Forecasting using a large number of predictors: is Bayesian shrinkage a valid alternative to principal components?

DOI10.1016/j.jeconom.2008.08.011zbMath1429.62659MaRDI QIDQ299225

Christine De Mol, Domenico Giannone, Lucrezia Reichlin

Publication date: 22 June 2016

Published in: Journal of Econometrics (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.jeconom.2008.08.011

zbMATH Keywords

ridge regression; principal components; Bayesian shrinkage; Bayesian VAR; large cross-sections; Lasso regression

Mathematics Subject Classification ID

62P20: Applications of statistics to economics

62H25: Factor analysis and principal components; correspondence analysis

62J07: Ridge regression; shrinkage estimators (Lasso)

91B84: Economic time series analysis

Related Items

Stock return predictability: A factor-augmented predictive regression system with shrinkage method, Model-based approach for scenario design: stress test severity and banks' resiliency, Random autoregressive models: A structured overview, Forecasting crude oil prices: do technical indicators need economic constraints?, Post-selection inference of generalized linear models based on the lasso and the elastic net, Bayesian principal component regression with data-driven component selection, IPAD: Stable Interpretable Forecasting with Knockoffs Inference, High-Dimensional Posterior Consistency in Bayesian Vector Autoregressive Models, Joint Structural Break Detection and Parameter Estimation in High-Dimensional Nonstationary VAR Models, Efficient Estimation with Many Weak Instruments Using Regularization Techniques, Determining the number of factors with potentially strong within-block correlations in error terms, Factor Extraction in Dynamic Factor Models: Kalman Filter Versus Principal Components, High-Dimensional Vector Autoregressive Time Series Modeling via Tensor Decomposition, Sparse Identification and Estimation of Large-Scale Vector AutoRegressive Moving Averages, BAYESIAN DYNAMIC VARIABLE SELECTION IN HIGH DIMENSIONS, Approximate factor models with weaker loadings, When are Google Data Useful to Nowcast GDP? An Approach via Preselection and Shrinkage, Lasso Inference for High-Dimensional Time Series, Regularized estimation in sparse high-dimensional time series models, Forecasting economic time series using targeted predictors, Sparse and robust normal and \(t\)-portfolios by penalized \(L_q\)-likelihood minimization, The three-pass regression filter: a new approach to forecasting using many predictors, Asymptotic analysis of the squared estimation error in misspecified factor models, Asymptotics of the principal components estimator of large factor models with weakly influential factors, Infinite-dimensional VARs and factor models, Strong selection consistency of Bayesian vector autoregressive models based on a pseudo-likelihood approach, Revisiting useful approaches to data-rich macroeconomic forecasting, Forecasting inflation and GDP growth using heuristic optimisation of information criteria and variable reduction methods, Real-time factor model forecasting and the effects of instability, Estimation and forecasting in vector autoregressive moving average models for rich datasets, Tactical sales forecasting using a very large set of macroeconomic indicators, Dynamic variable selection with spike-and-slab process priors, High-dimensional inference for linear model with correlated errors, Nowcasting with large Bayesian vector autoregressions, Sparse restricted perceptions equilibrium, Nonconcave penalized estimation in sparse vector autoregression model, Sequential testing for structural stability in approximate factor models, Consistent estimation of high-dimensional factor models when the factor number is over-estimated, Factor models in high-dimensional time series: A time-domain approach, Theory-coherent forecasting, Forecasting by factors, by variables, by both or neither?, Factor models with local factors -- determining the number of relevant factors, Structural inference in sparse high-dimensional vector autoregressions, Choosing between identification schemes in noisy-news models, Unnamed Item, Forecasting macroeconomic variables in a small open economy: a comparison between small- and large-scale models, Forecasting using targeted diffusion indexes, Sparse and stable Markowitz portfolios

Uses Software

• PDCO

Cites Work

• Unnamed Item
• Unnamed Item
• Forecasting economic time series using targeted predictors
• Least angle regression. (With discussion)
• The generalized dynamic factor model consistency and rates
• Atomic Decomposition by Basis Pursuit
• Confidence Intervals for Diffusion Index Forecasts and Inference for Factor-Augmented Regressions
• Forecasting Using Principal Components From a Large Number of Predictors
• An iterative thresholding algorithm for linear inverse problems with a sparsity constraint
• Forecasting and conditional projection using realistic prior distributions
• Tests of Conditional Predictive Ability
• OPENING THE BLACK BOX: STRUCTURAL FACTOR MODELS WITH LARGE CROSS SECTIONS
• Inferential Theory for Factor Models of Large Dimensions
• Determining the Number of Factors in Approximate Factor Models
• The Generalized Dynamic Factor Model
• Benchmark priors for Bayesian model averaging.
• The elements of statistical learning. Data mining, inference, and prediction