Published: Jan. 1, 2024
Language: Английский
SSRN Electronic Journal, Journal Year: 2023, Volume and Issue: unknown
Published: Jan. 1, 2023
This paper examines whether ChatGPT can identify useful news content for the aggregate stock market and macroeconomy, using headlines alerts on front pages of Wall Street Journal. We find that information extracted by is highly related to macroeconomic conditions. Investors tend underreact positive contents, especially during periods economic downturns, high uncertainty novelty news, which leads significant predictability ChatGPT. By contrast, negative only associated with contemporaneous returns, it cannot predict future market. Traditional methods textual analysis, such as word lists or small large language models (LLMs) like BERT, barely any in either nor news. In short, appears best its kind capable discerning economic-related drive
Language: Английский
Citations
19
Published: Jan. 1, 2025
This study assesses the impact of GPT-4-generated fund prospectus summaries on investor comprehension and decision-making through an online experiment. A total 305 participants were given designed with varying levels complexity, labeled as "easy" "hard." The findings reveal that easy-to-understand significantly enhance text accessibility by approximately 13% investment willingness 8%. improvement is observed regardless participants' self-reported or objectively measured financial literacy. also finds individuals' selfassessed competence plays a more crucial role than their actual literacy in interacting information making decisions. These results highlight capability advanced language models like ChatGPT to simplify condense complex disclosures, thereby potentially broadening participation enhancing engagement.
Language: Английский
Citations
0
Journal of Advanced Research, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 1, 2025
Language: Английский
Citations
0
Advances in computational intelligence and robotics book series, Journal Year: 2025, Volume and Issue: unknown, P. 151 - 200
Published: March 28, 2025
This study examines the potential of ChatGPT v4.0 for stock market analysis, particularly within Indian market. The research uses prompt engineering to assess ChatGPT's ability recommend stocks, construct portfolios, provide fundamental and understand price movements. combines sentiment analysis news data with historical forecast future Results demonstrate effectiveness in identifying investment opportunities providing insights decision-making. also evaluates accuracy predictions, finding an 88.89% rate based on a 94% when integrating data. Despite limitations, including constraints handling large datasets, exhibits promise supporting investors Future should focus refining models addressing ethical considerations ensure responsible AI integration finance.
Language: Английский
Citations
0
IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 134041 - 134061
Published: Jan. 1, 2024
This paper comprehensively examines Large Language Models (LLMs) in sentiment analysis, specifically focusing on financial markets and exploring the correlation between news Bitcoin prices.We systematically categorize various LLMs used highlighting their unique applications features.We also investigate methodologies for effective data collection categorization, underscoring need diverse comprehensive datasets.Our research features a case study investigating prices, utilizing advanced analysis methods to demonstrate practical application of LLMs.The findings reveal modest but discernible price fluctuations, with historical patterns showing more substantial impact Bitcoin's longer-term than immediate events.This highlights LLMs' potential market trend prediction informed investment decision-making.
Language: Английский
Citations
2
Cogent Education, Journal Year: 2024, Volume and Issue: 11(1)
Published: Oct. 7, 2024
Using a quasi-experimental method and content analysis as technique, this study tests ChatGPT, in its version 4, by assessing textual characteristics overall understanding regarding the recognition criteria of provisions under International Accounting Standards (IAS) 37, issued Board (IASB). For purpose, it uses set questions (input) from IASB's illustrative examples to compare answers (output) IASB ChatGPT two distinct strategies: with without prompting. The findings indicate that ChatGPT's are wordier, have higher magnitude levels, more predominantly inserted Business Finance. no-prompting strategy is globally negative subjective, while prompting one improves answers' focus readability, also presenting diverse tones characteristics, similar what was found answers. However, some were not accurate both strategies. These provide insights into how most disseminated artificial intelligence tools, can be used accounting professionals educators, being aware potential risks benefits strategies underlying experiment. Then, considering those aspects, practitioners, including accountants managers, but investors use understand matters issues assessment given situation, well sources consider when preparing financial statements or making decision. Academics open up discussions promote students' critical thinking skills classroom environment.
Language: Английский
Citations
1
Administrative Sciences, Journal Year: 2024, Volume and Issue: 14(7), P. 152 - 152
Published: July 16, 2024
Purpose: From an exploratory perspective, this paper aims to assess how well ChatGPT scores in accounting proficiency exam Portugal, as its overall understanding of the issues, purpose and context underlying questions under assessment. Design/methodology/approach: A quasi-experimental method is used study. The from by Portuguese Order Chartered Accountants (OCC, acronym) served input queries, while responses (outputs) were compared with those OCC. Findings: findings indicate that ChatGPT’s able deduce primary issue matters assessed, although some inaccurate or imprecise. Also, tool did not have same score all matters, being less accurate requiring more professional judgment. also show pass exam, it was close doing so. Originality: To best authors’ knowledge, there little research on accuracy exams, first such study Portugal. Practical implications: can be useful professionals understand may for practitioners, stressing could assist them improve efficiency, but cannot, at least now, replace human professional. It highlights potential use additional resource classroom, encouraging students engage critical thinking facilitating open discussion guidance teachers. Consequently, prove beneficial academic purposes, aiding learning process.
Language: Английский
Citations
0
Revue Internationale des Économistes de Langue Française, Journal Year: 2024, Volume and Issue: 9(1)
Published: Aug. 2, 2024
L’objectif – Le but de l’article est d'identifier les publications concernant ChatGPT dans le domaine la comptabilité publiées depuis lancement jusqu’à présent (mai 2024) afin d'avoir une image approximative popularité du concept en tant que sujets d'études recherche. L’analyse des domaines recherches, ainsi applications potentielles et défis d’utilisantion suit revue littétature. Conception/méthodologie/approche Une analyse contenu bases données Scopus Web of Science Scholar a démontré un intérêt croissant parmi cherchers scientifiques pour IA ChatGPT, notamment version 3.5 30 novembre 2022. 16 articles publiés 2023 2024 liés à l’utilisation ont été examinés. méthodologie recherche descriptive, comprenant littérature, appliquée d’atteindre l’objectif déclaré. Résultats Six sont études empiriques pendant autres dix constituent recherches théoretiques. Les résultats indiquent (1) tendance croissante académiciens quant possibilités l’éducation ; (2) grand intêret parmis practiciens lié l’application nouvelles technologies d’AI financière l’audit (au sein cabinets comptables), reporting ESG gestion. Originalité/valeur L’article contribue littérature comptabilité, constituant synthèse tout ce qui publié Sciences sur sujet.
Citations
0
Published: Jan. 1, 2024
Language: Английский
Citations
0