A0778
Title: On business confidence as an indicator for industrial production: Evidence from the EC survey
Authors: Marco Malgarini - ANVUR (Italy) [presenting]
Stefano Fantacone - CER (Italy)
Petya Garalova - CER (Italy)
Eleonora Mazzoni - CER (Italy)
Abstract: Business surveys are usually considered as good indicators for industrial production, being not revised, not in need to be filtered and released in advance with respect to quantitative data. However, evidence of a possible break in the relationship among survey data and industrial production (IP) has recently emerged: we look closely at this relationship using Eurostat and EC data for EU countries, analyzing the rolling correlation among soft and hard data and looking at coherence of the two series at cyclical turning points. We also check for the capability of business surveys of causing IP in the sense of Granger and evaluate the evolution of this relationship over time with rolling methods. Emerging differences in cyclical behavior may be either linked to statistical problems in sampling selection during the crisis, or to changes in the way agents form expectations. If long term perceptions of normal levels of output are lower than in the past, it is well possible for opinion variables to show a favorable trend even if the underlying quantitative variable does not show remarkable changes.
