Jason Appleson, Eric Parsons, and Andrew F. Haughwout
In our recent post on the state and local sector, we argued that structural problems in state and local budgets were exacerbated by the recession and would likely restrain the sector’s growth for years to come. The last couple of years have witnessed threatened or actual defaults in a diversity of places, ranging from Jefferson County, Alabama, to Harrisburg, Pennsylvania, to Stockton, California.
But do these events point to a wave of future defaults by municipal borrowers? History—at least the history that most of us know—would seem to say no. But the municipal bond market is complex and defaults happen much more frequently than most casual observers are aware. This post describes the market and its risks.
The $3.7 trillion U.S. municipal bond market is perhaps best known for its federal tax exemption on individuals and its low default rate relative to other fixed-income securities. These two features have resulted in household investors dominating the ranks of municipal bond holders. As shown below, individuals directly hold more than half, or $1.879 billion, of U.S. municipal debt; when $930 billion in mutual fund holdings is included, the household share rises to three-quarters. Although the low default history of municipal bonds has played a key role in luring investors to the market, frequently cited default rates published by the rating agencies do not tell the whole story about municipal bond defaults.
Two large bond rating agencies, Moody’s Investors Service (Moody’s) and Standard and Poor’s (S&P) provide annual default statistics for the municipal bonds that they rate. S&P reports that its rated municipal bonds defaulted only 47 times from 1986 to 2011. Similarly, Moody’s indicates that its rated municipal bonds defaulted only 71 times from 1970 to 2011. As shown in the table below, this record of defaults compares very favorably with the corporate bond market, especially given the larger number of issuers in the municipal bond market.
However, not all municipal bonds are rated and the market’s rated universe only tells part of the story. We have developed a more comprehensive municipal default database by merging the default listings of three rating agencies (S&P, Moody’s, and Fitch) with unrated default listings as tracked by Mergent and S&P Capital IQ. Rather than confirming Moody’s 71 listed defaults from 1970 to 2011, our database shows 2,521 defaults during this same period. Similarly, our database indicates 2,366 defaults from 1986 to 2011 versus S&P’s 47 defaults during this same period. In total, we find 2,527 defaults from the period beginning in the late 1950s through 2011. (We don’t have complete information on the number of issues, so we can’t compare default rates.)
Our findings raise the question, What causes such markedly different default frequencies between rated and unrated municipal bonds? Our answer: Not all municipal bonds are created equal. Different types of municipal bonds are secured by very different revenue sources with varying levels of predictability and stability. Furthermore, we believe that rated municipal bonds tend to be self-selected: issuers are less likely to seek ratings if their municipal bonds are not likely to achieve investment grade ratings.
To provide some background, we note that the municipal market is bifurcated into general obligation (GO) bonds and revenue bonds. GO bonds carry the broad full faith and credit pledge of a state or local government.; GO pledges are considered among the strongest type of security because municipalities have the authority to levy taxes; GO defaults like those in Jefferson County and Harrisburg are relatively rare. Revenue bonds, on the other hand, are backed only by a pledge of revenues raised from a specific enterprise, such as an airport, toll road, hospital, or school. GO bonds are therefore stronger than revenue bonds, because the revenue base for general obligations is much broader (for example, the ability to levy taxes on property or citizens within a specific geographic area) than the typical revenue bond
revenue base (for example, a specified stream of revenue or enterprise).
The default risk of a revenue bond varies with the essentiality of the service provided by the enterprise. For example, water and sewer utilities provide essential services and thus have a strong ability to generate revenue. By contrast, the services offered by an alternative energy plant, pollution control facility, or other corporate-like entity may not be considered essential, because of the availability of other energy sources. Thus, these enterprises may have less potential to generate revenue. Alternative energy plants, pollution control facilities, and other corporate-like enterprises are all examples of industrial development bonds (IDBs). IDB financing projects using new technologies or projects with no historical track record tend to make up a majority of unrated IDB defaults.
It is also not surprising that revenue bond sectors, such as IDBs, experienced more defaults than GO bonds because they have made up almost two-thirds of municipal new issuance since the mid-1990s. While total municipal issuance has bounced around over the past sixteen years, revenue bonds have remained between 60 percent and 70 percent of new issuance.
To clarify how the number and behavior of municipal bond defaults compare with those of corporate bond defaults, the graphs below plot each series against the Philadelphia Fed’s national coincident index, a proxy for GDP. In the case of corporate bonds, defaults occur at a higher frequency during recessionary periods. In contrast, the pattern of municipal bond defaults appears less tied to recessionary periods. The absence of a clear pattern leads us to believe that municipal bond defaults may be more a function of idiosyncratic factors associated with individual sectors or issues than the result of broad macroeconomic developments.
The second graph also illustrates one of the main points of our post: the untold story of municipal bonds is that default frequencies are far greater than reported by the major rating agencies. Again, this is partly because the rating agencies’ default numbers only cover bonds that they rate, and the unrated portion of the market can be home to municipal bonds of lower credit quality, exhibiting a higher frequency of defaults. Until recently, investors could take some comfort from the fact that many municipal bonds—both rated and unrated—carried insurance that paid investors in the event of a default. But now that bond insurers have lost their AAA ratings, they no longer play a significant role in the municipal bond market, increasing the risks associated with certain classes and certain issuers of municipal debt.
In a coming post, we will explore a municipal bond sector—industrial development bonds—where defaults have been relatively common, highlighting specific examples of defaults.
The views expressed in this post are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.
Andrew F. Haughwout is a vice president in the Office of the Directors Function of the New York Fed’s Research and Statistics Group.