Compare Quadratic Loss Function with Linear Loss Function in Rational Expectation Hypothesis on Analysts' Earnings Forecasts

In order to examine the robustness of Basu and Markov' s findings, we estimate the rationality of earnings forecasts issue based on the UK data. We employ Ordinary Least Squares (OLS) and Least Absolute Deviations (LAD) regression to test the hypothesis and obtain rejection of the null hypothes...

Full description

Bibliographic Details
Main Author: Wan, Fang
Format: Dissertation (University of Nottingham only)
Language:English
Published: 2008
Online Access:https://eprints.nottingham.ac.uk/22197/
Description
Summary:In order to examine the robustness of Basu and Markov' s findings, we estimate the rationality of earnings forecasts issue based on the UK data. We employ Ordinary Least Squares (OLS) and Least Absolute Deviations (LAD) regression to test the hypothesis and obtain rejection of the null hypothesis based on the statistical significant. However, consistent with Basu and Markov' s studies, we emphasize on the importance of the economic significance and consider that analysts are economically rationality on account of the linear loss function. Moreover, we provide critical analysis of our methods and conclusions in the end of the article. In addition, we suggest that although analysts appear to be not able to meet rational expectation theory in the statistical level, their forecasts may still be the optimal prediction of future.