Measuring Value-Relevance in Accounting-Based Variables Without Reference to Market Prices

37 Pages Posted: 1 Jun 1998

See all articles by Ilia D. Dichev

Ilia D. Dichev

Emory University - Department of Accounting

Date Written: December 1997

Abstract

Properly anticipated equity values (e.g., prices in efficient markets) summarize all available information, and change only in response to new information. New information is by definition unpredictable, which implies that unpredictability of changes is a fundamental characteristic of all proper measures of equity value. Thus, investigating the unpredictability of changes of accounting-based measures of equity value (e.g., EBO values and earnings) is potentially useful in assessing their value-relevance. A major advantage of the unpredictability of changes approach is that it allows for value-relevance investigations which are impossible or problematic with market-based data. To explore whether the unpredictability approach is suitable for practical applications, I formulate predictions based on existing market-based research. Using first-order autocorrelation in changes as a proxy for predictability, I obtain results which are consistent with the findings of market-based research. This agreement in findings provides further assurance about existing results, and suggests that the unpredictability approach is powerful and flexible enough to be of practical use. The paper concludes with a section that illustrates how the unpredictability approach might serve the needs of standard setters in accounting.

JEL Classification: M41, G12

Suggested Citation

Dichev, Ilia D., Measuring Value-Relevance in Accounting-Based Variables Without Reference to Market Prices (December 1997). Available at SSRN: https://ssrn.com/abstract=94328 or http://dx.doi.org/10.2139/ssrn.94328

Ilia D. Dichev (Contact Author)

Emory University - Department of Accounting ( email )

1300 Clifton Road
Atlanta, GA 30322-2722
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
1,244
Abstract Views
3,908
rank
15,587
PlumX Metrics