Peter Gulia

ERISA 405(a)(3) makes an observing fiduciary "liable for a breach of fiduciary responsibility of another fiduciary with respect to the same plan ... if he has knowledge of a breach by such other fiduciary, unless he makes reasonable efforts under the circumstances to remedy the breach." This is why those stuck in a co-fiduciary role try to avoid receiving unwelcome information. In some roles and services, it's difficult or impractical to avoid knowledge. For example, if a recordkeeper and a directed trustee are the same person, it's hard not to know that a participant contribution has not been paid over to the trust.

Brett Ashley's query supposes that the employer is a not-for-profit corporation, and the context of the doubted information suggests that the not-for-profit might be a charitable organization. Could the query be about whether an employer with fewer than two employees is a non-group?

SearchLight, thank you for the good help. Fear of co-fiduciary responsibility concerning an employer's irresponsibility or ineptness keeps service providers away from the employers that need help the most. But could that be an argument in favor of a separate QDRO service provider so it would deliberately lack knowledge beyond its assigned function?

Would it be simpler for the different businesses (even if they are parts of one employer) each to maintain its distinct plan? While those plans might be aggregated for coverage, non-discrimination, and top-heavy testing, is there anything precludes each business from maintaining its plan with its particular provisions?

Thank you, responders, for the helpful information. For those who are providing a QDRO-review service, what steps do you use to defend the idea that your review is merely clerical or ministerial, that you do not in practical effect make the fiduciary's decision, and that your service is not the practice of law?

Based on a recent court decision, it seems that Fidelity's fee for a review of an order is $300 if the order was generated using Fidelity QDRO Center and not altered, or $1,200 if the order is not so generated or is altered. Thoughts about this?

Focusing on the task of deciding whether an order submitted to a plan for treatment as a qualified order is or isn't a QDRO, how often, in your experience, does the employer/administrator make an incorrect decision? How often does the employer/administrator decide that an order is a qualified order when the TPA or recordkeeper knows the order is not a QDRO? How often does the employer/administrator decide that an order is not a qualified order when the TPA or recordkeeper knows that it is a QDRO?

Anyone have experience with Fidelity's procedure for ministerial decisions on whether an order is a QDRO?

Perhaps the IRS was wise in closing its examination without any handling of an examiner's disguised-contributions theory. However, that exam was before the Roth-ing era.

Fearing co-fiduciary responsibility is the strongest reason for not taking on a fiduciary role. Has anyone seen a QDRO-processing service designed to be "ministerial" so that it's non-fiduciary?

QDROphile's point about too-high interest resulting in a disguised contribution is real. A plan's administrator (without getting advice from anyone) had considered that a participant loan is a hybrid between an unsecured loan and a semi-secured loan (semi-secured because the participant's investments can lose value, and the set-off right is delayed until the participant's severance from employment), and had interpreted the idea of "loans [that] would be made under similar circumstances" to look to credit-card rates. Later, an IRS examiner informally raised the idea that high interest on repayment of a participant loan allowed participants to put more money into tax deferral. I'll never know whether the IRS would have pursued that idea; the examination closed without any handling of that issue.

Paragraph (6) refers to "[e]xpenses for the repair of damage to the employee's principal residence that would qualify for the casualty deduction under section 165 (determined without regard to whether the loss exceeds 10% of adjusted gross income)." If the plan's provision follows the rule's text, could an administrator interpret this to recognize a hardship if the expense is of a kind for which someone (not necessarily the participant) could get an IRC 165 deduction (if one meets or ignores the 10% condition)? Apart from that point, one might want some evidence that the claimant paid, incurred, or would incur the expense, and that it is reasonable for the claimant to pay for repairs to something she does not own. Without that, one might question whether the claimant has an "expense" or a "need".

I am interested in learning more about the several QDRO-outsourcing services available in the retirement-services markets, and hope to get information from BenefitsLink mavens’ wide experience. Of the big recordkeepers, which ones offer a service of deciding whether an order is a qualified domestic relations order? For those that offer a service, does the service provider accept responsibility as a fiduciary to the extent of its QDRO-or-not decisions? Or does the QDRO service provider get the plan’s named fiduciary to instruct the service provider to follow a written procedure it designed so that the service provider is not a fiduciary? I have seen QDRO service procedures that authorize the service provider to approve as a QDRO only an order that is identical (but for the names and addresses, and filling-in the amount or percentage to set over to the alternate payee) to a specified form of order. All else the service provider turns back to the plan’s administrator. How common is this way of doing a QDRO service? If a QDRO service procedure is not so limited as described in the preceding paragraph, what techniques does the procedure use to get rid of discretion? Do other recordkeepers or third-party administrators offer a service of deciding whether an order is a qualified order? Again, does one design it to be fiduciary or non-fiduciary? Are there are any “stand-alone” QDRO service providers that are not a part of or affiliated with a recordkeeper or third-party administrator? What methods do they use? If your customer says it wants to outsource QDRO decisions, what service provider do you suggest to your customer?

ERISA 408(b)(1)(D) and 29 C.F.R. 2550.408b-1(a)(1)(iv) call for "a reasonable rate of interest". The rule's subsection (e) and its examples suggest that a fiduciary that sets the participant loan interest rate should see to it that the plan (practically, the borrowing participant's account) doesn't get a too-little investment return. But nothing in the rule precludes an interest rate that is more than what a hypothetical bank would get. A fiduciary might consider whether the exclusive purpose of providing retirement benefits to participants could better be met by a higher interest rate. Depending on the plan's particular facts and circumstances, participants (as a class) might benefit from a higher interest rate. Another plan's participants might benefit from the lowest rate that the fiduciary can defend as meeting the conditions of the statutory prohibited-transaction exemption.

AlbanyConsultant, two thoughts: Your observation about a possible multiple-employer plan is clearer if you delete the last five words and focus on the first six words. Consider whether, even if the governors might not result in common control under 26 C.F.R. 1.414©-5(b), there might be some opportunity for permissive aggregation under -5©(1) if the organizations sufficiently coordinate their activities.

When efforts of the kind that SearchLight describes are unavailing and a would-be alternate payee really digs in, I've seen plan administrators assert a position that an order that would require the administrator to compute a payment or set-aside using an amount that can't be retrieved from the plan's records cannot be a qualified order. As a part of this, the administrator asserts that it need not look to records that, after applying ERISA and Internal Revenue Code retention requirements, it was not required to keep (and in fact discarded). What do BenefitsLink mavens think about this?

The Internal Revenue Service's Notice 2015-17 in its Q&A 2 states some non-enforcement relief. It also states: "If an S corporation maintains more than one such arrangement [to pay for individual health insurance premiums] for different employees (whether or not 2-percent shareholder-employees) ... all such arrangements are treated as a single arrangement covering more than one employee so that the exception in [internal Revenue] Code section 9831(a)(2) does not apply."

Thank you for the pointer. It leads to another question: If a participant does not consent to the terminated plan's distribution and the employer's other plan by its terms precludes a transfer-in (a common provision in my experience), how does the administrator of the terminated plan complete the termination?

Is it feasible to apply each employee's actual 401(k) election that was in effect for each period in which the employee performed the overtime that is a subject of the late-paid wages? If so, is it feasible to correct all W-2s and other tax and information returns for all affected periods?

Back to the originating query, if the employer will amend the plan to discontinue contribution accruals and provide a final distribution, could the employer also provide that for any eligible rollover distribution for which the distributee doesn't specify what he or she wants the default distribution is not a payment of money but rather a direct rollover to the employer's other plan?

Thank you for the help.

Let me ask a question (and please understand that it really is my open question from a lack of payroll knowledge): Once the employer has, on each employee, figured the period's wages, applied section 125 and section 401(k) reductions, counted net wages, computed withholding taxes, applied deductions, and counted payable wages, shouldn't the elective-deferral amounts to be paid over to the plan be known? What further steps are there before an employer is ready to transmit data and remit money to the plan's service provider?

So were the plan's provisions and testing procedures applied consistently across all participants? And if so, were they applied consistently over the years?

If the retirement plan's administrator relies on an incorrect measure of compensation for this participant, how would it affect the administration of the retirement plan? Would a too-low measure of compensation result in this participant getting an allocation of a nonelective contribution less than she is entitled to? Would a too-low measure of compensation result in this participant getting an allocation of a matching contribution less than she is entitled to? Would a too-low measure of compensation distort an ADP test, ACP test, or other measure of coverage or non-discrimination? Is something else affected? Focusing on what plan-administrator error could result from a mistaken measure might help your client spot a problem or exposure (or see a need for a plan revision). Or if the measure of compensation doesn't affect anything (which might be so concerning plans in some circumstances), it might be unnecessary to use any measure of compensation.

Flyboyjohn, thank you for the helpful information. And thank you for politely assuming I know more than I really do. leevena, the point of my hypothetical is to imagine an insurance-buying employer that is less likely to have engaged an employee-benefits lawyer or an actuary, and so looks to other sources of information.