Analysis of Determinants of Split Ratings and Rating Conservativeness between Japanese and US Credit Rating Agencies

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.

An Empirical Investigation of New Bond Issue Yield Spreads, Default Risk and Split Ratings

This paper attempts to explain the yield spreads charged to new corporate debt issues by comparing the initial yields of a set of 3,287 securities issued over eleven years in the US. We use the measure of constant maturity Treasury rates on the day of issue against the Moody’s Aaa Corporate Bond index for the week prior to the issue, and the yield on a daily index of long-term Treasury securities on the issue date. The influences of credit ratings and disagreement between rating agencies as reflected in split ratings and the interactions between these characteristics are measured. The contributions of sinking fund provisions, call or refunding status, overseas issue and contractual security arrangements are evaluated separately. The results support the view that the higher yields are observed when ratings of agencies differ and that factors associated with the issues also are significant drivers of the yield difference.