60456 EAP DRM KnowledgeNotes Working Paper Series No. 23 Disaster risk Management in east asia and the Pacific Disaster risk Financing: case stuDies By Sandeep Poundrik The accelerated frequency and severity of disasters has made disaster risk financing integral to the ability of a country to compete in the global market. Without ad- equate planning and preparation, post disaster financ- ing needs are often met by a diversion of funds from development budgets or by taking expensive loans, both of which harm the long term growth prospects of a country. The extant literature on disaster risk financing has created a body of good practices and indicated that geo-political realities are fundamental elements of good design. Financing risk, therefore, requires a process of evaluating the models and matching the elements best suited to the individual country. WHat is cOVereD in tHis nOte? In this note, the instruments supporting risk retention and risk financing at the national and regional/international levels will be the sole focus. Risk transfer, including insurance, is a very broad subject and requires a separate discus- sion; however, when a hybrid solution involves a transfer of risk or an instrument at the household or community level, a brief explanation of the instrument ensues. To this end, the note reviews examples generally considered to be good practices in the sector and seeks to elucidate well-regarded risk retention instruments and financing. concept of risk Layering The risk layering concept says that risk financing instruments should be selected on the basis of the frequency and severity of disasters. Risks with high frequency and low severity (e.g., floods) can be self-financed by the insured party (government or affected populace). Disaster reserve funds or budgetary allocation would be appropriate instruments in this case. On the other hand, risks with low frequency and high severity are likely to cause extensive damage and should be transferred to better-equipped third parties. Catastrophe bonds and catastrophe insurance pools fall into this cat- egory. Based on the frequency and severity of the disaster, risk layering has been classified in a model with five groups.1 This working paper series is produced by the East Asia and Pacific Disaster Risk Management Team of the World Bank, with support from the Global Facility for Disaster Reduction and Recovery (GFDRR). The series is meant to provide just-in-time good practice examples and lessons learned from projects and programs related to aspects of disaster risk management. 2 Disaster risk Management in east asia and the Pacific risk Layering Pyramid Probability of Damage as risk retention: event or return Proportion of emergency/ category Period gDP Budgetary Funds 5% risk Group 1 or Up to 3% Financing: Up to 20 years emergency/ contingent 3.33% Loans Group 2 or Up to 5% 20-30 years reinsurance: insurance Pools 1% Group 3 or Above 5% 30-100 years 0.5% catastrophe Bonds Group 4 or Above 5% 100-200 years Below 0.5% grants/aid: international Group 5 or Above 5% Organizations, Other countries Above 200 years Risk retention: Disasters in Groups 1 and 2 are high ages at the country level may not be feasible. Such risks frequency­low severity disasters, meaning that they are are best shared widely. Countries can cover these risks highly predictable and cause mild damage. Recurring by purchasing insurance from professionally managed floods and hurricanes in many countries can be put in insurance companies that diversify risk globally. One these categories. The risk transfer premium for these point is clear; countries should opt for insurance or risk types of disasters normally charged by the insurers is a transfer only when the expected losses are beyond their 1.5 to 2.5 risk­cost multiple (i.e., the ratio of the premi- means. In other cases the risk transfer costs may be very um payment for the insurance to the expected losses for high. the insurance policy period). Thus, it makes more sense Apart from the frequency and severity of disasters, other for the insured parties to retain these kinds of risks as factors have a bearing on the selection of instruments. the cost of insurance outweighs the benefits. Liquidity: While a country may be able to retain risk up to a certain threshold, problems of liquidity may intended audience arise after disasters due to concerted disbursements of The intended audience of this note is county and provincial funds. Risk financing or insurance are prudent choices governments, international agencies, and development orga- in this case, since these instruments are intended only to nizations, working in the disaster risk management field with cover the liquidity gap. an interest in risk financing strategies and instruments. Political economy: Risk retention instruments are not Risk pooling or financing: The risk of damages from immune to pressures from the political economy. For larger disasters (also in Group 1 &2) can be covered by example, the annual appropriation to a reserve fund pooling the risk within a group of countries or states comes under pressure when a surplus has built up for lack of disaster. Legislatures would readily divert the susceptible to disasters. A common Reserve Fund is es- appropriations to more visible schemes. Risk transfer tablished by the members, and then drawn upon when offers the government an attractive and pragmatic solu- affected by a disaster. tion: the government executes binding contracts with Risk transfer: Groups 3 to 5 are high severity­low insurance companies for a fixed number of years and frequency disasters, where provisions for possible dam the issue largely disappears from the legislative agenda. Disaster risk Financing: case studies 3 Procedural issues: How the funds flow takes through Borrowing capacity: Countries with a large debt bur- government system can also affect the choice of instru- den would be well advised It to choose risk transfer in- ments. Government financial rules and procedures in struments rather than risk financing instruments even many countries may not be conducive to a fast flow of for moderate frequency disasters. The reason being, funds. Budget allocations and reallocations normally their revenue resources may not be able to sustain the need legislative approval, which can be slow and drawn additional debt. out. Some reserve funds try to side step this requirement Another way of mapping risk financing tools is by so- by setting a provision for the accumulation of funds at cial/geographical relevance and financing modes. The the end of the financial year. It also matters how the tools for the national or regional levels are obviously government classifies funds received from international different from those for the individual household. The organizations; if they are treated as revenue receipts, the following table shows the tools useful for different lev- process for disbursal can be very time consuming. els of targeting. risk Layering 14 12 10 Losses in % 8 6 4 2 0 Risk Transfer 0 20 40 60 80 100 120 Probability of Occurance in Years Financing Modes savings credit investment insurance Poor Households Micro Finance Micro insurance small Businesses Levels Middle income households community social Funds catastrophe Pools national regional/ international Source: Pro Vention Consortium (2009). 4 Disaster risk Management in east asia and the Pacific natiOnaL Disaster FunD, MeXicO FONDEN covers the total cost of repair of federal public infrastructure; however, the state and municipal The 1985 earthquake in Mexico City killed 6,000 people, infrastructure is covered by co-financing, which ranges injured 30,000, left about 150,000 victims, and caused from 50:50 to 20:80 depending on the type of assets. direct economic losses to the tune of US$4 billion. As a States and municipalities are expected to cover their direct result, many new institutions were created to re- share from their own sources of funding. spond to disasters, including FONDEN in 1996. At the outset, FONDEN served as a budgetary tool to allocate funds on an annual basis through three separate reserve Funds funds (i) infrastructure fund to repair uninsured public Reserve funds are disaster risk financing instruments infrastructure, (ii) agriculture fund, to support low income that cover small and recurring losses caused due to natural disasters. Normally, they are funded from an- farmers, and (iii) assistance fund, to provide relief to vic- nual budget allocations. At the financial year end, re- tims of disasters. In 1999 a separate catastrophe fund was maining balances may accumulate as in the case of created within FONDEN to accumulate the unspent di- FONDEN (Mexico) or lapse like the National Calamity saster budget of each year. At present, FONDEN mainly Fund (Philippines). provides financial support to repair public infrastructure through the FONDEN program and assistance to low income households (e.g., shelters, food, primary health The repair work of federal infrastructure is supervised care) through FONDEN Prevention Fund. The third by the respective technical ministries, that is, monitor- instrument, FONDEN Trust Fund manages the assets ing the work and verifying the bills; contractors are paid of FONDEN, including its risk transfer strategy. directly by FONDEN. In the case of state/municipal infrastructure, FONDEN transfers funds to the con- In addition to the FONDEN trust at federal level, trusts cerned state fund only after the state/municipality has have been set up in all the 32 states of Mexico. While transferred its share of funds. Local departments over- the federal fund is supported by the federal government, see the execution but contractors are paid directly by the the state funds get support from the state governments local funds. as well as the federal fund. time for claim assessment by FOnDen A federal technical agency Technical agency provides Ministry of Interior is certifies the occurrence of the State Government with to ensure that: 1. the natural disaster and informs a technical and financial requested assistance State Governments evaluation is related to natural disaster; 2. the damaged infrastructure has not benefited from FONDEN on earlier occasions and if so, ask for proof of insurance; 3. formally approve co-financing of.. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 23 25 The State Government informs Federal Government. Ministry of Interior issues a declaration of state of disas- The outer time limit for ter and Ministry of Finance completing the procedure authorizes FONDEN to release partial contribution State Government sets up a technical committee to assess damages Source: Timeline made from data collected from World Bank note on FONDEN3 Disaster risk Financing: case studies 5 At present the federal government is trying to promote and the bulk of the funds were later provided from the insurance of federal and state infrastructure to decrease surplus in oil revenue. With the fall in oil production its dependence on FONDEN. FONDEN is also try- in Mexico and the downward trend of oil prices, this ing to transfer its risk by purchasing reinsurance and stream of funds has vanished. FONDEN is now expe- issuing catastrophe bonds. The Government of Mexico riencing a resource crunch and is exploring alternative has issued two catastrophe bonds: the first in 2006 for sources. A continuous revenue stream for the disaster US$160 million and the second in 2009 for US$290 fund, made legally binding in order to deflect political million.3 pressures, is a must. Another challenge is the time taken to complete the Political economy issues repair of infrastructure schemes. While a strict time- of risk Financing in FOnDen line of 23 days has been prescribed for the release of Why should the federal government rely on oil surplus funds, not all states and municipalities have the req- and not provide a regular budget allocation to FON- uisite capacity. Many schemes have been pending for DEN? The reason seems to be the difficulty in get- more than five years even though the completion time ting approval for the large fund allocation for disaster was one year. Ensuring a financing mechanism is neces- management. Since world oil prices and production in Mexico were generally above annual estimates, the sary but not sufficient: Executing capacity is required resultant annual revenue surplus was given to FON- to complement the funding provided by reserve funds. DEN. This bypassed discussion and approved in leg- islature, and FONDEN was received the funds neces- The existence of a national reserve fund and the pos- sary to keep it operational. sibility of getting money from it generate a tendency to exaggerate the scale of disaster from its actual lev- el. The pressure to declare disaster and release funds, strengths even when in actuality the natural disaster is below the threshold, is considerable. FONDEN relies exclusively The reserve funds provide the government a ready in- on the certification of federal agencies to consider as- strument to respond immediately to disasters. While sistance. But the states also cite lack of instruments in these funds are not meant as a solution to high sever- many places and, thus, a difficulty in identifying disas- ity disasters, they are normally the first line of response ters affecting small geographical areas. A transparent, against small and recurrent disasters like floods and credible, and reliable system to identify and declare the The main strength of the system is the fact that it hurricanes. occurrence of disaster is required, and devised to with- separates the disaster funds from normal budgetary stand political and popular pressures. operations and designates oversight of the funds to an Another question facing FONDEN is to what ex- tent it should finance projects. A large part of infra- institution with systems for rapid dispersal in times of structure damaged by disasters is old (e.g., buildings or disaster. bridges), so it make sense to improve the specifications A major positive aspect is the fact that any funds un- and design during repairs, a technique known as "build spent at the end of fiscal year do not revert to treasury. back better." Insurance companies do not finance the improvement even if the assets are insured. Currently, FONDEN is funding this component but the scarcity scope for improvement of funds has placed considerable pressure on it desist. The main problem faced by FONDEN today is a scar- The scope of risk financing needs to be clearly defined city of funds. FONDEN used to receive federal funds with guidelines on what will and will not be included. of about US$600­650 million annually, although not from annual budgetary allocation. Usually a notional al- location was made for FONDEN in the annual budget 6 Disaster risk Management in east asia and the Pacific POst Disaster risk Financing CNE is given five years to complete the reconstruc- MecHanisMs, cOsta rica tion phase using emergency procedures. In other words, emergency procedures cannot be used endlessly to by- The legislative framework for budget management and pass normal procedures. the institutional organization for the post disaster recov- ery process in Costa Rica have evolved over four decades. In 1969 Costa Rica established a mechanism for declar- national emergency and ing a state of emergency. It also established the National risk Prevention Law of 2005 Risk Prevention and Emergency Management Commis- The National Emergency and Risk Prevention Law of sion or CNE (Comision Nacional de Prevencion de Riesgos 2005 introduced the concept of "humanitarian as- y Atencion de Emergencias) and the National Emergency sistance." The law enables CNE to use the State of Exception without a declaration of emergency. CNE Fund or FNE (Fondo Nacional de Emergencia). Its most can acquire and deliver humanitarian assistance as recent reform was the National Emergency and Risk well as procure machinery to clean affected areas for Prevention Law of 2005, which addresses the problem of a maximum 100 hours. The mayor of an affected mu- disasters from a more comprehensive development point nicipality can contact CNE directly to request help. of view. The end result of this long-standing practice of continuously updating the framework is well-designed mechanisms to respond to disasters. institutional arrangement The institutional organization is covered by a national state of exception risk management system. CNE has the primary role in As per Article 180 of the Constitution, a State of Ex- the system and is responsible for planning, coordinat- ception can be declared for emergency situations, in- ing, managing, and supervising the response. This in- cluding public disasters, war, and civil unrest. The gov- cludes initial response to affected populations, damage ernment can issue a decree specifying the events that and needs assessment, formulation of recovery plans, have occurred, the general features of the crisis, and the coordination with other agencies, and procurement for area covered by the declaration. The main effect is that implementation of rehabilitation and reconstruction CNE can access the State of Exception, which enables projects. it to: As per the law, other institutions and local governments Receive special treatment in times of budget con- are obliged to act under the direction of CNE and to straints participate and support works assigned by CNE over Use fast-track procedures to procure goods and ser- their regular work. Public institutions are required to vices provide technical assistance for the assessment and im- Hire personnel by emergency appointments. plementation phases. The condition is that there should be direct causal re- CNE appoints Executing Units to implement the proj- lationship between these actions and the emergency, ects. These units are normally comprised of the depart- and they are subject to subsequent economic, legal, and ments with the technical competence to implement and fiscal controls. The State of Exception also increases monitor the projects. This measure ensures technical the government's powers to impose such restrictions supervision and consequent responsibility by the appro- as expropriations and demolitions temporarily. For ex- priate sector. ample, works executed under a State of Emergency are exempted the procedures and required clearances of en- vironmental laws. sources of Funding The law defines three phases of emergency manage- The National Emergency Fund (FNE) finances all the ment: response, rehabilitation, and reconstruction. activities related to prevention and mitigation, emer- Disaster risk Financing: case studies 7 gency management, rehabilitation, and reconstruction. Supplementary budget: All public entities are required The main sources of funding are: by law to contribute 3 percent of their profits/surplus to FNE at the end of the fiscal year. If an emergency Current budget: The current budget allocation for has been declared, public entities can transfer money FNE is generally meant for prevention activities to be to FNE for emergency operations. These public units carried out by CNE. The allocation ranged from US$4 are normally appointed as Executing Units (agencies to US$5 million until 2006 and was increased to over which actually carry out or implement the schemes) for US$18 million in 2007 to meet the growing responsi- the projects funded by these funds. The supplementary bilities of CNE. budget is the main source of funding for emergency op- erations. CNE Current Budget Allocation (2001­2007) 20 Transfer of Resources from Public Institutions and 18 Donations to the FNE by Emergency Decrees 16 14 50 42.3 43.1 US$ Million 12 40 10 US$ Million 8 30 6 20 4 10.6 10 2 0.75 0 0 2005 2006 2007 2008 2001 2001 2003 2004 2005 2006 2007 Years Years Source: World Bank note on FONDEN3 Allocation for risk management: A 2005 law requires budget, etc. and is prepared by CNE in collaboration with all public entities to make allocations for risk manage- the relevant public entities. This plan too is approved ment in their annual budgets. by the Board of Directors, which also approves the ap- pointment of an Executing Unit. Based on the approved Planning and investment instruments: There are two Investment Plan, CNE carries out the procurement of instruments for project planning and implementation: goods and services. After this, the Executing Unit takes the General Emergency Plan and the Investment Plan. . up the technical implementation of the project. The General Emergency Plan defines, prioritizes, and organizes the actions required to manage a specific cri- sis. It includes a sector-wide diagnosis of the situation, Delineation of responsibilities designates activities for the response, rehabilitation, and across agencies reconstruction phases, delineates institutional responsi- The friction between disaster management agencies bilities, and calculates the required funds as well as any and technical lines ministries over fund allocation and necessary additional resources. The Board of Directors control is a general problem. In Costa Rica, the Ex- of CNE approves and authorizes the implementation of ecuting Unit system and a provision of fund transfer to CNE seem to solve the problem to a large extent. A the plan within two months of declaring an emergency, mutually beneficial arrangement, CNE receives addi- and the plan is binding on local governments and public tional funds for disaster management activities while entities. the public entity gets access to simplified procure- ment procedures without losing technical control over The second instrument, the Investment Plan, is triggered the project. Generally, the public entity that transfers for each project proposed in the General Emergency funds to CNE, is appointed the Executing Unit of the corresponding projects. Plan. This document consists of all the technical details, 8 Disaster risk Management in east asia and the Pacific strengths Pan-african Disaster risk Pool for Food Clarity of roles and responsibilities: The legal frame- security work clearly defines the roles and responsibilities of This pool is still in the conceptual and discussion various public agencies, including CNE. The central stage but the idea is very appealing. Preliminary role of CNE is important for speedy procurement and findings indicate a 50 percent savings from diversi- design while technical supervision is provided by con- fication of drought-related losses across Africa. This means that if African countries were to pool their cerned departments/agencies. drought risk, the pool's capital requirement would be half the sum of each country creating its own reserves Coordination: The overall framework streamlines ­ making a Pan-African Disaster Risk Pool an attrac- the process for CNE and allows coordination across tive financing mechanism in support of African food all stakeholders. The fund flow mechanism described security. At present, the United Nations World Food earlier incentivizes the departments to coordinate with Program (WFP) is the primary responder to droughts in Africa in terms of food security. But the process of CNE, while CNE has the authority and mechanisms responding to such disasters is designed in such a to monitor implementation and ensure joint co-respon- way that the relief comes much later than required. For sibility. example, lack of rain in the October­April rainy sea- son clearly signifies impending disaster during next Strong credibility: Owing to its extensive experience, October to December, or "the hungry period." If steps CNE enjoys high technical credibility with the popula- are taken to ensure food security in April­May, the selling of assets and a decline in consumption can be tion as well as confidence in the transparency of its use avoided. Due to established procedures, WFP can- of reconstruction resources. not take proactive measures and can only carry out a needs assessment after the onset of drought in No- vember­December. The whole process (assessment, scope for improvement issue of an appeal, procurement, etc.) takes time, with delivery of food relief long after the actual requirement The main challenge with this system is that CNE may and at a much higher cost. be used just to avoid the normal safeguards like procure- Based on experience of Caribbean Catastrophe Risk ment or environmental safeguards. While the mutually Insurance Facility, the Pan-Africa Disaster Risk Pool beneficial arrangement of fund transfer and appointment would help prevent this delay and spread risk among countries. The idea is to take ex ante measures by of an executing agency may support coordination, it also having better early warning systems and making a gives rise to the possibility of use (or abuse) of emergency common contingency pool to quickly respond to provisions to bypass the safeguards. Establishing provi- droughts. The World Food Program has developed a methodology to assess the impact of weather events sions to prevent this possibility is a necessity. on food security across Africa: Africa RiskView is a The applicability of the State of Exception should be software platform that translates satellite-based rain- fall information into near real-time estimates of poten- as per requirement and a phased approach may be ap- tial emergency responses. The Africa Risk Pool would propriate. CNE can also balance its risk management be the instrument to provide ready liquidity should the strategy to put more emphasis on ex ante measures and early warning system predict a drought.13 transfer risk to the appropriate entities. tries can access funds from the facility if they declare a catastrOPHic risk DeFerreD state of emergency as a result of a natural disaster. DraWDOWn OPtiOn, WOrLD Bank Pre-approved, Cat DDO disburses quickly if and when The Catastrophic Risk Deferred Drawdown Option (Cat the borrowing government declares an emergency. The DDO) is a financial product offered to middle income loan amount is limited to US$500 million, or 0.25 per- countries by the World Bank. Its purpose is to make fi- cent of GDP (whichever is smaller), because Cat DDO nancing immediately available after a natural disaster. It provides short term liquidity (rather than reconstruc- is intended to fill the gap while other sources of funding, tion financing) following the disaster. It does not pre- such as emergency relief aid, are being mobilized. Coun- clude other borrowings. Disaster risk Financing: case studies 9 Cat DDO is available for three years and can be re- nonetheless, was disbursed by the World Bank, as soon newed up to four times. There is a single front-end fee as the fee was paid. As of December 2009, Cat DDOs of 0.5 percent of the approved amount, and each subse- had been approved for Costa Rica, Colombia, and Gua- quent renewal entails a fee of 0.25 percent. The interest temala.6 The CAT DDO proposal for Peru is in the is set at the IBRD rate prevailing when the funds are process of being approved. disbursed. The funds can be repaid at any time before the closing date, and the amount paid back would still be available for subsequent borrowing. Borrowers must, strengths however, have an adequate macroeconomic framework Cat DDO is a very cheap source of funds. In fact, it is in place when the loan is approved, and a disaster risk cheaper than the reserve funds, as the cost of maintain- management program monitored by the World Bank. ing reserve funds is higher due to high liquidity require- The world economic crisis in 2008 had an adverse im- ments. pact on the program because countries needed fund- Cat DDO ensures immediate liquidity in the after- ing for their regular programs. Given that Cat DDO math of disasters. The mutually agreed targets for di- comprises part of a country's overall credit limit (called saster mitigation measures between the government and exposure limit), using this facility would have reduced the World Bank work beneficially to encourage disaster the availability of regular funds. Thus, borrowers were preparedness and implementation of stronger financial not keen on availing this facility. With some positive mechanisms. Better disaster preparedness and planning change in the economic situation, the World Bank fore- also reduce the impact of disasters. sees one or two contracts in Year 2011. scope for improvement cat DDO The World Bank modified the Cat DDO to eliminate At present, this contingent credit line is only available the instrument's unintended implicit disincentives. for middle income countries. This facility could be es- Countries used neither the 2001 Development Policy tablished for low income countries. Not only would low Loan DDO nor the 2008 CAT DDO for two main rea- income countries be provided with an additional source sons. First, uncertainty surrounded the disbursement of funds, because concurrent Bank reviews were re- of liquidity but they also would be incentivized to better quired to drawdown funds. Second, DDO was more prepare for disasters. expensive than regular IBRD loans: a higher commit- ment fee was charged in addition to a surcharge for a A loan repayment holiday or loan forgiveness provi- longer maturity. To maximize the potential of the instru- sions would make the program more attractive. In ment, the Bank now monitors a borrower's economy the aftermath of a disaster, countries could be given the continuously to facilitate disbursement upon request. The funds may be drawn down at any time unless the benefits of either a repayment holiday or forgiveness for Bank gives prior notification to the borrower that one or a period of three to four years. The Cat DDO would act more of the drawdown conditions have not been met. as an additional liquidity source for affected countries. The second change aligned the pricing of the DDO with standard IBRD terms, thus eliminating the com- If the contingent line of credit is beyond a country's mitment fee and surcharge for the longer maturity.7 CAS envelope, countries will have more incentive to take advantage of the facility. Costa Rica is the second most exposed country to mul- tiple natural hazards (Natural Disaster Hotspots 2005) and was the first to have a Cat DDO approved. In Sep- tember 2008, US$65 million was allocated. Domestic politics delayed the payment of the front-end fee until a 6.2 magnitude earthquake struck on January 8, 2009, causing damage estimated at US$100 million. The loan 10 Disaster risk Management in east asia and the Pacific cariBBean catastrOPHe risk retains some of the risks transferred by the participating insurance FaciLitY countries through its own reserves and transfers some of the risks to reinsurance markets where this is cost- The Caribbean Catastrophe Risk Insurance Facility effective. This structure results in a particularly efficient (CCRIF) was established in 2007 basically to solve the risk-financing instrument that provides participating short term liquidity problems of Caribbean govern- countries with insurance policies at approximately half ments in the aftermath of disasters. It is an exempted the price they would obtain if they approached the re- company under Cayman Islands laws, holding an insur- insurance industry on their own (structure of CCRIF ance license, and is governed by a trust deed. CCRIF is and reduction in premium with increase in number of a joint reserve mechanism that provides participating participating countries shown in illustrations below). governments with coverage in case of disasters. Commercial insurance is available in the Caribbean re- Primarily an instrument to pool resources in order to gion, yet the total premium that businesses paid aver- buy parametric insurance, CCRIF also covers risk from aged about 1.5 percent of GDP between 1970 and 1999 its own reserves. It is being discussed here as an example while losses (insured and uninsured) amounted to only of risk pooling. To understand CCRIF, one could con- about 0.5 percent of GDP. This justifies the purchase sider a system through which several countries agree of parametric insurance jointly, where the cost comes to combine their emergency reserve funds into a com- down substantially. mon pool. If each individual country were to build up its own reserves to sustain a catastrophic event, the sum Donors provided US$67 million in start-up capital, of these country-specific reserves would be much larger and 16 member governments paid in US$22 million. than the actual needs of the pooled countries in a given Governments purchased parametric insurance, paying year. Considering that, on average, only one to three CCRIF about US$20 million in premiums for para- Caribbean countries is/are affected by a hurricane or an metric insurance coverage roughly totaling US$450 earthquake in any given year, a pool holding only the million. The Facility retains responsibility for the first reserves for three potential payouts should be sufficient US$20 million of payout (backed by its capital) and for for all pool participants. Each year, as the pool is depleted, participating countries would replenish it in Structure of the CCRIF proportion to their probable use. The CCRIF works in a similar manner by combining the benefits of pooled reserves from participating countries with the finan- cial capacity of the international financial markets. It Reinsurance/ART (Purchased on international financial markets) Advantages of Risk Pooling Country 1 100% Country 1 Premium (100% = one country) 90% Premium Payout 80% Country 1 Insurance 70% Premium 60% Country 1 Growth 50% 40% 30% Insurance Country 1 Payout 20% Reserves 10% 0% Country 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Number of participating countries Country 1 The greater the number of countries the lower the premium Source: Esquivel8 Disaster risk Financing: case studies 11 transfers the remaining exposure through reinsurance strengths and catastrophe swaps that the World Bank intermedi- ates. Donors expect the Facility's capital and reserves to CCRIF allows the participating countries to pool grow and, thus, to be self-sustaining. their risks and resources, creating a more diversified portfolio. While this increases the funds at the disposal Within two weeks of the November 2007 earthquake, of affected country as a retention mechanism, it also the most severe in the eastern Caribbean in 30 years, reduces the cost of insurance. In the case of CCRIF, the Facility paid about US$1 million to St. Lucia the premium is reduced by half as compared to what a and Dominica . It paid US$6.3 million to the Turks country would pay for individual insurance. and Caicos Islands after Hurricane Ike in September 2008.1,2,5 There have also been disasters that did not CCRIF is backed by donor funds held by the World trigger the parameters set: Hurricane Dean in 2007 Bank in a multi-donor trust fund. These additional caused considerable damage in Jamaica because of resources help reduce the CCRIF's dependence on the rain, but no payout occurred because wind speed was reinsurance market by increasing its own reserves. the parametric trigger. Similarly, the cumulative effect of the 2008 hurricanes in Haiti was devastating, but scope for improvement the winds were not strong enough to trigger a pay- out. The 7.0 magnitude earthquake that struck Haiti A review of parametric trigger mechanisms is needed. on January 12, 2010, was of sufficient magnitude to Non-payment of any damages in Hurricane Dean in Ja- trigger the full policy limit for Haiti`s earthquake cov- maica has shown that only wind speed measurements as erage purchased under the Facility. Based on calcula- triggering devices are not sufficient and more frequent tions from the preliminary earthquake location and events need to be covered giving wider coverage. magnitude data, Haiti will receive US$7.8 million, the The procedures for a multi-country fund need to be maximum payout under its earthquake policy. This is streamlined. Another important requirement in a multi- about 20 times its premium for earthquake coverage country fund is to have very clear and transparent proce- of US$385,500. Although shaking was felt in Jamaica, dures with high visibility and communication at all levels. another CCRIF-covered country, it was insufficient to This precludes any confusion in the aftermath of disasters generate any loss under the parametric index. as to whether a country could have received payments. cOncLusiOn Participating countries can decide on coverage as follows: The choice of instruments in disaster risk financing de- Perils Covered: Hurricane (protection against hur- pends on many factors. One way of looking at it is to ricane wind damage) and earthquake. classify the disasters in terms of their expected severity and frequency. More frequent disasters with low expect- Coverage Limit: The maximum possible payout for all claims in any one policy period, not to exceed ed severity (e.g., recurring floods) are better financed US$100 million and specified for each insured by retaining the risk, as the cost of transferring such peril. risk will be disproportionately high compared to the ex- Attachment Point: Level of government loss that pected damages or payments. On the other hand, risk triggers the policy, being the equivalent of a de- associated with low frequency­high severity disasters ductible. It can also be expressed in terms of the (e.g., major earthquakes) is best transferred to the in- return period of experiencing the level of loss or more. The Attachment Point cannot be set at loss ternational reinsurance market, as government may not levels with return periods of less than 15 years. (For have the capacity and resources to sustain the damages CCRIF insurance purposes, the only valid method caused by such disasters. to calculate the amount of government loss is the parametric loss model specified in the policy itself.) Another way of looking at the issue is to map the risk fi- nancing tools by social or geographical relevance. Tools 12 Disaster risk Management in east asia and the Pacific for the individual or household level naturally will differ 5. Ghesquiere, F. and F. Ramirez Cortes. 2008. Mechanisms for the Mobilization and Allocation of Financial Resources from those for national governments. in Post-Disaster Solutions in Costa Rica. Washington, DC: Taking in to consideration the scope of the topic, this World Bank. note focused only on risk retention and financing tools 6. Sanghi, A. et al. 2010. Unnatural Disasters, Natural Hazards: Effective Prevention through an Economic Lens. Washington, relevant for national governments or international bod- DC: World Bank. ies. The mode chosen was to study successful examples 7. Background Note on CAT DDO available at World Bank in the field and to assess their strengths and weaknesses. website To the extent that this review is fungible, the countries http://web.worldbank.org/WBSITE/EXTERNAL/NEWS/0 will benefit from mixing and matching the models and ,,contentMDK:21670477~pagePK:64257043~piPK:437376~t instruments to their specific geo-political context. Fi- heSitePK:4607,00.html nally, as the case studies in this note indicate, the uptake 8. Esquivel, M. Caribbean Catastrophe Risk Insurance Facility of risk retention and risk transfer mechanisms has made (CCRIF): A Joint Reserve Mechanism for Island States in the Caribbean. Washington, DC: World Bank. countries more resilient to natural disaster. 9. World Bank. 2008. A Review of CCRIF's Operation after Its First Season. Washington, DC. reFerences 10. World Bank. 2008. The Caribbean Catastrophe Risk Insurance Facility: Providing Immediate Funding after Disasters. 1. Dumitru, D. 2010. Mitigating the Adverse Impacts of Natural Washington, DC. Disasters on the Philippines Economy: A Study of Disaster 11. Cummins, J.D. and O. Mahul. 2008. Catastrophe Risk Risk Financing Options. Mimeo. Financing in Developing Countries: Principles for Public 2. World Bank. Catastrophe Risk Financing in LCR: Recent Intervention. Washington, DC: World Bank. Solutions and Challenges Ahead. IBRD Results Working 12. O'Donnell, I. 2009. Practice Review on Innovations in Paper. Washington, DC. Finance for Disaster Risk Management. Geneva: ProVention 3. World Bank. FONDEN: National Disaster Fund in Mexico. Consortium. Note Prepared by World Bank Team for Indonesia Mission. 13. Climate and Disaster Risk Solutions-Equipping Africa with Washington, DC. Technology and Tools to Manage Natural Disaster Risk, 4. Interview with Rubem Hofliger Topete, Director General, Rockfeller Foundation and UNWFP. FONDEN. Disaster risk Financing: case studies 13 anneX 1: instruMents OF risk ManageMent1,2 type of speed of cost of instrument examples comment instrument Disbursement Financing The funds in FONDEN, National Disaster Fund Mexico accumulate over (FONDEN), Mexico Reserve Funds Fast 1-2 years, while funds in NCF, National Calamity Fund, Philippines lapse at the end Philippines of the financial year. Budgetary re-allocation is used by most countries to Budgetary Moderate 1 get funds from other budget heads. Procedures, level of Reallocation approval, and time required varies across countries. Risk Tax increase as a response to disasters is difficult, as it A Retention adversely affects much-needed investment and is not Tax Increase Slow 1 popular. Connect it with the disasters is also difficult, as well as assessing to what degree this instrument is used as a resource. Normally donor assistance is available only in high severity disasters with international exposure and not for Donor Slow 0-1 low severity­high frequency disasters. It is also normally Assistance slow to come, sometimes with conditions attached, and does not solve the immediate liquidity needs. Catastrophic Risk Only 3 countries have used it Contingent Deferred Drawdown through December 2009. Fast 1 Credit Line Option (Cat DDO) from Risk the World Bank. B Financing Loans from Loans from international organizations are normally slow International Slow 1 to come and useful mainly for reconstruction. Organizations The only pooled fund in operation is CCRIF, which Multi Country/ emphasizes parametric insurance over risk retention, C Risk Pooling State Reserve Fast 1 but this concept of combined reserve funds can be very Funds useful for counties with similar disaster vulnerabilities. (Cost of Financing: The cost of financing here means out investing them to the expected losses from the di- the ratio of funds to be paid as a premium in the case of saster for the coverage period. Reserve funds have a cost insurance, interest in the case of catastrophe bonds, or of financing more than 1.0 because the funds cannot be the opportunity cost of funds to be kept in reserve with- invested in long term instruments and so cause losses.) east asia and the Pacific region The World Bank 1818 H St. NW, Washington, D.C., 20433 http://www.worldbank.org/eap GFDRR is able to help developing countries reduce their vulnerability to natural disasters and adapt to climate change, thanks to the continued support of its partners: ACP Secretariat, Arab Academy, Australia, Bangladesh, Belgium, Brazil, Canada, Colombia, China, Denmark, Egypt, European Union, Finland, France, Germany, Haiti, India, Ireland, Italy, Japan, Luxembourg, Malawi, Mexico, The Netherlands, New Zealand, Norway, Portugal, Saudi Arabia, Senegal, Spain, South Africa, South Korea, Sweden, Switzerland, Turkey, United Kingdom, United States, Vietnam, Yemen, IFRC, UNDP, UN/International Strategy for Disaster Reduction and The World Bank.