Selecting Funds with Pairwise Elimination and a Fund Confidence Set

Summary

The document summarizes a fund-selection method that combines portfolio holdings and historical returns to estimate predictive alpha. It first screens for funds with enough data and positive performance persistence, then compares candidates pair by pair. A fund is eliminated when its risk-adjusted return is clearly inferior; when two funds have highly correlated returns, the weaker performer is removed as their return sources may be similar. The survivors form a Fund Confidence Set (FCS), which is then used to build a portfolio with mean-variance weights.

The summary says the underlying empirical study found higher risk-adjusted returns for portfolios formed after defining the FCS than for portfolios based on simple alpha rankings. It provides no sample details, numerical results, or implementation specifics, and the linked paper is not reproduced in the document. The claimed advantage should therefore be read as a reported finding rather than enough evidence to assess robustness, costs, or performance in other markets and periods.

Key ideas

• The method estimates predictive alpha using fund holdings and historical returns.
• It screens for sufficient data and positive performance persistence before comparing funds.
• Pairwise comparisons eliminate the weaker fund based on risk-adjusted returns and return correlation.
• The surviving Fund Confidence Set receives portfolio weights through mean-variance optimization.
• The document reports better risk-adjusted returns than simple alpha ranking but gives no supporting details.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.