The risk of muni default and bond insurance
Standard & Poor’s upgraded bond insurer MBIA Inc on Tuesday, saying it expects the company to gain market share and resume its prior role as a strong player in guaranteeing U.S. municipal debt.
The agency raised MBIA’s rating to A- from BBB and upgraded National Public Finance Guarantee Corp, the company’s main unit for insuring municipal bonds, to AA- from A…
…S&P also raised the rating on bond insurer Assured Guaranty on Tuesday to AA from AA-, citing its strong competitive position relative to its peers.
Assured was the only bond insurer left standing after the crisis and remained the top municipal issuer in 2013 by backing $7.38 billion of debt in 466 deals, Thomson Reuters data shows.
S&P rates Assured at AA and National Public Finance at AA-, which means that they can be paid to insure municipal bonds rated A or lower. I wondered what percentage of municipal ratings are A or below. And how often do municipal bonds default at rating levels below AA? How much real market is there for municipal bond insurance?
Here is the ratings distribution for Fitch’s U.S. public finance ratings (page 1). that shows about 40 percent of municipal ratings are A or lower (Fitch is the smallest of the major raters).
How risky are bonds rated single-A or lower? Do they default so often that they need to be insured? We answer that question with the historical default statistics, Here is Fitch’s (page 4) historical default data:
For municipal bonds rated single A by Fitch 0.05 percent of the bonds defaulted after ten years, an exceptionally small number. After 10 years BBB rated bonds defaulted at a 1.28 percent rate a somewhat riskier proposition. And junk rated bonds had significant and increasing risk of default as shown in the data.
Moody’s default data for 2012 shows that their ratings embody more default risk than Fitch at the Ba (BB) level and lower:
Oddly, Standard & Poor’s doesn’t appear to have published historical default data for 2012, instead publishing “transition matrixes” that convey a different set of data. The SEC requires that 2013 default data be published 90 days after year-end, so I’ll watch it to see what is put out.
Looking at historical data, it doesn’t appear that bonds rated “A” have enough risk to warrant bond insurance. Of course there is always the risk that a bond might drop in quality over time, so insuring it for that risk might have value. I’m lukewarm on bond insurance.
Note: After this post was published, S&P provided me with its 2012 public finance default report, which can be found here.