What distribution fee should a plan charge for an early-out withdrawal?

Some BenefitsLink neighbors have observed that a recordkeeper will incur considerable expenses to tool up for new distributions and other features SECURE 2.0 permits. And some have observed that, even if one knew or estimated that many or most plan sponsors don’t want the newly permitted provisions, the expense to build a capability is almost unavoidable, because there will be some current and prospective service recipients that want a provision (or the opportunity and flexibility to choose it).

Further, new kinds of distributions—such as, an emergency personal expense distribution (not to be confused, or cost-accounted for, with a distribution from an emergency savings account) and an eligible distribution to domestic abuse victim—might involve increased complexity, might generate increases in transactions, and so might increase attributable or allocable costs.

(Without discussing specific amounts or anything else that could be price-fixing, collusion, or anti-competition:)

How should plan fiduciaries and recordkeepers together seek to allocate these expenses?

Compared to a fee for processing a normal distribution after severance from employment (or age 59½) and assuming one’s only reasoning is an attempt to allocate a cost to those who generate the cost:

Should a fee for processing an emergency personal expense distribution be less than, the same as, or more than the fee for processing a normal distribution? Why?

Should a fee for processing a domestic-abuse distribution be less than, the same as, or more than the fee for processing a normal distribution? Why?

Should a fee for processing a hardship distribution be less than, the same as, or more than the fee for processing a normal distribution? Why?

Should a plan’s sponsor—using its non-fiduciary settlor powers to decide a plan’s provisions, including charges—favor or disfavor some kinds of distributions?

Should a sponsor disfavor an emergency personal expense distribution by charging a higher distribution fee?

Should a sponsor disfavor a hardship distribution by charging a higher distribution fee?

Just some general thoughts:

It seems a little early to be making decisions on these items, without truly knowing what is involved in processing such transactions, for both the recordkeepers and the TPA's. And how is a fiduciary supposed to make an informed decision on such matters until these things are known, and communicated to the fiduciary?

That said, I generally am squeamish about the idea of a plan sponsor/fiduciary disfavoring such distributions by charging a higher fee. However, that's just my general bias, not backed up by any specific information. I also generally favor keeping distribution fees the same for different types of distributions if possible, for administrative ease and clarity as much as anything else. Again, I just feel like it is too early to have a reasonably informed opinion - the recordkeepers/TPA's are going to have to charge as much as they need to, and this could be structured in many different ways, to THEN be analyzed by the fiduciary as to whether the fees are reasonable and appropriate.

I dunno - I'm just blathering here...

P.S. - I do seem to recall that when Roth first came out, some recordkeepers/TPA's charged more for a Roth distribution than for a taxable distribution. Seems like that is pretty much gone now.

Belgarath, thank you for your helpful thoughts, practical observations, and fact observation.

For a big or mid-size plan, about a year is a project timeline for the many operational, legal, and communications steps needed to implement a change. I’m now negotiating some plans’ service agreements for terms that begin on or a little before January 1, 2024—just when the new early-out distributions take effect. That a sponsor or fiduciary should make informed choices is why their counsel is thinking about it now.

Is there really much about the new distributions that’s unknown? For each kind, we know: whether it allows a before-severance payment; whether a claim is self-certifying; whether the distribution is treated, for tax reporting and withholding, as not an eligible rollover distribution; whether the distributee has a three-year option to restore the amount to the plan; and other essential points.

Thank you for your observations about keeping charges for distributions constant, and not using a charge to influence participants’ behavior.

BenefitsLink neighbors, if one does not use a charge to influence participants’ behavior but seeks only to allocate a cost to those who generate the cost, how would you feel about charging more for a particular kind of distribution if the recordkeeper’s cost accounting shows clearly that it costs substantially more to process that kind of distribution?

Or, once a service provider has built capacity, do distinctions about costs become blended or at least less clearly attributable?

Is there anything different or special on administration of these other distributions (abuse, emergency, etc.)? Is there adjudication or are these self-certification? If it's only a matter of whether the 10% premature distribution tax applies, I don't see a huge reason for charging a higher fee. Personally, I think the majority of service providers will be increasing their fees in general because of this, not to mention general inflation, although this is all my opinion from the outside as I do not directly work on the inside nuts and bolts of these plans.

CuseFan, thanks.

The two new early-out distributions, and a hardship distribution, can be self-certifying; and my assumption (which I neglected to state in my originating post) is that the plan or its administrator would make them self-certifying.

Might a plan’s sponsor or administrator set a lower charge on some kinds of distributions?

For example, imagine the plan’s undifferentiated charge for a distribution is $100.

That charge might not attract attention when applied on a $200,000 severance-from-employment distribution.

But imagine a participant wants to use an emergency personal expense distribution to meet her need to raise $200. The participant would need to claim $300 to raise $200. (Assume no withholding for income taxes.) The $100 processing charge then is 50% of the net amount the distributee receives.

Because this kind of distribution is for $1,000 or less, some processing charges participants tolerate for bigger distributions might seem disproportionate.

Others might say that’s what happens when one uses a retirement plan as an any-purpose savings account.

I don’t advocate for or against any view. Rather, my queries are about illustrating the difficult choices service providers and plan fiduciaries face, and that there might be many (and differing) interests to consider.

That is a spot on observation Peter. It is a sensitive situation for sure, but not really anything new. Certainly price points could change and reflect the amount of work that goes into processing a distribution to the extent it is impacted by the amount being paid, but people don't know how to value those new pieces yet. We see that in the DB/CB world for the over/under $5,000 (so to be $7,000) lump sums.

Surprisingly, according to the Plan Sponsor Council of America’s 65^th Annual Survey, over 1/3 of all plans that responded to the survey pay 100% of all recordkeeping/administrative fees.

In my last plan sponsor role, all the way back in the mid 1990’s, we eliminated hardship distributions and added an admin fee (basis point charge) that was sufficient to pay all recordkeeping/administrative costs. In 2008, we changed the admin fee to a monthly, per capita fee, plus transaction fees. That change was prompted in part by a change to add automatic features in 2007. Auto features increased the processing requirements and the percentage of employees who were contributing by ~20%. Seemed inappropriate to have individuals with a lifetime of savings shoulder the increased costs from a massive increase in participation. The change reduced the disproportionate impact on participant-paid admin fees for those whose account contained a lifetime of savings (e.g., an account balance of $500,000 had been paying $500/year (10 bps), but prospectively, paid ~$36 per year (~$3/month) after the change).  

Our per participant monthly fee covered the cost of all regular processing – contributions, investment transfers, etc. We added specific transaction fees to cover processing costs on all loans and distributions (except for Required Minimum Distributions).

With respect to in-service distributions, my first recommendation is always to avoid them or curtail them - to minimize leakage. Only a small minority of distributions are repaid/re-contributed.

In terms of varying the charge for different distributions, the only ones to favor with a lower or no fee (to ignore the cost to process) are mandatory distributions. Otherwise, the fees should reflect 100% of the processing costs – so as to intentionally favor electronic processing/banking.   

When it comes to satisfying the participant’s need for liquidity, I favor “liquidity without leakage up to and throughout retirement”. That is, I favor adding/updating plan loan processing to 21^st Century functionality (line of credit structure, electronic banking, various behavioral economics prompts, processes, strategies and concepts). For example, when executing the plan loan application, the participant should have to authorize the loan and agree to repayment – effectively signing twice, once as the borrower and second as the lender (future self). Plan loans are not leakage unless they are not repaid. Most plan loans are repaid whenever employment continues throughout the term of the loan. And, because plan loans must be repaid, they dampen utilization. Because of substandard processing, most plan loans default where the individual terminates employment.

My recommendation is that before plan sponsors make any changes, and add any of these new leakage opportunities, that they consider the need to update liquidity and fee structures.

My experience with fee structures for larger service providers is they are driven fundamentally by profit.  A service provider calculates the cost of operations based on the same factors as any other business - salaries and benefits, physical facilities, operational expenses, technology, consumables...  They then set a target profit and add the two together to get a "required revenue" number.   Statistics about utilization of various services are compiled and an estimated price is assigned to each service.  These two factors combine to see if required revenue can be achieved.  For example, it is assumed that every plan will need a 5500, and the cost of a 5500 for a large plan differs from the cost of a 5500-SF or -EZ for a small plan.  Similarly, there are assumptions about the number of payments and types of payments, loans, QDROs, investment menu complexity, disclosure requirements, compliance testing complexity, managed accounts...  Starting with this a la carte menu and pricing, the service provider may assume that on average 10% of participants will take a payment during the year, or 5% will take a loan. This all is akin to an underwriting process.

The process now allows for a lot of flexibility in negotiating the terms of a service agreement with a plan sponsor.  If the plan sponsor is amenable to an all-in fee, then the fee agreement will bundle all of the services for the a equal to the required revenue number (and likely converted into a basis point charge on the investment funds).  If the plan sponsor wants to disincentivize a certain transaction type, the that transaction type gets a participant-paid fee assigned to it.  For example, the plan sponsor may wish to discourage loans so there is a loan setup fee (amortization schedule and promissory note), a loan check fee and a loan maintenance fee.  The service provider adjusts reduces the required revenue number by the projected fees to be collected from participants who take loans, and the plan sponsor pays or there is a basis point charge or both.  If the plan sponsor does not want a particular service, then the expected revenue from that particular service is subtracted from the required revenue.

(The one participant fee that I have seen that is more blatantly self-serving for the plan sponsor than most is charging a plan administration fee solely on the accounts of terminated participants who leave their balance in the plan.)

Deciding on a fee for a transaction type may be rationalized in many different ways, but at the end of the day profit margin drives the whole process.