Modelling Structural Breaks

A key aspect of the financial crisis was the lack of fore-knowledge, so analyses of forecasting in the face of breaks, during breaks and of breaks themselves remain a central focus. Importantly, if agents were to use such devices, a new basis for both economics and empirical models must be forged. New data sources, such as Google Trends, provide disaggregated high-frequency information on a vast scale, both allowing the possibility of improved ‘nowcasts’ for flash estimates of GNP, inflation etc., and providing incipient information on agents’ reactions to changes to help predict breaks and during breaks.