Einfluss der Medienberichterstattung auf die Risikowahrnehmung : eine empirische Untersuchung der impliziten Risikoprämie auf dem deutschen Aktienmarkt

I developed key indicators for ambiguity based on findings in the fields of ambiguity-, media- and communication research. These indicators describe information states where an increased risk perception exists or may arise. These indicators are: Too little, too much, many undifferentiated, contradictory, negative information. To explain and justify ambiguity aversion, I used psychological theories like the well-established competence hypothesis and aspects of dispute resolution. To measure risk perception in the stock market, I used the residual income model following Claus/Thomas (2001) to calculate the implicit risk premium. In the empirical part of my thesis, I used a comprehensive and detailed dataset of media reporting regarding the general economic situation in Germany and regarding companies of the major German indices DAX, MDAX and SDAX. In total, I included 175.410 news reports during the period 01.01.2001 to 04.14.2004 which were categorized into positive, neutral, and negative tendency to calculate the developed ambiguity indicators. This approach is to my knowledge the broadest and detailed empirical work in this area of research. The influence of media reporting on public perception is established in the context ot the agenda-setting function of the media. But the interrelation between media reporting and creation or amplification of risk perception on the capital market is still disputed. The results of my thesis add to this debate and permit the conclusion that the media influences the risk perception in the market. Main results are: First, there exists a bias in media reporting in the capital market. News reports regarding the general economic situation show a negative bias, whereas media reporting regarding DAX companies is balanced. In contrast, media reporting regarding MDAX and SDAX companies has a positive bias. Second, it is possible to predict the directional change of consensus earnings estimates with content analysis of news reports by using a simple filter system that eliminates noise in the data. Third, the developed ambiguity indicators point to a significant correlation between the tendencies (degree of contradiction, share of undifferentiated news…) of news reports regarding each of the analysed DAX companies and the respective implicit risk premium. This result is of high interest, because this represents evidence of ambiguity on the capital market. The identified ambiguity premium ranges from 7,4 to 14,1 percent of the implicit return of the respective stocks. This result is within range of ambiguity premiums that have been demonstrated in experimental setups on an individual level. Fourth, the prediction of the change of the implicit risk premium is possible based on the ambiguity indicator negative news. The results of statistical tests were significant but very weak.