Factor modeling for high-dimensional time series: Inference for the number of factors

Abstract

This paper deals with the factor modeling for high-dimensional time series based on a dimension-reduction viewpoint. Under stationary settings, the inference is simple in the sense that both the number of factors and the factor loadings are estimated in terms of an eigenanalysis for a nonnegative definite matrix, and is therefore applicable when the dimension of time series is on the order of a few thousands. Asymptotic properties of the proposed method are investigated under two settings: (i) the sample size goes to infinity while the dimension of time series is fixed; and (ii) both the sample size and the dimension of time series go to infinity together. In particular, our estimators for zero-eigenvalues enjoy faster convergence (or slower divergence) rates, hence making the estimation for the number of factors easier. In particular, when the sample size and the dimension of time series go to infinity together, the estimators for the eigenvalues are no longer consistent. However, our estimator for the number of the factors, which is based on the ratios of the estimated eigenvalues, still works fine. Furthermore, this estimation shows the so-called "blessing of dimensionality" property in the sense that the performance of the estimation may improve when the dimension of time series increases. A two-step procedure is investigated when the factors are of different degrees of strength. Numerical illustration with both simulated and real data is also reported.

arXiv commentPublished in at http://dx.doi.org/10.1214/12-AOS970 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (1)

• Identifying the finite dimensionality of curve time series

Cited by in corpus (69)

• Challenges of Big Data Analysis
• Projected principal component analysis in factor models
• Constrained Factor Models for High-Dimensional Matrix-Variate Time Series
• Statistical Inference for High-Dimensional Matrix-Variate Factor Model
• Gaussian and bootstrap approximations for high-dimensional U-statistics and their applications
• Posterior contraction in sparse Bayesian factor models for massive covariance matrices
• Networks, Dynamic Factors, and the Volatility Analysis of High-Dimensional Financial Series
• Principal component analysis for second-order stationary vector time series
• Identifying the number of factors from singular values of a large sample auto-covariance matrix
• Modelling matrix time series via a tensor CP-decomposition
• Estimation of the Number of Spikes, Possibly Equal, in the High-Dimensional Case
• Rank Determination in Tensor Factor Model
• Modeling High-Dimensional Time Series: A Factor Model with Dynamically Dependent Factors and Diverging Eigenvalues
• Extreme value analysis for the sample autocovariance matrices of heavy-tailed multivariate time series
• Determining the number of factors in a forecast model by a random matrix test: cryptocurrencies
• Localizing Changes in High-Dimensional Vector Autoregressive Processes
• Expandable Factor Analysis
• Sequential testing for structural stability in approximate factor models
• Semi-parametric TEnsor Factor Analysis by Iteratively Projected Singular Value Decomposition
• Tensor Factor Model Estimation by Iterative Projection
• Factor Models for Matrix-Valued High-Dimensional Time Series
• Factor Models for High-Dimensional Tensor Time Series
• Robust high dimensional factor models with applications to statistical machine learning
• Consistent estimation of high-dimensional factor models when the factor number is over-estimated
• High dimensional VAR with low rank transition
• Statistical Inferences Using Large Estimated Covariances for Panel Data and Factor Models
• High-Dimensional Low-Rank Tensor Autoregressive Time Series Modeling
• Generalized probabilistic principal component analysis of correlated data
• Optimal Subspace Estimation Using Overidentifying Vectors via Generalized Method of Moments
• Scalable marginalization of correlated latent variables with applications to learning particle interaction kernels
• Modeling High-Dimensional Unit-Root Time Series
• Moment approach for singular values distribution of a large auto-covariance matrix
• High dimensional deformed rectangular matrices with applications in matrix denoising
• Detecting Latent Communities in Network Formation Models
• Estimation of False Discovery Proportion with Unknown Dependence
• Threshold factor models for high-dimensional time series
• A Two-Way Transformed Factor Model for Matrix-Variate Time Series
• Projected Estimation for Large-dimensional Matrix Factor Models
• Spectral Analysis of High-dimensional Time Series
• Factor modelling for high-dimensional functional time series
• On singular value distribution of large dimensional auto-covariance matrices
• Generalized Yule-Walker Estimation for Spatio-Temporal Models with Unknown Diagonal Coefficients
• Heterogeneity Adjustment with Applications to Graphical Model Inference
• Finite Sample Theory for High-Dimensional Functional/Scalar Time Series with Applications
• On singular values distribution of a large auto-covariance matrix in the ultra-dimensional regime
• FarmTest: Factor-Adjusted Robust Multiple Testing with Approximate False Discovery Control
• Gaussian approximation for the sup-norm of high-dimensional matrix-variate U-statistics and its applications
• Exact Cluster Recovery via Classical Multidimensional Scaling
• On the estimation of locally stationary functional time series
• The eigenvalues of the sample covariance matrix of a multivariate heavy-tailed stochastic volatility model
• Interactive Effects Panel Data Models with General Factors and Regressors
• Sufficient Forecasting Using Factor Models
• Modeling Multivariate Time Series with Copula-linked Univariate D-vines
• Augmented Factor Models with Applications to Validating Market Risk Factors and Forecasting Bond Risk Premia
• Disentangling and Assessing Uncertainties in Multiperiod Corporate Default Risk Predictions
• Bi-cross-validation for factor analysis
• Determining the dimension of factor structures in non-stationary large datasets
• Prediction in latent factor regression: Adaptive PCR and beyond
• Adaptive Estimation for Non-stationary Factor Models And A Test for Static Factor Loadings
• Tight Risk Bound for High Dimensional Time Series Completion
• A QR Decomposition Approach to Factor Modelling: A Thesis Report
• Homogeneity and Sub-homogeneity Pursuit: Iterative Complement Clustering PCA
• Factor Models for High-Dimensional Functional Time Series
• Inverse Moment Methods for Sufficient Forecasting using High-Dimensional Predictors
• Banded Spatio-Temporal Autoregressions
• Forecasting high-dimensional functional time series with dual-factor structures
• Functional Linear Regression: Dependence and Error Contamination
• Factor Modelling for Clustering High-dimensional Time Series
• A Structural-Factor Approach to Modeling High-Dimensional Time Series and Space-Time Data