zimbo

It is the time of year to revisit the DB rules on RMDs (post 70.5 variety). A few that need refreshing: 1. When using the "term certain" method, assuming no additional benefit accruals since the initial calculation was done, is it necessary to recalculate the prior years' periodic term certain payment due to changes in the 417e interest rates, or does the certain payment from prior years simply continue unaltered? 2. When calculating a term certain annuity for an RMD in 2008, do we value the benefit using the new 417e segment rules and divide by a term certain amortization factor using the 3 appropriate segment rates (analogous to the amort base for funding shortfalls)? 3. When calculating the initial RMD in a DB, assuming the initial distribution calendar year is 2008, is the benefit determined as of the beginning of 2008 or is it determined as of the end of 2008? The regs do say that for increases in benefits occurring after the distribution calendar year, you can wait until the following year to reflect them but the regs are SILENT on the initial distribution calendar year. Any thoughts or opinions would be welcome!!

Why not allocate the Target Normal Cost in proportion to each employee's target normal cost. Then, allocate the remaining portion of the minimum (the Funding Shortfall amort and, perhaps, interest or carryover balance adjustments etc...) in proportion to each participant's funding target. It has a certain logic to it although it is somewhat more work.

I think the plan is out of compliance with QJSA rules. You cannot offer any benefit (except for small lump sums under $5,000) without offering the QJSA as well. So offering immediately lump sums versus QJSA only at NRA does not comply. I would therefore show the immediate QJSA for relative value purposes and then change the plan as well.

I tend to agree with you on #1. In analyzing the Appendix A QNEC correction and whether Otherwise Excludable testing options can apply, I assume that we can rely on the wording in .03 of Appendix A of Rev Proc 2008-50 which says that "a plan may not be treated as two separate plans....in order to reduce the number of employees eligible to receive QNECs". The implication being that they are limiting the employer's options to allocate the QNECs only to non-excludables only while not interfering with the employer's option to run the ADP/ACP tests separately for non-excludables and otherwise excludables in order to derive the applicable QNEC percentage? I sure wish these things were clearer.

I have a client who never did ADP tests for 2005 or 2006. In 2006, they failed ADP Test. Their Plan Document calls for Prior Year Testing. Since this plan had only a 3 month wait for eligibility, we used to Otherwise Excludable option to run the ADP test and this made the extent of the failure less severe. But, since no refunds were taken by 12/31/2007, we need to self correct. I have 2 questions: 1. Using the "One to One" Correction Method in Appendix B, the Rev Proc. says that "the plan may not be treated as two separate plans with one covering excludables…” . Does that mean that I must re run my test without the otherwise excludable option or may I keep the test the same, but share the resulting "One to One" employer contribution with ALL eligibles and not exclude otherwise excludables from sharing? 2. Can I make a QNEC as allowed in Appendix A, even if I used Prior Year Testing? If so, can I derive my QNEC percentage from my ADP Test that excluded "Otherwise Excludables" as long as I share the resulting QNEC contribution percentage with ALL eligibles?

My understanding is that, in accordance with the proposed regs, you project the 12/31/2007 (or 1/1/2008) hypothetical account balances at the plan's applicable accumulation rate (for the 2008 plan year, if you are doing 1/1/2008) until NRA and then this lump sum is discounted back to 1/1/2008 at the appropriate segment rate. That is done for each participant and the sum of all of these numbers is the Funding Target as of the valuation date. In other words, I don't think the equivalent monthly accrued benefits are particularly relevant for this purpose. I have heard this referred to as "funding whipsaw" because you bring up to NRA at one rate and discount back to AA at another rate. The results can be higher FT than the hypothetical account balances or lower depending upon the relationship between the plan's accumulation rate and the particular segment rates. I think mostly the FT will be less than the account balances.

I don't see any relief whereby you do a BOY for AFTAP and an EOY Val for 430 costs, without redoing your AFTAP based upon your final EOY val. Then you have the same materiality issues and risks, don't you? (see 2nd to last paragraph of today's IRS notice).

One possible strategy might be to do our 2008 AFTAPs using 12/31/2007 val results. When we get to run the actual 12/31/2008 val, if we notice "material changes" (i.e. we are now subject to benefit restrictions), we could instead do a 1/1/2008 val as permitted in 2008-73 and be fairly assured that those results would not be materially different than 12/31/2007. This is bulky and awkward, but may be workable. Not yet sure.

If you read this carefully it is far less than meets the eye. Basically, we cannot rely on our 12/31/07 val date AFTAPs for certifying 2008 if the eventual recertification based upon the real 12/31/08 val has a "material change". Material Change in the 436 regs basically occurs when the subsequent certification takes a plan with no benefit restriction and puts it into a category requiring benefit restrictions (or vice versa, I guess). The way plan asset values have gyrated during the course of 2008 (usually on the minus side) means that we cannot know if there will be a material change for many if not most plans. I might agree that if your 2007 AFTAP is over 100%, perhaps you can chance this but in all other cases, is this a risk worth taking? Would it be better to simply no certify and do freeze notices for Title I Plans with a cover letter explaining that the freeze will likely be retroactively reversed? Opinions?

And let's not forget #6, AFTAPs and EOY Valuations!!!!

I agree with you. The result is 0 for the Ave Benefit Test and for the Rate Groups. Thanks to all. Now if only we could all solve the AFTAP dilemma!!

Good point all around. I do see the section in the a4 regs defining employee and that is clear and convincing to me. But, I did not see the language you refer to in 1.410(b)-5(d)(5). In (b)-5© it does refer to employees and to nonexcludable employees. Those terms are defined in 1.410(b)-9 in a way that differs from the a4 regs in that it does not require that an employee benefit under the plan in order to be considered for the rate group test. Do you think that could be interpreted to mean that such a non-benefitting non excluable employee could have an accrual rate for Average Benefit Test purposes while not having one for non discrimination rate group purposes?

I am running an average benefit test for a plan that is narrowly failing the 70% rate group coverage test. Although the plan is a DB, the same question would pertain to a DC plan. I am using the "accrued to date" measurement period for the average benefit accrual rates. I have a participant who terminated in the plan year with more than 500 hours but less than 1,000 hours and who had no benefit accrual for the year but did accrue benefits in the prior 3 plan years. Does he still have a normal benefit accrual rate using the "accrued to date" measurement period? I would think that he would because he does have an accrued benefit and he is not an excludable employee for the year being tested. Any thoughts pro or con?

Basically the choice is to wait until after the 12/31/08 plan asset values are in or, if you are quite sure that the plan would have a significant variable rate premium, consider using the $5 per participant squared approach and file using that amount as the variable rate premium. In that case, the forms can be filed as early as now. But, you'd better be sure about the variable rate premium otherwise you may be costing your client money in higher premiums.

Notice 2008-73 appears to say that by changing to BOY Valuation for 2008, you will NOT invalidate your reliance on 2008-21 which you used to produce your 2007 AFTAP (using 2006 val. and Sch B info). That's all it says as far as I can see. It then hints that we may be able to change val dates again in 2009 (back to EOY val, I assume) but does not come out and say it. I see this as fairly useless, unless you are committed to doing BOY Vals for 2008 and quoting costs based upon beginning of year data.

Thanks for the link. It was right on the topic and confirmed that an acquisition of an additional interest in the primary residence qualifies for the extended loan repayment period.

A participant purchased a primary residence with his girlfriend. They each owned 50%. Now, six months later, he is acquiring the other 50% from his girlfriend and wishes to borrow from the plan for that purpose. Does that qualify as a loan used to "acquire a primary residence" which would allow for a greater than 5 year repayment period?

Depending upon the terms of the PPA plan document will indicate whether the reinstatement is automatic or requires an amendment. At this point, I would guess you operate the plan in the way in which you would want the eventual PPA amendment to read. So, assume automatic reinstatement and then amend the plan in that manner some time before the end of 2009.

Do you think it is permissible in a Cash Balance Plan to only credit compensation as a participant (from date of entry) in the first year a participant joins the plan? Generally this rule is only used for DC Plans and is not used for DBs, especially final average pay DBs. But since a Cash Balance Plan accumulation formula is commonly based on each year's pay, and since the 401a4 Regs clearly allow testing comp to be done this way, I believe it should pass muster. Any thoughts?

I have not seen formal guidance on this but at a May ASPPA NY Benefit Council meeting, Sal Tripodi indicated that PPA language in the plan will provide basically 2 options: One option is that any benefit accrual that was limited due to 436 benefit restriction will be recredited automatically when such restriction is lifted. The other option is for the removal of the freeze to only be prospective in nature. Of course, the automatic re-crediting of previously frozen benefit accruals due to the lifting of 436 restrictions is considered an increasing plan amendment (if the benefit freeze period exceeds 12 months) and cannot be done until the AFTAP is at least 80%. Assuming these rules are correct, then I would guess that we could operate the plan now in accordance with the PPA language that will eventually be written by the end of the 2009 plan year.

What about 403b salary deferrals. Would they count in the average benefit test? Would it matter whether the 403b is an ERISA or non-ERISA plan?

Thank you Tom. So, it seems that it is permissible to exempt existing employees who have already made an election for the K Plan (even if that election was NOT to contribute) but it is NOT permissible to limit the automatic enrollment only to employees first eligible after the plan is amended to provide automatic enrollment. The plan would need to subject existing employees who have NOT made an election previously to automatic enrollment in order to be a EACA. Do you agree with that conclusion?

I have a plan that wants to institute automatic enrollment as of 1/1/2009. They only want it to apply to participants who first become eligible for the plan on or after 1/1/2009. I believe that such a provision would cause the plan to not be considered an "Eligible" Automatic Contribution Arrangement (EACA) and therefore the plan would not be allowed to offer permissible withdrawls within 90 days for new automatically enrolled participants. Am I correct in my assumption that an EACA can only exclude those participants who have made affirmative elections as to a contribution percentage (including 0%)? If so, am I correct in saying that a QACA could be limited only to new participants but it would then no longer be an EACA and would therefore not be able to allow the permissible withdrawls described above?

As to the ERISA status, what if employer has 2 403b plans. The ERISA plan is a new comp PS allocation with 401(a)(4) testing and the other plan is the non-ERISA employee deferral only plan. Do you believe that we would still test those deferrals in the average benefit test? You seem to believe that we generally test in the same manner as a qualified plan. While I agree with that overall concept, I'm still not totally sure that the employee deferrals (ERISA or non-ERISA) fall into the tests since those deferrals are NOT subject to non discrimination testing under the universal availability rulese.

I have an existing 403b plan with an Profit Sharing provision. We are amending the allocation provisions of the PS portion of the plan to become a New Comp (previously it was a straight proportion of pay PS). In running the a4 non discrimination tests, do I include the 403b deferrals in my average benefit test? Would the answer be any different if my 403b deferrals were in a Non ERISA 403b plan?