Current browse context:
econ
Change to browse by:
References & Citations
Economics > Econometrics
Title: Optimal Portfolio Using Factor Graphical Lasso
(Submitted on 1 Nov 2020 (v1), last revised 1 Apr 2023 (this version, v5))
Abstract: Graphical models are a powerful tool to estimate a high-dimensional inverse covariance (precision) matrix, which has been applied for a portfolio allocation problem. The assumption made by these models is a sparsity of the precision matrix. However, when stock returns are driven by common factors, such assumption does not hold. We address this limitation and develop a framework, Factor Graphical Lasso (FGL), which integrates graphical models with the factor structure in the context of portfolio allocation by decomposing a precision matrix into low-rank and sparse components. Our theoretical results and simulations show that FGL consistently estimates the portfolio weights and risk exposure and also that FGL is robust to heavy-tailed distributions which makes our method suitable for financial applications. FGL-based portfolios are shown to exhibit superior performance over several prominent competitors including equal-weighted and Index portfolios in the empirical application for the S&P500 constituents.
Submission history
From: Ekaterina Seregina [view email][v1] Sun, 1 Nov 2020 06:01:08 GMT (122kb,D)
[v2] Fri, 6 Nov 2020 07:25:02 GMT (112kb,D)
[v3] Wed, 14 Apr 2021 01:12:53 GMT (146kb,D)
[v4] Sun, 4 Sep 2022 21:02:50 GMT (170kb,D)
[v5] Sat, 1 Apr 2023 18:41:13 GMT (879kb,D)
Link back to: arXiv, form interface, contact.