Completion Evaluation

Project area

The project was designed to cover 15 out of 19 outer atolls in the Northern and South-Central regions. The project area comprised: Haa Alifu, Haa Dhaalu, Shaviyani, Noonu, Raa, Baa and Lhaviyani, in the Northern atolls; and Vaavu, Meemu, Faafu, Dhaalu, Thaa, Laamu, Gaafu Alifu and Gaafu Dhaalu in the South-Central atolls.

The project atolls contain 16 800 households with 97 700 inhabitants. This represents not less than 73 % of the total population in the outer atolls.

Project objectives and design

Target group

The project aimed to assist 3 250 atoll households directly from credit access and to benefit 6 000 indirectly through improved earning opportunities. The target group comprised households with a per capita income below Rf 2 000 per annum, equivalent to Rf 1 000 per month for a family of six persons. This poverty line was based on the average household income in the outer atolls derived from the various surveys. In 1993, the income ceiling for the target group was raised to Rf 2 000 per month reflecting the trends in economic and social change.

Objectives and components

The project was designed as the first phase of a long-term program to build up the Bank of Maldives (BML) as a development finance institution and help it to widen its geographic coverage to provide financial service to the population in outer atolls. The project was costed at USD 6 million with IFAD contributing 3 million and GOM 2.5 million and UNDP 0.5 million.

The project set out to: (i) lay the foundations of development banking in the Maldives and develop a credit delivery capability in the outer atolls; (ii) reduce income disparities between the outer atolls and Male by increasing the employment opportunities and income levels, with a particular focus on the lower income groups; and (iii) develop a framework for supporting future programs to improve nutritional standards in the outer atolls.

The project comprised five components:

• Creation of a Development Banking Cell (DBC) in BML to establish credit delivery mechanism and mobilise savings in the outer atolls;
• Credit to individuals and groups to finance investment in fishing boats, fish processing, agricultural activities, trades and cottage industries;
• A pilot nutrition education program in an atoll to initiate a process of nutrition demonstration an education integrated with agriculture and extension;
• Training of the project beneficiaries, assistance to NGOs for training in agriculture, handicrafts and other income generating activities; and
• Provision of consultants for setting up a monitoring and evaluation system and to assist in design and analysis of baseline and impact evaluation surveys.

Ministry of Finance and Treasury (MOFT) was the co-ordinating agency. BML through DBC was the implementing agency. Ministry of Atolls Administration, the Ministry of Fisheries and Agriculture and the Department of Women's Affairs were designated as co-operating agencies to implement the training and nutrition components.

The Project was designed to establish the Project Co-ordinating Committee with the Minister of State for Finance as the Chair and officials from the Ministry of Planning and the Environment, Atolls Administration, Fisheries and Agriculture as members. The Project Co-ordinating Committee was expected to co-ordinate co-operating agencies, set policy guidelines for the project, review the project progress and approve the annual lending program.

Expected effects and assumptions

The project design reflected a series of assumptions about constraints affecting production and trade in the outer atolls in the Maldives archipelago. Limited access to credit and extension services was considered as the main obstacle impeding income generation. Due to insecurity, irregular transportation and perceived high transaction costs, existing financial institutions were not ready to operate outside of Male. Furthermore, the outreach of public sector extension services in the outer atolls and islands was very limited. The project assumed that access to credit services would enhance support for islanders to initiate income-generating activities. Implicitly, the project design assumed that empowerment coupled with the provision of agricultural credit would contribute to women's income generating opportunities. Empowerment and training of low-income producers at least implicitly was considered a prerequisite for credit uptake for the most vulnerable or most resource poor households.

Evaluation

Methodology of evaluation

Evaluation purpose and scope

An IFAD mission conducted a completion evaluation (CE) of the Atolls Credit and Development Banking Project in the Maldives from 28.2.1999 to 29.3.1999. This CE set out to analyse the final performance of the project compared to the objectives set out at the time of project design. Further, its lessons should be used to improve the performance of the current Southern Atoll Development Project (383-MV) as well as future similar IFAD projects.

Approach and methodology

The mission used several sources of data to analyse the performance of credit and other components. The data used by CE were: BML financial data, economic indicators of the Government of Maldives, and the Poverty Assessment Data collected by UNDP Maldives. The mission conducted the Rapid Impact Assessment Survey (RIAS) with the co-operation of MOFT and Bank of Maldives (BML) to measure the social-economic impact and target focus of the project. The mission designed three questionnaires, for beneficiaries, non-beneficiaries (control group), and to capture island socio-economic profile. The survey was conducted with 114 households in 18 islands representing all the 15 atolls serviced under the project. Fifty-eight beneficiaries and fifty-six non-beneficiaries were interviewed.

Implementation context

The project loan agreement became effective in January 1990 with a project duration of four years. Project activities were initiated towards the end of 1990, ten months after the loan agreement signed. During the first year no disbursements were made. Disbursements exceeded appraisal targets except for the first year of project implementation. The project was closed in December 1996. By this time, 96% of the total loans had been disbursed.

Project achievements

Credit delivery

The total value of DBC loan disbursements exceeds the SAR estimates. At the end of the project period, a total of Rf 116 million was disbursed, while the estimate of credit disbursement in SAR was Rf 77 million. The target group loans accounted for 43% of the total loan amount disbursed and 80% of the number of loans disbursed.

By the end of the project, the total target group loans amounted Rf 51 million in 2 515 accounts. In other words, 2 515 households, representing 70%-80% of the total number of households in the project area would have obtained credit. These households would not have been able to access credit without the project intervention.

The economic activities funded by DBC comprise fishing, small enterprises, trade/transport (boats) and agriculture. Lending to fisheries related activities was considerably lower than expected. The share of agriculture, small enterprise, trade and transport increased significantly reflecting growing importance of tourism related activities.

Institutional development

BML in 1990 established a Development Banking Cell (DBC) to provide outer atoll banking services. This was the first step in a long-term process engineered by the project to provide sustainable financial services to the low-income population in the outer atolls. In 1996, DBC was re-named as North Central Atolls Regional Unit (NCARU). DBC was set up as an autonomous operation within BML with separate accounting and reporting formats. This separate accounting and reporting system was an important precondition for measuring progress and contributed to successful performance.

The project successfully implemented the institutional strengthening component to undertake development-banking activities. DBC/BML established four branches and started mobile banking services for islands in the outer atolls.

In the original project design, DBC/BML was to open three branches (Kulhudhufushi, Naifaru and Mulee) and one sub-branch (Hihadoo) during the first year of the project. DBC/BML wisely decided to proceed more cautiously. It opened branch offices only when it was assured that it would have trained staff in place. It opened four branches over four years in Kulhudufushi, Naifaru, Mulee, and Fonadoo. The DBC/BML head office in Male has also extended development-banking services to Alifu and Kaafu atolls. DBC/BML built all branches on own premises using IFAD funds.

To provide financial services to some 50 islands, each branch office started mobile banking operations. Yacht dhonis equipped with safes were provided to each branch. The mobile banking staff was expected to visit islands periodically, collect loan applications, sanction loans and receive deposits and repayments.

Presently, the branches use the mobile banking service only to deliver and recover loans. The exception is the Naifaru branch; it collects savings from two near-by islands through the yacht dhoni banking service. Almost all savings at branch office level, or about 80%, are collected within the same island where the branch office is located.

The yacht dhoni operation turned out to be expensive. Operating and maintaining the yacht dhonis is expensive. A new dhoni costs around RF 1.2 million; it may have to be replaced every five years. To curtail costs, DBC/BML reduced the frequency of visits by the dhonis. It also withdrew the mobile loan officers (MLOs), who were based in atolls; instead, the MLOs were located at each branch office.

Effects sssessment and sustainability

Income generation

The credit furnished under the project is associated with significantly higher income of beneficiaries. The average initial monthly income of the target group beneficiaries was Rf 1 670; with or after the loan, it rose to Rf 3 990; this represents an increase of 138%. The median income for the target group beneficiaries before and after taking the loan was
Rf 1 700 and Rf 3 000, respectively; this reflects an increase of 76%.

Institutional sustainability

DBC/BML has consistently moved towards attaining sustainability of operations. The credit recovery of DBC/BML is excellent. The recovery performance hovers around 90%. Non-performing assets constitute less than 1% of the total loan disbursement. All branches mobilise savings though at a limited rate and make profit.

The interest rate on loans was negative in real terms, or minus 1.5%, during the initial phase of the project. But in 1994, DBC/BML raised the lending interest rate to 12% for its development banking lending; profit margins improved further since the inflation rate declined from 10% in 1994 to 6% in 1996; this meant a positive interest rate on lending of 6%.

The DBC/BML employed two credit delivery systems. In the first, in line with common commercial lending practice, loan applications are screened and sanctioned based on the bankable collateral. Borrowers provide collateral security equivalent to 200% of the loan value. Hypothecation of assets acquired with the loan is to form a part of the security. Assets such as boats, boat engines, and houses were accepted as collateral.

In the second system, credit is not collateral based. In most countries, for non-collateral credit, banks adopt group lending or joint liability approaches to cover risks. Instead, DBC/BML policy was to sanction loans up to Rf 15 000 without collateral or a group guarantee; but it carefully screened loan applications and the DBC central office controlled the approval process.

The recovery rate of loans under both systems of these credit delivery systems has remained high (around 90%) and it is important to understand the reasons. Four factors explain the high recovery rate: (i) the "graduation principle" built into the lending policy and procedure: the loan size is determined by economic activity as well as trustworthiness of the client; (ii) the "incentive" mechanism: prospect of receiving a new possibly larger loan, and with a longer repayment period; (iii) the rigorous credit recovery procedures; and (iv) the "sanctions" in the form of restrictions on further borrowing and legal actions resulting in auction of assets.

Saving mobilisation

DBC/BML has encouraged savings mobilisation and the performance reflects trust and the building up of social capital. Yet, saving mobilisation has not progressed as projected. Saving mobilisations in the outer atolls is associated with high transaction costs explained by saving promotion efforts and transport to collect savings in outlying islands. Maldives Monetary Authority (MMA) decrees that the interest rate on savings be 7 percentage points lower than the interest rate on loans (or 5%). At this rate, the interest rate on savings is negative. The public has little or no incentive to save apart from " the security aspect". Moreover, islanders already find cash availability for withdrawal to be restricted, since DBC/BML yacht dhonis visit islands occasionally and unpredictably. In short, DBC/BML find it easier to mobilise savings from commercial branches in Male rather than from outer atolls.

Factors explaining performance

Four factors stand out in explaining the performance of DBC/BML. First, it has rigorously screened the loan applications. Second, it has carefully monitored the portfolio and loan recovery. Third, the DBC/BML has inculcated a propensity for the islanders to save in financial instruments. Fourth, and finally GOM should be praised since it has provided BML the necessary autonomy, and independence of operation, for the latter to conduct its day-to-day operations without interference. In combination, this institutional setting has raised confidence and created trust among its customers.

This record contrasts sharply with the typical setting for credit projects attached to financial institutions that are controlled by governments. Internationally, development banking operations for government controlled banks have a record of poor profitability. Typically, such banking operations lend at subsidised rate of interest and incur high delivery and recovery costs. Loan recovery is mostly neglected. However, in this project, the DBC/BML operations have generated solid profits. DBC/BML has maintained its progression in outreach and the current level of operations. There is no upward trend in the cost of credit delivery. With current interest rates for loans and deposits, and cost of funds, the DBC/BML operation is institutionally sustainable.

Main issues and recommendations

One fundamental flaw in implementation – reflecting inadequate efforts also during design - affected negatively impact in poverty alleviation. GOM and the project authorities gave very limited attention to measures to empower, mobilise and train the most vulnerable households so as to permit also them to avail themselves of credit. The most vulnerable population was expected to become "aware" and receive sufficient training so as to link them to the emerging financial services. Although, the "lower" or "middle" poor households benefited from the project, the majority of the "bottom poor" or "most vulnerable" households were not assisted by financial services. Their loan demand remained limited for income generating activities.

The training component was only partially implemented for three reasons, first GOM's limited capacity to co-ordinate co-operating agencies; second, the target group definition was not precise. In turn, the absence of analysed socio-economic data contributed to the insufficient understanding of the extent of social and economic deprivation. Finally, the BML was not provided with financial incentives with which to compensate the higher transaction costs associated with reaching out also the poorer or most vulnerable households.

The modalities of co-financing are vital for success of the project but these modalities were not sufficiently developed at the time of design. The project was appraised and approved with merely an "in principle agreement" with UNDP. The project design allocated important components vital for assisting the "poor/vulnerable" households with limited entitlements, such as beneficiary training and pilot nutrition programme. But UNDP in consultation with GOM deleted the pilot nutrition component and prepared a separate project for technical assistance, training and beneficiary training. UNDP utilised the funds allocated for technical assistance and training. Beneficiary training was not conducted.

Lessons learned

Sustainability of Development Banking

Non sustainable elements in design

The absence of a proper policy context has caused immense damage to the sustainability of development-banking institutions and financial services across developing countries. Across development banking projects, two negative features stand out that have precluded financial sustainability and consequent institution building. First, development loans commonly have continued to be provided at low or negative rates of interest in real terms. The implicit or prevailing assumption was that a lower interest rate would increase the demand for credit and uptake by the poorer sections of the society. But this meant that financial margins have remained negative or insufficient to cover interest rates. Second, as a corollary, the already low or negative interest rates set for lending have provided little or no scope for positive interest rates with which to provide incentives for savings mobilisation. Conversely, the support for savings mobilisation is non-existent or weak. The banking institutions have not mobilised savings. Nor have they contributed to building up social capital through inculcating incentives in support of a behaviour among their clients towards regular savings and a longer term continuous transactions with the banking institution built on trust.

This adverse policy context, by and large, set the stage at the outset also for the current project. This project design following the custom at the time endorsed lower interest rates for target group lending in general and for agricultural loans in particular. At the time of project start-up, the lending rate in real terms (net of inflation) was even negative at a rate
of –1.5%, since at the outset the interest rate was 10.5%, whilst the inflation rate was 12%.

On the other hand, the design represented an improvement compared to other projects in that the DBC/BML would operate a weekly mobile banking service to outlying islands to deliver and recover credit as well as to mobilise savings. But the transaction costs of this mobile banking service were high.

Engineered shift towards sustainable financial services

The most important lesson learned from this project is the need of ensuring financial sustainability of banking operations at the time of project design. Most projects fail in this regard. Those in charge of design typically overlook sustainability aspects and the need for the development bank to generate positive financial profits. They accept the populist arguments for continued credit delivery credit at artificially low interest rates that do not cover total costs (fund cost, transaction cost and cost of bad debt).

The performance over time of this project amply demonstrates that the benefits of charging positive interest rates far outweigh the hypothetical negative effects such as public resentment and low growth in credit uptake. For smaller producers, access is more important than a subsidised cost of credit. It is necessary at the outset to build in critical features that will ensure sustainability, foremost charging rates of interest to cover total financial costs of development banking operations. DBC/BML took three measures in the right direction. First, it rigorously monitored the profitability of individual branch offices. Second, interest rates were raised. Third, it reduced the supervision and transaction costs for the mobile banking service.

Rigorous branch monitoring

DBC/BML established an unusually rigorous branch monitoring system. Its branches have continued to report on a monthly basis disbursements, recovery performance and profitability. To wit, this reporting system demonstrated the costs of providing loans at negative real rate of interest and of providing weekly mobile banking services.

Changes in interest rate policy

DBC/BML initial interest rate of loans was 2% less than the commercial rate of interest charged on loans granted by the commercial branches. The DBC/BML, in 1996 raised its interest rates, charging a uniform rate of interest across all types of loans. This was the first major step in introducing sustainability dimension into development banking.

Rationalising outreach and mobile banking service

DBC/BML improved the cost-effectiveness of its outreach services. First, it reduced the frequency of visits of the mobile dhonis from weekly to monthly. Second, it withdrew atoll based loan officers and posted them in the branch offices. The Loan Officers start from the branch office and return to the branch after completing their field trip. Costs of operation were cut and profitability improved.

Strategy for outreach

Analysis of lending and savings mobilisation

DBC/BML has followed a cautious policy of expansion and plans with which to establish volume of lending with which to cover costs of operations. To begin with, branches have been established first to cover the regional level, and second to cover atolls and islands with a minimum level of population and economic activity. At present, the four DBC/BML branches cover 13 atolls. Each branch covers three to four atolls and about 50 islands. The level of activity in the islands with branches and with close proximity to the branch office is high.

The data that have been analysed by the CE show that one third of the islands covered by the branches contribute not less than 80% of the target group lending and 70% of that to the non-target group. Savings are exclusively generated within the atoll in which the branch is situated. These findings have important implications for formulating the future strategy for expansion of branches and mobile banking.

Strategy for branch office expansion

First, DBC/BML has established branches at the regional level and these operations should now be consolidated. The next step is to open branches at the atoll level. But DBC/BML cannot afford to open branches in each and every island. Instead, a strategy is required with which to establish criteria for decisions to open low cost branches in the islands that have already possess a substantial client base. To this end, the break-even size of loan volume needs to be derived with which to generate a minimum level of revenue. The CE with actual data for the portfolio of the Kulhudufushi branch pursued this analysis. This analysis shows that a loan portfolio of Rf 4.2 million generates the minimum revenue for a branch to break, or about Rf 500 000.

The Kulhudufushi branch at present serves three atolls. The portfolio in each of these three atolls exceeds Rf 5.0 million, which is higher than the break-even portfolio volume of Rf 4.2 million. Opening branches in each of these atolls is viable. Similarly, it is possible to locate those atolls that are already served by other branches, and which have attained a minimum portfolio equal to or exceeding Rf 4.2 million.

Strategy for Mobile Banking with Dhonis (faster boats without branch expansion)

The benefits of mobile banking are more limited than first understood. The main thrust in outreach is based on the opening of new branches and developing manpower capabilities with which to expand the branch network. The mobile banking service is an effective complementary tool in providing financial services to islands located in the geographical proximity of the branch office. It is not effective in delivering financial services to islands located far away. Mobile banking with dhonis generates savings only within the atoll in which the branch is situated. It will not assist the branch office to mobilise savings in the more distant atolls, which is essential in the long term when the soft loan sources dry up. Finally, the costs of mobile banking need to be contained. The DBC/BML is considering buying faster boats. But faster boats are expensive to procure and also to operate.

Testing instruments for outreach

Maldives has 200 inhabited islands and most of these islands only comprise about 100 households. The DBC/BML cannot open branches in all the islands to provide financial service and mobile banking has its own limitations. Other complementary instruments need to be tested in the delivery of financial services to islands with a limited number of households. This project is now closed but any future support to the DBC/BML or similar programmes should contain incentives in the form of grants to experiment with new approaches.

Credit union or savings and credit society

A case can be made for establishing credit unions or co-operative societies in the more distant islands that are located far from the branches of DBC/BML. Most islands visited by the mission have women's committees, which undertake income-generating activities. These committees could form a nucleus for a financial service organisation. Any new organisation to be developed needs to come under the umbrella of DBC/BML for the latter to provide necessary backstopping including audit of account. DBC/BML will have to pilot test and then adjust this approach to suit Maldivian conditions. Three pre-requisites for implementing this approach emerge. They are: (i) preparing a legal framework for establishing a credit union/co-operative society; (ii) defining linkages with DBC/BML; and (iii) capacity building at DBC/BML to pilot test this approach.

Self-help groups / rotation savings and credit groups

In islands with a limited number of households it is appropriate to test the concept of self-help groups and rotation savings and credit schemes. This would help the islanders to access credit from their own savings that they mobilise. Such mechanisms are important in assisting food insecure households to develop coping mechanisms to overcome periods of temporary distress. This approach does not require a legal framework to begin with. But capacity building is required at the level of DBC/BML to operationalise the importance of this instrument to assist the "bottom poor/vulnerable". The methodology need to be defined for creating and supporting the self-help groups / rotation savings and credit groups. The next step is to train the staff.

Regulatory environment

The Government holds a majority equity position in BML, yet it has in no way misused this position of financial control. Contrary to the situation in many other countries, the Government has not used its position to intervene in decisions on lending and in day-to-day banking operations. BML was permitted the necessary autonomy: it enforced sanctions by auctioning the assets mortgaged, suspended disbursement to under-performing sectors and inculcated a habit of repaying the loans. The DBC/BML provided successively larger loans to clients who repaid promptly. The legal environment was supportive of the BML's actions to repossess and auction assets of the defaulters. The strict monitoring of branch office operations contributed to this excellent record.

MOFT and MAA remain the GOM's regulators to oversee banking activities of BML. The regulators allowed BML full autonomy in its operations. GOM did not seek to enforce any populist or ill-conceived agenda to seek to lower DBC/BML lending interest rates or to ask it to provide loans to favoured clients. As a result, despite being a government owned bank, the general public considers it as a commercial bank. It is reported that BML even refused granting loans to island electrification under government guarantee once a few island committees defaulted. GOM did not put any pressure on BML to sanction these loans. This autonomy is currently an unwritten understanding between MOFT and BML.

The need for financial and development banking institutions to have full autonomy in providing financially profitable and sustainable services is a must. This is one of the necessary preconditions for success. Such best practices related to autonomy of the financial institution will have to be codified and incorporated as conditions for funding future financial services projects. This is lesson learned from the Maldives Atolls Banking Project.

Improving targeting modalities in financial services projects

Identifying target group and the "bottom poor/vulnerable"

An overall broad target group definition is a not an effective tool to reach the bottom-poor in a financial services project context. Future projects need to sharpen the poverty focus in identifying target groups. Targeting criteria need to be simple and location specific. Generally, four sequential steps will have to be followed to sharpen poverty focus in targeting. They are: (i) a community-based wealth ranking to identify the resource poor households; (ii) a diagnostic survey mapping the resource endowments of these households, and the causes to their poverty; and (iii) selecting a set of indicators associated with/or explaining poverty (household size, number of working members, material used to construct walls, receipt of "zakaath" etc.). Four income strata should be distinguished: bottom-poor, lower, middle and upper income groups.

In a financial service project context, the upper income group will have to be excluded from the purview of IFAD assistance. On the other hand, the financial institution need to serve also this clientele from its other resources, or the savings mobilised. The income strata consisting of the "bottom poor/vulnerable" and the lower and middle income groups form the overall target group in a financial service project. The project needs to have separate target group definitions for three socio-economic strata that fall under the overall target group of IFAD. This is essential so as to design precise interventions oriented to reach and assist the "bottom poor/vulnerable".

Complementarity of instruments to reach the "disadvantaged or vulnerable" households

The project at the outset used a narrow target group definition of Rf 1 000 per year and set out that lending to this group should represent 70%. In 1993, this proportion was reduced to 45% of total DBC lending. The earlier too narrow target group definition excessively restricted lending to the non-target group, overall lending and the profits with which to break even and permit sustainable outreach services to be set up. This narrow definition was coupled with the absence of effective measures and funding with which to empower and provide training to the target group and of incentives for staff to accelerate lending to this group.

DBC/BML revised the target group definition. The income ceiling was raised by 518; after this adjustment, the target group represented about one fourth of all households. But the other complementary measures, i.e. social mobilisation and skill training were not instituted. These interventions were to have been provided through the abortive UNDP funding. This was a missed opportunity: the project because of this omission did not reach the most vulnerable households.

Set of interventions to assist the "bottom poor/vulnerable" households in financial services projects

The "bottom-poor/vulnerable" households face many constraints and provision of credit is not sufficient to improve their standards of living. They have few economic activities and their demand for credit is limited. A low standard of education and few skills worsen their situation. This is evident even in this project, which is one of the most successful credit projects in the region. Targeting the "bottom poor/vulnerable" households requires a fuller complement of empowerment and support services. GOM support is necessary for identifying profitable activities and for conducting skill development training. Once such a platform is created, the demand of this group for financial service will increase. This in turn requires capacity building at BML and other development partners and also provision of assistance in the form of budgetary support or grants to shoulder the development responsibilities.

Incentives coupled with targets

Defining the target group alone with not be of much use to ensure flow of assistance to the "bottom poor/vulnerable". Financial institutions in general tend to give less attention to the "bottom poor/vulnerable" and the low-income groups. This is on account of the perceived high risk and cost of delivering financial service to these groups.

Most projects do not design incentives to the participating financial institution to reach the "bottom poor/vulnerable". But development finance theory certainly would permit practices to be introduced where incentives can be designed to cover at least initial costs and induce the financial institutions to reach the "bottom poor/vulnerable" and low-income groups. GOM should provide a lower on-lending rate to the DBC/BML on loans to the "bottom poor/vulnerable". Moreover, DBC/BML costs should be covered for appointing a person to exclusively deal with the "bottom poor/vulnerable" and to co-ordinate with other development agencies. Temporary grant funding for this purpose should be considered. Finally, precise targets need to be defined for lending to this socio-economic strata. Targets should be reviewed based on the experience gained during implementation.

Co-financing

Modalities of co-financing that are fully owned by stakeholders so as to become operational are vital for successful project implementation. In this project, co-financing modalities were not finalised prior to the project appraisal. The project was appraised and approved with a mere "in principle agreement" with UNDP. The project design allocated important components vital to "bottom poor/vulnerable" such as beneficiary training and pilot nutrition programme. But UNDP in consultation with the GOM deleted the pilot nutrition component and prepared a separate project for technical assistance, training and beneficiary training. UNDP utilised the funds allocated for technical assistance and training. Beneficiary training was not conducted. The activities funded by IFAD and UNDP were undertaken as stand-alone activities; an integrated service delivery was not achieved. Integrating the activities funded by the co-financier into the core components of the project requires joint preparation, joint appraisal, explicit agreements formulated at the time of loan negotiations and close consultation during project implementation.