Bruce Steiner

Let me try to explain this more clearly. The issue of separate shares does not arise if there is a spouse and the IRA goes either to the spouse or to a credit shelter trust of which the spouse is a beneficiary. The issue arises where there is no spouse. I'll use a simple example to illustrate this point. Suppose the IRA owner is the surviving spouse, she has two children, A and B, and she wants to leave both her IRA benefits and her other assets to or in trust for A and B (with the share of a deceased child going in trust for the deceased child's issue). Now for the drafting. If she names as her IRA beneficiary (if her spouse does not survive her) something like "my issue per stirpes, except that each person's share [the share for each grandchild or more remote issue] is to be payable to his/her trust under my Will," that avoids this "separate share" issue. But this can be a problem with some IRA custodians. To deal with these problem IRA custodians, one solution is to name as her IRA beneficiary (if her spouse does not survive her) something like "the Trustees of the Family Trust under my Will." To coordinate this with her Will, her Will could say that if her spouse survives her, she leaves the credit shelter amount to the Family Trust, and if her spouse does not survive her, she leaves her residuary estate to the Family Trust. If her spouse survives her, the Family Trust would contain the usual credit shelter trust provisions. If the spouse does not survive, the Family Trust under her Will (which in that case would also receive her IRA benefits) would be divided into separate shares for her issue. The result is the same. A and B each receive 50% of her IRA, with the share for a deceased child going in trust for the deceased child's issue. The only reason for this approach is to satisfy problem IRA custodians. As to whether IRA benefits are payable to children outright or in trust, it shouldn't really make any difference as far as the "separate shares" issue goes, assuming that it's otherwise possible to stretch out the benefits over the child's life expectancy when using a child's trust as beneficiary (a different issue which has itself been the subject of considerable discussion, and which does not appear to have been resolved by the final regulations). Some people provide for their children in trust rather than outright for various reasons, including protection against creditors, predators, estate taxes in the children's estates, and protection against spouses. These reasons apply equally well in the case of IRA benefits as other assets (except to the extent IRA benefits are protected from creditors even if they pass outright). While not related to the issue at hand, a credit shelter trust is not generally difficult or expensive to administer. Also, while there are cases in which an independent trustee is desirable, there are also cases in which the family is able to handle the credit shelter trust without an independent trustee, although if the spouse ever wants discretionary distributions (other than as limited by an ascertainable standard), she will need a co-trustee. Whether one or more of the children are appropriate co-trustees of the credit shelter trust will vary from case to case.

If this turns out unfavorably, it would really glorify form over substance, and make it harder to cope with the less flexible financial institutions. If there is no spouse (e.g., at the surviving spouse's death), most clients leave their estates (including their IRA benefits) to trusts for their issue. There are two ways to draft the beneficiary designation. One way is to designate IRA owner's issue per stirpes (if there is no spouse or the spouse does not survive), except that each person's share goes to his/her trust under IRA owner's Will (rather than to him/her outright). That works fine even under the more restrictive interpretation of the regulation. But some of the less flexible IRA custodians will have trouble with it. The other way is to name the credit shelter trust under the IRA owner's Will as the IRA beneficiary if the spouse does not survive. The Will can then provide (a) for the typical credit shelter trust if the spouse survives, or (B) if there is no spouse or the spouse does not survive, then the trust goes to the issue per stirpes, with each person's share held in further trust. The drafting would look be much like a revocable trust. The drafting isn't that much harder, and sometimes is necessary in order to deal with some of the less flexible IRA custodians. There is no logical reason for the result to be different, assuming that the trustees have no discretion.

If the penalty applies, the penalty is 10% of the prior distributions. In addition, since the penalty is imposed in the year of the modification, there is an interest charge, so as to put the IRS in the same position as if you had to amend the back years. I, too, have a ruling request pending on this issue, and I know of another lawyer who is about to file one. Given the Enron collapse, it's possible that there may be more such cases. To date, I'm not aware of any case law or rulings directly on point. While it may not be much consolation, the IRS has approved hybrid methods in which taxpayers used the amortization or annuity method (which generally permits substantially higher distributions than the MRD method), while also permitting the use of the current account balance each year. PLRs 200051052, 200020063. For more than you ever wanted to know about avoiding the penalty on pre-59.5 distributions, see my article on this subject in the January 2000 issue of Estate Planning. The lawyer who handles your estate planning should subscribe to this publication.

Yes, you can invest in a hedge fund in an IRA. We have helped several clients do this. There are some IRA custodians/trustees which will permit investments in hedge funds. There are some hedge funds which do not generate unrelated business taxable income.

Where is the lawyer who is handling the estate? That's what he/she is getting paid for.

I've had clients do this.

Will they tell us before year-end whether beneficiaries of persons who died before 2000 can use the new rules in calculating their required distributions for 2001?

An unmarried participant in a 401(k) plan had a loan outstanding at her death. She did not name a beneficiary, and under the terms of the plan, the default beneficiary was her estate. Most (though not all) of her account consisted of appreciated employer securities. Soon after her death, the plan treated the loan as being in default, and issued a Form 1099 in her name (not in the name of the estate). 1. What generally happens upon the death of a participant with an outstanding loan? 2. How does this affect the requirement of a lump sum distribution in order to elect NUA treatment?

Suppose instead the employee had a loan against her account, and after her death the plan satisfied the loan out of the account. If the beneficiary is taxable on the outstanding loan, does that prevent the beneficiary from taking a lump sum in a subsequent year? (Suppose the employee died before 70 1/2, the beneficiary is the estate, and wants to take a lump sum near the end of the 5 years to use NUA.)

With one exception (there may be others, but I'm only aware of one), there isn't anything you can do other than to pay the tax or not invest in foreign stocks. Under the US-Canada income tax treaty, there is no Canadian tax on dividends received by a qualified plan. But even this treaty does not contain a similar exemption for IRAs.

Prop. Treas. Reg. § 1.401(a)(9)-1 Q&A-2 says that: "The distribution rules of section 401(a)(9) apply to *all* account balances and benefits in existence on or after January 1, 1985. Sections 1.401(a)(9)-1 through 1.401(a)(9)-8 [the new proposed regulations] apply for purposes of determining required minimum distributions for calendar years beginning on or after January 1, 2002." The preamble to the new proposed regulations says: "For determining required minimum distributions for calendar year 2001, taxpayers may rely on these proposed regulations or on the 1987 proposed regulations." On the other hand, the preamble also says: "For distributions for the 2001 calendar year, IRA owners [note the absence of the word "beneficiaries] are permitted, but not required, to follow these proposed regulations in operation, ...." So the new proposed regulations are helpful as to required distributions for IRA beneficiaries, but not ideal. The more controversial issue has been beneficiaries of IRA owners who died before 2000. As to a beneficiary naming a beneficiary, the balance has to go *somewhere* if the beneficiary dies before receiving all of the IRA benefits. (Of course, except for benefits payable to spouses, I prefer to have IRA benefits payable in trust rather than outright, for the same reasons that I prefer to have assets generally pass in trust rather than outright.) But if you leave an IRA outright to someone other than a spouse, about half the time the person will die before his/her life expectancy. For the most part, it makes very little *substantive* difference whether the beneficiary selects the succeeding recipients by Will or by designation with the IRA trustee/custodian. But here the new proposed regulations are helpful. Prop. Treas. Reg. § 1.401(a)(9)-5 Q&A-7(d)(1) says: "If the plan provides (or allows the employee to specify) that, after the end of the calendar year following the calendar year in which the employee died, any person or persons have the discretion to change the beneficiaries of the employee, then, for puropses of determining the distributino period after the employee's death, the employee will be treated as not having a designated beneficiary. However, such discretion will not be found to exist *merely because a beneficiary may designate a subsequent beneficiary for distributions of an portion of the employee's benefit after the beneficiary dies."

I agree that if you have a pecuniary marital/credit shelter (the form I generally prefer for the reasons set forth in my article in the October 1986 issue of the Journal of Taxation), the IRA benefits passing to the credit shelter won't accelerate the IRD because it's the residuary share. It sounds like you shouldn't have any trouble getting a favorable ruling on the spousal rollover if the spouse has the power to withdraw all of the assets of the marital trust. While the issue is not entirely free from doubt, the spousal rollover ought to reduce the risk that the use of IRA benefits to fund the pecuniary marital might accelerate the IRD. If the spouse disclaimed his/her interest in the credit shelter, then presumably it passes to or in further trust for the children, in separate shares. So (ignoring the possibility that the children take in trust and don't have any issue so that each child is the presumptive remainder beneficiary of the other child's trust and thus may be treated as a measuring life), you may be able to use each child's life expectancy for that child's share. In this regard, new Prop. Treas. Reg. § 1.401(a)(9)-8 Q&A 2(B), which allows separate shares to be established by the end of the year following death, may be helpful. Again, keep us posted on how your ruling goes.

Whether using the IRA to fund Trust B accelerates the IRD has been the subject of some discussion and differences of opinion. Some people have argued that the income should not be accelerated because Sections 72 and 408 override the general principles of acceleration when using IRD to fund a pecuniary bequest. It is also possible that the IRS may not spot the issue. Obviously this issue could have been avoided in several ways. While fractional share estates are difficult to administer, the trust could have used a fractional share formula. Or there could have been a fractional share formula in the beneficiary designation, which would have applied only to the IRA. Or, what is probably the simplest approach (though a disclaimer trust is less flexible than a mandatory credit shelter trust), the IRA owner could have named the spouse as primary beneficiary, and the credit shelter trust as contingent beneficiary, and the spouse could have decided whether to disclaim. In addition to the issue of whether the IRD is accelerated, there is also an issue as to whether the portion going to Trust A will qualify for a spousal rollover. There have been numerous PLRs on this, some favorable and some unfavorable, which I discuss in my article in the October 1997 issue of Estate Planning.

You can easily have a discretionary trust. For example, most credit shelter (bypass) trusts give the trustees discretion to distribute income and principal to the spouse or the issue. In that case, assuming the spouse is older than the issue, the spouse will have the shortest life expectancy, and his/her life expectancy will be used. Perhaps under the new proposed regulations, the trustees will be able to divide the trust (for example, to set aside a portion for the issue, of which the spouse will not be a beneficiary) by December 31 of the year following the IRA owner's death. But the trustees might not want to do so absent the spouse's disclaimer or consent (after all, the purposes of the credit shelter trust are to keep the money available for the spouse's benefit, and ofte to permit the spouse to control the ultimate disposition of the trust assets to or in further trust for the issue upon his/her death. The real issue comes up where you have two or more children, and you leave each child's share to him/her in trust rather than outright. Let's assume that the trustees have discretion to distribute the income and principal of each child's trust to that child or the child's issue. Ignoring the possibility that the siblings could be treated as beneficiaries (if a child dies without issue), each child is the oldest beneficiary of his/her trust. If the division into shares takes place in the beneficiary designation, then each child's trust uses the life expectancy of the child who is the beneficiary of the trust (again assuming that the child is the oldest beneficiary of the trust, and that there is no other reason that you couldn't look through to the beneficiaries of the trust). To do this, you would need a beneficiary designation along the lines of "my issue, per stirpes, except that each person's share shall not be payable to him or her, but instead shall be payable to the trust for his or her benefit under my Will." Since some IRA trustees/custodians may not like this, because it won't fit in the small space provided on the beneficiary designation form, some people instead have a beneficiary designation that says something like "the family trust under my Will," and then provide in the Will that if the spouse does not survive, the family trust is divided per stirpes and held in separate trusts for their benefit. There is no economic difference whether the division takes place in the beneficiary designation form or in the Will (or other trust instrument). And even under the old proposed regulations, assuming each child was at least 10 years younger than the IRA owner, there wouldn't have been an opportunity to manipulate the system even if IRA owners had to take distributions from separate IRAs or separate shares pro rata (which wasn't even the case). But there were some PLRs before the new proposed regulations that drew this distinction, at least where the IRA owner died after the required beginning date. Since reg_h2b is the lawyer seeking the PLR, I'll let him/her do the research to find them. In most cases, if the children were close in age, it may not have been worth the effort to create a more complicated beneficiary designation form and look for a financial institution that would accept it. Presumably the new proposed regulations will give the fiduciaries until December 31 of the year following death to set up the appropriate beneficiary accounts. Reg_h2b: Please keep us posted on your PLR.

Drafting is often the easier part of the lawyer's job. The harder part of the lawyer's job is often the decision making process -- applying the law to a specific set of facts, making the legal judgments and making recommendations to the client. Indeed, for a number of reasons, estate planning tends to be a particularly difficult area of law. This does not in any way take anything away from the role of other advisors, be they actuaries, accountants, investment advisors, insurance advisors, bank trust officers, or others. The client's other advisors can be helpful in the estate planning process. Often one or more of the client's other advisors are the catalysts in getting the client to deal with his/her estate planning. But ultimately the client has to sit down with his/her lawyer and develop (and periodically update) his/her estate plan. The curriculum in law school is geared more to analyzing cases than to the practical aspects. That's why I said that if Ryder wanted to become knowledgeable in tax/estates, after going to law school and getting an LL.M. in tax law, he/she should work for someone (most likely in the tax/estates department of a law firm) for a while.

With all due respect to the accountants, they're not necessarily tax/estates lawyers, so it's quite possible you know more about tax/estates law than many accountants. Law school is 3 years full-time, or 4 years part-time. Then you should get an LL.M. in tax, which takes 1 year full-time, or 3 years part-time. Then you'll have to work for someone for a while. So it probably will take you more than 8 years to become proficient in tax/estates law. If it's any consolation, it would take at least as long to become proficient in certain medical specialties.

Estate planning for retirement benefits can be quite complicated, and your clients should be getting their legal advice from competent counsel, not "sales representatives." I think that among the biggest challenges and issues are Roth conversions for large retirement benefits, and what to do when the client does not have sufficient nonretirement assets to fully fund the credit shelter trust.

It would be more efficient to first determine what you are trying to accomplish, and then to look for (or ask the lawyer who handles your estate planning to recommend) how best to accomplish your objectives, rather than asking about the merits of a particular technique.

Barry, I know. I tried to put a "g" in brackets after each sentence of my message, so that you wouldn't take it seriously, but they didn't appear in the posted message. I was trying to call attention to the fact that many people make a great effort to try to get under $100,000 in the year they convert to a Roth IRA, but might generally be over $200,000, so they'd have to pay the $2,250 IRS user fee rather than the $600 user fee. In either case, I would expect the legal fees in obtaining a PLR to be more than $2,250.

I guess it's easier to get one's income under $200,000 to qualify for the lower user fee for a PLR than to get one's income under $100,000 to qualify for a Roth conversion . But I assume that income from a Roth conversion counts for purposes of determining the user fee for a PLR even if it doesn't count for purposes of qualifying for the Roth conversion .

Surely there were far more important issues than the reporting requirements, especially now that for most people, the minimum required distributions will be based upon the old MDIB rules, which are easy to calculate based upon the IRA owner's age. Did anyone address any of the other issues?

A portion of the distributions will be income for trust accounting purposes, and a portion of the distributions will be principal for trust accounting purposes. The allocation between income and principal may vary depending upon state law. This is relevant in the case of a marital (QTIP) or "A" trust, since the spouse must receive (or at least have the right to demand) all of the income of the trust. But in the case of the credit shelter or "B" trust, it shouldn't make any difference how much is income and how much is principal for trust accounting purposes, since distributions from a traditional IRA are all taxable (ignoring any basis if any nondeductible contributions were made), and distributions from a Roth IRA are all nontaxable, and if the Will is well drafted the trustees will have discretion to distribute both income and principal or to accumulate income, taking into account income taxes and all other relevant facts and circumstances.

It's hard to know what your parents should do without knowing what they were trying to accomplish. Living trusts are tax neutral. They are helpful in some cases (for nontax reasons), but in most cases they don't make the estate administration significantly easier, and (as may be the case here) they merely serve as a distraction. Depending upon the terms of the trust, it may be possible for the survivor to do a rollover. There have been quite a few PLRs involving similar situations. Some of them permitted a rollover, and some of them didn't. I analyzed this in greater detail in my article on this subject in the October 1997 issue of Estate Planning. Your lawyer should subscribe to this publication. Notwithstanding the loss of the income tax benefits of a rollover (and the potential Roth conversion in the case of a traditional IRA), some clients leave IRA benefits to the spouse in a marital (QTIP) trust rather than outright, to control the ultimate disposition of the principal. You didn't say whether this was one of your parents' objectives. A more common plan would be to have Wills with a marital share (either outright or in trust) and a credit shelter trust, and for the spouses to leave their IRAs to each other. However, there may have been particular facts and circumstances in your parents' case that caused their attorney to recommend and them to select the plan that they selected. Bruce Steiner, attorney NYC and Hackensack, NJ also admitted in FL

Scenario 4 is the one I had in mind. It could make a big difference if child could use his/her life expectancy, rather than his/her parents' life expectancies. Even if child were the beneficiary from the start, the calculation is more favorable under the new rules, though the difference is not as great as in Scenario 4.

Barry: have you changed your view on this? Prop. Treas. Reg. § 1.401(a)(9)-1 Q&A 2 says that the distribution rules apply to all benefits in existence on or after 1/1/85, and the new proposed regulations apply for purposes of determining required minimum distributions beginning in 2002. The preamble says taxpayers may rely on the new rules for 2001. But you were the one who raised the question as to whether the new rules applied to beneficiaries of participants and IRA owners who died before 2000. Perhaps the IRS will clarify this before the end of 2001, so beneficiaries will know how much they have to withdraw this year.