Hierarchical Shrinkage in Time‐Varying Parameter Models

Belmonte, Miguel Angel Gonzalez; Koop, Gary; Korobilis, Dimitris

doi:10.1002/for.2276

Cited by 99 publications

(39 citation statements)

References 32 publications

(51 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The prior hyperparameters for the persistence of the latent log variances are fixed at a 0 = 10, b 0 = 2.5 for the idiosyncratic volatilities and a 0 = 2.5, b 0 = 2.5 for the factor volatilities; note that the parameters of the superfluous factor are only identified through the prior. The shrinkage hyperparameters are set as in Belmonte et al (2014), i.e. c i = c j = d i = d j = 0.001 for all applicable i and j.…”

Section: Simulation Studiesmentioning

confidence: 99%

Sparse Bayesian time-varying covariance estimation in many dimensions

Kastner

2019

Journal of Econometrics

View full text Add to dashboard Cite

We address the curse of dimensionality in dynamic covariance estimation by modeling the underlying co-volatility dynamics of a time series vector through latent time-varying stochastic factors. The use of a global-local shrinkage prior for the elements of the factor loadings matrix pulls loadings on superfluous factors towards zero. To demonstrate the merits of the proposed framework, the model is applied to simulated data as well as to daily log-returns of 300 S&P 500 members. Our approach yields precise correlation estimates, strong implied minimum variance portfolio performance and superior forecasting accuracy in terms of log predictive scores when compared to typical benchmarks.

show abstract

Section: Simulation Studiesmentioning

confidence: 99%

Sparse Bayesian time-varying covariance estimation in many dimensions

Kastner

2019

Journal of Econometrics

View full text Add to dashboard Cite

show abstract

“…The normal prior on ; which implies a gamma prior on 2 , is more innocuous in that it allows weight around zero. In an in ‡ation forecasting exercise, Belmonte et al (2014) remove the variable selection parameter ( ) and use shrinkage priors on the prior variance of the initial state (e 0 ) and the state variance (e ) in the non-centered speci…cation. The prior uses a Lasso prior for the standard deviation.…”

Section: Approaches To Overparameterizationmentioning

confidence: 99%

Bayesian State Space Models in Macroeconometrics

Chan

Strachan

2020

SSRN Journal

View full text Add to dashboard Cite

State space models play an important role in macroeconometric analysis and the Bayesian approach has been shown to have many advantages. This paper outlines recent developments in state space modelling applied to macroeconomics using Bayesian methods. We outline the directions of recent research, specifically the problems being addressed and the solutions proposed. After presenting a general form for the linear Gaussian model, we discuss the interpretations and virtues of alternative estimation routines and their outputs. This discussion includes the Kalman filter and smoother, and precision based algorithms. As the advantages of using large models have become better understood, a focus has developed on dimension reduction and computational advances to cope with high-dimensional parameter spaces. We give an overview of a number of recent advances in these directions. Many models suggested by economic theory are either non-linear or non-Gaussian, or both. We discuss work on the particle filtering approach to such models as well as other techniques that use various approximations -to either the time t state and measurement equations or to the full posterior for the states -to obtain draws.

show abstract

“…The scales and a are specified such to obtain a prior mean of 0.1 2 and 0.01 2 for, respectively, and a . In empirical work we often find strong evidence that intercepts are significantly time-varying while evidence for time-varying autoregressive coefficients is much weaker (Belmonte, Koop and Korobilis, 2014). Against this background, we additionally compare the Full model with a model including stochastic volatility, switching intercept and constant autoregressive coefficients (Intercept and SV henceforth).…”

Section: Alternative Modelsmentioning

confidence: 99%

Constrained interest rates and changing dynamics at the zero lower bound

Bäurle

Kaufmann

Kaufmann³

et al. 2019

Studies in Nonlinear Dynamics &Amp; Econometrics

View full text Add to dashboard Cite

The interaction of macroeconomic variables may change as nominal short-term interest rates approach zero. In this paper, we propose to capture these changing dynamics with a state-switching parameter model which explicitly takes into account that the interest rate might be constrained near the zero lower bound by using a Tobit model. The probability of state transitions is affected by the lagged level of the interest rate. The endogenous specification of the state indicator permits dynamic conditional forecasts of the state and the system variables. We use Bayesian methods to estimate the model and to derive the forecast densities. In an application to Swiss data, we evaluate state-dependent impulse-responses to a risk premium shock identified with sign-restrictions. We provide an estimate of the latent rate, i.e. the rate lower than the constraint on the interest rate level which would be state- and model-consistent. Additionally, we discuss scenario-based forecasts and evaluate the probability of exiting the ZLB region. In terms of log predictive scores and the Bayesian information criterion, the model outperforms a model substituting switching with stochastic volatility and another including intercept switching only combined with stochastic volatility.

show abstract

Hierarchical Shrinkage in Time‐Varying Parameter Models

Cited by 99 publications

References 32 publications

Sparse Bayesian time-varying covariance estimation in many dimensions

Sparse Bayesian time-varying covariance estimation in many dimensions

Bayesian State Space Models in Macroeconometrics

Constrained interest rates and changing dynamics at the zero lower bound

Contact Info

Product

Resources

About