Presents a reply to commentaries on a study about accuracy of earnings forecast. Clarification on the objectives of the study; Criticism on the model used in the study.

The "Comment and Extension" of R. Picur and J. McKeown (hereafter, P and M) examines three aspects of "Alternative Income Concepts and Relative Performance Evaluations" (hereafter, KMT). The differences with P and M per their respective classifications is noted with respect to theoretical, methodological and statistical terms. The statistical extension of P and M, in particular, is completely invalid. The form of current value reporting used in KMT was shown to have theoretical support both in academia and in the profession. It does not agree with the proposal of P and M. However, the purpose was not at all to advocate a specific current value form, but to consider whether a typical current value form would indicate a performance difference among firms in an industry as compared to a typical historical cost form. The P and M proposed adjustment of liabilities to reflect changing market rates clearly would change the performance measures of the sample of KMT. However, there is no general agreement that liabilities should be so adjusted in current value financial statements, any more than there is general agreement that annuity depreciation should be used for historical cost reports.

In this study, the question of forecast accuracy by examining forecasts released by firms during 1970 and 1971 is considered. Since these forecast releases were voluntary on the part of the firms involved, there is some question about the validity of drawing inferences regarding overall accuracy of mandatory forecasts; thus, some results are included which are merely descriptive, in addition to certain statistical tests addressed more to the issue of inference. A second aim of the study was to examine the accuracy of forecasts made by outsiders (security analysts) relative to the forecasts of firms. Ideally, one would expect corporations to be able to forecast their earnings more accurately than outsiders, even when those outsiders are professional analysts. In short, based on the empirical results, forecast accuracy does not appear to be highly impressive for either group. Given the potential disruptive effects of poor forecasting on investment decisions and, as a result, on current stock market mechanisms, perhaps it would be wise to declare a moratorium on pressures for mandatory published forecasts by firms until forecasting techniques have been refined sufficiently to assure considerably better accuracy than apparently exists.

This article presents response of the author on comments made by scholar Shu Liao on the paper "An Intra-Industry Comparison of Alternative Income Concepts and Relative Performance Evaluations" that was published in the October 1974 of the periodical "The Accounting Review." Liao examines the research methodology in three sections--statistical methods, components of income, and test of impact. First, however, it is pointed out that some of his statements might be interpreted to be full quotes of the study when, in fact, they are not. For example, in the second sentence of his note, he states the authors have concluded that each income concept could be used as a surrogate for other concepts. The statement is that, in many instances, each income concept could be used as a surrogate for the other concepts. The degree of agreement among different income concepts as a result of statistical tests is shown in each cell. He has made valuable suggestions for a number of extensions and for different approaches to studying the practical impact of alternative income concepts.

This article presents an intra-industry comparison of alternative income concepts and relative performance evaluations. While the amount of research on basic income concepts continues to be voluminous, questions of practical impact only recently have begun to receive much attention. In order to contain the research to a manageable size, a single industry, real estate investment trust (REIT), was selected for investigation. Also, only one of the three basic forms of real estate investment trusts, the equity real estate investment trust, was selected. There are several reasons for choosing this particular industry and type of trust. REITs are a homogeneous grouping of companies with a relatively limited history. Thus, for these companies it was expected that financial data would be available from the first year of operations to the current period. Such data are desirable in order to adjust properly for price-level changes. Since REITs are a homogeneous group, the application and selection of specific price-level adjustments are not as difficult as with less homogeneous companies.