Using principal component analysis to estimate a high dimensional factor model with high-frequency data

Yacine Aït-Sahalia; Dacheng Xiu

doi:10.1016/j.jeconom.2017.08.015

Using principal component analysis to estimate a high dimensional factor model with high-frequency data

Yacine Aït-Sahalia, Dacheng Xiu

Research output: Contribution to journal › Article › peer-review

120 Scopus citations

Abstract

This paper constructs an estimator for the number of common factors in a setting where both the sampling frequency and the number of variables increase. Empirically, we document that the covariance matrix of a large portfolio of US equities is well represented by a low rank common structure with sparse residual matrix. When employed for out-of-sample portfolio allocation, the proposed estimator largely outperforms the sample covariance estimator.

Original language	English (US)
Pages (from-to)	384-399
Number of pages	16
Journal	Journal of Econometrics
Volume	201
Issue number	2
DOIs	https://doi.org/10.1016/j.jeconom.2017.08.015
State	Published - Dec 2017

All Science Journal Classification (ASJC) codes

Economics and Econometrics
Applied Mathematics

Keywords

High-dimensional data
High-frequency data
Latent factor model
Portfolio optimization
Principal components

Access to Document

10.1016/j.jeconom.2017.08.015

Cite this

@article{e7528e1fd197447cb2b3194b9737c7ce,

title = "Using principal component analysis to estimate a high dimensional factor model with high-frequency data",

abstract = "This paper constructs an estimator for the number of common factors in a setting where both the sampling frequency and the number of variables increase. Empirically, we document that the covariance matrix of a large portfolio of US equities is well represented by a low rank common structure with sparse residual matrix. When employed for out-of-sample portfolio allocation, the proposed estimator largely outperforms the sample covariance estimator.",

keywords = "High-dimensional data, High-frequency data, Latent factor model, Portfolio optimization, Principal components",

author = "Yacine A{\"i}t-Sahalia and Dacheng Xiu",

note = "Publisher Copyright: {\textcopyright} 2017 Elsevier B.V.",

year = "2017",

month = dec,

doi = "10.1016/j.jeconom.2017.08.015",

language = "English (US)",

volume = "201",

pages = "384--399",

journal = "Journal of Econometrics",

issn = "0304-4076",

publisher = "Elsevier B.V.",

number = "2",

}

TY - JOUR

T1 - Using principal component analysis to estimate a high dimensional factor model with high-frequency data

AU - Aït-Sahalia, Yacine

AU - Xiu, Dacheng

PY - 2017/12

Y1 - 2017/12

N2 - This paper constructs an estimator for the number of common factors in a setting where both the sampling frequency and the number of variables increase. Empirically, we document that the covariance matrix of a large portfolio of US equities is well represented by a low rank common structure with sparse residual matrix. When employed for out-of-sample portfolio allocation, the proposed estimator largely outperforms the sample covariance estimator.

AB - This paper constructs an estimator for the number of common factors in a setting where both the sampling frequency and the number of variables increase. Empirically, we document that the covariance matrix of a large portfolio of US equities is well represented by a low rank common structure with sparse residual matrix. When employed for out-of-sample portfolio allocation, the proposed estimator largely outperforms the sample covariance estimator.

KW - High-dimensional data

KW - High-frequency data

KW - Latent factor model

KW - Portfolio optimization

KW - Principal components

UR - http://www.scopus.com/inward/record.url?scp=85026913044&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85026913044&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2017.08.015

DO - 10.1016/j.jeconom.2017.08.015

M3 - Article

AN - SCOPUS:85026913044

SN - 0304-4076

VL - 201

SP - 384

EP - 399

JO - Journal of Econometrics

JF - Journal of Econometrics

IS - 2

ER -

Using principal component analysis to estimate a high dimensional factor model with high-frequency data

Abstract

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this