Industrial Confidence Indicator (ICI) is a composite indicator derived from Business Survey in manufacturing industry. ICI is calculated (in accordance with the Joint Harmonised EU Methodology of Business and Consumer Survey) as a simple arithmetic average of seasonally adjusted balances of three variables: production expectations, order books and stocks of finished products (the last with inverted sign).

Business survey questioner usually has three replay options (positive, equal, negative), so a vector X for each question (variable) contains relative frequencies of this three answer options. The most common way of presenting business survey data is the balance. If P, E and M denote percentages of respondents' chosen options: positive, equal and negative (respectively) with the sum equals 100 for each variable X, the balance is the difference between P and M. In Croatian business survey, weighted counting of answers is used. It means that for each firm in survey weighting coefficient is determined in accordance with its size measured with its turnover. Based on empirical results it can be seen that ICI correctly predicts changes in industrial production with one or two quarters ahead.

The aim of this paper is to suggest two modifications in calculation of Croatian ICI. In accordance with the specificity of Croatian economy, the new Indicators were proposed. The first one is calculated as a simple arithmetic average of seasonally adjusted weighted relative frequencies of negative answers on three variables (components of ICI). The second one is a simple arithmetic average of seasonally adjusted weighted relative frequencies of positive answers on the same variables. We analyse the correlations between Croatian industrial production (expressed as a growth rate compared to the same quarter of the previous year) and original ICI and proposed modified indicators.

Various econometric models were estimated using the quarterly data from the first quarter 2005 to the first quarter 2010. The obtained results suggest that the econometric models with new Composite indicators are more appropriate to explain and predict changes in Croatian industrial production with one and two quarters ahead in the period before and during recession.

Keywords: Business survey; Industrial Confidence Indicator; Econometric model; Croatia

Biography: Mirjana Cizmesija was born in Zagreb, Croatia, in 1967. She is an Associate Professor at the Department of Statistics, Faculty of Economics & Business, University of Zagreb, Croatia. Mirjana Cizmesija graduated in 1990 (Diploma in Economics). She received M.Sc. degree in 1993 and Ph.D. degree in 2001, both in Economics. The Title of her M.Sc. thesis was Bayesian Decision Models with application in Marketing and the title of her Ph.D. thesis was Statistical and Methodological base of Business Survey. Her field of scientific interest is Business statistics.