Since cost allocation has been widely viewed as an arbitrary process, the theory of choice among allocation alternatives is little advanced. This paper illustrates some practical problems that result in the absence of well-established guidelines for choosing among cost allocation alternatives. Specifically, two appeals before the Armed Services Board of Contracts Appeals involving the issue of cost allocation are discussed and analyzed.

This paper introduces a model in which the firm's returns depend on trading volume when the firm defers disclosure, because market makers use volume to draw inferences about better-informed investors' private information on firm value. In addition, we show that a firm's commitment to a policy of timely disclosure of a broader range of outcomes dampens the slope coefficient on volume in a Taylor expansion of the log of the absolute value of returns on volume. The reason for this is that when a firm commits to a policy of timely disclosure of a broader range of outcomes, the deferral of a report indicates that the outcome is extreme. This heightens adverse selection. In turn, this makes informed trade more costly and hence less likely, thereby rendering returns less dependent on trading volume as a source of information about firm value. This result predicts that firms committing to more disclosure should experience a smaller slope coefficient on trading volume. Thus, the slope coefficient offers a potential tool for measuring the economic consequences of shifts in disclosure regimes.

We present a simple model of managerial reporting bias for a setting in which the capital market is uncertain about the manager's reporting objective. In this setting, the manager's reporting bias reduces the value relevance of the manager's report; that is, it adds noise to the report. Through comparative static results, our model yields insights into factors that affect the slope and intercept terms in a regression of price on earnings. Specifically, we find that the information content of the manager's report, as captured by the earnings slope coefficient, falls as the private cost to the manager of biasing reports falls, and as the uncertainty about the manager's objective increases. We also find that the magnitude of the adjustment for the expected amount of bias, as captured by the absolute value of the intercept, falls as the uncertainty about the manager's objective increases. Finally, to highlight conditions under which managers would lobby to retain an option to bias reports (i.e., retain reporting flexibility), we analyze the effect of the option to bias on the manager's welfare. For example, we show that the ex ante benefit from biasing the report is positive if there is sufficient uncertainty about the manager's reporting objective.

[We reexamine the "uniformity vs. flexibility" debate by considering the consequences of varying the amount of discretion managers have in reporting current period expenses. We study the effects of altering GAAP on both the "internal agency problem" between current shareholders and their manager and the "external agency problem" between current and prospective shareholders. We show that the internal agency problem is ameliorated by expanding discretion, and we exhibit examples where expanding discretion is undesirable when the internal and external agency problems are present concurrently and the nonexpense-related components of earnings are measured with error. We establish that expanding discretion is welfare-enhancing if either the manager's contract is publicly observable or else the nonexpense-related components of earnings are measured without error. These results demonstrate the limitations of evaluating GAAP on a piecemeal, or issue-by-issue, basis.]

This paper presents a partially revealing rational expectations model of competitive trading to identify two effects of information releases; an informedness effect and a consensus effect. The informedness effect measures the extent to which agents become more knowledgeable, and the consensus effect measures the extent of agreement among agents at the time of an information release. We demonstrate that informedness and consensus generally occur jointly when information is disseminated, and that unexpected price changes and trading volume are each influenced by both informedness and consensus. Thus, interpretations of unexpected price changes and volume associated with information releases are conceptually similar. Since informedness and consensus each affect both the variance of price changes and volume, our paper provides an economic rationale for examining both price and volume effects at the time of information releases.

[This paper presents a partially revealing rational expectations model of competitive trading to identify two effects of information releases; an informedness effect and a consensus effect. The informedness effect measures the extent to which agents become more knowledgeable, and the consensus effect measures the extent of agreement among agents at the time of an information release. We demonstrate that informedness and consensus generally occur jointly when information is disseminated, and that unexpected price changes and trading volume are each influenced by both informedness and consensus. Thus, interpretations of unexpected price changes and volume associated with information releases are conceptually similar. Since informedness and consensus each affect both the variance of price changes and volume, our paper provides an economic rationale for examining both price and volume effects at the time of information releases.]

We empirically examine standard agency predictions about how performance measures are optimally weighted to provide CEO incentives. Consistent with prior empirical research, we document that the relative weight on price and non-price performance measures in CEO cash pay is a decreasing function of the relative variances. Agency theory speaks to the weights in total compensation (annual total pay and changes in the CEO's equity portfolio value), however, and we document that very little of CEOs' total incentives come from cash pay. We also document that variation in the relative weight on price and non-price performance measures in CEO total compensation is an increasing function of the relative variances. The conflicting results using total compensation indicate that existing findings on cash pay cannot be interpreted as evidence supporting standard agency predictions. Based on our results, we suggest approaches for future research on performance measure use in CEO total compensation.

Prior literature predicts a positive relation between firms' expected pre-tax rates of return and investor-level capital gains tax rates. We show that this relation is more nuanced than suggested by prior literature and that in three circumstances the relation can actually be negative. The first circumstance is when a firm's systematic risk is very high. The second circumstance is when the market risk premium is very high. The third circumstance is when the risk-free rate of return is very low. The circumstances arise because, in addition to reducing investors' expected after-tax cash proceeds, capital gains taxes reduce the risk that investors associate with the expected after-tax cash proceeds.

Prior literature predicts a positive relation between firms' expected pre-tax rates of return and investor-level capital gains tax rates. We show that this relation is more nuanced than suggested by prior literature and that in three circumstances the relation can actually be negative. The first circumstance is when a firm's systematic risk is very high. The second circumstance is when the market risk premium is very high. The third circumstance is when the risk-free rate of return is very low. The circumstances arise because, in addition to reducing investors' expected after-tax cash proceeds, capital gains taxes reduce the risk that investors associate with the expected after-tax cash proceeds.