CMStatistics 2017: Start Registration
View Submission - CFE
Title: Cointegration, cobreaking and cotrending for trending time series Authors:  Josep Lluis Carrion-i-Silvestre - Universitat de Barcelona (Spain) [presenting]
Abstract: A set of variables with two types of non-stationary features, stochastic trends and broken linear trends is considered. We develop tests that can be used to see if there exists a linear combination of these variables under which the non-stationary features can be cancelled out. We provide a comprehensive treatment of cointegration testing with structural breaks that is missing in the literature. In particular, the first test we develop can be used to see if stochastic trends can be eliminated and thus cointegration holds, regardless of whether structural breaks in linear trends are eliminated. The second test can be used to see if both stochastic trends and breaks in linear trends are simultaneously removed and thus both cointegration and cobreaking simultaneously hold. The third test can be used to see if not only breaks in linear trends but also linear trends themselves are eliminated along with stochastic trends and thus both cointegration and cotrending hold. We provide the asymptotic null distributions of proposed test statistics and some Monte Carlo simulation results to assess the adequateness of our asymptotic distributions in samples with common sizes.