[The degree to which earnings changes are thought to be permanent may determine the magnitude of the security price response to a given earnings announcement [Brown et al., 1985; Kormendi and Lipe, 1987; and Easton and Zmijewski, 1987]. This study extends this literature by examining the impact on stock prices of quarter ahead and year ahead forecast revisions, as well as forecast errors. Results indicate that analysts' forecast revisions provide significant incremental explanatory power in a pooled time-series, cross-sectional regression of abnormal returns on forecast errors and analyst forecast revisions. In addition, the fourth quarter announcement appears to provide more information to analysts and investors than interim announcements.]

[This paper examines the issue of model specification in cross-sectional capital market research. In particular, it is argued that the decision of whether to select a price level or price change (return) specification is a joint function of (1) the economic model of equilibrium that is assumed; and (2) the nature of the econometric properties of the data that cause OLS assumptions to be violated. The historical origins of the model specification debate are traced to focus on three compelling reasons for selecting a returns-based methodology. We conclude that the relative advantages of returns and levels methodologies are dependent upon the set of assumptions maintained by the researcher regarding the pricing relation and the econometric properties of the data used for estimation. We provide examples and an interpretation of extant accounting capital market research to help illustrate our conclusions.]

Wayne R. Landsman, Aswath Damodaran, Using Shrinkage Estimators to Improve Upon Time-Series Model Proxies for the Security Market's Expectation of Earnings, Journal of Accounting Research, Vol. 27, No. 1 (Spring, 1989), pp. 97-115

[This study provides evidence on the relevance and reliability of option pricing-based value estimates for bonds and their components, i.e., conversion, call, put and sinking fund features. Findings reveal component value estimates represent large fractions of bond par value, and a fundamental components approach to account for corporate debt results in key financial statement amounts significantly different from those presently recognized. Component value estimates and financial statement amounts vary significantly with component estimation order. Thus, bond value estimates potentially meet the FASB's relevance criterion. However, estimate variation across component estimation orders and comparisons of estimates to available benchmarks indicate bond value estimates lack reliability.]

[This study provides evidence that fair value estimates of loans, securities and long-term debt disclosed under SFAS No. 107 provide significant explanatory power for bank share prices beyond that provided by related book values. In contrast to Eccher et al. (1996) and Nelson (1996), we consistently find incremental explanatory power for loans' fair values. Relatively stronger findings are obtained using a set of significant conditioning variables, including nonperforming loans, and interest-sensitive assets and liabilities. The joint significance of these loan-related variables and loans' fair values indicates that loans' fair values do not reflect completely loan default and interest rate risk. The loans' coefficient is significantly larger for banks with higher regulatory capital, which is consistent with market participants discounting unrealized gains on loans disclosed by less healthy banks. The findings are robust with respect to the inclusion of additional explanatory variables and to a first differences formulation.]

The degree to which earnings changes are thought to be permanent may determine the magnitude of the security price response to a given earnings announcement [Brown et al., 1985; Kormendi and Lipe, 1987; and Easton and Zmijewski, 1987]. This study extends this literature by examining the impact on stock prices of quarter ahead and year ahead forecast revisions, as well as forecast errors. Results indicate that analysts' forecast revisions provide significant incremental explanatory power in a pooled time-series, cross-sectional regression of abnormal returns on forecast errors and analyst forecast revisions, in addition, the fourth quarter announcement appears to provide more information to analysts and investors than interim announcements.

This paper examines the Issue of model specification in cross-sectional capital market research. In particular, it is argued that the decision of whether to select a price level or price change (return) specification is a joint function of (1) the economic model of equilibrium that is assumed; and (2) the nature of the econometric properties of the data that cause OLS assumptions to be violated. The historical origins of the model specification debate are traced to focus on three compelling reasons for selecting a returns-based methodology. We conclude that the relative advantages of returns and levels methodologies are dependent upon the set of assumptions maintained by the researcher regarding the pricing relation and the econometric properties of the data used for estimation. We provide examples and an Interpretation of extant accounting capital market research to help illustrate our conclusions.

Portfolios are formed directly and exclusively upon residual return behavior in the months prior to portfolio formation. The empirical behavior of residual return in the post-formation period is then examined. Based upon the overall time period studied (1932 through 1977), the average residual return is essentially zero in the months subsequent to the portfolio formation. However, systematic (i.e., non-zero) residual behavior is observed in particular years. Moreover, the results suggest the possibility that ‘abnormal’ returns observed after certain events (e.g., earnings announcements) may at least in part reflect more general phenomena associated with being ‘winners’ and ‘losers’ in terms of residual returns in the months previous to the event.

This paper examines the extent to which stock prices fully reflected the off‐balance‐sheet item net over‐ (under‐) funded pension liabilities during the four‐year period 1979–82. The evidence suggests that the market was informationally inefficient and that it displayed a general underreaction relative to the net pension liability. Significant excess returns could have been earned by using the net pension liability as an instrument for portfolio selection. The empirical analysis fails to identify the existence of confounding variables such as beta risk, size, or industry effects. Résumé. Les auteurs examinent dans quelle mesure le cours des actions a reflété l'évolution du poste net hors bilan de l'excédent (ou du déficit) de provisionnement de la dette au titre des régimes de retraite, pendant la période de quatre ans écoulée entre 1979 et 1982. Les constatations des auteurs laissent supposer que le marché était inefficient sur le plan informationnel et qu'il a enregistré une réaction générale mitigée à la dette nette au titre des régimes de retraite. L'utilisation de la dette nette au titre des régimes de retraite comme critère de sélection de portefeuilles aurait permis d'obtenir des rendements excédentaires importants. L'analyse empirique ne parvient pas à déterminer l'existence de variables contradictoires telles que le risque Beta, la taille ou les effets du secteur.

This paper examines changes in the information content of earnings over the past three decades using the two metrics from Beaver [1968]: abnormal trading volume and abnormal return volatility. We find no evidence of a decline in the information content of earnings announcements over the past three decades, as measured by both abnormal trading volume and return volatility around quarterly earnings announcements. If anything, our results suggest an increase over time in the informativeness of quarterly earnings announcements. Variables reflecting changes in firm‐specific factors account for a portion of the observed increase.