J Simmons

The terms "shared payment" or "separate interest" are not statute or regulation terms of art. I've seen a single DRO bear characteristics of both terms, and yet meet the statutory and regulatory standards to be a QDRO. For example, the DRO might specify that at any time the EE reaches the age for early retirement distribution if EE were then separated from service, the Ex may elect to take 50% of the benefit accrued during marriage, in the form of a single life annuity based on the Ex's life or as a lump sum. That would be what I would term a 'separate interest', if elected. The same DRO might specify that if the Ex has not so elected by the time of the EE's annuity starting date, that the benefits accrued during marriage will be payable as a QJSA on the joint lives of the EE and Ex, with the Ex to receive 50% of each payment otherwise payable to the EE until EE dies and then Ex receives the survivor benefits off of those benefits accrued during marriage. That is what I would term 'shared payment'. In such a DRO, if the Ex has elected the 'separate interest' before the EE's annuity starting date, then the remaining benefits accrued during marriage would be payable to the EE free of any claim or right, including as a surviving spouse, by the Ex. I think awarded benefits can be in the vein of a 'separate interest' even if the language of the QDRO does not have the common "all right, title and interest to the assigned benefit" buzz words.

Both QCIO and QDIA relief provisions require 30 days advance notice. Perhaps the ER could leave the Fund A directives in place as to new payroll, but give an notice now expressing the ER's concerns about Fund A to those that have a standing directive for investment of their payroll contributions, explaining that Fund B has been added (if it was not already among the choices), and urging each to choose and give a directive for upcoming payroll contributions into Fund B or another investment option under the plan. For those that do not do so by the next payroll, the presumption would be that they wanted to keep their payroll going into Fund A despite the ER's concerns. Or, if you already have a QDIA noticed up properly, you could notify the EEs that Fund A will no longer be an option effective immediately, that they may affirmatively direct into something other than the QDIA (or the now closed Fund A), but if they do not, they will be considered not to have a directive in place for future payroll deferrals and so such will be invested in the QDIA.

404c1 offers fiduciary relief for investing assets as directed by the EE under certain circumstances. 404c4 offers fiduciary relief incident to changing investment options of similar type, if certain procedural steps are taken. It would appear that while you are leaving 'old money' in Fund A, no new money will be allowed into Fund A. For those that have given a direction that they want their benefits (or a portion) invested in Fund A, they will as to 'new money' have it re-directed into Fund B. Are you not wanting to give EEs affected by this advance notice and opportunity to change their A Fund directives, such as directing the investment of their 'new money' into Funds C, D or E, but instead force that new money into Fund B?

Yes. A 'qualified change' is not limited to just to existing assets. Joint Committee Taxation (J.C.T. REP. NO. JCX-38-06), Act 621:

I've seen 24 months elapsed method from date of hire used. More frequently, two eligibility years, which eligibility years defined as 12 consecutive month periods during which the EE has a plan-specified number of hours or more (plan may not specify more than 1,000 hours for this purpose). The first 12 month measuring period ends on the first anniversary of the date of hire. Subsequent 12 month measuring periods may either end on subsequent anniversaries of hire or on the last day of plan years including anniversaries of hire (i.e., the 12 month measuring periods shift to plan years). Shifting to the plan year can take away much of the advantage that comes with the 2 eligibility year requirement that comes at a cost of not imposing any vesting requirement on ER contributions. It depends on what time in the plan year that the EE is hired. For example, suppose a calendar plan year and Joe is hired December 27, 2008. With the shift-to-the-plan-year, Joe might earn his second eligibility year on December 31, 2009--just a year and 4 days after hired. Joe enters the plan on 1/1/2010. Now suppose Joe had been hired 5 days later, on January 2, 2009. He could not have his two eligibility years until December 31, 2010, and his entry date would then be 1/1/2011.

Freeze the 401k effective the end of 2008 so that no further contributions are made or benefits accrue under that plan. You can then start your SIMPLE Treas Reg § 1.401(k)-4©(1). Then let the loan repayment schedule run its course before terminating the frozen 401k plan.

Had Treas Reg § 1.401(a)(9)-5, Q&A-1(b) not been included as part of the regulatory set, I would wholeheartedly agree. I just don't think it's proper to apply Treas Reg § 1.401(a)(9)-3, Q&A-1(a) in isolation, and there is nothing of which I'm aware in the 401a9 regs that suggests that Treas Reg § 1.401(a)(9)-5, Q&A-1(b) is subordinate to and must yield to Treas Reg § 1.401(a)(9)-3, Q&A-1(a).

Nor does Pub 575 specify that the RMD for the reader's "starting year" (the calendar year in which age 70 1/2 is reached) doesn't have to be made if you die after reaching age 70 1/2 and before April 1 of the next calendar year.

What is meant by a trust being set up under state law? Under most (maybe all) state laws, a trust is not set up until the trustee is holding something for a beneficiary. This corpus requirement does not, however, have to be met until after the first plan year has ended and the employer makes its first funding of the plan. Rev Rul 81-114, IRB 1981-15, 7. So having a signed trust instrument that meets the trust requirements of state law other than having a corpus would satisfy #(ii).

With all due respect, the problem inheres from the facts that the RMD from DC plan benefits is triggered for years in which the employee is 70 1/2 years or older (Treas Reg § 1.401(a)(9)-5, Q&A-1(b)) and the RBD is delayed to April 1 of the calendar year after the one in which the employee reaches age 70 1/2 (Treas Reg § 1.401(a)(9)-2, Q&A-2(a)). So it is possible as the OP posits (albeit in the IRA rather than DC context) that an employee may live to trigger a first RMD and yet die before his RBD, the date that first RMD must be paid or face the 50% penalty. Regarding an employee that dies at any time before his RBD, Treas Reg § 1.401(a)(9)-3, Q&A-1(a) provides that all benefits must be paid within the 5 year rule (or life expectancy exceptions). This is so whether the employee dies at age 45, 62, or March 31 of the calendar year after that in which he reached age 70 1/2. Notice 2007-7, A-17, quoted by Sieve, is addressing the RMD's triggered by the employee's death, whether per the 5 year rule or the life expectancy exceptions. Neither of those as triggered by the employee's death requires an RMD for the year of the employee's death. However, what -5, Q&A-1(b) provides is a different RMD trigger. Not the employee's death, but instead his having reached age 70 1/2 before dying. Treas Reg § 1.401(a)(9)-5, Q&A-1(b) specifies that the year that the employee reaches age 70 1/2 is a distribution calendar year, for which an RMD is required. This provision specifically deals with the calendar year the employee reaches age 70 1/2. Also note, -5, Q&A-1(b) does not indicate that the employee must live to at least April 1 of the next calendar year for the year of reaching age 70 1/2 to be a 'distribution calendar year'. Canons of interpretation would suggest that the more specific (-5, Q&A-1(b)) is favored over apparently conflicting broader provisions (-3, Q&A-1(a)). That would favor there being an MRD being triggered by reason of the employee reaching age 70 1/2 before dying, per -5, Q&A-1(b). However, here, both may be accommodated in a single interpretation, see post #12 above, which is another interpretive preference. The two provisions are not logically inconsistent. The triggering of an RMD under -5, Q&A-1(b) because the employee reached age 70 1/2 before dying does not prevent his death beneficiary from also complying with the 5 year rule or life expectancy exceptions to that rule under -3, Q&A-1(a). I understand the posts pertaining to IRA situations. There is the IRS statement from Publication 590 mentioned above in post #6. And the IRA regs provide that RMDs from IRAs are generally to mimic or piggyback off of the Treas Regs § 1.401(a)(9). I'm not sure that the clarity of a regulation about RMDs in the DC context (-5, Q&A-1(b)), i.e. that an RMD is triggered for the year the employee reaches age 70 1/2, may be ignored on the strength of a publication statement about RMDs from IRAs. The stakes being a penalty tax equal to 50% of a missed RMD amount, I'd want more basis for the DC context than the statement in a publication statement about RMDs from IRAs to ignore a regulation entitled, Required minimum distributions from defined contribution plans. Maybe just my conservative nature, but before I'd hazard that 50% penalty tax I would want something a bit more authoritative.

Gathering the info all the way back to 1998 and preparing now 10 late Forms 5500 is not only a difficult one, but could span quite some time. I'm wondering if there is a preliminary notification one can make under DFVC to put DoL on notice that such are being worked and will be filed, so that in the interim there is no audit/penalty exposure.

Right, but giving those with less than 3 vesting years earned at the time of the change is not required by the regs, but something you might permit voluntarily. Also, for the sake of clarity, this cut at 3 vesting years for those that must be given the option between staying on the old vesting schedule or going to the new one and those that need not be given this option is unrelated to the fact that the old vesting schedule was 3 year cliff.

Thanks, jevd and again mjb. For DC plans, Treas Reg § 1.401(a)(9)-5, Q&A-1(b) provides that the calendar year an employee reaches age 70 1/2 is a 'distribution calendar year' for which an RMD must be made even though RBD is not until April 1 of the next calendar year (and which RMD would be on the employee's life expectancy, for example 1/16th). Treas Reg § 1.401(a)(9)-2, Q&A-2(a) provides that one's RBD (unless not a 5% owner and yet working) is April 1 following the end of the calendar year in which he or she reaches age 70 1/2. Treas Reg § 1.401(a)(9)-3, Q&A-1(a) provides that if the employee dies before his/her RBD, then all the benefits must be paid out under the 5 year rule (unless over the beneficiary's life expectancy beginning the year following death, except that if the beneficiary is a surviving spouse of the employee, then over the surviving spouse's life expectancy when she/he reaches age 70 1/2). I don't see these provisions as incompatible, but as capable of a congruous interpretation. Employee dies after reaching age 70 1/2 but before April 1 of the following calendar year. Treas Reg § 1.401(a)(9)-5, Q&A-1(b) calls for 1/16th RMD for the year of death since he had reached age 70 1/2, that can be paid as late as April 1 of the following calendar year. Treas Reg § 1.401(a)(9)-3, Q&A-1(a) calls for all the benefits to be paid out under the 5 year rule or its exceptions. The other 15/16th's on my example could all be paid out under the 5 year rule or its exceptions.

Because it would clearly be against one's interest to elect the new, graded 2-6 year schedule if you already have earned 3 vesting years by the time of the change, wouldn't it trigger reasonable suspicion by the plan administrator that someone that made the election did not understand what he was doing? Making the election vulnerable to be set aside on that basis anyway?

That emphasized portion is for MRDs based on the beneficiary's life expectancy (as an exception to the 5 year rule), not MRDs based on the employee reaching age 70 1/2. What relieves the year in which the employee reaches 70 1/2 from being a distribution calendar year because he then dies before April 1 of the next calendar year, as clearly delineated in the first two sentences of Treas Reg § 1.401(a)(9)-5, Q&A-1(b)?

Thanks, mjb. Realizing that the OP is about an IRA, do you know if the IRS has a similar pronouncement re RMD if the benefits are in a DC plan at the time of death?

Good point, JanetM. If this happens to be a DB plan or a MPPP--or any plan with a fixed contribution obligation--there would likely be major problems with scrapping it and starting over. It would, in JanetM's vernacular, be FUBAR.

My understanding is that those with 3 or more vesting years earned at the time of the change must be given the option, as to the post-change accruals, to remain with the old vesting schedule (3 year cliff in your case) or switch to the new vesting schedule (graded, 2-6). All amounts vested pre-change must remain vested. Since it would be clearly against the interest of the EE's with 3 or more vesting years earned at the time of the change to go with the new graded, 2-6 year vesting, most such amendments are designed to keep those EE's in the 3 year cliff.

Thanks, Appleby. That seeming bright line is clouded a little however by the definition of distribution calendar year in the context of a DC plan, as set forth in Treas Reg § 1.401(a)(9)-5, Q&A-1(b) defined to include the calendar year that one reaches age 701/2.

Well, Chaz, I'm not anyone else, but maybe you could redesign it as a carrot rather than a stick, and have easier enforcement. Perhaps you could pay a larger part of the married EE's premium if he proves his spouse is covered under another policy than if the spouse is enrolled under your plan.

The other practitioner's response was just recently made to the IRS. No reply yet from the IRS reviewer.

ERISAnut, what's the citation for no RMD for year of death after reaching age 70.5 but before RBD?

Chaz, would the total cost for spousal coverage (including the surcharge) be more than the cost of the coverage (either the extra premiums to a 3rd party health insurer or under self-funded coverage)? If not, then it would seem all of that cost ought to be allowed tax-free. To the extent that it exceeds the actual cost of coverage for the spouse, then the excess is really not for health coverage and ought not be tax-free. If a spouse is eligible for Medicare as that other coverage, I think you'd have a problem. Otherwise, I think from a regulatory basis you can do what you describe. But as a practical matter, how can you enforce it? You have to depend on the 'honor system' to get the information that is against the interest of the employee for whom you are dependent for the info about whether the spouse is or is not eligible for other coverage.

No, but I am assisting another pension practitioner deal with an IRS reviewer on a Form 5310. Must be a new, trainee at the IRS using the Worksheet, not quite understanding it. First request listed 14 'problems' with the plan documents. Supplied with a copy of the documents submitted with the Form 5310, in a couple of hours time I found the provisions in the last restatement (GUST II) and subsequent amendments for each of the 14 issues. The reviewer accepted 11 and by a second round threw the other 3 back at the practitioner. We expanded the explanations on the 3 for the response--again all were properly addressed in the plan documents submitted--with perhaps one caveat. The prototype sponsor had adopted a good-faith EGTRRA amendment back in December 2001, using language word-for-word out of IRS Notice 2001-57, section 2.2.1. about changes in the definitions for determination of top-heavy status. As amended by EGTRRA, IRC 416(g)(3) (and the IRS Worksheet for 5310 review) uses the newer term, 'separation from service', but the IRS' Notice 2001-57 in the model language for a good-faith EGTRRA amendment (and such amendment to the prototype at issue) uses the older term 'severance from employment'. It's nice when the IRS will not even accept its own, suggested model language.

For plans that will not be frozen or terminated before 2009, for ongoing contributions after 12/31/2008 it is a prudent thing for the ER to get an ISA signed with a vendor before sending post-2008 contributions to that vendor. But if no new contributions will go to that vendor, not having an ISA with that vendor by 1/1/2009 will not equate with noncompliance.