large plan valuations administration

[Guest lip]

We are in process of taking over a 600 life DB plan.

Most of our plans are 2-100 life plans.

My sense is other than pre retire mortality and turnover assumptions;the valuation work will be similar to a small plan.

I know there are other issues like more FAS reporting and PBGC reporting;but am I being to optimistic re larger plan?

[david rigby]

Yes.

[Effen]

I agree, different animal, different problems. Just becuase you are a good rabbit hunter doesn't make you a good deer hunter. You should take a look at the actuarial standards of practice and make sure you are qualified to do the work under the standard.

Any you probably should increase your E&O coverage.

[SoCalActuary]

Expect a more intensive relationship, with issues on financial economics, HR policy, responsiveness to client, constant training of client personnel due to their turnover. Audit issues are more intense.

Responsibility for decisions shifts within the plan sponsor's politics. HR will try to assert authority, but the CFO will have more power.

Every decision will need to be better researched and documented. Assume every decision will have to be explained to legal counsel at some point.

And bill by the hour.

[Guest lip]

I agree, different animal, different problems. Just becuase you are a good rabbit hunter doesn't make you a good deer hunter. You should take a look at the actuarial standards of practice and make sure you are qualified to do the work under the standard.

Any you probably should increase your E&O coverage.

Thanks.

That's why we asked question.

What makes it a different animal in actuarial level of expertise terms?

thanks again

signed

rabbit hunter

[Guest lip]

Yes.

Do you know what would make it from a technical stanpoint actuarially more involved?

ty

[Effen]

One reason would be the need to use of multi decrement fully iterative systems to determine plan liabilities. Typically firms that do only small plans use software designed for small plans (ASC, DATAIR, etc.) Firms that do work for larger plans use software designed for larger plans (Proval, Lynchval, etc.). These larger plan systems are generally fully iterative. In other words the liability for any expected decrement can be isolated in any future year.

Therefore, when the client asks you to determine the value of a change in the early retirement or disability factor you can determine the true cost of the change. In addition, you can change the retirement scale to account for the increased or decreased utilization. Also, keep in mind that you are required to have assumptions that are independly reasonable. In other words, if the plan has a disabiilty benefit you are suppose to have a reasonable disability assumption. It is difficult to know your assumptions are reasonable if you can't isolate them.

I know the small plan software providers will argue their systems will do it, but they generally use approximation techniques and do not do it directly.

Another more basic reason is what SoCal pointed out. If you only deal with small plans you won't know what you don't know regarding larger plans. You will be getting a different type of question from various levels inside the company. The president of the company, the CFO, the auditor and the lawyer could all ask you the same question, but each would have a different reason for asking and you would need to know how to talk to all of them.

No, the math and funding concepts aren't really any different, but the application and information required can be dramtically different.

[david rigby]

Comments from SoCal and Effen are "spot on". A "small plan software" package may be perfectly reasonable for this plan.

Other issues that could arise may include: other plans within a controlled group, non-qualified plan(s), minimum coverage, time spent responding to auditor, completion of 5500, different benefit for different groups/divisions, etc.

One issue that will likely be irrelevant: top-heavy.

[Guest lip]

One reason would be the need to use of multi decrement fully iterative systems to determine plan liabilities. Typically firms that do only small plans use software designed for small plans (ASC, DATAIR, etc.) Firms that do work for larger plans use software designed for larger plans (Proval, Lynchval, etc.). These larger plan systems are generally fully iterative. In other words the liability for any expected decrement can be isolated in any future year.

Therefore, when the client asks you to determine the value of a change in the early retirement or disability factor you can determine the true cost of the change. In addition, you can change the retirement scale to account for the increased or decreased utilization. Also, keep in mind that you are required to have assumptions that are independly reasonable. In other words, if the plan has a disabiilty benefit you are suppose to have a reasonable disability assumption. It is difficult to know your assumptions are reasonable if you can't isolate them.

I know the small plan software providers will argue their systems will do it, but they generally use approximation techniques and do not do it directly.

Another more basic reason is what SoCal pointed out. If you only deal with small plans you won't know what you don't know regarding larger plans. You will be getting a different type of question from various levels inside the company. The president of the company, the CFO, the auditor and the lawyer could all ask you the same question, but each would have a different reason for asking and you would need to know how to talk to all of them.

No, the math and funding concepts aren't really any different, but the application and information required can be dramtically different.

Thanks very much;we really appreciate the information(and we do use datair!)

[AndyH]

Ingrain the words "At Risk" in your head and figure out how to avoid it.

[Ron Snyder]

A belated comment: The large-plan (seriatim) valuation package is a necessity if a large plan design. Small plans use valuation software that values only the normal retirement benefit at the normal retirement age. Other benefits are drafted as and assumed to be related to the "accrued benefit", a ratable portion of the NRB. Actuarial assumptions attempt to estimate the overall extent of the actuarial gains or losses by applying salary scale, mortality, turnover and morbidity assumptions.

Large plans use valuation software that uses the actuarial assumptions to separately estimate the probability and cost of each potential benefit event (NRB, early RB, disability benefit, vesting benefit) and value that portion of the benefit. They then total up all of the values to put the valuation together.

Neither method is "correct" or "incorrect". The valuation results from the 2 methods may be similar or quite different. However, the primary differences in results are more a function of plan design rather than differences in methodologies. Are ancillary benefits a ratable portion of the NRB, or are they developed separately. Is there a "window" benefit to value?

The other difference between the 2 is that larger plans employ "group" actuarial cost methods (ie, FIL and aggregate cost) rather than individual cost methods (individual aggregate, entry age normal). Both use unit credit or projected unit credit methods, although for different reasons.

Datair is not capable of providing a large plan or seriatim valuations. You may wish to have a service bureau provide the actuarial valuation numbers for you. You can continue to use Datair for plan administration, benefit tracking and statements, but export the data files to transmit them to outsource running the valuation numbers. As a small plan actuary who did several large plan valuations, that is what I had to do.