Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

THE SUPERIORITY OF ANALYST FORECASTS AS MEASURES OF EXPECTATIONS: EVIDENCE FROM EARNINGS

THE SUPERIORITY OF ANALYST FORECASTS AS MEASURES OF EXPECTATIONS: EVIDENCE FROM EARNINGS The Journal of Finance I. A. EXPERIMENTAL DESIGN Statistical Evaluation of Forecast Methods Without direct information on the costs of imperfect forecasts to forecast users, comparative forecast accuracy is usually evaluated by comparing the error distributions of different forecast methods statistically. However, statistical comparisons in past studies ([9], (11)) utilize test statistics improperly, particularly Theil's V [25] and Student's t. In this section, after discussing the defects of these statistics for evaluating two or more forecast methods, the alternative statistical methods used in this study are introduced.i Theil's V-statistic (applied to earnings) is the square root of where Pijt Ail = change in = predicted actual earnings per share of firm i from t - 1 to t, change in earnings per share of firm i from t - 1 to t by forecast method j, and T= total number of time series observations. For its computation, it requires time series data on a firm's earnings changes? Given forecast method j and earnings time series data on firm i, Theil's V compares the forecast accuracy of method j to that of a naive, no change, earnings forecast model.i' Since analysts' earnings forecasts are currently available only in short http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journal of Finance Wiley

THE SUPERIORITY OF ANALYST FORECASTS AS MEASURES OF EXPECTATIONS: EVIDENCE FROM EARNINGS

Loading next page...
 
/lp/wiley/the-superiority-of-analyst-forecasts-as-measures-of-expectations-jzmvthAmo6

References (17)

Publisher
Wiley
Copyright
1978 The American Finance Association
ISSN
0022-1082
eISSN
1540-6261
DOI
10.1111/j.1540-6261.1978.tb03385.x
Publisher site
See Article on Publisher Site

Abstract

The Journal of Finance I. A. EXPERIMENTAL DESIGN Statistical Evaluation of Forecast Methods Without direct information on the costs of imperfect forecasts to forecast users, comparative forecast accuracy is usually evaluated by comparing the error distributions of different forecast methods statistically. However, statistical comparisons in past studies ([9], (11)) utilize test statistics improperly, particularly Theil's V [25] and Student's t. In this section, after discussing the defects of these statistics for evaluating two or more forecast methods, the alternative statistical methods used in this study are introduced.i Theil's V-statistic (applied to earnings) is the square root of where Pijt Ail = change in = predicted actual earnings per share of firm i from t - 1 to t, change in earnings per share of firm i from t - 1 to t by forecast method j, and T= total number of time series observations. For its computation, it requires time series data on a firm's earnings changes? Given forecast method j and earnings time series data on firm i, Theil's V compares the forecast accuracy of method j to that of a naive, no change, earnings forecast model.i' Since analysts' earnings forecasts are currently available only in short

Journal

The Journal of FinanceWiley

Published: Mar 1, 1978

There are no references for this article.