Consider the situation where settlement accounting is required when total lump sums for a year exceeds the sum of the service cost and interest cost for that year. When is the settlement recognized?

I believe that in this situation it is quite common to reflect the settlement at the end of the year regardless of when the lump sums were actually paid. In normal years, the settlement gain or loss at year-end would not differ substantially from a more precise method. However, for 2008, an X% recognition of the accumulated loss at any point in the year would probably be better than the same X% recognition at 12/31/2008.

Thoughts?

Two thoughts:

(1) You will never (not even when Tuesday falls on Thursday) obtain guidance from the Company auditor.

(2) Since it's likely no one will understand the computation, year-to-year consistency of methodology is more germane than actual method. So, it's recommended not to change the methodology used previously if this issue has ever arisen. From a strict perspective, changing methodology is an accounting procedure change that should be disclosed.

Respectfully, I disagree with (1). I have recognized settlements at different points in time, and the auditor always signed off (although I may have made the suggestion first).

Absolutely. Auditor's sign off. They simply will not advise what is appropriate. At least that has been my experience.

Thanks for the replies.

Regarding the recognition during the year, assume a plan had 6 lump sums and the fourth one (made 5/31) puts the YTD total lump sums over SC + IC. Also assume the other 2 lump sums were made 8/31 and 10/31. Would I have 3 settlements (5/31, 8/31 and 10/31)? Obviously this could be more complicated with more lump sums, which is why I think most practitioners recognize 1 settlement at the end of the year.

I agree that consistency from year-to-year is important.

It may be reasonable to assume that most acturaries, most plan sponsors and most auditors will prefer the simplest method. But it also depends on how significant interim (quarterly) reporting is, and even whether or not there are other significant events included in such interim reporting. Thus, put the alternatives in front of the sponsor and auditor, and let them decide.

I would agree. I've gone so far as to calculate the first settlement in the quarter the threshhold was exceeded and then additional settlement pieces each quarter thereafter. But remember, this means a new measurement date each time and new assumptions and expense from that point forward each time. Not fun. I would definitely discuss with the sponsor in general terms as this approach yields significant consulting expenses.

They simply will not advise what is appropriate. At least that has been my experience. It depends on how significant interim (quarterly) reporting is, and even whether or not there are other significant events included in such interim reporting
regards,
barbie

_______________
[edited to remove spam link]