Maximum Likelihood Estimation and Inference for Approximate Factor Models of High Dimension

Bai, Jushan; Li, Kunpeng

doi:10.1162/rest_a_00519

Cited by 98 publications

(85 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Doz, Giannone, and Reichlin (2012) proved the consistency of the common factor estimators for their true underlying values, while Bai and Li (2016) have also obtained rates of convergence and asymptotic variances under some restrictions. Doz, Giannone, and Reichlin (2012) proved the consistency of the common factor estimators for their true underlying values, while Bai and Li (2016) have also obtained rates of convergence and asymptotic variances under some restrictions.…”

Section: Conclusion and Extensionsmentioning

confidence: 94%

“…The extension of our methods to models in which N∕T is nonnegligible would also constitute a very valuable addition with potentially interesting empirical applications. Doz, Giannone, and Reichlin (2012) proved the consistency of the common factor estimators for their true underlying values, while Bai and Li (2016) have also obtained rates of convergence and asymptotic variances under some restrictions.…”

Section: Conclusion and Extensionsmentioning

confidence: 94%

See 1 more Smart Citation

Dynamic specification tests for dynamic factor models

Fiorentini

Sentana

2019

J of Applied Econometrics

View full text Add to dashboard Cite

Summary We derive computationally simple expressions for score tests of misspecification in parametric dynamic factor models using frequency domain techniques. We interpret those diagnostics as time domain moment tests which assess whether certain autocovariances of the smoothed latent variables match their theoretical values under the null of correct model specification. We also reinterpret reduced‐form residual tests as checking specific restrictions on structural parameters. Our Gaussian tests are robust to nonnormal, independent innovations. Monte Carlo exercises confirm the finite‐sample reliability and power of our proposals. Finally, we illustrate their empirical usefulness in an application that constructs a US coincident indicator.

show abstract

Section: Conclusion and Extensionsmentioning

confidence: 94%

Section: Conclusion and Extensionsmentioning

confidence: 94%

Dynamic specification tests for dynamic factor models

Fiorentini

Sentana

2019

J of Applied Econometrics

View full text Add to dashboard Cite

show abstract

“…They showed that estimation under independent Gaussian errors still lead to consistent estimators for the large approximate factor model, even when the true model has cross-sectional and time series correlation in the idiosyncratic errors. Bai & Li (2012b) studied related issues for dynamic factors and cross-sectionally and serially correlated errors estimated by the maximum likelihood method.…”

Section: The Maximum Likelihood Methodsmentioning

confidence: 99%

Econometric Analysis of Large Factor Models

Bai

Wang

2016

Annu. Rev. Econ.

Self Cite

View full text Add to dashboard Cite

Large factor models use a few latent factors to characterize the co-movement of economic variables in a high dimensional data set. High dimensionality brings challenge as well as new insight into the advancement of econometric theory. Due to its ability to effectively summarize information in large data sets, factor models have been increasingly used in economics and finance. The factors, being estimated from the high dimensional data, can help to improve forecast, provide efficient instruments, control for nonlinear unobserved heterogeneity, and capture cross-sectional dependence, etc. This article reviews the theory on estimation and statistical inference of large factor models. It also discusses important applications and highlights future directions.

show abstract

“…As we have shown that our algorithm has a nonincreasing property, we terminate the algorithm when the reduction of the value of objective function is less than 10 −6 . We take the consistent estimates of Σ e and Λ given by Bai and Li (2016) as the initial values in our iterative procedure.…”

Section: Empirical Studymentioning

confidence: 99%

On the penalized maximum likelihood estimation of high-dimensional approximate factor model

2019

View full text Add to dashboard Cite

In this paper, we mainly focus on the penalized maximum likelihood estimation (MLE) of the high-dimensional approximate factor model. Since the current estimation procedure can not guarantee the positive definiteness of the error covariance matrix, by reformulating the estimation of error covariance matrix and based on the lagrangian duality, we propose an accelerated proximal gradient (APG) algorithm to give a positive definite estimate of the error covariance matrix. Combined the APG algorithm with EM method, a new estimation procedure is proposed to estimate the high-dimensional approximate factor model. The new method not only gives positive definite estimate of error covariance matrix but also improves the efficiency of estimation for the high-dimensional approximate factor model. Although the proposed algorithm can not guarantee a global unique solution, it enjoys a desirable non-increasing property. The efficiency of the new algorithm on estimation and forecasting is also investigated via simulation and real data analysis.

show abstract

Maximum Likelihood Estimation and Inference for Approximate Factor Models of High Dimension

Abstract: An approximate factor model of high dimension has two key features. First, the idiosyncratic errors are correlated and heteroskedastic over both the cross- Monte Carlo simulations show the method is easy to implement and an application to the U.S. yield curves is considered.

Cited by 98 publications

References 30 publications

Dynamic specification tests for dynamic factor models

Dynamic specification tests for dynamic factor models

Econometric Analysis of Large Factor Models

On the penalized maximum likelihood estimation of high-dimensional approximate factor model

Contact Info

Product

Resources

About