Projected principal component analysis in factor models
From MaRDI portal
Publication:5963521
Abstract: This paper introduces a Projected Principal Component Analysis (Projected-PCA), which employs principal component analysis to the projected (smoothed) data matrix onto a given linear space spanned by covariates. When it applies to high-dimensional factor analysis, the projection removes noise components. We show that the unobserved latent factors can be more accurately estimated than the conventional PCA if the projection is genuine, or more precisely, when the factor loading matrices are related to the projected linear space. When the dimensionality is large, the factors can be estimated accurately even when the sample size is finite. We propose a flexible semiparametric factor model, which decomposes the factor loading matrix into the component that can be explained by subject-specific covariates and the orthogonal residual component. The covariates' effects on the factor loadings are further modeled by the additive model via sieve approximations. By using the newly proposed Projected-PCA, the rates of convergence of the smooth factor loading matrices are obtained, which are much faster than those of the conventional factor analysis. The convergence is achieved even when the sample size is finite and is particularly appealing in the high-dimension-low-sample-size situation. This leads us to developing nonparametric tests on whether observed covariates have explaining powers on the loadings and whether they fully explain the loadings. The proposed method is illustrated by both simulated data and the returns of the components of the S&P 500 index.
Recommendations
- Robust projected principal component analysis for large-dimensional semiparametric factor modeling
- A self-reliant projected information criterion for the number of factors
- Using principal component analysis to estimate a high dimensional factor model with high-frequency data
- Directed principal component analysis
- Principal component analysis in very high-dimensional spaces
Cites work
- scientific article; zbMATH DE number 4048298 (Why is no real title available?)
- scientific article; zbMATH DE number 3765004 (Why is no real title available?)
- scientific article; zbMATH DE number 1911755 (Why is no real title available?)
- A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix
- A factor model approach to multiple testing under dependence
- A general framework for multiple testing dependence
- Asymptotics of sample eigenstructure for a large dimensional spiked covariance model
- Consistency of sparse PCA in high dimension, low sample size contexts
- Correlated \(z\)-values and the accuracy of large-scale statistical estimates
- Covariance regularization by thresholding
- Cross-dimensional inference of dependent high-dimensional data
- Determining the Number of Factors in Approximate Factor Models
- Determining the Number of Factors in the General Dynamic Factor Model
- Dynamic factor models with infinite-dimensional factor spaces: one-sided representations
- Efficient semiparametric estimation of the Fama-French model and extensions
- Eigenvalue ratio test for the number of factors
- Estimating false discovery proportion under arbitrary covariance dependence
- Estimation of (near) low-rank matrices with noise and high-dimensional scaling
- Exact matrix completion via convex optimization
- Factor modeling for high-dimensional time series: inference for the number of factors
- Forecasting Using Principal Components From a Large Number of Predictors
- GLS estimation of dynamic factor models
- Heteroskedasticity and Autocorrelation Consistent Covariance Matrix Estimation
- Improved penalization for determining the number of factors in approximate factor models
- Inferential Theory for Factor Models of Large Dimensions
- Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
- On the distribution of the largest eigenvalue in principal components analysis
- PCA consistency in high dimension, low sample size context
- Principal components estimation and identification of static factors
- Risks of large portfolios
- Sparse PCA: optimal rates and adaptive estimation
- Sparse principal component analysis and iterative thresholding
- Statistical analysis of factor models of high dimension
- Supervised singular value decomposition and its asymptotic properties
- The Generalized Dynamic Factor Model
- The high-dimension, low-sample-size geometric representation holds under mild conditions
- Time series modelling with semiparametric factor dynamics
Cited in
(46)- Dynamic Semiparametric Factor Model With Structural Breaks
- Subfactor projections.
- Factor Extraction in Dynamic Factor Models: Kalman Filter Versus Principal Components
- Autoencoder asset pricing models
- Supervised singular value decomposition and its asymptotic properties
- Semi-parametric single-index panel data models with interactive fixed effects: theory and practice
- Robust covariance estimation for approximate factor models
- Nonparametric fixed effects model for panel data with locally stationary regressors
- A self-reliant projected information criterion for the number of factors
- Estimating latent asset-pricing factors
- A semiparametric latent factor model for large scale temporal data with heteroscedasticity
- High dimensional minimum variance portfolio estimation under statistical factor models
- Time varying factor models with possibly strongly correlated noises
- Non-asymptotic properties of spectral decomposition of large Gram-type matrices and applications
- Directed principal component analysis
- Learning Latent Factors From Diversified Projections and Its Applications to Over-Estimated and Weak Factors
- Estimation and inference in semiparametric quantile factor models
- Predicting the VIX and the volatility risk premium: the role of short-run funding spreads volatility factors
- Sparse online principal component analysis for parameter estimation in factor model
- Cross-Validated Loss-based Covariance Matrix Estimator Selection in High Dimensions
- Varying-Coefficient Panel Data Models With Nonstationarity and Partially Observed Factor Structure
- Augmented factor models with applications to validating market risk factors and forecasting bond risk premia
- Recent developments in high dimensional covariance estimation and its related issues, a review
- Continuum directions for supervised dimension reduction
- Intraday cross-sectional distributions of systematic risk
- Interpretable Sparse Proximate Factors for Large Dimensions
- Sequential Scaled Sparse Factor Regression
- State-Varying Factor Models of Large Dimensions
- Dynamic Peer Groups of Arbitrage Characteristics
- Quasi maximum likelihood analysis of high dimensional constrained factor models
- Rejoinder
- Large covariance estimation through elliptical factor models
- Sufficient forecasting using factor models
- Uniform predictive inference for factor models with instrumental and idiosyncratic betas
- Changes in the span of systematic risk exposures
- Factor Models for High-Dimensional Tensor Time Series
- Large sample correlation matrices with unbounded spectrum
- Large dimensional latent factor modeling with missing observations and applications to causal inference
- An integrated panel data approach to modelling economic growth
- Robust projected principal component analysis for large-dimensional semiparametric factor modeling
- Least squares estimation of large dimensional threshold factor models
- GMM estimation for high-dimensional panel data models
- Target PCA: transfer learning large dimensional panel data
- Supervised multiway factorization
- Heterogeneity adjustment with applications to graphical model inference
- Doubly debiased Lasso: high-dimensional inference under hidden confounding
This page was built for publication: Projected principal component analysis in factor models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5963521)
