| dc.description.abstract | The current literature on this subject is sparce. One study by Peavy 18 centered on the building of a model for classifying the debt of bank holding companies into the catagories used by the major rating agencies. Peavy found that certain financial variables (total assets and total interest/total revenues) were especially significant in explaining bond ratings.
Hardy 8 cites confidentiality, flexibility, and nominal transactions costs as reasons why many investors use the private market. Zinberg 36 augmented the Hardy study by citing other private-market advantages including "negative clauses", which allow the lender to monitor the borrowers financial structure during the holding period, and "blind spots", which give the borrower a "grace period" before the commencement of debt service.
Shapiro and Wolf 26 note that when "switching" between the two markets does occur, it is usually based on interest-cost differentials, while Rea and Brockshmidt 22 found that other factors, such as those mentioned above, cause borrowers to forfeit the lower interest cost of the public market in lieu of nonpecuniary advantages associated with the private market.
The present study takes the Peavy paper a step further by rating private market issues with a model used to classify public bonds in an attempt to determine if there is any credit-quality differences between the two markets.
The statistical sample was gathered using data presented in Irvine Trust's Corporate Securities Issued: Commercial Banking 1 and included public and private bonds issued by banks from 1972-1979. The sample was then screened to include only those banks whose financial data is on the COMPUSTAT 3 computer-based tape series. Ratings were obtained from Moody's Investor Service 17. The Multiple Discriminant Analysis (MDA) method of multivariate statistics was used to find a model which was effective in predicting the ratings of bonds. Then, this same model (in various forms) was used to assign a rating to the private issues. By observing the differences in credit-quality ratings of the two markets, an inference can be made as to which market, if any, is financially stronger. The results of the analysis are as follows:
1. Little difference is observed between the credit quality of bank debt issues in the public and private markets-an equivalent amount of low and high quality issues was found in both markets.
2. The MDA model performed the poorest in trying to distinguish between A and AA rated issues.
3. The most significant financial variables were total assets (variable x12), borrowed funds (variable x11), and purchased funds (variable x10). Relatively insignificant financial variables were return on assets (variable x4), equity to loans (variable x7), and U.S. Treasury holdings (variable x9)
4. The important variables in number 3 (above) were significant in rating issues in both the public and private markets. | en |
