Publication Type

Master Thesis

Publication Date



Recent years have seen a large increase in the volume of financial news available to investors daily. What has traditionally been restricted to print media has now evolved to include the internet and satellite television as important media sources for financial news. With this overwhelming flow of information available to investors, the impact of financial news on market prices is at best uncertain. In this paper, a computational text-scoring methodology will be employed to uncover behavioral responses by investors to negative news. The empirical methodology employed in this paper will consist of three parts. Firstly, through the General Inquirer (GI) content analysis software, a sentiment score is derived from daily news articles published in the Wall Street Journal. The second part will be an analysis of the sentiment time series which was obtained, where comparison will be made to existing barometers of market sentiment and market volatility. The final part of the modeling methodology which will be presented is a predictive model of market implied volatility using daily news scores as the main input. In conclusion, it is found that high negative news scores do not necessarily predict negative abnormal returns in the S&P 500 across a 1-day to 5-day window. However, high negative news scores are highly correlated with higher market volatility. Given that the negative news is published prior to the market‟s trading start in the morning; we are able to utilize this information to construct a predictive model of the CBOE VIX index.


text scoring, news effect, stock market

Degree Awarded

MSc in Finance


Portfolio and Security Analysis


Chiyachantana, Chiraphol New

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.