MWeddell

It depends on what your plan document says. Most of the documents I've read would allow someone still employed within the controlled group to take an in-service withdrawal.

Providers who I've read or been told have this feature already in production: Fidelity CitiStreet JP Morgan Putnam Vanguard Hewitt T. Rowe Price Strong It's safe to say that nearly all major recordkeepers either have this feature and plan to implement it during 2004.

Yes, it can, assuming it satisfies 401(a)(4) general testing. I've seen this once before.

Prohibited transactions is an area that's complicated enough that I'd tell the client to get a legal opinion if they want to do it.

If you want to know what's happening in 2003, you've only got anecdotal evidence for now. However, in 2002, the anecdotal evidence suggested that match cuts or suspensions were commonplace and it wasn't so. Surveys comparing the average match rate in 2002 to the average match rate in 2001 showed no change. Surveys I've read are from Principal (specializes in small plans), New York Life Investment Management (specializes in mid-size plans) and Hewitt (specializes in very large plans).

Also, you might question why bother to change the plan year for the pharmacy's plan. Because the 402(g) limit is always a calendar year limit and because most payrolls report information more readily on a calendar year basis, I don't see why they'd care what the plan year is. Under today's plan design, there are no discretionary employer contributions where they first want to see year-end financials before deciding how much to contribute.

Yes, you'll need to file a Form 5500. There'll only be testing if there are contributions on December 31 itself. You might get lucky and not have any contributions then. If you do have contributions, the testing might not be so horrible because anyonw who's hit the 402(g) limit already will be 0%, which should help your test. No vesting issue if you're on elapsed time. If you're using 1,000 hours method, then Tom Poje's comment already addresses that. Other issues - You'll need your amendment to be signed before hand. One can't retroactively change a plan year. - Eventually you'll need an SMM or new SPD.

They should be able to elect to transfer their accounts to the ERISA plan. However, the employer cannot required that the amounts be transferred.

You've probably made a distribution that violates the distribution restrictions of Code Section 401(k). There is no remedy in the regulations. This is regarded as a qualification issue. You need to consider whether the Self-Correction Program or other correction from Rev. Proc. 2003-44 applies.

If you split the employee population, then consider having all HCE dollars contributed to the 403(b) plan. Having some of the HCE deferrals (to the extent not expected to exceed the ADP test limit) made to the 401(k) with the rest contributed to the 403(b) might not be allowed per Notice 89-23, Section V.B.3.d.

If you're summarizing the plan language, it sure doesn't sound like there is a hardship withdrawal event yet.

The answer is yes, although getting your recordkeeper & payroll to do this for only part of the plan might be a challenge. Perhaps your real question is whether having automatic enrollment for only part of the plan constitutes a benefit, right, or feature subject to testing under Treas. Reg. 1.401(a)(4)-4. I've concluded no, that what needs to be tested is the rate of available deferrals and match or the availability of investments, and automatic enrollment doesn't make a greater rate of deferrals or match or more investment choices available to participants. However, the resolution to this issue is not all that clear, so you might ask your own attorney.

It probably won't matter. If the NHCE that incorrectly was considered ineligible for the plan has been credited with QNECs equal to the deferrals and match that the average NHCE had in 2002, then the employee doesn't change the average ADP and ACP for NHCEs.

No, one does not need to match catch-up contributions. If one does indeed match catch-up contributions, then the match is subject to ACP testing.

There are a fair number of biases in the various surveys regarding the average participation rate. The PSCA survey tends to be biased toward those employers who are enthusiastic about their 401(k) plans. To me this is clearest when one examines the average level of contributions in the respondents' plans -- very high compared to other surveys and/or my subjective impression. The Hewitt survey tends to include large plans to a greater percentage than is true throughout the 401(k) industry. Given that many studies have shown that large plans tend to have lower participation rates than other plans, that could bias the average participation rate. For what is essentially a data question, I suggest that surveys based on data straight from the recordkeepers are more accurate than surveys that are compiled based on sending questionaires to plan sponsors and counting on them to submit accurate responses. Given that the big EBRI / ICI survey doesn't include this data, the next best source is Fidelity's surveys because they recordkeep plans of a wide variety of sizes. I realize that by focusing on plans of just one recordkeeper there's a different type of bias introduced, but I think this is the best one can do. Fidelity's Building Futures Volume III reports that as of the end of 2000 the average participation rate was 73%.

Well, I'd guess the opposite, that the IRS notice overrides any inconsistent plan provisions. During years that the plan uses the safe harbor method to avoid the ADP / ACP tests (except for any after-tax contributions), then the plan is treated as using the current year method for those plan years. This would be true even if the plan document seems to imply that the prior year method is elected.

The plan doesn't allow employees to make after-tax contributions, right? Otherwise, (1) employees who weren't active on the last day of the plan year are start eligible to make 401(m) contributions, specifically after-tax contributions, during the plan year and hence are benefitting and (2) the last day of the plan year condition for the match becomes a BRF problem, not a coverage testing problem. Also, the above answers tend to assume that you are not using snapshot data as of the last day of the plan year for coverage testing. If instead you use only employees on the last day of the plan year, it might be easier to satisfy coverage testing even if you now need to hit a ratio percentage of 77% instead of 70%. If all of this sounds unfamiliar to you, review the snapshot testing rules from Rev. Proc. 93-42. Good luck!

Under Code Section 401(k)(4)(B), tax-exempt organizations may include a 401(k) arrangement as part of a retirement plan, but governmental employers cannot. Between 1986-1996, tax-exempt organizations could not add a 401(k) arrangement to a plan, but that prohibition was lifted effective in 1997.

I disagree with the above. A salaried employee is paid for an unspecified number of hours subject to a minimum such as 40 hours per week. The person is not acting as an unpaid volunteer. Take a look at Labor Reg. 2530.200b-3©(1). Why would the regulations provide for equivalencies, including 45 hours per week, if one could merely limit hour of service credit to only 40 hours per week as suggested by the above post. Even though I disagree, the issue is not very clear from the regulations. Consider checking with your legal counsel.

There are several permissible equivalencies for when records of actual hours are unavailable. For example, one may assume that employees earn 45 hours of service per week. You'd have to read the plan document and regulations as the above post advised to see if the plan is using one of those equivalencies.

IRS Notice 97-45 requires consistency across plans regarding the calendar year election and the top 20% group election for determining HCEs.

Also, if the same match formula applies to catch-up contributions as applies to elective deferrals, then there's no benefits, rights, or features testing on the match on catch-up contributions. I believe it's easier to match all deferrals instead of matching only traditional elective deferrals and not matching catch-up contributions. The additional cost under most plan's matching fomulas is slight if any. It also gets rid of the problems of adding a match later if money originally thought to be catch-up contributions when contributed turns out not to be catch-up contributions when re-examined at the end of the year. However, most plans with catch-up contributions don't match those contributions.

The links in Demosthenes' post are to the prior year's CFO survey. The first half alphabetically of the survey (called the 401(k) Buyer's Guide) is already on www.cfo.com and the second half probably will be posted May 1 or so. The Pensions & Investments issue I mentioned before is 10/14/2002.

I'm unaware of any IRS guidance on precisely how to measure the 6-month (or the old 12-month) suspension period. Note that unless you're dealing with a plan intended to satisfy the 401(k) and/or 401(m) safe harbors, the suspension may be longer than 6 months as long as that complies with your plan document. If you're concerned about the issue, just make sure your plan document describes how the 6 months are measured and comply with it.

Using the prior year testing method for the year after the safe harbor plan rules no longer apply should be fine if you've been using the current year method for 5 consecutive years. The authority cited by the above posts concerns changes in the middle of a plan year.