MGB

Section 643, "Rollovers of After-Tax Contributions." In (a), it adds statements at the end of 402©(2), which in (A) allows it to go to a qualified DC plan (which agrees to separately account for it) in a trustee-to-trustee transfer; and in (B) allows it to go to an "eligible retirement plan described in clause (i) or (ii) of paragaraph (8)(B)." This last reference (402©(8)(B)) is referring to (i) an IRA and (ii) a 408(B) individual retirement annuity.

The IRS only posted the adjusted numbers in an Information Release; it did not require their action. The Code itself defines what the new numbers each year are and no action by the IRS is needed (although the rounding of percentage adjustments to a certain number of decimal places came from procedures established by the IRS through their Information Releases). The Cal. (and numerous other states) statutes refer to the Internal Revenue Code as it stood at a particular pre-EGTRRA date (e.g., 1/1/1999). The Code at that time stated how to calculate inflation-adjusted figures into the future. Therefore, the numbers are still around, and yes, the 2002 402(g) limit would be $11,000 without EGTRRA (they will start to diverge next year unless inflation picks up dramatically).

I do not have the example in front of me (it is under one of these stacks), but I can "guess": They are in the 15% tax bracket and otherwise would owe $3000 in taxes. They contribute $4000 and get a $2000 tax credit. However, they have also done it on a pre-tax basis, reducing their taxable income. That reduction saves 15%*$4000=$600 in taxes. 3000-2000-600=400

As Pax states, it is the termination of the plan, not the bankruptcy filing that triggers the 100%. Many companies file for reorganization (a type of bankruptcy) and come out of it as a viable employer again. This would not trigger the 100%. However, in many bankruptcy situations, numerous people get laid off. The termination of a large group of participants can be viewed as a "partial plan termination," which also triggers 100% vesting. The laws on partial plan terminations are very vague and whether or not one has occurred has only been settled in courts. But, given all of the history of court decisions, if a large proportion (e.g., 30%+) of nonvested participants have been laid off, you have one. It does not have to be one layoff, it can be multiple. However, all of the layoffs need to be associated with the same "event" (heading into bankruptcy).

Yes it is allowable to have more restrictions for HCEs than NHCEs. By coincidence, my employer (one of the top 10 consulting firms) uses exactly the limitations you describe for our own plan (I don't see the reasoning for limiting NHCEs to 25% - why would we limit them at all???).

I have asked this question repeatedly for over 20 years and have repeatedly gotten different answers depending on who I ask. The only "insight" I've ever gotten was from an ex-IRS employee that claimed that "flat benefit" is referring to a formula such as 50% of final pay, no matter what the service at NRA. Of course, this design has gone through so many changes since the nondiscrimination rules under TRA'86 that the description on the 5500 no longer makes sense. Unit benefit is the opposite: e.g., 2% of final pay times years of service. I presume the fixed benefit is a dollars-per-year type plan such as $20 per year of service (but I'd also consider this a unit benefit). For most plan designs nowadays, I can't imagine trying to figure out which of these categories various designs fit into. They need a category called "none of the above."

There is no such thing as "previously designated catch up contributions." All deferrals are regular deferrals until they exceed some limit no matter what the plan sponsor/participant calls them as they are made to the plan. It isn't until then that the excess becomes a true catch up contribution. If a plan sponsor is having the employees designate them as catch up contribuitons ahead of time (actually, my employer is doing this, too), that is only for their own internal naming of it; it has nothing to do with the law and whether or not it will actually become a catch up contribution. So the answer is yes, they are reclasified as a regular deferral. However, technically, they were never classified as a catch up contribution to begin with and were always regular deferrals.

Actually, EGTRRA did not do what you said it did. The law does not require any time length of nonparticipation following a hardship withdrawal. It does require the plan administrator to determine that there is an "immediate and heavy financial need" and that the distribution is needed to meet that need. In an IRS regulation, they created a safe harbor approach to this determination. Part of that safe harbor approach is to suspend the participation for one year (if they are in such dire need of the money, they presumably wouldn't be able to afford to defer more money over the next year). Note that this only applies if you are making use of the safe harbor approach; a plan need not do this if they have other procedures in place to determine an immediate and heavy financial need. EGTRRA did not directly change this. It contains a statement that directs the IRS to change their regulation to state that the safe harbor approach only use six months instead of twelve months (if the IRS decided not to do it, there is no way Congress could force them to do it). In Notice 2001-56, the IRS acknowledged that six months is now correct (even though they have not changed the regulation) for those using the safe harbor approach. In Notice 2002-4 (released about 12/20, for publication in an upcoming Internal Revenue Bulletin). the IRS rescinded the requirement that applied a single 402(g) limit to two years worth of deferrals around the distribution. Again, this is only for those plans making use of the safe harbor. Note that this is what is "allowed;" changing to six months is not mandatory, nor is the application of the two-year rule. If the plan document states that 12 months and the two-year rule applies, it needs to be amended with a good-faith amendment in 2002. This is not something that can be carried through to 2005 as part of the remedial amendment period because it is not a mandatory change.

This becomes a standard distribution that, once elected, is the forma of distribution in retirement. I.e., the participant must choose from all of the available forms (as long as the all meet the minimum distribution requirements) along with proper J&S notices, etc. The PVAB and the valuation of it are irrelevant; this is a defined benefit plan.

Yes and yes. But, if catchups are to be effective for the 2001-2002 plan year, the plan would need to be amended by the end of that plan year. My reference to end of year calculations was with respect to ADP testing. A person contributing $12,000 in total in 2002 across two plan years could conceivably have catch up in the second plan year without the first plan year having a catch up provision in the plan yet.

Taxable year is never the plan year. Taxable year refers to the individual's tax year, which is 99.99% of the time the calendar year. A noncalendar-year plan can institute catchups on 1/1 for the plan year that began in 2001. Note, however, that it is very cumbersome to coordinate the catchups between the partial years within the calendar year. (No one should have noncalendar year 401(k) plans - they create too many major administrative headaches.) Catchups are computed at the end of the plan year (it doesn't make any difference what you call the deferrals as they go in). Therefore, the only way to get a catchup in 2002 for a noncalendar year plan is to allow them in the 2001-2002 plan year.

The $35,000 is for limitation years both beginning and ending in 2001 (two years apply here because the $35,000 does not change under the old law's rounding to $5,000 increments). The $40,000 is only for calendar year plans in 2002. (What will be confusing is that this will reverse to "ending in 2003" when we get the 2003 figure (probably $41,000), which means the $40,000 will never be used by a noncalendar year plan.) Here is paragraph 3: The limitation for defined contribution plans under Section 415©(1)(A) is increased from $35,000 to $40,000 by Section 611 of EGTRRA effective for limitation years beginning after December 31, 2001. However, the limitation for defined contribution plans with non-calendar limitation years beginning before January 1, 2002, and ending after December 31, 2001, remains unchanged at $35,000. (See Q&A-9 of Rev. Rul. 2001-51)

Released today. Information Release 2001-115.

QSLOB rules have no applicability to the universal availability rules. Everyone in the controlled group must have them, you cannot limit to just a QSLOB.

I do not understand your analysis. Although the election makes the 403(B) an employer-provided defined contribution plan for 415, there is no longer a 415(e). So, the 403(B) should have no effect on the defined benefit plan and vice versa, even though it is an employer-provided DC benefit. If the employer wants to have a DC plan in addition to the 403(B) plan, that is a different matter. However, absent an actual C election, they should still not be aggregated and the 403(B) should be considered under the control of the employee. I originally thought EGTRRA changed this (as I wrote in our Legislative Information Bulletin issued soon after EGTRRA), but have since come to the conclusion that they are still considered separate from the employer.

What about safe harbor plans that will now use the match to satisfy the top-heavy contribution? I think this would be a cutback if not amended before 2002.

OBRA'86 and the proposed regulations under it made it clear this is not allowed (it was allowed before this).

Nothing actually changed under EGTRRA. EGTRRA ordered the IRS to revise their regulation to change the 12 months to six months (this provision is not in the law, only in an IRS regulation). Until the IRS actually changes the regulation (they are like an obstinate child when it comes to the Congress telling them to do something - it won't happen soon), your questions are unanswerable. The IRS has not indicated what the correct answer is yet. I am sure that everyone will interpret this whatever way makes sense to them until it happens.

I agree. In a phone conference with Elizabeth Drigotas (one of the lead Treasury attorneys on the proposed regulations), she stated that administrative limitations that are not in the plan document do not qualify. An even bigger issue is that she stated that the Treasury believes that these kind of restrictions are illegal because they are not definitely determinable benefits. But, rather than broaching that subject, they left it cloudy in the preamble to the proposed regs. However, expect them to issue something in the future that says this practice is no longer allowed.

Well, I'll sort of vascillate one more time. The IRS publishes a number with the other numbers, that is true. But, we feel it is wrong and I know of at least one large state that uses our "correct" number. The pre-OBRA'93 number was an unrounded number based on 4th quarter CPIs. The change in OBRA'93 to $150,000 also changed the adjustment period to 3rd quarter AND created a $10,000 rounding. There is also the statement that governmental plans can continue to use what they were using before if it was in the plan on July 1, 1993 (some states had no limit and are still operating that way). However, there is no reference that anyone using this grandfather should make use of any change in the indexing or rounding. In the releases from the IRS, they use 3rd quarter indexing (retaining the original 4th quarter base period, though). The reason I know this is I have a 5-page memo from Jim Holland from 1998 that shows the unrounded number for 1999. That is exactly what a 3rd quarter calculation produces. Then, the releases round it to $5,000. If it were rounded to $10,000, I could see where they are coming from...they are applying the post-OBRA'93 adjustment and rounding rules, even though OBRA-'93 did not say to do that. With the use of a $5,000 rounding convention, it makes no sense what they are doing. Without there being any guidance from them on why they are doing this, we have chosen to ignore it and continue to produce a 4th quarter unrounded amount for clients that are using it. Using the IRS's methodology on 3rd quarter, the unrounded amount is $295,460, which they will round to $295,000. Although that is what they will release, we will not use it. (Note the $295,460 here is greater than the $295,340 in my earlier post because the CPI went down from September to October.) In our view, technically, if a governmental plan makes use of the rounded 3rd quarter numbers that the IRS produces, they are not following the terms of the plan as of July 1, 1993 and therefore are no longer eligible for the grandfather.

I apologize for the mistaken post (I almost got through the year without a mistake!), this number is not released with the others (they usually come out in late October after the September CPI release). This number is based off of Oct-Nov-Dec CPI, so the number is not calculable until late January (the December CPI is released about January 18). If Nov and Dec CPI remain unchanged from the Oct CPI, the number would be $295,340, up from $289,260 in 2001. I would expect it to be slightly higher.

The GUST remedial amendment period for ALL plans was extended to 2/28/02 due to the terrorist attacks. See Rev. Proc. 2001-55.

Not yet. They issue all of the CPI'd numbers in one release. There are technical problems with a handful of others (e.g., minimum compensation for a SEP) that don't have correct references to a base period. They have decided to go forward with the release even though these issues remain, but it may take a couple more weeks.

This was covered in the 2001 Gray Book (IRS Q&As) from the Enrolled Actuaries Meeting. You officially continue on with aggregate. That is the funding method at this point. However, if you want to go back to FIL, that is a change in method that requires approval.

One other thing to note: You can fill out this certification with any provider of documents. You do not need to actually adopt the same document that is described in this certification. The certification just buys you additional time. So, if you can find another provider that knows what they are doing (e.g., Kemper again) and can fill this out with them, you can still come back and use the Smith Barnery document later. Even individually-designed plans can make use of this certification. Before the time limit expires for actual adoption (one year after the volume submitter gets its letter), the plan sponsor can revoke the certification and adopt any plan document.