Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

Cross-sectional Variation of Measurement Error and Predictability of Earnings and Stock Returns

Abstract

In capital markets research, market expectation of future earnings plays a vital role. However, almost all proxies inevitably measure the market expectation of future earnings with error, which results in unsatisfactory empirical outcomes in prior research (e.g., small empirical values of earnings response coefficient and poor quality estimates of expected rates of return). Using analysts' consensus forecasts, this study investigates how noisy measurement of the market expectation of future earnings affects the predictability of future earnings and stock returns. Based on the errors-in-variables approach, this study first provides a framework to capture cross-sectional variation of the measurement error in analysts' consensus forecasts. With this framework in place, this study documents that analysts' consensus forecasts with more measurement error have less ability to predict future earnings and stock returns, and that incorporating information about cross-sectional variation of the measurement error can improve the predictability of future earnings and stock returns. These findings will be useful to accounting research that relies on the market expectation of future earnings and to practitioners seeking to forecast profitability and stock returns.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View