Studies of Mutual Funds in 401(k) Plans

  • Yue Ye

Student thesis: Phd

Abstract

This thesis investigates the investments in 401(k) plans, one of the most popular types of defined contribution (DC) plans in the US. It consists of three essays related to this topic. In the first essay, I study the quality of 401(k) menus provided by plan sponsors. I find that the relationship between the Sharpe ratios of 401(k) menus and the number of funds on the menu is inverted U-shaped. With regard to the adequacy of 401(k) menus, I employ three spanning tests, the Wald, Likelihood Ratio (LR) and Kuhn-Tucker (KT) tests. The KT test shows that 42% of the 401(k) menus in my sample can span the eight indices, which is the highest spanning rate, and which does not come close to the spanning rate achieved in the existing literature (e.g. Elton, Gruber, and Blake, 2006, 53%; Tang, Mitchell, Mottola, and Utkus, 2010, 94%). One possible reason for my finding of a lower spanning rate compared to the existing literature is a time variation in the spanning rate. Over the 8-year sample period, the spanning rate ranges from 11% in 2010 to 80% in 2015. My findings suggest plan sponsors add bond indices to 401(k) menus. In the second essay, I investigate the performance of US domestic equity funds selected by 401(k) plan sponsors. I construct a set of control funds comparable to funds included in 401(k) plans by matching their fund size, age, and risk loadings. I find that, on average, there is no superior performance of 401(k) plans, compared to control funds. I then sort funds into size quintiles and find that the 20% largest plan funds significantly underperform control funds by 0.91% per annum. I show that this underperformance is partly due to higher expense ratios of the largest plan funds, compared to control funds. However, my further investigation of the 401(k) asset allocations suggests that plan participants tilt away from large funds with the worst performance. In the third essay, I examine the performance of target-date funds (TDFs), which change their asset allocation periodically to match the changing risk preferences of the investors as they approach retirement. I sort TDFs by their horizons, i.e., the time period to a fund's target date. I find that TDFs with long horizons underperform TDFs with short horizons, which is consistent with the previous literature. Then I match plan TDFs and control TDFs by their horizon. My results show that TDFs selected by 401(k) plans cannot outperform control funds at any horizon. This result is robust to two different risk-adjusting processes. I further investigate the 401(k) asset allocation and show that the allocation-weighted performance of plan TDFs is superior to their equal-weighted performance. My results suggest that this outperformance is due to the total expense ratios of TDFs, indicating that plan participants prefer low-cost TDFs.
Date of Award1 Aug 2023
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorAlex Taylor (Supervisor) & Stefan Petry (Supervisor)

Keywords

  • fund performance
  • 401(k) plans
  • mutual funds
  • TDFs

Cite this

'