Index-Based Weather Insurance in Rwanda

Transcription

1 In association with: Index-Based Weather Insurance in Rwanda Review of AFR s support of Kilimo Salama 03/06/2015 ii

2 VERSION 1.0 Author: Jeremy Leach Tyler Tappendorf Tel: Fax: Physical address: Regent House, Farm 2, 99 Jip De Jager Road, Bellville, Cape Town, 7530, South Africa Postal address: PO Box 5966, Tygervalley, i

3 Table of Contents Abbreviations... iv 1. Introduction Background on Access to Finance Rwanda Report one objective Index-based weather insurance context Kilimo Salama History of Kilimo Salama Expansion to Rwanda Summary of various insurance models in Rwanda Direct sales model Referral model Loan-linked model Portfolio-lite model Portfolio model Comparative cost of the models Organisational analysis of Kilimo Salama Management and staff Processes Funding Future plans for Kilimo Salama/ACRE Demand-Side perspectives Positive feedback Challenges cited Good scenario vs. bad scenario Key issues with IBWI in Rwanda Basis-risk issues Consumer protection issues Tax considerations Limited competition Conclusions on the impact of AFR funding for IBWI Is there a need for IBWI in Rwanda? Has AFR helped facilitate access to IBWI? Has AFR encouraged innovation and learning? Potential for scalability and sustainability Scalability/sustainability at product level Scalability/sustainability at organisation level Recommendations for AFR Conclusion Bibliography Appendix 1: List of Interviews Appendix 2: Calculation of Impact of VAT/WHT Appendix 3: Sustainability and Scalability of Kilimo Salama Products ii

4 List of Tables Table 1: Models of Index-Insurance Initiatives...7 Table 2: Market Opportunities for Agriculture Insurance in Rwanda (2012)...9 Table 3: Sample premium calculations for Kilimo Salama Rwanda products Table 4: Projections for for Kilimo Salama in Rwanda Table 5: Focus Group Summary Table 6: Ranking Distribution Approaches for Sustainability and Scalability List of Figures Figure 1: Summary of Kilimo Salama Model and Impact as of January Figure 2: Models of Agriculture Index Insurance in Rwanda Figure 3: Portfolio insurance theory of change Figure 4: FGD Quotes on Positive Feedback Figure 5: Challenges Cited in FGDs Figure 6: Locations of FGDs Figure 8: Smallholder agriculture lending by KCB in Rwanda Figure 9: Sustainability and scalability representation of various Kilimo Salama products iii

5 Abbreviations 1AF ACRE AFR BFA Cenfri FGD GIIF IBWI IFC IRI KCB MFI MINAGRI MINECOFIN RAB RRA SACCO SFSA UOB VAT WHT One Acre Fund Agriculture and Climate Risk Enterprise Access to Finance Rwanda Bankable Frontier Associates Centre for Financial Regulation and Inclusion Focus Group Discussions Global Index Insurance Facility Index-based weather insurance International Finance Corporation International Research Institute for Climate and Society Kenya Commercial Bank Microfinance institution Rwanda Ministry of Agriculture Ministry of Finance Rwanda Agricultural Board Rwanda Revenue Authority Savings and credit cooperatives Syngenta Foundation for Sustainable Agriculture Urwego Opportunity Bank Value Added Tax Withholding Tax (on reinsurance premiums) iv

6 1. Introduction 1.1. Background on Access to Finance Rwanda As part of the network of DFID-supported Financial Sector Deepening organisations across the African continent, Access to Finance Rwanda (AFR) plays an important role in increasing financial inclusion in Rwanda. Through supporting greater access to financial services such as savings, credit, insurance and payments, AFR seeks to improve the livelihoods of Rwandans. Within AFR s operations, one key objective is the support of the agriculture sector, specifically facilitating access to agriculture finance. Smallholder farmers in Rwanda are often times subject to weather shocks like drought and heavy rainfall that impact their farming outputs. This has made it challenging for farmers to maintain access to valuable credit that they need to purchase seeds and fertilizers, undermining their ability to continue farming and threatening their livelihoods. Banks, microfinance institutions (MFIs), and savings and credit cooperatives (SACCOs) have realised that without some sort of guarantee or insurance, agriculture lending is too risky. Therefore, index-based weather insurance (IBWI) has been viewed as an important and often compulsory tool to help protect both farmers and the financial institutions backing them. In early 2013, AFR signed an agreement with the Syngenta Foundation for Sustainable Agriculture (SFSA) in order to provide financial support to SFSA s Kilimo Salama project 1 as it expanded into Rwanda. Kilimo Salama Rwanda, building on its success in Kenya, sought to provide IBWI to cover crop inputs and livestock for smallholder farmers in Rwanda. The nearly $400,000 USD investment by AFR was intended to support the operational and staffing costs of Kilimo Salama Rwanda. In addition, AFR funded the acquisition and maintenance of automated weather stations (AWS) necessary for the IBWI product Report one objective In early 2014, AFR commissioned the Centre for Financial Regulation and Inclusion (Cenfri) and Bankable Frontier Associates (BFA) to address two distinct outputs linked to AFR s recent investment in agriculture microinsurance: 1. The first objective was to review AFR s investment in Kilimo Salama with the aim of assessing the impact on poor farmers the market and making recommendations around how to drive these models to scale. 2. The second objective of the project was to dive more deeply into a specific issue that was seen to be affecting the current agriculture insurance initiatives - value added taxes (VAT) and reinsurance withholding taxes (WHT). In this portion of the project, an assessment is made around the impact of these taxes on the market to inform recommendations for government. 1 In June 2014, SFSA s Kilimo Salama project became a standalone company called ACRE (Agriculture and Climate Risk Enterprise). More information about this change can be found in Section

7 In addition to the focus on agricultural insurance, a third outcome of the project was a brief on the current state of microinsurance in Rwanda as well as basic recommendations about how AFR can play a role in developing the insurance market moving forward 2. This report deals with the first objective of the three above, looking at AFR s current and future involvement in agriculture insurance initiatives. The report begins with a brief discussion on agriculture IBWI in Section 2, followed by an overview of Kilimo Salama and an organisational analysis in Section 3. In Section 4, feedback is presented on focus groups conducted with current and future clients of Kilimo Salama s insurance products providing insight from the ground level on how well the product is working. Section 5Error! Reference source not found. brings together key issues identified from both the supply- and demandside in regards to the current agriculture insurance products offered in Rwanda. The report concludes in Section 6 with a look at the impact of the funding provided by AFR as well as a few key recommendations for how AFR can play the most effective role in terms of promoting the resiliency of smallholder farmers moving forward, in Section Index-based weather insurance context Index-based weather insurance (IBWI) for agriculture is an innovative approach to solving the complicated issue of addressing the impact of weather on farmers and providing relief as quickly as possible. Typically linked to drought or excess rainfall 3, IBWI products use various triggering mechanisms like satellite imagery, weather station data, or area crop yields to serve as a proxy for weather impact in a geographic area. For example, a satellite can provide an estimate of the amount of rainfall that has fallen over a 100 km² (10x10Km) pixel with some models purporting to bring this down to 25km². The insurance policy would then stipulate that if more 4 than a certain amount of rainfall occurs in that area over the course of a certain time period, then a payout automatically flows from the insurance company to the beneficiary. As IFAD notes in their 2010 report on the subject, index initiatives occur at three levels: macro, meso, and micro. These models are further described in Table 1, and mapped to some of the current initiatives underway with IBWI in Rwanda. Although offering a unique solution to a challenging risk, the success of IBWI has been relatively limited to those products that provide macro-level cover (typically government level schemes) and those products which leverage large subsidies to cover the premium or infrastructure costs. IBWI initiatives targeting low-income groups or smallholders farmers have remained relatively stuck in pilot phase except for a few cases in India and Mexico, which have been able to reach scale (Hazell, et al., 2010) (Fuchs & Wolff, 2011). Those trials which have experienced moderate success have often done so as a result of subsidies. Nevertheless, some products have proven valuable to clients, when basis risk 5 issues can be mitigated (Sandmark, et al., 2013). 6 2 This report will remain internal with AFR. 3 Other index products are designed to cover the impacts of hurricanes, high winds, hail, and other quantifiable weather events. 4 In the case of excess rainfall cover. In the case of drought cover, the trigger would look at whether less than a certain amount of rain has fallen over a given time period. 5 Basis risk with index insurance is when there is a mismatch between the loss indicated by an index and the actual losses experience by the client. In some cases, client s actual losses may be higher than the insurance payout received from the index triggers. In other cases, the client might receive a higher value claim payout as compared to actual value lost. For more 6

8 Macro Meso Micro Channel Index Insurance for Disaster Relief Government Government protects itself against shocks: early liquidity/ first relief outlays Relief Agency Banks/MFIs/ Lending Institutions Financial Service Provider Farmer Association Input Supplier NGO Farmer Funds its operations through an index-based risktransfer contract or provides coverage through an index trigger contingent voucher Government purchases insurance on behalf of lending institutions to cover their portfolios Government could use banks, FSPs, input suppliers, farmers associations and NGOs to distribute vouchers for catastrophes Farmer receives explicit, redeemable, predictable coverage against a welldefined shock, and the premium is paid for mostly by government Table 1: Models of Index-Insurance Initiatives Source: Adapted from IFAD, Index Insurance for Development Government reinsures insurer through guarantees or similar instruments Lending institutions buy portfolio insurance to cover their own portfolio without directly informing the endlevel borrower. FSP buys portfolio insurance or group insurance to retail to farmers, linked to credit Farmers association buys group insurance to retail to farmers, linked to credit Input supplier buys group insurance to retail to farmers, linked to input purchases NGO buys group insurance to provide / retail onward to farmers Farmers buys insurance as part of a package(e.g. credit and other financial services, technology, agricultural information This category added to IFAD model by this project team Where weather index insurance initiatives in Rwanda currently operate information on basis risk see Sandmark, Debar, and Tatin-Jaleran s paper, The emergence and development of agriculture microinsurance, which serves as a useful further overview for a discussion on index-based agriculture insurance products. 7

9 3. Kilimo Salama 3.1. History of Kilimo Salama The Syngenta Foundation for Sustainable Agriculture (SFSA), based in Switzerland, initiated a project in 2009 called Kilimo Salama meaning safe farming in Swahili, which was tasked with exploring opportunities for agriculture insurance targeting smallholder farmers in Africa. The project first targeted Kenyan farmers, setting up headquarters for its operations in Nairobi. The project s first products were retail-based, targeting individual farmers, an approach which would eventually evolve into a distribution model that linked up with cooperatives, micro finance institutions and agriculture input providers in order to more easily reach clients in a cost effective manner. Products leverage technology like weather stations and satellites to calculate the impact of weather events on rural farmers. Products covering crops provide varying levels of benefits based on the point in the crop cycle when weather events occur. For example, farmers can purchase germination cover (protecting them from events during the first 21 days), drought cover, flowering, prolonged rainfall, storm cover, and area yield insurance. Maximum payouts range from 40% during germination stage to 100% during vegetative and flowering cover stage from days after planting. Kilimo Salama also began offering livestock insurance in Kenya, an indemnitybased product which links up with animal care initiatives to reduce risk. The same livestock product is intended to be extended in Rwanda. 99% of Kilimo Salama s customer base farms on 2-acre or less of land indicating the degree to which their products truly fit the bill as microinsurance products. Nevertheless, the product does also provide coverage to some larger farms, which account for roughly 50% of the premiums collected in Kenya. In Rwanda, as will be discussed below, almost all of the farmers covered are smallholder farmers. 7 Figure 1: Summary of Kilimo Salama Model and Impact as of January 2014 Source: (Kilimo Salama, 2014) 7 Kilimo Salama also operates in Tanzania, although not discussed in this report, and has plans to expand to into select southern and western African countries in the next three years. 8

10 3.2. Expansion to Rwanda In 2012, through the support of AFR and the Rwandan Ministry of Agriculture (MINAGRI) 8, SFSA undertook a study to explore the feasibility of crop and livestock insurance in Rwanda. The report identified a few key obstacles to agriculture insurance in the country namely a lack of structured ways to distribute crop insurance, crop value-chain risks that are more prevalent than climatic risk, and a lack of available data for the livestock sector. Despite these risks, the report ranked four crops as highly feasible to insure, as shown in Table 2 below. Sector Market potential Number of farmers Potential agriculture investment to be covered by insurance Maize High 75,000 2,940mil RWF Livestock High 28,877 23,786mil RWF Potatoes High 2,500 1,400mil RWF Tea High 9,562 1,719mil RWF Coffee Medium 19, mil RWF Rice Medium 6, mil RWF Beans Medium 67, mil RWF Wheat Low 2, mil RWF Cassava Low 104 6mil RWF Sorghum Low 50, mil RWF Table 2: Market Opportunities for Agriculture Insurance in Rwanda (2012) Source: (SFSA, 2012) Building off of previous relationships between the management of Kilimo Salama and MINAGRI, along with leveraging the Kenyan partnership between Kilimo Salama and agriculture extension provider One Acre Fund (1AF), Kilimo Salama was able to enter the Rwandan market relatively quickly, for the 2013A season which began in September Through AFR funding and by leveraging off 1AF s distribution model, a modicum of scale was achieved rapidly. Rather than operating as a project within SFSA, the Rwanda initiative was set up to build its skills as an insurance intermediary rather than just trying to capacitate insurance companies as was originally done in Kenya. In 2013, Kilimo Salama received approval from the insurance regulator in Rwanda 9 to operate as an insurance agent in the country, allowing them to generate revenue from their role as a market maker between insurance companies and client aggregators. 10 A number of partners play a role in the insurance product value chain in Rwanda, including: 8 MINAGRI plays a vital role in the Government of Rwanda s goal for the agricultural sector to professionalise itself, with 50% of the land cultivated using improved practises by MINAGRI also has been a key proponent of agriculture index-insurance with goals to see over 200,000 smallholder farmers covered by Beyond insurance, MINAGRI also focuses on key priorities like building access to agriculture extension services, collecting reliable weather data from meteorological services, and enabling regulatory frameworks that facilitate agriculture growth and protect farmers. 9 The insurance regulator sits within the Non-Bank Financial Institution department within the Central Bank. 10 It should also be noted that Kilimo Salama recently received in Insurance Surveyor license in Kenya, allowing it to collect revenue from its operations in that country as well. 9

11 Kilimo Salama, as the main insurance facilitator, which links the insurance company with the lending institutions and cooperatives who purchase the product. Kilimo Salama is also responsible for and designing the pricing and structure of the insurance products, before linking with the insurer and reinsurer. SORAS, as the local insurer. As one of the largest insurance providers in Rwanda, SORAS also provides some degree of brand strength to the product. They play a very limited role in terms of product design and pricing and only retain a small portion of the total risk 11. Swiss Re Corporate Solutions, as the reinsurer. Although a certain amount of local reinsurance is required to be held in Rwanda, this option has been waived leading to Swiss Re assuming upwards of 90% of the insurance premiums. Swiss Re, as one of the leading reinsurers playing in the index insurance space, also provides a degree of product design and pricing support. It is also noted that Swiss Re is the lead reinsurer for all projects supported by the IFC s Global Index Insurance Facility (GIIF) which also support Kilimo Salama, as noted below. Key partner institutions which help facilitate the distribution of the insurance product. These include lenders like Urwego Opportunity Bank (UOB) and Kenya Commercial Bank (KCB) and agricultural input providers like One Acre Fund (1AF) and Rwanda Agricultural Board (RAB). Key donor institutions which have helped to cover operational costs and premium subsidies. These donors include both AFR who covered the operational costs of setting up in Rwanda as well as the GIIF who provided the premium subsidy Summary of various insurance models in Rwanda One of the most challenging aspects in regard to understanding and analysing the operations of Kilimo Salama Rwanda is to distinguish between the varying operating models and partnerships that the organisation employs in the country. Although conversations with Kilimo Salama management indicate that the organisation utilises a portfolio or aggregated insurance model rather than a retail model, this description only partly reflects reality. It is true that in most cases the insurance product is not marketed directly to individual farmers; however there are instances in Rwanda where it is marketed directly to small farmer groups or cooperatives (such as Umbumwe Cooperative who the project team met with). These arrangements, in essence, function similar to a direct model with relatively high costs of distribution compared with using aggregators such as banks or agricultural input providers % of the premiums are reinsured with Swiss Re. 10

12 Based on conversations with Kilimo Salama Rwanda and MicroEnsure as well as a number of the various organisations who are purchasing or linked to the insurance in Rwanda, five distinct models were identified with four of these currently in operations by KSR as indicated in Figure Figure 2: Models of Agriculture Index Insurance in Rwanda Source: Author s own Direct sales model In the direct sales model, the insurance product is marketed directly to individual cooperatives. The product is not mandatory and in some cases, the cooperatives who purchase this insurance product are not even utilizing loans to finance their agriculture inputs. The money flows directly between the cooperative and insurer, both in the case of premiums and claims payouts. Kilimo Salama takes responsibility for training the cooperative about the product, a function which generally happens at the level of the cooperative president or executive committee. The direct sales approach is relatively limited in Kilimo Salama s overall operations reportedly representing approximately 5% of sales and is not their preferred approach. Conversations with Kilimo Salama indicate that many cooperatives which purchase the product directly have shifted from the loan-linked model discussed further on Box 1: The example of Umbumwe For the 2014A season, Umbumwe Cooperative insured their beans to the sum of RWF 210,000 for a net premium of RWF 63,231 or 30.1%. After VAT and WHT were added, the gross premium was RWF 74,612 or 35%. After the IFC s premium subsidy, the net premium paid by Umbumwe was RWF 44,769 or 20%. Source: Interview with Umbumwe 12 It is noted that there is a difference between KS stated approach to managing payouts and what we experienced in the field. KS state that they advise 1AF on which contracts triggered / which areas and the magnitude. 1AF then also uses its network of field staff to assess losses in the field. From there, payouts are determined. However, in our field work, 1AF stated that the payouts are divided equally across affected districts. 11

13 in this report. This shift certainly demonstrates a positive view of the product by the cooperatives.. Limited evidence indicates that if unsubsidised, the premiums paid by these groups can reach as high as 30% before taxes. Although it is understood that the experience of the cooperative the project team encountered which paid over 30%was most likely an outlier 13, the fact remains that there are clients who are paying extremely high premiums including taxes, unsubsidised premiums reach 35%. These premium amounts raise the important question of whether insurance is actually the best mechanism to protect these farmer groups. In the case of the Umbumwe Cooperative located in Kamonyi district with which the project team met, the premiums paid represent a less than 1-in-3 year loss scenario a ratio not often associated with insurance products. This group, specifically, was cited as having such high premiums due to delays in planting dates 14. The decision by this cooperative to move forward with the purchase of the product despite the high premium cost indicates either high client appreciation for the product, or a limited ability to discern whether the price is too high. Marketing, training, and distribution costs for this model remain much higher than the other models. Although these costs do not explain an unsubsidised net premium of 30% as Kilimo Salama do not currently price this into the product, the direct sales model does incur greater costs to operate than other models which will need to be factored in. In the other distribution models, clients are aggregated through either agricultural input or credit providers. These aggregators reduce marketing and training costs, whereas a direct sales model requires Kilimo Salama to spend time and staff capacity reaching fewer numbers of clients over a greater period of time Referral model The referral model, which through May 2014 had only been used by MicroEnsure through its coverage of loans clients from Bank of Kigali and Réseau Interdiocesain de Microfinance (RIM), involves farming groups being referred to the insurance companies by banks or MFIs who require that cooperatives receiving agricultural credit must have weather insurance. These lending institutions do not necessarily dictate where the clients purchase insurance 15 nor do they intermediate premiums or claims payments. Instead, clients seek out the insurance on their own simply with the need to show proof of coverage. This model comes with many of the same caveats as the direct sales model, although the insurance facilitator does not have to invest as heavily in selling the product to clients, as it becomes a product that s bought due to the mandate for coverage attached to the loan. Training responsibility still occurs direct to the cooperatives from MicroEnsure Loan-linked model The loan-linked model is when lending institutions both require insurance on agricultural credit and also facilitate the interaction with the insurer. Currently, there are a number of 13 Kilimo Salama indicates that the vast majority of contracts are less than 15%, with average premium rates across all distribution models around 6%. It is noted that this was not picked up in the research. 14 The impact of planting dates and related agricultural supply chain issues is discussed further in Section In early 2014, the two weather-index insurance providers were MicroEnsure and Kilimo Salama, although as of June 2014, MicroEnsure was discontinuing their products due to sustainability issues. 12

14 savings and credit cooperatives (SACCOs), as well as an MFI, Urwego Opportunity Bank (UOB), and Kenya Commercial Bank (KCB) which facilitate relationships between farmer cooperatives and insurers like Kilimo Salama and MicroEnsure. Premiums are deducted from the loan amount provided to the clients, leading to interest charges on the insurance premiums. Although the payment of premiums from the loan amount provided reduces cost and complexity of collection, the end-users ultimately end up paying the interest rate on the insurance premium which is capitalised into the loan, despite the fact that the premium is paid in one lump sum rather than over the course of the loan. This practice has been disallowed in some markets, including South Africa, due to potential unfair costs it adds to the end-user (LOA, SAIA, 2008). Claims payouts are funnelled directly into the cooperative account at the MFIs and SACCOs. The link to the loan accounts not only reduces the cost and complexity of premium payment, but it also simplifies the channel through which the claims are paid out. When a related weather event triggers an insurance payout, the money flows directly into the cooperative s account, which can then be used to reduce the outstanding balance of its loan Portfolio-lite model The predominant way in which IBWI has been able to flourish is when it is embedded directly in services or products offered to large aggregated groups of clients either by insuring an entire lending institution s portfolio or even providing cover at a government level. In the case of Kilimo Salama in Rwanda, the organisation s partnership with 1AF has allowed them to essentially tap directly into a base of over 50,000 clients. Whilst in theory, 1AF is encouraged to be silent to the beneficiary about the cover to prevent potential misunderstandings 16, in practice 1AF informs the farmer around the cover. This has led the project team to deem this approach portfolio-lite, rather than a full portfolio model as discussed in Section below. All 1AF clients who receive agricultural training and inputs for crops are required to pay a flat rate of 19% over the course of six months on their agricultural input loans. The insurance is embedded into this fee. The insurance contract is held between 1AF and the insurance company, rather than the individual farming groups. Nevertheless, the farming groups remain aware of the insurance product and ultimately receive a payout at the discretion of 1AF when the insurance triggers in their district regardless of whether they experienced actual losses or not. Training responsibility is assumed by 1AF, rather than by Kilimo Salama. 1AF uses an approach for its agricultural initiatives whereby village agents interact with various farming groups on a regular basis to accompany farmers through the input distribution, planting and harvesting processes. These agents not only coordinate the distribution of seeds and fertilizer, they also provide valuable training on farming practices. Rather than rely on Kilimo Salama s limited staff in Rwanda to provide the training on insurance, 1AF s village agents 16 In some cases where portfolio cover is disclosed to clients, there is a risk that clients will make a choice not to pay back loans following weather events, even if they have the financial means. Instead, they may rely on the insurance to pay out. In other cases, farmers who know their crops are insured may try to rely on this coverage even for non-weather related problems like poor farming practices or insects. Thus, some lending institutions who provide portfolio cover that is not silent to the beneficiary have seen loan delinquency actually increase. 13

15 assume all training responsibility around insurance. Although this reduces costs, it also introduces risk into the process. Conversations with 1AF management as well as observations in the focus groups indicate that certain agents understand the insurance product better than others, leading to uneven levels of education around the product. For example, one focus group respondent believed the cost of the insurance was 19% of the sum insured value whereas this covered the whole bundle of services as well as the loan. This also goes against KS s preferred approach which was to be silent to the beneficiary or not informing the farmers about the product, so as not to increase moral hazard or raise expectations which may not be met if there is localised basis risk. Premium costs can be significantly reduced through this method of distribution. By tapping into an existing base of clients through 1AF, Kilimo Salama is able to achieve scale with relatively less effort than the other distribution models. Rather than signing individual contracts with farming groups, master contracts can be signed with 1AF. Similarly, with claims payouts, all monies flow directly between the insurance company and 1AF. As mentioned in the previous section, training responsibility and cost also sits with 1AF. Finally, and perhaps most striking, is that 1AF facilitates the complete agriculture value chain for farmers meaning that seeds, fertilizers, and training are all coordinated to ensure the optimal planting times and practices, ultimately reducing the risks insured. These all combine to allow Kilimo Salama the lowest possible premium through the portfolio-lite approach Portfolio model The portfolio approach is not currently in practice by Kilimo Salama, but should be considered in the future for its ease of operation, lower costs, and reduction in basis risk issues. In a pure portfolio approach, an insurance relationship exists directly between the insurer and the lending institution, for example, to cover their portfolio. The end user is not informed of the product. Instead, banks and MFIs can simply use any claims payments to cover losses experienced on their portfolios. Training and distribution costs are nearly completely eliminated. The risk of clients not paying back their loan because of unmet expectations around an insurance payout is also reduced. As sophisticated organisations operating at scale, the impact of basis risk is more limited as it will likely be a minor part of any overall portfolio. This model appears to be the most sustainable in the long-term and to offer ex ante benefits to the value chain, as illustrated in Figure 3. 14

16 Figure 3: Portfolio insurance theory of change Source: (BFA, 2013) However, whilst this model would address the basis risk issues and cost of delivery, certain drawbacks do come with this approach including: 1. Although the clients assume the cost of the premiums, the benefits in many ways lie predominantly with the lending institution (Magnoni & Budzyna, 2013) (Magnoni, et al., 2012). 2. The end user does not become familiarised with the insurance product, a key secondary benefit that comes with more direct sales models, especially in countries where the insurance market is relatively underdeveloped. This awareness has been the reason for much of the positive impact that has been picked up in impact studies of index insurance (Hazell, et al., 2010). 3. Any benefit to the end-user can be significantly delayed as compared with more direct distribution models. For example, lending institutions will continue pushing for loan repayment until it is 100% clear that the cooperatives/groups cannot pay due to weather losses. 4. There is a risk there may be a limited number of willing buyers of portfolio cover. Some institutions may only cover the portfolio most exposed to weather events which will not be sustainable and would not support scale Comparative cost of the models In considering the above models, one needs to consider the typical costs that each approach incurs. Conversations with Kilimo Salama indicate that premium rates charged are not affected by the mode of selling insurance, but rather solely by the risks assessed that depend on historical data, planting dates, crops involved and geographic location. The organisation has yet to factor in the costs of the mode of distribution into the premiums. Thus one can assume that all the models have a similar starting point in terms of pure risk 15

17 premium as long as the models all cover similar crops and are spread across the same geographic areas. In reality, the project team witnessed a wide variance in premium prices during its incountry visit in May A visit to the Umbumwe cooperative to which the product is sold directly revealed they were paying a premium rate of over 30% before taxes, although this is cited by Kilimo Salama as an outlier due to an extremely late planting date. On the other hand, the project encountered groups paying 14%, 11%, and 9%. Kilimo Salama indicates that the vast majority of clients pay between 8-14%, with some rates as low as 3% which the project team did not see in practice. Although Kilimo Salama may not build distribution costs into the premiums charged, without a doubt certain models cost more to operate than others. The direct sales approach incurs much higher marketing and training costs as compared to a heavily aggregated, portfolio-lite approach as with One Acre Fund. It can therefore be assumed that there is an implicit cross subsidiy between the more efficient and less efficient models. All approaches incur 18% VAT, 15% WHT on reinsurance premiums paid to foreign reinsurers, and a 0.5% document fee increasing the overall premium cost. Those products linked to loans further incur interest charges on their premium amounts, although this doesn t affect the direct sales approach or the portfolio-lite, whereby premium costs are not deducted upfront from the value of the loan disbursed. Table 3 uses examples of cases encountered during the in-country visit, although it cannot be assumed that these examples are indicative of all clients served by Kilimo Salama. Loan amounts have been equalised to , for comparative purposes. Example 1: Direct Example 2: Loan-linked Example 3: Loan-linked Example 3: Portfolio-Lite Sum insured Net premium rate 30% 13% 11% 9% Net Premium ADD TAXES VAT Taxes Paid Reinsurance Tax (WHT) Document Fee Total Taxes Gross Premium Gross Premium rate 40% 17% 15% 12% Tax Impact on premium rates (% of sum insured) +10% +4% +4% +3% ADD INTEREST No loan Loan Loan Loan Interest charge on premium* Total Cost after Interest Total cost % 40% 19% 16% 12% 16

18 REMOVE SUBSIDY Premium after 40% subsidy % after subsidy 24% 11% 10% 7% Notes: Sum insured is just a sample figure to equate all four distribution models Net premium rates are examples based on sample policy documents obtained from Kilimo Salama, and would vary by the individual cooperative, planting date, geographic location, and crops insured. Assuming VAT of 18% on all premiums Assuming Reinsurance Tax of 15% on all reinsured premiums (Assuming 90% of insurance premiums are reinsured, with 10% retained by local insurance company) Assuming Document Fee of 0.5% Assuming 18% average APR on loans Assuming that the premiums are capitalised on the referral and loan-linked models. Table 3: Sample premium calculations for Kilimo Salama Rwanda products Source: Author s own with assumptions based on Kilimo Salama documents 3.4. Organisational analysis of Kilimo Salama Management and staff Kilimo Salama has built a strong foundation of capable management and staff on which to grow their products. Although the staff on the ground in Rwanda is limited to three people, most of the core functions operate from Kilimo Salama s headquarters in Nairobi including product development and pricing/actuarial, finance, IT and monitoring and evaluation teams comprised of 43 total staff as of May The split between Kenya and Rwanda allows a centralised staff to become more specialised, while providing the necessarily support to the local Rwandan staff who are more focused on operationalising the products. One key area of concern moving forward is the recent departure of the technical coordinator of Kilimo Salama. The former director played a vital role in developing the vision and operations of Kilimo Salama over a number of years. Although having built a capable team of staff as support, research indicated that a high degree of responsibility for the day-to-day operations fell on the technical coordinator. Thus, it will be key moving forward to ensure that Kilimo Salama finds a suitable replacement who can enter the company and continue the momentum which was built by the former leader Processes The organisation has a strong set of processes in place for product development and operations. Feasibility studies and pilot testing help guide new products into the market and allow for refinement prior to and after launch. Pricing of products is led by the Kilimo Salama team, but verified by Swiss Re Corporate Solutions, providing actuarial verification. The 17 Upon finalisation of this report, it was learned that Kilimo Salama/ACRE has hired Dr. CJ Jones to lead the organisation moving forward. 17

19 recent acquisition of an insurance agent licence in Rwanda and an insurance surveyor licence in Kenya will ensure that Kilimo Salama is able to both operate as regulated insurance bodies while also allowing it collect revenues and commissions. The monitoring and evaluation team sets clear metrics for its products which allows the company to track progress over time and ensure both financial and social objectives are achieved. Finally, the company has undergone external financial audits, giving a high degree of confidence in its representation of its current financial status Funding When it began in 2009, Kilimo Salama relied heavily on SFSA to fund its operations. Over its five years, however, its funding base has broadened to include a wide range of donors most notably the Global Index Insurance Facility (GIIF), a multi-donor trust fund finance by the European Union, Japan, and the Netherlands and implemented by the IFC and World Bank. Although the project has been bringing in revenue for a number of years, this revenue has not been enough to cover the cost of operations. In addition to receiving their insurance agent and surveyor licenses in Rwanda and Kenya, respectively, Kilimo Salama also recently became a standalone company in Kenya no longer operating solely as a project within SFSA. This new company known as ACRE (Agriculture and Climate Risk Enterprise Ltd) has Syngenta Foundation and three other impact investors as shareholders and members of the Board of Directors. The Rwandan organisation is now fully owned by Kenyan based ACRE Future plans for Kilimo Salama/ACRE Certainly one of the cornerstones of ACRE is their strong vision for the future and willingness to venture into products where other organisations have not. Their mission has strongly revolved around being innovative thought leaders in the field of agriculture microinsurance and this has been evident in the number and types of products they have offered both in Kenya and Rwanda over recent years and their innovation around channels. The below represent a few of the highlighted plans for the near future along with growth projections for Rwanda from in Table 4. Plans to expand from crop insurance to livestock insurance in Rwanda are currently being developed. In collaboration with 1AF, KSR has been working to design and launch a livestock insurance product to cover dairy cows. This product would be contingent on a partnership with an organisation like 1AF, which ensures the proper vaccination and training regimens for livestock. The dates of the launch of this product remain unclear currently due to challenges with reinsurance, and should be followed-up with closely by AFR as there are two milestones and associated funding for this product development and launch. Although the livestock product was identified as a high-potential product for Rwanda, KSR have faced some challenges implementing the product. Discussions were held with MINAGRI to include insurance vouchers in fertilizer bags, allowing individual farmers to activate insurance policies via mobile phones. This product would shift the distribution of agriculture insurance to the individual level and would allow insurance policies to be activated at the time of planting a key component to ensure accurate pricing and risk coverage. Kilimo Salama launched a similar product in Kenya in 2013 whereby 10,000 bags of fertilizer contained SMS vouchers for insurance activation. In Rwanda, subsidized fertilizer through MINAGRI currently reaches 200, ,000 farmers, indicating 18

20 the potential for a fertilizer-linked insurance product to reach scale quickly, although issues in terms of basis-risk, claims handling and political will could become more prominent and may prevent this initiative from being launched. It is noted that following the completion of the project that this initiative is unlikely to go forward Total Farmers Insured 155, , , ,391 Sum Insured (USD) 7,409,680 16,693,892 22,968,905 47,072,477 Premiums (USD) 468, ,570 1,293,613 2,718,281 Average Premium Rate 6% 6% 6% 6% Margins (USD) 98, , , ,279 Table 4: Projections for for Kilimo Salama in Rwanda Source: Kilimo Salama Sustainability of Kilimo Salama is still in question. Whilst Kilimo Salama has met some success in rolling out products across multiple channels, there is still a significant mismatch in terms of deemed revenue 18 vs operating costs which will create challenges as they transition to a standalone company unless they are able to drive operational efficiency. Whilst the new investors in ACRE will take some of this risk, the challenge for KS in reaching scale will remain the thin margins, the need to drive for efficiency on a complex high touch model, the continued challenge of basis risk which drives costs up in terms of the need for ground proofing and ex gratia payments and overall the fact that index insurance is not yet proven as a commercial model. Whilst these challenges could to a large extent be addressed through an enhanced portfolio-lite and portfolio models, we are still to see these models come to fruition. 4. Demand-Side perspectives In addition to supply-side interviews, which were conducted with a wide swath of people and organisations involved in the distribution of agriculture insurance in Rwanda, the project team also coordinated demand-side research in the form of focus groups with current and potential clients of Kilimo Salama. The focus groups were conducted by two Rwandan consultants, Marcienne Umubyeyi and Janine Ampulire, over the course of four days from May 2-6, Following a pilot test of the focus group questions and approach, six focus groups were conducted in three different provinces of Rwanda. Each focus group consisted of 7-10 participants. A total of 51 farmers participated with 29 men and 22 women. The groups were segmented into insurance clients who have received claims payouts, insurance clients who have not received claim payouts, and non-clients. A description of the focus group composition can be found in Table Whilst in the past they were not allowed to receive income due to their status as a project of SFSA, they did run virtual books which tracked the revenue they would earn as they transition to a regulated intermediary in their different countries. 19

21 FGD Province District Type Via... Number of Age # respondents Men Women Eldest Youngest 1 South Huye Clients no payout 1AF South Nyamagabe Clients payout 1AF East Kayonza Non clients None East Kirehe Clients - payout UOB East Ngoma Clients no payout SACCO West Nyabihu Non clients KCB Table 5: Focus Group Summary Source: Author s Own 4.1. Positive feedback Insurance, as a concept, is viewed fairly positively across all focus groups. Unlike many developing markets where either negative or no experiences with insurance has led to negative perceptions of insurance, in Rwanda, the feedback received from focus groups was that insurance is understood and widely appreciated. Numerous groups cited the Mutuelle de Sante, the national health insurance initiative targeting low-income groups in Rwanda, as having shaped their thinking around insurance. Clients understand the need to put money away in order to save for unforeseen events in the future. Similarly, the groups were asked to compare the weather insurance to an animal, in which case, the majority of groups cited a cow indicating that a cow helps provide for their families in many different ways. On radio, they tell us about global warming, so we are deciding to insure our crops Nyabihu District People love [insurance] because they see the progress you are making. -Nyamagabe District Insurance payouts can take about a week, but to us that s very quick -Kamonyi District Figure 4: FGD Quotes on Positive Feedback Source: Focus Group Discussions Premium levels are not prohibitive in terms of insurance uptake. One of the main concepts tested in the focus groups was whether the insurance premiums are viewed as too expensive, and thus prohibiting some farmers from purchasing the product. The feedback received was that the amount charged for the insurance is generally affordable for the farmers 19. One farmer noted that new clients will always push back and complain about the cost of the insurance until it ultimately comes time to make the purchase, in which case he is 19 This feedback contradicts the comments from many of the aggregators that complain to Kilimo Salama on behalf of clients that premiums are too high. 20

22 able to pay the premiums. For many of the farmers to which the insurance is linked to a loan, the ultimate objective is to get access to the capital which they will use to farm. Therefore, they will pay whatever the cost necessary for insurance so as to have access to financing for fertilizer and seeds. Similarly, with 1AF clients, the focus groups demonstrated the high value of 1AF support services to the clients and the willingness to pay whatever the insurance costs so as to retain access to the broader set of training, inputs, and assistance. Despite the relatively little negative feedback which was received in regards to premium costs, it can be deducted that should the cost of the premium decline, this would be of great benefit to the farmers. While one might not expect a reduction in premiums to lead to a large influx of new insurance clients, it could be assumed that farmers already insured might be able to cover a larger portion of their inputs or crops, leading to more appropriate levels of cover Challenges cited Basis risk issues stand out as the largest concern amongst focus group participants, who cite numerous mismatches between the insurance payouts and the losses on the ground. In nearly every group who purchased the insurance, examples were given whereby losses occurred to the farmers, but the insurance did not trigger a payout. While the issue of basis risk is a problem with most IBWI products which reach down to a consumer level, the issue appears to be even more pronounced in Rwanda due to the mountainous landscapes. With rolling hills throughout the majority of the country, Rwanda experiences a phenomenon known as microclimates, whereby weather can vary greatly in very small geographical areas. Because the satellite pixels leveraged for the IBWI are not more localised than 10km x 10km, many times the insurance will not trigger a payout even though a number of farms within that pixel might have experienced drought or excess rainfall. On the other hand, at times, farmers experience no weather losses, yet they still receive an insurance payout. Clients are very aware of these microclimate issues, with numerous focus groups citing the need for onthe-ground verification staff rather than just reliance on satellite models and weather stations. On the whole, clients desire a greater degree of education, training, and information about the insurance products. When asked to rank their biggest concern about the insurance initiative, 3 out of the 4 client groups ranked the sales process as the biggest concern 20. Although education for non-1af clients occurs directly between Kilimo Salama and cooperative leaders, there appears to be a gap in how that information flows back to the individual members within cooperatives and farming groups. Numerous individuals claimed that their cooperative leaders had not informed them of how the insurance product works, and some of the leaders indicated that they, too, did not really understand the product. In some groups there was hesitation amongst the leaders to filter down product information to the individuals because they didn t really feel comfortable with the product information themselves. Almost universally, those who participated in the FGDs were requesting more information and more training so as to be able to better determine for themselves whether the products they are purchasing are a good value for money. 20 The other options ranked by clients were premium costs, payouts, value-for-money, and the insurer. 21

23 Insurance is like a goat, because it doesn t produce much milk or fertilizer to help you along the way, but if you get really desperate you can sell it to help in an emergency Nyabihu District We had a meeting and they told us about the insurance which we didn t even understand then a law was passed from the district that anyone who isn t insured will not get fertilizers. Kirehe District If rain has fallen, let them visit individuals & see how it was distributed as some people may have made a loss, while other had good yields. Ngoma District Figure 5: Challenges Cited in FGDs Source: Focus Group Discussions Related to training, many clients did not know how to calculate either premiums or payouts. Many individuals interviewed were not able to say exactly how much their premiums cost, although this did improve when just considering the cooperative leaders. With 1AF clients, both FGDs conducted cited 19% as the insurance cost, despite the fact that this amount is the full cost of all 1AF services including interest on agricultural input loans. Although the insurance is one part of that 19%, no clients interviewed were able to separate that cost out, nor did they necessarily view the product as something separate from the complete bundle of 1AF services. Figure 6: Locations of FGDs Source: Map from Government of Rwanda, FGD locations added by author. While knowledge about the cost of the premium was limited, understanding about how to calculate insurance payouts was completely absent. Not a single client who participated in the FGDs admitted to understanding how the insurance company decides the amount that clients receive following a weather event. There were some instances where clients cited onthe-ground assessors would come to determine payouts, although this is not an approach 22

24 currently utilised in the Kilimo Salama model. More than likely, these on-the-ground assessors were either village agents (in the case of 1AF) or Kilimo Salama staff, which were assessing the situation without necessarily linking these visits to insurance payouts. Many clients cited that they must just wait for the computers and technology to determine what amount of money they would receive, although clients would prefer to be able to determine this amount themselves in order to increase confidence that the correct amount is being paid out 21. Whilst KSR states that they engage with the management of the cooperatives in this process, it is clear that there remained confusion both at the level of the management and the members of the cooperative Good scenario vs. bad scenario One Acre Fund clients appear the most satisfied with the insurance, whereas those who purchase through SACCOs and UOB appear the least satisfied. The clients who participated in the focus groups can be divided into three broad categories based on their appreciation for the insurance product. Those clients who were most satisfied were clients interviewed from 1AF. These clients viewed the insurance product as just one aspect of a larger, very important bundle of services provided by 1AF. Although they were not fully clear on the premium and claim payout amounts, these clients remained largely positive about the potential benefits of the insurance products. Clients cited both the willingness of 1AF to provide farmer with the knowledge necessary to farm more effectively as well as the flexibility to repay their input loans on terms which aligned with their harvests a luxury not usually afforded on loans from other institutions. Overall, it appeared the positive reviews of 1AF as an organisation spilled over onto its insurance initiatives. Those clients who purchase the insurance product directly, like the Umbumwe Cooperative (which participated in the pilot FGD) remain relatively optimistic about the insurance product. While these direct sales groups requested more information and training, they ultimately felt comfortable to continue purchasing the insurance on their own without any mandate requiring the product. However, they did indicate their desire to better understand the value of the product so as to make the most informed decision on the product moving forward one person did raise a concern about the 30% net premium, but was silenced by other members of the group. The final group of clients are those who are required to purchase the insurance product due to rules imposed by lending institutions like SACCOs and UOB. These clients had more mixed feedback about the IBWI product. In some cases, the potential value of insurance was cited but a lack of payouts made clients sceptical. In other cases, clients said that if they were not required to purchase the insurance, they would not do so in the future. These clients, although heavily reliant on the loans linked to the insurance product, do not necessarily see high value in the product. 5. Key issues with IBWI in Rwanda The following section highlights a number of the key challenges for the product based on a combination of both supply- and demand-side conversations. These issues ultimately feed into the conclusion and recommendations for AFR in Sections 7 and Kilimo Salama states they review the payout calculations with cooperative management during regular contract review, but not with individual cooperative members. 23

25 5.1. Basis-risk issues As is the case with numerous IBWI products, basis risk is perhaps the most pressing issue with agriculture IBWI in Rwanda. This problem manifests itself in a number of ways: The landscape of Rwanda makes it challenging for satellites to accurately capture rainfall data. As discussed in Section 4.2, Rwanda experiences many microclimates whereby the weather on one side of a mountain may be sharply different from the other side, leading to very different farming experiences by farmers in close proximity. While one would hope to narrow the size of the satellite pixel that is used to determine a trigger, in the case of Rwanda this approach appears less likely to ultimately solve the issue of basis risk. Although automatic weather stations (AWS) can provide perhaps more accurate local weather data in comparison to satellites, AWS are limited both by cost, challenges in maintenance and microclimates. Rwanda has not been able to use AWS for IBWI to the degree that Kenya has, where there are over 80 AWS operated by Kilimo Salama. In Rwanda, AWS data collection stopped completely in 1994 for years due to the genocide, making it challenging to build historical weather models for insurance. Nevertheless, Kilimo Salama does currently employ 34 weather stations in Rwanda with maintenance costs of $1,200 USD per year per AWS. Theft of these AWS, however, has been an issue in recent months. AWS are more accurate in indicating excess rainfall as compared to satellites, but nevertheless, they are limited by the microclimates of Rwanda in a similar way to satellites. Although not directly a basis-risk issue, Kilimo Salama believes that clients are currently under-purchasing to get a price which aligns with their limited purchasing power. When it comes time for claims payouts, however, client needs are not being met by the payout amounts. It is in this light that Kilimo Salama makes the argument that a reduction in taxes and thus a reduction in premium prices could lead to fewer people under-purchasing, which would drive better results from the insurance products Consumer protection issues Consumer protection issues vary by the method of distribution, but include the following: Product education does not currently filter down to the individual level, leaving many without an ability to determine whether the product they are purchasing is a good value for money. Pricing disclosure, especially in the case of 1AF, is not always present. For example, 1AF clients do not receive any specific insurance contract, having the premium costs instead built into their overall loan charge 22. Questions can be raised in some cases about whether the price of the insurance product is fair, with premiums rising above levels considered acceptable for insurance products, albeit this may be in limited instances. 22 Premiums also vary by pixel so if a farmer were charged the specific premium for their area instead of taking an average rate, then 1AF would have to charge 80+ different loan rates. 24

26 Interest charged on lump-sum premiums (through linked loans) increases the cost of the insurance. Clients do not understand how to calculate payouts, leaving little ability for them to ultimately determine whether the insurance product is functioning correctly and providing good value for money. Although termsheets are provided to cooperative management the calculation is complex and often hard even for management to understand Tax considerations Although the subject of taxes is discussed in greater depth in the second report in this series of three reports, it should be clearly noted that taxes are a large factor in increasing the cost of the insurance product to the end user. With 18% VAT and 15% WHT on reinsured premiums, the cost of the premium rises by 31.5% 23. Rwanda stands out amongst both its regional and overseas peers in regards to the amount of taxes placed on agriculture insurance premiums. Kenya, Uganda, and Tanzania all waive VAT and WHT taxes on agriculture, although all three tack on minimal fees between 0.5% - 7%. Premiums in Zambia and Malawi incur rates of 16% and 17% VAT, respectively. Mexico, India, Brazil, Nicaragua, Philippines, Turkey, and Uruguay are just a few of the many countries, which waive taxes on agriculture premiums all together. (Mahul & Stutley, 2010). Despite the relatively high taxes on agricultural insurance, the amount of income collected by the Rwanda Revenue Authority (RRA) from Kilimo Salama equated to $104,777 USD in This amount represents just 0.011% of the estimated $981 million USD revenue collected through taxes in the country (RRA, 2013). The tiny percentage impact this amount has on overall revenue pales in comparison to the much greater impact the taxes have on the smallholder farmers. Although focus group discussions did indicate that farmers do not feel the premium prices are restrictively high, the taxes still lead to approximately $1 USD additional premium for the average insurance client Limited competition Competition within the market is relatively limited on many different levels, as discussed below: Reinsurers: Currently, very few large international reinsurance companies play in the microinsurance space. In the case of the IBWI products in Rwanda, only Swiss Re Corporate Solutions has thus far played a role in reinsuring the risk. Although Swiss Re must be commended for its willingness to enter into this market and for its involvement in the development of index-based microinsurance solutions in a number of other markets, until greater competition enters Rwanda, reinsurance premiums are unlikely to decrease. Although mandatory local reinsurance cessions are required, these have been waived in the case of Kilimo Salama s products in Rwanda. Domestic reinsurers (like Africa Re) are not yet confident in their ability to 23 Assuming 90% of the premiums are reinsured by foreign reinsurers. 24 See Appendix 2 for calculation. 25

27 assume the risk of an IBWI product and there is not yet a business case for them to invest in this space 25. Insurers: Similar to the case of limited reinsurers, only one local insurance company has been willing to assume a portion of the risk of the Kilimo Salama project. SORAS, Rwanda s biggest general insurance underwriter, now purchased by Sanlam from South Africa, has been the main local underwriter both for Kilimo Salama and MicroEnsure s former products, but they have ultimately retained very little risk (10% in the case of Kilimo Salama). Interviews with SORAS cite limited capacity to develop and price IBWI products, despite SFSA investing heavily in training for SORAS over the last 2.5 years. Thus they rely heavily on Kilimo Salama and Swiss Re Corporate Solutions to assume most of the product development roles. Although competition from other local insurance companies is unlikely to bring down the price, increasing the capacity of local insurers and being able to shift some of the pricing and administrative functions to local insurers would allow for a greater sharing of costs. Nevertheless, it should be noted that as ACRE is now registered as an insurance agent in Rwanda, they are legally limited to working with one insurance company at a time. Insurance facilitators: At the time of the drafting of this report, Kilimo Salama is the only current facilitator of IBWI in Rwanda, with its previous competitor, MicroEnsure suspending operations in the country as of early 2014 and pulling out altogether in June Conversations with MicroEnsure indicated that they were unable to find a sustainable product model in Rwanda due to basis risk issues, and initially postponed new insurance contracts until they could find a suitable redesign (most likely involving some form of ground-truthing) before subsequently withdrawing. Having both MicroEnsure and Kilimo Salama in the market was beneficial to consumers, as in many cases like with the UOB-linked loans, clients were presented with competitive options from both organisations. Similar insurance faciltators in other markets like Planet Guarantee have also cited challenges with IBWI, indicating more widespread issues with provision of this type of insurance. 25 Africa Re does assume some of the risk with Kilimo Salama s products in Kenya, as is legally stipulated. 26

28 6. Conclusions on the impact of AFR funding for IBWI In order to examine the impact of AFR funding given to Kilimo Salama, one can apply a fivepart framework as illustrated below. The first four questions are answered in Section 6 while recommendations are covered in Section 7. Figure 7: Framework for evaluating impact of AFR funding on IBWI in Rwanda Source: Author s Own 6.1. Is there a need for IBWI in Rwanda? Conversations with individual farmers certainly indicate that weather risks play a major factor in determining farming outputs and ultimately livelihoods of millions of Rwandans. With nearly 80% of the population involved in agricultural activities, the stability of the farmers driving this sector is key to overall economic stability within the country. Similarly, interactions with many of the lending institutions providing vital financing for smallholder farmers indicate that weather risks also pose a major threat to their ability to distribute agricultural credit. From large commercial banks to smaller SACCOs, these institutions are requiring some form of assurance that the funds they lend to smallholder farmers will not be lost to floods or MicroEnsure s withdrawal of its index insurance product meant KCB cancelled the disbursement of loans of RWF282 million out of RWF 320 million. droughts in the country. Most pointedly, a conversation with KCB indicated just how significant that institution s reliance on agriculture insurance has become. KCB partnered with MicroEnsure to provide cover to thousands of its clients. Unfortunately during the recent farming season, MicroEnsure cancelled its product offering mid-season due to issues with satellite data and basis-risk. This stoppage led KCB to make the decision that it could no longer comfortably issue loans to smallholder farmers without some sort of insurance backing, having only distributed 38 million RWF ($56,000 USD) of loans before the insurance stopped. With original plans to distribute a total of 320 million RWF ($472,000 USD) for that season, hundreds of thousands of dollars of agricultural financing went undisbursed and thousands of farmers were left struggling to determine how they would finance their seeds and fertilizers, as shown in Figure 7. 27