Andy the Actuary

No question -- the PBGC did it foist !

I agree with you Mike that it's no big deal. The only difference is now we could have different AFTAP's floating around. E.g., that which is certified and that with appears in the Annual Funding Notice.

Thank you but real question is not answered by your response and I apologize for not asking more specifically: If you did certify in May, would you have to compute a preliminary average asset value that necessarily would differ from the average asset value used to determine the minimum contribution?

Would appreciate comments on whether or not I'm getting this correct: No credit balances are maintained. A client plans to contribute the last piece of their 2009 calendar year plan contribution - $100,000 - on 9/15/2010 The plan sponsor (i.e., plan administrator) has elected to determine the actuarial value of assets as average value. So, for purposes of computing the 2010 minimum contribution (officially after 9/15/2010), we determine the fair market value of assets on 1/1/2010 by including the discounted value of the $100,000, which in turn is used to determine the average value. However, suppose I will certify the 2010 AFTAP in May 2010. We cannot include contributions that have not been made as of the certification date. Consequently, for purposes of calculating the AFTAP, I must not include the (discounted value of the ) $100,000 in determining the fair market value that is used to determine the average value. Consequently, I must determine two average asset values as of 1/1/2010 !!! Worse, the AFTAP determined in May will be 79%. If I am requested (and I will be) to recalculate the AFTAP in September after the plan sponsor makes the $100,000 contribution, the AFTAP will be 81%. The 2009 AFTAP was 73%. So, lump sums will be restricted to 50% effective April 1 through September whenever and thereafter unrestricted (other than for HCEs). Further, (unless I plan to croak before October 1) there is no purpose in certifying the AFTAP in May since it is already presumed to be less than 80%.

My favorite question. In a standard plan termination, not only must all rights and features be preserved but also (assuming the employer is ongoing), the annuity contract must continue the same plan provisions (i.e., allow for lump sum) or if a lump sum is offered, it must assume that participant will grow into any subsidies or rights he would have had if the plan had continued. Thus, the cost of plan termination may surprisingly be much greater than the FT as the lump sum would assume the subsidy whereby the participant might otherwise leave before satisfying the conditions for the subsidy. See rev rule 85-6, which while it addresses early retirement subsidies, the line of thinking would be believed to apply to your situation. IRS_Rev._Rul._85_6_Early_Retirement_Subsidy_Termination.PDF

CAUTION: Avoid at all costs doing what makes sense.

Ah, another question in the "I thought PPA was supposed to fix pension funding?" category. No where in PPA does it state, "Notwithstanding the regulations, actuaries should follow the standard code of conduct and value benefits properly." In short, I believe your FT is $1.6 million. The position of $1.75 million implies 417(e) is relevant when you said (and know) it isn't. You can always advise the client that there is room to contribute and deduct up to whatever you believe the lump sum will actually be in conjunction with how much the client wants to contribute and the market value of assets (where you've directed current plan assets to be diversified among several mattresses).

Assume Plan that covers about 150 participants provides lump sum as greater of PPA and 71GA/PBGC. A former HCE is age 60 and can retire at age 62 with unreduced benefit. As of 1/1/2009, AFTAP was 73%, which included assumption (right or wrong) that employee would elect lump sum at age 62. Given the near zero probability that the Plan will be 110% funded in two years, the HCEs benefit will likely be distributed in an annuity form. Question: Should the FT be valued assuming a lump sum will be paid or should it more realistically assume a monthly annuity form of distribution?

Could you not issue a range certification?

Sounds like a forfeiture of accrued benefits to me.

Thank you. This is valuable and helpful information.

Question: There are numerous posts such as "We made a call in to the PBGC." Would you please discuss what "a call" entailed. How many calls? How did you finally get to the "right" person? Did you have to disclose who your client was to get a response in writing? About how long did this entire process take in terms of weeks/months, back and forth communications?

I get into arguments with friends about how much bigger a 16 in pizza is than a 12 inch. They say 33%; I argue 78%, because after all pi are squared and not round.

Do not worry about me. I'm just as sane as the day I applauded as I watched the pigs eat my little sister.

A friend told me he would pick us up at 6:43PM to go to dinner. He pulled this number out of the air because I commonly state weird pick-up times ("I'm a round number iconoclast"). 643 has somewhat interesting properties, I observed. 643 is a prime number. 13 = 6 + 4 +3 is a prime number 7 = 6 + 4 -3 is a prime number 5 = 6 - 4 + 3 is a prime number 67 = 64 + 3 is a prime number 61 = 64 - 3 is a prime number 37 = -6 + 43 is a prime number There may be other prime numbers that can be formed by the digits 6, 4, and 3 while respecting the order but 6:43 was approaching so I had to put my pencil down.

For sake of illustration, assume 6 (n items) identical items, x, going into 4 (r boxes) boxes. Let's assume no vacant boxes for the time being. Then, we look at the number of intervals as x|X|X|X|X|X. So, there are 5 intervals, or in the general sense, n-1 intervals. Now, let's just look at the objects partitioned into one arrangement into the four boxes: xxx|x|x|x. Note, there are three partitions, or in the general sense r-1 partitions. Then if we look at the number of combinations of n-1 intervals taken r-1 at a time, we have the answer when blank boxes are not permitted. This is n-1Cr-1. To see this it may be helpful in the above arrangements to blank out the x's in your mind and think of the intervals in the second paragraph as mail box pidgeon holes and the partitions as marbles. Now, suppose blank boxes are permissible. Then, rather than looking at n items, we're really looking at n+r items where we will distributing r blank items as if they were real items. So, the answer is n+r-1Cr-1

In the ancient books I read, n! is represented as | n

Let x = 1, then the answer is: ((x+1) x 10(x+1) + (x+2) x 10x + (x+3)2 x 100 / 2) x 101 Right? x=1 ==> (2 x 10^2 + 3 x 10 ^ 1 +16 x 1/2] x 10 = (200 + 30 + 8) x 10 = 2,380 Yes, I had, but was to0 dense to see if your expression had anything to do with how you solved the problem or you solved the problem in some other fashion and then you simply wanted to be cute.

We are talking about arranging 13 indifferent objects [kids] among 5 boxes [families] with blanks [no pun intended] permitted. Then, evaluate [n+r-1]C[r-1]: [13+5-1]C[5-1]=17C4=2,380 Best to test the formula with small numbers, like 5 objects among 3 boxes = [5+3-1]C[3-1]=7C2=21. It's fairly simple to write out the cases and see that you get 21.

You done good and in New Math (Math, Nu?) circles, we award an "A" for recognizing that though likely only about one sixth billionth of the US Natioinal Debt, © is a huge number

A census taker visits some number of families and determines that on average they have 2.6 children. (a) What is the least number of families in this census? (b) Given that the least number of families was polled, how many children in total did the families have? © So, (a) and (b) are too easy? Agreed. So, in how many ways may these children have been distributed among the families (e.g., one family could have had all of them, or they could have been split among two families, etc.)?

You should be able to amend the DB plan to max out benefits at 415 if not already being done.

Perplexing question. It is presumed you read and reread the most recent document and cannot find direction. I would vote for the documet when she finally broke service since she was still an active employee at that time. Now, hears some porridge to sip: Suppose that plan provides $100 month pension is amended to provide for $125 month pension. The amendment should be careful so that new level applies to the appropriate group. Otherwise, you might have to apply it retroactively! On the other side, suppose participant had terminated in 1977 and the Plan said he could get a lump sum at 65 and specified the lump sum interest rate was 8%. Now, the person is retiring today and so we must use the PPA stuff to determine the minimum lump sum. Of course, in this case, new law has superseded. It is interesting that when you terminate a plan and purchase a deferred annuity with a lump sum option that the insurers will price and guarantee for the current interest/mortality basis and will continue to use it regardless of published interest rates. While this doesn't seem kosher, they all seem to do this.

Case A - Election box checked - "I am not married" Case B - Election box checked - "I am married but cannot locate my spouse" In either event, have you had direct experience with this situation?

A DB Plan provides that the pension cannot be deferred beyond age 65. Participant Dick Putz is given a timely election package. The Company's records show that Dick is unmarried. A year later a legally married wife from whom Dick has been separated pops out of the woodwork demanding that she had never signed off on his election and that the pension should have been distributed under the automatic J&S form, which was a J&100% survivor annuity. It is determined that the separation is not a legal separation whatever the reasons. Case I Dick elected and received lump sum payment. Case II Dick did not return his election package and so after attempting to communicate with Dick (via telephone and certified mail which is accepted), the Plan started his pension under the life annuity form believing he was unmarried. Interested in comments from anyone who has experience with situations like this on how the issue was resolved. Is the response any different if say "n" (n>5) years rather than one year has elapsed?