Muni Bonds Not as Safe as Thought

Municipal bonds, widely seen as one of the safest investments, actually default more often than most people realize, according to research issued Wednesday by economists at the Federal Reserve Bank of New York.


The economists said the widely held belief that municipal bonds almost never default is based on only a narrow slice of the market — the safest part, consisting of bonds that are graded by the main ratings agencies when brought to market. When the researchers looked at a much broader sample, which included unrated bonds, they found there have been about 36 times as many municipal defaults over the past 40 years as the conventional wisdom suggests.

For example, Moody’s Investors Service has reported that from 1970 to 2011, there were only 71 municipal bond defaults. But the Fed report counted 2,521 defaults in that time.

Officials of the ratings agencies said that investors should not look at the Fed’s tally and conclude that large numbers of cities are on the brink of default. The data suggests that less than 1 percent of the bonds rated by Moody’s defaulted, while more than 4 percent of the Fed’s bigger, combined pool of bonds defaulted. And individual investors — who hold about half of the $3.7 trillion debt in the municipal bond market — are more likely to buy rated bonds, which are less likely to default.

Robert Kurtter, a managing director of Moody’s Investor Service, said he agreed with the basic conclusions of the report but questioned parts of the Fed economists’ methodology. He said the report was not surprising because it is widely known that the municipalities that come to the agencies for ratings are confident that they will get investment grades. Those municipalities that do not seek ratings typically involve “smaller and weaker credits with a higher risk profile,” he said.

Steve Murphy, a managing director with Standard & Poor’s, agreed. He said, “The lion’s share of the defaults occur in the unrated market, and they have for many years.”

The report found a similarly vast gap in the raw numbers of defaults in Standard & Poor’s data. The Fed’s combined database indicated 2,366 defaults from 1986 to 2011, compared with S.& P.’s 47 defaults during this same period.

The Fed team of Jason Appleson, Eric Parsons and Andrew Haughwout said they had done the research in light of a recent unusual string of municipal bankruptcies, which has raised the question of whether a wave of municipal defaults was looming.

“History — at least the history most of us know — would seem to say no,” they wrote. “But the municipal bond market is complex, and defaults happen much more frequently than most casual observers are aware.”

The researchers said the debt backed by a city’s general obligation pledge seldom defaulted, while debts backed by revenues generated by individual projects were more uncertain. Bonds to create housing constituted 17 percent of the defaults in the Fed’s sample, for instance, and bonds to finance nursing homes and health care projects accounted for 12 percent and 11 percent.

The biggest portion of defaults was associated with industrial development bonds, a type issued by a government authority on behalf of a company. That type of bond accounted for 28 percent of the defaults.

The report did not provide dollar values with the defaults.

Mr. Kurtter, of Moody’s, said he was not sure that industrial development bonds belonged in the Fed’s database. Normally a municipal credit analyst does not count such bonds as municipal bonds, he said. Although industrial development bonds are issued by a municipality, the entity standing behind the debt is a profit-making company, so analysts rate them as corporate credits, he said.

While the Fed economists said they could not compare municipal default rates because they did not have complete information on the number of bonds, the numbers they offered implied a default rate of about 4.5 percent on the large sample of rated and nonrated bonds.

Mr. Kurtter said the agency had rated about 11,000 municipal “obligors” since 1970, a number that he said was smaller than the total number of rated bonds. Seventy-one defaults in a group of that size suggests a default rate of less than 1 percent.

When comparing the defaults with those of corporate bonds, the Fed analysts noted that corporate defaults happen more often during periods of recession, while municipal defaults are less tied to economic downturns.

In addition to debunking the conventional wisdom about municipal bonds, the Fed team’s report appeared to undercut positions taken by members of Congress in the wake of the financial collapse of 2008. Then, lawmakers often complained in hearings that the ratings agencies had been treating municipalities unfairly by rating their credit too low, given how seldom they defaulted. That, in turn, was said to be forcing municipalities to waste money on unnecessary bond insurance, and to pay excessive interest on their debts.

Moody’s has since recalibrated its methods, and it now applies the same criteria to corporate debt and municipal debt, Mr. Kurtter said. Initially, the changes raised some municipalities’ ratings, but many of the ratings have fallen back down.

Mr. Murphy said Standard & Poor’s had not changed its rating system for municipal credit in light of the Congressional complaints.