Investment with New Sentiment Analysis in Japanese Stock Market: Expert knowledge can still outperform ChatGPT
This extensive paper proposes novel market-driven sentiment analysis methods based on customized financial lexicons for Japanese stocks and futures.
What it examines
The study develops a sentiment analysis method using expert-derived keywords and market return data to enhance stock investment decisions in Japan. It compares traditional methods, such as MeCab and ChatGPT, with its novel custom approach, addressing text complexities in non-English financial news.
What it concludes
The paper shows that integrating expert knowledge with market-based sentiment analysis outperforms generic language models. Results suggest applications in enhancing stock investment strategies and risk management. Future work may refine dictionary methods and extend use to other markets.
Evidence objects
The study reveals that integrating expert financial insights with 11-year market return data significantly enhances sentiment analysis, merging tailored finance-specific lexicons with advanced segmentation to improve investment decision-making in Japan.
key_findings bullet 1 · key_findings · validation V0
Surprisingly, a custom-built polarity dictionary using finance-specific keywords and refined segmentation techniques outperforms generic methods, delivering superior predictive accuracy for Japanese stock market trends and enhancing investment sentiment analysis remarkably.
key_findings bullet 2 · key_findings · validation V0
The research contributes innovative methods by extracting key financial terms and incorporating market premiums, while highlighting trends like mean reversion after negative news, despite noted segmentation challenges for future improvement.
key_findings bullet 3 · key_findings · validation V0
Integrating market return data with sentiment analysis, this study constructs innovative polarity dictionaries that outperform ChatGPT for the Japanese stock market. Blending expert knowledge and textual analysis, the paper introduces a novel methodology $$\text{Innovative-Approach}$$, directly incorporating financial market dynamics, thus offering a compelling perspective for financial news and sentiment evaluation.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
- … become a major research topic as natural language processing (NLP) (eg, [6–11]). … , which can be easily applied to stock markets in other countries with slight modifications…
Source row: 1129 · abstract type: snippet