Working Paper: CEPR ID: DP7883
Authors: Marta Babura; Domenico Giannone; Lucrezia Reichlin
Abstract: We define nowcasting as the prediction of the present, the very near future and the very recent past. Key in this process is to use timely monthly information in order to nowcast quarterly variables that are published with long delays. We argue that the nowcasting process goes beyond the simple production of an early estimate and it consists in the analysis of the link between the news in consecutive data releases and the resulting forecast revisions for the target variable. We describe an econometric framework that allows us to mimic, via a coherent statistical model, the judgemental process of nowcasting traditionally conducted in policy institutions and used, alongside the judgemental procedures, in many central banks. To illustrate our ideas, we study the nowcast of euro area GDP in the fourth quarter of 2008.
Keywords: Factor model; Forecasting; News; Nowcasting
JEL Codes: C33; C53; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| timely monthly data (Y10) | accuracy of quarterly GDP forecasts (E37) |
| unexpected news in industrial production (L69) | revisions in GDP estimates (E01) |
| unexpected news in survey data (C83) | revisions in GDP estimates (E01) |
| timely monthly data (Y10) | forecast revision (C53) |
| new data releases (Y10) | nowcast estimates (C51) |
| nowcasting process (C53) | GDP forecasts (F17) |