High-Dimensional Dynamic Covariance Models with Random Forests
Preprint
Abstract
This paper introduces a novel nonparametric method for estimating high-dimensional dynamic covariance matrices with multiple conditioning covariates, leveraging random forests and supported by robust theoretical guarantees. Unlike traditional static methods, our dynamic nonparametric covariance models effectively capture distributional heterogeneity. Furthermore, unlike kernel-smoothing methods, which are restricted to a single conditioning covariate, our approach accommodates multiple covariates in a fully nonparametric framework. To the best of our knowledge, this is the first method to use random forests for estimating high-dimensional dynamic covariance matrices. In high-dimensional settings, we establish uniform consistency theory, providing nonasymptotic error rates and model selection properties, even when the response dimension grows sub-exponentially with the sample size. These results hold uniformly across a range of conditioning variables. The method's effectiveness is demonstrated through simulations and a stock dataset analysis, highlighting its ability to model complex dynamics in high-dimensional scenarios.Keywords
All Related Versions
This publication has 24 references indexed in Scilit:
- Regularized covariance matrix estimation in high dimensional approximate factor modelsStatistics & Probability Letters, 2024
- Predicting the Global Minimum Variance PortfolioJournal of Business & Economic Statistics, 2022
- Nonparametric estimation of large covariance matrices with conditional sparsityJournal of Econometrics, 2021
- High-Dimensional Dynamic Covariance Matrices With Homogeneous StructureJournal of Business & Economic Statistics, 2020
- A new semiparametric estimation approach for large dynamic covariance matrices with multiple conditioning variablesJournal of Econometrics, 2019
- Estimation and Inference of Heterogeneous Treatment Effects using Random ForestsJournal of the American Statistical Association, 2018
- Large Dynamic Covariance MatricesJournal of Business & Economic Statistics, 2017
- Dynamic Covariance ModelsJournal of the American Statistical Association, 2016
- Positive-Definite ℓ1-Penalized Estimation of Large Covariance MatricesJournal of the American Statistical Association, 2012
- Estimating the Global Minimum Variance PortfolioSchmalenbach Business Review, 2006