RMD's from non-Spousal Inherited IRA

[Mike Preston]

I have a plan with a deceased participant whose beneficiaries were subject to 1.401(a)(9)-5;Q&A7. The section specifies the life expectancy for continuing required minimum distributions from a qualified plan in the case of multiple beneficiaries. For simplicity's sake, the Plan Administrator elected not to establish separate accounts. Hence, the life expectancy of the oldest beneficiary was used to determine ongoing RMD's from the plan.

Some of the beneficiaries wish to roll over their share of the participant's remaining benefits to non-spousal IRA's. The plan has no objection and wishes to accomodate the beneficiaries. To the extent necessary, the plan was amended to ensure that it allowed for non-spousal rollovers.

The beneficiaries are now wondering whether the election made by the plan administrator not to establish separate accounts requires that ongoing RMD calculations from the separate inherited IRA's be determined based on the single (shortest) life expectancy, as was the case while the monies were in the qualified plan.

Notice 2007-7 states quite clearly that non-spouse inherited IRA's should determine the RMD .... " if the employee dies on or after his or her required

beginning date, the required minimum distribution under the IRA for any year after the year of death must be determined using the same applicable distribution period as would have been used under the plan if the direct rollover had not occurred. "

If my description isn't clear, please ask for clarification.

[Bird]

I think that all the benes must continue to use the life expectancy of the oldest, at least that's what I think the cite says. (I'm wondering if the IRA custodians will be able to manage this automatically.)

[Mike Preston]

I was afraid you'd say that. Is there any option for the plan administrator to change their position with respect to the separate accounts? Certainly, everybody took minimums which are equal to or greater than the minimums that they would have taken had the plan administrator set up separate accounts.

[Belgarath]

I couldn't find any regulatory support for changing it. I think they are stuck. This sounds like one of those situations where you read a tax court decision that says something like, "While we sympathize with the defendant, we are bound to follow the clear requirements of the regulation..."

For anyone who is interested, here's the citation in 2007-7 Mike was referring to:

Q-19. After a direct rollover by a nonspouse designated beneficiary, how is the required minimum distribution determined with respect to the IRA to which the rollover contribution is made?

A-19. Under §402©(11), an IRA established to receive a direct rollover on behalf of a nonspouse designated beneficiary is treated as an inherited IRA within the meaning of §408(d)(3)©. The required minimum distribution requirements set forth in §401(a)(9)(B) and the regulations thereunder apply to the inherited IRA. The rules for determining the required minimum distributions under the plan with respect to the nonspouse beneficiary also apply under the IRA. Thus, if the employee dies before his or her required beginning date and the 5-year rule in §401(a)(9)(B)(ii) applied to the nonspouse designated beneficiary under the plan making the direct rollover, the 5-year rule applies for purposes of determining required minimum distributions under the IRA. If the life expectancy rule applied to the nonspouse designated beneficiary under the plan, the required minimum distribution under the IRA must be determined using the same applicable distribution period as would have been used under the plan if the direct rollover had not occurred. Similarly, if the employee dies on or after his or her required beginning date, the required minimum distribution under the IRA for any year after the year of death must be determined using the same applicable distribution period as would have been used under the plan if the direct rollover had not occurred.

[Mike Preston]

I guess my point is that if the Plan Administrator were to sympathize and ask what the minimums would have been had they taken the position that there were individual accounts, we would respond that they were all less than the distributions actually taken. Hence, nobody would have violated the minimum. Further, since everybody took the same amount, the accounting would yield a result such that each beneficiary has the same interest in the account. It seems to me that if, upon distribution, the Plan Administrator writes a letter to each beneficiary declaring that the individual account method was adopted to determine the absolute minimum for each participant but that each participant received a minimum distribution which kept everybody's account identical, then each beneficiary could determine their own minimum accordingly. Does this go under the definition of no-harm, no-foul?

[masteff]

One thing to consider before going down that road is does the plan in question have any other deceased participants w/ multiple benes and if so were those handled w/ one or separate accounts? What does the plan document say... could it be construed to require separate accounts for benes? (Though that might imply an operational failure.) How different are the ages of the benes (which is to say, how significant is the impact to the younger benes)?

[Belgarath]

Mike - I always get nervous when revisionist history is proposed. I agree that in common sense terms, this foolishness shouldn't be an issue. I just don't know how it stands up upon audit, or what the "real life" risk is even if it were picked up. First, depending upon the amounts involved, the excess they already received over the RMD (assuming the RMD had originally been calculated using separate accounts) is a "regular" distribution, not an RMD. Did the participants request a distribution over the RMD? I can't imagine that the plan permits the Administrator to distribute these funds absent a specific request from the beneficiaries, so now there's an operational error involved as well as Masteff suggested. Why is the Administrator willing to potentially violate, or take any risk whatsoever? That seems odd to me, unless these are family members.

On the other hand, I can't really see the IRS disqualifying a plan for this "mistaken" interpretation, and certainly the beneficiaries aren't going to complain.

Also, seems like a foolish risk for the beneficiaries, since the penalty for not receiving the higher RMD is draconian, and if it goes sour, you can bet they will come back on the Administrator. So I guess that while I agree with you that this SHOULDN'T be a problem since it really is sort of a no harm no foul, I don't think the IRS would agree, and I sure wouldn't sign off that it is ok using MY name.

[Mike Preston]

Yes, these are all family members. No, there are no other participants and there won't be as the plan is terminating. Yes, the election forms allow the beneficiaries to elect the amount of the distribution, subject to being increased if it doesn't satisfy the 401(a)(9) minimum required distribution amount. Yes, they each ELECTED the amount that is consistent with the oldest beneficiary (the sibling that is in charge of this process "asked" each of them to elect the same amount so as to keep the accounting simple - and therefore retain the same exact interest in the plan for all beneficiaries). I'm thinking that the issue is one that can go either way, based on how they document the file.

[Belgarath]

I do not question that proper documentation in the file would perhaps make it possible for this problem go away. Family businesses often have wonderfully correct documentation in such circumstances! I just still wouldn't want my name on the stamp of approval, but I'm pretty cowardly on such things.