The association between stock price volatility and financial news – a sentiment analysis approach
Purpose The purpose of this study is to develop the dictionary with grammar and multiword structure has to be used in conjunction with sentiment analysis to investigate the relationship between financial news and stock market volatility. Design/methodology/approach An algorithm has been developed for calculating the sentiment orientation and score of data with added information, and the results of calculation have been integrated to construct an empirical model for calculating stock market volatility. Findings The experimental results reveal a statistically significant relationship between financial news and stock market volatility. Moreover, positive (negative) news is found to be positively (negatively) correlated with positive stock returns, and the score of added information of the news is positively correlated with stock returns. Model verification and stock market volatility predictions are verified over four time periods (monthly, quarterly, semiannually and annually). The results show that the prediction accuracy of the models approaches 66% and stock market volatility with a particular trend-predicting effect in specific periods by using moving window evaluation. Research limitations/implications Only one news source is used and the research period is only two years; thus, future studies should incorporate several data sources and use a longer period to conduct a more in-depth analysis. Practical implications Understanding trends in stock market volatility can decrease risk and increase profit from investment. Therefore, individuals or businesses can feasibly engage in investment activities for profit by understanding volatility trends in capital markets. Originality/value The ability to exploit textual information could potentially increase the quality of the data. Few scholars have applied sentiment analysis in investigating interdisciplinary topics that cover information management technology, accounting and finance. Furthermore, few studies have provided support for structured and unstructured data. In this paper, the efficiency of providing the algorithm, the model and the trend in stock market volatility has been demonstrated.