Over the last two weeks, we've participated (marginally) in a discussion of the Investment Company Institute's recent survey of costs borne by 401(k) plans and participants. Elements of that discussion--including our initial post on the topic--have focused not just on the findings of the survey, but on ICI's descriptions thereof.
Writing at BrightScope's blog, Ryan Alfred advanced the debate by examining the ICI survey's sample. That sample leaned heavily on large plans, which tend to be less expensive than small plans, though there's enormous variance even among the biggest ones.
Yesterday, the ICI posted a comment on Ryan's response at BrightScope's blog and a longer response at its own website. We're pleased to see this discussion continue. It's important. And judging by the progress we've already made, it should help sponsors, participants, service providers, policy-makers, and reporters better understand the 401(k) marketplace.
Here are a few thoughts on the ICI's response to BrightScope...
Sampling. Here's ICI on this key factor:
This is a perfectly reasonable approach. As we noted in our initial item on this topic, the biggest problem isn't with the sample per se; it's with the ICI's description of the report's findings. About those adjectives: We'll grant ICI "extensive," but we won't quite go for "comprehensive," and we certainly wouldn't grant them the most important descriptor (which they don't use)...representative.
What ICI Has Said About the Survey. ICI notes that in his description of the report's findings, ICI Chairman John Murphy said this (emphasis added in bold by ICI):
In fact, however, ICI's estimate isn't an "all-in" fee, at least if you consider "fee" and "expense" practical synonyms in this context. By excluding trading costs (more on that in a moment), ICI compares its incomplete picture of what participants pay with others' claims about plans' and participants' true and total costs. In addition, for all the reasons we cited back on May 8th, ICI's "0.72%" excludes more than just implicit fund expenses. Bottom line: We need to get everyone on the same page, talking about the same elements of plan expenses in the same terms.
What Do 401(k) Participants Pay? The most important point made in ICI's response is that while most plans are small, most people participate in large plans. Weighting aggregate, system-wide estimates of expenses by assets or participants is fine. But it can't obscure the fact that, as an anonymous ICI employee posted at BrightScope's blog, "small and tiny plans" need help. And it shouldn't obscure the fact that a lot of non-small, non-tiny plans need help too.
Trading Costs. In a short summary of its response to BrightScope, posted at BrightScope's blog, ICI calls trading costs "a complicated issue." Indeed. But it's not so complicated that we can't get serious about understanding it and--just as important--including it in all discussions of plans' and participants' true and total costs. Here's what we know:
- Trading costs are, as ICI rightly points out, unavoidable.They can also be very substantial indeed.
- All else equal, they tend to be higher for funds in relatively illiquid asset classes (due to larger bid-ask spreads).
- All else equal, they tend to be higher for funds with more assets under management (due to larger market impact).
- All else equal, they tend to be higher for high-turnover funds (due to trading commissions and the fact that high turnover raises spread and impact costs).
To the extent that plan participants and fiduciaries care about implicit costs, as they should, they can compare investment options on the basis of asset class liquidity, fund size, and turnover. The point here isn't that big, high-turnover funds in less-liquid asset classes are always and everywhere bad choices. It's that sponsors and participants, guided by fiduciary service providers, should understand the nature and impact of these costs, and, armed with that knowledge, make prudent decisions about their investment options.
Big Ideas. The ongoing debate about expenses is important. The sooner we can establish a shared framework for comparing plans' and participants' true and total costs, the sooner we can agree on a new regulatory framework and industry standards to match. But we shouldn't let this talk about expenses and transparency distract us from the fundamental objective: to give participants and their beneficiaries a better shot at a decent retirement. Simply put, it's all about outcomes.
We care about expenses because, all else equal, lower expenses will produce better outcomes. We care about transparency because we think more and better information will help sponsors and participants correct the informational asymmetry that continues to plague the 401(k) marketplace. And we care about the 401(k) marketplace because it has profound implications for participants' economic outcomes.
So let's get disclosure right. Let's eliminate unnecessary expenses and shrink the rest. Let's replace the industry's selling machine with a thoroughly fiduciary model. Let's do all those things. But let's also remember that for the vast majority of participants, long-term outcomes are more a function of their investment programs--their original formulation, their ongoing maintenance, their periodic changes--than of the costs they pay to participate in these plans.
As we've noted in this space, a true fiduciary plan features low expenses and good decision-making. Unfortunately, the current model--in which participants (even those in the least expensive plans) are presented with a menu (which is often incomplete) of mutual funds (which are often mediocre) and left to fend for themselves--tends to produce all sorts of dysfunctional behavior and sub-optimal outcomes.*
So above all, let's organize the entire defined-contribution model around the inherently fiduciary concept of improving participants' long-term outcomes. That means modest expenses, full transparency, and easy access to superb, portfolio-centered (as opposed to fund-driven) investment programs.
* See this post from late 2007, "Real Participants, Real Problems," for our take on The Uncertain 1/n Hedger, The Shameless Performance-Chaser, and other categories of participants behaving badly.