Funding Target and Target Normal Cost in Frozen DB Plan

[JBones]

1) If a plan's benefit accrual formula has been frozen, but the amendment was not a hard freeze, and benefits increase due to an increase in average compensation, should those increases be funded through the target normal cost or do they get lumped into the funding target?

2) If a plan is frozen but a participant continues to be a active past normal retirement age and their benefit increases due to actuarial adjustments for late payment, does that increase end up in TNC or FT?

[Andy the Actuary]

1) If a plan's benefit accrual formula has been frozen, but the amendment was not a hard freeze, and benefits increase due to an increase in average compensation, should those increases be funded through the target normal cost or do they get lumped into the funding target?

2) If a plan is frozen but a participant continues to be a active past normal retirement age and their benefit increases due to actuarial adjustments for late payment, does that increase end up in TNC or FT?

1) I would vote TNC because were looking at the difference between the benefit at the end of the year in excess of the benefit at the beginning of the year.

2) If actuarial assumption is that those who are past NRA retire on the valuation date, then the increase would go into next year's FT. If the actuarial assumption is that those who are past NRA on the valuation date retire in a year (or in x years), then the increase would go into the TNC in accordance with (1).

[SoCalActuary]

1) If a plan's benefit accrual formula has been frozen, but the amendment was not a hard freeze, and benefits increase due to an increase in average compensation, should those increases be funded through the target normal cost or do they get lumped into the funding target?

2) If a plan is frozen but a participant continues to be a active past normal retirement age and their benefit increases due to actuarial adjustments for late payment, does that increase end up in TNC or FT?

1) I would vote TNC because were looking at the difference between the benefit at the end of the year in excess of the benefit at the beginning of the year.

2) If actuarial assumption is that those who are past NRA retire on the valuation date, then the increase would go into next year's FT. If the actuarial assumption is that those who are past NRA on the valuation date retire in a year (or in x years), then the increase would go into the TNC in accordance with (1).

I disagree over point #2. You had an experience change because your prior valuation assumed that retirement would occur, but in fact it did not. Only the current accrual which exceeded the actuarial equivalent of the prior year benefit is considered a new benefit accrual subject to TNC. The FT is the value on the valuation date of the prior year accrued benefit as adjusted to the new assumed retirement age.

[Andy the Actuary]

I will stand pat.

I am now age 66 and NRA is 65. The valuation assumption, for example, is those past NRA are assumed to retire one year later. It could be argued that the actuarial increase from 66 to 67 is in fact accrual because the Plan could so provide for a suspension of benefits, give the appropriate notice, and then there would not be any increase.

I doubt anyone will ever go to pension prison on this point.

[david rigby]

I agree with Andy. Both times.

[Mike Preston]

And you would be 1/2 right, David. Or some reasonable approximation thereto.

Long message to follow (sometime in 2011 if you can wait that long).

However, since no additional service or compensation is required to have the EOY benefit completely determinable, you will never convince me that there is anything other than Funding Target at play here.

[FAPInJax]

No additional accrual of benefits (just a difference in age of valuation). 1,000 at 65 is worth 1075 at 66.

I will take my license plate ACT2-AIRE <GGG>

[Effen]

Sorry Andy, but I think I'm going with SoCal on issue Q2.

Mabye I will just look at all of the guidance we have so I can know for sure....oh ya, we have none.

[Andy the Actuary]

Well, if you don't include it as accrual in your TNC, then you have an actuarial loss because your FT increases for more than just passage of time. How can you have an actuarial loss when your assumption (retire one year later) is met?

[Mike Preston]

I think if you work through the math you will not see a loss. The issue is one of assumed retirement date. You can only have one associated with a given portion of the liability in any one valuation.

[Andy the Actuary]

I think if you work through the math you will not see a loss. The issue is one of assumed retirement date. You can only have one associated with a given portion of the liability in any one valuation.

Suppose there is no TNC. Assume FT evaluated at 6% for current and next year and payment form is 20 years certain. Monthly pension is $100 at x. a120=91.17.

PV_x = $100 x 91.17=9,117. If no actuarial increase, PV_x+1=9,117 as well. However, benefit is actuarially increased and new payment is 100 x 1.06 = $106. So, PV_x+1=106 x 91.17 = 9,664.

Are you suggesting 9,664 - 9,117=547 is not a loss? If so, I'd appreciate your commenting on the flaw in this analysis?

[My 2 cents]

Wouldn't the normal approach be a straight life annuity? Say annuity value at age x = 100 and at age x+1 = 98.

current benefit = $100. With actuarial increase for a year, $108.16.

value now of immediate benefit = 100*100 = 10,000

value now of one year deferred benefit = 108.16 * 98/1.06 = $10,000

No gain or loss (OK, I did massage the factors a bit, but they are reasonably representative of what you might see)

Wouldn't a 20 year certain only form without an actuarial increase be a bit, shall we say, unfair? Plus, when would you ever see that as the normal form?

[Andy the Actuary]

Wouldn't the normal approach be a straight life annuity? Say annuity value at age x = 100 and at age x+1 = 98.

current benefit = $100. With actuarial increase for a year, $108.16.

value now of immediate benefit = 100*100 = 10,000

value now of one year deferred benefit = 108.16 * 98/1.06 = $10,000

No gain or loss (OK, I did massage the factors a bit, but they are reasonably representative of what you might see)

Wouldn't a 20 year certain only form without an actuarial increase be a bit, shall we say, unfair? Plus, when would you ever see that as the normal form?

The example was for the sake of illustrative ease. In your example, we would have:

PV_x=100x100=10,000

PV_x+1=100x98=9,800

PV_x+1 of increased benefit=108.16 x 98 = 10,600

What is 10,600 - 9,800 = 800?

[SoCalActuary]

Wouldn't the normal approach be a straight life annuity? Say annuity value at age x = 100 and at age x+1 = 98.

current benefit = $100. With actuarial increase for a year, $108.16.

value now of immediate benefit = 100*100 = 10,000

value now of one year deferred benefit = 108.16 * 98/1.06 = $10,000

No gain or loss (OK, I did massage the factors a bit, but they are reasonably representative of what you might see)

Wouldn't a 20 year certain only form without an actuarial increase be a bit, shall we say, unfair? Plus, when would you ever see that as the normal form?

The example was for the sake of illustrative ease. In your example, we would have:

PV_x=100x100=10,000

PV_x+1=100x98=9,800

PV_x+1 of increased benefit=108.16 x 98 = 10,600

What is 10,600 - 9,800 = 800?

Unless you have special provisions, $9800 is irrelevant.

PV is $10,000 and one year later it is $10,600. The $100 benefit payable at x is identical in value to the 108.16 payable at x + 1, because you did not receive the payments.

Andy, your example twists up the normal perception of late retirement equivalence.

Is it because of your perception that Assumed retirement age is x + 1,

and by delay in retirement the next valuation assumes payment starts at x+2?

I have a participant eligible at age x, and I assume they will retire at x + 1.

The assumed benefit of $100 at x must be converted into 108.16 at x + 1.

I grant the exception where someone is at their 415 limit,

in which case you will have a gain when they don't take the required payments.

If I assume at x + 1 that the $108.16 is payable at x + 1, then my assumptions are met.

You are implying that valuation at x + 1 assumes payment at x + 2.

So you should be valuing the equivalent of the $108.16 (roughly $117) payable at x + 2.

[My 2 cents]

If you are assuming that the person will commence benefit receipt in one year, then your current valuation will take into account one year's worth of actuarial increases if promised under the plan and will not do so if the plan pays no actuarial increases.

Say the plan provides an actuarial increase. This year's value will equal the value, discounted back for one year, of the anticipated benefit inclusive of the actuarial increase that would be in effect if your one-year-deferral assumption is met. So this year's value might be (to use my example of yesterday) equal to $108.16 [the amount anticipated to be payable if the person retires as assumed] X 98 [the anticipated value per dollar of benefit at assumed retirement date] / 1.06 [discounting back from assumed retirement date to present] = $10,000. Next year's value, if the person does retire, will be $108.16 [actual benefit amount, equal to actuarial equivalent of $100 payable now] X 98 [value as of valuation date at next year's valuation of $1 of immediate benefit] = $10,600. Nothing having been paid in between last year's valuation date and the current valuation date [relative to next year], the expected liability would be last year's $10,000 brought forwards one year with interest, or $10,600. Ergo, no gain or loss.

You get the same sort of result if the plan provides no actuarial increase. You would anticipate retirement next year with a benefit then of $100 (instead of the above example's $108.16). When you perform next year's valuation, you would then value the immediate $100 benefit, and observe no gain or loss.

So long as your valuation correctly takes into account whether the delayed commencement will entail an actuarial increase, you would not have any gain or loss.

[Andy the Actuary]

Okay, okay. The gist of all the comments is that the actuarial increase is part of the accrued benefit and therefore should be attributed to the funding target. Whether or not agreed, RIP.