Description:
The present paper investigates the reaction of common stock returns to rating changes: it is the first empirical evidence for the Italian case. The analysis focuses on a sample of 299 rating actions announced by Fitch, Moody's and Standard amp; Poor's with respect to Italian firms in the period January 1991-August 2003. Rating changes announcement are classified according to the their direction (i.e. upgrades versus downgrades), their anticipation (i.e. rating actions following additions to a credit watch list), the presence of concurrent news, the reason of the rating action, the sector of the issuer. Additions to credit watch lists were also examined as separate events. The results showed significant average excess returns in the event window only for additions to the negative credit watch list and for actual downgrades. Additions to positive credit watch list and upgrades show statistically significant and positive abnormal returns after the event window. These evidence is consistent with the results found by the previous literature. Expected rating actions exhibit a greater impact on market prices than unexpected ones: such result may in part be explained by the fact that most of the expected rating actions are also anticipated by a concurrent disclosure. Contaminated events are in fact associated with a higher stock price reaction; this is also true for the sample of downgrades following mergers and acquisitions, which almost overlap with the contaminated sample. Finally, the impact of rating changes seems to be different depending on the sector of the issuer only for downgrades: the results show that negative abnormal returns are lower for financial than for industrial firms. Overall, the abnormal returns, if any, seem to be driven mainly by concurrent disclosure concerning the reason underlying the rating action. The striking result is, however, the absence of pre-announcement abnormal returns even for the contaminated subsample: this might be indirect evidence that rating agencies react promptly when to the news which is already in the public domain. On policy grounds, the evidence lends support to those advocating that ratings cannot be usefully employed as an instrument for investor protection