Authors: Hiroshi Ishijima; Takuro Kazumi; Akira Maeda
Addresses: Graduate School of International Accounting, Chuo University, 1-18 Ichigaya-Tamachi, Shinjuku-ku, Tokyo 162-8478, Japan ' CyberAgent, Inc., Akihabara Dai Building, Sotokanda 1-18-13, Chiyoda-ku, Tokyo 101-8608, Japan ' Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro, Tokyo 153-8902, Japan
Abstract: This study analyses the sentiment towards the Japanese economy that might appear in daily news articles between 1 January 2007 and 30 September 2012. To quantify such a sentiment, we created an index that accounts for the frequency of occurrence of words that affirmatively or negatively describe the current economic situation. Articles were taken from the Nikkei, a popular newspaper in Japan comparable with the Wall Street Journal in the USA. Using a cutting-edge text mining technique, we counted the numbers of 'positive' as well as 'negative' words in the newspaper articles. Constructing a daily summary index, we then performed statistical analysis to examine correlations between the sentiment index and Tokyo Stock Exchange prices. One interesting finding is that the index significantly predicts stock prices of three days in advance.
Keywords: sentiment analysis; text mining; stock price predictability; Japan; stock markets; newspaper articles; Tokyo Stock Exchange; stock prices.
Global Business and Economics Review, 2015 Vol.17 No.3, pp.237 - 255
Received: 08 May 2021
Accepted: 12 May 2021
Published online: 21 May 2015 *