Micro Agriculture Finance Institutions of South Africa

Transcription

1 Expenditure and Performance Review Micro Agriculture Finance Institutions of South Africa Final May 2014 Research commissioned by the National Treasury

2 Version control of drafts submitted to National Treasury Date submitted Draft number shown on cover Name of file Draft v MAFISA EPR Report v Draft v MAFISA EPR Report v Draft v MAFISA EPR Report v.3 edited Draft v MAFISA EPR Report v Draft v MAFISA EPR Report v Draft v MAFISA EPR Report v Final MAFISA EPR Report Final - Full Project team Jonathan Carter Dominic Mitchell Conrad Barberton Carmen Abdoll jonathan@cornerstonesa.net dominic@realconsulting.co.za conrad@cornerstonesa.net carmen@cornerstonesa.net Senior Researchers with Cornerstone Economic Research

3

4 Key EPR Observations Micro Agriculture Finance Institutions of South Africa (MAFISA) Purpose of MAFISA 1. The objective of MAFISA is to provide financial services for empowering smallholder/micro level producers and processors in the agriculture, forestry and fisheries sectors. MAFISA is the financial support pillar of the Comprehensive Agricultural Support Programme (CASP). 2. When government announced MAFISA, it indicated that R1 billion would be allocated to it. DAFF appoints financial intermediaries on three- to five-year agreements, and gives them a specified amount of this MAFISA capital to make micro loans to beneficiaries. The beneficiaries pay 8 per cent interest on the loans. The financial intermediaries carry the risk should the beneficiaries default, unless there are special circumstances such as disasters in which case DAFF carries the risk. Outcomes from the costing model 3. The costing model shows that DAFF and the provincial departments of agriculture do not need to spend a lot on managing MAFISA. However, their roles in making the scheme work are absolutely critical and there is currently much scope to improve their performance. 4. The costing model shows that managing the rate of default on MAFISA loans is the most important variable in the design of the scheme, as it has the largest impact on how prudent financial intermediaries manage MAFISA loans. 5. The model indicates that MAFISA is only viable at default rates of less than 5 per cent. This means there is very little margin for error. Therefore prudent financial intermediaries must either set strict access conditions or cut the smallholder farmers out of the production process. 6. Financial intermediaries that have adopted a more relaxed approach (see below) to lending stand to incur very significant losses. The following table shows the outcome of modelling the operations of four financial intermediaries and the losses they are likely to incur on their MAFISA lending: SUMMARY OF MAFISA MANAGEMENT COST Scenario of Agreements with Financial Intermediaries write off Year 1 Year 2 Year 3 Year 4 Year 5 Net MAFISA - financial intermediary profit/loss Total Cost over five year agreement NERPO Profit/Loss from MAFISA 20% R R R R R R Kaap Agri Profit/Loss from MAFISA 3% R R R R R R SASA Profit/Loss from MAFISA 0% R R R R R R ECRDA Profit/Loss from MAFISA 70% R R R R R R Profit/Loss from MAFISA R R R R R R The intention of MAFISA is to reach smallholder farmers. This means default rates will be fairly high. At present the financial intermediaries carry the full risk of non-performing loans. A better approach would be for DAFF to take the first risk of 10 or 20 per cent, for example, and anything above that to be carried by the financial intermediary. This would recognise the inherent risks of lending into the micro-agricultural market but still incentivise prudent lending. Other key findings from the performance and expenditure analysis 8. DAFF needs to strengthen its capacity to appoint financial intermediaries, manage the MAFISA database and guide provincial departments in providing support to MAFISA. 9. Mechanisms need to be put in place to review the MAFISA product every two years to ensure it remains attractive to beneficiaries in the changing interest rate environment. 10. There is a perception among beneficiaries that grants offered by the Department of Rural Development and Land Reform compete with MAFISA loans. It was reported that certain beneficiaries choose not to take MAFISA loans because they feel it is unfair they should be required to repay loans while government gives grants to other small farmers. There is a policy co-ordination problem that needs addressing. 11. The expenditure analysis shows that DAFF has been allocating the MAFISA funds to entities not related to MAFISA. The circumstances surrounding some of these transfers warrant a forensic audit to establish whether or not there have been any irregularities.

5 12. Recent action by National Treasury to withdraw MAFISA funds can be extended to wind-up MAFISA. This could allow for more suitable products to be extended through the Land Bank.

6 1 Table of Contents LIST OF ABBREVIATIONS... 1 EXECUTIVE SUMMARY INTRODUCTION TOPIC PURPOSE AND SCOPE KEY QUESTIONS INFORMATION USED IN CONDUCTING THE REVIEW LIST OF DOCUMENTS THAT MAKE UP THE REVIEW BACKGROUND TO MAFISA MAFISA S OBJECTIVE TARGET MARKET INTENDED IMPACT DISBURSEMENT MODALITY LOAN SIZE AND INTEREST RATE DATA GATHERING ON MAFISA S PERFORMANCE COMPETING PRODUCTS OFFERED BY GOVERNMENT ANALYSIS OF THE MAFISA MANAGEMENT PROCESSES ROLE-PLAYERS IN THE DELIVERY OF MAFISA OVERVIEW OF THE FLOW OF MAFISA FUNDS LIABILITY FOR BAD DEBTS KEY FINDINGS REGARDING THE MAFISA MANAGEMENT PROCESSES THE PERFORMANCE ANALYSIS EXISTING INFORMATION ON MAFISA PERFORMANCE MEASURES INFORMATION ON MAFISA PERFORMANCE LOANS ISSUED THE CHAIN OF DELIVERY AND PROPOSED INDICATORS FOR MAFISA KEY FINDINGS REGARDING THE PROGRAMME ANALYSIS EXPENDITURE ANALYSIS TRANSFERS TO THE MAFISA FUND TRANSFERS FROM THE MAFISA FUND USE OF MAFISA FUNDS BY THE FINANCIAL INTERMEDIARIES OTHER GOVERNMENT EXPENDITURE THAT SUPPORTS MAFISA OTHER GOVERNMENT EXPENDITURE THAT APPEARS TO COMPETE WITH MAFISA KEY FINDINGS FROM THE EXPENDITURE ANALYSIS COSTING MODEL STRUCTURE OF THE COSTING MODEL COSTING MODEL OUTCOMES The cost to government of managing MAFISA The cost to the Land Bank of managing the MAFISA account The cost to financial intermediaries of managing MAFISA loans KEY COST DRIVERS KEY FINDINGS EMERGING FROM THE COSTING CONCLUSION: OPPORTUNITY TO REVIEW MAFISA ANNEXURE A INTERVIEWS CONDUCTED FOR THIS EPR ANNEXURE B: TRANSFER OF MAFISA FUNDS TO LAND BANK ANNEXURE C: TRANSFERS FROM THE MAFISA CALL ACCOUNT ANNEXURE D: TRANSFERS FROM THE MAFISA DEPOSIT ACCOUNT ANNEXURE E: INFORMATION ON MAKHATHINI COTTON COMPANY... 44

7 1 List of abbreviations CASP CIPC DAFF DPME ECRDA ENE EPR FI GEP GGP LRAD MAFISA MEGA MGK MoU NERPO NGO PDAs PDI SASA SEFA SLAG Comprehensive Agricultural Support Programme Companies and Intellectual Property Commission Department of Agriculture, Forestry and Fisheries Department of Performance Monitoring and Evaluation Eastern Cape Rural Development Agency Estimates of National Expenditure Expenditure and Performance Review Financial intermediary Gauteng Enterprise Propeller gross geographic product Land Redistribution for Agricultural Development Micro Agricultural Financial Institutions of South Africa Mpumalanga Economic Growth Agency Magaliesberg Grain Corporation Ltd Memorandum of understanding National Emergent Red Meat Producers' Organisation Non-governmental organisation Provincial departments of agriculture Previously disadvantaged individual South African Sugar Association Small Enterprise Finance Agency Settlement/Land Acquisition Grant

8 2 Executive Summary The stated objective of MAFISA is to provide financial services for empowering the smallholder/micro level producers, processors, working poor, micro entrepreneurs and producers in the agriculture, forestry and fisheries sectors. MAFISA is the financial support pillar of the Comprehensive Agricultural Support Programme (CASP). In 2005 government allocated to R1 billion to DAFF to establish MAFISA. DAFF appoints financial intermediaries on three to five-year agreements and gives them a specified amount of the capital with which to make micro loans to beneficiaries. The beneficiaries pay 8 per cent interest on the loans, of which 7 per cent is retained by the financial intermediary and 1 per cent is returned to the MAFISA account on expiry of the agreement period. The financial intermediaries carry the risk should the beneficiaries default. DAFF appointed the Land Bank to manage the MAFISA account. The Land Bank only disburses funds on DAFF s instructions. Key findings regarding the MAFISA management processes It is evident that there is significant divergence between the intentions of the MAFISA policy framework and how MAFISA is being managed in practice: 1. When MAFISA was introduced DAFF appointed nine financial intermediaries to provide services across the nine provinces. One of the intermediaries has since withdrawn from the scheme and two intermediaries contracts have ended and not been renewed. There are currently no financial intermediaries with offices located in the Northern Cape, North West, Free State or the Western Cape. These provinces are therefore under-served by MAFISA despite a large presence of smallholder farmers. 2. DAFF s allocation of contracts to financial intermediaries has in some instances restricted access to MAFISA loans for smallholder farmers to particular sectors, such as the red meat or sugar sector, since the intermediaries do not offer services to smallholder farmers in other sectors. 3. The policy framework clearly states that there should be a close working relationship between the provincial department of agriculture and the financial intermediaries. The primary reason for putting these working relationships in place is to ensure that extension officers assist potential beneficiaries in applying for loans, and that there is co-operation in providing technical extension support to the beneficiaries so as to improve the likelihood of success (and so mitigate the risk of default). However, very few provinces actively promote or support MAFISA. 4. The way in which the financial intermediaries manage the MAFISA loans impacts on the realisation of the policy objectives. a. The majority of loans extended by the ECRDA are based on unsecured lending to what appear to be subsistence producers. The expected default rate of 70 per cent suggests that few of these producers are progressing. Instead, the loan is pushing them into a debt trap. b. In KwaZulu-Natal the SASA finances sugar production on communal or smallholder land by exclusively using and paying contractors to execute all the production activities. If there is a margin, this is paid to the landowner. This approach reduces the risk of default to virtually zero. However, the land is effectively being rented using MAFISA funds, and the objective of empowering smallholder farmers and growing their capacity to engage in the agricultural sector is being missed. c. NERPO only lends to smallholder livestock farmers. As a red meat producer s organisation, NERPO is able to control the sale of the produce of beneficiaries. This gives NERPO some control over the cash flow of MAFISA beneficiaries and therefore the repayment of loans. NERPO has achieved an 85 per cent repayment rate. It is also worth noting that the Land

9 3 Bank recommended that the Wholesale Finance Facility, which is a product similar to MAFISA, be piloted by NERPO. d. Kaap-Agri, which served farmers in the Western Cape and Northern Cape, has generally extended larger loans to fewer beneficiaries. Unlike SASA and NERPO, Kaap-Agri is unable to control the distribution of the produce of beneficiaries and there have been two instances where beneficiaries have defaulted on their loans. 5. The provisions governing bad debts have not been tested, largely because most of the agreements with financial intermediaries are still current. It is known that Kaap-Agri applied to the DAFF in 2013 to have two loans written-off against the MAFISA capital, i.e. for the cost of the bad loans to be carried by DAFF and not the financial intermediary. DAFF is processing these requests, and a final decision has not been given yet. Whatever decision is made will have significant implications on the management of MAFISA going forward, and even on the financial viability of certain intermediaries such as the ECRDA. 6. MAFISA was initiated in 2004/5 when prime interest rates in the commercial market were 11 per cent. The rates then rose to 15.5 per cent in 2008 before dropping significantly in to 9 per cent. An interest rate of 8 per cent when the programme was piloted was very attractive; however in the current environment MAFISA loans do not have the same appeal. Any loan product needs to be able to respond to the prevailing interest environment in order to remain relevant. Key findings regarding the programme analysis 7. It is important to conduct proper planning before proceeding from a pilot phase to full-scale implementation. In the case of the MAFISA the transition to full-scale implementation was very rapid. Consequently, certain key implementation modalities were not adequately addressed. For instance, there are no standard operating procedures for the relevant role-players, the systems to collect and collate data and the capacity to monitor and analyse performance needed strengthening. 8. The programme was launched as a pilot and proceeded to the roll-out phase without adequate reporting and data capturing systems being embedded and standardised. The Land Bank and the financial intermediaries have been submitting data to DAFF, but the data has not been captured on a properly organised database or used to monitor MAFISA. There have also been problems with the standardisation of data, as well as non-compliance in the submission of information. Now, eight years into the programme, DAFF is taking steps to introduce improved data gathering by financial intermediaries. However, it is proposed that this system should be reviewed by a data analyst specialist before being fully implemented. 9. Neither CER nor the team conducting the impact assessment have received any meaningful information or analysis of the performance of MAFISA from DAFF. This is unusual, as we would expect DAFF to be publishing information on the performance of MAFISA on a regular basis, given the importance of the programme to government. 10. According to the ENE expenditure figures, MAFISA loans were issued between 2006/07 and 2011/12. However, information collected directly from the MAFISA financial intermediaries at the end of 2013 indicates that only about MAFISA loans have been issued since the scheme s inception. This is almost six times less than the figure reported in the ENE documents. According to DAFF officials the difference is due to an error in reporting: the higher numbers in the ENE documents are based on the number of drawdowns against loans issued and not on the actual number of individual loans issued. 11. The success of the MAFISA scheme depends on there being co-operation between the provincial departments of agriculture and the financial intermediaries. From interviews conducted, it would appear that only one province is providing any meaningful support to MAFISA. Officials from DAFF noted that MAFISA policy documents and the agreements with financial intermediaries do spell out the roles and responsibilities of provincial departments of agriculture, but also noted that DAFF cannot enforce compliance with these roles and responsibilities because agriculture is a concurrent

10 4 function. However, spelling out roles and responsibilities is not enough. DAFF ought to go further and put in place a standard operating procedure to guide provinces on how to support the scheme. 12. When DAFF embarked upon the roll-out of MAFISA, provincial departments of agriculture were asked to nominate a MAFISA champion who would take responsibility for promoting and coordinating the programme within the provincial department, as well as communicating with DAFF. Ideally provincial departments should have nominated the manager of the farmer support services directorate because this is where the extension officers responsible for implementing CASP and providing technical support to farmers are all located. Indeed DAFF should have insisted on the MAFISA champion being located in this directorate. The failure to do so means that in certain provinces DAFF is not communicating with the managers best situated to take MAFISA forward. Key findings from the expenditure analysis 13. DAFF has not been forthcoming with information regarding the use of MAFISA funds. The most recent information received from DAFF dates from December Of concern is that recent information received from the Land Bank regarding the management of the MAFISA accounts indicates that there is reason to be seriously concerned about how MAFISA funds are being used. 14. The funds paid into the two MAFISA accounts managed by the Land Bank do not reconcile with the amount allocated to DAFF for MAFISA through the budget. The information from the Land Bank also shows that R415.9 million more was transferred out of the two MAFISA accounts than into them. The information provided does not reflect either interest payments or repayments of capital by financial intermediaries. The Land Bank should be instructed to provide a full reconciliation of all funds flowing into and out of the two MAFISA accounts. 15. DAFF has transferred R608.8 million from the MAFISA accounts held by the Land Bank to approved financial intermediaries for lending out as MAFISA loans, and a further R899 million to non-mafisa entities. 16. When Parliament allocated funds to MAFISA through the budget process, the allocations were made specific and exclusive for MAFISA. This means neither DAFF nor National Treasury may vire or reallocate MAFISA funds for other purposes without going back to Parliament for approval. Usually this would be done through an adjustments budget. We could not find any evidence that Parliament has approved alternative uses for the MAFISA funds through an adjustments budget. 17. The above implies that, in all instances where DAFF has used MAFISA funds for non-mafisa purposes, these expenditures are unauthorised in terms of the PFMA. Where DAFF has transferred funds to buy a liquidated company (Makhathini Cotton) and allowed the asset to be vandalised, this amounts to fruitless and wasteful expenditure in terms of the PFMA. In line with the provisions of the PFMA, these expenditures need to be investigated and, where appropriate, steps taken to recover the funds, which may include recovering them from the official/s who authorised these expenditures, as well as laying charges of financial misconduct against them. 18. The most recent transfers to Peulwana are suspicious. The MAFISA agreement between DAFF and Peulwana only approves R20 million be transferred, but R208 million has been transferred to date. This suggests that either this agreement has been altered or DAFF is acting contrary to the agreement. Peulwana s website does not suggest that it is a very large operation. It is therefore not clear why DAFF has transferred such large amounts to it. 19. The latest transfer to NERPO of R50 million does not appear to be in accordance with the MAFISA agreement. The timing of the transfer on 21 December 2011 also raises questions. 20. The information received from the Land Bank suggests the following: a. DAFF has been treating the MAFISA funds (and other funds) as slush funds, and allocating them to whatever (pet) projects they choose without any regard for the intended use of the funds as determined by Parliament through the budget process.

11 5 b. DAFF may have been channelling departmental funds into the MAFISA accounts and then transferring them to other entities as a mechanism to avoid normal oversight processes. This may explain why more funds appear to have been transferred into the MAFISA accounts than the allocations indicated in the ENE. 21. Concerns about the management of the MAFISA funds have resulted in National Treasury recalling the unutilised MAFISA funds on 19 October Acting on National Treasury s instruction of 19 October 2013, the Land Bank returned R136.4 million of the unutilised MAFISA funds to the national revenue fund. This effectively caps MAFISA to the funds currently being managed by financial intermediaries. 22. When the MAFISA agreements with the financial intermediaries terminate, they are required to repay the capital plus 1 per cent interest to the MAFISA account at the Land Bank. It is not clear whether National Treasury s instruction of 19 October 2013 covers these funds as well. Given the concerns around DAFF s management of these funds, it is proposed that National Treasury should instruct the Land Bank to also return these funds to the national revenue fund. 23. Annual spending by provinces on farmer support is nearly three times greater than the original R1 billion allocation to MAFISA. In light of this, these programmes contribution to supporting smallholder farmers is far greater than MAFISA, even though access to financing is an important supporting service. 24. There is anecdotal evidence that MAFISA beneficiaries feel it is unfair that they have to repay their loans plus interest while other small-scale farmers receive free starter-packs from the provincial departments of agriculture or through the land reform grants. This points to the need for better communication with beneficiaries regarding the MAFISA loans and the other grants, as well as the need for greater policy synchronisation. Key findings emerging from the costing 25. The costing analysis shows that the key cost drivers for the different MAFISA role-players are: a. For DAFF and the Land Bank, the cost of employment. b. For the provincial departments of agriculture, the cost of employment, travel, the extent of support provided to MAFISA applicants and the extent of extension support provided to MAFISA beneficiaries. c. For the financial intermediaries, there are a number of processes that affect the cost of managing the MAFISA loans, and the resources spent on each process varies substantially between intermediaries given how the use of the MAFISA loans differs widely. 26. The following table sets out the most important service delivery parameters and how they relate to the cost of managing MAFISA loans. Table: Cost drivers for financial intermediaries in managing MAFISA loans Cost driver Risk of default The interest rate Number of beneficiaries applying for MAFISA loans Management of MAFISA portfolio within the financial intermediary Implications The model has separate risk rates for capital repayments and interest repayments. Increasing the risk rate increases the losses from loans. Capital losses have the largest effect on the likelihood that an intermediary will make a loss. The interest rate on MAFISA loans determines their competitiveness with other loan products, as well as the maximum return the financial intermediaries can earn from participating in the scheme. The rate was set at eight percent at the inception of the scheme and has not changed since. The more beneficiaries who apply for MAFISA loans, the more is spent assisting beneficiaries to complete applications. This is modelled separately for the PDAs and financial intermediaries It is possible to reduce the impact by changing assumptions around the proportion of beneficiaries who need assistance and the level of assistance required. The model allows for any size of management structure dedicated to MAFISA to be modelled. Generally, MAFISA is one of many of the job responsibilities of a few employees at an intermediary. However, increasing the number of people and time spent on MAFISA will impact on the personnel and operation costs of MAFISA.

12 6 Overall conclusions A Dealing with the risk of default 27. Interviews with financial intermediaries and DAFF showed that, while according to current data there is no significant default trend with either financial intermediaries or beneficiaries, this is not a reflection of the reality on the ground. As the loan agreements expire or need to be renewed, loan defaulting will become apparent. The ECRDA indicated it anticipates a 70 per cent or R80 million default book a significant percentage of the whole MAFISA product. SASA, by contrast, reports zero defaults. 28. The risk, or the rate of default, has the greatest impact on how prudent financial intermediaries manage MAFISA loans. Our modelling suggests that MAFISA is only viable at write-off rates of less than 5 per cent. This means there is virtually no margin for error, and so prudent financial intermediaries have either set strict access conditions to minimise the risk or cut the farmer out of the process completely by requiring that all payments go to contractors that farm the land. Where financial intermediaries have adopted a more relaxed approach to lending, they stand to incur very significant losses. 29. If the intention of MAFISA is to reach smallholder farmers, the issue of how write-offs are dealt with needs to be re-examined. The costing model allows for a risk sharing approach to be modelled for instance, 50 per cent of write-offs can be attributed to the financial intermediary and 50 per cent to DAFF. A better approach may be that DAFF takes the first 10 or 20 per cent and anything above that gets attributed to the financial intermediary. This approach would incentivise prudent lending by the financial intermediaries. B Changing interest rate environment and competing products 30. Six of the original nine MAFISA-accredited financial intermediaries 1 are financial service providers in their own right, and as such have their own loan or loan-type products to offer beneficiaries. This means that they may be competing with the MAFISA product. In some cases their own products have lower interest rates when compared with MAFISA loans, and more favourable repayment terms. The MAFISA interest rate has not changed since its inception in 2005/06, even though interest rates have dropped significantly since then and were at historic lows until very recently. Any loan product needs to be able to respond to the prevailing interest environment in order to remain relevant. C Opportunity to review MAFISA 31. National Treasury s instruction of 19 October 2013 to the Land Bank to return unutilised MAFISA funds to the national revenue fund has effectively put a cap on the MAFISA scheme. If, as proposed National Treasury were to instruct the Land Bank to return all MAFISA funds that the financial intermediaries repay at the conclusion of their agreements to the national revenue fund, this will result in MAFISA being wound-down and closed. 32. Given the serious concerns about DAFF s fiduciary management of MAFISA funds, and the need to investigate the probity of certain transactions, it is therefore probably advisable to close MAFISA. It needs to be assessed as to whether this will leave a significant gap in the agricultural micro-finance market, as there is a range of government grants and other loan products available to small-scale farmers. 33. If the various concerns with DAFF s management of MAFISA can be sorted out, then there is no reason why a remodelled MAFISA should not be re-launched sometime in the future. Alternatively, it may be advisable to build-up a product such as the Wholesale Finance Facility that has been developed by the Land Bank. 1 According to DAFF officials SASA, NERPO and Peulwana only provide MAFISA loans.

13 7 1. Introduction This Expenditure and Performance Review (EPR) was commissioned by the National Treasury. It forms part of a broader initiative to understand the planning, implementation and financing dynamics of some forty policy initiatives of government. It is envisaged that EPR studies such as this will allow for greater value for money in government expenditure and improved implementation of policy initiatives. 1.1 Topic This EPR examines the Micro Agricultural Financial Institutions of South Africa, otherwise known as MAFISA. 1.2 Purpose and scope There is frequently a large gap between policy development and implementation planning and practice. Many policy initiatives do not have good plans with clear, well-developed logic models or theories of change, baseline indicators, clear outcomes or impact indicators, and targets. There are serious concerns that MAFISA is not performing well. This prompted the Department of Performance Monitoring and Evaluation (DPME) to commission an impact assessment of MAFISA in The aim of this study was to evaluate the extent to which MAFISA has succeeded in reaching smallholder farmers, and whether or not it has assisted them in terms of increasing scale of production and penetration of markets. The purpose of this EPR is to explore the roles and responsibilities of the different role-players responsible for delivering and supporting MAFISA, as well as these institutions cost structure for supporting and managing MAFISA. This study is complementary to the DPME impact assessment in that it explores some of the key factors that may have affected MAFISA s actual performance. In terms of scope, this EPR aims to analyse the relevant role-players responsibilities in managing and supporting the MAFISA scheme by providing: a detailed analysis of the implementation logic of MAFISA, enabling critical performance measures to be developed across all components of the scheme; detailed understanding of the cost elements, expenditure drivers and expenditure associated with managing and supporting MAFISA; and proposed improvements for linking policy intent and programmatic design, based on a better understanding of the costs and causal linkages. It was agreed that the project would examine the operations of four of the financial intermediaries involved in providing MAFISA loans, namely: National Emergent Red Meat Producers' Organisation (NERPO), Eastern Cape Rural Development Agency (ECRDA) 2, Kaap-Agri, South African Sugar Association (SASA). 2 According to a statement issued by the Eastern Cape MEC for Rural Development on 26 June 2013 the Eastern Cape Rural Development Agency was formed by amalgamating the Eastern Cape Rural Finance Corporation and Asigisa-EC. See:

14 8 1.3 Key questions With the above in mind, this project aims to examine the following key questions: i. Are the role-players responsible for supporting MAFISA providing the required support? ii. What performance measures should be tracked to ensure the role-players contribute meaningfully to the success of MAFISA? iii. What are the key cost elements and expenditure drivers of managing and supporting MAFISA? iv. Is MAFISA appropriately designed? 1.4 Information used in conducting the review The following sources of information were used in conducting this review: Title MAFISA Credit Policy Framework, 2012 MAFISA Pilot Project Credit Policy, 2005 Strategic Plan for Smallholder support, /14 Strategic Plan 2011/12 to 2014/15 for the Department of Agriculture, Forestry and Fishers Strategic Plan 2012/13 to 2016/17 for the Department of Agriculture, Forestry and Fisheries Strategic Plan 2013/14 to 2017/18 for the Department of Agriculture, Forestry and Fisheries for Land and Agrarian Reform in South Africa, 2007 Presentation: 07 August 2012: Department Of Agriculture, Forestry And Fisheries Micro Agricultural Financial Institutions Of South Africa MAFISA Land Reform Policy Discussion document: June 2012 Interest rates in South Africa Selected historical rates Estimates of National Expenditure; 2005 to 2013 Source Department of Agriculture, Forestry and Fisheries Department of Agriculture Department of Agriculture, Forestry and Fisheries Department of Agriculture, Forestry and Fisheries Department of Agriculture, Forestry and Fisheries Department of Agriculture, Forestry and Fisheries Naefa Khan: Centre for Policy Studies The Portfolio Committee On Agriculture, Forestry and Fisheries Department of Rural Development and Land Reform ectedhistoricalexchangeandinterestrates.aspx National Treasury The team from Cornerstone Economic Research also conducted interviews with officials and representatives of the financial intermediaries responsible for managing MAFISA. Information was also shared with the consultants conducting the DPME Impact Assessment. Annexure A sets out the list of people interviewed for this project. 1.5 List of documents that make up the review The project outputs that comprise the review are: Step 2 and MAFISA ELEMENTS - Final this Excel workbook sets out the responsibilities of the departments in the different spheres of government,

15 9 the policy processes, a chain of delivery for each policy process, and the performance indicators. Step 4a MAFISA Vulindlela Expenditure Analysis this looks at historical spending by DAFF and provincial agriculture departments on activities related to MAFISA Step 4b MAFISA EPR Costing Model Final a costing model of MAFISA that allows for the cost structures of financial intermediaries to be modelled. Step MAFISA EPR Report Final this report. 2. Background to MAFISA Micro Agricultural Financial Institutions of South Africa (MAFISA) was conceptualised and established in It s envisaged purpose was to facilitate the provision of affordable and sustainable financial services to economically active rural poor communities. In line with the recommendations of the Strauss Commission, financial support was offered to beneficiaries through financial intermediaries (also known as Retail Lending Institutions) and not the Department Agriculture (now the Department of Agriculture, Forestry and Fisheries (DAFF)). Figure 1: Key MAFISA dates 2005 MAFISA pilot launched in three provinces the management of the Land Bank. 2008/9 MAFISA rolled out nationally under management of the Agricultural Finance and Co-operative Development Directorate (now called the Development Finance Co-ordination Directorate within the Chief Directorate of Development Finance, which reports to the CFO) within DAFF 2009 MAFISA transferred to the Chief Directorate: Development Finance DAFF MAFISA evolves as a pillar of the CASP programme MAFISA commenced as a pilot programme launched in three provinces in 2005, namely Limpopo, the Eastern Cape and KwaZulu-Natal. The main intention of MAFISA is to focus on smallholders in communal land areas and other small scale producers located on smallholdings with a variety of tenure systems. It is also intended to support and finance value-adding activities. MAFISA provides short- to medium-term production loans, savings mobilisation and capacity building for individuals and member owned co-operatives to enhance agricultural, forestry and fisheries activities. DAFF appointed a number of institutions operating in the agricultural finance sector to manage the MAFISA lending to beneficiaries. These are referred to as financial intermediaries. It was envisaged that MAFISA would leverage financial resources from the market through forging alliances with strategic partners based on their commitments in terms of Financial Sector Charter. In reality, this has only been partially realised. There are cases such as Kaap-Agri where MAFISA is seen as a loss-leader and part of the Financial Sector Charter or corporate social investment category of activities. There are a number of other examples where the MAFISA product is in direct competition with other financial services products offered by the financial intermediary. MAFISA focused solely on agriculture until Due to the restructuring of the ministry, it now also includes forestry and fisheries, though a very limited number of loans have been made to beneficiaries in these sub-sectors to date.

16 10 Also noteworthy is the envisaged linkage between MAFISA and the CASP. CASP has six focus areas, or pillars, as follows: information and technology management; technical and advisory assistance, and regulatory services; marketing and business development; training and capacity building; on/off farm infrastructure and product inputs; financial support. MAFISA is regarded to be the financial support pillar of CASP. According to the CASP strategy, farmers may receive grants, for example, for infrastructure and information technology, from CASP, but would be expected to take loans from MAFISA for production and input costs. 2.1 MAFISA s Objective MAFISA s stated objective is to provide financial services for empowering the smallholder/micro level producers, processors, working poor, micro entrepreneurs and producers in the agriculture, forestry and fisheries sectors. According to the updated policy documents, the facility is aimed to improve their livelihoods, reduce poverty, create employment opportunities, develop viable businesses, and assist them into developing into larger commercial businesses. As mentioned above, when the forestry and fisheries functions were shifted to the Department of Agriculture, the scope of MAFISA was extended to producers in these sectors. 2.2 Target market The updated 2012 MAFISA policy document identifies the following target markets for the MAFISA loan product: communal producers; women and youth and people living with disabilities; subsistence producers; smallholder producers; small agri-businesses; farm workers; user-owned self-help groups; community based organisations, NGOs and cooperatives involved in the agriculture, forestry and fisheries sectors; artisanal fisheries and small scale fisheries; forest growers and processors; aquaculture community-based projects; freshwater aquaculture producers and processors; aquaculture feed manufacturers/suppliers.

17 Intended impact MAFISA is intended to have the following high-level development impact in the sector/sub-sectors in which it is active: Entrepreneurial development: through the identification and servicing of market opportunities in the agricultural sector; Enterprise development: through stimulating new agri-business activities and enterprises that are economically viable, and expanding some existing businesses; Employment creation: through financing business activities that demonstrate the ability to create, sustain and increase employment; and Economic growth: through increasing turnover and asset value in the borrowing entities. 2.4 Disbursement modality Production loans and/or working capital is supplied to clients through a national network of accredited financial intermediaries. These financial intermediaries have to be registered as financial service providers in terms of the National Credit Act. They enter into an agreement with DAFF, usually for a three to five-year period, in terms of which they have access to a specified amount of the MAFISA capital, which is paid out to them when they require it. The Land Bank has been appointed to hold all MAFISA funds and disburse these funds to the financial intermediaries on instruction from the Director General of the DAFF. There was a laid-down and agreed selection process for selecting the financial intermediaries in line with the Public Finance Management Act. At the outset of the programme, there were nine financial intermediaries that aimed to provide an even spread of access across the provinces. This number has now fallen to five. Potential loan beneficiaries are informed of the existence of the loan facility by either the financial intermediaries or the provincial departments of agriculture. These bodies should support the potential beneficiaries to develop business plans and submit loan applications. There is no standard business plan or application process. Each of the financial intermediaries uses their own internal processes and procedures to evaluate and approve loan applications, as well as manage the disbursement of the funds and manage the process of repayment. 2.5 Loan size and interest rate The maximum secured loan a beneficiary can be granted is R per individual. It is proposed in the revised MAFISA policy that the maximum unsecured loan may be up to R per individual. In cases where the project requires more than the maximum loan amount, the financial intermediary may exceed these limits, provided the decision is based on the viability and profitability of the business. The interest rate is 8 per cent per annum, of which 7 per cent is retained by the financial intermediary and 1 per cent is returned to the MAFISA account at the Land Bank on expiry of the agreement period. 2.6 Data gathering on MAFISA s performance The policy framework envisages DAFF receiving monthly, quarterly and annual reports from the Land Bank and the financial intermediaries so as to enable maintenance of a comprehensive database on MAFISA s performance, monitor the performance of the individual financial intermediaries and analyse the success or otherwise of the scheme. 2.7 Competing products offered by government The following is a list of grants that the Department of Rural Development and Land Reform provide to smallholder farmers. These farmers form part of the target market for MAFISA loans. From our research it emerged that beneficiaries are of the view that these grants compete with MAFISA loans, and that

18 12 target beneficiaries choose to not take out MAFISA loans as they do not need to pay interest or repay any capital on these grants. i. Land Redistribution for Agricultural Development (LRAD) Grant. This grant allows black South African citizens to access land specifically for agricultural purposes. This grant can be accessed, on an individual basis, using a sliding scale from a minimum of R to a maximum of R , depending on the participants own contribution. The grant is used to cover expenses such as land acquisition, land improvements, agricultural infrastructure investments, capital assets, short-term agricultural inputs and lease options. ii. LRAD Planning Grant. This grant provides financial assistance for project planning to applicants applying for grant financing under the LRAD sub-programme. iii. Settlement/Land Acquisition Grant (SLAG). This grant is currently set at a maximum of R per qualifying household to be used for land acquisition, enhancement of tenure rights, investments in infrastructure, home improvements and farm capital investment. This grant is now subject to policy revision. iv. Settlement Planning Grant. This grant is to be used to enlist the services of planners and other professionals in assisting applicants for SLAG to prepare grant applications and post-transfer support. It is also subject to policy revision at the time of writing. v. Restitution Discretionary Grant. This grant is set at a maximum of R3 000 per restitution beneficiary household, where the original land is to be restored or where compensatory land is to be granted, through means of a negotiated settlement of the restitution claim. vi. Land Restitution Grants. These are intended for families or groups of families who successfully lodge land claims in the land claims court, and the land in question, or alternative land, is purchased by the state. vii. Grant for the Acquisition and Development of Land for Municipal Commonage. This grant is to enable primary municipalities to acquire land in order to extend or create a commonage, and provide infrastructure on the land to be acquired, or on existing commonage, for the use of qualifying persons. This grant does not compete with MAFISA. viii. Grant for the purpose of determining land development objectives. This grant provides for underresourced, poor or rural local authorities to undertake a strategic planning process to set land development objectives (under Section 28 of the Development Facilitation Act, Act No 67 of 1995). This grant also does not compete with MAFISA. 3. Analysis of the MAFISA management processes The MAFISA policy framework clearly describes the different role-players and their respective roles and responsibilities in managing and supporting the MAFISA scheme. By design, these roles and responsibilities are structured so as to (i) reduce the cost and risk associated with the MAFISA loans, and (ii) ensure the beneficiaries are empowered so that they can expand their businesses. 3.1 Role-players in the delivery of MAFISA The following figure shows the role-players involved in the delivery of MAFISA:

19 13 Figure 2: Relationships between the MAFISA role-players National Treasury Provincial Departments of Agriculture DAFF Development Finance Directorate Accreditation committee Land Bank MAFISA A/C Beneficiaries 3rd Party Input Suppliers Financial Intermediaries Commercial Banks (accounts with MAFISA funds not lent out) i. National Treasury manages the disbursement of funds allocated through the budget process, and is therefore responsible for transferring the MAFISA funds to DAFF. National Treasury is also responsible for exercising oversight of the Land Bank, since it is an entity that falls under the Minister of Finance like all other development finance institutions. This means National Treasury exercises oversight of the MAFISA funds in terms of its general finance oversight responsibility and through its oversight of the Land Bank. ii. The Development Finance Directorate in DAFF is responsible for the development and updating of the MAFISA policy, and managing its implementation. This involves the following activities: Convening the accreditation committee to select suitable financial intermediaries to manage the disbursement of the MAFISA loans. The policy envisages that financial intermediaries will be appointed to ensure reasonably equal access to MAFISA in all provinces. The last time this committee met, it consisted of representatives from DAFF, National Treasury, the Land Bank and SEFA (Small Enterprise Finance Agency). Entering into agreements with the selected financial intermediaries. Exercising oversight of their performance. Issuing instructions to the Land Bank to transfer MAFISA funds to financial intermediaries. Collecting, collating, checking and analysing data on the financial intermediaries management of MAFISA loans, as well as the benefits that beneficiaries derive from the loans. Providing induction and support to the provincial departments of agriculture regarding to their role in supporting MAFISA. iii. The provincial departments of agriculture provide extension support to farmers, some of whom are existing and/or potential MAFISA beneficiaries. Extension officers are supposed to inform potential beneficiaries of MAFISA and assist them in applying for MAFISA loans. They should also provide technical support to beneficiaries funded by MAFISA loans in the implementation of their business plans. There should be close working relationships between the provincial

20 14 departments of agriculture and the financial intermediaries operating within the province with regular reporting between them. (Note that MAFISA beneficiaries are a subset of the total group of beneficiaries supported by the provincial departments of agriculture.) iv. The Land Bank manages the MAFISA account. This involves investing the MAFISA funds that have not be disbursed to financial intermediaries, transferring funds to financial intermediaries on the instruction of DAFF, and receiving funds from the financial intermediaries at the end of their agreement periods. v. Financial intermediaries enter into agreements with DAFF and issue loans to beneficiaries and then collect the loan repayments according to the policy framework. They are required to submit reports on the performance of their MAFISA loans to DAFF. They take on the risk of lending. (Note that MAFISA is only one of the many products financial intermediaries offer to farmers). vi. Beneficiaries apply for the MAFISA loans, enter into loan agreements with financial intermediaries and are responsible for honouring the loan agreements. vii. Commercial banks: both the Land Bank and the financial intermediaries keep their MAFISA funds in money market accounts. viii. Third party suppliers: in terms of the MAFISA loan agreements, the financial intermediaries generally do not pay the loan funds to the beneficiary. Instead, the beneficiaries may purchase approved supplies or services from third party suppliers who the financial intermediary then pays from the beneficiaries MAFISA loans. In some case the financial intermediaries may designate the third party suppliers, but this is not required by the policy. For the purposes of the costing component of this project, all activities illustrated by solid lines in Figure 2 above are analysed and costed (see section 6). The following table summarises the roles and responsibilities of the different role-players in relation to the actual delivery of MAFISA: Table 1: Roles and responsibilities of the main MAFISA role-players DAFF Provincial Department s of Agriculture (PDAs) Roles and responsibilities Policy development Internal capacity development for MAFISA management Plan and cost implementation Budget (verb) to manage development finance directorate MAFISA fund capitalisation (with the Land Bank) Set criteria, advertise and select financial intermediaries Information dissemination Develop standard operating procedures for provincial departments Monitor performance and compliance of financial intermediaries Maintain and analyse beneficiary and loans database Reporting Manage relationship with Land Bank Establish and maintain standard operating procedures with financial intermediaries Disseminate information Oversee the implementation of the MAFISA scheme Support the implementation of the MAFISA scheme Land Bank F Manage and account for MAFISA fund Receive and execute payment instructions to financial intermediaries Receive repayments and interest from financial intermediaries Calculate and levy service charges to DAFF Strategy to roll out MAFISA loans (coupled with own products) i n a n c i

21 15 Beneficiarie s Roles and responsibilities Establish and update standard operating procedures with PDA Cost and budget to administer MAFISA Disseminate information Process MAFISA loan applications Finalise and sign loan agreements Manage and support procurement for beneficiary and pay suppliers Make beneficiary payments (up to 10%) Provide technical support to beneficiaries Manage loans and repayments Manage non-performing MAFISA loans Report to Land Bank, DAFF and PDAs Make MAFISA application Request technical assistance if required from PDA/FIs to complete forms Identify suppliers and get quotes Make loan repayments 3.2 Overview of the flow of MAFISA funds The following figure shows the flow of funds between the MAFISA role-players: Figure 3: Flow of funds between the MAFISA role-players Provincial Budget Equitable Share i Division of Revenue Provincial Departments of Agriculture (provide support but no funds) DAFF ii Voted funds Once off MAFISA allocation National Budget Beneficiaries Land Bank service fee Land Bank MAFISA A/C Conditional cash <10% iv Interest (8%) and capital payments Funds for loans (not annual) iii 1% interest and capital repayment Interest (mkt) 3 rd Party Input Suppliers Payment for inputs in business plan Financial Intermediaries Commercial banks Funds not on-lent held in MM A/C

22 16 The numbered items in the above figure highlight the following fund flows: i. Transfers through the Division of Revenue Act, the national budget and the provincial budget: a. The provincial equitable share is transferred to provinces to fund their budgets. The CASP grant is transferred to DAFF, which then transfers it to the provincial departments as a conditional grant. Provincial departments use funds from the provincial budgets and CASP to fund the farmer support programme that provides extension support to smallholder farmers. b. Provincial agricultural departments finance the farmer support directorate from own funds provided by the provincial budget. The officials who give extension support to smallholder farmers, including MAFISA beneficiaries, work from this directorate. The provincial departments do not manage any of the MAFISA loan funds. c. The MAFISA funds were transferred to DAFF in tranches from the national budget between 2005/06 and 2009/10. In addition, DAFF uses voted funds from the national budget to fund the operational costs of the Development Finance Directorate. ii. iii. iv. DAFF transfers funds earmarked for MAFISA to the MAFISA account in the Land Bank. DAFF also pays the Land Bank an administration fee for managing the MAFISA account. This fee is paid out of the MAFISA account 3. The Land Bank transfers MAFISA funds to the financial intermediaries on instruction from the DAFF. The financial intermediaries are required to pay 1 per cent annual interest on the MAFISA funds they lend out to beneficiaries. They are also required to hold the MAFISA funds in a money market account with a commercial bank, and the interest earned on these funds gets paid back to the MAFISA account at the Land Bank. 4 Capital repaid by beneficiaries is also held in the accounts the financial intermediaries have with commercial banks. At the end of their agreements with DAFF, the financial intermediaries are required to return all capital held to the MAFISA account. Financial intermediaries enter into loan agreements with beneficiaries underpinned by a business plan. The financial intermediaries can transfer up to 10 per cent of the loan value to the beneficiaries. The rest of the loan is transferred by means of inputs purchased on behalf of the beneficiaries from third part suppliers by the financial intermediaries. These inputs are identified by the beneficiary in their business plan. The beneficiaries have to repay the capital plus 8 per cent annual interest on the funds loaned to them. 3.3 Liability for bad debts The logic underpinning the MAFISA policy framework is that financial intermediaries must take the risk of their lending activities, and their financial reward for doing so is the difference between the rate at which beneficiaries borrow from them (8 per cent) and the rate they must pay for having access to MAFISA funds (1 per cent). The financial intermediaries are expected to exercise due diligence when entering into agreements with beneficiaries and so minimise the risk of bad debts. The policy framework provides that DAFF will only write off bad debts caused by natural disasters. If a financial intermediary is not repaid a loan for any other reason, it bears the cost of the bad debt. 3 See information received from the Land Bank on all transactions on the MAFISA account in Annexures B, C and D. 4 This is according to officials from DAFF. Information received from the Land Bank does not show any such interest payments being made into the MAFISA account. It may be that this interest only gets repaid to the MAFISA account when accounts are settled at the end of the financial intermediary s agreement period.

23 Key findings regarding the MAFISA management processes It is evident that there is significant divergence between the intentions of the MAFISA policy framework and how MAFISA is being managed in practice: 1. When MAFISA was introduced DAFF appointed nine financial intermediaries. Between them they provided services across the nine provinces. However, since then one of the financial intermediaries has withdrawn from the scheme and two intermediaries contracts have ended and not been renewed. Kaap-Agri, whose contract has ended more recently, does not know if or when a decision around extending their contract will be made. The other three intermediaries interviewed currently have active contracts. Due to the fact that NERPO has a national footprint and that Puelwana is active in five provinces (North West, KwaZulu-Natal, Free State, Northern Cape and Gauteng), all provinces are technically served by MAFISA. However, no financial intermediaries with current contracts have offices located in the Northern Cape, North West, Free State or the Western Cape. These provinces are therefore under-served by MAFISA despite a large presence of smallholder farmers in all four provinces 2. DAFF s allocation of contracts to financial intermediaries has in some instances restricted the access to MAFISA loans for smallholder farmers in particular sectors. For instance, the SASA manages MAFISA loans in KwaZulu-Natal, but only to smallholder sugar farmers. Other smallholder farmers can access MAFISA loans from NERPO or Puelwana, but it is doubtful if these financial intermediaries have a sufficiently large footprint to reach smallholder farmers in the province. Again, the result is that MAFISA does not appear to be adequately reaching the targeted beneficiaries. 3. The policy framework clearly states that there should be a close working relationship between the provincial department of agriculture and the financial intermediaries. The policy envisages that a formal working structure must be in place. In addition, the financial intermediaries operating in a province should provide monthly, quarterly and annual reports to the provincial departments of agriculture. These working relationships and associated reporting communications exist in some provinces, but not in others. Our interviews indicated that there are good working relationships in KwaZulu-Natal and the Western Cape, while in the Eastern Cape there is a need for improved coordination. 5 Note that the primary reasons for putting these working relationships in place is to ensure that extension officers assist potential beneficiaries to apply for loans from the financial intermediaries, and that there is co-operation in providing technical extension support to the beneficiaries so as to improve the likelihood of success (and so mitigate the risk of default). 4. How the financial intermediaries manage the MAFISA loans also impacts on the realisation of the policy objectives: a. In the Eastern Cape the majority of loans extended by the ECRDA are based on unsecured lending for inputs and small equipment to what appear to be subsistence producers. The default rate of 70 per cent suggests that very few of these producers are progressing. Instead, the loan is pushing them into a debt trap. b. In KwaZulu-Natal SASA finances sugar cane production on communal or smallholder land by exclusively using and paying contractors to execute all the production activities. The produce is then sent to the mill, which pays SASA for the produce. If there is a margin, this is then paid to the landowner. This approach of using professional contractors to do the farming reduces to the risk of default to virtually zero. However, the land is effectively being rented using MAFISA funds, and the objective of empowering smallholder farmers and growing their capacity to engage in the agricultural sector is being missed. 5 Note that the remaining provinces fell outside the scope of this project, so we are unable to comment on the nature of the relationships in those provinces.

24 18 c. Kaap-Agri, which services farmers in the Western Cape and Northern Cape, has generally extended larger loans to fewer beneficiaries. MAFISA loans are processed along with all other loans provided by Kaap-Agri. Applications are submitted to the credit controller, who then prepares an application for the credit committee. The credit committee processes MAFISA applications along with all other applications for loan finance. Kaap-Agri offer loans for a wide range of produce through MAFISA. However, members of the credit committee have the expertise to evaluate the feasibility of business plans and make informed decisions before approving a loan. This strategy has been successful in moving most of these beneficiaries forward, in that they are now able to access commercial loans for their farming operations. Unlike SASA and NERPO, Kaap-Agri is unable to control the distribution of the produce of beneficiaries and as a result there have been two instances where beneficiaries have defaulted on their loans. Another beneficiary came close to defaulting, but after careful management of the client and assistance from Kaap-Agri, the client appears to be willing to repay the loan. d. NERPO only lends MAFISA funds to smallholder livestock farmers. The process NERPO follows when approving applications is similar to that of Kaap-Agri, but the credit committee only evaluates business plans for livestock farming. As a red meat producer s organisation, NERPO is able to control the sale of the produce of beneficiaries. This gives NERPO some control over the cash flow of beneficiaries and therefore the repayment of loans. NERPO has achieved an 85 per cent repayment rate and has not yet attached the assets of any beneficiary. It is also worth noting that the Land Bank recommended that the Wholesale Finance Facility, a product similar to MAFISA, be piloted by NERPO. 5. The provisions governing bad debts have not been tested, largely because most of the agreements with financial intermediaries are still current. It is known that Kaap-Agri applied to DAFF in 2013 to have two loans written-off against the MAFISA capital, i.e. for the cost of the bad loans to be carried by DAFF and not Kaap-Agri. DAFF has not given a decision on this application yet. Whatever decision is made will have significant implications on the management of MAFISA going forward, and even on the financial viability of a financial intermediary such as the ECRDA, which is facing a 70 per cent default rate. 6. MAFISA was initiated in 2004/5 when prime interest rates in the commercial market were 11 per cent. The rates then rose to 15.5 per cent in 2008 before dropping significantly in to 9 per cent. An interest rate of 8 per cent when the programme was piloted was very attractive, but in the current financial climate MAFISA loans do not have the same appeal. In spite of these serious fluctuations in commercial prime rates, the interest rate attached to MAFISA has remained fixed at 8 per cent. This shows a lack of market agility. Any loan product needs to be able to respond to the prevailing interest environment in order to remain relevant. 4. The performance analysis The performance analysis builds on the roles and responsibilities of the respective main role-players identified in Table 1. One of the challenges of the EPR methodology is that the outputs of the analysis are best viewed in Excel. This is particularly true of the performance analysis. The detail of the performance analysis is presented in MAFISA Programme Elements. 4.1 Existing information on MAFISA performance measures The following official documents related to MAFISA were reviewed to identify performance measures and targets that have been put in place: i. MAFISA Credit Policy Framework describes the terms and conditions of loans and the criteria that must be used to evaluate whether or not beneficiaries can apply for MAFISA loans.

25 19 ii. Strategic Plan for Smallholder Support /15 mentions MAFISA as an important component of the department s strategy, but no targets related to MAFISA are provided. iii. The DAFF Strategic Plan for 2011/12 to 2014/15 sets out performance indicators for MAFISA as shown in the table below: Table 2 Performance indicators for MAFISA in DAFF Strategic Plan for 2011/12 to 2014/15 Service objective: Facilitate access to loans by farmers to improve production Key Service: Facilitate production loans and savings mobilisation Current standard Desired Standard 2008/ / / /12 Quantity A minimum of two Accredit a minimum of two Review MoUs based on Bring in member-based intermediaries per province more intermediaries per the performance of the financial institutions as accredited through which clients can access financial services. province. intermediaries in order to improve the effectiveness and efficiency of the intermediaries in all third-tier service providers (intermediaries at local level) to extend outreach. Quality Value money A minimum of two Accredit a minimum of intermediaries per two more intermediaries province accredited per province. through which clients access financial services. for Production loans received at lower interest rate compared to commercial banks. Production loans received at lower interest rate compared to commercial banks. provinces. Review MoUs based on the performance of the intermediaries in order to improve the effectiveness and efficiency of the intermediaries in all provinces. Production loans received at lower interest rate compared to commercial banks. Look into ways of reducing administrative costs. Bring in member-based financial institutions as third-tier service providers (intermediaries at local level) to extend outreach. Production loans received at lower interest rate compared to commercial banks. Reduce administrative costs. iv. Department of Agriculture, Forestry and Fisheries Strategic Plan for 2013/14 to 2017/18 describes MAFISA and its role in enabling access to finance and supporting the production of small farmers is mentioned, but no performance targets are listed. v. Department of Agriculture, Forestry and Fisheries Strategic Plan for 2012/13 to 2016/17 sets out a single performance measure related to MAFISA, namely: Assessment of DAFF programmes (LandCare; Ilima/ Letsema; MAFISA, etc.). The target is Approved framework for assessment of DAFF programmes (quarter one) and then Funded programmes and Assessment report with recommendations (quarter three). vi. The chapters in the ENE for the Department of Agriculture, and later DAFF, were also reviewed. In the ENE 2006, the performance targets list Rollout of MAFISA as an activity. The performance indicator is number of farmers awarded loans, and the target is farmers. In the ENE 2007 the action is: accelerate implementation of MAFISA products; the performance indicator is: number of farmers and rural entrepreneurs accessing MAFISA products; and the performance target is farmers and entrepreneurs. In the ENEs for 2009 through 2013 the same performance indicator is listed, namely the number of farmers awarded production loans under the MAFISA scheme. In 2009 the target per year of the MTEF ranges from to beneficiaries. In 2010 the range is to beneficiaries; 2011 is to 7 000; 2012 is to These targets are presented along with other targets related to the production of smallholder farmers. The impact of government support for small farmers could be inferred from those targets; however the link between MAFISA and them is not discussed.

26 20 The performance measures noted above only refer to the number of loans and the number of financial intermediaries appointed. There is no reference to the impact of the MAFISA policy. During the first round of meetings conducted as part of this EPR, the DAFF mentioned that the MAFISA policy framework was under review and that a draft would be made available when it was ready. By the time this EPR was concluded, that document had not yet been made available. Therefore we could not assess whether new performance indicators had been developed. 4.2 Information on MAFISA performance loans issued Various ENEs provide information on the number of MAFISA loans issued as follows: The 2009 ENE states that loans were issued between 2006/07 and 2008/09. The 2012 ENE indicates that a further loans were issued between 2009/10 and 2011/12. Thus, according to these ENE figures, MAFISA loans were issued between 2006/07 and 2011/12. The team contracted by the DPME to conduct the MAFISA impact assessment collected data directly from the financial intermediaries through a snap survey. Financial intermediaries were asked how many loans they had disbursed since the inception of MAFISA. Seven of the nine financial intermediaries provided the information shown in Table 3 below. The remaining two financial intermediaries, MEGA and Hlhanganani, refused to participate in the survey. Therefore their performance is estimated based on a comparison of the value of the MAFISA funds disbursed to them and the performance of similar intermediaries. Table 3: Responses to: How many loans were disbursed since the inception of MAFISA? Number of Financial Intermediaries that participated in the Snap Survey NERPO loans 915 ECRFC GEP 18 Kaap Agri 29 MGK 456 Peulwana 322 SASA 180 Total Financial Intermediaries that refused to participate in the Snap Survey MEGA - estimated based on comparison to MGK 450 Hlhanganani - estimated based on comparison to Peulwana 400 Estimated total MAFISA loans issued This suggests that an estimated MAFISA loans have been issued since the scheme s inception. This is almost six times less than the figure reported in the ENE documents. The difference points to problems in data capture and reporting, but more particularly to issues with the interpretation/analysis of MAFISA information. According to DAFF officials the difference is due to an error in reporting: the higher numbers in the ENE documents are based on the number of drawdowns against loans issued and not on the actual number of individual loans issued. 4.3 The chain of delivery and proposed indicators for MAFISA Using the information gathered from interviews and documents, the EPR process requires the development of an ideal set of programme level outputs, outcomes and impacts for MAFISA, along with the proposed indicators. These are presented in Table 4 below:

27 21 Table 4: Programme level outputs, outcomes and impacts for MAFISA Micro Agricultural Financial Institution of South Africa (MAFISA) DAFF - Development Finance Directorate Impact Indicator / Measure Data source Outcomes Indicator / Measure % contribution to GGP by smallholder farmers, small-scale forestry, artisinal fisheries and aquaculture No. of people employed by MAFISA beneficiaries Number of MAFISA Average % increase in beneficiaries marketing beneficiary enterprise produce or livestock turnover Per cent of MAFISA beneficiaries that graduate to accessing commercial finance products Stats SA/DAFF ii database MAFISA database MAFISA database MAFISA database MAFISA database Number of MAFISA beneficiaries Value of MAFISA loans Number of loan applications refused Spread of loans by value across different commodities Grow th in value of beneficiary productive assets measured against baseline Spread of loans by value across provinces Data source MAFISA database MAFISA database MAFISA database MAFISA database MAFISA database MAFISA database Outputs Indicator / Measure Economic grow th in the agricultural sector is supported through fostering increased capitalisation of small holders and businesses, leading to employment creation and long term income growth. i Loan finance is accessible and is used to grow smallholder farmers, small-scale forestry, artisinal fisheries and aquaculture operations iii A national network of financial intermediaries implementing loans and and providing capital to targeted producers iv Number of financial intermediaries Percentage of provinces w ith active MAFISA intermediaries Number of beneficiaries by province Number of appointed financial intermediaries w ho w ithdraw from the MAFISA product No. of loan defaulters as percentage of total no. of beneficiaries Data source MAFISA database MAFISA database MAFISA database MAFISA database MAFISA database Provincial Departments of Agriculture v Impact Intended MAFISA beneficiaries are supported to access provincial economy Outcomes Indicator / Measure Proportion of beneficiaries PDA assists w ith applications that receive MAFISA loans Number of MAFISA beneficiaries given extension support Value of MAFISA loans issued in the province Data source MAFISA database PDA database MAFISA database Outputs Indicator / Measure Number of MAFISA loan beneficiaries in the province No. of farmers in PDA settlement programmes accessing MAFISA loans Provincial databases include information about access to MAFISA loans Data source MAFISA database MAFISA database DAFF Assessment Outcomes Indicator / Measure Proportion of MAFISA beneficiaries selling produce through off take agreements Proportion of MAFISA beneficiaries able to access commercial credit after paying off MAFISA loan Number of jobs created by beneficiaries of MAFISA loans Data source FI Report FI Report FI Report Outputs Indicator / Measure Intended MAFISA beneficiaries that apply/qualify for MAFISA loans and receive them Provincial support for MAFISA Financial Intermediaries Intended MAFISA beneficiaries are able to finance production by accessing micro finance Financial intermediaries issuing and managing MAFISA loans Number of MAFISA loan beneficiaries per financial intermediary vi Value of MAFISA loans lent by financial intermediary Value of turnover enabled by access to MAFISA loans Repayment rates on MAFISA loans Data source FI Report FI Report FI Report FI Report

28 22 With regards to the above table, note the following: i. The proposed impact statement is based on the overall objectives set out in the MAFISA policy framework. The proposed performance indicators are based on information that can be collected (if it is not already being collected) through the processes already in place for monitoring MAFISA beneficiaries. ii. The references to the MAFISA database as the primary source of information for most of the indicators/measures envisages a complete overhaul of the existing database (which DAFF has commenced), and the establishment of strong data management and analytical capacity. iii. To date the focus of measuring the performance of MAFISA has been on the number of MAFISA loans issued. However, this does not give a complete picture of whether the loans are accessible or not. Therefore it is proposed that a range of other measures should be used, such as Spread of loans by value across different commodities and The spread of loans by value across provinces. However, an important test of accessibility is to monitor the number of loan applications that get turned down. Not all applications should be approved, but if a very large number of applications get turned down, it raises questions about the accessibility of the loan. This may point to the need to either improve the quality of support provided to applicants or reevaluate the eligibility criteria for the loan. iv. DAFF is not responsible for issuing MAFISA loans. However, it does need to ensure that there is a network of competent and accessible financial intermediaries in place that are issuing loans. This output statement and the associated performance measures therefore describe what DAFF is responsible for putting in place in order to ensure that the outcomes and impacts described above are realised. v. It is important to monitor the number of financial intermediaries that withdraw from the MAFISA scheme and to investigate their stated reasons for withdrawing. In this regard, it would be useful if an independent third party were to be commissioned by DAFF and shared with National Treasury to conduct an exit interview with each financial intermediary whose contract ends or who withdraws from the scheme. vi. The design of MAFISA presupposes that the provincial departments of agriculture will disseminate information about MAFISA to potential beneficiaries, support them in making applications and provide technical extension services to successful beneficiaries in co-operation with the financial intermediaries. These performance measures are intended to provide DAFF with a mechanism for monitoring and comparing the provincial departments performance in this regard. Table 5 on the following page shows the different role-players responsibilities with regards to the implementation of MAFISA and the associated intermediate outputs and related performance indicators.

29 23 Table 5: Intermediate outputs and performance indicators for MAFISA DAFF PDAs Land Bank Financial Intermediaries Roles and responsibilities Intermediate outputs Indicator measure Data source Policy development of micro-finance pillar of Policy developed and implemented through Policies revised in accordance w ith current Commissioned CASP the Land Bank and FIs. operating environment Assessment Internal capacity development for MAFISA Team w ith required capacity to implement No. of key positions in MAFISA unit of DAFF Management management MAFISA not filled reports Implementation planning and costing Costed implementation plan Implementation plan for MAFISA DAFF Budget (verb) to manage unit responsible Budget for internal management team in Budget allocation for MAFISA Chief DAFF for MAFISA place Directorate MAFISA fund capitalisation Capitalised fund at Land Bank Regular reporting on MAFISA fund status Land Bank Report Set criteria, advertise and select FIs FIs selected and agreements in place No. of contracts w ith FIs in place MIS Information dissemination Well-informed provincial departments of agriculture and public Grow th in applications for MAFISA Loans Consolidated reporting from FIs Standard operating procedures for provincial departments Standard operating procedures for provincial departments' support of MAFISA Compliance of provincial departments w ith MAFISA standard operating procedures Assessment by DAFF Monitor performance and compliance of FIs Compliant and effective FIs, Land Bank and beneficiaries No. of FIs submitting monthly reports on time Management reports Maintenance and analysis of beneficiary and loans database Compliant and effective FIs, Land Bank and beneficiaries MAFISA database updated monthly AG Report Reporting Compliant and effective FIs, Land Bank and Timely monthly reporting by DAFF to National National Treasury beneficiaries Treasury MI Manage relationship w ith Land Bank Effective relationship w ith Land Bank Percentage of transfers to FIs transferred Land Bank Report w ithin required timeframe Establish and maintain standard operating MAFISA is integrated into farmer support Number of MAFISA loan beneficiaries in the MAFISA database procedures w ith FIs strategies in provinces province Information dissemination Working relationships betw een provincial No. of farmers in provincial settlement MAFISA database departments and FIs programmes accessing MAFISA loans Oversee the implementation of the MAFISA scheme Up-to-date information about MAFISA performance at provincial level Provincial databases include information about access to MAFISA loans DAFF Assessment Support the implementation of the MAFISA Provincial departments actively participate in PDAs scheme implementing MAFISA roll-out Manage and account for MAFISA funds MAFISA transfer system in place Number and value of transfers made to FIs Land Bank reports Receive and execute payment instructions to FIs MAFISA account management system Value of repayments received from Financial Intermediaries Land Bank reports Receive repayments and interest from FIs MAFISA account management system Value of repayments received from FIs Land Bank reports Calculate and levy service charges to DAFF MAFISA account management system Management fee Land Bank reports Strategy to roll-out MAFISA loans (coupled w ith ow n products) All FIs have a MAFISA Management plan Percentage of FIs w ith MAFISA operational strategy DAFF Assessment Establish and update standard operating procedure w ith PDA Standard operating procedure in place Percentage of FIs w ith agreements or w orking relationships w ith PDAs DAFF Assessment Costing and budgeting to administer MAFISA Staff and budget in FIs to implement MAFISA Dedicated capacity at FIs to implement FI Report loan MAFISA Information dissemination Informed stakeholders Grow th in applications for MAFISA Loans FI Report through FI Process MAFISA loan applications Applications for MAFISA loans evaluated No. of Applications received FI Report Finalise and sign loan agreements MAFISA loans aw arded No. and percentage of loan applications that FI Report are successful Manage and support procurement for Facilitated flow of inputs and w orking Number and value of transactions effected FI Report beneficiary and pay suppliers capital per month Make beneficiary payments (up to 10%) No. of payments not made w ithin SOP FI Report timeframes. Technical support to beneficiaries Farming operations of beneficiaries implemented according to appropriate business plan Percentage of projects/farmers supported by MAFISA operating according to business plan FI Report Manage loans and repayments Well managed loan books and accounts Percentage of interest and loan repayments FI Report made in line w ith loan agreements Manage non-performing MAFISA loans Non-performing loans w ritten off Value of w rite off requests granted as a FI Report percentage of total loan books Reporting to Land Bank, DAFF and PDA Reporting systems Proportion of reports submitted on time DAFF Assessment Make MAFISA application Properly completed application forms No. of applications received FI Reports Request technical assistance if required Technical support provided to beneficiaries No. of technical support visits to FI Reports from PDA/FIs to complete forms beneficiaries. Identify suppliers and get quotes Flow of inputs and supplies Value of goods and services delivered to FI Reports beneficiary Make loan repayments Cash flow from sales Value of sales made by beneficiaries Beneficiary financial reports Note: FIs = financial intermediaries Beneficiaries

30 Key findings regarding the programme analysis The preceding programme analysis highlights the importance of doing proper planning before proceeding from a pilot phase to full-scale implementation. This ensures there are costed plans in place, the necessary capacity, standard operating procedures for the relevant role-players, systems to collect and collate data, and the capacity to monitor and analyse performance. Many of these aspects are absent in the MAFISA scheme. 1. The programme was launched in the pilot phase, and continued in the roll-out phase without adequate reporting and data capturing systems being embedded and standardised. The Land Bank and the financial intermediaries have been submitting data to DAFF, but the data has not been captured on an adequately structured database. There have also been problems with the standardisation of data. Now, eight years into the programme, DAFF is taking steps to introduce improved data gathering by financial intermediaries. However, it is proposed that this system should be reviewed by a data analyst specialist before being fully implemented. There are many examples of data inefficiency in the existing database, and it would be unfortunate if they were not properly addressed in the new one. The most oft-cited issue relates to the number of loan beneficiaries, where multiple loans to single beneficiaries, or direct payments to suppliers, have been captured as separate loan extensions. According to the DAFF officials, there is a reporting template, but they did not share this with the team and we are therefore unable to comment on it. 2. DAFF needs to develop the capacity to monitor and analyse the available data relating to MAFISA. From interviews conducted with officials at the provincial departments, the four financial intermediaries and the Land Bank, it is clear that a substantial amount of data on the MAFISA scheme is sent to DAFF on a monthly and quarterly basis. All of the participants commented that they do not know what DAFF does with the data. Reporting is a tedious process, and if the people doing the reporting do not understand how the data is used or even why they are reporting in the first place, they have little incentive to ensure the data is accurate. DAFF officials indicated that they use the information to compile monthly reports for the department and quarterly reports to National Treasury. They also indicated that there are report back meetings with the provinces and the financial intermediaries twice a year where the information is shared. 3. Indeed, both this project and the team conducting the impact assessment have not received any meaningful information or analysis of MAFISA s performance from DAFF. This is unusual, as we would expect DAFF to be publishing information on the performance of MAFISA on a regular basis given the importance of the programme to government, but also to fulfil basic transparency and accountability requirements. 4. The success of the MAFISA scheme depends on there being co-operation between the provincial departments of agriculture and the financial intermediaries. From interviews conducted, it would appear that only one province is providing any meaningful support to MAFISA. Officials from DAFF noted that MAFISA policy documents and the agreements with financial intermediaries do spell out the roles and responsibilities of provincial departments of agriculture, but they also noted that DAFF cannot enforce compliance with these roles and responsibilities because agriculture is a concurrent function. However, spelling out roles and responsibilities is not enough. DAFF ought to go further and put in place a standard operating procedure to guide provinces on how to support the scheme. 5. When DAFF embarked upon the roll-out of MAFISA, provincial departments of agriculture were asked to nominate a MAFISA champion who would take responsibility for promoting and coordinating the programme within the provincial department, as well as communicating with DAFF. Ideally provincial departments should have nominated the manager of the farmer support services directorate because this is where the extension officers responsible for implementing CASP and providing technical support to farmers are all located. Indeed DAFF should have insisted on the MAFISA champion being located in this directorate. The failure to do so means that in certain

31 25 provinces DAFF is not communicating with the managers best situated to take MAFISA forward. This lack of clear communication lines with the right directorate within provincial departments can partly explain the lack of implementation at the provincial level. 5. Expenditure analysis The analysis and mapping of expenditure in this EPR depended on obtaining information from various sources., which in certain instances was not made available. DAFF did not share any useful financial information regarding the management of the MAFISA funds or the performance of financial intermediaries. Certain of the financial intermediaries were also reluctant to share details of their MAFISA-related activities, claiming that the information on their loan books is confidential business information. However, other financial intermediaries shared their information freely. The Land Bank also shared relevant information on the management of the MAFISA accounts. We were able to get some financial information, based on the snap survey, from the team conducting the impact assessment. Furthermore, the team analysed the Vulindlela data to identify government spending that either supports the implementation of MAFISA or competes with the scheme. 5.1 Transfers to the MAFISA fund When government announced the establishment of the MAFISA fund in 2004, it indicated that R1 billion would be allocated to the fund. The funds were transferred, through the national budget, in tranches between 2005/06 and 2009/10. The process involved two sets of transactions: (i) transfers from the national revenue fund to DAFF and (ii) transfers from DAFF to either of the MAFISA accounts held with the Land Bank. Information for the first set of transfers was obtained from the actual audit information in the ENEs for various years. Information for the second set of transactions was provided by the Land Bank. The following table compares the information on transfers to the MAFISA fund from these two sources. Table 6: Transfers to the MAFISA fund R thousands 2005/ / / / / / / / /14 1. Information from ENE - MAFISA fund transfers to DAFF Highest amount promised in MTEF Actual transfers to MAFISA Cumulative transfers to MAFISA Information from Land Bank - MAFISA fund transfers to Land Bank Transfers to MAFISA Call Account Transfers to MAFISA Deposit Account Cumulative transfers to MAFISA Difference (1-2) The above table shows that there are some notable differences between what was transferred to DAFF and what DAFF transferred to the MAFISA accounts held by the Land Bank. On balance, DAFF appears to have transferred R56.6 million more to the MAFISA accounts held by Land Bank than it received from the national revenue fund. This requires some explanation. Table 7 and Table 8 below show that when the MAFISA Call and Deposit Accounts where closed in November 2013, more funds had been transferred out of these MAFISA accounts than Table 6 shows as having been transferred into the accounts. The differences are summarised below: MAFISA Call Account Transferred in R Transferred out R Difference -R MAFISA Deposit Account Transferred in R Transferred out R

32 26 Difference -R The information made available to us shows that R415.9 million more was transferred out of the two MAFISA accounts than was transferred into them. This indicates that the information in Table 6 above is incomplete. Two possible sets of payments may account for some of the differences: (i) interest earned on the balances in the accounts, and (ii) the return of capital (and interest) loaned to financial intermediaries. We are aware of at least three financial intermediaries that say they have returned funds to Land Bank, but these transactions do not appear in the data we analysed. However, the fact remains that the current information does not reconcile, and so further information is required. 5.2 Transfers from the MAFISA fund The following two tables show transfers made from the two MAFISA fund accounts held at the Land Bank. In line with the management agreement with DAFF, the Land Bank only transferred funds out of these accounts under instruction from DAFF. Table 7: Transfers out of the MAFISA Call Account Rands Total Transfers to accredited financial intermediaries ECRFC Gauteng Enterprise Propeller Hlanganani Farming Finance Kaap Agri Bedryf Ltd MEGA Nerpo Financial Services Peulwana SASA Sub-Total Transfers to Land Bank Administration Fees Land Bank: Administration Fees Other transfers to Land Bank Land Bank Land Bank for Rumibyte Land Bank: WFF Sub-Total Other transfers Agricultural Research Council AGRO Landbou Dienste ASGISA-EC Khula Credit Guarantee Nieuwoudtville Rooibos Rovic and Leers Temo Agri Services Ubongwa Cotton Dev. Services Sub-Total Account Closure National Revenue Fund (Closure) Total i. The transfers to the accredited financial intermediaries can mostly be cross-referenced against other information gathered by the team conducting the impact assessment. An important exception in this regard is the R206 million transferred to Peulwana Financial Services in 2012 and According to other information, the MAFISA agreement between DAFF and Peulwana allocates R20 million to Peulwana. However, in 2012 DAFF instructed the Land Bank to transfer R178 million to Peulwana in three separate transfers: R30 million on 08 May 2012, R50 million on 28 November 2012 and R98 million on 18 December Then DAFF instructed the Land Bank to make a further transfer of R28 million on 05 April According to the snap survey,

33 27 Peulwana reported that it has only lent out R35 million. This means Peulwana is holding R192.5 million. This is contrary to sound fiduciary management of these funds, and completely at odds with how DAFF has managed the disbursement of funds to all other financial intermediaries. There seems to be prima facie something unusual with these transfers that requires investigation. ii. The Land Bank has had two agreements with DAFF to manage the MAFISA account. The first agreement was signed in March Under this agreement the Land Bank charged an administration fee of 1 per cent of transfers into the MAFISA accounts. In March 2009 the Land Bank charged government an administration fee of R5.94 million under this first agreement. It would appear that this amount was paid by DAFF from departmental funds, as Land Bank does not report it being paid from either of the MAFISA accounts. The second agreement was signed on 11 December Under this agreement Land Bank charges government an administration fee calculated as a percentage of transfers. The rate was 0.5 per cent for the first year (December 2008 to December 2009), and this grew by 0.1 per cent each year thereafter. Therefore, from December 2012 to December 2013, the administration fee was 0.9 per cent. Under this second agreement, the Land Bank has charged government a total of R8.09 million. These charges were levied between 2010 and 2013 against the MAFISA Call Account. In December 2013 the Land Bank charged an administration fee of R to close both the MAFISA accounts. According to the data received from the Land Bank, it has managed the processing of 72 transfers out of the MAFISA accounts between March 2006 and December 2013, for which it has charged government R14.4 million. That works out to almost exactly R per transaction. Admittedly, there are other responsibilities involved in managing the accounts, but none of them require the Land Bank to take on any direct financial risk. iii. The transfer to the Land Bank labelled WFF refers to the Wholesale Financial Facility, which is a facility through which the Bank provides small loans within the agriculture sector. However, we are not aware of how this relates to MAFISA. We do not have any further information on the purpose or final destination of the other R77.3 million shown under the other transfers to Land Bank. Further information is required in this regard. iv. The table shows that DAFF instructed Land Bank to transfer R537 million to other entities. None of the MAFISA documents reviewed mention any these entities as being part of the MAFISA scheme, either as a financial intermediary or in any other role. Most of the entities do offer financial services, but according to information provided by DAFF none of them are contracted to provide MAFISA loans. Of particular concern is the R337 million transferred to the Agricultural Research Council. Apparently these funds are for a tractor mechanisation programme". Firstly, this is something that should not be funded from MAFISA funds and secondly, a programme of this nature should be managed and delivered by the provincial departments of agriculture and not the Agricultural Research Council. v. On 19 October 2013, National Treasury instructed the Land Bank to return all funds in this MAFISA account back to the national revenue fund. A balance of R44.7 million was returned.

34 28 Table 8: Transfers out of the MAFISA Deposit Account Rands Total Transfers related to Makhathini Land Bank (for the Liquidation of Makhathini Project) ESKOM-for the Re-installation of electricity supply at Makhathini Khawuleza Protection-for security services at Makhathini Salaries-Makhathini Cotton Staff Sub-total Other Transfers Deneys Reitz NERPO Sub-total Account closure National Revenue Fund (Closure) Total With reference to the above table: i. The funds transferred under the heading Transfers related to Makhathini relate to transfers made from the MAFISA deposit account to a private company called Makhathini Cotton (Pty) Ltd. According to CIPC, information this company went into provisional liquidation on 14 January A statement table in Parliament on 5 March 2013 (see Annexure E) indicates that the Land Bank took the company into liquidation. It also indicates that DAFF purchased the company from the liquidators in 2010 for R30 million with a view to establishing it as a co-operative to benefit small-scale cotton farmers on the Makhathini Flats, and subsequent expenditures have been made to secure the assets. The statement concludes that DAFF needs to explain to Parliament and South Africa why R36 million that was committed to an important project has yielded nothing. It is evident that at least R32.9 million of MAFISA funds have been allocated to this project. This raises a number of questions: Which persons or entities benefited when DAFF bought the company from the liquidators? Did DAFF pay a fair price for the liquidated assets? Why were MAFISA funds used for this purpose? ii. According the other information received from DAFF, the agreement with NERPO provides that it will receive an allocation of R50 million. However, DAFF has instructed the Land Bank to transfer a total of R104.8 million from the MAFISA accounts. The timing of the latest transfer of R50 million on 21 December 2012 also raises questions. iii. On 19 October 2013, National Treasury instructed the Land Bank to return all funds in this MAFISA account back to the national revenue fund. A balance of R91.6 million was returned. 5.3 Use of MAFISA funds by the financial intermediaries Table 9 on the following page shows the status of contracts with the financial intermediaries, MAFISA loans approved per intermediary, total disbursed to each intermediary and the balance held by each intermediary in December This is the most up-to-date information provided by DAFF. Alongside this is the information received from the Land Bank and information gathered by the snap survey.

35 29 Table 9: Summary disbursements from MAFISA Source of information DAFF (12/2011) Land Bank (11/2013) Snap Survey (10/2013) Intermediary Status of contract MAFISA Capital Loan Approved Total Transferred Intermediaries Balance Total Transferred Number of loans issued Value of loans issued Average size of loan issued NERPO Extended EFRCF Extended Kaap-Agri Closed SASA Running Peulwana Running MGK Closed no mention GEP Extended Hlanganani Running MEGA Running Total With reference to the above table: i. DAFF has instructed the Land Bank to exceed the agreed funds transfers to four financial intermediaries, namely NERPO, EFRCF, Peulwana and Hlanganani. In two of the cases the additional amounts are very substantial and not supported by the level of lending. ii. The R50 million transferred to MGK is not reflected in the information provided by the Land Bank. iii. As noted previously, Hlanganani and MEGA refused to participate in the snap survey, hence the gap in data. iv. The average value of loans made by different financial intermediaries varies significantly. Table 10 below summarises the data collected by the team conducting the impact assessment for DPME. This is the most up to-date data they could access. Data that supported an analysis of settled loans was provided by DAFF, and this is only available for four intermediaries. Some of the data in Table 9 is repeated to enable comparisons. Table 10: Summary of data related to loan performance NERPO ECRFC Kaap-Agri SASA MEGA MGK Loans issued Loans settled Value of loans issued R R R R Value of loans settled R R R R Loans settled/loans issued 4% 36% 59% 1% Average size of loan issued R R R R Average size of loan settled R R R R Age analysis of settled loans Age shown in months Average age Oldest Loan Settled Youngest Loan Settled Funds returned to MAFISA R R R The contracts with MGK and Kaap-Agri have been closed, and it appears these intermediaries have returned unused funds. The contract with MEGA is ongoing, but this intermediary has returned unutilised funds to the MAFISA account. It is assumed they are unable to lend these funds out to beneficiaries. That these intermediaries returned funds should not be seen as an indication of poor performance. It is more likely a sign that they followed due diligence when extending loans.

36 Other government expenditure that supports MAFISA Table 11 shows the expenditure by provincial departments of agriculture on the farmer support programme. Table 11: Provincial expenditure that supports MAFISA R thousands 2005/ / / / / / / / /14 PDA Farmer Support CASP (actual payments) This is the budget where provinces show expenditure related to technical extension support to farmers in the province. The assistance that provincial extension officers provide to farmers in preparing business plans and completing applications for MAFISA loans is one of the many activities funded from this budget. The aftercare support that extension officers provide is also funded through this programme. This budget programme is supplemented by the CASP grant. Therefore the top row PDA Farmer Support includes transfers to provinces through CASP. MAFISA is one of the pillars of CASP. In 2008/09 provinces spent R2.5 billion in total through this sub-programme. In 2009/10 they spent R2.8 billion on farmer support, which is nearly three times as much as the total value of MAFISA. One can argue that this funding competes with MAFISA. Many small farmers would rather receive inputs from government for free through the farmer support programme than repay the interest on loans they obtain from financial intermediaries. However, the two are not directly comparable, as farmers who are able to access MAFISA loans will probably receive the inputs more reliably and more quickly than they would if they were hand-outs from government. This is highlighted by research on cotton production trends on the Makhathini Flats: Only a few farmers are able to finance cotton inputs out of their own pockets and from other income sources and most farmers remain dependant on some kind of assistance. It can also be argued that these farmers who are able and willing to finance their own inputs are the real cotton farmers. In 2005/06 for instance the KwaZulu Natal Agricultural Department supplied input hampers to farmers on the Flats, which explains the jump in farmer numbers from 548 in 2004/05 to 2169 in 2005/06. In 2006/07 when no freebies were available farmer numbers again fell to The Vulindlela data for the above programmes was analysed and no references to MAFISA were found. This is not surprising for two main reasons. Firstly, it was established during our interviews that most of the provinces do not provide any support to MAFISA. Secondly, if an extension officer provides support to a farmer to access a MAFISA loan and/or provides support to a MAFISA beneficiary, then that support is provided as a routine part of the package of support provided by provinces. Furthermore, MAFISA is one of the many products extension officers will help farmers access. Therefore the process of providing support would be described as technical support rather than something such as MAFISA implementation. 5.5 Other government expenditure that appears to compete with MAFISA The expenditures frequently cited as competing with MAFISA are the land reform grants. Government has two funding streams related to land reform. These are shown in Table 12 on the following page. 6 Michel Fok, Jean-Luc Hofs, Marnus Gouse, Johann Kirsten, Contextual appraisal of GM cotton diffusion in South Africa, published in: Life Science International Journal, Vol. 1, No. 4, Pages

37 31 Table 12: Funds that compete with MAFISA R thousands 2006/ / / / / / / /14 Land Reform Grants Agricultural Land Holding Account DSD social relief transfers The Agricultural Land Holding Account is the account in which funds are held for payment of farms bought by government either through the land restitution or the land reform programme. In both these processes, there are usually unpredictable delays between the time land is identified for purchase and the time the payment for the land is transferred. The holding account enables the department to keep money set aside for land to be purchased and then make payment once the land is ready for transfer. Land reform grants are used to recapitalise farms purchased through either programme. This programme funds purchases of farming equipment that the beneficiaries require to ensure they can use the land productively. People interviewed during the research said that many MAFISA beneficiaries and/or potential MAFISA beneficiaries see other farmers receiving inputs for free through the land reform programme and so feel they should not have to repay their loans. The above information is anecdotal. MAFISA loans, like the farmer support provided by provincial departments, are intended to support the land reform programmes. Therefore if smallholder farmers see land reform grants as competing with MAFISA, it is evidence that the departments are handling perceptions badly. If, however, they are actually competing with each other, then the departments are handling their products badly. The Department of Social Development occasionally makes social relief transfers. As the name suggests, these are intended to provide relief and are not part of the department s ongoing social assistance grants. Beneficiaries of these relief transfers receive, amongst other things, food parcels and sometimes inputs to start home gardens. One can argue that a farmer would prefer to receive a food parcel for free rather than go through the effort of applying for and repaying a MAFISA loan. Such an argument is insensitive to the plight of people in need of social relief. 5.6 Key findings from the expenditure analysis 1. DAFF has not been forthcoming with information regarding the use of MAFISA funds. The most recent information received from DAFF dates from December Of concern is that recent information received from the Land Bank regarding the management of the MAFISA accounts indicates that there is reason to be seriously concerned as to how MAFISA funds are being used. 2. The funds paid into the two MAFISA accounts managed by the Land Bank do not reconcile with the amount allocated to DAFF for MAFISA through the budget (as reflected in the ENEs for various years). Also, there are more funds in these accounts than can be accounted for by the allocations to MAFISA. The information made available shows that R415.9 million more was transferred out of the two MAFISA accounts than was transferred into them. The information also does not reflect either interest payments or repayments of capital by financial intermediaries. The Land Bank should be instructed to provide a full reconciliation of all funds flowing into and out of the two MAFISA accounts. 3. DAFF has transferred R608.8 million from the MAFISA accounts held by the Land Bank to approved financial intermediaries for lending out as MAFISA loans, and a further R899 million to non-mafisa entities. 4. When Parliament allocated funds to MAFISA through the budget process, the allocations were made specific and exclusive for MAFISA. This means that neither DAFF nor National Treasury may vire or re-allocate MAFISA funds to other purposes without going back to Parliament for approval. Usually this would be done through an adjustments budget. However, we could not find any evidence that Parliament has approved alternative uses for the MAFISA funds through an adjustments budget.

38 32 5. The above implies that in all instances where DAFF has used MAFISA funds for non-mafisa purposes these expenditures are unauthorised in terms of the PFMA. Where DAFF has transferred funds to buy a liquidated company (Makhathini Cotton) and has allowed the asset to be vandalised, this should be regarded as fruitless and wasteful expenditure in terms of the PFMA. In line with the provisions of the PFMA, these expenditures need to be investigated and, where appropriate, steps taken to recover the funds, which may include recovering them from the official/s who authorised these expenditures, as well as laying charges of financial misconduct against them. 6. The most recent transfers to Peulwana are suspicious. The MAFISA agreement between DAFF and Peulwana only approves R20 million be transferred, but a total of R208 million has been transferred. This suggests that either this agreement has been altered or DAFF is acting without an agreement in place. Peulwana s website does not suggest that it is a very large operation. Therefore it is not clear why DAFF has decided to transfer such large amounts to it. 7. The latest transfer to NERPO of R50 million does not appear to be in accordance with the MAFISA agreement. The timing of the transfer on 21 December 2012 also raises questions. 8. The information received from the Land Bank suggests the following: a. DAFF has been treating the MAFISA funds (and other funds) as slush funds, allocating them to whatever (pet) projects they choose without any regard to the intended use of the funds as determined by Parliament through the budget process. b. DAFF may have been channelling departmental funds into the MAFISA accounts and then transferring them to other entities as a mechanism to avoid normal oversight processes. This may explain why more funds appear to have been transferred into the MAFISA accounts than the allocations indicated in the ENE. 9. Concerns around the management of the MAFISA funds resulted in National Treasury recalling the unutilised MAFISA funds on 19 October Acting on National Treasury s instruction of 19 October 2013, the Land Bank returned a total of R136.4 million to the national revenue fund. This effectively caps MAFISA s scope to the funds currently being managed by financial intermediaries. It is therefore probably advisable to close MAFISA. 10. When the MAFISA agreements with the financial intermediaries terminate, the financial intermediaries are required to repay the capital they borrowed plus 1 per cent interest to the MAFISA accounts at the Land Bank. It is not clear whether National Treasury s instruction of 19 October 2013 covers these funds as well. Given the concerns around DAFF s management of these funds, it is proposed that National Treasury should instruct the Land Bank to return these funds to the national revenue fund as well. 11. Annual spending by provinces on farmer support is nearly three times greater than the original R1 billion allocated to MAFISA. In light of this, these programmes contribution to supporting smallholder farmers is far greater than MAFISA, even though access to financing is an important supporting service. 12. There is anecdotal evidence that MAFISA beneficiaries feel it is unfair that they have to repay their loans plus interest while other small-scale farmers receive free starter-packs from the provincial departments of agriculture or through the land reform grants. This points to the need for better communication with beneficiaries regarding the MAFISA loans and the other grants, as well as the need for greater policy synchronisation.

39 33 6. Costing model The costing model is designed to allow for a high degree of flexibility in the specification of the different variables. This enables users of the model to explore different options with regards to these variables. The model allows users to test the cost implications of different modes of extension support by province and/or the financial intermediaries. Users can model different structures and risk profiles of the loan books to test whether it is financially viable for financial intermediaries to offer MAFISA loans. The model includes assumptions regarding the number of beneficiaries applying for MAFISA loans per province or financial intermediary, the amount of support beneficiaries need when applying for funds, and the level of aftercare provided. Changes to these variables in the provincial sheets changes the cost implication for provinces, and changing them in the financial intermediary sheets affects the expected profit from participating in the scheme. The management processes identified in Table 1 provide the basis for what is costed in the model. In many instances, policy processes were grouped together under a specific heading and so are not costed separately. The Service Description worksheet provides a breakdown of the policy processes costed on each worksheet. 6.1 Structure of the costing model The following figure sets out the structure of the costing model. Figure 4: Spreadsheets in the MAFISA costing model The model consists of seven main worksheets: 1. General Assumptions this sheet has assumptions related to the rates of inflation and salary growth; salary levels for all levels of staff used in the model; the number of days and hours worked per year; the cost of travel and accommodation. 2. Totals & Scenarios this is the key output sheet of the model, consolidating the costings from all other sheets. The user can test the effects of changing the interest rate at which financial intermediaries loan out MAFISA funds, the interest rate financial intermediaries must pay for use