دانلود رایگان مقاله عقب رفتن؟ احساسات در مقالات روزنامه ها و قیمت سهام در هلند

abstract

This study investigates emotions in Dutch newspaper articles and their effects on, and responses to, opening prices of 21 stocks listed on the Amsterdam Exchange index for twelve years (2002–2013). With regard to the financial context, we employed a selection of the Dutch Linguistic Inquiry and Word Count dictionary to automatically content analyze emotional tone in news articles (N = 128,507). Neither positive nor negative emotions in news articles show consistent effects on the opening prices of stocks the following day. Granger causality tests suggest, however, that newspapers rather reflect movements on the stock market the following days by using more negative emotional words after an increase in the change of the opening prices.

5. Discussion

Reoccurring short-term fluctuations of stock market prices as a reaction to unexpected news, and newspapers reporting on these events the other day, suggest a strong interrelation between media coverage and the stock market. In trying to explain this reciprocal relationship, the results of this study give a comprehensive overview to what extent emotional words in newspaper articles affect opening prices of stocks in the Netherlands, and vice versa. Concerning the salience of stocks in the news, we could not find media attention to explain a considerable amount of the change in the opening prices of stocks the following days. The few findings suggest media attention to have a small negative effect on the opening prices, whereas in the reversed direction no consistent effects could be identified. Similarly, we only evidenced indications that an increase in both positive and negative emotional words in articles dealing with a stock leads to a decrease of the opening prices of the stocks. In this regard, it is also not surprising that we did not identify a moderation effect, implying the absence of a stronger effect of emotions in news articles on stock market prices when these stocks are more prevalent in the media. These findings might contradict previous research (e.g., Akhtar et al., 2012), but are in line with studies that could not detect media to affect stock market prices after all (e.g., Campbell, Turner, & Walker, 2012).