TABLE 2 - uploaded by Allen Huang
Content may be subject to copyright.
Source publication
We document that textual discussions in a sample of 363,952 analyst reports provide information to investors beyond that in the contemporaneously released earnings forecasts, stock recommendations, and target prices, and also assist investors in interpreting these signals. Cross-sectionally, we find that investors react more strongly to negative th...
Contexts in source publication
Context 1
... A of Table 2 provides the descriptive statistics for our key variables. Overall, we find that the average length of the reports in our sample is 57 sentences, of which 31 sentences are neutral, 19 are positive, and 7 are negative. ...
Context 2
... 1 show that textual opinions generally become less optimistic after 2001, consistent with the impact of regulatory changes, including the enactment of Reg FD in 2000 and the Global Settlement reached in 2003, on the financial analyst industry. Table 2, Panel B describes textual opinions by report type. Specifically, our average OPN decreases by 40 percent from 0.231 to 0.138 from a BUY to a HOLD recommendation and by a further 36 percent from 0.138 to 0.089 from a HOLD to a SELL. ...
Context 3
... the results in Panel B of Table 2 show that average OPN drops by 19 percent from 0.229 to 0.186 from an upgrade to a reiteration (i.e., a report that contains the same stock recommendation level as in the last report issued by the same analyst), and by more than 46 percent from 0.186 to 0.101 from a reiteration to a downgrade, suggesting that the level of OPN reflects revisions in the recommendations. This finding is consistent with analysts discussing what has changed at the company since their last report and providing detailed explanations for the revisions of the quantitative summary measures. ...
Similar publications
We examine whether gold is an effective hedge against inflation over different time horizons. Using a stationary test with a flexible Fourier function, we consider all possible structural breaks with unknown forms and find that real gold returns over horizons ranging from 1 month to 15 years are nonlinear stationary processes. Although the real gol...
Research Summary
Despite the proliferation of lists and rankings that recognize firms for superior performance, empirical studies have been limited in their ability to causally evaluate how inclusion for the marginal firm influences shareholder value. We address this limitation by examining how investors responded to firms that were barely included...
This paper discusses the optimal portfolio analysis of banking stocks for the period January 2022 to December 2022 using the Markowitz method and a single index model. The aim is to provide guidance for investors to develop an optimal portfolio of banking stocks by considering aspects of risk, return, and correlation to the market. The data used ar...
Untuk meningkatkan kegiatan investasi, maka setiap daerah tak terkecuali kabupaten Lombok Timur berusaha untuk mempromosikan potensi ekonomi dan beragam peluang investasi yang dimiliki termasuk berbagai kemudahan dan insentive yang diberikan kepada para investor. Upaya tersebut menjadi semakin penting mengingat perekonomian kabupaten Lombok Timur s...
The paper investigates the Shariah compliant and conventional portfolios in financial settings of Pakistan during the period 2009-2019 by using Markowitz Minimum-Variance (MMV) framework. Using daily excess returns, we first investigate the impact of Shariah screening criteria on stock returns then we evaluate the overall risk of Shariah compliant...
Citations
... This strategy enhances the overall tone, making the disclosed information more optimistic [29]. In recent years, research on strategic management disclosure have focused on annual report/10-K/10-Q filings [31][32][33], media news [34], earnings press releases [29,35], analyst reports [36], and conference calls [37,38]. ...
Digitalization is anticipated to substantially improve information transparency and enhance the capital market’s information environment. However, during the initial phase of digital transformation, companies face challenges in achieving high-quality information disclosure. This is because digital technology implementation and organizational adjustment are still at early stages. Concurrently, investors face difficulties in assessing the accuracy of disclosed information. This challenge provides management with the opportunity to overstate the benefits of digital transformation. This study investigates the impact of corporate digital transformation on management’s tone manipulation behavior, using a sample of Chinese A-share listed companies from 2012 to 2021. Findings indicate that corporate digital transformation significantly fosters management’s tone manipulation during its exploration phase. Media slant positively moderates this relationship. Further analysis supports the paper’s hypothesis: companies with weaker financial flexibility and lower risk information disclosure levels show a stronger positive correlation between digital transformation and tone manipulation. Concurrently, mechanism analysis reveals that management overconfidence partially mediates the relationship. This suggests that digital transformation increases managerial overconfidence, thereby promoting tone manipulation. The conclusion offers new insights into enhancing management discussion and analysis information disclosure quality from a corporate strategic transformation perspective. It serves as a valuable reference for accurately identifying misleading management signals.
... Analysts, as information intermediaries, oer stock valuations that may provide insights into rms' information uncertainty. Specically, sell-side analysts gather and interpret costly information, which they use to issue stock valuations that assist investors in their decision-making (Huang et al., 2014;Hilary & Hsu, 2013). However, given that analysts have commercial incentives (Cowen et al., 2006;Jackson, 2005) and often issue optimistic forecasts (Malmendier & Shanthikumar, 2014;Lim, 2001), it remains uncertain whether their valuations eectively illuminate a rm's information uncertainty. ...
This study investigates whether analysts provide informative forecasts for stocks issued by firms with greater information uncertainty. As firm-specific information uncertainty is not directly observable, the research highlights the role of analysts' forecasts and reports in offering valuable insights to investors. It also investigates whether forecast quality is sufficiently captured by forecast bias. The findings indicate that forecast quality tends to be lower for firms with greater information uncertainty and that forecast bias alone does not fully reflect the informational content of analysts' forecasts. Overall, the results suggest that analysts' forecasts possess positive informational value.
... The NLP has favored the carrying out of studies that use qualitative measures, such as research that deals with narratives (Beattie, 2014;Fisher et al., 2016). Thus, some studies have begun to explore the potential of tweets and news as data sources for predicting market returns [see the works by Brown (1999), Baker and Wurgler (2006), Garcia (2013), Huang et al. (2014), Han and Li (2017), Jiang et al. (2019), Liang et al. (2020), Shapiro et al. (2017), Huynh et al. (2021) and Obaid and Pukthuanthong (2022)]. ...
... Concerning textual information, studies commonly analyze two main attributes: (a) readability, defined as the syntactic complexity of communication, reflecting the level of text comprehension (Lehavy et al., 2011;Beattie, 2014); and (b) tone or content, referring to the semantics of the communication process. This involves the interpretation of information, with potential optimistic or pessimistic implications based on the words used in the text (Beattie, 2014;Huang et al., 2014). ...
... This analysis employs natural language processing techniques, enabling the identification and classification of emotions conveyed in the text, such as positivity, negativity, and neutrality. Therefore, a considerable amount of textual information has been the focal point of empirical investigations, encompassing diverse sources such as newspapers [refer to the works of Tetlock (2010), Griffin et al. (2011), Dougal et al. (2012, Garcia (2013), Liu et al. (2017) and Hendershott et al. (2015)]; management reports or specific documents (Li, 2008;Loughran & McDonald, 2011;Huang et al., 2014); chat messages, Twitter, Facebook, and spam messages related to assets (Antweiler & Frank, 2004;Bollen et al., 2011;Hu et al., 2010;Karabulut, 2013); and analyst forecasts (Twedt & Rees, 2012). More recently, information has also been sourced from Chat-GPT (Lopez-Lira & Tang, 2023). ...
The rapid advancement of artificial intelligence, exemplified by tools such as Chat-GPT, has significantly transformed the landscape of stock market analysis. This paper aims to leverage these technological developments to predict the daily returns of the Ibovespa by utilizing predictors derived from technical indicators and sentiment indices extracted from textual data and Chat-GPT-generated sentiment indices. Our findings reveal that the Chat-GPT-based sentiment index does not enhance the out-of-sample prediction of Ibovespa returns. Conversely, the sentiment index derived from financial news data, utilizing a time-varying dictionary, demonstrates improved out-of-sample predictive accuracy for the Ibovespa. Notably, the predictor based on the technical indicator Accumulation–Distribution (AD) outperforms the historical average benchmark, establishing itself as the superior forecasting model. This study contributes to the ongoing discourse on the integration of artificial intelligence and traditional financial analysis, offering insights into the efficacy of sentiment indices and technical indicators for forecasting stock market returns in the Brazilian context.
... If they do not hide and decide to frankly raise the concerns to the investors, then they will disclose more risk factors related to the upcoming and ongoing crisis. The expression ways such as the sentiment and the level of details when disclosing the risks should also show different states as the crisis evolves (Huang et al. 2014). For example, when a financial institution faced the subprime crisis, its disclosures' sentiment may be more negative. ...
This paper studies whether the risk factors disclosed in financial reports are as informative as SEC requires. The subprime crisis is used as a typical real-risk event to test their informativeness by text mining methods. By analyzing the textual attributes and specific contents of the risk disclosures in 14089 Form 10-K statements of 1,685 financial firms from 2006 to 2022, the results show that firms tend to disclose risks in more specific and negative language and significantly reduce stickiness and boilerplate during the subprime crisis. This paper also finds that the changes of risk factors are consistent with the actual risk profile before, during and after the subprime crisis. The failed institutions during the crisis have disclosed more risks related to the subprime crisis using a more negative tone than the survived institutions. These findings all provide evidence that the risk factors in Form 10-K are informative and not invariable boilerplate.
... Online consumer reviews are a rich source of information about purchase-related decision-making (Huang et al., 2014). However, there is little exploration of quality perceptions from the customer perspective in the literature. ...
Purpose
This study aims to understand holistic consumer perceptions of quality and their effect on re-purchase intentions by measuring the latent characteristics of online Amazon reviews.
Design/methodology/approach
Data was collected from entries in the Amazon customer review data set, which explicitly mentions quality, economic evaluation and future purchase intention. The analyses included natural language processing, structural topic modeling and econometric analysis. The study used real-time customer reviews to determine the overall perceived quality, the impact of perceived quality on re-purchase intention and the mediating roles of price consciousness and customer satisfaction.
Findings
Consumers’ perception of overall quality includes product- and service-related dimensions. Perceived quality influences re-purchase intentions through the mediating role of customer satisfaction. While price consciousness impacts the link between perceived quality and customer satisfaction, it does not affect re-purchase intention.
Practical implications
The managerial implications emphasize multiple dimensions of quality in the online environment and the role of customer satisfaction in consumers’ online re-purchase intentions. The results also illustrate that price effects are insignificant in influencing re-purchase intentions. Thus, while price cuts may encourage initial purchases, quality and customer satisfaction are vital to stimulate re-purchase.
Originality/value
The e-commerce literature lacks a comprehensive and rigorous understanding of the components of consumers’ perceived quality. This research develops a thorough understanding of what impacts overall e-commerce quality based on real-time customer reviews, avoiding the biases arising from traditional methods, including surveys.
... Researchers also focus on the modal word frequency and sentiment analysis of corporate disclosure tone by examining earnings conference calls (Li et al., 2021), mandatory SEC filings (Huang et al., 2023), press releases (Henry and Leone, 2016), analyst reports (Huang et al., 2014), critical audit matters (Liu et al., 2022) and social media (Booker et al., 2023). While many articles have explored the ML application in various firms' disclosures, the ML application in ESG disclosures, known as nonfinancial disclosure, has less evidence. ...
... For example, researchers use ML to predict firms' financial fraud behavior (Perols, 2011;Bao et al., 2020), bankruptcy (Barboza et al., 2017), misstatements (Bertomeu et al., 2021), effective tax rates (Guenther et al., 2023), auditor switches (Hunt et al., 2021) and selecting directors (Erel et al., 2021). Another trend of ML in accounting research focuses on textual analysis of corporate disclosures (Li, 2010;Huang et al., 2014Huang et al., , 2023Henry and Leone, 2016;Li et al., 2021;Liu et al., 2022;Booker et al., 2023). In the ESG setting, several researchers simply measure firms' ESG practices by an indicator variable to capture whether a firm has ESG disclosure (Dhaliwal et al., 2012;Hoi et al., 2013;Dai et al., 2023). ...
Purpose
Environmental, social and governance (ESG) factors have become increasingly important in investment decisions, leading to a surge in ESG investing and the rise of sustainable investment assets. Nevertheless, challenges in ESG disclosure, such as quantifying unstructured data, lack of guidelines and comparability, rampantly exist. ESG rating agencies play a crucial role in assessing corporate ESG performance, but concerns over their credibility and reliability persist. To address these issues, researchers are increasingly utilizing machine learning (ML) tools to enhance ESG reporting and evaluation. By leveraging ML, accounting practitioners and researchers gain deeper insights into the relationship between ESG practices and financial performance, offering a more data-driven understanding of ESG impacts on business communities.
Design/methodology/approach
The authors review the current research on ESG disclosure and ESG performance disagreement, followed by the review of current ESG research with ML tools in three areas: connecting ML with ESG disclosures, integrating ML with ESG rating disagreement and employing ML with ESG in other settings. By comparing different research's ML applications in ESG research, the authors conclude the positive and negative sides of those research studies.
Findings
The practice of ESG reporting and assurance is on the rise, but still in its technical infancy. ML methods offer advantages over traditional approaches in accounting, efficiently handling large, unstructured data and capturing complex patterns, contributing to their superiority. ML methods excel in prediction accuracy, making them ideal for tasks like fraud detection and financial forecasting. Their adaptability and feature interaction capabilities make them well-suited for addressing diverse and evolving accounting problems, surpassing traditional methods in accuracy and insight.
Originality/value
The authors broadly review the accounting research with the ML method in ESG-related issues. By emphasizing the advantages of ML compared to traditional methods, the authors offer suggestions for future research in ML applications in ESG-related fields.
... Bao et al. (2020) employed ensemble learning to develop a fraud-prediction model that demonstrated superior performance compared to the logistic regression and support vector machine models with a financial kernel. Huang et al. (2014) used Bayesian networks to extract textual opinions, and their findings showed that they outperformed dictionarybased approaches, both general and financial. Ding et al. (2020) used insurance companies' data on loss reserve estimates and realizations and documented that the loss estimates generated by ML were superior to the actual managerial estimates reported in financial statements in four out of the five insurance lines examined. ...
This study provides a comprehensive review of machine learning (ML) applications in the fields of business and finance. First, it introduces the most commonly used ML techniques and explores their diverse applications in marketing, stock analysis, demand forecasting, and energy marketing. In particular, this review critically analyzes over 100 articles and reveals a strong inclination toward deep learning techniques, such as deep neural, convolutional neural, and recurrent neural networks, which have garnered immense popularity in financial contexts owing to their remarkable performance. This review shows that ML techniques, particularly deep learning, demonstrate substantial potential for enhancing business decision-making processes and achieving more accurate and efficient predictions of financial outcomes. In particular, ML techniques exhibit promising research prospects in cryptocurrencies, financial crime detection, and marketing, underscoring the extensive opportunities in these areas. However, some limitations regarding ML applications in the business and finance domains remain, including issues related to linguistic information processes, interpretability, data quality, generalization, and the oversights related to social networks and causal relationships. Thus, addressing these challenges is a promising avenue for future research.
... Taking the analysis further, Garcia (2013) examined two New York Times columns spanning over a century by analyzing the frequency of positive and negative words in the text; and found that language tone is correlated with future stock returns, particularly during recessions. Huang et al. (2014) report an asymmetric effect, where investors react more strongly to negative texts than to positive ones. A rich source on textual sentiment literature is the review article by Kearney and Liu (2014) and references therein. ...
... Such an asymmetric response to media sentiments, but constructed from news reports and articles, is consistent with findings reported in studies such as Huang et al. (2014), Heston and Sinha (2017), Bajo and Raimondo (2017), Huang et al. (2018), and He et al. (2024). Consequently, our results confirm the negativity bias documented in the climate finance literature. ...
This article develops multiple novel climate risk measures (or variables) based on the television news coverage by Bloomberg, CNBC, and Fox Business, and examines how they affect the systematic and idiosyncratic risks of clean energy firms in the United States (US). The measures are built on climate related keywords and cover the volume of coverage, type of coverage (climate crisis, renewable energy, and government and human initiatives), and media sentiments. We show that an increase in the aggregate measure of climate risk, as indicated by coverage volume, reduces idiosyncratic risk while increasing systematic risk. When climate risk is segregated, we find that systematic risk is positively affected by the \textit{physical risk} of climate crises and \textit{transition risk} from government and human initiatives, but no such impact is evident for idiosyncratic risk. Additionally, we observe an asymmetry in risk behavior: negative sentiments tend to increase idiosyncratic risk and decrease systematic risk, while positive sentiments have no significant impact. These findings remain robust to including print media and climate policy uncertainty variables, though some deviations are noted during the COVID-19 period.
... The corpus is then used to construct a financial vocabulary (FinVocab) for pretraining BERT. The Financial PhraseBank, FiQA Task 1 [38], and AnalystTone datasets [41] are then used to fine-tune the pretrained model, resulting in three different versions of FinBERT. The experimental results showed that their FinBERT models have higher performance compared with the generic BERT models. ...
Predicting the directions of financial markets has been performed using a variety of approaches, and the large volume of unstructured data generated by traders and other stakeholders on social media microblog platforms provides unique opportunities for analyzing financial markets using additional perspectives. Pretrained large language models (LLMs) have demonstrated very good performance on a variety of sentiment analysis tasks in different domains. However, it is known that sentiment analysis is a very domain-dependent NLP task that requires knowledge of the domain ontology, and this is particularly the case with the financial domain, which uses its own unique vocabulary. Recent developments in NLP and deep learning including LLMs have made it possible to generate actionable financial sentiments using multiple sources including financial news, company fundamentals, technical indicators, as well social media microblogs posted on platforms such as StockTwits and X (formerly Twitter). We developed a financial social media sentiment analyzer (FinSoSent), which is a domain-specific large language model for the financial domain that was pretrained on financial news articles and fine-tuned and tested using several financial social media corpora. We conducted a large number of experiments using different learning rates, epochs, and batch sizes to yield the best performing model. Our model outperforms current state-of-the-art FSA models based on over 860 experiments, demonstrating the efficacy and effectiveness of FinSoSent. We also conducted experiments using ensemble models comprising FinSoSent and the other current state-of-the-art FSA models used in this research, and a slight performance improvement was obtained based on majority voting. Based on the results obtained across all models in these experiments, the significance of this study is that it highlights the fact that, despite the recent advances of LLMs, sentiment analysis even in domain-specific contexts remains a difficult research problem.
... The differences between our setting and short sellers extend to third-party producers of firm-specific information, such as analysts (e.g., Huang et al. 2014), journalists (e.g., Bushee et al. 2010), and firms that tweet about their peers (Cao et al. 2021). Our setting's activists have different economic incentives: they have a positive ownership stake in the firm, substantial ownership-conferred influence over management, and private information about their and management's intent. ...
Activist investors in a firm often voluntarily release information about their governance intentions to the public. Voluntary disclosure theory suggests that an activist investor will disclose when she expects other investors to respond positively and support her in upcoming corporate control contests. We find that activists’ disclosures are accompanied by positive abnormal returns, reductions in bid-ask spreads, and increases in future earnings relative to similar targets without voluntary activist disclosures. Disclosures by activists who demand a board seat (the most common demand) have the highest announcement returns, and disclosers also win proxy contests and directorships more frequently than non-disclosers. These findings suggest that the activist’s beliefs about investor response in both pricing and voting are an important driver of her disclosure choice.
