We show that scale in active equity portfolios is understated by at least 65% because the majority of mutual funds have “twin” institutional vehicles (IVs) managed under the same strategies. Omitting these IVs can severely skew crucial estimates in asset management research: by including IV assets, diminishing returns to scale of active investments is significantly reduced, and dollar value added of active strategies is more substantial and persistent than previously suggested. We further show that IV assets meaningfully influence managers’ portfolio decisions. In addition, these measurement issues apply to common flow measures and extend to passive funds and bond funds.

We show that mutual fund investors rely on simple signals and likely do not engage in sophisticated learning about managers’ alpha as widely believed. Simplistic performance chasing best explains aggregate flows to the mutual fund space and flows across funds. These results hold for both actively managed and passive index funds. Empirical patterns commonly interpreted as reflecting learning about managerial skill also appear in falsification tests and are mechanical. Our results are consistent with the view that, on average, households are homo sapiens with limited financial sophistication rather than hyperrational alpha-maximizing agents, as often assumed in the literature.

We show that mutual fund ratings generate correlated demand that creates systematic price fluctuations. Mutual fund investors chase fund performance via Morningstar ratings. Until June 2002, funds pursuing the same investment style had highly correlated ratings. Therefore, rating-chasing investors directed capital into winning styles, generating style-level price pressures, which reverted over time. In June 2002, Morningstar reformed its methodology of equalizing ratings across styles. Style-level correlated demand via mutual funds immediately became muted, significantly altering the time-series and cross-sectional variation in style returns.