Mutual Fund Performance : Using Bespoke Benchmarks to Disentangle Mandates, Constraints and Skill

While no two mutual funds are alike in terms of their mandates and constraints, benchmarks used to evaluate managers fail to account for these differences by typically using unadjusted indices as performance benchmarks. In this paper, we develop a methodology to evaluate a portfolio manager's performance relative to a conditional multifactor benchmark that explicitly incorporates the fund's mandate and constraints. After accounting for the impact of the mandate and constraints, we can provide an accurate measure of a portfolio manager's skill.At the center of our study are the mandates and constraints that shape the investment strategy underlying a sample of active funds. Mandates and constraints take on many different forms: investment universe, limits on borrowing/lending – being fully invested, concentration limits, risk exposures. In order to gather the constraint data for our funds, we intend to review the prospectus for each one of our sample funds. Specifically, we plan to hand collect data from the 485A and 485B Prospectus of Security (POS) as well as the Statement of Additional Information (SAI) submitted to the SEC to record constraint data on 4 dimensions: security holdings, investment level, other securities and benchmarks.Our analysis and results have important implications for mutual fund investors, mutual fund advisors/boards, and academics. From an investor perspective, having a mechanism to evaluate the performance of an active manager given their mandates and constraints is invaluable. Being able to gauge mandate/constraint adjusted performance may be the difference between lost time and money in a poor fund and the receipt of positive alpha from a skillfully run fund. Similarly, mutual fund advisor firms and mutual fund board of directors should be evaluating portfolio managers based on mandate/constraint adjusted performance for compensation and retention purposes to optimally execute their fiduciary responsibility to their shareholders. Finally, academics should be incorporating these practical frictions into our standard asset pricing models