Muni Bond Defaults and Rated vs. Unrated Bonds
With all of the high profile municipal bond defaults and scares in the last few years from Vallejo, CA to Harrisburg, PA, it is easy to get spooked about the normally staid and safe muni bond market. So with that in mind, I wanted to put some numbers out there regarding municipal bonds, and help to quantify the risks for these securities.
Bear in mind all of the figures I report here in this article come from one of four sources: Standard & Poors, Moody's, the NY Fed, and the MSRB (Municipal Securities Ruling Board).
Let's start by talking about rated municipal bonds -- that is, muni bonds that are rated by one of the three major rating agencies: S&P, Moody's, and Fitch. These bonds are usually for larger issuers, and they pay lower yields than unrated bonds. As most of you reading this probably know, the higher the rating, the less the bond yields, all other things (like maturity) the same.
Between 1970 and 2011, there were 71 defaults on rated municipal bonds, 46 of these occurred after 1986. During that period, roughly 11,000 municipal bonds were issued. This means that, on average, rated municipal bonds defaulted at a rate of about 0.64%, or roughly 1 bond out of every 150 issued. This is for all municipal bonds, investment grade and non-investment grade, and we are talking about the default rate over the entire life of the bond, not in any particular year.
Now of course, this also does not mean that if a person held a bond that defaulted, then they would lose all of their money in the bond. In fact, even when municipal bonds do default, recovery rates are very high. (I will go over this more in a future article.) Nor does mean that a muni bond holder couldn't lose money even if their bond didn't default -- interest rate risk is still a very real problem for those holding long maturity fixed rate bonds. (I will also go over variable rate muni bonds in a future article.) But what these statistics do suggest is that credit risk is very, very low in muni bonds.
So if large, rated muni bonds are safe, are unrated small municipal issuers also safe? Can an investor earn extra returns without taking on too much extra risk by holding unrated bonds from small municipalities?
Well, maybe… after all, getting a bond rated by the rating agencies is expensive… frequently about $500,000. So it makes sense that a small city or a school district might not want to pay to have their bonds rated -- it could simply be too expensive, even though the city or school will end up paying a little more in interest.
But it's also possible that unrated bond issuers are mainly risky issuers who don't want to pay the rating agencies to rate their bonds, because they know they will end up with low ratings and high bond yields anyway. These risky muni issuers may decide it is easier to simply skip the rating altogether and issue unrated bonds instead.
So which is it? Are unrated issuers small but high quality municipalities, or are they larger, riskier ones? The answer most researchers have come to is that unrated issuers are probably a combination of the two groups. This sets up what's known as a lemons problem in economics. When an investor is looking at an unrated bond, he can't tell if it's a good bond from a small but safe municipality, or if it is a bad bond from a risky municipality (that may also be small). As a result, the investor has to play it safe and demand a high yield from the municipality just in case the bond is risky.
So how risky are these unrated municipal bonds? Well, if you recall, there were 71 defaults on the 11,000 rated bonds between 1970 and 2011. In contrast, there were 2,366 defaults on 43,500 unrated bonds between 1970 and 2011.
This means that the rated bonds had a default rate of 0.64%, while the unrated bonds had a default rate of 5.4%. Thus of every 19 unrated bonds, about 1 will default at some point during its life. Combined, the total municipal bond market for both rated and unrated bonds has a default rate of about 4.5%. Again, all of these default rates are over the entire life of a bond.
The other significant thing here is that there are a lot more unrated bonds than rated bonds out there (11,000 vs. 43,500). As a result of this difference in default rates, unrated municipal bonds frequently pay rates that are 5 times or more what some rated bonds pay. It is not particularly unusual for unrated bonds to pay 8-12% tax free.
So the key, if you want to buy unrated bonds and earn higher yields on your muni bonds, is to have a statistical model that will help you determine which bonds are safe and which ones aren't safe. Further, you also want to be sure the unrated bonds you buy are backed by good collateral, or that you have a reasonable chance of getting a decent recovery rate in the event that they do default. Again, statistical analysis can help enormously with this. I will talk more about how to do this kind of modeling and what tools big institutions use to do it in the future.