Do Newspaper Articles Predict Aggregate Stock Returns?

Abstract

We analyze whether newspaper content can predict aggregate future stock returns. Our study is based on articles published in the Handelsblatt, a leading German financial newspaper, from July 1989 to March 2011. We summarize newspaper content in a systematic way by constructing word-count indices for a large number of words. Word-count indices are instantly available and potentially valuable financial indicators. Our main finding is that newspaper articles have provided information valuable for predicting future DAX returns in and out of sample. We find evidence that the predictive power of newspaper content has increased over time, particularly since 2000. Our results suggest that a cluster analysis approach increases the predictive power of newspaper articles substantially.

Do Media Data Help to Predict German Industrial Production?

“…While media indicators are helpful for Switzerland, it is not the case for Germany. Ammann et al (2014) compute the number of mentions of a lexicon of 236 words in the online archive of Handelsblatt with the aim of predicting yields of the German stock market DAX index. They show that newspaper content is a valuable predictor of future DAX returns.…”

Using clustering of rankings to explain brand preferences with personality and socio-demographic variables

“…Forming clusters is our approach to summarize the ranking information. Occasionally, cluster analysis is used for reducing more complex information for use in explanatory methods (e.g., [3]), but the specific requirements of such clusterings are rarely discussed. [35] calls such a pragmatic clustering task "constructive", meaning that the aim is not to find true underlying "real" clusters, but rather to organize the data in a suitable way for the requirements of the specific application.…”

“…Section 2 summarises the relevant literature on market segmentation. Section 3 gives some details about the two empirical studies. Section 4 presents that statistical methodology that was applied here.…”

News media analytics in finance: a survey

“…For example, researchers have studied the influence of news at the firm level (Tetlock et al ., ; Ferguson et al ., ), industry level (Li et al ., ; Smales, ), market level (Tetlock, ; Wei et al ., ), commodity level (Clements and Todorova, ; Maslyuk‐Escobedo et al ., ), between currencies (Nassirtoussi et al ., ), across countries (Griffin et al ., ) and at the global market level (Uhl et al ., ). Time horizons vary from intraday (Groß‐Klußmann and Hautsch, ; Ho et al ., ), daily (Tetlock, ; Garcia, ), weekly (Lu and Wei, ; Sinha, ), monthly (Ammann et al ., ; Cahan et al ., ) and over several years (Hillert et al ., ; Kraussl and Mirgorodskaya, ).…”

“…feature transformation . Examples include Latent Semantic Indexing (LSI) (Tetlock, ), hierarchical clustering (Ammann et al ., ), probabilistic LSI, latent Dirichlet allocation and linear discriminant analysis. See Lee et al .…”