Government Credit Crisis Over – So Where are the Upgrades

The sovereign and municipal debt crisis of the early 2010s is finished. Overblown predictions of a credit meltdown among European sovereigns, US states and cities, and other advanced economy governments have not been realized. Yes, there have been a few high profile defaults – Greece, Detroit, Harrisburg, Stockton and San Bernardino all come readily to mind because their insolvencies received so much coverage. But many other predicted defaults – Italy, Spain, California, Illinois, San Jose – failed to materialize and the overall default rate among government issuers has been only a few basis points annually. Meredith Whitney’s 2010 forecast of dozens of major municipal defaults is now fully beyond resuscitation – even by Michael Lewis.

Muni bond market bears set their 2014 hopes on Puerto Rico, but this month’s successful $3.5 billion bond sale makes the odds of a near term default or restructuring remote. Last year, both major pension systems received major overhauls with many current employees taking reduced benefits. Most of Puerto Rico’s debt is long term and annual deficits are relatively low, so the Commonwealth’s intermediate term financing needs are modest.

The end of the default “wave” and its limited magnitude leave credit rating agencies in an awkward position. Having repeatedly downgraded government credits, their current ratings are inconsistent with those that prevailed at the beginning of the apparent crisis. Also, their government credit ratings are now even more inconsistent with their ratings for corporate and structured – asset classes that have more underlying risk because issuers cannot levy taxes.

In 2013, Fitch recently announced that it downgraded twice as many US public finance credits as it upgraded in 2013. Moody’s 2013 transition report has yet to appear, but weekly accounts of its upgrades and downgrades at MunicipalBonds.com suggest a similar pattern. This preponderance of downgrades is occurring despite the overall improvement in state and local government finance. Renewed economic growth is yielding more income and sales tax revenue, rising home prices are swelling property tax receipts and a buoyant stock market is shrinking unfunded pension liabilities. But because Moody’s decided to use a lower rate of return assumption for pension fund assets, it has created the perception of increased credit risk, despite the absence of such. The blizzard of downgrades has largely offset the effects of a 2010 municipal ratings rescaling that had been undertaken in the wake of a lawsuit by Connecticut’s attorney general.

Meanwhile, the high profile states of California and Illinois remain at single-A despite the marked improvement in their prospects. Since Moody’s last downgraded California, the state has swung from deficit to surplus and seen a substantial decrease in its unemployment rate. Illinois, downgraded in mid-2013 due to a temporary failure to pass pension reform, has yet to see a compensatory upgrade now that the reform has been enacted. My own view was that neither state had material default risk in the medium term, given their low debt service requirements relative to projected revenue.

Markets appear to be rejecting the drumbeat of dire rating actions. In the same week that Puerto Rico successfully sold its non-investment grade issue, Chicago placed $884 million in securities on the heels of two Moody’s downgrades (a three notch reduction from Aa3 to A3 in July 2013 followed by a further one notch cut to Baa1 this month).

Perhaps markets have started to ignore ratings because they have become so rudderless. Ratings have inconsistent meanings because they are products of human discretion. If, instead, they were the outcome of stable, empirically based algorithms, ratings would more likely have the same meaning across time and between categories. Unlike human analysts, computer models don’t have to worry about criticism that they are being soft on politicians or inadequately mindful of unfunded pension liabilities – which are rarely associated with actual bond defaults anyway.

Finally, it is worth noting that inconsistent, incoherent ratings are not merely a sin of US rating agencies. Dagong, which commanded respect for issuing a sub-AAA rating to the US back in 2010, has not covered itself in glory since. After the end of the October 2013 government shutdown it inexplicably downgraded the US rating to A-.

The Chinese rating agency, apparently unaware that partial shutdowns are a familiar part of the US political scene, suggested that the October incident reflected an unprecedented level of risk. Contrary to political and media hyperbole, there was never a serious risk of a Treasury default arising from either a shutdown or a delay in raising the debt ceiling. While I agree that an issuer that engages in kabuki theatre over its credit obligations cannot warrant a top rating, it is absurd to place the world’s most powerful government a few notches above junk amidst declining deficits and accelerating economic growth. Further, we should all take pause from the fact that the Fed has proven capable of buying the lion’s share of new Treasury issuance with freshly printed money and without triggering price inflation.

Dagong’s goal appears to be to convince the world that the US is a worse credit than China. That’s a hard case to make given the latter’s relatively short history as a market participant, its lack of transparency and the risk that its single party political system cannot be sustained over the long term.

But regardless of the ratings themselves, Dagong’s process is disturbingly similar to that of the Western incumbents – discretionary ratings subject to political pressure and human biases. This is unfortunate for a rating agency that hopes to displace the ruling ratings triumvirate.  By declining to offer a superior analytical product, Dagong leaves investors little choice but to stay with the incumbents.

Wikirating also needs to review its rating procedures in light of the improvement in sovereign credit. In 2012, the organization introduced two sovereign rating methods: one based on an index of macro indicators (the Sovereign Wikirating Index or SWI) and a poll-based method through which individuals vote on ratings and an average is reported. These methods assign the US ratings of BB- and BB respectively – levels that are totally out of synch with financial market wisdom.

The SWI method requires backtesting and re-calibration to ensure that the index usually assigns low ratings to countries that subsequently default and relatively high ratings to countries that continue to service their debts. This task is on our agenda for 2014. We have already started to gather historical data for various index components so that we can see what SWI levels would have been prior to 2012 and correlate these with actual default history.

The poll method also needs to be reimagined. While it is attractive to aggregate large numbers of individual opinions, revisions are needed to control biased and uninformed voting. For example, a user who has no knowledge of sovereign credit and who objects to US foreign policy can express his or her attitude by assigning the US a D rating in the Wikirating poll.  This view is then averaged into the final rating. If enough users adopt this approach, the overall rating can be seriously biased. One possible solution – adopting an Intrade-style method of requiring users to make a small cash wager – has become difficult to implement due to new US regulations.

Wikirating and all other rating agencies need to adopt to a new regime in which government credit crises are the exception rather than the rule. While there will always be a few difficult cases like the Ukraine, improved economic performance and persistently low interest rates in the advanced economy have rendered the odds of a generalized government credit crisis as very remote indeed.