Measuring The Usefulness Of Information For Investors Inferences: The Case Of Quarterly Earnings Announcements

Abstract

In this paper, we propose a new method for assessing the usefulness of information, its inferential value. In the context of accounting and finance, we define the inferential value of information about a firm as how efficaciously the information enables investors to draw correct inferences regarding its future financial performance. On the basis of this definition, we develop a stylized model to measure the proximity of a firm’s future realized rates of return to the estimated rates of return implied by its current stock price. We then use the new measure to test the hypothesis that quarterly earnings announcements have a higher inferential value than other information arriving during interim (non-earnings announcement) periods. Our empirical findings suggest that investors are able to make more informative inferences about a firm’s future profitability based on quarterly earnings announcement than based on information available during interim periods. However, our findings also suggest that, in general, investors do not correctly anticipate future losses. Finally, we find that earnings announcements are as important in anticipating future profitability for larger firms as they are for smaller firms.