Accured Benefit on Plan effective date

[flosfur]

A plan's benefits are based on Service using average comp during "Participation".

Since avg comp on the plan's effective date is zero, would it be correct (or reasonable) to say that on the plan's effective date the accrued benefits, PVABs and the Current Liability etc are all zero?

[david rigby]

Absent other provisions that might define an alternative formula, that would be my interpretation of your description. However, close inspection of all terms of the plan document is always appropriate.

[Guest DFerrare]

Is the benefit service-based but accrued over years of participation?

[AndyH]

If your plan provides benefits based upon service but excludes comp prior to participation, it is my understanding that such an exclusion is not permitted until such person has pay at least equal to the averaging period. But maybe I assume too much? More detail would help.

[Guest flogger]

It certainly common to have a plan state that average comp is determined over 3 yrs (for ex) of Participation, and if the Participant has less than 3 yrs, then avg comp is determined over that period of time. Therefore, the avg comp at the date of participation in that case would be the rate of pay in effect at that time. If past service is credited, then a PVAB would exist.

[AndyH]

I'm not sure if we're saying the same thing or not. I don't think that a plan that is a safe harbor can be written as flosfur's first sentence states, unless it is a career average plan.

Example, a plan can define average comp as high 3 starting at DOP provided that average comp until the participant has 3 years of participation is averaged from date of hire.

[flosfur]

Is the benefit service-based but accrued over years of participation?

Both the Projected and the Accrued benefits are based on service.

[flosfur]

If your plan provides benefits based upon service but excludes comp prior to participation, it is my understanding that such an exclusion is not permitted until such person has pay at least equal to the averaging period.  But maybe I assume too much?  More detail would help.

Andy, all the prototype (standard & non-standard) and volume submitter documents I have ever seen and worked with don't prohibit using Service for benefit formula & Avg Comp during participation or alternatively using Partipation for benefit formula & Avg Comp during Employment.

You can't get anymore Safe-Harbor than a Standarized Prototype document provisions?

[AndyH]

I would urge you to re-read them with this interpretation in mind. I agree that you are allowed to use average compensation only based upon participation, but only once the averaging period has expired.

Have one with some language you could post? I'll look also.

[AndyH]

From the ERISA Outline Book 2004 edition chapter 9 part A.6 (page 9.93)

"Compensation History taken into account ..........

"The compensation history instead may begin on a date which is later than the employee’s original employment date, such as the effective date of the plan or the date the employee’s participation in the plan commences, so long as the compensation does not end earlier than the current plan year. However, if the compensation history taken into account is for a period which is less than the stated averaging period, compensation before the start of the designated compensation history period is taken into account to the extent needed to equal the averaging period. If the employee’s entire period of employment is less than the stated averaging period, then average compensation is determined over that entire period of employment. These rules are found in Treas. Reg. §1.401(a)(4)-3(e)(2)(i)."

[flosfur]

....Therefore, the avg comp at the date of participation in that case would be the rate of pay in effect at that time.  If past service is credited, then a PVAB would exist.

An observation - at the plan effective date, YOP =0, therfore if there was no prior DB plan, $Max=0 for all participants. But they may have S415 deminimis ($10k proated for service) depending on whether or not the sponsor has (or had) a DC plan.

So PVAB @ plan effective date would pretty limited.

[AndyH]

I don't agree. The numerator of the fraction always equals at least one, so 1/10th of the full limit is always available.

[Blinky the 3-eyed Fish]

Andy is right.

[mwyatt]

Interesting. Your accrued benefit is based on YoS, including service prior to the effective date of the Plan, but you are basing your AMC solely on compensation earned as a participant. One way to look at this:

12:01 AM 1/1/2004: 5 Years of Service at x% of AMC, but I haven't earned anything yet.

12:01 AM 1/2/2004: Guess what, I got paid for New Year's Day. Looking at Relius VS language where you code for participation, AMC covers from date of participation through date of termination (determination).

I guess we can get the angels out to dance on the head of our pin here, but I would have a real hard time explaining to a layman who has 5 YoS going in why he has a $0 accrued benefit starting out...

[AndyH]

Mike, that contradicts my point. Can you upload that language? I would think that a careful reading would support my point. No?

[mwyatt]

Hey Andy:

I ran a dummy DB case on Relius Autodoc this morning. Here are the results of the definition of Average Monthly Compensation:

First, AMC language with Accrued Benefit coded for fractional ratio on YoP, NRB based on YoP:

"Average Monthly Compensation" means the monthly Compensation of a Participant averaged over the three (3) consecutive Plan Years from date of participation which produce the highest monthly average. If a Participant has less than three (3) consecutive Plan Years of service from date of participation to date of termination, the Participant's Average Monthly Compensation will be based on the Participant's monthly Compensation during the Participant's months of service from date of participation to date of termination. Compensation subsequent to termination of participation pursuant to Section 3.4 shall not be recognized.

Now, rerun with Accrued Benefit, NRB based on service:

"Average Monthly Compensation" means the monthly Compensation of a Participant averaged over the three (3) consecutive Plan Years from date of participation which produce the highest monthly average. If a Participant has less than three (3) consecutive Plan Years of service from date of participation to date of termination, the Participant's Average Monthly Compensation will be based on the Participant's monthly Compensation during the Participant's months of service from date of participation to date of termination. Compensation subsequent to termination of participation pursuant to Section 3.4 shall not be recognized.

Note that both definitions are identical in the basis for <3 YoP is from entry to determination date. Of course, again at the end of the first day of entry, have a basis to determine compensation average (on 1-day of pay). So it goes back to my original point: may try to say that despite granting pre-entry service, the argument that this is rendered invalid as of the first day of entry since your comp definition is zero is fairly shaky. Looking at this language, at the end of the first day you do indeed have a non-zero basis for average, and hence a non-zero accrued benefit.

Think that this may be a glitch in Corbel's language. I would have thought that their def would have changed for <3 YoP if the benefit YoS basis changed to eliminate this conundrum, but no.

[AndyH]

Thanks for the post, Mike. Either the EOB or Corbel appears to be wrong. I think it's the Corbel language.

[mwyatt]

No problem, Andy. Looks like this is one of those situations that noone really thought about too much. However, I'd still be hard-pressed to explain the logic in saying the benefits are $0 at entry date given the past service grant solely because of your average compensation definition excludes comp prior to entry.

[flosfur]

I don't agree.  The numerator of the fraction always equals at least one,  so 1/10th of the full limit is always available.

If the fraction at point 0 (plan effective date, with no participation) is 1/10th, then is the fraction 2/10th at the end of year 1?

[AndyH]

No, if you look at 415(b), the fraction is always at least 1/10 in the first year for both the service and participation limits. It is 1/10 at January 1 and December 31.

[mwyatt]

That exception at entry date/effective date for IRC 415 has always been there (otherwise, every DB plan granting past service for accrual has been in violation under ERISA).

[flosfur]

Hey Andy:

I ran a dummy DB case on Relius Autodoc this morning. Here are the results of the definition of Average Monthly Compensation:

First, AMC language with Accrued Benefit coded for fractional ratio on YoP, NRB based on YoP:

"Average Monthly Compensation" means the monthly Compensation of a Participant averaged over the three (3) consecutive Plan Years from date of participation which produce the highest monthly average. If a Participant has less than three (3) consecutive Plan Years of service from date of participation to date of termination, the Participant's Average Monthly Compensation will be based on the Participant's monthly Compensation during the Participant's months of service from date of participation to date of termination. Compensation subsequent to termination of participation pursuant to Section 3.4 shall not be recognized.

Now, rerun with Accrued Benefit, NRB based on service:

"Average Monthly Compensation" means the monthly Compensation of a Participant averaged over the three (3) consecutive Plan Years from date of participation which produce the highest monthly average. If a Participant has less than three (3) consecutive Plan Years of service from date of participation to date of termination, the Participant's Average Monthly Compensation will be based on the Participant's monthly Compensation during the Participant's months of service from date of participation to date of termination. Compensation subsequent to termination of participation pursuant to Section 3.4 shall not be recognized.

Note that both definitions are identical in the basis for <3 YoP is from entry to determination date. Of course, again at the end of the first day of entry, have a basis to determine compensation average (on 1-day of pay). So it goes back to my original point: may try to say that despite granting pre-entry service, the argument that this is rendered invalid as of the first day of entry since your comp definition is zero is fairly shaky. Looking at this language, at the end of the first day you do indeed have a non-zero basis for average, and hence a non-zero accrued benefit.

Think that this may be a glitch in Corbel's language. I would have thought that their def would have changed for <3 YoP if the benefit YoS basis changed to eliminate this conundrum, but no.

To see more examples of this, see prototype adoption agreements of any one of the following: Datair, PenDocs, The New England Financial & MassMutual.

I don't have these on the computer so I cannot post it here without typing in the Questions & the answer boxes available for all of them. But as an example, here is what the Datair's "Standardized" (non-Integrated) Prototype has:

Part II

C 4. Average Compensation - S1.2.7 provides that

a. Average Annual Comp

b. Average Monthly Comp

is based upon the comp which yields the highest average received during:

c. Plan Provisions - the 3 Consecutive Years of Service while a Participant.

d. any ____ Consecutive years (minimum of 3)

e. ______ years (min.. 3 ) of the last ___ years.

f. the final ____ years (min.. 3)

g. all years (career average).

h. annual (current) comp (applies only to a accumulation plans).

C 5. Counting of Average Annual (or Monthly) Comp - Include compensation received during:

a. Years odf Service

b. Years of Service while a participant.

Exclude compensation received: (Select all applicable.)

choices are: In the year: NRA is attained, NRA is attined and the ____ years prior to NRA, in which participant fails to earn a Year of Credited Service, break in svc, of participant's termination and so on.

D 8. Limits on Credited Service: ...

a. include all years with the employer

b. include years while a participant.

c. disregard years prior to __/___/___

d. ... disregard years in excess of ____.

Note: Must exclude years prior to fives years before the effective date of the benefit formula and cannot limit the number of years less than 25.

Note: (relates to 412(i) plans)

..............

_________________________________________

The less than 36 months (or any other averaging period) issue is addressed in the Basic document.

I do not see any Warning or Caution about using the same service for benefit service credits & comp averaging.

[flosfur]

Sorry folks - I was going to shorten the Quote to reduce the clutter but I clicked the Add Reply instead of the Preview Post button.

[flosfur]

............  However, I'd still be hard-pressed to explain the logic in saying the benefits are $0 at entry date given the past service grant solely because of your average compensation definition excludes comp prior to entry.

For what its worth ...

As it happens, I just came across 2 first year valuations prepared by two different actuaries using the ASC and PenTab (now Corbel) valuation softwares. Both plans credit past service for both the projected and the accrued benefits. For both, the accrued RPA & OBRA liabiliy is zero and the benefits accruing during the year are exactly equal to the Accrued @ EOY (both are EOY valuations).

I have seen the same results using the Datair system.

[mwyatt]

Hey Flosfur:

I am an ASC user, and I will caution you about the GIGO phenomenom (Garbage In, Gospel Out). Their system to date still hasn't been recoded to reflect the aforementioned minimum 1 numerator in the 415 limit. If you want, try running it with a real past service benefit (service, comp, etc.) without closely examining the default 415 assumptions in the Plan Specs screen and you'll see what I mean...

One other thing, after observing that you are referencing EOY first year valuations, is that both systems don't calculate the beginning of year accrued benefit; rather they look to the prior year accrued benefit field (which, surprise, surprise, is $0 being this is the first year of the plan). Don't take everything a computer tells you for granted.

[AndyH]

Thanks for the comments, Flosfur. The discussion is very useful, but multiple wrongs don't make it right. I would be more convinced of the errors of these documents if I saw the underlying prototype. That is where I would expect a "notwithstanding _____". And, yes, I recently took over a plan that used only participation comp in the first and second years. But I am convinced that it was just plain wrong. I think this is just one of those things that is just not commonly known.

Anyone out there able or willing to go back and check the LRMs? Problem is, I don't know when they would have been issued on this subject.