Text Analysis for IPO firms in Korea : Analysis of Korean Texts in Registration Statements via Machine Learning |
Yongseok Kim, Sung Wook Joh |
한국어 텍스트 분석과 적용 |
김용석, 조성욱 |
서울대학교 |
|
Abstract |
This paper analyzes the texts in registration statements that IPO firms filed when they issued securities in the Korean stock market during June 2009 to December 2017. We classify tones of texts as positive or negative via machine learning. An IPO firm with more negative tones in the risk factor section of registration statement tends to set its final offer price ratio higher than the initial offer price. However, the tone of the texts is not significantly related to the IPO’s initial returns. This is the first paper to analyze tones of texts in the natural language submitted by companies in the Korean financial market. In addition, this paper provides the program used for the analysis of non-standardized text to help scholars analyze Korean texts in the future. |
Key Words:
텍스트 분석,한국어 어조,머신러닝,증권발행신고서,IPO 기업,Text Analysis,Korean Texts,Machine Learning,IPO,Registration Statement |
|