Abstract
*Stanford University; tCornell University; +Northwestern University. We appreciate comments from workshop participants at Baruch College, Cornell University, Harvard Business School, the University of Iowa, the University of North Carolina, the University of Notre Dame, Ohio State University, Pennsylvania State University, the University of Rochester, Temple University, the University of Waterloo, the Financial Economics Conference at SUNY Buffalo, the American Accounting Association Annual Meeting, and the Financial Decisions and Control Seminar at Harvard Business School, and from four anonymous reviewers, Dan Collins, Sanjay Kallapur, Wayne Landsman, Charles Lee, Barbara Lougee, and Krishna Palepu. We also appreciate the research assistance of Don Cram, Philip Joos, and Peter Joos. Funding from the Financial Research Initiative, Stanford University Graduate School of Business, the Stanford Business School Class of 1969 Faculty Fellowship, Cornell's Accounting Institute, the Institute of Professional Accounting at the University of Chicago, Northwestern's Accounting Research Center, and Division of Research, Harvard Business School is gratefully acknowledged. We thank IIBIEIS for use of analyst forecast data. 1 The term multiple refers to either the coefficient on earnings in price regressions or the coefficient on earnings changes in returns regressions. We refer to firms with nondecreasing patterns of earnings as firms with increasing patterns. Fewer than 1% of the firm-year observations have a zero year-to-year change as part of an increasing earnings pattern.