Public Finance

Special Report

Municipal Default Risk Revisited

Analysts
David T. Litvack 1 212 908-0593 [Link]@[Link] Mike McDermott 1 212 908-0605 [Link]@[Link]

Outlook Fitch Ratings published its first study of municipal defaults in September 1999. The results of this study indicated that the 13- to 20year cumulative default rate on municipal bonds issued from 1979 1986, including industrial development bonds, was 1.49%. While the overall default rate was low relative to many fixed-income sectors, default rates varied significantly across municipal subsectors. Specifically, the incidence of defaults on tax-backed, water/sewer, and other plain vanilla municipal bonds was almost nil, with 13- to 20year cumulative default rates of less than 0.26%. Conversely, the risk was much higher for municipal bonds issued on behalf of corporations or by municipal entities vulnerable to bankruptcy and other elements of corporate risk; the 13- to 20-year cumulative default rate was 14.89% for industrial development bonds, 4.86% for multifamily housing, and 2.62% for health care. The significant changes in the economic and geopolitical landscape since 1999 a protracted recession, three-year bear equity market, war, increased threat of terrorism, and fear of Severe Acute Respiratory Syndrome (SARS) have caused revenue reductions, budget deficits, reduced general reserves, and, therefore, an erosion in municipal credit. From Jan. 1, 2002May 31, 2003, downgrades of Fitch-rated municipal securities outweighed upgrades by a ratio of 1.8:1. Therefore, Fitch believed it appropriate to update the study to determine if actual municipal default rates had changed. This report provides an update on the default rates presented in the 1999 report and also includes: a breakout of default rates within health care; discussions on various other revenue sectors; an analysis of default occurrence relative to economic cycles; recovery rates; and default rates by state. Fitch will continue to update its default analysis periodically, based on concerns in the market or as conditions warrant.
2003 Findings

Replication of the 1999 analyses with 3.5 years of additional data reveals no meaningful changes in overall municipal default risk, which is still very low, and no meaningful changes in default risk on the sectors studied previously, with the exception of environmental facilities, where default risk has increased, as had been anticipated given deregulation in that sector. However, additional analysis performed in the update study reveals some interesting new findings. For example, the update study shows a dramatic distinction in risk for health care by subsector, with much lower default risk for hospitals than long-term care and other facilities. Another new analysis shows a moderate correlation of default risk as a lagging indicator of economic cycles, explained not just by high defaults on industrial development
June 23, 2003
[Link]

Public Finance
Change in Cumulative Default Rates*
(%) Bond Issuance Years 19791986 19871994 19871997 19791994 19791997

By Dollar Volume of Defaults

1999 Default Study Update Change from 1999 Study 1.49 1.50 0.01 1.72 1.76 0.04 0.49 0.63 0.14 0.47 0.58 0.11 N.M. 0.56 N.A. N.M. 0.53 N.A. 0.87 0.96 0.09 0.95 1.04 0.09 N.M. 0.84 N.A. N.M. 0.91 N.A.

By Number of Defaults
1999 Default Study Update Change from 1999 Study

*Includes defaults occurring through October 2002. N.A. Not applicable. N.M. Not measured.

bonds (IDBs), but event risk affecting the much lower risk municipal credits; this adds a cautionary note to the still relatively low municipal default rates seen over the last two years. A third interesting finding comes from the analysis of recovery rates, which indicates that defaulted municipal bonds have relatively high recovery rates (about 67%). In the cases of defaults by most essential service providers, the issuers resume paying debt service. Even some of those municipal sectors that have relatively high default rates tend to have favorable rates of ultimate recovery. For example, multifamily housing and industrial development bonds the two sectors with the highest default rates are often secured by collateral that results in substantial recovery values.
Use of Third-Party Data

Default Risk Update

Cumulative Municipal Default Rates

In examining the performance of Fitch-rated municipal securities, only two payment defaults were found Prince Georges County (MD) Greater Southeast Healthcare System hospital revenue bonds (series 1993), which defaulted in 2002, and Tri-City Hospital Authority (GA) (South Fulton Medical Center, Inc.) revenue anticipation certificates (series 1993), which defaulted in 2001. These securities were originally rated BBB+ and BBB, respectively. Since two observations are obviously not sufficient for a meaningful study, Fitch used default data provided by Income Securities Advisors, Inc, the successor to Bond Investors Associates, which is the same source Fitch used for its 1999 study. Limitations inherent in the use of third-party data require using a different methodology for determining municipal default rates than the one Fitch typically uses in computing default rates on corporate bonds. A full description of the methodology Fitch employed in this study is detailed on pages 89.
Municipal Default Risk Revisited 2

The database Fitch used in its analysis included municipal defaults that have occurred since Jan. 1, 1980. Through October 2002, the cutoff date for this study, there were 2,339 cumulative municipal defaults, totaling $32.8 billion. As of March 31, 1999, the cutoff date used for the 1999 study, there were 1,885 defaults totaling $26.3 billion. Therefore, with 3.5 years of new data, both the number and dollar volume of cumulative defaults increased approximately 24%, which is somewhat more than the 18% increase in the period covered (22.75 years vs. 19.25 years). However, consistent with the 1999 study, the results of the update indicate that the overall default rate for municipal bonds remained low, relative to many other fixed-income sectors. The cumulative default rates for bonds issued during various periods are displayed in the table above and show that only small changes in the overall default rates have occurred since the 1999 study. The cumulative default rate for bonds issued from 1979 to 1986 inched up only slightly to 1.50% from 1.49% as they approached or reached maturity, while the cumulative default rate on bonds issued from 1987 to 1994 increased more significantly to 0.63% from 0.49%. As shown in the table on page 3, cumulative default rates are generally lower for bonds issued after 1986. Fitch attributes this to several factors: the Tax Reform Act of 1986, which placed restrictions on issuance of tax-exempt debt, particularly IDBs, the poorest performing municipal sector; better disclosure, shifts to accrual basis government accounting standards, and other improvements in issuer financial management practices; increased assistance and oversight by states of local

Public Finance
Cumulative Default Rates by Year of Issuance*
Dollar Amount of Defaults Number of Issues Dollar Amount ($000) Total Issued ($000) Default Rate (%) No. of Defaulted Issues Total Issues Default Rate (%) 1,087,257 63,158,000 1.72 58 6,931 0.84 1,190,742 73,618,600 1.62 75 7,689 0.98 880,578 80,576,800 1.09 66 6,652 0.99 1,454,915 120,568,700 1.21 121 9,095 1.33 2,055,721 119,196,700 1.72 225 8,672 2.59 1,576,258 132,949,800 1.19 182 8,183 2.22 3,486,892 228,799,800 1.52 302 11,082 2.73 3,168,119 172,684,600 1.83 156 9,058 1.72 987,398 125,544,300 0.79 76 8,319 0.91 1,100,372 140,981,400 0.78 115 9,895 1.16 1,340,063 154,601,000 0.87 107 11,606 0.92 1,089,489 162,631,900 0.67 76 11,608 0.65 1,225,554 217,244,000 0.56 54 14,657 0.37 858,341 277,561,100 0.31 57 16,419 0.35 1,326,606 339,602,500 0.39 65 18,131 0.36 2,336,098 205,326,900 1.14 58 14,949 0.39 807,982 198,328,800 0.41 54 14,841 0.36 1,051,537 226,709,400 0.46 81 16,155 0.50 827,829 266,930,700 0.31 76 16,576 0.46

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

*Includes defaults occurring through Oct. 31, 2002.

governments; greater scrutiny by rating agencies, bond insurers, and institutional investors; improved economic conditions during most of the 1990s; lower interest rates, which reduced the cost of debt; and timing, since more recently issued bonds have had less of a chance to default. For more detailed information on the improvements in the municipal marketplace, see Fitch Research on Credit Ratings in the 21st Century, dated March 6, 2000, available on Fitchs web site at [Link].
Sector Analysis

percentage of total defaults by sector against the percentage of bonds issued by sector. It is interesting to note that approximately 8% of total bonds issued, i.e. those for multifamily housing, all other health care, and industrial development, account for over 56% of defaults experienced, with the remaining 92% of bonds issued accounting for the other 44% of defaults. In particular, the education and general purpose sectors account for 46% of issuance but only 13% of defaults.
Health Care Risk by Subsector

The new study confirmed the varying default rates by sector found in the initial study (see table, page 4). Tax-backed and traditional revenue bonds remain low risk, with 16- to 23-year cumulative default rates of less than 0.25%, while IDBs (14.62%), multifamily housing (5.72%), and non-hospital related health care (17.03%) bonds exhibit a much higher level of default risk. A significant increase in default incidence was found in only one sector, environmental facilities, and this was anticipated given the deregulation in legal flow control that was introduced into the solid waste industry in the mid1990s. The cumulative default rate on environmental facility bonds issued from 1979 to 1994 increased to 0.89% from the 0.31% rate found in the 1999 study. The table on page 5 highlights both the par amount of defaults and the number of defaults by sector and expresses these as a percentage of the total default incidence. The table on page 6 compares the

When the health care sector is broken out into its component parts, hospitals are shown to have a much lower default risk than was previously perceived to be the case. This is because most of the defaults in the health care sector are on obligations of nursing homes, continuing care retirement communities (CCRCs) and congregate living centers, and other medical facilities such as clinics and treatment centers. Therefore, while all health care bonds issued from 1979 to 1986 have a cumulative default rate of 2.45%, these are heavily weighted toward nonhospitals, which have a cumulative default rate of 17.03% vs. hospitals, which have a cumulative default rate of only 0.63%. However, hospitals, as well as the other components of the health care sector, show a spike in defaults from 1997 to 1999 following the implementation of the Balanced Budget Act of 1997, which included major reductions in Medicare and Medicaid reimbursement rates.

Municipal Default Risk Revisited 3

Public Finance
Cumulative Default Rates by Sector*
(%) Issuance Years Change from 1999 Study 19791986 19871994 19871997 19791994 19791997 19791986 19871994 19791994 14.62 3.60 3.05 7.77 6.50 (0.27) 0.35 0.25 0.05 0.04 0.03 0.04 0.03 (0.00) 0.00 0.00 0.13 0.23 0.21 0.20 0.20 (0.00) 0.01 0.00 2.60 0.40 0.32 1.58 1.43 (0.17) 0.07 0.12 2.45 1.56 1.67 1.92 1.92 (0.17) 0.59 0.32 0.63 0.98 0.97 0.84 0.86 N.A. N.A. N.A. 17.03 6.63 7.01 10.98 10.11 N.A. N.A. N.A. 3.62 1.90 1.53 2.93 2.59 N.A. N.A. N.A. 0.26 0.01 0.06 0.15 0.16 0.08 (0.01) (0.95) 5.72 4.39 3.27 5.27 4.64 0.86 (0.20) 0.49 0.33 1.49 1.60 0.89 1.05 0.01 1.19 0.58 0.04 0.00 0.05 0.01 0.04 (0.00) 0.00 0.00 0.04 0.08 0.07 0.06 0.06 N.A. N.A. N.A. 0.03 0.04 0.04 0.04 0.04 (0.01) (0.02) (0.01) 0.25 0.55 0.42 0.45 0.38 (0.01) 0.02 0.01

Development Education Public Facility Electric Health Care Hospitals All Other Health Care Housing Single-Family Multifamily Environmental Facilities Transportation Municipal Gas Utilities Water and Sewer General Purpose

*Includes defaults occurring through October 2002. N.A. Not available because the sector was not measured in the 1999 study; based on dollar amount of defaults.

While it appears hospitals have a lower default risk than previously thought, it is still higher than that of other municipal sectors, with the exceptions of IDBs, multifamily housing, and environmental facilities. In most cases, hospitals are subject to many elements of corporate risk, such as labor shortages, increased supply costs, capital needs, and competition, which represent long-term credit concerns for the industry. For more detailed information on the health care sector, see Fitch Research on 2003 Nonprofit Hospitals and Healthcare Systems Outlook, dated Jan. 23, 2003, and Outlook for Continuing Care Retirement Communities, dated Feb. 12, 2003, available on Fitchs web site at [Link].
Transportation Risk

April 7, 2003, available on Fitchs web site at [Link]. The default study also shows a very low incidence of defaults on bonds issued for toll roads. Fitch believes, however, that the low reported rate masks several risks, including: the exclusion of a few major toll road defaults that occurred prior to 1979; defaults on privately owned toll roads that are not included in the dataset; and the current acceptance by the municipal market of project debt refinancing and restructuring. For more detailed information on historical defaults and expected defaults for toll roads, see Fitch Research on Ramp-Up at the Toll Plaza: A New Outlook for US Toll Facility Debt, dated May 25, 2000, available on Fitchs web site at [Link].
Public Power

Both the 1999 default study and this update report show a very low default rate for the transportation sector. In the airport subsector, no airport with significant commercial service has defaulted on its debt, except for airport-issued special facility bonds where lease payments from specific airlines are the primary security for the bonds. However, risk on both special facilities bonds and general airport revenue bonds has clearly increased since Sept. 11, 2001 due to the increased threat of terrorism, additional security costs, and fear of SARS, which have depressed passenger travel and caused severe credit deterioration in the airline industry. Many airport bonds have been downgraded recently, although most of the sector is still solidly investment grade. For more detailed information on the effects of these factors on the airport sector, see Fitch Research on War Increases Credit Risk at U.S. Airports, dated

A closer examination of the defaults in public power reveals that approximately 33% of the defaults on debt originated during the 19791997 period are attributable to municipal bonds issued on behalf of private electric companies. Since the default of the Washington Public Power Supply System (WPPSS) bonds 4 and 5, there have been only a limited number of cooperative utility defaults. While the percentage of public power defaults is likely to remain low and below that of investor-owned utilities, it is expected that municipal electric and cooperative utilities will be challenged by greater variability in energy and fuel prices in the near term. Certain public power systems are likely to be exposed to greater financial and operating risk due to

Municipal Default Risk Revisited 4

Public Finance
Composition of Defaults*
Sector Development Industrial Development General Industrial Development LTV Industrial Development Airlines Industrial Development Chemical Industrial Development Energy Industrial Development Executive Life Industrial Development Hotel/Motel Industrial Development Manufacturing Industrial Development Retail/Malls Industrial Development Transportation Lease-backed Other Development Education College/University Revenue School Districts Student Loan Revenue Public Facility Stadium/Auditorium Revenue Golf Course Revenue Correction Facility Revenue Electric Power Health Care Hospitals Medical Facilities Nursing Homes Retirement/Congregate Living Housing Single-Family Multifamily Environmental Facilities Industrial Development Pollution Control Solid Waste Transportation Highway Tax Revenue Parking Revenue Transit Revenue Port/Airport/Marina Revenue Toll Revenue Utilities Natural Gas Revenue Multiple-Utility System Revenue Sewer Revenue Utility Tax Revenue Water Revenue Water and Sewer Revenue General Purpose General Obligation (Unlimited Tax) General Obligation (Limited Tax) Assessment Districts Certificates of Participation Sales Tax Revenue Total Defaulted Par ($000) 7,877,811 839,813 684,400 595,245 108,915 1,080,090 1,850,000 706,765 1,044,271 222,208 106,235 636,369 3,500 190,110 117,070 16,595 56,445 249,270 45,730 78,680 124,860 3,986,560 6,475,109 2,448,533 227,500 2,313,479 1,485,597 6,680,173 366,286 6,313,887 2,693,675 1,966,100 367,115 463,745 6,600 178,445 4,440 252,210 22,050 192,665 21,220 4,800 18,825 8,995 81,870 56,955 3,953,376 325,015 834,434 2,451,223 341,809 895 32,762,494 % of Total 24.05 2.56 2.09 1.82 0.33 3.30 5.65 2.16 3.19 0.68 0.32 1.94 0.01 0.58 0.36 0.05 0.17 0.76 0.14 0.00 0.00 12.30 19.76 7.47 0.69 7.06 4.53 20.39 1.12 19.27 8.22 6.07 1.13 1.42 0.02 0.54 0.01 0.77 0.07 0.59 0.06 0.01 0.06 0.03 0.25 0.17 12.20 1.00 2.58 7.57 1.05 0.00 100.00 No. of Issues 700 101 33 14 11 46 8 100 196 57 9 123 2 18 12 3 2 27 6 10 11 46 576 100 29 317 130 588 35 553 57 47 10 23 2 5 2 9 3 26 6 2 4 1 5 8 281 4 5 239 32 1 2,339 Total 29.89 4.31 1.41 0.60 0.47 1.96 0.34 4.27 8.37 2.43 0.38 5.25 0.09 0.77 0.51 0.13 0.09 1.15 0.26 0.00 0.00 1.96 24.59 4.27 1.24 13.54 5.55 25.11 1.49 23.61 2.43 2.01 0.43 0.98 0.09 0.21 0.09 0.38 0.13 1.11 0.26 0.09 0.17 0.04 0.21 0.34 12.00 0.17 0.21 10.20 1.37 0.04 100.00

*Includes all municipal defaults between Jan. 1, 1980Oct. 31, 2002.

a growing reliance on natural gas-fired facilities, demand-driven volatility of natural gas prices over the near term, the gas industrys mixed success with drilling, and limited gas storage inventory. When incorporated with unstable climatic conditions, which have negatively affected hydroelectric operations,

particularly in the West, these risks can be exacerbated. As a result, Fitch has expanded its rating guidelines to include financial liquidity and energy risk management policies. Fitchs future discussions with public power managers will include more indepth inquiry into their hedging and risk management
Municipal Default Risk Revisited 5

Public Finance
Issuance and Default Occurrence by Sector
(%) Development Education Public Facility Electric Health Care Hospitals All Other Health Care Housing Single-Family Multifamily Environmental Facilities Transportation Utilities Water and Sewer General Purpose Total Issuance* 3.15 15.30 3.42 6.53 9.23 8.17 1.06 10.55 6.82 3.76 4.62 7.88 1.30 7.15 30.83 100.00 Defaults** 24.31 0.59 0.77 12.30 19.98 7.56 12.43 20.62 1.13 19.49 7.20 1.43 0.11 0.49 12.20 100.00

% Change in Nominal GDP vs. Default Occurrence

% Change in GDP 14 12 10 8 6 4 2 0
19 78 19 80 19 82 19 8 19 4 86 19 88 19 90 19 92 19 9 19 4 96 19 98 20 00 20 02
GDP Gross domestic product.

Defaults
($ Mil.)

(%)

6,000 5,000 4,000 3,000 2,000 1,000 0

*Includes bonds issued in the 19792001 period. **Includes defaults included in the dataset in the 19802002 period.

practices and how this fits into their long-term business strategy. More details regarding natural gas price volatility, utility hedging practices, and financial liquidity are included in the recent Fitch reports Fuel Factor Assessing the Risks of Public Powers Increased Exposure to Natural Gas and Liquidity Matters, both dated June 5, 2003, and available on Fitchs web site at [Link].
Environmental Facilities

experience lower default rates than obligations issued prior to deregulation. Defaults and GDP Fitch sought to determine if municipal defaults increased during weak economic periods. In fact, there does seem to be a moderate correlation between the percentage change in gross domestic product (GDP) and the dollar volume of defaults that occur in a given year; a one-year lag produces a somewhat higher correlation. In particular, the low growth years of the early 1980s and early 1990s correspond to the largest peaks in default occurrence. The chart above plots the annual percentage change in GDP versus defaults. The table on page 7 shows the correlation between defaults and the percentage change in GDP by sector. A correlation of negative one would represent a perfect inverse relationship; zero would represent only a random relationship. The data show that IDBs and single- and multifamily housing have the strongest correlation to a slowing in GDP growth. With a one-year lag, the correlation of IDBs and housing drops slightly, but the correlation factors for some other sectors increase, triggering a rise in the overall correlation. Except for the IDB and housing sectors, municipal defaults appear to be lagging economic indicators. Where defaults occur in more plain vanilla municipal sectors, they have often been due to events that are indirectly related to economic conditions, e.g. a reduced demand for power that led to the legal challenge and eventual default on WPPSS bonds 4 and 5 in 1983; the failure of Executive Life Insurance

The environmental facilities sector is made up of two components solid waste obligations and industrial pollution control obligations. Defaults increased in both subsectors as expected, as a result of deregulation. Some of the solid waste defaults were attributable to systems collecting industrial or hazardous waste systems not typically run by or associated with municipal entities. Therefore, the default risk for traditional municipal solid waste systems is likely lower than the overall sector rate would indicate. To mitigate the increased risks posed by deregulation in 1994, which eliminated legal flow control and increased competition, many issuers restructured outstanding debt, diversified revenue sources, implemented long-term contracts with waste providers, and developed public-private partnerships. As a result, obligations issued for solid waste facilities following these developments may
Municipal Default Risk Revisited 6

Public Finance
Correlation of Defaults to % Change in GDP by Sector
Correlation Coefficient Sector No Lag One-Year Lag Development (0.40) (0.18) Education (0.13) (0.36) Public Facility (0.36) (0.28) Electric 0.08 (0.30) Hospitals (0.25) (0.12) All Other Health Care (0.15) (0.20) Single-Family (0.43) (0.24) Multifamily (0.39) (0.37) Environmental Facilities (0.24) (0.29) Transportation (0.27) (0.37) Utilities (Includes Water and Sewer) 0.31 (0.08) General Purpose (0.23) (0.20) Overall Defaults (0.49) (0.56)
GDP Gross domestic product.

bond payment was paid by the insurer and the default was subsequently cured by Wilkes-Barre within four days. In this case, nonpayment occurred due to the bankruptcy of the intended payer and the inability of the cash-strapped city to honor its backing in a timely manner. This default was not reflected in the data because it occurred in December 2002, after the Oct. 31, 2002 cutoff. Because there appears to be a lag between an economic downturn and the occurrence of municipal defaults, it remains to be seen if additional municipal defaults will yet occur. Ultimate Recovery In the event of default, the bond indenture may provide for recoveries in one of two ways: the borrower may cure the default and resume paying full debt service or collateral backing the bonds may be liquidated. Debt service payments may also be made by a third-party guarantor, in which case the guarantor generally assumes the rights of the bondholder. Determining ultimate recovery requires a fundamental analysis of the debt structure, the holders rights and remedies, the ability of the debt holder to exercise those rights, and the quality of the collateral.
Resumption of Debt Service

Co. in 1991, which caused defaults on numerous municipal bonds in which proceeds were invested in Executive Life guaranteed investment contracts; and the sharp change in interest rates that eroded the speculative and extremely interest rate sensitive investments by Orange County in 1994. Credit quality has certainly deteriorated in the last several years as the economic recession has deepened, with numerous states and municipalities being downgraded since then. During the period from Jan. 1, 2002 to May 31, 2003, Fitch downgraded 210 municipal issuers while it upgraded only 116, for a downgrade-to-upgrade ratio of 1.8:1. While credit quality has declined, default rates in the low-risk municipal sectors have not yet increased significantly, although there were two noteworthy exceptions. One was a default in September 2000 on a $31.5 million revenue bond, backed by a moral obligation of Spokane, WA, for a downtown parking facility. Fitch did not rate the bond, but it was rated BBB by Standard and Poors. Because of low utilization, revenues collected by the facility proved insufficient to meet debt service and fully support operations. Spokane determined it could not honor its pledge to loan money to the authority, reasoning that since there was little chance it would be repaid, such a payment would be illegal under the citys investment guidelines. This case highlights the risk inherent in a moral obligation pledge that backs a project not serving a broad public purpose. The second exception was a temporary default by the City of Wilkes-Barre, PA on an insured general obligation-backed industrial development bond. The

Resumption of debt service payments is likely the means provided in the bond indenture to cure defaults on tax-backed debt, as well as revenue bonds for projects such as water/sewer facilities or toll roads issued through a trust estate. For these types of securities, the physical asset is public property and never pledged to bondholders. However, bondholders maintain a lien on revenues, and recoveries are often 100%, although in the case of toll-backed transportation facilities, repayment may be delayed by years, and sometimes only principal and nominal interest (not interest on interest) is repaid (for additional information on recoveries of toll-road defaults, see Fitch Research on Ramp-Up at the Toll Plaza: A New Outlook for US Toll Facility Debt, dated May 25, 2000).
Liquidation of Collateral

For other revenue bonds and many certificates of participation, bondholders have a right to the physical collateral in the event of default. If the collateral is plant and equipment, the lien must be perfected through a Uniform Commercial Code (UCC) filing. UCC liens are governed by state law and generally must be renewed annually. Mortgage liens do not usually need to be renewed. If the lien is not properly
Municipal Default Risk Revisited 7

Public Finance
Recovery Rates
(%) Average No. of Dollar Occurrences Weighted 65.30 74.76 72.25 91.01 59.33 54.25 52.86 36.37 54.59 67.46 82.50 86.90 47.88 62.20 44.88 54.69 68.33 48.42 72.12 89.00 85.88 57.33 62.44 15.85 90.27 66.92

Recovery Rates

Development Education Public Facility Electric Health Care Hospitals All Other Health Care Housing Single-Family Multifamily Environmental Facilities Transportation Utilities (Includes Water and Sewer) General Purpose* Total Averages

As shown in the table above, overall, municipal bonds have an average recovery rate of 68.33% based on the number of defaults and a 66.92% recovery rate based on the dollar weighted average, both of which are higher than public corporate bonds that have a long-term average recovery rate of approximately 40%. Of note, high default rates do not necessarily translate into low recovery rates. For example, multifamily housing and industrial development sectors that have higher-than-average default rates also have higher-than-average recovery rates, due to the collateral backing many of the transactions. Default Rates by State One additional analysis that Fitch performed in this update study was cumulative default rates by state. It is interesting to note that during the period studied, there have been no defaults on bonds issued in Vermont, while West Virginia has the highest cumulative default rate at 3.04%. The table on page 9 shows cumulative default rates by state for bonds issued during the 19791997 period. Some states conduct closer oversight of local credits than others. For one state, North Carolina, Fitch grants credit enhancement of one to two notches on debt ratings below AA for local government issuers under the control of the North Carolina Local Government Commission (LGC), because of the very strong oversight and control exercised by the LGC (see Fitch Research on North Carolina Local Government Commission, dated March 29, 1999). This rating enhancement does not apply to authorities and other issuers not controlled by the LGC. Other states, most notably New Jersey, have also exercised emergency powers to force municipalities to take actions such as raise taxes or reduce expenses when necessary to prevent defaults; however, no rating enhancement is granted except at very low rating levels, since these controls do not become effective until the issuer is close to default. Also, many states have formal state-aid intercept mechanisms to enhance the credit of local borrowers, most often school districts. In these cases, Fitch generally assigns program ratings notched off the states general obligation ratings. In the absence of statutory regulations or financial policies that directly impact credit quality (as in the case of the state intercepts or the North Carolina Local Government Commission), Fitch would not consider differing

*Includes defaults on certificates of participation and assessment districts, which represent the bulk of the general purpose defaults. Recovery rates on general obligation bonds were 100%.

perfected, then the bondholder may lose priority on the claim and possibly end up as a general unsecured creditor (see Fitch Research on Public Finance Ratings Unaffected by UCC Article 9 Revisions, dated Dec. 12, 2001, available on Fitchs web site at [Link]). In analyzing the recoveries that can be garnered from liquidating the collateral, Fitch considers the value of the collateral relative to debt outstanding, the bondholders legal security in the collateral, and the ability of the bondholder to gain access to the collateral in bankruptcy. In valuing the collateral, Fitch considers the volatility of the resale market, the obsolescence of the collateral versus the amortization of the debt, the liquidity of the resale market, and practical and legal restrictions on alternative uses.
Third-Party Guaranties

In rating securities backed by a third party, Fitch considers the credit rating of the guarantor and the nature of the guaranty. In the case of a bond insurer or bank, Fitch substitutes that entitys rating. For public entity guarantors, the analysis may be more complex if the guaranty obligation is not backed by the entitys full faith and credit, in which case the rating may be haircut by one or more notches, or ignored entirely, based on the possibility that the guarantor will not honor its obligation. Fitch may also discount the rating value of a third-party guaranty if there are potentially significant timing delays.

Municipal Default Risk Revisited 8

Public Finance
Cumulative Default Rates by State*
(%) Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming District of Columbia Puerto Rico 1.36 0.21 2.79 0.67 1.01 2.02 0.27 0.12 1.51 0.79 0.05 0.17 0.80 0.68 0.52 0.75 1.31 1.55 0.06 1.15 0.44 0.52 1.21 0.36 0.74 0.42 1.29 0.16 1.83 0.59 0.66 0.30 0.15 0.70 0.69 0.50 0.07 1.03 0.02 0.48 0.23 2.22 0.71 0.43 0.00 0.52 2.15 3.04 0.24 0.26 0.00 0.21

Advisors, Inc., the successor to Bond Investors Associates, which was the source used for Fitchs 1999 study. Each default record includes fields such as default date, issuance date, credit enhancement (if applicable), recovery percentage, and other relevant information. The database covers defaults that occurred from Jan. 1, 1980 to Oct. 31, 2002. It generally excludes purely technical defaults but includes defaults on issues with credit enhancement, even if the enhancement provider made the payment. Upon reviewing the default data, Fitch made some changes to classification codes and excluded certain records when information on the default was incomplete. Default rates are calculated by analyzing the performance of bonds in a cohort (bonds issued in a given year) over time. The default rates are based on the dollar amount of defaulted par or number of defaulted issues compared to issuance volume during that period. When looking at default rates based on dollar volume, a large dollar default can skew the results, as can a default by a single entity with a large number of outstanding bond issues. This report presents default rates based on both methodologies where appropriate. Recovery rates are calculated based on principal recovered after settlement. This differs from recovery rates in corporate bond studies, which are based on market value quotes 30 days after the default. While a large portion of the default records did not have associated recovery rates and had to be excluded, over 1,095 observations were used to analyze aggregate recovery rates. Fitch calculated total and sector recovery rates using two approaches: the arithmetic mean; and a weighted average based on the size of the default. The overall rates under either approach can be skewed by the size of a particular default in the case of the weighted average and by the number of issues associated with a particular entity in the case of the arithmetic mean. The source of bond issuance information is various editions of the Bond Buyer Yearbook/Securities Data Company Yearbook. All editions of the yearbooks contain the number of issues and par amount of issues for bonds and notes. Since earlier editions of the yearbooks did not include sector or state information on note issuance, Fitch applied a pro rata approach to estimate note issuance by sector for the years 19791982 and by state for the years 1979 1986.
Municipal Default Risk Revisited 9

*Includes defaults occurring through October 2002; bonds issued 19791997.

default rates by state to be meaningful for rating local borrowers. Methodology The source of the default information used in this study is a database compiled by Income Securities

Public Finance

Copyright 2003 by Fitch, Inc., Fitch Ratings Ltd. and its subsidiaries. One State Street Plaza, NY, NY 10004. Telephone: 1-800-753-4824, (212) 908-0500. Fax: (212) 480-4435. Reproduction or retransmission in whole or in part is prohibited except by permission. All rights reserved. All of the information contained herein is based on information obtained from issuers, other obligors, underwriters, and other sources which Fitch believes to be reliable. Fitch does not audit or verify the truth or accuracy of any such information. As a result, the information in this report is provided as is without any representation or warranty of any kind. A Fitch rating is an opinion as to the creditworthiness of a security. The rating does not address the risk of loss due to risks other than credit risk, unless such risk is specifically mentioned. Fitch is not engaged in the offer or sale of any security. A report providing a Fitch rating is neither a prospectus nor a substitute for the information assembled, verified and presented to investors by the issuer and its agents in connection with the sale of the securities. Ratings may be changed, suspended, or withdrawn at anytime for any reason in the sole discretion of Fitch. Fitch does not provide investment advice of any sort. Ratings are not a recommendation to buy, sell, or hold any security. Ratings do not comment on the adequacy of market price, the suitability of any security for a particular investor, or the taxexempt nature or taxability of payments made in respect to any security. Fitch receives fees from issuers, insurers, guarantors, other obligors, and underwriters for rating securities. Such fees generally vary from US$1,000 to US$750,000 (or the applicable currency equivalent) per issue. In certain cases, Fitch will rate all or a number of issues issued by a particular issuer, or insured or guaranteed by a particular insurer or guarantor, for a single annual fee. Such fees are expected to vary from US$10,000 to US$1,500,000 (or the applicable currency equivalent). The assignment, publication, or dissemination of a rating by Fitch shall not constitute a consent by Fitch to use its name as an expert in connection with any registration statement filed under the United States securities laws, the Financial Services and Markets Act of 2000 of Great Britain, or the securities laws of any particular jurisdiction. Due to the relative efficiency of electronic publishing and distribution, Fitch research may be available to electronic subscribers up to three days earlier than to print subscribers.

Municipal Default Risk Revisited 10