Determining the number of factors in approximate factor models with global and group-specific factors

By F. Dias, Maximiano Pinheiro, Antonio Rúa. 2008

567 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Principal Components and Regularized Estimation of Factor Models (2017, methodology): “Bai and Ng (2002) suggests a class of consistent factor selection criteria, that is, rules that yield r such that prob( r = r ) → 1 as N, T → ∞ . b b While criteria in this class have little diﬃculty singling out the dominant factors, they tend to be too liberal.”
Dynamic factor models with infinite-dimensional factor spaces: One-sided representations (2015, methodology): “Criteria to determine r consistently are given in Bai and Ng (2002) (see also Alessi et al. 2010).”
Estimation and Inference of FAVAR Models (2014, methodology): “One approach is to first estimate the total number of factors based on Z (denoted as r̂) by the information criterion proposed by Bai and Ng (2002), and then get r̂1 = r̂ − r2 where r2 is the number of known factors.”
Regional Inflation Dynamics within and Across Euro Area Countries and a Comparison with the Us (2006, background): “Statistical information criteria for the selection of the number of factors, in particular those proposed by Bai & Ng (2002), are hardly applicable in our context due to the rather small sample size and the presence of some correlation across the idiosyncratic components.”
A simple panel unit root test in the presence of cross-section dependence (2007, unknown): “To this end we first applied the factor selection procedure advanced in Bai and Ng (2002), using their information criteria, PCi and ICi, i D 1, 2, 3, to the standardized first differences of the series, assuming the maximum number of factors, kmax, to be 1, 4 and 6.”
Theory and methods of panel data models with interactive effects (2014, background): “If not, the information criterions developed by [11] can be used to determine it.”
Panel Data Models with Multiple Time-Varying Individual Effects (2013, methodology): “Bai (2005) also shows that the methods of Bai and Ng (2002) can be used to consistently estimate the true number of factors.”
Understanding Systematic Risk: A High‐Frequency Approach (2020, unknown): “However, a shortcoming of the characteristic-sorted approach is that 2The general case of a static large-dimensional factor model is treated in Bai (2003) and Bai and Ng (2002).”
HIGH DIMENSIONAL COVARIANCE MATRIX ESTIMATION IN APPROXIMATE FACTOR MODELS (2011, background): “Bai and Ng (2002) determined K = 2 common factors.”
A bootstrap procedure for panel data sets with many cross‐sectional units (2008, unknown): “This is indeed of practical significance given the recent work of, among others, Stock and Watson (1998), Bai and Ng (2002), Bai (2003), Bai and Ng (2004) and Pesaran (2002).”
Constrained Factor Models for High-Dimensional Matrix-Variate Time Series (2017, background): “The factor models in Tsai and Tsay (2010) and Tsai et al. (2016), following the definition in Bai (2003), Bai and Ng (2002, 2007), Forni et al. (2000, 2004), attempt to separate the common factors that affect the dynamics of most original component series from the idiosyncratic series that at most…”
A Testing Procedure for Determining the Number of Factors in Approximate Factor Models With Large Datasets (2010, methodology): “Other rigorous methods for doing this, in a variety of factor models, have been developed previously by Bai and Ng (2002, 2007), Amengual and Watson (2007), and Hallin and Liska (2007).”
TESTING HYPOTHESES ABOUT THE NUMBER OF FACTORS IN LARGE FACTOR MODELS (2009, methodology): “We set the maximum number of static factors at 10 and, in Bai and Ng (2005) notation, δ = 0.”
A parametric estimation method for dynamic factor models of large dimensions (2009, unknown): “which is a special case of the specification in (1), even though by not taking into account the particular structure of the A matrix and the reduced rank of the error process we are losing in terms of efficiency.”
Synthetic controls with imperfect pretreatment fit (2016, methodology): “This is also consistent with the literature on linear factor models, which shows that these models can be consistently estimated in large panels (e.g., Bai (2003), Bai and Ng (2002), Bai (2009), and Moon, Roger, and Weidner (2015)).”
Maximum Likelihood Estimation and Inference for Approximate Factor Models of High Dimension (2016, background): “” (See, for example, Bernanke and Boivin, 2003, Bernanke et al. 2005, and Kose et al. 2003.) The purpose of this paper is to provide an inferential theory for the estimated parameters of high dimensional approximate factor models. The notion of approximate factor models is proposed by Chamberlain an”
Determining the Number of Factors from Empirical Distribution of Eigenvalues (2010, methodology): “…nonrejected null, or to rmax if all the nulls are rejected.9 Our first experiment consists of generating 1,000 replications of data produced by the data-generating process used in Bai and Ng (2002): Xit = r∑ k=1 ΛikFkt + √ θeit eit = ρei,t−1 + vit + ∑ 1≤| j|≤8 βvi−j,t; Fkt , Λik , vit ∼ i.i.d.”
Principal Component Analysis of High-Frequency Data (2016, methodology): “That being said, PCA is often employed to analyze (approximate) factor structures (Ross 1976; Chamberlain and Rothschild 1983), including estimating the number of approximate factors (see Connor and Korajczyk 1993; Bai and Ng 2002; Amengual andWatson 2007; Kapetanios 2010;Onatski 2010, and Trapani 2”
Testing for Structural Stability of Factor Augmented Forecasting Models (2014, methodology): “As discussed above, the number of factors is estimated based on the full sample of data, using the approach of Bai and Ng (2002).”