Research
Publications
“Tensor Principal Component Analysis” with Andrii Babii and Eric Ghysels, submitted
Abstract: In this paper, we develop new methods for analyzing high-dimensional tensor datasets. A tensor factor model describes a high-dimensional dataset as a sum of a low-rank component and an idiosyncratic noise, generalizing traditional factor models for panel data. We propose an estimation algorithm, called tensor principal component analysis (TPCA), which generalizes the traditional PCA applicable to panel data. The algorithm involves unfolding the tensor into a sequence of matrices along different dimensions and applying PCA to the unfolded matrices. We provide theoretical results on the consistency and asymptotic distribution for the TPCA estimator of loadings and factors. We also introduce a novel test for the number of factors in a tensor factor model. The TPCA and the test feature good performance in Monte Carlo experiments and are applied to sorted portfolios.
Recommended citation: Babii, Andrii, Ghysels, Eric, and Pan, Junsu. “Tensor Principal Component Analysis.” arXiv preprint arXiv:2212.12981 (2022).
Working Papers
“Missing Financial Data: Filling the Tensor Blanks” with Andrii Babii, Eric Ghysels, and Jiaxi Li
“Dynamic Portfolio Selection with Regularization”
Work in Progress
“Conditional Asset Pricing Factor Models with Firm Characteristics Tensor Data” with Andrii Babii and Eric Ghysels
“Identification and Estimation of Factor Models Through Coskewness Tensor” with Andrii Babii and Eric Ghysels