One of the Democrats' proposals for solving the problem underscored by the Enron Corp. 401(k) losses is to require that defined contribution pension plans be run by a joint board of trustees with equal employee and management representation. It's a little unclear how this proposal would work. Apparently, the idea is that joint trustees would have final say over, at least, what sorts of funds are offered under the plan or, more generally, how the plan is run. How employee representatives would be elected is left for the secretary of labor to determine.
My first reaction to this proposal was negative. When I think about macro pension policy, I have only one goal in mind - improving retirement income. And it seemed to me that, unfortunately, participants are part of the problem.
It is well-known that single-employer defined benefit plans, which for the most part are managed exclusively by employers, outperform single-employer DC plans, largely because in the latter the critical investment decision - asset allocation - is made by the participant. The fact is - and there is a lot of data on this - participants typically don't make the best choices about how to invest their money. Generally, they are too risk averse, although, perversely, they also are likely to overallocate to single security investments, like company stock.
My concern is that putting participants in charge of picking fund menus and monitoring investment managers will only exacerbate this problem. They would introduce a degree of amateurishness, sentimentality and private agendas into the defined contribution plan investment process.
But advocates of joint trusteeship make several points that are hard to rebut. One, it's the participants' money. In contrast to defined benefit plans, the performance of defined contribution plans drops to the participants' bottom line. So why shouldn't they have some say in what funds they can choose and, generally, how the plan will be managed? Two, while there are a lot very well-run DC plans out there, some of them are not well-run at all, with critical choices being made by individuals who have no particular expertise, based on vague and ephemeral criteria. Three, employers have their own private agendas and sometimes use the investment of DC assets to advance them.
As I considered the problems in our current defined contribution system that joint trusteeship advocates have pointed out, I came to two conclusions.
First, we should have some sort of standards for service on a plan's fiduciary committee. A lot of fact-finding and listening will have to be done before we can adopt specific standards. Must a person be a financial professional to make these kind of decisions? What about participant sentiment - should it be indulged, even if it is unjustifiable from the point of view of a financial professional? And if so, how is it to be represented? Our goal is clear, though: making sure the people who make critical DC fiduciary decisions are adequately equipped to do so.
Second, and most important, we should have a set of standards for whether a plan is doing a good job for its participants. Here's one approach:
Create a benchmark set of investment options. Pick whatever panel of Nobel-prize-winning economists you want to develop this benchmark. I assume what you would get with this suite of options is some sort of approximation of the efficient frontier. Then pick a set of standard investment profiles - young/not-young, aggressive/not-aggressive - and allocate among the benchmark funds based on these profiles and your Nobel-team asset allocation model. From this exercise you get something approximating a benchmark rate of return for the different profiles.
Then apply this benchmark to the plan in two ways. First, rate the plan assuming your abstract profiles were optimized among the choices offered in the plan. That tells you how well a participant could do under the plan.
Second, calculate the actual rate of return for plan participants under the plan and compare that with the performance of the benchmark (taking into account the plan's demographic makeup). I realize that gathering this data would be a huge task. And you would get, at best, only a very rough comparison between the ideal benchmark and the real life of the plan. But only by doing so will we be able to judge the effectiveness of the decisions made by those managing the plan - the investment choices offered and the education provided to participants to assist them in making those choices.
By scoring plans this way you find out two very important things: whether it is possible for plan participants to get a reasonable rate of return, and what kind of return participants are actually getting. If this method is too difficult to implement, how about starting with something simple - have each plan provide three pieces of data: average return, average risk and average cost.
Employers are capable of adding incredible value to the management of 401(k) plan assets, by bundling participants' investment buying power, shaping a plan to fit employees' needs, providing expertise in the selection and monitoring of its options, and assisting participants in making asset allocation decisions. But we have to introduce some kind of discipline to this process - otherwise, employers, with the best of intentions, have no idea whether they are doing a good job or not. This means we must provide some kind of basis for making apples-to-apples comparisons. Finally, while I would very much like to be wrong about this, I think that some sort of legislation will be necessary in obtaining the sort of data we need to make these comparisons - I think of it as a 10-K for the plan.
I offer these suggestions as a first step. Before we can talk about who should serve on a joint board of trustees, before we can talk about what standards should be required of fiduciaries serving on DC fiduciary committees, we must have some established, objective measure of what a well-managed DC plan is.
Michael P. Barry is president of Plan Advisory Services Group, Chicago.
