Co-integration and common trends analysis with score-driven models : an application to the federal funds effective rate and US inflation rate

Co-integration and common trends are studied for time series variables, by introducing the new t-QVARMA (quasi-vector autoregressive moving average) model. t-QVARMA is an outlier-robust nonlinear score-driven model for the multivariate t-distribution. In t-QVARMA, the I(0) and I(1) components of the variables are separated in a way that is similar to the Granger-representation of VAR models. The relationship between the co-integrated federal funds effective rate and United States (US) inflation rate variables is studied for the period of July 1954 to January 2019. The in-sample statistical and out-of-sample forecasting performances of t-QVARMA are superior to those of the classical Gaussian-VAR model ; Blazsek and Licht acknowledge funding from Universidad Francisco Marroquín. Escribano acknowledges funding from Ministerio de Economía, Industria y Competitividad (ECO2016-00105-001 and MDM 2014-0431) and Comunidad de Madrid (MadEco-CM S2015/HUM-3444).