2025/11/14 by Christian Bongiorno, Efstratios Manolakis, Rosario N. Mantegna +1 · 1 citation
Decision Sciences · Economics, Econometrics and Finance · #Financial Markets and Investment Strategies #Risk and Portfolio Optimization #Stock Market Forecasting Methods #cs.LG #q-fin.PM
paper · pdf · doi:10.1145/3768292.3770370
openalex created_date 2025/11/14 · openalex publication_date 2025/11/14 · openalex updated_date 2026/07/30
This paper introduces a compact reformulation of a modular, end-to-end neural network for global minimum-variance portfolio optimization that decouples model complexity from both look-back window length and universe size. A five-parameter hyperbolic weighted moving average combined with a saturating exponential replaces the original 2,400-parameter lag-transformation layer, and a bidirectional gated-recurrent-unit eigencleaning module together with a streamlined marginal-volatility network reduce total learnable parameters from 39,586 to just 2,175. In out-of-sample tests against state-of-the-art nonlinear-shrinkage and risk-parity benchmarks, the compact network attains the lowest realized portfolio variance without compromising expected return. Under long-only constraints, the variance reduction supports substantially higher leverage while maintaining comparable drawdown control. Validation in a high-fidelity trading simulator that incorporates realistic margin-call dynamics confirms enhanced over-leverage resilience. These findings demonstrate that end-to-end variance-minimization architectures can achieve substantial parameter efficiency and robust capital-efficiency gains without sacrificing risk-adjusted performance.