Peter Gulia

ESOP Guy, thank you for your good observations. My experience too is that those who have practical experience with recordkeeping operations spot issues that others miss. I'm interested in what you think about my idea for an element in an employer's QDRO procedures: Before sending an approval of a "live" order that the administrator believes is a QDRO, give the instruction to the recordkeeper. Next morning, check the system to see whether all the changes transacted. If all processed, put the QDRO-approval letters in that afternoon's mail. If anything didn't process, talk with the recordkeeper. It's likely that the recordkeeper's reason for not processing the administrator's instruction will reveal at least one reason for deciding that the order is not a QDRO. What do you think?

Belgarath, thank you for the further explanations. Count your writing here today as a valuable pro bono contribution, supporting mine, to a charity that's doing great things.

David Rigby, both of your observations are good. In formal terms, a recordkeeper doesn't preclude an administrator from doing something; rather, a recordkeeper chooses which service it offers (or doesn't offer). If an employer doesn't have big-plan buying power, what recordkeepers offer often is a practical restraint on what an administrator chooses. You're right that a service agreement would state the add-on services. My work as counsel to administrators is about employee-benefit plans with tens of thousands of participants and enough buying power to custom-negotiate each service agreement. But my inquiry here is about unusual work for a small plan, and they won't ask for the QDRO add-on unless I advise them that it's worth considering. So that's why I ask BenefitsLink mavens: Is a recordkeeper's QDRO-review service worthwhile? (I'm hoping to see the usual sharpness of different viewpoints.)

Belgarath, thank you for this super helpful information. So the only way an employer can get comfort about whether its document states a plan that (if administered according to its provisions) gets the Federal income tax treatment of Internal Revenue Code section 403(b) is to get its lawyer's or other practitioner's written advice?

So if either expense allocation is possible, does a recordkeeper allow an employer to specify which allocation the employer prefers?

If an employer in 2015 states a new plan by using a document that now is waiting for approval and later becomes IRS-approved, does the user get retroactive reliance on the IRS's letter?

Some 401(k) recordkeepers offer an optional service, for an incremental fee, under which the recordkeeper will review a domestic-relations order to decide (but for the plan administrator's rubber stamp) that an order is a QDRO. For those that offer such a service, does the recordkeeper allow the plan's administrator to specify how the expense is allocated among participants' accounts? Can the allocation be proportionate across all participants' balances? Does a recordkeeper require or permit an allocation of these QDRO-service expenses only to the account that is the subject of a division or proposed division? Does the recordkeeper provide an indemnity to stand behind the accuracy of its decision or recommendation? Is this kind of service worthwhile?

David Rigby, thank you for that reminder. But isn't the problem of an employer stealing participant contributions concentrated mostly in plans with few enough participants that an independent qualified public accountant is not engaged?

Belgarath, thank you for the information. Do you know which providers have submitted a document to the IRS?

We seem to have two votes for a non-fiduciary doing no more than informing the employer that failed to pay the contribution. Any different views?

Of the big 403(b) providers, which of them offers an IRS-preapproved prototype or volume-submitter document (or a document so intended)?

Should a recordkeeper or TPA try to get an employer to pay a past-due participant contribution? A recent BenefitsLink forum topic shows how the duty to collect a contribution often is set with someone who decided not to pay the contribution (or his or her subordinate). http://benefitslink.com/boards/index.php/topic/57803-special-trustee-responsible-for-contribution-deposits/ Imagine that you’re such a plan’s recordkeeper or third-party administrator (and you’ve designed your business to be non-fiduciary at every turn). What do you do if a participant contribution is long past due and has not been paid to any trustee, custodian, or insurer? Is it okay to make no effort to get the employer to pay the past-due contribution? Or should you “do something” to try to get the employer to pay? If so, what is the “something” you do?

AndrewZ, thank you for the helpful information. Is the condition that the plan must specify a natural person as the "special trustee" stated on the Adoption Agreement or in its written instructions? If so, I might seek more information about whether the Internal Revenue Service requires adherence on that point as a condition to reliance on the IRS advisory letter. If the condition is not so stated, a user might consider and evaluate the possibilities.

Without disagreeing with My 2 cents’ and Andy the Actuary’s observations, a few further thoughts: If the plan’s administrator accedes to the independent qualified public accountant’s suggestion, there might be an awkwardness about what the administrator reports on Schedule C. Under the Schedule C instructions, “indirect compensation” is “[c]ompensation received from sources other than directly from the plan or plan sponsor[.]” Yet “direct compensation” is only “[p]ayments made directly by the plan[.]” If the investment manager’s fee is paid by the plan sponsor, correct responses to Schedule C’s part I line 2 might be: (d) direct compensation -0- (e) indirect compensation No (f) Did indirect compensation include eligible indirect compensation? [blank] (g) total indirect compensation -0- (h) formula instead of an amount? No To the computer, this might look like the service provider had no compensation. One wonders whether the software would display an “are you sure” message: “Do you intend to report as a service provider a person that had no compensation?” If that happens and the administrator starts thinking about how to “trick the system”, the administrator might reconsider whether to accede to the independent qualified public accountant’s suggestion. If it reconsiders, the administrator might recall that there is a little off-rule guidance that might support a reporting position. The Schedule C FAQs state (in Q 26), not in an answer but rather in a question as a conclusion EBSA assumes: “Plan administrators are not required to report on Schedule C information with respect to service providers receiving less than $5000 in total compensation (direct and indirect) from the plan.” For the auditor, there is an easy retreat. The AICPA tells an auditor to “read the other information of which the auditor is aware [such as Form 5500 and its Schedules] in order to identify material inconsistencies[.]” If the plan’s financial statements show no payment to, and no payable due to, the investment manager, those statements would be not inconsistent with the investment manager’s absence from Schedule C. Likewise, there should be no need for a disclosure reconciling differing amounts between the financial statements and the other elements of the Form 5500 report.

I've enjoyed the benefit I asked for, getting several practitioners' thoughts. Thank you all for helping me focus my thinking.

Mojo, it's easy to agree with your business-sense observations about choice of law for matters other than the terms of an ERISA-governed IRC 401 retirement plan. Your last paragraph shows why an ERISA-governed plan's administrator shouldn't need any State's law.

In fairness, EBSA's FAB 2008-01 said only that a trustee can't be relieved of a duty concerning collection of contributions unless some document provides for the duty to be allocated to another fiduciary (most often, a non-trustee that directs the trustee). The provision AndrewZ describes might be a way to accomplish the goal of allocating the responsibility to somebody, but it is not the only way. And it happens not to be what the Field Assistance Bulletin suggested. (The FAB explained that a document could provide for a non-trustee fiduciary to direct the trustee.) To respond to AndrewZ's query, what do BenefitsLink people think about checking the "other" box and filling-in on its blank line "the Administrator"? If the typical choice for "Administrator" is the plan's sponsor, this might make the organization, rather than a particular natural person, responsible to decide whether to attempt collection of a contribution.

hr for me, My 2 cents, MoJo, thank you for the helpful thoughts. But I consider the spouse and beneficiary examples as illustrations of why some employee-benefits lawyer feel it's at least unnecessary, and perhaps unwise, to mention a State's law. For example, if a retirement plan's administrator needs to decide whether a participant has a spouse such that the participant may not elect a distribution other than a qualified joint and survivor annuity without her spouse's consent, the word "spouse" has the meaning given by ERISA section 205 and the plan (insofar as the plan is consistent with ERISA titles I and IV). And if the plan's terms (which include any required under ERISA section 205) are ambiguous, or the application of the plan's terms to a set of facts is unclear, a typical plan grants the administrator discretion to construe, interpret, and apply the plan's terms, and even to make discretionary findings of facts. It might be strange to apply a Plan-specified State law to a particular set of facts. Imagine a plan sponsor that is a California corporation and in its retirement plan specifies California law. Imagine that an opposite-sex couple have both always lived and worked in Pennsylvania. Does it seem strange to apply California law to resolve a question about the existence or non-existence of a marriage of two people who have always lived and worked in Pennsylvania? More importantly, instead of plan-document language that someone could argue tells the administrator to look to a particular State's law, might an administrator prefer to use its plan-granted discretion? (A court defers to an ERISA plan administrator's exercise of discretion, even if the administrator's decision is contrary to what would have resulted under State law.) MoJo, my query didn't consider State law concerning a trust, because every documenting situation I advise on has a trust agreement that is separate from the plan document. For a trust, even a trust for an ERISA-governed plan, a choice of State law might have a consequence. In many situations I see, the choice of State law in the trust agreement is different from the plan sponsor's choice of law in the plan.

If ERISA governs a plan (and ERISA preempts State law), under what circumstances would a choice of State law be needed?

My 2 cents, Belgarath, and hr for me, thank you for your considerate help. Belgarath, you are right that some preapproved documents include a default: for example, saying a blank line results in a specification of the State in which the plan's sponsor is incorporated or similarly organized. (That's why sometimes a response might be writing-in the word "none" as an attempt to negate such a default.) A CCH/Fort William volume-submitter document I saw recently did not include such a default. I am not worried about the Internal Revenue Service tax-disqualifying a plan. Instead, I'm thinking about whether an employee-benefits lawyer can advise his client that leaving blank (or negating) the State-law specification does not undo the user's reliance on the preapproved document's IRS letter. If one can't be confident about that conclusion, a lawyer advising his or her client about its use of a preapproved document might feel a need to explain the advantages and disadvantages of specifying a State law and the relative probabilities of the risks involved. The IRS's resolve to provide fewer opportunities for a user to get a determination will make more important questions about whether a user's adoption agreement is within, or strays from, a preapproved document's confines. BenefitsLink mavens, any further thinking on whether a fair reading of the Revenue Procedure allows a user to rely on the preapproved document's IRS letter if the user omits a State-law specification that the document asks for?

A typical prototype or volume-submitter document's adoption agreement often includes a fill-in-the-blank choice for specifying a State's law to govern whatever ERISA doesn't preempt (if anything). Imagine that a user prefers not to fill in this blank. (The plan's sponsor is worried that a specification could be argued to constrain the administrator's discretion to a narrower range than would apply if the plan states nothing about State law.) The rules for relying on the IRS's letter for a preapproved document call for staying within the confines of what the IRS approved. But the Revenue Procedure suggests that a user might vary some "administrative" provisions without losing reliance on the IRS's letter. What do you think? If a user's adoption agreement leaves blank the State-law line (or responds "none"), does the user keep or lose reliance on the preapproved document's IRS letter?

jpod, thank you for the helpful learning!

Among other conditions, the exemption requires approvals by a fiduciary who is independent of the person who engages in the covered transaction. The exemption's definitions and special rules states: "A plan fiduciary is independent of a person only if the fiduciary has no relationship to or interest in such person that might affect the exercise of such fiduciary's best judgment as a fiduciary."

Could the administrator omit a report on each of 2013 and 2014 based on an interpretation that the plan had no reportable asset on any day in either year? (Because none of us knows the facts about the "expense recovery", I'm imagining the possibility that the situation that did not produce a payment until 2015 might be one that generally accepted accounting principles might treat as one that had no value to be recorded as at 2013 or 2014.) And could filing Form 5500 reports (perhaps with lots of zero entries for 2013 and 2014) and furnishing summary annual reports to participants help protect the administrator and perhaps other fiduciaries by setting up knowledge to run a statute-of-limitations period? I would at least ask the client to decide what accounting to participants and regulators the client prefers.

While agreeing with much of the practical suggestions, is it so clear that a Form 5500 report would be too much burden or plan-administration expense? If the plan had only 75 participants as at the first day of a plan year and the plan's one asset is derived from assets that were qualifying assets, perhaps no independent qualified public accountant's report should be required. While there might be a recordkeeper's or TPA's fee to prepare a report, it should be not imprudent to incur a reasonable fee for what's necessary to file a report required by the statute.