Abstract
Nkwe Enterprise Finance, a financial service to fund microenterprises in the North West province, suspended its operations after just over two years of lending. This happened against the background of three significant events. First, a new apex for pro-poor microfinance was designed, which resulted in the need to reconsider the nature of Khula's ongoing role as wholesaler to microfinance institutions. Second, the Financial Services Charter was signed in October 2003, in which banks committed to targets for so-called empowerment lending, including black small and medium enterprises. Third, a large number of microfinance initiatives continued to fail to reach sustainability and scale. This article investigates the reasons for the suspension of Nkwe's activities and draws conclusions on the lessons that should be learnt for the future development of microenterprise funding in South Africa.
1. Background
Nkwe Enterprise Finance (henceforth, Nkwe) was initially conceived by its originator, the Millennium Development Fund (MDF), as a financial service organisation to fund microenterprises in the North West province (NWP). The MDF called for tenders and awarded a management contract to Vulindlela Development Finance Consultants. Vulindlela's concept of a commercially oriented model of individual, microenterprise lending was appealing to the National Business Initiative (NBI) that was managing the MDF on behalf of the North West government. There have been very few, if any, South African success stories with these three characteristics. After just over two years of lending, Nkwe suspended operations in October 2003.
The suspension of Nkwe's activities happened against the background of several recent events, in particular:
| 1. | The design of a new apex for pro-poor microfinance, which resulted in the need to reconsider the nature of the ongoing role of Khula as wholesaler to the microfinance institution (MFI) sector. | ||||
| 2. | The signing of the Financial Services Charter in October 2003, in which banks in particular, committed to targets for so-called empowerment lending, including black small and medium enterprises (SMEs). Even though not all banks may undertake microenterprise finance directly, the impact of recent experiences in this sector will influence the types of initiatives supported. | ||||
| 3. | The continued failure of a large number of microfinance initiatives in South Africa to reach sustainability and scale. Several new ‘microfinance’ initiatives by commercial banks also ran into difficulties or were closed during the period of Nkwe's brief operations. | ||||
These events make it even more important that the reasons for the suspension of Nkwe's activities and the lessons that can be learnt from it are fully understood. For these reasons, and at the original request of the Nkwe board, FinMark Trust agreed to fund this study. Its aim in doing so is to ensure that ongoing learning takes place. While the purpose is not to build a counter-factual scenario specific to Nkwe, as this would be academic, the main question here is: which factors, choices or decisions made by key players would result in a different outcome in future cases involving the start-up and development of microenterprise-focused MFIs in South Africa?
The present document is based on an informal enquiry, carried out over one week, which seeks to provide an independent summary of reasons for the suspension of Nkwe's activities. The team carried out a series of structured interviews with key institutions involved, and reviewed a range of relevant documentation and correspondence obtained primarily from Khula, Vulindlela and FinMark Trust. Several other knowledgeable individuals were interviewed to obtain reference points from outside the circle of individuals and institutions that were directly involved.
During this process, the team reviewed the history of Nkwe to establish the main reasons for the suspension of activities, in part verifying and commenting on the allegations made in The Nkwe Story, written by the board and the original promoters. The team was most impressed by the complete access to information provided by the key parties, and their willingness to participate in the spirit of learning from the Nkwe process. At every turn, the team encountered openness and eagerness to address the difficult issues that were raised through this effort.
Several key parties were interviewed and used as a reference group to comment formally on the draft report prior to finalisation. (See Annexure A for contact details and positions held in their respective organisations.) Nevertheless, given the nature of the assignment and the content of this article, the conclusions remain the exclusive responsibility of the authors and should not be imputed onto any of the members of the reference group.
2. Nkwe operational performance
The original business plan projections showed an individual lending operation reaching operational sustainability, 7 500 active clients and a R13 million (US$1,95 million) outstanding portfolio midway through Year 5. Vulindlela's initial market analysis of planned branch locations in North West province suggested sufficient demand to proceed. Lending started in September 2001 in one branch office and, by April 2002, had expanded to a further branch and satellite office. Initial performance exceeded targets, but within seven months Nkwe started to miss key targets related to portfolio growth and quality ().
Table 1. Lending performance against targets
In a report to the Nkwe board in June 2002, the general manager set out the following reasons for missing targets: lower than expected market demand; loan officer turnover; lower than expected client renewals (repeat loans); and delays in the implementation of the management information system (MIS). Subsequent market analysis by Accion indicated that the demand for individual loans was weaker than had initially been thought. Following initial repayment problems, Nkwe also further restricted its target market by excluding hawkers, and by changing to a credit bureau that included client information from retail and consumer lenders. The latter measure resulted in bad credit records being highlighted for a significant proportion of clients.
Nkwe's performance never really recovered since first missing key targets in the second quarter of 2002. By as early as June 2002 – less than a year after lending began – write-offs were running as high as 10 per cent (Nkwe operated a strict 90-day write-off policy). Vulindlela proposed expansion to parts of Gauteng province (including Soweto) that represented more active and densely populated client markets, and market research took place in early 2003. However, this expansion was never put into effect, as Nkwe's primary donor, Khula, indicated its reluctance to fund it.
The selective timeline presented in outlines key events and actions relating to Nkwe's operations and funding, and helps illustrate indicative relationships between operations on the one hand and funding on the other. Six banks and a similar number of donors were approached for funding. Although four donors (Hivos, Khula, NBI and MDF on behalf of the North West province) provided a total of R12,45 million (US$1,86 million) in funding, with about two-thirds as subsidised loans and one third as grants, no bank came forward with an investment or loan. This was despite one bank (ABSA) going as far as conducting due diligence and receiving investment committee approval for a R7,2 million investment and a R28 million loan subject to conducting a satisfactory due diligence on the Nkwe operation.
Table 2. Selective timeline of key funding events and operational performance
Nkwe had ceased operating by the end of 2003 – just over two years since it had initiated lending operations. In that time, over R9 million was spent on technical assistance, management fees, systems development and operational subsidies. This compares to a total of R2,9 million (US$430,000) lent to 2 185 clients (as of June 2003). The ratio of funding received to loans disbursed is therefore over 4 : 1, or R5 400 (US$800) per loan.
3. Analysis of original concept and execution
The project was never ultimately put together as it was originally conceived, i.e. as a ‘sustainable microbank’ funded by private investment capital with investors as promoters with rights to convert to shareholding once profitability was achieved, and who would take out dividends. The original company was a hybrid non-governmental organisation (NGO) company structure to facilitate receipt both of donor grants and of investment capital, with the intent ultimately to convert into a private financial intermediary. Nkwe would constitute a caretaker board of directors consisting of promoters until investors could be put on. The management contract was to be given to Vulindlela to set up this entity and participate in its capital structure. The focus was on the NWP, and later, on national expansion to be rolled out after some years.
Funding was raised from NBI/MDF, Hivos and, provisionally, from USAID on the basis of this new approach to microfinance in South Africa. This approach and the potential involvement of bank partners as the private investors in this new institution was a primary interest of Accion, which had secured a relationship with Vulindlela when it took over Calmeadow Foundation's portfolio in Africa. In fact, Accion probably envisioned a greater potential role for ABSA Bank in the project than Vulindlela, which envisioned Nkwe as an independent MFI that would receive funding support and investment from the bank – rather than the closer ‘agency’ type of relationship Accion has moved towards in recent years.
In a number of countries, banks have participated in microfinance, through linkage mechanisms, as investors and directly as microlenders. The inability of Nkwe's promoters to obtain a partnership with a bank should be considered a prime cause of the ultimate failure of the organisation:
| 1. | USAID's grant was not triggered, and thus the funds were not forthcoming, | ||||
| 2. | Hivos's funding was predicated on this model, and was ultimately suspended in substantial measure because of the inability of the promoters to carry out the project as conceptualised. | ||||
| 3. | Accion pulled out of its relationship with Vulindlela after it became clear that ABSA was not going to participate. | ||||
Nkwe and the individual lending methodology it used could perhaps still have prospered, given a more viable market (the initial market analysis of NWP turned out to have been overoptimistic, but expansion to Gauteng was subsequently proposed by Vulindlela); a more constructive relationship with principal funders; and significant additional grant and/or loan funding. Nkwe's difficult relationship with Khula is central to these other factors, as will be outlined in this article. One could still argue, however, that if Nkwe had successfully achieved its original vision, in conjunction with a private bank investor, Khula's funding would have been marginal in its importance, and therefore these issues would not have been critical ones. As it turned out, in addition to significant financing from NBI and Hivos, Nkwe also needed Khula's funding to survive. Khula therefore became the principal funder, a situation that was never envisaged.
From a document review, it is clear that Khula never became comfortable with the hybrid core proposition Nkwe represented. There was always a fundamental mismatch. While it can be argued that Khula should, or ought to be able to support a proposition such as that of Nkwe, Khula had at the time never undertaken to support an initiative that included an operational subsidy to a ‘for profit’ entity engaged in socially important programming. Khula also seemed to have faced some constraints imposed by the European Union (one of its main funders) to providing support to a microfinance organisation organised as a section 32 company.
While it is not clear that any bank in South Africa would have been ready at the time to form such a partnership as was being proposed, it is clear that Nkwe's promoters did not proceed along standard practice norms for seeking bank involvement. Typically, the involvement of any bank in microfinance has to be negotiated and supported from the very top of the organisation. This is the manner in which bank downscaling, joint ventures and other initiatives have been carried out worldwide. Without strong top-level support, microfinance is too vulnerable to the vicissitudes of decision making by very senior executives who do not understand this activity or how it fits within the bank's core commercial strategy. This is particularly true in the South African context where a number of banks have tried and failed to successfully execute microfinance programmes. They have consistently undervalued the international experience and misjudged the market.
ABSA shares some important responsibility in the Nkwe story. Initially, ABSA claimed exclusivity on the Nkwe board of directors and retained for itself the right to be the sole bank partner for the project. Subsequently, related perhaps to substantial failures in its retail banking business, and in spite of a reportedly favourable result from the due diligence exercise, it abruptly pulled out, leaving the project without a bank partner and creating a real funding gap. Clearly, its staff either did not in reality have the mandate to proceed with Nkwe that they appeared to have, or senior management did an abrupt about face as a result of problems in other parts of the bank's system.
At times, there was a great deal of confusion as to the exact nature of the expected relationship between ABSA and Nkwe. For some time, Nkwe operated without adequate staff in the critical management positions of operations manager and finance manager. When it had difficulty filling these positions with locally based staff, and as a condition for securing its participation in funding Nkwe, ABSA planned to fill those vacancies. Recruitment was not timely, which left the organisation without critical capacity for too long a period.
Management staff at ABSA and Nkwe met a number of times during the process to address matters such as the launch of a joint savings product, the integration of MIS, and new payment mechanisms. The projected relationship seems to have swung from one extreme of being a relatively passive funder to the other of becoming a potential joint venture and strategic partner and, as such, the dominant force behind Nkwe. This continual search for an acceptable and workable model, in the context of actually launching operations, certainly hindered Nkwe's ability to build a coherent and cohesive plan and management team.
Vulindlela was not well placed in South Africa to obtain access to very senior levels of bankers to demonstrate the profitability and desirability of taking on microfinance. It does not have a track record in managing commercially viable microfinance operations (except through its association with Calmeadow), nor do its principals have the necessary contacts at high levels in the banking community. These two characteristics would have been important in an environment where banks have traditionally not participated in socially motivated lending. They would be doubly important later on in the process, after a couple of banks lost substantial sums in ‘microfinance’ and became skittish about the viability of consumer loan-driven retail banking and microfinance in general.
This access could have been provided by successive Banking Council representatives who sat on the board of Nkwe from the beginning. While the Banking Council seems to have played a limited role during the first two years of the process, in contacting Vulindlela with representatives from several banks, these efforts did not pay off in establishing any agreements. In the case of ABSA, it neither identified the fact that Vulindlela was dealing with bank staff who were too junior, nor did it intervene to make direct contact in support of the programme at the more appropriate levels of chief executive officer or managing director.
When the Banking Council representative (the third to be named) did ultimately intervene at very high levels, it was largely for the purpose of damage control and to bring ABSA back to the negotiating table. While this role was vital and appropriate, it might well have been too late in the process to save the relationship, as it came after the first rejection by ABSA. Although the representative was successful in getting ABSA re-engaged to the point where it developed a Memorandum of Understanding and vetted Nkwe's proposal at the level of its Investment Committee, by then Nkwe was facing serious operational difficulties and loss of donor support.
Early on, Accion International could also have played this role. Alternatively, at the very least, it could have organised the right individuals to contact banks and promote the model. However, Accion focused more on providing high-quality technical assistance in methodology, procedures and staff training. It has extensive experience working with commercial banks in Latin America, and certainly knows the importance of obtaining buy-in at the most senior levels of the organisation if microlending is to be adequately supported.
Yet, Accion did not pursue bank relationships beyond that of ABSA. Here, Accion seemed to follow Vulindlela's lead and operate primarily at the middle management level, on the understanding that its ABSA contacts had a mandate to move forward. When ABSA first rejected Nkwe's proposal, Accion did not subsequently accompany Vulindlela in pursuing other options. By mid-2002, Accion had probably concluded that Nkwe was in serious trouble and did not merit additional support. By the end of 2002, Accion had withdrawn from any involvement with Nkwe.
Finally, the calculated decision to develop such an innovative and ‘commercially driven’ concept in such a small target market, made to take advantage of the initial tender offer from NBI/MDF, set the initiative up for a weak start. First, the initiative was less exciting to potential bank partners due to the small size of the project and the time it would take to grow. Any bank that would engage seriously in microfinance would have wanted a far more aggressive growth strategy built in from the beginning. Second, when the market turned out to be weaker than expected, the project was set up for underachievement, and a generalised feeling developed that it was failing.
Vulindlela's management felt that it had enough of the pieces put together to ensure success – it had signed a contract with NBI, it had support in principle from Khula and Hivos, and a letter of commitment from ABSA, and had performed all the basic market research and business plans. It is hard to second-guess their decision to move forward and execute the project. In execution, however, Vulindlela made a series of critical misjudgements that doomed the project, almost from the start. The most critical of these, which caused the majority of ill feelings towards the project on the part of its partners, were the confused roles Vulindlela played as promoter, contractor to NBI, technical assistance provider, holder of a management contract with the Nkwe board, and equity investor. Vulindlela (and to an extent also Nkwe's board) never understood nor accepted the concerns expressed by Khula, Hivos and others about appropriate checks and balances, rights and responsibilities, and accountability.
4. The role of Khula (government and donor-funded apex)
From the documentation it is clear that Khula was never comfortable with the fundamental proposition that was Nkwe. Khula did not like the structure of a management contract between Nkwe and Vulindlela; it did not like what it regarded as a high level of fees for the promoter/technical assistance provider (Vulindlela), or the weakness of the governance structure and the pre-eminent role played by the project's promoters on the board. It never expressed concern about the lack of bank partners and it seems clear that its primary motive for supporting Nkwe was to establish more microenterprise lending in the NWP. Khula was used to supporting stand-alone retail financial institutions (RFIs) with strong independent boards with funds to support operating costs and loan funds. While it has had a technical assistance fund, it does not seem to support internationally based technical assistance provision to RFIs and, in the view of some in the industry, underutilises this facility.
Khula had supported a number of RFIs, most of whom had not been successful either in obtaining large volumes of lending, or in reaching financial viability. It had recently written off substantial sums for bad debt, and in response, seemed to have developed a fundamental distrust of RFIs and their management. This distrust is evidenced in the manner in which Khula structured its relations with RFIs. For example, it built its support around the mechanism of monthly disbursements of funds to cover operating losses. This gave Khula a constant means through which to exercise pressure on management not only to comply with the business plan, but also to attend to any other issues that Khula felt needed to be addressed.
Delayed Khula funding tranches, followed by the suspension of tranches in the third quarter of 2002 due to Nkwe missing portfolio targets, pushed the hiring of loan officers behind schedule. This, in turn, made it more difficult to keep up with production targets. While this does not normally represent an insurmountable challenge in an MFI, it can set off a vicious circle of always having to try to catch up, with the attendant mistakes that inevitably come from operating under extreme pressure. To Nkwe's management, Khula did not seem to recognise its own role in hindering Nkwe's achievement of agreed-upon performance targets. In addition, Khula engaged in a level of ‘second-guessing’ management decisions all along, even when the activities being undertaken by management fell within the business plan that Khula had approved and included as part of their funding agreement.
This unhelpful tendency was reinforced by the way Khula apparently participated through its position on Nkwe's board of directors. When issues of importance to Khula were being discussed at board level and decisions were being taken, Khula remained relatively silent. Subsequently, it would receive the communication from Nkwe about the decisions taken by the board and then would respond negatively through formal communications. This style provoked frustration on the part of Vulindlela and Nkwe's board and certainly degraded their view of Khula's effectiveness and ultimate intent with respect to supporting Nkwe. At certain points, this approach caused sufficient doubt in the minds of Nkwe's promoters and they questioned Khula administrators as to whether Khula was interested in continuing with its support. A more helpful approach would have been to exercise its governance role in shaping Nkwe policy and responses to important issues while participating actively on the board, and then exercise its ultimate power if it considered that Nkwe's board-level decisions failed to correct fundamental flaws in the performance of the institution.
In principle, Khula's performance-based approach is sound and in line with international best practice. Specific business plan targets were used as the basis of the funding relationship. An external consultant was contracted by two of Nkwe's funders, Hivos and Khula, to conduct joint monitoring and reporting on the basis of those targets – which is also in line with international good practice for multi-donor funding. In practice though, the relationship between Khula and Nkwe/Vulindlela did not work well. Khula held Nkwe disbursements ‘hostage’ to performance on business plan targets, which almost inevitably tend to be overoptimistic, as project promoters are usually intent on pleasing the donor agency's need to disburse a maximum amount and reach full sustainability quickly. Most projects underperform in relation to business plan targets.
The Consultative Group to Assist the Poor (CGAP) has adopted and recommends a minimum threshold approach where the donor agency sets minimum performance standards below which it intends to withdraw support, if there is not a compelling justification for missing the threshold. This offers the donor the possibility of allowing management room to make mistakes or adjust to unexpected circumstances, and yet sends a clear message about management's accountability for performance. Had Khula taken this approach, far more management time could have been spent solving the operational challenges facing Nkwe and less time in constant negotiations with Khula to try to keep funds flowing. This situation was compounded by the monthly disbursement schedule for operating support. While Khula was right in noting the operational underperformance, the manner in which it engaged Nkwe took vital energy and opportunity away from Nkwe and Vulindlela's management to focus on key issues and, ultimately, contributed to putting Nkwe out of business prematurely.
On repeated occasions, Khula raised the same issues of weak governance of Nkwe's board of directors; conflict of interest between Vulindlela and Nkwe; lack of transparency in the relationships among key players (Accion, Vulindlela and Nkwe); and the high cost of Vulindlela's technical assistance/management team. It then signed and disbursed a contract after asking for certain amendments to the contracts between Vulindlela and NBI. These amendments were effected and the board issued a clarifying letter regarding its relationship with Nkwe. Private conversations suggest that Khula felt pressure to move forward with this investment in order to increase microfinance coverage in an underserved province.
Had Khula been more experienced in the practices surrounding equity investments and partnerships, it might have been in a position to suggest ways in which many of these issues could have been addressed. In fact, the management contract between Vulindlela and Nkwe does not conform to a typical contract of this nature. It does not protect against overdependence on one particular technical partner (Vulindlela) in the event that the manager needs to be replaced. It also ignores Vulindlela's responsibility for obtaining a partnership with a bank as a key performance indicator (although an early version of the contract reportedly did have a clause addressing this before it was revised).
Moreover, there is a lack of clarity in the way Vulindlela was to provide staff support to Nkwe in return for the management fee. The content of the management contract appears more like a technical assistance contract although, in practice, Vulindlela acted like a management contractor and eventually as an investor too. In fact, to the stakeholders, Vulindlela seemed to retain almost total operational control over Nkwe management and decisions throughout the entire period of the project. Ultimately, the lack of clarity surrounding these relationships and the legal documents supporting them, combined with several of Vulindlela's management practices, generated an aura of a lack of transparency. This, in turn, generated further distrust on the part of Vulindlela's partners, including Hivos, and provided a bone of constant contention. Had board governance and oversight been stronger, the perception of Vulindlela operating without sufficient checks and balances could perhaps have been addressed.
A far more important factor, however, is that the documentation suggests that Khula was never on board with the fundamental Nkwe proposition. Khula should not have signed a contract with Nkwe if it was not comfortable with the resolution of certain issues. These issues arose in monitoring reports again immediately following the signature and disbursement of the first tranches. There is no mention of a negotiated solution that Nkwe or Vulindlela subsequently failed to honour, leading to the conclusion that they had never been adequately addressed. If verbal assurances were given to Nkwe about any resolution of outstanding issues during the negotiation process, these did not find their way into the written record. This is particularly important given the high level of staff turnover in Khula whereby several different staff members dealt with Nkwe within a two-year period. We might logically conclude that each newly assigned staff member would refer to the written record in setting out his or her relationship with the organisation.
Vulindlela and Nkwe reasonably assumed that if Khula had signed an agreement, then its initial concerns had been satisfactorily responded to during the preceding correspondence. By subsequently playing out its core disagreements through a drawn-out process of constant renegotiation and withholding of tranches, did not ensure a positive and constructive relationship. This and other evidence about the length of time it generally takes Khula to negotiate any agreement with its RFIs suggests that Khula needs to sharpen up its management capacity to drive contentious issues to closure. Each potential investment will have its own peculiarities and Khula clearly needs to improve its ability to meet internally, discuss options, come to a management decision about what it is willing to accept, and then negotiate a definitive solution. One suspects that the ‘communication’ issues so often mentioned in reference to RFI relations with Khula are actually management issues. Staff cannot communicate clearly if decisions are not made clearly; and decisions, once made, are not backed up upon execution.
5. The role of Vulindlela as local promoter
Vulindlela receives relatively high marks for technical competence and creativity in seeking to establish a new microlending paradigm in South Africa. In fact, the fundamental concepts behind the Nkwe project represent leading ideas from the industry outside of the country, although the original promoter, MDF/NBI, played a less central role than is the case in similar models. NBI was a valuable funder for Nkwe, and also stepped in with additional loan funds at a late stage when Nkwe desperately needed them, but it did not play a conventional promoter role after contracting Vulindlela. Vulindlela had as its partner a world leader, Accion, in bank-RFI linkages. It was able to include on its staff individuals with significant experience in microenterprise lending in southern Africa. Vulindlela showed considerable commitment to the Nkwe proposition, including investing its own equity, and effectively playing the promoter role in the absence of an active early role from MDF/NBI.
In spite of the difficulty Khula had in accepting the role of promoter, technical assistance and management/governance that Vulindlela eventually took on, several leading international microfinance promoters follow a variation of this model and consistently garner support from the donor community. The best known is the German consulting firm IPC, although this approach is also used by Accion International and LFS in Bonn, Germany. Variations of the model have been used by Opportunity International, Shorebank, Finca and others. These organisations typically develop the initial proposal in response to a concrete opportunity (in some cases, the conversion of one of their own NGO microlenders into a regulated entity); promote the initiative among potential investors; obtain technical assistance funds to support the new institution; and sit on the board of directors. Many also place management directly into the new institution or closely control operations through resident advisers. In this model, promoters who are so committed as to also invest either hard cash or sweat equity into an initiative are actually preferred. A number of donors have accepted and funded this model.
The model is not without its detractors in the donor community, though; and in this Khula is not alone. Many in the donor community, used to non-profit organisations, are suspicious of giving a promoter too much control and look for nominal checks and balances in the complete separation of different roles among technical assistance providers, investors, donors, managers and boards of directors. Whether this attitude actually leads to better performance in the NGO sector remains open to question, but clearly the model initially conceived by MDF/NBI and developed by Vulindlela bundled far too many of these functions together for the comfort level of the donors involved. In its execution, the set of agreements defining the relationships among parties were not put in place.
Vulindlela and the board also failed to put in place all the checks and balances that would be required in the proposed institutional arrangements, a failure which, in turn, generated substantial discomfort on the part of virtually all of its partners. It had tried to retain an NGO structure while moving towards a hybrid structure of some sort, without ever seeming to understand the full nature of the adjustments that would be needed. Accion only played a marginal role, yet has vast international experience in the transformation from NGOs to national business finance institutions, which could have been of far greater assistance in sorting out the process.
Unfortunately for the project, none of the partners successfully thought through the diverse issues arising from the hybrid nature of the proposition. While all share blame for not sorting out conflict of interest, governance and accountability issues in their respective bilateral relations, Vulindlela shoulders the ultimate responsibility as the de facto promoter. As a result, the actual implementation phase appears as a series of (albeit well-intentioned) moving of targets, shifting of funds, shuffling of staff, changing of business plans and, even more fundamentally, transformations in the fundamental nature of the proposition.
In the relatively short period of 18 months:
| 1. | The project concept evolved from a form of joint venture with private investors to create a sustainable microbank paying dividends, to a typical donor-subsidised and ultimately unsustainable NGO. | ||||
| 2. | The target area moved from the NWP, to a national business plan, to a pilot in Soweto. | ||||
| 3. | Vulindlela's role became more and more central, with the organisation becoming increasingly committed to Nkwe, and the distinction between it and Nkwe becoming increasingly blurred. Vulindlela moved from providing technical assistance with seconded staff as temporary managers, to direct on-staff management. | ||||
| 4. | A portion of funds from Khula and Hivos was switched from their original intended purpose in the way they were actually utilised, related to funding shortages. | ||||
| 5. | The number of actual clients fell far short of the projected number of clients to be served according to the business plan. | ||||
While each of the steps in the transformations and modifications are accompanied by reasonable explanations, the sum of the parts leaves the impression that the end result just was not what any of the actors had signed on for. Whatever the reasons, the sheer number of non-performance items, from the inability to find enough clients, the inability to hire adequate managers for Nkwe, the failure to secure a bank partner, to the ultimate inability to appease the sole donor agency that held the key to the financial future of Nkwe, all add up to a reasonable and grounded loss of confidence in the project on the part of its partners. Vulindlela was not able to take this project concept through to its final execution. In the final analysis and with perfect hindsight, Vulindlela should probably not have undertaken to start operations until more of the pieces to the puzzle were in place.
Most importantly, Vulindlela did not have enough funding signed up and committed by donor organisations to survive the uncertain approval and disbursement processes involved. Khula reportedly regularly takes a year and a half to actually sign agreements with organisations it supports, and difficulties with its disbursements are well known within the industry. Clearly, Vulindlela understood the signals it was getting from its partners to be affirmative and supportive of the plan of action upon which they embarked, yet did not fully prepare for what it should have expected in terms of donor process and disbursement delays.
While Vulindlela assumes the ultimate responsibility in this, each partner must also examine why things progressed so far along, with clear indications of support from their side, only to have decisions subsequently reversed. One suspects that the unravelling of Nkwe is also related to the quality of relations among the partners, and of the partners with Vulindlela management. There was clearly a creeping sense of discomfort and loss of confidence in management, which Nkwe's board did not effectively respond to or address.
While there were no suggestions of actual impropriety on management's part, to the outsider reader, Vulindlela's repeated responses to the conflict, internal control and governance issues do not suggest that it was trying to move closer to the core of the concerns being expressed. The tenor of the responses on both sides (i.e. Vulindlela/Nkwe and Khula) shows frustration and a lack of respect for, or understanding of the position of the other. There was not much movement over the life of the relationships and the overarching impression is one of partners being stuck in their relative positions, even in the face of changing circumstances. While the promoter struck out seeking almost desperately one solution after another to save the initiative, ultimately it was creating a confidence crisis rather than actually moving the process forward. Although Nkwe's board increasingly took responsibility for communication with Khula in place of Vulindlela, this gap in trust and effective communications was never fully bridged.
6. Conclusions: generalisable lessons learnt
The list of lessons offered here is neither complete nor detailed, but rather is intended to provide positive and forward-looking insights from this case in key areas of relevance to the microfinance sector in South Africa.
6.1 For RFIs and their promoters
Generalisable lessons for RFIs and their promoters are difficult, as much about the Nkwe case could have turned out differently given different decisions, timing and context. Decision factors for banks in South Africa to expand into microlending have also changed since even two years ago. The following lessons are offered, with these important caveats in mind:
| 1. | If engaging a commercial bank is an objective, securing top management buy-in is virtually essential. | ||||
| 2. | Unless using a community banking or low-level group model, start in a relatively strong market before expanding into secondary, less dynamic and more challenging markets. (In this case, Accion indicated that pre-launch market research was permeated by wishful thinking, because it was driven by the funder, the NWP.) | ||||
| 3. | Relationships with funders are key to an RFI's survival and performance. The management and board need to invest sufficient time in these relationships and be willing to respond to their concerns. (In this case, it is not clear whether concerns were effectively addressed.) | ||||
| 4. | An RFI must have an approach and the capacity to deliver, and must mobilise funds around that proposition. Constant changes in approach, albeit seemingly justified, provoke misunderstanding, distrust and a sense that the RFI does not know what it is doing. | ||||
| 5. | Technical assistance, whether from local or international sources, is an essential ingredient in the development of new financial products and can be a relatively expensive initial investment. (In this case, the cost of technical assistance was confused with the cost of management by Vulindlela, creating the impression that the technical assistance was even more expensive than it needed to be, and the consequence was resistance by Khula.) | ||||
| 6. | Structure and clarity of relationships within the organisation are essential. (In this case, agreements between parties were either not formalised, and therefore could not form the basis for project operation, or were not fully accepted by all parties even where formal agreements existed.) | ||||
| 7. | Management information systems underpin RFI operations and should be installed and operationalised at an early stage. (In this case, Vulindlela and Accion imported a system that turned out to be problematic to install and had higher-than-projected installation and maintenance costs.) | ||||
6.2 For wholesale funders
According to a study commissioned by CGAP (Citation2002), common factors worldwide causing failure of wholesale funding mechanisms to deliver, and which are relevant to this case, include:
| 1. | Insufficient separation from local or state government priorities and influences | ||||
| 2. | Pressure to disburse too large a volume of funds relative to the absorptive capacity of RFIs | ||||
| 3. | Not allowing sufficient flexibility for RFI management, within a framework of agreed performance criteria and effective monitoring | ||||
| 4. | Inability to accept and support innovative models | ||||
| 5. | Crowding-out (rather than facilitating) commercial investment in RFIs. | ||||
Positive lessons learned from the Nkwe case for wholesale funding mechanisms are:
| 1. | Use a grant-based capacity-building facility to complement loan facilities in building RFI absorptive capacity. Technical assistance from such a facility should be designed in partnership with RFIs, not prescribed centrally. | ||||
| 2. | Set minimum threshold performance standards for RFIs below which support will be withdrawn (rather than optimistic performance targets), in the absence of a powerful justification for missing the thresholds. | ||||
| 3. | Where more than one donor or investor finances an RFI, joint monitoring and reporting based on common performance targets (as was the case with Khula and Hivos) can avoid laying an excessive reporting burden on the RFI, and will improve communication. | ||||
| 4. | Within the framework provided by those standards, allow RFI management sufficient room to make mistakes or adjust to unexpected circumstances. | ||||
| 5. | Offer guarantees to facilitate commercial loans, and only lend for RFI portfolio needs if there is no possibility that a commercial bank might otherwise do so. | ||||
| 6. | Ensure that apex staff and management are sufficiently well skilled and free from political or government influence to be able to take informed and calculated ‘risks’ on supporting innovative approaches to microfinance. | ||||
| 7. | Improve apex decision making, taking clearer action earlier and communicating this as early as possible. | ||||
| 8. | Treat RFIs as ‘clients’ and never as subcontracted implementation agencies. | ||||
6.3 For commercial funders or investors
For this group, the lessons are as follows:
| 1. | Microfinance institutions and promoters can possess valuable client knowledge and lending expertise, while banks tend to have the back office capacity, legal resources, branch networks and access to funds that are required for a large and efficient operation. The ‘service agent’ model offers a way to harness the respective strengths of each. Here, the bank owns the microloan portfolio and pays a fee to a specialised microfinance entity, in which it may or may not have an ownership stake, to manage that portfolio. Accion is an expert in this approach, although in this case the model was not followed. | ||||
| 2. | Potential investors need to be clear about their demands right from the start, including appropriately skilled managers, reporting requirements in line with international accounting standards, loan portfolio performance, an appropriate organisational legal structure, board membership, etc., to avoid later delays and misunderstandings. | ||||
| 3. | A bank can profitably lend to RFIs that have an established track record of sustainable operations and excellent portfolio quality, without the more complicated engagement of an equity investment. | ||||
Reference
- Consultative group to assist the poor (CGAP) 2002 Apex institutions in microfinance CGAP Donor Brief No. 5, July 2002. Available online at:, http://www.cgap.org/docs/DonorBrief_05.pdf
Annexure A: Persons interviewed
Cas Coovadia, Banking Council and Nkwe Board Chair
Wagied Allie, Programmes Manager, NBI and Nkwe Board
Peter Roussos, Managing Director, Vulindlela
Barbera Calvin, Senior Director, Vulindlela
Beth Rhyne, Accion
Xola Sithole, Chief Executive Officer, Khula
Sandile Luthuli, General Manager: Operations, Khula
Ndumiso Mpofu, Investment Officer, Hivos
Sonja van Vliet, Manager: New Business Opportunities, ABSA Flexi Banking Services
Mike Alman, General Manager: New Business, ABSA Flexi Banking Services
Gabriel Davel, Chief Executive Officer, Microfinance Regulatory Council
Jennifer Hoffmann, Managing Director, Teba Bank
Roland Pearson, Managing Director, Siana Strategic Advisors (Pty) Ltd
Gerhard Coetzee, Director, ECI Africa
John de Wit, Director, Small Enterprise Foundation
Wessel Venter, Director, Beehive Enterprise Development Centre