MWeddell

Also consider these points: - An SPD must have the name of the plan and the plan number. Hence, if the name of the plan changes due to a plan merger, a new SPD or an SMM must be distributed within 210 days after the end of the plan year in which the change is made. - If there are different provisions in the new 401(k) plan that may affect participants, then some courts may find that ERISA imposes a duty on fiduciaries to disclose the provisions. Otherwise, participants who take detrimental action in reliance on the old provisions may be able to sue for damages. See Panaras v Liquid Carbonic Indus. Corp. (1996, CA7).

Per month and per plan-year quarter options are also available. Here's Q&A-2 from IRS Notice 2000-3: Q-2. Can a 401(k) safe harbor plan match elective and employee contributions on a payroll-by-payroll basis (instead of on an annual basis) without making additional contributions at the end of the year to take into account the total amount of an employee's compensation for the plan year? A-2. Notwithstanding section VII.A. (or any other provision) of Notice 98-52, the requirements of sections V.B.1. and VI.B. of Notice 98-52 that relate to matching contributions may be met for a plan year by meeting such requirements either (1) with respect to the plan year as a whole, or (2) if the plan so provides, separately with respect to each payroll period (or with respect to all payroll periods ending with or within each month or plan-year quarter) taken into account under the arrangement for the plan year (the "payroll period method"). If the payroll period method is used, however, matching contributions with respect to elective or employee contributions made during a plan year quarter beginning after May 1, 2000 must be contributed to the plan by the last day of the following plan year quarter. Accordingly, in the case of a calendar year plan that uses the payroll period method, matching contributions with respect to elective or employee contributions made during the calendar quarter beginning July 1, 2000, must be contributed to the plan by December 31, 2000. The payroll period method applies only for purposes of satisfying the ADP safe harbor matching contribution requirements of section 401(k)(12) (section V.B.1. of Notice 98-52) and the ACP safe harbor matching contribution requirements of section 401(m)(11) (section VI.B. of Notice 98-52).

I agree with Tom's result but disagree with his reasoning. Rev. Proc. 2003-44, Section 6.02(2) makes clear that the correction methods listed in Appendices A and B are not the only possible correction methods. Hence, telling a client that its desired correction method isn't listed in the Appendices doesn't completely win the argument. Instead, you want to look at Section 6.02(2)(b). It contains a weasel word "typical" but otherwise disposes of your client's argument: The correction method for failures relating to nondiscrimination should provide benefits for nonhighly compensated employees. For example, for Qualified Plans, the correction method set forth in § 1.401(a)(4)-11(g) (rather than methods making use of the special testing provisions set forth in § 1.401(a)(4)-8 or § 1.401(a)(4)-9) would be the typical means of correcting a failure to satisfy nondiscrimination requirements. Similarly, the correction of a failure to satisfy the requirements of § 401(k)(3), § 401(m)(2), or § 401(m)(9) (relating to nondiscrimination), solely by distributing excess amounts to highly compensated employees would not be the typical means of correcting such a failure. Another argument might be that if all of the correction methods specified in the 401(k) regulations were permitted under the SCP then it would make useless the 12-month correction deadline in the 401(k) regulations. IRS guidance shouldn't be interpreted to make part of it to have no effect.

It's permitted by law pursuant to Rev. Ruling 71-224. However, most plan documents do not allow this, so you'll need to check your plan.

Although the issue's resolution isn't clear, I tend to disagree with mbozek's response. If you think about it, the employer (or an outside payroll provider hired by the employer) will be using the 403(b) deferral data in order to compute the what the match is in the 401(a) plan. This might be going beyond the "maintaining records" provision of Labor Reg. 2510.3-2(f)(3)(iv). Also, the match may change the "completely voluntary" nature of a non-ERISA arrangement in Labor Reg. 2510.3-2(f)(1). Others may come to a more aggressive conclusion (I already admitted the resolution isn't clear), but I tend to give my clients conservative advice on this issue because the worst result is for the client to treat the arrangement as non-ERISA and later have it determined that it is subject to ERISA. Here are some other threads on this issue although neither one really got a full-blown debate going: http://benefitslink.com/boards/index.php?s...t=0entry22263 http://benefitslink.com/boards/index.php?s...t=0entry64104

What's the source for Tinman's answer to question 1 above? Do others agree that it's not too late to amend a plan to change from prior year to current year testing method?

Just to be clear about it, the vesting schedule you have is not permitted for a post-2001 matching contribution. The modification you suggest (20% vesting in year 2) would work. Obviously, I've not addressed what to do about past distributions to those with 2 but less than 3 years of vesting service who earned an hour of service after the end of your 2001 plan year.

This approach is aggressive, but defensible based on the regulations. I'd do it but disclose to the client. Your example is somewhat simplified, helpful to make your point but don't overlook something. Because of the way post-1996 ADP tests are corrected (by reducing HCEs starting with the highest dollar amount), your ADP test numbers after you make refunds won't by 3.5% and 5.5% but rather the HCE average will be somewhat higher than 5.5%. The corrective method of refunds will deem you to have passed the ADP test, but your average percentages won't quite be the same as shown in your example. Sorry if I'm not explaining this very well.

I agree with the above answer as long as we're assuming that the 403(b) is not a church plan defined under Section 403(b)(9).

On the general topic of what records must be retained by an employer or plan administrator, see Rev. Proc. 98-25 (and also IRS Notice 99-1, Law and Analysis, fourth paragraph). More specifically regarding the initial post, the QDRO Answer Book (by Panel Publishing, part of Aspen Publishing) has a discussion on the topic that is worth checking out. The short answer is that it's unclear.

I've seen this plan design a few times. In essence, it's communicated as a match where the additional benefit is provided in the pension plan, but it's subject to 401(a)(4) not 401(m) testing. It only makes sense for large employers -- otherwise the actuarial and administrative issues it introduces offset any advantages.

There is no maximum for an enhanced safe harbor match. There are some restrictions including (i) discretionary contributions can't exceed 4% of pay, (ii) no match on deferrals above 6% of pay, (iii) no HCE gets a better match rate than any NHCE. Of course 415 limits apply too. For example, if a plan document were to specify that the match is 1000% of the first 6% of pay deferred into a plan, that plan could still meet the enhanced matching 401(k) safe harbor formula. In practice, I've never seen any match nearly so generous. The 401(k) safe harbor rules are mostly in Notices 98-52 and 2000-3. The July 2003 proposed 401(k) & 401(m) regulations consolidate the 401(k) safe harbor rules, but they aren't effective yet and occasionally change the rules.

Treas. Reg. 1.401(k)-1(a)(3)(ii) and (iii).

(a), although the plan document could be more strict about the timing and require (b) or ©. The entire $1,000 is available in cash to the employee after the later of when he or she becomes eligible and when the cash or deferred election was made.

To the extent it covers your questions, the annual EBRI / ICI survey at http://www.ici.org/stats/res/per09-05.pdf is the largest most representative sample out there. Otherwise, I think the Fidelity survey will be your best source.

Looks like there's a rebuttal to Joel's article on 403bwise.com. The rebuttal is here - http://www.403bwise.com/features/nysutrebuttal_lm.html - and at the top of that page is a link to Joel's article, in a much easier to read form than a message board post.

Joel, That was a bit of a conversation stopper. I liked your Frank v Arronson posts better.

Medium and large 401(k) plans are more likely to select lower-priced funds and to qualify for lower-priced class of shares than the average retail investor. Of course jumbo 401(k) plans often have 1 or more asset classes than no longer use mutual funds but instead will use lower-priced institutional alternatives.

Note that the average expense ratio doesn't include sales loads, so it understates the expenses that the average retail investor pays. On the other hand, if you're talking about the average cost of mutual funds in a 401(k) plan, then that's specific to the size of plan you're talking about and using the Morningstar averages will overstate the typical expense ratios for medium or larger plans. Use those averages carefully.

The proposed regulations don't require IRS permission to switch from the current year to the prior year method after 5 or more consecutive years of using the current year method. Prop. Treas. Reg. 1.401(k)-2©(1).

We're assuming that a match at one time in the past was made, right?, so that the 3% rule can't be used. You're right that the solution is to use the current year method and to stay on it for five years (if not permanently).

Yikes, this thread did meander around a bit since the last time I checked in. I agree that a match of 100% on the first 4% of pay deferred and 500% on the next 2% of pay deferred is not permitted in a plan designed to meet the 401(k) or 401(m) safe harbors. If we're talking instead about a plan that must meet 401(k) and 401(m) testing, obviously whether the match is actually credited to HCEs in a discriminatory fashion is governed by the 401(m) test. In my opinion, there is no benefit, right, or feature testing issue because it is a uniform matching formula that is available to all eligible participants. While it is true that in the ERISA Outline Book, Ch. 11, Section XII, Part E.3.b.1 it says that the tiered match might need to be tested for effective availability, if one follows the cross reference to Ch. 9, Section X, Part C and also reads all of Treas. Reg. 1.401(a)(4)-4©, you'll see that there are no facts that support that there's a benefit, right, or feature problem. The match is in fact currently available to all, but one suspects that it is less likely that NHCEs will use the full match -- that's a 401(m) concern not a BRF concern. The examples in Treas. Reg. 1.401(a)(4) concern cases where NHCEs can't grow into the conditions regarding availability of the feature or the feature isn't communicated to them, circumstances not present in this example. Blinky, I don't really have experience with the IRS on this. Yes, I've obtained a detemination letter for a client with a tiered match where the match increases, but since there's no BRF testing issue in my opinion, I didn't ask the IRS to specifically rule on the possible BRF issue. If no NHCE defers more than 3%, then the 401(m) test is likely to fail and will need to be corrected but it's still not a issue regarding whether the match was available to NHCEs. RButler, does the plan you're talking about allow employee after-tax contributions? If not, then I agree with KJohnson's reply that it's a 410(b) issue (and one that sounds like it might fail for most employers). If it does allow employee after-tax contributions, then the fact that the match is available to only some of the 401(m) eligible participants is a BRF issue (and again passing it may be a challenge). I agree with Tom Poje's responses to Earl's questions.

It strikes me as extremely unlikely that a match on the first 6% of pay would be found to violate the effective availability test. The examples in the regulation indicate the test is designed to prevent much more outlandish practices. I doubt that you need to do any BRF testing to justify that this match formula is permissible. An increasing tiered match is prohibited for plans designed to meet the 401(k) & 401(m) safe harbor rules but otherwise are not prohibited.

The regulations could be clearer on this point, but if a plan has a minimum participant loan amount > $1,000, then the participants who have a large enough account balance to obtain a loan must meet the nondiscriminatory classification test in order that the loan not be considered a discriminatory benefit, right, or feature.

WDIK points us to one thread but there have been a lot of them over the years on this topic. My own opinion is that assuming that the NHCEs in the plan are allowed to contribute at least 6% in a calendar year plan ($12.000 / $200,000) (or 12% for a non-calendar year plan), then you needn't worry about the $200,000 limit at all because the 402(g) limit is more restrictive so that even after the fact when one considers only $200,000 of compensation, no HCE has a higher available rate of deferrals than was available to the NHCEs. This is assuming that the plan document is properly drafted.