Premium
Corrélation tendance‐cycle, discontinuité, et estimation de la tendance et du cycle dans le PIB canadien.
Author(s) -
Basistha Arabinda
Publication year - 2007
Publication title -
canadian journal of economics/revue canadienne d'économique
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.773
H-Index - 69
eISSN - 1540-5982
pISSN - 0008-4085
DOI - 10.1111/j.1365-2966.2007.00422.x
Subject(s) - bivariate analysis , univariate , econometrics , monte carlo method , business cycle , economics , recession , shock (circulatory) , statistics , multivariate statistics , mathematics , macroeconomics , medicine
. Univariate correlated trend cycle models are highly sensitive to the specifications of breaks in the data. This paper argues, using Monte Carlo experiments, that a bivariate correlated unobserved components (UC) framework with breaks delivers substantially more accurate results for the trend‐cycle parameters than the corresponding univariate frameworks in a finite sample size. The paper estimates stochastic trend and cyclical fluctuations in Canada from a bivariate UC model. Results show a fairly volatile stochastic trend after the drift break and the negative trend‐cycle shock correlation are accounted for. The estimated cyclical component is large, persistent, and consistent with ECRI denoted Canadian recessions.