社交媒体信息披露如何影响审计师选择？——基于公司新浪微博的经验证据

作为互联网信息技术下的代表性产物，社交媒体的涌现对市场参与者的经济行为产生了重要影响。从审计师选择视角，研究公司微博信息披露如何通过影响市场参与者进而作用于管理层的决策行为。研究发现，微博信息披露越多的公司越倾向于选择高质量的审计师，并且微博信息披露主要通过外部关注压力（分析师、媒体）和内部监督压力（董事会、审计委员会）两种渠道促使管理层选择高质量的审计师；异质性分析发现，对于投资者信息需求更高、内部控制水平更高、市场竞争更激烈的公司而言，微博信息披露对审计师选择的影响更为明显。该研究结论有助于投资者和监管方更好地理解社交媒体信息的作用机制，对进一步完善信息披露制度促进资本市场健康发展具有一定的启示。     

监测加密数字货币操纵行为的机器学习模型

加密数字货币匿名性、去中心化、跨区域、全球“7*24”小时交易、监管缺乏等特性促使其市场操纵行为频发。同时，匿名网络社交平台和自动化交易机器人的盛行使得抬价出货操纵行为公开化和短暂化，传统监管手段已无法适用于加密数字货币市场。首先，采用无监督学习模型对加密数字货币交易所分钟级的价量数据进行建模，对异常的量价数据进行快速的监测和预警。然后通过网络实时爬虫技术获取网络社交平台中抬价出货操纵行为的文本信息，与交易所秒级订单数据匹配构建训练集和测试集，采用有监督学习模型对抬价出货操纵行为进行事后识别。模型结果显示：无监督学习模型监测抬价出货的准确率高达84.07%。有监督学习模型在测试集上的精准率和召回率分别为82%和93%，且有监督学习模型的AUC曲线（Area Under Roc Curve）得分为0.83。两种模型均为加密数字货币交易所监管抬价出货行为提供了相关参考。     

企业避税与年报可读性

避税活动加剧了企业的信息不对称，管理层在信息披露中是否会采用晦涩的文本信息掩盖避税行为？本文利用2008—2017年中国A股上市公司数据，考察企业避税对年报可读性的影响及其机制。结果发现，避税活动越多，企业年报采用的复杂词汇就越多，年报可读性就越差。在运用工具变量弱化内生性问题、更替年报可读性指标与企业避税指标、考虑递延所得税信息披露和税收政策影响等一系列稳健性检验后，避税行为降低企业年报可读性的结论依然成立。机制分析发现会计信息质量在避税行为对年报可读性的影响中发挥了部分中介作用，避税活动通过降低会计信息质量削弱了年报可读性，信息披露中文本信息与数字信息相配合的观点从企业避税视角得到了验证。此外，在无税收优惠、递延所得税负债较多和外部治理环境较差的企业中，避税降低年报可读性的现象更为明显。因此，规范企业税收制度能够减少企业避税、限制管理层寻租行为，促进税收透明化，从而提高企业的信息披露质量。     

Economists in the 2008 financial crisis: Slow to see,fast to act

We study the economics- and finance-scholars’ reaction to the 2008 financial crisis using machine learning language analyses methods of Latent Dirichlet Allocation and dynamic topic modelling algorithms, to analyze the texts of 14,270 NBER working papers covering the 1999–2016 period. We find that academic scholars as a group were insufficiently engaged in crises’ studies before 2008. As the crisis unraveled, however, they switched their focus to studying the crisis, its causes, and consequences. Thus, the scholars were “slow-to-see,” but they were “fast-to-act.” Their initial response to the ongoing Covid-19 crisis is consistent with these conclusions.     

Do enhanced derivative disclosures work? An informational perspective

Firms use derivatives both for hedging and nonhedging purposes. The Statement of Financial Accounting Standards No. 161 (SFAS 161) requires firms to disclose the purposes of their derivatives usage, thereby helping investors to evaluate the effects of derivatives usage on firm performance. Using a hand-collected sample of US listed firms and a difference-in-differences research design, we find that, compared with nonderivative-users, derivative-users compliant with SFAS 161 experience a significantly greater reduction in stock illiquidity and the probability of informed trading in the post-SFAS 161 period, and such impact is evident only for firms with a high degree of investor attention.     

Do Jobseekers Value Diversity Information? Evidence from a Field Experiment and Human Capital Disclosures

We examine how information about the diversity of a potential employer's workforce affects individuals’ job-seeking behavior. We embed a field experiment in job recommendation emails from a leading career advice agency in the United States. The experimental treatment involves highlighting a diversity metric to jobseekers. Our results indicate that disclosing diversity scores in job postings leads jobseekers to click on firms with higher diversity scores, with such effects varying across jobseeker demographics. A follow-up survey provides evidence on potential explanations for why jobseekers value diversity information. We then examine how jobseekers’ preferences for diversity relate to disclosure choices under the U.S. SEC Human Capital Disclosure requirement. We find that firms in industries characterized by higher jobseeker responsiveness to diversity information tend to voluntarily disclose diversity metrics in their 10-Ks under these new disclosure requirements.     

ESG disclosure and technological innovation capabilities of the Chinese listed companies

The main objective of this study is to explore how ESG disclosure effectively promotes technological innovation capabilities (TIC) and also in different industries (green vs. high-tech). Further, examine the role of financing constraint (FC) in the relationship between the ESG disclosure and TIC. We employed the panel regression model, Causal step approach, Bootstrap mediation effect test, 2SLS, and GMM model. We used Bloomberg’s ESG disclosure score of China’s A-share listed companies from 2011 to 2019 (1); we found that the ESG disclosure has a significant relationship with corporate innovation indicators (OTI, STI, NSTI) and play a significant role in promoting TIC at different levels of corporate innovation (2) ESG disclosure of non-green (high-tech) industry is more effectively promote TIC than green (non-high tech) industry (3) ESG disclosure can promote corporate innovation by reducing the level of corporate financing constraints, and FC has a partial intermediary role between ESG and TIC.     

Feasibility analysis of machine learning for performance-related attributional statements

We investigate the feasibility of machine learning methods for attributional content and framing analysis in corporate reporting. We test the performance of five widely-used supervised machine learning classifiers (naïve Bayes, logistic regression, support vector machines, random forests, decision trees) in a top-down three-level hierarchical setting to (1) identify performance-related statements; (2) detect attributions in these; and (3) classify the content of the attributional statements. The training set comprises manually coded statements from a corpus of management commentary reports of listed companies. The attributions include both intra- and inter-sentential attributional statements. The results show that for both intra- and inter-sentential attributions, F1-scores of our most accurate classifier (i.e., support vector machines) vary in the range of 76% up to 94%, depending on the identification, detection and classification levels and the content characteristics of attributions. Additionally, we assess the hierarchical performance of classifiers, providing insights into a more holistic classification process for attributional statements. Overall, our results show how machine learning methods may facilitate narrative disclosure analysis by providing a more efficient way to detect and classify performance-related attributional statements. Our findings contribute to the accounting and management literature by providing a basis for implementing machine learning methodologies for research investigating attributional behavior and related impression management.