Nowcasting , Business Cycle Dating

A modelling framework is proposed in which the real time informational context of decision-making is properly reflected. Comparisons are drawn with ‘standard’ estimated models that incorrectly omit market-informed insights on future macroeconomic conditions and inappropriately incorporate information that was not available at the time. An analysis of quarterly US data 1968q4-2006q1 shows that neither diagnostic tests applied to the standard models nor typical impulse response analysis are able to expose the misspecification clearly. Estimated real time models considerably improve out-of-sample forecasting performance, provide more accurate ‘nowcasts’ of the current state of the macroeconomy and provide more timely indicators of the business cycle. A case study highlights the use of information in recognising the US recessions of 1990q3 — 1991q2 and of 2001q1 — 2001q4.