D. Paul Newman

The Demand for Accounting Conservatism for Management Control

We show that conservative financial reporting arises naturallyin principal-agent settings as a means of efficiently motivating agentswhen the penalties that can be imposed on agents are limited. Weconsider an accounting system whose reports are used for contracting andwhose parameters are controlled by the principal. One advantage of our model is that the information system we describe has the accountingcharacteristic of mapping unbiased underlying information about the firminto a reduced message space. The principal can choose how that mappingoperates, i.e., conservatively, liberally, or neutrally. When penaltiesare sufficiently limited (a limited liability setting), we show that theaccounting system designed by the principal is always conservative. Alternatively, in an unlimited liability setting, any bias in the systemdepends on random circumstances, and we would not expect accountingconservatism to arise as a pervasive and enduring phenomenon.

Contracts without memory in multiperiod agency models

Abstract Settings are considered in which optimal multiperiod contracts can have no memory, i.e., where second period payments do not need to depend on first period outcomes. If contracts have no memory, a repeated agency game can be played myopically; there are no gains to long-term relationships. Conditions on preferences for a no memory contract are presented. In an agency game with moral hazard on the act selection, preference separability and domain additivity imply the existence of a no memory contract. In a setting without moral hazard but with asymmetric information on the outcome, domain additivity implies no memory.

Auditor Liability and Client Acceptance Decisions

The accounting profession has raised concerns that excessive liability exposure renders audit firms unwilling to provide audit services to risky clients, limiting the prospective clients’ ability to raise external capital. In this paper we address this concern in a model where the auditor evaluates the riskiness of the client before accepting the client engagement. We consider a setting where a shift to stricter legal liability regimes not only increases the expected damage payments from the auditor to investors in case of audit failure but also increases litigation frictions such as attorneys’ fees. The main finding is that the relationship between the strictness of the legal regime and the probability of client rejection is U-shaped. Our model therefore suggests that in environments with moderate legal liability regimes the client rejection rate is lower than in environments with relatively strong or relatively weak legal regimes. The analysis also generates empirical predictions relating the client’s riskiness to equilibrium audit quality, rejection rates, and audit fees. ∗We wish to thank Bill Kinney and Evelyn Patterson for their very helpful comments. †University of Texas at Austin, Department of Accounting, volker.laux@mccombs.utexas.edu ‡University of Texas at Austin, Department of Accounting, paul.newman@mccombs.utexas.edu.

Ex Ante Randomization in Agency Models

We consider how a principal can use randomized strategies in designing optimal contracts in agency settings. We distinguish between ex post randomization (over fee schedules following act selection by the agent) and ex ante randomization (over fee schedules before act selection). We show that ex ante randomization may be efficient in both full information and private information settings under certain assumptions regarding preferences and the production technology. It is particularly significant that additive separability and concavity of the agents utility function as well as concavity of the production function do not rule out efficient ex ante randomized contracts in private information settings, although, as is known, they rule out ex post randomization.

Prospect theory: Implications for information evaluation

Abstract The foundation of formal models of information evaluation is expected utility theory. Although generally accepted as a normative model, expected utility theory is an inadequate descriptor of behavior in many settings, leading to the development of prospect theory to o explain various “paradoxes” which yield empirical violations of expected utility maximization. The implications of this theory are explored in a simple example, contrasting the effects of expected utility and prospect theory on choices between alternative information systems. Since accounting information can be ranked differently by prospect theory vis-a-vis expected utility theory, academicians, practitioners, and policymakers, in their roles as information evaluators and decision makers, are particularly affected by this alternative decision model.

Costly Public Information: Optimality and Comparative Statics

The value of public information in markets has been analyzed in a variety of settings. Ohlson and Buckman [1981] provide an overview and synthesis, demonstrating that the efficiency criterion satisfied by an information structure is crucially dependent on the market regime, available securities, and endowments. For example, if securities span the state space, (public) information is redundant. Similarly, the Pareto dominance of information structures (over, for instance, null information) is inextricably related to endowments, the number of trading points, homogeneity of market participants, and other variables.