Leading indicators on construction and retail trade sectors based on ISAE survey data

According to the Commission recommendations, ISAE has recently restructured the methodological framework of its survey on firms operating in the Italian construction sector and in retail trade. The innovations specifically regard the sampling design and the weight system for both sectors; this last revision, in particular, allowed the reconstruction of the ISAE historical series. In the light of the changes introduced, first of all the aim of this paper is to analyze the cyclical features and to evaluate the "leading" performances of the new ISAE series with respect to the quantitative ISTAT data. Finally, we are going to build a "leading indicator" for both construction and retail trade in Italy. We first apply the NBER methodology in order to establish the main cyclical features of the series. Then we use cross-correlation analysis to estimate the extent to which the ISAE variables and the ISTAT series are correlated. Later on the Granger causality and out of sample tests were used to evaluate the forecasting performance of the ISAE series. On the basis of the results obtained, we finally build a leading indicator for both sectors, and test its performance comparing the results to those of the confidence index elaborated by ISAE.