My 2 cents

Nothing formal. Wouldn't there be diminishing returns - most of the people inclined to cash out would have done so, and most of the people who declined to cash out would probably tend to continue to do so. What you may experience would be heavy anti-selection. That is, of the people who would not cash out two years ago, anyone whose health has deteriorated would be among the few choosing to cash out now, leaving an unusually healthy cohort behind. Greater than normal longevity should perhaps be expected for those who have stayed behind. there be before the sponsor's accountant decides that future windows should be built into the accounting calculations?

Is it safe to say that the plan sponsor, as the plan's fiduciary, is (a) wasting no time and (b) sparing no expense to verify the accuracy of the account balances? Seems to me that the burden of proving that the amount in any given account is not correct belongs entirely to the sponsor/plan administrator. The last thing the sponsor should be able to do is question whether the amount in someone's account actually belongs there (and, based on that assumption, declining to comply with withdrawal requests) without having first had the account balance reconstructed from payroll deductions and investment returns. Also, is the embezzler so stupid and unimaginative as to have put the money into actual accounts with the intention of trying to access it later? Not even using the names of minor characters from Harry Potter or John le Carre novels? It is, throughout, apparently being assumed here that this is not a conspiracy, where the account holders were in on it all for a cut of the stolen money. If the employee requesting the withdrawal is not personally under suspicion, how can the sponsor refuse to just release whatever is in the account?

I am only really familiar with the DB world. In the DB world, one can satisfy the RMD by starting lifetime benefit payments by the April 1 following the year of attainment of age 70 1/2 (with increased periodic payments starting with that year's 12/31 and any other 12/31 if additional benefits will have accrued). I suspect that the circumstances under which RMDs are required are the same for both, even if acceptable ways to deal with RMDs may differ. And IRAs are subject to slightly different RMD rules than either DB or DC (such as it not mattering whether one is a 5% owner or still employed).

Is this HR department exempted from the requirement that the terms of the plan be followed? It shouldn't be a matter of meeting a notice requirement. If for a given year, it is determined that the plan cannot pass testing unless the HCEs are limited to 8%, then it must be done (even if prior salary reduction amounts have to be disgorged by the plan, at which time, one presumes, the affected individuals would be fully informed), but such a limitation must either come out of an explicit plan provision OR the results of a test that cannot otherwise be passed. It may be OK for HR to "encourage" HCEs to keep their contributions down to 8%, but without a plan provision to that effect or a test result forcing that, HR should not be able to enforce such a limitation.

On occasion, TPA firms can get it wrong. My understanding* is that the person having reached age 70 1/2 in 2016, there must be a 2016 RMD taken out of any total distribution made from the DB plan on or after 1/1/16. Further, as the distribution is to be made in 2017, the 2017 RMD must also be carved out. So both the 2016 and 2017 RMDs must be taken from the terminating plan and not be allowed to be rolled over to an IRA. * I have been known to sometimes get things wrong.

Out of curiosity, what needs to be corrected? Is it bigger than a breadbox?

Strictly speaking, it's not a seven-day yield, it's an annual yield based on projecting what happened in the last week out for 51 more weeks.

Is it even possible to put that many zeroes between the decimal point and the first non-zero place (i.e., 7-day yield = 0.000000003%)? Why would one want to show such a figure?

Seems to me that if you need to bring in a terminated employee to pass (and, even then, give them a contribution of 90% of pay!), perhaps you are cutting things a bit tighter than may be workable.

Inevitable question: What does the plan say? If it is clear, do what it says. If not, then the plan administrator must make a formal interpretation. Did nobody think of this when the plan was drafted to exclude overtime?

I may be mistaken, but I think the rule is that there would be an RMD for 2016 (even if the distribution is made in 2016 and the required beginning date is April 1, 2017). I don't think that it matters whether the distribution takes place before or after attainment of age 70 1/2. 2016 is the year of attainment of age 70 1/2 and the participant had separated from service, so there must be an RMD for 2016. It would be my understanding of the rules (imperfect as that may be) that you have to process an RMD and only allow the balance after that to be rolled over.

[semi-satirical comment] Would the selection of the independent qualified public accountant then be a fiduciary action, subject to litigation if the fees were higher than those of some other accountants?

Good luck on finding everyone!

Explicit mention in the instructions or not, surely they have to be reported on the Schedule H if they are not indirect expenses? Wouldn't "investment advisory and management fees" be the most logical place to put them?

I would expect that, as far as the regulatory authorities are concerned, the plan provisions governing loans ("allows too many loans and has constant battle keeping them straight") must be strictly followed, MONEY BEING NO OBJECT (like GMK, I apologize for shouting). Operational failures concerning the granting and administration of loans cannot be excused because it would have cost more to do it right. As I recall, the costs of plan disqualification are pretty substantial, and it seems to me also that a jury could well hold the plan fiduciaries liable if a participant claimed to have been damaged by the consequences that arose from the plan administrator having failed to restrict access to loans above the plan's limits. Think how Ford made out, having explicitly weighed the costs of fixing the design of the Pinto versus their assessment of how much would have to be paid out as a result of litigation.

It was my understanding that for purposes of implementing the Section 436 benefit restrictions when the plan sponsor is in bankruptcy, effective in 2015 NONE of the funding relief provisions can be reflected. That is, the AFTAP has to be 100%+ using the non-relief segment rates. One cannot use the originally schedule MAP-21 relief rates at all.

The IRS requires the first day of the plan year language in prototype plans, so for those plans, it would be the first day of the plan year of entry, without question. Many plans do not contain the "first day of" qualifier, and following the terms of those plans (which have not been denied determination letters) would permit forfeiture of benefits if the participant terminates after the first day of the plan year and the 5th anniversary of the actual plan entry date. My suspicion is that the original intent was to distinguish which entry date prevails if a person becomes a participant, ceases to be a participant (say due to broken service), becomes a participant again, and so on. The 5th anniversary should be measured from the very earliest date of entry. That would not expand upon ERISA's clear language to base NRA on the 5th anniversary of plan entry, merely clarify it.

Please note that the cited section of the regulation does NOT use the term "plan year", and the statutory language refers only to the anniversary of plan entry.

If they are excluded from the plan, then they are 3 HCEs not in any given rate group (I think). Example: Say there are 30 HCEs in all (including the 3) who are not excludable (exclusion from the plan does not make someone excludable*). If 15 HCEs are in a specific rate group, and these 3 plus another 12 (out of the total of 30) are not, then HCE coverage percentage for the rate group is 50% (15 in out of 30 HCEs). I think. *Only things like too young, not enough service, non-resident alien etc. make someone excludable. Being in a group ineligible for the plan does not.

The IRS has issued proposed guidance on 417(e) issues. Is it my imagination, or has the IRS rejected outright the conclusions of the 6th Circuit Court of Appeals that if a plan's pre-retirement death benefit calls for no forfeiture of benefits in the event of death before retirement (e.g., a cash balance plan paying the full account balance in the event of the death of the participant before benefit commencement) then no pre-retirement mortality discount is permissible under Section 411? The IRS seems to be saying that (a) Section 411 permits amending out death benefits other than the mandated QPSA and (b) accordingly, discounting for pre-retirement mortality is generally permissible since there is no requirement to value ancillary death benefits. The IRS does assert, however, that benefits attributable to mandatory employee contributions cannot be discounted for pre-retirement mortality. Are others seeing what I think I am seeing?

Did I hear somewhere that if the plan is under investigation by the DOL, DFVCP correction of the failure to file Form 5500's is not available?

If it has to be taken out as an RMD, isn't it inevitably a taxable distribution?

Wouldn't the compensation limit involve combining compensation from all controlled group sources? Ditto for 415 limitations, 401(k) limitations, etc.?

Or inaction, as the case may be!

Some of us only care about the 8955-SSA filings. The down period includes the filing deadlines for plans with plan years beginning between February 2 and April 1. And yes, this happened last year (at least). Naive question (since I am not involved with 1099-R filings): If one has the 1099-R filing ready to go before January 16 (is that even feasible?), how hard is it to shoot it through the FIRE system in the fortnight after it reopens? How quickly are the 1099-R forms themselves usually ready to go? I expect that the forms must be issued before the FIRE filing can be submitted. 5 weeks does seem to be a long time for "scheduled maintenance", however.