Comparer des méthodes

	Modèle Bayésien de Correction d'Erreur Vectoriel (Bayesian VECM)	Modèle Bayésien ARIMA
Domaine	Économétrie	Économétrie
Famille	Regression model	Regression model
Année d'origine	2002–2005	1970s (ARIMA); Bayesian extension prominent from 1990s
Auteur d'origine	Kleibergen & Paap; Villani	Pole, West & Harrison (Bayesian treatment); Box & Jenkins (ARIMA foundation)
Type	Bayesian multivariate time series model	Bayesian time series model
Source fondatrice	Kleibergen, F., & Paap, R. (2002). Priors, posteriors and Bayes factors for a Bayesian analysis of cointegration. Journal of Econometrics, 111(2), 223–249.	Pole, A., West, M., & Harrison, J. (1994). Applied Bayesian Forecasting and Time Series Analysis. Chapman & Hall. ISBN: 978-0412416903
Alias	Bayesian VECM, B-VECM, Bayesian cointegrated VAR, Bayesian vector error correction	Bayesian ARIMA, BARIMA, Bayesian Box-Jenkins model, Bayesian integrated time series model
Apparentées	5	6
Résumé	The Bayesian VECM combines the classical Vector Error Correction Model — which captures both short-run dynamics and long-run cointegrating relationships among non-stationary multivariate time series — with Bayesian prior distributions over the cointegrating rank and coefficient matrices. This allows principled uncertainty quantification, incorporation of economic theory as priors, and coherent inference even in small samples.	The Bayesian ARIMA model combines the classical Box-Jenkins ARIMA framework with Bayesian inference. Instead of obtaining single point estimates for autoregressive and moving average parameters, it places prior distributions over them and uses observed data to update beliefs into a full posterior distribution, enabling coherent uncertainty quantification and probabilistic forecasting.