Maximum likelihood estimation for dynamic factor models with missing data

Jungbacker, Borus; Koopman, Siem Jan; Wel, Michel van der

doi:10.1016/j.jedc.2011.03.009

Cited by 66 publications

(51 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Looking at the e¤ects of parameter uncertainty when constructing intervals for estimated factors in empirical applications is within our research agenda. Also, the analysis of real data systems can be extended to consider unbalanced data bases by using, for example, the computationally e¢ cient procedures by Jungbacker et al (2011) and Jungbacker and Koopman (in press).…”

Section: Discussionmentioning

confidence: 99%

“…However, the number of parameters that need to be estimated increase with the cross-sectional dimension in such a way that ML estimation is unfeasible for moderate systems. Jungbacker et al (2011) and Jungbacker and Koopman (in press) propose a computationally feasible device to deal with large dimensional unobserved component models using the Kalman …lter. However, if the cross-sectional dimension is large, this procedure is only feasible if the idiosyncratic noises are serially uncorrelated.…”

Section: Kalman …Lter and Smoothingmentioning

confidence: 99%

See 1 more Smart Citation

Small- Versus Big-Data Factor Extraction in Dynamic Factor Models: An Empirical Assessment

Poncela¹,

Ruiz²

2016

Advances in Econometrics

View full text Add to dashboard Cite

In the context of Dynamic Factor Models (DFM), we compare point and interval estimates of the underlying unobserved factors extracted using small and big-data procedures. Our paper differs from previous works in the related literature in several ways. First, we focus on factor extraction rather than on prediction of a given variable in the system. Second, the comparisons are carried out by implementing the procedures considered to the same data. Third, we are interested not only on point estimates but also on confidence intervals for the factors. Based on a simulated system and the macroeconomic data set popularized by Stock and Watson (2012), we show that, for a given procedure, factor estimates based on different cross-sectional dimensions are highly correlated. On the other hand, given the cross-sectional dimension, the Maximum Likelihood Kalman filter and smoother (KFS) factor estimates are highly correlated with those obtained using hybrid Principal Components (PC) and KFS procedures. The PC estimates are somehow less correlated. Finally, the PC intervals based on asymptotic approximations are unrealistically tiny. Keywords January 2015 AbstractIn the context of Dynamic Factor Models (DFM), we compare point and interval estimates of the underlying unobserved factors extracted using small and big-data procedures. Our paper di¤ers from previous works in the related literature in several ways. First, we focus on factor extraction rather than on prediction of a given variable in the system. Second, the comparisons are carried out by implementing the procedures considered to the same data. Third, we are interested not only on point estimates but also on con…dence intervals for the factors. Based on a simulated system and the macroeconomic data set popularized by Stock and Watson (2012), we show that, for a given procedure, factor estimates based on di¤erent cross-sectional dimensions are highly correlated. On the other hand, given the cross-sectional dimension, the Maximum Likelihood Kalman …lter and smoother (KFS) factor estimates are highly correlated with those obtained using hybrid Principal Components (PC) and KFS procedures. The PC estimates are somehow less correlated. Finally, the PC intervals based on asymptotic approximations are unrealistically tiny.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Kalman …Lter and Smoothingmentioning

confidence: 99%

Small- Versus Big-Data Factor Extraction in Dynamic Factor Models: An Empirical Assessment

Poncela¹,

Ruiz²

2016

Advances in Econometrics

View full text Add to dashboard Cite

show abstract

“…The algorithm has been shown to be computationally efficient and feasible even with high-dimensional data. Recent results by Jungbacker et al (2011) and Jungbacker and Koopman (2015) show how computational efficiency can be further improved.…”

Section: Insert Figure 3 Herementioning

confidence: 99%

“…See Giannone et al (2008); Aruoba et al (2009);Camacho and Perez-Quiros (2010);Jungbacker et al (2011);Bańbura and Modugno (2014), and for surveys Bańbura et al (2011Bańbura et al ( , 2013.…”

Section: Introductionmentioning

confidence: 99%

Nowcasting Business Cycles: A Bayesian Approach to Dynamic Heterogeneous Factor Models

D’Agostino¹,

Giannone²,

Lenza³

et al. 2015

FEDS

View full text Add to dashboard Cite

We develop a framework for measuring and monitoring business cycles in real time. Following a long tradition in macroeconometrics, inference is based on a variety of indicators of economic activity, treated as imperfect measures of an underlying index of business cycle conditions. We extend existing approaches by permitting for heterogenous lead-lag patterns of the various indicators along the business cycles. The framework is well suited for high-frequency monitoring of current economic conditions in real time -nowcasting -since inference can be conducted in presence of mixed frequency data and irregular patterns of data availability. Our assessment of the underlying index of business cycle conditions is accurate and more timely than popular alternatives, including the Chicago Fed National Activity Index (CFNAI). A formal real-time forecasting evaluation shows that the framework produces well-calibrated probability nowcasts that resemble the consensus assessment of the Survey of Professional Forecasters.JEL Classification: C11, C32, C38, E32, E38.

show abstract

“…They argue that, even though the principle components approach has been used extensively in the literature, maximum likelihood estimation can lead to greater efficiency gains, even when the DFM is misspecified. Jungbacker et al (2011) use a similar maximum likelihood approach for DFMs and extend it to account for missing data.…”

Section: Introductionmentioning

confidence: 99%

A Generalized Dynamic Factor Model for Panel Data: Estimation with a Two-Cycle Conditional Expectation-Maximization Algorithm

Zirogiannis

Tripodis

2013

SSRN Journal

View full text Add to dashboard Cite

Abstract:We develop a generalized dynamic factor model for panel data with the goal of estimating an unobserved index. While similar models have been developed in the literature of dynamic factor analysis, our contribution is threefold. First, contrary to simple dynamic factor analysis where multiple attributes of the same subject are measured at each time period, our model also accounts for multiple subjects. It is therefore suitable to a panel data framework. Second, our model estimates a unique unobserved index for every subject for every time period, as opposed to previous work where a temporal index common to all subjects was used. Third, we develop a novel iterative estimation process which we call the Two-Cycle Conditional ExpectationMaximization (2CCEM) algorithm and is flexible enough to handle a variety of different types of datasets. The model is applied on a panel measuring attributes related to the operation of water and sanitation utilities.

show abstract

Maximum likelihood estimation for dynamic factor models with missing data

Cited by 66 publications

References 22 publications

Small- Versus Big-Data Factor Extraction in Dynamic Factor Models: An Empirical Assessment

Small- Versus Big-Data Factor Extraction in Dynamic Factor Models: An Empirical Assessment

Nowcasting Business Cycles: A Bayesian Approach to Dynamic Heterogeneous Factor Models

A Generalized Dynamic Factor Model for Panel Data: Estimation with a Two-Cycle Conditional Expectation-Maximization Algorithm

Contact Info

Product

Resources

About