MWeddell

Ask your new and old recordkeepers this same question. They likely handle mid-plan year conversions all the time. Besides the issues you mentioned in your question, consider that you'll probably need a one-month statement and then a two-month statement from the recordkeepers instead of the usual 3-month statement covering the first quarter of 2000. Also, you should identify who is going to report the distribution taxation for payments made during January 2000.

The plan may not use the pre-SBJPA correction (leveling) method for plan years beginning after 12/31/1996. Hence, an amendment will be needed for that purpose and any ADP/ACP test refunds should have been computed in compliance with the new law.

Whether you use permitted disparity rules as part of the employer nonmatching contribution really shouldn't have any impact on the 401(k) portion of the plan. I suppose those NHCEs who understand the allocation method might perceive that it favors the well-paid and therefore might become skeptical of the whole plan, but that sounds somewhat unlikely and definitely not quantifiable. Regarding the integration level, note that as the social security wage base rises from $72,600 to $76,200 in 2000, that the $15,000 dollar amount will become <= 20% of the wage base. Hence, the maximum excess allowance will go from 4.3% in 1999 to 5.7% in 2000 if you keep the $15,000 integration level. Note that 411(d)(6) regulations will force you to make any plan amendments to the allocation formula effective next plan year if any employees have satisfied all of the plan's conditions for receiving this year's profit-sharing allocation (i.e. if there's no "last day of the plan year" condition).

Responding to Off_the_record: The requirement that one must take any nontaxable loans before one becomes eligible for a hardship withdrawal applies regardless of which method the plan uses to pass the resources test. See Treas. Reg. 1.401(k)-1(d)(2)(iii)(B)(4). While this language describes the contents of the employee's written representation if the employer chooses to rely on the representation, the flush language after that makes clear that the ability to borrow from the plan on a nontaxable basis is a resource that the employee must first use under the general test. I would think that checking to see if those who had hardship withdrawals first exhausted their opportunities to take loans would be fairly easy for an IRS agent to catch. The problem can be detected just be looking at a recordkeeping valuation report. The requirement to take a loan first is accurately stated in the 1994 401(k) examination guidelines, although it's not listed as a specific action step. I wouldn't call the odds of being caught "astronomical." In any case, I though JSamuelson was inquiring about what the correct answer was, not the likelihood of being caught if he or his client chooses to violate an IRS regulation.

Short answer is that stating that the maximum elective deferrals is the Section 402(g) limit ($10,000 in 1999; $10,500 in 2000) is permitted. One doesn't have to state a maximum contribution percentage Some reasons why having a maximum contribution percentage might be a good idea: - Your 401(k) recordkeeper might require that its plans state a maximium contribution percentage. - To keep contributions (especially from nonhighly compensated employees) rising when employees' compensation rises, you might prefer employees to make elections in terms of percentages of compensation, not dollar amounts. Elections as a percentage of pay may be more consistent with your communications campaign. - Elective deferrals in excess of the 415 limits may be refunded only for certain reasons. I'm paraphrasing the regulation very loosely, but in general one can only correct 415 amounts if the error wasn't reasonably foreseeable. This is the reason most plans set the maximum contribution percentage at no higher than 25% minus the percentage of pay of all other defined contribution plan contributions and forfeitures.

A few years back (circa 1993) the IRS made clear that while it supported Congressional action to change cross-testing plans, it felt that 1.401(a)(4)-8 faithfully interpreted the current Code. Any change would have to come from Congress and there's been no proposals that I recall to change the law in this respect.

I agree with Dan Ashley's post. You can do this, but it's a matching contribution subject to 401(m) testing. You also have to have a definite allocation formula in the plan document and consider whether the Section 415 contributions are due to the type of error that allows for corrections.

The prohibited transaction area with all of its exemptions is pretty complicated. I'd tell the company to name a new broker of record or pay (or have the broker pay) for legal counsel to research whether it's legal. [This message has been edited by MWeddell (edited 11-05-1999).]

The IRS probably does not regard this as a separation from service sufficient to permit a 401(k) distribution. Check out whether any of the other distribution events, especially those listed in Code Section 401(k)(10) apply, but probably allowing distributions that are rolled over isn't allowed. Transferring assets will work but the new employer must preserve optional forms of benefit payment. Take a look at the filing instructions for Forms 5310-A. There's an exception for defined contribution plans that usually applies if there are no suspense accounts. If Forms 5310-A are required, check out the 414(l) regulations to make sure you're not violating them and then a Form 5310-A would be filed for each of the plans involved in the transfer of assets, so each employer would be responsible for a form. [This message has been edited by MWeddell (edited 11-05-1999).]

One must take any nontaxable loans from the plan first before the hardship withdrawal request would be valid. Although the employer can deny a loan request (thereby enabling the participant to then apply for the hardship withdrawal), it sounds like the only grounds would be if the employer knows that the employee has no intention of repaying the loan. This might be hard to do if the employee signed an irrevocable consent to have loan payments withheld from paychecks. In short, I'd be careful before letting the employee change his mind and get a hardship withdrawal instead of a loan.

Whether the difference is de minimis is a facts and circumstances determination under Reg. 1.414(s)-1(d)(3)(v), so nothing other than an IRS ruling for your client (typically as part of a determination letter application) is authoritative. That being said, the rule of thumb I've seen used is no more than 2% if de minimis, but I don't see any justification at all for that figure in the regulations. You may of course allocate discretionary contributions as you currently are doing but use a safe harbor 414(s) definition of compensation and perform general 401(a)(4) testing of the contribution. You'd have to perform general 401(a)(4) testing instead of 414(s) testing, but would be more certain about whether the results would be acceptable to the IRS. [This message has been edited by MWeddell (edited 11-05-1999).]

Just to clarify a couple things, it sounds like the plan design aims to give employees the chance to hit the 25% / $30,000 415© limit if the employee contributes 15% to the 401(k) plan. In practice, employee deferrals will vary so not everyone will reach 25% of pay in total contributions. Also, let's clarify that we've got two plans here because the money purchase plan can't include a 401(k) arrangement. Yes, you may use compensation from date of participation (including a one-year of eligibility service requirement) for any employer matching or nonmatching contributions but use compensation from date of hire or whatever earlier eligibility date you establish for the elective deferrals. If your testing software doesn't accommodate this, then run the (k) test separate from the (m) test and manually combine them to see if the multiple use limit passes. Of course, you won't need testing software (assuming there's no employee after-tax software) for 401(k) / 401(m) testing because it sounds like you'll design the plan to meet the safe harbor rules. Section VIII(H) of Notice 98-52 lets you view the portion of the plan benefiting employees with less than one year of service separately, which means the 3% nonelective safe harbor contribution doesn't need to be allocated to those with less than one year of service.

I guess the employer figures the office employees aren't about to unionize, eh? Besides the fact that the plan design seems intended to tick off office employees and field employees who did contribute, I'd be concerned that it violates the contingent benefit rule in the 401(k) regulations. Treas. Reg. 1.401(k)-1(e)(6)(i). The employer also has issues about whether the amount exceeds employees' IRA limits and (considering they are eligible for a 401(k) plan) whether employees make deduct IRA contributions. The employer might also convert the IRAs into an ERISA plan (there's some DOL guidance that addresses what the borderline is). Not a great idea, all in all.

I'll try to clarify. The company I'm working with currently has a cash balance plan contribution credit of 4% per year. Because interest credits are computed based on T-bill rates and the plan assets are assumed to earn more than that (and the plan has been experiencing investment returns in excess of the assumptions, although this doesn't have an immediate impact on the expense), actual cost to the employer is more like 2% currently. My client is considering replacing it with a profit-sharing contribution. The problem is that a 4% profit-sharing contribution will cost 4% of course. Well, after a few years there'll be some forfeitures so in the long run it'll cost somewhat less than 4%, but the cost is still nearly double. I'm just offering this as an illustration: these figures will vary from one employer to the next. In my client's case, if we're to compare plans with the same expense, we'd compare the 4% cash balance plan to a 2% profit-sharing plan. The point I was trying (unsuccessfully?) to illustrate is there's a trade-off involved. Even if one agrees with jlf's point that employees should prefer to bear the investment risk, the stated contribution rate will fall significantly, by 50% in my client's situation, if we want to compare plans of the same expense. Simply observing that a 4% annual profit sharing contribution is better than a 4% cash balance plan contribution, not that jlf expressed the comparison in those precise terms, doesn't end the debate.

You're right about the determination letter. The fact that one employer received a favorable letter provides little evidence that it was valid. I shouldn't have even mentioned it but thought it relevant that I'd at least seen it once in practice. Dowist, let us know if you find anything else to support your comment #2 above. Otherwise, if the nondiscriminatory classification test regulation says we can't enumerate employees by name, it implies to me that it's all right for plans that satisfy 410(B) through the ratio percentage test. Maybe the IRS has an informal position against listing employees by name that can't be detected through regulations, notices, etc.

I haven't seen that clause before. Typically the first drafts of the agreements will have a dollar limit (often much too low) on liabilities, so it's not too much more egregrious than what others propose. Assuming that such a limit wasn't in the proposal and hence not in the "handshake deal" the parties reached before BISYS' lawyers got involved, it may not be too hard to talk them out of it or tell them the client will walk away if that's really needed.

Responding to one comment by Dowist, as far as I know one can enumerate the employees by name who are eligible to participate. One would have to pass the ratio percentage test because it would flunk the reasonable classification part of the nondiscriminatory classification test. I've seen one law firm plan that listed employees by name and received a determination letter.

Watson Wyatt received the same answer when we posed the question verbally to the IRS. The $85,000 amount applies to 1999 compensation when one is identifying highly compensated employees for plan years beginning in 2000.

KJohnson, you have correctly summarized my position. The second point you raise is quite interesting. I don't know that I'll change my mind but it's an excellent point.

Assuming that the plan only allows a participant to take one loan at a time, I think the participant's strategy suggested by mam would work. Note that the participant will have income tax withheld and owed on the $9,000 hardship withdrawal and hence could request more to compensate for the taxes on the withdrawal itself.

If you're offering an opinion (regardless of whether your viewpoint is well-informed and well-reasoned), then it's an editorial. There's no need to take offense when someone begins a comment with "interesting editorial." Moving on to the more substantive points, I've not heard of any litigation to compel an employer to convert a DB plan to a DC plan. I doubt such litigation would be successful. As long as the plan complies with applicable law, the employer is free to offer whatever type of retirement plan, if any, it wants to. In regard to your comment that an employee is entitled to the full economic value of an employer's contributions, keep in mind that if we compare plan designs with the same expected long-term cost to the employer, then there's a link between who is enjoying the investment returns and what those contributions will be. I'm involved currently in a plan redesign project for a large company with a cash balance plan. The company is considering changing to an all defined contribution approach but to provide benefits at the same accounting cost as the cash balance contribution credits (for which the employees get a fixed rate of return that's fairly modest), it would have to halve approximately the level of the stated contributions. Hence, giving investment direction to the employees doesn't just shift the investment risk for better or worse but also significantly decreases the level of contributions.

Piggybacking on the earlier postings, is there a Web site or other reference material that shows all of the common mortality tables?

Well, you have a 401(a)(4) general testing situation, but I wouldn't consider it too much of a problem. If both the NEC and the QNEC contributions are allocated in proportion to compensation, you'll have two rate groups: one for employees receiving both the NEC and the QNEC and one for employees receiving only the QNEC. The first rate group probably passes readily because you stated that there is no 410 problem (prior to the decision to allocate the QNEC) and the employer nonelective portion of the plan is subject to separate 410(B) coverage testing. In general, the same arithmetic used to show the plan passed coverage testing is now used to show that the NEC+QNEC rate group passes 401(a)(4) testing. If you exclude this group of employees from the QNEC, make sure that decision complies with the plan document, which generally shouldn't allow for employer discretion of this sort unless your plan has multiple types of QNECs each with its own allocation formula.

If you're amending the plan document, be aware that the IRS believes that amending the QNEC allocation formula after an employee has satisfied the conditions to earn an allocation (which occurs no later than the end of the plan year) isn't permitted because the retroactive cutback violates 411(d)(6). This is sort of documented by looking in the definition of "other right or feature" near the end of Reg. 1.411(d)-4. I think the IRS might have issued some other weaker form of authority more clearly setting forth its position. However, because the IRS allows multiple types of contributions each with their own allocation formula, you can probably add a new class of QNECs with its own allocation formula. It's an aggressive position: I think it works but check with your (or your client's) attorney. On its face, it's a benefit enhancement, not a retroactive cutback, so doing it for effective for a plan year that has already ended will work.

The gap period refers to the period from the end of the plan year until the date of distribution. Investment earnings credited for the gap period are refunded only "if the plan so provides." Treas. Reg. 1.401(k)-1(f)(4)(ii)(A). Hence the answer is that you have to read the plan document to determine whether gap period earnings are owed on corrective distributions of excess contributions. The date 2-1/2 months after the plan year ends (March 15 for calendar plan years) is relevant for determining the participant's tax consequences and whether the plan sponsor must pay an excise tax but doesn't affect gap period earnings computations (or at least I've never read a plan document where that date affected the gap period earnings computations).