Leading Indicators of Economic Sentiment - an Econometrical Analysis

Abstract

Quality of three most popular Polish leading indicators of economic sentiment has been investigated by wide range of analytical tools. An additional index has been constructed on the base of industry - specific indicators of the Central Statisitcal Office of Poland. Eurostat's coincident index has been used as a benchmark. All of the analysed indicators are capable of forecasting economic sentiment for the next 1-2 months. Econometrical models that involve leading indicators as explanatory variables are of good quality. The phenomenon of cointegration can be observed for all the indicators except the Ipsos one. Economic cycles in those three indicators have the same pattern as in the coincident indicator. Usually phases of expansion are signalised earlier than phases of contraction. Analysis of autocovariance functions proved that patterns of economic cycles in both leading and coincident indicators are equal. The similarity between the leading indicators and the coincident indicator is statistically significant. Forecasts derived from econometrical models that use leading indicators as explanatory variables proved to be of good quality. Moreover, including the leading indicators in the models leads to higher adequacy of the forecasts. As the result of the analysis the adequacy of indicators of IRG SGH and BIEC can be assessed as the highest. (short original abstract)(original abstract)