ABSTRACT As corporate disclosures increasingly report figures in the millions, billions, and trillions, large number innumeracy may systematically distort how investors interpret firm performance. This cognitive limitation arises when individuals compress large magnitudes on the mental number line, overestimating absolute values while underestimating differences between them. We introduce large number innumeracy as a distinct perceptual construct in the accounting literature and synthesize cognitive science and accounting research to explain how it may affect investor interpretation of corporate disclosures and why susceptibility likely varies across investor types. We develop a mitigation framework organized by how directly each mechanism addresses the problem: direct approaches alter how values are represented, contextual supports aid interpretation without changing magnitude representation, and interpretive intermediaries shift processing to outside actors or automated tools. We conclude with a research agenda on prevalence, capital market consequences, and mitigation, with implications for disclosure design, investor protection, and market efficiency. Data Availability: Data are available from the public sources cited in the text. JEL Classifications: M41; G41; D91; D83.
ABSTRACT This paper considers two related earnings/capital management opportunities available to banks as they transitioned to ASU 2016-13. Under the FASB’s transition guidance, banks estimate credit losses under the new standard and take the initial adoption adjustment (IAA) directly into retained earnings. Shifting pre-adoption provision accruals to their IAA buckets enables banks to increase their reported income in the period before adoption while avoiding regulatory scrutiny, since their allowances are increased promptly. Consistent with this hypothesized behavior, we find that banks making timely adoptions of ASU 2016-13 on 1/1/2020 reported more negative abnormal provision accruals in 2019 than nonadopting banks. Negative abnormal provision accruals in 2019 were associated with larger IAAs in Q1 of 2020. We also formulate and test competing hypotheses about banks’ decisions to delay adoption under the CARES Act (U.S. Congress 2020). We find that adoption delays are associated with size-related adoption challenges rather than earnings management. Data Availability: Data are available from the public sources cited in the text. JEL Classifications: M42; G21.
ABSTRACT A large political cost literature suggests that managers alter disclosures to avoid political costs. This inference crucially relies on the assumption that managers’ disclosure choices can influence realized political costs, yet this assumption is untested. To address this gap, I investigate whether increases in mandatory disclosures affect realized political costs. I identify exogenous increases in firm disclosures using the staggered promulgation of 14 FASB standards and examine their effects on a highly general measure of realized political costs based on restrictions imposed on firms by the U.S. government’s 375 departments and agencies. I find that restrictions increase following FASB standard promulgations, providing new evidence that disclosures influence political costs. This evidence supports a crucial assumption underlying the political cost literature and, more broadly, demonstrates that accounting disclosures designed to inform investors can generate economically consequential information externalities through their use by noninvestor stakeholders. Data Availability: Data are available from the sources identified in the text. JEL Classifications: F52; G34; G38; K22; M41.
This study examines the impact of firms' environmental, social, and governance (ESG) disclosures on the pricing and nonpricing terms of banks' debt contracts. Using ESG disclosure scores and bank debt contract data for S&P 1500 firms between 2009 and 2021, we find strong evidence that banks impose more covenants on firms with lower ESG disclosure scores, particularly those with higher informational uncertainty, and weaker evidence that banks charge higher interest rates. Banks include restrictions on excess cash flow sweeps, asset sales sweeps, and collateral requirement general covenants when ESG disclosures are low. ESG disclosures remain a significant determinant of covenants after controlling for ESG performance and accounting for simultaneity between pricing and nonpricing terms of debt contracts. ESG disclosures provide banks with incremental information on firms' default risk beyond what is available from financial data. Our findings suggest that increasing ESG disclosures can yield favorable debt contract terms for firms.
Accounting standards exclude most securities gains and losses from net income until the securities are sold, providing incentives to sell securities based on the gain/loss positions. We revisit prior literature and deepen understanding about this behavior among banks in a variety of ways. First, we find that what the prior literature calls earnings "smoothing" is more precisely characterized as boosting low earnings; banks boost low earnings via gain selling but do not materially reduce high earnings via loss selling. Second, we find this behavior more aligns with opportunism than with signaling, and a specific opportunistic motive is to meet the regulatory guideline for dividend payments. Finally, we uncover additional tendencies of banks, which include selling larger portions of gain positions than loss positions, more aggressively using gain selling to offset a given amount of loss selling than vice versa, and selling securities in a pattern that conforms to prospect theory.
