"continued from the previous page"

Box 3 The relationship between interest-rate differential and rating

Differences in the rating of and the credit risk on bonds are naturally reflected in the interest-rate differentials observed in the market, cf. Chart 1. The widening of the differential between high- and low-rated issuers when the remaining maturity increases corresponds to the expectations of default over time shown in Tables 3 and 4. The marginal change in the interest-rate differential as a consequence of a rating adjustment by one grade is also greater for issuers with a "speculative grade" rating (BB+ or below) than for issuers of "investment grade". The chart also shows that the widening of the interest-rate differential between high- and low-rated issuers subsides after approximately 10 years. An interesting aspect of the default frequency is thus that while there is a growing risk of default over time for issuers with high ratings, the opposite is the case for issuers with low ratings, where the annual default frequency is declining. The explanation is that issuers with the lowest ratings either default (relatively early) or gradually move up the rating scale, whereas high-rated issuers tend to gradually slide down the rating scale, whereby the risk of default is increased. Finally, on determining the risk premium on an issuer with a low rating investors must take into account that an increase (drop) in an issuer's rating entails a capital gain (loss) on the issuer's bond debt. This will tend to make B-rated issues more interesting, since there is a chance of a higher rating despite the high risk of default. On the other hand, AAA issuers can only see their rating reduced. For 8-year periods starting in 1983-89 Moody's has calculated that only 53 per cent of the AAA issuers retained their rating throughout the period. With regard to B issuers, in the same period 27 per cent would have their rating upgraded, while only 12 per cent would see it downgraded.

Table 5 Development in Standard & Poor's ratings of Asian countries, 1997/98

Country	Rating on June 30, 1997	Rating on June 30, 1998	Change in number of grades
South Korea	AA-	BB+	- 7
Malaysia	A+	A-	- 2
Philippines	BB+	BB+	0
Singapore	AAA	AAA	0
Hong Kong	A+	A+	0
China	BBB+	BBB+	0
Thailand	A	BBB-	- 4
Indonesia	BBB	CCC+	- 8
Source: Bloomberg.

20 grades, a rating is not an exact measure of the credit risk. For comparison, investors can adjust the required bond yield on a continuous scale.

Irrespective of the aforementioned good general correlation between the rating and the yield on US issuers it is usually the case for the individual issuer that a rating is changed after the market has adjusted the yield in the same direction. This has often given rise to criticism. However, it should be stated that not only rating agencies, but also banks and supervisory authorities have often found it difficult to predict deterioration (or improvement) in the payment ability of an issuer/borrower in due time.

The rating agencies and Asia

The financial and economic crisis affecting a number of East-Asian countries over the last year has given rise to a lot of criticism of the rating agencies.⁸⁾ The following is a review of the most significant elements of this criticism.

As previously stated, ratings must be perceived as long-term assessments of the issuers' ability to pay. The rating of a country or a business enterprise should therefore not be changed frequently. Moreover, the movement over a short period should only in exceptional cases exceed one grade up or down on the rating scale. In view of the repeated downgrading of certain Asian countries in 1997-98 the agencies cannot be said to have fulfilled their own objective in this area.

In this connection South Korea is a particularly striking example. The country's sovereign ceiling had for many years been stable at A1/AA- at respectively Moody's and Standard & Poor's, but during the 4th quarter of

1997 the sovereign ceiling was reduced by 6-7 grades to Ba1/BB+. The picture is the same for other countries, cf. Table 5. The development in the aforementioned country ratings was accompanied by equivalent lowering of other issuers' ratings of these countries.

It has been stated that the agencies reacted too late to the development on the financial markets in Asia. In several cases the changes of the ratings were not announced until after the onset of the financial crisis and when the markets had already reacted by widening the credit differential. With regard to Korea this is illustrated in Chart 2. It should be noted that Moody's downward adjustment from A1 to A3 on November 27, 1997 occurred approximately 3 weeks after the price differential betweeen a South Korean dollar-denominated bond and an equivalent US government bond had widened by over 10 price points. It is likewise remarkable that during the first 9 months of 1997 the price differential between the two bonds widened gradually. This was in accordance with the growing concern at the development in Asia, including South Korea.

The problems faced by the rating agencies in Asia are particularly attributed to a lack of insight into the fundamental political, cultural and economic conditions in these countries. Among other reasons this may be due to a lack of reliable information on which the rating agencies can base their assessments.⁹⁾ For example, many market participants, including the agencies, were surprised that a significant part of South Korea's foreignexchange reserve was placed with domestic banks. In practice this impeded use of this part of the reserve. Equivalently, the problems in Thailand were not exactly reduced when it became clear that the published foreignexchange reserve had been eroded by the central-bank's forward purchases of Thai baht.

Furthermore, the rating agencies' assessment of the banks in the Asian countries has been influenced by the fact that only relatively few banks in the area were assessed. This implies inadequate market familiarity. This familiarity is a requirement for evaluation of individual banks. As a consequence the agencies did not in time realize and make public the danger that the countries' banking sector as a whole would face difficulties.

In their defence the rating agencies have stated that already before the downgrading of Asian ratings got under way they had in several cases issued assessments and analyses drawing attention to some of the conditions contributing to the development of the crisis¹⁰⁾, including the countries' large volume of short-term debt. The agencies have also stated that the plummeting prices for e.g. Korean bonds at the end of 1997 were excessive in relation to the credit risk on the bonds, cf. Chart 2.¹¹⁾ The substantial movements in bond prices in relation to the agencies' perception of the credit risk can, however, also be taken to indicate that at the close of 1997 investors no longer found ratings in the East-Asian countries to be reliable. For this reason, since the end of 1997 investors have not found ratings to be of any great significance on evaluation of investments in East-Asian securities.

It has also been argued that the rating agencies were excessively influenced by issuers and authorities in the relevant countries. For their part, the agencies have put forward examples of how critical statements concerning conditions in the Asian countries were counteracted in statements from e.g. major international investment banks which drew a considerably more optimistic picture of the situation.

Finally, it should be remarked that on starting up in a new market the quality of the rating agencies' work cannot really be evaluated until actual default frequencies can be compared with the ratings for a sufficiently large number of debtors in the relevant market, over a suitably long period. Since the agencies have only been active in Asia for a short time and cover relatively few issuers, this has so far not been possible.

Conclusion

The application of ratings to a capital market is closely associated with business enterprises' financing by issuing bonds. Several European capital markets are characterized by a close relationship between banks and business enterprises, where the enterprises by and large obtain financing only via banks and other financial institutions. This explains the so far relatively limited application of ratings in Europe. In step with the growing internationalization of the European capital markets, however, the conditions for the application of ratings appear to have changed. This tendency will presumably be amplified on the introduction of the third stage of EMU.

However, a number of conditions must be fulfilled before the ratings can function as a valuable supplement to the investors' own assessment of the credit risk on bond issuers and borrowers:

• Publicly available information that forms the basis for ratings must be reliable. This has not been the case in Asia.
  
  
• The rating agencies must have in-depth knowledge of the market. This seems to be the case in the USA, but apparently not in Asia. Disregarding the greater similarity between US and European financial markets differing legislation and accounting practices between otherwise closely related countries should not be underestimated.

If investors doubt the quality of the agencies' analyses and ratings, this will have a negative impact on the speed at which ratings are applied in Europe.

With regard to Denmark, a business structure with many small companies makes an increase in demand for ratings from borrowers unlikely. Only very few business enterprises besides those already rated by the agencies have a borrowing requirement which makes the issue of rated bonds interesting. In this connection the mortgage-credit institutes' great significance to the long-term borrowing of business enterprises plays a central role. Not many Danish companies are likely to be able to improve their terms of borrowing by issuing bonds directly in the market rather than by raising mortgage-credit loans.

On the other hand, among for example institutional investors growing use of ratings may be expected, on approximately the same scale as in other European countries, when they invest in European bond issues, cf. the above reservations, however. Any introduction of requirements from the authorities regarding placement of assets by institutional investors, e.g. minimum requirements of the rating of the issuers, will naturally make a significant contribution to this development.

Fodnoter