Scholarlink

<p>This study investigates the correlation between Bitcoin returns and sentiment scores derived from finance-related news headlines. The aim is to understand how emotional cues in trusted news sources influence cryptocurrency market behavior, particularly during volatile periods such as the Covid-19 pandemic. A total of 309,000 news articles were filtered using finance-specific terms, and sentiment analysis was performed on the headlines using three transformer-based models: FinBERT, XLNet, and DistilRoBERTa. The resulting sentiment scores were then correlated with Bitcoin returns using the Ordinary Least Squares (OLS) regression method to model their relationship. The study established a significant relationship between Bitcoin returns and the variance of sentiment scores of financial news headlines at 10% level of significance. Consequently, the realized volatility of Bitcoin had a significant relationship (at 1% level of significance) with the news count i.e., number of news articles released in different news media.<!--Query ID="Q1" Text="Please confirm if the author names are presented accurately and in the correct sequence (given name, middle nameinitial, family name). Author 1 Given name: [Mahesh Kuma] Last name [Morampudi], Author 2 Given name: [Chinmaya Kuma] Last name [Swain], Author 3 Given name: [M. Vijaya] Last name [Prabhagar]. Also, kindly confirm the details in the metadata are correct." Resolved="yes"--> The correlation is particularly evident during the Covid-19 period, suggesting that media sentiment plays a significant role in shaping investor behavior in high-volatility environments. This research is among the few to combine transformer-based sentiment models with financial market analysis during a global crisis. It highlights the impact of emotionally charged financial news on cryptocurrency markets and validates the application of advanced NLP models in financial sentiment analysis. The study is limited to headlines of English-language news articles from renowned publishers and focuses solely on Bitcoin. The influence of multilingual sentiment, social media sentiment, or other cryptocurrencies is not explored and may be considered for future work.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analyzing the impact of the sentiment of Covid-19 news on bitcoin through transformer-based models

  • 期刊论文

  • 发表日期：

    2025年12月30日

查看全文
  • Mahesh Kumar Morampudi,
  • Chinmaya Kumar Swain,
  • M. Vijaya Prabhagar,
  • Sundar Rengasamy

摘要

This study investigates the correlation between Bitcoin returns and sentiment scores derived from finance-related news headlines. The aim is to understand how emotional cues in trusted news sources influence cryptocurrency market behavior, particularly during volatile periods such as the Covid-19 pandemic. A total of 309,000 news articles were filtered using finance-specific terms, and sentiment analysis was performed on the headlines using three transformer-based models: FinBERT, XLNet, and DistilRoBERTa. The resulting sentiment scores were then correlated with Bitcoin returns using the Ordinary Least Squares (OLS) regression method to model their relationship. The study established a significant relationship between Bitcoin returns and the variance of sentiment scores of financial news headlines at 10% level of significance. Consequently, the realized volatility of Bitcoin had a significant relationship (at 1% level of significance) with the news count i.e., number of news articles released in different news media. The correlation is particularly evident during the Covid-19 period, suggesting that media sentiment plays a significant role in shaping investor behavior in high-volatility environments. This research is among the few to combine transformer-based sentiment models with financial market analysis during a global crisis. It highlights the impact of emotionally charged financial news on cryptocurrency markets and validates the application of advanced NLP models in financial sentiment analysis. The study is limited to headlines of English-language news articles from renowned publishers and focuses solely on Bitcoin. The influence of multilingual sentiment, social media sentiment, or other cryptocurrencies is not explored and may be considered for future work.