How many principal components for optimal portfolio selection?

Details

Supervisors

Lassance, Nathan

Faculty

Louvain School of Management

Degree label

Master [120] : ingénieur de gestion, à finalité spécialisée

Abstract

enStandard mean-variance optimization is highly sensitive to estimation errors in high-dimensional markets, justifying the use of principal component analysis (PCA) to stabilize covariance matrices by filtering out idiosyncratic noise. However, since traditional methods for selecting the number of retained components (K) rely on purely statistical heuristics, this thesis proposes a new portfolio-driven cross-validation framework that dynamically optimizes K to explicitly minimize out-of-sample risk.