Essays on taxation and mutual funds

Abstract

The first chapter of this dissertation examines the behavior of long- and short-term capital gains distributions after the passage of the Tax Reform Act of 1997, which lowered the maximum tax rate on long-term gains. Using a panel of mutual fund data, I examine the ratio of long-term to total gains distributions around the time of the Act, and find that fund managers appear to tilttheir distributions towards the long-term after 1997. This behavior is consistent with the hypothesis that managers are tax-sensitive, and the estimates are robust to the inclusion of fund-level fixed effects and other controls. I also examine fund capital gains patterns in a difference-in-differences framework, comparing actively managed to index funds; this technique gives a lower-bound estimate of the increase in the fraction of gains paid-out as long-term after the Act of five percentage points. The second chapter examines equity mutual fund dividend yields. Dividends are highly-taxed for individual investors, but tax-favored for corporate investors. Consequently, I hypothesize that corporate investors may prefer to hold higher-dividend yield funds than non-corporate investors. I use institutional funds as a proxy for corporate, trust, or non-profit shareholders, and find that these funds do have systematically higher dividend yields than their retail counterparts. These results are consistent with the tax clientele hypothesis, and are robust to the inclusion of a number of fund- and industry-level controls.

(cont.) Chapter three, co-authored with James Poterba, documents the increasing use of redemption fees throughout the mutual fund industry. These fees are levied against investors who liquidate their positions before a specified time, and are paid back into the fund to compensate existing investors for the negative externalities of redemptions. We find that foreign and sector funds are more likely to impose redemption fees, and that institutional funds are less likely to have fees. Using data from the SEC, we find that the largest funds with redemption fees collected nearly $58 million in fees in 2000. We also find that funds with redemption fees appear to have lower turnover and higher returns than their no-fee counterparts, controlling for time, fund objective, and other characteristics.