Auto Everything—Including 401k Auto Decumulation

auto features

Image credit: © Olivier Le Moal | Dreamstime.com

A recent 401kSpecialist.com story about Schlomo Bernartzi’s presentation at the 2022 NAPA 401(k) Summit is interesting. He said:

“… I’m trying to create a decumulation process and solution for humans” … “We have to start democratizing income planning” … the process should not start with a product but a process to create plans for everyone, regardless of how much money they have … “

It was consistent with a February 12, 2019, 401kSpecialist article titled Shlomo Benartzi: Fairness Key to Participant Annuity Use where he said:

“While the accumulation of retirement savings can be improved using non-personalized nudges such as auto-enrollment in a company’s 401k and auto-escalation of contributions,” …  “decumulation requires a highly personalized approach, in which the solutions are tailored to reflect the goals, circumstances, and preferences of the individual.”

Is it Time for Full Auto[i]—including auto decumulation? Maybe.

Despite having minimal knowledge of each worker’s personal financial situation, plan sponsors forged ahead and adjusted their choice architecture to deploy automatic features for new hires—folks they knew little about, individuals who have little in common other than a recent hire date.

Comparatively, plan sponsors have greater knowledge of retiring workers, including two features all workers have in common—items that might form the basis for an auto decumulation “default.”

Today’s decumulation “default’ 

Taxpayers are ultimately responsible for complying with the Required Minimum Distribution (RMD) rules. However, to be tax-qualified, a “…plan must provide that the interest of each employee will begin to be distributed to the employee not later than the required beginning date which means, in general, April 1 of the calendar year following the later of the calendar year in which the employee:

The first feature all plans have in common is this “default” distribution requirement. Some plans provide for a lump sum once the participant reaches Normal Retirement[iii] or the Required Beginning Date (RBD). Some require the involuntary distribution of small account balances.[iv]

Other plans have installment payout features designed to issue RMD-compliant payments – where the initial payment upon reaching age 72 would be approximately 3.6% of the prior year-end account balance.[v]  

The second feature that almost all workers have in common is that they have no reason to defer the commencement of Social Security benefits once they have reached age 70.

Post-SECURE, are you considering adding an in-plan, guaranteed income option?

Dr. Bernartzi was also quoted as saying:

”(a) one-size-fits-all approach won’t work with decumulation. The individual situations are too different.

… It’s going to be very difficult to create a personalized solution unless we engage people.”

Home equity and Social Security provide a significant level of guaranteed income for many lower and middle-income retirees.[vi] Some studies suggest most households should not annuitize any accumulated wealth[vii].

Certainly, a “one-size-fits-all” approach will not work for all participants. Neither will a “one-product-fits-all” in-plan annuity or other guaranteed income option.

Because the median tenure of American workers has been less than 5 years for the past 5 decades[viii], and because almost every 401k allows for distributions while employed and soon after separation from employment, most participants encounter one, two, or many plans while employed and an even greater number of payout decisions – most far in advance of retirement—however you define retirement.[ix]

My plan sponsor experience confirmed that fewer than one in 20 new hires remained with my employer until they retired. Tenure trends suggest my experience was typical. So, why take on new fiduciary risk exposure and expense to add in-plan guaranteed income options given minimal take-up rates as a percentage of all plan participants?

And, given the mismatch between a “one-product-fits-all” approach and the plethora of products available in the individual insurance marketplace, why take on new fiduciary risk exposures and expenses by adding a product that does not match the diversity of your participants’ needs?

Tomorrow’s decumulation ‘default

Where there is a need for more retirement income, participants may want to consider the only cost-effective, well-priced, guaranteed, inflation-indexed retirement income opportunity available to everyone—deferring commencement of Social Security.[x]

To prompt consideration, you may want to consider changing your plan’s “default” distribution process. And, since most decumulation communication occurs as part of the SPD that few read or coincides with a payout election, why not introduce the new decumulation default far in advance of retirement?

It is important to remember that a decision to amend your plan to adjust your choice architecture and deploy automatic features is generally not a:

Here is an outline of one possible decumulation default a plan sponsor could voluntarily adopt. It is designed to provide an annually increasing payment of 401k assets as a “bridge” to meet income needs until commencing Social Security at age 70 and a level income (adjusted for inflation) through the commencement of required minimum distributions:

Service providers would make available a Social Security optimization calculator and electronic banking capability to facilitate a monthly income frequency. The service provider would also make the “default” estimator available as a tool – ensuring participants can hyper-customize a payout stream – potentially adjusting for:

If “income is the outcome,” if it is a top priority for your plan, you may find that a “one-size-fits-all” default decumulation process offers superior value by prompting lower and middle-income retirees who have modest 401k assets to (re)consider their readiness to retire and their retirement income needs. Individuals with larger account balances and greater income replacement needs can confirm that this default is not adequate to meet their needs and that they may need to consider individual insurance products.

I always appreciate your criticisms, suggestions, improvements, etc. Feel free to contact me at:  jacktowarnicky@gmail.com

Disclaimer No. 1: My comments are my own based on my experience in plan sponsor and consulting roles – they do not necessarily reflect those of any employer, group or association I have been employed by or affiliated with, past, present, or future.

Disclaimer No. 2: This information was provided by individuals with knowledge and experience in the industry and not as legal or tax advice. The issues presented here may have legal implications and you should discuss this matter with legal counsel prior to choosing a course of action. This article is intended to be informational only. It is not (and you/others should not use it as a substitute for) legal, accounting, actuarial, or other professional advice. Any advice contained in this article was not intended or written to be used, and cannot be used by anyone for the purpose of avoiding any Internal Revenue Code penalties that may be imposed on such person [or to promote, market or recommend any transaction or subject addressed herein]. You (others) should seek advice based on your (their) particular circumstances from an independent tax advisor.


[i] J. Towarnicky, Is It Time for Full Auto? Benefits Quarterly, 1st Quarter 2020.

[ii] Internal Revenue Service, Accessed 4/26/22 at: https://www.irs.gov/retirement-plans/a-guide-to-common-qualified-plan-requirements#11

[iii] Internal Revenue Service, Accessed 4/26/22 at: https://www.irs.gov/retirement-plans/plan-participant-employee/when-can-a-retirement-plan-distribute-benefits

[iv] Internal Revenue Service, Accessed 4/26/22 at: https://www.irs.gov/retirement-plans/plan-participant-employee/401k-resource-guide-plan-participants-general-distribution-rules#required-distributions

[v] Internal Revenue Service, Accessed 4/26/22 at: https://www.irs.gov/pub/irs-prior/p590b–2021.pdf

[vi] J. Towarnicky, Annuitization Anyone? 5/1/19, Accessed at: https://www.psca.org/news/blog/annuitization-anyone

[vii] F. Reichling, K. Smetters, Optimal Annuitization with Stochastic Mortality and Correlated Medical Costs, American Economic Review, November 2015. “The conventional wisdom since Yaari (1965) is that households without a bequest motive should fully annuitize their investments… We modify the Yaari framework by allowing a household’s mortality risk itself to be stochastic due to health shocks. A lifetime annuity still helps to hedge longevity risk. But the annuity’s remaining present value is correlated with medical costs, such as those for nursing home care, thereby reducing annuity demand, even without ad-hoc liquidity constraints. We find that most households should not hold a positive level of annuities, and many should hold negative amounts.” Accessed at: https://www.aeaweb.org/articles?id=10.1257/aer.20131584

[viii] Bureau of Labor Statistics, Accessed 4/23/22 at: https://www.bls.gov/news.release/tenure.t01.htm  C. Copeland, Employee Benefits Research Institute, Trends in Employee Tenure, 1983–2018, 2/28/19, Accessed at: https://www.ebri.org/content/trends-in-employee-tenure-1983-2018

[ix] Author’s note: I arbitrarily defined retirement as: (1) Discontinuing all full-time employment, (2) Eligibility for Social Security old age benefits, AND (3) Commencing payout of retirement plan and/or IRA assets. According to Form 5500 data, more than 1 in 5 defined contribution plan accounts belong to individuals who are not “active participants”. That is a 10-fold increase in percentage terms or an 81-fold increase in number when compared to 1975! Department of Labor, Employee Benefits Security Administration, Private Pension Plan Bulletin Historical Tables and Graphs, 1975-2019, September 2021. 1975: 11,507,000 participants, 11,217,000 active participants. 2019: 109,096,000 participants, 85,504,000 active participants. Accessed at: https://www.dol.gov/sites/dolgov/files/ebsa/researchers/statistics/retirement-bulletins/private-pension-plan-bulletin-historical-tables-and-graphs.pdf

[x] W. Pfau, J. Tomlinson, S. Vernon, Viability of the Spend Safely in Retirement Strategy, May 2019, Accessed at: https://longevity.stanford.edu/wp-content/uploads/2019/07/Viability%20SSiRS%20Final%20SCL.pdf  See also: A. Munnell, G. Wettstein, Would 401(k) Participants Use a Social Security “Bridge” Option? December 2021, Accessed at: https://crr.bc.edu/wp-content/uploads/2021/12/wp_2021-27.pdf

[xi]  See: Lockheed Corp. v. Spink, 517 U.S. 882 (1996) “… Lockheed and the board of directors, as plan sponsors, were not acting as fiduciaries when they amended the Plan. Given ERISA’s definition of fiduciary and the applicability of the duties attending that status, the rule that this Court announced with respect to the amendment of welfare benefit plans in Curtiss-Wright Corp. v. Schoonejongen, 514 U. S. 73, applies equally to the amendment of pension plans.” See: Curtiss-Wright Corp. v. Schoonejongen, 514 U.S. 73 (1995), citing Adams v. Avondale Industries, Inc., 905 F.2d 943, 947 (CA6 1990) (“[A] company does not act in a fiduciary capacity when deciding to amend or terminate a welfare benefits plan”). Author’s Note: Of course, fiduciary rules apply to the plan administrator in administering the terms of the plan.

[xii] Author’s Note: The plan investment fiduciary would want to consider the need for a default investment allocation during the default payout period. Importantly, for those with minimal or modest account balances, the default would confirm how little guaranteed, inflation-indexed income is available.

[xiii] Where the participant’s account balance is insufficient to bridge income all the way to the required beginning date, the estimator would produce a payout that would exhaust the account balance in the calendar year in which the participant reaches age 70. Similarly, where it is insufficient to bridge to age 70, the estimator would produce a payout sufficient to bridge to a pre-age 70 Social Security commencement date.

[xiv] Author’s calculations. This example assumes the individual needs $22,608 per year (or more) to maintain their pre-retirement standard of living. In the “default,” Social Security benefits start at age 70, and $1,745 is the estimated annual RMD payment at age 72. By age 84, the “default’ provides the largest amount of guaranteed, inflation-indexed income, compared to commencing Social Security at age 62 or age 67. Different assumptions produce different results.

[xv] Author’s Note: SECURE lifetime income disclosures are more likely to mislead than to inform and less likely to prompt increased saving compared to use of automatic features. Had the SECURE required disclosures been in effect since ERISA, an individual age 67 today might have received 45 or more annual statements (depending on how many qualified plan account balances they may have).  Those statements would vary substantially due to significant changes in purchase rates, account balances, distributions (hardship, in-service and post-separation withdrawals) as well as rollovers to IRAs (where the mandated disclosures do not apply). See J. Towarnicky, Discordant Disclosures, 9/8/19. “Standardized, simple, and least confusing is easy. There is already a retirement savings individual account distribution mandate in place today – enforced by a 50% tax for non-compliance. It is called required minimum distributions (RMD).” Accessed at: https://www.psca.org/news/blog/discordant-disclosures.

Exit mobile version