Publications:
Portfolio Selection under Systemic Risk [Link]
with Jose Olmo and Abderrahim Taamouti (2025)
Journal of Money, Credit and Banking (ABS 4)
Portfolio Selection under Non-Gaussianity and Systemic Risk: A Machine Learning Based Forecasting Approach [Link]
with Abderrahim Taamouti (2024)
International Journal of Forecasting (ABS 3)
Working Papers:
[New] Systemic Growth-at-Risk and Growth Spread Measures [SSRN Link]
with Abderrahim Taamouti (2025)
(Under Review)
Abstract:
This paper develops two forward-looking macro-financial metrics – Systemic Growth-at-Risk (GaR) and Growth Spread Measures – to assess how growth risks propagate across countries and to quantify the net benefits of regional integration. Applied to 18 OECD EU countries, we estimate time-varying growth distributions using GARCH-Copula and GARCH-Dynamic Conditional Correlation (DCC) models, capturing both idiosyncratic shocks and cross-country interdependence. We find that integration enhances shared growth potential but increases systemic risk exposure. Growth dividends from EU membership are heterogeneous and correlate with trade openness, fiscal stance, development level, and global uncertainty. The framework extends GaR to multi-country settings, showing economic unions as both stabilizers and amplifiers of risk.
Enhancing Portfolio Resilience to Systemic Risk: A Neural Network Approach [SSRN Link]
with Abderrahim Taamouti (2025)
(Revise & Resubmit to Journal of Empirical Finance)
Abstract:
This paper aims to enhance the classical mean-variance portfolio selection model by using machine learning techniques and accounting for systemic risk. The optimal portfolio is obtained through a three-step supervised learning model. Firstly, the Smooth Pinball Neural Network is employed to predict return distributions of individual assets and the market. Secondly, we use a copula to model dependence between assets and the market, based on which we simulate return scenarios. Lastly, we maximize an ex-ante Sharpe ratio conditioning on systemic events. We train our models using a comprehensive dataset of nearly 600 U.S. individual stocks over 37 years; however, the portfolio analysis and backtesting are conducted on several representative portfolio sets with sizes up to 50 assets. This design allows the models to learn from a broad cross-section of firms while focusing the evaluation on computationally tractable portfolio dimensions. Our set of predictors includes 94 firm characteristics, 14 macroeconomic variables, and 74 industry dummies. The backtesting results demonstrate the superiority of our proposed approach over popular benchmark strategies including a GARCH-based model. This outperformance is statistically significant and robust to the inclusion of transaction costs.
Work in Progress:
Robust Estimation and Inference for Stock Return Predictability
with Hao Xu and Abderrahim Taamouti