Volkswirtschaftliche Fakultät - Digitale Hochschulschriften der LMU

The present work contributes to the literature on uncertainty by using empirical methods to measure uncertainty and volatility, and empirical and theoretical methods to analyze two transmission channels through which uncertainty is linked to the business cycle. The first chapter looks into the problem of measuring inflation uncertainty and proposes to use common information contained in a variety of different uncertainty proxies. In the second chapter, we develop measures of firm-specific volatility and quantify the effect of heightened volatility on the price setting behavior of firms and analyze whether this link changes the effectiveness of monetary policy. The third chapter compares the effects of heightened idiosyncratic uncertainty on credit spreads in bank-based and market-based financial systems.

This dissertation addresses two different kinds of agents, professional forecasters and manufacturing firms, and their means of dealing with information in forming expectations. The first two parts evaluate the macroeconomic forecasting performance of component-wise boosting, a variable selection device to deal with high-dimensional data. We find that boosting not only outperforms the autoregressive benchmark in most cases, but that it is also a serious competitor to other state-of-the-art methods such as forecast averaging and factor models. However, the forecasting accuracy of boosting depends on the method to determine its key regularisation parameter, the number of iterations, with resampling methods dominating information criteria. The third part discusses the role of news media in the expectation formation process of firms. We investigate empirically on a micro level whether the intensity of news coverage and its evaluative tone have an effect on the expectation updating of enterprises that exceeds the impact of actual economic developments. Our findings suggest that a firm’s propensity to update business expectations increases when media coverage becomes more intense. This volume effect is the stronger the more unusual the economic situation is, and becomes especially relevant in economic downturns. The overall tone of news reporting, however, does not play a role for a firm’s decision whether or not to update its business expectations. The final part of the dissertation analyses empirically on the macro level whether mass media as an important transmitter of aggregate news affect comovement of sectoral business expectations as well as of sectoral production. We do not find evidence for the hypothesis that the more intense media coverage of economy-wide news is, the more do business expectations or production comove across sectors because the latter share a greater common basis of information. However, our results suggest that sectoral business expectations become more synchronised in reaction to a negative news tone shock, which is also reflected in a delayed increase of sectoral output comovement.