A robust Bayesian dynamic linear model for Latin-American economic time series: “the Mexico and Puerto Rico cases”

Abstract

The traditional time series methodology requires at least a preliminary transformation of the data to get stationarity. On the other hand, robust Bayesian dynamic models (RBDMs) do not assume a regular pattern or stability of the underlying system but can include points of statement breaks. In this paper we use RBDMs in order to account possible outliers and structural breaks in Latin-American economic time series. We work with important economic time series from Puerto Rico and Mexico. We show by using a random walk model how RBDMs can be applied for detecting historic changes in the economic inflation of Mexico. Also, we model the Consumer Price Index, the Economic Activity Index and the total number of employments economic time series in Puerto Rico using local linear trend and seasonal RBDMs with observational and states variances. The results illustrate how the model accounts the structural breaks for the historic recession periods in Puerto Rico.Keywords Robust Bayesian dynamic model Á Outliers and structural breaks Á Latin-American time series Á Consumer Price Index Á Economic Activity Index Á Total number of employments JEL Classification C11 Á C40 Á G17 Á N16

Modified Recursive Bayesian Algorithm for Estimating Time-Varying Parameters in Dynamic Linear Models

“…ϕ is assumed to be distributed inverse-gamma a priori, while we estimate Ω t via discounting method which is explained later in this section. The difference between this model and the one stated in Fuquene et al (2013) is that the observational variance is presumed to be fixed and the evolution variance is estimated by the method of discounting unlike the use of Wishart prior which is common in literature. Also, in contrast to the Box-Jenkins methodology, which still plays an important role in time series analysis today, the specified dynamic linear model approach allows for structural analysis of univariate as well as multivariate problems without initial differencing or log transformation of the observed series.…”

“…In particular, Fuquene et al (2013) proposed a dynamic linear model which is specified by a normal prior distribution for the p-dimensional state vector for macroeconomic modeling with prior θ 0 ) as follows:…”

“…Consequently, researchers require a better way of structuring the evolution variance which previous studies have failed to effectively address. The aim of this study, therefore, was to modify an existing RBA of Fúquene et al (2015) for estimating TVPs in DLMs. Discounting is proposed as an alternative way of coping with the system evolution variance of economic series in order to portray a clearer picture of the volatility of the parameters in the model under study over time.…”

Dynamic Linear Models for analysing time series data in coastal environmental monitoring

“…Petris et al (2009) described a model for outliers and structural breaks that allows for the identification and treatment of outliers and structural changes for all model components including seasonality in a single estimation process. Fúquene et al (2015) published the results of a robust version of this type of model.…”