"Been There, Done That."
Recently, in the aftermath of the bond insurance crisis, there have been cries from members of Congress and state and local finance officers that the recent low default rate on governmental issuers of municipal bonds relative to all global borrowers begs for a wholesale upgrading of muni bond ratings. More than likely, bond ratings on state and local government obligations would substantially be lifted to triple-A and double-A levels.
The truth is that the early bond raters of the 20th century felt much the same way. Prior to 1929, most cities and states were rated triple-A and double-A because the conventional wisdom at the time treated them much like the equivalent of the credit risk for sovereign governments. That thought process by rating agencies changed during the Great Depression, when state and local government credits, large and small, defaulted left and right.
According to the George Hempel study, "The Postwar Quality of Municipal Bonds," there were as many as 3,252 total municipal bonds in default at the peak of the Depression in 1935. The peak number of recorded defaults on municipal bonds by category and date during that period can be seen in the accompanying chart.
According to Hempel's findings:
* Nearly 90% of the 310 cities with populations of over 30,000 were rated triple-A in 1929. Nearly 98% of this same group were rated double-A or better.
* Of all defaulting issues, 48.1% were triple-A rated in 1929 and 78.0% were rated double-A or better.
* Nearly 80% of the total dollar value for all defaulted issues were rated triple-A in 1929 and 94.4% were rated double-A or better.
* Municipal bond ratings were not the only quality measure that was not indicative of the defaults during the "Great" Depression era. Proponents of market yield as a measure of municipal bond quality were likewise disappointed. Rather than being lead indicators of defaults, long-term market yields of municipal bonds tended to rise significantly only after default had occurred.
The Hempel study was first written as an unpublished dissertation in 1964 at the University of Michigan, Ann Arbor. It was later used as the seminal research piece to actuarially stress-test worst-case municipal default possibilities for the bond insurance business. The importance of this study in the context of the debate taking place in 2008 cannot be understated.
Ratings applying to municipals today have a foundation based on the total 20th century default history.
Modern day ratings may be more stringent than they would otherwise be if the major rating agencies had not had the experience of being essentially "surprised" by the large number of defaults during the Depression. Prior to 1929, municipal general obligation bonds were almost all triple-A and double-A rated bonds.
The Hempel study documented the bias towards the highest ratings prior to the Depression. Overall, the analysis determined that if the indebtedness of the very short and controversial defaults were eliminated from the study, approximately 15.4% of the average net municipal debt outstanding in the early 1930s went into default. (It is worth noting on the positive side that traditional government municipal debt that defaulted had a fairly high recovery rate during the Depression. By one count, less than 1% was never repaid.)
It appears that it was only after so many double-A and triple-A municipal bonds defaulted that the rating agencies recalibrated their criteria to set tougher rating standards. These points should cause those who advocate sweeping upgrades to municipal government ratings (presumably to triple-A and double-A levels) to pause and consider the fact that the market has "been there, done that" once before and many investors paid the price of too much optimism.
While the federal government's bailout mentality is certainly more prevalent today than in the political and economic context of the 1930s, there is no legal guarantee built into municipal credits that the federal or state governments will stand behind a default. It is also important to remind contemporary critics who contend that state and local debt is underrated to the disadvantage of state and local borrowers that today's rating system has its foundation based on a longer experience record than just the default record since the 1970s. There exists a history of real and absolute municipal defaults that caused the agencies to revise the rating process.
It should also be noted that a much higher percentage of the municipal bond market involves nonessential-purpose governmental credits and not-for-profit credits like health care facilities, housing, student loans, and industrial revenue bonds, which have a higher potential to default with lower recovery prospects.
Many proponents of using a global-equivalent rating scale based on recent history should read Hempel's warning below, which is prophetic to almost all debt crises that have taken place:
"Problems develop when the expectations of the borrower and lender prove to have been too optimistic. The danger of over-optimistic expectations is probably the greatest during periods of sustained economic prosperity. At that time the expectations for future income and the prospects for selling assets are at a maximum. Such optimistic expectations may lead many individual borrowing units into accepting future commitments dangerously large in relation to their resources and without sufficient allowances for possible future emergencies. For the total economy, there is a distinct danger of an overexpansion of economic capacity during such optimistic periods. This overcapacity is undesirable and may contribute toward and increase the severity of a subsequent recession."
Complacency of positive expectations could leave municipal credit prospects by investors in a vulnerable position. Since 1970, the U.S. economy has had its ups and downs, but on the whole economic growth has been strongly positive and a major reason why America dominated the global economy over the past 50 years. These were not decades that one would say have retested the tumultuous years of the 1930s.
While no one is suggesting that a repeat of Depression-level defaults is imminent or inevitable, there are reasons to be concerned about municipal credit prospects. Existing debt levels, pent up infrastructure demands, pension, and other post-employment benefit obligations have the potential to weigh down government balance sheets for years to come, especially in places and regions whose local economic bases are likely to grow much slower than in years past. Just as important, there is a fairly good chance that average growth in the gross domestic product going forward is not likely to be as high in the next 50 years as it was in the last.
Like the legacy obligations that were promised by older industries like auto and steel during their peak growth years, many older cities and regions are no longer growing fast enough to absorb the commitments that are coming due without imposing a substantial burden. The "fixed" cost obligations and commitments for state and local governments are increasing in aggregate as a percentage of their discretionary budgets, thereby diminishing the spending power for current obligations.
Combining the growing long-term liability for state and local debt, pensions, and OPEBs is much like the budget squeeze created by the federal government's obligations to pay interest, social security, and Medicare - except local governments do not have the benefit of a "printing press." That said, state and local governments have a much better chance of meeting these obligations as long as economic growth remains robust.
Is the current rating system unfairly penalizing taxpayers? There is little support for this concern, especially if one believes that more stringent rating criteria actually puts pressure on state and local government to show evidence of prudent financial practices. Lending standards that are too lenient are not as likely to help hold down debt or taxes.
Proponents of a wholesale upgrading of nearly all municipal GO bonds point to their very low default history based on default studies done by the rating agencies that generally cover a relatively prosperous period in U.S. economic history from 1970 to the current day. The point of my comments is that one should not ignore the reasons why the current rating differentials came into being.
Raising all to the highest levels leaves little room for error, especially in a forward view of economic times that might not be as favorable as in recent years. This is a particularly timely concern given what has happened to triple-A structured products relating to subprime loans. While reform may be needed to create a better basis for comparison across ratings of all fixed-income classes, caution should be the watchword in moving too fast on this analysis.
Richard A. Ciccarone is managing director and chief research officer at McDonnell Investment Management LLC.