Models and Data for Bank Ratings
Here, and further in this section, ordered probit/logit econometric models were used to forecast rating grades (for example, see Peresetsky and Karminsky (2011)). Numeric scales for ratings were also used as a result of the mappings mentioned in Fig. 1. For the main international CRAs, nearly 18 corporate rating grades were used.
The original databases for different classes of entity were used. There were two different databases used separately for banks for both international and Russian ones. The first database was obtained from Bloomberg data during the period 1995— 2009. The database includes 5,600+ estimations for 551 banks from 86 countries. The data contains the banks from different countries including more than 50 % from developed and 30 % from developing countries. Russian banks are also included in the sample and form nearly 4 %.
The second database was constructed from the data for Russian banks according to Russian financial reporting. It contains 2,600+ quarterly estimations from 2006 until 2010 for 370 Russian banks.
We carried out model choices from different points of view for three agencies simultaneously. We determined which financial explanatory variables were the most informative ones. Then we considered quadratic models, using macro, market and institutional variables, as well as dummies. We used a rating grade as a dependent variable where the lower numbers were associated with a better rating. So a positive sign in the coefficient related to a negative influence on the ratings, and vice versa.
For the international bank models, an accurate forecast was generated in nearly 40 % of the cases. The forecasting power may be estimated by the mistakes of the models, which in the case of no more than two grades gave the probability of
Notes: *, **, *** represent 10%, 5%, 1% levels of significant, respectively. Italic texts were connected with statistical summaries of the tables.
1-2 %. These results were comparable with the previous models, but extended to three international rating agencies simultaneously.
The signs for all the models were almost equal and could be easily explained from a financial point of view. Coefficient sign analysis allowed us to make the following conclusions:
• The size of the bank is positive for a rating level increase, as are capital ratio and asset profitability as the retained earnings to total assets ratio.
• Such ratios as debt to asset and loan loss provision to total assets have a negative influence on the rating grade.
• Macro variables are also important for understanding the behavior of bank ratings, and are presented with a negative sign for the corruption index and inflation.
We also constructed the models for Russian banks ratings using a Russian database, and have concluded that the influence of financial indicators is mainly the same (Vasilyuk et al. 2011).