The Relationship of Asset Flow Measures to Bond Ratings
In recent times there has been increased interest in the potential decision usefulness of asset flow measures (i.e., income, income plus depreciation, working capital from operations, and cash flow from operations). In particular, some individuals and regulatory bodies have proposed to increase the disclosures and/or the analysis of corporate cash flows (Drtina and Largay , FASB [11, 12], Heath , Ijiri [22, 23, 24], Largay and Stickney , and Thomas ). The FASB  responded by issuing Statement of Financial Accounting Standards No. 95, which requires corporations to issue a cash-flow statement. While it is known that cash flow from operations is an empirically distinct construct from the other three asset flow measures (Gombola and Ketz [17, 18]), there is not much evidence as to whether cash flow is an important construct in decision contexts.
One area that has been investigated is the role of asset flow measures in the prediction of financial distress. Both Gentry, Newbold, and Whitford  and Casey and Bartczak  found that cash flow did not offer incremental significance to the prediction of corporate failure. Casey and Bartczak found that earnings performed quite well and were an important predictor of financial distress.
The present study is similar in spirit to the above research although the arena is different. We examine the ability of the asset flow measures to predict corporate bond ratings. The purpose of the study is to assess which of the asset flow measures best classifies corporate bond ratings.
Bond ratings are important to users because they provide a signal about the corporation's probability to default on the debt (Altman and Nammacher ) and because of their high correlation with debt-security yields (Foster [14, p. 509]). Models have been built to predict corporate bonds ratings by Horrigan , Poque and Soldofsky , West , Pinches and Mingo , Kaplan and Urwitz , and Belkaoui [4, 5]. Essentially they all have found that earnings is an important variable in the discrimination among the ratings. None of them, however, has investigated the ability of cash flow to discriminate among the bond ratings.
The results of this study are important to creditors since they will inform them of the importance of cash flow measures to bond raters. The results are also important to debtors since they will be interested in how their firm's debt is being analyzed and, in particular, whether cash flow numbers are important in the evaluation process. Finally, the FASB (and other standard setters) will be interested in the results since they reflect upon the appropriateness of SFAS 95.
DESCRIPTION OF THE DATA SHEET
Firms Employed in Study
The initial data set is taken from Maher , who investigated the impact of pension variables on bond ratings. His data set consisted of 182 observations in 1980, 212 in 1981, and 200 in 1982. We decided to eliminate duplicate data points (i.e., if firms had more than one bond rating, we chose one randomly). A few additional observations had to be eliminated because they did not contain the data to compute all of the independent variables. The resulting data set has 167 observations from 1980, 191 from 1981, and 185 from 1982.
All of the companies are industrial firms with SIC codes between 2000 and 3999. Each company had a bond issue outstanding that is listed in the December edition of the S & P Corporate Bond Guide of that year. The S & P bond rating was obtained from that reference book. Six bond ratings were noted in this study, one each for AAA, AA, A, BBB, and BB. The last group consists of B and CCC ratings because they separately had a small number of observations. The breakdown of the sample by year and by bond ratings is shown in Table 1.
Table : TABLE 1
FREQUENCY OF BOND RATINGS
S & P RATING 1980 1981 1982
1 (AAA) 12 11 12
2 (AA ) 37 43 42
3 (A ) 79 84 70
4 (BBB) 24 34 39
5 (BB ) 9 6 10
6 (B and CCC) 6 13 12
Total 167 191 185
Choice of the independent variables is problematic in this study because of the lack of a formal theory linking accounting numbers to corporate bond ratings. …