封面来源:ChatGPT-4 DALL·E

Abstract: Growing AI readership (proxied for by machine downloads and ownership by AI-equipped investors) motivates firms to prepare filings friendlier to machine processing and to mitigate linguistic tones that are unfavorably perceived by algorithms. Loughran and McDonald (2011) and BERT available since 2018 serve as event studies supporting attribution of the decrease in the measured negative sentiment to increased machine readership. This relationship is stronger among firms with higher benefits to (e.g., external financing needs) or lower cost (e.g., litigation risk) of sentiment management. This is the first study exploring the feedback effect on corporate disclosure in response to technology.

摘要:随着人工智能(AI)读者数量的增加(以机器下载量和拥有AI设备的投资者为代表),企业逐渐调整其披露文件,使其更便于机器处理,同时降低算法不利解读的语言语调。通过Loughran and McDonald(2011)以及自2018年开始使用的BERT事件研究,证实了机器读者增加导致负面情绪减少的现象。这一关系在具有较高外部融资需求或较低情绪管理成本(如诉讼风险)的企业中表现得更为明显。这是首项探讨企业披露受技术反馈影响的研究。

引用:Cao, S., Jiang, W., Yang, B., & Zhang, A. L. (2023). How to talk when a machine is listening: Corporate disclosure in the age of AI. The Review of Financial Studies, 36(9), 3603-3642.

文献:PDF

JEL: D83, G14, G30

Introduction

The annual report (and other regulatory filings) is more than a legal requirement for public companies; it provides an opportunity to communicate financial health, promote the culture and brand, and engage with a full spectrum of stakeholders. How those readers process this wealth of information significantly affects their perception of and, hence, participation in the business. Increasingly, companies realize that the target audience of their mandatory and voluntary disclosures no longer solely consists of human analysts and investors. A substantial amount of buying and selling of shares is triggered by recommendations made by robots and algorithms that process information with machine learning tools and natural language processing kits.1 【年报(以及其他监管文件)不仅仅是上市公司的法律要求;它为企业提供了一个机会,用以传达财务健康状况、推广企业文化和品牌,并与广泛的利益相关者互动。这些读者如何处理这些丰富的信息,将显著影响他们对企业的认知,进而影响他们参与企业的活动。企业越来越意识到,其强制性和自愿性披露的目标受众不再仅限于人类分析师和投资者。大量的股票买卖交易是由使用机器学习工具和自然语言处理工具包处理信息的机器人和算法给出的建议所驱动的。】

年报和披露的多重功能(年报不仅仅是法律要求,它还是企业传达财务健康、推广品牌文化、与利益相关者沟通的机会)→→→读者的认知影响(读者如何处理这些信息,将直接影响他们对企业的认知和对企业的参与程度)→→→受众的变化(企业逐渐意识到,其披露文件的受众不再仅仅是人类分析师和投资者)→→→自动化交易的兴起(相当数量的股票买卖是基于机器人和算法处理信息后做出的建议,这些算法依赖于机器学习和自然语言处理技术)。

1. 例如,Gara (2018)提到领先的对冲基金The Man Group已开始使用人工智能和算法交易管理其大部分资产(Satariano and Kumar 2017)。

Both the technological progress and the sheer volume of disclosures make the trend inevitable: $\color{blue}{Cohen\ et\ al.\ (2020 JF)}$ document that the length of 10-Ks increases by five times from 2005 to 2017. Companies that wish to accomplish the desired outcome of communication and engagement with stakeholders need to adjust how they talk about their finances, brands, and forecasts in the age of AI. In other words, they should heed the unique logic and techniques underlying the rapidly evolving analysis of language and sentiment facilitated by large-scale machine-learning techniques, such as automated computational processes that identify positive, negative, and neutral opinions in a whole corpus of firm disclosures that is beyond the processing ability of human brains. While the literature is catching up with and guiding investors’ rising aptitude to apply machine learning and computational tools to extract qualitative information from disclosures and news, there has not been an analysis exploring the feedback effect: how companies adjust the way they talk knowing that machines are listening. This paper fills this void. 【技术进步和大量信息披露的趋势使这一趋势不可避免:Cohen、Malloy 和 Nguyen(2020)记录了10-K报告的篇幅从2005年到2017年增加了五倍。那些希望与利益相关者实现预期沟通效果的公司,必须在人工智能时代调整其财务、品牌和预测的表述方式。换句话说,企业应该关注大规模机器学习技术所促成的语言和情感分析的独特逻辑和技术,例如自动化计算过程,这些过程能够在超过人类大脑处理能力的企业披露语料库中,识别正面、负面和中立的意见。虽然现有文献正在逐步跟上并引导投资者运用机器学习和计算工具来提取披露和新闻中的定性信息,但还没有研究探索反馈效应:公司在意识到机器正在“倾听”时如何调整其表述方式。本研究填补了这一空白。】

技术进步和信息披露增加(由于技术进步和披露量的增加,企业披露文件的长度和复杂性显著提高,例如10-K报告的篇幅在短时间内增加了五倍)→→→沟通的必要调整(为了达到与利益相关者沟通的预期效果,企业需要在人工智能时代调整其表述方式,包括财务、品牌和预测的传达)→→→大规模机器学习的崛起(大规模机器学习技术迅速发展,能够通过自动化的计算过程,分析和识别公司披露中正面、负面和中立的情感和语言)→→→文献滞后于投资实践(尽管文献逐渐赶上了机器学习在披露定性信息提取中的应用,但缺乏对反馈效应的分析)→→→尚未有研究探讨企业如何调整其表述方式以适应机器读者的需求

Our analysis starts with a diagnostic test that connects how machine-friendly a company composes its disclosures (measured by Machine readability following $\color{blue}{Allee\ et\ al., (2018JAR)}$and the expected extent of machine readership for a company’s SEC filings on EDGAR, for which we develop multiple proxies. The first variable, Machine downloads, is constructed by tracking IP addresses that conduct downloads in large batches. Machine request is a precursor and a necessary condition for machine reading, and the sheer volume of machine-downloaded documents makes it unlikely for them to be processed by human readers alone. Because the SEC Log files used to construct Machine downloads became available to the public in 2015, our analyses implicitly assume that firms were aware of the extent of machine readership before the exact numbers of machine downloads became public. To relax the assumption, we also construct a measure based on share ownership by institutional investors with AI capabilities, AI ownership, tracked from their AI-related job postings. Finally, we proxy investor technology capacity by calculating the ownership-weighted AI talent supply available to institutional investors, based on the state-year-level proportion of the working-age population with IT degrees where the investors are headquartered. Because asset manager headquarters were mostly chosen before the AI era and bear no direct relation to portfolio firms, the last variable is likely to be orthogonal to omitted variables explaining Machine readability. 【我们的分析从一个诊断测试开始,该测试连接了公司披露文件的机器友好性(根据Allee、DeAngelis和Moon 2018的测量标准)与公司在EDGAR上提交的美国证券交易委员会(SEC)文件中预期的机器阅读量。我们为此开发了多个代理变量。第一个变量是机器下载量,通过追踪大量下载的IP地址来构建。机器请求是机器阅读的前提和必要条件,机器下载的文件数量庞大,这使得它们不可能仅仅由人类读者处理。由于构建机器下载量的SEC日志文件自2015年起向公众开放,我们的分析隐含假设公司在具体的机器下载数量公开之前已经意识到机器读者的规模。为了放松这一假设,我们还构建了一个基于具有AI能力的机构投资者持股比例的度量,即AI持股量,通过这些投资者的AI相关招聘信息进行跟踪。最后,我们通过计算机构投资者所拥有的技术能力,使用AI人才供应作为代理变量,这是根据投资者总部所在地在特定年份拥有IT学位的适龄劳动人口比例计算的。由于资产管理公司的总部大多在AI时代之前选定,并且与投资组合公司没有直接关系,最后一个变量很可能与解释机器可读性中的遗漏变量正交。】

机器友好性与机器读者的关系(分析的重点是公司披露文件的机器友好性和预期机器读者数量之间的联系)→→→机器下载量作为代理变量(通过跟踪大量下载的IP地址,机器下载量被用来衡量机器读者的存在及其影响)→→→机器下载是机器阅读的前提(机器请求是机器阅读的必要条件,机器下载的文件数量表明这些文件不可能仅由人类处理)→→→分析假设的放宽(为了解决公司在机器下载数据公开之前对机器读者规模的认知问题,研究使用AI持股量和AI人才供应作为额外的代理变量)→→→AI持股量与AI人才供应作为替代(AI持股量和AI人才供应用于测量机构投资者的技术能力和机器阅读对披露文件的影响,特别是与机器友好性相关的影响)

We show that, in the cross-section of filings with firm and year fixed effects, a one-standard-deviation change in expected machine downloads is associated with a 0.24-standard-deviation increase in the Machine readability of the filing. On the other hand, other (nonmachine) downloads do not bear a meaningful correlation with machine readability, validating Machine downloads as a proxy for machine readership. The alternative proxies AI ownership and AI talent supply bear similar economic and statistical significance. We further validate the economic mechanism underlying our main variables by showing that trades follow more quickly after a filing becomes public when Machine downloads is higher, with even stronger interactive effect with better Machine readability. Such a result demonstrates the real impact of machine processing on information dissemination.【我们表明,在公司和年度固定效应的横截面分析中,预期机器下载量变化一个标准差,与披露文件的机器可读性提高0.24个标准差相关。另一方面,其他(非机器)下载与机器可读性之间没有显著的相关性,这验证了机器下载量作为机器读者代理变量的有效性。替代变量AI持股量和AI人才供应具有相似的经济和统计显著性。我们进一步验证了主要变量背后的经济机制,显示当机器下载量较高时,在文件公开后交易速度加快,并且当机器可读性更高时,这一互动效应更加显著。该结果证明了机器处理对信息传播的实际影响。】

机器下载量的变化与机器可读性增加的关联(一个标准差的机器下载量变化与披露文件的机器可读性提高0.24个标准差相关)→→→非机器下载与机器可读性无关(其他非机器下载没有显著相关性,进一步验证了机器下载量作为机器读者代理变量的有效性)→→→AI持股量和AI人才供应的相似显著性(AI持股量和AI人才供应这两个替代变量在经济和统计上也具有相似的重要性)→→→验证经济机制(当机器下载量较高时,交易速度加快,并且当文件的机器可读性更好时,互动效应更加显著)→→→信息传播的实际影响(更高的机器下载量和更好的机器可读性共同加快了信息传播的速度,显示出机器处理对市场交易的实际影响)

After establishing a positive association between a higher AI reader base and machine-friendlier disclosure documents, we next explore how firms manage the “sentiment” and “tone” perceived by machines. It is well documented that corporate disclosures attempt to strike the right sentiment and tone with (human) readers without being explicitly dishonest or overtly noncompliant ($\color{blue}{Loughran\ and\ McDonald,\ 2011JF}$; $\color{blue}{Kothari\ et\ al.,\ 2009JAR}$ ). Hence, we expect a similar strategy catering to machine readers. While researchers and practitioners have long relied on the Harvard Psychosociological Dictionary (especially the Harvard-IV-4 TabNeg file) to count and contrast “positive” and “negative” words to construct “sentiment” as perceived by (mostly human) readers, the publication of $\color{blue}{Loughran\ and\ McDonald,\ (2011JF)}$ ( “LM” hereafter)presents an instrumental event to test our hypothesis pertaining to machine readers. This is not only because the paper presented a specialized finance dictionary of positive/negative words and words that are informative about prospects and uncertainty but also because the word lists that came with the paper have served as a leading lexicon for algorithms to sort out sentiments in both the industry and academia2. The differences in both the timeline and the context of the new dictionary allow us to trace out the impact of AI readership on sentiment management by corporations. 【在建立了更高的AI读者群体与更具机器友好性的披露文件之间的正相关关系后,我们接下来探讨公司如何管理机器感知到的“情感”和“语调”。已有文献表明,企业披露文件会尝试在与人类读者交流时保持适当的情感和语调,而不会明确欺骗或公然违规(Loughran and McDonald 2011;Kothari、Shu and Wysocki 2009)。因此,我们预计公司会采取类似的策略来迎合机器读者。尽管研究人员和实践者长期依赖哈佛心理社会学词典(尤其是Harvard-IV-4 TabNeg文件)来计算和对比“积极”和“消极”词汇,以构建人类读者感知的“情感”,Loughran和McDonald(2011,以下简称“LM”)的出版为测试我们关于机器读者的假设提供了一个重要事件。这不仅因为该论文提出了一个专门的财务词典,包含积极/消极词汇以及关于前景和不确定性的信息词汇,还因为该论文附带的词汇表已成为行业和学术界用以分类情感的主要词典。时间线和新词典的语境差异使我们能够追踪AI读者群体对企业情感管理的影响。】

机器友好性和AI读者基础的关联(确认AI读者数量增加与更具机器友好性的披露文件之间的正相关性后,接下来分析情感和语调的管理)→→→企业与人类读者的情感管理(企业在披露文件中会平衡适当的情感和语调,避免明显不诚实或违规行为)→→→同样适用于机器读者(预计公司会对机器读者采取类似的策略,以管理情感和语调)→→→LM词典的重要性(Loughran和McDonald(2011)的财务词典为测试关于机器读者的假设提供了关键事件,因其提出了专门用于财务文件的情感分类词汇)→→→AI读者对情感管理的影响(通过分析LM词典的时间线和语境差异,能够追踪AI读者如何影响公司情感管理策略)

As a first step, we establish that firms which expect high machine downloads avoid LM-negative words but only post-2011 (the publication year of the LM dictionary). Such a structural change is absent with respect to words deemed negative by the Harvard dictionary. As a result, the difference, LM – Harvard sentiment, follows the same path as LM sentiment. For a tighter identification, we further confirm a parallel pre-trend in LM – Harvard sentiment between firms with high and low (top and bottom terciles of) machine downloads up to 2010. Post-2011 saw a clear divergence where the “high” group significantly reduced, relative to the “low” group, the use of negative words from the LM dictionary as opposed to those from the Harvard dictionary. Given the quasi-randomness of the exact timing of publication, the difference-in-differences in the sentiment expression is more likely to be attributable to firms’ catering to their AI readers than to an alternative hypothesis that the publication was a side show of a preexisting and continuing trend. 【作为第一步,我们确立了那些预期有大量机器下载的公司在2011年之后(即LM词典出版年份)避免使用LM词典中的负面词汇。这样的结构性变化在哈佛词典定义的负面词汇中并不存在。因此,LM词典与哈佛词典之间的情感差异(LM – Harvard情感)与LM情感呈相同路径。为了更严格的识别,我们进一步确认了2010年之前高机器下载量公司和低机器下载量公司(前1/3和后1/3分位数)在LM – Harvard情感上的平行趋势。2011年后,”高”机器下载量组相较于”低”组,显著减少了使用LM词典中的负面词汇,而不是哈佛词典中的负面词汇。鉴于出版时间的准随机性,情感表达中的差异-差异分析更可能归因于公司为了迎合AI读者,而不是现有持续趋势的附带现象。】

公司对机器读者的反应(2011年之后,预期有较多机器下载的公司开始避免使用LM词典中的负面词汇)→→→结构性变化(这一变化仅存在于LM词典的负面词汇中,而不适用于哈佛词典)→→→情感差异的变化路径(LM与哈佛情感的差异遵循与LM情感相同的变化路径)→→→前期平行趋势的确认(2010年之前,高机器下载量和低机器下载量的公司在LM – Harvard情感上具有相似的趋势)→→→2011年后差异的产生(2011年之后,高机器下载量组明显减少使用LM词典中的负面词汇,低下载量组则没有类似变化)→→→差异-差异分析的原因(由于LM词典发布时间的准随机性,这一变化更可能是公司为了迎合AI读者的结果,而不是持续趋势的附带现象)

The documented relation raises intriguing equilibrium implications. If firms can “positify” language without cost and constraint in order to impress machine and human readers, the signals would quickly lose relevance. To remain in an equilibrium in which investors extract information from disclosures, we hypothesize that firms derive and incur heterogeneous benefits and costs from managing sentiment and tone. On the benefit side, we find that firms facing imminent external financing needs are more likely to suppress $\color{blue}{Loughran\ and\ McDonald\ (2011JF)}$ negative words and to disclose in more machine-readable format so as to ensure that the positive signals are well received. On the cost side, firms facing higher litigation risk are more moderated in their word-mincing. 【已记录的关系引发了有趣的均衡含义。如果企业能够在没有成本和约束的情况下通过积极化语言来取悦机器和人类读者,这些信号将很快失去其意义。为了保持一种投资者能够从披露中提取信息的均衡状态,我们假设企业在管理情感和语调时会产生异质的收益和成本。从收益方面来看,我们发现面临迫切外部融资需求的企业更有可能抑制使用LM(2011)负面词汇,并以更便于机器阅读的格式进行披露,以确保正面信号能够被很好地接收。从成本方面来看,面临较高诉讼风险的企业在遣词造句上会更加谨慎。】

信号的潜在失效(如果企业可以毫无成本地将语言“积极化”以取悦机器和人类读者,信息信号可能会失去其相关性)→→→均衡的保持(为了维持一种投资者仍然能够从披露中获取有效信息的均衡状态,企业需要在情感和语调的管理上进行不同的权衡)→→→收益与成本的异质性(企业在管理情感和语调时面临不同的收益和成本)

  • 收益:有外部融资需求的企业更倾向于减少使用负面词汇,并以机器友好的方式披露信息,以确保正面信号传达给投资者。
  • 成本:面临较高诉讼风险的企业在使用措辞上会更加谨慎,避免过度修饰语言。

The rapid evolution of AI technology, even during the writing and revision of this paper, provides “out-of-sample” tests to affirm that the relation we identified off the publication of $\color{blue}{Loughran\ and\ McDonald\ (2011JF)}$ is not a lone incidence. First, we resort to the emergence of Bidirectional Encoder Representations from Transformers (BERT) developed by Google in 2018 ($\color{blue}{Devlin\ et\ al.,\ 2018}$), the state-of-the-art for machine processing of textual data. We show that BERT-measured negative sentiment drops more post-2018 for firms with higher AI readership, measured by AI ownership and AI talent supply. Second, we take the study about “how to talk when a machine is listening” literally into the speech setting. Earlier work ($\color{blue}{Mayew\ and\ Ventakachalam,\ 2012JF}$) finds that managers’ vocal expressions, as assessed by vocal analytic software, can convey incremental information valuable to analysts covering the firm. Thus, managers should recognize that their speeches need to impress bots as well as humans. Applying the software to extract two emotional features well-established in the psychology literature, valence and arousal (corresponding to positivity and excitedness of voices), from managerial speeches in conference calls, we find that managers of firms with higher expected machine readership exhibit more positivity and excitement in their vocal tones, echoing the anecdotal evidence that managers increasingly train or even seek professional help to improve their vocal performances along the quantifiable metrics. 【人工智能技术的快速发展,即便在本文撰写和修订期间,也为我们提供了“样本外”测试,以验证我们基于LM(2011)发布所识别的关系并非个例。首先,我们利用了由Google在2018年开发的双向编码器表示(BERT),这是当前处理文本数据的最先进技术。我们展示了,BERT测量的负面情绪在2018年后对那些具有较高AI读者群体(通过AI持股量和AI人才供应衡量)的公司下降得更多。其次,我们将“如何在机器倾听时进行沟通”的研究引入了语音场景。早期的研究(Mayew和Ventakachalam 2012)发现,经理人的语音表达经过语音分析软件评估后,可以传达对分析师有价值的增量信息。因此,经理人应认识到,他们的演讲不仅要打动人类,还要打动机器。通过应用软件提取心理学文献中广泛认可的两个情感特征——愉悦度(valence)和激发度(arousal)(对应于声音的积极性和兴奋度)——从公司电话会议中的管理层发言中,我们发现那些预期拥有更多机器读者的公司,管理层表现出更多的语音积极性和兴奋度。这与传闻相呼应,越来越多的经理人接受培训,甚至寻求专业帮助,以提高他们在可量化指标上的语音表现。】

AI技术进步提供验证机会(AI技术的快速发展为本文验证LM(2011)发布后识别的关系提供了“样本外”测试)→→→BERT的应用(通过使用2018年Google开发的BERT技术,我们发现具有较高AI读者的公司,其BERT测量的负面情绪在2018年后下降更多)→→→语音分析的延伸(将“如何在机器倾听时进行沟通”延伸到管理层的语音表达分析,发现经理人的语音情感特征(愉悦度和激发度)与机器读者的预期数量相关)→→→演讲的机器化调整(经理人需要通过积极性和兴奋度来打动机器读者,越来越多的经理人为提高其语音表现寻求专业帮助)

Our study builds on an expanding literature on information acquisition and dissemination via SEC-filing downloads ($\color{blue}{Bernard\ et\ al.,\ 2020}$; $\color{blue}{Chen\ et\ al.,\ 2020}$; Cao et al. 2021; Crane, Crotty, and Umar 2022), opting for a new angle on the consequences of and human reactions to machine processing. A central theme from the rapidly growing literature on textual analysis is that qualitative information from and the writing quality of disclosures predict asset returns and corporate performance.3 The computational textual analyses have been steadily advanced by more-modern machine-learning techniques (see a survey article by Cong et al. 2021), and have been extended to nontext data, such as the audio of conference calls (Mayew and Ventakachalam 2012) and video of startup pitch presentations (Hu and Ma 2021). Our study departs from the extant literature as we explore managerial disclosure strategies in response to the growing presence of AI analytical tools in both the industry and academia. 【我们的研究基于不断扩展的文献,这些文献探讨了通过SEC文件下载进行信息获取和传播的方式(Bernard, Blackburne, and Thornock 2020; Chen et al. 2020; Cao et al. 2021; Crane, Crotty, and Umar 2022),并从一个新的角度探讨了机器处理的后果及人类的反应。快速增长的文本分析文献的核心主题是,披露文件中的定性信息和书写质量能够预测资产回报和企业绩效。随着现代机器学习技术的进步(参见Cong等2021年的综述文章),计算文本分析逐渐扩展到非文本数据领域,如公司电话会议的音频(Mayew和Ventakachalam 2012)和创业公司演示的录像(Hu和Ma 2021)。我们的研究不同于现有文献,因为我们探讨了管理层在应对AI分析工具日益广泛应用时的披露策略。】

信息获取与传播的现有文献(通过SEC文件下载进行信息获取和传播的研究不断扩展,已有文献探讨了这一过程中机器处理的后果及人类反应)→→→文本分析的核心主题(文献表明,披露文件的定性信息和书写质量可以预测资产回报和公司绩效)→→→机器学习技术的进展(随着现代机器学习技术的进步,计算文本分析不仅限于文本数据,还扩展到了非文本数据,如电话会议音频和创业公司演示视频)→→→我们的研究创新(与现有文献不同,我们探讨的是管理层在AI分析工具日益广泛应用的背景下,如何调整其披露策略)

Cao et al. 2021;

$\color{blue}{\ et\ al.,\ 2012}$

Crane, Crotty, and Umar 2022

$\color{blue}{\ et\ al.,\ 2012}$

$\color{green}{picture\ superiority\ effect}$

$\color{red}{在照片显著的时期,照片会从文字中捕获注意力}$

$\color{blue}{\ and\ ,\ 2004}$

$\color{blue}{\ et\ al.,\ 2012}$

$\color{blue}{,\ 2007}$

References

  1. Cohen, L., Malloy, C., & Nguyen, Q. (2020). Lazy prices. The Journal of Finance, 75(3), 1371-1415.
  2. Allee, K. D., DeAngelis, M. D., & Moon Jr, J. R. (2018). Disclosure “scriptability”. Journal of Accounting Research, 56(2), 363-430.
  3. Loughran, T., & McDonald, B. (2011). When is a liability not a liability? Textual analysis, dictionaries, and 10‐Ks. The Journal of Finance, 66(1), 35-65.
  4. Kothari, S. P., Shu, S., & Wysocki, P. D. (2009). Do managers withhold bad news?. Journal of Accounting Research, 47(1), 241-276.
  5. Devlin, J., M. Chang, K. Lee, & K. Toutanova. (2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
  6. Mayew, W. J., & Venkatachalam, M. (2012). The power of voice: Managerial affective states and future firm performance. The Journal of Finance, 67(1), 1-43.
  7. Bernard, D., Blackburne, T., & Thornock, J. (2020). Information flows among rivals and corporate investment. Journal of Financial Economics, 136(3), 760-779.
  8. Chen, H., Cohen, L., Gurun, U., Lou, D., & Malloy, C. (2020). IQ from IP: Simplifying search in portfolio choice. Journal of Financial Economics, 138(1), 118-137.