Examining the Predictive Values of Accounting Summary Numbers in Annual Reports

Abstract

A key objective of financial reporting is to provide useful information for external users to make economic decisions. Accounting summary numbers are cost-effective ways of conveying firm-specific information, helping users predict future earnings and future cash flows that will be used as input in their decisions. But users also collect information from a myriad of other sources throughout the year, raising an intriguing question about the role of accounting summary numbers in capital markets in the digital age. To address that question, this study examines the relationships between the predictive values and market variables. The predictive values are defined as the ability of accounting summary numbers to predict future earnings or future cash flows and are measured as −1 times the absolute value of residuals from the regressions of one-year-ahead earnings or one-year-ahead cash flows on contemporaneous values of earnings, cash flows, and book values in annual reports. The study documents fewer forecast errors, smaller forecast dispersions, smaller bid-ask spreads, and lower cost of capital for firms with higher predictive values. The study also documents inter-temporal variations in the relationships during more digitally intensive periods, highlighting the role of accounting summary numbers in capital markets in the digital age. The study finds that the predictive values are positively correlated with financial reporting quality (FRQ) and exhibit properties similar to FRQ in multiple regression analysis, suggesting that the predictive values can be used as alternative measures of FRQ.