AMUCSS, Cáceres. PAPER No. 35 MICROREINSURANCE APPLICATIONS FILLING SUPPLY AND DEMAND GAPS. Alex Bernhardt, Guy Carpenter

Transcription

1 AMUCSS, Cáceres PAPER No. 35 MICROREINSURANCE APPLICATIONS FILLING SUPPLY AND DEMAND GAPS Alex Bernhardt, Gy Carpenter Jly 2014

2 ACKNOWLEDGEMENTS The athor is heavily indebted to Peter Wrede for his collaboration and inpts on earlier drafts of this paper. He wold also like to thank Denis Garand and Pranav Prashad for their editorial inpts and each of the organizational representatives who contribted to case stdies featred in this report. Copyright International Labor Organization 2014 First pblished 2014 Pblications of the International Labor Office enjoy copyright nder Protocol 2 of the Universal Copyright Convention. Nevertheless, short excerpts from them may be reprodced withot athorization, on condition that the sorce is indicated. For rights of reprodction or translation, application shold be made to ILO Pblications (Rights and Permissions), International Labor Office, CH-1211 Geneva 22, Switzerland, or by pbdroit@ilo.org. The International Labor Office welcomes sch applications. Libraries, instittions and other sers registered with reprodction rights organizations may make copies in accordance with the licences issed to them for this prpose. Visit to find the reprodction rights organization in yor contry. ILO Cataloging in Pblication Data Microreinsrance applications: Filling spply and demand gaps Bernhardt, A.; Carpenter, G. and Company, LLC International Labor Office - Geneva: ILO, p. (paper no.35) ISBN: (web pdf) International Labor Office Key Words - microinsrance / reinsrance / microreinsrance ILO Cataloging in Pblication Data The designations employed in ILO pblications, which are in conformity with United Nations practice, and the presentation of material therein do not imply the expression of any opinion whatsoever on the part of the International Labor Office concerning the legal stats of any contry, area or territory or of its athorities, or concerning the delimitation of its frontiers. The responsibility for opinions expressed in signed articles, stdies and other contribtions rests solely with their athors, and pblication does not constitte an endorsement by the International Labor Office of the opinions expressed in them. Reference to names of firms and commercial prodcts and processes does not imply their endorsement by the International Labor Office, and any failre to mention a particlar firm, commercial prodct or process is not a sign of disapproval. ILO pblications and electronic prodcts can be obtained throgh major booksellers or ILO local offices in many contries, or direct from ILO Pblications, International Labor Office, CH-1211 Geneva 22, Switzerland. Cataloges or lists of new pblications are available free of charge from the above address, or by pbvente@ilo.org Visit or website: 2

3 contents Acknowledgements 2 Glossary and acronyms Introdction 6 2 The microreinsrance vale chain (Re)insrers Intermediaries Donors Delivery channels 10 3 Forms of MicroreinsranceFE Excess of loss Per occrrence Aggregate Pro rata Qota share Index-based soltions Basis risk Triggers 21 4 Core fnctions of microreinsrance Capacity Srpls relief Reslts stabilization Catastrophe protection Market entrance or exit Expertise 28 5 Microreinsrance spply and demand Spply Demand 32 6 Conclsion The ftre of microreinsrance 35 7 References 37 8 Appendix 38 3

4 LIST OF TABLES Table 1. Top ten global reinsrers by gross written premim, life and non-life (US$) Table 2. A basic balance sheet acconting scenario analysis demonstrating the ability of QS reinsrance to provide cedants with srpls relief (US$) Table A1. Reinsrance organizations active in microreinsrance LIST OF FIGURES Figre 1. The microreinsrance vale chain comprises several main actors, each flfilling niqe fnctions Figre 2. Basic classes and forms of reinsrance. Boxes otlined in red are most relevant to microinsrance. Figre 3. Fnctionality of excess of loss reinsrance in six distinct claim scenarios. Note that reinsrance coverage only kicks in above a certain claim level. Figre 4. Fnctionality of 40% qota share reinsrance in six distinct claim scenarios. Note that the loss amont in each scenario is shared proportionally 60/40% between the insrer and the reinsrer. Figre 5. Fnctionality of index-based reinsrance in six distinct claim scenarios. Figre 6. Types of basis risk associated with typical catastrophe index insrance contracts Figre 7. MiCRO s basic corporate strctre Figre 8. A stylized view of the relative complexity/development cost and basis risk associated with certain index-based trigger types verss indemnity contracts Figre 9. In recent years reinsrers have paid a high proportion of insred losses arising from major catastrophic events worldwide Figre 10. Assessing commercial spply of and demand for microreinsrance Figre 11. Reinsrance cession rates for select regions worldwide sing non-life market statistics LIST OF BOXES Box 1. ICMIF LARG Box 2. Aga Khan Agency for Microfinance, Pakistan Box 3. Orchard Insrance Grop: A QS microreinsrance case stdy Box 4. Microinsrance Catastrophe Risk Organisation, SCC (MiCRO): An index-based property microreinsrance case stdy Box 5. Mnich Re CLIMBS initiative: An index-based property microreinsrance case stdy Box 6. Total global microreinsrance market potential: A demand perspective 4

5 GLOSSARY AND ACRONYMS Attachment The point at which an excess of loss reinsrance coverage begins; the top end of the XOL coverage retention Basis risk The risk of mismatch between the payot of a non-indemnity insrance contract and actal losses incrred by the contract holder Cedant The prchaser of a reinsrance agreement CSR Corporate social responsibility Facltative Reinsrance for individal insrance policies as opposed to portfolios of policies (see Treaty) GIZ Detsche Gesellschaft für internationale Zsammenarbeit (the German government international development agency) GWP Gross written premim ILS Insrance-linked secrities LPT Loss portfolio transfer LR Loss ratio Microinsrance Insrance which is oriented toward low-income poplations or grops otherwise exclded from the formal financial system Microreinsrance The provision of reinsrance prodcts, services and expertise to spport microinsrance programmes MFI Microfinance instittion PPP Pblic private partnership Pro rata A reinsrance agreement involving the proportional sharing of risk from the first dollar of loss QS Qota share; the main type of pro rata reinsrance Retention The portion of a reinsrance agreement retained by the cedant Retro Abbreviation for retrocession or reinsrance for reinsrers RSBY Rashtriya Swasthya Bima Yojana (a cashless national health insrance scheme for poor people rn by the Government of India in partnership with private insrers) Social insrance Insrance which sally caters to low-income poplations and is spported significantly by sovereign or sb-sovereign pblic entities throgh premim/loss sbsidies and/or mandatory prchase reqirements; a prominent example wold be the RSBY programme in India Treaty Reinsrance of portfolios of insrance policies XOL Excess of loss 5

6 1 INTRODUCTION Over the past 200 years the amont and scope of risk insred worldwide has grown rapidly amid widespread economic development. This growth has felled demand among insrers for a secondary reinsrance market 1 where catastrophic exposre and capital constraints can be readily alleviated and large risks or risk accmlations 2 can be transferred and better (that is, more efficiently) managed. While some localized or discrete market segments have a tradition of co-insring among insrers to address sch capital management isses, most of the secondary market for primary insrance risk today is handled by professional reinsrers and, recently, a growing variety of alternative secondary risk transfer providers spported by the capital markets have become involved (for example dedicated insrance-linked secrities (ILS)). 3 Table 1 shows the top ten global traditional reinsrers ranked by 2012 gross written premim (GPW), all lines. Table 1: Top ten global reinsrers by gross written premim, life and non-life (US$) Reinsrance premims written Life & non-life Non-life only 2013 ranking Company Gross Net Gross Net Total shareholders fnds Mnich Reinsrance Co. a Swiss Reinsrance Co. Ltd Hannover Reckversicherng AG a Lloyd s b,c Berkshire Hathaway Inc. d Reinsrance Grop of America Inc. China Reinsrance (Grop) Corp. Korean Reinsrance Co. e PartnerRe Ltd Total top 10 Next 40 Total top For a good review of the historical origins of reinsrance, see Swiss Re (2010, p. 10). 2 In this docment, the term risk will be sed in two ways: (1) to describe the probability of something going wrong (as in the risk of an accident happening or the risk of an earthqake ); and (2) to describe a nit of risk transfer, i.e. something that is insred (as in pooling a large nmber of comparable risks, which means providing insrance to many people, bildings, etc.). This dal se of the word is cstomary. 3 The recent and ongoing convergence between the traditional reinsrance markets and the capital markets via ILS is a hot topic in the market today thogh addressing this key trend is beyond the scope of this analysis. A short definition of ILS, however, may be sefl for some readers: In general, an ILS is an investment whose nderlying performance and risk of loss is tied to an insrance risk, often a catastrophic one. An ILS can originate in the form of a bond, a reinsrance contract, an eqity investment in a special prpose vehicle, or a derivative instrment, sch as an option. The ILS market is made p of catastrophe bonds, collateralized reinsrance, sidecar vehicles, and ILWs (indstry loss warranties) (Conning, 2013). a Net premims written data not reported, net premims earned sbstitted. b Lloyd s premims are reinsrance only. GPW for certain grops within the rankings also may inclde Lloyd s Syndicate GPW when applicable. c Total shareholders fnds incldes Lloyd s members assets and Lloyd s central reserves. d Shareholders fnds incldes non-reinsrance sbsidiaries of Berkshire Hathaway. e Fiscal year-end 31 March Sorce: Best (2013). 6

7 Professional reinsrers have niqe expertise in the analysis and management of large or complex risks and in the layering and parcelling of complicated risk profiles. They are also well positioned to spport risk pooling on a global scale: since reinsrers transact with other risk indstry professionals, sch as insrers, reinsrance brokers and other reinsrers, they are generally sbject to different (read less stringent) (inter) national reglations than insrers. This facilitates geographical diversification, which in trn allows reinsrers to better harness the law of large nmbers for certain big or complex risks to manage catastrophic or covariate risk. 4 Nevertheless, while reinsrance negotiations and arrangements are a cstomary component of conventional insrance vale chains, this is generally not tre in microinsrance. This paper contends that, as in traditional insrance markets, where secondary risk transfer is essential to the operational solvency of primary insrers both large and small, microreinsrance is, and will contine to be, important to the long-term growth and sstainability of the global microinsrance market. However, microreinsrance is not the same as traditional reinsrance; in many instances it is, and shold be, sed differently from traditional reinsrance (or not sed at all), in order to accommodate the niqe aspects of microinsrance programmes. This is often left nacknowledged microreinsrance is a specialist field and shold be treated as sch. Moreover, at this still nascent stage in the development of a global microinsrance market, there is a gap between the spply of and demand for microreinsrance. This gap is most prononced in jrisdictions where insrance programmes lack significant governmental spport 5 and is in part enabled by a break in knowledge between practitioners on either side of potential transactions. We hope this paper helps to close that gap. Some of the other main reasons for this market imperfection inclde, bt are not limited to, the following: many microinsrance portfolios are written by commercial insrers and are still too small to threaten their bottom lines if things go wrong; simple and predictable credit life bsiness, where the need for reinsrance is limited, still makes p the lion s share of microinsrance bsiness worldwide; many non-credit-life microinsrance programmes, where reinsrance might be more sefl, are still in pilot mode, with premim adeqacy and prodct design being fine-tned and with limited exposre; the profit potential in microinsrance remains relatively minimal for both insrers and reinsrers today. The above notwithstanding, microinsrance is developing rapidly. Globally, lives covered by microinsrance were p nearly 300 per cent in 2011 since jst and are projected to doble by the end of this decade. 7 Yesterday s pilot projects are providing the key lessons for tomorrow s scalable offerings, and the potential market is growing every day with the advent of innovative distribtion methods and the inflx of capital from donors and (to a lesser extent) investors to spport experimentation and scaling. Extending crrent trends into the near ftre makes it very clear that the microinsrance market will need to embrace the reinsrance market and vice versa. In an effort to bring these two market segments closer together, this paper will explain what reinsrance is, why it is relevant to microinsrance and how it is being sed in microinsrance today. It will also attempt to identify gaps between crrent microreinsrance spply and demand, with the primary goal of devising strategies for the development of more reinsrance activities spportive of microinsrance market growth and sstainability. 4 Some backgrond on probability theory may be sefl here for some readers. One of the best-known principles of probability theory is the law of large nmbers, which, roghly speaking, means that the more comparable risks yo bndle together in a pool, the better yo will be able to predict the total otcome of the pool, since the degree of ncertainty that makes any individal risk intolerable is redced for a large enogh collection of risks. Stated anecdotally, it is impossible to say if a particlar person will have an accident dring the next 12 months, bt it is possible to predict within tolerable bondaries how many persons ot of 10,000 will have an accident. The term covariate risk is sed to describe the possibility of the same adversity happening to a large nmber of persons or assets in the same geographical area. Some well-known cases of covariate losses are earthqakes, hrricanes, epidemics and droghts. 5 Big government-sponsored programmes with the objective of social protection or economic development already have significant bsiness potential for reinsrance companies and have drawn significant attention therefrom. Sch social insrance schemes, thogh important, are not the central focs of this paper. 6 Lloyds estimated 135 million covered lives in 2009 (Lloyds and Microinsrance Centre, 2009). The ILO estimated 500 million covered lives in 2011 (ILO, 2012). 7 Microinsrance Network (2013). 7

8 2 THE MICROREINSURANCE VALUE CHAIN Figre 1. The microreinsrance vale chain Intermediaries can play a key role in facilitating risk transfer at each stage in the process. Donors promote vale chain development, fnctionality and scaling. Donor / Pblic entity Retrocessionaire Microreinsrer Microreinsrer Delivery channel Intermediary Intermediary Intermediary Micro(re)insrance programme vale chains are often complex. The reasons for this are many thogh they inclde the newness of many microinsrance bsiness models and concepts, the diversity of delivery channels, the involvement of varios stakeholders otside the direct vale chain (for example donors and intermediaries) and the idiosyncrasies of certain types of microinsrance bsiness. Moreover, the relationships and contracts between delivery channels and insrers, and insrers and reinsrers are often determined by other factors in the vale chain that are beyond their control, sch as the size and natre of an nderlying risk profile, the vale-added capabilities of an intermediary or the availability of retrocession protection (or retro for short). 8 To nderstand the importance of microreinsrance to a given vale chain, it is important to nderstand broadly the role of each vale-chain constitent (RE)INSURERS Reinsrers and insrers are essential to any micro(re)insrance vale chain. Withot appropriately licensed and willing risk-bearing entities, risk cannot be assmed from low-income consmers. 9 Insrers generally set the criteria for the assmption of a given risk and will transfer the said risk on to the secondary market where necessary, thogh for some types of risk and in certain markets reinsrers either effectively or actally set the price or determine the capacity available for primary risk transfer (more on this later). Reinsrers can also spport insrers by helping them to develop their nderwriting capacity for certain types of risk (for example, index-based weather insrance). 8 Retro, pt simply, is reinsrance for reinsrance companies and, where needed, typically represents the last stage in a potential risk transfer vale chain. As retrocession markets and nderwriting criteria are distinct from reinsrance eqivalents and the perceived need for retro coverage in microinsrance is de minimis, at present this paper will not focs on retro in any degree of detail. A retrocessionaire appears in Figre 1 for illstrative prposes only. 9 While the athor acknowledges the widespread existence of so-called informal microinsrance schemes ngoverned by national insrance spervisory athorities (and the significant body of literatre devoted to their analysis), one of the key barriers to the expansion of sch informal schemes is their general inability to access reinsrance and related expertise. To qote from the paper Isses in reglation and spervision of microinsrance (IAIS and CGAP Working Grop on Microinsrance, 2007): Sch arrangements may have negative implications at the instittional and client levels even thogh their services might be the only one that a poor hosehold has access to. First, these providers operate on an neven playing field, which goes along with market distortions and a lack of transparency, and favors arbitrary condct by providers. Second, they have little chance to grow, attract investors or partners, or access risk transfer mechanisms like reinsrance. Third, they may have little or no reserves, and be rather inefficient. Conseqently, there is a lack of risk sharing and adeqate technical spport from reinsrers for nderwriting and claims administration. If they were licensed nder the insrance law, sch microinsrers wold be able to access reinsrance (and reinsrers technical spport) and cold manage their capital more efficiently. This being said, it shold also be noted that reinsrance, when sed prely as financial risk management, is rarely the sole or main impediment to the growth of microinsrance schemes, informal or otherwise. 8

9 Insrers generally set the criteria for the assmption of a given risk and will transfer the said risk on to the secondary market where necessary, thogh for some types of risk and in certain markets reinsrers either effectively or actally set the price or determine the capacity available for primary risk transfer (more on this later). Reinsrers can also spport insrers by helping them to develop their nderwriting capacity for certain types of risk (for example, index-based weather insrance). The process by which insrers determine the price for assming a given risk is referred to as nderwriting. 10 The types of information insrers might obtain from prospective insrance prchasers vary by line of bsiness, thogh for life and health risks (re)insrers will want to collect information abot the mortality and/ or morbidity of the insred, sch as age, gender and occpation. For property risks the type of information collected generally pertains to the likelihood of a given property to sffer damage, which is determined by its location, constrction qality, occpancy and so on. Portfolio or treaty reinsrers (defined in section 3, Forms of microreinsrance ) nderwrite by obtaining aggregate information relative to the portfolio of risks being considered INTERMEDIARIES De to the many machinations involved in vale-chain formation, intermediaries with expertise in certain aspects of programme development and management are often relied pon to fit the pieces of the pzzle together. Reinsrance intermediaries can provide both reinsrers and cedants (another term for reinsrance prchasers) with a variety of vale-added services, sch as negotiating the details of cover, performing rotine administrative or complex analytical tasks and even spporting aspects of prodct development. In general, reinsrance intermediaries have the same aspirations as insrance brokers: to be advocates who nderstand the needs of both parties in a transaction and negotiate a fair deal. Bt in reinsrance, they often have an additional role, namely breaking down large risks and distribting them among a variety of programme participants a process called risk syndication. In traditional reinsrance contexts many reinsrers only want to take a small share (for example, 10 per cent) of a single reinsrance cover, typically when large property catastrophe risks are covered. In this scenario ten similar reinsrers wold be needed to provide the insrer with the desired level of risk transfer. A reinsrance broker is well eqipped to find them all and collect their commitments. Obviosly, reinsrance brokers reqire remneration for their services and therefore some (re)insrers avoid them. As the global market for reinsrance has grown larger and more complex, the vale of the services provided by an intermediary has increased. In the US market over 80 per cent of reinsrance bsiness is now condcted throgh intermediaries. 11 Sometimes, brokers deal with both insrance and reinsrance. A few traditional intermediaries have applied their bsiness models to microinsrance, and provide both primary distribtion and reinsrance placement services to insrance companies, which are thereby reqired to play only a passive role in the assmption of microinsrance risk. Other brokers specialize in either the primary or secondary transference of risk. Notable specialist microinsrance intermediaries inclde MicroEnsre and PlaNet Garantee. The services these intermediaries provide are similar to those provided by wholesale intermediaries or managing general agents in developed world markets. They act as a bridge of sorts between insrers and a variety of retailers (known more readily as delivery channels in microinsrance). In addition they provide insrers with specialist prodct design expertise and administrative spport. 12 They may even sometimes be involved in the placement of programme-related microreinsrance or alternative risk transfer soltions. 10 Views on the etymological origins of the term nderwrite as sed in insrance are somewhat mixed, thogh the most common one refers to the early days of insrance at Lloyd s of London in the seventeenth centry, where risk-takers wold literally write their names at the bottom of a policy nder the risk information provided, thereby assming financial responsibility for the risk in exchange for a premim. 11 Conning w/ A.M. Best data, Based on market share of US direct (18 per cent) and broker (82 per cent) reinsrers. 12 For more on microinsrance intermediaries see Chrchill and Matl (eds) (2012, chapter 23). 9

10 2.3. DONORS In many micro(re)insrance programmes, donors play a crcial role. As many sch programmes are pre-commercial (meaning they have little chance near-term of trning a profit, owing to insfficient technology, limited distribtion or ntested prodct designs), donor spport for technical development and sometimes premim sbsidy dring the initial stages is essential ntil the programme in qestion can become self-sfficient. Learning to work with and alongside donors to achieve programme sstainability is not always the easiest of transitions for commercial reinsrers to make. The reasons for this are varios thogh they relate primarily to the emphasis by donors on development impact and the measrement thereof. This is a very different yardstick for performance measrement than many commercial players are sed to. To track sch performance, grants from donors also generally involve extra application and reporting reqirements, which can be cmbersome DELIVERY CHANNELS It is also important for commercial (re)insrers to recognize the importance of delivery channels to sccessfl micro(re)insrance programme administration. Delivery channels are a diverse bnch ranging from microfinance instittions (MFIs) to retailers and mobile network operators and are the first point of contact for low-income insred people when it comes to policy issance and claims settlement. Given their direct relationships with low-income consmers, delivery channels are also a sorce of invalable information relative to risk characteristics and historical losses. The importance of their role in the vale chain cannot be nderestimated. As sch, it is vital for any given micro(re)insrer to invest in the capacity development of its delivery channel(s), which will probably be nfamiliar with the nances of insrance administration. It will also be important for wold-be micro(re)insrers to embed their operations at the delivery channel level as far as possible; meaning some degree of interconnectivity between the delivery channels systems/reporting lines and the micro(re)insrers wold be advisable. To achieve this will sally reqire some sort of capacity bilding work on the part of the micro(re)insrer (for example, training of trainers) to ensre the delivery channel develops sfficient instittional knowledge to inform programme administration. Sch work increases relationship stickiness and shold improve the overall qality and smoothness of programme administration, thogh it cold also reslt in a form of overreach for commercial (re)insrers that are not sed to investing in capacity-bilding activity with distribtion partners FORMS OF MICROREINSURANCE Pt simply, reinsrance is insrance for insrance companies. Reinsrance transactions come in a variety of shapes and sizes and are sally adapted to specific sitations. This section discsses all major reinsrance options sitable for microinsrance, as well as the main ses of reinsrance in the context of microinsrance programme design and sstainability. Particlar emphasis is placed on those forms of reinsrance of most relevance to microinsrance bsiness. 13 For more on microinsrance delivery channels see Chrchill and Matl (eds) (2012, chapter 22). 10

11 To redce nderwriting risk to levels commensrate with their risk appetite, professional insrance companies sally prchase an array of types of reinsrance cover. Their composition depends on the natre of an insrer s bsiness, an insrer s risk inclination and its reinsrer(s) preferences and specialisms, as well as on some ancillary considerations indirectly related to risk transfer, which will be discssed in more detail later on. The varios types and forms of reinsrance generally available are otlined in Figre 2. While not all the types shown are applicable to microinsrance contexts, others are highly relevant. Generally, reinsrance arrangements will be made p of one or more of the basic reinsrance prodcts shown in Figre 2 (in matre markets, more sophisticated forms of reinsrance are available bt they are seldom jstified in microinsrance and will not be covered here). 14 Figre 2. Basic classes and forms of reinsrance. The boxes otlined in red are most relevant to microinsrance. Reinsrance Treaty Facltative Prorata Index-based soltions Excess of loss Prorata Index-based soltions Excess of loss Qota share Srpls share Per risk Per occrrence Aggregate Broadly speaking there are two classes of reinsrance treaty and facltative (the latter is often called fac for short). Treaty reinsrance is sed to protect portfolios of mltiple nderlying risks, whereas facltative reinsrance is sed for single risks that are too large, asymmetrical or heterogeneos to be retained in their entirety by an individal insrer and may fall otside the scope of the said insrer s treaty arrangement(s). Microinsrance programmes are generally defined by large homogenos portfolios of risk hndreds, thosands or even millions of low-income consmers with reasonably similar risk profiles obtain reasonably similar or the same standard coverage for a given risk. From a reinsrance perspective sch portfolio-level characteristics lend themselves particlarly well to treaty nderwriting. 15 Ths, for the prposes of this review we will focs on treaty reinsrance concepts and ignore facltative ones. 16 Within the treaty category, there are three main types of coverage offered. They are presented in Figres 3, 4 and Among the more complex types of reinsrance transaction that will not be discssed in any detail in this paper are excess cessions agreements (hybrid agreements whereby an insrer cedes a portion of its risk excess of a certain limit threshold at a predefined price and receives a ceding commission in retrn); loss portfolio transfers (LPTs) (sed to transfer a book of claims which have already occrred off of an insrer s balance sheet protecting against adverse reserve flctations sed also to spport novation, or the complete transfer of an in force book of bsiness); block acqisition transactions (where the reinsrer advances part of the present vale of ftre profits from a book of long-term (sally life) insrance); strctred or finite reinsrance; and srpls share reinsrance. LPTs are toched on in section Somewhat perversely these same characteristics impede demand for treaty reinsrance since the predictability of portfolio otcomes is increased by their size and relative homogeneity. See section 5.2, Demand, for more on this sbject. 16 Some microreinsrance contracts are written facltatively, thogh these tend to reinsre a master or grop policy whereby a single policy and named instittion is the original insred nderlying the reinsrance agreement. However, this named instittion (e.g. an MFI) is sally extending coverage in some form or another to its individal low-income membership. While the reinsrance contracts pt in place to protect sch master policies cold technically be determined facltative since only one policy is being reinsred, the risk transfer concepts being applied relate to a portfolio of risks a decidedly treaty approach to nderwriting rather than a tre individal risk (e.g. a very large skyscraper). 11

12 Figre 3. Fnctionality of excess of loss reinsrance in six distinct claim scenarios. Note that reinsrance coverage only kicks in above a certain claim level. Non-Proportional / Excess of Loss Reinsrance total claim reimbrsed by reinsrer retained by insrer Figre 4. Fnctionality of 40% qota share reinsrance in six distinct claim scenarios. Note that the loss amont in each scenario is shared proportionally 60/40% between the insrer and the reinsrer. Pro Rata Reinsrance total claim reimbrsed by reinsrer retained by insrer 60% 60% 60% 60% 60% 60% Figre 5. Fnctionality of index-based reinsrance in six distinct claim scenarios. Index-Based Reinsrance total claim index payhot basis risk Note that the index payot does not always eqal the loss amont, reslting in a basis risk gain or loss for the reinsred. Also note that the sage of index-based microreinsrance depends on the natre of the nderlying policies being reinsred. If the nderlying policies are themselves index-based, the spporting reinsrance contract can simply be strctred as an excess of loss or qota share of the nderlying parametric liabilities. However, if the nderlying policies are indemnity-based, the index-based reinsrance agreement shold fnction in mch the same way as an index-based insrance prodct and shold be strctred to mimic the actal experience of the reference risks (in this case a collection of insrance policies rather than crop yields, for instance). 12

13 Excess of loss (XOL) coverage is often described as providing prchasers with vertical protection hence coverage above a retention. Pro rata coverage on the other hand is described as applying horizontally or sideways, since pro rata reinsrers participate in the losses of their nderwriting partners, starting from the first dollar or from the grond p. Index-based coverage provides coverage as soon as the index criteria are met (for example, rainfall of x millimetres or less in location y), thogh more complex strctres are available and all index-based soltions involve some level of basis risk or the risk of mismatch between the contract payot and the actal losses incrred by the contract s beneficiary. Index-based reinsrance at the portfolio level fnctions in mch the same way as its does at the individal level thogh both versions are designed to match the loss experience of the risk pool rather than that of an individal. 17 Each of these treaty coverage forms is discssed in more detail in sections 3.1. to EXCESS OF LOSS Excess of loss (XOL) contracts come in many sb-forms inclding per risk, per occrrence/event and aggregate (also known as stop loss). Per risk XOL tends to be sed to provide protection for lmpy or heterogenos portfolios of risk with some small and some large individal exposre. Since microinsrance programmes are generally homogenos meaning they tend to comprise many small risks with similar characteristics per risk XOL cover is often not the first choice as a form of protection. However, portfolios of small microinsrance risks can still be exposed to catastrophic loss, against which per occrrence XOL is ideally sited to protect. Aggregate XOL or stop loss protection can be very sefl for many types of microinsrance programmes since it provides a fairly definitive cap on adverse aggregate loss experience. Aggregate XOL contracts are often strctred sing loss ratio figres (aggregate programme claims divided by aggregate programme premim) to provide clients with bdgetary and solvency assrance. A strctre wold be something like this: 50 per cent loss ratio points of coverage in excess of an 100 per cent loss-ratio-point retention (or, in short hand, 50% LR xs 100% LR cover). Sch contracts can be especially sefl for new programmes in new areas with limited catastrophic risk, where ncertainty arond price adeqacy and ltimate loss ratio otcomes can be qite high (for example, large new life and health schemes). 18 However, in exchange for relative certainty arond primary otcomes, reinsrers can charge hefty premims for sch cover. A more detailed overview of per occrrence and aggregate XOL contracts follows. Regardless of form, the selection of attachment and exhastion points for any given XOL cover shold be based on risk analysis of the nderlying portfolio. Ideally this risk analysis wold be probabilistic in natre, providing estimates of both loss freqency and severity, thogh simple deterministic measres can be sed as well (e.g. a LR of > x% will case programme insolvency, so prchasing aggregate protection below x% wold be advisable). Microinsrance programme managers shold also be wary of over-reinsring their risk. This can manifest in XOL by attaching too low and bying nnecessary protection or paying more than warranted for the risk ceded. Reinsrance can be a very sefl financial tool bt it is not always the most efficient or effective means of financing risk detailed risk analysis is integral to appropriate XOL protection design. 17 It shold also be noted that portfolios of individal-level index-based insrance prodcts can be reinsred sing pro rata or excess of loss strctres as well. 18 It is rare for reinsrers to provide XOL coverage for new health schemes thogh demand for sch protection may still be there. 13

14 Per occrrence Per occrrence reinsrance contracts are best and most often sed to protect against catastrophic losses. As low-income poplations tend to be more exposed and vlnerable to catastrophes than their higher-income conterparts, it wold seem logical for per occrrence microreinsrance to be endemic in the microinsrance market. However, according to research condcted by the athor, there are very few, if any, instances of programme-specific per occrrence reinsrance being sed by microinsrers. The reasons for this are varied, thogh they inclde the following: Most catastrophic risk is dealt with via index-based risk transfer soltions. 19 Catastrophic risk is most apparent in property and very few property microinsrance programmes operate on an indemnity basis today. A large portion of global microinsrance bsiness is written by very large commercial insrers, who simply do not reqire reinsrance spport for their microinsrance programmes. Most commercial insrers prchase per occrrence XOL reinsrance protection corporately for their entire portfolio. To the extent that sch insrers engage in non-life microinsrance, these programmes are most likely sbject to their corporate catastrophe cover and are ths technically reinsred on a per occrrence basis (however, while technically reinsred, they may not be effectively reinsred since corporate commercial cover tends to be designed to protect traditional risks and so may attach at a mch higher level than cold be effected by a microinsrance-specific loss). The same is generally tre for life microinsrance, whereas for health there is very little per occrrence XOL available for either microinsrance or commercial insrance, an operational definition of pandemics being one obstacle. Non-life microinsrance programmes with catastrophic exposre tend to be small, in start-p mode, geographically concentrated or share some combination of these factors. It is very difficlt for a reinsrer to nderwrite a small portfolio of risks on an XOL basis economically since per occrrence XOL cover is generally most effectively/efficiently implemented to cover infreqent severe events, ths sharply diminishing the available ceded premim volme by simple virte of the prodct s inherent strctre (the higher the attachment point, the less risk transferred and the lower the premim rate for the excess risk ceded). Moreover, withot at least some nderlying portfolio diversification, the risk of loss to the bsiness can often simply be too high or volatile for the cost of risk transfer to make sense for the cedant. It is not clear whether the strctral barriers to obtaining programme-specific per occrrence microreinsrance mentioned above are acting as a hindrance to the growth of the indemnity-based property microinsrance market. It is more likely constrained at the front-end by difficlties srronding distribtion, loss assessment and local nderwriting expertise. Regardless, there are some interesting active microreinsrance models which seem to be addressing the conndrm of non-life microreinsrance access from both sides throgh pblic private collaboration and the creative se of pooling mechanisms or new prodcts. One sch model has been sccessfl at removing some of the strctral barriers to obtaining per occrrence and other reinsrance protection for a consortim of small Latin American cooperatives (see Box 1 ICMIF LARG). Another has developed a hybrid index and indemnity protection for MFIs (see Box 4 MiCRO). 19 Again, while reinsrance contracts protecting index-based microinsrance programmes can be written on an XOL basis (e.g. only triggered losses above a certain threshold are covered, p to a specified limit), we make a deliberate distinction here between per occrrence XOL and index-based microreinsrance, based on the nderlying characteristics of the microinsrance bsiness reinsred. 14

15 Box 1: ICMIF LARG The Latin American Reinsrance Grop (LARG) hosed at the International Cooperative and Mtal Insrance Federation (ICMIF) consists of 16 members, all of which are insrance cooperatives based in Latin America. The impets behind the formation of the LARG was gronded in a market failre each member organization strggled to access non-life reinsrance independently, owing mainly to their relatively small size and perceived lack of sophistication. With assistance from the ICMIF and other stakeholders, these individal insrers banded together and developed a consortim approach to the prchase of reinsrance in the form of LARG. At present only six of the 14 LARG members are bying reinsrance jointly, thogh the remaining members intend to by in the near ftre and have joined the grop in advance thereof to gain access to training, nderwriting spport and other apprtenant membership benefits. Reinsrance prodcts prchased by LARG inclde property catastrophe XOL; per risk, motor and miscellaneos working cover XOL. American Agricltral is LARG s lead reinsrer and Willis Re is the facility s placing broker. For those members that do prchase reinsrance jointly, they save on reinsrance premim compared to what they wold pay individally for eqivalent protection at market rates. These disconts are attribtable both to the statistical merits of exposre volme and risk diversification and to the administrative efficiency associated with combining member portfolios from a variety of jrisdictions, as well as to the adherence to common nderwriting standards and discipline by LARG members. All LARG members pay an annal membership fee and reinsring LARG members pay an additional 2 per cent of gross premim as a contribtion towards the operating expenses of the LARG. Fees go to pay the salary of a fll-time coordinator, the costs of an annal seminar and other operating expenses of the grop. LARG s decision-making process is highly democratic, meaning members are asked to vote on most decisions related to reinsrance coverage, nderwriting parameters and training procedres. While LARG is not a licensed entity in any one contry at present, its members are interested in obtaining a licence so as to acqire a legal stats that will allow it to operate with all faclties. A step towards this formalization was taken as a reslt of a large fire loss sffered in 2006 by a LARG member, which hit LARG s property XOL reinsrance treaty and sparked some conversations among the grop and its reinsrers abot standard nderwriting criteria. LARG has since developed a set of joint nderwriting principles as a condition of membership, which are dly enforced. None of the LARG members define themselves as pre microinsrers per se, thogh if we apply as the main criterion for this stats an access-based definition of microinsrance (that is, insrance for the otherwise financially exclded or n(der)insred) instead of the traditional income-based definition (insrance for persons on a low income by local standards) then mch of the insrance provided by LARG members wold most likely be considered microinsrance. Frthermore, since many LARG members were formed (as many cooperatives are) to fill a commercial market gap, their very mission implies ipso facto an alignment with the above access-based definition of microinsrance. This being said, for LARG members offer prodcts that are specifically branded microinsrance, all of which are forms of life cover and so do not fall nder the prview of the LARG s non-life reinsrance programme. One of the LARG members, however, is working on a project to offer non-life microinsrance and was expected to lanch the prodct in Irrespective of the grop s crrent relevance to microinsrance, the LARG is an excellent example of how risk pooling can be sed effectively by small insrers (formal, informal or otherwise) to achieve cost-efficiencies by increasing the exposre volme, geographic diversity and administrative simplicity of a sbject portfolio. Several similar schemes have been mooted for microinsrance thogh none have to date gathered any steam. The reasons for this will be discssed in more detail in the conclsion to this paper, thogh some of the key lessons to be drawn from the LARG s sccess are as follows: A consortim approach to reinsrance prchasing increases the prchasing power of each member and reslts in valable premim cost disconts afforded by increased exposre volme, risk diversification, nderwriting discipline and administrative efficiency. The alignment of interest and nderstanding among consortim members is crcial for the solidarity of the grop and to maximize nderwriting efficiency, leading to the development of a common ideology and consenss on best practices/standards. Resorcing the consortim adeqately is crcial to ensre that prchasing and other ancillary activities are managed appropriately. 15

16 Aggregate Aggregate XOL, sometimes referred to as stop loss arrangements, are reinsrance contracts that provide annalized protection to cedants. Sch arrangements are attractive to cedants becase they provide protection against deviations both in the freqency and the severity of claims, thogh a stop loss arrangement is often difficlt to obtain since it has a nmber of disadvantages for the reinsrer. Specifically, reinsrers are reqired to assme a greater degree of volatility in potential loss otcomes since both catastrophic claims and/or a large nmber of small claims can expose the contract. For lines of bsiness where catastrophic claims are rare, sch as in health, stop loss arrangements are somewhat common, thogh in other lines of bsiness, sch as property, tre stop loss arrangements are harder to come by. Box 2: Aga Khan Agency for Microfinance, Pakistan When the Aga Khan Agency for Microfinance (AKAM) decided to develop microinsrance in Pakistan in 2006, its belief in the bsiness case for microinsrance, combined with an acceptance of the fact that prodct design and pricing had to be experimental in the beginning, gave the agency the conviction that it wold be able to gather sfficient data in the first few years of the programme to improve prodct design and minimize the risk of error. Bt as AKAM did not have an insrance license, it had to rely on existing insrance companies for policy issance and their appetite for health microinsrance was limited: the initial mispricing risk was high in the absence of prior experience and data, and there was little room for safety margins in the premim, given the inherent small ticket natre of microinsrance. To address the worries of its insrance partner, AKAM partnered with a large international reinsrance company to provide a stop loss cover for the health microinsrance bsiness. To overcome the reinsrer s initial relctance to take on sch a high level of risk, AKAM arranged a fnds provision and hold harmless agreement with the reinsrer in parallel to the stop loss reinsrance contract between the reinsrer and the insrer. Under that agreement, AKAM established a deposit with the reinsrer, from which the reinsrer was compensated for claims payments nder the stop loss in a form of retrocession. With this strctre, AKAM had very cost-efficiently replicated an insrance captive for its health microinsrance. In other words, AKAM arranged insrance that was nderwritten by NJI Life Pakistan, which in trn reinsred the risk with Swiss Re. AKAM also arranged the fnds provision and hold harmless agreement with Swiss Re, which compensated the reinsrer for any payments they might make nder the reinsrance agreement with NJI Life. As the stop loss paid only 90 per cent of covered claims p to a limit, the insrer retained some risk and alignment of interest. Bt as knowledge of this arrangement spread along the vale chain, some microinsrance programmes came to view it as a generos (if complicated) donor sbsidy mechanism, and nderwriting reslts deteriorated as alignment of interest diverged. In other words, since losses were ltimately paid by AKAM, local stakeholders involved in the scheme inclding the First Microinsrance Agency and its vale chain partners like NJI Life, were not inflenced by the risk of nderwriting losses and were incentivized to increase sales and provide generos healthcare to their commnities rather than being incentivized to control claims costs. Given the delay in obtaining sefl data from remote areas and implementing the necessary corrective action throgh long vale chains, these isses cold not be resolved satisfactorily, and eventally the stop loss and hold harmless agreement were not renewed PRO RATA Qota share Qota share (QS) is the simplest form of reinsrance, where a reinsrer agrees to pay a cedant a predefined percentage of every claim in retrn for the same percentage of premim for the reinsred book of bsiness. 20 To compensate the cedant for its costs of acqiring the bsiness sbject to the reinsrance contract, reinsrers will sally offer cedants a ceding commission taken as a percentage of ceded premims. So if Insrer A bys a 50 per cent qota share from Reinsrer B with a 25 per cent ceding commission, for every dollar of premim A cedes to B it will receive 25 cents in retrn. 20 Qota share is similar to co-insrance contracts in long-term life markets. 16

17 On a net basis, after acconting for ceding commission income, A will retain five-eighths of total gross programme premim bt only half the gross risk. While this income differential shold be largely commensrate with the cedant s related acqisition costs, in practice a ceding commission is a negotiated rate 21 and it is possible for insrers to make risk-free commission income on the portion of the book ceded. QS reinsrance fndamentally provides cedants with capacity relief; that is, it frees p solvency capital reqired by the insrer to be sed elsewhere. QS also protects against the risk of initial programme mispricing and the risk of an exogenos factor affecting nderwriting otcomes. It can also be spportive of growth, thogh it does not protect against covariant risk. For more on the ses and benefits of QS reinsrance see section 4.1, Capacity relief. Box 3: Orchard Insrance Grop: A QS microreinsrance case stdy 22 Orchard insrance Grop is an insrance sbsidiary of the African Alliance Grop and sister company to Select Africa, a leading pan-african MFI. Orchard received its licence to operate in Swaziland in 2010 and was formed primarily for the prpose of writing microinsrance bsiness to cover Select Africa s microlending clients. Orchard s crrent prodct site consists of two main prodct lines: credit life and fneral. Credit life is offered only to microlenders on a mandatory grop/portfolio-wide basis, whereas fneral is offered both as a mandatory and a volntary grop prodct 23 and on an individal basis (the latter retail offering was lanched late in 2011). Prior to Orchard s formation, Select was offering its borrowers credit life and fneral microinsrance prodcts via partnerships with local insrers. In this way Select took no nderwriting risk, bt instead was compensated for its role in the vale chain as a delivery channel. After several years of operating in this fashion, Select s management began to recognize the vale of microinsrance prodcts to its individal clients and its instittional credit risk management. They also recognized that the loss history associated with the Select credit life and fneral bsiness was reasonably stable over time and that managing it separately from other cases of credit defalt increased their comfort with the risk. After developing a firm fondation of nderstanding in microinsrance nderwriting and experiencing some volatility in prodct pricing, as qoted by third-party insrers, Select staff and management became interested in increasing the company s control over its insrance bsiness. This prompted the development of a fll-service bsiness model for the delivery of microinsrance prodcts to Select s clients Orchard represented the clmination of this exercise. Select s sister microinsrer has since expanded its client base beyond Select and intends in the short term to extend its reach into other lines of short-term bsiness which it is approved to transact in Swaziland. As a small start-p insrer owned by a predominantly non-insrance financial enterprise, Orchard came into existence with de cation and relied heavily pon relationships with reinsrers from its inception. With the assistance of reinsrance intermediary Gy Carpenter and Company, in early 2011 Orchard prchased a credit life and fneral QS from a panel of international reinsrers, covering 90 per cent of the company s risk. The terms and conditions of this reinsrance agreement informed the company s initial rating and nderwriting practices and bolstered its capital base sbstantially, allowing it to write significantly more bsiness than it wold have been able to otherwise. With the fll technical and financial spport of its reinsrance partners, the company is now moving towards greater risk retention, more nderwriting athority and prodct diversification. 21 The insrer may not always have the bargaining power to negotiate the best rate, in which case it may be best to associate with another cedant and/or solicit the services of an intermediary 22 See also Chrchill et al. (2012, Box 10). This paper can be fond here: 23 A volntary grop prodct involves an insrer writing a single policy covering an association of some description bt allowing the individal members to decide whether to sign p for the coverage, which is not a prereqisite for membership. 17

18 Many MFIs are overwhelmed by the prospect of adding microinsrance prodcts to their portfolios, irrespective of the manner in which sch prodcts are managed and delivered. Partner agent relationships are a common and effective way for MFIs to enter the microinsrance marketplace withot assming any nderwriting risk. For more sophisticated or larger MFIs, or those with some specific in-hose experience relative to insrance, the fll-service model is a compelling and viable alternative. Absent a significant controlled portfolio of bsiness, ample internal nderwriting expertise and/or a significant initial investment in nderwriting capital to spport risk retention, starting a new insrer often reqires reinsrance spport, both technical and financial. Lessons to be drawn from Orchard s experience inclde the following: The fll-service model of microinsrance makes sense for more sophisticated, experienced and/or larger MFIs since it allows them to control more aspects of the prodct offering. It can also represent a profitable new bsiness line so long as nderwriting risk is managed prdently. Owing to the nderwriting risk involved in the fll-service model, MFIs interested in starting their own microinsrance bsiness shold evalate the many tangible and intangible merits of obtaining reinsrance spport for sch an effort, and this evalation shold be renewed periodically. Beyond internal concerns arond risk levels, new insrers are commonly reqired by reglators in their domicile of choice to incorporate certain reinsrance arrangements into the bsiness plan accompanying their license application. The natre of sch a reglatory reqirement will natrally vary depending on the insrer s initial capital levels and target lines of bsiness INDEX-BASED SOLUTIONS Microinsrance prodct development inevitably enconters a variety of challenges. For commercial insrance companies to access the low-income marketplace, entirely new bsiness models are reqired so as to streamline distribtion and attendant administrative processes withot compromising prodct and service qality. Sch an innovation imperative erects a significant barrier to market entry this is particlarly tre for natral catastrophe risk. Index-based or parametric insrance prodcts represent a very promising avene for the implementation of effective and efficient catastrophe risk transfer. Sch prodcts can benefit the poor and be spportive of economic development, mainly de to their ability to streamline programme administration, redce associated expenses and increase the speed of payots to low-income policyholders after a natral disaster. Since index insrance payots are priced and triggered on the basis of an objective parameter, they do not reqire any of the detailed nderwriting and loss adjstment processes sed for traditional indemnity prodcts. 24 However, while administrative expenses may be redced by the se of index-based prodcts, these can prove rather difficlt and expensive to design. Moreover, the implementation and/or expansion of index-based microinsrance programmes may also remain hampered by a nmber of other external limiting factors. These inclde data shortage, model development complexity, and a dearth of technical capacity in emerging markets, all of which lead to diminished or non-existent catastrophe risk transfer markets. Recognizing this, a nmber of aid agencies and government-fnded initiatives have begn to focs their efforts on promoting the significant research and development necessary and nlocking the reqired international resorces to deploy effective, efficient and innovative index-based catastrophe cover. These inclde the ILO Microinsrance Innovation Facility, the Global Index Insrance Facility (GIIF), the Index Insrance Innovation Initiative (I4) and the Mnich Climate Insrance Initiative (MCII), to name bt a few. 24 While data on expense ratios in microinsrance are sparse, some life and health microinsrance programmes have posted total expenses above 40 per cent of total programme premims. This is somewhat comparable to the US property and casalty insrance indstry, which posted an aggregate nderwriting expense ratio of 28% and a loss adjstment expense ratio of 12% in 2012 (Sorce: SNL Financial, US P&C Underwriting Analysis, accessed in Nov 2013). With appropriate prodct design, some scale and a well-fnctioning vale chain, index-based catastrophe programmes shold theoretically be able to achieve single-digit expense ratios inclsive of nderwriting and loss adjstment costs. 18

19 Mch of the expertise needed to design index-based prodcts and mch of the risk capital needed to spport the transfer of catastrophic risk from the developing world resides in the international reinsrance market. It is for this reason that mch of the reinsrance market s activity in microinsrance has taken place in the index-based catastrophe arena. Unsrprisingly a nmber of notable reinsrance organizations have affiliated with the above-mentioned donor initiatives, either as contractors or as co-fonders, in an effort to develop soltions for the management of catastrophic risk among global low-income poplations. As this section will reveal, sch reinsrance-driven pblic private partnerships (PPPs) play a pivotal role in the establishment and expansion of many index-based micro(re)insrance schemes; to the extent where many reinsrers can be seen acting in a capacity well beyond the traditional scope of their operations, advising on microinsrance prodct design or development and assisting with the technical capacity bilding of delivery channels. With this in mind, mch of the section that follows will focs on primary index-based microinsrance innovations and challenges, given the heavy inflence of reinsrers on this market. Moreover, focsing prely on the reinsrance of index-based microinsrance schemes wold not be very informative for readers since sch arrangements are generally placed on a simple pro rata or XOL basis and many are strctred as pre pass-throghs, where the insrer transfers per cent of its risk to the reinsrer Basis risk 25 Despite the promise of index insrance and the high-profile initiatives mentioned above, few programmes have attempted to address head-on what is argably the most notable hindrance to the widespread adoption of parametric catastrophe microinsrance programmes: basis risk. Basis risk or the risk that an index-based contract payot will be greater than the actal loss incrred by the insred ( basis risk gain ) or less than it ( basis risk loss ) is navoidable to some degree in index insrance. However, it tends to be prononced in a microinsrance context becase of the comparative lack of detailed long-term weather and exposre data in low-income markets. When index prodct payots are insfficient to cover the losses incrred by a low-income individal or commnity, the risk of irrecoverable reptational damage to the prodct is significant withot adeqate client edcation copled with any nmber of additional financial and/or social basis risk management methods. Figre 6 presents a more technical portrayal of basis risk. Figre 6. Types of basis risk associated with typical catastrophe index insrance contracts The total width of the diagram (A+B+C) represents the total risk faced by the insred. The total basis or ninsred risk nder a standard index insrance contract will be the sm of the idiosyncratic risk pls the design effect (A+B). The insred risk is represented by the width of area C. The design effect reflects the imperfect ability of any index to predict the average otcomes in the covered zone. The size of B will vary loss by loss. Total natral hazard risk Idiosyncratic risk A B Design Effect Correlated risk C Basis risk Index-based insred risk Sorce: Adapted from Carter (2011). 25 The introdctory paragraphs of this sb-section were paraphrased from Bernhardt and Yong (2012). The article can be fond here. 19

20 Of the many index-based microinsrance pilots worldwide, few have been able sccessflly to address the isse of basis risk directly. However, many innovative programmes have been lanched in recent years with significant technical and financial spport from reinsrers in an attempt to increase catastrophe insrance penetration among the poor and to nlock international reinsrer capacity for the prpose. Sch schemes tend to be driven from the top down by reinsrance practitioners since the skills to develop and nderwrite index insrance reside predominantly in the reinsrance indstry, where catastrophe modelling is de riger and science-based analytics are already well nderstood and sed. Boxes 4 and 5 otline two sch programmes. Box 4: Microinsrance Catastrophe Risk Organisation, SCC (MiCRO): An index-based property microreinsrance case stdy In March 2011, the lanch of the Microinsrance Catastrophe Risk Organisation (MiCRO), a licensed insrance company domiciled in Barbados, was annonced by its fonding partners (see below for details). MiCRO began operations in Haiti in Janary 2011, spporting the provision of mandatory grop catastrophe insrance 26 to approximately 58,000 microentreprener borrowers of the MFI Fonkoze. Althogh initially lanched in Haiti, MiCRO intends to expand globally, providing catastrophe cover distribted throgh organizations sch as MFIs. The following case reflects on MiCRO s first implementation of microinsrancein Haiti. This microinsrance programme has since been discontined in favor of a portfolio cover for Fonkoze, and MiCRO is crrently developing new prodcts in Central America. The basic strctre of the initial MiCRO partnership is shown in Figre 7: Fonkoze Haiti s largest MFI and Mercy Corps are eqity investors in the facility, with spport for operations provided by a host of major donors inclding the UK Department for International Development (DFID) and the Swiss Agency for Development and Cooperation (SDC). Swiss Re provided parametric reinsrance (technically retrocession) to the facility, and CaribRM and Gy Carpenter & Company s GC Micro Risk Soltions nit were the contracted operators of the facility (responsible for technical prodct design and financial risk management/risk transfer placement, respectively). Reinsrer Mlti-Donor Trst Fnd Micro Local Insrer Basis Risk MFI 1 MFI 2 MFI 3 26 Cover is provided for wind, rainfall and earthqake. 20

21 This catastrophe cover provided by MiCRO endeavored to leverage the lower cost and expediency of parametric insrance while bffering the MFI and its clients against basis risk by offering them the option to prchase comprehensive basis risk cover, sbject to certain policy limits. Cover was comprehensive in that it inclded both idiosyncratic losses and losses (or gains) cased by the design effect portrayed in Figre 6. This niqe basis risk approach was developed to provide instittional clients with a hedge against the risk of mismatch between actal losses and parametric payots, which is a significant impediment to the development of viable parametric micro(re)insrance prodcts worldwide. MiCRO s parametric prodct in Haiti was designed to align with Fonkoze s Kore W prodct, which eliminated borrowers debt and provided them with an emergency expense payment of 5,000 Haitian gordes (US$125). Overall MiCRO paid ot nearly US$9m in parametric and basis risk claims to Fonkoze in its first 2 fll years of coverage, well over the amont of associated premim collected. In addition, becase Fonkoze was not flly insred for the total vale of the Kore W prodcts otstanding with clients, the MFI absorbed a significant portion of the losses. An open qestion abot the sstainability and scalability of this model remains the role of donor capitalization for the provision of basis risk coverage and the pricing of the associated risk. As MiCRO expands to reach additional contries and to provide different types of cover, it will need to adapt the model to reflect the lessons learned in Haiti and redce its dependence on donor fnding over time. If it is able to develop a prodct that is both affordable and financially sstainable, while maintaining the efficacy of its operations in order to transition away from grants, MiCRO cold play a key role in extending disaster microinsrance coverage to many more low-income commnities. Lessons learned from MiCRO s experience in Haiti inclded the following: Basis risk can be insred against if techniqes can be developed to qantify the risk, thogh insring both types of basis risk idiosyncratic and design-effect with a single policy can be extremely challenging since sch risks are very different. Design-effect basis risk losses in particlar have a high potential to create an adverse reslt, given the extreme vlnerability of low-income clients to catastrophe events and the difficlty of designing affordable parametric index prodcts that captre all aspects of this vlnerability appropriately. Innovating can be expensive and time consming. After less than 3 years of operations and US$9m in claims payments, MiCRO s coffers need to be replenished to allow the company to contine nderwriting basis risk coverage. Finding a sstainable new bsiness model in a difficlt frontier market can take years, ample hman resorce and a significant amont of cash. Sorce: Adopted and pdated from ASEAN (2012) and spplemented with the first-hand knowledge of the athor Triggers The main difference between index-based microinsrance contracts and traditional indemnity or qasi-indemnity 27 contracts is the objective determination of loss of the former. To determine whether a loss has occrred, index-based contracts rely on what is called a trigger or a predefined threshold at which a payot becomes de. Types of triggers which are or cold be sed for index-based micro(re)insrance contract design inclde the following: 27 Qasi-indemnity cover is a contract that is triggered by some actal loss bt does not compensate insreds dollar for dollar as with fll indemnity contracts. An example wold be a rbric-based damage assessment cover whereby the amont of the payot depends roghly on the extent of damage actally incrred by the insred (for example, assessed total loss as predefined eqals US$100 payment, partial loss as predefined eqals US$50 payment, etc.). 21

22 Parametric loss: A parametric loss contract ses an objective measrement of a parameter sch as wind speed, rainfall amont or earthqake grond motion to determine loss amonts. Events measring above the stated parameter(s) will reslt in loss payots. Pre parametric triggers where the parameter itself determines the occrrence and magnitde of loss are differentiable from parametric index contracts, where a parametric reading is fed into an indexation formla to determine the loss amont. In either instance parametric loss triggers often involve sbstantial basis risk. Indstry or area loss: Indstry or area loss triggers depend on actal aggregate losses experienced in a given commnity and reqire an independent third party to verify commnity loss amonts. Contracts will typically trigger if aggregate losses exceed a certain predefined threshold, and can involve sbstantial basis risk if the insred portfolio in a given commnity does not reflect the exposre characteristics of the whole commnity. The smaller the commnity sed for index loss determination, the lower the associated basis risk. Yield-based or mlti-peril crop insrance cover cold be classed as area loss triggers insofar as they are set at the commnity level or some higher administrative level, for example. Modelled loss: A modelled loss trigger ses parametric inpts from an event to simlate the event s characteristics and its effects on the covered exposre portfolio. Models are sally escrowed (that is, held in stasis and validated by an independent third party) and rn post-loss by a modelling firm (and validated by the escrow agent) to determine the magnitde of payots. Generally today sch complex trigger designs are not warranted or practical in the context of micro(re)insrance, thogh this cold change as technology improves and the reach of existing catastrophe models extends frther into developing markets. Each of these trigger types has advantages and disadvantages, which shold be evalated on a case-by-case basis. Generally evalation of trigger design sage shold take place along two dimensions: the cost and complexity of trigger development and the maintenance and the associated basis risk (see Figre 8). Coarser index prodcts will generally have a higher degree of basis risk, thogh often the cost of developing a more granlar (and more complex) version of the prodct otweighs the associated benefits in terms of basis risk minimization. Figre 8. A stylized view of the relative complexity/development cost and basis risk associated with certain index-based trigger types verss indemnity contracts Relative Basis Risk Relative Cost/Complexity Low High Indemnity Modeled Loss Indstry/ Area Loss Parametric Index Pre Parametric Low High 22

23 Box 5: Mnich Re CLIMBS initiative: An index-based property microreinsrance case stdy In 2011, Cooperative Life Insrance and Mtal Benefit Services (CLIMBS) entered into a pblic private partnership with Mnich Re, with spport from the Detsche Gesellschaft für internationale Zsammenarbeit (GIZ), to strengthen the sstainability of member MFIs in the Philippines. As an mbrella cooperative with more than 2,000 cooperative primaries and federations contrywide, CLIMBS was able to provide a sizeable distribtion network for the extreme weather event insrance prodct designed principally by Mnich Re. The parametric index for the prodct was strctred at the mnicipality level and categorized wind speed and rainfall into 10-, 15-, and 20-year events. Using these benchmarks as payot triggers, CLIMBS was eqipped as the local insrer to make payments to local cooperatives based on the percentage of their loan portfolios affected. Severe weather events were monitored in real time via satellite across the entire contry and timely reports were made available online to all member cooperatives. Mnich Re acted as the reinsrer to the project, providing reinsrance capacity to enhance the long-term sstainability of the programme. Loss payments (if any) were meant to strengthen the cooperatives lending operations and to keep them active shortly after loss occrrence a critical time for local disaster recovery. With stronger loan portfolios, the cost of microcredit shold have also been redced. Frthermore, by a binding statement of principles, the local cooperatives committed themselves to passing on any insrance proceeds to those individal members most affected by a given catastrophic event. This innovative instittional-level claims methodology placed the ons for eqitable distribtion of loss fnds on the local cooperatives and relied on the mtality of member cooperatives to serve the best interests of their clients. This in trn served to mitigate the potential adverse impact of basis risk on individals. Insrance payots were intended for se by the individal beneficiaries to assist them, for example, in rebilding their hoses or replacing livestock or other assets after a tropical cyclone. Althogh the reslts of the CLIMBS project have yet to be flly assessed, and the programme is presently dormant, it does offer some insight into the vale of pblic private partnerships that enhance disaster resilience by bringing advanced disaster risk financing tools and technology to bear on behalf of individal farmers and landholders. It also highlights the potential seflness of mtal insrance strctres to the downscaling of index-based insrance programmes and to the eqitable management of basis risk. Lessons learned: Wholesale aggregation of risk increases the size and potential scalability of the programme significantly verss partnering with individal cooperatives. A mtal operating strctre can facilitate the management of basis risk among individal members since the effects of any mismatch in actal losses verss indexed payots are smoothed across a cooperative grop and the rles/mechanisms for this smoothing are mtally determined among grop members, thereby minimizing the risk of discontent over otcomes. Sorce: Mnich Re (2010). Given the experiences reconted in Boxes 4 and 5 above, phased programme development is important to the sstainability of any index insrance programme. In many cases there is a great deal of concentration of resorces and effort on lanching a pilot programme thogh very little follow-throgh with respect to growing the pilot and placing it on a firm commercial fondation where programme volmes are sfficient to spport all risk transfer vale-chain stakeholders. It takes time (sometimes years) for the ltimate beneficiaries of index insrance prodcts to begin to trly appreciate the benefits of the cover, for delivery channels to bild their sales and administration capacity and for (re)insrers to adjst and improve the prodct so as to better attend to client demand. All in all, developing sstainable markets for index insrance reqires a long-term investment/commitment on all sides. In the longer term, the microinsrance market for property and agricltre risks may transition from parametric index-based soltions to area-loss or even indemnity soltions, where basis risk is minimized and risk transfer maximized. Broadly, the bsiness model innovations, technology, data collection and 23

24 validation efforts and cltral change reqired to see sch a revoltion come abot are probably many years away. In the meantime, the index-based initiatives featred in this report and other similar initiatives are making significant contribtions toward the mitigation and minimization of basis risk, and the lessons learned from each project will inform many similar innovations as time goes on. Reinsrers are well positioned to contine to drive this innovation as they seek new bsiness in alternative, emerging jrisdictions and contine to possess the lion s share of commercial risk indstry expertise with respect to index-based prodct design and catastrophe risk modelling and management. 4 CORE FUNCTIONS OF MICROREINSURANCE Microreinsrance, like traditional reinsrance, serves one or several of six main fnctions, each of which is explained in this section. To provide some backgrond to this, it is important to explain the broad types of risk that reinsrance is generally prchased to protect. These are: Risk of random flctation: This is the risk of deviation of actal claims from the estimated expected vale when by chance a particlarly high or low nmber of insred events occr and/or there is a particlarly high or low nmber of individal claims. Covariant risk, or the risk that an event will affect a large nmber of the insred items/people at the same (for example, an earthqake, flood or pandemic), is perhaps the most commonly reinsred form of random flctation. Risk of error: This is the risk of deviation that reslts from an incorrect assessment of a given probability distribtion. The risk of error is most present when an insrer expands into new fields. While there is plenty of global mortality data, there is very little data on the mortality of, say, low-income people between the ages of 20 and 65 in a particlar region and even less on their morbidity. The same holds tre for property or crop damage in new markets as well. Based on this reality, errors in pricing for new microinsrance programmes are likely. Reinsrers are often willing to assme this risk, especially if the profit potential of the new market is promising. However, they try to limit their loss positions by participating in the primary nderwriting of new cover, reqiring stringent monitoring by insrers and/or delivery channels to ncover errors as early as possible, and reserving the right to review reinsrance conditions as necessary. 28 Understandably, reinsrers often have more information on new perils or coverage forms than local insrers in less matre markets, given their experience from other markets. It is often reinsrers who bring new insrance prodcts to new markets and are more comfortable assming the risk of error than relatively inexperienced local insrers. However, in many lines of microinsrance bsiness the risk of error is likely to be a major deterrent for reinsrers. This may be becase there is little room for margins intended to compensate for pricing errors in microinsrance premims and an aggravating factor may be that the risk of error is not diversifiable, and grows with growing insrance volmes; therefore, the larger the anticipated scale of a microinsrance ventre, the more worrisome this risk becomes for (re)insrers, sbject of corse to the potential velocity of catching and implementing corrections as needed (which, owing to the typically long vale chains in microinsrance, can be slower than for conventional insrance). Risk of change: This is the deviation from the estimated vale of a given loss distribtion de to an nanticipated exogenos change that took place after the distribtion was generated impacting the slope of the probability distribtion. All insrance is sbject to the risk of change, or the risk of external factors affecting the profitability of a given programme. Common examples of sch risk inclde an adverse change in a salient macroeconomic interest rate or a political act or event that has an adverse effect on a local microinsrance plan. The only remedy for sch risk is timely corrective action, which is feasible in the absence of any long-term garantees. This is a strong argment to avoid long-term garantees in experimental new lines of bsiness, and microinsrance cold be inclded in this category. Bt even when no policies with a dration of longer than 1 year are sold, it can take time to realize that the premim has become inadeqate becase of the long and often manal process of microinsrance data collection, collation and distribtion, from the field to the nderwriter and to prescribe and implement a remedy becase of the long vale chain in microinsrance, which often reqires agreement from several partners (especially in the instance of often complex PPPs). In this sitation appropriately strctred reinsrance can ensre the financial srvival of a microinsrer or a particlar microinsrance programme. 28 While reinsrers may have the necessary technical design knowledge, in microinsrance it sally makes more sense to determine what clients want with market research rather than imposing prodcts from the top down. 24

25 Generally speaking, reinsrance is not only sed for the prposes of risk transfer: sometimes the so-called intangibles of partnering with a reinsrer can weigh significantly on a cedant s decision to engage with it. Sffice it to say that a reinsrance transaction rarely flfills a nilateral prpose cross-fnctionality within single transactions is common and sometimes a combination of reinsrance and non-reinsrance (for example, conslting) arrangements is warranted, to achieve more complex or holistic aims. The core fnctions of microreinsrance are otlined in this section CAPACITY Reinsrance can enable insrers to write large individal risks, or portfolios of risk, which may be larger than they cold take on alone, owing to reglatory restrictions (for example, no individal risks greater than 10% of srpls allowed), rating reqirements (for example, premim-to-srpls or other financial leverage ratios) or capital constraints SURPLUS RELIEF One of the reasons insrers prchase reinsrance is to achieve a degree of srpls relief, most often via some form of pro rata protection. The motivation behind the prchase of pro rata reinsrance is generally tied to the internal cost of capital considerations and the strain which growth can place on internal capital ratios. For instance, a start-p microinsrer with a large distribtion channel (for example, a mobile network operator) and/or limited initial capital may find itself qickly on the wrong side of certain financial metrics that are sed as gages of insrance company solvency, sch as premim-to-srpls (a generally accepted benchmark for this ratio in non-life bsiness is 3 to 1). A QS reinsrance transaction with a well-rated and legally admitted conterparty 29 enables a cedant to access the balance sheet of a reinsrer and to limit the potential negative impact of new bsiness on its own finances. The case stdy of Orchard Insrance Grop in Box 3 provides a good example of the effective se of QS reinsrance by a microinsrer. Table 2. A basic balance sheet acconting scenario analysis demonstrating the ability of QS reinsrance to provide cedants with srpls relief (US$) QS reinsrance and capacity relief demonstration Assets Net premim Ceding commission income Other No reinsrance % QS reinsrance w/30% ceding commission Total assets Liabilities Net nearned premim reserve Other Total srpls Premim to srpls ratio 3:33 1:11 Difference in srpls Local reglations mst allow cedants to take credit for the cession of risk to a given conterparty in order for QS risk transfer to have the desired effect on the cedant s capital position. The acceptance of a given conterparty s credit will generally depend on its rating and reglatory stats (e.g. admitted/licensed, A+ or better, etc.). These reqirements vary by jrisdiction. 25

26 Many reinsrers prefer to take a pro rata position on new or small acconts since it increases their likely premim income more than XOL coverage alternatives. Moreover QS coverage increases reinsrers control of sbject bsiness by placing them in a first loss position and often making them the majority risk bearer with respect to the nderwriting reslts of a new reinsred book RESULTS STABILIZATION Reinsrance can be sed to stabilize nderwriting reslts. Reslts are at risk of instability either from acte risks, which manifest qickly and dramatically (see section 4.4 for more detail) or chronic risks affecting aggregate portfolio qality over time, sch as premim mispricing (a.k.a. the risk of error) or an nexpected shift in exogenos circmstances (a.k.a. the risk of change). Reinsrance can be sed to smooth nderwriting reslts by ctting off peak losses or diminishing the aggregate effect of a given portfolio s deterioration on overall company reslts. XOL (per occrrence or aggregate) is most commonly sed when a cedant wishes to assre its stakeholders of reslts within certain limits. In some circmstances QS and XOL reinsrance may be sed together to achieve the degree of stabilization desired; for instance, a QS coverage applying to a cedant s gross portfolio with an XOL protection applying to the cedant s net position CATASTROPHE PROTECTION Sstainable nderwriting reslts are characterized by a microinsrer s ability to manage the varios above-mentioned chronic risks, sometimes withot reinsrance. Bt even large homogeneos risk pools in stable reglatory environments with adeqate premim can face rinos acte losses when catastrophes hit. Earthqakes or floods of previosly nimaginable dimensions as well as intentional or nintentional manmade disasters can in one day case more insrance claims than in years of normal experience, and withot reinsrance some sch events cold wipe ot even the strongest insrance companies in the world. To some extent, this risk is diversifiable: disasters seldom affect all regions of a contry with the same strength, so a microinsrer who has a good geographic spread of bsiness may be able to compensate for catastrophic losses in the north with profits in the soth to some extent (and vice versa). However, a microinsrer operating in a small contry (for example, any of several dozen small island states) verss a large one (for example, India) has mch less diversification potential. Moreover profits in microinsrance are seldom large enogh to compensate for big regional losses, and most microinsrers have geographic concentrations sch is the general natre of microinsrance distribtion today in low-income commnities. Moreover, qantifying the benefit of diversification reqires expertise that not all single-contry or regional insrers, reglators and rating agencies have. It is in the catastrophe risk market that the expertise of reinsrance practitioners in decomposing and redistribting risk on many balance sheets has possibly the biggest impact. This is evidenced by recent catastrophic loss events in markets as diverse as Japan, Astralia, New Zealand, Chile and Thailand, where reinsrers paid significant portions of overall insred losses between 40 and 90 per cent for the events sampled (see Figre 9). While by no means a comprehensive stdy, this ratio implies a heavy reliance on reinsrance capital to spport the primary assmption of catastrophic risk in many developed and developing jrisdictions. 26

27 Figre 9. In recent years reinsrers have paid a high proportion of insred losses arising from major catastrophic events worldwide Billions of 2011 Dollars $40 $35 $30 $25 $20 $15 $10 $5 $ Japan Earthqake Tsnami (Mar 2011) 40% Reinsrance share of total insred loss 73% New Zealand Earthqake (Feb 2011) Reinsrer Share Primary Insrer Share (Sorce: Insrance Information Institte from reinsrance share percentages provided in RAA, ABIR and CEA press release, 13 Jan Economic loss figres from CRED EM-DAT.) Crrent catastrophe (re)insrance markets, however, do not yet cover low-income poplations extensively. This is primarily de to the fact that the vast majority of (potential) catastrophe victims in emerging markets remain n(der)insred. Somewhat paradoxically, the poor tend to live in contries exposed heavily to natral catastrophes, and are often forced to live in the most perilos areas in the least protected homes. Becase of this, as microinsrance penetration increases, so too will the exposre of insrers bottom lines to catastrophes. So not srprisingly, the first reinsrance cover many microinsrance programmes desire is catastrophe XOL. 60% Thailand Floods (Ag-Nov 2011) 95% Chile Earthqake (Feb 2010) 44% Astralia Cyclone/Floods (Jan-Feb 2011) 4.5. MARKET ENTRANCE OR EXIT This fnction of reinsrance is intimately tied with the other fnctions discssed in this section, since, for example, a reinsrance transaction spporting market entrance or exit will often provide capacity relief or be copled with essential nderwriting expertise (see section 4.6). As mentioned in the discssion of risk of error above, reinsrers may be better placed than local insrers to develop and deploy certain new prodcts since they may have more experience in a specific class of bsiness gained from global operations or more resorces for research-and-development-type activities. On the side of market exit, reinsrers often have expertise either in claims management or bsiness processes to spport the more efficient rn-off or management of existing portfolios. 30 Reinsrance in general plays an important role in filling market gaps and developing insrance markets in emerging economies. Knowledge is transferred to local insrers throgh their engagement with international reinsrance companies and can go as far as reinsrance companies taking a leading role on the primary side (inclding throgh prodct development or policy form drafting, for example). Cedants willingness to enter the microinsrance market beyond standard life or accident prodcts is often only reached once the availability of reinsrance has been confirmed. Moreover, many insrance markets reqire certain amonts of reinsrance coverage by law. 30 Reinsrers will sometimes acqire closed books of long-tail bsiness throgh transactions known as loss portfolio transfers or LPTs. In these sitations the reinsrer assmes the reserves associated with a book of bsiness where claims have already been incrred and bets effectively on its ability to administer these claims more efficiently or aggressively than the ceding insrer cold. Exit can also be achieved throgh novation of an active book of bsiness, whereby the assming party (sally another insrer in this case) acqires all the rights associated with the active administration of the ceding party s bsiness, in the hopes of growing the book. 27

28 4.6. EXPERTISE In addition to the five main fnctions above, reinsrance flfills a sixth, often nsng, role in the form of spplying intellectal capital or expertise. Reinsrers hone valable skills and bild p associated services by transacting bsiness with mltiple clients in many contries. This gives them access to invalable data and experience which allow them to help (or reqire) local cedants to meet global best practices in their nderwriting and risk management procedres. They can offer this expertise to insrers to assist them with entering (or exiting) lines or classes of bsiness, sch as microinsrance, by spporting the development of new prodcts. 5 MICROREINSURANCE SUPPLY AND DEMAND Every peril that can be insred can be reinsred. Matre insrance markets have developed a considerable diversity of lines of bsiness, and the characteristics for insrability carry over to reinsrability. Not all perils demand the same degree of reinsrance, however. Large, stable portfolios of small risks with reasonably similar and small sms insred and/or broad geographic spread tend to need less risk transfer than, say, a small portfolio of highly concentrated and very large risks. The first portfolio description in the paragraph above represents a good portion of microinsrance written today, specifically the many credit life books written by and throgh MFIs. In markets where microfinance is prominent or well nderstood, local and mltinational insrers are generally sfficiently well eqipped to nderwrite and manage sch risk and ths do not have mch need for microreinsrance to spport engagement in this line. Credit life portfolios are also not generally exposed to covariant risk to the same extent as other microinsrance prodcts, which diminishes the demand for a key form of microreinsrance protection: catastrophe or per occrrence XOL. On the other hand some credit life portfolios might be nable to access reinsrance spport if needed, owing to important spply-side considerations, sch as a lack of scale or scalability. As shold by now be apparent, reinsrers have a variety of different preferences. Some apply to reinsrers in general. For example, most reinsrers prefer to be long-term partners when spporting the development of new lines of bsiness sch as microinsrance, so that their initial investment can be amortized and early experimentation (with highly ncertain reslts) paid for. They also like to see that a programme has the potential to reach scale or critical mass to spport sch amortization. And they prefer to do bsiness in contries with predictable political systems and rle of law. Some preferences, on the other hand, are specific to a particlar reinsrer, sch as a focs on life insrance or on very large, very rare events (for example, catastrophic risk). Some preferences are reglarly pdated according to the reinsrer s assessment of the market cycle. 31 Beyond credit life bsiness, the market for microreinsrance is not easy to nderstand or navigate. Becase of this opacity, the development of a robst microreinsrance market is hampered, reslting in gaps in commercial spply of and demand for related prodcts and services. Overall the reasons for these gaps vary. For the prposes of this paper, spply and demand are assessed across three different basic dimensions: (1) Size and sophistication of cedant: Insrer size and sophistication are sally positively correlated, meaning that as the size of an insrer increases so too does its level of sophistication. As a cedant grows in size and sophistication, its need for microreinsrance and vale-added services tends to diminish. Paradoxically the willingness and ability of reinsrance conterparties to transact bsiness with them increases. (2) Size of programme: A lack of crrent or potential scale in individal microinsrance programmes is a hindrance to gaining effective commercial microreinsrance spport. Very few programmes, excepting those in large contries spported heavily by government sbsidy (which by some definitions woldn t even qalify as microinsrance bt rather as social insrance ), have reached sfficient scale to attract open market microreinsrance spport. 31 De mainly to relatively low barriers for market entry, reinsrance is considered a highly cyclical bsiness: after a year of very bad insrance reslts (following, for example, a major natral catastrophe) the demand for reinsrance grows in line with the capital depletion of insrers, driving p reinsrance premim rates (the start of a hard market ) and attracting new reinsrers. This reactive srge of spply starts eroding premim margins, leading to a soft market which eventally depletes reinsrers collective capital levels, forcing some to exit ntil another exogenos event starts the cycle over again. For a graphic view of the cyclical natre of the global reinsrance market, see 28

29 (3) Line of bsiness: This is what might be referred to as the microinsrance complexity paradigm (or possibly paradox): as microinsrance prodcts increase in complexity (for example, move down the continm towards index-based and health prodcts), the amont of vale reinsrers can add to a given programme increases. Ironically, since sch prodcts reqire greater investments in time and resorce to develop and nderwrite and tend to be more difficlt to distribte and administer (and therefore scale), reinsrers general willingness and ability to provide risk financing spport is inversely related to their ability to add vale when evalated on a prely commercial basis. Figre 10 illstrates how these three dimensions relate generally to demand for and spply of microreinsrance. While there are some nances to any given sitation, and while the diagram is far from scientific, significant gaps clearly persist in the marketplace at present. Based on this illstrative assessment, the strongest indicator of market flidity is the size of a given programme, followed by the complexity of the prodct in qestion. The size of the cedant matters less thogh is still a factor. These findings and others will be discssed in more detail in sections 5.1 and 5.2. Figre 10. Assessing commercial spply of and demand for microreinsrance Sch assessments for new and/or complex prodcts like microreinsrance are challenging. This diagram attempts to distill the reslts of research condcted for this report, as well as the athor s experience, to determine where market gaps exist. The green, yellow and red icons in the third section respectively represent either excessive, adeqate or insfficient spply to meet demand. (Sorce: Athor, qalitative research). Briefly to synthesize the findings from the analysis presented in Figre 10, reinsrance is most in demand from small cedants, with small programmes in relatively complex and catastrophe-prone lines of bsiness. Spply on the other hand is most readily available for large cedants, with large programmes and in relatively simple lines of bsiness. Donors have been instrmental in bridging spply demand gaps with grant fnding and some reinsrers have taken it pon themselves to sbsidize their involvement in this brgeoning market sector. Until property, agricltre and health programmes begin to scale p, by and large these gaps will persist. 29

30 5.1. SUPPLY While there are several notable exceptions (many of them featred in this paper), and while interest appears to be high, there is a general lack of action among commercial reinsrers regarding microinsrance bsiness despite its apparent market potential. This is srprising since microinsrance wold seem a very compelling opportnity for commercial reinsrers in the aggregate (see section 5.2 on demand). According to research condcted by the athor, of the top ten global reinsrers in the microinsrance market (see Table 1) only three Mnich Re, Swiss Re and Hannover are appreciably involved in spporting transactions at present. Others have dipped their toe in the water (for example, Lloyd s), have taken a prely passive role (for example, SCOR) or have already come and gone (for example, PartnerRe). Some smaller reinsrers (for example, Achmea Re) have a notable presence and one reinsrance broker (Gy Carpenter) contines to be involved thogh overall the reinsrance market s role in spporting the hndreds of microinsrance programmes in existence is limited. Of these reinsrers it seems only Swiss Re and Hannover approach the microinsrance market on a largely commercial basis. Others, notably Mnich Re and Achmea, rn their microinsrance activities alongside or within their sister fondations and categorize most of their related corporate activities as corporate social responsibility (CSR) as opposed to a bsiness line (see the appendix for a more detailed smmary of reinsrance organizations active in microinsrance). Despite the potentially large volme of new bsiness for reinsrers, in its crrent form the microinsrance market is diffse, and individal programmes (otside India and China, and not heavily sbsidized by governments) are generally qite small. This can make it difficlt for reinsrers to jstify the investment of time and resorce necessary to nderwrite innovative new programmes in new territories, which in microinsrance tend to have a limited history and limited potential to reach scale in the short term, and operate in relatively small markets. Another concern cited by certain reinsrers for their hesitance is the potential inverse relationship between the demand for microreinsrance and the size of a given programme. As microinsrance portfolios grow to their anticipated scale, the need/demand for secondary capacity may diminish as the risk in the programme becomes more predictable by virte of the law of large nmbers and the insrer becomes more savvy. In this scenario reinsrers fear that whatever intellectal capital they have invested in the programme s growth at or near inception will have been absorbed by the insrer, along with the majority or all of the nderwriting risk, leaving the reinsrer with nothing to show for its investment in the long term. Taking the above challenges into consideration, a given microinsrance prodct or project is sbject to evalation by a potential microreinsrance partner from varios perspectives, ranging from the prely commercial to the prely social. Some reinsrers will take a prely commercial view. In this instance an opportnity is deemed by the nderwriter to be either commercially viable or not. The criteria for sch an assessment can be complex thogh ptting it in simple terms, a reinsrer evalating microreinsrance from a short-term commercial perspective will expect a payback on its initial investment to fall within overall corporate profitability expectations. While there may be some degree of leniency towards microinsrers regarding profitability standards (for example, as compared to, say, Florida property catastrophe bsiness) and perhaps an allowance made for the innovative natre of microinsrance and the benefit of diversification that it wold bring, if a reinsrer does not expect a retrn on its initial investment within a reasonable timeframe (considering ongoing administrative and capital charges) and expects in the interim to be spporting a project far afield and on an ncertain path towards scale and sstainability, it is nlikely to elect to provide spport. Since microinsrance is still very mch an evolving field and individal programmes have rarely achieved scale withot ample government intervention and/or sbsidy, most programmes remain qite small and will stay as sch for the foreseeable ftre, sbject to improvements in ntried and ntested bsiness models, cltral shifts and/ or the development of innovative high-volme delivery channels. 30

31 Given this sitation, approaching microreinsrance in general is a difficlt proposition from a prely commercial angle. This is frther the case since there is very little precedent for (or at least very little pblic disclosre of) the profitability of existing microinsrance bsiness, even programmes which have scaled. This makes it impossible for reinsrers (and other wold-be capital providers) to make an informed microreinsrance investment decision based pon traditional investment criteria. In frther digesting this point, it is instrctive to think of reinsrance as an alternative sorce of capital sed by insrance companies to complement debt and eqity. From this angle reinsrers cold then be considered a type of investor. It therefore may be sefl in the recent historical context of the microinsrance market s development to analyse briefly where microinsrance investments have been made to date. The following is a non-exclsive list of some recent notable transactions: Sweden-based BIMA, a specialist mobile insrance technical service provider, received an investment of US$7 million in UK-based MicroEnsre, a specialist microinsrance administrator and intermediary, received an investment in 2012 after pioneering mobile microinsrance in Ghana and expanding to other markets. 33 Maritis-based MFS Africa, a specialist mobile money and vale-added service provider, received an investment of US$2.5 million in 2011 and is in the process of raising an investment of a frther US$10 million. As is clear, the above transactions all centre on the promise of distribtion by mobile network operators. While this sb-set of the overall microinsrance market appears to be in the process of becoming commercially viable, secondary risk transfer opportnities therein are qite limited since the prodcts presently being distribted via mobile phones (with the notable exception of the Kilimo Salama programme operated by Syngenta and Safaricom in Kenya, which is spported by Swiss Re) tend to be homogenos short-term life prodcts with relatively low catastrophic loss potential. While it may be impossible for the time being to jstify an investment in microreinsrance throgh the promise of meaningfl short-term nderwriting income, some reinsrance organizations have been able to profit from their expertise throgh conslting. The availability of fnds to spport sch conslting work can serve to nlock corporate resorces for specific projects, thogh very rarely does the conslting income alone meet the profit expectations of reinsrers. For all of these reasons it is nlikely that a reinsrer will enter the microreinsrance market withot reqiring at least some potential long-term gains and/or non-commercial benefits. With a few notable exceptions, many reinsrers are having considerable difficlty sccessflly straddling the divide between traditional (re)insrance and micro(re)insrance and finding a way to profit from micro(re)insrance in the long term hence certain market exits (see the appendix for details of the activity of specific reinsrers). This in trn reslts in prodct/service access difficlties on the demand side. By and large there is a lack of creativity regarding bsiness models in microreinsrance regarding both the programmes already in existence and those being developed, as well as a general nwillingness to forego a short-term retrn in favor of investment in a long-term prospect. Withot some sort of bsiness model innovation, the only ways for commercial reinsrers to write microreinsrance sccessflly are by targeting large programmes, of which there are very few

32 On the other hand, by entering into PPPs (which sally reqire significant staff time and entail some form of in-kind sbsidy for technical assistance) or by simply agreeing to forego short-term retrns in favor of long-term and intangible gains, reinsrers can achieve market share, gain a first-mover advantage and learn a great deal abot an emerging market segment. For some reinsrers the pll of these potential benefits have proven so compelling that they have engaged in the market pro bono, with CSR goals acting as a primary driver. 34 There are of corse opportnity costs associated with any sch investment and these will need to be evalated on an individal basis DEMAND As regards reinsrance, there appears to be a general lack of knowledge of good practices and available secondary risk transfer options among microinsrance practitioners. This leads to low awareness srronding the potential benefits both tangible (financial) and intangible (non-financial) associated with microreinsrance, ths hindering demand ndly. It also impedes active and prospective programmes from engaging with reinsrers appropriately, if at all. Demand is also sppressed by the bsiness mix in microinsrance, which again is dominated by credit life, where large amonts of reinsrance are simply nnecessary becase of the natre of the bsiness. On the other hand demand for reinsrance is high in the agricltre or natral disaster realms, thogh the overall volme of sch bsiness is limited. For those nascent programmes where a need for microreinsrance prodcts or services has been acknowledged, in order to access sch spport it is essential for the programme manager to develop a convincing bsiness model and growth plan. In the absence of obvios short-term commercial growth prospects, government premim spport or a commitment from a donor, reinsrers are nlikely to be interested in spporting the project. If growth prospects are long term then it may be essential for a pilot programme to establish a PPP with a donor (if sch a relationship does not already exist) to spport bsiness development and programme expansion, in order to attract commercial reinsrer involvement. Overall, if demand for microreinsrance eventally catches p with demand for traditional reinsrance, the market shold become mch more flid than it is today. The market s potential size is estimated in Box 6. Box 6: Total global microreinsrance market potential: A demand perspective With over 1 billion lives expected to be covered by and US$40 billion in potential ltimate direct written premims 2 worldwide, the market for microreinsrance and related services is potentially qite large. To contextalize the potential opportnity for reinsrers in microinsrance, the following is an estimate of the potential size of the global microreinsrance market. Reinsrance sage varies across contries and regions depending on diverse factors inclding reglations, cltral tendencies and market size. One common way to measre sage of reinsrance in a given market or by a certain company is to calclate the related cession ratio (also sometimes referred to as the reinsrance ratio ), which is the total amont of reinsrance premims ceded to reinsrers over the total amont of insrance premim written. While this is a coarse measre, distorted in some instances by large or niqe reinsrance transactions, legal reqirements and other factors, when it is calclated over a wide enogh area, the effects of otliers and market distortions shold be more or less netralized. 1 ibid., footnote 7. 2 Referring respectively to the ILO s estimate of global microinsrance policy issance circa April 2012 ( and Swiss Re s estimate of matre microinsrance market size circa 2010 ( 34 Achmea and Mnich Re are perhaps the best-known proponents of this approach. See Paginas/defalt.aspx and 32

33 Figre 11 shows weighted average reinsrance cession ratios for select regions worldwide based on an index of component contries compiled and selected by the athor. As is evident from the variance across regions or grops of contries, the sage of reinsrance depends pon the economic stats of a given jrisdiction, the political regime, the geographic particlarities and the risk-taking cltre. This regional variance is most manifest in the BRC grop, where the cession ratio is mch lower than might normally be expected of middle-market contries exhibiting significant economic growth. Diving into this nmber a bit deeper, the standalone cession ratio for China over the time period srveyed is 6.5 per cent, compared to 14 per cent and 16 per cent in Brazil and Rssia respectively. One reason for China s low sage of reinsrance may be the dominance of large state-owned insrers on the local primary market, which generally have less need for secondary sorces of risk capital becase of their large balance sheets and implicit (or explicit) sovereign garantees. Brazil s cession ratio is also somewhat low compared to that of other non-developed markets. This may have something to do with the geographic characteristics of the market Brazil does not have a high incidence of catastrophic risks, thereby diminishing demand for catastrophe XOL reinsrance, a key non-life reinsrance prodct. Figre 11. Reinsrance cession rates for select regions worldwide sing non-life market statistics (Sorce: A.M. Best) 33

34 Taking all of the above into consideration, generally speaking reinsrance sage tends to be higher in developing markets than in developed ones. 35 This reflects the relatively fast growth of developing market economies. Sch growth drives demand for insrance, often otpacing the growth in the necessary capital base of local insrers and creating a need for srpls relief from external capital providers. This need is then often componded by the high catastrophe exposre seen in many sch contries. From Figre 11 it is also possible to see the downward trend in reinsrance sage that has taken place over time on a more or less worldwide basis. This may reflect the relative strengthening of insrer balance sheets over the same time period, thogh regardless of case, decreased sage of reinsrance overall indicates an increasing pressre on reinsrers top-line growth. Only in spporting the development of new markets and prodcts like microinsrance will reinsrers be able to grow their existing bsiness meaningflly. In the context of microinsrance it is nclear whether microreinsrance will be sed to the same extent as traditional reinsrance. However, in the absence of a better gide it is fair to assme crrent traditional reinsring trends wold persist and apply in this emerging bsiness class. In this instance, taking a range of potential ltimate cession ratios of 15 to 30 per cent, the total potential market for microreinsrance is in the realm of US$6 billion to US$12 billion worldwide, based on the primary microinsrance market estimate of US$40 billion promlgated in 2010 by Swiss Re and cited above. 35 Increased sage does not eqate to a larger market. While on a relative basis emerging market insrers may cede more risk than developed market insrers the overall premim volme in emerging markets is mch lower than in developed markets. 34

35 6 CONCLUSION THE FUTURE OF MICROREINSURANCE Treaty portfolio nderwriting principles prevail. The need for facltative reinsrance in microinsrance is minimal since many of the hallmarks of typical microinsrance programmes lots of small policies with similar characteristics lend themselves well to treaty reinsrance nderwriting techniqes. By virte of these same hallmarks certain treaty coverage strctres are largely irrelevant in the microinsrance market. A good example is per risk XOL. Since microinsrance policies are generally limited to the same or a similar small size, there is generally no need to protect against large losses arising from single risks. Microreinsrance vale chains are complex. Often microinsrance programmes are lanched and fostered with donor assistance and via PPPs. This adds a nmber of parties who are external to the central flow of risk transfer and who have alternative vested interests in project otcomes (for example, development impact) that can be very different from the typical retrn-on-investment criteria sed by commercial reinsrers to evalate transactions. Reinsrers may also be srprised by the extra effort reqired of them in locating and training adeqate delivery channels, which are essential to sccessfl programme design thogh they may initially be new to insrance altogether. Reinsrance can address the risk of random flctation, the risk of error and the risk of change. It can also provide insrers with extra capacity, srpls relief, reslts stabilization, catastrophe protection and spport for market entry or exit. However, despite this litany of balance sheet benefits, the hottest reinsrer commodity in the microinsrance market appears to be their general depth of expertise in relation to the management of key risks, especially natral disasters. Since demand for microreinsrance services revolves primarily arond the delivery of reinsrer expertise as opposed to capital this has created the beginnings of a market for reinsrance practitioners qa consltants to advise on programme/prodct development, thogh it has not yet led to a vibrant market for microreinsrance transactions. Behind expertise, demand for catastrophic microreinsrance protection follows. However, sch protections are hard to come by on the open market since most catastrophe-exposed property or agricltre microinsrance programmes remain small and ths well below the commercial radar of most reinsrers. Withot donor involvement or government sbsidy to promote programme lanch and scale, reinsrer involvement is generally not possible. In the absence of programme-specific coverage another option to promote reinsrer involvement wold be to develop new bsiness models or risk-sharing mechanisms which allow for the artificial achievement of scale and to create an incentive-based form for the sharing of best practices. Specifically, the pooling of microinsrance programmes within or between nations offers a compelling opportnity for microinsrers to engage the reinsrance commnity and access reinsrance capital for the first time or more efficiently. The ICMIF s LARG facility is an excellent example. Its individally small members have sccessflly achieved a level of scale and coordination that is attractive to the reinsrance market by banding together across national bondaries and sing the law of large nmbers and geographic diversification to their collective advantage. This solidarity has also led to the important development of best practices in nderwriting, which increases the homogeneity of their collective risk profile, making it more palatable and probably more predictable for secondary nderwriting partners. This type of arrangement offers excellent lessons and promise for the microinsrance commnity. With greater collaboration at the regional, continental or even global level, individal microinsrance programmes cold more easily obtain competitive microreinsrance cover by pooling their liabilities. Mltinational microinsrance organizations, sch as the Microinsrance Network, the Access to Insrance Initiative or certain development banks might prove excellent conveners/administrators for sch a facility given appropriate resorces. Alternatively a professional reinsrance organization with global reach and spport from a donor might be able to pll something like this together. 35

36 While the volme of microinsrance bsiness worldwide is appreciable and expected to eclipse 1 billion lives covered by the end of this decade, 36 it is highly dispersed throghot the global market, with very few individal programmes having achieved scale. Given this market dynamic, very few reinsrers have developed a commercial bsiness model that allows them to nderwrite what amonts to a nmber of small and diverse microinsrance programmes on a worldwide basis. Swiss Re, the reinsrer with probably the largest share of the natral catastrophe microreinsrance market, appears to nderwrite sch programmes on a commercial basis thogh it is nclear if this bsiness has been profitable for the company. Hannover, the reinsrer with probably the largest share of the life and health microinsrance market, has focsed primarily on large so-called social insrance schemes operated with heavy government sbsidies and mainly in India, and again profitability is nclear. The potential ltimate size of the global microreinsrance market is estimated somewhere between US$6 billion and US$12 billion in ceded premim. An ntapped market of this size warrants significant attention from the reinsrance commnity, given downward global trends in reinsrance prchasing and other pressres on top-line growth. Yet, while the global reinsrance indstry is flsh with capital, the spply of microreinsrance is qite limited and the cover that is provided is concentrated among jst a few players. Spply of microreinsrance is constrained by a variety of market forces thogh perhaps most notably a lack of bsiness model creativity on the part of reinsrers. In the same way that primary microinsrance bsiness development reqires innovative approaches, so too does bsiness development in the secondary market. This is often left nacknowledged microreinsrance is a specialist field and shold be treated as sch. The case stdies in this report srveying varios existing or discontined microreinsrance programmes reveal some interesting insights into the potential ftre of microreinsrance. Several index-based insrance programmes have pshed or are pshing the frontier of accepted standards of insrance delivery and are developing innovative ways of addressing the basis risk dilemma. The se of mobile network operators as distribtion channels is showing very positive signs of enabling certain lines of bsiness sch as short-term life to scale and achieve sstainability. Before long in certain pockets of the microinsrance market the gap between spply of and demand for microreinsrance prodcts and services will close and first-movers will be best positioned to capitalize on the growth of this new market. 36 Ibid, footnote 7 36

37 7 REFERENCES A.M. Best Best s Special Report: Global Reinsrance Segment Review, Agst (Oldwick, A.M. Best). ASEAN Advancing disaster risk financing and insrance in ASEAN member states: Framework and options for implementation. Volme 2: Technical appendices. Available at: Appendices_Jne2012.pdf. Bernhardt, A.; Yong, S Catastrophes and climate change: Innovations in index-based micro(re) insrance, Rio+Soltions, 30 May. Available at: Carter, M.R Innovations for managing basis risk nder index insrance for small farm agricltre, FERDI Policy Brief B41 (Clermont-Ferrand, Fondation por les étdes et recherches sr le développement international). Chrchill, C.; Matl, M. (eds) Protecting the poor: A microinsrance compendim. Volme II (Geneva, ILO and Mnich Re Fondation). Chrchill, C.; Dalal, A.; Ling, J Pathways towards greater impact: Better microinsrance models, prodct and processes for MFIs, Microinsrance Paper No. 18 (Geneva, ILO). Conning w/ A.M. Best data Property-Casalty Reinsrance Distribtion Year-End 2013 (Hartford, CT). Conning Insrance-linked secrities: As alternative risk transfer goes mainstream, the reinsrer bsiness model is changing (Hartford, CT). IAIS and CGAP Working Grop on Microinsrance Isses in reglation and spervision of microinsrance (Basel, International Association of Insrance Spervisors). ILO Annal report 2011 (Geneva, ILO Impact Insrance Facility). Lloyds; Microinsrance Centre Insrance in developing contries: Exploring opportnities in microinsrance, Lloyds 360 Risk Insight (London, Lloyds). Microinsrance Network Microinsrance: 1 billion covered by 2020, 30 Janary. Available at: by_2020.pdf Mnich Re Protecting cooperatives and their low-income members in the Philippines against extreme weather events throgh microinsrance, Press release, 11 October (Mnich). Available at: Swiss Re The Essential Gide to Reinsrance (Zrich). 37

38 8 APPENDIX Table A1. Reinsrance organizations active in microreinsrance Reinsrance organization Description of main microreinsrance activities to date Crrent stats Mnich Re Fonder of the Mnich Re Fondation, an important micro(re)insrance research and development organization; fonder of the Mnich Climate Insrance Initiative (MCII); development and reinsrance spport of select weather index microinsrance pilots in varios jrisdictions Some ongoing reinsrance spport of active microinsrance programmes thogh the blk of Mnich Re s activity takes place at its sister fondation Swiss Re Dedicated microinsrance professionals within Global Partnerships division; significant involvement in index-based weather insrance projects worldwide throgh Swiss Re Grop s Reinsrance and Corporate Soltions divisions; retained as technical spporter to IFC Global Index Insrance Facility (GIIF) project management nit; reinsrance spport for varios life, accident and health microinsrance schemes worldwide. Investment in LeapFrog s second fnd. Annonced several microinsrance commitments at Clinton Global Initiative with partners. Headline-grabbing SIGMA microinsrance research Market leader in index-based micro(re)insrance programme development and risk-bearing spport; involvement in or fostering of PPPs with partners inclding mltilaterals, donors, MFIs and NGOs; engages with a commercial approach related to the development of mltiline microinsrance programmes Hannover Reinsrance of pblic health microinsrance programmes, predominantly in India (e.g. RSBY); co-development and reinsrance spport of weather-index pilots in varios jrisdictions; fonding investment in PlaNet Garantee, a micro(re)insrance intermediary and lead spporter of PlaNet-Garantee-led microreinsrance pool Focs on social health insrance in Asia, particlarly India; diverse book of bsiness in Africa (mostly francophone contries); se PlaNet Garantee and other brokers as main bsiness development channel Achmea Re Lloyd s (Liberty Syndicates) Achmea Reinsrance provides reinsrance coverage for the projects managed by the Achmea Insrance for Development department; since the start of these projects (in 2004) Achmea Re provided reinsrance coverage for projects in contries sch as India, Senegal, Nepal, Sri Lanka and the Philippines. Crrently, Achmea Re provides protection for livestock insrance, health insrance, crop insrance and life insrance. Achmea Re performs its microinsrance work nder Achmea s CSR banner and will sometimes not charge reinsrance premim for the first 3 years of a new programme Weather index deal in Philippines Actively spporting a diverse array of prodcts on a CSR basis Unknown 38

39 Reinsrance organization Description of main microreinsrance activities to date Crrent stats Allianz Re RIICE RIICE project is ongoing. No other active projects to speak of PartnerRe Several weather-index pilot projects with World Bank No longer active in microinsrance SCOR Flagstone Re (since prchased by Valids) Caisse Centrale de Réassrance Mapfre Re PlaNet Garantee Gy Carpenter & Company, LLC Investment in LeapFrog s first fnd; some spport for index-based insrance in India Spported at least one social insrance transaction ot of India; investment in LeapFrog s first fnd Following market spporter of PlaNet Garantee credit life reinsrance pool Following market spporter of PlaNet Garantee credit life reinsrance pool Administrator of microreinsrance pool led by Hannover Global reinsrance intermediary with a division dedicated to micro(re)insrance called GC Micro Risk Soltions ; mltiline expertise thogh with a focs on index-based insrance prodct development and micro(re)insrance vale chain development; reinsrance broker for several established microinsrance programmes Passive involvement in microinsrance via LeapFrog transitioning to more active involvement in index insrance Passive involvement in microinsrance via LeapFrog Unknown Unknown Involved in the placement of microreinsrance to spport local microinsrance programmes Involved as a consltant in the development of greenfield donor-fnded projects; reinsrance broker for microinsrance programmes Sorce: Internet research by athor. This is a representative, non-exclsive list; the athor makes no warranty as to its completeness. 39

40 IMPACT INSURANCE FACILITY Hosed at the International Labor Organization, the Impact Insrance Facility enables the insrance indstry, governments, and their partners to realise the potential of insrance for social and economic development. The Facility was lanched in 2008 with generos spport from the Bill & Melinda Gates Fondation, and has received sbseqent fnding from several donors, inclding the Z Zrich Fondation, Mnich Re Fondation, the IFC, USAID and AsAID. Facility, Lindmeier impactinsrance@ilo.org