Services W Final Report Prepared for the United States Agency for International Development Mary M. Miller, Development Alternatives, Inc. John Cann, Development Alternatives, Inc. ' Arelis Gomez, IMCC Mario B. Lamberte, Philippine Institute for Development Studies March 2004 RECEIVED APR (I 6 2004 USAID : PHIL§. ORP Developanent Altemmta'ves, Pnc. 7250 Woodmont Avenue, Suite 200, Bethesda, Maryland 20814 Tel: (301) 71 8-8699 Fax: (301) 71 8-7968 Email: infoBdai.com BACKGROUND AND IN'~'IZOL)UC~'ION ................................................................................ 1 ACHIEVEMENT OF PROJECTARGETS .................... ..... ................................... I . Special Activities ............... ........ ............................................. 6 Other Measures ........................ .... ... ................................................. 7 The NABS Approach .................................................................................... 8 Measures of Success for MARS ......................................... ......................... 9 MABS Broader Effect on Rural Banking ........................................................ 10 Issue: Low Volume .................... ................................................................ 11 ClieniRetentiori .......................... .. .............................................................. 13 Internal Controls ............................................................................................... 13 I ! STREAMLINED MAB S APPROACH ............................................................................... 14 RBAP .......................................................................................................................... 15 MABS SERVICE PROVWERS .................................................................................... 16 RE32000 ......................................................................................................................... 19 General Findings ............................. ................................................ 21 Specific Feedback on RB2000 ..................................................................... 22 . . .. ................................................ \ The Overall Commercial Viability of RB2000 23 .. Progress Towards Achievement of MARS Goals and Objectives ................... 24 POLICY AND REGULATORY ENVIRONMENT ................................................................... 25 CONTINUATION OF MABS AND RECOMMENDATIONS ................................................. 28 RECOMMENDATIONS .................... ........................................................................... 29 Before end of project 30 September 2004 ........................ ........................... 29 Possible Phase Three Activities ..................................... ... ............................... 30 RB2000 and MIS Recommendations ........................................................................... 32 Before end of project 30 September 2004 ................................ .... .................... 33 Possible Phase Three Activities ................... ................ ............................... 34 ANNEX A: PERSONS INTERVIEWED ....... .................................................... A-1 ANNEX B: MICROLOAN PORTFOLIOS AND STATISTICS AS OF 12/31/03 FOR SELECTED BANKS .................................................. B-1 ANNEX C: MABS EVALUATION: REGULATORY COMPONENT ............. C-1 ANNEX D: PERFORMANCE 2003 ROLLOUT., ................... ,.. ...... ...,..... D-1 ANNEX E:PERFORMANCE AFTER YEAR ONE ROLLOUT ...................... E-1 ANNEX F: DEFINITION OF OPEN SOURCE ................................................... F-1 ANNEX G: ASSESSMENT OF RB2000 COMPONENTS .................................. G-1 Table 1 Microfinance Lending Reported to the Central Bank ......................... .. ........ 3 ABC ARMDEV ARMM ASEC A'TM BAP BSP CAAM I CGAP CIBI CPLP EAGLE ISSEC MABS MAB STeR mDCo MSP NCC PAR PDA PhP PMS 1 Punla RB AP RBRDFI TWG US AID Academy for Banlung in the Countryside Associated Resources for Management and Development, Inc. (MSP) Autonomous Region of Muslim M-indanao Automated Systems & Equipment Colp. (RB2000 developer and roll out firm) Automated teller machine Bankers Association of the Philippines Banglco Sentral ng Pilipinas (central bank) Conflict-Affected Areas of Mindanao Consultative Group to Assist the Poorest (World Bank microfinance office) Credit investigation and background investigation Credit Policy Improvement Project Earnings, Assets, Growth, Liquidity, Efficiency (bank rating) RB2000 roll out firm Microenterprise Access to Banking Services MABS Technical Resource Mindanao Economic Development Council MABS Service Provider National Credit Council Portfolio at Risk Personal digital assistant Philippine peso; exchange rate as or January-February 2004 was PhP Performance Management System Punla Sa Tao Foundation (MSP) Rural Bankers Association of the Philippines Rural Bankers Research and Development Foundation, Inc. Technical Working Group United States Agency for International Development vii The Microentesprise Access to Banking Services (MABS) Program began in 1998, as a joint effort of USAD, the Mindanao Economic Development Council (hEDCO), and the Rural Bankers Association of the Philippines (RBAP) to bring financial services to microentrepreneurs in Mindanao. Technical assistance in the implementation of the project was provided by Cheinonics International. This multiyear project was to provide the rural banks with microfinance lending technology so that they could provide credit and deposit services to the target clientele. The expectation was that the banks would learn that they could do microlending, relying on character and cash flow of borrowers and not looking at hard collateral, on a profitable basis, and would henceforth look at microfinance as a core line of business. The project was later expanded to all areas of the Philippines, and the contract was extended in 2002. The closeout date for the project is now September 30, 2004. This evaluation was intended to pasticularly cover the second phase of the project, which focused on expanding and strengthening the program and the number of both banks and client pasticipants, particularly among the "Muslim" areas of Mindanao. Phase Two also addressed on instituting "anti-backsliding" procedures to ensure that the banks would continue to use the MABS approach to lending, MABS is a well-designed, well-timed, and well-implemented program that particularly benefited both from a good distribution network and from a strong enabling environment. The rural banks are the smallest size category of regulated and supervised banks that tend to be located in smaller municipalities and rural areas, and are appropriate size for the rnicroentrepreneur clientele. As banks they already had lending and deposit-taking powers, and the project was only to provide technical assistance and no funding or loan guarantees. Further, the government of the Philippines and particularly the Bangko Sentral ng Pilipinas has been interested in enabling access to credit for microentrepreneurs, and has been revising policies and regulations to accommodate cash flow-based lending. As of December 31,2003, the program was operating in 117 branches of 41 banks, and was serving approximately 45,000 active loan clients. The project has largely exceeded the contractual goals set for Phase Two, and is currently managing to higher participation targets informally set with USAID staff. MABS is a systematized approach to microlending that incorporates the best practices of microfinance into a rigorous set of steps to be closely followed, including doing market research, product development, credit investigation and background investigation (CIBI) on prospective and current borrowers, doing a cash receipts analysis to determine debt service capacity, arid closely monitoring performance. Almost all the banks' MABS portfolios are profitable. In addition to enjoying profits from microlending, the banks describe the MABS technology as having a significant effect on their other lending. For instance, MABS employs a "zero tolerance" approach to past due payments, i.e., late and non-payments are immediately addressed and followed. The banks are stat-ting to apply this policy to all loans, rather than relying on collateral value to ultimately make the loan good. Another example of ... Vlll the MABS influence is the spread of portfolio at i-isk (PAR) as the measure of postfolio risk rather than just the past due amount of loans. The only major criticism that the evaluation team found of the program is that the average number of loans per account officer, at 99, is low by intesnational standards. While a strong initial focus on quality and the fact that many of the loans are on a daily payment basis offer some explanation for the low productivity, overall volumes should be increased. Some additional attention should also be given to client retention and to strengthening internal controls. As a part of expanding the MABS program more widely and quicltly than the project could do on its own, three training fisms designated as MABS Service Providers (MSPs) were trained and licensed to provide MABS training and technical assistance on a fee basis for rural banks. Two of these firms are actively soliciting and providing MABS service. These rims were well chosen, and their work is highly regarded. A corollary benefit of bringing in the MSPs, as well as of the "streamlined" approach that the project took to providing the banks with lil.le MABS technology (i.e., providing some of the MABS technology through training programs that were attended by several banks in lieu of one on one assistance) is that training and technical assistance materials are now highly standardized and available to implementers and participants on compact disc. The project osiginally was designed to work hand in hand with the RBAP, and eventually to have all technical assistance and certifications work to be handed over to it. However, RBAP's direction is effectively dependent upon the interests of the president, who is elected for only a year at a time. RBAP therefore has not provided the constancy that the program sought. However, in the long run the MABS approach is expected to be sustainable because of the profitability of the approach to the MABS banks, thus their interest in it. The MSPs, as well as other programs such as the resource person program (the MABSTeRs) will also provide for sustainability and continuity. A major product of the MABS program has been the development of the RB2000 software package, which was developed to give the rural banlts reliable, user-friendly software that would meet reporting and accounting needs for a microfinance loan portfolio, as well as handle deposits and general ledger. RB2000 is high quality software that is stable and appropriate for the market, and is available for free to any rural bank that will cover the installation costs, which are modest by international standards and which should be cost￾effective for the banks. A significant feature for the rural banks is that RB2000 is certified to meet the central bank's reposting requirements. As mentioned above, the central bank has made microfinance a priority, and in the recent past has amended its regulations to allow the banks to malce microfinance loans, loans of up to PhP150,OOO to rnicroentrepreneurs for business purposes without hard collateral (real estate) but made in accordance with a lending procedure. While the MABS portfolio are only 10% of the self-reported microfinance loans, the MABS technology has been an important model for the central bank of individual lending. While not endorsing MABS, the central bank does refer banks to MABS for individual credit technology, and asks MABS representatives to speak to examiner training sessions. The central bank also acltnowledges MABS9s role in popularizing PAR as a measure of portfolio risk, which it has recently started to require that the banks report. We are recommending that MABS be extended for eighteen months to two years, in large part because we feel that the program is now realizing significant momentum, which should be suppoi-ted and nurtured. We are also recommending that specific steps be taken to increase productivity within cestain banks and across the board. Other recommendations include - developing a micro agri-loan product, as well as investigating technology such as smart cards and Palm Pilots that: can reduce the cost and increase access to banking services for the project target clientele. The Microenterprise Access to Banking Services (MABS) Program was initiated by USAID in conjunction with the Mindanao Economic Development Council (MEDCo) and the Rural Bankers Association of the Philippines (RBAP) in 1998 as a way to assist microentrepreneurs and small producers access formal financial services including both lending and deposit services, and to spur rapid and widespread economic growth in the - Philippines, particularly in Mindanao. The project premise was that micro and smaller businesses could grow faster if they had access to more reasonably priced financial services, and that banks would find that lending and providing other banking services to the micro sector could be profitable lines of business, if they applied appropriate best practices to lending. To date the banks had found administrative costs of providing small loans to be too high, and only provided such credit out of a sense of social responsibility. The project only was intended to provide the banks with new lending technology, and did not include loan funding or guarantees. ! The program was initiated through the rural banks, which are small, privately owned banks, supervised and regulated by the central bank, that typically operate in the rural areas, These are the smallest of four categories of banks in the Philippines, with limited powers and lower capital requirements than other bank categosies. There are 780 rural banks in the country, with approximately 1,800 branches. This programming decision was a good one, as MADS appears to have been the right program at the right time, with the right participants. The rural banks are spread throughout the country, and are the right size, and in the sight locations, to be candidates for microlending. Further, the government of the Philippines, particularly the central bank, is supportive of microenterprise lending, and several regulatory and policy changes have recently been instituted that support the banks7 entry into cash flow based microlending. The first phase goals of the MABS program, which included introducing a microlending technology, "MABS," into 20 banks and reaching 8,000 boll-owers and 15,000 depositors I were fully achieved: in fact, the program was operating in 60 bank branches, and 18,000 depositors and 45,000 microdepositors were being served. Significantly, the banks using MABS were finding microlending to be profitable, and embraced the technology. The second phase of the MABS program, which this evaluation is reviewing, was designed to expand and strengthen the program and number of participants throughout the Philippines, including particular attention to the "Muslim areas" and areas with substantial Muslim populations, as well as instituting a11 effective "anti-backsliding" program to ensure that the banks continue to adhere to the MABS approach. RBAP capability was also be developed so that it could install and strengthen the MABS approach in the rural banks. This evaluation was performed in January and February 2004. by a four person team including microenteprise operations analyst and team leader Mary Miller, strategic planning specialist Arelis Gornez, policy/regulatory specialist Mario Lambei-te, and information Microertterprise Access to Banking Services (MABS) Progrant Evaluation management specialist John Cann. The team reviewed program documents and reports, and interviewed staff from MABS, the RBAP, MABS service providers ARMDEV and Punla Sa Tao Foundation, RB2000 service firms ASEC and ISSEC, as well as representatives of the central bank, the Bankers Association of the Philippines and its subsidiary BAP Credit Bureau Inc., and CPIP. Most important, the team visited with some nineteen banks that are using the MABS approach to microlending, had additional visits or discussions with six more bankers whose banks are using or training to use MABS, and met with an additional three banks that are using the RB2000 although are not using MABS. A list of the persons interviewed for the evaluation is attached in Annex A. Statistics for the banks visited, as well as summary statistics for all of the MABS participants, are contained in Annex B. Additional evaluation steps are also described in the section concerning the evaluation of the RB2000 development project. The MABS project is being implemented with technical assistance provided by Chemonics Internationail, under contract with USAID. The current expiration date of the project is 30 September 2004. Note: At the time of this evaluation the exchange rate was approximately PhP55 = US$l. The approximate equivalents of some of the amounts and costs in the report include: PhP3,OOO - PhP 5,000 Typical starting loan size, $55 - $91 PhP25,OOO - PhP 50,000 Typical bank maximum microfinance loan size, $450 - $900 PhP7,004 Average MABS loan size, $127 P11P150,000 Maximum microfinance loan per central bank definition, also approximate cost of MABS training and technical assistance from MSP, $2,725 PhP350,OOO - PhP500,OOO Estimated full cost of cost of purchasing hardware, operating system software, a data base, and peripheral equipment for installation of RB2000, $6,360 - $9,100 MABS is a well-designed, well-implemented and effective program. While there is room for improvement in areas of productivity and scale, overall the program can be rated as a successful one that potentially can be replicated worldwide. Key factors in success appear to be the choice of the rural banks as target intermediaries, highly standardized training and technical assistance packages, an effective approach based on best practice standards, zero tolerance on non-payments, and close monitoring. The project has also addressed bank management and financial practices, and has contributed to a strong enabling environment. The banks cite a range of reasons for entering the MABS program, including the interest in microfinance technology, opportunity for a profitable new line of business, interest in social Microerzterprise Access to Banking Services (MABS) Program Evaluation outreach, and a desire for free or subsidized training and/or equipment and vehicles. The banks have also been interested in developing microfinance experlise because the central bank has put a moratokum on branching, except for branches that will do a majority microfinance business. (To qualify the banks have to demonstrate that they have a rnicrofinance technology in place, as well as trained staff and directors.) Virtually all the banks report that microfinance lending is profitable for them, and one case the bank reposts that over 50% of its profit is from microfinance, despite only having 12% of its loan portfolio devoted to microlending. Microlending generally is defined the provision of small unsecured loans to the poor and low income households for microenteryrises and small busihesses, in amounts of PhP150,OOO or less, that are based on the projected cash flow of the borrower. It is important to recognize that MABS is not the only microfinance technology in use in the Philippines. The other microfinance approach that is widely known is the Grarneen technology, which is available through a training course offered by Card Bank, a Grameen affiliate. However, most microfinance in the country is homegrown, and does not necessarily particularly follow an established fdraiula. The chart below shows i he microfinance lending reported to the central bank as of 31 October 2003, with MABS affiliates' microlending shown as a comparative figure: Table 1: Microfitlance bending Reported to the Central Bank 31 October 2003 Total (122 Banks) --.---- L .- 2,835.68 ( 462,691 1 6,1291 I-- MABS Banks (inc While the MABS loans are only about 10% of the acknowledged microlending by regulated banks in the country, MABS's influence is greater than the concentration figures indicate. The MABS technology is recognized by the central bank as a substantial, effective technology for microfinance, and the central bank does refer banks that are interested in microfinance loans for individuals to MABS. (It will refer inquiries on group lending to Card Bank; however, the central bank does not endorse any lending technology per se.) Microenterprise Access lo Banking Services (MABS) Program Evaluation Listed below are the targets that were established for the second phase of the project, and commentary on progress towards completion. Note that the original numerical targets specified in the contract have mostly been reached and exceeded, and over the course of the project Chemonics and USAID have agreed to manage towards greater targets. Expansion of the number of participating bank ~mits from 60 to 100; As of December 31, 2003, the program has 11 7 active bank branches, and 180 branches has been selected as a new target. In the remaining months of the project some 32 new banks, some with multiple branches, are slated to join the program, and there should be some expansion into new branches by some of the current participants. While the 180 branch target may not be reached, the total branch number should be close. Expansion of the total number of microborrowers from 18,000 to 50,000 (with at least 12,000 of the new microboirowers coming from new participating banks); As of December 31, 2003, the total number of' active borrowers was 43,216, including 8,237 aclive borrowers f'rorn the new participating banks. 57,000 borrowers has been set as a new target. This total should be achieved and exceeded by September 30, 2004. Expansion of the total number of microdepositors from 45,000 to 90,000 (with at least 15,000 of the new microdepositors coming from new participating banks); As of December 31, 2003, the total number of microdepositors was 188,669, including 8,779 depositors f'rom the new participating banks. The new target of 200,000 should easily be reached. Installation of the "MABS Approach" in at least three new participating banks in the "Muslim areas" of Mindanao (i.e., in the ARMM Region or in non-ARMM municipalities with substantial Muslim populations); Three new banks in the ARMM-CAAM region have been added, bringing the total number of banks serving substantial Muslim populations to eight. Significant expansion of the number of Muslim clients of at least five already participating banks in non-Muslim areas of Mindanao; Banks in the region as necessary have been provided with additional institutional support, as well as training to expand lending to religious and ethnic minorities. Separate statistics on amounts and numbers of loans to minorities are not available. However, one of the banks reports that over 60% of its clients are Muslim. The remaining four banks report a 10% to 20% participation by Mtislim clients. Microerilerprise Access to Banking Services (MrlBS) Program Evaluation I ! A11 effective "anti-backsliding" program aimed at helping assure that banks which have "graduated" from the MABS Program continue to adhere to MABS Approach will be in place and being implemented by RBAP; Developnzent of anti-backsliding nteasures through a variety of nzeclzanisnzs, and not only througlz the XBAP, has been and continues to be the major focus oftlze MABS project at present. Anti-baclcsliding measures include the certijicatiorz program, tlze EAGLE award, royzdtables andfora for MABS barzks, as well as perj5onnarzce nzoizitoring. Notwithstanding the aforenzentioned, the most eflective anti backsliding tool lzas proved to be tlze MABS approach. itseyand its positive impact on the banks' operations. All banks interviewed conjhned their interest in continuing the application of MABS approach and principles into the bank's operations, because of profitability, benefit to the conznzunity, and tlze positive eflect on other bank operations. Full development of the capability of RBAP to install and strengthen the "MABS Approach" in Rural Banks in Mindanao and throughout the Philippines. Af.11 capability to deliver the MABS approaclz has been developed in tlze Microfinance Service Providers (MSIJs). At present there are two for-profit consulting Jcirnts that are providing training and technical assistance to fourteen banks, with more banks in the pipeline, thus the MABS in,rtallatiorz capacity has been substantially expanded past the project capabilities. Additional capacity in Mindanno may be appropriate. The changing role of tlze RBAP in the course oftlze project is discussed below. The principal secondary targets are: Development and dissemination of more appropriate bank management software; MABS has worked with local software developnzent firnzs lo develop tlze RB2000 bank management sofiware, which to date lzas been installed in 61 rural banks. This is a quality product tlzat is appropriate for the market, and a good value for a very conzpetitive cost. The developnzent and achievements of XU2000 are discussed in a separate section below. Bringing,about a regulatory envirorlment more supportive of the spread of nlicrofinance in banks; The changes in the regulatory environinent have been a great success story for nzicrofinance in tlze Philippines in recent years, and MABS has been a signijkant good exanzple of rnicrofinance lending for the industry and the central bank. Iiegulatory reform is discussed extensively in Annex C. Microenterl)rise Access to Banking Serl~ices (MABS) Program Evnlrcatiort a The development and utilization by participating banks of a "credit reference bureau." This is a major on-going activity for the MABS project, which is working with the Bankers Association of the Philippines to incorporate the rural banks into the "blacklist" dntabase to access and provide irzformatiorz on borrowers. This is an ongoing project in which MABS is assisting the RBAP and the rural banks to lzave representation and input in the eventual development of a fill1 service credit bureau, however, it is not a major focus of the project. Special Activities In an effort to attain several of the key targets listed above, the contractor has undertaken two "special activities" during this second phase of the MABS Program. These are: An effort to/'streamlinen and reduce the costs associated with the installation of "the MABS Approach" in banks wishing to offer microfinance services. This involves, principally, a detailed assessment of the actual need for and value of each aspect of the training and technical support provided by the Program to each new participating bank, with an eye to eliminating all training and technical support that is not shown to be essential to the effort; and a reduction, whenever it seems possible, of the amount of "in-bank technical assistance" provided by the Program to pasticipating banks as they learn the MABS Approach-with "classroom training" for multiple banks substituting for the "in-bank technical assistance." The "streamlining" effort has been successfil, with the banks of diflerent rollout groups pei-Jorming comparably. An effort to accelerate the pace with which additional Rural Banks can be provided the assistance they need to install the MABS Approach-principally through having the Program train personnel from several organizations in the theory and practices associated with installing the MABS Approach into interested banks, and then "licensing" those other organizations to do so. (With existing resources, the Program is able to assist only about 20 banks per year to install and implement the MABS Approach. Given that there are literally hundreds of Rural Banks that wish to use the MABS Approach, it could take decades to respond to all requests. By training and licensing other organizations to also install the MABS Approach into interested banks, it is hoped that some 60-100 new banks per year can be brought into the Program). Three MSPs were trained to deliver the MABS approach on a for-proBt basis, and two of these firms are actively soliciting and working with bank clients. The active MSPs lzave . skilled stajfand afliliated consultmzts who are getting strong reviewsfor their training and conssdting skills. Microenterprise Access to Banking Services (MABS) Prograrn Evaluation Other Measures Cost Effectiveness I~istorically this lzas been measured by calculating n "cost per borrower, jJ determined by dividing total project costs by the curnulalive number of new borrowers. Through 31 December 2003, project cost totaled about $1 O,5OO,OOO, and total cumulative borrowers was 11 1,302, for a cost of about $94 per borrower. The projected cumulative number of - new borrowers tl~rough 30 September 2004 is 125,000, and comparing tlzis to the total contract amount oj'$12,000,000 suggests a cost per borrower of about $96. We note tlzat this figure is only meanirzgful in comparison to similar USAID projects, does not include benefits such as the development of tlze RB2000 software or the eflects of the MABS technology for the benefit of other bank lending programs, and does not consider tlzat the lending technology will be used for many years in the future, efictively 'yree" for all new borrowers. Interestingly, virtually all the account oflicers for tlze bnnlcs ' microlending programs are nzen, and some of tlze banlcs indicated tlzat they only hire nzen for tlze job, which includes riding a nzotorcycle and going to neighborhoods described as dangerous. In most countries would be cause for concern, however, as tlze great n~ajority of the MABS clients (80%+) are women, this is not considered an issue. Sustainability As is described elsewhere in tlzis report, our view is tlzat "the MABS approach" to microfinance will be widely used afler tlze close of the project. While the project originally contemplated housing MABS in tlze RBAP aj?er the close of tlze project, there is concern about the appropriateness ofthe RBAP to be tlze MABS advocate, both from the central bank which would prefer tlzat tlze technology not be identified with a particular trade group, and because only a minor number of RBAP members are MABS-certijied. While strengthening a local institutional counterpart is the standard for a USAID project, in tlzis case the MABS technology lzas been adopted by a host of rural banks tlzat will continue to use it because it is profitable for thenz. This type of sustainability is actually the most important and valuable lcind Microenterprise Access to Banking Services (MABS) Prograni Evaluation THE MABS APPROACH The MABS approach generally comprises the following steps: Market survey to determine potential market Market research to determine loan term, amount preferences Lending steps: Orientation-one to two horns to introduce new borrowers to the program. Infosmation sheet completed to qualify clients (e.g., in business for one year, over 21 years old, within 30 minutes of branch). Qualified clients complete an application form with the account officer, including identifying a co-borrower and two guarantors. / Credit investigation and background investigation (CIBI) completed Cash receipts analysis-estimate of cash flow available for loan repayment-completed. CIBI and analysis typically takes a week for new clients; loans for repeat clients are processed in a day. Promissory note and interest disclosure statements signed; client may also pledge household goods as collateral for loan. Loan is disbursed. Account officer collects loan payments; repayment monitored. MABS loans are usually individually-based, but there is some group lending. Additional steps for group lending include several group meetings before loan disbursement and weekly meetings thereafter. Initial loans are usually PhP3,OOO - 5,000, and loan limits are usually PhP40,OOO - 50,000. Increased loan amounts for subsequent cycles are based on good repayment history and cash flow coverage. Once the program ceiling amount is reached the loan will make greater loans, but will take hard collateral (real estate). Loans are made for greater amounts if the overage is covered by the deposit balance. Interest rates range from 2 to 3.5% per month flat rate, plus a loan fee of 2.5 - 3.0%. Loan terms range from two to six months. Repayments are most commonly done on a daily basis for first loans, but may be switched to a weekly or even semi-monthly or monthly basis after the first few cycles. Typically the account officers collect payments from clients, rather than clients coming to the branch to make payments. In many respects the loans parallel the experience of "516" or "Bumbai" lenders, who typically lend on a daily repayment basis and collect at the bossower's place of business, who Microenterprise Access to Barzkirag Services (MABS) Program &valuation were the only loan source for most of the borrowers prior to the MABS program. The borrowers' greatest appreciation for the program is the relatively lower interest rates: the 516 lenders charge 10 - 20% per month. (The "516" name comes from lending five, and collecting six later.) The MABS program also includes a 5 - 10% forced savings program. Savings are not available to borrowers until they finally withdraw from the program. Per the banks, the borrowers do not wish to liquidate the savings, prefessing to borrow against them (e.g., a loan . of PhPr75,000 secured in past by PhP25,OOO of savings) and treating them as a nest egg to be used for non-business puxposes such as children's education. In addition to the forced savings MABS has worked with the banks to mobilize savings more broadly. This has been an unexpected benefit for the banks, which historically had riot sought such grassroots sources of monies to lend. Some of the banks do borrow from the Peoples Credit and Finance Corporation, a government lender, for onlending both for group lending and for individual microfinance loans. There is also some boirowing from the central bank's rediscount wihdow. One of the major questions for MAIfS has been the provisioli of measures for "anti￾backsliding," i.e., assurance that the banks will continue to follow MABS principles and provide a MABS product after the end of the project. The major factor in preventing backsliding is in the profitability and performance of MABS for the bank. The microfinance loans are seen as highly profitable. The interest ir~corne is higher than in "regular" loans (2% to 3% flat monthly vs. 1% to 2% monthly). A 3% to 4% service fee, mandatory savings of 5% to lo%, and daily repayments result in very high effective interest rates and, thereby, income. While the program does require considerable staffing, salaries are low enough that the product profits, even based on low loan amounts, are ample to cover costs. The participant banks' portfolio concentration in microfinance varies substantially, from minuscule percentages to almost 50%. The banks that have been the most successful in MABS typically have a loan portfolio percentage of around 10% or more. This percentage both represents a diversity in assets for the bank, and a portfolio concentration that is significant enough to warrant the sort of attention that the development of a new line of business deserves. Another measure of success, or failure, for the banks in accepting microfinance can be measured by the Portfolio at Risk or PAR. The project has been emphasizing quality, and those banks with a lower PAR continue to be the most enthusiastic about the program. In one case a bank with a rising PAR appears to be letting its portfolio run off, apparently due to a disenchantment with the program. A cursory review of credit files suggested that the bank was not following the CIBI and cash flow analysis steps, and needs to get back to basics, Microenterprise Access to Bnrzkirzg Services (MABS) Z'rogrant Evalz6ation MABS BROADER EFFECT ON RURAL BANKING A major outcome of the MABS program, aside from the increase in lending to ~nicroentrepreneurs, has been the effect of the MABS policies on the management and policies of the banks. This includes: Estimation of projected cash flow, hence ability to pay debt service (rather than only looking at collateral value) PI Use of management reports for planning and review; better tracking of profitability, including by product line Zero tolerance towards past due payments I. 8 ! Measurement of portfolio risk by calculation of portfolio at risk (PAR) rather than merely counting past due payments Market research prior to opening a new branch or introducing a new product to sample the market for interest, volume estimates, and desired loan terms Reducing loan terms and repayment to coincide with cash flow 1 Zero tolerance is perhaps the most important message in MABS both for microfinance lending, and for the lending operations of the rural banks in general. Zero tolerance refers to not tolerating late payments, i,e., immediately following up on any client who misses a payment or otherwise does not perfolm. A measure of zero tolerance is a low PAR, a measure of the risliness and likelihood of non-payment of the loan portfolio. Historically it appears that the banks overly relied on their collateral positions, and did little to immediately pursue any non-payments or late payments. The MABS emphasis on zero tolerance has impressed upon the banks the need to actively seek payments on all past due loans. Measurement of PAR has been emphasized along with zero tolerance, but it is important to understand that these are not the same thing. In the past the banks were only recording past due amounts of loans, thus were underestimating the riskiness of their portfolios. Jt is significant that they are recognizing that PAR is a much more accurate measure. However, some of the banks seem to have the attitude that zero tolerance means that they should always have a 0% PAR, and not take any risk at all in approving loans. While a prudent approach to lending is expected, and embodied in MABS, risk of non-payment is endemic to lending, thus some PAR is reasonably expected. An overly cautious approach to lending suggests that some good loans are not getting made, thus the lender is foregoing income and profits. It is also expected that the increased level of profits will more than offset some low level of loan losses. Microenterprise Access to Bankit~g Services (MABS) Program Evaluation i 1 Issue: Low Volume The lack of volume is really the only significant criticism of the MABS program. As of December 2003, for instance, on average each loan officer was handling 99 loans, including group loans. This is quite low by international standards. The Microbanking Bulletin (April 2001) cites field officer productivity averages ranging from 134 (Asia Pacific, n=6) to 420 (Asia Large, n=420) to 452 (Asia South, n=9) for an overall average of 306. the Mix Market statistics for 2002 on staff productivity cite the following figures: EMT, Cambodia387 FINADEV, Benin 284 FINCA Uganda, 213 Compartamos, Mexico 195 Note that these are average staff productivity figures, which are lower than account officer productivity statistics, because staff also includes support personnel. While lower productivity figures are expected for individual lending in remote areas, the rural banks are generally lending in the local marketplaces, hence enjoy a concentration of borrowers. We note that while MABS is generally an individual loan product, a substantial number of group loans are included in the total borrower numbers. Two reasons are cited for the low volume, the zero tolerance policy and the fact that many loans are on a daily payment amortization. These arguments do have some merit. The bias towards quality at the start of the program was understandable, given that the banks7 existing postfolios had substantial past due loans, and it was critical to orient the banks to the importance of prompt collections, particularly because the new microloans would not be collateralized. The daily payments also take five times the time to collect as a weekly payment loan. Notwithstanding, these are not overarching reasons for low portfolio volumes. It is worth noting that target portfolio volumes seem to always be set low. Punla indicated that it encouraged banks to set a maximum loan volume target per account officer of 120 loans; ARMlDEV suggested that 150 should be a maximum for group loans and 60 for individual loans. At least one of the banks targets 150 loans, and noted that one of its account officers got to a 170 loan port~olio by insisting that borrowers accept a weekly or bi-weekly payment schedule. USAD noted this as a sllortcoming and last November and asked Chemonics to have a consultant review the reasons for low loan volume. It appears that the recommendations will be publicized at the next annual RBAP convention, in March. Additionally, MEDCo has requested that a day long meeting be held to discuss the productivity issue and find ways to address it. w The program could be improved, both from USAID7s point of view because more borrowers could be served, and from the banks' perspective, because they would realize more income and profits. There are several ways this could be done, as described below. It will be important for the banks to do their own planning reviews, as well as to survey bosrowers, to determine what combination of steps will be the most effective on an individual basis. Microerzterprise Access to Bankirtg Services (MABS) Progrant Evaluation Go from daily to weekly payments. Many of the banks indicate that they do make this change after two or more loan cycles, but there is a convenience to borrowers in that they only need to accumulate small amounts to make a payment, so many borrowers are still on a daily payment program even after several loan cycles. None of the banks seem to offer a reason to borrowers to pay weekly, such as a lower interest rate, but this might give the needed incentive. - Eliminate duplicative steps in the loan approval procedure. As a part of the initial desire for assurance of high quality the program focused on instituting several steps in the CIBI process, in part because it was not sure what the best assurances of quality would be in the market. All of the steps are still in place (which may be appropriate) and additionally the banks frequently have instituted additional screens. Overall the effect is that, for many banks, the approval process is redundant, which costs time and money but does not realize any additional benefit. A less cumbersome approval process would allow more time to make new loans, and likely would attract more boi-rowers. t For other banks, the PAR is going up (and loans per officer down, because of attention to past due loans) because approval procedures are not being followed. Loan files have been reviewed that indicate that after several cycles CIBI procedures are not being performed, and cash flow analysis is not being calculated. In these cases a "back to basics" approach sho~ild result in more, better quality loans. Increase loan sizes. Many of the banks have put a limit of PhP25,000 - 50,000 on the microfinance program, and thereafter seek hard collateral for loans. (Note that the central bank has designated PhP150,OOO as the top limit for microfinance loans, Le., loans that do not require hard collateral.) The banks appear to be limiting their lending opportunities, despite having proven strong borrowers. Many of the banks do appear to be reconsidesing these limits, (Note that allowing larger loans will not increase the loans per officer statistic, but will increase the amount of loan portfolios.) Market limits. Many of the banks are only soliciting loans in the central markets, and at this point are relying only on word of mouth to develop new borrowers. Both the sales approach and the geographical targets could be reconsidered. Collection points. In most cases the banks follow the 516 lenders' policy of collecting from the borrower at their place of business, but we note that they usually insist that borrowers who live more from a certain distance from the branch (3 - 5 kilometers) come to the bank to make payments. It would take less time for the loan officers to do collections, especially from dispersed clients, if boi~owers could be incentivized to come to the bank to make payments. Account officer incentive programs. Incentive programs are regularly used to increase loan officer portfolio volumes while requiring quality. Some of the banks are using incentive formulas which call for portfolio peso volume, number of loans andor number II/Iicroewterrise Access to Banking Services (MABS) Progra~n Evnluatiorz of new loans in the month, and low PAR, based on a formula suggested by MABS.' At present the number of loans required to get the maximum payment is set low, at 100 loans; some of the banks are revising these upwards. CLIENT RETENTION A related issue to the question of volume is client retention. When asked the banks claim that - client retention is high, but anecdotal infomation indicates that this is not the case. Reasons for leaving the program include a desire to get use of forced savings (which are typically available only when the client ceases borrowing, not when individual loans are repaid), low loan ceilings, and the rigorous requirements for co-borrowers and guarantors.for relatively low loan amounts. The retention question does deserve more attention, and the banks should be individually encouraged to do client research to first measure, then increase, client retention. All of the reasons cited above for client defection can justify appropriate changes to lending pol'cy to continue to serve borrowers. A maxim of any business is that it is much / less costly to retail1 a client than to recruit new ones, and this is certainly true in a business with high administrative costs such as microlending. MABS has developed a comprehensive internal control module specially designed for microfinance operations. Among other features, the module includes a warning (or "red flags") system aimed at alerting managers and supervisors on potential fraud, and specific control measures for daily collections. Internal auditing functions arc present, pasticularly among the largest banks, and even some small ones have hired internal auditors. Overall, the banks interviewed affirmed their internal controls and audit functions have been substantially strengthened under MABS. Some banks report fraud cases, usually in small amounts, detected at early stages because of effective internal controls. The majority of these cases are linked to payment collection. Only two cases involved relatively large amounts and weak internal controls. Effective internal controls do not eliminate fraud but allow the banks to prevent andlor detect it at early stages. Notwithstanding, as the bank portfolios expand strong internal controls become pasticularly important given the account officers' role in payment collection. Most banks see the on-site payment collection as an important service to the client. The use of Palm Pilot technology, under pilot testing by some banks, is expected to make substantial contribution towards assurance of payment collection. However, MARS should continue pressing the banks to move clients, particularly repeat ones, to pay directly to the bank on a weekly basis, so that payments are not only handled by the account officer. - ' Interestingly, the formula includes a sharing percentage for other branch staff, so that performance is a broader responsibility than just a single account officer, which we applaud. Microenterprise Access to Banking Services (MABS) Program Evaluation As of December 2003, MABS has eight participating rural banks from ARMM-CAAM. Three banks have been added since August 2003. Their combined portfolio totals PhP48,131,505 for 6,653 clients. With the exception of one bank with a rate of 14.15%, the PAR >30 is fairly low, ranging from 0.6% to 3.5%. Two of the banks received Eagle Awards for 2002 and will receive it for their performance during the year 2003 as well. - As pa11 of MABS integration efforts, the banks participate actively in the managers and supervisor forums. One of the banks recently participated in an inteinational seminar in Morocco. Some of the banks also participated in the visit to microfinance institutions in Bolivia, along a large group of participating rural banks. Following MABS recommendations, some banks have hired experienced managers from the commercial banking sector. All banks interviewed affirm MABS has made major contributions to operations, profitability and image. They all were in the process of expanding their MABS approach lo other branches (two to three new brancheshetail outlets incaverage), existing and planned. M~BS'S major contributions, as expressed by the ARMM-CAAM banks include the following: I MIS: RB2000; Microbanker; Palm pilot technology Improved image in the banking community Improved image in the community Staff professionalization Infrastructure improvement Improved operations: higher performance of regular portfolio; increased deposits; increased profits Five rural banks, Isulan, Koronadal, Tacurong, Kapatagan and ~alabdn, are Christian-owned but serve ethnic and religious minorities, specifically Muslim clients. our of them have branches in predominantly or significantly large Muslim populated areas. Tacurong has branches in Sultan Kudagaran, and Maguindanao; Kapatagan has branches in Iligan and Pagadian; Malaban is located in a predominantly Muslim municipality. STREAMLINED MADS APPROACH , I The MABS Approach is a four-module training course and technical assistance program aimed at preparing participating rural banks to provide microfinance services (loans and deposits) profitably. It follows an eight-step process spread over a period of six month. Initially the provision of technical assistance and training was cal~ied out on an individual basis under what is referred to as the Traditional Approach. In 2002, in an effort to streamline Microenterprise Access to Wanking Services (MABS) Program Evaluation the installation of the program, MABS introduced the Streamlined Approach. The one-on￾one bank training was substituted by classroom training, some aspects of the technical assistance were standardized into training modules, and in-bank technical assistance was limited to aspects that needed to be tailored to the banks needs. The streamlined approach became a highly standardized, systematized version of the "traditional" approach. The standardization has made the MABS Approach a systematic, well-structured, more efficient approach to training and technical assistance. It has given MABS legitimacy, - effectively a trademark, as a lending approach. In terns of cost savings, the accelerated approach allows for cost reductions by (a) eliminating the redundancies in a less systematic traditional approach; (b) substituting some aspects previously addressed through technical assistance with classroom sessions; and (c) facilitating scale and replication. The installation of the streamlined approach, as with the predecessor, takes a period of six months. While the installation period remains unchanged, the new version aims at more effectiveness and consistency. It has been good in both counts. The high level of standardization of the new version has been a major contributor to the establishment of the NSPs and the MABSTeRs. It has facilitated the provision of the MABS approach through multiple providers (institutions or individuals) in a uniform manner and quality. Because it covers all aspects of the implementation of microfinance services, from product design to management and control, the streamlined approach leaves little room for deviation, interpretation and/or improvisation. The new version could be used to replicate the MABS Approach, and results, in other USALD progsams worldwide. The streamlined approach has been used with the fifth and sixth rollout banks2. A coniparison of portfolio performance by rollout-groups (Annex D) shows that the performance by rollout has improved over the years. As of December 2003, the PAR >30 days, is lower for fifth and fourlh rollouts and increasingly higher for second, first and pilot rollouts. These results indicate a growing improvement in the MABS approach rather than of a better approach. Performance by rollout after one year in operations (Annex E) shows a similar pattern but with better. performance overall for the PAR >30 days. The original plan for MABS contemplated housing the MABS technology in the Rural Bankers Association: RBAP staff would have the expertise to bring the product to the rural bank members, would monitor MABS portfolios, and othelwise would be the home for individual lending technology expertise in the Philippines. Over time, these roles and responsibilities, initially handled by the MABS project, would be handed over. In reality however, RBAP's interest in MABS has depended upon the interest of the president of the association, who is only elected for a year at a time. (A p1a.n to also elect a president-elect, who will serve on the RBAP board for a year befose serving as president, is ---- The sixth rollout banks started in December 2003. Their statistics are not relevant to the analysis. Microerzterp~ise Access to Barzking Services (MABS) Program Evaluatiort now under consideration to give the organization more continuity.) While MABS was strongly supported by the RBAP leadership in its first years, a recent president did not feel that the MABS program was benefiting the wider membership, and essentially eliminated the staff positions that had been created to take over the MABS functions. The cuuent president of RBAP has recommitted to MABS, and has reinstituted technical assistance into RBAP, by appointing two program managers who are responsible for MABS and rollout of other technical assistance. The program managers and eight account officers started training on the MABS Approach. The account officers manage 39 rural bank federations nationwide and are responsible for disseminating information, program coordination, and information gathering. The program managers are also expected to be responsible for monitoring the activities of the MSPs, Learning Centers, and MABSTeRs as well as implementing the EAGLE and MABS certification components. In'general the staff seems interested and committed to bring MABS to RBAP. Past and cur-rent issues over MABS permanency under RBAP do not seem a major concern. Notwithstanding its renewed commitment to MABS, RBAP is not ready to take over MABS responsibilities yet. The program managers, while personally qualified and capable, are new, and have only been in place since November. There is not enough time from here to the current project end to transfer the knowledge and skills developed over the past five years. In order to oversee the work of the MABSTeRs and MSPs they would need to have the knowledge and capacity to evaluate and provide articulated recommendations. Furthermore, the Rural Banks see RBAP as too politicized to manage the program transparently and effectively. All the banks interviewed expressed this opinion, and indicated that they would simply expect to relain the MABS approach on their own. While the lack of a consistent relationship with RBAP has been a disappointment we do see a future role for RBAP as a professional services provider and institution builder for the rural banks. We feel that the stronger and more professionally run the rural banks are, the better they are able to serve the ultimate project target market, the rural microenterprises. We therefore do encourage MABS to continue working with RABP on instituting and improving constituent services. MABS SERVICE PROVIDERS A significant new element of second phase of the project was to train and license other organizations to provide the training and technical assistance to install the MABS approach in rural banks, so that a greater number of banks could adopt the methodology. This became particularly needed when the RBAP president for 2002 decided to disown the program. Several training and consulting firms were screened for inclusion in the MABS Service Providers (MSP) program, and five firms were invited to consider participating in the program. Three of these firms sent representatives (at firm cost) to a training for MSPs conducted last fall, and all three were certified as MSPs. Since that time one of the firms has dropped out of the program and two remain, Associated Resources for Management and Development, Inc. (ARMDEV) and Punla §a Tao Foundation. As is common with consulting firms, consultants are both on staff and independent, although committed to the company. It Microenterprise Access to Banki~zg Services (MABS) Program Esaluation is significant that most of the consultants associated with the firm that dropped out have re￾associated with the two firms that remain as MSPs. These are also two firms that are known for training in microfinance in the Philippines (one is the CGAP microfinance trainer for the Philippines, for instance), and were particularly interested in learning the MABS approach. The MSPs had an opportunity to showcase themselves in a series of "roadshows" that RBAP conducted across the Philippines this past fall, designed to promote participation in MABS. The MSPs selected their first and second choices for pasticipation, so that they would have clients concentrated in particular areas. 'This also reflected ties that they already have to rural banks, and of course they relied on their existing network of contacts to develop rural bank clients. At present some fourteen banks are in the trainingltechnical assistance process between the two service proviclers, and more banks have been signed up in the pipeline. The training and technical assistance provided by the MSPs are getting strong reviews, and it is interesting to note that some of the banks are opting to have board members attend modules other than just the senior mdnagement program. We think the choice to go to MSPs was a good one, and will serve the rural banks well in the future. A couple of issues have arisen concerning the MSPs which should be mentioned. One of these is the cost of the MABS technology, which runs around PhP150,OOO plus pass-throughs of about PhP30,OOO. Some of the banks are objecting to this cost, considering it high for a training program. (Note that MABS is subsidizing about half of this cost.) We think this is largely a problem of marketing. A discussion with several presidents of MABS participants suggested that the program needs to be promoted not as just another training program, but effectively as buying a new culture, a totally new approach to lending. Another approach to the cost question was broached by central bank officials, who suggested that they considered that the banks purchasing the MABS approach were really buying a whole new technology that would be used over several years, an expenditure that would not recur on an annual basis. For this reason they suggested that they would find it acceptable for the banks to capitalize the cost of the MABS package and amorlize it over a three-year period, like buying new software. This tax treatment is attractive to the banks because it will smooth emiings, and should be investigated further. While ovcrall we do not consider cost a real issue, we do note that the smaller banks which are less likely to be able to pay are in Mindanao, and costs to deliver the program there are the highest. One of the MSPs suggested that MABS conduct an additional training of trainers there, so that there would be more local consultants that the MSPs could draw upon, and thus keep costs down. We think this idea has merit, and should be pursued. A tie-in may also be possible with the Learning Center at the Rural Bank of Sto. Tomas, which is providing a range of MABS and non-MABS training. Microenterprise Access to Banking Services (MABS) Progrant Evaluatiort As discussed above, the strongest anti-backsliding measure for the program is the pel-foimance and profitability in the banks themselves. As long as the technology continues to make a profit for the banks anti-backsliding should not be an issue. MABS has established a set of additional anti-backsliding measures aimed at assuring that participating banlts continue to adhere to MABS Approach after "graduation." MABS has implemented a contract/agreement by which graduated banks commit to continue reporting MABS statistics, adhere to MABS principles, and accept MABS monitoring. They also commit to participate in the EAGLE ratinglaward program, In return, the banks keep the right to participate in MABS training and seminars for a nominal fee. They are invited to attend fora and roundtable meetings. The agreement is renewable and valid for two years (or project end). All graduated banks have signed the agreement. We note that the banlts find it particularly conducive to performance to be required t'o report their statistics on a monthly basis. Roundtable Forums: take place every six months with the participation of the banks' senior management. The main purpose of the forum is to exchange experience. The interviewed banks deemed these fora as a key vehicle for exchanging experience and dissemination of information. When asked on ways of improving these fora most cornments centered on the need to have these fora more often. Similar fora have been established for the participation of supervisors. The EAGLE award: a rating system based similar to CAMf3L or PEARLS enable MABS to assess compliance with MABS principles while acting as an effective incentive/motivational tool. The EAGLE award is highly regarded and anticipated by the rural banlcs. MABSTeRs is a program to train selected staff around the banks to have a particular expertise in MABS, to be a major resource for their own banks and to occasionally provide training for others. Some years of experience and a recommendation from the bank is required for entry to the program, and it is regarded as prestigious and career￾building. We are impressed with the knowledge and performance of the MABSTeRs, and have recommended to individual banks that they get staff into the program. The MABSTeRs training follows the same four-module course of the MABS Approach but this time ad.justed to be a training of trainers. The five-day intensive course is provided by MABS staff and other MABSTeRs. Its quality is high and likely to increase the participants' capabilities and skills to provide advisory services. This could also actually be very effective as refreshment course for supervisors, most of who are responsible for the training of new recruits. In fact, most of the participants when asked about the reasons to attend gave their interest in improving their supervision and training skills as the main reason. Microenterprise Access to Banking Services (MABS) Program Evnlualiorz I / Leasning Centers are an effort to establish banks as institutional resources for other banks, other than direct competitors. The intent here was to have banks that were willing to host other banks interested in adopting the MABS technology. This program more reflects the individual interests of the bank. For instance, the Rural Bank of Sto. Tomas has effectively created a side business as a training center, and will incorporate training rooms in its new building, now under construction. The major issue concerning post-project backsliding is where MABS will be housed post- - project. The MABS participants have indicated that while the RBAP was originally designated for this role, they consider the organization too political, and do not wish for it to perform this role. One suggestion was that it be placed with the RBAP's Research and Development Foundation (RBRDFI), specifically in the Academy for Banking in the Countryside (ABC). We note, however, that ABC only exists on paper, and also note that the profit-making aspect of the MABS installation has already been transferred to the MSPs. We generally hold the view that MABS does not have to have a home for use of the MABS technology to'continue and prosper. As discussed above, ultimately MABS is being embraced by the practitioner banks because they are realizing profits from it. The MSPs also have a strong grounding in MABS, and of course also have financial reasons to continue promoting the technology. We do consider fisther development, as well as some deviations from the current MABS approach, as inevitable, but this reflects the natural process of change that should occur. There is an interest from some of the MABS banks in forming their own professional development/practitioner group, which is being informally referred to as the "Council of Eagles." This group could effectively become the "keeper of the flame," which we would expect could continue the symposiums, forums, and refresher training that support MABS adherence. We would also expect that the MABSTeRs would be trained, and continue their activities and communication under the auspices of this group. We would, however, expect this group to be formed and funded through the initiative of the group itself. While the group could ask MABS for support (e.g., training MABSTeRs, connecting the group with international microfinance sources) we do not think that the group's activities should be underwi-itten by MADS or USAID, that it is time for profit-making banks to spend their own funds for their own betterment. When MABS and USAID set out to develop a software package to support rural banks in March 1999, their goal was to develop a system that was user friendly, feature rich, flexible, scalable, and above all, affordable to the majority of small, medium and large rural banks. With Rural Banker (RB) 2000's introduction of Version 1.0 in March 2002 and subsequent release of the BSP certified Version 1.1 in 2003, MABS and USAID have more than achieved their goals. Microenterprise Access to Bartlcirtg Services (MABS) Program Evaluatiort RB2000 is a customizable, fully integrated banking software system that can support bank operations from single branch institutions running one computer up to multi-branch networks with multiple concurrent users. The software consists of basic banking modules for deposit, lending, and general ledger (GL) accounting management. The system also has modules for financial product design (FPD); automatic teller machines (ATM); charges, fees and taxes; and system interfaces. Additionally, MABS is developing and piloting system modules to support the integration of Personal Digital Assistants (PDA), and a Bankers Association of the Philippines (BAP) Credit Bureau Negative File Information System (NFIS). Under USAID's sponsorship, the MABS project developed RB2000 using the Microsoft Visual Basic 6.0 Development Tool and operates on Microsoft SQL Server Data Base Management System (DBMS). The system runs on Windows 2000 Server and supports a variety of desktop Windows operating systems. At the time, the decision to develop on,the Microsoft platform was the best choice based on the availability of prograininess in the market and the stability of the Microsoft product line. In addition $0 the RB2000 system, MABS has implemented some other support activities to ensure the success of the rural banks in the area of MIS. One such effort is the deployment of a Performance Monitoring System (PMS) to MABS participants. The PMS enables banks to provide timely and accurate monthly perfosmance reports to MABS. The system was built in Visual Basic 6.0 Development Tools and runs on an Access data base. The PMS tracks both MABS parlicipant bank performance across numerous indicators as well as monitor Portfolio at Risk (PAR). As a part of the MABS evaluation, an assessment of the RB2000 system was conducted through meetings and information gathering sessions with the MABS team, rural banks, and the roll out companies. The time for the project was short and we needed to cover as much information as possible in order to make an assessment and draw some conclusions. To accomplish this task, we organized the analysis to capture as much information as possible in the shortest amount of time. For each area of study we looked at the following factors: Applications . Data Base Hardware/Cornmunications Back up/Recovery Securily Development Methodology Reports In addition to the interviews, we also spent time examining the physical software and functionality critical to the r~rral banks. This included several product demonstrations conducted by the two rollout companies (ASEC and ISSEC) and observing the operations of RB2000 in a selected number of rural banks. Finally, we examined the licensing agreement drafted between USAID and the Rural Bankers Association of the Philippines (RBAP). The spirit of the license is intended to make Microenterprise Access to Banking Services (MAIPS) Program Evalrcatiorr RB2000 and its source code available for free to any rural bank that wants the software and can afford to pay for installation and maintenance. GENERAL FINDINGS RB2000 is high quality software that appears stable and appropsiate for the market. Relative to other packages in its price range, RB2000 is comparable if not superior in terms of its - features and functionality. The software supports the vast majority or processes and procedures a rural bank would need to operate a microfinance lending business and to take deposits. The fact that the central bank has certified that the software meets all of its data processing and reporting requirements for rural banks is a testimony to the rigor with which this software is constructed. The above praise is even more remarkable when considering that RB2000 is provided free of charge to any bank willing to pay for installation and maintenance. The banks we interviewed dad mostly positive comments and were very happy with the software and roll out companies. Certainly the cost of purchasing hardware, operating system software, a data base, and peripheral equipment requires some investment on the part of the banks, but at a price range of PhP350,OOO to 500,000 the offering is quite reasonable by international standards.' The licensing agreement also allows banks to have the source code. To date, only three banks have requested the code. Two of the requesters are holding companies of multiple rural banks; one requester is a bank with an associated IT company. After reviewing the licensing agreement signed by RBAP and USAID in 2001, there is serious cause for concern, Clearly the spirit of the agreement is to make RB2000 widely available to rural banks at a reasonable price and to ensure that the banks are not beholden to one specific vendor for installation and suppost. The other objective is to guard against any one vendor profiting unreasonably at the expense of USAID'S investment. However, the wording of the license is so loose that any entity with noble or ill intentions could obtain the I, ,) 1' software, modify it, and sell it on the open rna~ket.~ The agreement does not even include protection for the use of the name, RB2000. ', The licensing agreement does not meet the criteiia for definition of Open Source software, nor is it restrictive enough to protect the member banks, roll out companies, or the reputation of USAID.' While this repost is not a study in intellectual property law it is clear that making the source code available for unfettered distribution and licensing carries risks and unintended consequences. For example, changes to the source code by a bank or third party could jeopardize the BSP certification. The BSP informed MABS of this risk when the -- 3 Figures for installation of RB2000 were derived from rollout companies and quotes from Mabalacat Rural Bank and Bangko Santiago de Libon. 4 Clause 2.3B of the licensing agreement gives member banks and "other requestors" a no-cost nonexclusive license to use the software, with the right to modify, distribute or license the software to any other party. See Annex F for definition of open source software. Microenterprise Access to Bardzing Services (MABS) Program Evali~atiort certification was granted in 2003. If such altered software is allowed in the market under the name RB2000 it could potentially create confusion among institutions and regulators. Such a liberal license is also a disincentive to the roll out companies to support RB2000 as their market will be cannibalized by member banks going it alone and by too many other IT companies entering the market. Even if a company re-labels the software, it is not clear that USAID'S intention was to allow companies to unfairly profit from their investment while the roll out companies are forced to distribute the software free of charge. - The Chemonics development and roll out contracts with ASEC and ISSEC go slightly further in controlling the redistribution and re-labeling of the software up to one year after acceptance of Version 1.0 of RB2000, but these contracts are not software licenses and do not detail the conditions of use and distribution of the software in sufficient detail to protect the intellectual property rights of USAD, the roll out companies, or the me'mber banks. A literal interpretation of these contracts leads to the conclusion that ASEC and ISSEC are only bound to offer the software free of charge for one year after acceptance. There is no license in place for the software itself that can be passed on to member banks or other requesters by RBAP, ASEC or ISSEC. There is also no license in place between RBAP and the roll out companies to distribute the software. Pinally, the documents reviewed for this assessment refer to Version 1.0 of RB2000. Considering that ASEC and ISSEC are now distributing Version 1.1, it is not clear if any valid software license is in place governing the use of this most current release. The feedback on RB2000 from the rural banks was generally positive. All of the banks reported a high level of satisfaction with the software once the conversion process and clean up of minor problems were addressed. Some of the common areas of satisfaction included: User friendly interface that flows logically with the work of the bank, Ijighly integrated modules streamlining data entry and eliminating redundancy, A high number of accurate and easy to produce reports, Generally good suppol-1 from the rollout companies, and Increased efficiency and flexibility. Of course, not all of the feedback was completely positive. The major complaints included concerns over the cost of the third party software and installation. A number of problems were reported in relalion to conversion from the old systems to the new system. This is largely due to two factors surprisingly unrelated to the software; errors in data entry or incompatibility with the bank's chart of accounts, and incorrect setup of system parameters mostly due to a lack of detailed and comprehensive training. A third common complaint included poor performance when requesting reports and when the number of records in the data base became larger over time. A summary analysis of the RB2000 software components is included in Annex G. Microenterprise Access to Banking Services (MABS) IJrogrum Evaluation The long term commercial viability of RB2000 is unclear at this time. Substantial time and effort was made to ascertain the financial position of the product line within the two roll out companies. Unfortunately, since both ASEC and ISSEC did not track revenue and expenses by product line, accurate financial statements were not available. The findings and conclusions of commercial viability are therefore only based on information provided by - ASEC and ISSEC in conversations. ASEC did provide some financial analysis of costs and proposals recently written to customers, but again these documents were illustrative only. From the information provided by ASEC and ISSEC it appears that both companies are not making profit on installing and supporting IiB2000. ISSEC sells the installation for PhP25,000 per module (there are three modules). Their cost basis is PhP15,OOO per person per month for labor. lSSEC has eight full time staff responsible for RB2000 with two staff committed full time to installation and conversion. ISSEC claims that if the roll out is performed w:thin the one month budgeted time, the company can make a profit. Unfortunately the majority of installations are taking longer than one month. ISSEC also does not charge a maintenance fee, but does charge PhP2,500 per day for customizations. So far, only one bank has requested custornizations beyond what is included in the base price of installation. ASEC reports that a complete rollout of the software costs the company PhP112,OOO per branch. ASEC is cussently charging PhP100,OOO per branch. While they do charge a maintenance fee of PhP1,OOO per module per month, the company is losing money on the installations. Both companies report that the IiB2000 business gives them an entry into other product sales wch as ATM and passbook printing machines. Both ASEC and ISSEC want to be able to make money on the software installations but feel pressure to keep costs down, an astifact of the original agreement with RBAP. RB2000 could be commercially viable if the companies were able to price their services based on real costs and market rates. If ASEC and ISSEC are free to charge what it costs to implement RB2000 the price would likely temporarily rise until new market entrants forced the companies to find new ways to sell the installations at a lower cost while still making a profit. Currently there is not enough competitive pressure enabling market forces to work. Despite the fact that ASEC and ISSEC are not reporting a profit, the cost of installing RB2000 is quite reasonable by international standards. The total investment of $5,000-8,000 is very low compared to what similar banks pay in other countries for comparable software. Even in the Philippines, RB2000 is a cost effective alternative to Microbanker which is generally 35-40% higher in price for the software only.' he latest price quotes for Microbanker is PhP135-147,000 for the software only, compared to PhP75- 100,000 for RB2000 fully installed. Microenterprise Access to Banking Services (MABS) Program Evaluation While there are other products available globally that could be purchased off the shelf, it is I ( not apparent that any other software vendors had a presence in the Philippines at the time \ RB2000 was conceived. Accessibility to affordable banking solutions was low several years 'I ago. Consequently, the development of RB2000 was a good decision in 2000 if only to provide another low-end banking solution to the local market. i PROGRESS TOWARDS ACHIEVEMENT OF MABS GOALS AND OBJECTIVES As previously stated, MABS has exceeded its goal of creating a viable low cost solution for I' I the rural banking market. The project not only delivered the software in 2002, but enhanced the software with a new release, Version 1.1 which contains significantly greater \ functionality and stability over the first release. MABS has gone beyond the core modules of %. 1 RB2000 and created interfaces for ATMs, PDA, and the credit bureau negative information file. I 1 MABS has drained two additional roll out coinpanies for RB2000-ISSEC and Microbanker. While Microbanker has yet to participate in any implementations of RB2000, both ASEC and ISSEC have installed the software in 61 rural banks throughout the country. MABS continues to work towards achievements of their goals and targets for 2001. As it is I I\ i 1 early in the year, the project still has time to accomplish what it has set out to do, but there are some challenges. Below is an item by item assessment of MABS progress: /' i Increase the number of RB2000 installations from 61 rural banks to a total of 125. This goal requires that ASEC and ISSEC install 30 more banks (not branches) by September 2004. This means that each company will need to sell and install 4.28 banks per month. The companies report that it takes 3 months to complete a sale at a bank. Each company has approximately 20 banks in its pipeline now. Each company has the capacity to manage five installations per month without much diflculty. The average installation takes 3-4 months. While it is possible to achieve this goal, it is not probable that the target of 125 banks up and running by September can be realized. MABS should not stop trying to reach this stretch target. The project can work with ASEC, ISSEC and RBAP to continue promoting the benefits ofRB2000. However, the reality that MABS practically has no control over the rate of sales and installations cannot be ignored. Implement a performance and quality monitoring system and develop a web site with user support, a knowledge base, and frequently asked questions. MABS has made great progress on the performance monitoring system (PMS). The application is up and running in all the MABS offices and rural banks are contributing perfonizance data weekly and monthly. The next step is to deploy the software in the banks Microenterprise Access to Banking Services (MABS) Progrant Evaluation so they can use tlze facility to monitor their own progress after the MABS project ends. There is no foreseeable reason why this objective cannot be achieved by September. MABS is also making progress on the web site. Current activities include linking to the ASEC site that hosts irzfornzation on RB2000, identifying frequently asked questions, and designing the knowledge base to be deployed on the site. This objective is also achievable by September. Conduct the RB2000 Users Conference to provide users the venue to share information, experiences, concerns, and solutions. Preparations are currently underway for the User Conference. While the initial date of late February is not likely to be achieved, arrangements are underway and MABS should be able to conduct at least one and maybe two Conferences by the end of the year. Cerlairzly the first Corzference will be possible by mid-2004. %i-ain at leht three additional rollout companies. This goal is neither achievable nor advisable and should be reevaluated by USAID and tlze project. MABS has already trained two conzpanies in RB2000-ISSEC and Microbanker. The latter has not committed to the sofbvare package. Three additional conzl~anies will saturate tlze market loo much. There are 780 rural banks in the Plzilippines. Assuming the bottom 20% legitimately does not have the capacity to purchase tlze software and another 20% have allernative software, this leaves a market of468 batiks. If there are six providers in the market, this leaves only 78 banks per provider. This number of potential customers is too low to keep the companies engaged with rural banks and MABS runs tlze risk offlight from the lower end of tlze markel. Therefore, one additional provider can reasonably be absorbed into the market while leaving enough customers to make the venture attractive to the companies. Also, since it takes lime to train a new conzpany (realistically six months or more), MABS and ASEC would only have time for one new company in 2004. POLICY AND REGULATORY ENVIRONMENT MABS's charge to bring about a regulatory environment more supportive of the spread of microfinance in banks cannot be divorced from the objectives of another USAID-supported program, the Credit Policy Improvement Program (CPIP), which provides assistance to the National Credit Council (NCC), because they interact and complement each other at various levels. While CPP has a decided advantage in assisting the NCC in formulating microfinance-friendly policies and regulations, MABS has a clear advantage in bringing to the authorities concerned the perspectives of microfinance institutions on certain policies and regulations that directly or indirectly affect them. A history of the recent regulatory changes affecting microfinance is detailed in Annex C. Microenterprise Access to Banking Services (MABS) Program Evaluation The opportunities for MABS to contribute to the formulation of regulatory framework for microfinance started in 2001. In the last three years, the BSP Institute has been inviting MABS to give a two day lecture to the BSP staff on best practices on microfinance, the MABS approach and experiences of MABS participating banks. More than a hundred staff, mostly from the supervision and examination office, attended each session. For most of them, this was the first occasion wherein they were exposed to the concept of and best practices in microfinance, as well as to how banks put into practice proven microfinance technologies. When a new government was installed in 2001, it organized the National Socioeconomic - Summit to help it formulate its economic agenda and specific measures to be adopted to sustain the economic recovery. The recommendation to encourage more thrift and rural banlts to engage in microlending by using proven technologies, such as the MABS approach, and to assign such responsibility to the BSP was adopted by the Summit. With the passage of the General Banking Law, the BSP did indeed develop a regulatory framework and an advocacy program to implement this recommendation. However, it did not promote solely the MABS approach since it is its policy to remain technology neutral. Nonetheless, bringing microfinance, in general, and the MABS approach, in particular, into the Summit's agenda is a big accomplishment in that it made various stakeholders aware of the need to develop the microfinance market as well as the existence of proven microfinance technologies. MABS helped in the formulation of the regulatory framework for MFIs through RBAP, which was an official member of the TWG that was organized by the NCC. As a member of the TWG to develop the performance standards for all types of MFIs in the Philippines, MABS lent its expertise and wealth of experience to draft such standards and complemented CPIPYs work. It had substantial influence in the crafting of Circular 409 through the Microfinance Unit and interaction with CPIP, which is frequently consulted by the Unit on matters related to microfinance regulation. What MABS has done with its participating banks insofar as monitoring portfolio-at-risk (PAR) and appropriate provisioning of microfinance loans became a vital input to the formulation of such circular. As far as the BSP supervision and examination f~~nction is concerned, the MABS approach has added discipline to the participating banks. BSP examiners have observed that MABS participating banks have generally performed better than non-participating banlts in terms of satisfying prudential norms. The BSP examiners think that the regular monitoring done by the MABS office of MABS participating banks' performance greatly help in maintaining financial discipline. Also, the use of RB2000 by MABS participating and non-participating banks has greatly eased their work in examining banks. As mentioned earlier, the BSP has been going around the country conducting seminars to promote microfinance. Some officers of MABS participating banks have been invited to these seminars. As resource persons, they readily share their experience in using best practices in microfinance and the IT system they put in place to cope with the growing demand for small loans and deposits. In the process, they show to participants a model microfinance-oriented bank. This has certainly made BSPYs advocacy work easier to do. The BSP officials have expressed satisfaction with the assistance provided by MABS in crafting appropriate procedures and standards for examining microfinance operations for Microenterprise Access to Banking Services (MABS) Program Evalzration banks and training its staff how to use them. However, in providing similar assistance in the future, it is important to consider one thing; that is, BSP views MABS as a facility for RBAP, and it does not like to receive assistance from individuals or institutions that can cast doubts to its independence. Thus, an arrangement similar to the one forged between USAID and the local contractor that insures independence of the consultants from rural banks in developing a regulatory framework for banks involved in microfinance would be an appropriate one to deal with BSP's need for technical assistance in the future. - Having enjoyed extesnalities from the MABS program, the BSP has raised some concerns regarding its sustainability. MABS participating banks are still few relative to the total number of rural banks in the system and the proportion of their microfinance loans to the total loan postfolio is still small. There is also a danger that some of them may revert back to their traditional banking operations once MABS support and monitoring system cease. Considering that microfinance has been made a flagship project of the BSP and that there are still few banks involved in microfinance, the continuation of the MABS program can greatly complement the task of the BSP in promoting microfinance. Thus, the BSP officials hope that the MAB'S program can be institutionalized and be made sustainable to ensure its continuity. They, however, have some concerns regarding institutionalizing it in the RBAP I considesing the experience in the past that the RBAP can sometimes be engaged in unhealthy internal political squabbles that endanger the viability and sustainability of an otherwise good program such as the MABS. There is a need for MABS to take a closer look at the propensity of MABS participating banks to access the rediscounting window. Occasional borrowing from the rediscounting 1 window is not at all bad. However, frequent access to the rediscounting facility raises the question of whether MABS participating banks have been able to acquire the capacity to mobilize more deposits to finance their microfinance loans. MABS has directly and indirectly assisted the BSP in crafting its policy and regulatory framework for banks involved in microfinance as well as its advocacy program for microfinance. Indeed, institutionalizing and making MABS operate on a sustainable manner will bring more benefits to the BSP. However, it is not appropriate for MABS to be :,I institutionalized at RBAP because its effectiveness can be undermined by internal political squabbles that sometimes occur within the Association. The development of a relationship between RBAP and the BAP Credit Bureau Inc. has been an important focus of recent MABS work. In developing credit bureau access for the rural banks the MABS project chose to work with an existing credit bureau rather than to develop a stand-alone operation, which we consider to be the better option. In the long run the rural banks will be better served through access to financial infosmation from a range of sources rather than only from the rural banks themselves. The other benefit, of course, is that partnering with an existing organization is an easier and less expensive choice than starting a greenfields operation. Microenterprise Access to Bartlcing Services (MABS) Program Evaluatiort The BAP Credit Bureau Inc. is a for-profit subsidiary of the Bankers Association of the Philippines, the trade group for the commercial banks, and currently has a "black list" of borowers, drawn from bank accounts closed for overdrafts, closed credit cards, and legal suits filed. Membership cost is PhP3,300 per annum, plus PhPll per inquiry. Subscribers include 29 (of 41) commercial banks, 12 (of 100) thrift banks, 9 financelleasing companies, 14 credit card companies, 8 rural banks, and an SME loan guarantee program. MABS has concluded a Memorandum of Understanding with the BAP Credit Bureau to expand its use among the rural banks. For the time being the tural banks are permitted to draw on the database for credit checks without contributing information. The inquiries have produced a surprisingly high hit rate: for instance, in January 23% of the 8 12 inquisies produced a match. It appears that many of the problem borrowers who can no longer borrow from the commercial banks are seeking credit from the rural banks. However, to date the rural banks have been reluctant to use the credit bureau services, generally objecting to the cost, because the database does not yet include microfinance borrowers. It is hoped that this attitude willlbe overcome in time, so that the rural banks will come to enjoy the benefits that a credit bureau can bring to the CIBI process. As noted elsewhere in this report ASEC is in the process of developing a credit bureau intesface for RB2000, which will enable users to easily upload client information, as well as make inquiries, to the credit bureau. The BAP Credit Bureau is planning to become a more full service credit bureau. The availability of positive credit information is supported by the central bank, which is considesing requiring that banks provide credit information on their clients for credit check references. There is a privacy legal issue, however, which cursently is an impediment and which is being addressed. Access to and regular use of a credit bureau would enhance both the MABS program and the lending perfolmance of the rural banks. It will be significant for MABS, and RBAP, to keep the rural banks tied into the credit bureau development, particularly as this influence is likely to assure that regional areas (rather than just metro Manila) are covered earlier rather than later, and that threshold levels for entry of data (i.e., data will cover small botrowers, not just borrowers with larger loans) will be set low. ON TI NU AT ION OF MABS AND COMMENDATIONS We would be remiss to not note the tremendous contribution the MABS project team have made to the success of the MABS program. All of the banks and others with whom we talked had only favorable comments and opinions about the project team, and indeed went out of their way to credit successful implementation of the MABS technology to them. We were also impressed with their microfinance knowledge and practical experience, as well as the personal strengths and skills that they bring to MABS. We expect that the staff will continue to be a vital part of microfinance scene in the Philippines in the years to come. We also note Microenterprise Access to Bunking Services (MABS) Program Evaluution ~. '! ' lj j ;'\ i I \,..,I , I:" ! -i), r ,,,. b. ... .; i I that the USAID counterparts have been effective facilitators of the program, providing strong support and direction, thus also deserve a vote of thanks for this superior project. We do believe that the MABS project should be continued for some time, perhaps 18 months to two years. At this point the acceptance of the MABS technology has tremendous momentum, and we believe it would be a mistake to suspend the project at this point, just when it is having a substantial impact on microfinance in the Philippines, and on the operations of the rural banks. At the same time, we think the attention of the project can be - shifted to special projects and particular areas of interest that are more focused on institution￾building. The pastnership with RBAP, as discussed above, was well-intentioned at the stai-t of the project, and was the best institutional choice at the time the project was designed. However, overall RBAP is perceived as too political an organization to house or keep the MABS technology when the project finally is completed, and we do not recommend that monitoring activities, certification, or granting of the EAGLE awards be transferred to them. This matter is discussed ih length above so will not be repeated here. Set forth below are recommendations concerning both activities until the end of the current project, and for a possible Phase 3. Note that recommendations for activities to be addressed prior to the end of the project are only elaborated below if not already discussed above in the report. A separate set of pre- and post-September 30,2004, recommendations with respect to RB2000 are shown at the end. Before End of Project 30 September 2804 1. Focus on increasing volume of loans in individual banks. There are a few MABS banks that are relatively big players, and if their participation in the program can be increased there will be a substantial direct benefit for the project beneficiaries, the microenterprises in the rural areas. While the course of the project has been towards a streamlined approach, in this case we feel that some one-on-one technical assistance will specifically pay off in substantial volume. 2. Conduct additional MABSTeRS training. This program is ensuring that the banks have resident MABS expertise in their banks, and an effort should be made to get all or most of the banks to have a MABSTeR. 3. Conduct a training of trainers in Mindanao to qualify more MABS consultants there. 4.. Continue promotion of credit bureau participation to the rural banks. Microenterprise Access to Banking Services (MABS) Progrant Evaluation Possible Phase Three Activities 1. Focus on scaling up and increasing volume in the banks, as well as making recommendations to the MSPs concerning immediate or a quicker transition to volume lending by new MABs banks. Not all banks are ready for scaling up. Some are at early stages of the program implementation and others have experienced PAR increases as the number of clients expanded. While all banks will require further strengthening in order to increase productivity and scale, efforts should be focused on those with the highest potential (likely four of five banks). In addition to maximize results, MABS would create a demonstration effect that the remaining banks could follow and adapt. In this regards, eligible Banks must meet or should have the capacity to meet the following criteria: Top management commitment to scale Proven ldnding product, competitively priced and efficient (transaction periods lower than five days for first loans and two days for repeat loans) A large number of branches and/or service outlets A MIS that accounts for the increased volume and generates all portfolio reports in a timely, simple and accurate manner. A deposit base and expansion strategies that ensure financing of portfolio expansion. A market that can absorb rapid expansion A adequately trained staff Adequate/proven staff incentive system Adequate internal controls The implementation of the scale up activities must follow a written plan that includes: Assessment of bank capabilities Market assessment Expansion strategies Assessment of staff training and T.A needs Staff training plan Microenterprise Access to Banking Services (MABS) Program Evaluation MABS training and T.A schedule Weekly and monthly outreach (number of clients), productivity per account officer, supervisor and branch. Profitability analysis Budget So~irces of funds Monitoring tools and indicators to assess progress towards targets. During the expansion phase MABS needs to closely monitor progress and mistakes commonly made because of pressure to pesform. Among these commonly made mistakes are: / Deviation from credit procedures (e.g. minimized credit investigation), and policies (e.g. recommending loan amounts and terms larger than advised). Staff insufficiently trained or experienced are given targets beyond capabilities. Weakened monitoring on the part of supervisors and managers because of the increased number A0 to supervise workload. Weakened analysis for repeat loans/ tendency to grant automatic loan increases without adequate attention to cash flow and repayment capacity. Develop anti-backsliding training components. The best anti-backsliding strategy is to have strong microfinance operations in the participating banks. Banks with lending operations that are highly profitable in large scale s 'I will not shy away from the MABS approach. Deviations usually occur when there is dissatisfaction with a particular product. All training and technical assistance efforts towards consolidation and expansion of the rural banks microfinance operations will also have a positive effect in the sustainability of the MABS approach. Notwithstanding, any follow-on phase of MABS should consider the development of training modules aimed at improving the banks' capacity to introduce modifications and adjustments to policies and procedures loan products in a strategic manner. In this regard, bank managers should have the capacity to analyze the effect of any modifications to credit policies and proced~~res oncosts, productivity, scale, risks and profits. 3. Develop a micro agri-lending product. Microerzterprise Access to Bartking Services (MABS) Program Evalualiort This would be a major new activity for MABS, and we expect would be developed in a fashion similar to the MABS microenterprise lending technology, although with substantial participation in the planning process by the active MABS banks. There is a substantial interest in developing such a product, and several of the banks have already taken some steps in this area on their own. 4. Work with the RBAP to professionalize its operations and increase its technical services offerings. While we realize that MABS is not and is not intended to be a bank strengthening program per se, the rural microfinance clients are better served, especially as depositors, by stronger, better-run banks. We do recommend, however, that MABS respond to RBAP requests, rather than initiating activities, and that MOUs specifying performance expectations by RBAP be created for each activity. 5. Continue and expand deposit mobilization activities. We see a great opportunity to more broadly serve microenterprise clients through providing a greater range of deposit products, which will have the advantage to the banks of supplying them with more lendable funds. This would mean that they would no longer need to borrow from the PCFC or from the rediscount facility at the central bank. Addressing the form of forced savings may also help in increasing client retention. There are a range of technologies that may be applied to give depositors more flexibility in accessing funds, which in turn will benefit the banks by bringing more deposits to them, that should be investigated. These are described a bit more in the section below on recommendations for technology. 6. Continue credit bureau expansion and support activities. Rl32000 AND MIS RECOMMENDATIONS Before End of Project 30 September 2004 1. Revisit and revise the software license for RB2000. The license for RB2000 is the most important software issue for the rest of the year. Currently the license is too loosely written allowing the source code to be obtained and potentially abused without controls in place that protect the reputation of USAID and RBAP, as well as the financial security of ASEC and ISSEC. USAID can still make the source code available. The Agency can still ensure that the software will always be free. In fact, USAID can do just about anything with the license. They simply need to determine exactly what they want to do with the software in terms of who can use, modify, and license the software. USAID then must craft the license to reflect its desires. Microenterprise Access to Banking Services (MABS) Program Evaluation RB2000 is a significant achievement for USAID and MABS. 'This software can easily be made available worldwide to USAID missions and beneficiaries if the Agency puts in place the proper controls and mechanisms. It is not necessary to give the software only to RBAP. That will only ensure that RBAP will keep the software in the Philippines. USAID should retain the intellectual property sights, draft a license that dictates exactly what it wants to have happen to the software, and then license as many other distributors as it would like around the world. MABS should hire a local intellectual property attorney to work with USAID and the project to draft the license which will go with RB2000. 2. Put in place licenses between the rural banks and RBAP Currently only a user agreement or contract is in place for the software. A legal licensing agreement needs to be singed by each bank currently using the software and RBAP. This license would be more or less the same one as above. ISSEC and ASEC also need to be signatories to the new license. 3. Complete development of PDA collections modules and credit bureau interface. These modules also represent a significant feature to RB2000 as both PDA and credit bureaus are becoming a critical component of microfinance programs globally. Having these features increases the value of RB2000 as a comprehensive solution that supports a wider variety of institutional needs. MABS must continue with this work and set a goal to complete development and piloting by September 2004. 4. Increase the capacity of ASEC and ISSEC outreach and communication of the benefits for RB 2000. The analysis performed on RB2000 banks indicates that the cost of the software is less of a fact01 than desire on the part of rural banks to make the investment. Several of the banks we interviewed were small in assets and number of bran~hes.~ Despite their size, these banks had the capacity to invest upwards of PhP500,OOO in RB2000. Therefore, it is reasonable to conclude that the resistance to adopting the software is one of mindset rather than financial cap acity. Certainly USAID has the resources to fund hardware and third party software acquisition thus making the investment smaller on the part of the rural banks. It is not advisable at this time to fully fund all the hardware and third party software expenses for the rural banks since it will set a destructive precedent that will trap USAID and the roll out companies in a cycle of subsidy and hand outs. Rather, MABS has put in place an incentive program whereby the project will subsidize the difference in price from the original Microsoft software license and the new cost of the SBS suite. Bangko Santiago de Libon is a RB2000 bank but has only three branches and less than PhPGO million in assets. Microenterprise Access to Banking Services (MABS) Program Evaluation MABS can provide technical assistance to the roll out companies by developing communication strategies, training programs, and strengthening capacity to increase penetration of the software in the market. 5. Develop a RB2000 training program for rural banks to be delivered by roll out companies Many of the problems identified by rural banks with regard to RB2000 were not inadequacies in the software itself but rather the result of insufficient training of bank staff and users. Common problems centered around set up of the software and creating parameters in the Financial Product Definition (F'PD) module. Many banks said that training was informal, disjointed, and haphazard. A rigorous user orientation and detailed training program is in order. Such a program would contribute to smoother transitions to RB2000, greater banks satisfaction, and fewer problems in conversions. The cost of training is not likely to add more than PhP5,OOO-15,000 to the cost of installation. MABS should take the lead in developing and introducing RB2000 training to the roll out companies. MABS should also monitor progress of the training in new rural bank installations. 6, Train one more roll out company As previously mentioned, there is room for one more company in the RB2000 market. MABS should begin to train this company in association with ASEC. The company must be made proficient in all aspects of the software including but not limited to application and data base maintenance, customization and enhancement, installation, conversion, and user training of the hardware, software, and network. Possible Phase Three Activities U? USAlD decides to continue with MABS after September 2004, there are a few initiatives MABS can undertake to increase the effectiveness of RB2000. -I.' Conduct a feasibility study and pilot different delivery mechanisms for RB2000. While the cost of RB2000 is low by international standards, there are a few things that the roll out companies can do to reduce the barriers to entry. One possible solution could be to adopt a service bureau model for data processing. A company would set up a data center running a number of servers loaded with RB2000. The rural banks would connect to the data center remotely over DSL or other high band width connection (microwave, satellite, or cellular). Each bank would have their data base at the center with complete security. All data entry and system access would occur via the Internet or secure dial up connection. The center might also provide services such as BSP reporting and submissions to the BAP Credit Bureau. MABS can take the lead in exploring these new delivery methods by conducting the feasibility studies and piloting innovative approaches. Microenterprise Access to Banking Services (MABS) Program Evaluation 2. Increase RB2000 penetration in southern regions Currently the majority of RB2000 installations are concentrated in the nosthern regions of the Philippines. MABS can provide assistance to RBAP and the roll out companies to increase installations in the south by implementing awareness and promotional campaigns. MABS may also consider a company from the Mindanao as the one additional company trained in RB2000. MABS can then provide, additional support to that company as they build the - installation base throughout the region. 3. Explore more peripheral technologies to complement RB2000. MABS should increase the features and functionality of the PDA technology to include loan applications, more robust background investigations, and mobile credit scoring modules. MABS should also look into other technologies such as fingerprint identification and smart cards. This exploration would be conducted in the form of feasibility studies and pilot tests of various so1uti"ons available in the market. 4. Develop Management Infosmation Systems (MIS) training for rural banks. This MIS training would go beyond simply how lo use the software. Rural banks would benefit from instruction on how to maximize the use of management information contained within RB2000 and to apply this infomation strategically to their business practices. The training could consist of a combination of technical and business training course for branch and bank management. Microenterprise Access to Banking Services (MABS) Progrant Evaluation ANNEX A PERSONS INTERVIEWED PERSONS INTERVIEWED Agency l Company - USAlD MABS Project Rural Bankers Association of the Philippines (RBAP) Credit Policy Improvement Program Department of Finance National Credit Council Automated Systems & I Equipment Cor'poration Bankers Association of the PhiIJpines --- -- BAP Credit Bureau, Inc. - Associated Resources for Management and Development, Inc. (ARMDEV) Punla Sa Tao Foundation Mindanao Economic Development Council -- Rural Banks: --- ---- Bangko Kabayan -- Rural Bank of Mabitac - The Countrv Bank Rural Bank of Pagbilao First United Farmers Bank Rural Bank of Dulag Rural Bank of Sto. Tomas -- Network Bank -- Rural Bank of Montevideo Rural Bank of Panabo - Centurv Rural Bank Cantilan Bank - Robert F. Barnes Ma. Teresita F. Espanilla John V. Owens Meliza H. Agabin Full MABS Staff --- Daniel R. Arcenas Ma. Alicia P. Libo-On - Mary Jane Ramos Maria Piedad Geron Joselito Almario Florentine A. Roque Teodoro Gener, Jr. Edgardo N. Burgos Rosanna B. Tiongson Delfiri Samaniego Leonilo G. Coronel Manuel R. Bataliones Elda M. Montera ---- Aristides S. Lineses Cristopher G. Lomboy Femy G. Calderon Atty Francis Ganzon Ernesto Mane Virailio Coronado Senen Glorioso General Huberto Rebong Natividad Yu Col Mariano S. Solis Sr Federico G. Pineda Ill Rosele Relator Solis Ma. I-ourdes S. Pineda --- Alex V. Buenaventura Percival Cantiller Jason Daauiaan " Cedric Petalcorin Rodolfo Del Rosario Gen. William K. Hotchkiss Economic Growth Advisor Proiect Manaaement specialist - TO for MABS Chief of Party Deputy chief of Party President, FY 2003-2004 Program Manager program ~anager -- Chief of Party, Credit Policy &rovementaram -- (CPIP) Deputy Executive Director President General Manager President Senior Manager Vice President - Executive Director Manager President/Chief Executive Officer Consultant - Director, Capacity Exchange Group Assistant Secretary, Vice￾Chairman, MEDCo President President -- Former President President President President Chairman President /Manager Vice President/Compliance officer President General Manager Microfinance Manager -- Microfinance Manager President President - Chairman ? L Bgbnby /$$brirb#jny ' :: )Jgiinb . -' Title .* %,, . . L 2 Green Bank Joseph Omar 0. Andaya President and Chairman of the Board Rural Bank of Mabalacat George K. Dycaico President I I Silahis Bank, Inc. Mario A. Punzalan AVP-Compliance Officer Agnes Dycoco President Banco Santiago de Libon Prescilla A. Besa Branch Manager Shanele S. Secillano Loans Bookkeeper Rural Bank of Guinobatan Paolo I-lonrado Assistant to the President Rural Bank of Makati - Rural Bank of Koronadal Rural Bank of Man Rural Bank of Datu Paglas Rural Bank df Tacurong Rural Bank of Kapatagan Cooperative Bank of Misamis Oriental Shirley Tan Comptroller John Oropesa President Jose Laaon President - Lala ~bella Vice President of Operations Dado lbrahim Paglas Chairman -- Justino Marquez General Manager Emile Lagoon -. 'vice President Nicolas Lim President Eve Cane I Manager lsidro 0. Lico Chairman Mirna A. Sescom Famare Duque General Manager Supervisor First Macro Bank Ricky Oampo Program Manager -- Rural Bank of Mabitac - Rural Bank of Siam Cecile Tanael I Vice President Rural Bank of Maranao I Datu Faisal Rasuman I Chairman I Francis Ganzon ( President Bangko Kabayan Rogelio Ceradoy General Manager Nanette Castillo Loans Manager Bangko Sentral na Pilipinas (BSP; central bank) Nestor Espenilla Wilfredo Dom-ong Alfonso J. Mendoza Rogelio A. Encinas Rollando Alexander0 Q. Agustin Mario Panopio Assistant Deputy Governor I Supervisory Reports and Studies - . - - - - Office - - - Director Supervision and Examination Department Deputy Director Supervision and Examination Department Bank Officer Ill Supervision and Examination Department Director Department of Loans and I Credit -. Deputy Director Department of Loans and Credit I Agency / Company Cora Purugganan Eduardo Jimenez Pia Roman 1 Name MIS Technical Staff Head Assistant Title Credit Department of Loans and Credit ANNEX B / MICROLOAN PORTFOLIOS AND STATISTICS AS OF 12/31/03 FOR SELECTED BANKS ANNEX C: MAW EVALUATION: REGULATORY COMPONENT MABS EVALUATION: REGULAWRY COMPONENT 1. FOCUS AND APPROACH QFTHE EVALUATION This section of the Evaluation Report focuses on one of the "secondary" targets of the MABS Program, which is to bring about a regulatory environment more supportive of the spread of microfinance in banks. This target cannot be divorced from the objectives of another USAID￾supported program, the CPP, which provides assistance to the National Credit Council (NCC), because they interact and complement each other at various levels. While CPIP has a decided advantage in assisting the NCC in formulating microfinance-friendly policies and regulations, MABS has a clear advantage in bringing to the authorities concerned the perspectives of microfinance institutions on certain policies and regulations that directly or indirectly affect them. However, in the discussions that follow, more emphasis will be given to the MABS' contributions to developing a regulatory environment more supportive of the spread of miyofinance in banks. I Figure 1 outlines the approach in evaluating the MABS relative to the above-mentioned target. Recent changes in policy, regulatory and institutional frameworks will first be discussed, after which, an evaluation of MABS' contributions to these changes will be assessed. 11. CHANGES IN POLICY, REGULATORY AND INSTITUTIONAL FRAMEWORK FOR ~ICROF~NANCE A. Policy and Regulatory Framework NCC's main f~~nction is to help the government develop a market-oriented financial and credit policy environment which will promote efficient financial markets and will help micrhinance institutions (WIs) broaden and deepen their microfinance services. In 1997, the NCC issued the National Strategy for Microfinance, which outlines the vision and objectives for developing the microfinance market. The Strategy has subsequently been incorporated in the Social Reform and Poverty Alleviation Act (SRPAA) of 1997 (RA 8425). Thus, the policy framework for developing the microfinance market has now been enshrined in a law. More specifically, Section 13 of SRPAA defines the following thrusts: 1. Development of a policy environment, especially in the area of savings generation, supportive of basic sector initiatives dedicated to serving the needs of the poor in terms of microfinance services; 2. Rationalization of existing government programs for credit and guarantee; FIGURE I. USAlD INTERVENTION IN THE MICROFINANCE MARKET (R.A. 8425) i sRpAA . e ......................................................... ~*~~,.............~...'.....=~.........~.....~............=.......~s......* Legend: SRPAA - Sociai Reform and Poverty Alleviation Act GBL - General Banking Law of 2000 (R.A. 8791) BSP - Bangko Sentral ng Pilipinas E.O. - Executive Order R.A. - Republic Act MABS - Microenterprise Access to Banking Services CPlP - Credit Policy Improvement Program 3. Utilization of existing government financial entities for the provision of microfinance products and services for the poor; and 4. Promotion of mechanisms necessary for the implementation of microfinance services, including indigenous microfinance practices. In line with SRPAA, Executive Order (EO) No. 138 was issued in 1999 that directs government non-financial agencies and government-owned corporations to immediately - rationalize, in accordance with NCC guidelines, all directed credit programs they are implementing. One of the early accomplishments of this EO was the stoppage of budgetary allocations for credit programs managed by government non-financial agencies. . A technical working group (TWG) composed of representatives from various microfinance stakeholders was formed and tasked to assist NCC to formulate a regulatory framework for MFIs. Said framework was approved by the NCC in July 2002. Among others, the framework outlines the institutional set-up for the regulation of microfinance. Subsequent to this, a TWG has formed consisting of concerned stakeholders and was mandated to assist NCC to formulate and develop a uniform set of pel-formance standards that will cut across all types of institutions involved in microfinance. 'The TWG will meet on 20 February 2004 to finalize the draft that outlines the proposed performance standards. The General Banking Law (GBL) that was enacted in 2000 contains 3 provisions that provide a general regulatory framework for banlts that will be involved in microfinance. The Bangko Sentral ng Pilipinas (BSP) subsequently issued circulars to implement the 3 provisions of the GBL. Circular No. 272 (30 January 2001) provides a definition of microfinance loans, the interest rate that could be charged on microfinance loans which should not be lower than the prevailing market rates, exemption of microfinance loans from Section 4 1 of the GBL on unsecured loans, and allowing microfinance loans as an alternative means for complying with the required loans to small and medium enterprises as stipulated in the Republic Act No. 6977. In the wake of the 1997 financial crisis, BSP issued a moratorium on the licensing of new banks. However, there was a strong demand for allowing microfinance-osiented banks to enter the market. Thus, RSP issued Circular No. 273 (27 February 2001) that provides for a partial lifting or the moratosium on the issuance of new thrift and rural banks and branches to accommodate microfi~~ance-oriented banks, the minimum capital requirement and other licensing requirements. This was subsequently amended by Circular No. 340 (30 July 2002), which sets the rules and regulations for the establishment of brancheslloan collection and disbursement points (LCDPs) of microfinance-oriented banlts established under Circular No. 273 and the establishment of microfinance-oriented brancheslLCDPs of banks that are not microfinance-oriented. This was further amended by Circular No. 365 (23 January 2003) and Circular No. 369 (17 February 2003) that relax branching regulations for thrift, national cooperative and microfinance-oriented banks. The BSP has opened a rediscounting facility to refinance the loan portfolio of microfinance￾oriented banks. The guidelines governing the rediscounting facility are contained in Circular No. 282 (19 April 2001) and Circular No. 324 (12 March 2002). Most recently, BSP issued Circular No. 409 (14 October 2003) which prescribes the rules, regulations and standards that shall govern microfinancing operations of banks. Among others, it defines portfolio-at-risk (PAR) and specific allowance for probable losses on microfinance loans. The circular took effect on 1 January 2004. Since the changes in the BSP regulatory framework for microfinance, 122 banks have been involved in microfinance, of which 6 obtained a license as microfinance-oriented banks under Circular No. 273. With the changes in the regulatory framework and growing number of banks involved in microfinance, the BSP saw the need for developing appropriate examination policies and procedures to better measure the institutional risks and performance of the microfinance operations of banks. The draft containing the suggested amendments to the existing manual of bank regulations and examination procedures has just been completed and is curreptly being reviewed by the Microfinance Committee before endorsing it to the Monetary Board for approval. There are two ticklish issues in the BSP's regulatory framework for microfinance. One is the opening of rediscounting facility for microfinance-oriented banks. The microfinance community feels that this goes against one basic principle of promoting microfinance, that is, banks involved in microfinance should mobilize deposits and should not be dependent on external funds especially if they are cheap. The BSP, on the other hand, feels that it must open its rediscounting window to microfinance-oriented banks just as it does for other types of banks. However, since microfinance loans are not backed up by hard collateral, the BSP finds it proper to develop another set of rediscounting rules for its rediscounting facility for microfinance-oriented banks and, at the same time, addresses the concerns of the microfinance community that banks would not use the microfinance rediscounting facility as a cheap source of funds. Thus, the BSP has decided to charge a rediscount rate for microfinance loans equivalent to the 91-day Tbill rate of the last auction date of the preceding month. This is one-percentage point higher than the regular rediscounting rate. In addition, the BSP has imposed stricter eligibility requirements for its microfinance rediscounting facility than those for the ordinary rediscounting facility. These are: 1. At least 1 year track record in microfinance; 2. At least 500 active microfinance borrowers; 3. A past due ratio of not more than 5 percent for microfinance; 4. A collection ratio of not less than 95 percent during the preceding months; 5, Officers and staff responsible for microcredit operations who have completed training course and with at least one year experience in microlending activities; and 6. A manual of operations pertaining to microfinance operations. The other ticklish issue has something to do with branching regulations. Circular 340 sets the guidelines for microfinance-oriented banks and banks that are not microfinance-oriented but wish to establish microfinance-oriented branches. This is definitely a step in the right direction in support of developing further the microfinance market. I-Iowever, there are restrictive provisions in that circular. One is that a branch can be established only in places not fully served by existing microfinance banking offices. It would be extremely difficult to determine whether a certain place is fi~lly served by existing microfinance banking offices. Incumbents may use this provision in the circular to lobby against new entrants. The policy should instead be to make every place, especially rural areas, a contestable microfinance market but strengthen prudential regulations for banks involved in microfinance. Another provision is that the geographical location where the new branch may be opened is overly restrictive, especially for small microfinance-oriented banks. It effectively closes the possibility that a small microfinance-oriented rural bank located in Luzon can establish a branch in one of the rural towns in Mindanao not being served by a microfinance-oriented bank. Some hnks that are owned by NGOs like Card Rural Bank can circumvent this regulation by setting up an NGO branch in other localities in competition with other microfinance-oriented banks operating there. However, this will only make the playing field uneven to microfinance players. There are also various requirements with respect to security and hours of operation that can raise the minimum fixed costs of opening a minimum scale rural branch beyond the anticipated revenues and thus depsive rural residents in these areas of the variety of financial services that a bank branch can provide. B. Institutional Framework To institutionalize microfinance within its organization, systemize its regulation of banks involved in ~nicrofinance and to send a strong signal to the financial market that it supports the development of the microfinance market, BSP has introduced important modifications to its organizational structure. More specifically, it has created a Microfinance Committee headed by no less than the Deputy Governor for Supervision and includes as members , representatives from various bank supervision and examination departments, Supervisory Reports and Studies Office (SRSO), and Department of Loans and Credit (DLC). Two members of the Monetary Board act as advisers of the Committee. The MFI Committee recommends to the Monetary Board appropriate policy and regulatory initiatives in support of microfinance within the banking sector. It is assisted by the Microfinance Unit, which coordinates relevant activities such as briefings, exposures, training, workshops arid other advocacy activities to promote sustainable arid viable microfinance programs in the Philippines. Currently, the Unit is manned by a consultant and a non-permanent staff. However, the BSP has decided to make the Unit into a permanent one to be headed by a director with 5-6 permanent staff. The Monetary Board is expected to approve the staffing pattern and the compensation scheme for this unit within the year. To strengthen its supervision and examination of microfinance-oriented banks, the Supervision and Examination Department (SED) IV created the Microfinance Core Group, which initially was composed of volunteers who are interested in microfinance. This was later replaced by the Microfinance Division under the same department. Aside from these organizational changes, the BSP has also embarked into capacity building to ensure that its supervision and examination staff have the needed capacity to supervise and examine microfinance-oriented banks. Its staff, therefore, undeiwent training on microfinance, in general, and on the use of the proposed manual of bank regulations and examination procedures for microfinance banks. Many banks especially in the provinces have not yet fully appreciated the difference between microfinancing and traditional banking, and therefore do not know that microfinancing can be very profitable when done properly using best practices. To promote the development of sustainable microfinance operations and institutions, the BSP has been conducting a one-day information seminar on microfinance in the provinces. The seminar discusses the basic concepts of microfinance, the policy environment, the different methodologies and best practices as well as many stories of lives that have been helped through microfinance operations. $0 far, the BSP has conducted such seminars in Cebu City, San Fernando City, La Union, Legaspi City and Davao City. Another one will soon be conducted in Tacloban City. Each of these seminars was attended by over 150 institutions from the banking, cooperative and NGO sectors. 111. MABS' CONTRIBUTIQNS TO THE REGULATORY ENVIRQNMENT A. Context Unlike CPIP which provides assistance to an important policy-making body, the NCC, the MABS Program was designed primarily to provide technical assistance and training to Mindanao-based rural banks in microfinance best practices. The MABS head office was then based in Mindanao. It was only after successfully completing the first phase targets that MABS included in its second phase targets the secondary target of bringing about a regulatory environment supportive of the spread of microfinance in banks. Also, the MABS transferred its head office to Manila within the premises of RBAP as it has expanded its client banks to include other rural banks in the country while retaining rural banks in Mindanao as the dominant focus of the Program. MABS' main approach in meeting this target has been through the provision of direct technical assistance to the BSP in the development of an appropriate supervision and examination manual for microfinance operations of banks, training of BSP examiners on microfinance best practices on the use of newly developed manual, and active involvement in TWGs, fora, and workshops that discuss policy regulatory issues affecting microfinance. It is in this context that MABS' contributions to the formulation of regulatory environment must be evaluated, B. Contributions MABS has had no contribution to the formulation of the first few major policies and regulations affecting the microfinance market discussed earlier. This can be attributed to at least 2 factors. First, the concern about regulatory framework was articulated only in its second phase targets. Second, MABS initially focused its activities in Mindanao whereas , most of the discussions on appropriate policy and regulatory framework for microfinance took place in Manila where technical working groups were formed and most fora/workshops were conducted. The oppostunities for MABS to contribute to the formulation of regulatory framework for microfinance have changed since 2001. In the last 3 years, the BSP Institute has been inviting MABS to give a 2-day lecture to the BSP staff on best practices on microfinance, the MABS approach and experiences of MABS pai-ticipating banks. More than a hundred staff most of them coming from the supervision and examination attended each session. To most of them, this was the drst occasion wherein they were exposed to the concept of and best practices in microfinance as well as to how banks put into practice proven microfinance technologies. When a new government was installed in 2001, it organized the National Socioeconomic Summit to help it formulate its economic agenda and specific measures to be adopted to sustain the ecoiiomic recovery. The recommendation to encourage more thrift and rural banks to engage in microlending by using proven technologies, such as the MABS approach, and to assign such responsibility to the RSP was adopted by the Summit. With the passage of the CBL, the BSP, as discussed above, did indeed develop a regulatory framework and an advocacy program to implement this recommendation. However, it did not promote solely the MADS approach since it is its policy to remain technology neutral. Nonetheless, bringing microfinance, in general, and the MABS approach, in particular, into the Summit's agenda is a big accomplishment in that it made various stakeholders aware of the need to develop the microfinance market as well as the existence of proven microfinance technologies. MABS helped in the formulation of the regulatory framework for MFIs through RBAP, which was an official member of the TWG that was organized by the NCC. As a member of 1 V , the TWG to develop the performance standards for all types of MFIs in the Philippines, MABS lent its expeitise and wealth of experience to draft such standards and complemented CPIP7s work. It had substantial influence in the crafting of Circular 409 through the Microfinance Unit and interaction with CPP, which is frequently consulted by the Unit on matters related to microfinance regulation. What MABS has done with its participating banks insofar as monitoring portfolio-at-risk (PAR) and appropriate provisioning of microfinance loans became a vital input to the formulation of such circular. One of the most impostant contributions of MABS, which cannot be officially attributed to it for reasons discussed below, is the formulation and implementation of a program of technical assistance and training to internalize and institutionalize microfinance supervision with the BSP. MABS helped in drawing up the terms of reference (TOR) and in selecting the appropriate consultants. Due to confidentiality, the Evaluation Team has not been given access to the memorandum of understanding (MOU) between the USAD and the BSP and the detailed reports and recommendations of the two foreign consultants. However, the interviews conducted by the Team with BSP officials yield the following observations: The teims of reference (TOR) for the consultants truly reflects what the BSP expected of the program. There were no amendments made to capture unanticipated issues. The expestise and experience of the chosen consultants were appropriate to the tasks indicated in the TOR. The process that was followed from beginning to end of the program (i.e, from conduct of the review of the financial statements, infosmation systems, and operations manuals of 8 banks, drafting of the amendments to the Manual of Examinations, field-testing and revision of the proposed amendments, to training of bank examiners) was appropriate. The institutional set-up that was aimed at maximizing the cooperation between the consultants and BSP (MFI Committee, MFI Unit, the Microfinance Core Group and other staff from the Supervision Examination Sector) greatly facilitated the implementation of the program. The consultants' suggested amendments to the manual of examinations have been substantially adopted. The training conducted by the consultants for the BSP staff regarding the use of the amended manual of examinations was well received and appreciated. In a nutshell, the BSP is very much satisfied with the process and outcomes of the program. More importantly, the officials of SED IV have expressed greater confidence in supervising microfinance-oriented banks now that they have a regulation and examination procedures manual for such types of banlts, and that their staff have a better understanding of microfinance operations of banlts and adequately trained to use such manual. One of the ticklish issues discussed above is the opening of the rediscounting facility for microfinance-oriented banks. With greater emphasis on savings mobilization, MABS pasticipating banks are not expected to greatly depend on such facility. Between 2002 and 2003, the BSP lent a total of PhP185.5 million to 13 microfinance-oriented banks through this facility. As of 3 1 January 2004, the microfinance rediscounting loans stood at PhP20.1 million, which is only 0.3 percent of the total outstanding peso rediscounting loans. This is indeed a relatively negligible amount on the aggregate. However, some rediscounting loans are relatively large by rural bank standards, ranging from PhP15 million to PhP21 million. Moreover, out of the 13 microfinance-oriented banlts that have tapped this facility, 9 are MABS pai-ticipating banks. Moreover, of the 9 banks, 6 have bossowed from the facility for two consecutive years (i.e., 2002 and 2003). One could argue that these banks are actually paying a penalty rate for borrowing from the rediscounting window because of the huge differential between rediscounting rate and deposit rates (6.5% vs. 3.0%). However, this can be offset by the differential in the transaction costs between accessing the rediscounting window in bulk and mobilizing a myriad of small deposits. According the staff of the Department of Loans and Credit, it takes only a day for them to process an application for rediscounting so long as the loan documents submitted by banks are in order. Considering the cussently huge spread between the rediscounting rate and lending rate, banks could potentially generate more profits by tapping the BSP microfinance rediscounting facility. Indeed, it is wort:hwhile for MABS to take a closer look at this issue. C. Externalities The MABS program has apparently generated some externalities benefiting the BSP. There are two areas where BSP has enjoyed some externalities from MABS: BSP supervision and examination function; and advocacy work. As far as the BSP supervision and examination function is concerned, the MABS approach has added discipline to the participating banks. BSP examiners have observed that MABS participating banks have generally performed better than non-participating banks in terms of satisfying prd'dential norms. The BSP examiners think that the regular monitoring done by the MABS office of MABS participating banks' performance greatly help in maintaining financial discipline. Also, the use of RB2000 by MABS participating and non-participating banks has greatly eased their work in examining banks. As mentioned earlier, the BSP has been going around the country conducting seminars to promote microfinance. Some officers of MARS participating banks have been invited to these seminars. As resource persons, they readily share their experience in using best practices in microfinance and the IT system they put in place to cope with the growing demand for small loans and deposits. In the process, they show to participants a model microfinance-osiented bank. This has certainly made BSP's advocacy work easier to do. B. BSP's Concerns The BSP officials have expressed at least two major concesns about MABS. One is that MABS is viewed as a RBAP facility, which is expected to serve first the interest of RBAP members. While the BSP welcomes comments and suggestions from various quarters on how to structure its regulatory framework, it is very much concerned about preserving its independence. Thus, it exercises caution when it comes to receiving free assistance offered from any individuals or institutions. The case in point, which is discussed above, is the assistance it received for improving its manual of bank regulations and examination procedures and for training its staff. It did not want MABS to directly provide such assistance for fear of regulatory capture. As the Evaluation Team was made to understand, the USAID provided such assistance to the BSP, but MABS was not explicitly made a party to the memorandum of understanding (MOU) between USAlD and BSP. Chemonics In'c., which is USAID'S local contractor for MABS, had to subcontract ICC Inc. to deliver the services stipulated in the MOU. Although MABS helped in preparing the TOR and in scouting for qualified consultants who are independent from its consultants working with rural banks, it has never bden given access to the report and recommendations of the consultants. Although all parties seemed to have been satisfied with this arrangement, nevertheless it shows the constraints facing MABS if it directly extends to the BSP technical assistance that has major implications on policies or regulations affecting RBAP members. Having enjoyed externalities from the MABS program, the BSP has raised some concerns regarding its sustainability. MABS participating banks are still few relative to the total number of rural banks in the system and the proportion of their microfinance loans to the total loan portfolio is still small. There is also a danger that some of them may revert back to their traditional banking operations once MABS suppost and monitoring system cease. Considering that microfinance has been made a flagship project of the BSP and that there are still few banks involved in microfinance, the continuation of the MABS program can greatly complement the task of the BSP in promoting microfinance. Thus, the BSP officials hope that the MABS program can be institutionalized and be made sustainable to ensure its continuity. They, however, have some concerns regarding institutionalizing it in the RBAP considering the experience in the past that the RBAP can sometimes be engaged in unhealthy internal political squabbles that endanger the viability and sustainability of an otherwise good program such as the MABS. IV. CONCLUDING REMARKS AND RECOMMENDATIONS This section of the report focuses only on one of the secondary targets of the MABS program - bringing about a regulatory environment more supportive of the spread of microfinance in banks. MABS' main approach in meeting this target has been through the provision of direct, technical assistance to the BSP in the development of an appropriate supervision and examination manual for microfinance operations of banks, training of BSP examiners on microfinance best practices on the use of newly developed manual, and active involvement in TWGs, fora, and workshops that discuss regulatory issues affecting microfinance. MABS brings to the members of the TWGs and the participants of various fora/workshops its expertise and experience in promoting best practices in microfinance and the perspectives of the practitioners, especially the MABS participating banks. Thus, the abovementioned secondary target appears to have been satisfactorily met. In providing technical assistance to the BSP to reform its policy and regulatory framework, it is important to consider one thing; that is, BSP views MABS as a facility for RBAP, and it does not like to receive assistance from individuals or institutions that can cast doubts to its independence. Thus, an assangement similar to the one forged between USAID and the local - contractor that insures independence of the consultants from rural banks in developing a regulatory framework for banks involved in microfinance would be an appropriate one to deal with BSP's need for technical assistance in the future. There is a need for MABS to take a closer look at the propensity of MABS participating banks to access the rediscounting window. Occasional borrowing from the rediscounting window is not at all bad. However, frequent access to the rediscounting facility raises the question of whether MABS participating banks have been able to acquire the capacity to mobilize mod? deposits to finance their microfinance loans. MABS has directly and indirectly assisted the BSP in crafting its policy and regulatory framework for banks involved in microfinance as well as its advocacy program for microfinance. Indeed, institutionalizing and making MABS operate on a sustainable manner will bring more benefits to the BSP. It is not appropriate for MABS to be institutionalized at RRAP because its effectiveness can be undermined by internal political squabbles that sometimes occur within the Association. ANNEX D PERFORMANCE 2003 MQLLOUT ANNEX E PEIXFQRMANCE AFTER YEAR ONE ROLLOUT PERFORMANCE AFTER YEAR ONE ROLLOUT . et change in number of deposit ANNEX F DEFINITION OF OPEN SOURCE DEFINITION OF OPEN SOURCE (Courtesy of the Open Source Initiative) Version 1.9 The indented, italicized sections below appear as annotations to the Open Source Definition (OSD) and are not a part of the OSD. A plain version of the OSD without annotations can be found here. In froduction Open source doesn't just mean access to the source code. The distribution terms of open-source software must comply with the following criteria: 1. Free Redistribution The license shall not restrict any party from selling or giving away the software as a component of an aggregate software distribution containing programs from several different sources. The license shall not reqyire a royalty or other fee for such sale. Rationale: By constraining the license to require free redistribution, we eli~ninate the temptation to throw away many long-term gains in order to make a few short-term sales dollars. If we didn't do this, there would be lots of pressure for cooperators to defect. 2. Source Code The program must include source code, and must allow distribution in source code as well as compiled form. Where some form of a product is not distributed with source code, there must be a well-publicized means of obtaining the source code for no more than a reasonable reproduction cost-preferably, downloading via the Internet without charge. The source code must be the preferred form in which a programmer would modify the program. Deliberately obfuscated source code is not allowed. Intermediate forms such as the output of a preprocessor or translator are not allowed. , Rationale: We require access to un-obfuscated source code because you can't evolve programs without modifying them. Since oi~rpurpose is to make evolution easy, we require that modification be made easy. 3. Derived Works The license must allow modifications and derived works, and must allow them to be distributed under the same terms as the license of the original software. Rationale: The mere ability to read source isn't enough to support independent peer review and rapid evolutionary selection. For rapid evolution to happen, people need to be able to experiment with and redistribute modifications. 4. Integrity of the Author's Source Code The license may restrict source-code from being distributed in modified form only if the license allows the distribution of "patch files" with the source code for the purpose of modifying the program at build time. The license must explicitly permit distribution of software built from modified source code. The license may require derived works to carry a different name or version number from the original software. Rationale: Encouraging lots of improvement is a good thing, but users have a right to know who is responsible for the software they are using. Authors and maintainers have reciprocal right to Inow what they're being asked to support and protect their reputations. Accordingly, an open-source license must guarantee that source be readily available, but may require that it be distributed as pristine base sources plus patches. In this way, "unofficial" changes can be made available but readily distinguished from the base source. 5. No Discrimination Against Persons or Groups The license must not discriminate against any person or group of persons. Rationale: In order to get the maximum benefit from the process, the maximum diversity of persons and groups should be equally eligible to contribute to open sources. Therefore we forbid any open-source license from locking anybody out of the process. Some countries, including the United States, have export restrictions for certain types of software. An OSD-conformant license may warn licensees of applicable restrictions and remind them that they are obliged to obey the law; however, it may not incorporate such restrictions itself. 6. No Discrimination Against Fields of Endeavor The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research, Rationale: The major intention of this clause is to prohibit license traps that prevent open source from being used commercially. We want commercial users to join our community, /lot feel excluded from it. 7. ~istribdtion of License The rights attached to the program must apply to all to whom the program is redistributed without the need for execution of an additional license by those parties. Rationale: 7-his clause is intended to forbid closing up software by indirect means such as requiring a non-disclosure agreement. 8. License Must Not Be Specific to a Product The rights attached to the program must not depend on the program's being part of a particular software distribution. If the program is extracted from that distribution and used or distributed within the terms of the program's license, all parties to whom the program is redistributed should have the same rights as those that are granted in conjunction with the original software distribution. Rationale: This clause forecloses yet another class of license traps. 9. License Must Not Restrict Other Software The license must not place restrictions on other software that is distributed along with the licensed software. For example, the license must not insist that all other programs distributed on the same medium must be open-source software. Rationale: Distributors of open-source software have the right to make their own choices about their own software. Yes, the GPL is conformant with this requirement. Software linked with GPLed libraries only inherits the GPL if it forms a single work, not any software with which they are merely distributed. "1 0. License Must Be 7Bchnology- Neutral No provision of the license may be predicated on any individual technology or style of interface. Rationale: This provision is aimed specifically at licenses which require an explicit gesture of assent in order to establish a contract between licensor and licensee. Provisions mandating so￾called "click-wrap" may conflict with importan t methods of software distribution such as FTP download, CD-ROM anthologies, and web mirroring; such provisions may also hinder code re￾use. Conformant licenses must allow for the possibility that (a) redistribution of the software will take place over non-Web channels that do not support click-wrapping of the download, and that (6) the covered code (or re-used portions of covered code) may run in a non-GUI environment that cannot support popup dialogues. Origins: Bruce Perens wrote the first draft of this document as "The Debian Free Software Guidelines," and refined it using the comments of the Debian developers in a month-long e-mail conference in June, 1997. He i ANNEX G ASSESSMENT OF RE2000 COMPONENTS ASSESSMENT OF RB2000 COMPONENTS As mentioned in the main body of the report, RB2000 is a customizable, fully integrated banking software system that can support bank operations from single branch institutions running one computer up to multi-branch networks with multiple concurrent users. The software consists of basic banking modules for deposit, lending, and general ledger (GL) accounting management. The system also has modules for financial product design (FPD); automatic teller machines (ATM); charges, fees and taxes; and system interfaces. Some other features of the system include: . Security at the data base, application, and transaction level; Clmgeable user level access rights; Flexible p,roduct and transaction attributes; Audit trail for all transactions, users, and workstations; and Automated GL posting of accounting entries. Additionally, MABS is developing and piloting system modules to support the integration of Personal Digital Assistants (PDA), and a Bankers Association of the Philippines (BAP) Credit Bureau Negative File Information System (NFIS). Under USAID'S sponsorship, the MABS project developed RB2000 using the Microsoft Visual Basic 6.0 Development Tool and operates on Microsoft SQL Server Data Base Management System (DBMS). The system runs on Windows 2000 Server and supposts a variety of desktop Windows operating systems. At the time, the decision to develop on the Microsoft platform was the best choice based on the availability of programmers in the market and the stability of the Microsoft product line. Perliaps one of the most significant features of RB2000 is the parameter dsiven architecture that allows users to create and modify deposit and lending products without the intervention of a software programmer. These parameter tables (the FPD module) are user friendly screens that enable rural banks to define all the characteristics of their products including, but not limited to interest rate, tenor, individual or group, collateralized or cash flow-based, etc. The parameter tables also allow the banks to define the data processing business rules for each product including grace periods, fees and penalties, and what to do at various stages of arrears. Another significant aspect of RB2000 is the BSP certification. In 2003, the BSP certified that RB2000 meets all of its data processing and reporting requirements for rural banks in the Philippines. This is a significant development as the certification eases the burden on the rural banks to comply with BSP requirements and it also lends credibility to the software itself. Regarding the operations of RB2000, the software is good quality and stable. The banks we interviewed reporting few problems. Of the problems communicated to the evaluation team, the majority of the complaints were related to some common sources: The Functionality Gap. There is often a gap between how a bank does business and how the software supports data processing. This problem arosc whcn tlie chart of accounts in RB2000 was not consistent with the chart of accounts at the bank. Some effort was required to resolve the gap through customization and minor programming. In some cases, banks had vastly - different modes of operation that made it difficult for the software to fully support the business. For example, one bank did not charge penalties for late payments, did not use interest rates for their loans, and did not accrue interest on their deposits. In such a case, RB2000 is not the problem. The issue is one of customizing the software to fit the way the bank wishes to do business. Parameter Setup. Since all of the data processing is driven by business rules established in the software, proper set up of the parameters during installation and product definition is essential. Idmany cases, these parameter settings were incorrect, creating errors and miscalculations. The root cause of the problem was largely due to a lack of training and user knowledge of how to set up the parameters during installation. In no cases was it evident that the software itself was the cause of the difficulties. Minor Data Processing Items. In a few cases, the system needed to be adjusted due to how the software calculated some internal figures. For example, the software rounds off at a predefined number of digits when constructing an amortization schedule. If manually developed schedules do not round off, then the software will produce a different schedule than that of the bank. In cases where customers made their payments on time, the software can still create an arrear for that account. Another example is the inability of the software to support mixed mode placements (e.g., parlial cash and check). The software did not support compound entries. A few banks reported some isolated problems with different balances between end of day and beginning of day. These incidents only occurred once on a specific day of the month. Some banks also reported complex procedures for time deposits and special savings deposit accounts. Early withdrawals and add on deposits were not easily accomplished in the system. RB2000 prevented adding additional monies to time deposits and early withdrawals required liquidating the entire deposit and then recreating a new time deposit for the remaining balance. This problem is being addressed by ASEC and an enhancement will be forthcoming. Finally, a common complaint focused on end of day processing. If the bank already ran end of day and the users had additional transactions to make, the system needed to be restored and end of day run again to sweep in the additional transactions. Users could not simply add transactions after end of day processing was completed. I None of the above issues are cause for concern. As software is being developed and rolled out to institutions, changes and modifications are commonplace. ASEC and ISSEC are aware of the problems and are responsive in addressing the issues. These fixes should in no way be intespreted as deficiencies in RB2000. 'The data base for RB2000 is Microsoft's SQL Server. This data base management system (DBMS) is one of the most popular data bases for enterprise-wide data processing applications. SQL Server is scaleable and extensible. The DBMS is also intelligent as it can be programmed with stored data management procedures directly into the data base. There is every reason to assume that the rural banks would have no problems associated with SQL Server and its suitability to their operations. Unfortunately, there were some reports where system performance had degraded as the size of the data base grew over time. These / performance problems occurred particularly at times where all of a bank's users were on line with the system and a large number of report requests were sent to the printer. There are a number of possible hardware, network, and software reasons for the decline in system performance or slow response time. The worst case scenario would be that the data base itself is poorly or inefficiently designed. This would be a case where data base tables and their relationships to each other are not optimized for efficient data processing. It is also possible that there are some coding errors in the stored procedures, but this problem is not as likely because such coding errors would be indicated by incorrect reports, errors in data in the data base, or erroneous results from some data processing routines. We were informed and assured by ASEC that the data base was fully normalized and optimized for efficiency. Given the fact that the DBMS is fully efficient, it is likely that the performance problems could be either related to hardware (referenced below), or the log files. Since RB2000 keeps an audit trail of every transaction automatically, the log file can grow substantially over the course of operations. If this log file is not cached off to a separate drive in the computer, performance can often decline over the course of business hours. The roll out companies may want to,consider advising the banks to have a separate computer hard disk and storage for the audit files. This will free the production drive store only the original transactions - thus possibly enhancing performance. Other than the performance, problems, no other issues were reported regarding the data base. Perhaps the most likely cause for performance degradation is hardware. The evaluation team was sui-prised to discover that different banks were running RB2000 on different sizes of servers. Some banks were running Pentium 111, some Pentium IV, some on Celeron Processors. Also most of the banks were running the applications, data base, log files, and printers off of the same server. Alternatively, the banks would have two servers sharing a single network switch. In such a case, if a large number of reports were sent to the printer, performance of the application and data base would decline. This happens for two reasons; (I) the data base is tied up with processing reports and cannot be accessed for normal transactions, and (2) the hardware is insufficient to process the large number of instructions required to pei-fom all of the immediate requests. Unfortunately, the roll-out companies have not been diligent about enforcing the - recommended hardware specifications for installing RB2000. While they do communicate .the recommended server configuration, many banks have chosen to implement RB2000 on lower powered servers to save on cost. The roll-out companies have capitulated with providing minimum specifications for servers. In the future the roll-out companies need to refrain from this practice and make the recommended server specifications 'the minimum. Disaster recovery is less a reflection of the software than it is a matter of good technology management on the past of the bank. Each institution has the responsibility to back up its data files regularly - preferably every day. If the system goes down and transactions are lost, the bank can restore the previous day's files and resume business. In the banks we examined, the data base was backed up to tape or hard disk at the end of every day. That day's backup remained in the bank and then was over-written by the next day's backup. While this practice is the minimum amount of disaster recovery by industry standards, there are some serious deficiencies in this mode of operation. First, the back up remains on site at the bank. If there is a fire in the building overnight, the bank risks permanently losing its data. The backup should be removed to an offsite location. Also, there is no historical record of transactions. The bank cannot go back and research transactions that were made more than one day prior. Each institution would be better off backing up their data to CD-ROM and storing several days, weeks, or even months of transactions for historical puyoses. This back up and recovery issue is not a software problem, it is an operations problem. The blame for this is not with the banks either, but with the roll-out companies. ASEC and ISSEC should do a better job of advising their clients on the proper procedures for backing up their data per industry standards. The only really serious issue with RB2000 is that of system security. While the features support user level data base, application, and transaction access, the system itself lacks security enforcement. The minimum number of user ID and password characters required to access the system is one. Industry standard is six alpha-numeric characters for IDS and passwords. The system also does not require users to have their own passwords - but supports multiple users accessing the system through the same logon. It was common to see all of a branch's staff sharing the same ID and password for RB2000. Additionally, RB2000 does not enforce changing passwords periodically, While this is less of an issue than sharing passwords, the practice of enforcing periodic changes is becoming industry standard. ASEC and ISSEC should look more closely at this issue and develop more rigorous security - procedures for RB2000. They must also do a better job of educating the banks on the importance of a strong security policy with enforcement. After reviewing the system requirements and programming specifications, it is clear that the developers of RB2000 used an adequate system development methodology for building RB2000. ~hddocumentation did contain standard industry deliverables such as data flow diagrams, context diagrams, some rudimentary entity relationship diagrams for data base design, and data definition tables. Although no formal named methodology was used, these practices appear to have been sufficient to produce a well documented software package. The procedures and methodology for system support also appear to be well-developed. Both ASEC and ISSEC use problem reports or user incident reports to track problems with the software in each installation. These reports contain a description of the nature of the problem, when and where is occurred, and what the resolution wa.s. If resolution could not be achieved locally, the companies used an escalation process to send the problem to the developers in home office. This appears to be an effective and sufficient system to support RB2000 installations. RB2000 supports all of the reporting needs of the rural banks to efficiently and effectively manage their operations. The software comes standard with a long list of pre-defined, pre￾programmed reports that can be accessed and invoked by users at any time of the day and with a minimum of effort. The most common praise received for RB2000 was the large number of reports and the ease with which they could be generated by bank staff. In addition to the standard reports, RB2000 also contains the eleven required MABS performance monitoring reports required by the project, and all of the BSP regulatory reports for the Central Bank.