The risks of municipal default and bond insurance: Part 2

March 27, 2014

When I put up a post about bond insurance and default rates last week, I expected pushback from proponents of municipal insurance. I got some.

Events in the last seven years show anecdotal evidence for and against municipal bond insurance. The bankruptcies of Jefferson County and Detroit and the workout of Harrisburg, Pennsylvania demonstrate the value of a bond insurer that makes full interest and principal payments for defaulted bonds. Investors undoubtedly benefit from this continuity of payments.

In contrast, we have the financial collapse of 2008, when most bond insurers were downgraded below the level of the issuers they had insured. The issuers started trading on their underlying ratings, and as far as I know, they were not refunded for the insurance premiums they had paid upfront. In a big financial crisis, bond insurance is a bust.

Mark Palmer of BTIG addressed another dimension of bond insurance; the lowered borrowing cost that AA-rated bond insurers (Assured, MBIA and BAM) can pass onto the new bonds that they insure. Palmer writes:

Bond insurance prevents payment interruptions for insured investors and allows them to avoid the protracted and difficult recovery process following defaults when they do occur, which are rare for higher-rated bonds. However, the primary purchasers of bond insurance are the issuers of municipal bonds – the municipalities themselves. And municipalities gain several benefits from insurers’ wraps, including lower borrowing costs due to improved execution, capital market access – a key benefit for issuers that otherwise would have limited or no ability to float bonds without credit enhancement – and the surveillance and intervention that insurers provide.

I conceptually agree with Palmer, but I don’t agree with his data. More from Palmer:

The difference between the average ‘AA’ municipal bond rated by Moody’s is 4.03 percent, while the average ‘A’ muni bond rated by Moody’s is 4.72 percent, according to Bloomberg. Saving those 69bps, less the cost of the insurer’s wrap, is certainly beneficial to municipalities. That spread represents funding that a strained city or county could devote to personnel costs or infrastructure.

I’m assuming that Palmer is quoting spreads on 30-year general obligation bonds from Bloomberg when he claims a savings of 69 basis points. Thomson Reuters Municipal Market Data, though, shows a 51 basis point spread between 30 year GO AA and A yields. See below:

Of greater interest to me is that the yield for 30-year insured bonds averages 4.37 percent, or a mere 10 basis points less than single A GOs, according to MMD. The insured rate is a blended yield for a wide variety of bonds, but it suggests that the secondary market, where bonds are traded after being issued, doesn’t give insured bonds the same lowering of yield they receive when they are first issued and insured in the primary market.

Of course the most important and unknown variable in this discussion is the premium that insurers charge a municipality to insure their bonds. If you have data, please send it over.

Returning to the default discussion, Standard & Poor’s sent me the 2012 default data. S&P also had low default rates over the long term:


This discussion needs more data and analysis. There are clearly instances where issuers would benefit from having their bonds insured, but there are likely other instances where it is unnecessary and has no economic benefit. Stay tuned.

One comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

>>Of course the most important and unknown variable in this discussion is the premium that insurers charge a municipality to insure their bonds.

Doesn’t the insurance premium effectively get paid by the bond holders? For example, an insured bond should have a lower coupon and/or higher price than a non-insured bond, thus that differential between it and a non-insured equivalent should be compared to what the out of pocket expense for paying the insurance premium would be. The offset might not be $1 for $1, but hopefully pretty close. In particular, I believe bonds that get wrapped in the secondary market work this way in that the bond holders agree to, say, give up 15bps of coupon or yield (via price adjustment when they surrender to old uninsured bonds for the new ones) in order to pay for the insurance.

But staying with your point, if future expected default rates are low then one would think the cost of insurance to be correspondingly low. Looking forward to any insurance premium data you can locate…

Posted by LongTailFinance | Report as abusive