Peter Gulia

I’d like to learn more about what TPAs and recordkeepers do in processing distributions under a retirement plan that includes Roth and non-Roth amounts. Do you permit a participant (or a beneficiary) to specify his or her preference on which portion of a distribution is taken from Roth amounts? Do you require a claimant to direct an allocation between Roth and non-Roth amounts? For example, I have a client that’s considering whether its plan documents should specify that a claim will be denied unless the claim specifies the claimant’s allocation between Roth and non-Roth amounts. If a participant isn’t permitted or fails to specify a preference, what ordering rule do you use to allocate a distribution between Roth and non-Roth amounts? What were your reasons for choosing that ordering rule? How much of the ordering rules is in the plan document, and how much is the plan administrator’s interpretation? Is there any particular kind of distribution for which you don’t permit a claimant to request his or her preferred allocation between Roth and non-Roth amounts? If so, why do you preclude the choice? If a qualified domestic relations order doesn’t state details on which portion of the alternate payee’s portion is to be credited as Roth amounts, what ordering rule do you use? Is that rule stated in the plan document? Stated in the QDRO procedure? A rule you interpret for practical administration?

Those who think that the IRS might be overreaching in regulating tax preparers (and I don't express a view) might want to read your neighbors’ reactions to my not-so-hypothetical in “Must a Form 5500 preparer tell her client that the plan is tax-disqualified?” http://benefitslink.com/boards/index.php?s...c=46333&hl=

I based the hypothetical on a real situation, but with some facts varied to further protect my client’s (and her prospective client’s) confidences and to sharpen the tensions involved in similar situations. (As some of my friends know, a significant part of my law practice is advising a lawyer, actuary, or certified public accountant about his or her professional conduct.) In the real situation, the practitioner had not yet accepted the Form 5500 engagement, but through her procedures for vetting prospective clients already had knowledge of the plan-document defects. I posted the hypothetical after I had advised my client, hoping to learn whether my advice was or wasn't consistent with community norms. Moreover, I believe that the problem illustrated by the hypothetical is frequently recurring, at least for those who work with small-business employers. What did I advise? Although my client asked about using her engagement letter to limit sharply the scope of the representation, I advised that the expense, delay, and difficulty of getting the prospective client’s informed consent to such a scope limit would be disproportionate to the nature and normal fee of the proposed Form 5500 engagement. (Although this client has no professional credentials, my general advice to her is to follow conduct principles and 31 C.F.R. Part 10 as though she belonged to a recognized profession.) After my client told me that she was certain that the business owner would refuse to correct the plan documents, I suggested that she politely decline to accept the proposed engagement.

A somewhat related question: Do some of the businesses that do a "dependent eligibility audit" do so on a contingency fee; for example, $yy multiplied by the number of ineligible persons removed?

Thank you everyone for a nice collection of useful ideas.

rcline46, thank you for helping me think about a business-conduct problem. When you refer to a duty even if law doesn't impose a requirement, do you mean a moral duty or something else? Does it matter that this practitioner might not be an actuary, lawyer, or certified public accountant, and might not have any license or credential? Would it affect your view if the engagement letter had stated explicitly that the practitioner would not do anything beyond what's absolutely necessary to prepare the Form 5500? Please understand that I also think that the practioner must mention the defect, but I seek to understand more about why we feel that way.

I’d like to know what the BenefitsLink mavens think about this hypothetical: A corporation (100% owned by one shareholder) engages a practitioner to prepare a small retirement plan’s Form 5500. The plan has never had an audit or any level of CPA service concerning the plan’s financial statements. For every previous Form 5500, the financial reporting was on the cash-receipts-and-disbursements method of accounting. The practitioner’s engagement letter says that she may rely on information furnished by the employer or any financial institution unless she has actual knowledge that the information is false. Although the practitioner didn’t ask for it, the corporation furnishes to the practitioner a copy of the plan’s document. Against good business judgment, the practitioner doesn’t send it back with a letter saying that she didn’t look at the document. Rather, she does glance at it, and immediately notices that the document hasn’t been amended for several law changes that were required to be in the plan’s document before the year to be reported. Based on the plan’s financial information and consistently using the cash method of accounting, it’s possible to answer truthfully every required item of Form 5500 without ever mentioning that the plan isn’t tax-qualified. The practitioner would prefer not to say to her client that the plan is tax-disqualified. Why? She believes that mentioning the point would lead to unbillable telephone time with the client’s owner. Leaving aside the wisdom of the practitioner’s reluctance even to mention what she noticed, is there any Treasury department rule or other Federal law that makes it improper for her to prepare the Form 5500 without mentioning the tax-qualification defect?

On the query about whether a son is or isn't some kind of "dependent", the new rule says: Restrictions on plan definition of dependent. With respect to a child who has not attained age 26, a plan or issuer may not define dependent for purposes of eligibility for dependent coverage of children other than in terms of a relationship between a child and the participant. Thus, for example, a plan or issuer may not deny or restrict coverage for a child who has not attained age 26 based on the presence or absence of the child's financial dependency (upon the participant or any other person), residency with the participant or with any other person, student status, employment, or any combination of those factors. In addition, a plan or issuer may not deny or restrict coverage of a child based on eligibility for other coverage, except that paragraph (g) of this section provides a special rule for plan years beginning before January 1, 2014 for grandfathered health plans that are group health plans. (Other requirements of Federal or State law, including section 609 of ERISA or section 1908 of the Social Security Act, may mandate coverage of certain children.) On risking the hazards and inconvenience of litigation, there might not be a much better way to get a conclusion (if anyone challenges a denial). The plan administrator's lawyer will read the same statute, interim rule, and other law sources that Chaz and the rest of us read. Uncertainty about the right answer to a question isn't a sufficient reason for failing to decide a properly presented question. The alternative of not taking any risk on failing to offer coverage to adult children could mean taking a risk of harming the plan by providing coverage and payment methods that need not have been provided and in so doing worsening costs or benefits for everyone else. If not required by the plan's documents, risking those harms might be a fiduciary breach. A plan administrator's task often involves interpreting ambiguous law. That's why courts defer to an administrator's interpretation unless it was so obviously wrong that it must have been an abuse of discretion.

On Chaz's hypo, perhaps the plan's administrator might send son a claims-denial letter that explains the administrator's discretionary interpretation of the plan that only an employee may pay (and only by wage reduction) for the non-employer portion of any cost of coverage. This claims-denial letter would describe the plan's procedure for review of a denied claim. If the son seeks review, the next denial would explain that he has ehausted the plan's claims and review procedures (and so may sue in Federal court). One imagines that the probabilities are against son finding a lawyer who would successfully argue that the administrator's interpretation was an abuse of discretion.

Even if the target is confident that she could prove that she isn't and wasn't an administrator, is it possible that she was a trustee (or otherwise was a fiduciary)? Did the corporation's shareholder (and, one guesses, a director) have power to appoint or remove the administrator, arguably making the shareholder a fiduciary to the extent of that power? If the DoL could find any fact to argue that the target is or was a fiduciary, imagine that the DoL could argue that a fiduciary who has knowledge of a co-fiduciary's (the administrator's) breach has duties to take prudent steps to correct or remedy the co-fiduciary's breach. bzorc, I have some suggestions (turning on where the facts lead) on steps that the target's lawyer could take to persuade EBSA to withdraw its demand. For tactical reasons, it's unwise to show the ideas on a public website that anyone (including government people) could read. Please feel free to call me

Even if one assumes that the investment house is wrong: If the investment house furnishes the prototype documents without a separate or incremental fee, might the business owner's expense in adopting a revision be relatively modest? If there are several clients in the same situation with the same documents set, perhaps a practitioner could allocate his or her reading time among the clients so as to bill each only an allocable portion of that time, plus the advising and hand-holding time that's particular to each client. Shouldn't a practitioner prefer that a client make its choice about whether it wants whatever arguable protection could be gotten from doing the revised documents. Conversely, in the absence of an IRS determination letter or clear reliance on a prototype, it might be troublesome for a practitioner to assure a client that an absence of a revision couldn't lead to a disqualification.

My suggestion is based on an assumption that the business owner prefers to file Form 5500-EZ on old-fashioned paper. In my limited experience, a paper filing not only reduces what information gets into the computers but also slows down the agency's processing of the return. I've handled situations in which a taxpayer, even after responding to a specific deficiency notice, never heard from the IRS again - thus practically escaping several years' penalties.

The fillable .pdf form on the IRS website allows typing dates in the first line. In the clean-up situation that you describe, I'd use 2009 forms for the old years' returns.

I’d be reluctant to advise a client to omit any year’s return based on a statute-of-limitations presumption. For the plan’s trust, it’s the act of filing a Form 5500 (including Schedule P for those years that called for it) that starts the running of the statute of limitations. Neither a three-year nor a six-year limitations period applies if the trustee hasn’t filed a return. IRC § 6501©(3). [With the right facts and circumstances, it’s possible for a trustee’s Form 5500 that omitted a Schedule P to start running a limitations period. See Martin Fireproofing Profit Sharing Plan and Trust v. Commissioner, 92 T.C. 1173 (1989). But I don’t know how one does it without any return at all.] A client might consider omitting a year’s Form 5500 if she knows, or in good faith estimates, that plan assets were less than $250,000 or $100,000 at the relevant time. But despite that reporting relief, many plan trustees feel more comfortable filing a return to trigger the running of limitations periods. Once a trustee (with the work of her trusty CPA) is reconstructing enough of the plan’s and trust’s records to file Form 5500 for some of the years, it’s not too much more work to pull together returns for all years. Moreover, looking at the year-to-year internal consistency of the whole timeline can be a practical way to spot errors or illogical estimates. If the records are incomplete, I’d focus more energy on making the contribution numbers consistent with those shown in the business and individual tax returns. Along with this, I’ve seen returns in which the beginning and ending balances were disclosed as estimates, and the year’s investment gain or loss likewise was disclosed as an estimate interpolated as a subtraction of the contributions and the estimated beginning balance from the estimated ending balance.

Understand the reasons for a service provider to get rid of the beneficiary designations. My further thinking-out-loud is about whether a service provider should or shouldn't present a particular suggestion on how the plan administrator keeps the records. For the plan administrator that must keep the beneficiary designations, might there be a better filing system than putting the beneficiary designations mixed in with employer records that aren't the retirement plan's records and are likely to have different retention times? Is this at least a question that the plan administrator must or should evaluate? A service provider that isn't a plan fiduciary usually doesn't have a duty to give uncompensated advice to the plan administrator. But could a send-'em-back letter suggest that the plan administrator ask for its employee-benefits lawyer's advice about how to keep plan records? Would such a suggestion, even if ignored, be a stronger defense for a service provider than a volunteered too-specific suggestion? Or is there a business problem that a service provider's client will be offended by a suggestion that it might need or want advice about how to keep plan records?

Bill Presson's and austin3515's ideas that a service provider might prefer not to have possession of records (and get rid of any that are unwelcome) make sense (at least for a business that chooses not to provide a service of checking the internal consistency of the beneficiary designations). Should we be concerned that putting a retirement plan beneficiary designation in the participant/employee's personnel file (if someone doesn't render more guidance) might result in mistakenly destroying the record sooner than ERISA permits, and possibly sooner than the record could be needed to decide a claim to a death benefit? Many employers destroy a personnel file about four years after the former employee's employment ended. Because a retirement plan might not require a beneficiary to take a death benefit until five years after the participant's death, could a beneficiary designation that's amid a whole personnel file get scrapped before the retirement plan's administrator needs to look at that record?

Without remarking on which is a good or bad practice, my experience is that some small TPAs do check beneficiary designations, and big recordkeepers don't. Along with Ted Triska's query, some of us might be interested in a related question: assuming that a service isn't free and might be paid for from the plan's (rather than the employer's) assets, is it a smart use of plan assets to check beneficiary designations? Why or why not?

The link is to the public version of unannotated Pennsylvania Statutes that West Publishing provides for free under its contract with the Commonwealth. But some computers and Internet browsers block it as a potential security threat unless you revise your security settings to allow cookies and treat Westlaw sites as trusted. If that's inconvenient or ineffective, the attached .doc files are downloads from that site. The annotated Lexis or Westlaw editions don't provide much more. 72_PS_4521_1.doc 72_PS_4521_2.doc 72_PS_4521_3.doc 72_PS_4521_4.doc

An S corporation maintains a defined-benefit pension plan for its only employee, who also is the corporation’s only shareholder. The plan provides a pension that’s designed to meet exactly the IRC § 415(b) limit. This business owner has flexibility in setting her salary: for example, she might pay herself as little as $50,000 or as much as $150,000. (For this inquiry, assume that she could defend anything in that range as no less than, and no more than, reasonable compensation for the owner’s leadership of the business.) If feasible, this hypothetical client would prefer to get an actuary’s work only once for a year, and before she decides how much salary she wants to pay herself for the year. Moreover, deciding how much income to devote to pension funding rather than other investments is a part of the business owner’s financial planning. Assuming that all other amounts and facts are constant and the only variable is the participant’s salary, could it really be as simple as saying that the amount needed to fund the current year’s accrual of the pension varies proportionately with the salary? Would this funding amount needed on a salary of $100,000 be simply double the funding amount needed on a salary of $50,000? My small brain worries that the idea that funding falls in a line following the salary is too facile. But I’m hoping that the BenefitsLink mavens can show me why it isn’t that simple.

Matt, at least one of Kathy Moran's questions concerns whether the township has power to pay a contribution beyond those elected as a salary reduction. It's a serious question because a State's law might not empower a municipality to provide a matching or nonelective contribution. A political subdivision has only as much power as the State granted. (In more than a quarter-century of working with governmental deferred compensation plans, I've seen many that were void.) Although in an ideal world a municipality would check State and local law before doing a new act, it's not unheard of for a municipal official to neglect asking for a lawyer's advice.

Concening a Pennsylvania political subdivision's power to provide deferred compensation, the relevant statute is at 72 P.S. § 4521.1 to -.4. For a free version, http://weblinks.westlaw.com/result/default...TC&vr=2%2E0. The township should want the advice of its regular and expert lawyers about possible constructions and interpretations of the statute. Remember that a political subdivision's act that exceeds the power that the State granted to the political subdivision is void. Please feel free to call me if you're interested in my views.

A township of which State? State laws vary widely concerning whether a contribution beyond elective salary-reduction deferrals is permitted or precluded. Did your client explain anything about why it wants a matching contribution to a retirement plan to relate to a contribution to a plan that, at least in form, states its purpose as higher education?

The most straightforward way is to work from the proposed rules [in the Federal Register of August 6, 2007]. The proposed rules include a subparagraph captioned "Written plan requirements" [page 43947] that states ten requirements. The proposed rules use the word "optional" to refer to plan provisions that are permitted but not required.

Has the plan's administrator asked the employer whether the name, TIN, and other identifying information presented for the potential distribution is consistent with the information presented for Form I-9 and on Form W-2 for the years that contributions were made? Also, consider that a database might report one or more addresses with little or no background information on who reported the address or how it was associated with a name or TIN. Likewise, the IRS and other government agencies receive submissions in which a filer reported an address other than the subject's correct address (or a name other than the subject's name).

Thanks for the kind words. It's also wise to remind the business owner to get his or her lawyer's advice about possible disadvantages of a one-participant plan, including that it might not be ERISA-governed and so might result in less protection against creditors. If that protection, or some other value from ERISA governance or preemption, is of value to the business owner, he or she might design the second plan so that it includes one trusted employee beyond the owner (and owner's spouse).