Edward Herbst

Sequential Monte Carlo Sampling for DSGE Models

We develop a sequential Monte Carlo (SMC) algorithm for estimating Bayesian dynamic stochastic general equilibrium (DSGE) models, wherein a particle approximation to the posterior is built iteratively through tempering the likelihood. Using three examples consisting of an artificial state-space model, the Smets and Wouters (2007) model, and Schmitt-Grohe and Uribes (2012) news shock model we show that the SMC algorithm is better suited for multimodal and irregular posterior distributions than the widely-used random walk Metropolis- Hastings algorithm. We find that a more diffuse prior for the Smets and Wouters (2007) model improves its marginal data density and that a slight modification of the prior for the news shock model leads to drastic changes in the posterior inference about the importance of news shocks for fluctuations in hours worked. Unlike standard Markov chain Monte Carlo (MCMC) techniques, the SMC algorithm is well suited for parallel computing.

Evaluating DSGE model forecasts of comovements

This paper develops and applies tools to assess multivariate aspects of Bayesian Dynamic Stochastic General Equilibrium (DSGE) model forecasts and their ability to predict comovements among key macroeconomic variables. The authors construct posterior predictive checks to evaluate the calibration of conditional and unconditional density forecasts, in addition to checks for root-mean-squared errors and event probabilities associated with these forecasts. The checks are implemented on a three-equation DSGE model as well as the Smets and Wouters (2007) model using real-time data. They find that the additional features incorporated into the Smets-Wouters model do not lead to a uniform improvement in the quality of density forecasts and prediction of comovements of output, inflation, and interest rates.

Effective Monetary Policy Strategies in New Keynesian Models: A Re-Examination

We explore the importance of the nature of nominal price and wage adjustment for the design of effective monetary policy strategies, especially at the zero lower bound. Our analysis suggests that sticky-price and sticky-information models fit standard macroeconomic time series comparably well. However, the model with information rigidity responds differently to anticipated shocks and persistent zero-lower bound episodes - to a degree important for monetary policy and for understanding the effects of fundamental disturbances when monetary policy cannot adjust. These differences may be important for understanding other policy issues as well, such as fiscal multipliers. Despite these differences, many aspects of effective policy strategy are common across the two models: In particular, highly inertial interest rate rules that respond to nominal income or the price level perform well, even when hit by adverse supply shocks or large demand shocks that induce the zero-lower bound. Rules that respond to the level or change in the output gap can perform poorly under those conditions.

SEQUENTIAL MONTE CARLO SAMPLING FOR DSGE MODELS: SEQUENTIAL MONTE CARLO SAMPLING FOR DSGE MODEL

We develop a sequential Monte Carlo (SMC) algorithm for estimating Bayesian dynamic stochastic general equilibrium (DSGE) models, wherein a particle approximation to the posterior is built iteratively through tempering the likelihood. Using three examples consisting of an artificial state-space model, the Smets and Wouters (2007) model, and Schmitt-Grohe and Uribes (2012) news shock model we show that the SMC algorithm is better suited for multi-modal and irregular posterior distributions than the widely-used random walk Metropolis-Hastings algorithm. Unlike standard Markov chain Monte Carlo (MCMC) techniques, the SMC algorithm is well suited for parallel computing.(This abstract was borrowed from another version of this item.)

Estimating (Markov-Switching) VAR Models Without Gibbs Sampling: A Sequential Monte Carlo Approach

Vector autoregressions with Markov-switching parameters (MS-VARs) fit the data better than do their constant-parameter predecessors. However, Bayesian inference for MS-VARs with existing algorithms remains challenging. For our first contribution, we show that Sequential Monte Carlo (SMC) estimators accurately estimate Bayesian MS-VAR posteriors. Relative to multi-step, model-specific MCMC routines, SMC has the advantages of generality, parallelizability, and freedom from reliance on particular analytical relationships between prior and likelihood. For our second contribution, we use SMCs flexibility to demonstrate that the choice of prior drives the key empirical finding of Sims, Waggoner, and Zha (2008) as much as does the data.

Monetary Policy, Real Activity, and Credit Spreads: Evidence from Bayesian Proxy Svars

This paper studies the interaction between monetary policy, financial markets, and the real economy. We develop a Bayesian framework to estimate proxy structural vector autoregressions (SVARs) in which monetary policy shocks are identified by exploiting the information contained in high frequency data. For the Great Moderation period, we find that monetary policy shocks are key drivers of fluctuations in industrial output and corporate credit spreads, explaining about 20 percent of the volatility of these variables. Central to this result is a systematic component of monetary policy characterized by a direct and economically significant reaction to changes in credit spreads. We show that the failure to account for this endogenous reaction induces an attenuation bias in the response of all variables to monetary shocks.