Lorraine Dorsa

Inclusion in the testing group depends on eligibility (age, service, hours worked, etc), not on the fact that you ended up with an allocation of $0 (zero is a legitimate number, just like 1 or 10 or 10,000). The coverage rules of IRC 410(B) define includable and excludable employees, not the employer's decision to give them an infinitesmially small (= zero) contribution.

There has never been any guidance on this and the IRS has never mentioned it at any conference or seminar I have attended. Most practitioners I know either ignore it since there is no guidance re what it means or, like me, assume that if you use an "or 5" for normal retirement age (such as "65 or 5th anniversary of participation") everyone will be funded over at least 5 years and this provision will be satisfied. If anyone else has any more comforting answer or analysis, I am interested in hearing it.

I would set up the plan this way: Plan will be sponsored by one of the entities and the other will be a participating employer. Therefore, use the EIN of the sponsoring employer for 5500 purposes, etc. Since Mr. X works for 2 employers, he is treated for contribution calculation purposes as 2 participants--one under the sole proprietorship and the other under the corporation--and his contribution is calculated separately for each employer. Each employer deposits the computed contribution and the sum of the two amounts is allocated to Mr. X. (Just be sure that the sum of the compensation amounts used to compute his contribution does not exceed $160,000 and his allocation does not exceed the 415 limit.)

I think this might be a HR designation granted by SHRM (Society for Human Resources or something like that). SHRM is a major HR organization so I'm sure your HR people can help.

There was a previous question similar to yours and the reply was the Society of Actuaries (located in Schaumberg, IL or at soa.com).

Yes, it was also my guess that he was thinking of the option to recalc one spouse but not the other, but he insisted it was not that, so I thought I'd post the question to see if anyone else had an idea what he was talking about. I agree that if you don't recalc one you don't get the absolute minimum amount possible, but the difference is often not enough to make a difference. It is my understanding that the general wisdom is to recalc participant, but not recalc spouse. This way, if the spouse dies, you can continue to use spouse's life expectancy (it reduces by 1 each year from start date) in the minimum distributions. If you recalc'd the spouse's life expectancy, it drops to zero upon death which results in a large increase in the minimum distribution.

The is a method for "ageless" (or DC general test with imputed permitted disparity) plans which could provide for extra $ to a younger owner. However, it generally requires a fairly significant contribution (about 13% of pay as I remember) to a group of NHCEs. With the right demographics, it's better than integrated for the HCE, but a lot more complicated. Due to the need to provide a large contribution to a group of NHCEs (you need enough NHCEs in this group to pass the DC general test--at least 22.5% to as much as 70%), this is not a common plan design. Essentially, you define classes of employees--HCEs, favored NHCEs who will get the high contribution and other NHCEs who will get the minimum contribution. Then you allocate the minimum to the other NHCEs, the large contribution to the HCE and a contribution equal to the HCE's % less 5.7%. Then you run the DC general test with imputed permitted disparity as proof that this allocation passes.

One of my clients has asked me about a method of calculating minimum distributions in which the IRA is somehow split between the husband (IRA owner) and wife (beneficiary) and separate calcs are done on each part. Apparently, as he understands it, you could use joint life expectancy for his portion and single life expectancy for his wife. I've never heard of this and asked him to send me the information he received from his source describing this method. Has anyone heard of this?

I agree with pax, "reasonable" is what we as actuaries agree is reasonable for a given situation. However, I've not seen a case in which I could justify 3% or 4% as reasonable. It is not impossible for 3% or 4% to be reasonable, just unlikely given the recent historical and expected future rates.

No, I haven't received mine yet. For some reason I wasn't expecting to get it until the 1st quarter of 1999. Maybe that's when they were sent for the last cycle?

I've done Beta testing for the Pentabs DB system with regard to Y2K issues. Just about everything can go wrong if the system is not compliant. The most obvious areas to have problems are service and age computations which in turn affect benefit computations and funding, which is results in just about everything going wrong. Then there are all sorts of not so obvious areas, such as PIA calcs, 415 limits, various interest rates (GATT, PBGC), report printing, etc., etc., etc. If your system is not Y2K compliant, I would be very hesitant to use it for anything other than a preliminary proposal which I then thoroughly checked by hand.

I'm looking at Corbel's DB checklist and I don't see an age exclusion. Are you interpreting the "highest X years in the last ten, excluding the five years preceding NRD" as an age exclusion? I don't see this as such because NRD, depending on how it is defined, could be a single age or a range of ages (e.g. later of 65 or 5th anniversary of participation). Since a DB is designed to pay benefits at NRD, if a reference to NRD is deemed to be an indirect reference to age, almost every provision in the plan could be deemed to be indirectly contingent upon age. All the salary averaging definitions on the checklist are commonly used and to my knowledge have not been challenged as being indirect references to age and therefore impermissible under this regulation.

With the recent significant decrease in GATT rates, I think your issue is going to become a common one. Under the PBGC rules, you must follow an elaborate ballet of notices, filings, etc. all keyed to specific dates. Failure to meet any of the requirements voids the termination for PBGC purposes. It seems to me that simply failing to proceed with the next appropriate step in the termination would void it for PBGC purposes. The IRS follows completely different rules. For their purposes, a plan is not terminated unless assets are distributed within a "reasonable" amount of time, generally defined as 1 year. Again, failing to distribute assets within a reasonable period of time would void the termination. Practically speaking, I would have the plan sponsor adopt a resolution rescinding the termination of the plan (I don't think he could rescind the 100% vesting), notify the participants of such and continue to administer and fund it as a frozen plan. I have seen situation which went as far as receiving an IRS favorable determination letter upon termination which later rescinded the termination and continued as frozen plans and then later received another favorable determination letter at the time of actual termination. Two issues to consider: 1) if the plan termination stated that new participants would enter the plan and no further benefits would accrue and both of these plan provisions continue in force, you need to keep an eye on 401(a)(26) and 410(B) issues. There is an exception to 401(a)(26) for frozen plans unable to terminate in a standard termination and generally no one accrues a benefit so 410(B) is not an issue. 2)if the plan is top-heavy, top-heavy benefits must continue to be provided. In some cases in which there is a DC plan with contribution of 5% or more, the plans are amended such that the top-heavy benefit is provided in the DC plan. Otherwise, top-heavy benefits in the DB plan must continue, which might delay the date at which the plan would be sufficient to terminate.

Under the first method, the funding is recomputed each year. Under the second method, the annual cost is the sum of the normal cost (computed using the plan's funding assumptions) and the 1 year term cost of insurance. Therefore, the annual cost is larger than the cost of a similar uninsured plan by the amount of the 1 year term cost. This total cost is then split between insurance premiums and a side fund contribution.

The plan must be adopted on or before the last day of the plan year to create the plan but it is not clear if the plan must be funded by the year end. The issue regarding funding was that a trust cannot exist without a corpus so an initial contribution (if even only $1) was required to create the trust. This was an active issue a number of years ago, but I have not heard any discussion about it lately. I know many practitioners have the client sign the document on or before the last day of the plan year and then make the contribution at some later date.

Remember that there is no automatic approval of a change in valuation date to the end of the plan year. To make this change, an application must be filed with the IRS and approval received. (ASPA has commented to the IRS on the most recent change in funding method guidance and has suggested that automatic approval be extended to any change in valuation date.)

I will be holding a 2 day review course for the ASPA C-2DB exam on Monday and Tuesday November 16 and 17 in Jacksonville, FL. I know its the last minute, but if you are interested please email or call me at 800 361-4635.

Since benefits must be definitely determinable, I don't feel comfortable in saying the interest rate is the actuary's choice. If I am put in the situation where option 1 is the option the Plan Administrator decides to use, I make the Plan Administrator interpret the plan and tell me the rate.

My firm specializes in providing retirement plan administration and design services for small closely held firms. Please feel free to contact me via email (lda@leading.net) or phone 904 249-9171.

The PBGC rates are one of the reasons I prefer recomputing the lump sum (option 2 above) to simply increasing the shortage by an interest rate. I don't feel I could justify (except maybe in the last few months) assuming a 4% rate of return on assets as making the participant whole.

This issue has been discussed at various conferences and meetings, generally in the context of a small plan (often 1 participant) in which the principal is at this 415 limit so excess assets are not reallocable, but must be reverted with all the applicable excise and income tax consequences. There are organizations which arrange, as I understand it, the sale of the business (including the plan). The sale price takes into account in some fashion the assets in the plan so the original owner realizes more than he would have under the reversion scenario. I have not reseached this in any way, but I have heard several reputable practitioners state that this has worked for their clients. I do not recommend or not recommend this option, I just point it out for your reference. If you do consider an option like this, I agree with Wessex that all aspects need to be considered.

Plans are not required to comply with 401(l) for qualification. 401(l) is called permitted disparity because it works just that way--you are permitted to ignore the disparity in benefits for non-discrimination purposes. In most cases, plans which use 401(l) permitted disparity do so because they want to have a benefit which will comply with the safe harbors of 401(a)(4) and therefore will not require non-discrimination testing using the general rule. If a plan provides integrated style benefits and does not comply with 401(l), either intentionally or unintentionally, it must be tested for non-discrimination re the amount of benefits under the general test of 401(a)(4).

There are two ways to look at this. One is to say that the lump sum was underpaid by $X and bring $X forward with interest. The other is to say that the lump sum represented only a partial distribution which is actuarially equivalent to $Y per month at retirement. Therefore, the participant still has an accrued benefit of $Z (=total accrued benefit less $Y) and is entitled to an additional lump sum which is the actuarial equivalent of $Z. In practice, if the error is discovered relatively soon after the initial payment, the interest rate (if you are using option 1 above) is not particularly significant. Many practitioners use the actuarial equivalent rate. I've also heard a case made for the fund's actual rate of return. I personally like option 2 since it avoids the need to select an interest rate (and then justify that choice to the participant). If the error is discovered a significant time later, you may want to go in for a VCR ruling and, if you choose option 1, have the IRS "bless" your selection of interest rate.

My concern is with ongoing plans and what benefits are protected, how they are protected and what is reasonable for funding purposes? For example, as of 12/31/99 (assumed freeze date), a participant has an accrued benefit under the plan formula of $10,000/month, NRA=65, AE = 5% 1983 GAM (blended). Final implementation date is 1/1/00 and plan selects frozen + future method. As of 12/31/2000, his accrued benefit under the plan formula has increased to 415 $ limit, assumed to be $10,500. 2000 GATT rate = 6%. What is the lump value of his benefit on 12/31/00? Is it 1) the sum of the present value of $10,000/mth @ 5%/1983GAM + the present value of $500/mth @ 6%/1983GAM or 2) the preceding amount, but not more than the present value of $10,500/mth @ 6%/1983 GAM? How do I fund for this benefit if he is going to retire in 2005 and is expected, as many small business owners do, to take a lump sum? Should I assume his benefit will be worth 1), 2) or 3) a lump sum computed at what I expect the GATT rate to be in 2005?

At last week's ASPA meeting, Jim Holland spoke on Rev Ruling 98-1. All of those I spoke to who attended the session, including myself, are just as confused as we were before about exactly what options are available, how to select which option is best for a particular plan or participant, how to calculate the benefits described and how to write these rules in a document. I would like to hear from others, whether or not you attended the session, about how you are addressing this issue.