PBGC Variable Premium

[Guest Don N]

The General Rule calculates liabilites for the current year at the required interest rate and the Alternate Method adjusts the prior year's liabilities from Schedule B to the current year and the required interest rate; does anyone know if it's acceptable to apply the Alternate Method to the current year's Schedule B liabilities ?

[Guest Ray Williams]

NO!!!!! The Alterntive Method is the Method as described in the instructions. Any attempted variation would be a method not in accordance with the PBGC instructions and therefore not acceptable. The PBGC allows no diviation, however slight, from its rules, regulations and instructions.

Why would you think that that you could alter the detailed instructions from PBGC?

[Guest Don N]

Allow me to rephrase the question -has anyone interpreted the PBGC's "General Rule" instructions in a way that you would "in effect" be applying the Alternative Method methodology to current year "current liabilities" ?

[This message has been edited by Don N (edited 12-14-1999).]

[richard]

Are you suggesting that one might calculate (or interpret the PBGC rules to allow one to calculate) the current year's PVVB using the "roll-forward and adjust" methodology from the prior year's PVVB instead of actually calculating the current year's PVVB (based on the current year's employee data)?

If so, it's an interesting suggestion although I don't think that would be allowed. The idea behind the PBGC's Alternate Method was a response to actuaries who were unable to calculate exactly the current year's PVVB in time for the PBGC's filing due date. (Yes, in large plans, the actuarial valuation for a January 1 plan year is sometimes completed between September and December of the year.) (Also, this saves the small plan actuary the hassle of calculating an additional liability -- not a big deal today with software packages.)

While one virtually never does any actuarial calculation "exactly" (there are always data assumptions and valuation shortcuts), the "rollforward and adjust" approach does not use current year's data, and therefore (in my occasionally humble opinion) would not be acceptable.

[Guest Don N]

Here's what I'm saying - the end result of both the Gen. Rule & Alt. Method is the valuation of plan liabilities at the Required Interest Rate (RIR) for the current premium year - the Gen. Rule accomplishes this directly by valuing liabilities at the RIR & the Alt. Method uses power factors to adjust prior year's Sch. B liabilities (at the CL rate ) and also includes the 1 year accrual factor of 1.07; if we apply the same power factors to this year's Sch. B liabilities ( at the CL rate) but omit the 1.07 accrual adjustment, don't we end up with current year liabilities at the RIR ?

[david rigby]

I've often wondered the same thing, but I don't think it is supported by the instructions to the form.

I think richard's second paragraph above is accurate.

[mwyatt]

Let me see if I get this straight (assume a calendar year plan). For example, you are doing the Premium for the 1999 year beginning 01/01/99, so the following liabilities would need to be calculated:

1) If General Test, use Accrued Benefit as of either 12/31/98 or 01/01/99, run on your Current Liability assumptions and the PBGC interest rate

2) If Alternative Calculation, take results from prior 1998 Schedule B as of 01/01/1998, which is the 01/01/1998 Accrued Benefit valued on Current Liability assumptions, then adjusted using the methodology described in the instructions to Schedule A.

By virtue of what you are proposing, you already have the 01/01/99 Accrued Benefits calculated and valued on Current Liability assumptions. I think that the biggest reason (as previously mentioned) for using the Alternative Calculation method is that this benefit has not yet been calculated; if you already have the 01/01/99 benefit determined, I wouldn't think it would be such a big deal to run using the PBGC required interest rate.

[Guest Ray Williams]

Again, the PBGC instructions for the Alternative Calculation specify the liability that must be used, and the interest rate. For a 1999 plan, use must use the Present Value of the Vested Benefits as of the first day of the 1998 Plan Year.The interest rate must be the interest rate reported on the 1998 Schedule B. If you have already run the current year valuation and have the current year liabilities, you can then use the General Method. It would appear from your question and clarification, that your real concern is to not have to have to pay an EA to sign the Schedule A.

[Guest Don N]

Our dialoque has helped me think this through & I think what I was suggesting overstates the liabilities; recall, that I was suggesting applying the Alternative Method factors to current year val. numbers at the BIR rate ; what this is doing is not only adjusting to the RIR but advancing liabilities 1 year which is what the Alt. Method was designed to do but it was meant to be used with last year's numbers & if it's used with current numbers it overstates-

thanks for the dialoque!!

[richard]

Don N's idea of applying the Alternative Calculation methodology to the current year's Schedule B liabilities raises an interesting thought.

Using the General Rule, do we have to run an actual valuation at the RIR using current year data to calculate a PVVB, or can we adjust other valuations that use the same current year data to arrive at the same result? (It is critical that we use the current year data.)

For example, let's say these calculations are for the 2000 PBGC Variable Premium. Assume the RIR is 6%, and we already have computed PVVB at 5.5% and 6.5% (perhaps for funding FASB or other purposees), using 1/1/2000 data.

Alternative #1 - Rather than running an additional valuation at 6%, could we average the PVVBs calculated at 5.5% and 6.5%? Perhaps.

Alternative #2 - Assume instead, we have already computed PVVB at 8% (for funding purposes), and have run no other valuations. Could we "adjust" the 8% PVVB to 6% using the PBGC's adjustment factors? Perhaps.

In my oft-laughed-at humble opinion, we could make a case for using either alternative #1 or #2 subject to issues of materiality and bias. (I personally would be comfortable with #1 but not with #2; though as a practical matter, I would simply perform the additional valuation at 6%.)

PVVBs are not exact amounts; they are based on 1/1/2000 data (which can have hopefully immaterial errors in them). [Note that the PBGC reserves the right to audit the underlying data and require a revised calculation of PVVB.] These PVVBs are calculated using valuation systems that make approximates insofar as employee data, and use rounding conventions, and other calculation experiencies. In general, these are (hopefully) immaterial and unbiased.

Would an approximation under #1 be likely to cause any material errors? Is it biased? I don't think so, hence an argument could be made to use it.

Would an approximation under #2 be liked to cause any material errors? Is it biased? It can be, particularly in valuing actives and terminated vesteds. The 15 year deferral inherent in the PBGC's liability adjustment formula understates the adjustment needed in a young (weighted by liability) employee group, and overstates the adjustment needed in an old (weighted by liability) employee group. The 0.94 adjustment factor for the immediate annuity could be biased depending on the form of annuity, the underlying interest rates, and the normal retirement age. In both cases, the biases could be material especially if there is a significant diference between the interest rate used and the RIR (in my example, 8% vs. 6%). Therefore, I believe this approach could be acceptable if used carefully (e.g., adjusting from 6.1% valuation rate to 6.0% RIR), but could also be improper.

Enjoy.

[david rigby]

Whether or not you "could" adjust in Alternative 1 or 2 as suggested by richard, it seems to me that the Enrolled Actuary is "putting it on the line" here, by signing that the liabilities shown are accurate.

There is the "macro" approach where the plan is easily fully funded, resulting in no variable premium. In that case, how much fine tuning is important?

There is also the other perspective that the PBGC just may want a good estimate of the liability (for its own analyis purposes) even when a variable premium is not at issue.

Probably up to the EA to provide best estimate of the information that is being requested. The method used by the EA to do that is the decision, and responsibility, of the EA. If there are extenuating circumstances, let them know, but no games please.

[mwyatt]

I could understand the need to approximate the PBGC variable run if this was back in 1983 when we all were using "state of the art" IBM PCs with 256k ram and two floppy disks, but this is 1999 with a little more computer power available to all for a fraction of the cost.

I'm not trying to be flippant, but I really don't understand why you don't just run the numbers using the PBGC interest rate. Most of the software packages I've seen provide for all of the various runs to be processed at one time anyway. If you have the current benefits calculated, why is it such a problem to do one extra valuation run (if not already calculated with the regular valuation?).