This past summer, the Massachusetts Institute of Technology, the University of Pennsylvania, and Yale, Columbia, Duke, and New York Universities, along with at least five other major colleges, were served with lawsuits alleging failure of fiduciary duty toward participants in their retirement plans. The retirement plans at all of these institutions either have or previously had a large number of investment options. In addition, some of the investment options were too costly, given the plan size. The lawsuits claim that offering too many investment options — in some cases, more than a hundred — is imprudent, and that the universities failed to use their size to negotiate lower fees.
The lawsuits will no doubt spark a wave of changes in the design of higher-education retirement plans. As colleges summon their lawyers and consultants to revisit their plans, it is worthwhile to review what academic economists know about the principles that should guide the design of retirement plans.
The past 20 years have produced economic research that should carry weight not only in how the courts decide the merits of the lawsuits, but also in how administrators at universities not yet hit with a lawsuit should continue reshaping their retirement plans.
Retirement plans are likely to benefit from simplifying the array of options they offer.
Here are some key principles to guide them:
Offer access to key asset classes.
These include stocks, bonds, and money-market accounts, among others. Access to different asset classes enables employees to invest in assets that match their risk preferences. For example, while younger workers may want to invest more in stocks, older ones may find bonds more appropriate. Economists have long recognized that asset allocation plays a key role in the risk and return of a portfolio. They have also recognized that asset allocation may need to vary with age. There is less agreement on what constitutes a key asset class. While few economists would object to U.S. stocks as a key asset class, there will be less agreement with regards to emerging markets or real estate.
Minimize cost.
Plans should offer access to key asset classes at the lowest possible cost. Generally, the least costly way to invest in a particular asset class is to buy an index fund that tracks that particular asset class. Index funds buy all securities in an index rather than actively managing which securities to buy and which to sell. One thing that economists know for sure is that active management is a game in which the average investor loses. As Dartmouth’s Kenneth R. French said in a 2008 address to the American Finance Association, “The typical investor would increase his average annual return ... if he switched to a passive market portfolio.” Nevertheless, plans should recognize that not all index funds are created equal, and should carefully monitor their costs.
Don’t chase returns.
Plans should not change the investment menu too often. Economists know that past returns don’t predict long-term future returns. Thus, selecting funds based on their past performance is silly. One study found that plans usually add funds that do no better than the funds they dropped. The good news is that if plans invest in low-cost index funds as outlined above, their performance will match their benchmarks. Index funds do not under-perform or out-perform benchmarks — they are the benchmarks.
Keep it simple.
Plans should have few investment options, and what they do have should be simple. Some of the most influential work in economics has been in the area of making participation in defined-contribution retirement plans more straightforward. This research gave rise to the increasing use of options like automatic enrollment, qualified default investments, and automatic escalation in contributions.
A related line of research showed that participants in defined-contribution plans are largely passive, don’t take full advantage of their employer’s matching contributions, and have a hard time differentiating among investment options. Thus, simplifying plans and providing participant education is essential.
Avoid conflicts of interest.
Plans should hire advisers who are free of conflicts of interest — and starting next year, anyone advising retirement assets will have fiduciary duty, under a new Labor Department rule. Economists have long recognized that financial markets are plagued with such conflicts. For example, economists found that advisers who receive indirect compensation give conflicted advice, that advisers’ own 401(k) plans are cheaper than those of their clients, and that mutual-fund companies that serve as trustees of 401(k) plans favor their own funds. A more subtle conflict of interest arises when advisers over-treat their clients by adding complexity to the products they offer.
Even in the absence of indirect compensation, advisers have an incentive to provide (and bill for) more services than are necessary. For example, they may be able to justify higher fees for a plan with many options and frequent changes in the menu than for a plan with a few simple and stable options. It is best to avoid the more complicated approach and stick to the principles outlined here.
Be careful about annuities.
Plans should be careful about including annuities. Although economists have long recognized the benefits of life annuities in retirement, the best use of annuities within a defined-contribution plan is still the subject of much research. The problem is that, unlike with mutual funds, the pricing of annuities is far less transparent. This is particularly relevant for the 403(b) plans prevalent in the higher-education market since many of them include annuities.
Without question, there is room for improvement in higher-education retirement plans. In particular, they are likely to benefit from simplifying the array of options they tend to offer. An example of a well-designed plan is that of Stanford University. The plan includes only six core funds, in addition to target date/lifecycle funds and annuities. The funds are all low cost and passively managed, and the recordkeeping costs are covered through a flat annual charge to participants.
This is far more transparent and equitable than the typical practice of embedding recordkeeping costs in the expense ratios of different funds. The inclusion of annuity contracts is limited to three options, and they are clearly marked as distinct from the core options. While Stanford has the advantage of being a multibillion-dollar plan, the key features of its design can be replicated in a plan of any size.
I have no doubt that administrators at colleges and universities have the well-being of their faculty, staff, and retirees in mind. By using the fruits of economic research as outlined above, they may make more informed decisions as well as effectively respond to and be protected from litigation.