My 2 cents

May be off-track, but if the person turned 70 1/2 in 2014, don't you have to look at the 12/31/13 balance for the minimum distribution for 2014 (due 4/15/15)? If the person turned 70 1/2 after 12/31/14, you don't have a minimum distribution due until 4/15/16, but it would be for 2015 and could impact any attempts to roll over proceeds during 2015.

In today's news - TurboTax has completely shut down state form filing (all states) pending enhancement of their security.

Real estate trusts with liquidity issues do not strike at the security of insurance companies that made the investment available. The insurer, if the legal documents and sales methods are handled properly, with proper warnings about the potential for illiquidity, could never be called upon to pay anything to those who could not get their money out. The lack of the real estate trust's liquidity should have nothing to do with the financial condition of the sponsoring insurer. Mutual Benefit (taken over in 1991) is ancient history! Back then, top ratings turned out to be inadequate indicators. Aren't the ratings agencies (assuming there are still ratings agencies) more careful in their analyses these days? And Wikipedia says all of their policyholders were made whole. That sort of thing is why I think that the plaintiffs' claims about losing security do not hold up. Other insurance companies would necessarily step in to cover the losses were the big, secure company from whom the annuities were purchased ran into financial difficulties. Concerning your second point, do state garnishment laws apply to annuities purchased when a plan terminates? Are annuities purchased to settle ERISA plan liabilities not accorded treatment as still subject to federal jurisdiction since all benefits and rights under the plan must be protected? Why would there be any distinction at all between annuities purchased by a non-terminating plan and those purchased by a terminating plan?

In today's BenefitsLink Retirement Plans Newsletter, there is an article on the assertions being made by the Verizon retirees who have been fighting against the de-risking transaction under which annuities were purchased for their benefits from a large, secure insurance company, to reduce the size of the Verizon plan. If I understand it correctly, there were two areas in which the assertion of harm is being made: 1. The annuities are not considered as secure as continued coverage by the PBGC. 2. The annuities, no longer being paid from an ERISA plan, do not provide the same degree of protection from creditors in the event of personal bankruptcy. In thinking that both are entirely wrong, am I missing something? I just don't understand why the lawsuit is not considered frivolous. As I understand it, the plaintiffs are the people for whom annuities were purchased, not those left behind. 1. Considering especially the self-declared tenuousness of the PBGC's finances, wouldn't proper fiduciary care have resulted in the selection of an unquestionably strong insurer at least as capable of standing behind the benefit promise as the PBGC, and even if things were to turn sour in the future for that insurer, wouldn't the state guaranty associations substantially reduce the risk to the retirees (or are we contemplating some sort of asteroid-strike-sized economic calamity, in which case would the PBGC survive)? 2. I am not a lawyer, but surely any annuity purchased to settle benefit rights under an ERISA pension plan would have to provide the same iron-clad protection against creditors as would arise were the benefit obligation still under the pension plan, wouldn't it? Despite the assertions in the article, wouldn't the purchased annuities still be covered by ERISA? If the plan sponsor decided to terminate the plan outright (an act by the sponsor which presumably cannot be legitimately challenged by the participants if not covered as a result of collective bargaining), when considering just those for whom annuities were purchased in the de-risking transaction, when the termination was completed and annuities were purchased for the retirees, how would they be any better off than they are under the de-risking transaction? Certainly, if annuities are to be purchased for retirees in pay status as part of a full plan termination, the retirees have no right to demand a lump sum payment instead! In nearly every plan termination of a defined benefit plan, the current retirees are given no choices. An insurer is selected and those in pay status are taken care of by an annuity purchased from that insurer.

How would it be handled if everyone wants to roll their piece over? If everyone wants to let it sit undistributed? Clearly, if everyone takes cash, no problem.

Not really familiar with the rules governing this sort of thing, but I would expect that they would require the RMD be allocated proportional to the distributions. Is this not so? I would not expect that the RMD would be affected by the methods chosen to make the distributions. If the split of the account is equal, I would expect the RMD to be split equally.

1. Changing from basing the funding on distribution under the normal form (say straight life annuity) to basing the funding on anticipated distribution as a lump sum is never, in my opinion, a change in method. Absolutely, 100% of the time, it is a change in actuarial assumptions. 2. The plan having been amended (presumably) to permit distribution as a lump sum (which will only happen with the approval of the participant and, if the participant is married, the spouse), the law requires that the funding be based on the enrolled actuary's best estimate as to the percentage of benefits to be paid as lump sums. If an assumption of 100% utilization works for you, that's what you use. 3. Adding a lump sum option does not have a material impact on the funding target, unless the plan calculates lump sums using the greater of a fixed (low) discount rate or the 417(e) rates. If the plan specifies that the lump sums are to be based solely on 417(e) rates (it does specify that, right?), assuming that benefits will be paid as lump sums has virtually no impact on the funding target as a result of the substitution rule (the requirement that the expected lump sums used for the funding target are to be determined using the funding discount rates - yes, those under HATFA - irrespective of how much more will be payable to anyone who actually takes a lump sum payment). Given the substitution rule, the only real difference between assuming an annuity and assuming a lump sum is that the former is based on sex-distinct mortality and the latter is based on blended unisex mortality. 4. As for needing IRS approval for a change in assumptions that reduces the funding shortfall by $50 million (that's the benchmark for needing IRS approval, right?), don't give it a thought unless the plan is huge and covers a population that is almost all female or if the change in assumptions includes other changes that will materially reduce the funding target. And, yes, if there was no shortfall before the change, making the funding target lower will not reduce the shortfall.

Not an expert on these matters, but for the plan to reverse the distribution, wouldn't the participant have to return 100% of the distribution to the plan (including the amount withheld for taxes)? If she could do that, she could get out of the situation cleanly by rolling that amount over to an IRA. Or am I missing something here?

Agreed!

For what it's worth, it is my understanding that,under most circumstances, collectively bargained plans get pretty much a free ride on the non-discrimination and coverage rules.

I thought that if you give out the election materials based on a specific ASD and the QJSA notice, both before that ASD, then the ASD holds if the election is made less than 180 days later. No RASD language is needed under those circumstances. Example: Election materials and QJSA notice are given out mid-December 2014 for a 2/1/15 ASD under a plan that does not allow RASDs. The election is returned, signed by the participant in mid-April, and payments based on the 2/1/15 ASD (i.e., no further actuarial adjustments) commence 5/1/15, with a catch-up payment for February, March and April, with appropriate interest. No violation has occurred.

Still, if the plan defines service as elapsed time (as it should when the sponsor cannot track hours), wouldn't you use an elapsed time standard to determine who is an OE under 410(a), not an hours-based standard?

1. Probably would do no good in this particular instance (assuming that it would not pass), but isn't it optional to split the plan into "main" and OEs? 2. Aren't you concerned that a sponsor who can't track hours accurately and a plan that bases eligibility etc. on hours worked is an operational defect waiting to happen? If they can't do it, the plan should call for the use of elapsed time, not hitting an hours benchmark.

Agreed! Under any qualified defined benefit plan, if a vested terminated participant reaches NRA, only one of three things can happen: 1. Start payments at once. 2. When payments start later, issue back payments from NRA 3. When payments start later, actuarially increase the amount payable from NRA. The burden is not on the participants to come in (although doing so is helpful), and failure to file for them cannot cost the participant post-NRA amounts. I remember seeing plans that specified that nothing was payable between NRA and the participant making a benefit claim, but that is in flat violation of ERISA.

The Academy had written to the PBGC because the PBGC, in a couple post-termination audits, had indicated that there were issues in those plans. Those plans had specified a pre-retirement mortality assumption in the definition of actuarial equivalent. The Academy, pointing out that there are three components of an actuarial increase for deferral beyond NRA. Simplified, they are (a) accrual of interest on the value of the benefit as of NRA, (b) the fact that annuity factors at NRA are higher than those at the deferred retirement date, and © the results of forfeitures due to mortality between NRA and the deferred retirement date. The Academy letter asserted that even if there is an assumed pre-retirement mortality assumption, it would not be appropriate to require plans to factor the mortality component in if there is no possibility of forfeiture. The PBGC response says that they do not actually require that unless the plan provisions actually call for it. No details were provided in the PBGC response as to what plan provisions came into play for those plans. Consider a plan (with a pre-retirement mortality equivalence assumption) whose pre-retirement death benefit is fully subsidized. The plan may even contain language making that assertion. If so, should the "cost" of the post-NRA pre-retirement death benefit, if no forfeiture is possible, be passed along to the participant through diminished actuarial adjustment factors? Suppose a plan's pre-retirement death benefit is a distribution equal to the full value of the accrued benefit for married participants and $0 for unmarried participants. Would it be correct to apply higher actuarial increase factors to the unmarried participants than to the married participants if the plan asserts that the pre-retirement benefit is fully subsidized?

1. What does the plan say? As noted by others above, it depends on whether the plan says suspension notices must be provided and that actuarial increases do not generally apply. Remember - if the plan says to give suspension notices and they are not given out, that is an operational failure! 2. Please be aware that an actuarial increase for deferred retirement is, under no circumstances, considered an accrual. So freezing a plan cannot shut off the application of actuarial increases (as provided under the plan) for deferred commencement of benefits after NRA.

Concerning the lump sum calculation, it depends. What is the plan year? What is the stability period. If the stability period containing the annuity start date begins in 2015, then it certainly seems that the lump sum should be based on the 2015 table. What discount rates were used? The lump sum amounts as of early 2015 (if based on late 2014 segment rates and 2015 mortality) are higher than lump sum amounts as of late 2014 (if based on late 2013 segment rates and 2014 mortality), so it is important not to use lump sum rates that are not current. But it all depends on the stability period for that particular plan. Also, is it not the case that if the participant is provided all the proper paperwork before an annuity starting date of (for example) October 1, 2014, then even if the election is not made until now, the annuity starting date would remain as October 1, 2014 until 180 days after the QJSA notice was given?

I don't work on 401(k) plans, but could someone explain to me how the investment earnings on the employer match would be calculated if the employer match is not invested promptly? The approach described above does involve employer matches generating investment earnings similar to those produced by the salary reduction amounts, doesn't it? As for Tom Poje's suggestion ("I have taken my copy of the regs, and, in crayon (I used green, I hope that is ok, not sure what color is permissible) under 1.401(m)-2(a)(4)(iii)© which says to be included in the ACP test 'The contribution is actually paid to the trust no later than the end of 12-month period immediately following the year that contains that date' I have added 'Unless the employer only chooses to fund the match once someone terminates'. Hopefully that will suffice and at least I will have something to point to if the IRS asks."), I suggest disappearing ink. Then you could respond to the IRS "I'm sure it was here somewhere!".

Please take note that, with respect to defined benefit pension plans, despite the fact that the IRS regulations do not require plans to reach back before the issuance of the Windsor decision, a suit has been filed demanding qualified pre-retirement survivor benefits in a case where the participant died shortly before the Windsor decision was issued. The suit is based on the Windsor decision establishing that the relevant provisions of DOMA were, from inception, unconstitutional. Remember - the IRS not requiring retroactive application of the Windsor decision cannot prevent individuals from suing for benefits asserted to be due under ERISA. Which is not to say that sponsors must amend their plans beyond what is required by law and regulations, just that litigation may ensue.

Has the 60 day period after you filed the PBGC Form 500 elapsed? If so, you can offer participants their choice between a lump sum and annuity purchase. If not, here are reasons why you cannot do a lump sum window for anyone: The plan would have to be amended to establish a lump sum window other than as part of the final distribution of plan assets. It is not permissible to adopt such an amendment after the effective date of the plan termination. The PBGC does not permit any accelerated distributions between the date the employer decided to terminate the plan and the date 60 days after the 500 was filed, especially if they were not being made as part of the normal administration of the plan. Unlike a lump sum window under an ongoing plan, the choice would have to be between an annuity purchase and a lump sum. It would make no sense to offer a lump sum window where the choice was between a lump sum and the status quo (with another choice being offered, between an annuity or a lump sum, when the termination was to be completed). Like all choices involving lump sums, to obtain proper spousal consent, you would have to offer an immediate QJSA. Lump sum now versus deferred QJSA is not an acceptable choice. If not distributing all benefits at the same time, you still need to watch out for 25-high restrictions. And watch out if paying actives before the distribution of all plan benefits - the plan's qualification could be jeopardized if you make in-service distributions ahead of finishing the termination, if for some reason the termination cannot be completed. It's happened. What if the discount rates used in insurance company purchase rates drop 150 points and everyone decides to demand an annuity purchase (which is indisputably outside the control of the sponsor), raising the additional assets needed to complete the termination by 100% or more?

The SOA tables are fully projected (that is, the age 65 mortality rate depends not just on the year of projection but also on the year of birth. The age 65 mortality rate is lower for someone now age 50 than it is for someone now age 60). Do you think that the tables that will be mandated will use the currently-age-65 age 65 mortality rate irrespective of the participant's current age, the way that the currently mandated tables do?

Suppose that in 2016 or 2017 the IRS decides to mandate the use, for minimum funding and for lump sum calculations, mortality tables based in some fashion on the new Society of Actuaries mortality tables. Does anybody have any comments, yes or no, on what they expect with respect to the following? 1. Will the IRS tables include mortality improvements up to the year to which they apply, with no further mortality improvements for subsequent years (i.e., project mortality for 2018 to 2018 but static thereafter)? Or will the mandated mortality rates involve projected mortality improvements all the way out? The currently mandated tables do not require future mortality improvements for funding and do not involve future mortality improvements for lump sum determinations. 2. Is it safe to assume that the mandated lump sum mortality table will be a 50/50 unisex version of exactly the same sex-distinct mortality tables mandated for minimum funding? That is, both with assumed mortality improvements stopped at the current year or both with assumed mortality improvements all the way out? Any other thoughts on the subject?

Would action by an affiliated employer to participate in a plan be considered an amendment?

The election form and the plan definition of compensation should coordinate. I would imagine that a good 401(k) would define earnings in terms of taxable compensation plus salary deferrals (or prior to salary deferrals). The intention in your example surely is that the person, who would, but for the salary deferrals, receive $10,000 and who elects to defer 10% would in fact wind up deferring $1,000, right? Make the plan document, the intention and actual administration all match up. Are there 401(k) plans out there (I do not work on 401(k) plans) that do NOT include salary deferrals in the definition of earnings? At best, if deferrals are not counted in the definition of compensation, don't you wind up with some kind of recursive mess? Let's see: Pay before deferrals is $10,000, elects to defer 10%. That would be $1,000, but then earnings =$9,000, so 10% = $900, but then pay before deferrals is $9,100 so deferral is $910, but wait, then pay = $9,090 and deferral = aw forget it. Just make sure that the definition of pay includes deferrals.

HATFA modified the benchmark for applying Section 436 restrictions when the sponsor is bankrupt. Starting in 2015, to pay accelerated benefits in a plan sponsored by a company in bankruptcy, the AFTAP must be certified as at least 100% based on non-relief segment rates. Here are two or three questions (assume that the sponsor is in bankruptcy and that the plan year is the calendar year): 1. To what extent would the 2014 AFTAP as certified continue to apply between 1/1/15 and 3/31/15? Is there any carryover for that period, or must there be a fresh 2015 AFTAP certification (range or otherwise) to justify payment in 2015 of any accelerated benefits? Would it matter if the enrolled actuary is willing to certify that the 2014 AFTAP, determined without regard to the relief segment rates, would have been 100%, or must a certification using 2015 assets and 2015 liabilities be issued? 2. Suppose that the PBGC has decided to take over the plan based on a plan termination date in 2014 related to the corporate bankruptcy filing. Would the plan's limitation status for 2014 become permanent because the plan terminated before 2015 (i.e., given that the AFTAP for 2014 was certified as being at least 100%, would IRC Section 436 never become applicable, notwithstanding the fact that in 2014 the non-relief AFTAP would have been lower than 100%)? Would it matter if the retroactive termination date chosen by the PBGC is earlier than the date (prior to October 1, 2014) on which the 2014 AFTAP was certified by the enrolled actuary? Any thoughts concerning these issues would be welcome.