COVID19-MLSF: A multi-task learning-based stock market forecasting framework during the COVID-19 pandemic

Yuan, Chenxun; Ma, Xiang; Wang, Hua; Zhang, Caiming; Li, Xuemei

doi:10.1016/j.eswa.2023.119549

Cited by 18 publications

(6 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhang et al found that social media data may cause investor fear and the propagation of false information, which in turn increases market volatility and decreases stock prices [2]. Similar to what we see here, P H and Rishad discovered that negative sentiment in social media data is linked to increased market volatility and falling stock prices [14]. Hence, these studies underscore the necessity of monitoring and interpreting social media data for market participants and imply that inactive social media data might have a negative influence on the stock market.…”

Section: Positive Relationshipsupporting

confidence: 82%

“…The connection between social media sentiments and stock market volatility has also been explained from the standpoint of behavioral finance theory [14]. According to this hypothesis, market volatility is influenced by investors' subjective experiences, including their emotions, prejudices, and sentiments.…”

Section: Theoretical Perspectives On Using Social Media and News Sent...mentioning

confidence: 99%

See 1 more Smart Citation

The Relationship Between Social Media Data and Stock Market Volatility

Yang

2023

AEMPS

View full text Add to dashboard Cite

In recent years, the influence of social media data on the stock market has been deeply concerned and studied by academics and professionals, because the information reflected in these data has become increasingly important in contemporary society. By offering a complete synthesis of current research, evaluating the contributions made by earlier studies, and highlighting knowledge gaps warranting additional inquiry, this literature review seeks to examine the complicated link between social media data and stock market volatility. In order to accomplish this, the paper delves into a wide range of topics, such as where to get social media data, how to gauge stock market volatility, and what kinds of approaches have been taken to examine the possible connections between the two. Some research suggests that social media sentiment may forecast market patterns, while other research suggests a lower effect, as shown by our study, which concludes that there is either no relationship or a positive one between these factors. More study should focus on perfecting sentiment analysis methods, delving into nonlinear correlations, exploring variations in social media platforms, and addressing ethical problems in data utilization. We hope that these efforts will help us better comprehend the relationship between social media data and stock market volatility, allowing us to create more precise forecasting instruments.

show abstract

Section: Positive Relationshipsupporting

confidence: 82%

Section: Theoretical Perspectives On Using Social Media and News Sent...mentioning

confidence: 99%

The Relationship Between Social Media Data and Stock Market Volatility

Yang

2023

AEMPS

View full text Add to dashboard Cite

show abstract

“…However, the authors found significantly better results when some or all of the parameters of the stock models are not shared. Since then, much more research has been done on multi-task learning in finance, but much of this is focused on forecasting forward returns, ranking assets and not portfolio construction (Ma & Tan, 2022;Kang et al, 2022;Yuan et al, 2023).…”

Section: Literature Review and Research Gapmentioning

confidence: 99%

Constructing Time-Series Momentum Portfolios with Deep Multi-Task Learning

Ong

Herremans

2022

SSRN Journal

View full text Add to dashboard Cite

“…Among these elements, the COVID-19 pandemic is the most signifcant factor afecting the fnancial performance of businesses worldwide. References [6,9] explore that the coronavirus illness (COVID- 19) has posed a signifcant threat to both the public and business worlds. COVID-19 began in December 2019 in China, and the entire economic and social worlds are grappling with this issue [10].…”

Section: Introductionmentioning

confidence: 99%

Overflow Effect of COVID-19 Pandemic on Stock Market Performance: A Study Based on Growing Economy

Qadri

Raza²,

Qadri

et al. 2023

Discrete Dynamics in Nature and Society

View full text Add to dashboard Cite

Purpose. The purpose of this study is to evaluate the effect that COVID-19 has on the performance of the stock market in emerging economies. The findings as a whole demonstrate that the stock market does not react significantly. We believe that the outcomes of this study provide knowledge that is useful for decision makers in financial markets and policy all around the globe. Methods. Data for 140 companies are obtained from the official websites of the companies listed on the Pakistan Stock Exchange. The Wilcoxon signed-rank test was used to determine whether there was a significant difference before and after the pandemic era. In addition, the ARDL model was chosen because of the variable integration mix (order 0 and 1). Implications. Theoretically, it added to the information that was already available about the implications of the COVID-19 pandemic on the performance of stock markets in emerging economies. In a purely practical way, it will help those in charge of making policies come up with plans to deal with things that cannot be predicted. Results. The findings of the analysis demonstrate that the COVID-19 pandemic has caused a 52.85% reduction in the volatility of Pakistan’s net profit returns. The study shows a statistically insignificant negative association between the COVID-19 pandemic and Pakistan’s stock returns. The empirical results of ARDL models provide the first conclusion of the analysis, indicating that the number of long-term connections was greater than short-term connections. Conclusion. According to the findings of the research, the pandemic caused by COVID-19 has a bigger impact on the financial performance of enterprises (both positively and negatively). Some companies are able to maintain their place in the market even if the bulk of the firms see their performance suffer during a pandemic. Originality. We are the first to use the ARDL model to evaluate the effect of the new COVID-19 pandemic on a stock market in a developing nation.

show abstract

COVID19-MLSF: A multi-task learning-based stock market forecasting framework during the COVID-19 pandemic

Cited by 18 publications

References 41 publications

The Relationship Between Social Media Data and Stock Market Volatility

The Relationship Between Social Media Data and Stock Market Volatility

Constructing Time-Series Momentum Portfolios with Deep Multi-Task Learning

Overflow Effect of COVID-19 Pandemic on Stock Market Performance: A Study Based on Growing Economy

Contact Info

Product

Resources

About