Knowledge that Transforms

Public accounting firms are owned by all equity partners, but day-to-day management is generally delegated to a team of leaders. Using data from China, this study examines which equity partners are selected to the firm's national leadership team, and whether firmwide audit quality is related to leadership attributes. We find that a partner is more likely to be selected as a leader if the partner is more experienced in public company auditing and has attracted more new clients to the firm. Firmwide audit quality is higher when leaders are more experienced in public company audits or have a past record of larger audit adjustments, and is lower when leaders have attracted more high-risk new clients to the firm. Leadership attributes exhibit a relatively strong (weak) association with audit quality at the headquarters (branch offices). Moreover, audit quality is higher when a firm has more leaders in an audit-quality role.

We investigate the effect of tick size, a key feature of market microstructure, on managerial learning from stock prices. Using a randomized controlled tick-size experiment, the 2016 Tick Size Pilot Program, we find that a larger tick size increases a firm's investment sensitivity to stock prices, suggesting that managers glean more new information from stock prices to guide their investment decisions as the tick size increases. Consistently, we also find that changes in managerial beliefs, as reflected in adjustments of forecasted capital expenditures, respond more strongly to market feedback under a larger tick size. Additional evidence suggests the following mechanism through which tick size affects managerial learning: a larger tick size reduces algorithmic trading, in turn encouraging fundamental information acquisition. Increased fundamental information acquisition generates incremental information about growth opportunities, macroeconomic factors, and industry factors, with respect to which the market has a comparative information advantage over management.

This study investigates whether tax-based proprietary costs associated with being a public firm (i.e., costs resulting from increased visibility to the tax authority) discourage public listing. I exploit the introduction of a mandatory disclosure requirement (FIN 48) which generated a signal to the government regarding the uncertainty of public firms’ tax positions, allowing for more carefully targeted audits. I hypothesize and find evidence of an increased propensity to go private by public, tax aggressive firms following the enactment of the disclosure rule but prior to its adoption. Cross-sectionally, the effect is stronger among firms that are more sensitive to tax-based proprietary costs. Moreover, IPOs by tax aggressive firms exhibit a relative decline after FIN 48, consistent with the disclosure requirement deterring private, tax aggressive firms from going public. Overall, my findings suggest that mandatory disclosure rules imposing tax-based proprietary costs may discourage some firms from operating as public entities.

While Sani, Shroff and White (2023) examine a plausibly exogenous shock to the information acquisition landscape that arguably changes informed traders’ incentives to generate private information without the focal firm changing its disclosure policy, the causal chain underlying their analyses shares the implicit assumptions employed in the extant empirical literature that tests the Learning Hypothesis. Thus, the focus of our discussion is to make explicit the implicit assumptions and highlight the resulting challenges faced by authors who wish to interpret Investment Sensitivity to Price as evidence of managerial learning. In the end, we suggest future research work to extend our understanding by exploring in greater detail and more directly the What and the How managers learn from outsiders.

We examine the effect of a large dividend tax cut on corporate investment efficiency by exploiting the 2003 personal taxation reform in the U.S. as a quasi-natural experiment. Using a difference-in-differences approach based on the probability that a firm’s marginal investor was an individual investor, we show that the 2003 dividend tax cut significantly improved the investment efficiency of U.S. listed firms. However, we find no evidence that the dividend tax cut increased the level of investment of U.S. listed firms. Further, we show that the tax cut increased investment efficiency by mitigating agency problems associated with the excessive free cash flows of overinvesting firms and by relaxing the financial constraints of underinvesting firms.

SEC-mandated, machine-readable structured filings are an alternative source to Compustat for companies' accounting data. Discrepancies between as-filed and Compustat data, potentially a result of Compustat's standardizations, are more pronounced for firms with complex financial reporting. We show that these data discrepancies affect inferences in four research settings: (i) properties of accrual accounting, including accruals-cash flow relationships and abnormal accruals; (ii) real earnings management; (iii) the existence and magnitude of six of 21 accounting-based anomalies examined, including the accruals anomaly; and (iv) disclosure quality assessments based on the hierarchical structure of financial statement items. FactSet data also exhibit significant and often larger discrepancies from as-filed data. Our findings demonstrate the importance of these data discrepancies for the interpretation of empirical tests.

We study the roles of the head office (HO) and the business units (BUs) of a multinational corporation (MNC) in reducing income tax and tariff payments through internal transfer prices in international trades. Using confidential transfer price data of a large MNC, we analyze how the different elements of internal transfer prices set by the HO and BUs vary differently from external prices with income tax rates, tariff rates, and the tradeoff between the two. Absent severe agency conflicts, we find that the BUs contribute more to tax planning than the HO, despite that explicit incentives to do so are not included in the compensation schemes. The roles of the HO and BUs vary with product market competition, the risk of conflicts with tax and customs authorities, and agency problems within the firm. Moreover, we provide evidence of strategic trade cost allocations among BUs to reduce income taxes.

This paper examines whether mandatory disclosure affects the extent to which firms learn from external market participants. Conventional wisdom suggests that mandatory disclosure should increase the total amount of information in financial markets. However, disclosure can also reduce investors' incentives to acquire and produce information. Using the JOBS Act to identify variations in disclosure requirements, this paper finds that firms with reduced disclosure requirements attract more informed investors and learn more from financial markets than those with stricter disclosure requirements. This learning is concentrated among firms that attract sophisticated investors, particularly those with industry expertise, and weakens once firms are forced to disclose more information. Overall, the results suggest that one benefit from regulators’ recent efforts to reduce U.S. firm disclosure requirements is an increase in firm learning.

We examine how the partisan alignment between chief executive officers (CEOs) and the United States president influences corporate disclosure outcomes. We predict and find evidence that more partisan-aligned CEOs display greater optimism in their corporate disclosures. CEO partisan alignment is positively associated with the likelihood of issuing a management earnings forecast and issuing overly optimistic forecasts. More partisan-aligned CEOs also use a more optimistic tone in the corporate disclosure and exhibit a lower level of accounting conservatism. Our findings suggest that time-varying beliefs of CEOs formed by the political environment can shape corporate disclosures.

We examine the relation between plant-level predictive analytics use and centralization of authority for more than 25,000 manufacturing plants using proprietary US Census data. We focus on headquarters' authority over plants through delegation of decision-making and design of performance-based incentives. We find that increased predictive analytics use is associated with reduced delegation of decision-rights to local managers, increased centralization of control over data gathering and reduced plant managerial payrolls. In terms of incentives, predictive analytics use is associated with more accurate targets and tighter linkages between rewards to workers (performance-based bonuses, promotions and firings) and measured performance. Overall, our findings suggest that predictive analytics use is associated with increased centralization of authority in headquarters.