For a plan that pays the full value of the accrued benefit on death at any time prior to actual retirement, how should the actuarial increase be calculated from age x to x+1 for a deferred retiree? It always seemed to me that the accumulation should be done using interest only, since there is no forfeiture on death. To accumulate with interest and mortality would be a double dip. But other people have recently told me otherwise, saying that the post-retirement actuarial equivalence factors include interest and mortality, so you have to accumulate with both because that's what the plan says.

I agree with both of you. I think.

I think it is a double dip if you include a mortality increase for a period where death does not bring about a forfeiture.

I think you have to follow the terms of the plan, and if it is drafted to give a double dip, then a double dip it shall be.

Gotta watch that 415 thingy, though.

Merlin, is your question about a normal small plan with no pre-retirement mortality and post-retirement mortality of whatever table? If so, I have never seen a document that specifically describes the calculation of the actuarial increase for post-retirement, rather it just states the terms "actuarial equivalent increase" or some other like term. That being said, with no forfeiture upon death, the mathematics behind the calculation are clear to me there is no mortality adjustment.

Of course as Mike mentions, you could have some strange doc language.

No strange language, Blinky. It's just as you describe. Mike, the 415 thingy is what led me to the conclusion in the first place. Thank you both for your answers.l

The 411 actuarial equivalence factors set forth in the document are determinative. It can and should be done either way depending on the doc. If the plan indicates a pre-retirement mortality discount, the PVAB is calculated at age 58 using both discounts. The same calculation at age 59 would therefore be greater for interest and mortality both.

I'm not an actuary, but since the benefit is expressed as a lifetime annuity (not as an accumulation), so that the life expectancy determines the term of payment, don't you have to take into account the mortality factor to determine the value of the benefit, whether payment occurs before or after retirement. Maybe the plan already takes that into account in how the lifetime benefit is expressed. Under common factors, there is a 1/15th reduction in the benefit for each year that payment begins before 65 (to 60) - another way to express that is that there is a 1/15th increase for every year between 60 and 65. That 1/15th increase primarily reflects the fact that the number of payments is decreased by one year (rather than the increased earnings) so the remaining payments should be greater to get to the same value.

The question really is...Before actual retirement but after NRA do I use pre-retirement or post retirement assumptions.

Absent something in the plan directly to the contrary I would argue for pre-retirement assumptions since, in this plan, there is no forfeiture at death until actual retirement.

Having said that, I have heard Jim Holland say that actuarial equivalence post-NRA requires a mortality increment. Most in the room disagreed. In your plan there is no forfeiture at death or benefit of survivorship until actual retirement

Recommended reading: comments above from Mike, Blinky, and vebaguru.