The study inquires whether ratings and rating changes in particular have anyimpact on bond prices. In doing so, the price data of a number of corporate vanillabonds were investigated over corresponding rating change event periods.For various reasons, the South African Bond Market could not deliver a largesample of plain vanilla bonds that experienced a rating change. Therefore, tocomplement the study’s main sample, a small sample of callable corporatevanillas was also studied. In both cases the eventual sample sizes necessitatedan investigation based on single issues. Neither of the samples contained“serious” rating changes; the lowest rating studied was of lower-mediuminvestment grade quality and none of the rating changes was by more than onenotch.Stationarity and unit-root tests were conducted on the non-callable and callableissues’ individual price series to determine whether the issues experiencedsignificant price impacts in response to their respective rating-changeannouncements. Based on the test-results, only a number of non-callable issuesexperienced significant impacts. The number shrunk considerably aftercontamination could be proven in some cases, and when the directions of theyield spread movements were taken into consideration. An even smaller numberof callable issues registered significant impacts. Contamination was not thatcustomary in the case of callable issues, but direction of movement still played arole.Overall, out of the ten non-callable issues, none appear to have been impacted bytheir respective rating changes; out of the nine callable bonds, only one may havebeen impacted by a rating change, conditional on the issue’s call option. In somecases the rating changes were preceded by rating warnings – whether ratingoutlooks or rating watches; the study did not investigate whether the marketresponded to these warnings.The study also concludes that issue-specific characteristics play an important rolein determining whether ratings can potentially impact prices. With regards to this –iiTime-to-Maturity is a good example. Additionally, it appears as if rating datashould be a lot more issue-based - as opposed to being issuer-based - in order tobe of any use, at least in the case of bonds that is.Under the Literature Review, enormous attention is paid to bond pricedeterminants, the impact thereof on credit ratings, and thus the resultant ability ofcredit ratings to impact bond prices. The review manages to unravel thecorrelation between credit ratings and bond prices, to some extend. In addition,the study offers new insights into the pricing of risky bonds, particularly through analternative modelling of default losses.
