shrinkage in the time varying parameter model framework using the r package shrinktvp

• DOI: 10.48550/arxiv.1907.07065
• Publication Date: 2021-01-01
• Authors: Knaus, Peter; Bitto-Nemling, An...; Cadonna, Annalisa; Frühwirth-Schnatt...
• Keywords: time-varying parameter (TVP) models; FOS: Computer and information sciences; 101018 Statistik; 502025 Ökonometrie; Bayesian inference; Econometrics (econ.EM); Statistics - Computation; 01 natural sciences; FOS: Economics and business; QA76.75-76.765; markov chain monte carlo (mcmc); Gibbs sampler; Markov chain Monte Carlo (MCMC); gibbs sampler; 0101 mathematics; log predictive density scores; Computation (stat.CO); Economics - Econometrics; 102022 Softwareentwicklung; 101018 Statistics; bayesian inference; Statistics; 102022 Software development; HA1-4737; Bayesian inference, Gibbs sampler, Markov chain Monte Carlo (MCMC), normal-gamma prior, time-varying parameter (TVP), models, log predictive density scores; 502025 Econometrics; normal-gamma prior; HA29-32; time-varying parameter (tvp) models

Time-varying parameter (TVP) models are widely used in time series analysis to flexibly deal with processes which gradually change over time. However, the risk of overfitting in TVP models is well known. This issue can be dealt with using appropriate global-local shrinkage priors, which pull time-varying parameters towards static ones. In this paper, we introduce the R package shrinkTVP (Knaus, Bitto-Nemling, Cadonna, and Fr��hwirth-Schnatter 2019), which provides a fully Bayesian implementation of shrinkage priors for TVP models, taking advantage of recent developments in the literature, in particular that of Bitto and Fr��hwirth-Schnatter (2019). The package shrinkTVP allows for posterior simulation of the parameters through an efficient Markov Chain Monte Carlo (MCMC) scheme. Moreover, summary and visualization methods, as well as the possibility of assessing predictive performance through log predictive density scores (LPDSs), are provided. The computationally intensive tasks have been implemented in C++ and interfaced with R. The paper includes a brief overview of the models and shrinkage priors implemented in the package. Furthermore, core functionalities are illustrated, both with simulated and real data.