This paper revisits the classic problems of equity fund selection and portfolio construction through a novel perspective that emphasizes practical implementation. The authors propose a framework that diverges from conventional approaches that typically construct long-only portfolios from large universes of stocks or macro factors. Instead, their method focuses on creating long-short portfolios from a smaller, more targeted pool of stocks derived from mutual fund top holdings data. This approach aims to provide investment professionals with an easy-to-implement solution that can be readily applied in real-world settings.

The methodology is grounded in statistical evidence, requiring careful monitoring of model validity and the development of repair strategies to maintain performance over time. The research demonstrates that this focused approach can generate impressive results, potentially outperforming traditional methods while offering greater practicality for investment applications. The paper contributes to both theoretical understanding and practical implementation in quantitative finance by bridging the gap between academic portfolio theory and real-world investment constraints.