Pre-GATT mortality for lump sums

[John A]

If a plan has not yet adopted GATT and is computing a lump sum, is the mortality table used required to be 1) mortality defined in the plan document, 2) old PBGC adjusted UP84 mortality, or 3) can it be either 1) or 2)?

[david rigby]

The issue is a bit different.

I believe it is the "definitely determinable" requirement that requires plan to include a definition in the plan. It is IRC 417(e) that states a plan must use a minimum for its definition of a lump sum actuarial equivalent.

GATT has changed the definition of this minimum in 417(e). The plan must include this minimum its definition if it uses another method for determining actuarial equivalence.

For example, a plan could define the lump sum using the UP84 mortality table and 2% interest. In all likelihood this will produce a greater amount than the GATT minimum, but the plan's defintion (that is, the langauge in the document) will be inadequate unless it also includes the GATT defintion and specifies that will be a minimum.

[John A]

Prior to the existence of GATT, what mortality was required by 417(e) in determining the actuarial equivalent lump sum?

[Guest Steve C]

Prior to GATT, 417(e) addressed only the interest rate. No specific mortality basis was mandated. UP-84 was a common choice, but by no means was it the only choice.

[John A]

In other words, prior to GATT, any mortality was acceptable as long as the same q's were used for both males and females (no matter what table was used, it was applied on a unisex basis), correct?

[david rigby]

Of course, using a unisex mortality basis is required.

[Guest]

But an employer could "blend" the unisex table to reflect its retirement population. For example, an employer in the construction industry may have a singel table derived from a population of 95% males and 5% females. This was perfectly legitimate before GATT (and some would say it is still legitmate after GATT). The IRS, however, mandates a unisex table with a 50/50 male/female mix. This mandate can greatly increase the cost of lump sums for our hypothetical construction company (ignoring the impact, if any, of using an interest rate equal to 30-year treasuries) . . .

[mwyatt]

Actually, the pre-GATT mortality table used for IRC 417 purposes was the mortality table specified in your definition of actuarial equivalence. So if your plan defined actuarial equivalence to be the 1983 Individual Annuity Mortality Table (pre-retirement only) at 5.0% per annum (a fairly generous set of assumptions but common in small plans), the PBGC minimum would be calculated using the PBGC interest rates with the plan specified mortality table.

This of course pointed out the problem with REA in the first place. The PBGC rates are developed in conjunction with the UP84 tables and try to mimic the current market cost of providing an annuity. Of course, most insurance companies aren't using 25+ year old mortality tables in pricing annuity contracts. Therefore, in recent years the PBGC has had to "floor" their interest rate to somewhat match current market conditions. However, note that REA only had you use the interest component; if you had more liberal mortality assumptions in your plan document, the 417 minimum benefit really went off the charts. This is one of the primary reasons that the GATT rates tie the interest and mortality assumptions together.

[John A]

mwyatt, now I am back close to my original question. Are you saying that, pre-GATT, a plan could NOT use old PBGC adjusted UP84 mortality to calculate actuarial equivalent lump sums unless it was the mortality table specified in the document for actuarial equivalence? If the mortality table specified in the definition of actuarial equivalence was 83IAM, did the plan have to use 83IAM to calculate lump sums, or did it have to use the PBGC UP84 adjusted table, or could it use either 83IAM or the PBGC UP84 adjusted table, provided whatever was used was used consistently? If your answer is that 83 IAM was required, do you have a cite?

Thanks for your help.