Alejandro Rodriguez Dominguez on Portfolio Construction and Model Risk

Video

Building More Resilient Portfolios Through Model Diversity

At Advanced Model Risk Europe, Alejandro Rodriguez Dominguez explores a new approach to portfolio construction, reframing diversification as a multi-hypothesis prediction problem. He discusses how linking model diversity directly to out-of-sample risk can enhance robustness, improve generalisation, and create more resilient portfolios under stress.

Apr 13, 2026

Alejandro Rodriguez Dominguez, Head of Quantitative Analysis and Artificial Intelligence, Miraltabank

Tags: Model risk

The views and opinions expressed in this content are those of the thought leader as an individual and are not attributed to CeFPro or any other organization

In this interview, Alejandro Rodriguez Dominguez discusses how traditional approaches to portfolio construction can be reimagined by treating assets as hypotheses within a structured predictive framework. By establishing a formal connection between diversity in predictive models and portfolio-level risk diversification, he highlights how institutions can better control risk outcomes before optimisation even begins.

He also explores the practical implications of this approach for both quantitative teams and risk practitioners, including improved robustness, stronger performance under stress, and greater alignment between modelling techniques and real-world risk management objectives. The discussion offers a forward-looking perspective on how model-driven portfolio design can evolve in increasingly complex market environments.

Alejandro Rodriguez Dominguez Bio

Alejandro Rodriguez is Head of Quantitative Analysis and AI at Miraltabank, where he leads the R&D of quantitative solutions across risk management, portfolio construction, credit analysis, data analytics and AI. He has published peer-reviewed work in causal portfolio optimization, correlation change-point detection, AI-driven risk diversification, and multi-hypothesis prediction. He is also a Quant Advisor at Inspiration-Q, a startup providing quantum-inspired optimization tools for portfolio management and arbitrage strategies. Previously, he worked in London at Société Générale and Nomura as a financial engineer, and earlier at BBVA and BNP Paribas as a sales trader. He holds an M.Eng in Engineering and several MSc degrees in Statistics, AI, and Finance, and is currently a PhD candidate in AI at the University of Reading.