Some properties of portfolios constructed from principal components of asset returns

Thomas A. Severini*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Principal components analysis (PCA) is a well-known statistical method used to analyze the covariance structure of a random vector and for dimension reduction. When applied to an N-dimensional random vector of asset returns, PCA produces a set of N principal components, linear functions of the asset return vector that are mutually uncorrelated and which have some important statistical properties. The purpose of this paper is to consider the properties of portfolios based on such principal components, know as PC portfolios, including the efficiency of PC portfolios, the use of PC portfolios to reduce the return variance of a given portfolio, and the properties of factor models with PC portfolios as factors.

Original languageEnglish (US)
Pages (from-to)457-483
Number of pages27
JournalAnnals of Finance
Volume18
Issue number4
DOIs
StatePublished - Dec 2022

Keywords

  • Dimension reduction
  • Efficient frontier
  • Factor models
  • Minimum-risk frontier
  • Portfolio theory

ASJC Scopus subject areas

  • Finance
  • Economics, Econometrics and Finance(all)

Fingerprint

Dive into the research topics of 'Some properties of portfolios constructed from principal components of asset returns'. Together they form a unique fingerprint.

Cite this