Short-Term Estimates of Euro Area Real GDP by Means of Monthly Data

By Gerhard Rünstler, Franck Sédillot. Social Science Research Network, 2003

180 citations recorded by the source corpus.

The first official data releases of quarterly real GDP for the euro area are published about eight weeks after the end of the reference quarters. Meanwhile, ongoing economic developments must be assessed from various, more readily available, monthly indicators. We examine in the context of univariate forecasting equations to what extent monthly indicators provide useful information for predicting euro area real GDP growth over the current and the next quarter. In particular, we investigate the performance of the equations under the case that the monthly indicators are only partially available within the quarter. For this purpose, we use time series models to forecast the missing observations of monthly indicators. We then examine GDP forecasts under different amounts of monthly information. We find that already a limited amount of monthly information improves the predictions for current-quarter GDP growth to a considerable extent, compared with ARIMA forecasts. JEL Classification: C22, C53

Subfields: Economics and Econometrics

View Paper: https://www.semanticscholar.org/paper/ab0665376134741f2b4867235a017c0184b9e4c1

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

Citations in the corpus

Introducing the Euro-STING: Short Term INdicator of Euro Area Growth (2010, methodology): “As in these proposals, we diverge from the euro area univariate bridge equations employed by Runstler and Sedillot (2003) and Diron (2006) and from those which try to measure high-frequency objects (as real-time activity) on a daily or hourly basis, such as Aruoba, Diebold and Scotti (2009).”
Macroeconomic Forecasting With Mixed-Frequency Data (2008, methodology): “output growth often include AR terms, as in the ADL models of Stock and Watson (2003). Including AR dynamics in models that sample the explanatory variables at a higher frequency clearly is desirable.”
A Look into the Factor Model Black Box: Publication Lags and the Role of Hard and Soft Data in Forecasting GDP (2007, unknown): “Studies in general report that soft data contain little information beyond the real activity data, but they mostly use quarterly data or ignore publication lags (e.g. Rünstler and Sédillot, 2003; Forni et al., 2003; Stock and Watson, 2003; Banerjee et al., 2005).”
Maximum Likelihood Estimation of Factor Models on Data Sets with Arbitrary Pattern of Missing Data (2010, methodology): “This is the case for e.g. bridge equation models (see e.g Rünstler and Sédillot, 2003, for the implementation of bridge equations to forecast euro area GDP).”
Short-term forecasting of GDP using large monthly datasets - A pseudo real-time forecast evaluation exercise. NBB Working Papers. No. 133, 17 June 2008 (2008, methodology): “However, following Rünstler and Sédillot (2003) and Giannone, Reichlin and Small (2005) we take account of publication lags in the individual monthly series and consider a sequence of forecasts to replicate the fl ow of monthly information that arrives within a quarter.”
Early Estimates of Euro Area Real GDP Growth: A Bottom Up Approach from the Production Side (2008, background): “The former class of models can probably be described as the most frequently used ones (for examples see e.g. Rünstler and Sédillot (2003) and Diron (2008)), but more recently the latter type of models has gained increasing popularity as well (see e.g. Angelini, Camba Mendez, Giannone, Reichlin and…”