Abstract:
While knowing there is a financial distress when you see it might be true, it is not particularly helpful. Indeed, central banks have an interest in understanding more systematically how their communication affects the markets, not least in order to avoid unnecessary volatility; the markets for their part have an interest in better deciphering the message of central banks, especially of course with regard to the conduct of future monetary policy. In this paper we use a novel approach rooted in textual analysis to begin to address these issues. Building on previous work from textual analysis, we are able to use quantitative methods to help identify and measure financial stress. We apply the techniques to the European Central Bank's Monthly Bulletin and show that the results give a much more complete and nuanced picture of market distress than those based only on market data and may help improve how the Central Bank's communication is designed and understood.