Abstract The study provides large‐scale descriptive evidence on the timing and nature of corporate financial tweeting. Using an unsupervised machine learning approach to analyze 24 million tweets posted by S&P 1500 firms from 2012 2020, we find that are more likely tweet information around significantly negative or positive news events, such as earnings announcements filing statements. This convex U‐shaped relation between likelihood posting materiality accounting events becomes stronger...

In this paper, the dynamic CoVaR method is used to measure changes in systemic risk financial industry during COVID-19 pandemic. We find that, first, after outbreak of pandemic, increased significantly. Second, impact pandemic on securities was greater than that banking and insurance industries.

Heroes embody the essence of national spirit and serve as catalysts for rejuvenation. Narratives play a pivotal role in conveying heroic tales upholding ethos heroism. While traditionally set domain mainstream media, responsibility crafting disseminating narratives has increasingly shifted to social media response societal evolution. This study examines innovative approaches on Chinese platform Bilibili. It highlights how Bilibili transformed traditional storytelling by adopting concise...

This research applies panel vector autoregression method (PVAR) to analyze the annual data of Chinese A-share listed companies over period 2009–2020, study relationship between enterprise ESG performance and dynamic financial behavior, by using Sino-Securities' data. The main findings include: first, there is a positive interaction investment performance. Second, factors (e.g., cash flow) have effect on performance, while has fundamental sales revenue). above significant effects are...

This study investigates whether CEO political contribution, as a measure of ideology, is associated with firm’s financial reporting policies in accounting conservatism. Using sample federal-level contributions by CEOs S&P 500 firms, we find that firms Republican-leaning CEOs, who tend to have conservative are higher degree conservatism than Democratic-leaning CEOs. We further show changes ideology around turnovers the policies. Our results robust battery robustness tests. Taken together,...

Our research explores how the COVID-19 pandemic has influenced asymmetric spillover effects in oil and gold markets. Through a VAR(p)-BEKK-AGARCH(1,1) model fitted to daily price data, 1) we find evidence of only from market that this effect is stronger during 2) conclude negative information shock larger impact on return volatility compared positive intensified pandemic.

This study investigates the relationship between corporate site visits (CSVs) and firms’ real earnings management. Using a unique dataset of to Chinese firms listed on Shenzhen Stock Exchange from 2009 2016, we find that such are negatively associated with The results robust using alternative CSV measures, controlling for communication channels, propensity score matching method. In cross-sectional analyses, negative association management is stronger more complex greater information...

Using a machine learning approach to analyze 12.8 million tweets posted by S&P 1500 firms from 2012 2016, we find that time financial around earnings announcements, accounting filings as well other important corporate events, and are more likely use media (images or video) links in those tweets. The above pattern holds for both good bad news. Moreover, feedback Twitter users encourages future of links. These results collectively suggest make discretionary choices timing presentation format...

Using a machine learning approach to process 11 million tweets posted by S&P 1500 firms from 2011 through 2016, we find that poor corporate social responsibility (CSR) performance tweet more about CSR activities and use are shorter, with passive voice extreme tone. Good less CSR, yet gain twice followers per than firms. also experience greater decrease in institutional ownership along higher increases bid-ask spread stock return volatility after joining Twitter do Our findings suggest play...

This article uses transaction-level fund trading data from the United States to study information advantage of institutional investors. Our research design follows a two-step procedure. In first step, we identify funds that sell shares in firms before their unexpected revelation stock option backdating (BD) investigations, and thus establish fund–firm pairs interest. second focus on takes place at other times find are more likely make correct trades earnings announcements paired performance...

Outdoor physical activity duration is a key component of outdoor behavior older adults, and therefore, an important determinant their total levels. In order to develop successful program, it identify any heterogeneity in preferences for patterns among adults. addition, more insight needed the influence environmental characteristics on choice creating supportive neighborhood environments matching individuals' preferences. To this end, mixed multinomial logit model estimated based one-week...

This study uses five machine learning algorithms (Stochastic gradient descent (SGD), Decision tree, Random forest, Gradient boosting decision tree (GBDT), and Convolutional neural networks (CNN)) to explore their prediction effects on China's stock market. It constructs a monthly rolling model for return prediction. Selecting stocks of the CSI 300 index from January June 2021 as specific samples classifying three factors – fundamentals, volatility(risk) technical indicators, results...