RESEARCH IN PROGRESS

Selection Bias in Mutual Fund Advertisements (w/ M. Mercer)

A financial company that operates multiple mutual funds may try to appeal to investors by selectively advertising the performance of its most successful funds.  If so, a question arises as to whether investors appreciate that fund performance data that a company cherry picks for inclusion in its ads are diagnostically inferior to data that arise from a complete sampling of funds.  We investigate these “selection bias” issues in one archival study and two experimental studies.  Study 1 finds that management companies selectively advertise their high-performing funds while ignoring low performing products.  Study 2 finds that both notice and expert investors are completely insensitive to selection bias in mutual fund advertisements unless the advertisement includes information that makes the selective nature of the performance data transparent.  As a result, investors are relatively anxious to invest in unproven companies and products.  Study 3 replicates and extends Study 2 by showing the identical pattern of what we call “partial sensitivity to selection bias” when the advertising company is a large well-known company.  We conclude that financial companies engage in selective advertising, and that investors are likely to overuse the performance data contained in these advertisements unless the potentially biased nature of the data selection process is made transparent.

.