The name is absent

probit and random effects ordered probit specifications. The latter is the best procedure
for panel data as it considers the existence of an additional normally distributed cross-
section error term. This approach allowed both to determine the cut-off points
throughout the rating scale as well as assessing whether a linear quantitative
transformation of the ratings is actually more appropriate than a possible non-linear
transformation.

Our main findings in the panel random effects framework allowed us to detect a set of
core variables that are relevant for the determination of the ratings: per capita GDP;
GDP real growth rate; government debt; government effectiveness; external debt and
external reserves; sovereign default indicators. Moreover, the importance of fiscal
variables appears stronger than in the previous existing literature.⁷

The ordered probit analysis confirmed the overall estimation results from the linear
panel regressions. Interestingly, there is some evidence for different approaches of the
agencies with regard to the distance between ratings thresholds. For instance, for
Moody’s the estimated thresholds point to a relatively large jump between the ratings
for BBB- and BBB. This suggests that countries close to the non-investment grade
rating are given a wider range before they actually cross that threshold. For Fitch, the
hypothesis of equal distances is strongly rejected as the thresholds for higher ratings are
further apart than those of the lower ratings. In this case the kink lies at the A rating. On
the other hand, no clear switching pattern emerges for S&P.

The panel sample we used is quite comprehensive, which allowed for a sub-period
analysis and for a differentiated high and low rating analysis. While the results are
roughly stable across agencies, time periods and ratings levels, some additional
interesting results emerge. For instance, for the low rating levels, external debt and
external reserves are more relevant. On the other hand, for the early sub-period, 1996-
2000, the current account balance was more important, while external reserves were
possibly somewhat more important in the later period, 2001-2005 (for Moody’s and

⁷ We performed additional analysis from some different perspectives. For instance, we used the
information on credit rating outlooks but no relevant improvement on the fit of the models occurred. In
addition, we assessed also whether different exchange rate regimes added information to the rating
determination, but that was not the case.

29

<<   26   27   28   29   30   31   32   >>

More intriguing information