This paper presents evidence indicating that the magnitude of the post-earnings-announcement drift is positively related to the direct and indirect costs of trading. Share price and annual dollar trading volume are chosen to proxy for the inverse of direct and indirect costs of trading, respectively. Drift is found to be inversely related to both these variables and these relations subsume the previously documented inverse relation between drift and firm size. The paper's evidence suggests that transactions costs are an important determinant of the efficiency of capital markets.

This paper develops a model of information collection about publicly traded firms in an economy. The supply noise is modeled as the variability of liquidity-motivated trading in the shares of the firm. The paper theoretically examines the influence of various firm characteristics on the amount of information collected about a firm and on the marginal information content of announcements made by it. Empirical work focuses on the marginal information content of quarterly earnings announcements made by firms. The empirical results are generally consistent with the model's predictions.

This paper provides additional evidence on the relation between the actual and asymptotic standard errors under generalized least squares (GLS). Simulations are conducted in the event‐study framework using both daily and weekly returns. The case of simultaneous calendar time and industry clustering is also considered. The major finding is that the asymptotic standard errors under GLS can significantly underestimate the true standard errors when the number of parameters estimated is not small compared to the number of observations available to estimate them. To minimize the possibility of incorrect inferences, use of GLS should in general be combined with standard errors obtained from other techniques such as bootstrap. Résumé. L'auteur apporte de nouvelles preuves de la relation entre les erreurs‐types réelles et asymptotiques dans le cadre de l'utilisation des moindres carrés généralisés. Il procède à des simulations dans le contexte d'une étude d'événements, en utilisant les rendements à la fois quotidiens et hebdomadaires. Le cas de simultanéité de la date et du groupage par secteur d'activité sont également pris en considération. Principale conclusion de l'auteur: les erreurs‐types asymptotiques dans le cadre de l'utilisation des moindres carrés généralisés peuvent entraîner une sous‐estimation importante des erreurs‐types véritables lorsque le nombre de paramètres estimés n'est pas restreint par rapport au nombre d'observations disponibles pour formuler ces estimations. Si l'on veut réduire au minimum le risque d'inférences inexactes, l'utilisation des moindres carrés généralisés devrait, de façon générale, être combinée aux erreurs‐types obtenues à partir d'autres techniques comme celle de l'amorce.

[In this article I develop a model that accounts for interdependence between trading costs in various asset markets arising from the optimizing behavior of liquidity traders. The model suggests that noise trading is an important determinant of the liquidity of asset markets and provides a positive theory for diversified asset holdings by risk-neutral liquidity traders.]

In this article I develop a model that accounts for interdependence between trading costs in various asset markets arising from the optimizing behavior of liquidity traders. The model suggests that noise trading is an important determinant of the liquidity of asset markets and provides a positive theory for diversified asset holding by risk-neutral liquidity traders. Article published by Oxford University Press on behalf of the Society for Financial Studies in its journal, The Review of Financial Studies.

We analyze myopic trader models of noisy prices in financial markets. Unlike extant analysis, such as De Long et al. (1990a), a classical equilibrium exists in our analysis, e.g., a riskless perpetuity is priced by arbitrage and its price does not vary with noise. A unique noisy equilibrium exists only when i) noise traders' beliefs are rational regarding volatility and irrational regarding expected returns, and ii) noise traders can hold infinite positions. In the absence of these strong assumptions, multiple noisy equilibria can coexist with the classical equilibrium, but these equilibria exhibit conflicting comparative statics. Furthermore, the price of a long-lived asset with risky cash flows can vary with noise even when investors are not myopic. One conclusion is that myopia is neither a necessary nor a sufficient condition for noisy prices. A second is that it is difficult, if not impossible, to use myopic trader models to derive implications for investment or regulatory policy.