Is it worse than we thought? The Washington Post reports:
Economists at the agency counted 2,521 muni bond defaults since 1970, whereas ratings agency Moody’s Investors Service, for instance, reported 71.
The new number it 35 times higher than the old one. That’s a large undercount. These numbers feed into default and risk models so if the new data is right those will have to be rewritten and financial companies will need to adjust their positions and hedges. Lots of work for August.
There was this bit of confused reasoning:
Muni bonds often act as an investment haven for ordinary Americans
They shouldn’t. Tax-advantaged bonds offer the greatest benefit to those with the highest tax rates (who also often have the most to invest) so bonds are priced according to that rate. This means that those with lower tax rates are overpaying for the bonds and taking on too much risk for their money. “Ordinary Americans” shouldn’t be buying these investments and should stick to non-tax-advantaged corporate bonds.
But how could such a large difference in the data exist?
Ratings agencies only track the behavior of the bonds they rate
That’s a huge selection problem right there and it’s amazing no one pointed this out before now (I’m guilty of the oversight as well). Remember the first rule: Know Your Data!
There’s this bit which is nice from a risk-management perspective but bad from a modelling perspective:
Researchers found no pattern of spikes in defaults during recessions, rather defaults appeared to be a “function of idiosyncratic factors associated with individual projects,” according to the study.
It’s nice to know that these aren’t a cyclical investment but “idiosyncratic” factors, or noise, are notoriously hard to model. There’s also another key point about diversification:
eneral-obligation bonds, issued by municipalities, rarely fail because they are backed by tax revenue. But the Fed found bonds that finance hospitals, stadiums and nursing homes default at much higher rates because they have a narrower income stream. A sports stadium, for instance, needs to sell tickets, otherwise it may not generate enough to meet its debt obligations.