Retirement Income Comes into Focus for Plan Sponsors

Anne Lester: From a plan sponsor perspective, there's been an evolution from focusing on accumulation and not necessarily focusing on decumulation. It's partly because for many years, the baby boomers were being trained by auto-enrollment to just keep doing what they were doing and remained happily aware that their assets were piling up. It was really a one-way street.

Now, we're hearing loud and clear from plan sponsors that participants are thinking about retirement and they're getting calls about How do I do this? What's the right way to do this? What are my risks? Suddenly there are these unanswered questions, and the products and services inside of most strategies are not properly equipped.

Drew Carrington: We are past the realization about the large demographic bulge approaching retirement. Furthermore, plan sponsors are starting — as they get better at analyzing data in their own plans — to see that not only are a lot of participants in that zone of approaching retirement, but they have most of the money in the plan, and there's a lot of money in the plan that could walk out tomorrow. Whether it's people over 59-1/2 or vested former employees, the plan could dramatically shrink. So they want to make sure that they're serving that population for a variety of reasons, and want to make sure that they're providing the kinds of tools and plan design changes that make the plan easier to use for near retirees and folks who are entering that transition phase.

Rob Reiskytl: The issues that Anne and Drew have mentioned are spot on. There's still a lot of focus on participation rates, savings rates, investment mix, all of that. Those are leading indicators of the ultimate long-term result. And even though plan sponsors do continue to focus on the leading indicators, we are seeing more interest on what they lead to. What it can lead to is not having enough or being ready to retire. The importance of learning to live on the assets and helping individuals convert from an accumulation phase to one in which they are living on the assets and figuring out how to do that efficiently is the dynamic that we're seeing unfolding here for corporate America.

Lester: Define enough. The problem is that the 401(k) plan was structured and designed as a supplemental savings program that is not intended to provide an outcome. What we've done is essentially jury-rigged a number of things onto it, such as auto-enrollment and auto-escalation, that will improve the probability that somebody hits an outcome. You can get people in, you can get them increasing their savings. But the final leg of the stool is how do you help people understand how to take money out? Because in accumulation, people are much more homogeneous, that's an easier problem to solve from an investment perspective than decumulation. And you've got to remember all of this is voluntary.

Reiskytl: If you look back a decade, the defined contribution was used primarily as a supplement or an “extra” plan. In fact, many defined contribution plans are still called that. It's a supplemental savings plan, even though it's the only vehicle many people have for retirement savings. So defined contribution is now primary and very central, and it needs to be taken very seriously and perhaps thought of differently than the way that it was five or 10 years ago. What we're seeing over time now is a broader view of outcomes and a result that someone can live on. We're continuing to see interest in both primarily the first piece, which is the accumulation, and a growing interest in the spend down, or the decumulation piece.

In 2017, Aon released findings from their survey of plan sponsors about what they offer to help people convert their savings into lifetime income. Two-thirds said that they offered online modeling tools to help participants see how much they might be able to spend each year in retirement, and half offered an automatic payout over a set period, such as 10 or 20 years. That's not truly dealing with the question at hand, which is, How can you live on your income for life and not run out and have enough until you die?

Only 12% say that they facilitate the purchase of annuities outside the plan, 8% offer an annuity purchase as part of the fund lineup within the plan, and 3% promote a qualified longevity annuity contract within the plan. Those are very small percentages, and we'd love to see them be higher.

Reiskytl: We asked about that. Plan sponsors we surveyed said the No. 1 reason that they didn't have in-plan income solutions was fiduciary concerns — 46% said that. Another common answer, given by 41%, was that they were waiting to see the market evolve, and 40% said operational or administrative concerns. Finally, 33% cited participant utilization concerns or difficulty with communications.

Carrington: One of the problems with the annuity purchase for individuals is that it seems like a really big one-time decision, and it's a decision that individuals don't get any practice on. You think about most financial decisions that we make over the course of our lives, we get to make them multiple times. We get to learn from our mistakes. Think about mortgage interest rates. We have ideas about what is a good rate or a bad rate. Should I do a variable or adjustable-rate mortgage? We have all had multiple opportunities to make those decisions. Sometimes they work; sometimes they don't, but we learn from those choices.

Annuity purchases are often irrevocable and that becomes scary. We can pull away from the big cliff and make it smaller, with bite-size purchases or some sort of dollar-cost averaging. Whether it's an insurance-based product or a capital markets product, you do it inside the plan, and it's a lot easier to deal with. That's one way to serve the population that's not yet retired.

Reiskytl: Twelve percent of plan sponsors we surveyed said they facilitate the purchase of annuities outside the plan. Only 8% said they do an annuity purchase or insurance products as part of the fund line-up within the plan. And only 3% said yes, they provide that within the plan.

A common solution is a plan distribution option that allows participants to elect installment payments from the plan over a specific period of time such as 10 or 20 years. With that, you're spreading your value out and receiving that somewhat evenly during that time period. But what it doesn't get at is the chance that you could die before the end of the payout period or live well beyond it. These types of payments are not as helpful as truly having an income stream that you can't outlive. The more appropriate payment options, in my view, are the ones that truly provide income for life, either for the participant or the participant plus a named beneficiary.

The most important solutions are the ones that are the least frequently offered right now. The 12%, 8%, 3% I just mentioned. Those are very small percentages. We'd love to see them be higher.

Reiskytl: Unfortunately, a lot of employees take their cues from their plans and not from their financial reality. If you look at savings patterns within our defined contribution system, the most common savings rate is almost always the place where the matched savings stops. If an employer matches dollar-for-dollar at 6% of pay, the most common savings rate tends to be 6% of pay. If you have automatic escalation, that might eventually get up to 10% of pay; but if that's where the escalation stops, that's where the savings also tend to stop. Employees may end up getting there and stopping, and never really paying attention to whether that's the right number or not, given their circumstances.

Lester: Yes. There are a lot of different decisions that happen as you approach retirement. Are you relocating or not relocating? Have you paid off your house or do you have a mortgage? Are you taking your 401(k) money and buying a small business franchise?

There are all kinds of things that people do that would require larger or smaller pools of capital, and from what we observe, relatively few people do the 4% thing and stay within a very stable withdrawal band. We follow about 60,000 people to see how they're spending money, and only about one in five stay within 20% of what they spent the year before.

There's a lot of bumping around; so what should the right default setting be? I think one of the reasons you've seen relatively slow movement in product development is that nobody's quite sure what anybody wants. Part of the reason that nobody knows what they want is that we haven't been here yet. This is all new for everybody. It's a very individual choice.

Carrington: There's also a lot of variation around how much each individual is relying on his or her 401(k). Some people have earned a defined-benefit pension from a prior employer, either a corporation or a public-sector employer. Other individuals have never had a pension. If they're married, then they're part of a household and there are potentially even more accounts. Data on the population that's over 50 and participating in a 401(k) plan shows that odds are they're married and their current 401(k) plan represents less than half of their current household liquid wealth, not including home equity.

What does that mean? It means that the plans, when we communicate with those participants, need to remember that this is only one piece of the puzzle. They almost certainly have other assets outside of what we can see.

Lester: We have done a lot of research and modeling on behaviors and distributions looking at this, and what we found in a recent whitepaper we published was a little bit surprising to us. When you just run the math and model infinite time series of return, the worst asset strategy to pursue is one that starts at a high level of equity risk and the risk declines as you approach retirement. That's because you are essentially putting a huge amount of your capital at risk at the moment in which you start selling and lock in those losses, and then next year you'll have less equity to recover as you need to grow that back. You're sort of doing the opposite of dollar-cost averaging. We think that's remarkably counterproductive.

The optimal answer in our view would be to start with an equity level between 15% and 20% at the point of retirement and then increase that back up. There are a number of problems with that, including that it's extraordinarily counterintuitive for most people, and rerisking feels uncomfortable. So we landed at an equity allocation of about 33%. We also launched a companion product, which delivers to an investor an annual communication on the amount we think can be spent, which we re-optimize annually based on the length of time remaining in our capital market assumptions. We also re-optimize the amount of risk we're taking in the portfolio, so it's not a predetermined glide path. It will vary between 30% and 50% equity depending on our capital market assumptions and the time horizon remaining.

Carrington: There's actually a wide array of data that suggest that participants in their 50s are more engaged with their plan. People who are close to retirement are more likely to call the call center, they're more likely to log onto the website on a regular basis and they're more likely to rebalance their portfolio. If they were defaulted into a target-date fund, they're more likely to diversify that position. They're more likely to use the managed account provision. The data is really compelling on that front.

Reiskytl: The communication also depends on the industry. If you're working in pharmaceuticals or energy or oil or gas, where private employers generally provide stronger benefits and higher compensation, those individuals might be in a very different place than a group in retail or healthcare, where the compensation and benefits might be more modest. The messages that you share might be different for each group. We're obviously not trying to have people over-save, but at the same time, we want to have people essentially build up a reasonable retirement asset so that they have financial choices, flexibility and financial freedom once they reach retirement.

Carrington: The common shorthand is the three tiers: tier one is the do-it-for-me participants, tier two is the do-it-with-me participants, and tier three is the I'll-do-it-myself group. Our idea of the tiers is that they're not just a list of investment options, but a collection of things: targeted communications, plan-design changes and tools to help people, depending on where they are.

We think of a retirement tier as serving participants from around age 50, when they're starting to make catch-up contributions. At that point, people really start paying attention; they're much more engaged.

So for a retirement tier, that could mean plan design changes, such as allowing for partial, ad hoc withdrawals once a participant is separated from service. Individuals don't necessarily go from working and saving to their 65th birthday. They don't necessarily immediately retire and land in a rocking chair on the front porch, drawing down their qualified plan in a level and steady way.

We think that there's more of a pause in between the accumulation phase and the more dedicated draw down phase, where participants, maybe they're working part-time, maybe their spouse is working. So they're not contributing, but they're not drawing down either. So the plan design change might be for ad hoc withdrawals, or maybe it's a tool like a Social Security optimizer that helps participants think about their strategy for claiming Social Security, particularly at the household level.

Lester: The reality is that on average, people's spending declines as they age in real terms, and the wealthier they are, the faster it declines. That has a whole bunch of implications. One implication is that I suspect we are over-estimating how much money some people will need in retirement. The second implication is that as we think about building products, as we think about providing advice and guidance for that drawdown phase, we need to be mindful that we're not causing people to underspend when they're young and healthy and overspend when they're older.

Lester: The risk is getting it wrong. Time is not your friend anymore. Time is your enemy in terms of what happens, when it happens and how you react to it. Something goes wrong in your portfolio in your 30s, and it's just noise by the time you are 60. Something goes wrong in your portfolio when you are 70, it's going to change the course of your life. There is a lack of formal, regulatory guidance and frankly, a lack of social consensus around what the right answer is. So most people are muddling through with a combination of Social Security and — for now — a defined benefit cash balance or annuity. But the number of retirees with a pension is going to be cut in half in the next five to 10 years.

Without a set of common assumptions around how to help people convert their 401(k) into an income stream, there's a real reluctance to make irrevocable decisions, such as triggering an annuity purchase, which may not be something an individual wants to do.

Reiskytl: Many plan sponsors are concerned as fiduciaries. They know they are not required to provide retirement income solutions, and they might be criticized for the way they do it, so many choose to not offer these solutions at all. We would like to see more employers taking action in this area. What we should be saying is “Let's find ways to get beyond the fiduciary concerns and find products and solutions that work better.”

Carrington: You're running a plan, and there are risks associated with it. That's the reality. There has been this view historically that retirement income — particularly guaranteed products — have some special radioactive risk that if you touch them as the plan sponsor you're at a much greater risk. That's not the case. There's already a safe harbor around the selection of these products.

You can manage your risks as a plan sponsor. You can make the decision to outsource some of the decisions to another fiduciary in a 3(38) arrangement; you can share those fiduciary decisions with an adviser or consultant; you can carefully document the process that you go through around thinking about these things and mitigate all of those risks that you may or may not face. There is nothing particularly exceptional or dangerous about those choices vs. other choices that you might make in the plan.

As an industry, we migrate toward binary framing: It's retirement income or it's not. The framing here should be that there are things that I can do now that are relatively low risk and can move the ball forward. I don't have to do all the things at once, but I can get started. I can start serving this population and communicating with them, and giving them tools.

Carrington: Going back to the binary framing, declaring success or failure is a little grandiose. I think you have to be more modest in terms of whether you declare success or failure for your participants. These are longer-time-horizon problems, and the goals are more individualized. It's simple to measure the number of people who used a Social Security optimizer. But if you're able to see that you reduced the number of people claiming Social Security at 62, that might also be a measure of success.

Reiskytl: A lot of what employers are doing is still focused on getting people to save and pay attention to their retirement plans. Maybe they do it through automation, maybe they focus on things like financial well-being and the infrastructure of economics that people face.

But it is both about helping individuals build up a reasonable, appropriate amount toward a future retirement event and then eventually helping them live on that efficiently. It's about both spaces, and you need to spend time in both.

We also would like to see more plan sponsors understanding that providing lump-sum payouts and assuming employees will “self-insure” during their retirement years tends to be much less efficient than partial annuitization, where individuals benefit from the pooling of longevity and investment risk. We are also intrigued by solutions that pair up professional investment management with deeply deferred annuities, such as qualified longevity annuity contracts.

Carrington: I think it's important to get started and know that you don't have to do everything at once. My hope is that in a year, we've seen adoption, and that plan sponsors have gotten past the hurdle that it's all too big, too scary, too risky to do anything. I hope it's no longer a conversation about what we might do, but instead, we're looking at the data and evaluating the success of what people have done.

Lester: It's possible to make great improvements in both the information that we're able to share with participants as well as the way we shape product choices. There are many, many things we can be doing that will materially help people and improve the choices that they're making. That's what we need to focus on: Can we make it better than it is today? Not whether it's definitionally perfect in every respect.

I don't think there are any silver bullets in this space. I think target-date funds and other QDIAs are pretty darn good for most people, and so you get most of the way there. I do think it's possible to get close with decumulation as well. We just have to start doing it.