My 2 cents

If a participant has terminated employment with a small vested benefit/account balance, the law gives the plan administrator (consistent with applicable plan provisions) the authority to force them out. Best is for the participant to say how they want it distributed. Failing that, the plan administrator, having made a suitable effort to find and/or communicate with the participant, has the right to cash the small amount out. Unless the plan administrator fails to exercise due prudence in the selection of the default IRA provider (which might not even be subject to the same degree of scrutiny as other fiduciary acts), the expenses charged by the default IRA provider are of no consequence to the plan administrator. Remember, if the amount payable is over $200, there would have been an opportunity provided for the participant to choose between a lump sum or a direct rollover, and only if there is no timely response can the money be sent on the IRA provider, so the participant, by failing to keep in touch with the sponsor and/or to respond to the election that was offered, is at least partially at fault. The plan will not afford veto power over the distribution to the participant - these are involuntary by nature.

Based on my understanding of how this is handled in defined benefit plans: I am not sure, but surely the plan can mandate that all benefits worth less than $1,000 be paid out in a single sum without there being a default IRA provider. To involuntarily cash out benefits worth more than $1,000, there must be an arrangement with a default IRA provider. If the default IRA provider will allow default rollovers down to just over $0, why shouldn't it be acceptable to make it so that smaller values will be subject to default rollover in the absence of an explicit election by the participant to the contrary (for a lump sum, net of mandatory withholding, or for a direct rollover to an institution or plan of the participant's choosing if the amount is at least $200; under $200, no choices need be given)?

How much is he making from fishing to make setting up a 401(k) plan worth the effort and expense? How much could the tax savings be? Also, something about a 15-year old with earned income using it for neither college savings (and PLEASE don't say that the 401(k) plan would be used for that purpose!!!! UGH!) nor to help pay for clothes, electronics, movies, and other niceties strikes me (for what that's worth) as excessively prudent. Mid-20's, by almost any rational perspective, is soon enough to be saving for retirement. 401(k)s are ONLY supposed to be used for retirement.

My answer is based on an IRS publication from 1998 (referred to as "Chapter 6"), which I presume is essentially still true. As noted, I am have not had to deal with such situations myself. The general method for determining RMDs is to divide the present value as of the Valuation Date in the year prior to the RMD year in question (in this case, December 31, 2016) by a suitable life expectancy. In the case of a defined benefit plan paying non-decreasing monthly benefits for at least the life of the participant, continued payment of such amounts (or the purchase of an annuity contract to make such payments) would be considered to satisfy 401(a)(9). The regulations as summarized in that document do not appear to divide the RMD year into portions. With respect to the first RMD year and each subsequent RMD year, commencement of those regular payments by April 1 of the year after the first RMD year would seem to satisfy the RMD requirements (presuming no changes to the amount or frequency of the regular payments). Therefore, it would appear that the 85-year old's RMDs are to be considered satisfied by the life annuity payments through the end of the year before the change to a lump sum approach. If the payment methodology is being changed this year, it is my interpretation of the regulations as summarized that this year's RMD is NOT based on special rules applicable to defined benefit plans paying uniform amounts for at least the lifetime of the participant. The rules say to me that this year's RMD is based on the value of the benefit as of the end of last year divided by a suitable life expectancy factor. They also say to me that the first 6 monthly payments count towards the satisfaction of this year's RMD. If the RMD net of those 6 monthly payments is greater than $0, that excess cannot be rolled over to an IRA (which would, perforce, not have an RMD until next year, since the 12/31/16 balance was $0). As there is not an RMD for the first few months of the year and another for the last few, one cannot conclude that the amount that cannot be rolled over to an IRA is merely equal to the sum of the 6 monthly payments that are not going to be made. Having reviewed this Chapter 6 document concerning the requirements of 401(a)(9), I consider my earlier "opinion" to be consistent with the regulations (as indicated above, assuming no material subsequent changes in the regulations as they would apply to this situation).

Incidentally, whether someone is an HCE or not is NOT plan specific. The top-25 HCEs subject to that restriction may ALL be non-participants in the plan in question, so a plan, even a plan with HCEs, may in effect be able to ignore the top-25 restrictions because none of the top 25 have any benefits in that plan.

I may be wrong, but I think that "Is this person a highly compensated employee?" has one answer for all possible purposes. To the extent that there is any give in how one determines whether someone is a highly compensated employee, pick one method and use it. There should be no such thing as someone who is an HCE for one purpose and not for another (or vice versa). For example, if it is possible, by limiting the HCEs to the top 20% by election, to wind up with fewer than 25 HCEs and there is someone who, but for that election, would have been an HCE, by virtue of the top 20% election is not an HCE, and would have been one of the 25 highest-paid people in the history of the sponsor (with earnings high enough, other than the top 20% election, to meet the initial definition of an HCE), then that person is not subject to the 25 highest-paid limitations (I think).

My guess (for what that is worth) is that the RMD to be distributed outside of the rollover is equal to the lump sum present value of the life annuity as of the end of the prior year minus the 6 monthly payments already received, and as the monthly payments are stopping, the RMD can no longer be based on 12 equal monthly payments under the special rules for a defined benefit plan. Of course, not having had to deal with such a situation myself, I could be all wet, but you asked for an opinion.

Addressing only the first question: I know of no rules that would prevent assignment of more than 50% to the alternate payee. As a QDRO cannot increase the plan's liability, it would seem that 100% of the account balance to the alternate payee is the absolute limit for a QDRO. Perhaps the participant is keeping the $250,000 house, and the divorce agreement may (among other things) give the $30,000 car, 100% of the 401(k) balance, and other assets to the alternate payee to reach an equitable distribution of the joint assets. Presuming that both sides agree to the distribution of assets, it is not for the plan administrator to worry about the QDRO giving most or all of the plan balance to the alternate payee. Concerning the second, and not being an attorney, I would be a bit leery of a request to provide such a draft QDRO, especially if I did not know for sure that it was consciously agreed to by both sides. Wouldn't want to aid one side in defrauding the other! Would you still get paid if you sent the draft to both parties, pointing out the change in the cover letter?

My view is that asking "what RMDs are required when allowing 85 year old retirees to elect to switch from a lifetime annuity to a lump sum when a plan terminates?" is asking the wrong question. Asking "does it ever make sense to offer 85 year old retirees the opportunity to elect to switch from a lifetime annuity to a lump sum when a plan terminates?" is the place to start. Not a rational way to save the sponsor money and relatively unlikely to improve the quality of life for the retirees, whose ability to successfully manage the proceeds may be impaired by age.

Offering lump sums to people 20 years out of the workforce is playing with fire. If the participant elected a straight life annuity in 1997 with spousal consent, would spousal consent be required in 2017 to switch the payment from a life annuity to a lump sum? Does your answer depend on whether the participant's current spouse is the same person who waived the QJSA in 1997? The participant may have been single then but is married now. If the participant was married, was divorced, and remarried, it is possible that the consent of both the former spouse (if living) and the current spouse would be required. It would, without question, be required if the elected benefit was a joint and survivor form (with the spouse as of 1997 being the joint annuitant). Please don't tell me that people with joint and survivor forms would not be able to elect lump sums if people with life annuity forms could! How can you make sure that the participant (and spouse, if married) completely understand the consequences of the election? The participant has been long retired and is advanced in age. Can an election to change the payment form from a life annuity to a lump sum be made by someone with a power of attorney or guardianship or can such an election only be made by the participant (and spouse) themselves? If someone is living in a nursing home that has guardianship over the retired participant, any idiot can easily see that the nursing home being able to act to switch from a life annuity to a lump sum is fraught with conflicts of interest. If it were up to me, the sponsor should bite the bullet and buy annuities, however much more they would cost. Making the offer should only be done after obtaining an IRS PLR that says it's OK.

Are you saying "no" to the idea behind ERISA preemption (to facilitate a highly desirable uniformity of administrative practices between jurisdictions) or to the question as to whether states can require specific coverages under an ERISA health plan? It has always been my understanding that ERISA preempts most state laws so that similarly situated participants in different states can be administered exactly the same way. A company doing business in a particular state should not have to handle a claim differently just because a participant has retired and moved to a different state. That would place an excessive burden on the plan administrator.

There was a recent Montana ruling on a matter of a conflict between state law and the plan concerning divorce and beneficiary rights. In that case, the spouse was named as beneficiary under an ERISA life insurance plan. The participant and spouse were divorced and no action was taken to change the life insurance beneficiary (when is there ever?). State law terminated spouse's interest in such benefits upon divorce. After the divorce, the participant changed his will to name his two children as being entitled to the entire estate. When the participant died, both the ex-spouse and a daughter filed suit claiming the insurance proceeds. As reported in the October 20 BenefitsLink newsletter, the court ruled in favor of the ex-spouse, saying in effect that to administer the ERISA plan, the plan administrator needs to follow the terms of the plan and should not have to worry about state laws that may conflict with the plan provisions. It was not clear whether there was diversity of state residence involved, but it would not appear that if all parties lived in Montana that the court would have ruled differently. As a general rule, ERISA preemption exists so that a plan sponsor need not hire a battery of attorneys to make sure that various state laws are being followed. The case cited by Fiduciary Guidance Counsel above involved a national insurance company (whose employee health plan had participants and beneficiaries in all 50 states) fighting against having to isolate claim date for Vermont residents and reporting that information, to enable Vermont to judge whether state law rules for health plans should be changed. I am trying to remember - can states impose specific coverage requirements on ERISA health plans or is that sort of authority preempted?

All of this depends very much on the actual plan language. 1. There are three kinds of service (which may, depending on plan language, overlap): Eligibility service (used to determine eligibility to become a participant and, depending on plan language, possibly to determine eligibility for early retirement or disability), accrual service, and vesting service. The plan should clearly indicate how each one works. Eligibility service will often be measured in terms of the first 12 months of employment, usually switching thereafter to plan years (beginning with the plan year beginning during that 12 month period if the person did not complete enough hours during the first 12 months of employment). Benefit accrual service and vesting service would normally be measured in terms of plan years, possibly with partial accrual of service in the first or last plan year of employment. 2. As the individual would surely have entered the plan prior to the most recent date of termination (2015, after nearly 3 years of employment), eligibility service is no longer relevant - resumption of status as an active participant will either be immediate or, after a full year from date of rehire, retroactive to date of rehire). 3. For vesting purposes, as at no point did the person incur a break of at least 5 consecutive years, all service (including the 2008-09 period) must be taken into account for purposes of determining vesting service. It's a good 30 years since the rule of parity operated to disregard a year of service preceding a 3-year break. Nothing can be disregarded unless there has been a 5-year break.

If the participant cannot afford to continue the loan repayments (as stated in the original post), how will they be able to afford the income taxes (and 10% excise tax) on the deemed distribution? No getting out from under that, no matter what their financial situation may be. The plan sponsor, without question, will report the resulting distribution to the IRS, and there will have been $0 withheld to cover those taxes.

How do you get around the ERISA/regulatory mandate that in-service withdrawals can only be made if the hardship rules are met? Statements from the loan recipient under penalty of perjury that they do not intend to convert the loan to a distribution by stopping repayments while employed?

Despite many years working on pension plans, am I naïve to believe that plan documents that don't specify exactly what is to be considered a break in service (including what the measurement period is) ought not to exist? How can something that fails to specify what a "break in service" is even be considered to be a plan document?

I don't work on health insurance, but surely the benefit structure in effect when the employee terminated must be preserved for COBRA purposes. That means no increases in the deductible effective after termination of employment would be acceptable (except as a lower-cost alternative to the straight COBRA coverage). The cost of the COBRA coverage will be set by the insurance company, based on what is required to be included, right? So if the insurance company chooses to push premiums up 20% in the absence of a higher deductible, the employee opting for COBRA coverage is pretty much stuck, right? Is it permissible for HRAs to be offered to former employees? Certainly, the COBRA rules would not appear to mandate continuation of such a program after termination of employment even if it had been in place (which isn't even the case here).

If they were in Europe, they could schedule it for August, when everybody is on vacation for the month. Not too sure that they are obliged to treat "you were closed when the filing was due, as it always is" as an ironclad excuse. The uncharitable question they might raise is "why weren't you ready?", but the excuse might be instead that it could not be filed since the necessary contributions, not yet overdue, had not been completed. Don't you wish you could block off a month-long period to be unavailable? Do they really need that much time for "planned maintenance"? If it were an interstate highway, there would be a detour (rather than "road closed - you can't get there from here"). If waiting until the system were back up was really acceptable, then why don't they say so in advance? Sorry if I sound a bit crabby about this, but I still remember having to hustle the first time they did this to get the handful of necessary filings in before the gates swung shut so they would not be late.

In general, I don't think that an erroneous 5500 filing would subject the plan, otherwise exempt, to ERISA coverage. Why are they a non-ERISA plan? If the sponsor is a governmental entity, I would think that filing a 5500 would have no effect. A governmental plan would remain a governmental plan, exempt from coverage by ERISA. I also recall that a non-electing church plan would have to do more than file a Form 5500 to effectively elect to be subject to ERISA (i.e., there would have to be an explicit election of coverage - merely filing a Form 5500 would not be considered an election of coverage).

As good an explanation as any!

Is it possible to distinguish between "insurance services" being paid from the plan and "fiduciary services" paid to the same recipient? If paid to the same person/organization, aren't both considered together as being subject to fiduciary standards? Not that I would know, but it would surprise me if they were subject to different standards.

I have trouble seeing the deduction to repay the loan as "optional". Agreeing to payroll deductions was a condition precedent to receiving the loan in the first place. If the participant signed an agreement to repay the loan based on a specified repayment schedule (to be handled via payroll deductions), there should not be a state law that would interfere with the terms of that agreement being carried out, and unless the loan agreement gives the participant the right to cease making repayments at will, the participant should be treated as bound by the agreement to repay. If forced to allow the loan recipient to stop repaying this loan, the sponsor ought to consider amending the 401(k) plan to prohibit such a choice for all future loans. If the participant cannot afford to repay the loan, it should be considered from the start as a withdrawal (and subject to all in-service restrictions on withdrawals). 401(k) plans are not supposed to be used to pay current expenses (except for permissible hardship situations) - they are supposed to stay in place to provide retirement income/assets.

Not a lawyer, not an expert on 401(k) loans, but... Isn't the idea that administrators of ERISA plans are not required to be conversant with the laws of numerous states, ergo ERISA preempts state laws? What is a plan with active participants in 35 states supposed to do, administer the repayment of loan provisions 35 different ways? The best approach would probably be put something explicit in the plan document concerning the handling of such requests (absolutely not or whatever the participant wants, with favorable determination letters having been obtained for the plan document) and follow it without regard to the laws of the state the participant lives in. If it's a prototype, there ought to be a yes or no checkbox in the adoption agreement concerning the ability of participants to stop making required repayments. State laws cannot require that the provisions of a federally regulated pension plan be violated, due to ERISA preemption. The handling of loan repayment requirements in 401(k) plans should automatically be treated as outside the scope of state wage payment laws. Those should not be treated as deductions from wages the way salary reduction amounts to generate 401(k) employer contributions are.

If the plan permits it, so be it, but then the participant would have to deal with two transactions (since there is no chance that moving the money back from the IRA to the 401(k) plan would negate the rollover from the 401(k) plan to the IRA): 1. Distribution from 401(k) plan (direct rollover to IRA) 2. Distribution from IRA to 401(k) plan (do IRAs have direct rollovers?)

OK - when measuring whether a cash balance plan meets the 133 and 1/3 rule (generally the one to go with when assessing whether the plan meets the accrual requirements of ERISA), measure the monthly life annuity expected to be provided at NRA by each year's pay credit, with interest credits assumed for every future year based on the plan provisions (using the current year's rate if it's variable - and please don't ask me how to handle a market loss when tying the interest credit to actual investment performance, since I had no experience with such things). Feel free to ignore what I said earlier about considering the amount of the pay credit if it confused the issue. My comments with regard to a flat pay credit may have just represented a shortcut, not explicit in the law or regulations. Accrual of interest on the current balance in every future year is part of the "accrued benefit" even in cash balance plans defining the accrued benefit as the current balance. Participants are always entitled to future interest credits, and I go along with the idea that the plan cannot be amended to reduce the rate of future interest credits (except as permitted by the regulations when the interest crediting rate must be modified to meet the regulations, which is, I believe, something that had to have been done already).