NATIONAL FLOOD INSURANCE PROGRAM Actuarial Rate Review Thomas L. Hayes, ACAS Actuary Mitigation Division Emergency Preparedness and Response Directorate NOVEMBER 30, 2003 FEMA Contents Purpose of This Document 1 Overview 1 A Brief History of the NFIP 1 Financial Structure of the NFIP 3 Borrowing Authority 3 Operating Expenses 3 Premium Structure 4 Impact of Other Public Policy Objectives on the NFIP 5 Ratemaking 6 Subsidized Rates 7 Pre-FIRM Subsidized Rates 7 Special Post-FIRM Classes That Are Subsidized 8 Actuarial Rates 8 Overall Rate Level Indications 9 Target Level Premium Analysis 10 Rate Review Results 12 Expense Constant 13 Federal Policy Fee 13 Impact of Community Rating System 14 B, C, and X Zones Experience 14 Preferred Risk Policies (PRPs) 14 X Zone Standard Policies (non-PRP policyholders) 15 V Zone Experience 15 Deductibles 16 Increased Cost of Compliance (ICC) Coverage 16 Mortgage Portfolio Protection Program (MPPP) 16 Exhibits 17 Exhibit A. Effects of Revisions on Written Premium 18 Exhibit B1. Key Underwriting Components by Year, 1978-2002 19 Exhibit B2. Detailed Underwriting Experience by Year for the Latest 10 Years 20 Exhibit B3. Detailed Underwriting Experience Aggregated by Experience Period 22 Exhibit B4. Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1978-2002 23 Exhibit B5. Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1986-2002 25 Exhibit C. Calendar/Accident Years 1978-2002 Experience for the Larger Risk Zones 27 Exhibit D. Average Expenses per Policyholder 28 Exhibit E. Projected Annual Premium Requirements Based on 1978-2002 Loss Experience vs. Projected Written Premium 29 Appendix. Actuarial Rate Formula A-1 N A T I O N A L F L O O D I N S U R A N C E P R O G R A M Actuarial Rate Review NOVEMBER 30, 2003 Purpose of This Document An annual review of the National Flood Insurance Program (NFIP) underwriting experience, with accompanying Program revisions, is an integral part of maintaining the Program's goal of a fiscally sound rating and coverage structure. The purpose of this document is to share the results of the latest actuarial review of the rating structure in the context of the history and goals of the Program. Overview Floods have been, and continue to be, the nation's most destructive natural hazard in terms of economic loss. Since the inception of the Program in 1968, the Federal Government has had to assume a major financial role in easing the impact of flood damage on individuals and communities. Studies indicate that, although insurance does not and probably cannot respond to all the needs of disaster victims, insurance is the most efficient and equitable method of providing disaster assistance (GAO Report, PAD-80-39). As a result, the U.S. Congress established the National Flood Insurance Program (NFIP) with the passage of the National Flood Insurance Act of 1968. The NFIP provides the means by which flood insurance is made available through the cooperative efforts of the Federal Government and the private insurance industry. The NFIP is part of a coordinated, three-pronged approach developed to (i) identify those areas within local communities that are most at risk of flooding, (ii) minimize the economic impact of flooding events through a combination of mitigation efforts and floodplain ordinances, and (iii) make flood insurance available to help individuals and small businesses recover following a flood. The NFIP can provide the flexibility for flood insurance to be based on workable methods of pooling risks, minimizing costs, distributing burdens equitably among those protected by flood insurance and the general public, and structuring rates to support mitigation and floodplain ordinance efforts. A Brief History of the NFIP The National Flood Insurance Act of 1968 created the NFIP, which since 1979 has been part of the Federal Emergency Management Agency (FEMA). Earlier this year FEMA became part of the newly created U.S. Department of Homeland Security (DHS). Within FEMA, the NFIP has been historically administered by the Federal Insurance Administration, which has been more recently consolidated with other operations of FEMA and renamed the Mitigation Division. The basic structure of the NFIP was established by the 1968 Act and that structure continues today. The NFIP is a Federal program enabling property owners in participating communities to purchase insurance as a protection against flood losses in exchange for State and community floodplain management regulations that reduce future flood damages. Participation in the NFIP is based on an agreement between communities and the Federal Government. Flood insurance is made available within a community when it adopts and enforces a floodplain management ordinance to reduce future flood risk to new construction in floodplains. To encourage participation in the NFIP, the Flood Disaster Protection Act of 1973 expanded the authority of FEMA to grant premium subsidies as an additional incentive to encourage widespread state, community, and property owner acceptance of program requirements. For the next 7 years, the heavily subsidized premium charges remained in effect. During that period, nearly every community with a flood hazard joined the NFIP, and the insurance policy count increased dramatically, reaching 2 million by 1979. States also responded: governors appointed floodplain management coordinators to assist local communities' governments in working with FEMA on Program matters. These actions resulted in establishing, for the first time, a nationwide response to address the flood peril. In 1981, with the NFIP firmly established, FEMA initiated rating and coverage changes through the mid-1980s that placed the Program on a fiscally sound basis with significantly less subsidy being provided. In establishing a fiscally sound program, which was achieved in 1988, FEMA has stressed that, as opposed to the traditional insurance definition of fiscal solvency, the NFIP's intent is to generate premium at least sufficient to cover expenses and losses relative to what is called the "historical average loss year." The National Flood Insurance Reform Act of 1994 reinforced the objective of using insurance as the preferred mechanism for disaster assistance by expanding mandatory flood insurance purchase requirements and by effecting a prohibition on further flood disaster assistance for any property where flood insurance, after having been mandated as a condition for receiving disaster assistance, is not maintained. These measures were added in recognition of the fact that loan or grant programs, to the extent that they parallel the insurance mechanism, can undermine the ability of the insurance program to operate efficiently and equitably. More recently, Congress has focused its attention on the impact of repetitive loss properties on the National Flood Insurance Fund. Since 1999, there has been proposed legislation on this issue. On November 20, 2003, the U.S. House of Representatives passed H.R. 253, the Flood Insurance Reform Act of 2003. The bill introduces a 5-year pilot project that (1) defines severe repetitive loss properties, (2) allocates additional funds for mitigation projects, and (3) mandates a 50% increase in premiums for property owners who decline a mitigation offer, along with an appeal process. Repetitive loss properties are a significant NFIP issue that has caught the attention of a number of Members of Congress. It remains to be seen whether the Senate will pass H.R. 253 during this session. But, even if Congress doesn't, it appears probable that the repetitive loss issue will continue to be debated by Congress, and that the next few years will see some form of legislation enacted on this issue. Financial Structure of the NFIP Borrowing Authority The Program has not been capitalized and pays losses and operating expenses out of policyholder premiums. The result is that during less-than-average-loss years the Program generates surplus, while during higher loss years that accumulated surplus is used to pay the amount by which insured flood losses exceed that year's net premium revenue. The NFIP has borrowing authority with the U.S. Treasury to cover losses in the event that policyholder funds and investment income are inadequate. Initially, the NFIP was granted a $1 billion borrowing authority, but in 1996 legislation was passed (and subsequently extended) providing an increase in borrowing authority from $1 billion to $1.5 billion in order to provide a greater cushion against potential losses. As of the end of FY 2003, the National Flood Insurance Fund had a positive balance of just over $660 million, which should be more than sufficient to pay the outstanding claims from Hurricane Isabel, which occurred during September 2003. During the last decade, however, the NFIP has exercised its borrowing authority three times. Following the Midwest Flood of 1993, the Program borrowed $11 million, which was quickly repaid. The Program borrowed again as a result of the heavy flood losses during 1995 and 1996 that were at twice the historical average. That borrowing peaked at $922 million during FY 1998, but was completely repaid by June 2001. However, Tropical Storm Allison (June 2001)—the first $1 billion storm in the history of the NFIP—required the Program to borrow $650 million. That amount was repaid as of October 31, 2002. Operating Expenses From 1987 through 1992, the Congress, rather than appropriating tax dollars for Federal staff salaries and the costs of flood studies and floodplain management as had been done previously, instead transferred policyholder premiums to salary and expense accounts and the emergency management program accounts of the Federal Emergency Management Agency (FEMA). These expenses were not authorized to be included in the insurance premium charges. The current value of this transfer and the resulting loss of investment income and increased borrowing is effectively a reduction in loss reserves in the National Flood Insurance Fund of about $572 million. This has made the fund more vulnerable to the need for exercising the NFIP's statutory borrowing authority in order to cover losses arising out of a large flood event. FEMA believes that most of the salary, study, and floodplain management costs delineated above in the discussion of fund transfers are Federal in nature and benefit taxpayers as a whole through programs that reduce future flood losses and resultant Federal expenditures. However, the Congress legislated, with the Budget Reconciliation Act of 1990, that the full funding of these expenses would be borne by flood insurance policyholders through a Federal Policy Fee. To keep this charge as low as possible, the legislation specifically states that the fee is not subject to agent commissions, company expense allowances, or State or local premium taxes. Therefore, although in this rate review the Federal Policy Fee is included in exhibits and analyses of rate level indications, for accounting and Write Your Own (WYO) company reporting purposes, the fee is not considered to be premium. Premium Structure In establishing a fiscally sound program, which was achieved in 1988, FEMA has stressed that, as opposed to the traditional insurance definition of fiscal solvency, the NFIP's intent is to generate premium at least sufficient to cover expenses and losses relative to what is called the "historical average loss year." The underwriting experience period has, to date, included 7 heavy-loss years . Despite these heavy-loss years, the absence of extremely rare but very catastrophic loss years leads to the conclusion that the historical average is less than what can be expected over the long term. The establishment of this target level of premium income for the Program as a whole accommodates the combined effect of the portion of NFIP business paying less-than-full-risk premiums (a subsidy provided by statute) and the portion of the business paying full-risk premiums that contemplate in their rates the full range of loss potential including catastrophic levels. The distribution of business written in 2004 is anticipated to be 27% at subsidized rates and 73% at full-risk premium rates. FEMA estimates that, were the catastrophic contingency contemplated in establishing all rate levels, the Pre-FIRM subsidized portion of the business would have to pay about two and a half times the current premium, and the overall target level for premiums would have to increase on the order of 50% to 75%. The most recent changes were effected on May 1, 2003. These resulted in an average rate increase of 3.6% for actuarially rated policies and 1.9% for subsidized policies, with the average Program-wide rate increase being 2.9%. There were minor rate increases for most zones, with the largest increase (2% to 9%) falling on Pre- and Post-FIRM V-Zones. Also, the Expense Constant was eliminated, and the potential loss of revenue was offset by raising the basic limits rates. The Federal Policy Fee (FPF) for Preferred Risk Policies (PRP) was increased to $10 from $5. Finally, the limit of liability under Increased Cost of Compliance (ICC) coverage was increased from $20,000 to $30,000. This year's Actuarial Rate Review recommends that the actuarial based rates increase 0.1% and the subsidized rates increase 5.1%, corresponding to an overall premium increase of 2.2%. A breakdown of the proposed rate increases by category is shown in Exhibit A. The largest of these increases are again in the Pre- and Post-FIRM V-Zones. In addition, five other changes are recommended. First, it is recommended that the contents coverage amounts be increased for 1-4 Family PRPs, with the building coverage amounts remaining the same. Second, a contents-only coverage PRP is being recommended to be introduced for all Residential PRPs. Third, the introduction of two Non-Residential PRPs, one with both building and contents coverage and the second with contents-only coverage, is being recommended. The fourth recommendation is that, for all PRPs, the Federal Policy Fee (FPF) should be increased to $11 from $10. Fifth, it is recommended that the Increased Cost of Compliance (ICC) premium, which applies only to building coverage, be decreased from $6 to $1 for PRP policies. Impact of Other Public Policy Objectives on the NFIP The Program's financial status must be addressed in a context that is broader than the focus of this rate review. While low loss experience can provide opportunities to rebuild surplus from policyholder premiums, other measures and public policy issues must also be explored. For example, FEMA has developed a strategy for addressing repetitive loss properties, prioritizing them, and seeking ways to increase mitigation assistance and reduce the extremely large levels of subsidy provided to such high-risk, older properties. Implementation of this strategy began in 1999 with the start of a new Special Direct Facility to handle the policies on these properties. The degree to which funds are available to mitigate repetitive loss properties has a strong bearing on the acceptability of premium and coverage changes for such properties. In addition, a technical study, directed by the 1994 NFIP Reform Act, to examine the economic effects of eliminating the subsidy was released by FEMA during FY 2000. FEMA drafted a multiyear plan to substantially reduce the subsidy and had completed a first round of vetting that plan with other agencies, Congressional staff, and other NFIP stakeholders. The Presidential FY 2002 and FY 2003 Budget proposals contained slightly different subsidy-reduction proposals, neither of which was enacted by Congress. Although the President's FY 2004 Budget proposal was silent on this issue, FEMA continues to refine measures that would reduce the NFIP's level of subsidy. Other public policy objectives that have a bearing on the Program's financial status must be accommodated by the NFIP. It is sound public policy to maximize the number of people who have flood insurance, so as to lessen the reliance on disaster assistance. In recent years, policyholder growth has been only 1% to 2%. This slower policy growth is not due to a lack of new business, but to a high non-renewal or lapse rate. To increase this growth rate, the NFIP is undertaking a new marketing campaign that, while continuing to market to new customers, will also focus on retaining existing policyholders and attracting back those individuals who previously had flood insurance. Although the growth in policyholders has slowed during recent years, average amounts of insurance purchased have also increased, which increases the potential dollar amounts borrowed, even if those amounts are small relative to overall premium volume. And apart from the Pre-FIRM subsidy, it is public policy to encourage the purchase of flood insurance in areas that are known to be experiencing temporary conditions of heightened flood risk, although a 30-day waiting period reduces some of the effects of this adverse selection. The possibility of borrowing funds would be present even if all NFIP policyholders paid full-risk premiums. Twenty-seven percent of policyholders paying significantly less than full-risk premiums impedes the NFIP's ability to generate surplus or to repay borrowed funds, which depends on levels of annual losses that are highly variable. Funding of the Program from policyholder income or potentially from other sources must be addressed in the context of the long-term governmental goals for the NFIP, including its substitution for disaster relief and its encouragement of floodplain management. Subsidized insurance for older construction, built to lower standards in regard to the flood risk and for which full-risk premiums could be unreasonably high, was the quid pro quo for local community adoption of ordinances controlling new construction in the floodplain. It is also a means by which owners of older construction can prefund at least part of their disaster recovery. The NFIP's standards for new construction are now saving an estimated $1.1 billion annually in flood damage avoided. Additionally, it should be recognized that, in fiscal years 1986 through 2003, the NFIP paid out, from policyholder funding, about $10.2 billion in insurance claims, which otherwise would have greatly increased taxpayer-funded disaster relief. Ratemaking Generally accepted actuarial principles require at a minimum that a rating system provide protection against the economic uncertainty associated with chance occurrences by exchanging the uncertainty for a predetermined price. This price for insuring the uncertain event must: • Protect the insurance system's financial soundness; • Be fair; and • Permit economic incentives to operate and thus encourage widespread availability of coverage. For the purpose of setting prices, the broad grouping of risks with similar characteristics is a fundamental precept of a financially sound and equitable system. Because each property at risk is different, a rating system that attempts to identify and reflect in prices every risk characteristic is usually unworkable and costly. The basic features that must be present in sound risk groupings in order to meet the above criteria are: • The system should reflect cost and experience differences on the basis of relevant risk characteristics. • The system should be applied objectively and consistently. • The system should be practical, cost-effective, and responsive to change. • The system should minimize anti-selection. • The system should be acceptable to the public. Also, in the case of flood insurance authorized under Public Law 90-448 (National Flood Insurance Act), the system of insurance and pricing must further the purposes of the Act, which include, among other things, to "(1) encourage State and local governments to make appropriate land use adjustments to constrict the development of land that is exposed to flood damage caused by flood losses, and (2) guide the development of proposed further construction, where practicable [emphasis added], away from locations that are threatened by flood hazards." In order to give practical meaning to these objectives, the standard of a 1% annual chance of flood is now used by virtually all Federal, State, and local agencies and participating communities in the administration of floodplain management programs. The risk of experiencing a flood of this magnitude or larger is one chance in four during a typical 30-year mortgage period. In terms of flood insurance, this standard yields reasonably priced insurance protection to the property owner. The use of a lesser standard approximating pre-1969 building practices would expose future risks to a better than 50% chance of being flood damaged during a 30-year mortgage period and result in insurance rates three to four times those reflecting the "1% annual chance of flood" standard. It was just this consideration of unaffordable full-risk premium (actuarial) rates that prompted Congress to "grandfather" existing construction at subsidized rates. The National Flood Insurance Act of 1968 separated the flood insurance ratemaking process into two distinct categories, namely, chargeable premium (subsidized) rates and estimated- risk premium (actuarial) rates. Subsidized Rates These are countrywide rates by broad occupancy type classifications, which produce a premium income less than the expense and loss payments incurred for the flood insurance policies issued on that basis. The funds needed to supplement the inadequate premium income are provided by the National Flood Insurance Fund. Pre-FIRM Subsidized Rates FEMA has promulgated subsidized rates for use in two cases. The first case is for the Emergency Program (added to the NFIP in 1970). Subsidized rates are also used in the Regular Program on construction or substantial improvement started on or before either December 31, 1974 , or the effective date of the initial FIRM, whichever is later. Exhibit E details the relationship between the amount of subsidized premium to be collected and the amount of premium required to fund the historical average loss year. The Pre-FIRM properties that pay less than full-risk premium are estimated to pay between 35% and 40% of the full-risk premium needed to fund the long-term expectation for losses. Special Post-FIRM Classes That Are Subsidized There are three other cases where classes of business are being subsidized either statutorily or by agreement with Congressional oversight committees. The first of these is the class of risks located in Zone A99 areas that are subject to the 1% annual chance flood but for which structural protection that will protect to that level is at least 50% completed. By statute, rates are charged as if that protection were already in place. A second case, added by statute in 1998, is the class of risks located in Zone AR areas. These are areas for which structural measures have been decertified as no longer providing protection to the "1% annual chance of flood" standard. If the areas meet certain criteria pertaining to a scheduled restoration of protection levels, then rates for new and existing construction are capped at the Pre-FIRM subsidized level. After careful consideration of several public policy issues, FEMA set the initial rates for AR Zones at levels equivalent to X Zone rates. Such rates are substantially lower than the cap allowed by statute. The third case is the class of risks comprised of Post-FIRM construction in the V Zones built between 1975 and 1981. These buildings were built to NFIP standards that accounted for stillwater flood elevations but not the associated wave heights, which were not determinable by the engineering state-of-the-art of the time. In October 1981, the NFIP was able to make use of the latest engineering developments and began to require new construction to be built to more stringent standards and to charge rates that took into account the risks posed by the waves associated with the Base Flood . Because the previously compliant construction would be subject to very high rates if held to the same new standards, discussions with Congressional oversight committee members led to the decision to "grandfather" the 1975-81 construction with less than the full-risk premium rates indicated by the latest knowledge of the risk. Actuarial Rates These rates are promulgated by FEMA for use under the Regular Program (the phase of the National Flood Insurance Program that a community may enter after the initial publication of the FIRM). The actuarial rates are applied in the rating of Post-FIRM construction and second layer limits of insurance on all construction (e.g., in the case of 1- to 4-family residences, amounts of insurance in excess of $35,000). Actuarial rates are based on consideration of the risk involved and accepted actuarial principles. An overview of the actuarial rate calculations utilized in developing the indicated rates can be found in the Appendix. The formula described there follows in principle the "hydrologic method of estimating flood damage risk" outlined in the 1966 U.S. Department of Housing and Urban Development (HUD) report Insurance and Other Programs for Financial Assistance to Flood Victims. There are a few risk zones (Zones A, B, C, D, AO, AH, X, and V) where costs to obtain the hydrologic and topographic information needed to develop flood magnitude and frequency relationships would be extremely high in relation to the floodplain management benefits. Average rates based on actuarial and engineering judgments and underwriting experience have been promulgated for these zones. Overall Rate Level Indications It is important to note that the 1966 HUD report described the "hydrologic method" of ratemaking as a method that "uses available data on the occurrence of floods and damage, but is considerably more sophisticated than merely averaging losses over a period of time." This method of ratemaking, when coupled with special financial arrangements to protect the insurance company pool members against the risk of severe underwriting losses , eventually led to the legal requirements for actuarial rates under the National Flood Insurance Act of 1968. This marriage of ratemaking and financial arrangement with private sector insurers was a necessary outcome. While the actuarial formula is the only valid estimate of flood damage over a very long period of time, the annual provision for flood insurance losses and loss adjustment expenses cannot be accurately predicted with any high degree of certainty. In fact, the estimated amount of losses in any future 1-year period is so uncertain that it can be provided for only by having available large loss reserves and replenishing those reserves by accumulating funds during good years to offset the drain on the reserve during bad years. Since the chargeable rates for so many policyholders are less than the actuarial rates by statute , the ability to accumulate loss reserves during the good years is impeded. However, the achievement of the goal of collecting sufficient premium to cover at least the historical average loss year now allows for some accumulation of reserves during years with loss volume less than the historical average. In view of the catastrophic loss potential, the current statutory method of providing borrowing authority to finance the long-term loss and loss adjustment provision of the flood insurance program makes a good deal of sense. Even though the Federal Government became the sole insurer in 1978, the funding mechanism has essentially remained the same. The NFIP experience over the years 1970 through 2002 clearly demonstrates the uncertainty in the average loss and loss adjustment cost per policy. The annual results are shown in the following table. AVERAGE COST ($) Accident Year Untrended Trended to 05/1/05 Accident Year Untrended Trended to 05/1/05 1970 16.29 1987 53.09 86.55 1971 35.00 1988 25.55 41.53 1972 87.60 1989 311.96 475.28 1973 204.68 1990 74.63 109.62 1974 72.51 1991 148.76 217.01 1975 195.65 1992 289.34 413.48 1976 53.08 1993 254.39 345.90 1977 96.59 1994 148.85 200.02 1978 146.87 406.44 1995 416.14 536.49 1979 311.40 769.74 1996 243.44 305.10 1980 124.92 276.90 1997 142.34 174.53 1981 68.57 138.37 1998 225.14 271.10 1982 110.68 212.07 1999 188.89 219.62 1983 240.31 459.13 2000 60.62 67.70 1984 138.67 250.15 2001 303.11 329.31 1985 199.08 352.35 2002 95.47 102.08 1986 64.60 110.05 In lieu of strictly establishing an overall rate level indication based on historical loss ratio data adjusted to current rate levels and further adjusted for trends impacting on loss costs per policy, the rates for the different classifications are developed by the use of the mathematical models described in the Appendix, or by appropriate selection of rates based upon judgment and review of underwriting experience. FEMA has employed mathematical and computer simulation approaches to define average annualized losses and the concurrent catastrophe loss requirements. With these analytical tools, criteria have been developed to measure the prospective underlying pure premium, to project the probabilities of various levels of borrowing needed to meet catastrophe losses for which prefunded loss reserve has not been established, and to estimate capability to repay borrowed funds. Target Level Premium Analysis In 1981, FEMA established the goal of becoming self-supporting for loss year levels at least equivalent to the historical average loss year. This was accomplished by 1988. Qualifying the target as the historical average as opposed to the long-term expected annual losses is an important distinction. Because NFIP experience since 1978 does not include any loss years of catastrophic levels for the Program, the historical average is significantly less than that which can be expected over the long term where the influence of extremely large loss years would be felt. The importance of targeting the historical average should not be discounted, however. It is the level around which the great preponderance of loss years will concentrate and allows for the accumulation of reserves in years where losses are less than that level to help fund losses in years where they exceed that level. The target level premium established by the historical average loss year allows FEMA to make a judgment during each rate review as to how well the NFIP's self-supporting status is being maintained overall. This "historical average loss year" approach to setting rates accommodates the statutory mandate that premium charges for Pre-FIRM risks, if less than full-risk premiums, must be reasonable. It provides a mathematical basis for determining rates for Pre-FIRM risks, which in the past were determined solely on a political basis, and provides an important framework for making accurate estimates of fiscal soundness. In following through on this approach, the premium charges for the two major categories of business, actuarial and Pre-FIRM subsidized, are developed very differently. Actuarially rated policies are charged premiums that consider the probabilities of the full range of possible losses, including catastrophic levels. Thus, these premiums are targeted at the true long-term average. Written premiums for actuarial policies will generally be greater than those that would be based on the historical average loss year. This is consistent with the expectation that the long-term average annual losses will be higher than the historical experience to date because of the influence of relatively infrequent but catastrophic loss years. Subsidized policies are defined as a category of business that does not make an adequate contribution to the loss reserve pool. These risks are charged premiums that are based on political and statutory considerations that override actuarial considerations. The probabilities of expected and/or catastrophic losses are not contemplated in the rates, which are established for Pre-FIRM construction as rate caps (limitations on chargeable rates) by occupancy type and flood risk zone. FEMA estimates that the premiums for policyholders in this category are between 35% and 40% of what would be charged if the premiums were developed like those charged to the actuarially rated policies. Use of the premium requirements indicated by the historical average loss year as a target level provides a means by which the NFIP can objectively assess its self-supporting status. Typically, during the rate review, it is first determined whether the actuarial rates need to be adjusted. The effects of any such adjustments on maintaining the overall target level are then projected. Adjustments to policy coverage or premiums for Pre-FIRM risks will likely be proposed to make up any overall shortfall so that, once again, the combination of actuarial and subsidized business can generate written premium at least to the level of the NFIP's self-supporting target. This methodology was particularly pertinent during the years leading up to achieving the self-supporting target and the first few years afterward. It is important to note that the historical average is not a static target. If all factors influencing NFIP experience remained constant but for the addition annually of another year to the experience period, the historical average could be expected to rise as it approaches the true long-term average. Other influences that have specific importance in projecting the target level are related to inflation and the expected types of policies to be written, particularly in regard to those paying full-risk premiums versus those that will be subsidized. Even without any shortfall in the overall target level, proposals regarding Pre-FIRM subsidized rates and coverage may be made in order to gradually reduce the amount of subsidy. This has been an important consideration in more recent years, as the NFIP has moved toward maintaining written premium at a level somewhat above the level needed to fund the historical average loss year. The level of subsidy provided in the Program has been the subject of much Congressional debate, and the NFIP reform legislation directed FEMA to study the economic effects of charging actuarially based premium rates for Pre-FIRM structures. PriceWaterhouseCoopers was contracted to conduct this study, and FEMA released the results during FY 2000. FEMA drafted a multiyear plan to substantially reduce the subsidy and had completed a first round of vetting that plan with other agencies, Congressional staff, and other NFIP stakeholders. The Presidential FY 2002 and FY 2003 Budget proposals contained slightly different subsidy-reduction proposals, neither of which was enacted by Congress. Although the President's FY 2004 Budget proposal was silent on this issue, FEMA continues to refine measures that would reduce the NFIP's level of subsidy. Rate Review Results Costs based on the 1978 through 2002 underwriting experience and expected NFIP activities were projected to the 2004-2005 cost levels. Exhibit E shows the premiums required by these projections, the expected average written premiums, and the relationship of the written premium to the historically indicated premiums for flood insurance coverage excluding the premiums for Increased Cost of Compliance coverage. The written premium based on all rate and rule changes through May 2004 is expected to be 124% of the level needed to fund the historical average loss year. The rate and rule changes recommended for May 1, 2004, implementation would result in an overall premium increase of 2.2% and include the following major points: ? No changes to the rates of standard policies in B, C, and X Zones, AR Zones, and A99 Zones. ? An increase in rates for V-Zone categories as follows: Pre-FIRM V Zones, 6%; Post- '81 Post-FIRM V Zones, 5.5%; and Pre-'81 Post-FIRM V Zones , 8%. ? An increase in the rates of AE Zones and A Zones of about 3%. ? There are major proposed revisions to the Preferred Risk Policy (PRP), which are described in more detail below. The chief revisions are: - The PRP is recommended to be extended to additional segments of the NFIP beyond the building/contents combinations of coverage that are currently available only to 1-4 Family buildings. The proposed new versions of the PRP are (1) a contents-only PRP that will be available to all occupancies other than Condominium Unit Owners, and (2) building/contents combinations of coverage for Non-Residential buildings. The current underwriting eligibility requirements will be extended to all new PRP offerings as well. - The contents coverage for the existing PRP that is available to 1-4 Family buildings is proposed to increase to 40% of the stated building amount of coverage. In order to generate sufficient revenue to pay for this increase in contents coverage, we propose to increase premiums by $5 for all coverage options. - The ICC premium for PRPs is proposed to decrease to $1 from the current $6. - The Federal Policy Fee is proposed to increase to $11 from the current $10. Exhibit A provides, by risk zone category, the average increases in premium projected as a result of the May 2004 rate and rule recommendations. Expense Constant Prior to May 1, 2003, FEMA used an Expense Constant—a flat charge per policyholder—to cover certain acquisition costs and general expenses of the NFIP. On May 1, 2003, the Expense Constant was eliminated in a revenue-neutral manner that included increases in the basic limits rates designed to generate approximately the same amount of revenue that the Expense Constant previously did. Starting with this year's rate review, proposed increases to rates have been calculated without a bifurcation of expenses between fixed and variable costs; all expenses are now loaded as variable expenses, with a slight variation of how those expenses are loaded into basic limits and additional limits rates. Federal Policy Fee The expense of flood insurance studies, floodplain management, and FEMA administrative costs is charged to policyholders through the Federal Policy Fee. Under the RCBAP, the fee varies according to the number of units in the building. Preferred Risk Policies will be charged an $11 fee per policy, an increase from the current $10. Other non-RCBAP policies are now charged a fee of $30. On the basis of recent historical trends, the Federal Policy Fee is expected to produce about $106 million in income in 2004-2005. Impact of Community Rating System Policyholders in communities that participate in the Community Rating System (CRS) are eligible for premium discounts based on the creditable activities undertaken by their communities. The impact is considered in the target premium level projections and in their comparison with expected written premium. The success of CRS—both in terms of number of communities and policyholders and in terms of activities undertaken and losses avoided—has continued to grow. Currently, nearly two-thirds of all NFIP policyholders are in participating CRS communities, with discounts ranging from 5% to 40%. As a result of CRS communities' improving their risk classes by adopting additional creditable activities, SFHA policyholders in the participating CRS communities should receive an average premium discount of 12% in 2004. B, C, and X Zones Experience Both standard policies and PRPs in the X Zone had been subjects of scrutiny in the 1996 and 1997 Actuarial Rate Reviews. Preferred Risk Policies (PRPs) Close examination of the PRP led to the conclusion that the poor experience was due, in part, to heavy flood years occurring early in that product's experience period. In addition, the following two requirements necessary to write a PRP policy, implemented in 1998, have tightened the PRP underwriting rules: ? The insured property must be in the X Zone at the time of the policy inception and at each subsequent renewal; hence, no "grandfathering" is allowed. ? The insured property's flood history must meet additional requirements regarding paid insured losses and Federal Disaster Relief payments. The intent of these requirements is to screen out certain repetitively flooded properties from being eligible for the Preferred Risk Policy. As part of this year's rate review, major revisions are proposed for the PRP. Recommendations include (1) increasing contents coverage for the existing PRPs to 40% of the building amount of insurance, (2) extending the PRP to Non-Residential buildings that include building/contents coverage combinations up to the maximum Program Limits for Non-Residential buildings, and (3) introducing a contents-only PRP that will be available to all occupancies, but will not be available to Condominium Unit Owners. We estimate that existing X Zone policyholders who satisfy the underwriting requirements for one of the new PRP options under (2) and (3) will realize premium decreases that will average about 32%. This assumes that they will renew for the same amount of coverage. Other changes are also being proposed for the PRP. As a result of the additional PRP underwriting requirements (described above), ICC premiums are proposed to decrease to $1 from the current $6. However, total PRP premiums for current policyholders will remain the same, since the $5 decrease in ICC premiums will be exactly offset by a $5 increase in the rest of the PRP premium to pay for the increase in contents coverage. Finally, as a result of the expected conversion of existing X Zone "standard" (i.e., non- PRP) policyholders to these new PRP options, revenue generated from the Federal Policy Fee (FPF) will decrease. To at least partially offset that decrease in FPF revenue, the Federal Policy Fee for PRPs will increase to $11 from the current $10. The net result of the changes described in this paragraph is that existing PRP policyholders will see their total bill increase by $1. X Zone Standard Policies (non-PRP policyholders) For standard X Zone policies, rates are adjusted so the premium level relates to the historical indicated premium level at least in the same way as for actuarially rated AE Zone policies. This has resulted in premium increases for the last 5 years that ranged between 3% and 9%, with a cumulative increase during that time period of 35%. Although the relationship of current X Zone premium to historical indicated premium is 127%, while that same relationship for AE Zone policies is 146%, we are proposing no change to X Zone rates. We will continue to closely monitor this relationship and take future increases as needed to achieve that desired relationship. V Zone Experience The increased risk of flooding brought about by erosion has been an area of concern for the NFIP. The 1994 NFIP reform legislation directed a study of a series of possible policy changes to address erosion hazards within Federal programs. The Heinz Center for Science, Economics, and the Environment was contracted to perform this analysis, and the study was released in June 2000. The study results demonstrated that the risk of flooding in those areas of V Zones that are susceptible to erosion will dramatically increase (a two- to three-fold increase in the risk in various areas of the country) during the next 30 to 60 years. The NFIP's ratemaking methodology for V Zones has not directly addressed this increased flood risk brought about by erosion. FEMA is currently investigating ways to do so in the flood maps and the flood rates. The Flood Insurance Rate Maps could be refined to delineate erosion zones. However, that will depend upon funding, development of mapping standards, and political acceptance of higher premiums targeted at those subject to the increased flood risk due to erosion. In May 2001, to partially address the hazard of erosion, the NFIP began a multiyear plan to increase rates for all V Zone policies. The fourth round of increases, which will be part of the May 1, 2004, rate changes, varies between 5% and 8%. Deductibles As part of the May 1, 2003, rate changes, higher deductibles were introduced for Non- Residential policyholders and for RCBAP policyholders. This year, as part of the May 1, 2004, rate changes, slight revisions to some of the relativities for those higher deductibles are proposed. Increased Cost of Compliance (ICC) Coverage The 1994 National Flood Insurance Reform Act (NFIRA) mandated a new coverage to compensate policyholders when they are required to bring their insured structures into compliance with local floodplain ordinances as a result of being substantially damaged by a flood. NFIRA required this new coverage to be actuarially sound, but placed a $75 limit on what any policyholder could be charged. In compliance with these directives, FEMA introduced Increased Cost of Compliance (ICC) coverage in 1996 that provided up to $15,000 of coverage. That amount was subsequently increased, first to $20,000 in 2000, and then to $30,000 in 2003. These increases in coverage were based on analyses of the expected claim frequency under this coverage. FEMA will continue to monitor our ICC experience to assure optimal use of this coverage. In addition, independent of this year's rate review, FEMA is exploring possible additional utilization of the ICC coverage in connection with mitigation offers to policyholders whose insured buildings meet certain substantial-damage or repetitive-damage criteria. We should be able to report on this analysis as part of next year's rate review. Mortgage Portfolio Protection Program (MPPP) The Mortgage Portfolio Protection Program (MPPP) was introduced in 1991 as an additional tool to assist the mortgage lending and servicing industries in bringing their mortgage portfolios into compliance with the flood insurance requirements of the Flood Disaster Protection Act of 1973, as amended. Since the lender or servicer issuing the MPPP policy does not have many of the underwriting data available to it, a policy written through the MPPP requires less underwriting data. As a result, FEMA has target MPPP rates at levels that will compensate us for the greater uncertainty in these risks. Effective May 1, 2003, MPPP rates were increased for the first time in several years. In a continuing effort to assure that these rates are in line with those charged to our non- MPPP policyholders, we are again increasing these rates. Exhibits The following Exhibits include the information below. A. Effects of Revisions on Written Premium B. Insurance Underwriting Experience C. Calendar/Accident Years 1978-2002 Experience for the Larger Risk Zones D. Average Expenses per Policyholder E. Projected Annual Premium Requirements Based on 1978-2002 Loss Experience vs. Projected Written Premium Exhibit A. Effects of Revisions on Written Premium Exhibit B1. Key Underwriting Components by Year, 1978-2002 Exhibit B2. Detailed Underwriting Experience by Year for the Latest 10 Years, Page 1 Exhibit B2 (cont'd.). Detailed Underwriting Experience by Year for the Latest 10 Years, Page 2 Exhibit B3. Detailed Underwriting Experience Aggregated by Experience Period Exhibit B4. Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1978-2002, Page 1 Report: ARPCRPBA FEDERAL EMERGENCY MANAGEMENT AGENCY Exhibit B4 Rundate: Mar 17, 2003 NATIONAL FLOOD INSURANCE PROGRAM Page 2 ACTUARIAL INFORMATION SYSTEM LOSS AND EXPENSE EXPERIENCE Accident Period 1978 - 2002 VE,V1-V30 Pre-FIRM Post-FIRM A Zone AE,A1-A30 AO & AH Emergency Subsidized Program Pre 10/81 Pre-FIRM Pre-FIRM Pre-FIRM Program Totals Totals 1) Earned Exposures 1,179,035 219,513 3,943,701 14,504,691 1,161,075 3,200,517 24,208,532 67,126,327 2) Average Earned Premium $403.87 $350.56 $312.64 $364.81 $373.42 $112.11 $325.09 $249.21 3) Number of Paid Losses 25,415 3,194 71,825 306,737 6,049 104,772 517,992 860,809 4) Average Loss Payment $16,615.70 $19,937.63 $13,603.08 $14,564.69 $12,029.73 $5,641.60 $12,730.67 $13,442.16 5) Loss Ratio 0.89 0.83 0.79 0.84 0.17 1.65 0.84 0.69 6) Loss Frequency per 100 Policy Contracts 2.5 2.1 1.8 2.2 0.5 3.3 2.2 1.4 7) Average Loss Cost per Policyholder $358.16 $290.10 $247.75 $308.01 $62.67 $184.68 $272.40 $172.38 8) Other Expenses (Average per Policyholder) a) Servicing Facility/WYO Operating Allowance $61.11 $57.51 $54.95 $58.47 $59.05 $41.42 $55.79 $51.52 b) Agent Commission $60.58 $52.58 $46.90 $54.72 $56.01 $16.82 $48.76 $37.38 c) Loss Adjuster $12.89 $10.26 $10.01 $12.11 $3.17 $10.45 $11.14 $7.13 d) Total $134.58 $120.36 $111.86 $125.30 $118.23 $68.68 $115.69 $96.04 9) Operating Surplus/(Deficit)* per Policyholder on Paid Basis ($88.87) ($59.89) ($46.96) ($68.49) $192.52 ($141.26) ($63.00) ($19.21) 10) Total Operating Surplus/(Deficit) ($104,785,184) ($13,147,087) ($185,202,664) ($993,431,067) $223,527,589 ($452,108,660) ($1,525,147,073) ($1,289,560,548) * The operating surplus is the policyholder contribution in periods of relatively better loss experience toward reserves used to fund high-loss years. Exhibit B4 (cont'd.). Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1978-2002, Page 2 Exhibit B5. Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1986-2002, Page 1 Exhibit B5 (cont'd.). Detailed Underwriting Experience by Zone and by Actuarial vs. Subsidized, 1986-2002, Page 2 Exhibit C. Calendar/Accident Years 1978-2002 Experience for the Larger Risk Zones Exhibit D. Average Expenses per Policyholder Exhibit E. Projected Annual Premium Requirements Based on 1978-2002 Loss Experience vs. Projected Written Premium APPENDIX Actuarial Rate Formula Actuarial Rate Formula Actuarial rates are applied in the rating of Post-FIRM construction and additional layer limits of insurance on all construction. This Appendix provides an overview of the actuarial rate formula that is utilized in developing these rates. The actuarial rates are based on consideration of the risk involved and accepted actuarial principles. The actuarial rate formula may be expressed as follows: Where: Min = minimum elevation relative to lowest floor at which flood damage occurs. Max = elevation relative to lowest floor at which flood damage approaches a maximum. The variable PELV is the probability of a particular water surface elevation relative to the 100-year Base Flood Elevation (BFE). For example, in Zone A10, the probability of water's rising to or above an elevation 1 foot less than the 100-year flood elevation is 1.6%, and 1 foot or more above the 100-year flood elevation is 0.6%, whereas the probability of water's rising to or above BFE is 1%. There are many risk zones, and they are based on information gathered and calculations made by engineers and hydrologists. Various Federal agencies, such as the U.S. Army Corps of Engineers, and private engineering firms are performing detailed risk zone and elevation studies of all major flood-prone areas. The flood risk zones are determined from these detailed studies and PELV values are assigned to these zones. The results of these studies are published on a Flood Insurance Rate Map (FIRM) showing zones and, where appropriate, BFEs. The assignment of PELV values must be accomplished in such a way as to keep the rating of policies as simple as possible and still distinguish expected average cost differences among the rate zones. There are 30 numbered A Zones for which different sets of PELV values may be assigned. However, there are three main technical reasons for combining risk zones for rating purposes : ? Lowest Floor Elevations are measured to the nearest foot. ? Due to the difficulty in estimating the extremely rare flood, the base frequency curves are truncated at about the 350- to 500-year event. ? The BFEs are approximations based on the best available data about the major sources of flood. As a practical approach, in 1982 five risk zone combinations were established reflecting 1.0 foot elevations, and a minimum elevation difference of 1.5 feet between the maximum flood level and the BFE was established for the risk zones that had the lowest flood hazard factors. Considering the relative variance in flood levels that can occur because of conditions that affect a particular building site during an actual flood, even more averaging for insurance rating is reasonable for buildings constructed with a Lowest Floor Elevation of –1.0 foot or above, relative to the BFE (the elevation of a flood with an exceedance probability of 1%). In 1983, the transition to a single rate schedule was approved. This approach has provided the NFIP with the means for simplifying FIRMs. Since 1985, all new FIRMs have shown at most ten zones. These are A, AE, V, VE, AH, AO, AR, A99, X, and D. Zone AE includes all zones formerly designated as A1-A30, and Zone VE includes all those formerly designated as V1-V30. Zone X encompasses areas formerly shown as Zones B or C. To assure consideration of the maximum flood level that might damage a building located in a Special Flood Hazard Area (even though elevated to the BFE or higher) and to recognize a minimum price associated with the risk transfer, the use of a minimum insurance rate has been continued. This is virtually mandated when adverse selection and the uncertainty of risk elevation are factors as important as they are in flood insurance. The minimum rate is $.16 per $100 of basic limits building coverage. The need to establish minimum values also can be found in the manner that the Flood Insurance Study process treats hydrologic uncertainties. The accepted methods used in the studies tend to underestimate the calculated flood frequencies when there is little or no recorded flood data. Generally, recorded data relating to flooding events exceeding the 100-year event are sparse. The number of years of recorded flood data rarely exceeds a 30-year period. Even in those instances where longer records exist, changes in floodplain characteristics partly invalidate the usefulness of the data. It is generally accepted that the uncertainties involved in calculating the 500-year flood level are significant. Statistical analysis of these calculations has been published in the American Society of Engineers Proceedings. It has been projected that complete reliance on the traditional flood frequency tables in the calculation of insurance rates would produce only about one-half the insurance premium required to meet the insured risk. The variable DELV is the ratio of the flood damage to the value of the insurable property and is obtained from depth percent damage tables. These tables are subject to experience checks by FIMA from a review of actual flood insurance claim files. The DELV values are calculated by weighting the actual insurance claims experience and the previously established depth percent damage values. The weighting is accomplished by using standard actuarial techniques (credibility). The variable LADJ is the loss adjustment expense factor expressed as a percentage of losses (claim payments to policyholders). This provides funds for the payment of loss adjusters' fees and special claims investigation costs that are required to determine the appropriate insurance value of the flood damage and the amount due the policyholder under the terms and conditions of the flood insurance policy. The value of LADJ is currently projected to be 4.2% under the adjuster fee schedule that was implemented on May 1, 1997. The variable DED is the deductible offset. This variable is required to reflect the insurance policy condition that the first $500 of damage does not qualify for an indemnification payment. The factor DED is based on size of claim data produced from insurance claim files. The variable UINS is the under-insurance factor and is included in the formula because flood insurance policyholders do not always insure to value. This requires that the impact of the DELV values in the formula be adjusted to account for the difference between property values and the amount of insurance purchased within basic and additional coverage limits for each category of risk. The value of UINS is determined by a review of insurance claims data. The variable EXLOSS is the expected loss ratio and serves to load the actuarial rates for insurance agents' commissions and other acquisition expenses incurred in the selling of flood insurance policies and a small contingency loading. The contingency loading is 5% in nonvelocity zones and 10% in velocity zones. This concept of targeting premium levels to the "historical average loss year" is explained in more detail in the section entitled "Premium Structure" on page 4. These 7 loss years are 1979 (Hurricane Frederic), 1983 (Hurricane Gloria), 1989 (Hurricane Hugo), 1992 (Hurricanes Andrew and Iniki), 1993 (the Midwest Flood), 1995 (the May New Orleans Flood and a smaller Mississippi Flood), and 2001 (Tropical Storm Allison). This estimate of 27% is composed of 25% Pre-FIRM and 2% other categories. For a more complete discussion of the various subsidized rates categories, see the "Ratemaking" section on pages 6-9. A "FIRM" is a Flood Insurance Rate Map, an official map of a community on which FEMA has delineated both the Special Flood Hazard Areas (SFHAs) and the risk premium zones applicable to the community. "Pre-FIRM" pertains to a building for which construction or substantial improvement occurred on or before December 31, 1974, or before the effective date of an initial FIRM. This additional "grandfathering" was added to the NFIP in 1973. The Base Flood is the flood associated with the Base Flood Elevation (BFE). In other words, there is a 1% chance in any given year that a flood will occur that equals or exceeds the Base Flood. The chance still remained that another severe hurricane like Hurricane Betsy or Camille could have wiped out the private insurers' pledged capital. By statute, all structures in the SFHA that were built before December 31, 1974, or the effective date of the initial FIRM, whichever is later, are to be charged less than actuarial rates. These policies are referred to as Pre-FIRM Subsidized. "Pre-'81 Post-FIRM V Zones" refers to the class of risks comprised of Post-FIRM construction in the V Zone built between 1975 and 1981. These buildings were built to NFIP standards that accounted for stillwater flood elevations but not the associated wave heights, which were not determinable by the engineering state-of-the-art of the time. In October 1981, the NFIP was able to make use of the latest engineering developments and began to require new construction to be built to more stringent standards. "B, C, and X Zones" is abbreviated to "X Zone" throughout this section and elsewhere in the document. As mentioned in the Appendix, since 1985 all new FIRMs have shown a reduced number of zones, with one of those being an X Zone. The X Zone encompasses areas formerly shown as Zones B or C. Some of the factors that increase flood hazard (e.g., local urban drainage problems and urbanization of other parts of the watershed) are virtually impossible to quantify if the Flood Insurance Study process is to remain cost effective. NFIP Actuarial Rate Review November 30, 2003 17 i 1 NFIP Actuarial Rate Review November 30, 2003 18 NFIP Actuarial Rate Review November 30, 2003 19 NFIP Actuarial Rate Review November 30, 2003 27 NFIP Actuarial Rate Review November 30, 2003 28 NFIP Actuarial Rate Review November 30, 2002 A-1 NFIP Actuarial Rate Review November 30, 2003 29 A-1 NFIP Actuarial Rate Review November 30, 2003 A-5 NFIP Actuarial Rate Review November 30, 2003 A-3