Real-Time Simulation of Alternative Specification of Euro Area Turning Points Detection

• Activity Number: 464
• Type: Topic Contributed
• Date/Time: Wednesday, August 4, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Gian Luigi Mazzi*+ and Monica Billio and Rosa Ruggeri Cannata
• Companies: EUROSTAT and Università di Venezia and EUROSTAT
• Keywords: Coincident Indicators ; turning points detection ; non linear modelling ; real-time simulation

Abstract

In the recent years, Eurostat has developed a couple of coincident indicators for dating euro area turning points for classical business cycle and growth cycle. Those indicators are based on univariate Markov Switching models fitted to each component and the final result is a weighted average of the filtered probabilities returned by each model. This paper presents a real time simulation of the two indicators against alternative specifications based on SETAR models. The simulation is carried out using the PEEIs vintage database of Eurostat. The main outcome of the paper is that while SETAR models return a more timely detection of turning points, especially for the classical business cycle, they produce a higher number of false signals than the Markov Switching models based indicators. The application of a censoring rule does not change significantly the results.