Forecasting Economic Growth in the Euro Area During the Great Moderation and the Great Recession

Abstract

, as well as an anonymous referee for helpful comments, and Nikita Perevalov and Domenico Giannone for sharing elements of their Matlab code. Abstract 4 Non-technical summary 5 © European

The Role of Survey Data in Nowcasting Euro Area GDP Growth

“…Moreover, in the light of the results in Boivin and Ng (2005), static factors estimated via principal components seem to perform systematically better when more complicated but realistic error structures are considered. In addition, imposing an autoregressive structure may induce some rigidities in the model structure, which lead to inaccurate forecasts in the presence of sudden changes in the target series over the forecast horizon (see, for example, Lombardi and Maier, 2011). Moreover, Alvarez et al (2012) show that dynamic methods to estimate factors result in similar problems to the static one when the set of predictors is large, corroborating our choice of focusing on the static method alone as the 'representative method' of the extraction of factors in order to simplify result reporting.…”

“…Focusing on the required data transformation (see middle part of Table I), we follow the practice applied by comparable publications. As in Lombardi and Maier (2011), Giannone et al (2009) and Bulligan et al (2012), all survey variables are treated as stationary in levels, so that we do not impose transformations on them. Real activity data series, by contrast, are expressed as percentage changes, apart from the two unemployment rate series which are differenced, this being in line with the approach taken by Lombardi and Maier (2011), as well as Bulligan et al (2012).…”

“…As in Lombardi and Maier (2011), Giannone et al (2009) and Bulligan et al (2012), all survey variables are treated as stationary in levels, so that we do not impose transformations on them. Real activity data series, by contrast, are expressed as percentage changes, apart from the two unemployment rate series which are differenced, this being in line with the approach taken by Lombardi and Maier (2011), as well as Bulligan et al (2012). In respect of financial data, we follow Bulligan et al (2012), expressing commodity prices, monetary aggregates and exchange rates in percentage changes, bond yields and interest rates in first differences while spreads (as well as stock market volatility indices) are in levels.…”

Influence of the Intermediate Goods Market on the Success of Economic Activity Forecasts

“…Vzhledem k časovému odstupu opatření a výsledných efektů je pro hospodářské politiky klíčová představa o stavu ekonomiky v budoucích obdobích odvozované od znalosti budoucí úrovně HDP závislé na současných proměnných. Otázka volby informační základny pro projekci ekonomické výkonnosti je řešena buď omezeným datovým fondem zejména v podobě indexu nákupních manažerů (PMI), či širokým datovým fondem, jež je vstupem pro dynamický faktorový model (Lombardi-Maier, 2011).…”

“…Vzhledem k měsíční periodicitě PMI fl exibilněji reaguje na náhlé změny ve srovnání s modelovým přístupem. I přes svoji jednoduchost PMI poskytuje v případě eurozóny a jejích členů spolehlivější základnu pro projekci HDP ve srovnání s přesností dynamických modelů (Lombardi-Maier, 2011).…”