Why Illinois' Bond Ratings Don't Add Up

Illinois sold $1.3 billion in general obligation bonds on Wednesday, but the sale was not much of a bargain for state taxpayers. All bond yields rose last week after Ben Bernanke signaled a tapering of the Federal Reserve's quantitative easing (i.e., money-printing) program. But Illinois bonds are worse off after being downgraded by Fitch and Moody's earlier this month. The Prairie State will be paying 1.5 percent higher annual interest on its 10-year bonds than the rates incurred by the safest states and cities. These safe bond issuers carry the coveted AAA-bond rating-six steps above lowly Illinois in the ratings pantheon.

If all of the state's debt were refinanced at this 1.5 percent spread, taxpayers would be looking at an extra $500 million in annual borrowing costs-but is Illinois debt really so dangerous that it warrants a risk premium on such a large scale? I have investigated this question by creating a risk model for Illinois state bonds. The model is described in my new working paper published by the Mercatus Center.

My study doesn't produce a rating; instead it estimates an annual probability of default-failure to pay-on a scale of 0 percent to 100 percent. If the results were converted into ratings, they would place Illinois somewhere between AA and AAA-well above the state's present level. If this analysis were to convince investors, they would demand much lower rates to buy state bonds.

This raises the obvious question: With all the bad news about state pensions, how can one possibly conclude that Illinois bonds are not as risky as rating agencies and the markets seem to think? The reason for the different conclusion is the confusion produced by the rating scales themselves.

As I report in the study, Illinois last defaulted on bond payments 170 years ago. In fact, no state has defaulted on general obligation bonds since 1933, and most states weathered the Depression scot-free. This means that state general obligation bonds have a very strong historical track record.

How does that compare to other types of bonds-especially those rated AAA? In the years before the financial crisis, rating agencies assigned AAA ratings to tens of thousands of mortgage-backed securities and collateralized debt obligations. At the time, these asset classes had only been in existence for 20 years or less, so none of these types of "structured finance" obligations had established a payment record comparable to state general obligations.

During the crisis, many of these AAA-rated structured finance bonds defaulted, yet states continued to pay their general obligation bondholders on time and in full. Today, issuance of structured finance assets is rebounding and many still receive AAA ratings-yet Illinois bonds are thought to be so risky that they merit only a single A. Notably, this ratings inconsistency doesn't only involve structured finance. Some kinds of companies also seem to be rated more leniently. Perhaps the strangest example of this phenomenon involves companies that insure municipal bonds.

Before the crisis, Ambac and FGIC-two municipal bond insurers-were rated AAA. During the crisis, both went bankrupt. A third AAA-bond insurer, MBIA, was downgraded well into the "B" range-substantially below where Illinois sits today. In the pre-crisis world, these firms rented their AAA ratings to states and cities. Government bond issuers would pay the insurers a premium to guarantee payment on their bonds; allowing their debt to trade as if it were AAA rated.

It was a great business for the bond insurers, since so few municipal bonds defaulted. In fact, these companies felt so confident about the safety of government bonds that they levered themselves to the hilt. According to hedge-fund titan Bill Ackman, MBIA had 139 times more insurance than capital back in 2002. Ackman tried to convince rating agencies to downgrade MBIA, but his arguments fell on deaf ears.

The municipal bond insurance industry might have survived the financial crisis and continued to profit from the misalignment of municipal and corporate ratings if the companies had stuck to their core government bond business. Instead, they diversified into insuring structured finance securities-perhaps becoming bedazzled by the inflated ratings of these instruments. Ultimately it was defaults among insured mortgage-backed securities that marked the undoing of these formerly AAA rated firms.

Returning to my findings, I don't dispute that Illinois has uniquely bad fiscal policies and the worst outlook of any state. However, because of constitutional limitations, most states have relatively little debt. Illinois' state debt is only 6 percent of economic output; the analogous ratio for the federal government is 74 percent. Pension underfunding is a serious issue, but because state workers and teachers only account for about 2 percent of the Illinois' population, it is a much more limited problem than the underfunding of social security and Medicare at the federal level. Finally, unlike depressed cities such as Detroit, a large state such as Illinois has a very diverse and growing revenue base.

Just as the worst swimmer at the Olympics is probably better than anyone at the neighborhood pool, the worst state bonds are probably a lot safer than bonds issued by entities that can't guarantee revenue by taxing their customers. If rating agencies are going to rate the Olympian and the weekend warrior on the same scale, they need to keep this in mind.