NEWS OR NOISE? HOW NEWS DRIVES COMMODITY PRICES Completed Research Paper

The efficiency of electronic markets is affected by the availability of information. While some information comprises quantitative data, the majority is of qualitative nature. Although qualitative facts are essential, they are difficult to decode. As a status quo, knowledge on information processing of human agents facing qualitative news is still marginal and mostly unknown. Accordingly, it is crucial to understand how different decision makers process qualitative information in commodity markets. In this paper, we show that sentiment analysis facilitates research in qualitative information processing. We provide empirical evidence that abnormal returns in commodity markets can be explained, up to a large extent, by news sentiment. Further, we contribute to the question of how investors process novel information. Our results suggest that the influence of sentiment is asymmetric – with commodity markets being mostly driven by negative news. In addition, we show that words exhibit strong changes in polarity over time