The State of Floridas

Property Insurance Market

2nd Annual Report

Released January 2013

for the

Florida Legislature

by

The Florida Catastrophic

Storm Risk Management Center

1

TableofContentsExecutive Summary ........................................................................................................................ 3Introduction ..................................................................................................................................... 8

Framework .................................................................................................................................. 8Stakeholders ................................................................................................................................ 8Current Marketplace Challenges ................................................................................................. 8

Property Insurance Market History ................................................................................................. 9Legislative and Regulatory Activity ......................................................................................... 10State Involvement in Property Insurance .................................................................................. 13

Citizens Property Insurance Corporation .............................................................................. 13The Florida Hurricane Catastrophe Fund ............................................................................. 14Florida Insurance Guaranty Association ............................................................................... 15

Primary Market Structure ............................................................................................................. 15Benefits of Private Market ........................................................................................................ 18

Pricing of Risk ...................................................................................................................... 18Incentives to Mitigate ........................................................................................................... 18Diversification beyond Florida ............................................................................................. 19

Exposure ................................................................................................................................... 19Exposure Trends ................................................................................................................... 20

Pricing ....................................................................................................................................... 23Rate Level Trends ................................................................................................................. 23Role of Catastrophe Models.................................................................................................. 24Reinsurance Pricing .............................................................................................................. 25

Regulatory Environment Factors .............................................................................................. 26Impact of Mandatory Mitigation Credits .............................................................................. 26Non-windstorm Losses: Sinkholes ....................................................................................... 29Non-windstorm Losses: Lengthening of the Property Claims Tail ...................................... 31

Private Market Performance ......................................................................................................... 32Competition............................................................................................................................... 32Private Insurance Availability ................................................................................................... 34Number of Companies and Their Structure .............................................................................. 35Premiums and Surplus .............................................................................................................. 36

Profitability ........................................................................................................................... 38Market Size and Capitalization ............................................................................................. 39Market Concentration ........................................................................................................... 41Leverage ................................................................................................................................ 42

Reinsurance Market Current Structure ......................................................................................... 46Benefits of Reinsurance to the Florida Marketplace ................................................................. 46

Pricing of Risk ...................................................................................................................... 47Access to Capital Markets .................................................................................................... 47The Citizens 2012 Catastrophe Bond Issue .......................................................................... 48

Measuring the Net Effects ............................................................................................................ 48Insurance Market Impact: Changing Exposure Levels ............................................................. 48

State Aggregate Exposure ..................................................................................................... 49

2

Private Market Exposure....................................................................................................... 49FHCF Exposure .................................................................................................................... 50

Consumer Impact: True Cost to Florida Consumers ................................................................ 50Assessment Structures and Processes ................................................................................... 51Subsidies in the Assessments ................................................................................................ 62

Economic Impact: Where do we go from here? ....................................................................... 62Conclusions ................................................................................................................................... 63Appendix A Data Sources & Data Limitations .......................................................................... 64

Quarterly Supplemental Reports ............................................................................................... 64NAIC Database of Insurer Annual Reports .............................................................................. 64

Appendix B Company Listings.. 65 Appendix C -- Citizens Policies and Occupied Housing Units by Legislative District.... 78 Appendix D References ............................................................................................................. 84

3

Executive Summary The number and magnitude of catastrophic events in recent years, such as the September 11, 2001 terrorist attacks, Hurricane Katrina, and the March, 2011 Japanese tsunami, have inspired a rise in the attention devoted to catastrophes. The impact of catastrophes on insurance markets, in particular, has been a subject of attention from policymakers, the media, academics and the general public in recent years. Beginning with Hurricane Andrew in 1992, Florida has been at the forefront of discussions related to natural catastrophes and insurance markets. The state, given its loss history and exposure to hurricanes, provides an excellent setting to analyze the effect that catastrophe exposure and catastrophe losses have on property insurance market operations and performance. This report serves as an update to The State of Floridas Property Insurance Market 2011 report completed by the Florida Catastrophic Storm Risk Management Center in December, 2011. First, the report provides a brief history of events, legislative and regulatory activity, and the state of Floridas involvement in residential property insurance. Second, a thorough analysis of the primary property insurance market in Florida is provided, including information on exposure, pricing, and regulatory factors that have significant effects on insurer performance. Third, the report discusses the data regarding competition, insurance availability, supply-side size and market structure, as well as financial performance of Florida insurers, and provides a comparison to other hazard prone states. Fourth, the reinsurance and other capital markets available to Florida are reviewed. Fifth, the report evaluates the impact of all of these factors by examining the states exposure levels and the effect on consumers of the potential post-loss assessments. Finally, the report poses some open questions regarding the economic impact of the market structure that is currently in place in the state.

The most prominent challenge facing Florida is the current and future exposure to catastrophic windstorms. The current condition of the Florida property insurance market has developed from a confluence of natural and man-made events that have taken place over the last three decades. Hurricane Andrew in 1992 and the combined effects of the 2004/2005 storm seasons, population growth and changing demographics, the evolution of the catastrophe modeling industry, management of catastrophe exposure by insurers/reinsurers, and legislative/regulatory actions in Florida have all contributed to the current market conditions. The lack of significant incentives to mitigate, the dependence on coastal development as an economic driver and the potential impact of climate change on storm activity and damages are all challenges facing the marketplace.

The Historical Context In 1995, as one of several responses to the devastating effect of Andrew on the State of Florida, Florida State Budget Specific Appropriation 1743A created the Academic Task Force on Hurricane Disaster Insurance to review Floridas market system for storm disaster insurance. The Task Force final report, Restoring Floridas Paradise, (popularly known as the Collins Report)

4

proposed a Balanced Equation approach with the goal to reestablish a healthy, competitive private insurance market within two to five years. Characteristics of a healthy, competitive private insurance market mentioned in the Collins Report:

1. Coverage by financially strong private companies of most of Floridas homeowners for hurricane risk;

2. Affordable, competitive rates consistent with widespread coverage of homeowners by private companies;

3. Low numbers of the truly uninsurable in one remaining public insurance entity; 4. A strong Florida Hurricane Catastrophe Fund; and 5. A limit on market share for any one company in high-risk areas.

The proposed plan called on private homeowners insurance companies to continue to pay for most hurricane catastrophe damages. It further recommended cooperation and mutual responsibility among private insurers, individual homeowners, banking industry, private capital markets, and local, State and Federal governments for the health of the private insurance system. The Collins Report outlined major components of its Balanced Equation in an effort to establish the abovementioned healthy, competitive private market characteristics. (The full report is available at www.stormrisk.org.) Since the time of the Collins Report, Florida has seen major changes in its property insurance market system, most of which have not resulted in a healthier insurance market. The state, however, has made strides during this time toward reducing the underlying risk, both through improvements in knowledge of the risk and through strengthened building codes. Internal risk reduction through improvement in knowledge of the risk. Florida leads the other 49 states in the use of catastrophe models to estimate the risk, and created the Florida Commission on Hurricane Loss Projection Methodology (Commission) to evaluate the accuracy and reliability of the modeling process used to set base insurance rates for the homeowners market. Florida requires residential property insurers to use loss costs from a catastrophe model approved by the Commission for residential insurance pricing. The Commission collaborates with modelers to continue improvement of risk knowledge. External risk reduction through strengthened building codes. Early 2012, the Insurance Institute for Business and Home Safety (IBHS) released a report rating building codes in 18 hurricane-prone states located along the Gulf of Mexico and the Atlantic Coast based on an IBHS-directed study. The study measured the building regulatory systems that hurricane-prone states have in place to promote life safety and property protection. Florida rated highest out of the 18 states with an overall score of 95 out of 100.

5

Unfortunately, risk reduction on existing construction has not met with the same success as with new construction due to problems within the My Safe Florida Home (MSFH) and Insurance Mitigation Discounts programs (discussed in detail in The State of Floridas Property Insurance Market 2011 available at www.stormrisk.org). And also unfortunate, Florida must be given low marks for the achievement of a healthy, competitive private residential insurance market. This report focuses primarily on the health status and continuing challenges of the insurance marketplace. Marketplace Challenges The most prominent challenge to the Florida residential insurance market is the states current and future exposure to catastrophic windstorms. Catastrophe exposure has in fact driven the evolution of Floridas property insurance market. The series of actions and reactions that have followed Hurricane Andrew in 1992 all stem from attempting to address this basic underlying driver. Attempts at effective public policy are made more challenging by:

1) The lack of homeowner-perceived incentives to mitigate; 2) Reliance on coastal development as an economic driver; 3) The potential impact of climate change on storm activity; and 4) Affordability of property insurance premiums.

But effective public policy must be accomplished for a return to Florida as a state where citizens and taxpayers have confidence in its disaster insurance market to be possible. Floridas property insurance market, simply put, does not meet the Collins Report healthy market standard on any front. Lack of coverage by financially strong private companies. A shift in private insurance market domination from large, geographically diverse companies to small domestic insurers has taken place since the time of the Collins Report and private market capacity and performance have declined. High numbers and high market share concentrated in one public insurance entity (Citizens). Rather than having low numbers of the truly uninsurable in one remaining public insurance entity, high numbers of both the insurable and uninsurable are today covered by Citizens. Citizens is the primary writer of new insurance policies in Florida, with a total number of new policies written greater than the combined total of the other 9 companies in the Top 10. The growth of the states role in providing residential property insurance may be both evidence of a malfunctioning private insurance market and also a cause of that malfunction. The catastrophe fund. The Florida Hurricane Catastrophe Fund tries to provide financial stability to Citizens and the private market. It relies heavily, however, on its ability to borrow and to assess policyholders in order to pay its costs in the event of a disaster.

6

Questionably affordable, competitive rates inconsistent with widespread coverage of homeowners by private companies. The emphasis of public policy related to property insurance seems to have been placed on insurance affordability, so much so that since the time of the Collins Report, overall homeowners insurance premiums declined while insured values did not. Rates in the private market are now effectively set by Citizens, and it is unknown whether the true price of the insurance is competitive, fair or affordable since potential post-loss assessments could be quite high as well as result in subsidies across Florida policyholders. Information in this Report Capital is needed to support the windstorm risk present in the state of Florida. While there appears to be ample capital to support catastrophic exposure at higher levels of the loss distribution (e.g., reinsurance and capital market products), the availability of capital to support catastrophic exposure at the primary insurer level appears to be less healthy. This can be seen via the:

High number of insurance companies exiting the Florida market and a lack of companies entering to replace them;

Growing number of insurance companies no longer writing new policies in Florida; Lack of new insurance company formation in Florida; and Slow growth of capital supporting premiums written by primary insurers in Florida.

To establish evidence for the concerns mentioned herein, this report analyzes the Florida insurance market in two ways. First, looking at the Florida market individually, the report studies the four major categories of residential property insurers: Citizens, Florida domestic insurers, pup companies, and multiline national insurers. Florida is a market dominated by Citizens and relatively small monoline domestic property insurers. This type of structure raises questions as to the ability of the private market to withstand significant storm activity.

The second analysis compares Florida to other hurricane-prone states. This analysis shows that Floridas structure is unique among similarly exposed states in terms of both the states involvement and the number and market share of independent insurers. Further, it reveals that the insurance marketplace is weaker, overall, than that of other hurricane-prone states.

Data from The State of Floridas Property Insurance Market 2011 have been updated to include 2011 and 2012 data (where available). Much of the media coverage of Floridas property insurance market in 2011 and early 2012 focused on Citizens, so this report includes a significant discussion of the states residual property insurer. Indeed, four updates from The State of Floridas Property Insurance Market 2011 focus specifically on Citizens:

1. The evaluation of Citizens policy counts and assessments now includes a brief analysis of Citizens policy counts by legislative district (See Table 12);

7

2. The discussion of Floridas post-loss assessment structure reflects the legislative changes to Citizens assessments (See Figure 3 and Table 13);

3. A discussion of Citizens 2012 catastrophe bond issue has been added (See the end of Section V Reinsurance Market Structure); and

4. A discussion of Citizens depopulation activities has been included, although it is too early to draw any conclusions about their success or failure (See Section II Property Insurance Market History).

In addition to the added discussion of Citizens, many of the trends noted in The State of Floridas Property Insurance Market 2011 continue:

1. Citizens continues to dominate new policy writing in Florida (See Table 3); 2. While Citizens and the FHCFs financial pictures are better than in 2011, the FHCF may

still not be able to bond adequately to cover potential liabilities (See Table 14); 3. The percentage of PMLs that would be funded through post-loss assessments by Citizens

and the FHCF continues to decline as the number of years without a land-falling storm continues to increase (See Table 14);

4. The role of catastrophe models and model verification continues to remain an issue in pricing in both the primary and reinsurance markets (See Section III Primary Market Structure);

5. Reinsurance (and the FHCF) is still vital to Floridas private primary insurers as retention ratios are continuing to hover around 40% (See Chart 7);

6. The concentration and leverage of Floridas private property insurers is still a significant concern (See Table 5 and Chart 6);

7. Overall exposure levels in Florida have remained relatively level as measured by total insured values (See Chart 1) and PML (See Table 10); and

8. Rates continue to slowly increase but for most market segments remain well below 2005 levels (See Chart 2).

Open questions still remain regarding a variety of issues in the Florida market. Should the state be involved in providing insurance or reinsurance? If so, what is the participation strategy achieves the healthiest private market? How much post-loss financing (if any) should be utilized within the public strategy? What are the economic impacts of the current and future insurance market structures? These questions need to be answered and public policy implemented accordingly if stakeholders are to once again have confidence in the Florida property insurance marketplace.

8

Introduction

Framework This report serves as an update to The State of Floridas Property Insurance Market 2011 report completed by the Florida Catastrophic Storm Risk Management Center in December, 2011, available at www.stormrisk.org. First, the report provides a brief history of events, legislative and regulatory activity, and the state of Floridas involvement in residential property insurance. Second, a thorough analysis of the primary property insurance market in Florida is provided, including information on exposure, pricing, and regulatory factors that have significant effects on insurer performance. Third, the report discusses the data regarding competition, insurance availability, supply-side size and market structure, as well as financial performance of Florida insurers, and provides a comparison to other hazard prone states. Fourth, the reinsurance and other capital markets available to Florida are reviewed. Fifth, the report evaluates the impact of all of these factors by examining the states exposure levels and the effect on consumers of the potential post-loss assessments. Finally, the report poses some open questions regarding the economic impact of the market structure that is currently in place in the state.

Stakeholders There are a variety of stakeholders in the Florida property insurance market. Clearly Florida property owners and primary insurers are stakeholders; however, because of the potential post-loss assessments and economic importance of a functioning insurance market, all current and future Floridians are stakeholders in this market. In addition to the primary insurers, reinsurers and investors in capital market products such as catastrophe bonds are also stakeholders in this market. Finally, public policy makers at the local, state and federal level are stakeholders who need to be considered.1

Current Marketplace Challenges There are a variety of challenges facing the Florida property insurance market. The most prominent is the current and future exposure levels to catastrophic windstorms. Catastrophic exposure levels have driven the evolution of Floridas property insurance market. The series of actions and reactions that have followed Hurricane Andrew in 1992 all stem from attempting to address this basic underlying driver. The lack of significant incentives to mitigate, the reliance on coastal development as an economic driver, and the potential impact of climate change on storm activity are all challenges facing the marketplace. These issues combined with a challenging regulatory environment, insurers and insurer rating agencies focusing on catastrophic exposure, and an unclear cost of capital associated with investing in catastrophic risk make the Florida property insurance market a unique and challenging marketplace.

1 The potential for federal government disaster relief funds makes all Americans potential stakeholders. The analysis presented here does not explicitly account for federal post-loss intervention.

9

Affordability of property insurance premiums is still a significant issue in the state of Florida and it appears that this will continue for some time. Recent efforts to reduce the states role in the insurance market by shrinking the size of Citizens Property Insurance Corporation will be discussed later in this report.

Property Insurance Market History The current condition of the Florida property insurance market has developed from a confluence of natural and man-made events that have taken place over the last three decades. Hurricane Andrew in 1992 and the combined effects of the 2004/2005 storm seasons, population growth and changing demographics, the evolution of the catastrophe modeling industry, management of catastrophe exposure by insurers/reinsurers, and legislative/regulatory actions in Florida have all contributed to the current market conditions. Prior to Hurricane Andrew, Floridas property insurance market was similar to many other states in that the market was dominated by a few large multistate, multiline insurers. Following Andrew, insurers re-evaluated their books of business and reduced their exposure to catastrophes while Florida continued to see substantial population growth and an increase to its exposure to catastrophic damage. These two trends have spurred a series of legislative, regulatory, and insurer actions/reactions that have significantly altered the property insurance landscape in Florida. An analysis of the state of the property insurance market in Florida requires a significant examination of these crucial developments in Florida. The growth of the property insurance residual markets in Florida may be a cause of some private market disruptions, but also may just be a symptom of other underlying issues. The implementation of mitigation credits in Florida in an attempt to reduce property insurance premiums has had a significant impact on insurer strategy and performance, as well as on the size of the residual market. The availability of capital to support catastrophic exposure at the primary insurer level appears to be dwindling. This can be seen via the:

High number of insurance companies exiting the Florida market and a lack of companies entering to replace them;

Growing number of insurance companies no longer writing new policies in Florida; Lack of new insurance company formation in Florida; and Slow growth of capital supporting premiums written by primary insurers in Florida.

It does appear that, despite a weak primary market, there is capital to support catastrophic exposure at higher levels of the loss distribution as the markets for reinsurance and capital market products appear to be healthy.

10

Legislative and Regulatory Activity Regulation and legislation have both played important roles in premium setting during the last decade. The most influential legislative activities:

House Bill 1A, passed in 2007, rolled back Citizens rates, froze rates going forward, allowed policyholders to purchase Citizens policies without first being rejected by the admitted market, and expanded the capacity of the FHCF; and

The Citizens glidepath legislation, passed in 2009 (effective 2010), limited premium increases on any individual Citizens policy to 10% per year.

Recent legislative activity has not had as large an impact on insurance rates. In 2011, Citizens was required to file with Floridas Office of Insurance Regulation (OIR) actuarially fair rates for sinkhole coverage and any rate increases due to sinkholes were to be implemented outside the 10% glidepath restriction on rate increases. In 2012, legislation altered the assessment structure of Citizens. Within Floridas residential property insurance market, the OIR affects premium setting in two primary ways. First, by setting Citizens rates, the OIR implicitly sets market rates, as private insurers are in direct competition with Citizens. Second, and more directly, the OIR is charged with rate approval for private insurers. Recent regulatory activity that had significant effects on premiums (in addition to Citizens rate setting and approval/disapproval of rate increases) involved mitigation credits. These activities included:

Creation of the mitigation discount tables in 2003 by rescaling the ARA mitigation credit study from the average structure being the base house to the weakest structure being the base house, thus ensuring that most inspected homes can be eligible for a mitigation credit and being a worse than average risk results in no surcharges; and

Full implementation of the mitigation credit structure in 2006-2007 without allowing insurers to adjust their base rates to reflect the fact that the weakest structure is the base house. (Insurers used the average house as their base structure to compute their base rates.)

In 2011, Citizens filed for sinkhole rate increases of 447% for HO and 631.5% for Dwelling. The OIR limited Citizens rate increases for sinkhole coverage to 32.8% for HO and 96.5% for Dwelling.

11

Figure 1 shows major legislative and regulatory activities since 1970 affecting Floridas residential property insurance market. Additional legislative and regulatory activities are discussed in more detail throughout this report.

12

13

State Involvement in Property Insurance Florida has developed several property insurance mechanisms.2 In 1970, the Florida Windstorm Underwriting Association (FWUA) was enacted by the Florida Legislature to offer wind only coverage in Monroe County and the Florida Keys. The FWUA was gradually expanded to provide wind coverage in 29 of Floridas 35 coastal counties. Since this initial attempt to provide a public policy response to catastrophic windstorm risk, three entities have evolved with expressly different purposes: Citizens Property Insurance Corporation, the Florida Hurricane Catastrophe Fund, and the Florida Insurance Guaranty Association. Each is briefly introduced here.3

Citizens Property Insurance Corporation After Hurricane Andrew in 1992, the Florida Legislature met in a special session to address problems in the residential insurance market. Several insurers had become insolvent, and others were concerned about increased insolvency risks. The Legislature addressed the need for homeowners insurance policies that provided full (multi-peril) coverage rather than wind-only policies offered by the FWUA. The Florida Residential Property and Casualty Joint Underwriting Association (FRPCJUA) or (JUA) was created in 1992, and later combined with the residual market mechanism that insured commercial residential or condominium and apartment buildings (the Florida Property Casualty Joint Underwriting Association).

The Florida Legislature merged the FWUA with the FRPCJUA, creating Citizens Property Insurance Corporation (Citizens) effective August 1, 2002. Citizens has three distinct accounts: the Personal Lines Account, the Commercial Lines Account, and the Coastal (formerly High-Risk) Account. The Coastal Account consists of policies from the FWUA territories.

When any of these three accounts has a deficit, Citizens may levy assessments. These assessments are not only against its policyholders but also against the policyholders of private insurers in almost all lines of property casualty insurance. A more detailed discussion of the assessments is found in Section VI (Measuring the Net Effects) of this report. While developed as a residual market entity, Citizens now accepts nearly all insurable applicants. Legislative changes in 2007 (based on House Bill 1A, or HB1A) removed eligibility restrictions on Citizens as a competitor with private insurers. Prior to HB1A a policyholder had to be rejected in the private market or show that admitted insurers were charging 25% more or higher than Citizens to be eligible for Citizens coverage. HB1A removed the requirement that a policyholder must be rejected by the private admitted insurance market to be eligible for Citizens coverage. After the passage of HB1A, the minimum premium difference requirement for 2 For a complete discussion of residual market mechanisms and their development in Florida see Newman (2010) and Newman (2009). 3 See Cole et. al. (2009).

14

Citizens eligibility was reduced from 25% to 15%. It does not appear, however, that there are any enforcement mechanisms in place to verify the premium difference requirement. Table 1 shows the growth in Citizens policy count from end-of-year 2003 through 2011. The table also indicates an overall decrease (or at least lack of growth) in Citizens policy count from end-of-year 2011 through November, 2012.

Table 1: Citizens Property Insurance Corporation Personal Residential Policy Counts PLA

(Multi-peril)

Coastal (Wind Only)

Coastal (Multi-peril)

Total Personal Residential Policy

Count Nov. 2012 888,423 240,184 165,729 1,294,336 Apr. 2012 984,621 242,689 176,056 1,403,366 Dec. 2011 1,003,856 245,506 173,798 1,423,160 Dec. 2010 829,406 248,328 154,663 1,232,397 Dec. 2009 609,652 251,287 114,561 975,500 Dec. 2008 629,467 328,775 67,672 1,025,914 Dec. 2007 845,857 421,505 24,676 1,292,038 Dec. 2006 743,592 403,509 1,147,101 Dec. 2005 407,387 399,418 806,805 Dec. 2004 416,529 453,765 870,294 Dec. 2003 383,280 433,056 816,336 Source: www.citizensfla.com, various financial statements As of December 2011, Citizens insured in excess of 1.4 million Florida personal residential policies. By the end of January 2012, Citizens policy count peaked at nearly 1.5 million policyholders.4 More than 1 million of these policies were in the multi-peril PLA category, meaning they covered non-coastal properties. As seen in Table 1, the number of policies in the Coastal account has been relatively stable, while the number of policyholders in the PLA account more than doubled between end-of-year 2005 and 2011. This upward trend was partially driven by sinkhole issues, as Citizens has added nearly 75,000 policies (a 65% increase) in Hernando, Hillsborough, and Pasco counties between the third quarter 2009 and April 2012. The Citizens market share (based on policy count) in those counties increased from 23% (third quarter 2009) to more than 42% (end-of-year 2012) in two years. Strategies to reduce the policy count in Citizens may be working as Citizens policy count reduced by more than 150,000 policies by November, 2012 from its high point in January, 2012.

The Florida Hurricane Catastrophe Fund The Florida Hurricane Catastrophe Fund (FHCF) was created by the Florida Legislature in 1993 to provide additional insurance capacity and help stabilize the property insurance market in Florida (Fla. Stat. s. 215.555(1)). The FHCF provides reimbursement for a portion of a property

4Citizens had 1,475,000 policyholders as of January 31, 2012.

15

insurers hurricane losses above the amount retained by the insurers. Insurers enter into contracts with the FHCF and pay a premium. The FHCF is able to accumulate premium payments on a tax- free basis as it is exempt from federal income taxation. The FHCFs total potential obligation is set by statute. For the 2012-2013 storm year, the FHCFs total potential obligation for the mandatory layer is $17 billion. There is also some exposure for the FHCF in the Temporary Increase in Coverage (TICL) layer. The exposure in this layer will depend on insurance companies opting to purchase a contract in that layer (they have until June 1, 2012). As of early May, the FHCF had only $317 million in exposure in the TICL layer.5 In the event that the FHCFs losses exceed its surplus, the FHCF is authorized to collect assessments on policyholders in almost all lines of property casualty insurance. The amount of coverage available from the FHCF, the cost of the coverage, and the potential assessments are significant factors in the state of the insurance market. A more detailed description is found in Section VI (Measuring the Net Effects) of this report.

Florida Insurance Guaranty Association Although not a residual market, a discussion of The Florida Insurance Guaranty Association (FIGA) seems appropriate at this point. FIGA was created by the Florida Legislature in 1970 to address concerns about the adverse effects of insolvent insurers. Its specific purpose is to provide a mechanism for the payment of covered claims under certain insurance policies to avoid excessive delay in payment and to avoid financial loss to claimants or policyholders because of the insolvency of an insurer. (Section 631.51(1), F.S.) FIGA does not accumulate funds in advance of an insurers insolvency but, similar to Citizens and the FHCF, obtains funds through pro-rata assessments levied by the Office of Insurance Regulation on companies subject to assessment. A more complete discussion of FIGA is found in Section VI (Measuring the Net Effects).

Primary Market Structure The market for residential property insurance in Florida has been scrutinized extensively since 1993, in the wake of Hurricane Andrew. Market health is determined largely by the interplay between prices and availability. If insurers determine they have underestimated the costs of providing coverage, it is imperative they are able to adjust estimates (and prices) upward to avoid potential insolvency (i.e., the inability to keep the insuring promise made to policyholders). If they cannot raise prices adequately to pay expected losses, they might leave the market. Conversely, if insurers determine they have overestimated the costs of providing coverage, it is imperative they are able to adjust estimates (and prices) downward before they begin to lose

5 See the May 10, 2012 Claims Paying Estimates report for the FHCF done by Raymond James.

16

market share to the competition, which is likely growing given the short-term possibility of excess profits. The state of Florida has more than $1.8 trillion in insured residential property exposure (see Chart 1). The state has more than $4 billion in expected average annual losses (AAL) due to windstorms ($4.17 billion according to RMS v11.0, $4.35 billion according to AIR 12.0.1) and nearly $60 billion in 1-in-100 probable maximum losses (PML) due to windstorms ($52.6 billion RMS v11.0, $61 billion AIR 12.0.1).6 To support this type of catastrophic exposure, Florida needs a large diverse capital base available to pay for losses if and when they occur. Table 2 shows the probable maximum losses (PML) per occurrence faced by residential policyholders in Florida.

Table 2: Probable Maximum Windstorm Loss Amounts for Florida by Return Period

Return Time (Years)

Gross PML

50 $38 B 100 $57 B 250 $87 B

Source: 2012 Annual report of aggregate net probable maximum losses, financing options, and potential assessments Per-occurrence PML amounts represent large loss amounts that are estimated to be exceeded only once during a return period. For instance, a PML of $38 billion with a 50-year return period can be interpreted to mean that loss costs from a single storm are estimated to exceed $38 billion only once every 50 years, over the long run on average. Another way to interpret the value is to restate the return period as a probability: There is a 2% (1/50) probability that the loss costs from a single storm will exceed $38 billion. When large losses occur, capital is needed to pay for the losses. Potential capital sources include residential property owners, the private insurance market, the State and its affiliated entities, the Federal Government, and tax payers (within Florida and nationally). Property owners in Florida are the first tier source of capital, as they will need to be able to pay for any uninsured damage as well as any deductibles on insured losses to their property as well as any losses over policy limits. The second tier of capital is Floridas primary insurance market; the current structure of which includes both the private insurance market and the States residual market. The private market for primary insurance includes insurers admitted by the OIR to compete in Florida for standard business as well as non-admitted insurers selling only in the surplus lines. We do not have

6The AALs and PMLs are for residential and commercial residential properties only. Commercial residential includes condominiums, apartment complexes, etc., including the common elements in those complexes.

17

significant data on the premiums collected or the amount of exposure currently being insured in the surplus lines market.7 This report focuses on the admitted insurers and Citizens and partitions the Florida property insurance market into four segments of primary insurers:

1. Citizens. Floridas state-sponsored property insurer. 2. Florida domestic insurers (domestics). Florida-domiciled insurers who write primarily

Florida property insurance. Many of these domestics began operating after Hurricane Andrew. The Insurance Capital Build-Up Incentive Program provided start-up surplus loans to 13 new property insurers in 2006-2007.

3. National company subsidiaries (pups). Florida-domiciled subsidiaries of national insurers. These pups are focused solely on property insurance in Florida while still members of the parents insurance group. The subsidiary structure allows national insurers to isolate the parents assets from Floridas catastrophic property exposure.

4. National insurers and others. These insurers are not Florida-based and do not fall in any of the above categories. Most are traditional national insurers with a multi-line focus.

A list of company names and National Association of Insurance Commissioners (NAIC) codes used in this report, along with assignments to these Florida segments, is included as Appendix B. Excluding Citizens and non-admitted (surplus lines) insurers, Floridas primary insurance market collected approximately $6.45 billion in residential property insurance premiums that include wind coverage in 2011. Citizens collected nearly $2.5 billion in residential property premiums in 2011 that include wind coverage. Residential property insurance includes homeowners, dwelling, condominium owners, and allied lines that included wind coverage.8 The total homeowners insurance premiums collected in 2011 were approximately $7.7 billion ($6.0B in private market and $1.7B for Citizens).9 While this is enough to cover the AAL ($4-

7 While most surplus lines business has traditionally been commercial lines coverages, the role of the non-admitted market has expanded greatly in providing personal lines property insurance coverage in hazard prone areas. The non-admitted market adds significant capacity in high risk areas, but their policies are generally not protected by the states guaranty funds. The Florida Surplus Lines Insurance Offices 2010 Annual Report shows that surplus lines premiums in the state of Florida were more than $3.8 billion in 2010. Of that, more than $1.3 billion (both commercial and personal lines, property and liability) in premiums were written in Miami-Dade, Broward, and Palm Beach counties (southeastern Florida). 8 Two main datasets were used in the analysis for this paper, see Appendix A. The first dataset is the QUASR dataset maintained by the Florida Office of Insurance Regulation. When data from that dataset is referenced the terminology will reference residential property insurance premiums. The second dataset utilized is the NAIC annual statement data. That dataset organizes data by line of business which is slightly different than the categories used by the QUASR data. The homeowners insurance line of business in the NAIC dataset does not include non-owner occupied dwelling policies, which are included in the QUASR residential property insurance premiums. In the NAIC data, non-owner occupied dwelling policies are included in Allied lines. This difference in categorization makes apples to apples comparisons between datasets difficult, but both datasets were used because the NAIC dataset allows for cross state comparisons that the QUASR data does not allow. 9 Source: NAIC annual statement database.

18

$4.5 billion) from windstorms, it does not approach the amount needed to cover the PMLs.10 Recall that total losses from the 2004-2005 storms were more than $35 billion and the 1-in-100-year PML is nearly $60 billion. Capital is needed to support the deviation from expected loss amounts that can occur. Through reinsurance transactions, the FHCF and the global reinsurance market are also suppliers of capital to support Floridas catastrophic risk.11 Citizens, the FHCF, and FIGA all utilize post-loss assessments to supply capital, which means that almost all property casualty policyholders in Florida ultimately supply a portion of their capital. The final supplier of capital may be the Federal Government and ultimately all taxpayers if federal aid is necessary (and available) to pay for losses.

Benefits of Private Market Many reports have discussed the value of having a properly functioning and solvent private market for property insurance. There are three main benefits to having the private market: accurate pricing of risk, incentives to mitigate, and diversification of risk beyond Florida.

Pricing of Risk One of the major advantages of having a private market for insurance coverage is the market forces that will work to accurately price risk. Competition will ensure that insurers cannot overcharge for coverage. If an insurer is charging too much, other insurers in the marketplace will be able to gain market share. In addition, insurers will not be able to undercharge for insurance and remain solvent. The accurate pricing of the risk will include the appropriate market determined risk load or return on capital. To entice investors to be willing to put their capital at risk in Floridas property insurance market, they must be adequately compensated. The rate of compensation is market determined and varies with market conditions, including (but not limited to) other investment options, supply and demand for capital, volatility of losses, and interest rates.

Incentives to Mitigate Accurate pricing of risk will also lead to proper incentives to mitigate. Although the insurance mechanism is not the only mechanism that could or should be used to incentivize mitigation, it is one of them. Accurate risk pricing provides policyholders with optimal location and mitigation incentives by internalizing the cost of catastrophic risk to those who are exposed to the risk. Mitigation will reduce the cost of risk and the expected cost of losses. Therefore, policyholders forced to bear the accurate cost of their risk can make rational mitigation decisions and invest in 10These premiums must also cover losses from all other causes of loss covered by a homeowners policy. 11Both insurers and reinsurers have a variety of capital market products that may be used to draw investors and capital to support catastrophic risk. These products can include: catastrophe bonds, catastrophe swaps, industry loss warranties (ILWs) and other insurance-linked securities. These markets, while growing, do not supply a significant portion of the capital in Florida or anywhere.

19

cost effective mitigation strategies. It is important to remember that avoidance is also a mitigation strategy. For Florida, this strategy could be seen as individuals chosing not to relocate to Florida or a reduction in new building in hazardous areas.

Diversification beyond Florida Private market insurers can benefit from the law of large numbers and diversification benefits of having a book of business that contains some property exposure in Florida along with exposures in other parts of the country (or world). This type of diversification ensures that the cost of risk can be minimized and the insurer has funds that can flow in and out of the State to pay for losses when they occur. State run insurers do not have the same ability to diversify across state lines.

Exposure The sources and cost of the capital available to support Floridas catastrophic exposure have evolved over the last 25 years. Figure 2 shows the major changes that have occurred in Floridas private residential insurance market since Hurricane Andrew in 1992. Formation of the pups and domestics occurred almost entirely during this period. Since 2006, several of the domestics have become insolvent. While some of these insolvencies can be traced directly to the 2004-05 storm seasons, the most recent insolvencies (in 2010-11) are not necessarily traceable to specific catastrophic losses. Citizens, the FHCF and FIGA are all relying on debt as a significant source of capital in the event of a major loss. As will be discussed in more detail in Section VI (Measuring the Net Effects), approximately 30% of 50- or 100-year return time losses in 2012 will be financed through assessments from Citizens and the FHCF. This does not include any assessments that may be necessary from FIGA for insolvent insurers. In Section IV (Private Market Performance), we discuss how the amount of capital available from the private, admitted insurers market has been decreasing while exposure has increased. Reinsurers provide a significant source of capital and diversification to the Florida insurance marketplace. As the private primary market has become one of smaller and less diversified companies, more reinsurance is necessarily being used. Citizens market share continued to grow. As of year-end 2011, the residual entity wrote 53% of the Dwelling/Fire, 40% of the Allied Lines, and 22% of the Regular Homeowners markets (multi-peril and wind only policies), respectively. Citizens now underwrites a combined total of 27% of the personal residential insurance premiums in Florida, excluding mobile homes.12

12 Source: OIR QUASR market share reports.

20

Exposure Trends Property insurers measure exposure to risk by units of insured value, which ultimately represents the size of the promises they have made in contracts. Insured values are based on principle of indemnity in the insurance contract, and should reflect replacement costs of construction, not market real estate values. Real estate values and construction costs, however, often increase in tandem when labor and materials are scarce. The housing depression in Florida has resulted in slack markets and only very slow inflation in replacement costs since 2007Q2, but not deflation of such costs. The path of statewide insured values is shown in Chart 1. As the personal lines policy count has remained roughly flat during the analysis window, at 5.7 million (plus or minus 100,000), the increase in exposures up through 2007 is almost entirely based on increases in the insured value of properties rather than growth in policy count.

The total insured value has been near $1.8 trillion since early 2008. Insured exposure values decreased slightly over the last two years and population growth has slowed significantly over the last few years coinciding with the drastic changes in the Florida housing markets.13 It is important to reiterate that insurance provides replacement cost coverage on structures, so market value of property is not a good measure of exposure. While construction costs (and therefore replacement costs) vary, they do not necessarily move with housing market prices; construction costs have increased only slightly. This, combined with slow construction, implies that exposure, as measured by replacement costs, has remained relatively stagnant during this time period.

13 See University of Florida Study on Florida Population (2011).

21

Chart 1: Trend in Florida Personal Residential Property Insured Values (2005-201214)

14 The insurance industry has generally become more sophisticated about monitoring insured values in accordance with the insure-to-value requirement of most homeowners insurance policies. The net rise since 2007Q2 to just under $1.9 trillion in insured value, despite the slack economy, likely reflects insurance-to-value efforts.

22

Figure 2: Timeline of Major Changes in Floridas Private Market for Residential Property Insurance

1992 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011WeatherEvents

HurricaneAndrew

HurricanesCharley,Frances,Ivan,Jeanne

HurricanesDennis,Katrina,Rita,Wilma

PupFormation

TravelersofFloridaestablished(FirstFloridianAuto&Home)AllstateFloridianestablished(CastleKeyInsurance)

StateFarmFloridaestablishedNationwideInsuranceCompanyofFloridaestablished

EncompassFloridianestablished(Allstate)

CapitalIncentivePrograms

InsuranceCapitalBuildUpIncentiveProgram

DomesticInsurerEntries

3 3 8 5 2 1 1 4 8 6 9 4 1

DomesticInsurerInsolvencies

AtlanticPreferredIns.Co.(PoeGroup)FloridaPreferredPropertyIns.Co.SouthernFamilyIns.Co.

VanguardFire&Casualty

CoralIns.Co.MagnoliaIns.Co.NorthernCapitalIns.Co.

HomewisePreferredIns.Co.(receivership)

23

Pricing There are a number of factors that affect pricing of residential property insurance in Florida. The legislative and regulatory factors previously discussed had a significant impact on rates and therefore premiums over the last six years. In addition, the growing role of catastrophe models and reinsurance pricing are all highly influential factors in pricing. This section will discuss each of these factors.

Rate Level Trends Rate levels actually earned by insurers are best measured by the ratio of premium to insured value premiums per unit of risk. Rate is the amount charged per $1,000 of insured value. Similar to exposure, changes in rates inherently have significant impact on insurance premiums. Florida has seen notable volatility in insurance rates over the last decade, including a relatively high 15% increase occurring between year-end 2005 and 2007.15 House Bill 1A, which took effect in early 2007, included the expansion of the FHCF along with increases to the minimum mitigation credit requirements and changes to Citizens rate structure and thereby reduced insurance rates. Rates declined through the end of 2009 to below year-end 2005 levels and still remain below that baseline on a statewide average basis. Rate levels for domestic companies and Citizens generally declined the most. This rate erosion appears to have bottomed out in late 2009 or early 2010. Rates in each of the four company categories have been climbing since 2010, but still need significant growth to reach 2005 levels across the board. By late 2012, it appears only rates for the pup and other (mostly national group) companies are close to regaining their end of 2005 levels. Rate decreases combined with stagnant insured values imply significant erosion in the statewide premium base. Chart 2 shows the rate trends over the last seven years in Florida.

15 Rates rose about 15% statewide during the run-up of 2006, but insured values rose by about twice as much (in percentage terms). Said differently, about two of every three dollars of premium increases seen by consumers in 2006-2007 were due to recognition of increased exposure, not increased rates.

24

Chart 2: Trend in Florida Personal Residential Property Insurance Rate Levels (2005-2012)

Source: QUASR

Role of Catastrophe Models The traditional rating methodology of using historical loss costs aggregated to territorial levels has proven unreliable with regard to catastrophic perils such as hurricanes or earthquakes (Nyce and Maroney, 2010). Because of the nature of these perils (i.e., the frequency, severity and correlation of exposures), traditional rating methods have been replaced by catastrophe models. These models are forward-looking computer simulations incorporating a variety of sciences (including, but not limited to engineering, meteorology, geology, geography, and oceanography) to develop average annual loss costs, not at a class or territory level but at the individual property level. These models do not rely on homogeneity of exposures to develop historic loss costs, but on scientific estimates of storm frequency and severity and engineering estimates of structural damage to determine a loss estimate at the property level. The catastrophe models are reviewed and approved for use for setting rates in Florida by the Florida Commission on Hurricane Loss Projection Methodology (Commission). The shift to catastrophe models would likely lead to a fundamental shift in the rate-making methodology for the windstorm portion of property insurance premiums. Insurers would not determine the rate they need for their book of business but the rate they need for each individual

25

property insured. Although catastrophe models generate average annual losses (AALs) at the property level, rating plans filed by insurers (in Florida and other states) still rely on territorial rating for homeowners and wind-only property coverage. A switch to this bottom-up rate making depends on the accuracy of the catastrophe models, the validation of the relevant construction features, and the verification of their existence on the property for windstorm risk rating. This type of approach raises a number of questions as to the role of the catastrophe models in setting wind rates. Are they a replacement for historical data? How accurate are they? Can the loss costs they generate be used as the sole determinant of wind rates? If so, then the roles of model verification (such as the Florida Commission on Hurricane Loss Projection Methodology) and data accuracy become increasingly important. The variation across models (Cole and McCullough, 2010) and the volatility found from one version to the next is problematic. For example, a recent study indicates significant differences in predicted average loss costs per $1,000 value depending on whether using:

1) AIR 12.0.1 or AIR 11.0; 2) RMS 11.0a or RMS 8.0a; and 3) AIR 12.0.1 and RMS 11.0a.16

These models represent the current best practices in prediction of windstorm losses and are vital to appropriate pricing of property insurance in Florida and other windstorm prone states. Catastrophe models are updated as technological advances and new information become available. These updates can have a significant impact on predicted loss costs and, therefore, on indicated rates and premiums. This is a relatively new and developing field. As with any new field there will be mistakes and significant changes or developments that have substantial impact. These changes can take place in any of the fields of science (meteorology, oceanography, etc.), engineering, technology, information or market evolution and will be continuing for the foreseeable future.

Reinsurance Pricing The evolution of the Florida property insurance market, from being a market dominated by large, diverse national insurers with significant surplus to becoming a market dominated by smaller, monoline (or virtually monoline) and geographically focused insurers has made the market increasingly reliant on the global reinsurance market as its primary source of diversification and risk capital. Reinsurers are the primary link between Florida property insurance companies and the broader capital markets that may be interested in making capital investments in the catastrophe risk market. In 2012, Citizens issued their first catastrophe bond without going 16 Rollins Analytics, Inc., in March, 2011, completed a comparative analysis of loss costs generated by recent models from AIR and RMS. The analysis compared predicted average loss costs for each wood frame, masonry and mobile home structure on a county-by-county basis.

26

through a reinsurer. Details on that catastrophe bond offering are discussed later in this report. The reliance on reinsurance affects rates and premiums in Florida in two ways. First, reinsurers can set risk-based rates using their choice of catastrophe models that may or may not have been approved by the Commission. The models used by reinsurers may result in significantly different rate indications than those used by Florida property insurers. Second, as the link between Florida windstorm risk and the capital markets, reinsurers are the best indicator of the cost of capital needed to entice investors to Floridas catastrophe risk market. Because this cost of capital is dependent on market conditions and other investment opportunities to which investors have access, it can fluctuate widely and have significant impact on reinsurance rates, premiums, and private market insurance stability in Florida.

Regulatory Environment Factors As noted in the Property Insurance Market History section, regulation and legislation have both played important roles in premium setting during the last decade. The most influential legislative activities were:

House Bill 1A, passed in 2007 , rolled back Citizens rates, froze rates going forward, allowed policyholders to purchase Citizens policies without first being rejected by the admitted market, and expanded the capacity of the FHCF; and

The Citizens glidepath legislation, passed in 2009 (effective 2010), limited premium increases on any individual Citizens policy to 10% per year.

In addition to the role that regulation plays in premium setting (setting Citizens rates and rate approval for private insurers), recent regulatory activity that has had a significant impact on premiums involved mitigation credits. These activities included:

Creation of the mitigation discount tables in 2003 by rescaling the ARA mitigation credit study from the average structure being the base house to the weakest structure being the base house, thus ensuring that most inspected homes can be eligible for a mitigation credit and being a worse than average risk results in no surcharges; and

Full implementation of the mitigation credit structure in 2006-2007 without allowing insurers to adjust their base rates to reflect the fact that the weakest structure is the base house. (Insurers used the average house as their base structure to compute their base rates.)

Impact of Mandatory Mitigation Credits One of the methods of reducing Floridas current exposure to windstorm damage is to mitigate the existing building stock. The reduction in average annual losses and probable maximum losses due to windstorms can be reflected in insurance premiums through mitigation credits applied to the property insurance premiums on deserving properties. The use of mitigation credits can serve as an incentive to property owners to undertake cost effective mitigation

27

activities such as installing storm shutters or other types of opening protection. However, an improperly structured mitigation credit program can reduce a property owners mitigation incentives or have other perverse effects on the insurance market. In 1993, the Florida Legislature enacted section 627.0629 of the Florida Statutes. This statute required rate filings for residential property insurance to include appropriate discounts, credits, or other rate differentials (or appropriate reductions in deductibles) for properties on which fixtures have been installed that are actuarially demonstrated to reduce the amount of loss in a windstorm (s.13, ch.93-410, Laws of Florida). In 1997, the Department of Insurance issued rule 69O-170.017 F.A.C., which required shutter discounts to be at least equal to the Insurance Services Office (ISO) discounts.17 In 2000, section 627.0629 of the Florida Statutes was amended to provide that rate filings for residential property insurance must include actuarially reasonable discounts, credits, or other rate differentials, or appropriate reductions in deductibles for properties on which fixtures or construction techniques demonstrated to reduce the amount of loss in a windstorm have been installed or implemented. New language added to the statute stated that:

The fixtures or construction techniques shall include, but not be limited to, fixtures or construction techniques which enhance roof strength, roof covering performance, roof-to-wall strength, wall-to-floor-to-foundation strength, opening protection, and window, door, and skylight strength. Credits, discounts, or other rate differentials for fixtures and construction techniques which meet the minimum requirements of the Florida Building Code must be included in the rate filing. All insurance companies must make a rate filing which includes the credits, discounts, or other rate differentials by June 1, 2002 (s.99, ch.2000-141, Laws of Florida).

The Legislature subsequently amended the law and changed the filing date to December 31, 2002, and then to February 28, 2003. In 2002, Applied Research Associates (ARA) conducted two studies to quantify wind loss reduction for wind mitigation construction features. Development of Loss Relativities for Wind Resistive Features of Residential Structures focused on single-family homes. Development of Loss Relativities for Wind Resistive Features for Residential Buildings with Five or More Units addressed condominium and renter occupancies in buildings with five or more units. The Florida Office of Insurance Regulation (OIR) issued Informational Memorandum OIR-03-001M on January 23, 2003. In essence, the Memorandum states that only premium credits should be offered. Thus, the results of implementation would be premium neutral or result in premium decreases, not premium increases. The Memorandum goes on to state, Credits were then

17 See Mitigation Credit Study (2010).

28

determined and tempered by 50%. This tempering was applied in view of the large rate changes which might otherwise be induced, the approximations needed to produce practical results (such as the specifications of the houses used for modeling and the number of rating factors used), and the potential for differences in results using different hurricane models. Rule 69O-170.017 F.A.C. was amended effective December 16, 2006, to require insurers to make new rate filings by March 1, 2007, to double the credits to 100% of the studys indicated value or provide actuarial justification for an alternative system. Informational Memorandum OIR-07-03M issued February 27, 2007, stated that the windstorm mitigation discount filing shall not include any modification of the rating factors or base rates for any purpose, including the offset of revenue impact on current business. To date, we are not aware of any insurer providing actuarial justification for an alternative mitigation credit system to the OIR. ARA updated the study of mitigation credits in 2008, but the recommendations and relativities of that study have not been implemented. The Uniform Home Grading System was to be implemented in 2011 to replace the current mitigation discount system. It would have required discounts to be based on a numerical score of the Uniform Home Grading System. The system has not been implemented. Following the implementation of the mitigation credits in 2007, the claims-paying ability (policyholders surplus) of Floridas homeowners insurers declined. The implementation also reduced the incentives of many policyholders to undertake mitigation activities. This may best be illustrated through an example.

Suppose Bob owns a home that has a hip roof and hurricane straps (both of which qualify for mitigation credits). Bobs insurance company is not aware of these mitigation features and charges him the premium for an average property. (This was the common practice in the industry as the base rates charged by insurers were based on the average property they insured.) As the property owner, Bob chooses to get a mitigation inspection, which reveals these features to his insurer. According to the loss relativities from the 2002 ARA mitigation study, these features should result in a premium credit as follows: The ARA loss relativity for the average house is 1.0 (the average loss amount), but the loss relativity for a house with a hip roof and hurricane straps is 0.91 (i.e., losses are expected to be 9% lower than the average home because of these features). Therefore, Bob should receive a 9% premium credit for mitigation features. Informational Memorandum OIR-03-001M issued by the OIR in 2003 created a mitigation discount table based on ARAs loss relativities table, but made the weakest structure rather than the average structure the base. In the ARA table the weakest structure represents a relativity of 2.37, meaning losses at the structure are expected to

29

be 2.37 times as bad as losses to the average structure (with its loss relativity equal to 1.0). The OIRs mitigation discount table divides every relativity by 2.37, such that the weakest structures loss relativity becomes 1.0, and Bobs house with the hip roof and hurricane straps warrants a relativity of 0.384 rather than 0.91. This indicates the losses at Bobs house are expected to be 62% (1-.384) lower than they would be at the weakest house. The OIRs mitigation table shows this 62% credit. Therefore, by having a mitigation inspection, Bob now receive a 62% discount on his wind premium. This would not be a problem if the base rates charged by the insurer were based on the weakest structure. Informational Memorandum OIR-07-03M issued by the OIR in 2007, however, has prevented insurers from adjusting their base rates (set at the average structure) to the weakest structure. In practice, the following occurs. Suppose that the base wind premium being charged by the insurer is $1,000. The hip roof and straps should reduce that premium to $910, a 9% premium credit. By resetting the mitigation credit table to the weakest structure and not allowing a change in the base rate, OIR-07-03M effectively changes the premium credit to 62% and the wind premium for the house to $380, well below what is actuarially indicated by the ARA loss relativities.

These types of legislative and regulatory interventions into the marketplace create significant uncertainty to private insurers and add to the cost of doing business in Florida. While there have been no studies that examine how significant these events have been in the decision-making process of insurers deciding to sell homeowners insurance in the state of Florida, there is substantial anecdotal evidence that it is important. Even within Floridas existing market of private insurers, those interested in taking out policies from Citizens (and thus contributing to depopulation of the residual insurer) may prefer uninspected policies-properties over inspected policies-properties if they determine that pricing within the category of inspected and credit eligible homes is inadequate due to the abovementioned distortions in the current mitigation credit program.

Non-windstorm Losses: Sinkholes In 2010, Citizens earned approximately $32 million in sinkhole premium and is expecting to pay out approximately $245 million in losses and loss related expenses ($19.6 million and $84 million respectively in 2009). Citizens 2012 rate filing had a nearly 450% sinkhole rate change indication. In December 2010, The Florida Senate Interim Report, Committee on Banking and Insurance, provided a discussion on the sinkhole issue. This followed the November 2010 Florida Office of Insurance Regulation Report on Review of the 2010 Sinkhole Data Call. The Florida Senate Interim Report does an excellent job of addressing the issues surrounding sinkholes. The relevant points discussed in the Senate Interim Report include:

30

Citizens rate change indication is driven by sinkhole claims in Hernando, Pasco, and Hillsborough counties, where claims have increased by 375%, 187% and 384% respectively between 2006 and 2009;

Private insurers have also seen their sinkhole claims and costs rise by double and triple-digit percentages over the past several years. According to the OIR Data Call report, the total reported sinkhole claims increased from 2,360 in 2006 to 6,694 in 2010. Over that time period, there were 24,671 total sinkhole claims for approximately $1.4 billion. Of those claims, 66% were in Hernando, Pasco, and Hillsborough counties;

Based upon the opinions of licensed geologists in Florida, there is no geological explanation for such a significant increase in the number of claims; and

Representatives from OIR, as well as insurers, believe that a major driving force for the significant increase in sinkhole claims is the fact that many policyholders are incentivized to file such claims because they can keep the cash proceeds from the claim instead of effectuating repairs to their home or remediating the land.

The failure of sinkhole claimants to make repairs or stabilize land has concerned property appraisers in several counties, particularly in Hernando and Pasco counties. They believe that this dilemma has had a damaging effect on the market values of affected homes, which could lead to financial instability of local governments. The Hernando County property appraiser has estimated that since 2005, the county has lost $173 million in total market value as a result of value adjustments to sinkhole homes. During the 2011 Florida Legislative Session, legislation specifically designed to address sinkhole insurance issues passed. While it is likely safe to assume that the changes effected by SB408 will reduce sinkhole claims, the magnitude of the reduction is difficult to quantify this soon. This issue was at the forefront of Citizens 2012 rate hearing and subsequent rate setting by OIR. In the press release outlining the Citizens 2012 rates, the OIR states:

In Senate Bill 408, the Legislature eliminated the 10% statutory cap on sinkhole rates, but at the same time it enacted fundamental changes to reduce sinkhole losses. The rates established by this order contemplate cost-savings of Senate Bill 408, but do not fully quantify the cost-saving effects due to the lack of data at this time.

There were a number of components to SB408 that were designed to address these trends. These components include (but are not limited to):

Limits on public adjuster solicitations and compensation; Limiting sinkhole coverage to the principal residence; Stricter definitions on what constitutes sinkhole damage; A two-year limit on claims notification; and Requiring repairs to sinkhole damaged property.

31

SB408 may actually increase sinkhole claims in the short run as policyholders may file claims under existing insurance contracts before the more restrictive policies issued following SB408 are implemented. This may take a few years since current policies are occurrence policies. This means that sinkhole damage that occurs this year, regardless of when the claim is reported, is covered by this years policy. Therefore, policyholders may still file claims in future years that will be covered under policies issued prior to SB408 (often referred to as the tail problem). In addition, there is a time lag from the time that SB408 was passed until its components are incorporated into every policy issued. For example, it will be nearly a year before a policy that was issued just prior to SB408 is renewed with the new SB408 rules included. It will take a minimum of three to four years for credible data to emerge regarding the impact of SB408 on sinkhole claims and the growth of Citizens market share in this area.

Non-windstorm Losses: Lengthening of the Property Claims Tail18 Property insurance claims have traditionally been considered short-tailed. In other words, property claims have generally been opened and closed within relatively short time periods. Property claims associated with catastrophes have generally had longer tails than have ordinary property claims. It takes more time to settle catastrophe claims because of the overwhelming nature of the catastrophe and the strain it puts on the claims process. Even with catastrophes, the amount of time it takes to settle property claims has been relatively short. In Florida, the tail on property claims, especially catastrophe claims, has significantly increased in the last 5 years. Following the 2004-2005 storm seasons, most claims were closed by the middle of 2007. Citizens experienced 303,000 total claims stemming from those two storm seasons, only 3,500 (approximately 1.1%) of which were still open in June 2007. Similarly, State Farm had 320,000 total claims with only 1,820 (approximately .6%) still open in June 2007.19 In other words, even with two storm seasons that placed substantial strain on the claims paying process, more than 99% of claims were closed by the middle of 2007. These numbers changed substantially in 2008 and 2009. Between March 2008 and June 2009, Citizens opened or re-opened 14,997 catastrophe claims (7,299 new claims and 7,698 re-opened claims) related to the 2004-2005 storms.20 This represents approximately 5% of all Citizens claims related to those storms. This type of claims activity significantly lengthens the tail of the claims distribution and substantially impacts the loss costs associated with property insurance. The trend of new and reopened insurance claims related to the 2004-2005 storms was not isolated to Citizens policyholders. The FHCF was required to issue bonds in the amount of

18 SB 408 addressed issues that may result in a shortening of the property insurance claims tail. The results of the Legislation will be evident over time. 19 See Diamond (2007). 20 See OPPAGA (2010).

32

$675.92 million, in 2010, to cover the additional reinsurance claims private insurers faced due to new and reopened 2004-2005 storm claims.21 Of the reopened Citizens catastrophe claims, 57% of the claimants were represented by public adjusters; this means 43% of the reopened claims were filed by claimants not utilizing public adjusters. Furthermore, only 19% of the newly opened Citizens catastrophe claims were represented by public adjusters.22 The role of public adjusters is not limited to catastrophe claims. Of the 61,324 non-catastrophe claims, 26% of the new claimants were represented by public adjusters and 40% of the claimants with reopened claims utilized public adjusters. In all cases, the use of a public adjuster lengthened the claims process. For example, the median time to close non-catastrophe claims was 41 days without a public adjuster compared with 115 days with a public adjuster. The lengthening of the claims tail for property claims greatly affects the loss costs of insurers. The additional time adds uncertainty to the settlement dates, reserve amounts, and ultimate paid amounts driving up insurers loss costs and ultimately premiums, as discussed below.

Private Market Performance

Competition Competition plays an important role in pricing in any market, including insurance markets. Florida property insurance markets have three types of competitors: licensed primary insurers, surplus lines insurers and Citizens. For competition to affect premiums, an insured must have the ability to shop for products from competitors to determine who is offering the best product/service for the best price. To shop for property insurance in Florida, an insured must find insurers who are selling new policies. Table 3, taken from the OIRs QUASR reports, shows that Citizens is the primary writer of new insurance policies in Florida. The nearly 1 million new policies issued in 2012 (as of the end of November) represent approximately a significant portion of the 5.7 million residential properties in Florida. Citizens total number of new policies is nearly equal to the combined total of the other 9 companies in the Top 10. Given Citizens dominance in issuing new policies, Citizens may be setting the premiums charged in Floridas private residential property insurance market. In other words, it may be Citizens premiums, rather than competitive practices within the private industry, that are impacting statewide property insurance premiums in Florida.

21 See http://www.sbafla.com/fhcf/Home/FHCFEmergencyAssessment/tabid/333/Default.aspx, accessed 8/25/11. 22 See OPPAGA (2010).

33

Table 3: Top 25 - New Policies Written 2012 (As of 11/30/2012) Personal Residential Property

Insurer Name # of New Policies

Type of Insurer

1 CITIZENSPROPERTYINSURANCECORPORATION

310,998

StateRunResidualMarket2 UNIVERSALPROPERTY&CASUALTY

INSURANCECOMPANY93,713 FloridaDomestic

3 SECURITYFIRSTINSURANCECOMPANY 56,051 FloridaDomestic4 AMERICANBANKERSINSURANCE

COMPANYOFFLORIDA50,569 Other

5 AMERICANMODERNINSURANCECOMPANYOFFLORIDA,INC.

32,509 FloridaDomestic6 AMERICANINTEGRITYINSURANCE

COMPANYOFFLORIDA30,695 FloridaDomestic

7 UNITEDPROPERTY&CASUALTYINSURANCECOMPANY

30,364 FloridaDomestic8 CASTLEKEYINDEMNITYCOMPANY 29,565 Pup9 FEDERATEDNATIONALINSURANCE

COMPANY27,310 FloridaDomestic

10 ST.JOHNSINSURANCECOMPANY,INC. 26,951 FloridaDomestic11 SOUTHERNFIDELITYINSURANCE

COMPANY23,673 FloridaDomestic

12 ASIPREFERREDINSURANCECORP. 22,456 FloridaDomestic13 TOWERHILLPRIMEINSURANCE

COMPANY22,301 FloridaDomestic

14 FLORIDAPENINSULAINSURANCECOMPANY

22,035 FloridaDomestic15 ARKROYALINSURANCECOMPANY 21,392 FloridaDomestic16 TOWERHILLSIGNATUREINSURANCE

COMPANY17,229 FloridaDomestic

17 CYPRESSPROPERTY&CASUALTYINSURANCECOMPANY

15,826 FloridaDomestic18 FLORIDAFAMILYINSURANCECOMPANY 15,603 FloridaDomestic19 GULFSTREAMPROPERTYANDCASUALTY

INSURANCECOMPANY14,964 FloridaDomestic

20 AUTOCLUBINSURANCECOMPANYOFFLORIDA

13,616 Other21 PEOPLE'STRUSTINSURANCECOMPANY 12,837 FloridaDomestic22 CAPITOLPREFERREDINSURANCE

COMPANY,INC.11,074 FloridaDomestic

23 OLYMPUSINSURANCECOMPANY 9,758 FloridaDomestic24 TOWERHILLPREFERREDINSURANCE

COMPANY9,626 FloridaDomestic

25 SAFEHARBORINSURANCECOMPANY 8,564 FloridaDomestic Top25Total(thru3Q2012)

929,679Source: QUASR market share reports quarters 1 through 3 of 2012.

34

Private Insurance Availability The availability of insurance is typically measured in multiple ways that, together, outline the sources and supply of capital that can be used to support risk. The number of carriers that choose to do business in the state, the structure of these participants, their premium sales volume and their surplus in the market are all important elements of availability. This section explores the availability of private homeowners insurance in Florida according to these four aspects of the market and benchmarks Florida against the availability of homeowners insurance nationally as well as in other Atlantic and Gulf Coastal states. Specifically, states selected for comparison include Alabama, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Texas and Virginia. Keep in mind information regarding the activity and performance of surplus lines insurers is not available so the discussion focuses on admitted insurers in these states. For purposes of benchmarking, Floridas private homeowners insurance supply is segmented by company structure type in two distinct ways:

1) whether an insurance decision center behaves in the market as an independent company or as a group interest; and

2) whether an insurer is a national carrier, a pup company (established by a national carrier separate from its other non-Florida or non-Florida-property business), or a domestic Florida company.

First, a common distinction in insurance is between independent or stand-alone companies and those insurance companies that are members of a group of insurers. In the analysis presented in this report, all insurance companies that are members of a group (for example, State Farm, Allstate, Progressive, etc.) are considered to be a single entity or decision center. This is an attempt to recognize that these companies may work in conjunction with one another to achieve the groups goal, rather than competing as individual entities. Each independent company is considered its own decision center. This allows for an apples-to-apples comparison across states since the definition of independent and group companies will remain consistent across all states. The second distinction between national carriers, pup companies and domestics is unique to Florida. Most states have not had the same type of domestic insurer development as Florida nor have they seen the same type of pup formation. These are important distinctions to Florida but do not translate well for interstate comparisons. Appendix B contains a list of all insurance companies used in the analysis in Florida along with their segments. The importance of these categorizations is related to the ability of a company to 1) raise capital and 2) diversify its business, and thus also its risks. These categorizations are similar although not identical ways of segmenting company structure type. Most group companies are considered national in scope, so the parallel between these two categories is close. Most domestics are set up

35

as independent companies although, again, the parallel is imperfect. For purposes of comparison to other states, the statistics and analysis here focus on statewide figures and conclusions.

Number of Companies and Their Structure As of year-end 2011, there were 74 total private (i.e., other than Citizens) decision centers writing homeowners insurance in Florida, more than in any of the other coastal states used for comparison. As indicated by Table 4, Texas is the next closest state with 70 decision centers. Compared to the national total of 397, the amount of competitive interest in the Florida market appears relatively high. Segmenting the private marketplace into independent companies and group interests, Table 4 reveals that the comparatively large supply of homeowners insurers doing business in Florida is largely due to the contribution of independent companies to the overall Florida total. Relative to the states compared, supply in Florida is significantly more dependent on independent insurers. The 23 independent insurers writing homeowners insurance business in Florida represent not only a higher absolute number than in any of the other states but also a higher portion of the total number of decision centers in the state.23

Table 4: Number of Homeowners Insurance Decision Centers by State in 2011 Type | State Natl AL FL GA LA MS NC SC TX VA Group 197 50 51 59 36 42 47 57 55 56 Independent 200 3 23 8 10 4 14 9 15 8 Total 397 53 74 67 46 46 61 66 70 64 Source: NAIC Annual Statements, Authors calculations The high number of independent companies is an important statistic regarding the nature of available homeowners insurance in Florida relative to other states and the nation. Generally, group companies are larger and have a larger geog