Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)




This study provides empirical evidence to support the FASB's contention that current cost data as specified by SFAS No. 33 may better aid financial statements users to assess future cash flows than historical cost financial information. Models derived from the theoretical and empirical literature were used to forecast 1981 operating cash flow. One set of models employed, as explanatory variables, historical cost data while another comparative set employed current cost data. A sample of 361 firms, representing 33 industries, was drawn from the FASB Statement 33 Data Bank. The firms were stratified by industry classification, then randomly assigned to sample subgroup (A) or (B). Sample group (A) was used to develop cross-sectional model parameters for the predictor models. Sample group (B) was employed to forecast 1981 operating cash flows. Vectors composed of the models' mean relative forecast errors were constructed. One set of vectors was used to simultaneously test the current cost models' predictive ability against that of the historical cost models. Another set of vectors was employed to test whether there was a significant difference in the predictive ability of current cost models whose explanatory variables include a measure of total replacement cost income (current cost income from continuing operations plus holding gain or loss) and current cost models whose explanatory variables include only the income from continuing operations component of total replacement cost income. The test statistic employed in tests of these hypotheses was Hotelling's T('2). Predictor models which employ total replacement cost more closely predicted subsequent cash flows than models employing only current cost income from continuing operations as a predictor variable. Only those current cost cash flow predictor models which employ total replacement cost income as a predictor variable, were more accurate than historical cost models. The current cost model whose forecasting ability was the most accurate, as compared to its historical cost counterpart, was a more accurate predictor of 1981 operating cash flow than a comparable forecast model whose predictor variable was 1980 operating cash flow.