MWeddell

No one knows. The legal standard is to use an interest rate "commensurate with the interest rates charged by persons in the business of lending money for loans which would be made under similar circumstances." I don't know of any persons in the business of lending money at a fixed rate of 15 years to individuals who do not have to demonstrate credit worthiness where the loan is fully secured by an account balance already on deposit with the lender. Hence, from the moment the loan regulations were issued 20 years ago, it is never been clear what interest rate to use. (Banks actually used to make passbook savings loans -- don't know whether they still do but in any case the interest rates for them are not widely communicated and I doubt they were long-term loans anyway.) When in doubt, I figure there is safety in numbers. Given that surveys indicate the most common interest rate charged by other DC plans is prime rate + 1% (and that it typically doesn't change depending on whether the plan is making a 5-year or a 15-year loans), then pragmatically, I would use that interest rate. That's not a theoretically satisfactory answer at all, just a pragmatic one. I can't see the DOL taking any enforcement action against using prime rate + 1% given their lack of useful guidance and their knowledge that that's what the plurality of plans are using. Others' opinions may differ.

I agree. This is a significant change. 2008 coverage tests are needed for the pretax and the match (although one would think that they would pass the ratio percentage test handily. Also, consider whether there are benefits, rights or features in the aggregated plan that must be tested under Treas. Reg. 1.401(a)(4)-4. No, they are not required to permissively aggregated the profit-sharing portions of the plan just because the 401(k) and 401(m) arrangements are permissively aggregated.

I agree with Austin.

There is not a hard and fast rule. The relevant criteria is from Revenue Procedure 93-42, from the section titled "Three Year Testing Cycle." The guidance recognizes that whether one has to rerun the tests annually or can rely on a prior year's test for an extra year or two depends in part by how narrowly the test passed in the first place.

Another possibility is to amend the plan so that the plan year ends September 30. It wouldn't be my first choice, but it's better than telling the client that they can't change the match until January 1.

jlmn, Can you describe more precisely the contributions that were late? Are they safe harbor matching contributions (at least 100% match on 3% of pay and 50% match on the next 2% of pay)? Or are they the 3% safe harbor nonmatching contributions? Right now, it's confusing to follow because the first two posts from this thread refer to safe harbor matching contributions, for which there is the three-month deadline if they are allocated for periods shorter than the whole plan year. On the other hand, your post uses the term "QNEC" which typically refers to nonmatching contributions.

Just to clarify the facts ... So if you combine the two types of matching contributions, then participants got a 200% match on the first 3% of pay and a 100% match on the next 1% of pay and no match on pay deferred in excess of 4%. Is that right? If so, then I agree with the above post. You should be fine and no ACP test need be run.

Due to the nuisance factor though, you probably come out behind.

I've taken the opposite conclusion as above, while warning the client that it is an aggressive position. I've interpreted the "same plan year" requirement to be satisfied if both of the plan years end on December 31. I don't know of any IRS or Treasury guidance that clarifies that both plan years must begin and end on the same dates to meet the "same plan year" requirement.

I doubt this helps much, but it is only a problem for the taxpayer in question. It's not a qualification problem for the 2 plans, assuming that are unrelated employers.

I think rcline46's response fits your facts more accurately. Because discretionary match has been made on deferrals above 6% of pay, I don't think this is a 401(m) safe harbor plan. Therefore all match must be tested. Caveat is that I didn't look up the regulations to verify this. I'll leave you with that homework.

Perhaps the plan document can be interpreted so that the ACP test is part of the allocation formula for the match if the results are known before the match is allocated. This doesn't sound likely to me. More promising idea is that some plan documents allow for contributions by HCEs to be capped as needed to pass the ADP/ACP tests. This provision is typically drafted thinking that cutbacks will happen during the plan year, but it might fit your situation. If those ideas don't work, yes, I think you're stuck.

I would think the answer is yes, that a non-US taxpayer can roll over money from a US qualified plan to a US IRA. Whether that makes sense from the taxpayer's perspective, I don't know.

If you are trying to satisfy the 401(k) / 401(m) safe harbor plan design conditions, this doesn't work. The entire plan must satisfy one of the safe harbor plan designs for the entire plan year.

I ran across a similar issue today, so I'm bumping this thread. It seems to me there are two problems: (1) Failure to disclose in SPD, as mentioned by J Simmons. There's no penalty for this and no DOL program to correct it. The solutiion is to update the SPD or provide an SMM and move on. (2) Failure to administer the plan in accordance with the written document. Presumably distribution packets failed to communicate the existence of an optional form of benefit payment. Failure to follow the plan document is regarded by the IRS as a potential qualification failure and can be corrected following the EPCRS. However, I don't see that there is concrete correction action that one can take because it's impossible to identify which participants if any would have elected the optional form of benefit payment if it had been properly communicated. I think the correction is to just fix it prospectively, starting to communicate it on a going forward basis only.

Agreed. The plan has to state over which time period compensation is being measured for contribution allocation purposes.

I agree with the above response. Sometimes the document also will have a provision that says one can use any other testing rules permitted by the 401(k) & 401(m) regulations, in which case you can decide on the fly whether to use the whole plan year's compensation or just compensation during the portion of the plan year that participants are eligible. Often times, you'll only have data for the whole plan year though.

Guidance on this issue is not 100% clear, but your suggestion probably does not comply with IRC 411, nor does it comply with the parallel provisions in ERISA. If both the pre-2009 and post-2008 contributions are the same type of contribution, such as employer non-matching, non-QNEC dollars, then the whole pot of money must comply with a vesting schedule that satisfies IRC 411. Having some money on 2-6 year graded vesting and some on 3-year cliff vesting does not satisfy IRC 411.

A plan either satisfies one of the 401(k) / 401(m) safe harbor designs for the whole plan year or it is not a safe harbor plan. More specifically, if you meet the QACA 3% nonelective safe harbor design for the first part of the year and then satisfy the older 3% nonelective safe harbor design for the rest of the plan year, this is NOT a safe harbor plan for that plan year.

Treas. Reg. 1.410(b)-4(b)(last sentence) provides that a plan that enumerates by name who is eligible does not satisfy with the nondiscriminatory classification test, which is part of the average benefit test. It strongly implies that including or excluding employees by name is fine as long as one is relying on the ratio percentage test method of satisfying Code Section 410(b). In other words, I agree with J Simmons.

I agree with Tom. It's unclear whether a notice mailed today would be received a "reasonable" period of time before the plan year began to afford participants with an opportunity to change their deferral elections in reaction to what the notice states. Once the notice is given < 30 days before the plan year begins, you don't know for certain that it was timely. How big is the employer? Is it possible that the employer could get the notice drafted and distributed ASAP and then contact each eligible employee individually to see if they had questions? If there was a certification in the file from a human resources dept. employee indicating that those phone calls were made and everyone understood both the notice and the deadline in case they wished to change their deferral elections, then it would be harder for the IRS or a participant to later assert that the notice wasn't timely considering all the facts and circumstances. Just a wild idea, for your consideration.

I disagree with the above post. The rules for loans are the same for profit-sharing only plans as they are for any other qualified defined contribution plan. The IRS regulations under Code Section 72(p) and the DOL prohibited transaction regulations apply to profit-sharing only plans the same way that they apply to any other DC plan. For hardship withdrawals, profit-sharing contributions can be withdrawn once the participant has experienced a hardship event, as defined in the plan document. One does not also have to satisfy the 2 year wait or 5 years of plan participation requirement. See Treas. Reg. 1.401-1(b)(ii), which permits distribution "upon the prior occurrence of some event such as layoff, illness, disability, retirement, death, or severance of employment," and Rev. Ruling 71-224, which interprets financial hardship as being such an event. Most plan documents will define "hardship" similar to how the 401(k) regulations permit one to define the term, but it doesn't have to be defined in that way. Of course, now that you have a vote on each side of the issue, you should probably wait for more opinions from other posters.

I agree, depending of course on what exactly your plan amendment says. Severance payments typically are excluded from plan compensation only if paid after one's employment has terminated.

In my experience it is more common for a plan not to force a distribution (above the $1,000 or $5,000 de minimis threshold) to a terminated participant before age 70½. One has to read the plan document to see if the plan is forcing out distributions upon age 62 or age 65 but that doesn't seem to be the majority practice.

I stand corrected. Sieve is right that IRC 72(t) does not require that the distribution be a lump sum. Of course the particular plan would have to allow partial distributions to participants who have separated from service.