We study the performance of actively managed US equity mutual funds using traditional models and, as in previous studies, find that they perform negatively. At the same time, we note that the investments in mutual funds increase each year. It thus doesn’t seem realistic to admit that mutual fund clients would continually accept negative performances. We put forth the idea that traditional measures of performance are misleading from a client’s point of view. Expenses are justified by managers as part of their information acquisition activity. If managers are successful, clients believe to be protected (at least partially) against information risk. It follows that, from the client’s point of view, the performance should be calculated as the mutual fund net realized return minus an expected return which only accounts for traditional risk premia factors (like systematic, size, book to market and momentum factors) and not for any information risk factor. We show in this paper that traditional mutual fund performance models are not in line with this idea because the traditional factors used in these models (market, size, book-to-market, momentum) embed an information risk premium. Based on the Merton (1987) model and on the Firm Specific Return Variation variable of Durnev et al. (2004), we compute an information risk factor. We then show that the traditional mutual funds performance models undervalue the funds alpha since they control the fund net realized return for an information risk premium when they shouldn’t from the client’s point of view. Finally, we propose a new methodology for measuring the mutual funds performance from the client’s perspective. We conclude with this new methodology that the performance of the US equity mutual funds is well explained.