Open AccessAnalysis of Determinants of Split Ratings and Rating Conservativeness between Japanese and US Credit Rating Agencies
Author(s) -
Yoshiki Shimizu,
Junghee Lee,
Hideki Takei
Publication year - 2013
Publication title -
international journal of accounting and financial reporting
Language(s) - English
Resource type - Journals
ISSN - 2162-3082
DOI - 10.5296/ijafr.v3i1.3507
Subject(s) - credit rating , bond credit rating , leverage (statistics) , market liquidity , multivariate statistics , actuarial science , asset (computer security) , business , bond , agency (philosophy) , multivariate analysis , econometrics , economics , psychology , credit risk , finance , statistics , credit reference , mathematics , computer science , philosophy , computer security , epistemology
In the previous paper, we confirmed the existence of the split ratings between Japanese and US credit rating agencies (CRAs). Our study did not support early studies suggesting that the split ratings were merely random occurrences. Rather, our findings suggested that the split ratings occurring between Japanese and US CRAs were not random and frequently occurring. The Japanese CRA assigned less conservative ratings than the US CRAs. In this paper, we performed the multivariate regression analysis to find variables which would differentiate the degree of rating conservativeness. Our samples were 192 Japanese companies which were assigned their ratings by Japanese and US credit rating agencies. We used 10-year bond ratings of these companies from 2000 and 2009. Our data sources were Nikkei NEEDS-Financial Quest for Japanese ratings and financial information and Thomson Reuters Datastream for US ratings. All financial data of the 192 firms were collected from Nikkei NEEDS-Financial Quest. According to our findings, Japanese agency seems to put higher weight on ROA than US agencies while all agencies seem to use variables such as asset, liquidity, and leverage to assign ratings. We assume that this is the main variable that has differentiated the degree of rating conservativeness.