Estimating US Monetary Policy Shocks Using a Factor-Augmented Vector Autoregression: An EM Algorithm Approach

Bork, Lasse

doi:10.2139/ssrn.1348552

Cited by 15 publications

(39 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Straightforward adaptation of their expressions to the missing data case results in the restricted estimate given by

vec (Λ_{r} (j + 1)) = vec (Λ_{u} (j + 1)) + (\sum_{t = 1}^{T} {double-struckE}_{θ (j)} [f_{t} {f'}_{t} | Ω_{T}] \otimes R (j)) {H'}_{Λ} \times {((), H_{Λ} ((), \sum_{t = 1}^{T} E_{italicθ ((), j)} ([], f_{t} {f'}_{t} true| Ω_{T}) \otimes R ((), j)) {H'}_{Λ})}^{- 1} (κ_{Λ} - H_{Λ} vec (Λ_{u} (j + 1)))

where Λ u ( j + 1) is the unrestricted estimate given by expression . Restrictions on the parameters in the transition equation H A vec( A ) = κ A can be imposed in an analogous manner (see Bork, )…”

Section: Econometric Frameworkmentioning

confidence: 97%

“…Bork () and Bork et al . () show how to modify the M‐step of Watson and Engle () in order to impose restrictions of the form H Λ vec(Λ) = κ Λ for the model given by –.…”

Section: Econometric Frameworkmentioning

confidence: 99%

“…One of the advantages of the maximum likelihood approach proposed here, with respect to nonparametric methods based on principal components, is that it allows us to impose restrictions on the parameters in a relatively straightforward manner. Bork (2009) and Bork et al (2009) show how to modify the M-step of Watson and Engle (1983) in order to impose restrictions of the form H Λ vec(Λ) = k Λ for the model given by (1)-(2). Straightforward adaptation of their expressions to the missing data case results in the restricted estimate given by…”

Section: Restrictions On the Parametersmentioning

confidence: 99%

“…• FAVAR models as proposed by Bernanke et al (2005). Bork (2009) has recently shown how to estimate this type of model by EM algorithm. • Mixed frequency models-for example, the approach of Mariano and Murasawa (2003) to joint modelling of monthly and quarterly variables requires imposition of restrictions on the factor loadings of the latter.…”

Section: Restrictions On the Parametersmentioning

confidence: 99%

See 3 more Smart Citations

Maximum Likelihood Estimation of Factor Models on Datasets With Arbitrary Pattern of Missing Data

2012

View full text Add to dashboard Cite

show abstract

“…Straightforward adaptation of their expressions to the missing data case results in the restricted estimate given by

vec (Λ_{r} (j + 1)) = vec (Λ_{u} (j + 1)) + (\sum_{t = 1}^{T} {double-struckE}_{θ (j)} [f_{t} {f'}_{t} | Ω_{T}] \otimes R (j)) {H'}_{Λ} \times {((), H_{Λ} ((), \sum_{t = 1}^{T} E_{italicθ ((), j)} ([], f_{t} {f'}_{t} true| Ω_{T}) \otimes R ((), j)) {H'}_{Λ})}^{- 1} (κ_{Λ} - H_{Λ} vec (Λ_{u} (j + 1)))

Section: Econometric Frameworkmentioning

confidence: 97%

“…Bork () and Bork et al . () show how to modify the M‐step of Watson and Engle () in order to impose restrictions of the form H Λ vec(Λ) = κ Λ for the model given by –.…”

Section: Econometric Frameworkmentioning

confidence: 99%

Section: Restrictions On the Parametersmentioning

confidence: 99%

Section: Restrictions On the Parametersmentioning

confidence: 99%

See 2 more Smart Citations

Maximum Likelihood Estimation of Factor Models on Datasets With Arbitrary Pattern of Missing Data

2012

View full text Add to dashboard Cite

show abstract

“…One of the advantages of the maximum likelihood approach, with respect to nonparametric methods based on principal components, is that it allows us to impose restrictions on the parameters in a relatively straightforward manner. Bork (2009) and Bork, Dewachter, and Houssa (2009) show how to modify the maximization step of EM algorithm described Watson and Engle (1983) in order to impose restrictions of the form H Λ vec(Λ) = κ Λ for the model described in equations (1) to (3). Bańbura and Modugno (2014) show how those restrictions can be imposed in the presence of an arbitrary pattern of missing data.…”

Section: Restrictions On the Parametersmentioning

confidence: 99%

A Global Trade Model for the Euro Area

Modugno¹,

D’Agostino²,

Osbat³

2015

SSRN Journal

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in Non-technical summaryTrade dynamics are important determinants of output growth and inflationary pressures coming from import prices and policy institutions as well as the private sector need to forecast trade developments conditional on macroeconomic scenarios. Two main approaches have been used for forecasting trade: large structural macro models and time series models, to which our work belongs. We propose a dynamic factor model and show that exploiting the co-movement between macroeconomic and trade variables is essential for obtaining accurate short-term forecast of trade quantities and prices. We use this model to predic developments of trade variables given scenarios for macroeconomic variables and to quantify the effect on euro area trade of changed macroeconomic conditions in euro area trading partners. Our factor model is estimated with the methodology of Ban´bura and Modugno (2014), who propose maximum likelihood estimation based on a modification of the expectation maximization (EM) algorithm that allows to exploit datasets characterized by arbitrary patterns of missing data. This approach also makes it straightforward to introduce restrictions on the parameters and we use this to identify the nature of the unobserved factors. We evaluate the model in a pseudo short-term out-of-sample simulation from January 2006 to April 2013 on a panel of trade and macroeconomic data and we find that it delivers accurate forecasts because it can fully exploit the co-movement in the panel and the earlier releases of the macroeconomic variables. Indeed, including real macroeconomic variables, confidence indicators and prices improves the forecast accuracy over a model that exploits only trade information. We also find, in contrast to Burgert and D´ees (2009), but in line with Marcellino, Stock, and Watson (2003) for other euro area macroeconomic variables, that the "bottom-up" forecast approach for euro area exports and imports delivers forecasts as good as those obtained with a "direct" approach. This result is important, because it allows us to disentangle the contribution to the extra euro-area forecast from different world regions, adding value to the interpretation of conjunctural trade developments as well as to scenario analysis. We also run a natural experiment and generate the dynamics of trade variables during the great recession conditional on the realized path of macroeconomic variables. We find that the model tracks tra...

show abstract

Monetary transmission mechanism and time variation in the Euro area

Bagzibagli

2013

Empir Econ

View full text Add to dashboard Cite

Estimating US Monetary Policy Shocks Using a Factor-Augmented Vector Autoregression: An EM Algorithm Approach

Cited by 15 publications

References 50 publications

Maximum Likelihood Estimation of Factor Models on Datasets With Arbitrary Pattern of Missing Data

Maximum Likelihood Estimation of Factor Models on Datasets With Arbitrary Pattern of Missing Data

A Global Trade Model for the Euro Area

Monetary transmission mechanism and time variation in the Euro area

Contact Info

Product

Resources

About