Rethinking Performance Attribution in the Age of Systematic Investing and AI

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Traditional Brinson attribution remains useful when a manager’s process follows identifiable asset allocation and security selection decisions, but it may provide an incomplete or misleading explanation when portfolios employ derivatives, systematic strategies, factor models, and artificial intelligence. This article explains how practitioners can adapt or supplement Brinson attribution with exposure-based, factor, risk, decision, and model-explainability analyses that better reflect how modern portfolios are managed. It also presents a practical framework for communicating results to clients and fiduciaries by explaining what drove performance, whether the portfolio behaved as intended, and what the analysis cannot establish. Generative AI can assist in producing clear and tailored explanations, provided its output remains grounded in validated analytics, traceable data, appropriate controls, and expert review.