Αριθμητική απεικόνιση της εταιρικής πιστοληπτικής ικανότητας
KeywordsΒαθμοί πιστοληπτικής ικανότητας ; Αριθμητική απεικόνιση ; Οίκοι Πιστοληπτικής Αξιολόγησης ; Μη κανονικές αποδόσεις μετοχών ; Μελέτη γεγονότων ; Επεξηγηματική μεταβλητή ; Σφάλμα μέτρησης ; Credit rating ; Numerical mapping ; Numerical representation ; Credit rating agencies ; Abnormal returns ; Event study ; Control variable ; Measurement error ; Errors-in-variables
The expansion of the financial markets and the existence of more participants into them, have grown the necessity of extra information about their credit quality. Large portion of this information is transmitted through credit ratings that are released from the Credit Rating Agencies, which are also included often into regression models that are aiming to explain various market factors. Simultaneously, it has been observed that the transition to each rating level, has different effect on the rated firm, so its representation in a linear scale with a constant step (AAA=1, AA+=2, … etc), possibly does not reflect correctly the transmitted information and might lead to a measurement error of the credit rating variable. The aim of this thesis, is to study these different effects, approached by the abnormal stock returns of the rated firms, because of the announcement of credit rating change, applying an event-study methodology. Ultimately, based on the abnormal returns, there is estimated a new numerical representation of the corporate long-term credit ratings, which is a proposed methodology to include credit ratings into econometric models.