What is the purpose of the out-of-sample forecast evaluation exercise?

The out-of-sample forecast evaluation exercise covers the period after the introduction of the euro and focuses on the 12 observed variables in the NAWM that are endogenously determined by the model.

What is the effect of a positive permanent technology shock on the growth rates of the real variables?

A positive permanent technology shock implies a permanent increase in the levels and a transitory increase in the growth rates of the real variables with a half-life of 3 quarters.

What is the view that the random walk is relatively accurate for forecasting macroeconomic time series?

Given the view that the random walk is relatively accurate for forecasting macroeconomic time series (in levels), the Minnesota prior is based on shrinking the VAR parameters towards univariate random-walk processes.20 ECB Working Paper Series No 1185 May 2010Moreover, VAR models may be considered as linear approximations of DSGE models.

What is the deterministic equation for the k-dimensional vector xt?

H ′ξt + wt, t = 1, . . . , T. (11)The k-dimensional vector xt is here assumed to be deterministic, while wt is a vector of i.i.d. normal measurement errors with mean zero and covariance matrix R.

What is the log predictive score for the random-walk and mean models?

In addition, the authors can compute the log predictive score for the random-walk and mean models by assuming that the 1-step ahead prediction errors are multivariate normal with mean zero and covariance matrix Σ.

What is the criterion for evaluating the goodness of fit of a model estimated?

Provided that the models the authors wish to compare are not subject to Lindley’s paradox (Bartlett, 1957; Lindley, 1957), the marginal likelihood is one important criterion for evaluating the goodness-of-fit of a model estimated with Bayesian methods: the larger the value of the marginal likelihood, the better a model fits the observed data.

Open AccessBook10.1093/OXFORDHB/9780195398649.013.0005

Forecasting With DSGE Models

Kai Christoffel, +2 more

- 08 Jul 2011

- Social Science Research Network

122

TL;DR: In this paper, the forecasting performance of the New Area-Wide Model (NAWM) that has been designed for use in the macroeconomic projections at the European Central Bank was examined.

Abstract: In this paper we review the methodology of forecasting with log-linearised DSGE models using Bayesian methods. We focus on the estimation of their predictive distributions, with special attention being paid to the mean and the covariance matrix of h-step ahead forecasts. In the empirical analysis, we examine the forecasting performance of the New Area-Wide Model (NAWM) that has been designed for use in the macroeconomic projections at the European Central Bank. The forecast sample covers the period following the introduction of the euro and the out-of-sample performance of the NAWM is compared to nonstructural benchmarks, such as Bayesian vector autoregressions (BVARs). Overall, the empirical evidence indicates that the NAWM compares quite well with the reduced-form models and the results are therefore in line with previous studies. Yet there is scope for improving the NAWM’s forecasting performance. For example, the model is not able to explain the moderation in wage growth over the forecast evaluation period and, therefore, it tends to overestimate nominal wages. As a consequence, both the multivariate point and density forecasts using the log determinant and the log predictive score, respectively, suggest that a large BVAR can outperform the NAWM.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What are the contributions in this paper?

In this paper, the authors discuss an algorithm for estimating the predictive distribution of the observed variables based on draws of the DSGE model parameters.

2. What are the future works in this paper?

If the authors wish to estimate quantiles, confidence regions or the probability that the variables reach some barrier, then they need a numerical algorithm for computing the predictive density since the integral in ( 12 ) can not be solved analytically.. This so-called sampling the future algorithm has been adapted by Adolfson, Lindé, and Villani ( 2007 ) to state-space models.. The DSGE-VAR approach was further enriched by Del Negro, Schorfheide, Smets, and Wouters ( 2007 ) into a framework for assessing the time series fit of a DSGE model.. On the other hand, if the forecast evaluation only requires moments from the predictive distribution, then such an algorithm is not needed since the moments can be estimated with high precision using draws from the posterior distribution of the parameters.

3. What is the main reason for the ranking of forecast models?

Since MSE-based measures are unable to account for the correlation between forecast errors at different horizons, the ranking of forecast models may depend on the choice of data transformation.

4. What are the two measures that are often used for evaluating multivariate forecast accuracy?

The trace and the log determinant are two measures that are often used in practice for evaluating multivariate forecast accuracy; see, e.g., Adolfson, Lindé, and Villani (2007).

Figure 11. Log predictive scores when forecasting annual changes of variables in first differences.

Figure 3. Recursive posterior mode estimates of selected structural parameters of the NAWM.

Figure 9. Log determinant statistics of the scaled MSE matrices when forecasting annual changes of variables in first differences.

Figure 7. Trace statistics of the scaled MSE matrices when forecasting annual changes of variables in first differences.

Table 2. Percentage share for squared mean errors of mean squared errors when forecasting quarterly changes of variables in first differences.

Figure 5. Scaled root mean squared forecast errors for 12 variables when forecasting annual changes of variables in first differences.

Citations

Repository

Forecasting: theory and practice

Fotios Petropoulos, +84 more

- 04 Dec 2020

- arXiv: Applications

TL;DR: A non-systematic review of the theory and the practice of forecasting, offering a wide range of theoretical, state-of-the-art models, methods, principles, and approaches to prepare, produce, organise, and evaluate forecasts.

...read moreread less

325

•Book Chapter•10.1016/B978-0-444-53683-9.00002-5

DSGE Model-Based Forecasting

Marco Del Negro, +1 more

- 01 Jan 2013

- Social Science Research Network

TL;DR: In this article, the authors provide a real-time assessment of the forecasting performance of the Smets and Wouters (2007) model data up to 2011, compare it with Blue Chip and Greenbook forecasts, and show how it changes as they augment the standard set of observables with external information from surveys.

...read moreread less

254

•Posted Content

DSGE model-based forecasting

Marco Del Negro, +1 more

- 01 Jan 2013

- Research Papers in Economics

...read moreread less

187

•Book Chapter•10.1016/B978-0-444-53683-9.00005-0

Forecasting and Policy Making

Volker Wieland, +2 more

- 01 Jan 2013

- Research Papers in Economics

TL;DR: In this article, the authors investigated the use of economic forecasting in policy making by discussing practical examples, providing new empirical evidence and computing forecasts using different macroeconomic models, and provided a detailed exposition of methods for producing forecasts.

...read moreread less

101

•Journal Article•10.17016/FEDS.2016.033

A Time‐Series Model of Interest Rates with the Effective Lower Bound

Benjamin K. Johannsen, +1 more

- 01 Apr 2016

- Social Science Research Network

TL;DR: A flexible time– series approach which includes a “shadow rate”—a notional rate that is less than the ELB during the period in which the bound is binding—without imposing no–arbitrage assumptions is proposed.

...read moreread less

...

Expand

References

•Journal Article•10.2307/1912017

Macroeconomics and reality

Christopher A. Sims

- 01 Jan 1980

- Econometrica

TL;DR: In this article, the authors argue that the style in which their builders construct claims for a connection between these models and reality is inappropriate, to the point at which claims for identification in these models cannot be taken seriously.

...read moreread less

13.7K

Journal Article•10.2307/2076916

Time Series Analysis.

Jorg Breitung, +1 more

- 01 Mar 1995

- Contemporary Sociology

12K

Journal Article•10.1016/0304-3932(83)90060-0

Staggered prices in a utility-maximizing framework

Guillermo A. Calvo

- 01 Sep 1983

- Journal of Monetary Economics

TL;DR: In this article, the authors developed a model of staggered prices along the lines of Phelps (1978) and Taylor (1979, 1980), but utilizing an analytically more tractable price-setting technology.

...read moreread less

9.1K

•Journal Article•10.1086/426038

Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy

Lawrence J. Christiano, +2 more

- 01 Feb 2005

- Journal of Political Economy

TL;DR: In this article, the authors present a model embodying moderate amounts of nominal rigidities that accounts for the observed inertia in inflation and persistence in output, and the key features of their model are those that prevent a sharp rise in marginal costs after an expansionary shock to monetary policy.

...read moreread less

Journal Article•10.2307/1913386

Time to build and aggregate fluctuations

Finn E. Kydland, +1 more

- 01 Nov 1982

- Econometrica

TL;DR: In this article, a general equilibrium model is developed and fitted to U.S. quarterly data for the post-war period, with the assumption that more than one time period is required for the construction of new productive capital and the non-time-separable utility function that admits greater intertemporal substitution of leisure.

...read moreread less

6.7K

...

Expand

Forecasting With DSGE Models

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What are the contributions in this paper?

2. What are the future works in this paper?

3. What is the main reason for the ranking of forecast models?

4. What are the two measures that are often used for evaluating multivariate forecast accuracy?

5. What is the purpose of the out-of-sample forecast evaluation exercise?

6. What is the effect of a positive permanent technology shock on the growth rates of the real variables?

7. What is the view that the random walk is relatively accurate for forecasting macroeconomic time series?

8. What is the deterministic equation for the k-dimensional vector xt?

9. What is the log predictive score for the random-walk and mean models?

10. What is the criterion for evaluating the goodness of fit of a model estimated?

Figures

Citations

Forecasting: theory and practice

DSGE Model-Based Forecasting

DSGE model-based forecasting

Forecasting and Policy Making

A Time‐Series Model of Interest Rates with the Effective Lower Bound

References

Macroeconomics and reality

Time Series Analysis.

Staggered prices in a utility-maximizing framework

Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy

Time to build and aggregate fluctuations

Related Papers (5)

Shocks and frictions in US business cycles: A Bayesian DSGE approach

An estimated dynamic stochastic general equilibrium model of the euro area

Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy

Large Bayesian vector auto regressions

Macroeconomics and reality