A convenient representation for structural vector autoregressions

2021

Appendix C in the book discussed the estimation of a DSGE model. These estimations are known as structural estimations, because they are built on the basis of a "structural model". An alternative way of doing macroeconomic forecasting is to use an approach that is agnostic about the model and works through unveiling the statistical properties of the data itself. This is the approach that was pioneered by Chris Sims and earned him, eventually, the Nobel Prize in Economics. This appendix will introduce you to the estimation of VARs and SVARs. In order to do so we will do three things. First, we will introduce the concepts from a theoretical point of view. Then we will review the concepts of reduced form vector autoregressions (VARs) and structural vector autoregressions (SVARs). To do so we will refer to a specific case, which is the estimation of these models applied to a two variable system including US GDP and US inflation. As we go along, we will provide the codes in R so that you can apply these methods to your own variables of interest.

Journal of Macroeconomics, 1992

Economics Bulletin, 2009

The purpose of this paper is to develop a new approach allowing us to identify the structural shocks in the SVAR model. This approach ameliorates substantially the decomposition methods of Bernanke (1986) and Bernanke & Mihov (1998) and improves in the same way the identification procedures pioneered by Blanchard & Quah (1989) and Blanchard & Perotti (2002).

2011

Oxford Bulletin of Economics and Statistics, 2015

A growing line of research makes use of structural changes and different heteroskedasticity regimes found in the data in a constructive manner to improve the identification of structural parameters in Structural Vector Autoregressions (SVAR). A standard assumption made in the literature is that the reduced form unconditional error covariance matrix varies while the structural parameters remain constant. Under this condition it is possible to identify the SVAR without needing to resort to theory-driven restrictions. With macroeconomic data, the hypothesis that the structural parameters are invariant to heteroskedasticity regimes is debatable. This paper proposes new identification schemes which exploit simultaneous discrete changes in the reduced form error covariance matrix and the structural parameters, in which different structural models are imposed on different heteroskedasticity regimes. This approach opens up interesting possibilities for practitioners and gives rise to new necessary and sufficient local identification rank conditions which generalize the ones which apply for SVARs with constant parameters. An empirical illustration shows the usefulness of the suggested identification strategy by focusing on a small monetary policy SVAR of the U.S. economy. Considering the case of two heteroskedasticity regimes before and after the 1980s, we estimate the effect of monetary policy shocks on output and inflation through some novel identification schemes.

Modeling Financial Time Series with S-Plus®, 2003

Scottish Journal of Political Economy, 1993

2003

2019

We show that the contemporaneous and longer horizon impulse responses estimated using small-scale Proxy structural vector autoregressions (SVARs) can be severely biased in the presence of information insufficiency. Instead, we recommend the use of a Proxy Factor Augmented VAR (FAVAR) model that remains robust in the presence of this problem. In an empirical exercise, we demonstrate that this issue has important consequences for the estimated impact of monetary policy shocks in the US. We find that the impulse responses of real activity and prices estimated using a Proxy FAVAR are substantially larger and more persistent than those suggested by a small-scale Proxy SVAR.

Econometric Institute Report EI …, 2004

Economic policy decisions are often informed by empirical economic analysis. While the decision-maker is usually only interested in good estimates of outcomes, the analyst is interested in estimating the model. Accurate inference on the structural features of a model, such as cointegration, can improve policy analysis as it can improve estimation, inference and forecast e¢ciency from using that model. However, using a model does not guarantee good estimates of the object of interest and, as it assigns a probability of one to a model and zero to near-by models, takes extreme zero-one account of the 'weight of evidence' in the data and the researcher's uncertainty. By using the uncertainty associated with the structural features in a model set, one obtains policy analysis that is not conditional on the structure of the model and can improve e¢ciency if the features are appropriately weighted. In this paper tools are presented to allow for unconditional inference on the vector autoregressive (VAR) model. In particular, we employ measures on manifolds to elicit priors on subspaces de…ned by particular features of the VAR model. The features considered are cointegration, exogeneity, deterministic processes and overidenti…cation. Two applications -money demand in Australia, and a macroeconomic model of the UK proposed by Garratt, Lee, Persaran, and Shin (2002) are used to illustrate the feasibility of the proposed methods.

Applied Mathematics and Computation, 1986

We discuss the Structural Econometric Modeling and Time Series Analysis (SEMTSA) approach put forward by Zellner and Palm, which provides a synthesis of econometric and time series methods in modeling economic time series. The approach aims at giving guidance for checking the data admissibility of the dynamic specification of a model in its various forms, in particular the transfer function form and the final equation form. We review the SEMTSA approach, discuss recent developments, and briefly compare the SEMTSA with other methodologies for econometric modeling. Finally some remarks are made about problems that remain to be solved.

This paper proposes a Bayesian, graph-based approach to identification in structural vector autoregressive (VAR) models. In our Bayesian graphical VAR (BGVAR) model, the contemporaneous and temporal causal structures of the structural VAR model are represented by two different graphs. We apply an efficient Markov chain Monte Carlo algorithm to estimate jointly the two causal structures and the parameters of the reducedform VAR model. We study the efficiency of the proposed BGVAR model through comparisons with some existing approaches widely used in applied econometrics, such as the Granger-causality approach, the PC algorithm, the Bayesian VAR (BVAR) model and the stochastic search variable selection (SSVS) procedure. The simulation results and the applications to real data show that the BGVAR model produces a better representation of the structural relationships than the Granger-causality approach and the PC algorithm, and has a higher predictive accuracy than the BVAR model and the SSVS procedure. The empirical applications to moderate-dimension time series previously considered in the macroeconomic and financial literature illustrate the effectiveness of BGVAR models and the advantages over alternative approaches. The macroeconomic application contributes to the identification of the relevant structural relationships among 20 US economic variables, providing a useful tool for policy analysis. The application also draws attention to the identification errors originated by the omission of relevant variables when using different inference approaches. The financial application contributes to the recent econometric literature on financial interconnectedness and on graphical models for financial time series analysis. That application finds further evidence of a strong unidirectional linkage from financial to non-financial super-sectors during the 2007-2009 financial crisis and a strong bidirectional linkage between the two sectors during the 2010-2013 European sovereign debt crisis. The financial application also shows that pairwise and conditional Granger-causality approaches, often used in the financial econometrics literature for inferring a network, may overestimate the number of linkages when compared to a graph-based approach, such as our BGVAR model.

2002

This paper discusses techniques for estimating structural vector autoregressions. Especially when monetary policy shocks are estimated, VAR residuals turn out to be leptokurtic. It is argued that this is no coincidence but follows directly from the properties of monetary policy decisions. The paper proceeds to suggest an independent components estimator (ICE) that works well with leptokurtic residuals. Furthermore, the ICE

2004

Bringing together a collection of previously published work, this book provides a timely discussion of major considerations relating to the construction of econometric models that work well to explain economic phenomena, predict future outcomes, and be useful for policy-making. Analytical relations between dynamic econometric structural models and empirical time series MVARMA, VAR, transfer function, and univariate ARIMA models are established with important application for model-checking and model construction. The theory and applications of these procedures to a variety of econometric modeling and forecasting problems as well as Bayesian and non-Bayesian testing, shrinkage estimation, and forecasting procedures are also presented and applied. Finally, attention is focused on the effects of disaggregation on forecasting precision and the new Marshallian macroeconomic model () that features demand, supply, and entry equations for major sectors of economies is analyzed and described. This volume will prove invaluable to professionals, academics and students alike.