COMPOSITE LEADING INDICATOR FOR TURKISH ECONOMIC ACTIVITY

This paper focuses on the construction of a composite leading indicator (CLI) for Turkish economic activity to provide earlier signals of turning points between expansions and slowdowns. For this analysis, Index of Industrial Production (IIP) is used as an indicator of economic activity and a broad set of economic indicators related with IIP are analyzed to find leading indicators that perform well both in forecasting and tracking cyclical developments of the economic activity. From the broad set of series, seven of them are selected as leading indicators. The selected indicators constitute a balanced subset of demand, supply and policy variables. While constructing CLI, growth cycle approach is used and the cyclical pattern of the series are obtained by eliminating seasonal, irregular and trend components via TRAMO/SEATS and HP filter. Selected leading indicators are combined into a composite leading indicator using the optimal weights derived from principal component analysis. It is found that CLI is anticipating the IIP and it performed well both in tracking and predicting previous turning points of IIP. JEL Classification: E32, C1