Macroeconomic forecasting and variable ordering in multivariate stochastic volatility models

Arias, Jonas E.; Rubio-Ramı́rez, Juan Francisco; Shin, Minchul

doi:10.1016/j.jeconom.2022.04.013

Cited by 15 publications

(3 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The fact that quantile regions are sensitive to alternative orderings is one limitation of Wei (2008)'s approach and reflects the difficulty of extending the concept of quantiles to a multivariate setting, for which a clear benchmark has yet to emerge. We refer to Section 4 for a more thorough discussion of the literature. Time‐varying parameter VAR models with stochastic volatility are also not ordering invariant, since the order of the variables affects the posterior distribution of the model parameters (see, e.g., Arias et al 2022; Cogley & Sargent, 2005; Primiceri 2005). Arias et al (2022) show that the ordering does not affect the point prediction but affects substantially the standard deviation of the predictive densities.…”

Section: Quantile Vector Autoregressionmentioning

confidence: 99%

“…We refer to Section 4 for a more thorough discussion of the literature. Time‐varying parameter VAR models with stochastic volatility are also not ordering invariant, since the order of the variables affects the posterior distribution of the model parameters (see, e.g., Arias et al 2022; Cogley & Sargent, 2005; Primiceri 2005). Arias et al (2022) show that the ordering does not affect the point prediction but affects substantially the standard deviation of the predictive densities. The reason OLS VAR are the ordering invariant is that the expected value is a linear operator.…”

Section: Quantile Vector Autoregressionmentioning

confidence: 99%

See 1 more Smart Citation

Forecasting and stress testing with quantile vector autoregression

Chavleishvili,

Manganelli

2023

J of Applied Econometrics

View full text Add to dashboard Cite

SummaryA quantile vector autoregressive (VAR) model, unlike standard VAR, traces the interaction among the endogenous random variables at any quantile. Quantile forecasts are obtained by factorizing the joint distribution in a recursive structure but cannot be obtained from reduced form estimation. Identification strategies and structural quantile impulse response functions are derived as generalization of the VAR model. The model is estimated using real and financial variables for the euro area. The dynamic properties of the system change across quantiles. This is relevant for stress testing exercises, whose goal is to forecast the tail behavior of the economy when hit by large financial and real shocks.

show abstract

Section: Quantile Vector Autoregressionmentioning

confidence: 99%

Section: Quantile Vector Autoregressionmentioning

confidence: 99%

Forecasting and stress testing with quantile vector autoregression

Chavleishvili,

Manganelli

2023

J of Applied Econometrics

View full text Add to dashboard Cite

show abstract

“…We accordingly rely on a recursive structure to identify orthogonal disturbances, with the results depending on the chosen order -see, for example,Arias et al (2023) orChan et al (2023). Placing inflation before money growth in the system allows the latter to react contemporaneously to the former (but not the other way around).…”

mentioning

confidence: 99%

A note of caution on the relation between money growth and inflation

Berger,

Karlsson,

Österholm

2023

Scottish J Political Eco

View full text Add to dashboard Cite

We assess the bivariate relation between money growth and inflation in the euro area and the United States using hybrid time‐varying parameter Bayesian VAR models. Model selection based on marginal likelihoods suggests that the relation is statistically unstable across time in both regions. The effect of money growth on inflation weakened notably after the 1980s before strengthening after 2020. There is evidence that this time variation is related to the pace of price changes, as we find that the maximum impact of money growth on inflation is increasing in the trend level of inflation. These results caution against asserting a simple, time‐invariant relationship when modeling the joint dynamics of monetary aggregates and consumer prices.

show abstract

Fast and order‐invariant inference in Bayesian VARs with nonparametric shocks

Huber,

Koop

2024

J of Applied Econometrics

View full text Add to dashboard Cite

SummaryThe shocks that hit macroeconomic models such as Vector Autoregressions (VARs) have the potential to be non‐Gaussian, exhibiting asymmetries and fat tails. This consideration motivates the VAR developed in this paper that uses a Dirichlet process mixture (DPM) to model the reduced‐form shocks. However, we do not follow the obvious strategy of simply modeling the VAR errors with a DPM as this would lead to computationally infeasible Bayesian inference in larger VARs and potentially a sensitivity to the way the variables are ordered in the VAR. Instead, we develop a particular additive error structure inspired by Bayesian nonparametric treatments of random effects in panel data models. We show that this leads to a model that allows for computationally fast and order‐invariant inference in large VARs with nonparametric shocks. Our empirical results with nonparametric VARs of various dimensions show that nonparametric treatment of the VAR errors often improves forecast accuracy and can be used to analyze the changing transmission of US monetary policy.

show abstract

Macroeconomic forecasting and variable ordering in multivariate stochastic volatility models

Cited by 15 publications

References 30 publications

Forecasting and stress testing with quantile vector autoregression

Forecasting and stress testing with quantile vector autoregression

A note of caution on the relation between money growth and inflation

Fast and order‐invariant inference in Bayesian VARs with nonparametric shocks

Contact Info

Product

Resources

About