Bird

It's supporting evidence that the five year rule is not intended to be used to delay a distribution, then invoke the life expectancy rule.

Thanks Masteff, that's the point I was trying to get across, not very well...

Appleby- Oops/my bad. You're right, I didn't cite that correctly. You can indeed rollover, but then you still have to take it out (of the IRA) within 5 years. My point was that you can't invoke the 5 year rule, then do a rollover and use the life expectancy rule from the rollover.

mjb, I did respond about the 5-year rule. I don't think it can be used to defer a distribution and then roll over. The five year rule says you have to take a complete distribution within 5 years, and I believe that means that any distribution taken after electing the 5 year rule is an RMD, ineligible for rollover. The IRS made that clear with non-spouse rollovers in a notice earlier this year, and I think the same logic applies. But, there was never any indication that the 5 year rule was elected or would be a default election in this case anyway so I don't think it's applicable. And I have given the cite, as has Masteff, for our position. And before you get testy about non-responsiveness, how about acknowledging the fact that you wasted my time, and anybody else who took the time to look it up, when you provided an inapplicable cite in an outdated publication? I have no problem with someone being wrong, myself included, but an "oops/my bad" doesn't hurt when applicable. Appleby, I've given it some thought and stand by my original position, PLR notwithstanding. It may sound arrogant and/or stubborn, but I think they got it wrong in allowing a rollover of the full amount. A quick recap, without cites, goes like this: Distributions must commence by 12/31/07. What's the minimum amount? Account balance divided by the applicable life expectancy. That amount is a required minimum and is ineligible for rollover. I recognize what you're saying about doing it before the end of the year, and I recognize I am applying some lifetime distribution logic to this situation, but I think it is appropriate. It's weird, at best, that doing a rollover on Dec 30 means no RMD is due but on Dec 31 means one is due. If that's not true, then you get odd results. If Fred had died the year before or the year after, either he or Wilma would have to take an RMD from his account in 2007. Just happening to die in 2006 means that the RMD is skipped for 2007? Logic doesn't always apply, but that doesn't make any sense to me.

mjb: Ah. Yes, it says no RMD is required. Unfortunately, they are talking about lifetime distributions to the owner so it is not relevant to this discussion. Clearly, we are talking about death before the required beginning date and we have been talking about death distributions all along. appleby: The PLR appears to support your position. There are some differences; in the PLR, the spouse is younger than the participant so distributions had to begin before the end of the year in which the spouse turned 70 1/2 (incidentally I messed up my condensation of the regs a few posts ago and will edit it to say that in the case we're talking about, distributions must begin by the end of the year following the year of death, not the end of the year in which the spouse turned 70 1/2). I'm not sure that distinction is all that relevant though, so you clearly have some good ammo there. Essentially it says you can beat the "distributions must commence by" clock by doing a rollover in the year during which they must commence as long as you do it before the end of the year. Frankly, I still have a problem with that...at the moment I'm not sure if I'll dig any deeper or not. As for using the 5-year rule, I didn't think it was possible to use the five year rule and then do a rollover, thereby delaying RMDs. If the PLR is right, then maybe you can (even though the 5-year rule isn't in the PLR, the same logic of "distributions must commence by" and rolling over to beat the clock might apply.)

mjb: Masteff gave it: Reg 1.401(a)(9)-3 Q&A-3. I was paraphrasing it. Could I ask what year's Pub you refer to? I checked both 2006 and 2007 and the page number is off; the relevant sections appear to be on pp 35-36 and I don't see what you said is there. Appleby: If you condense the cite down to the relevant words it comes out: distributions must commence on or before...the end of the calendar year in which the employee would have attained age 70½. EDIT: Oops, I meant to say: distributions must commence on or before...the end of the calendar year immediately following the calendar year in which the employee died P.S. I think Masteff's P.S. is accurate. It should still be titled as a beneficiary account in the plan. So they must commence on or before 12/31/07. If you can spell out for me how you think that should be interpreted, I might understand better, but I'm not following.

Are you saying that if she rolls over any day between Jan 1 and Dec 30, there is no RMD for 2007, but if she rolls over on Dec 31 there is? I disagree. Distributions must begin in 2007, as I described. If she rolled over the entire amount some of it was ineligible.

You're right to question it; I don't get it either. Maybe there is a simple answer that we're missing, but, while you could at least in theory have that in a document, I think you'd have to do annual testing and I can't see it in a N/S prototype. I think I'd call the document provider and ask what we're both missing.

Ah, your original post said the money was moved shortly prior to the end of 2006. I think I get it now, the 12/31/05 val was done and money was moved out fairly early in 2006? That make some sense. I agree with the consultant that the most proper correction would be to reallocate as if the segregated money was part of the pooled account. I don't think it's really a 411d6 issue if this means that some other participants get less; it was just a correction of a mistake, not a true cutback. But it's a major hassle and if not a 411d6 problem an employee relations problem. The alternate solution sounds ok to me; I don't know about the official correctness of it but it's very reasonable.

Are you sure that definition of compensation doesn't just apply to matching contributions? I can't imagine trying to allocate contributions based on each pay period, and then testing would have to be done on annual comp...I have to think that's flawed language or maybe a misinterpretation.

Definitely issue a corrected 1099-R. The participant didn't say anything about getting a 1099-R but no money?! If your plan permits segregation to a separate account, and it earned interest, I would just pay out that amount now. If forms were already signed, I would issue a new Tax Notice and complete the payment. (Again, the participant never asked about the payment, if forms were signed?) If the plan didn't permit segregation, then I might agree to just include those accounts as part of the general pooled fund and allocate earnings of the general fund to the terminee...for 2007, if the money stays in through the end of the year. If it was moved to a segregated account late in 2006, then I have to wonder why then and not early in 2007, when the amount to pay was known, and I have to wonder if the participant was credited with gains from the general account for 2006. But if they were credited general gains for most of 2006, then I wouldn't bother fixing a little bit of gain for the rest of that year. (But it doesn't make sense to me that the participant would be credited gains for most of, but not all of, 2006, at least in the context of the way we run a plan.) I'm not sure that the failure to pay is such a problem that it needs restorative payments and/or re-allocation of gains or losses. But the whole segregation timing and allocation thing isn't quite clear to me either.

Excess deferrals are reduced by excess contributions previously distributed (look for a thread in this forum titled "ADP Test Issue" from Oct 2005 - but that references old regs). I think this is the cite but I have to admit this is giving me a bit of a headache: Reg § 1.402(g)-1(e)(6). So yes, I think you can exceed the 402(g) limit and the excess doesn't get refunded if it has already been refunded for a failed test.

The more I look at this, the less I see what the Roth conversion has to do with anything - it's a separate issue entirely. Unless the goal is really to do a $200,000 Roth conversion, but effectively reduce the tax bite to $100,000, which isn't working.

jevd's right, proper reporting would be to reflect earnings on the improper rollover. I was thinking that since the RMD could have been taken as late as Dec 31, no harm/no foul, but yes, it did become an excess rollover the minute it went in. Much marginal hassle for little marginal accuracy; oh well, we're all used to that, right?!

I think there's a flaw in your reasoning that results in an RMD from Fred's account being skipped entirely for 2007. I'm pretty sure this is how it should shake out: Fred's required beginning date was 4/1/2007, so he must have turned 70 1/2 in 2006. He died in 2006, before his RBD, so Wilma must begin distributions from his account by the later of 1) the end of the calendar year immediately following the calendar year in which he died [he died in 2006 so that would be 2007], or 2) the end of the calendar year in which he would have reached age 70 1/2 [2006]. So RMDs to Wilma from Fred's account are required to begin in 2007. That means that if you rolled over the entire account, you should have first processed an RMD (based on her life expectancy in 2007, the year following his death) and then rolled over the rest. No worries; the fix for an ineligible rollover is to just take out the ineligible portion (i.e. the RMD) from her account. I think I'd just show the full rollover and then report the taxable RMD from her account (I believe correct reporting would show an RMD taxable amount from his account and the balance rolled over, and a non-taxable distribution of the excess rolled over coming back out of her account, but that's likely to get messed up and ultimately if you just take it from her account you get where you want to from a tax standpoint).

I understand and agree with TLGeers' points, but if they happen to not matter for a particular plan or the sponsor wants to barge ahead anyway, if you terminate a 403(b) plan, what exactly does that mean (with respect to individual annuity contracts)? Does each individual have the right to "do" something with his or her contract, whether it is surrender, roll to a new plan, or roll to an IRA? Considering that the contracts are just sort of "out there" without much if any supervision by the employer, does the employer tell the investment company "oh yeah, the plan is terminated, deal with it" or ...? Thanks/sorry if I am naive in this area.

It seems to me that in order to distribute assets, you have to terminate the plan. And a requirement to terminate the plan is to bring it up to date...I was going to say there's no cite but I'm guessing the latest rev proc on amending plans says something like the deadline for amending for PPA is the earlier of plan termination or [whatever the remedial amendment date is]. Having said that, if you submit for a determination letter you'll have a chance to fix any omissions even if assets have been distributed.

There are a couple of things that I don't understand... exactly which entity is making the contribution? The partnership? Does the partnership have any income? What does it mean to ",,,take a distribution...into his Roth plan?"

I'd say that's a permissible time frame for a distribution. If he wanted to lock in a value he should have converted to cash himself; it's self-directed right up to the point of distribution.

True. Just be careful to make sure you and the accountant are on the same page. If any company contributions are being made during the year, and the accountant has already expensed them before giving you the profit figures, then you don't want to subtract them again. (Actually it's best to add them back in to start with net profits before any plan contributions in that case.) And you need to know if the partners or sharing expenses equally or not. At the end of the day you just have to do the best with what you're given.

I don't think it's the same employer; I think that's determined by the controlled group rules and each individual sole prop is not part of a controlled group with the new corporation. That said, I would still look to amend or restate one plan and merge the other into it, or do a new plan and merge both into it.

Thanks all. It appears that all contributions were made in 2005 and 2006, and the rollover occurred in 2007, so the s/l has not run on the contributions and there is time to properly correct the rollover.

A plan was sponsored by an ineligible organization for a couple of years and they just realized the error of their ways. They sent a participant a letter stating that he should include contributions and earnings in income, and that he can't roll over the account. Just to make it interesting, he's already terminated and rolled over the money to an IRA. My experience with IRA custodians is that there's no way they're going to just return the money and not report it as taxable income, so if he does the right thing and picks up the income from when he made the contributions, he'll wind up with double taxation, or at least the appearance of it, if he takes the money from the IRA and it is reported on a 1099-R. Any thoughts on both legal and practical approaches (by practical I don't mean illegal, just good ideas)? Obviously if he's going to take money from the IRA and not have it reported as taxable income, there must be a code for disgorging an ineligible rollover, but I can't find it.

I think, as noted by others, that you should review any codes of conduct as applicable and client agreements. Having said that, I seriously doubt that you have any legal obligation to report this to the government and if you do, you're probably exposing yourself to legal action by the client. That's a sad state of affairs but I think that's the case.

Yes. That's buying the policy and there is a prohibited transaction exemption for it.