Bruce Steiner

If the IRA owner had left the IRA to the beneficiary in trust rather than outright, it would have avoided these complexities. The only complexity would have been that the stretchout would have been limited to the life expectancy of the oldest beneficiary of the trust.

If the IRA owner is still living, he/she should consult with tax/estates counsel. Based on what you describe, since the spouse is the oldest beneficiary, assuming the trust complies with the recent rulings (such as PLR 200228025 and 200235038), the spouse will be the measuring life. This plan will most likely destroy the rollover, destroy the possible Roth conversion after the IRA owner's death, leak the income to the spouse even if it's not necessary to elect QTIP for the entire IRA, and force the principal to the children, destroying the ability to keep it out of their estates and to protect it from their creditors. Here is a link to an article I wrote on trusts as beneficiaries of IRA benefits: http://www.bnatax.com/tm/tmm0903_steiner.rtf It appeared in the BNA Tax Management Estates, Gifts and Trusts Journal.

It's hard to tell without seeing your calculations. But it seems surprising that a 7 point difference in the tax rates would outweigh the benefits of the Roth conversion. What assumptions are you making as to the rate of return in your IRA, the rate of return after income taxes on your taxable account, your estate tax rate, how long you will live, the age of your beneficiary at your death, etc.?

Your ideas are quite good for someone who's not a lawyer. We've seen lawyers who haven't done as well. But instead of your trying to come up with the solutions, you should focus on your objectives and concerns, and let your lawyer advise you as to how best to accomplish your objectives. You must have some concern about your siblings if you are considering providing for them in the form of a 20-year annuity. and you want to make sure they don't cash in the IRA at once. But if you leave your IRA to them, you can't stop them from cashing it in (at least after your husband's death once they get the annuity). A better way to provide for them is in trust rather than outright, so the trustees will have some control. A 20-year annuity is extremely inflexible, and not very tax efficient. Again, it would be better to select the appropriate trustees, and give the trustees some discretion. If you leave the IRA to or in trust for your siblings, the stretchout is limited by their life expectancy. Is there some other way to get a longer stretchout? A life estate is also not very flexible. Suppose at some point your husband wants to move? Suppose he wants to refinance the mortgage (if there is one), or take out a home equity loan. Again, a trust is much more flexible. Finally, we rarely do codicils. To do a codicil, you have to make sure it meshes with the Will. After your death, everyone will both the Will and the codicil, whoever's interest was reduced by the codicil may be unhappy. If you do a whole new Will, no one would know what the changes were unless for some reason the prior Will came to light. Finally, a codicil is often done as a temporary stopgap, but once the codicil is done, redoing the Will may seem less urgent. So don't do a codicil; do a new Will. With the aid of word processing, it probably won't cost much more.

Decedent, unaware of the Section 691© deduction available to her beneficiaries, withdrew her entire IRA just before she died, to remove the income tax from her estate. The issue is whether her executors or the beneficiaries of her IRA can put the money back into the IRA, since the benefit of the stretchout would probably far outweigh the fact that the Section 691© deduction applies only to the Federal, but not the state, estate tax. Gunther, 573 F. Supp. 126 (W.D. Mich. 1982) says yes. Rev. Proc. 2003-16 (which lists death as an example) and the temporary regulations under former Section 4981A suggest yes. But the IRS said no in PLR 200415011. If the financial institution will let the executor put the money back (within the 60 days), and if the beneficiaries of the IRA are the same as the beneficiaries of the estate, the executor could put the money back, and then either seek a ruling (and take the money back out if the ruling is unfavorable), or run the risk of the excess contribution penalty. But if the financial institution won't let the executor put the money back, she'll need to get a ruling both on the rollover after death and the waiver of the 60 days. Even if the executor puts the money back without calling attention to the IRA owner's death, the financial institution may spot the issue when the beneficiaries set up their beneficiary IRAs. Does anyone have any thoughts, other than having the executor move to the Western District of Michigan? Was anyone in this group involved in PLR 200415011?

Why do you think that the dividends are exempt from Canadian tax?

Depending on the terms of the trust, the trustees may be able to distribute the trust assets to another trust for the benefit of the same beneficiaries.

Without knowing all the facts, it's hard to evaluate the transaction described in the New York Times article. It's hard to know what to make of the Swanson case. In one sense, it's strong, since it involved not the underlying issue, but rather the issue of attorneys' fees based on the IRS' taking an unreasonable position. But since the IRS conceded the underlying issue, we don't have the benefit of the court's analysis of the underlying issue based on the arguments of the opposing parties. There may also be special considerations for DISCs or FSCs. DOL advisory opinion 2000-10A is also a difficult one. Some people have expressed concern as to whether the DOL got the math right regarding the issue of 50% ownership, with attribution. In summary, the Section 4975 prohibited transaction rules are extremely complicated, and to try to understand their application to a particular case based only on a summary in a newspaper is not possible, at least for me.

There are lots of possible problems. But rather than trying to describe them all, it would be more efficient if the person would say what he/she is trying to accomplish, and then to determine how best to accomplish his/her objectives.

Assuming no change in the tax rates, John would be correct that it would come out a wash if you had to use the IRA money to pay the tax on the conversion. But that would not be the case if you had non-IRA funds with which to pay the tax on the conversion. Example, assuming a constant 25% tax rate: you have $100 in your IRA and $25 of other money. You convert and use your $25 of other money to pay the tax. Over some period of time, your $100 Roth IRA grows to $200. If you didn't convert, then over the same period of time your $100 traditional IRA would grow to $200 before income tax, or $150 after income tax. If you were always in a zero bracket, your $25 of other money would grow to $50, so you would be just as well off. But if your tax rate on your investment income is greater than zero, your $25 of other money will grow to something less than $50.

It's not enough that I kept you out of the soup. You want the analysis, too. Some people (i.e., some experts) think that if you use an IRA to satisfy a preresiduary formula marital or credit shelter bequest, then the IRA is immediately taxable, notwithstanding the stretchout that would otherwise have been available. Others (i.e., other experts) think that this should not be a problem. Why take a chance? If you leave the IRA to this trust, how is the spouse supposed to do a rollover? Obviously you can't for the portion that goes to the credit shelter. If the marital share passes in the form of a QTIP trust rather than outright, then obviously you can't for the marital share either. If the marital share passes outright, then you may be able to do the rollover for some or all of the portion of the IRA that will go to the spouse via the trust. This issue is far too complicated for a short answer. For a long answer, see my article on that subject in the October 1997 issue of Estate Planning. The lawyer who handles your estate planning should subscribe to this publication. But again, why look for a complicated solution that may work when there are simpler solutions that do not have these uncertainties?

While this matter may be urgent, it is not particularly difficult; and the estate is not particularly large. To consult with three lawyers would likely double the cost of the project. I suggest consulting with one good lawyer, but doing so right away. Why would he not convert to a Roth? Why has he not considered leaving the children's share in trust rather than outright? Suppose a child (i) has a taxable estate, (ii) has a creditor problem, (iii) goes into a nursing home and wants Medicaid, (iv) gets divorced, or (iv) outlives his/her spouse and remarries. With 3 children, there's a good chance at least one of these things will happen to at least one child.

I'm not sure what you mean by "his trust." If you mean a separate trust containing a marital/credit shelter formula, he could leave the IRA to that trust. But as Katherine hints at, there is a difference of opinion as to whether that might accelerate the income taxation of the IRA benefits. And it will sacrifice, or at least complicate, the rollover of the portion going to the marital share. If his objective is to use his IRA to the extent necessary to fully fund the credit shelter trust, after using whatever other assets are available, he can accomplish this in the beneficiary designation. The tradeoff is the loss of the spousal rollover. A simpler approach is to name the spouse as beneficiary, with a disclaimer/credit shelter trust as contingent beneficiary. The tradeoffs are (i) the decision becomes the spouse's rather than the IRA owner's, and (ii) a disclaimer trust is less flexible that a regular credit shelter trust. I will be presenting a paper on this at the BNA Tax Management advisory board meeting in New York on September 18th. If your lawyer can make it, he/she may find it useful. If not, he/she should watch for a copy of the paper in his/her BNA Tax Management Memoranda.

My article covers various situations in which the spouse was permitted to do a rollover even though he/she was not named as beneficiary. In doing the research for the article, I found two PLRs involving community property. I was not suggesting that they are determinative. (PLRs are not binding on the IRS in any event except with respect to the taxpayers to whom they are issued.) But I think these PLRs and my article may assist Mr. Foley in his research, and in deciding whether he thinks it is worth his effort to apply for a ruling in his case. I don't think the Service's view on this issue (spousal rollovers where the spouse is not named as beneficiary) will change by reason of the final regulations. If you don't subscribe to Estate Planning, you should be able to get a copy of the article from the lawyer who handles your estate planning, or from any law firm with a tax/estates practice with which you have a relationship.

See PLRs 9427035 and 8927042. Also see my article on this in the October 1997 issue of Estate Planning at page 369.

See my article in the October 1997 issue of Estate Planning on spousal rollovers where the spouse is not the named beneficiary. I recently obtained a favorable ruling in a similar case.

Godmom, the IRA owner can leave the IRA to the beneficiaries in trust rather than outright, in the same way that he/she can leave his/her other assets to the beneficiaries in trust rather than outright. Each beneficiary can have as little or virtually as much control over his/her trust as the client deems appropriate. The trusts can be created in the IRA owner's Will, or in a separate document. If the beneficiaries are the IRA owner's children, then most likely, depending upon the terms of the trusts, each trust will have to use the oldest child's life expectancy (see, e.g., PLR 200228025). But if the children are close in age, that shouldn't be a major concern. T. Bouman, if they don't tell you enough at the seminars you're going to, then perhaps you might want to go to some of the ones where I'm on the panel <g>. See PLR 200235038 for a possible road map (subject to the caveat that PLRs are not binding on the IRS except with respect to the taxpayer to whom they are issued, and the IRS won't rule on IRA distributions unless the IRA owner is already deceased). While perhaps more a subject for another day and another place, I think it's worth following up on the issue of revocable trusts. I know that some people believe in them with almost a religious fervor. But they do not save taxes, and my experience has been that they don't significantly simplify the estate administration. The tax planning, the estate tax returns, the appraisals, deciding what to do with what assets, deciding what to do with the IRA, and the planning for the beneficiaries, is all the same either way. What's the big deal about probate, or avoiding probate? In most cases, it's just a matter of submitting a few forms to the court along with the Will, a death certificate, and a small filing fee. Since most Wills contain a marital/credit shelter formula, trusts for children, and lots of pages of boilerplate, how many people's Wills are of interest to the public or the media? While there may be reasons for creating a revocable trust in some cases, in many cases they tend to be more of a distraction than anything else.

I think there's still some confusion here. I think it was silly of the IRS in PLR 200317041 to distinguish between leaving the IRA to separate trusts for each child in the beneficiary designation and leaving the IRA to a single trust in the beneficiary designation and then having the trust be divided into separate trusts for each child. But in most cases it won't have much practical significance other than making beneficiary designations more complicated. In most cases, each child will be a contingent remainder beneficiary of each of the other children's trusts (if a child dies without any issue and without having exercised his/her power of appointment). So, if PLR 200228025 is correct, each child's trust will have to use the oldest child's life expectancy anyway. There's no need for a separate trust instrument. The children's trusts can be created in the IRA owner's Will. In any case, why (unless there is some other reason in a particular case) create a revocable trust? And if there is some reason for creating a revocable trust in a particular case, why run the IRA benefits through the revocable trust? As noted previously, revocable trusts do not save taxes, and in most cases do not significantly simplify the estate administration.

It is not difficult to provide for beneficiaries in trust rather than outright. Indeed, our clients virtually always provide for their children in lifetime trusts rather than outright, to protect the children's inheritances from their creditors, predators and spouses, and to keep the children's inheritances from being included in their estates for estate tax purposes. These reasons apply to IRA benefits just as they do for other assets. It is not much more work to draft the Wills to provide for the children in lifetime trusts rather than trusts to a specified age or outright. Obviously there is a point at which the amount involved is too small to be worth the effort. And in the case of IRA benefits, you give up some flexibility (in terms of being able to have older individuals as contingent beneficiaries -- see PLRs 200228025 and 200235038). So if the IRA benefits are relatively small, it may not be worth the effort to leave them in trust. If each child's share is only $150,000, some people would consider this to be too small for trusts (unless the beneficiary has special needs). But others would go ahead with the trusts anyway, since the assets might grow over time. But it is not particularly expensive or burdensome to administer a trust. Family members may not charge trustees' commissions; and if professional management of the money is desired, you'll pay about the same amount for that regardless of whether the money is in a trust or not. A trust has to file an annual income tax return, but if the trust has a small number of accounts, that should not be particularly expensive. Income taxation of trusts can sometimes be complicated, but as a general rule, distributions carry out income to the extent of the trust's income, and if the income exceeds the amount distributed, the trust pays the tax on the income it retains. Items generally retain their character. There are different rules for required distributions and for capital gains. Most clients prefer to let the trustees decide whether to distribute to the beneficiaries the minimum required distributions (MRDs) from the IRA (or a greater or lesser amount), rather than requiring that the trustees distribute the MRDs to the beneficiaries. In this way, the trustees can consider both tax and nontax factors each year. You should discuss all of this with your lawyer, who can give you specific advice as to how best to accomplish your objectives.

1. Income taxation of estates and trusts can be complicated. But in general, distributions carry out income to the extent of the trust's income. In the example given, since the trust had $22,000 of income and distributed $10,000 to each of two beneficiaries, each beneficiary is taxable on $10,000, and the trust is taxable on the $2,000 it retained. There are different rules for required distributions and for capital gains. The ability to decide each year whether to make distributions to beneficiaries in low income tax brackets or to retain the income in the trust is a benefit of the trust, though probably not the main reason for leaving assets in trust rather than outright. Distributions from an estate or trust generally retain their character, so that traditional IRA benefits will generally remain taxable, and Roth IRA benefits will remain tax-free, even if they pass through a trust. 2. Even if each child's trust has to take distributions based upon the oldest child's life expectancy (in addition to PLR 200317041, see PLR 200235038), if the children are close in age, it shouldn't make a great deal of difference. There are lots of reasons for leaving assets in trust rather than outright, such as protecting against creditors, predators and spouses, and keeping the assets out of the children's estates. These reasons apply to IRA benefits as well as other assets. 3. It's possible to draft a grandchild's trust such that it can use the grandchild's life expectancy. However, the drafting will take some effort. See PLRs 200228025 and 200235038. 4. Leaving IRA benefits to a revocable trust may complicate things. There are generally other ways to accomplish the same result. (While not directly related to IRA benefits, revocable trusts tend to be overused. They do not save taxes. In most cases, they do not significantly simplify the estate administration or reduce the expenses of the estate administration.) 5. A trust can be designed to give the beneficiaries a great deal of control, or not, depending upon what is appropriate in the particular case. 6. The above is not intended as specific legal advice. IRA owners (including the previous posters in this thread) should consult with competent counsel, who can give them specific advice as to how best to accomplish their objectives.

At least two PLRs (9608042 and 9418034) allow a spouse who is the beneficiary of a qualified plan to transfer the benefits to a beneficiary IRA (rather than rolling the benefits over into his/her own IRA). I cannot think of any logical reason that the result should depend upon whether the beneficiary is the spouse. However, I have not found any authority one way or the other. Does anyone know of any authority either way as to whether a non-spouse beneficiary can do this. In PLR 200244023, the beneficiary of a terminating plan took the benefits in the form of an annuity. Perhaps if the beneficiary IRA were permissible, the beneficiary would have done that instead. But it's possible the beneficiary in that case did not consider that possibility.

For $3.5 million, it's worth a try.

It's also not quite correct to say that she's "limited" to the 3 methods cited. While the IRS listed these 3 methods in its notice, they are not exclusive. There have been a number of PLRs that have approved variations of these methods, and some rulings have approved hybrid methods. I discuss this in more detail in my article on this subject in the October 1997 issue of Estate Planning.

Reg: I think there's a big difference between being frustrated that this isn't clearly permissible and bringing a Tax Court case. If it makes enough difference, just draft it so the separate shares are established in the beneficiary designation (e.g., my issue per stirpes, except that each person's share is payable to his/her trust under my Will instead of to him/her). If the IRA custodian won't accept it, and it's important enough, then you can move the account to another financial institution.

Reg: Since there is no substantive difference between the two methods of drafting, it makes no sense to treat them differently for tax purposes. It becomes a problem because some IRA custodians can't deal with beneficiary designations that don't fit on one line. Shelton: a trust is not like a corporation in this regard. You can't give property to a trust, only to the trustees (in their fiduciary capacity).