RDP 2004-02: The Impact of Rating Changes in Australian Financial Markets 6. Conclusion

The results of our study provide reasonably clear evidence that announcements of rating changes (or that ratings are ‘on watch’ or ‘under review’ for change) do appear to be ‘news’, in that they affect prices in the Australian bond and equity markets. Bond spreads appear to widen in response to ratings downgrades and contract with upgrades. Equity prices tend to fall on days of downgrades and rise on days of upgrades. This finding is significant in light of the relatively limited role – in contrast to some other countries – that ratings play in the Australian regulatory framework. It is also noteworthy that we find significant effects for both upgrades and downgrades, and both bonds and equities. Many previous foreign studies have failed to find such consistent evidence for market impacts. We attribute our identification of such effects to the reasonable sample size that we have in the case of equity events, and in the case of bonds, to the quality of our data and our focus on spreads rather than prices.

However, our estimates of the price impacts associated with rating announcements are arguably fairly small, so they suggest that agencies are not generally viewed as consistently having access to important information that is not already in the public domain. Instead, the modest size of the price impacts perhaps suggests that markets at most perceive agencies as simply obtaining and processing information in a way that at the margin provides a summary measure of creditworthiness.

Of course, looking ahead to a regime where ratings might have some role in regulatory bank capital requirements, it is possible that the decisions of agencies could affect market valuations because they impact directly on the cost of funds. In addition, if ratings become more widely used in debt contracts – for example, through ratings triggers which call for higher interest rates, provisions of more collateral, or even the repayment of the debt – then we might expect to see decisions of agencies have a greater impact on market prices. However, recent evidence suggests that in 2002 only around 15 per cent of large Australian firms had ratings triggers built into some of their borrowing agreements and also had ratings near enough to these triggers to warrant concern that they could be subject to potentially destabilising rating downgrades (see Standard & Poor's (2002)). Hence, we suspect our results are not significantly influenced by the use of such provisions.

In the case of equity returns and downgrades, we find very strong evidence that rating changes tend to follow long periods of underperformance. On average, companies that are downgraded have underperformed the market by around 20 per cent in the preceding period. This suggests that downgrades are typically based on information that is well-and-truly in the public domain, which would be consistent with the finding of Ederington and Goh (1998) that most bond downgrades in the United States are preceded by declines in actual corporate earnings and in stock analysts' forecasts of earnings.^[20] This would help to explain why the announcement-window impacts of rating changes were quite small. Yet, like most other work on rating agencies and market prices, our study yields a few conflicting findings, and it is puzzling that we find no evidence of noticeable pre-announcement movements in prices in the bond market or before upgrades in the equity market. These results may partly reflect the smaller sample of events in these cases. However, the finding is also suggestive of some form of segmentation between the markets for corporate bonds and equities, whereby bond investors fail to respond to information that is already reflected in the more liquid equity markets. If there has indeed been such segmentation in the past, we suspect it will be less prevalent in the future given the growing use of market-based credit risk models, which link bond valuations directly to equity price movements (see Lowe (2002)).

More broadly, it might be asked how our results relate to the debate over the role of rating agencies in the financial system. Critics of the agencies often point to the belated reactions of agencies in cases such as Enron. In addition, they point to the lack of market impact as evidence that agencies do not provide new information to the market. In contrast, rating agencies would argue that their ratings are opinions of long-term creditworthiness, and their aim of providing ratings that ‘look through the cycle’ would suggest that they should not respond to all short-term price movements: indeed, this reluctance to change ratings may also contribute to the stability of the financial system. Accordingly, if a firm's prospects have altered to such an extent that a rating change is warranted, it would not be surprising if markets had already adjusted substantially. This points to the tensions that exist between the conflicting interests of different end-users, and the relative importance they place on timeliness and stability of ratings.

Footnote

However, it should be noted that Ederington and Goh (1998) find that analysts do revise their earnings forecasts downwards after downgrades (with only small revisions following upgrades), which is consistent with the market viewing agencies' announcements as conveying some information. [20]