This paper applies algorithmic analysis to financial market text-based data to assess how narratives and sentiment might drive financial system developments. We find changes in emotional content in narratives are highly correlated across data sources and show the formation (and subsequent collapse) of exuberance prior to the global financial crisis. Our metrics also have predictive power for other commonly used indicators of sentiment and appear to influence economic variables. A novel machine learning application also points towards increasing consensus around the strongly positive narrative prior to the crisis. Together, our metrics might help to warn about impending financial system distress.
