Composite Likelihood Methods for Large Bayesian VARs with Stochastic Volatility

Adding multivariate stochastic volatility of a ?exible form to large Vector Autoregressions (VARs) involving over a hundred variables has proved challenging due to computational considerations and over-parameterization concerns. The existing literature either works with homoskedastic models or smaller models with restrictive forms for the stochastic volatility. In this pa- per, we develop composite likelihood methods for large VARs with multivariate stochastic volatility. These involve estimating large numbers of parsimonious models and then taking a weighted average across these models. We discuss various schemes for choosing the weights. In our empirical work involving VARs of up to 196 variables, we show that composite likelihood methods have similar properties to existing alternatives used with small data sets in that they estimate the multivariate stochastic volatility in a ?exible and realistic manner and they forecast comparably. In very high dimensional VARs, they are computationally feasible where other approaches involving stochastic volatility are not and produce superior forecasts than natural conjugate prior homoskedastic VARs.

[1]  Massimiliano Marcellino,et al.  Large Bayesian vector autoregressions with stochastic volatility and non-conjugate priors , 2019, Journal of Econometrics.

[2]  J. Chan.  Large Bayesian Vector Autoregressions , 2019, Macroeconomic Forecasting in the Era of Big Data.

[3]  Christian Matthes,et al.  A Composite Likelihood Approach for Dynamic Structural Models , 2018, The Economic Journal.

[4]  Robert F. Engle,et al.  Fitting Vast Dimensional Time-Varying Covariance Models , 2017, Journal of Business & Economic Statistics.

[5]  Florian Huber,et al.  Sparse Bayesian vector autoregressions in huge dimensions , 2017, Journal of Forecasting.

[6]  Todd E. Clark,et al.  Measuring Uncertainty and Its Impact on the Economy , 2016, Review of Economics and Statistics.

[7]  Joshua C. C. Chan,et al.  On the Observed-Data Deviance Information Criterion for Volatility Modeling , 2016 .

[8]  G. Kastner.  Sparse Bayesian time-varying covariance estimation in many dimensions , 2016, Journal of Econometrics.

[9]  Todd E. Clark,et al.  Large Vector Autoregressions with Stochastic Volatility and Flexible Priors , 2016 .

[10]  A. Roche.  Composite Bayesian inference , 2015, 1512.07678.

[11]  Joshua C. C. Chan,et al.  Large Bayesian VARs: A Flexible Kronecker Error Covariance Structure , 2015, Journal of Business & Economic Statistics.

[12]  Michael W. McCracken,et al.  Real-Time Forecasting with a Large, Mixed Frequency, Bayesian VAR , 2015 .

[13]  Joshua C. C. Chan,et al.  Bayesian Model Comparison for Time-Varying Parameter VARs with Stochastic Volatility , 2015 .

[14]  Michael W. McCracken,et al.  FRED-MD: A Monthly Database for Macroeconomic Research , 2015 .

[15]  Allan Timmermann,et al.  Complete subset regressions with large-dimensional sets of predictors , 2015 .

[16]  Dimitris Korobilis,et al.  Model Uncertainty in Panel Vector Autoregressive Models , 2014 .

[17]  Ruey S. Tsay,et al.  High Dimensional Dynamic Stochastic Copula Models , 2014 .

[18]  Domenico Giannone,et al.  Conditional Forecasts and Scenario Analysis with Vector Autoregressions for Large Cross-Sections , 2014, SSRN Electronic Journal.

[19]  Allan Timmermann,et al.  Complete subset regressions , 2013 .

[20]  Marek Jarociński,et al.  Granger-Causal-Priority and Choice of Variables in Vector Autoregressions , 2013, SSRN Electronic Journal.

[21]  G. Koop.  Forecasting with Medium and Large Bayesian VARs , 2013 .

[22]  Todd E. Clark,et al.  Common Drifting Volatility in Large Bayesian VARs , 2012 .

[23]  Michele Lenza,et al.  Prior Selection for Vector Autoregressions , 2012, Review of Economics and Statistics.

[24]  Todd E. Clark.  Real-Time Density Forecasts From Bayesian Vector Autoregressions With Stochastic Volatility , 2011 .

[25]  Dirk P. Kroese,et al.  Handbook of Monte Carlo Methods , 2011 .

[26]  Troy D. Matheson,et al.  Analysing shock transmission in a data-rich environment: a large BVAR for New Zealand , 2010 .

[27]  G. Kapetanios,et al.  Forecasting Government Bond Yields with Large Bayesian Vars , 2010 .

[28]  Luca Onorante,et al.  Short-Term Inflation Projections: A Bayesian Vector Autoregressive Approach , 2010 .

[29]  Dimitris Korobilis.  VAR Forecasting Using Bayesian Variable Selection , 2009 .

[30]  A. Davison,et al.  Bayesian Inference from Composite Likelihoods, with an Application to Spatial Extremes , 2009, 0911.5357.

[31]  Antonello D’Agostino,et al.  Macroeconomic Forecasting and Structural Change , 2009, SSRN Electronic Journal.

[32]  Todd E. Clark.  Real-Time Density Forecasts from VARs with Stochastic Volatility , 2009 .

[33]  Rodney W. Strachan,et al.  On the evolution of the monetary policy transmission mechanism , 2009 .

[34]  G. Kapetanios,et al.  Forecasting Exchange Rates with a Large Bayesian VAR , 2008 .

[35]  J. Geweke,et al.  Optimal Prediction Pools , 2008 .

[36]  F. Diebold,et al.  Comparing Predictive Accuracy , 1994, Business Cycles.

[37]  R. Horn.  Topics in Matrix Analysis , 1991 .

[38]  M. Schervish,et al.  Characterization of Externally Bayesian Pooling Operators , 1986 .

[39]  J. Zidek,et al.  Aggregating opinions through logarithmic pooling , 1984 .

[40]  Deborah Gefang,et al.  Bayesian doubly adaptive elastic-net Lasso for VAR shrinkage , 2014 .

[41]  N. Reid,et al.  AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS , 2011 .

[42]  D. Giannone,et al.  Large Bayesian vector auto regressions , 2010 .

[43]  Dongchu Sun,et al.  Bayesian stochastic search for VAR model restrictions , 2008 .

[44]  S. Hall,et al.  Combining density forecasts , 2007 .

[45]  Marco Del Negro,et al.  Federal Reserve Bank of New York Staff Reports Time-varying Structural Vector Autoregressions and Monetary Policy: a Corrigendum , 2022 .