Barry’s Pickings: We Need Standardized Lifetime Income Disclosure, ASAP

Art by Joe Ciardiello

I have a strong bias against forcing other people, e.g., 401(k) plan participants, to do “what we think is right.” People generally should be (channeling Milton Friedman) free to choose. But in many cases it is useful, for instance, to require that they be provided clear and transparent information on the basis of which to choose.

Moreover, one of the main lessons (perhaps the lesson) of behavioral economics is that people have a biologically encoded tendency to underestimate their future needs and a specific tendency to discount their need for retirement income. As we have begun living into our 80s instead of dying in our early 40s, that bias (“hyperbolic discounting”) has been recognized a problem. Most individuals are likely to benefit from some gentle “nudging” (channeling Professors Thaler and Sunstein) towards more emphasis on future needs and risks and towards more savings.

With respect to two of the three major defined contribution (DC) plan adequacy challenges—getting participants to contribute enough and to make efficient asset allocation choices—these sometimes conflicting values have, in many respects, been synthesized in a regime of systematic disclosure and defaults that nevertheless preserves the right of the participant to opt out of the default regime.

With respect to the third DC plan adequacy challenge—getting participants to adopt a DC plan distribution strategy that will last for their remaining life—we have no such framework. At best we have a set of disparate, voluntary and ad hoc attempts by some employers to address the issue.

The Senate’s lifetime income disclosure solution

In this context, I want to put myself down as strongly in favor of a consistent lifetime income disclosure rule, as (at least) the beginning of an attempt to address this issue.

In this regard, there is (and has been for some time) bipartisan Senate legislation (e.g., the Retirement Enhancement and Savings Act of 2018 (RESA)) that would require DC plan administrators to provide participants (annually) a description of the lifetime income stream equivalent of their account balance. For this purpose, “lifetime income stream equivalent” would be the monthly amount of lifetime benefit that could be funded by the participant’s account balance, based on assumptions specified in rules prescribed by the Department of Labor (DOL).

Sponsor community objections

This proposal has generally been opposed by the sponsor community. Their objections focus on its (1) potential to confuse, (2) cost and (3) rigidity. As an alternative, the ERISA Industry Committee (ERIC), for instance, has proposed that DOL develop an “illustration tool” that participants can use, based on their own inputs, to draw their own conclusions about what sort of income stream their account balance will produce. And a requirement that plan administrators provide a notice about the availability of that tool.

I could be wrong on this, but, in my humble opinion that proposal is a non-solution to the problem we’re trying to solve.

Nothing that we know about participant behavior supports a conclusion (1) that many participants will ever go to a DOL website providing “lifetime income illustrations” or (2) that those who do will input realistic assumptions or (3) that participants who do go to the site and input realistic assumptions will actually act on them. And—by the way—I doubt whether whatever participants do with such a DOL website will reduce confusion.

Let’s take a step back and consider the underlying problem we’re trying to solve.

The DC plan retirement income challenge

DC plans provide a vehicle for the accumulation of an account balance through a process that is pretty intuitive: you make contributions, you invest them and your account balance grows. But, at retirement, all the participant has is that account balance, when what she needs is a lifetime income.

There are nuances here—the participant may want to leave a legacy, or provide for emergencies or may (e.g., for health reasons) anticipate a very short life. Or, the participant may for very good reasons—e.g., the existence of a generous pension benefit—not be expecting to have to depend on her DC plan account for a minimally adequate retirement income.

But generally, the point of retirement savings generally is to provide an adequate income for the participant in retirement.

And, when our participant contemplates her account balance, she may not be aware of the very significant “retirement income” risks she faces: Most critically longevity (the risk that she will live in retirement for an unexpectedly long time) and market performance (the risk that her investment return will not meet her expectations). More fundamentally, she may not even be aware of how long her balance would last, even if she did know how long she would live and how much she would earn.

And, indeed, it’s not just important for a retiring participant to understand these issues. They are also critical to the decisions younger participants make about, e.g., how much to save.

Participants—who are, in the DC plan system, making all the key decisions—need to know what sort of retirement income their account balance can buy.

What is the best way to solve it?

Many sponsors have already developed tools to address this issue with their participants. Without having done a survey of them, I think we can safely assume that they are all over the place—using different assumptions and methodologies to give participants some idea of the “lifetime income stream equivalent” of their account balances.

More than five years ago, the DOL floated an “Advance notice of proposed rulemaking,” proposing to create some sort of standardized disclosure mandate. That project seems to have foundered on the same sort of sponsor resistance we are seeing with respect to the Senate proposals.

In my opinion, if anything, the proposal in RESA isn’t rigid enough. After staring at the issues presented by the challenge of translating a DC plan account balance into a retirement income for the last 20 or so years, my conclusion is that the simplest, least confusing and best way to present this information is to tell the participant what sort of age-65 life annuity he or she could buy with his/her current account balance.

Indeed, when you think about it, it is kind of weird that there is (currently) very little price transparency in the life annuity market. Thus, the first step we should (perhaps) be considering is mandated annuity pricing disclosure. Wouldn’t it be useful to have a widely distributed annuity purchase index?

If that—“compelling” the annuity market to simply tell us what price they are charging—is regarded as a violation of the freedom of annuity carriers to keep their prices secret, then Plan B would simply be to have the DOL produce an annuity purchase rate based on some reasonable algorithm.

In any case, what this sort of regime—mandated annual disclosure of the age-65 annuity that could (at current rates) be purchased with a DC plan account balance—would do is provide every participant with a transparent metric. With appropriate caveats to the effect that prices in these markets can and will change.

But wait… Isn’t this violating everyone’s freedom of choice?

None of this even rises to the level of compulsion of a traffic signal. This is simply a matter of disclosure.

With respect to contributions (via automatic enrollment and automatic escalation) and asset allocation (via qualified default investment alternatives (QDIAs)) we have aggressively used defaults to encourage participant behavior that we believe is “best,” while in all cases preserving participant freedom of choice via an opt-out. We have done all of that out of a dissatisfaction with the results that mere talk produced.

No one is talking (yet) about implementing, e.g., an annuity distribution default in DC plans. The lifetime income disclosure that some in the sponsor community find objectionable is still just jawboning.

That they find it objectionable, and that we aren’t talking about defaults here, implies a belief that this isn’t that serious a problem.

Maybe they’re right. But here’s why I think they’re wrong: Even when interest rates and investment returns were in double digits, participants had a propensity to underestimate their retirement income needs. To underestimate the longevity and market performance risks they faced in retirement—where, in the drawdown phase, the market performance risks, especially, are more acute and qualitatively different than they were during the accumulation phase. And, as a result, to overestimate, in many cases, how long their “nest egg” would last.

Now we are in an era of much lower interest rates (the key driver of the increase in the cost of annuities), much lower returns and increased life expectancy.

I think it’s pretty rational to believe that participants need an update on just what sort of annuity—what sort of guaranteed lifetime income—their savings will actually purchase. Giving them a standard tool, that produces the same answer for them that it does for their neighbor, is an obvious next step.

Michael Barry is president of October Three (O3) Plan Advisory Services LLC. He has 40 years’ experience in the benefits field, in law and consulting firms, and blogs regularly http://moneyvstime.com/ about retirement plan and policy issues.