The SECURE Act’s safe harbor for 401(k) annuities was presumably well-intended. But it may be a case of too little, too soon. Let me explain.
Asset managers and life/annuity companies, of course, are more than ready to pitch income-generating products to the $7.4 trillion 401(k) plan market. They’ve been ready since IncomeFlex (Prudential) and SponsorMatch (MetLife and Barclays Global Investors) were born almost 20 years ago.
But, based on what I hear and read, most plan sponsors and their advisers still aren’t ready to agree to have so-called “in-plan” annuities embedded in their plans as Qualified Default Investment Alternatives. They have a lot of questions that the SECURE Acts don’t answer.
After talking to 401(k) industry veterans, including executives and business owners, I’ve created a list of issues whose lack of resolution makes even large-plan sponsors, those with the best advisers and consultants, cautious about contracting in-plan annuity fever. It’s a long list.
Softness of the SECURE Act. Safe harbors, in the literal sense, are typically walled in by granite and limestone jetties. The legal “safe harbor” for annuities in the SECURE Act of 2019 offered nothing so reassuring.
Written to admit the widest range of options, it doesn’t give plan sponsors much guidance in sifting through the many types of annuities and their often-complex product designs. With minimal direction from lawmakers, plan sponsors can only proceed slowly and with caution.
“The safe harbor provision basically requires that the insurer must have complied with state financial reporting and solvency requirements for the last seven years,” the American Academy of Actuaries wrote in 2020. “Some would view this safe harbor definition as being too low a bar by sweeping in all insurers who are eligible to operate within the state. The safe harbor does not include any criteria to distinguish financially strong insurers from those that aren’t as strong but are still in compliance with the safe harbor.”
Tablecloth too small for table. Policymakers should have addressed the retirement plan coverage shortfall long before passing the SECURE Act. At any given moment, about half of Americans lack access to a retirement plan at work. Some states have mandated auto-enrolled Roth IRAs, but that doesn’t solve the job turnover/leakage problem. (See below.) Data showing the average or median number of years that Americans spend participating in a retirement plan before they reach retirement would be useful. Yes, SECURE 2.0 increased incentives for small and mid-sized employers to sponsor new 401(k) plans; perhaps that’s beginning to move the needle on coverage.
Can’t win for losing. At age 65, the median savings (50th percentile) for a Vanguard participant, according to Vanguard’s annual “How America Saves” survey, is about $90,000. The average savings (representing the 75th percentile) is about $275,000. “Nudges” won’t increase the savings capacity for low-income workers at small plans, and employers have no incentive to contribute to the accounts of high-turnover workers. Assuming that many people will need to keep at least half their DC savings more or less liquid in retirement, there won’t be enough left to buy a meaningful guaranteed income stream.
Not all 401(k) plans are equal. No one talks about this. Great companies have great DC plans (i.e., low expenses, employer matches and, if you’re lucky, discretionary employer contributions). Small company plans often have none of that. Retirement security often depends on working for a great company, getting paid well, and participating in a generous plan for many years. But it’s hard to get a secure job in a great company.
Not all 401(k) income products are equal. Too much is being asked of plan sponsors. The SECURE Act “does not specify which annuity products could be offered within a defined contribution plan; consequently, there is a full spectrum of possible products,” according to the American Academy of Actuaries whitepaper. “If the goal of plan sponsors is to provide their employees with options for lifetime income, they need to know that some of the options they add to their plans may meet this goal better than others. [Emphasis added.] Annuities that have additional investment-based guarantees attached or that mix investments with insurance add complexity to the education process and may be less effective in providing lifetime income than simpler products.”
The friction of liquidity. In the retail annuity market, a portion of a deferred income annuity’s yield comes from its illiquidity; its issuer holds bonds to maturity and avoids the inefficiency of holding short-term liquid instruments. If liquidity reduced yield by 100 basis points per year (an actuary’s estimate), from a net 5% to a net 4%, on $100,000, starting at age 50 with a $5,000 annual contribution, the accumulation by age 65 would be ~$284,000 instead of ~$321,000. That’s a $37,000 cost of liquidity.
“Most people don’t use the liquidity when it exists. At least not during their working lives. They may want to use it at retirement. So if you an amass an extra 10% of assets, even if the surrender fee is 2.5%, you are still way better off using the less liquid version. These folks end up paying thousands of dollars for an unused feature!” a long-time industry actuary told RIJ.
Loss due to leakage. “Leakage” from tax-deferred savings during job changes reduces lifetime accumulations dramatically. The median job tenure for Americans is only 4.1 years. One industry veteran told me that 75% of 401(k) plan participants won’t be in their current plan when they reach retirement age.
Gender offender. Despite the fact that women on average live a couple of years longer than men, all else being equal, lifetime payout rates for 401(k) annuities must be gender neutral—i.e., the same for both. This could incentivize men to look for an annuity outside the plan. It could also create a communication challenge for plan sponsors.
The burden of providing education. Unless plan sponsors, plan advisers and plan providers invest money and effort in educating participants about retirement income planning, a lot of participants who have been defaulted into annuities will opt-out of them at retirement or before. It’s not even clear how to educate participants. If the best way to use a living benefit is to switch on the income payments immediately, but the annuity issuer would prefer that contract owners delay payouts, who will be candid with participants?
Possible blowback from benefits paid for but not received. “We have ongoing conversations with plan sponsor clients and other fiduciaries, about what is referred to as the potential for ‘unrealized benefits,” said Kelli Hueler, whose online platform, Hueler Income Solutions, enables participants nearing, at, or in retirement to shop for and buy low-cost income annuities. She’s referring to benefits that auto-enrolled participants may have paid for during the accumulation period but that—either because they change jobs or opt out of the annuity at retirement—they never receive. “It’s a serious issue that’s being discussed by astute plan sponsors and their consultants. They’d like to see lifetime income product designs that prevent this from happening.”
Soft assumptions. “Some of the recent TDF in-plan income products appear to rely on positive cash flow assumptions into the investment fund in order to provide guaranteed par value withdrawals into the future for retirees” Hueler told RIJ. “Importantly, it’s an assumption that hasn’t yet been tested.” Given that the products are designed for older participants who will take withdrawals relatively soon, the assumption seems questionable to her and some plan fiduciaries. For many years, Hueler also ran a company that collected data on stable value funds, including historical cash flow statistics, for fund evaluation and reporting to plan sponsors and other fiduciaries. Morningstar purchased that business in 2020.
Conflicts of interest. It strikes me as inherently challenging for plan fiduciaries, whose loyalty to plan participants is supposed to be undivided, to act in the participants’ best interests when choosing products that are designed, simultaneously, to serve the interests of a providers’ shareholders. The incentives and responsibilities are mismatched. Such conflicts of interests have consistently proven to create situations of “asymmetrical information” that disadvantage the clients. Clients typically don’t read, heed or understand verbal or written disclosures.
Paradox of QDIA annuities, Part I. Providers of in-plan 401(k) annuities need to be able to auto-enroll participants into qualified default investment alternatives (QDIAs such as TDFs, managed accounts or stable value funds) because they don’t believe that many participants will actively choose to invest in an annuity. Yet they seem to expect inertia or a newfound respect for annuities to propel these participants into actively choosing the annuity at retirement. I don’t understand this.
Paradox of QDIA annuities, Part II. Because retirement income planning involves long-term and sometimes-painful trade-offs, and because it requires discussions of both death and taxes, it requires substantial engagement on the part of a future retiree. But auto-enrolled participants are often the least-engaged participants. When, how and why will they engage?
Paradox of QDIA annuities, Part III. The smartest, best-paid participants, who are the ones most likely likeliest to reach retirement with the largest accumulations, are also the ones most likely to select their own investments. To the extent that that’s true, QDIA annuities will miss the largest, juiciest fruit on the 401(k) tree.
Portability issues still unsolved. The last time I checked, the folks at Retirement ClearingHouse were still making progress on its auto-portability initiative, which would forward the 401(k) assets of job-changers from their old plan to their new plan. I’m still trying to understand portability issues. It’s complicated. But it’s still a work in progress.
Patchwork regulation. The ongoing legal battle over the DOL’s “best interest” rule suggested to me that every player entering the defined contribution game would prefer to use their own playbook. They aren’t in the same book, let alone on the same page. Asset managers want to play by SEC rules, insurers want to play by state regulatory rules, employees rely on the Labor Department’s protections—and the judicial system appears vulnerable to gaming. Meanwhile, plaintiffs’ attorneys are waiting for plan sponsors to mess up. I’m not optimistic.
The ‘black box’ of insurance. Once insurers lock themselves into a guarantee, they must leave themselves some wiggle-room in which to manage their risks. At times, as I understand it, that can mean adjusting the costs or the benefits of a product in order to avoid a loss. Plan sponsors and participants may not be able to see into the “black boxes” where these adjustments are made. Yet fee transparency has to be a sine qua non for adoption by plan sponsors; they’re terrified of lawsuits.
Participants need more choices. Given the uniqueness of each retiree’s household balance sheet, and each one’s future income needs, it seems odd that a plan’s participants would not have several income-generating options at retirement. These might include non-guaranteed systematic withdrawal plans, for instance, as well as annuities.
“Any one in-plan income solution would be a mistake,” a DC plan veteran told me recently. “No one in-plan solution can accommodate that diversity of needs and goals. And, any two or more in-plan income solutions would add complexity far beyond what plan sponsors want to deal with, or everyday workers can use—without an advisor.”
Underwhelming demand from plan sponsors and participants. Despite what their proprietary surveys say, the sell-side of the 401(k) annuity market is still more interested in bringing this trend to fruition than the buy-side—i.e., the plan sponsors and participants. That’s to be expected, since the benefits to the sell-side are immediate and tangible, while the benefits to the buy-side are more remote and incalculable.
The takeaway
The push for 401(k) annuities may very well serve the business goals of DCIO (defined contribution investment-only) TDF providers. They partner with insurers to preserve assets-under-management levels in 401(k) plans and slow down the rollover-driven hemorrhage of TDF balances at retirement. But it’s not fair or efficient to use sponsors and participants as guinea pigs, or expect that, after a long shakeout period, the market’s invisible hand will deliver an optimal solution.
If the past is any guide, a few large asset managers and life insurers will try to impose a regime that serves their needs but not the public’s. That’s OK for hula-hoops or even computer operating systems, but this is a national program that participants finance and taxpayers subsidize. Too little has been done to address fundamental issues, leaving most plan sponsors not ready to commit.
© 2024 RIJ Publishing LLC. All rights reserved.
