Knowledge that Transforms

The pricing of Big 5 industry leadership in the U.S. audit market is investigated using audit fee disclosures for the 2000–2001 fiscal years and the joint nationalcity framework in Ferguson et al. (2003). There is a significant fee premium of 19 percent on those engagements where Big 5 auditors are both the nationally top-ranked auditor and the city-level industry leader in the city where the client is headquartered, indicating that national and city-specific industry leadership jointly affect auditor reputation and pricing. However, there is never a premium in any tests for auditors that are national industry leaders alone without also being city-specific industry leaders, indicating that national leadership by itself does not result in a premium. The evidence is mixed with respect to city-specific industry leaders alone that are not also national industry leaders. While there is a premium of 8 percent in the primary tests, this result is inconclusive as it does not hold in all sensitivity analyses.

Schrand and Walther's (2000) archival evidence suggests that managers strategically disclose prior-period benchmarks in current earnings announcements, which, in turn, influences investors' judgments. Using a controlled experimental setting, I present evidence confirming that a transparent description of a transitory prior-period gain or loss affects how investors apply prior-period earnings when evaluating currentperiod earnings. I also provide evidence that this effect is likely to be unintentional on the part of investors, resulting from limitations in their memory for the prior-period event. Overall, the experimental results suggest that a quantitative description of the transitory prior-period gain or loss in a current earnings announcement helps investors to evaluate company performance. The results also highlight the need for consistency in reporting non-GAAP financial performance measures.

In a market experiment, buyers respond to sellers' disclosures of one of the two potential outcomes of a risky prospect. Buyer reactions reflect two separate disclosure phenomena that heretofore have not been considered jointly. First, consistent with cognitive predictions that users tend to anchor on explicit one-sided disclosures, buyers pay more relative to expected value when the seller discloses the higher of the two potential outcomes than when the seller discloses the lower potential outcome. Second, consistent with the incentive-driven predictions of information economics, buyers systematically discount bids when sellers choose which potential outcome they wish to disclose, relative to bids in a control condition in which the seller is constrained to disclose randomly. Both findings are robust to each other, addressing qualifications in prior research about generalizing cognitive disclosure phenomena to a strategic disclosure environment. One implication supported by the data is that if cognitive information processing limitations are robust to a strategic environment, then strategic agents can exploit these limitations when choosing disclosures.

We develop an empirical method that allows us to evaluate the reliability of an expected return proxy via its association with realized returns even if realized returns are biased and noisy measures of expected returns. We use our approach to examine seven accounting-based proxies that are imputed from prices and contemporaneous analysts' earnings forecasts. Our results suggest that, for the entire crosssection of firms, these proxies are unreliable. None of them has a positive association with realized returns, even after controlling for the bias and noise in realized returns attributable to contemporaneous information surprises. Moreover, the simplest proxy, which is based on the least reasonable assumptions, contains no more measurement error than the remaining proxies. These results remain even after we attempt to purge the proxies of their measurement error via the use of instrumental variables and grouping. We provide additional evidence, however, that demonstrates that some proxies are reliable when the consensus long-term growth forecasts are low and/or when analysts' forecast accuracy is high.

This paper proposes and tests a risk model that explains how investors perceive financial risks. The model combines conventional decision-theory variables—probabilities and outcomes—with behavioral variables from psychology research by Slovic (1987), such as the extent to which a risky item is new, causes worry, and is controllable. To test our model, we conduct two studies in which M.B.A. students judge the risk of a broad range of financial items. Our results indicate that both the decisiontheory variables and Slovic's (1987) behavioral variables are important in explaining investors' risk judgments. Further, we demonstrate that information about the amount of potential loss outcome contained within mandated risk disclosures not only directly influences risk judgments, but also indirectly affects such judgments via its effect on some of Slovic's (1987) behavioral variables. By identifying this unintended consequence of current risk disclosures, these results have the potential to influence the way accounting regulators, firm managers, and academic researchers think about risk disclosure.

The SEC issued Staff Accounting Bulletin (SAB) No. 101 to address its concern that firms were masking true performance by managing earnings using accelerated revenue recognition. Critics of this Accounting Bulletin stated that it would eliminate industry-accepted revenue recognition practices and reduce the quality of reported earnings. The FASB's revenue recognition discussions echo these concerns stating that revenues recorded prior to the completion of the earnings process contain value-relevant information about future performance. This paper investigates these two hypotheses using a sample of firms that accelerated revenue recognition prior to SAB No. 101 adoption (SAB 101 firms) and a matched set of firms that were unaffected by this regulation (unaffected firms). Our earnings distribution tests indicate that SAB 101 firms are more likely to meet earnings benchmarks. Specifically, we find that, in the pre-adoption period, SAB 101 firms report fewer small negative and more small positive earnings than they do in the post-adoption period and than do unaffected firms in the pre-adoption period; SAB 101 firms report fewer small negative and more small positive earnings changes in the pre-adoption period compared to the post-adoption period. We also document that SAB 101 firms are more likely to have weaker corporate governance and more likely to have financial covenants, providing them with greater incentives to manage earnings. However, we find that the association between earnings and future cash flows and between unexpected earnings and earnings announcement period returns were higher for SAB 101 firms than for unaffected firms in the preadoption period, indicating higher earnings informativeness for SAB 101 firms. These associations declined for SAB 101 firms in the post-adoption period, suggesting that SAB No. 101 caused a decline in earnings informativeness. Overall, our results suggest that, although the revenue recognition practices targeted by SAB No. 101 have been used by some firms to manage earnings, the regulation's prohibition of revenue recognition prior to completion of the earnings process, on average, results in less informative earnings since these unearned revenues provide value-relevant information.

We examine how the market's ability to assess the truthfulness of management earnings forecasts affects how managers bias their forecasts, and we evaluate whether the market's response to management forecasts is consistent with it identifying predictable forecast bias. We find managers' willingness to misrepresent their forwardlooking information as a function of their incentives varies with the market's ability to detect misrepresentation. We examine incentives induced by the litigation environment, insider trading activities, firm financial distress, and industry concentration. With regard to the stock price response to forecasts, we find the market varies its response with the predictable bias in the forecast. The efficiency of the market's response, however, varies with the forecast news.

Applying a Burgstahler and Dichev (1997)/Degeorge et al. (1999) type methodology to quarterly data for the 1985–2002 time period, we show that, since the mid-1990s, but not before then, managers seek to avoid negative quarterly earnings surprises more than to avoid either quarterly losses or quarterly earnings decreases. Our findings suggest that the quarterly earnings threshold hierarchy proposed by Degeorge et al. (1999) does not apply to recent years, and that managers' claim that avoiding quarterly earnings decreases is the threshold they most seek to achieve (Graham et al. 2004) is inconsistent with their actions. We provide an intuitively appealing economic rationale for why the shift in threshold hierarchy occurred; since the mid-1990s, but not before then, investors unambiguously rewarded (penalized) firms for reporting quarterly earnings meeting (missing) analysts' estimates more than they did for meeting (missing) the other two thresholds. We provide several explanations for why investors unambiguously reward firms for reporting quarterly earnings that meet or beat analysts' estimates more than for meeting the other two thresholds late (but not early) in our sample period: increased media coverage given to analyst forecasts, more analyst following, more firms covered by analysts, and temporal increases in both the accuracy and precision of analyst forecasts.

Earnings press releases provide managers a forum to present their firm's quarterly financial information and perhaps influence perceptions of the firm's stakeholders. We explore the use of managerial emphasis as a disclosure tool and contribute to the debate over pro forma earnings. We examine (1) the determinants of emphasis placed on pro forma and GAAP earnings within quarterly earnings press releases, (2) whether there has been a shift away from emphasizing pro forma earnings toward GAAP earnings, and (3) whether stock market reactions to earnings news were influenced by emphasis placed on metrics within the press release. We find that firms emphasize metrics that are more value-relevant and portray more favorable firm performance. We also find that the extent of a firm's media coverage affects managers' emphasis decisions. Further, our results indicate a highly significant shift toward GAAP emphasis and away from pro forma emphasis in 2002 relative to 2001. Finally, our stock market tests suggest that greater emphasis on an earnings metric results in a stronger market reaction to the surprise in that metric. Overall, these findings are consistent with managers using emphasis in the earnings press release as a disclosure tool and this emphasis influencing at least one important stakeholder group—investors.

This paper examines the extent to which management makes accounting choices to record intangible assets based on their insights into the underlying economics of their firm. It exploits a setting in which management has accounting discretion to record a wide range of intangible assets. The results suggest that management's choice to record intangible assets is associated with the strength of the technology affecting the firms operations, the length of the technology cycle time, and propertyrights-related factors that affect the firm's ability to appropriate the investment benefits. These effects are more important than other contracting and signaling factors consistent with the underlying economics operating as a first-order effect as envisaged by GAAP. The results also indicate that the intangible assets management has a voluntary (unregulated) choice to record—identifiable intangible assets—are more highly correlated with underlying economic factors than the regulated classes, purchased goodwill and R&D assets. This result suggests that limiting managements' choices to record intangible assets tends to reduce, rather than improve, the quality of the balance sheet and investors' information set.