A Three-Dimensional Efficient Surface for Portfolio Optimization

Battle Count: 69

Highly relevant to quantitative trading and portfolio management. The framework directly extends Markowitz mean-variance optimization to incorporate network/systemic risk, which is critical for institutional portfolio managers and systematic strategies. The connectedness beta provides a practical diagnostic tool for identifying systemic hubs. The three-fund separation result offers a tractable implementation path. The empirical hypotheses (H1-H3) address practical concerns: tail risk reduction, factor timing, and implementation costs. However, the incomplete empirical section and strong theoretical assumptions (simultaneous diagonalization) temper immediate practical applicability.

Moderate to high complexity. Requires: (1) VAR estimation and FEVD computation for the connectedness matrix C (standard econometric procedure), (2) shrinkage covariance estimation for Σ, (3) solving quadratic programs with potentially long-only constraints (active-set or interior-point methods), (4) rolling-window re-estimation for dynamic surfaces, (5) three-fund decomposition and barycentric coordinate computation. The closed-form solution under simultaneous diagonalization simplifies implementation, but the general case requires numerical optimization. The active-set algorithm (Algorithm 1) is provided but requires careful implementation for large N.

The paper provides detailed theoretical proofs (Propositions 1-4, Theorem 1) and algorithmic pseudo-code (Algorithm 1 for active-set solver). However, the empirical section appears incomplete (references to 'Section??' and 'Table??' suggest missing content). Data sources are mentioned (CRSP, one-month Treasury bill) but no code repository or exact replication scripts are provided. The VAR-FEVD estimation procedure is described at a high level but lacks full implementation details.