Search Results
Working Paper
Capturing Macroeconomic Tail Risks with Bayesian Vector Autoregressions
A rapidly growing body of research has examined tail risks in macroeconomic outcomes. Most of this work has focused on the risks of significant declines in GDP, and has relied on quantile regression methods to estimate tail risks. In this paper we examine the ability of Bayesian VARs with stochastic volatility to capture tail risks in macroeconomic forecast distributions and outcomes. We consider both a conventional stochastic volatility specification and a specification featuring a common volatility factor that is a function of past financial conditions. Even though the conditional ...
Working Paper
Measuring Uncertainty and Its Effects in the COVID-19 Era
We measure the effects of the COVID-19 outbreak on uncertainty, and we assess the consequences of the uncertainty for key economic variables. We use a large, heteroskedastic vector autoregression (VAR) in which the error volatilities share two common factors, interpreted as macro and financial uncertainty. Macro and financial uncertainty are allowed to contemporaneously affect the macroeconomy and financial conditions, with changes in the common component of the volatilities providing contemporaneous identifying information on uncertainty. The model includes additional latent volatility ...
Working Paper
Tail Forecasting with Multivariate Bayesian Additive Regression Trees
We develop multivariate time series models using Bayesian additive regression trees that posit nonlinearities among macroeconomic variables, their lags, and possibly their lagged errors. The error variances can be stable, feature stochastic volatility, or follow a nonparametric specification. We evaluate density and tail forecast performance for a set of US macroeconomic and financial indicators. Our results suggest that the proposed models improve forecast accuracy both overall and in the tails. Another finding is that when allowing for nonlinearities in the conditional mean, ...
Working Paper
Common drifting volatility in large Bayesian VARs
The estimation of large vector autoregressions with stochastic volatility using standard methods is computationally very demanding. In this paper we propose to model conditional volatilities as driven by a single common unobserved factor.> This is justified by the observation that the pattern of estimated volatilities in empirical analyses is often very similar across variables. Using a combination of a standard natural conjugate prior for the VAR coefficients and an independent prior on a common stochastic volatility factor, we derive the posterior densities for the parameters of the ...
Working Paper
Bayesian VARs: specification choices and forecast accuracy
In this paper we examine how the forecasting performance of Bayesian VARs is affected by a number of specification choices. In the baseline case, we use a Normal-Inverted Wishart prior that, when combined with a (pseudo-) iterated approach, makes the analytical computation of multi-step forecasts feasible and simple, in particular when using standard and fixed values for the tightness and the lag length. We then assess the role of the optimal choice of the tightness, of the lag length and of both; compare alternative approaches to multi-step forecasting (direct, iterated, and ...
Working Paper
Large Vector Autoregressions with Stochastic Volatility and Flexible Priors
Recent research has shown that a reliable vector autoregressive model (VAR) for forecasting and structural analysis of macroeconomic data requires a large set of variables and modeling time variation in their volatilities. Yet, there are no papers jointly allowing for stochastic volatilities and large datasets, due to computational complexity. Moreover, homoskedastic VAR models for large datasets so far restrict substantially the allowed prior distributions on the parameters. In this paper we propose a new Bayesian estimation procedure for (possibly very large) VARs featuring time varying ...
Working Paper
The financial accelerator mechanism: does frequency matter?
We use mixed-frequency (quarterly-monthly) data to estimate a dynamic stochastic general equilibrium model embedded with the financial accelerator mechanism à la Bernanke et al. (1999). We find that the financial accelerator can work very differently at monthly frequency compared to quarterly frequency; that is, we document its inversion. That is because aggregating monthly data into quarterly data leads to large biases in the estimated quarterly parameters and, as a consequence, to a deep change in the transmission of shocks.
Working Paper
Tail Forecasting with Multivariate Bayesian Additive Regression Trees
We develop novel multivariate time series models using Bayesian additive regression trees that posit nonlinear relationships among macroeconomic variables, their lags, and possibly the lags of the errors. The variance of the errors can be stable, driven by stochastic volatility (SV), or follow a novel nonparametric specification. Estimation is carried out using scalable Markov chain Monte Carlo estimation algorithms for each specification. We evaluate the real-time density and tail forecasting performance of the various models for a set of US macroeconomic and financial indicators. Our ...
Working Paper
Capturing Macroeconomic Tail Risks with Bayesian Vector Autoregressions
A rapidly growing body of research has examined tail risks in macroeconomic outcomes. Most of this work has focused on the risks of significant declines in GDP, and it has relied on quantile regression methods to estimate tail risks. Although much of this work discusses asymmetries in conditional predictive distributions, the analysis often focuses on evidence of downside risk varying more than upside risk. We note that this pattern in risk estimates over time could obtain with conditional distributions that are symmetric but subject to simultaneous shifts in conditional means (down) and ...
Working Paper
Empirical simultaneous prediction regions for path-forecasts
This paper investigates the problem of constructing prediction regions for forecast trajectories 1 to H periods into the future - a path forecast. We take the more general view that the null model is only approximative and in some cases it may be altogether unavailable. As a consequence, one cannot derive the usual analytic expressions nor resample from the null model as is usually done when bootstrap methods are used. The paper derives methods to construct approximate rectangular regions for simultaneous probability coverage which correct for serial correlation. The techniques appear to work ...