How much alpha can public information generate in realistic settings? We simulate a highly skilled public-information user with an AI-powered analyst that selectively modifies mutual funds’ quarterly portfolios while preserving styles, risks, and costs. From 1990–2020, actual managers generate $2.6M in quarterly alpha. Our hypothetical analyst generates an incremental $12.7M and outperforms 90% of managers over their careers. It also generates higher information ratios and first-order stochastically dominates human managers, implying outperformance across non-standard objectives. These findings indicate that skilled public information trading is a significant driver of asset managers’ performance and provide a credible point estimate of its value.
| Speaker: | Dr Ed deHaan The MBA Class of 1963 Professor of Management, Stanford University |
| When: |
- |