Insurance, Reinsurance and Catastrophe Protection in the Caribbean

Transcription

1 Page 1 of 45 Insurance, Reinsurance and Catastrophe Protection in the Caribbean A Working Paper prepared in collaboration with the World Bank Organization of American States General Secretariat Unit for Sustainable Development and Environment USAID-OAS Caribbean Disaster Mitigation Project May 1996 TABLE OF CONTENTS Acknowledgements Glossary Executive Summary Section I: The Insurance Market Section II: Dependence on Foreign Reinsurance Section III: Disaster Mitigation and Risk Management Practices Appendices A. Major Natural Disasters in the Caribbean since 1899 B. Impacts of 1995 Hurricane Luis on Leeward Islands C. Insurance Companies' Key Fiscal Ratios D. Recommendations for Evaluating Reinsurance Security E. Puerto Rico Catastrophe Fund F. Insurance Regulatory Upgrading Proposals Notes ACKNOWLEDGEMENTS

2 Page 2 of 45 The principal author of this paper was Arthur F. Evans, consultant to USAID/OAS. The task manager for the World Bank was John Pollner. He was assisted by W. Ed Morris, consultant to the World Bank. Coordination and supervision was provided by Jan C. Vermeiren, project director for the USAID/OAS Caribbean Disaster Mitigation Project. Many Caribbean people from government, financial, and insurance circles kindly gave of their expertise and time during consultant visits in Their assistance and contributions are greatly appreciated. GLOSSARY OF INSURANCE TERMS Adjusting: The adaptation of a claim put in by an insured to conform it to the coverage conditions of the policy for settlement purposes. Catastrophe: A major disaster event causing damage to multiple interests. Captive: An insurance company owned by its policyholders and primarily concerned with insuring the owners' risks; industry trade associations often set up their own captive companies. Deductible: A policy provision reducing the amount of claim settlement, often expressed as a percentage of the sum insured. General Agency: An insurance agency empowered by the insurance company(ies) it represents to accept stipulated risks and settle claims on behalf of the insurer. Hazard: A natural phenomenon with the potential to cause damage, including its location and the severity and probability of the damage. PML: Probable Maximum Loss, a term used by insurers to signify the estimated amount likely to be claimed under policies, often expressed as a percentage of the full values insured. Property: A generic term to signify those classes of insurance covering real and personal property (buildings, equipment, contents, etc.) as opposed to automobile, liability, life/health, etc. Reinsurance: The mechanism by which one insurance company contractually passes a proportion of insured risks to another insurance company, the reinsurer. Reinsurance

3 Page 3 of 45 commission: The consideration paid by the reinsurer to the reinsured company. Reinsurance pool: A mechanism by which insurers combine their risks to purchase reinsurance collectively and/or the combined acceptance of reinsured risks. Risk: Estimate of probable loss expected from the effects of a given hazardous event on a given element, or all elements, in an area. Risk relates a hazard to vulnerability. Risk management: A technology (also a profession) whereby risk exposures are systematically identified and quantified and determinations are made about whether to eliminate or modify the risk, self-insure the risk, or purchase traditional insurance protection. Retention Level: The level (amount) of risk retained by an insurer before reinsuring selfinsurance. The amount of risk-taking borne by property owner. Vulnerability: Degree of loss that may be expected of an element (e.g., a structure) exposed to a hazard event with a given location and severity. EXECUTIVE SUMMARY Means of preventing or ameliorating the consequences of natural disasters in the Caribbean fall into two basic categories: (1) hazard mitigation and vulnerability reduction measures adopted before a hazard event to minimize losses, and (2) economic mechanisms such as insurance, which pre-finance the costs of reconstruction. The former are ultimately more efficient, although the latter can reduce economic volatility by spreading risks more evenly. The Caribbean insurance industry has been existence for some 25 years, principally as a component of the larger financial products and services industry. Relatively few independent insurers operate in the area. Caribbean insurance companies have not been motivated to generate and expose significant capital for underwriting catastrophe perils. Characteristically, they limit their catastrophe risk retention levels to under 15 percent, the remainder being ceded to reinsurers. The availability of reinsurance affects the profitability of Caribbean insurance companies, as it governs the ability to write policies and thus generate subsequent income from reinsurance commissions. Tight markets, i.e., high premiums, have a pass-through effect, including proportionately higher commissions. Policy coverage restrictions are generally designed and imposed by foreign reinsurers, and their effect falls on the policyholders rather than on the insurance companies. In recent years the Caribbean area has seen its share of natural disasters, and its insurance markets have been through one of their most difficult periods. As a result, Caribbean insurers need to evaluate their portfolio risks with greater precision so that they can prove to reinsurers their true exposures and thereby maintain reinsurance protection at reasonable cost. The trends towards liberalization of trade and commerce are meeting a mixed response from the local Caribbean insurance industry. On the one hand, companies welcome the ability to invest assets in harder-currency areas. On

4 Page 4 of 45 the other hand, they fear outside competition. It is as apparent that the insurance industry needs to be rationalized as it is evident that there are far too many companies in proportion to population and GDP. A parallel condition is general under-capitalization. Although substantiating data are very limited, there is some indication that the industry's capital reserves are inadequate to secure the 15 percent average local insurance coverage. A greater regulatory role for government is needed comprehensive research in the region to eliminate this weakness and to ensure that reinsurers are financially capable of meeting their liabilities fully and on time. One successful incentive used by several Caribbean countries is allowing locally dedicated insurance reserves to be tax-deductible. For broader reinsurance security a single Caribbean "clearing house" is recommended, perhaps an assigned responsibility of the Caribbean Association of Insurance Regulators. Ideally, the effective regulatory role for governments is to serve as a fair and proactive umpire between policyholders and insurers. To shelter a major portion of its risk exposure in the Caribbean, the reinsurance industry has forcibly introduced a combination of deductibles and "average" clauses on claims resulting from damage caused by natural events. The deductible clauses customarily require self-insurance for the first 2% of the full insurable value of the property. For the average clauses, claim payments may be limited by the proportion by which the insurance covers less than the full insurable value. As recently experienced, these limitations may significantly reduce compensation for damage and result in dissatisfied policyholders. Full disclosure, in parallel with the creative marketing of insurance policies whose premiums are adjusted according to vulnerability estimates, should significantly improve the public attitudes and participation in such risk insurance. Affordability is crucial to the sustainability of the Caribbean insurance industry. Caribbean insurance-industry statistics are inadequate. Where they do exist, they are not compatible. Undue reticence surrounds the disclosure of data required for public and regulatory demonstration of fiscal performance. Although the importance of hazard and vulnerability mitigation measures is broadly accepted, Caribbean insurance companies are not inclined to take any initiative. They rely on the underwriting requirements laid down by reinsurers and view leadership in vulnerability reduction as a government function. It is evident that to keep the industry sustainable and protect the public interest, all parties--governments along with the local and the reinsurance industries--must assume an active role in promoting vulnerability reduction measures. The development of these roles is an excellent target for technical and financial assistance from international aid agencies. Some larger institutions in the Caribbean have organized self-insurance programs that have been very effective in providing coverage and reducing costs. In addition, there is growing interest in establishing regional reinsurance pools to provide capital to cover major costs in the case of natural catastrophes. Critical to the success of these programs is excellent apolitical financial management of the funds. It is quite possible that external development agencies and financing institutions might support their initial liquidity. It is believed that reinsurance premium pricing and commission mechanisms can be adapted to encourage improvement in the risk quality of portfolios and hence to allow discriminatory premium pricing for vulnerability reduction measures by policyholders. What is first required is a set of meaningful and workable risk-quality criteria and their incremental cost implications. This would include (1) hazard-mapped locations, (2) building type, structural vulnerability characteristics, and building contents, and (3) engineering certification for individual risk characteristics. Regulatory measures for encouraging vulnerability reduction in the real sector include (1) non-structural measures such as identification of hazard-prone areas and limitations on their use, dissemination of hazard risk information, and use of incentives and disincentives to promote safer development; and (2) structural measures such as the use of building codes and materials specifications, the retrofitting of existing structures, and the use of protective devices. Strengthened insurance regulatory oversight in the region can also support the objectives of properly evaluating real-sector portfolio risks and ensuring the financial strength and solvency of the insurance industry.

5 Page 5 of 45 SECTION I: OVERVIEW AND FISCAL POSITION A. The Insurance Market and Outlook 1. Geography challenges the nations of the Caribbean to respond to the socioeconomic consequences of natural hazard catastrophes--hurricane, storm surge, flood, earthquake, and volcano 1. Responses fall into two basic categories: a. Hazard mitigation and vulnerability reduction measures adopted prior to a hazard event to optimize protection from damage. 2 b. Economic mechanisms aimed at pre-financing the repair of the damage caused by disasters. The mechanism of insurance is in this category. 2. The former can be considered ultimately more efficient than the latter, which does not prevent or minimize the impact of the damage. The use of mitigation as a primary strategy in facing the consequences of natural disasters is slowly, but increasingly, gaining acceptance in public and private sectors around the world. Insurance remains a key pre-funding mechanism, but its efficiency is limited, especially when coverage is incomplete or not affordable. 3. The Caribbean insurance market obtained its own identity some 25 years ago with the passage legislation governing the establishment of local insurance companies. Most companies then established had hitherto operated under general agency or branch structures of foreign (mainly UK) insurers. For the most part, these local operations continue today as insurance units of trading and financial companies whose profit strategies center on their roles as agents for a wide range of products and services. Only relatively few independent insurers operate in the area. 4. Caribbean insurance companies have not been motivated to generate and expose significant capital for underwriting catastrophe perils. Furthermore, foreign insurers and reinsurers, with longer and broader expertise, remain willing to shoulder the major risk via reinsurance. Characteristically, Caribbean insurance companies limit their catastrophe risk retention levels to under 15 %, the remainder being ceded to reinsurers. 5. While 85 % of gross property insurance premiums are transferred to reinsurers, the actual remittances to them are reduced first by the commission paid by reinsurers to local companies (e.g. 30% x 85% = 25.5%) and second by the reinsurers' share of claims payments to policyholders under an assumed-loss ratio of 50% of premiums (85% x 50% = 42.5%). Thus, for a normal year the net outflow from the region would be 32% (10.0% -25.5% %) of the original gross premiums contracted by Caribbean insurance companies. A year with abnormally high claims experience can of course result in a net remittance inflow from reinsurers. 6. Caribbean insurers have largely restricted their involvement to their own geographic markets. However, the active involvement of some governments--bermuda, The Bahamas, the Cayman Islands, and to a lesser degree Barbados--has succeeded in attracting sizeable offshore captive insurance company business. Strong financial and insurance regulation has been an essential underpinning to these successes. 7. Over recent years, the Caribbean area has seen its share of natural disasters and insurance markets in the area have been through one of their most difficult periods. The spate of world-wide natural hazard events over the last several years had, through 1994, created the tightest reinsurance market in memory, with very high prices for limited availability, and stringent coverage limitations. There has been some alleviation of these conditions so far in 1995, but a renewed frequency of hazard events in the summer and fall will probably cause a reversion to tight conditions for those islands hit by storms. 8. The availability of reinsurance affects the profitability of Caribbean insurance companies as it governs the

6 Page 6 of 45 ability to write policies and thus generate subsequent income from reinsurance commissions. Tight markets, i.e., high premiums, have a pass-through effect including proportionately higher commissions. Policy coverage restrictions are generally designed and imposed by foreign reinsurers, and their effect falls on the policyholders rather than the insurance companies. In fact, the insurance companies benefit by the reduced risk exposure from the coverage restrictions, which customarily include a claims deductible of 2 % of the amount insured and a reduction of the claims payment in the same proportion that the amount insured by the policy is less than the full insurable/market value (known as the "average" clause). 9. The reinsurance product available to Caribbean markets is in essence designed and priced by foreign reinsurers on the basis of their worldwide (rather than Caribbean-area) catastrophe experience. Caribbean insurers now face the need to evaluate their portfolio risks more precisely than before so as to demonstrate their true exposures to reinsurers, and thereby maintain reinsurance protection at a justified and reasonable cost. 10. The trends toward further liberalization of trade and commerce are being met with a mixed response by the local Caribbean insurance industry. On the one hand, companies welcome the ability to invest assets in hardercurrency areas. On the other hand, they fear outside competition, especially if such competition is free of local regulation and tax obligations and also focuses on the broader, better risks. Caribbean countries' laws vary on the validity of insurance contracted with companies licensed outside the region (non-admitted insurance). Harmonization of these provisions is recommended, perhaps following the Puerto Rico example which allows non-admitted insurance subject to regulatory approval and a significant premium tax (approximately 20%). Another facet of the liberalization issue is the likely need to rationalize insurance companies/agents/brokers in the region, where close examination (in the absence of reliable statistics) shows that there are far too many companies in proportion to population and GDP. 11. Historically, feelings as to the industry's role for active involvement in promoting hazard and vulnerability mitigation have been very mixed. While the insurance companies do not deny the inherent benefit of such measures, their concerns center on the complexities and costs of implementation, particularly as reinsurers are seen as unlikely to share in such costs. They view leadership role on mitigation measures as a government function. Caribbean insurance regulators who guard market solvency are rightly concerned with the paucity and untimely delivery of meaningful data from companies, many of which are perceived as undercapitalized. Regulators are also pushing for needed upgrading of their powers and modernization of insurance laws and regulations. 12. The insurance markets appear intensely competitive for property insurance--a competition primarily seeking reinsurance commission revenues rather than underwriting or "risk taking". The larger Caribbean insurance markets contain insurance companies (with sizable market shares) that form part of broader commercial groups. For these companies, the thrust for increasing insurance policy production is further motivated by gaining "cross sales" growth in their related services such as insurance agency/brokerage, claims adjusting, mortgage, real estate, or banking operations. Observation suggests that insurance-market participants could be saturated while undercapitalized. This, however, is not yet substantiated, because the unavailability of credible statistics makes it impossible to determine asset strength. 13. Apart from the inadequacy of industry statistics in almost all markets (particularly balance-sheet information), there is little commonality among information systems, in either individual markets or the area as a whole. Outside of the Caribbean, insurance is a highly information-intensive industry and software systems abound, including those for regulatory reporting purposes. 14. One of the key issues in the Caribbean insurance market is the reinsurance availability/affordability factor. Given the more recent easing of reinsurance prices, Caribbean insurers should see to it that long-term interests are best served by strengthening all elements of the industry infrastructure to meet the reasonable expectations of policyholders, taking into account market capacity for coverage at different premium price levels. 15. This strategy should include enhanced contributions by the insurance industry to the promotion of hazard and vulnerability mitigation measures--positive as well as negative incentives--through discriminatory premium

7 Page 7 of 45 pricing. Also envisioned is more effective regulation of the business to ensure, inter alia, adequate solvency margins. Both the development of mitigation measures and the improvement of regulation can be expected to qualify for technical and financial assistance from international aid agencies. B. Fiscal Position of Industry 3 Insurance Company Reserves (and other reserve practices) 16. Since premiums are paid to provide for the possibility of future claims, insurance companies in areas prone to natural hazards have a particular responsibility to display their stewardship of premium moneys received and their fiscal performance. In particular, the insurance regulator needs timely, accurate, and full disclosure of pertinent data so as to fulfill the solvency vigilance role (see Appendix C). Compared with other markets, most of those in the Caribbean are unduly reticent about disclosing the data required for public and regulatory demonstration of fiscal performance. 17. Visits and discussions during 1995 with regulators, companies, and the Insurance Association of the Caribbean (IAC) revealed that the data reported are inadequate to permit a description here of the industry's current fiscal position. In particular, data reporting was seen as unduly delayed (up to two years or more), and very scanty indeed on balance-sheet elements necessary to identify key fiscal ratios or early-warning signs of perilous trends. The present report has therefore been limited to basic fiscal considerations relating to natural hazard insurance. 18. Insurance-company reserves fall into two basic categories: first, the shareholders' capital and free "surplus" reserves, and second, the insurance or "technical" reserves. The latter are customarily tax-deductible, constituted for known liabilities such as pre-paid (annual unused) premiums or payment of reported but unpaid claims. In effect, the capital and free reserves represent the solvency margin and are intended as the last asset resource should the technical reserves prove inadequate. 19. Two factors in particular apply in relating these considerations to natural-hazard. First, although catastrophes are accepted as severe but infrequent events in accounting principle terms (Generally Accepted Accounting Principles, or GAAP), their timing or magnitude cannot be precisely forecast, so that it is difficult to determine the optimal amount and investment instruments for reserving purposes. Second, more than 85 % of the insured catastrophe liabilities fall under reinsurance contracts placed mostly outside the Caribbean. These two factors prompt a local insurer to figure out first how to prudently reserve for the net liability retention (under 15 %) before the event occurs, and second how to ensure that reinsurers are financially secure enough to meet their liabilities fully and on time for the lion's share of the liabilities (over 85%). 20. Until recently, Caribbean insurers were discouraged by existing tax laws from setting up specific reserve provisions for catastrophes before the event. Several Caribbean countries are now permitting tax deductibility for such dedicated reserves. Without this dispensation, very little of paid premiums becomes available to meet future catastrophe claims liabilities under the policies issued. In concept and very broadly, in the absence of a natural hazard event, less than 20% of premiums customarily see its way to an increase in free reserves. Operating expenses can characteristically consume 30% of premiums, income tax about 24%, and a dividend policy perhaps a further 17%, leaving about 29% plus investment income, passing to free reserves. Hence, in the policyholder's interest, it is recommended that dedicated, properly monitored tax-deductible catastrophe reserves be allowed. The Puerto Rican "Catastrophe Trust Fund" is worth examination in this respect (see Appendix E). 21. The second challenge, that of reinsurance security, is of greater consequence than the first and is exacerbated by the recent turmoil in major foreign reinsurance markets (e.g. the Lloyds market). Caribbean insurance companies, as never before, must exert every proactive effort at their disposal to determine the ability of reinsurers to deliver their reinsurance commitments fully and on time (see Appendix D). Inadequate reinsurance security is tantamount to gambling the solvency and survival of a primary insurance company. Reinsurance security similarly has to be a focal point for the regulator's scrutiny. A single Caribbean reinsurance security "clearing house" would be advantageous in terms of effectiveness and efficiency; perhaps the Caribbean

8 Page 8 of 45 Association of Insurance Regulators could examine a potential role for itself in this function. 22. Reserves held by insurers contemplate several self-insurance practices that do not involve customary insurance mechanisms, in a standard "risk taking" sense. The most basic form of self-insurance is concerned with claim deductibles under policies. These are customarily 2% of the full insurable value of the property and are normally mandatory under catastrophe peril policies. This represents a significant transfer of risk, particularly for homeowners, as the 2% deductible will often cover repair costs e.g. roof repairs. The 1995 Leeward Island storms demonstrated that policy deductibles can almost relieve insurance companies of providing any coverage at all. Self-insured deductibles can also be significant to commercial policyholders; furthermore, they find that they need to be largely self-insured for business-interruption loss (moss of profit"). Some larger and special-risk categories (e.g., power utilities), have also over recent years found it impossible to obtain full or in some cases any, affordable insurance. On occasions, such entities have voluntarily devised very high self-insured deductible levels aimed at covering the potential loss damage and separate self insurance funding for business interruption. These risk-management arrangements, the emulation of which is recommended, have served to attract greater levels of insurance / reinsurance cover for the higher, less exposed, risk levels. Furthermore, the insurance cost and availability difficulties have prompted trade associations (e.g., Caribbean Hotel Association, CHA), to employ risk-management techniques and/or arrangements with off shore insurance companies to buy reinsurance on a group basis. 23. The extent to which reserves exist for varieties of self-insurance is not known, but tax deductibility for commercial enterprises to pre-fund catastrophe provisions is conceptually the same issue as for insurance companies and is likewise recommended. For home properties, mortgage and other lending institutions can help finance the self-insurance portion resulting from the application of claims deductibles and underinsurance penalties so as to assure timely repairs on their collateral. Insurance Liabilities and Corresponding Assets 24. A catastrophe insurer's gross potential liabilities (the sums insured under its policies) can run into hundreds of millions, if not billions, in any currency. The gross liabilities are reducible to net liabilities though reinsurance. A critical management challenge to insurers is a meticulous assessment of the commitment liabilities, both gross and net of reinsurance. This assessment includes the closest scrutiny of reinsurance contract provisions, including the clarity and accuracy of the information supplied to reinsurers. 25. Catastrophe commitments deserve clear tabulation, with segregation both for disaster types (hurricane, earthquake, etc.) and for alternative severity scenarios (hurricane storm tracks, wind speed classes, and earthquake Richter-scale severities). Provision also needs to be made for so-called "Second Event" scenarios as reinsurance contracts customarily vary from primary policy provisions by placing limits on the amounts of protection available for second (subsequent) catastrophes during any one reinsurance contract period. The 1995 hurricane season has given examples of unexpectedly high storm frequencies, and some insurance companies are understood to be faced (as at October 1995) with difficult decisions regarding reinstatement of their reinsurance contract limits for the remainder of the period (most expire at year end). The reinstatement premiums quoted, if at all, will be at very high "adversity" levels. 26. Clearly, a company's capital and free reserves, plus any dedicated catastrophe reserves, need to be demonstrated as sufficient and readily realizable for these purposes. The assets held to cover a company's insurance or "technical" reserves should not be considered available for the payment of catastrophe claims because they are constituted customarily to cover prior known reported outstanding claims or unexpired pre-paid (annual) premiums. 27. Consideration also needs to be given to the assets and investment instruments employed for covering reserves set up to meet an insurance company's liabilities (net of reinsurance) under natural hazard policies. Since the liabilities emanate from insurance policy contracts, the corresponding reserves and their covering assets are held by the company in a fiduciary capacity for the benefit of policyholders. In consequence, the investment instruments selected should be those that best meet the purposes for which policyholders purchase their insurance.

9 Page 9 of 45 One consideration has to be the secondary potential of natural hazard events to affect local financial markets adversely at the time liquidity is required. It is recommended that specific regulation support the view that preferred instruments are found in hard-currency financial markets least likely to be affected by natural catastrophes. Likewise, instruments should be placed in open financial markets entirely on an "arm's length" basis (without any strings attached, as opposed to some existing regulations that allow up to 20% of assets to be kept in intra-group holdings). C. Claims Payments and Reinsurance Recoveries 28. The basic principle governing natural hazard claims settlement and reinsurance recovery practices is that claims should be met fairly, speedily, and openly. The key new challenge facing claims settlement involves restrictions on the scope of policy coverage. The 1988 Hurricane Gilbert claims in Jamaica were met with admirable speed and fairness with assistance from teams of imported adjusters and close supportive involvement by reinsurers. With reinsurers' agreement, Gilbert's claims, for the most part, were settled without applying the "average" clause; furthermore, few policies then contained significant deductibles. Today, insurers have to negotiate settlement agreements under which the amount of the damage suffered is reduced first by the 2% deductible and second by the "average" clause. 29. These two coverage limitations, as is their intent, act to reduce materially the final amount of any claim settlement and in some cases, as has been said, preclude any payment at all--where, for example, the deductible of 2 % is more than the cost of the damage suffered. It has to be accepted that the settlement process will be much more protracted and potentially acrimonious than formerly, as policyholders might express shock at these fineprint coverage limitations. Furthermore, reinsurers (who characteristically have over 85 % of the claims' total), will vigorously scrutinize the application of these limitations, which have a very high impact on their liability. In summary, astute planning to deal with this novel scenario is recommended in order to avoid procedural and public-relations nightmares. 30. Practice drills for all personnel likely to be involved in the claims process are recommended to insurance companies. It is important that internal systems be tailored to catastrophe eventualities, because the overall claims settlement and reinsurance recovery processes are very information-intensive. Practice drills will hone management, public relations, and other skills to allow the transparency of operations essential to maintaining all round good will. Particularly important is the advance marshaling and training of an effective staff of adjusters to enable a company to portray a proactive, rather than reactive, response to claimants. The public relations function merits special focus, since the post-event image of an insurance company will depend on the image gained during adversity. A trained and energetic 24 hour "help desk" function goes a long way toward this objective. Policyholders need to know how to put together the information they need to press their claims and where to get practical help to safeguard their belongings. 31. Planning the reinsurance recovery/remittance process is recommended as another essential. A company's claims-management function should have a close involvement in reinsurance negotiations, especially on the contractual wording referring to claims procedures. Agreements should be reached with reinsurers, in advance of an event, as to the precise sequence of procedures, documentation requirements, and provisional payment arrangements. Such agreements should be in writing and should be made directly with the reinsurers, rather than rely on representations of intermediary reinsurance brokers. Reputable reinsurers will respect primary insurance companies for attention to detail on these aspects. Delayed reinsurance remittances can wreak havoc with a primary insurer's efforts to satisfy policyholders' reasonable requirements. D. The Regulator's Role (including data management and voluntary disclosure) 32. In other major markets, natural hazard experiences have prompted insurance regulators to take greater initiative in advancing multi-sector catastrophe protection approaches. Regulators have devised novel techniques and procedures based mainly on information technology for the necessary fiscal integrity and compliance matters. Most revised approaches enjoy the support of the regulated companies which share with the regulator a common

10 Page 10 of 45 interest in publicly demonstrating key fiscal and compliance performance indicators using common, costefficient, computerized formats. 33. The effective professional insurance regulator will be perceived publicly as a fair and diligent umpire between policyholders and insurers. This regulator will also guard the broader public interest served by the insurance industry, particularly the industry role in national capital markets (see the Puerto Rico example in Appendix E). It is recommended that pertinent government circles more readily recognize in practice that the strength, growth, and social responsiveness of the industry will be directly commensurate with the strength, growth, and social responsiveness of the industry's regulatory function. 34. A natural hazard event has the characteristics of low frequency and high severity. The adage "When the wind doesn't blow, nobody wants (or needs), to know" may have had some historical validity. However, catastrophes around the world in recent years have prompted insurance regulators to reassess their roles. The regulatory attitude or driving force recommended is one of encouraging and enforcing efforts to maximize the economic mitigation and protection available through the insurance mechanism without denying insurers a fair expectation of profit. 35. There is a parallel with financial services regulation, where the "hazard" to be mitigated is a liquidity crunch with potential for a run on a banking system. A natural hazard could provoke a run on an insurance system, along with the secondary "hazard" of public recriminations. In terms of delivering on their policy commitments, the issue of reinsurance security deserves the closest attention (see Appendix D). Regulators could consider adopting a risk-management philosophy approach to this challenge. Such an approach holds that the risks to be managed must first be identified and measured, and then be eliminated, minimized or transferred by means of appropriate techniques. The methods employed include adept scrutiny, compliance controls, and asset verification/certification as well as respected, speedy enforcement powers. 36. From meetings with Caribbean regulators and companies, it was evident that significant activity aimed at modernizing insurance laws and regulations is going on. Several findings called for improved regulatory effectiveness covering the regulators' institutional framework, statutory status, and enforcement powers, as well as financial, human, equipment, and facility resources. While these initiatives are very welcome, they imply that, generally, the regulator's function has not kept up with the times and that existing tools are not perceived as adequate. Furthermore, there was the implication that the regulatory improvement process itself is too slowmoving. 37. It is recommended that, to be strong and effective, insurance regulation include the following: a. Minimum capitalization requirements for local carriers and brokers. b. Solvency and liquidity levels. c. Adequate asset/liability management (including maturity and currency matching where applicable), and insurance security. d. Incentives and requirements (including tax concessions) to build catastrophe reserve funds up to minimum levels. e. Minimum standards for non-ceded retention of local coverage. f. Accurate verification and valuation of companies' balance-sheet entries to ensure adequate financial capacity to cover claims. g. Increased allowance for overseas investments of insurance assets.

11 Page 11 of 45 h. Industry entry requirements, including the admission of foreign competitors. i. Requirements that the security and reliability of overseas reinsurers that take on portfolios of local coverage be verified. j. Linkage of insurance regulation to require compliance with building codes before insurance coverage can be provided. k. Monitoring and inspection techniques. 1. Conditions for revoking licenses and shutting down operations. 38. Recommendations for institutional strengthening of the regulatory function would include the following: a. Adequate funding for the effective regulatory role (see Appendix E). b. Focused training and contracting of adequately skilled staff to develop, implement, and enforce procedures supporting the execution of regulatory reform. c. Upgrading information technology capabilities and computer equipment to allow for more automated compilation and analysis of insurance industry data, as well as timely reporting. d. Setting up "twinning" arrangements with superintendencies that have undergone or are undergoing successful institutional and regulatory reform (e.g., in Trinidad and Tobago), in order to share best practices. 39. The initiative from Trinidad and Tobago is encouraging. The Government has commissioned (with World Bank funding) a respected Canadian public accounting firm to produce proposals for the management criteria, status, and practices of the insurance regulatory function (Appendix F). It is recommended that these methods be adopted by other Caribbean nations as a "best practices" approach. SECTION II: DEPENDENCE ON FOREIGN REINSURANCE A. Impact of Losses and Reconstruction Debt 40. The direct damage caused by Hurricane Gilbert in Jamaica in 1988, as estimated by the Planning Institute of Jamaica, amounted to US$956 million, with nearly 50% from losses in agriculture, tourism, and industry, 30% in housing, and 20% in economic infrastructure. In terms of indirect effects, revenue projections had to be adjusted dramatically to allow for expected losses of US$130 million in export earnings, and over US$100 million in tourism earnings. Instead of a GDP growth of 5%, a decline of 2% was adopted. Secondary effects induced by the disaster were expected increases in inflation (30%), government expenditures (US$220 million), and public sector deficit (from 2.8% to 10.6% of GDP). In September 1989, Hurricane Hugo's most severe damage was inflicted on Montserrat with a total damage estimate of US$240 million. Of the 98% of housing damaged, 50% was severely damaged and 20% completely destroyed. The port's concrete jetty was destroyed and debris littered all island roads. The three main hotels were put out of business for at least four months. Agricultural crops were destroyed, and the fishing sector lost boats, buildings, and pots. The total damage exceeded five years of GDP. 41. In August and September 1995, hurricanes Luis and Marilyn hit the Leeward Islands with direct damage estimated at US$149 million in St. Kitts, US$254 million in Antigua and Barbuda, and some US$175 million in

12 Page 12 of 45 estimated rehabilitation costs in Dominica. The impacts on the public, social, and economic sectors were widespread. The advance forecasting of a catastrophe's overall costs is indeed a complex exercise, but appropriate, so as to discern which sectors require and deserve priority focus for disaster mitigation efforts. 42. International aid and development funding agencies, besides sharing consternation at delays, disruptions, and increased costs, have the strong view that wisely planned hazard mitigation efforts and funding before a catastrophe pay excellent dividends in reducing the economic impacts; mitigation expenditures are a very small fraction of the funds spent on reconstruction in the aftermath. For this purpose, regional governments might consider implementing policies to (a) make vulnerability reduction a national strategy covering all sectors, (b) institutionalize vulnerability reduction at the operational level via a multi-sectoral cabinet-level council linked to national disaster management agencies and serving as an information "clearing house", and (c) develop comprehensive hazard maps of each country, available to business and home developers, architects, engineers, and insurance companies. 43. Caribbean insurance companies have traditionally assessed the potential impact of catastrophe losses in closest collaboration with their reinsurers, and generally the latter's views prevail. These approaches have included tabulating insured risks on catastrophe maps that display perceived hurricane tracks and seismic zones. Total exposures, split by distinctive catastrophe peril, are compiled. A Probable Maximum Loss (PML) percentage factor is applied to reflect assumptions as to the vulnerability of distinctive areas, construction types and structure occupancies. It is from this PML, or expected aggregate catastrophe risk exposure in an insurance company's portfolio of issued policies that reinsurers determine the premium amount they require. Since they are characteristically assuming over 85 % of the catastrophe risks, reinsurers' premium levels have a very high gearing effect on the premium level charged under the primary policy issued to the policyholder. As has been said, reinsurers have additionally seen fit to insist on policy restrictions such as deductibles and full insurance. 44. Methodologies for assessing insurance risk, while widely used (if not mandated by reinsurers), are imperfect in two aspects. First, they result in broad-brush applications for individual risk characteristics and premium ratings. The technology for the assessment of vulnerability to risk has advanced to allow much more accurate and meticulous estimations for particular locations and structures, but existing practices do not provide for rational discriminatory premium ratings to reward the better-protected risks and penalize the poorer risks. Second, the absence of such discriminatory pricing discourages policyholders from taking steps to reduce vulnerability. Active study to implement discriminatory catastrophe premium rating for individual risk characteristics is recommended in order to provide more accurate market pricing signals from both the insurance provider and the policyholder. B. Reinsurance Needs 45. Before getting into the question of sectoral needs, it is germane to point out that throughout the Caribbean, natural hazard insurance is perceived to be unaffordable by much of the population. Although statistics are unavailable, it is recognized that most homeowners, except perhaps in Barbados, do not carry insurance except when required to do so by lending institutions. This is also thought to be true of the majority of small and even, to a significant extent, mid-sized businesses. Additionally, underinsurance is widespread and in the early 1990s many policies were allowed to lapse in response to high premium increases and the coverage restrictions discussed above. The needs and demand for insurance (and reinsurance) protection exist and are not being met largely because of affordability constraints. 46. For the most part, government physical assets such as buildings, schools, libraries, roads, and some hospitals appear to be underinsured or uninsured. Exceptions include Barbados and the government-owned Insurance Corporation of Barbados responsible for insuring public assets. Additional exceptions are thought to include properties owned by statutory corporations such as port and airport authorities, as well as utility companies that have independent access to the insurance markets. To reduce insurance cost, utility companies are actively considering a regional self-insurance program with the Caribbean Development Bank (CDB), the Inter-American Development Bank (IDB), and the Caribbean Electric Utilities Service Corporation (CARILEC). Although in the conceptual stage, the program's principal elements are as follows:

13 Page 13 of 45 a. Each utility that joins the fund contributes an annual amount representing 80% or more of its calculated commercial catastrophe premium. b. The CDB and other multilateral organizations provide contingent financing in the form of a line of credit during the formative years of the fund. c. As the fund grows, the utility companies rely less on the line of credit until it eventually becomes a standby support to be used only if the fund is depleted from claims arising out of a catastrophe. 47. In addition to this regional initiative, individual utility companies are examining ways in which they themselves can reduce insurance costs and/or increase available coverage levels. For example, the Barbados Light & Power Company Limited (BL&P) has started its own self insurance program because transmission and distribution catastrophe insurance coverage was unavailable in 1993 and the subsequent price for obtaining it was extraordinarily high (estimated at 25% of the expected loss). By year-end 1995, the company's fund will be capitalized at a level recommended by an overseas risk-management firm based on an engineering study. The fund is composed of cash and committed lines of credit. As in the regional approach, the lines of credit are to be used only after the cash portion of the fund is depleted. BL&P is now able to look at its insurance needs from a more reasonably priced "excess of loss layer" above the level of its self-insurance fund. The company estimates very significant annual savings in the premium cost. The adoption of such sound risk-management approaches is recommended throughout the public and private sectors. 48. Jointly with the Caribbean Electrical Utility Services Corporation (CARILEC), the Caribbean Disaster Mitigation Project (CDMP) has recently completed a vulnerability audit for all installations (and transmission and distribution) of St. Lucia Electricity Services Ltd. (LUCELEC) and for the hydroelectric installations of Dominica Electricity Services Ltd. (DOMLEC). This audit is now being used as a pilot study to generate guidelines for all Caribbean utilities on how to reduce their exposure to natural hazards, and will be used by CARILEC in training utility engineers. A concerted effort by electrical utilities to reduce exposure and eventual disaster losses, as a means to significantly strengthen any self-insurance or catastrophe fund mechanism for the sector, is recommended. 49. In 1993, the Caribbean Hotel Association (CHA) retained a U.S.-based risk management firm to perform a pan-caribbean study of windstorm risks to its members' properties to see if there was some way of reducing the upward-spiraling costs of insurance. A computer-generated wind study performed by a sub-contractor of the riskmanagement firm provided a probable maximum loss profile of the region and divided the Caribbean into six different risk zones, suggesting that there appeared to be enough diversification of risks among these zones to allow a regional insurance company for the CHA properties. With the PML information as the starting point, and using its own financial modeling capabilities, the risk management firm determined a capitalization figure for a regional insurance company to sell "all risks" property insurance to each of the some 1,000 CHA members. The firm then created, and today manages, a Bermuda insurance company whose exclusive clientele are members of the CHA. The company is not a captive insurance company of the CHA: neither CHA nor its members own it. However, members of the CHA are policyholders. CHA members are not obligated to buy the product; they can shop for competitive rates and coverage. Similarly, the insurance company can market to third parties, although it is not currently doing so. The risk-management firm was successful in attracting international institutional investors, portfolio managers, and venture capitalists. In addition, it was able to obtain multi-year reinsurance support from the global reinsurance community. In total, the program took roughly 18 months to put in place. 50. For many homeowners, the answer to soaring insurance rates, over the last five years, has been to reduce or eliminate their coverage. With the lack of statistics, it is impossible to measure the degree to which capital stocks are underinsured, although industry and government officials interviewed agree that underinsurance is substantial. From an insurance industry perspective, underinsurance is business lost that would otherwise carry little incremental operating cost. Governments are, of course, concerned, since the more people who are uninsured (or underinsured), the longer it takes for the local economy to recover from catastrophic events. It is recommended that the insurance industry, through the Insurance Association of the Caribbean, create a forum, with other sectors, to explore ways to expand the availability and affordability of catastrophe insurance to homeowners.

14 Page 14 of 45 C. Regional Risk Retention Capacity 51. Although well-managed local insurers have enjoyed high profits during the last five years of high premium levels, most have continued their traditional practice of paying out profits in dividends rather than reinvesting them to increase net worth and capital structure to allow an increase in risk retention capacity net of reinsurance. Two local trends, as yet slowly evolving, may suggest some change in strategy in the markets: (a) tax-deductible dedicated catastrophic reserve funds are being set up by a few insurers, though, given current levels, it will take a long time to reduce reliance on reinsurance, and (b) there is some increase in the buying of "excess-of-loss" reinsurance, as opposed to the traditional "proportional" reinsurance 4. The former is perceived as more costly up front, making this strategy affordable only for the financially stronger companies. Other trends include several nations' regulatory reforms, which embrace very much-needed increases in minimum capital requirements and tighter solvency ratios for insurance companies. The immediate need to adopt legislation to increase capital requirements to realistic levels to ensure prudential and sustainable risk management practices, is strongly emphasized. 52. However, strengthening the financial health of local companies does not necessarily mean that they will opt to increase their risk retention. They could simply continue to write more business based on the required increased capital, and continue to reinsure the same level. Financial strengthening of the industry is recommended for the benefit of the industry itself, the policyholders, and the government sector in that the regional capital stock deserves the highest quality of insurance protection--for property, business, and economic development and related social impacts--against natural hazards. Regulations allowing tax deductibility of catastrophe reserves would also encourage insurance companies to retain a higher level of profits within their business. D. Creation of Regional Reinsurance Pools and Financial Management Options 53. The idea of creating regional reinsurance pools for local primary insurers has been debated often in the Caribbean. There are several options for their design, although key aspects to take into account are whether the creation of a pool gives incentives to primary carriers to reduce dependence on commercial international reinsurers and to what extent a pool itself would need to cede part of its coverage to international reinsurers (a retrocession agreement). Both these issues will help to determine the success and functionality of an insurance pool. In essence, it is recommended that the critical criteria should be the extent to which any pool arrangement is able to generate genuine additional capacity, rather than merely reshuffle existing capacity, and to promote more efficient market behavior. 54. In addition to these "structural" issues, there are various public-sector regulatory aspects of a reinsurance pool that would need to be coordinated among several regional governments, a process that is not always guaranteed to result in consensus, especially given the differences in insurance industry development and disaster-proneness in the various Caribbean countries. However, there are ways of basing contributions to a regional pool on objective criteria such as country-wide PML assessments and expert forecasts of disaster probabilities by subregions. Some of these techniques could be developed with technical assistance at a regional level, as described further below. 55. While premiums paid into the pool fund would support its capitalization, regional governments might also need to make initial contributions or guarantees to arrive at a minimum level of capital for insurance risk-taking requirements. Thus, governments might have initial shareholdings in the pool, which would diminish as premium income paid by private carriers increased its equity base. In the long run, it is recommended that the pool be privatized once public-sector financial support was no longer required. During the early years of capitalization, the pool could be backed up by a multilateral institution's contingent line of credit, which, if utilized, would leverage the pool until additional premium income and/or contributions paid this off while the pool was recapitalized. 56. In terms of financial benefits, the establishment of a regional reinsurance pool could potentially:

15 Page 15 of 45 (a) Help reduce premiums paid by primary carriers from the high commercial rates charged by reinsurers, provided that the insured portfolio risks contained in the pool were more meticulously evaluated than is now customary among international/national institutions; (b) Increase overall market capacity and avoid leaving large portions of the population and public-sector assets uninsured; (c) Encourage the local insurance industry to increase its capital base from premium cost savings, and thus possibly retain a larger portion of its coverage as a condition for purchasing reinsurance from the pool; and (d) Stabilize the volatility in premiums by delinking the influence of other international catastrophic events from the Caribbean market, and by spreading the regional financial risk in a pre-funded pool. 57. While the reinsurance pooling concept appears worth pursuing, harder dimensional information would be required to permit an informed opinion on the available options. It is therefore recommended that a separate technical study be commissioned seeking financial and technical support from a multinational institution. Suggested terms of reference for such a study would include the following: Approach: The purpose would be to identify sustainable methods of increasing the availability of affordable catastrophe insurance protection and risk-funding mechanisms while reducing the volatility in the price of coverage, and to increase the deployment of measures to reduce vulnerability to disasters. The study would adopt the following: Scope of Work: a. Consideration of the following structures, separately and in combination: 1. Insurance-company-specific "catastrophe reserve trust funds". 2. Reinsurance pooling for use by insurance companies within one country. 3. Multi-country/geographic-zone approach, pooling arrangements with more diversification of risks. 4. Sectoral-level reinsurance pooling or limited direct funding for uninsured public-sector and underinsured private-sector exposures, on a country-by-country or a regional basis. 5. Multi-layered financial arrangements including industry pooling, statutory reserve funds, public budgetary allocations/reserves, market cooperation arrangements, and capital-market-backed contingent credit lines or guarantees. b. Identification of dimensions for catastrophe exposure hazards, through comprehensive hazard mapping and actuarial input on risk probabilities (including

16 Page 16 of 45 event return periods) and estimation of probable loss parameters using engineering methods. c. Identification of priority coverage classifications for the existing shortage of catastrophe insurance, including funding provisions. Determination of reasons for inadequate market response or coverage availability/constraints/risks in such sectors. d. Identification of practical real-sector vulnerability reduction measures, and their potential impact in reducing catastrophe losses. Identification of strategies to promote such measures within a market-incentive framework with regulatory support for optimizing price/coverage availabilities. e. On the basis of exposure data, projection of the capital requirements, potential sources, and long-term financial scenarios of alternative funding/pooling/credit mechanisms. f. Identification of appropriate legal structures and administrative, financial, and operating environments (including public, quasi-public, and private entities). g. Recommendation of opportunities and their relative merits for improving existing reinsurance purchasing practices for the region, to ensure full market access to reputable reinsurance companies. 58. The financial implications of setting up a regional reinsurance pool, however, present some concerns from both the public and the private sectors. From the public-sector point of view, the establishment of a pool that would reinsure national carriers implies some sort of financing contribution to initially capitalize the pool. While such a commitment from various governments may be difficult to carry through, it might be accomplished with less resistance once the value of the contributions is shown methodically to be less than the value of post disaster debt and reconstruction costs (particularly where public assets or non-insured property need to be repaired following catastrophic events). A mechanism to consider for this purpose would be governmental guarantees of initial liquidity provided by institutional investors. The stabilization of risks via a financial mechanism such as a pool is especially beneficial if such a pool creates incentives for local carriers to insure previously uninsured property and/or other assets. The reinsurance pool need not necessarily compete against the existing commercial reinsurers--it might also fill a market niche for coverage currently perceived as not profitable for international reinsurers. 59. From the private-sector industry perspective, a regional reinsurance pool seems appealing only in that it could potentially reduce premium costs and avoid sharp fluctuations in premium charges following disaster events. However, some of the large primary carriers in the region feel that they have a comparative advantage in staying out of such a pool, for four reasons: a. They have already set up substantial reserves to cover against potential disasters, and thus are rather self-sufficient. b. Purchasing reinsurance from a pool would in many instances mean to them that they are subsidizing less-than-prudential policies of smaller carriers or brokers who would benefit the most from the pool. c. Large carriers generally have good business relationships with international commercial reinsurers from which they can obtain better-than-average premium contracts as part of long-standing customer relationships. d. Carriers in certain Caribbean countries perceive their risk of exposure to catastrophe as being less than that of other countries, and they think that even if the premium

17 Page 17 of 45 prices those other countries paid into the pool reflected the higher risks, there might still be some sort of subsidy that could be avoided by negotiating their own reinsurance directly. 60. A key aspect to consider would be whether the pool itself would retrocede some of the risk to international commercial reinsurers. Sufficient initial capitalization and a minimum ongoing level would be required of the pool to assure that it retains sufficient coverage to meet its objectives. The result of retroceding too much of its portfolio would probably be to invalidate much of the benefits of establishing a pool, unless of course the excessof-loss limits for such retrocession were above certain already high levels of coverage, thus avoiding excessively high premiums from international reinsurers and the associated volatility. Reinsurers, of course, will also need to be persuaded to take on any of the risks from a pooled portfolio at any level of coverage. However, this issue can be addressed by developing a priori and showing clearly the derivation of the probable loss values of the pooled risks, backed by methodologically robust techniques covering country-by-country and sector-by sector risk assessments, including damage and disaster event probability estimates (catastrophe return periods). In terms of the private-sector issues mentioned above, there are very valid arguments for some of the large and secure carriers not to participate at all in the pool, even though this would deprive it of substantial income to maintain its operations. However, since this concern is due in part to the segmented quality of the market, it might be feasible to establish a pool for a subregional country grouping (excluding the largest countries) in which the insurance industry has fewer large players and in which risk-sharing would be much more beneficial given the small sizes of their national economies. This would require, however, that the risks were sufficiently diversified in any specified subregion (e.g. OECS). 61. The pool could be administered by an independent party coupled with best-practices underwriting and financial policies and could establish conditions for the participation of local primary carriers, including meeting prudential practices such as maximum limits for ceding coverage (i.e., some rational minimum local retention) coupled with minimum capital requirements and portfolio risk assessments based on a commonly understood methodology (including property valuation, hazard zoning, disaster event frequencies, PML, land-use restrictions, and building-code compliance). Such conditions, as discussed later in this report, could be contingent on the availability of funding to guarantee liquidity in the pool before it is fully capitalized and to pay for the requisite technical assistance to assure that insurance regulatory measures can be enforced and disaster mitigation measures are in place. 62. In terms of financial management, the assets of a regional pool could be invested mostly overseas to achieve maximum returns and provide needed foreign exchange in the event of disasters. The administration of a pool would be much the same as that of a large insurance company, with the chief executive operating under authority delegated by a board of directors composed of representatives of pool participants. Individual Country 5 or Sectoral Disaster Funds 63. As an alternative and/or complement to regional reinsurance pools there exists a different niche that would primarily cover uninsured government-owned property and uninsured or underinsured properties, particularly for those poorer segments of the population who cannot afford current insurance rates and whose property would be subject to a possible government backstopping of liabilities related to events such as hurricanes or other disasters. For this latter area, consideration might be given to a public insurance fund that could also be used to provide coverage for government-owned assets such as buildings, water systems, roads, and other public goods. The capitalization of such a "national disaster fund" would need to be financed primarily out of the national budget. This in essence would be a self insurance mechanism, which in the long run would avert the need for emergency borrowing in the event of disasters. 64. Some degree of additional risk would be taken in providing coverage for poorer segments of the population by offering insurance at what might be slightly below-market rates to ensure affordability. Such low-cost insurance, however, would be coupled with strict hazard-mitigation policies required of the end beneficiaries, including building retrofitting, optimal land use, building-code compliance, and other measures. The development of sound methodologies to implement such a program might require the assistance of international development

18 Page 18 of 45 institutions, as discussed below. The modest premium income, coupled with the government contributions, would serve to capitalize the fund, with the possibility of obtaining guaranteed lines of credit from multilateral institutions to ensure sufficient liquidity to pay claims before the fund becomes fully self-capitalized. The level of government support, of course, would be based on PML valuations and disaster event frequency analyses, which would determine the actuarially driven requirements for financial contributions on an annual basis. 65. The fiscal impact of budgetary contributions to such funds usually deters governments from engaging in such schemes. However, it should be noted that while some of the public sector assets can be insured directly with private commercial insurers, there are other assets and there are poorer segments of the population that will not be covered by the private carriers, given the current state of the market. At the same time, the private local carriers are seeking regulatory concessions, including the tax-free accumulation of catastrophic disaster reserves and further flexibility to invest reserves overseas--conditions that are well in line with international insurance practices and would improve the capital of local carriers while stimulating increased retention of local coverage. If the governments allow the regulatory flexibility to meet these industry needs, they might potentially phase out the taxing of catastrophic reserves by setting aside a declining proportion of those taxes for a few years in order to help build up the national disaster funds that would finance the coverage of the low-income population. As with all the pooling or funding schemes mentioned, a critical ingredient is the identification of real financial incentives for all parties involved (public sector, private sector, and beneficiaries) to generate sufficient economic interest from all sides to implement new arrangements or institutions that, in the end, reduce long-term costs and volatility in the economy. 66. Other alternatives to national public insurance schemes include sectoral funds, which can be funded either publicly or by the industry concerned. At a regional level, the examples of the proposed CARILEC program and the dedicated insurance company of the Caribbean Hotels Association (see paragraphs 46 and 49 above) may be cited. Such arrangements might provide initial pilot models for more generalized funds at a regional or national level. However, as an alternative pilot approach, national disaster reserve schemes can also cover specific sectors given the homogeneous risk characteristics and thus more quantifiable risk forecasting. Priority candidate sectors might include agriculture, road transport, schools and hospitals, water systems, or housing. 67. To reiterate, the main selling point for establishing pools or self-insurance funds relates to the present value and rate of return of a fund's invested contributions and premiums, which, in the event of a disaster, will yield a higher coverage value and level of net proceeds than the alternative of contracting debt (even if concessionary) after the disaster. Such calculations, while not currently available in detail, could be developed during the initial stages of the pool concept to present the long-term cost/benefit outcome before any of the financial or institutional arrangements are implemented. Role for External Development Financing Institutions in Supporting Initial Liquidity 68. With the worldwide increase in the frequency of hazard events, international and bilateral donors have increasingly restricted their response to appeals for assistance. This has begun to manifest itself in donor countries' budget and foreign-aid allotments, and thus a number of developing countries have become aware of the need to undertake mitigation and self-insuring steps. In this regard, this report analyzes the role of the insurance industry in contributing to the stabilization of financial risks associated with natural disasters, as well as other mechanisms whereby regional governments and multilateral institutions can begin developing contingency plans to reduce the economic volatility associated with such unpredictable events. 69. The initial full capitalization for regional pools or disaster reserve funds in line with actuarial requirements poses some funding risks, particularly with respect to the payment of claims. For this purpose, it is envisaged that during the early years of the establishment of any such fund, a multilateral institution would be asked to provide a guarantee of financing or a contingent line of credit for quick disbursement if the pool or disaster fund's assets are insufficient to provide the requisite coverage. This would not necessarily imply that the first incident would exhaust all the capital of a given pool or disaster fund, but rather that the pool or fund would use the line of credit to complement its liquidity for the payment of claims while still keeping its minimum capital base to remain viable from a regulatory financial standpoint.

19 Page 19 of A line of credit from a multilateral institution would be favorable in that it would (a) be committed a priori, (b) have a longer maturity and thus a lower repayment load, (c) have a lower financing cost than comparable commercial lines of credit because of the multilateral institution's lower borrowing cost and higher returns on its own liquidity, and (d) be coupled with technical assistance in the areas of disaster mitigation, hazard mapping, building-code enforcement, and strengthening of the insurance regulatory institutions. Technical Assistance Requirements of Regional Pooling Arrangements 71. In establishing a regional pool and/or disaster reserve funds a number of analytical steps would have to be taken before specifying their financial structure and their framework for financial management and administration. From the perspective of multilateral institutions, any funding, contingent line of credit financing, or guarantee that might become available would most likely need to be coupled with real sector mitigation measures as well as insurance-sector regulatory measures, to ensure the lowest cost in premium payments and to maximize the coverage at the lowest risk. In addition, a prudent third-party financial manager would be desired to ensure effective investment of the assets and to enforce the premium structure on the basis of objective methodical studies that would determine the contribution levels by country, sector, or industry risk on a weightedaverage basis. 72. The technical assistance in key areas that might be needed before such a fund was financially structured and put in operation might therefore include the following: Mitigation in the Real Sector a. Hazard-mitigation measures and mechanisms such as developing and enforcing building-code regulations; b. Disaster planning through prudent land-use management and the establishment of legal restrictions for building on extremely risky land; c. Hazard mapping by country and vulnerable areas to determine the intensity of risks from hurricanes, other winds, flooding, earthquakes, etc., and to assist in proper land use management and PML valuation; d. Coupled with (a) above, the conduct of underwriting assessments of asset and property values, resistance to wind or other hazards, and likelihood of damage as reflected in PML estimates, based on probability data on the occurrence of such natural disaster events; and e. Given the importance of disaster data, additional compilation of event frequencies and estimated probabilities, as well as covariance and correlation studies among regions and between events, to feed into the valuations of property risk. Financial Management of Pools/Disaster Reserve Funds a. Mechanisms to institutionalize regional and/or national insurance or reinsurance funds with financial management and administrative control arrangements supported by legal contracts and executed by independent parties, to be financed in accordance with adequate procurement practices of multilateral lenders; b. Making actuarial projections and analyzing financial plans and options to build up the country-based disaster reserve funds or regional pools, while providing regulatory support for investments in appropriate financial markets;

20 Page 20 of 45 c. Developing probability and financial options models to show the rate of return of the pool or fund with respect to anticipated disaster events, and to compare this return with traditional pay-as-you-go international relief debt incurred after disasters; and d. Developing the legal and budgetary requirements and establishing autonomy of operation to ensure the institutionalization and successful implementation of the fund/pool. Insurance Regulatory Reform & Institution-Building, including the recommendations in Section I.D of this paper. SECTION III: DISASTER MITIGATION AND RISK- MANAGEMENT PRACTICES A. Vulnerability Reduction 73. Vulnerability reduction and mitigation measures are broadly accepted as the highest-impact mechanisms to reduce losses from catastrophes, the ensuing reconstruction debt, and other adverse consequences. Insurance does not, per se, neutralize damage; it is limited to providing monetary compensation for damage. In addition, only some of the insurance premium is available to compensate for losses; the rest goes for administrative/operating expenses and taxes, particularly under regional operations that have relatively low income-reinvestment practices but high dividend payouts. 74. Empirical data showing the cost/benefit yield potential from vulnerability reduction in the Caribbean area have not yet been compiled, but the anecdotal information is compelling: a. In Jamaica over half of the insured losses from Hurricane Gilbert in 1988 related to dwellings, most of which suffered roof damage. The resulting millions of dollars in claims payments could have been substantially reduced if inexpensive roof straps had been installed. Shutters would have been an additional safeguard. The Construction Resources Development Center of Jamaica has designed and is operating a campaign to help low-income householders retrofit their roofs. The campaign is now spreading to higher-income housing, and one of the major insurance companies is offering a 10% premium discount to homeowners who adopt such measures. The CDMP is implementing a similar safe-roofing campaign in Dominica and Saint Lucia, with a revolving loan facility to make investing in increased safety more affordable to low income families. b. Barbados Light & Power was reported in a recent interview to have spent US$250,000 on vulnerability reduction measures to make its properties capable of withstanding severe damage from a Class III (under 120 mph wind speed) hurricane. While these actions have fortunately not been put to the test of such a storm, the annual premium savings are reckoned as perhaps twice this amount. c. A Barbados civil engineer reported in April 1995, after five years of involvement in designing and implementing structural vulnerability reduction measures (including retrofitting), that many buildings can be made virtually invulnerable to Class III hurricanes at a cost equivalent to one to two years' catastrophe insurance premiums. d. The Chief Coordinator of the National Emergency Management Agency in