NAR and MVAR Calculation

[Guest PensionNW]

I only work on safe harbor DB plans but would like to get the calculation of the NAR and MVAR of 1.401(a)(4)-3(d) straight.

I found a posting from March of 2001 regarding the calculation of the NAR and MVAR if the normal form of the benefit were 10C&C.

In determining the NAR, the example indicates that the 10C&C accrued benefit payable at NRA is first converted to a life annuity using the plan actuarial equivalence factors and this would then be expressed as a dollar amount or a percentage of average annual compensation. Is this conversion from 10C&C to a life annuity correct and if so, why? It looks like section 1.401(a)(4)-3(d)(1)(i) indicates that the NAR involves the “increase in the employee’s accrued benefit (within the meaning of 411(a)(7)(A)(i)…” Section 411(a)(7)(A)(i) says that the accrued benefit is “expressed in the form of an annual benefit commencing at normal retirement age.” Does 411(a)(7)(A)(i) not consider a 10C&C benefit to be an “annual benefit”? If normal form were a 50% QJSA, would one convert this to a life annuity first in order to calculate the NAR?

In the posting from 2001 basically, the accrued benefit payable at normal retirement as a 10C&C benefit is converted to a QJSA benefit payable at the participant’s current age using plan equivalence and then this benefit is normalized. This looks right to me, kind of (see below). The posting indicates::

1. Using plan equivalence, calculate the lump sum equivalent of the 10C&C accrued benefit payable as of the participant’s current age.

2. Using plan equivalence, convert this lump sum to a 50% J&S benefit payable at the participant’s current age.

3. Using standard interest and mortality, calculate the lump sum equivalent of this 50% J&S accrued benefit payable as of the participant’s current age.

4. Standardize this lump sum by accumulating it to testing age using standard interest and then dividing by the APR determined using standard interest and mortality.

I still contend that if the QJSA under the plan is not subsidized and is determined from the normal form using plan equivalence that it is not more valuable than the normal form of the benefit and the MVAR would be the same as the NAR but the regulations don’t seem to agree with me, and they win. Does anyone else have an opinion on this?

[AndyH]

I was probably the one asking those questions. I find your question to be confusing. Your third paragraph seems to say convert to a SLA but your fourth and fifth seem to say convert to a QJSA.

Apparently in the third paragraph you are discussing the NAR and thereafter you are discussing the MVAR, where any subsidy might be captured. Right?

I just thought some clarification might help any responders.

Three years later, I've yet to find anything that conflicts with this method (as clarified) but there is clearly room for interpretation so I'd like to hear other views as well.

FWIW, I submitted one done exactly this way (it had a 10 C&C normal form) for a FDL and it was approved no questions asked.

[Guest PensionNW]

Hi Andy, thanks for responding to my posting. I believe I am responding to your original question posted March 21, 2001.

Yes, my third paragraph indicates that the when calculating the NAR the 10 C&C benefit is converted to a life annuity using plan equivalence factors at normal retirement age. This is what is shown in the March 23 response to your original posting. This calculation is labled "Normalized Life Annuity (LA) at Testing Age 65". My question is actually two fold; first, why is the 10C&C benefit converted to a life only annuity {this makes sense but I just want to know why because 1.401(a)(4)-3(d)(1)(i), Normal Accrual rate does not mentioin any conversion or normalization} and then, why is the conversion done using plan equivalence and not standard interest and mortality as is used to calculate a "normalized" benefit?

I apologize for not mentioning this (it was getting late) but yes, my calculations in the fourth and fifth paragraphs pertain to calculation of the MVAR. Here it appears that any optional form of benefit is first converted to a QJSA using plan equivalence and then this QJSA is normalized to NRA using standard interest and mortality.

Once again, it seems to me that if the optional forms are not subsidized but are calculated as being actuarially equivalent to the normal form, then the optional forms are not more valuable than the normal form and the NAR and the MVAR should be the same. But, I think the mechanics of calculating the MVAR take a different approach and mandate normalization of optional forms regardless of if they are subsidized or not.

[AndyH]

Good questions.

Regarding the NAR calculation

Look at (d)(2)(i) Consistency requirement "..If plan benefits are not expressed as straight life annuities .......they must be normalized."

Regarding the MVAR calculation and the need to convert to J&S, others here have expressed the same reaction, and I happen to agree, that it does not seem to make a lot of sense to convert to J&S if there is no subsidy. The answer seems to be that you need to do it that way because the regs say you need to, not because it makes any particular sense.

If you were to keep searching, there has been some discussion of this exact question. One person said he/she was not going to do it that way to which others responded that to do so would be at their peril. Regardless of whether or not you think it makes sense, it does seem that you must do it that way.

[Guest merlin]

This is a related question that keeps rearing its ugly head. For a plan that pays lump sums, do you have to recognize the 417e rates in computing the MVAR? A few years ago the answer according to Larry Deutsch and Carol Sears was "yes". In his DB/DC Combo Plans outline for this year's ASPPA conference, Kurt Piper says "no". Where are we?

[AndyH]

Merlin, I thought that issue was resolved conclusively. He waffles considerably in the outline. He says that IRS employees have said no, but that newer employees might say yes. It seems to me that this is unhelpful. I was not at his session or I would have questioned him on it. Did/were you?

There are other matters in that outline that contradict discussions that have taken place here as well.

[Guest pension222]

I have come to the same conclusion that although the equivalent QJSA is really not more valuable than the normal form, all optional forms need to be converted to a QJSA and then normalized in order to calculate the MVAR. It doesn't make any sense but is required by the regulations and is just a mechanics.

Regarding the 417 subsidized lump sum. I've come to think of it this way. 1.401(a)(4)-20 Q&A 16 says that in the case of a married participant, the QJSA must be the most valuable benefit.

Now if the lump sum is subsidized by 417, I don't believe that we need to then take the value of this subsidized lump sum and calculate an equivalent QJSA. The QJSA is equivalent to the normal form of the benefit.

If we do not need to back into a new QJSA based on the 417 subsidized lump sum, then we should not need to consider the 417 subsidized lump sum in the calculation of the MVAR.

[AndyH]

Thanks for your comments, pension222. I have spent the last few years collecting opinions on this issue, and this is the first such opinion I have heard.

Does anyone else out there still think that the 417(e) lump sum need not be considered in the MVAR calculation?

[ak2ary]

My recollection is that you would use the plan's assumptions to determine the benefit payable as a lump sum without consideration of 417(e), and then convert to a QJSA and normalize that benefit. Thus if a plan provides for a lump sum calculates at 83 IAM 6%, you would ignore the 417(e) rate in determining the most valuable benefit, but if the plan only used 417(e) for determining the lump sum, the 417(e) is the plans assumption set, and has to be recognized in the MVAR

[AndyH]

Merlin, now look what you've done!

I know, it's actually Mr. Piper who caused this to resurface.

How about an email to LD who is the one that originally took ak2ary's approach but two years ago said Jim Holland told him he was wrong; that the 417(e) lump sum must be in the MVAR?

[Guest merlin]

I've email'd the question to Larry.

[Guest merlin]

Larry's answer:

"Gee, I don't remember saying that you should reflect the 417(e) rates, especially since that is not how it is done. I did do a session with Newell Kimlin of the IRS, where he expressed that opinion, but has since acknowledged that he is not sure"

It is clear that the MVAR is determined by reference to the benefit payable in the form of the QJSA. If the plan does not reflect the 417(e) rates in this calculation, then the 417(e) rates do not impact the MVAR.

On the other hand, the regulations require that the QJSA benefit be the most valuable benefit form. There have been ongoing discussions, both inside and with the IRS about what to do if the lump sum is the most valuable form. There is an argument (and Jim has suggested that it might be required) that if the assumptions for determining lump sums are more favorable than the assumptions used to determine the QJSA, then the lump sum is the most valuable benefit form.

This leads to the conclusion that by default the QJSA benefit must be at least as large as the actuarial equivalent using the 417(e) rates (assuming that any benefit form under the planis subject to the 417(e) rates), even though the regulations specifically state that this is not the case. This issue came to a head with the issuance of the relative value regulations, when it became clear that under any set of "reasonable assumptions, where the interest rate was higher than the 417(e) rate, the lump sum would be the most valuable benefit. The current state of affairs is that the IRS and treasury are rethinking the problem".

Just as clear as it never was. BTW, i also put the question to Newell Kimlin. No response yet.

[AndyH]

Thanks, Merlin. Good thing for those tapes (I don't have them but their existence reassures me). From past discussions, I believe that you and I both heard him say no at ASPA 2001 and then reverse himself in ASPA 2002. In 2002, I think we were both in the "white freezer room" when LD went so far as to suggest avoiding including 417(e) in the MVAR by designing the plan to have no lump sum over a certain (low) limit until the owner goes to retire. And he admitted he was reversing his position because JH told him he was wrong.

And I believe you may have been in the 2001 General Testing session done by Carol Sears. Someone asked about this and she said yes and several people, myself included, responded by saying that such answer conflicted with comments LD had just made and she essentially said "so be it".

Why can't we hear a direct IRS opinion on something as basic as this?

p.s. I don't know who Newel Kimlin is, and I'm certain he wasn't a speaker in any of the sessions to which I refer.

[Blinky the 3-eyed Fish]

Newell is an IRS actuary in California. He is also the leading hater of certain DB offset arrangements. He often speaks at the LA Benefits Conference but I guess doesn't venture east.

Logically IMHO, it seems silly to test under 417(e) rates as they have expired once the testing can be run for all those who did not receive a distribution.

[AndyH]

What about a lump sum caused by a fixed rate in the plan? I don't see the argument against that.

[Blinky the 3-eyed Fish]

And neither do I. (We should hug now.)

[Guest merlin]

Kimlin agrees that this is an area over which there is some disagreement. His opinion is that the MVAR should be calc'd "...from the optional form that results in the largest accrual rate, including a single sum distribution. However, most general test submissions (Demo 6) that I have seen calculate the most valuable accrual rates based on the joint and survivor optional form of payment and without regard to the single-sum distribution option".

[Guest Harry O]

I have ignored the lump sum subsidy currently inherent in 417(e) and make sure to disclose it in the 5300 filing. Never heard a peep from the IRS . . . so far.

[Blinky the 3-eyed Fish]

Andy, I would like to see if we are on the same page. Can you prep a MVAR calculation not factoring in 417(e) or any special subsidies, but assuming the plan offers lump sums. I am hoping it is easier if you pick the figures to use. Thanks.

[AndyH]

Blinky, I use the method outlined in detail by Doug Goelz in a 2001 post that pensionnw references. Maybe somebody can link to it.

In addition, ASPA's C-4 study guide had a detailed example prepared by either Joan Gucciardi or Norman Levinrad (co-authors) which I learned from, except that the edition I was using had an error in it which Joan confirmed. That was a few years ago and I have not checked recent editions to see if it was corrected.

I'll dig out a sample calc when I get a minute. (Jury duty is over!)

[AndyH]

Here is the link:

http://benefitslink.com/boards/index.php?s...ic=9479&hl=mvar

[AndyH]

For those who frequent short attention span theater, the MVAR using 417(e) for a plan with no subsidized early retirement benefit would be the actual lump sum which would be paid by using the increase in the accrued benefit during the measurement period divided by the plan's J&S annuity rate at the current age, multiplied by the J&S annuity rate at the current age using testing assumptions, then projected to testing age at the testing interest rate, divided by the life annuity APR at testing age using testing assumptions. Then divide that by average pay and by testing service.

[Blinky the 3-eyed Fish]

Congrats on the hat trick, three posts in a row! We are on the same page.

[WDIK]

Congrats on the hat trick, three posts in a row!

Not to diminish AndyH's accomplishment, but it pales in comparison to the feat of one member who not only made three posts in a row in the same thread, but the three posts were also consecutive for the entire forum, on a weekday, during business hours.