The sovereign ceiling: Now a broad consensus on its permeability

Recently, Fitch, Moody's Investors Service, and Standard & Poor's announced several changes or clarifications to their ratings policies in regards to the sovereign ceiling and bank ratings. In June 2001, both Moody's and Fitch, and on July 23, Standard & Poor's, clarified their policies towards the sovereign ceiling.[i] While some of these changes apply to ratings generally, our concern in this comment is particularly with the impact they have on bank ratings.

But before exploring the significance of these changes, or clarifications as the case may be, it is helpful to take a broader look at the meaning of ratings and the rationale for the earlier policies. Ratings, of course, are evaluations of relative risk, distilled into a single symbol. They resemble a school grade in this respect, but unlike school marks, their purpose is not the evaluation of past performance — although this they take into account — but the estimate of the probability of future default and the loss in such case. The sovereign ceiling refers to a traditional view by rating agencies that the risk of default reflected in the sovereign debt rating of a particular country should be an upper limit, that is, a ceiling, on the debt ratings of other entities headquartered in the same jurisdiction.

The concept derives from the notion that as the sovereign exercises regulatory control over such entities, their risk of default is subject to the exigencies of the government. For example, in a foreign currency crisis, a government may impose a moratorium on a foreign exchange remittances, as the Philippine government did in the mid-1980s, compelling otherwise creditworthy issuers to default. As Fitch, then known as Fitch-IBCA, explained the concept in a 1998 report:

A sovereign's rating on its foreign currency obligations has traditionally been regarded as a ceiling on ratings for other issuers domiciled in the country. The operative assumptions are that a sovereign default could force all other domestic issuers to default or that a sovereign could force other issuers to default as a means of avoiding its own default. Circumstances leading to a national foreign exchange crisis -- including economic and political upheavals, balance of payments crises, trade shocks, and high inflation -- directly affect the debt servicing capacity of private borrowers. Countries facing default may impose exchange controls and other restrictive measures that impede access by issuers to the foreign currency necessary to service their obligations.[ii] 

Other rating agencies have traditionally maintained similar guidelines, applied more or less rigidly depending upon circumstances. The 1998 Fitch report noted that sovereign default did not always result in the default in the private sector. To take one example, it noted that in some circumstances, for example, Venezuela in the mid-1990s, a defaulting sovereign would encourage private borrowers to continue debt service. The agency's policy, therefore, would be flexible, depending upon a variety of criteria.  

As in any rating exercise, the issuer's standalone financial strength and business prospects were of paramount concern. Corollary considerations included access to foreign currency earnings and diversification of revenue streams. Local debt ratings were more apt to exceed the sovereign ceiling than foreign currency ratings, since payments in local currency have little effect on a nation's foreign reserves, and hence it is much less likely a government would restrict such payments. Finally, the track record of the issuer and the sovereign, the existence of external support relationships, such as a strong multinational parent, and government policy were key concerns to Fitch as to whether the sovereign ceiling should be pierced.

 Moody'sand Standard & Poor's historically have applied the sovereign ceiling concept in practice fairly strictly. Exceptions have long been recognized, however, in the case of securitizations where techniques such as the creation of offshore special purpose vehicles, guarantees, insurance and over-collateralization and other 'credit enhancements' enabled its circumvention. These exceptions, however, proved the rule in that use of an offshore entity as a conduit for funds flows or external support mechanisms kept the credit enhanced issuer or debt out from under the operation of the sovereign ceiling rule.

 Moody'spolicy statement[iii] this year does represent a significant step away from exacting application of the sovereign ceiling. The agency said its decision was based on recent experience in several countries where "governments in default may choose to allow foreign currency payments on some favored classes of obligors or obligations". Moody's went on to note that while the sovereign ceiling would apply in most cases, it would not be rigidly applied. Criteria to be considered in piercing the sovereign ceiling were: First, standalone creditworthiness and external support; second, the likelihood of the applicable government refraining from the imposition of a general payment moratorium; and third, the borrower's access to foreign currency.

S&P's clarification, issued this week, implicitly acknowledges the applicability of the concept commonly referred to as the sovereign ceiling, but favours factoring sovereign risk into issuer and debt ratings according to the circumstances.

While [the] relationship [between the sovereign risk and the risk associated with locally domiciled issuers and debt] varies by type of entity, when a sovereign is either in distress or in default economic and financial market conditions are, in general, most likely quite hostile for all issuers in the given country. Therefore, there is typically a link between an issuer's ratings and those of the relevant sovereign.

Like Moody's, S&P observes that empirical data supports the proposition that a sovereign default usually, but not always, results in private entity default. The exceptions are significant enough — both agencies indicate that in about one third of the cases, a sovereign default does not trigger a private entity default — so that a rigid rule is not appropriate. Cases where a private entity rating may be higher than the sovereign rating, according to S&P, tend to fall into one of the following categories (excluding credit enhancements):

• where the government is not likely to impose moratoria on  foreign exchange remittances as a matter of policy;
• where the government has 'dollarized' its currency;
• where the country belongs to a monetary union having a higher-rated central bank; 
• where the issuer is highly diversified geographically;
• where there is reliable parent company support.

Local currency ratings likewise may exceed that of the sovereign, but a different set of criteria generally not directly related to foreign exchange concerns applies to such situations. In respect to banks, S&P is loath to rate banks above the sovereign, because the nature of their operations tends to link them quite closely to sovereign risk. According to the agency:

With few exceptions, bank foreign and local currency ratings are generally not higher than the relevant sovereign's foreign and local currency ratings. While governments may take steps to continue to allow banks to make international payments during external debt difficulties, the underlying creditworthiness of these financial institutions is likely to be hard hit by a sovereign or economic crisis.   

One misconception about the sovereign ceiling is that it applies across the board to all types of ratings. This is not the case. It is clearly applicable to debt ratings, both local and foreign currency, but there has historically been a degree of ambiguity about how it applies to bank counterparty ratings, which at least in the case of Fitch and Moody's utilize an entirely different ratings scale.

While sovereign and debt ratings utilize the well-known AAA (or in the case of Moody's Aaa) to C scale, counterparty ratings—Moody's Bank Financial Strength Ratings and Fitch's Individual Ratings—use a single A to E scale. Such bank counterparty ratings do not rate particular obligations, but the standalone financial strength of banking institutions. It can safely be said that the sovereign ceiling is not applicable directly to these ratings, although there is certainly correlation between these counterparty ratings and the debt ratings of a financial institution. Fitch's most recent statement on the applicability of the sovereign ceiling attempts to further clarify this correlation.

The Fitch statement,[iv] which specifically addresses the applicability of the sovereign ceiling to banks, outlines the considerations to be taken into account as to whether the sovereign ceiling on debt ratings may be pierced. Among these considerations are: that the bank in question has 'low exposure' to the government; the bank is owned by a foreign bank, the agency is confident of the parent's support for its subsidiary, and the parent has an investment-grade, long term foreign-currency rating; the subsidiary bank has in principle an 'individual rating' of at least 'C'; or  the subsidiary bank has other credit-positive characteristics, such as a support rating of '2' or better, and is either 'systemically vital', or 'highly liquid'. While the report does not specify precisely what combination of characteristics will enable a particular institution to be rated better than the sovereign, it does add some transparency as to when this will occur.

Finally and separately from the issue of the sovereign ceiling, in March 2001,  Moody's added further refinements to its bank financial strength rating scale. In the past, the Moody's, Fitch and the now defunct BankWatch rating scales were very similar, although there were differences in methodology and supplemental (for example, support) ratings. All followed a nine notch A to E scale as follows:

This movement to allow for greater precision in ratings is to be welcomed. It addresses the problem of 'ratings compression', which occurs when the number of possibilities for ratings assignments is too limited, relative to the actual differences in creditworthiness within a particular market. It is prone to occur in low-rated countries where all financial institutions are clustered near the bottom end of the counterparty ratings scale. Ratings compression can introduce an element of arbitrariness into deciding whether to assign one rating or another, in what is ultimately a judgment call. BankWatch attempted to deal with this problem by creating a separate rating scale for emerging markets. That approach, however, introduced problems of its own. Refining the rating scale to allow a more comparable number of gradations to those made in respect of debt ratings is the better solution, and can hardly be viewed as less than a positive development.

As for the end of a rigid application of the sovereign ceiling, it is practically a non-issue as to whether this is a positive development or not; all the major agencies, it would appear, agree that a permeable ceiling is the proper policy. If ratings are intended — as there is little doubt they should  be — to reflect relative risk, and if empirical experience shows that there is less correlation between sovereign risk and the risk of default among private issuers, then there is scarcely a case to be made that this arbitrary historical rule should be maintained merely for the sake of tradition. Tradition may be a virtue in some contexts, but the evaluation of the risk of default is hardly one of them.  

To be sure, in the case of banks, there is some difference of opinion concerning the strength of the ceiling with regard to debt ratings, with S&P, perhaps, taking a more conservative view than Fitch, at least based on their most recent statements. An empirical study would have to be undertaken, however, to ascertain just how different the views of the three agencies are in practice with respect to the application of the sovereign ceiling to financial institutions. Certainly, bank counterparty ratings have never been explicitly subject to the rule, notwithstanding that it was observable some correlation existed between counterparty ratings and debt ratings. Fitch's latest report removes some of the ambiguity from correlation, which is something that Moody's broadening of its bank financial strength scale to more closely approximate the scale used for its debt ratings, tends to do as well.

Jonathan Golin is the author of The Bank Credit Analysis Handbook: A Guide for Analysts, Bankers and Investors, published John Wiley & Sons in June 2001. From 1997 until its acquisition in November 2000, Jonathan was associated with Thomson Financial BankWatch, a leading specialist bank rating agency, where as vice-president he covered banks in some of the region's most volatile economies. He can be reached at [email protected]