to reconfirm how difficult it is for the poor to save. Approximately two thirds of the families with IDA accounts were not able to take advantage of the matched funds. That is, they never left enough funds for enough time to qualify for a matched deposit. On the other hand, they did continue saving and maintained their accounts during the time studied. One third of those who entered the program did save enough to earn the grants. The reports showed that they invested their savings and earnings in home purchase, education, business creation, or in other long term projects. Proponents of the IDA program argue that a on third success rate i s an excellent record for helping unbanked families to accumulate assets and to become upwardly mobile. They suggest that in the long runthe government subsidy (through the corporate tax reduction) will be recovered as families' increased earnings will generate greater revenues. Helping the unbanked to Payfor Services and Money transfers 6. Insofar as it i s in the interests of governments for their citizens to receive remittances, and for their public companies to be paid, it i s also in government interest to assure that these transactions are easily available, safe and affordable. It remains an open issue whether government should provide the means for financial payments and for transfers, particularly when government agencies may be the principal benefactors, or whether private agencies should do so. A dialogue focused on transfer and payment services should weigh issues and compare the cost o f different alternatives. It may be a case of revising regulations to permit commercial outlets to offer payments and financial transfer services or providing payment access at all ATMs. Box 4-2 Municipal Smart Cards Facilitate Electronic Payments for Public Services The City of Guatemala has taken an interesting step to help its citizens pay costs of municipal services. Considering the enormous cost difference between personal financial transactions i.e. Payments to a teller, and electronic transactions, the city has provided its public service customers with smart cards and has installed ATM machines around the city to receive payments in cash or by transfer from personal bank accounts. Presumably the municipality i s recovering its costs through the service payments. The citizen smart card also can be used for voter registration and to obtain maps and latest municipal news. If a local government is interested in making such investments, the returns should be strong considering that use of the equipment can also be charged to banks or other financial institutions. 7. Consumption loans ranked high among financial services desired by the unbanked, along with unsecured credit - for personal projects besides housing and business development. Consumer lending promises to be a large business. A recent report on micro finance institutions claims that consumer lending in Latin America i s currently showing more dynamic growth than microenterprise credit (Christen, 2000). Non- 58 business loans are particularly important for low-income households and the poor who are particularly vulnerable to "economic stressors" (Sebstad and Cohen, 2000, Westley, 2002). In her review of successful MFIs Elizabeth Rhyne argues that "The key to successful lending was not how clients used the money but the fact that the loan was based on the client's existing ability (and willingness) to repay."56 Evidence from MFIs and low-income credit programs also find that households increase their income even with non-business loans. Expenditure on health, education, housing and technologies can lead to more productive individuals (Sebsdtad and Cohen 2000, Westley 2002.) Box 4-3 Mexican Banks introduce innovate ways to facilitate commercial credit Banco Azteca i s demonstrating that consumer and personal loans for low income persons can be a widespread and high profit business. I t has been noted that the unbanked pay a de facto surcharge on commercial goods insofar as stores which honor credit cards pass the credit card interchange fee (usually 10%) on to all customers, not only those who pay with credit cards. This raises the point about other benefits of consumer lending to the unbanked. Credit cards offer a convenience factor beyond the consumer loan - they allow cash-free shopping, and a delayed payment option which is paid for in the purchase price. Debit cards, which allow the same advantage of cash-free shopping, are part of a major new program being discussed by BANAMEX and CAME for a pilot program in Valle de Chalco in Mexico City. The Ford Foundation will be evaluating this experience could also determine what interest the unbanked have in the cash free aspect of debit card use. Furthermore, the evaluation should measure what percentage of users could qualify for delayed payments or for automatic credits. Depending on the answers, BANAMEX and CAME should be encouraged to develop debit cards allowing for short term delayed payment options (like first stage credit cards) and eventually automatic credit cards. 2. PRICE-BASEDAPPROACH 8. Research findings described in Chapter 3 show that the most frequent reason unbanked Research findings described in Chapter 3 shows that the most frequent reason unbanked persons give for not using formal banks are their high costs. Since the unbanked also indicated that they are most interested in savings products and service payments, items such as minimum balance, maintenance charges (both in relation to interest earned) must be considered in addition to transaction charges. In all cases, it was noted in Chapter 4 that the Mexican banks charge higher prices than U.S. banks. Of course conditions differ greatly between the two countries. The prices charged by Mexican banks reflect a different environment in terms of technological and entry costs, of perceived risks, of regulatory and structural obstacles, some of which were discussed in Chapter 4. A dialogue focused on ways to encourage more `affordable' financial services should look first at the factors which lead to present cost structure. The same dialogue should also review the degree of concentration and competition in the Mexican formal banking sector. While this may not be seen as a market risk, it may well affect the charges for banking services. 9. It is worth noting that price caps for electronic and telephone banking services have been under discussion recently in Mexico in response to the uncommonly high prices 56Rhyne, Elizabeth, "Microfinance Institutions in Competitive Conditions". October 2001. 59 charged by commercial banks for banking services (see chapter 4). These services are not generally used by lower-income customers, nor are they sought by the unbanked. They do, however, appear to affect enough clients with political clout to have come to the attention of bank regulators. Insofar as prices of financial services are already an issue for regulators, it could be an appropriate time to consider pricing of products which affect the unbanked and lower-income clients. Indeed, the prices could perhaps fall without capping if usage increased to reach the unserved market. Box 4-4 Induced cornpetiton can lead to lower prices for bankingservices In New York State a consumer `watchdog' agency tracks the fees, requirements and returns of financial service products offered by commercial banks. The results appear in a monthly newsletter which ranks banks according to affordability. How these reports affect pricing i s o f course a matter of speculation, but other analysts have noted that legislation to require "lifeline" banking products (at least one affordable deposit service per institution) was really not necessary in New York State, since banks already offered products which met and surpassed the legislation requirements (Doyle et al. 1999) IIn a similar vein CONDUSEF in Mexico reports that after it began publishing mortgage rates on its website, at least one bank immediately lowered its rate.57The incentive to financial institutions is, of course, to appear on the list as a low cost product and thus to draw clients. 3. NEIGHBORHOOD BASEDAPPROACH 10. As discussed in chapter 3 and 4, the presence of financial institutions i s associated with areas with a concentration of high income levels. The physical absence of banks in the lower-income delegations is certainly a factor in Mexico City's the high "unbanked" rate. The absence of banks from low-income and minority neighborhoods may also contribute to the circle of poverty, as it could well be a reaction to it. Clearly bankers and formal sector financial institutions associate high costs, both in construction, security and risks, with establishing new branches or service outlets in previously unserved or underserved neighborhoods. While technological advances (ATMs, "supermarket banking", etc.) can reduce the costs and even risks associated with setting up service outlets in previously unserved neighborhoods, their impact as such has not yet been felt inMexico. 11. Depressed or marginal neighborhoods and how to bring about their development have long been a concern to both the national government (SEDESOL,'FONHAPO), to the Central Bank (FOVI) and to municipal governments as well. To date, however, community development programs in Mexico rely solely on public and NGO support. A dialogue which focuses on broadening financial services to particular, underserved and depressed neighborhoods could dovetail with efforts to upgrade targeted communities. Households in marginal communities could presumably benefit from programs to build assets, as they could from programs to improve the level of urban services. B y the same token, the neighborhood approach could encourage private sector institutions to become financing partners in local community development "AsdoestheWashingtonPost 60 efforts. Box 4-5 Attitudes can cost money and loose clients inthe low income market In the United States, legislation to promote access to financial services in the US began on a geographic basis, mainly to break the common practice of "redlining" certain city districts which was viewed, and demonstrated to be, a form of racial and income discrimination. The passage of the Community Reinvestment Act in 1977 (hereafter "CRA5*") and other measures (including the creation of the Neighborhood Reinvestment Corporation) aimed to break discriminatory patterns and to get banks involved in neighborhood development by lending for property improvement and community projects. CRA established a point system which rates banks according to their provision of services, balancing lending and savings products, to low income neighborhoods. CRA points carry certain weight with regulators considering branch closures, the establishment of new branches, mergers, etc. and they have influenced lending patterns in support of involvement with poor and minority groups. In parallel, community organizations appear to have been strengthened by the new power they can exert over banks to force compliance with CRA. (Bates, Holyoke 2000). By the 1990's Community Development Banking had blossomed into a market of its own, with commercial banks trading for credits in providing financial services for low-income ' neighborhoods. Profitability for small business mortgage, home improvement, and community development lending under CRA has been shown, after an initial period of stagnation, to be virtually the same as non CRA lending as seen in the attached figures taken from US Federal Reserve Bank reports. Figure 4-1 Profitability of CFtA Lending Profitabilityof Community Development Lending (percentageof dollars originated) Unprofitable RProfitable 0Break Even Marginally I 58For a full history of CRA see Litan et. al. 2000, Haag, 2000, Squires, 1992 61 Figure 4-2 Profitability of CRA lending Profitability of Community Development Lending (percent of institutions) I 1 1 Marginally Unprofitable EU Break Even 0MarginallyProfitable 0Unprofitable Profitable 4. TECHNOLOGY BASED APPROACH 12. New technologies used in banking and financial services could open new for reaching marginal areas and unserved population groups. Combinations of change machines, payment machines and ATMs could overcome the problems mentioned earlier by bankers in reaching low income persons with ATM services. `Mobile banking' services such as ATMs (or ATMs in vans) provides a cost effective way to expand service and reach areas where branches may otherwise not locate. Box 4-6 Electronic Transfer payments can lower cost and lead to greater banking services Governments of the United States, Argentina and Bolivia, among others, have recently passed legislation requiring that all government employees be paid by electronic transfer. The change from cash payment to check to electronic payments, results in savings for government, but also promotes provision of banking services and the development of debit card accounts for many unbanked government employees. In the United States government has combined the interests of reducing its own costs with supporting unbanked and low income families to open bank accounts - in particular for savings. The federal government has encouraged banks to create low-cost savings accounts for people receiving federal benefit payments electronically, mainly social security payments. I 5. INFORMATION AND EDUCATION APPROACH 13. The research described here has pointed to many failures of information underlying the low levels of coverage o f financial services in Mexico. Considering that both the formal banking sector appears to lack information about the unserved population, and the unbanked have limited access to or understanding of the banking system, the term asymmetry does not appear applicable. Misunderstandings actually to be quite symmetrical. Thus it i s important that a dialogue on access to financial services review the needs for information, among public sector regulators, private sector suppliers (financial institutions) and the potential clientele (the unbanked.) "Financial literacy" has been emphasized in the United States as key in helping low-income 62 families to use financial services to build assets. In Mexico, CONDUSEF points to one of its operational pillars, "Diffusion of Financial Literacy and Culture" which supports financial education. Box4-7 Public disclosure andinformationas a tool to increase accessto financial services Much o f the success of the CRA and community banking movement in the United States has been attributed to the public information sharing and indeed the publicity surrounding community banking efforts which could demonstrate that the poor can be a very good business. However, without a conscious effort to gather, analyze and to share information, the findings will never be acted on and gains and losses will never be detected. The Mexican government has used information sharing to great advantage under the FOVI program. By reviewing and publishing comparative regulatory costs for housing in different states and cities, government enabled the private sector to know where its investments in home construction would have greater impact, and where it should pressure for deregulation. 6. NEXTSTEPS 14. This sector work has attempted to describe a problem affecting both the financial sector, and the population which exists without it. While it suggests a path for further work, it stops short of recommending actual solutions or even pilot projects because the next steps must be taken as a partnered effort. With government in the lead59and with the participation of financial institutions from the commercial banking sector, the social banking sector and the commercial outlets, as well as inputs from NGOs and academic communities which are working in the area, the World Bank team should support a dialogue and interchange of ideas about international experiences and possible applications to programs to broaden access to urban financial services in Mexico. 15. While the overriding objective, however, i s to broaden access to financial services and to reduce the number of unbanked through the strengthening of a positive regulatory and business environment, a conceptual framework must be developed in parallel with any directed efforts. The risks which could be engendered by a broader- based financial system must be identified and quantified. Thus a role for the public sector to accompany and to insure against the risks o f a more responsive financial sector could also be defined. 59We are pleasedto report that several representatives of Mexican government (BANIXCO, SHCP, CNVB, CONDUSEF, BANSEFI, SEDESOL) have worked with the team on developing this ESW and have reviewed this document. 63