Probate & Property Magazine, September/October 2009, Volume 23, Number 5

Letter to the Editor
Several Observations . . .

The article makes several references to the nongrantor insured spouse's having a withdrawal power over the trust. The author correctly notes that such an annual power must be limited to the greater of $5,000 or 5% of the trust value. The lapse of a power that exceeded those limits would have untoward estate tax consequences under Code § 2041. I have a more fundamental concern, however, that even a "five by five" withdrawal power might be considered a Code § 2042 incident of ownership in the hands of the insured nongrantor spouse. It is true that in PLR 9748029, discussed by the author, the nongrantor insured spouse had a "five by five" power and the IRS apparently did not regard this feature as an incident of ownership. As the author rightly cautions, however, private letter rulings issued to other taxpayers have no precedential weight. I at least would hesitate to give the nongrantor insured spouse a "five by five" power without the comfort of a ruling directed to my client on the Code § 2042 issue. Of course, given the user fee and other expense of procuring a ruling, it might be better simply to leave out the withdrawal power and rely on the independent trustee's discretion to distribute to the nongrantor spouse.

The article refers to the use of the insurance proceeds to pay the estate tax liability, and it may bear reiterating that while the trustee can and should be empowered to lend to or buy assets from the estate that needs liquidity for taxes, there must not be any legal obligation to provide funds for the taxes. The author does allude briefly to this point in his comments on PLR 9748029 toward the end of the article.

On page 62, the author recommends that if the second-to-die policy is already in place with the prospective grantor as owner, the policy should be sold to the trust, rather than transferred by gift, to avoid inclusion of the proceeds in the transferor's gross estate under Code § 2035 if he or she dies within three years. Whether the sale approach is effective to avoid Code § 2035 depends, however, on how one evaluates the merits and implications of United States v. Allen, 293 F.2d 916 (10th Cir. 1961), and the National Office's controversial position in TAM 8806004, extrapolating from Allen, that the adequate consideration exception to Code § 2035 is satisfied only if the policy is sold for an amount equal to its death benefit.

The author rightly stresses, for exam- ple in the third column on page 62 and the middle column on page 63, the need to avoid direct or indirect contributions by the nongrantor spouse to the trust. It would seem, however, that the inclusion problem arising from such a contribution would be under Code §§ 2036 and 2038, as noted by the author, but not under Code § 2042, as the author also suggests, and the inclusion would be in the gross estate rather than the taxable estate as the author indicates. Similarly, the issues implicated by a loan from the trust to the grantor, discussed at the end of page 62 and the beginning of page 63, arise under Code § 2036 rather than Code § 2042. The author discusses several criteria to be satisfied by such a loan, and these may well be sensible, but I am not aware of any authority specifically endorsing them.

Some clarification of the second full paragraph in the third column on page 62 is in order. Although the author implies that the distribution discretion of the trustee must be limited by ascertainable standards relating to the support or health of the nongrantor spouse, such limitations are not necessary if the trustee is independent. In the same paragraph, where withdrawal powers are discussed, there is another reference to potential inclusion in the taxable estate which should be to the gross estate. Moreover, although I agree that any withdrawal power should be constrained by the "five by five" limits, I would, as noted above, have a more basic reservation, from a Code § 2042 standpoint, about giving the insured nongrantor spouse even a "five by five" power, at least without the comfort of a ruling.

The bullet points in the third column on page 63 appear to suggest that Code § 2036 could apply to a nongrantor spouse, when in fact the estate tax issue for a nongrantor would be under Code § 2041 or § 2042.

Finally, although I agree with the author that the prudent course is not to have even the nongrantor insured spouse as trustee, but to use an independent trustee instead, it is probably worth the trust drafter's while to review Rev. Rul. 84-179, 1984-2 C.B. 195. The analysis in the revenue ruling could be pertinent if, for example, the spouse might become trustee at some point under a succession provision because of unexpected events. It does seem safest to preclude the nongrantor insured spouse from ever serving as trustee.

John R. Jones
King & Spalding
Atlanta , Georgia

Return to Probate & Property Magazine

I am writing with several observations on the article, The Spousal Access Irrevocable Life Insurance Trust Funded with a Second-to-Die Policy , printed in the May/June 2009 issue of Probate & Property. Although none of these comments is fatal to the very interesting strategy proposed by the author, there are a few technical misstatements in the article and, it seems to me, some additional issues that merit consideration before such a plan is implemented.

Advertisement