Centre De Referència En Economia Analítica Barcelona Economics Working Paper Series Working Paper Nº 92 Honey, I Shrunk the Sample Covariance Matrix Honey, I Shrunk the Sample Covariance Matrix

By Oliver Ledoit, Michael Wolf, Olivier Ledoit

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Communication-Avoiding Optimization Methods for Distributed Massive-Scale Sparse Inverse Covariance Estimation (2017, background): “(cid:29) This problem, known as sparse inverse covariance estimation, is an important one in statistics, and plays a role in a wide variety of real-world applications, including ﬁnance ( e.g. , [34, 32, 53, 27, 4]), biology ( e.g. , [22, 27, 38]), and sustainability ( e.g. , [54, 4, 3]).”
Forest Through the Trees: Building Cross-Sections of Stock Returns (2020, unknown): “, Ledoit and Wolf (2004), or using a lasso-type penalty in the mean-variance estimation to obtain sparse portfolio weights, e.”
Shrinking the Cross Section∗ (2017, unknown): “The scaled identity matrix proposed in Ledoit and Wolf (2004a) is most appropriate in our empirical setting of zero-β anomaly portfolios.”
Large Dynamic Covariance Matrices (2019, methodology): “DCC-L1: the portfolio (6.4), where the estimator Ĥt is obtained fromDCC based on the linear shrinkage of Ledoit and Wolf (2004b) with shrinkage target given by (a multiple of) the identity matrix.”
SSD: A Unified Framework for Self-Supervised Outlier Detection (2021, methodology): “Finally, we use the Ledoit-Wolf method (Ledoit & Wolf, 2004) to estimate the covariance of the OOD samples.”
Parameter uncertainty in portfolio selection: Shrinking the inverse covariance matrix (2012, methodology): “…counterpart Σ̂ and a shrinkage target Λ: SLW = (1− f)Σ̂ + fΛ (14) The set of typical shrinkage targets includes the identity matrix (Ledoit and Wolf 2004a), the constant correlations matrix (Ledoit and Wolf 2004b) and a covariance matrix that corresponds to a 1-factor model (Ledoit and Wolf 2003).”
Nonlinear Shrinkage of the Covariance Matrix for Portfolio Selection: Markowitz Meets Goldilocks (2017, unknown): “Despite is relative simplicity, Markowitz (1952) portfolio selection remains a cornerstone of finance, both for researches and fund managers.”
Cluster analysis for portfolio optimization (2005, background): “…1985), the realized Sharpe ratio of the global minimum variance portfolio (Jagannathan and Ma, 2003), the role of constraints in portfolio optimization (Eichhorn et al., 1998; Jagannathan and Ma, 2003), the shrinkage estimator of large dimensional covariance matrices (Ledoit and Wolf, 2004a,b).”
Portfolio Selection with Parameter and Model Uncertainty: A Multi-Prior Approach (2004, unknown): “We have made a conscious decision to use as a starting point of our analysis the static mean-variance portfolio model of Markowitz (1952) rather than the dynamic portfolio selection model of Merton (1971).”
Improved Estimates of Higher-Order Comoments and Implications for Portfolio Selection (2010, methodology): “Finally, γij denotes the squared error of the structured estimator, with a consistent estimator given in Ledoit and Wolf (2003, 2004).”
Federated Learning via Posterior Averaging: A New Perspective and Practical Algorithms (2020, unknown): “…ˆ θ (cid:96) } , let ˆ µ (cid:96) be the sample mean, ˆ S (cid:96) be the sample covariance, and ˆ Σ (cid:96) := ρ (cid:96) I + (1 − ρ (cid:96) )ˆ S (cid:96) be a shrinkage estimator [Ledoit and Wolf, 2004b] of the covariance with ρ (cid:96) := 1 / (1 + ( (cid:96) − 1) ρ ) for some ρ ∈ [0 , + ∞ ) .”