Knowledge that Transforms

We examine the relation between auditor characteristics (quality and tenure) and the cost of debt financing. Consistent with the hypothesis that audit characteristics are important to the capital markets, we find that (1) auditor quality and tenure are negatively and significantly related to the cost of debt financing, (2) the relation between auditor characteristics and the cost of debt is most pronounced in firms with debt that is noninvestment grade, and (3) both the insurance and information role of audits are economically significant to the cost of debt. Overall, our results suggest that, through their dual roles of providing information and insurance, auditor quality and tenure matter to capital market participants.

This paper examines the relation between accounting choice and U.S. institutional investor ownership in non‐U.S. firms. We predict that U.S. investors exhibit home bias in their preference for accounting methods conforming to U.S. Generally Accepted Accounting Principles (GAAP) because such methods are more familiar, reduce information processing costs, and are perceived as higher quality. We find that firms exhibiting higher levels (changes) of U.S. GAAP conformity have greater levels (changes) of U.S. institutional ownership. Lead‐lag regressions suggest that increases in U.S. GAAP conformity precede increases in U.S. investment, but changes in U.S. institutional holdings do not precede changes in accounting methods. We also find that the positive relation between U.S. GAAP conformity and U.S. investment holds regardless of a firm's visibility to U.S. investors (e.g., American Depositary Receipt listing, stock index membership, analyst following, firm size). However, we find that U.S. GAAP conformity has a significantly greater impact among firms already visible to U.S. investors.

We analyze the optimal behavior of an organization when its employees can manipulate the organization's accounting system to their private advantage. We find that the organization may benefit by helping its employees manipulate the system. This help can reduce the employees' private returns from devoting effort to further manipulation of the accounting system, which reduces the cost of motivating the employees to devote their effort to improving the real (rather than the measured) performance of the organization.

In this paper we study when it is advantageous to improve corporate transparency by allowing shareholders direct access to corporate information and when it is preferable to rely on a reporting system in which shareholders only gain access to information that management chooses to disclose. We show that in an agency model that allows for contract renegotiation, the desirability of a fully transparent reporting regime hinges on the stewardship properties of the information in question. Specifically, information that is mainly useful for predicting future events and of little use for evaluating past actions should only be made available to the public through management's self‐interested disclosures. Only if the information is useful for making inference about managerial actions can it be optimal to have full corporate transparency, so that outsiders have independent access to the same information as management.

We examine whether transient institutional investors (i.e., institutions that trade actively to maximize short‐term profits) have information that allows them to predict a break in a string of consecutive quarterly earnings increases and thereby avoid the economically significant negative stock price response associated with the break announcement. We show that transient institutions predict the break at least one quarter in advance of the break quarter. We also provide evidence that is consistent with transient institutions obtaining information regarding the impending break from private communications with management.

Do fees for non‐audit services compromise auditor's independence and result in reduced quality of financial reporting? The Sarbanes‐Oxley Act of 2002 presumes that some fees do and bans these services for audit clients. Also, some registrants voluntarily restrict their audit firms from providing legally permitted non‐audit services. Assuming that restatements of previously issued financial statements reflect low‐quality financial reporting, we investigate detailed fees for restating registrants for 1995 to 2000 and for similar nonrestating registrants. We do not find a statistically significant positive association between fees for either financial information systems design and implementation or internal audit services and restatements, but we do find some such association for unspecified non‐audit services and restatements. We find a significant negative association between tax services fees and restatements, consistent with net benefits from acquiring tax services from a registrant's audit firm. The significant associations are driven primarily by larger registrants.

We investigate factors associated with firms' decisions in 2002 and early 2003 to recognize stock‐based compensation expense under Statement of Financial Accounting Standards (SFAS) No. 123. We find that the likelihood of SFAS 123 expense recognition is significantly related to the extent of the firm's participation in capital markets, the private incentives of top management and members of the board of directors, the level of information asymmetry, and political costs. Although recognizing firms have significantly smaller SFAS 123 expense, we find no significant incremental relation between recognition likelihood and SFAS 123 expense magnitude after controlling for other factors that we expect explain the recognition decision. We also find positive and significant announcement returns for earlier announcing firms, particularly those stating that increased earnings transparency motivates their decision.

We examine the effect of higher order beliefs on the ability of decentralized decision makers to coordinate and take advantage of improvements in information transparency that can increase welfare. Theories that address this question have not been empirically explored. We study coordination in a laboratory experiment with privately informed decision makers. Economic outcomes in the setting depend both on agents' rational beliefs regarding economic fundamentals and on their rational beliefs regarding the beliefs of other agents. Increasing information transparency mitigates uncertainty about economic fundamentals but may increase strategic uncertainty, precipitating multiple equilibria and less efficient group outcomes. We provide evidence that sometimes the equilibrium attained by creditors is inferior from a welfare perspective to other available equilibria. Risk dominance appears to determine equilibrium selection in our setting.

We investigate whether a firm's intangible investments should be measured and separated from operating expenses. We find that the information extracted from accounting reports of investments and earnings is different when intangibles are measured and identified separately from operating expenses than when intangibles are left commingled with operating expenses. This difference in the market's information causes a change in the behavior of market prices, inducing changes in the firm's investments and cash flows. Thus, from a real effects perspective, measuring intangibles is not unambiguously desirable. We identify the conditions under which providing information on intangibles may be desirable. This study also shows the inadequacy of statistical associations between accounting numbers and prices as a basis for evaluating the desirability of measuring intangible investments. We show that the measurement of intangibles alters the very distribution of cash flows about which the measurement regime is seeking to provide information.

We examine client acceptance and client continuance decisions of a large audit firm to provide empirical evidence on the extent and nature of risk avoidance that the firm uses to purposefully manage its client portfolio. Our results support several key new inferences regarding audit firm portfolio management decisions. First, the results show that this firm is shedding the riskier clients in its portfolio, consistent with the risk avoidance theory of audit firm portfolio management. Second, the results show that the firm's newly accepted clients are less risky than its continuing clients. Although results of both the client continuance and client acceptance decisions imply a less risky portfolio emerging over time, there are greater differences in risk between continuing and discontinued clients than between continuing and newly accepted clients. Third, we find that audit risk factors are more important in audit firm portfolio management decisions than are financial risk factors. Finally, we find no evidence that audit pricing affects the client acceptance and continuance decisions of this firm, controlling for risk and other client characteristics.