Between her husband’s death, in 1971, and her own, in 1986, Ethel H. Kurz was the beneficiary of two trusts. Kurz received the income from each. She was entitled to as much of the principal of one (which we call the Marital Trust) as she wanted; all she had to do was notify the trustee in writing. She could take only 5% of the other (which we call the Family Trust) in any year, and then only if the Marital Trust was exhausted. When Kurz died, the Marital Trust contained assets worth some $3.5 million, and the Family Trust was worth about $3.4 million. The estate tax return included in the gross estate the whole value of the Marital Trust and none of the value of the Family Trust. The Tax Court held that Kurz held a general power of appointment over 5% of the Family Trust, requiring the inclusion of another $170,000 under 26 U.S.C. § 2041(a)(2). 101 T.C. 44, 1993 WL 270973 (1993); see also T.C.Memo 1994-221, 1994 WL 193870 (computing a tax due of approximately $31,000).
Section 2041(b)(1) defines a general power of appointment as “a power which is exercisable in favor of the decedent, his estate, his creditors, or the creditors of his estate”. Kurz had the power to consume or appoint the corpus of the Marital Trust to anyone she pleased whenever she wanted, and the Estate therefore concedes that it belongs in the gross estate. For her part, the Commissioner of Internal Revenue concedes that the 95% of the Family Trust that was beyond Kurz’s reach even if the Marital Trust had been empty was not subject to a general power of appointment. What of the other 5%? None of the Family Trust could be reached while the Marital Trust contained lit, and the Estate submits that, until the exhaustion condition was satisfied (which it never was), the power to appoint 5% in a given year was not “exercisable”, keeping the Family Trust outside the gross estate. To this the Commissioner replies that a power is “exercisable” if a beneficiary has the ability to remove the blocking condition. Suppose, for example, that the Family Trust could not have been touched until Ethel Kurz said “Boo!”. Her power to utter the magic word would have been no different from her power, under the Marital Trust, to send written instructions to the trustee.
The Tax Court was troubled by an implication of the Commissioner’s argument. Suppose the Family Trust had provided that Kurz could reach 5% of the principal if and only if she lost 20 pounds, or achieved a chess rating of 1600, or survived all of her children. She could have gone on a crash diet, or studied the games of Gary Kasparov, or even murdered her children. These are not financial decisions, however, and it would be absurd to have taxes measured by one’s ability to lose weight, or lack of moral scruples. Imagine the trial, five years after a person’s death, at which friends and relatives troop to the stand to debate whether the decedent was ruthless enough to kill her children, had enough willpower to lay off chocolates, or was smart enough to succeed at chess. The Tax Court accordingly rejected the Commissioner’s principal argument, ruling that raw ability to satisfy a condition is insufficient to make a power of appointment “exercisable”.
If not the Commissioner’s position, then what? The Estate’s position, 180° opposed, is that the condition must be actually satisfied before a power can be deemed “exercisable”. The Tax Court came down in the *1029middle, writing that a condition may be disregarded when it is “illusory” and lacks any “significant non-tax consequence independent of the decedent’s ability to exercise the power.” Of course, illusions are in the eye of the beholder, and we are hesitant to adopt a legal rule that incorporates a standard well suited to stage magicians (though some legal drafters can give prestidigitators a run for their money). No one doubts that the Kurz family had good, non-tax reasons for the structure of the trust funds. The only question we need resolve is whether a sequence of withdrawal rights prevents a power of appointment from being “exercisable”. Despite the large number of trusts in the United States, many of them arranged as the Kurz trusts were, this appears to be an unresolved issue. Neither side could find another case dealing with stacked trusts, and we came up empty handed after independent research.
For a question of first principles, this one seems remarkably simple. Section 2041 is designed to include in the taxable estate all assets that the decedent possessed or effectively controlled. If only a lever must be pulled to dispense money, then the power is exercisable. The funds are effectively under the control of the beneficiary, which is enough to put them into the gross estate. Whether the lever is a single-clutch or double-clutch mechanism can’t matter. Imagine a trust divided into 1,000 equal funds numbered 1 to 1,000, Fund 1 of which may be invaded at any time, and Fund n of which may be reached if and only if Fund n-1 has been exhausted. Suppose the beneficiary depletes Funds 1 through 9 and dies when $10 remains in Fund 10. Under the Kurz Estate’s view, only $10 is included in the gross estate, because Funds 11 through 1,000 could not have been touched until that $10 had been withdrawn. But that would be a ridiculously artificial way of looking at things. Tax often is all about form, see Howell v. United States, 775 F.2d 887 (7th Cir.1985), but § 2041 is an anti-formal rule. It looks through the trust to ask how much wealth the decedent actually controlled at death. The decedent’s real wealth in our hypothetical is $10 plus the balance of Funds 11 through 1,000; the decedent could have withdrawn and spent the entire amount in a trice. Whether this series of trusts has spend-thrift features (as the Kurz trusts did) or is invested in illiquid instruments (as the Kurz trusts were) would not matter. The Estate does not deny that Kurz had a general power of appointment over the entire Marital Trust, despite these features. If the costs of removing wealth from the trust do not prevent including in the gross estate the entire corpus of the first trust in a sequence (they don’t), then the rest of the sequence also is includable.
Wait!, the Estate exclaims. How did first principles get into a tax case? After consulting the statute, a court turns next to the regulations. 26 C.F.R. § 20.2041-3(b) provides:
For purposes of section 2041(a)(2), a power of appointment is considered to exist on the date of a decedent’s death even though the exercise of the power is subject to the precedent of giving of notice, or even' though the exercise of the power takes effect only on the expiration of a stated period after its exercise, whether or not on or before the decedent’s death notice has been given or the power has been exercised. However, a power which by its terms is exercisable only upon the occurrence during the decedent’s lifetime of an event or a contingency which did not in fact take place or occur during such time is not a power in existence on the date of the decedent’s death. For example, if a decedent was given a general power of appointment exercisable only after he reached a certain age, only if he survived another person, or only if he died without descendants, the power would not be in existence on the date of the decedent’s death if the condition precedent to its exercise had not occurred.
The Kurz Estate takes heart from the provision that “a power which by its terms is exercisable only upon the occurrence during the decedent’s lifetime of an event or a contingency which did not in fact take place or occur during such time is not a power in existence on the date of the decedent’s death.” Like the Tax Court, however, we do *1030not find in this language the strict sequencing principle the Estate needs.
This is the Commissioner’s language, and the Commissioner thinks that it refers only to conditions that could not have been satisfied. Regulation-writers have substantial leeway in their interpretation, see Shalala v. Guernsey Memorial Hospital, — U.S. -, -, 115 S.Ct. 1232, 1236-37, 131 L.Ed.2d 106 (1995), because the delegation of the power to make substantive regulations is the delegation of a law-creation power, and interpretation is a vital part of the law-creation process. See Homemakers North Shore, Inc. v. Bowen, 832 F.2d 408 (7th Cir.1987). A reading must of course be reasonable — must be an interpretation — else the rulemaker is revising the law without the requisite notice and opportunity for comment. Pettibone Corp. v. United States, 34 F.3d 536, 541-42 (7th Cir.1994). The Commissioner’s understanding of the regulation tracks its third sentence, which is designed to illustrate the second. It says: “For example, if a decedent was given a general power of appointment exercisable only after he reached a certain age, only if he survived another person, or only if he died without descendants, the power would not be in existence on the date of the decedent’s death if the condition precedent to its exercise had not occurred.” All three examples in the third sentence deal with conditions the decedent could not have controlled, at least not in the short run, or lawfully. The rate of chronological aging is outside anyone’s control, whether one person survives another does not present an option that may be exercised lawfully, and whether a person has descendants on the date of death is something that depends on the course of an entire life, rather than a single choice made in the administration of one’s wealth.
By contrast, the sequence in which a beneficiary withdraws the principal of a series of trusts barely comes within the common understanding of “event or ... contingency”. No one could say of a single account: “You cannot withdraw the second dollar from this account until you have withdrawn the first.” The existence of this sequence is tautological, but a check for $2 removes that sum without satisfying a contingency in ordinary, or legal, parlance. Zeno’s paradox does not apply to financial transactions. Breaking one account into two or more does not make the sequence of withdrawal more of a “contingency” — at least not in the sense that § 20.2041-3(b) uses that term.
No matter how the second sentence of § 20.2041-3(b) should be applied to a contingency like losing 20 pounds or achieving a chess rating of 1600, the regulation does not permit the beneficiary of multiple trusts to exclude all but the first from the estate by the expedient of arranging the trusts in a sequence. No matter how long the sequence, the beneficiary exercises economic dominion over all funds that can be withdrawn at any given moment. The estate tax is a wealth tax, and dominion over property is wealth. Until her death, Ethel Kurz could have withdrawn all of the Marital Trust and 5% of the Family Trust by notifying the Trustee of her wish to do so. This case is nicely covered by the first sentence of § 20.2041-3(b), the notice provision, and the judgment of the Tax Court is therefore
AFFIRMED.