Contextual Analysis of Tax Ownership

Ownership is one of the most fundamental concepts in tax law, yet it is remarkably confused. The uncertainty inhibits tax planning, leads to inconsistent responses from the government, and produces unexpected outcomes in the courts. There has been no shortage of scholarly attention to the issue, but most of the commentary has been either exceedingly narrow or focused on far-reaching reforms. As a result, the law of tax ownership lacks conceptual foundation. This Article attempts to remedy the deficiency. It proposes a comprehensive approach to tax ownership and demonstrates that the doctrine may (and should) be significantly clarified without a dramatic overhaul of the existing substantive law. The approach rests on dividing all ownership questions into four categories depending on the context in which the questions arose. Using this analytical framework, the Article allocates various tax ownership authorities to appropriate categories and develops the underlying principles guiding the analysis for each group. Because these principles differ among the categories, the Article suggests that the existence of numerous seemingly inconsistent tax ownership decisions should be understood not as a sign of a confused doctrine, but as an appropriate result reflecting the underlying conceptual differences. By rationalizing and organizing the law of tax ownership, the Article provides a framework for resolving future tax ownership controversies.


If one wanted to make a point that there is no such thing as a simple tax concept, ownership would be a perfect example. It is one of the central concepts of tax law, and it is as complex as it is confused. It generated controversy virtually from the inception of federal income tax, gave rise to countless disputes between the government and the taxpayers throughout the years, and remains as contentious an issue today as it has ever been. As one could expect from this ongoing battle, no clear winner has emerged. More importantly, no comprehensive doctrine has developed. Competing views of tax ownership have been expressed and tested in a multitude of fairly narrow unrelated factual settings. Particular types of transactions raised tax ownership questions at different times. Years, and sometimes decades, separated the judges considering novel tax ownership questions from the decisions that could have provided a useful insight. More often than not, remoteness and factual dissimilarity prevented consideration of the relevant authorities. As a result, the law of tax ownership today is a patchwork of rules that appear to lack a unifying principle (or set of principles).

In the absence of a conceptual foundation, novel ownership issues continue to raise considerable uncertainty at the planning stage, lead to confused and sometimes self- contradicting responses from the government, and produce highly unpredictable outcomes in the courts. Tax commentators have long identified the problem, but have not succeeded in remedying it. While the scholars agree that the current ownership doctrine is unsatisfactory, their work, for the most part, has been either fairly narrow and restricted to the analysis of particular transactions,[2]or extremely broad and focused on questions of fundamental reform.[3 ] This article attempts to fill in the gap by suggesting a comprehensive approach to tax ownership and demonstrating that the doctrine may (and should) be significantly clarified without a dramatic overhaul of the existing substantive law. By rationalizing and organizing the law of tax ownership, I hope to develop a framework that will be useful in resolving future ownership questions.[4]

My main argument is that confusion plaguing the law of tax ownership is largely unnecessary and may be considerably reduced if one recognizes that tax ownership issues should be analyzed differently depending on the context in which they arose. The contextual approach I propose here is based on an observation that two fundamental distinctions profoundly affect the tax ownership analysis. The first is the distinction between fungible and nonfungible assets. The second is the difference between an inquiry into the timing of an ownership transfer (I will call it the when case) and an examination of the substance of that transfer (I will call it the whether case). Taking these two distinctions into account results in four separate categories: (1) fungible when cases, (2) nonfungible when cases, (3) fungible whether cases, and (4) nonfungible whether cases.

Having identified these categories, I attempt to discern whether authorities in each group can be reconciled and, if so, on what basis. Unfortunately, cases are replete with generic references to a “bundle of rights,”[5]the benefits and burdens, and the never-failing arguments based on the taxpayers’ intent.[6] Only on rear occasions do authorities refer to the precedents that are relevant based on the proposed framework. On the other hand, cites to irrelevant decisions are commonplace. Separating the authorities into four groups helps to go behind the rhetoric and reveal the underlying approaches. Perhaps somewhat unexpectedly, authorities within each category are fairly consistent with each other, in outcomes if not in reasoning. Conversely, the approaches differ, sometimes dramatically, from one category to the next. Once the numerous cases are sorted out into the appropriate groups, and once the test applicable for each group is identified, the law of tax ownership becomes considerably more coherent and predictable.

There are many familiar examples in each of the suggested categories. A sale of an office building is a classic nonfungible when case. The question here is whether the tax ownership is transferred at the signing, the closing, or somewhere in between. The nonfungible when authorities purport to make this determination based on a multi-factor test. I argue, however, that rather than merely balancing the factors, most of these authorities answer the question by identifying, in essence, the point of no return. Thus, in nonfungible when cases, the sale takes place when neither the buyer nor the seller can back away from the deal.

A forward sale of IBM stock is a typical fungible when case. Timing of ownership transfer in this context hinges on the uncertainty about the shares being sold (uncertainty that is entirely absent in the office building case and that reflects the fundamental difference between fungible and nonfungible assets). Although the forward buyer gains full economic exposure to IBM at the inception of the contract, the ownership transfer is delayed until the specific shares are identified, usually by delivery to the buyer. While this conclusion is directly contrary to the widely applicable rule dealing with settlements of securities sales, I reexamine the rule’s history and argue that the rule might be viewed as a misguided attempt to follow the same principles that apply to all fungible when cases.

Nonfungible whether decisions, such as sale-leaseback authorities, turn largely on the economic considerations. The question here is whether the “sold” property was actually sold, or merely pledged as security for a loan. To be sure, title and possession

are inevitably cited along with economic exposure as major factors affecting the analysis. Despite the recitations, however, I suggest that title and possession usually have little weight because they are frequently split in a predictable and repeated fashion. Allocation of economic benefits and burdens, on the other hand, varies from case to case, often in nuanced ways.

A sale and repurchase of securities, or a “repo,” is a typical fungible whether case. In this context, the question whether ownership has been transferred at all is answered by deciding whether the original owner has transferred control over the securities to a repo buyer. Whoever ends up with control is the tax owner. On the other hand, economic exposure is not determinative. Repeated assertions by the courts and the government that economics matters in this setting confuse the analysis. On the other hand, lack of a detailed consideration of the subtleties of control has resulted in a considerable uncertainty surrounding tax treatment of popular market transactions.

Finally, fungibility itself is context-sensitive. Fungible assets such as publicly traded securities may lose their fungibility and start “acting” as if they were nonfungible. When this happens, the inquiry in either when or whether context shifts to that applicable to nonfungible assets. Understanding this phenomenon is critical to placing a particular dispute into the appropriate category, as I demonstrate using an example of a recent tax ownership controversy.

The remainder of the article consists of seven parts. Parts two through five address each of the four categories and (beginning with the second part) compare the principles governing each category with those underlying a related one. Part six compares when and whether authorities in general. Part seven considers a recent case raising a tax ownership issue. A brief conclusion completes the article.


A. Short Sales, Prepaid Forwards and the Trade Date Rule

The fungible when authorities, as all when cases, address the tax ownership question in circumstances where the nature of the transaction is clear: it is a transfer of an asset for consideration, i.e., a sale. What is not clear is the exact moment when this transfer occurs. The analysis used by courts and the Service to resolve this issue depends on whether or not the asset in question is fungible.

The fundamental difference between fungible and nonfungible assets was recognized by the Supreme Court several years prior to enactment of the federal income tax. In Richardson v. Shaw,[7]the Court considered whether a broker owns shares purchased by it in street name on behalf of its margin account customer. The Court concluded that the broker’s bankruptcy should not prevent the customer from withdrawing the shares because the customer, not the broker, owned the stock. Rebutting the argument that the broker should be viewed as the owner because it was free to deliver to the customer certificates other than those originally placed in the margin account, the Court focused on the fungible nature of the asset involved:

A certificate of the same number of shares, although printed upon different paper and bearing a different number, represents precisely the same kind and value of property as does another certificate for a like number of shares of stock in the same corporation. It is a misconception of the nature of the certificate to say that a return of a different certificate or the right to substitute one certificate for another is a material change in the property right held by the broker for the customer. As was said by the court of appeals of New York, one share of stock is not different in kind or value from every other share of the same issue and company. They are unlike distinct articles of personal property which differ in kind and value, such as a horse, wagon, or harness. The stock has no earmark which distinguishes one share from another, so as to give it any additional value or importance; like grain of a uniform quality, one bushel is of the same kind and value as another.[8]

Not long after the arrival of federal income tax, the Service concluded that the holding of Richardson v. Shaw regarding stock ownership applies for tax purposes and, therefore, the client, and not the broker, is an owner of dividends paid on the stock held in street name.[9 ] Even earlier, the Service tackled a tax ownership issue when it considered whether securities dealers should recognize gains and losses with respect to their open short sales at the end of their taxable years.[10]

In a short sale, a short seller borrows an asset, such as shares of stock, and sells it into the market. The short seller undertakes to return shares identical to those borrowed to the stock lender and usually posts the proceeds of the sale with the lender as collateral. Eventually, the short seller acquires shares identical to those borrowed and returns them to the stock lender, completing (or “covering”) the short sale. A short sale is a unique transaction that is important in both when and whether contexts. The Supreme Court held early on that the first leg of a short sale—lending of securities subject to the borrower’s obligation to return identical securities—transfers ownership of securities for tax purposes despite being called a “securities loan.”[11] This transaction is addressed in detail later in this article. Once it is concluded that a stock lender loses ownership at the inception of a short sale, it follows that the return of securities by the borrower is also an ownership transfer. Timing of that transfer presents a classic when question. The borrower’s obligation to return identical securities to the stock lender is similar to that of a taxpayer who agrees to sell the stock on a future date, i.e., enters into a forward contract. Both transactions raise a question whether an obligation to deliver stock in the future should lead to an immediate transfer of its ownership to the future purchaser.

Analyzing the tax consequences of a short sale in 1919, the IRS began its analysis by emphasizing a fungible nature of the stock sold short: “Shares of stock are fungible things, and their loan with an agreement to return things of the same class is the mutuum of Roman law . . . .”[12] Focus on fungibility is hardly surprising because the distinction between fungible and nonfungible assets underlies the very nature of a short sale. Generally, only fungible assets may be sold short. If the asset is fungible, the lender, being fully aware of the borrower’s intent to sell the specific asset borrowed, would accept the borrower’s promise to return an identical asset in the future for two reasons. First, it makes no difference to the lender whether the asset returned is the very same asset that was borrowed (the specific asset), or is different but indistinguishable from it (the identical asset). Second, it is reasonable for the lender to expect that the borrower would keep the promise because the asset in question is widely available.[13] Thus, in tax ownership cases, an asset is fungible if it is widely available, at least to the party relinquishing tax ownership (the forward seller or the stock borrower obligated to return the stock), and if the party acquiring tax ownership (the forward buyer or the stock lender expecting a return of the stock) is indifferent as to which particular asset is eventually delivered as long as it belongs to the same category of economically identical assets.

Having recognized the importance of fungibility, the IRS observed that as long as a short sale remains open (i.e., until the borrower returns specific shares to the lender), it is impossible to determine the tax consequences of the transaction. In a short sale (as in a prepaid forward contract discussed next), the amount realized from the sale is known before the tax basis of the asset sold. The amount realized is the amount received by the short seller upon a market sale of the borrowed asset. The adjusted basis of the asset sold, however, is that of the asset ultimately delivered to the lender to cover the short. Until the specific asset is delivered, its basis cannot be determined. In light of this uncertainty facilitated by the fungible nature of stock, the Service concluded that a short sale should not be given tax effect until it is closed by delivery of specific shares (with a known tax basis) by the borrower. The Supreme Court strongly embraced a broader principle that a transaction should not be viewed as closed for tax purposes until it is possible to ascertain its tax consequences—a rule that has become known as the “open transaction” doctrine.[14] In the tax ownership context, this principle means that a transfer of ownership should not be recognized until its tax consequences can be determined with certainty. Throughout the article I will refer to this principle as the “certainty rule.”

The decision reached by the Service in 1919 withstood eighty plus years of tax disputes. The Service confirmed it several years later and formalized it in the Treasury regulations in mid-1930s. These regulations remain in effect today in an essentially unaltered form.[15] They were upheld by courts and reaffirmed in several revenue rulings, the latest issued in 2004.[16]

It is not hard to see the wisdom of treating a short sale as an open transaction if the short seller has nothing that conceivably could be viewed as owned by the lender. But it is hardly self-evident that a taxpayer who sells a stock short while holding an identical stock (i.e., enters into a so-called “short against the box” transaction) should be treated as the continuing owner as well. After all, this taxpayer holds the stock of a type that it will ultimately deliver (and that could be viewed as sold when the short sale is entered into). Furthermore, the short seller completely immunizes itself from any economic exposure to that stock. On the other hand, the stock lender is fully exposed to the stock’s upside and downside. Yet, it is a black letter law that there is no difference whether a taxpayer enters into a short sale of securities while holding identical securities or acquires identical securities later. The short sale remains open in either case.[17] The only possible explanation for this complete separation of tax ownership from economics is that uncertainty about the specific asset being sold and the concomitant impossibility of determining the tax consequences of the sale override the stock lender’s full economic exposure to the stock. This results in what Edward Kleinbard termed “riskless ownership” of securities in his important article. Far from being unique to short sales or transactions involving securities, the most consistent conclusion of the numerous fungible when authorities is that transfer of economic benefits and burdens has no effect on tax ownership as long as no specific asset is identified.

This conclusion give taxpayers potent tax planning opportunities when it is combined with the so-called tax identification rules.[18] Taxpayers selling a portion of their stock may (and frequently do) use these rules to choose the specific shares being sold among a larger number of economically identical shares purchased by them at different times and prices. Present in tax law almost from inception, these rules are exceedingly generous to taxpayers because they allow “cherry picking,” i.e., choosing among several economically identical alternatives the one with the most favorable tax result. Relying on these rules, a taxpayer who enters in a short against the box transaction could choose whether some of her long shares held “in the box”, or the newly acquired shares, should be treated as covering the short position.[19] When the Service became aware of this trick, it concluded that taking advantage of tax identification rules in the short sale context was too much to ask for. Courts disagreed and upheld the taxpayers’ positions time after time.[20]

Eventually animated by a particularly large short against the box sale,[21]Congress stepped in and eliminated this tax planning technique, but it did so without raising the ownership issue. With the enactment of Section 1259, taxpayers entering into a short against the box trade (and several other transactions) are viewed as constructively selling their appreciated stock and realizing the built-in gain when they enter into the transaction. By creating a constructive sale regime, Congress indirectly reinforced the fundamental conclusion that a short against the box seller retains ownership of the stock held “in the box” under general tax principles.[22] A revenue ruling issued in 2003 confirms this conclusion.[23]

A principle that a fungible asset is not sold until it is identified has been recently confirmed when the IRS addressed tax ownership question raised by the so-called variable delivery prepaid forward contracts. As with short against the box sales, the government’s initial reaction was to disagree with the taxpayer’s intended treatment. A case filed by the IRS in the Tax Court in August of 2002 sent a mild shock wave through some Wall Street bankers, their tax advisors, and their clients. The Service asserted in Stevenson v. Commissioner[24]that the taxpayer should have realized gain on the stock when he entered into a prepaid variable share delivery forward contract to sell it, not upon its maturity. In other words, the government sought to defeat the entire purpose of this transaction.[25]

Under a typical variable delivery prepaid forward contract, a taxpayer holding a large amount of appreciated stock would undertake to deliver to the forward buyer a variable number of shares of this stock or their cash equivalent on a future date. In exchange, the forward buyer would immediately pay the taxpayer a large portion of the current price of the stock subject to the forward (hence the term “prepaid”). To secure its future obligations, the taxpayer-seller would pledge the maximum number of shares deliverable under the contract to the buyer, sometimes transferring the shares to an unrelated third party trustee. The seller would retain the right to vote the pledged shares and to receive the dividends paid on those shares during the term of the forward. In addition, the seller would retain the right to substitute other identical shares or other collateral such as U.S. Treasury securities with a value equal to a certain percentage of the value of the maximum number of deliverable shares, marked to market on a daily basis. In some cases, the seller may also allow the buyer to borrow the shares on terms that satisfy the requirements of Section 1058.

A typical contract would provide that the number of shares to be delivered by the seller when the forward matures would vary based on the stock price on that date (hence the term “variable delivery” forward). Although the prepaid forward in the Stevenson case had a more complicated payout, a typical pattern would be as follows. Assuming that the forward is on 100 shares and the stock is trading at $20 when the forward is entered into, if the stock price on maturity date is less than $20 per share, the seller will deliver all 100 shares. If the stock price at maturity of the forward is between $20 per share and $25 per share, the seller will deliver shares with a total value equal to $2,000 (as the stock price increases from $20 to $25, the number of shares decreases from 100 to 80). If the stock price at maturity of the forward exceeds $25 per share, the seller will deliver 80 shares.[26 ]Typically, the seller will have the right to settle the forward in cash.

Allegedly to attract young lawyers to practice of tax, or for another noble reason, the two changes in the payout pattern that occur if the stock price at maturity is $20 and $25 per share have been termed “kinks,” and the prepaid forward with these payout pattern is commonly referred to as a “kinky forward.”[27] From the seller’s point of view, the most attractive, if not kinky, feature of this instrument is that the seller typically receives a substantial portion of the current value of the underlying shares at the inception of the forward, substantially reduces its economic exposure to the shares, but does not expect to recognize any gain from sale of the appreciated shares until the forward settles several years later. In a sense, this is almost as good as a short against the box, so the government’s attention to this transaction is understandable.

Fortunately for Mr. Stevenson and many others, the Service quickly reconsidered its original position. In Revenue Ruling 2003-7,[28]the government concluded that a prepaid forward very similar to that contested in Stevenson does not result in an immediate sale of stock under general tax principles. The forward had the kinks at $20 and $25 and other typical features, including a three-year term, a pledge of stock by delivery to a third-party trustee, and a cash-settlement option. The ruling assumed that the seller intended to deliver some or all of the initially pledged shares when the time comes to settle the forward.

From the first sentence of the ruling—the statement of the issues—it is clear that the government is focusing on the “right” factors: “the shareholder . . . has the unrestricted legal right to deliver the pledged shares or to substitute cash or other shares for the pledged shares on the delivery date, and is not economically compelled to deliver the pledged shares.”[29] In these few words, the ruling identified two principal tax ownership issues raised by the prepaid forward: fungibility and a material condition precedent. As long as the seller has the right to deliver other shares and also has a wherewithal to take advantage of this right, for example by purchasing new shares and delivering them under the forward without recognizing gain on the existing shares, the pledged shares remain fungible and no identification occurs. Consistently with the short sale authorities, the ruling concluded that in the absence of identification the sale contract remained executory and the forward seller retained ownership of the shares.[30]

The idea that ownership doesn’t change hands until it is certain what specific fungible property will be ultimately delivered is difficult to reconcile with a well- established rule that exchange-traded securities are treated as bought and sold on the “trade date,” not on the “settlement date.”[31] A trade date is the date on which the contract to buy or sell the security is made; the settlement date is the date on which the security is delivered and the payment is tendered.[32] Admittedly, the two dates are just a few days apart,[33]but the temporal proximity provides no reason to choose either of the two as the date of sale. The trade date rule triggers an ownership change on the date when an executory contract is entered into, not when it is closed. While nothing could seemingly reconcile this outcome with the short sale and forward authorities, a historical detour might explain the reason for the diverging standards.

Quite simply, the rule used to be exactly the opposite. When the Service was first called on to decide when a seller should be treated as transferring the ownership of exchange-traded stock, it decided that the settlement date was the right moment in time.[34]

The Service reasoned that a loss from a sale should be evidenced by a closed and completed transaction and that no such transaction existed until the stock certificate was delivered to the purchaser (or its broker). A taxpayer who wanted to take his losses in the tax year when he placed a sell order sought relief in court and prevailed. In Ruml v. Commissioner,[35]the court concluded that a plaintiff who ordered his broker to sell all of his 4,000 shares in December of 1928 should be allowed to take a loss in that year even though he did not deliver 2,500 of the 4,000 shares to the broker until February 1929.

The taxpayer could not possibly deliver the shares in 1928 because they were pledged to a bank and the taxpayer was able to withdraw them only upon repayment of the bank loan in February of the following year.

It is not clear whether the Service based its argument in Ruml on the short sale and other fungible when authorities, but it is evident from the opinion that the court reasoned primarily relying on nonfungible when cases. The court’s analysis is worth quoting both because it gave rise to the only line of cases inconsistent with other fungible when decisions and because it was cited almost in full by the Service in reversing its original position and adopting the trade date rule in its current form.

It is clear that the petitioner intended to sell the specific shares he owned and only those. If the broker made a short sale to his customer, it was for his own account, not for the petitioner who had authorized no short sale for his account. It seems to us clear that the transaction between the petitioner and the broker was a sale by the petitioner to the broker of the specific shares pledged to the bank, under an understanding with the broker that certificates therefor were to be later delivered . . . . [W]hen the evidence of realization is a sale of personal property, it is not always necessary to deliver the property before there may be a deduction of a loss. It is enough that the obligation to deliver is so fixed that the loss is reasonably certain in fact and ascertainable in amount. Here the transaction was so far advanced in December that the petitioner was bound to deliver the stock to the broker at a price which was then determined by the sale the broker made. That sufficed to make the loss certain and established the amount. Moreover, the intention of the petitioner and the broker being that the particular shares then owned by the petitioner should be delivered to the broker, it follows that title to them passed to the broker in December.[36]

As the discussion in the next part will show, a question whether the transaction is advanced far enough to become irreversible is the test used for nonfungible assets in timing cases. The conclusion that the sale was completed because “the petitioner was bound to deliver the stock to the broker at a price which was then determined” by broker’s sale begs the question—what stock? To be sure, the amount realized from the sale was certain in December of 1928. But not the basis of the shares sold. After all, if Mr. Ruml could not repay the loan, the bank would not have released the shares and Mr. Ruml would have had to buy new shares in the market and deliver them to the broker to cover the open short position. It is entirely possible that in this scenario, Mr. Ruml would have recognized a gain, not a loss, from the sale. Needless to say, it is of little help that the court knew what had actually occurred. Concluding that the loss was sustained in 1928 appears to have flagrantly violated the certainty rule.[37]

Or maybe the departure from the certainty rule was not so blatant after all? The court began its reasoning by observing that “[i]t is clear that the petitioner intended to sell the specific shares he owned and only those,”[38]and concluded by reiterating that “the intention of the petitioner and the broker being that the particular shares then owned by the petitioner should be delivered to

the broker.”[39] The court used the taxpayer’s intent as a means of identifying the specific shares sold. Thus, it recognized the importance of identification. To be sure, it was hardly wise to rely on the taxpayer’s intentions. After all, we can always change our mind, especially when doing so would materially improve our tax position.[40 ]Setting aside reliance of the Ruml court on the taxpayer’s intent, its finding that a sale is complete when specific shares certain to be delivered are identified resonates with the rest of the fungible when authorities. Indeed, once identification occurs, tax consequences of the transaction can be determined and the certainty rule would not be violated if the ownership were transferred for tax purposes on the trade date.[41]

B. Commodity Futures, Forwards and Contracts for Deferred Delivery

At the same time as the Service addressed the fungible nature of securities when it considered taxation of short sales, it focused on the fungible nature of commodities at the urging of cotton and grain industries. Long before enactment of federal income tax, cotton and grain merchants started to hedge their businesses by purchasing and selling contracts for future delivery of their products. These futures contracts traded on many exchanges throughout the country and represented unconditional obligations to sell or purchase a fixed amount of cotton or grain on a specified future date. In keeping their books, the merchants reflected open futures contracts as parts of their inventories of physical commodities. When it became necessary to determine how these contracts should be taxed, the merchants requested that the Service allow them to follow their accounting treatment for tax purposes.

In a 1920 Appeals and Revenue Memorandum, the IRS concluded that under no circumstances can futures contracts be included in inventory: “There is in fact no profit or loss in the purchase of a commodity until the transaction has been completed by the sale of that particular commodity, nor is there any profit or loss in a transaction in ‘futures’ until the transaction has actually been closed.”[42] One year later the Service reiterated the point as follows:

[T]he Committee . . . holds it to be self-evident, . . . that any proposition to add to an inventory the value of a commodity the title to which is not at the time actually vested in the taxpayer, or to deduct from an inventory the value of a commodity the title to which may or may not be vested in the taxpayer but which is to be delivered only at some time in the future, cannot by any correct system of reasoning or logic be maintained[43]

The reasoning looks familiar: Because there is no way of knowing whether a merchant holding a certain commodity will deliver it under the futures contract, there is “no profit or loss” with respect to the commodity or the futures contract until this question is resolved. In part, the government relied on its own regulations that provided at the time that “goods merely ordered for future delivery and for which no transfer of title has been effected should be excluded”[44]from inventories. Treasury regulations currently in force contain almost exactly the same provision.[45] The conclusion of the 1920 ruling was confirmed in Revenue Ruling 74-227,[46]which repeated its reasoning virtually verbatim.

Finally, the Service concluded early on,[47]and has continuously adhered to the view,[48]that the effect of a futures contract on tax ownership does not depend on the manner in which the ownership of an underlying asset is acquired in the first place. Revenue Ruling 74-226 considered whether so-called “straddlers” were entitled to inventory their physical commodities. A straddler’s business was to purchase physical commodities and simultaneously sell the exact amount of this commodity forward by entering into futures contracts. They were arbitrageurs seeking market inefficiencies in the pricing of either the spot commodity or the futures contract, not merchants selling physical commodities and partially hedging the price risk. The ruling allowed the straddlers to inventory their holdings of physical commodities and to mark their futures positions to market (while not including them in inventories). Obviously, the straddlers were in a business quite different from that of cotton and grain growers and traders. They did not produce the commodity and did not acquire it in order to hold and sell when the prices move advantageously. They never had economic exposure to the asset, nor any reason to own it free and clear (if that should matter). That is, the straddlers were the ultimate riskless owners. Nevertheless, the IRS made no attempt to argue that they did not own the commodities for tax purposes.

Thus, the Service has consistently maintained that a futures contract to purchase a commodity does not make the holder of the contract an owner of the commodity. This is true even though such holder has virtually the same economic exposure to the commodity as its outright owner. The rule is the same on the sale side. A taxpayer does not cease to own a commodity by virtue of entering into a futures contract to sell it even though the futures contract completely eliminates her economic exposure to the commodity. In fact, her position is very similar to that of a taxpayer who enters in a short against the box transaction. For tax purposes, both taxpayers become riskless owners of a fungible asset. These outcomes are entirely consistent, of course, with the rules for shorts sales of securities and securities forwards.

Forward contracts are very similar to futures (the main difference being the forwards are not exchange-traded). While the Service was developing its approach to commodities futures mostly through a ruling process, a parallel development dealing with contracts for forward sale of commodities was taking place in courts. At a very early stage, two opposing views of taxation of commodity forward contracts were articulated, argued and adopted by different circuits. A typical factual setting involved a transaction in which a seller entered into a contract to sell a commodity to a buyer, but the contract has not been fully performed at the end of a tax period. For example, a taxpayer in United States v. Amalgamated Sugar Co.[49]operated several sugar refineries and sold the refined sugar wholesale. In making its tax return for the fiscal year ending on February 28, 1917, the taxpayer reported as sold almost two hundred thousand bags of sugar that remained in its warehouse at the end of the year. The Service argued that the sugar was not sold until the following year when the sugar was delivered, but the court agreed with the taxpayer. The court relied in part on the taxpayer’s intent to make a sale before the end of February of 1917 and in part on state contract law. But most importantly for our purposes, the court addressed head on the government’s argument that the contract was,

in essence, merely an executory contract:

[I]t is contended that the contracts were executory and that title remained in the company on February 28, 1917, because the property had not been segregated and identified in separate form. Beet sugar of a standard and uniform grade, in gabs of one hundred pounds each, is fungible property. In that respect it falls within the same class as flour, grain, or oil. Title to an unseparated part or unit of a larger quantity of fungible property passes under a valid contract of sale without separation, or segregation, if that is the intention of the parties.[50]

Two other circuit courts considered a similar issue a few years earlier in Haas v. Commissioner[51]and Brown Lumber Co. v. Commissioner.[52] In both cases, the buyers of fruit and lumber took a loss in the tax year when they contracted to purchase the goods, but did not receive them, in the amount by which the fair market value of the goods at the end of the year was less than their cost to the buyers. The government argued for the deferred treatment of the sale and won. In both cases the courts concluded that because the specific commodity was not identified at the end of the tax year, no ownership transfer took place.

The fact that seller had fruit at the close of 1920 to meet the contract, in its warehouses, as to time and grade, since none was segregated or set apart, is not material. The fungible mass doctrine has no application. The appellants were not to take part of a larger mass of like kind.[53]

Amalgamated Sugar and Haas made the choice abundantly clear. Both courts were writing on a clean slate. Amalgamated Sugar relied on non-tax state law cases dealing with ownership transfers. Haas had no cites in support of the quoted paragraph. Instead, it relied on the same Treasury regulations regarding purchases and sale of inventories that the Service cited in its cotton futures ruling in 1920. It is not clear why the Haas court concluded that the “fungible mass doctrine” did not apply. What is clear, however, is that Amalgamated Sugar was not followed and Haas was. The identification requirement prevailed.[54] The current Treasury regulations state that a seller remains a tax owner of inventory subject to a contract of sale as long as the property is not “segregated and applied to the contract.”[55]

An interesting, albeit a limited, example of how the principles discussed in this part apply in yet another context are the so-called “price later” contracts.[56] These contracts for sale of a commodity have one unusual feature—the purchase price is not determined at a time when the goods are delivered. Instead, a seller retains a right to designate (or “call”) as the selling price for its goods the market price of that commodity on any day within a specified period (the “call date”) that can last for as long as a year after the date of delivery. Keeping the price open exposes the seller to market fluctuations between the inception of the contract and the call date. Thus, price later contracts present a situation that is the exact opposite of a prototypical pattern discussed in this part. Ordinarily in fungible when cases, the economics is transferred while the title and possession are not. In a price later contract, the seller retains full economic exposure to the asset sold, but surrenders title and possession at the inception of the contract. Although none of the authorities discussing price later contracts refer to any of the learning discussed so far, their uniform conclusion is entirely consistent with this learning. Just as divesting of economic exposure does not transfer ownership of a fungible asset in a when case, retaining economic exposure by the seller does not delay the ownership transfer. As in most other fungible when cases, it is the identification, not the economic exposure, that determines the timing of ownership transfer.

Authorities addressing taxation of short sales, variable delivery forwards, cotton and grain futures, and contracts for deferred delivery of various commodities described above do not exhaust the list of the fungible when cases. The approach developed in these cases applies to determine the timing of ownership transfers in other settings where fungible assets are involved, even thought these additional authorities neither acknowledge their intellectual “heritage,” nor announce the suggested principle as the reason for their decisions.[57]

C. Some Preliminary Observations

Although courts and the Service formulated their decisions regarding commodities futures and forward contracts during the same period as they considered the proper taxation of short sales, they did not view these areas as related. Yet the same tax ownership issue is raised in all these contexts and the conclusions reached by the authorities in each case are entirely consistent. Upon gaining full economic exposure to a fungible asset the buyer does not become its tax owner. Rather, the seller remains a riskless owner of the asset. The delay in ownership transfer is caused by uncertainty regarding the specific asset being sold that exists because the asset in question is fungible. Fungibility means that the seller may freely choose which particular asset of the same type to deliver to the buyer who would be indifferent about the seller’s choice because the assets are identical, as far as the buyer is concerned. To ignore this fundamental uncertainty would violate the certainty rule, at least where the fungible asset is subject to tax identification rules that allow taxpayers to choose between economically identical assets with different tax attributes.

Tax identification rules are difficult to reconcile with the idea of fungibility. In a way, they take the distinction between capital assets and ordinary property to its logical (or illogical) limit. Ordinary property, or at least ordinary property that is accounted for as inventory and is relevant for the purposes of this article,[58]is subject to ordering conventions rather than identification rules. These conventions, such as the last-in-first- out or the first-in-first-out rules, determine which particular items are deemed to be sold regardless of which specific units are delivered to the buyer. Furthermore, some inventory pricing conventions, such as valuing inventories at lower of cost or market, or at market, make tax bases of inventory items, originally different due to varying purchase costs, either more uniform or entirely uniform.[59] Finally, gross income from inventory sales depends not on gain or loss from sale of each inventory item, but on the difference between the annual gross receipts and annual cost of goods sold and other expenses.

Ordering conventions, valuation rules, and the manner in which gains and losses from inventory sales are calculated make inventory items not only economically fungible, but fungible for tax purposes (or tax-fungible) as well.[60] Not so with capital assets. The amount and the long-term or short-term character of gain or loss from sale of a capital asset depends on the specific asset being sold. With some limited exceptions, there are no mechanisms that would erase the differences in the bases of capital assets in a manner similar to marking inventories to market. Thus, if a seller has several shares of the same issuer purchased at different times and prices, it will make a difference for him, assuming he is not a dealer in securities, which particular shares he delivers to the buyer, but it will make absolutely no difference to the buyer. The buyer is indifferent because the shares are fungible. The seller cares because they are not tax-fungible. Tax identification rules acknowledge this difference and allow taxpayers to use it to their advantage.

Resolving the uncertainty regarding the specific asset being sold is clearly important where this asset is not tax-fungible, such as in a short sale or a prepaid forward of stock held as capital asset. However, this uncertainty simply doesn’t exist for tax- fungible assets. Thus, the certainty rule would not be violated by treating an unconditional contract for sale of a tax-fungible asset as transferring tax ownership when it is entered into because no matter what particular units are ultimately delivered, the tax consequences of the sale would be the same. This is exactly the point made by the Amalgamated Sugar court. Why, then, was this reasoning not followed, at least for tax- fungible assets such as commodities? Why did courts keep demanding that the specific commodities be identified?

The most apparent reason for the identification requirement is a historic reliance on state contract law by the tax tribunals that developed the standard in the first place. There are other reasons, however, why the Amalgamated Sugar rule, even if restricted to tax-fungible assets, was not likely to survive. If adopted, the rule would make the tax ownership inquiry too context sensitive. It would only apply if the asset subject to the ownership change is tax-fungible. However, this depends not on the asset itself, but on the manner in which the seller treats it for tax purposes. The same security may or may not be a capital asset depending on whether its owner is a dealer or an investor. The same commodity item may or may not be subject to inventory accounting. Both issues, i.e., the distinction between capital and ordinary assets and the necessity of maintaining inventories have generated a considerable amount of controversy. Basing such a fundamental concept as tax ownership on resolution of other contentious disputes should have given pause to any decision maker looking for a workable rule. Furthermore, it would hardly be desirable to make buyer’s ownership treatment dependent on the seller’s tax accounting for the asset. Today, this would be particularly problematic because a seller that is a trader in securities or commodities may choose between an all-ordinary mark-to-market treatment of Section 475 and the historic capital and realization tax rules. If the trader-seller makes the election, her securities and commodities will be tax-fungible and the Amalgamated Sugar rule would call for a conclusion that a forward buyer immediately acquires tax ownership of a security held by such trader. If the trader-seller does not make the election, a forward purchaser would not become the owner until the specific asset is identified. No sane buyer would accept such regime.

Another problem that the Amalgamated Sugar rule and reliance on tax fungibility would create is caused by what Kleinbard called “many longs, one owner” phenomenon.[61] Looking at the Amalgamated Sugar itself, the buyer of sugar, being assured that it is about to receive the bags from the seller, could sell it forward to a third party. In this case, the sugar manufacturer, the buyer, and the third party would all be long sugar—the manufacturer would be long because it actually holds the commodity, and the buyer and the third party, because each is on a long side of a forward contract. The manufacturer and the buyer would also be short sugar—being on a short side of a forward contract.

Being long on a forward contract is just one example of risky non-ownership. A futures contract, a total return swap, a stock loan, or a combination of a call and a put option will all result in a long position that is economically identical to that of an outright owner. Of course, there will be a short position on the other side of the contract. Because one need not own the asset to enter into a derivative instrument replicating its economics, these long and short positions may multiply indefinitely. The overall number of shorts will always be one less than the overall number of longs, reflecting the fact that one, but only one, of the holders of the long position actually owns the asset. But this insight does not help in deciding which one of many longs is the “real” tax owner. A tax system in which more than one taxpayer can claim full ownership of a single tax- advantaged asset would simply have no chance. No one would pay any tax because everyone would be able to shelter income with any combination of numerous tax preferences associated with tax ownership of property, such as depreciation, depletion and amortization deductions, the dividends received deduction, tax-exempt interest, foreign tax credits, and so on.[62] An alternative tax system could be conceived of that would allocate ownership among several taxpayers without duplication of its benefits and burdens, but this solution has generally not been adopted by Congress and the courts. In order for the existing tax system to survive, it must identify a single owner of each asset. Throughout the rest of this article, I will refer to this principle as the “single owner rule.”

This brings us back to Amalgamated Sugar. A rule that would treat a buyer of a tax-fungible asset under an executory contract as a tax owner before the specific property has been identified would not violate the single owner rule only as long as the seller owns the property sufficient to satisfy its obligations under the contract at the time when the contract is entered into and at all times until the property is delivered to the buyer. This appears to have been the case in Amalgamated Sugar. Obviously, this would not always be the case. Authorities dealing with executory contracts take it for granted that a seller in such a contract need not own the property to transfer its economics using a derivative instrument, such as a forward. Even if she does, because the property is fungible, there are usually no limitations on what the seller may do with the property while the contract is outstanding. Thus, a seller who owns the property at the inception of an executory contract may sell it the next day. If the buyer under the executory contract is already treated as an owner, how should the tax system treat the party who purchased the property from the seller while the contract remained executory? The single owner rule plays a critical role in the whether cases dealing with fungible assets; it is of little importance in the when cases considered in this part. Quite possibly, we should thank the courts that declined to follow the Amalgamated Sugar doctrine for sparing us from an inevitable tension between this doctrine and the single owner rule.

Finally, the concept of tax-fungibility helps in understanding why possession, a feature referred to as a critical factor in tax ownership analysis,[63]does not have an independent significance. Instead, possession resolves a fungible when case only if it may serve as a proxy for the ultimate inquiry—identification. This happens only if the asset in question is tax-fungible. If a buyer obtains possession of a tax-fungible asset at the inception of the contract, neither buyer nor seller would bother to substitute an otherwise identical asset in lieu of the one already held by the buyer. The buyer would be indifferent because the asset is fungible, the seller—because it is tax-fungible. Because the seller in this case has no incentive to identify and ultimately deliver an asset other than that initially delivered to the buyer, and because even if the seller makes the substitution, no change in the seller’s or the buyer’s economic or tax position would take place, a rule that treats the initially delivered asset as irrevocably identified is entirely reasonable. Price later contracts are the case in point.

The same is not true for tax nonfungible assets. A seller of a capital asset may wish to substitute another asset for the one held by the buyer prior to the closing of the sale because the substitution may change the seller’s tax consequences from the transaction. As long as the seller retains this right, it would violate the certainty rule to treat the sale as closed. Hence, Revenue Ruling 2003-7 concludes that a forward seller retains ownership of shares despite pledging them to the forward buyer and delivering them to a third-party trustee. Similarly, Revenue Ruling 72-478[64]holds that borrowing of the short seller’s long stock by a broker who executed the short sale does not close the short sale. While the broker holds the client’s stock and while the client will ultimately have to deliver identical stock to the broker to cover the short, the transaction remains open because the client hasn’t designated the stock borrowed by the broker as used to close the short sale. Thus, possession is not a dispositive factor in tax ownership analysis within the fungible when category. Identification, on the other hand, always determines the timing of ownership transfer in this context.


A. The Basic Standard

While the discussion in this part starts with the Supreme Court decision in Lucas v. North Texas Lumber,[65]this widely cited opinion hardly laid a foundation for the analysis of nonfungible when authorities. In a very brief decision, the Court concluded that even though a prospective buyer who held an option to purchase a piece of real estate notified the seller about his decision to exercise the option in late December, and even though upon receipt of this notice the seller ceased its operations and withdrew from the land, the sale was not consummated until early January of 1917. The seller, the Court explained, “did not prepare the papers necessary to effect the transfer or make tender of title or possession or demand the purchase price in 1916. The title and right of possession remained in it until the transaction was closed. Consequently unconditional liability of vendee for the purchase price was not created in that year.”[66]

In the absence of a detailed analysis from the Supreme Court, lower courts kept offering various approaches to resolving the timing question. After several tests were proposed, the decision in Fordyce v. Helvering[67]enunciated perhaps the most complete and convincing approach, effectively putting an end to the doctrinal disagreements in the nonfungible when context. In Fordyce, the court considered a dispute regarding the proper time when the acquirer’s stock received by the taxpayers-sellers in a tender offer should be valued to determine the amount realized from the sale.[68] The purchaser corporation made a public offer to purchase up to a certain amount of the target’s stock upon acceptance of the offer by a minimum amount of target shareholders. The taxpayer delivered her shares to a depositary before November 12, 1929. At that time, not enough shares had been deposited to make the offer self-executing, but the purchaser reserved a right to accept a smaller number of shares prior to November 22. On November 13, the acquirer exercised its right and declared the tender offer effective. Under the terms of the offer, the acquirer had until November 26 to deliver its stock and cash consideration. The target shareholders, including the taxpayer, received acquirer’s stock and cash on November 27. The purchaser’s stock closed at $24.62 per share on November 13 and at $29.94 per share on November 27. Needless to say, the Service argued that the consideration received by the seller must be valued when actually received, resulting in a substantially larger gain (and tax) for the sellers. After reciting all of the events that took place on or prior to November 13, the court reasoned as follows:

This was all done on the 13th and established the rights and liabilities of the parties, and we think it of no consequence that [the purchaser] had until the 27th for delivery of the shares and cash. When the conditions of the bargain were all met, as they were [on] November 13th, and the contract became binding, [the seller] lost all of his right to and control over the [target] stock which he had delivered to the depositary, and at the same time had the unconditional promise of [the buyer] to make delivery of the shares and money he was to receive in exchange. The exchange or sale, by whichever name it is called, then and there became binding on both parties, and the rights of both became fixed . . . . It was then that the parties to the exchange were clothed with beneficial ownership.[69]

A review of numerous authorities leads one to conclude that transfer of ownership of a nonfungible asset occurs when the transaction progresses to a point when the seller (1) has a right to recover the purchase price and (2) is unconditionally obligated to deliver the property being sold upon buyer’s performance (i.e., seller has no right to rescind), and the buyer (3) has a right to the property (i.e., can demand specific performance) and (4) is unconditionally obligated to pay the purchase price upon seller’s performance. That is, a sale occurs when the buyer’s and seller’s rights shift from the rights to their original property (seller’s right to the asset and buyer’s right to consideration) to the rights to the counterparty’s property (seller’s—to consideration, buyer’s—to the asset) and neither buyer nor seller have a right to reverse the transaction unilaterally. Finally, the parties may contractually change either of their respective rights and obligations.[70]

To be sure, authorities do not go to the trouble of reciting the entire four-prong definition suggested above. The most likely explanation is that although the four-prong definition is comprehensive, it is also likely to be redundant. Unless some special circumstances exist, there is no reason to expect that any one of the four prongs will be met while any other will not be. After all, each of the prongs reflects the same underlying state of affairs—a situation when a contract of sale “became binding on both parties, and the rights of both became fixed.” A finding that any of the four prongs is met on a particular date should generally suffice to support a conclusion that ownership was transferred on that date.

Although many nonfungible when cases are decided based on the suggested analysis, they virtually always cite the test set forth in Segall v. Commissioner:[71]

There are no hard and fast rules of thumb that can be used in determining, for taxation purposes, when a sale was consummated, and no single factor is controlling; the transaction must be viewed as a whole and in the light of realism and practicality. Passage of title is perhaps the most conclusive circumstance. Transfer of possession is also significant. A factor often considered is whether there has been such substantial performance of conditions precedent as imposes upon the purchaser an unconditional duty to pay.[72]

One readily recognizes in this passage one of many multi-factor tests that are so dear, it appears, to the hearts of many judges and, therefore, so prevalent in the law of taxation. It is instructive to note, however, that Segall itself was decided without much consideration of the factors described above. Other than reciting North Texas Lumber and discussing the intent of the parties, the entire reasoning of the opinion is as follows:

An executory contract was made on October 2, 1931. Some of the purchase price was [ ] paid and a promissory note for broker’s commission given, but [the buyer] did not have an unconditional right to the execution of the documents transferring title until it delivered or tendered the promised debentures on January 2, 1932; nor had [the seller] an unconditional right to the [consideration] until it had delivered or tendered the bills of sale contemplated. It follows, we think, under the doctrine of the [North Texas Lumber] case, that for taxation purposes the sale herein did not occur until January 2, 1932.[73]

The Segall factors proved to be a helpful addition to the tax ownership analysis in nonfungible when context, but not because they provide a formula to determine the timing of sale. Rather, they are useful in ascertaining when the rights and obligations of the parties become fixed. It is worthwhile, therefore, to consider how each of these factors, as well as some additional considerations added by the courts over the years, are taken into account in the ownership analysis.

B. Title, Possession, Benefits and Burdens, and Other Factors

Transfer of title is an important tax ownership indicator in nonfungible when cases. However, it is equally clear that timing of a title transfer does not always determine the timing of a sale. Courts have recognized that title may be retained by sellers without delaying a sale, most frequently to secure payment of the purchase price.[74]

Authorities usually cite possession as another crucial factor in the analysis. This factor is very helpful when a complete transfer of possession occurs in a single moment in time. In many cases this does not happen. Instead, the seller may retain possession subject to restrictions contained in the sale contract which may be quite substantial.[75] A buyer may obtain certain rights with respect to the property prior to obtaining an outright possession, such as a right to enter upon the property and to inspect it, to survey the property, to begin substantial construction, or to control the property together with the seller.[76] Finally, possession is not a particularly helpful indicator of ownership transfer when neither buyer nor seller actually use the property, such as when the property is leased to a third party.[77]

The benefits and burdens of ownership (in a narrow sense of economic exposure, as this term is usually applied by courts) are always taken into account in determining the timing of a sale. Courts focus frequently on the moment when risk of catastrophic loss is transferred, most likely because it is often specifically negotiated by the parties.

However, a careful consideration reveals that economic exposure by no means dominates the analysis, and the ultimate timing of the sale often does not coincide with the transfer of the economics. Several reasons combine to account for this feature of the nonfungible when cases. First, primarily in real estate cases, economic exposure, sometimes referred to as “equitable title”, shifts at a fairly early stage such as when the executory contract is signed. Courts acknowledge this shift, but conclude that ownership has not been transferred until a later date.[78] Second, transfer of economics frequently occurs gradually and it is difficult to pinpoint a single moment when the benefits and burdens shift. In cases of this type, the buyer and the seller share the benefits and burdens for some period of time as the ownership transfer unfolds.[79] Third, in some cases the ultimate economic exposure remains dependent on the completion of the sale, making it a particularly poor indicator.[80] Finally, on occasion, benefits and burdens are transferred by a separate agreement while the sale contract determines transfer of all other attributes of ownership, including title and possession.[81]

As the Segall test suggests, another important consideration is satisfaction of conditions precedent. The rule is simple: if a meaningful condition precedent has not been fulfilled, the contract of sale will remain executory. Conversely, an insubstantial condition will not delay the sale. In any case, the inquiry is not how many of the conditions have been met relative to the number of the remaining ones, but an assessment of the importance of those that are still not satisfied.[82]

Transfer of ownership may be delayed for yet another reason—uncertainty regarding the underlying property. This uncertainty may exist because a buyer has not had a chance to inspect the property and verify the information provided about it by the seller.[83] Uncertainty may also exist because a seller retains the right to change the property until a fixed future date, or because the property may change for a number of reasons other than a catastrophic event. In these circumstances, a buyer may protect itself by obtaining a right to call the sale off. Because it is impossible to predict whether the seller will change the property and, if it does, whether the buyer will exercise its right to rescind the contract, courts are reluctant to find a completed sale until this uncertainty is resolved.[84]

Payment of the purchase price is yet another factor taken into account by the courts in determining the timing of sale. Although payment of consideration is relevant because it demonstrates performance by one of the sides, it is hard not to notice that many, if not most, of the decisions consider situations in which a portion of the purchase price was either prepaid (i.e., paid before the sale closed),[85]or, more frequently, deferred (i.e., paid after the sale closed).[86] Putting aside the question of proper accounting for the amounts received, the timing of payment appears to have a fairly limited effect on pinpointing the exact moment of the tax ownership transfer.

Overall, review of numerous nonfungible when authorities leads one to conclude that there is no magic rule or overriding consideration. It is clear that the ultimate inquiry is whether the rights of each party to the counterparty’s property have become unconditionally fixed. Answer to this question depends on many different features, including transfer of title, possession, economic risks and rewards, payment of the purchase price, resolution of uncertainties about the asset in question, and satisfaction of material conditions precedent. The cases are frequently fact-intensive and involve balancing because a buyer and a seller often share important attributes, such as control and economic exposure, to a varying degrees. If there is one clear rule that emerges from these authorities, it is that the analysis is flexible and no single factor is controlling.

C. “Nonfungible” Securities

At first glance, some decisions discussed in this part appear to contradict the conclusions of the short sale, commodities futures and other fungible when authorities. Fordyce, for example, involved a tender offer, i.e., a sale of publicly traded stock. The court spent no time, however, pondering whether any specific shares were sufficiently identified. Is this approach consistent with the short sale, commodity futures and other authorities discussed so far? A closer look reveals that ownership analysis must be even more context-sensitive than it might have originally appeared. It is not sufficient to conclude that some types of assets are fungible and other types are not and that legal theories of tax ownership are different for each type. One also needs to take the next step and consider whether the same asset, such a publicly traded stock or an inventoriable commodity, may be both fungible and nonfungible depending on a particular factual setting. The Fordyce court took for granted that in the situation before it, there was no meaningful uncertainty regarding which particular shares would the taxpayer ultimately deliver.[87] Without any uncertainty, publicly traded target stock have become, in substance, identified.

Another example of how fungible shares may lose fungibility is Bradford v. United States.[88 ]Mr. Bradford learned about an intriguing business opportunity to purchase a large block of shares of Knights Life Insurance Company of America, 55,348 shares at $62.50 per share, to be exact. While the price seemed advantageous, Mr. Bradford simply did not have the funds to make this purchase. The solution he and his companion (the sellers) devised was to find a buyer for the stock before they bought the stock themselves. They were successful and on January 8 they signed an agreement with American General Insurance Company (the buyer). The agreement obligated the buyer to purchase all of Knights Life shares held by the sellers for $67.50 per share if the sellers acquired at least 55,000 shares prior to January 31. The sellers undertook to use their best efforts to purchase up to 75,000 shares of Knights Life. To enable the sellers to take advantage of a reduced tax rate imposed on long-term capital gains, the agreement provided for a closing more than six months from the date of its execution. The delay also satisfied some of the buyer’s business objectives. The sellers were required to deliver their shares to the buyer’s nominee, with transfer stamps affixed and subject to no lien or encumbrance.[89] On the same date, the sellers obtained a commitment from two banks to provide sufficient funds to purchase the amount of Knights Life shares necessary to satisfy the sellers’ agreement with American General. The transaction went as planned, but the IRS refused to go along. It argued that Mr. Bradford did not acquire a long-term holding period in the stock and won.

The Bradford opinion is a classic example of nonfungible when analysis. The court observed that the buyer’s obligation came into existence once the sellers met the only material condition—to purchase at least 55,000 shares of Knights Life. As soon as they did, “as a practical matter the sale was consummated” because:

American General’s liability was, as a practical matter, unconditional since the shares were in the name of its nominee, in form for good delivery and nothing remained to be done except the payment of the purchase price. On the other side of the fence, [the sellers] had no right to cancel.[90]

The court observed that the sellers’ profit from the sale was fixed at the moment of acquisition, that American General was fully exposed to any appreciation or depreciation in value of the Knights Life stock and concluded that the sellers “holding period for the shares here in controversy began and ended with their acquisition.”[91]

The decision makes perfect sense if the stock were a nonfungible asset. Although the court did not address the issue, the facts strongly indicate that there was nothing fungible about the Knights Life stock. The sellers had to borrow a considerable sum in order to purchase about fifty six thousand shares of this stock. They were contractually obligated to deliver all of it. More importantly, the agreement with American General obligated the sellers to deliver all of their Knights Life shares up to 75,000. The amount offered at a favorable price was only 55,348. On these facts, the likelihood that the sellers would actually deliver shares other than those originally deposited with the buyer’s nominee was remote at best. As a practical matter, no uncertainty existed about the asset being sold, so the court was correct in concluding that as a practical matter, the specific shares became identified the moment the sellers acquired the stock.

A connection between fungible and nonfungible when reasoning that underlies Bradford, but is not expressly addressed in this decision, is found in an old Board of Tax Appeals opinion dealing with a fungible commodity. Barde Steel Product Corp. v. Commissioner[92]was one of a series of cases dealing with deferred delivery contracts described earlier, but it had an interesting twist that led to an entirely different reasoning. The court in Barde Steel did not need to resolve the central issue raised in similar controversies like Amalgamated Sugar and Haas, i.e., whether specific fungible goods were unconditionally appropriated to the contract. The buyer in Barde Steel claimed, and the court assumed, that the buyer’s certification of specific steel served to identify it. The identification question being resolved, the buyer was looking for a quick approval of its ownership claim. The Board, however, had an entirely different question in mind. It proceeded with the analysis consistent with that enunciated later in Segall, distinguished Amalgamated Sugar, and found that a material condition precedent (certification by the seller) had never been met and, therefore, no sale took place when the parties entered into the sales contract.

Barde Steel addressed expressly what other courts took for granted. Commodities such as steel in Barde Steel, or securities such as publicly traded stock in Bradford and Fordyce, do not change their fundamental characteristics depending on the particular circumstances surrounding individual transactions. But for tax purposes, these circumstance may result in a de facto identification. This happens when as a practical matter, the likelihood that anything other than the specific asset held by the seller at the inception of the transaction will be eventually delivered to the buyer is so remote that it should be disregarded. Furthermore, while it may appear from the discussion of the fungible when authorities that identification of a fungible asset subject to a sale contract means an immediate ownership transfer, this conclusion, while often correct as a practical matter, would be generally mistaken. Identification merely resolves the first, albeit a central, issue relevant in determining tax ownership. Once the specific fungible asset is identified, it loses its fungibility and begins to “act” like a nonfungible one. A different analysis—the one found in Segall and Fordyce—must be applied next to determine whether the ownership of this “nonfungible” asset has been transferred.

To be sure, in most cases dealing with fungible assets nothing impedes the ownership transfer once the asset has been identified. Not surprisingly, most fungible when authorities simply skip the second step of the analysis. Sometimes, however, the second step cannot be ignored, as Bradford and Fordyce demonstrate by treating shares of stock as nonfungible assets and considering when the respective rights and obligations have become unconditionally fixed. Another example of this (implicit) two-step analysis is the prepaid forward revenue ruling discussed above.[93] In the ruling, the IRS focused not only on the lack of identification of the specific shares to be delivered under the forward, but also on the possibility that the forward would be settled in cash. The government must have recognized that identification of the shares would not resolve the ownership issue because there was a condition precedent (taxpayer’s right not to deliver stock at all) that had to be resolved before the taxpayer could be viewed as unconditionally obligated to deliver any stock, identified or not. In sum, there is no inconsistency between Bradford and Fordyce on the one hand and the cases dealing with cotton futures and short sales of stock on the other. To the contrary, the two lines of authorities are mutually reinforcing.

D. Differences Between Fungible and Nonfungible When Authorities

As the discussion in this and the previous parts demonstrates, analysis in the timing cases differs depending on whether the asset in question is fungible or not. If it is, the crucial issue is the uncertainty regarding the specific asset that will eventually be sold. The certainty rule and the single owner rule result in an open transaction treatment at least until the identification occurs. Because transferring the asset’s entire economics to the seller has no bearing on identification, it is irrelevant for the purposes of the fungible when analysis. The approach is very different if the asset is nonfungible. No uncertainty exists as to what is being sold. Instead, the focus of the inquiry is whether the respective rights and obligations of the parties have become unconditionally fixed. Title, possession, economic exposure, conditions precedent, payment of the purchase price, and verification of the asset are all relevant to the analysis with no single factor controlling.

Another way to delineate the difference between the two types of cases is to consider the remedies upon breach. The default remedy in the nonfungible world is specific performance. This remedy does not make sense if the asset is fungible, even if it is identified, and it is not discussed in the fungible when cases. A buyer of a nonfungible asset wants to become the owner of a specific asset. A villa in Hamptons would not do if one wanted a cottage on Cape Cod. But a buyer of a fungible asset wants to become an owner of an asset of a specific type having paid a specific price. It would make no difference to the buyer which specific bale of cotton or share of stock it receives and whether the particular bale or share certificate comes from the seller, or from anyone else, as long as the buyer retains the economic benefit of the bargain. Money damages would completely compensate the buyer in these circumstances. On the other side of the fence, a seller of a nonfungible asset wants to cease being its owner, often by a particular point in time. Offering the seller a sum of money and an advice to go find another buyer would not necessarily make the seller whole. If the asset is fungible, the seller would easily find another buyer, and as long as it receives the same price (whether from the new buyer or from the new buyer in combination with damages from the original one), the seller is happy. Finally, it is highly unlikely that seller of a fungible asset would ask for a right to take back the specific asset sold if the buyer defaults because the seller could easily obtain an identical asset elsewhere. Precisely for that reason it is unlikely that the seller would be granted this right even if it asked. If the asset is nonfungible, a court is much more likely to consider returning the asset to the seller who would not be able obtain it from anywhere else.

Finally, a typical fungible asset is “simple.” In a way, it has to be in order to be fungible. Sale of an IBM share or a Euro banknote is unlikely to be delayed by governmental approvals, inspections of the asset, financing and other contingencies. Once the asset is identified, there is just not that much that can go wrong with it. On the other hand, a typical nonfungible asset is “complex.” Its transfer is usually subject to meaningful conditions precedent. The asset may change over time and the sale contract may provide for various remedies should this occur. The buyer may require an inspection of the asset which takes time and extends uncertainty regarding the ultimate sale. One example of this distinction familiar to any transactional lawyer is a difference between a public and a private acquisition. The former, such as a tender offer described in Fordyce, is a relatively straight forward transaction. Other than insuring that the stock delivered to the buyer is valid and gathering public information about the target, the buyer’s advisers are somewhat limited in what they can do. A simple condition precedent may delay the transaction slightly (as it did in Fordyce itself), but the delay is unlikely to be significant. Things are quite different in a private deal. Verification of the asset alone—the infamous “due diligence”—may last for weeks. Lawyers negotiate for all sorts of conditions to closing that, if breached, allow the parties to rescind the deal. In no circumstances are the rights and obligations of the parties fixed until the deal is closed. The complex nature of nonfungible assets necessarily complicates the ownership analysis. On the other hand, the simple nature of fungible assets means that once a fungible asset has been identified, it is much more likely that its ownership will be transferred immediately or soon after the identification. In sum, there is a substantial difference in the tax ownership analysis of fungible and nonfungible assets in the when setting.


A. The Basic Standard

Whether authorities are not concerned with identifying the specific moment in time when the tax ownership is transferred from one party to another. Instead, they focus on whether the particular transaction resulted in a transfer of tax ownership at all. In the vast majority of cases dealing with nonfungible assets, at least one possible characterization of the transaction is a sale coupled with an additional contract such as a lease, an option, a management or agency agreement. Alternative treatments of the overall transaction include a loan, a lease, an agency, or a sham.

The discussion of the whether authorities reverses the previously established order and begins with the nonfungible cases. While fungible and nonfungible when cases are based on mostly independent reasoning, the fungible whether authorities have an important similarity to their counterparts dealing with nonfungible assets, but present an additional question critical to the analysis. It seems logical, therefore, to start with the simpler, nonfungible case.

It is rather ironic to begin with an assertion that the nonfungible whether setting presents a “simpler” case. Controversies of this type gave rise to many Supreme Court opinions, produced an extensive scholarship, and, most likely, involved the largest dollar amounts of all four categories discussed in the article. Yet, as the discussion will show, a single factor predominates the analysis in these decisions—economic exposure to the asset. The parties dispute what should be counted in evaluating this exposure and what should be the result stemming from a particular spatial and temporal division of the economics, but there is virtually no disagreement about the underlying assumption that economic exposure holds the key to tax ownership in this particular context.

One classic example of a nonfungible whether case is a leveraged lease transaction frequently used to finance equipment purchases. For decades, the government and the taxpayers argued about the circumstances in which a lease should be treated as a sale.[94] If a lease were recharacterized as sale combined with a loan from seller-lessor to buyer-lessee, the lessee would be treated the owner of the property and would be precluded from deducting the entire amount paid to the lessor as rent. The lessor, in turn, would be denied depreciation deductions. Although litigation of these issues started as early as 1920s, the Service made a fairly comprehensive statement of the criteria it viewed as relevant to the inquiry only in 1955 when it issued Revenue Ruling 55-540 to address “many new and unique types of agreements.”[95]

The ruling listed the following factors as relevant to the analysis: whether a portion of rental payments are made specifically applicable to an equity to be acquired by the lessee, whether lessee will acquire title once it pays all of the stated rentals, whether the rents are excessive, whether rents payable in a short period amount to a large portion of the purchase price, whether a lessee has an option to acquire the property for nominal value, and whether a portion of rental payments is specifically designated as interest. All of these factors probe the same issue: does the economic arrangement result in an acquisition of the property by the lessee or compel the lessee to acquire the property at the end of the lease. The former result obtains, for example, if rents amount to the entire purchase price such that lessee takes title at the end of the term. The latter arrangement exists when rents exceed market rents or when lessee has an option to acquire the property for nominal value.

A revenue procedure issued by the Service takes a different approach and concentrates more on the other party to the contract—the lessor.[96] According to this procedure, the IRS will not issue an advance ruling guaranteeing the intended tax treatment of the lease unless, inter alia, the lessor has a certain amount of “at risk” investment at the inception, throughout the term, and upon expiration of the lease, the equipment is reasonably estimated to have a meaningful useful life at the end of the lease term, the lessee has no right to purchase the property from the lessor for less than fair market value, and the lessee does not spend such a considerable amount of funds on improving the property that it is economically locked into purchasing it at the end of the lease.

Similarly to the government’s announcements, litigation arising from leveraged lease transactions does not raise fundamental questions regarding the factors affecting tax ownership. Rather, the taxpayers, the government and the courts consider the same criteria and disagree whether in a particular case the taxpayers pushed the envelope too far, giving the lessor too little economic exposure to the asset.[97]

A related and even more celebrated area of controversy involves sale-leaseback transactions. These are used most frequently as a substitute for secured borrowing by the seller-lessee and the leased property is usually a depreciable asset, such as machinery, equipment, or commercial real estate. The Supreme Court established an early precedent by concluding that a sale of real estate coupled with leasing it back for 99 years with an option to renew the lease and to purchase the realty for a specified price was in substance a mortgage.[98] A much more recent Supreme Court opinion dealing with a similar transaction—Frank Lyon Co. v. United States[99]—has been widely criticized for its lack of clear standards and the uncertainty it created regarding the importance of taxpayers’ tax avoidance intent.[100]

Although the Frank Lyon decision is hardly a useful guide for a detailed analysis of a particular transaction, it is helpful indeed for the purposes of our discussion. The Court strongly confirmed the “general and established principles” (applicable in the specific context of a case before the court, we might add) that location of title is of minor importance, that form does not control, and that economics play a dominant role in determining tax ownership.

[T]axation is not so much concerned with the refinements of title as it is with actual command over the property taxed—the actual benefit for which the tax is paid . . . . In applying this doctrine of substance over form, the Court has looked to the objective economic realities of a transaction rather than to the particular form the parties employed. The Court has never regarded the simple expedient of drawing up papers as controlling for tax purposes when the objective economic realities are to the contrary.[101]

In a buildup to its holding, the Court offered a 27-factor summary that in large part was referring to the parties’ relative economic exposure. The Court took into account, for example, reasonableness of the purchase price, reasonableness of the rentals, substantiality of the option to purchase and resulting uncertainty regarding its exercise, absence of any side agreements regarding the purchase option, absence of lessee’s liability on lessor’s indebtedness incurred to purchase the asset as well as lessor’ risk of asset depreciation and lessee’s default.[102] Post-Lyon authorities do not deviate from Lyon’s reliance on the economic analysis in determining tax ownership.

One of such decisions gave birth to the most oft-cited tax ownership test. The test enunciated in Grodt & McKay Realty, Inc. v. Commissioner[103]provides precious little support for the argument advanced in this article—it makes no reference to fungibility and mixes the inquiries relevant in nonfungible when and whether contexts. On the other hand, the decision itself is entirely consistent with other nonfungible whether authorities.

In this case, the Tax Court disregarded a sale completely (rather than recast it as an alternative transaction). The taxpayer, a corporation engaged in real estate business, purportedly purchased cattle having paid about three percent of the purchase price in cash and the remainder in a nonrecourse promissory note secured solely by the cattle. As the “owner” of the cattle, the taxpayer claimed investment credits and depreciation deductions based on the high purchase price as well as deductions for management fees and interest paid under the promissory note. In short, this was a classic late 1970s tax shelter. Summarizing the factors considered by the courts in determining whether ownership has passed from one taxpayer to another, the court listed the following “Grodt & McKay factors:”

(1) whether legal title passes; (2) how the parties treat the transaction; (3) whether an equity was acquired in the property; (4) whether the contract creates a present obligation on the seller to execute and deliver a deed and a present obligation on the purchaser to make payments; (5) whether the right of possession is vested in the purchaser; (6) which party pays the property taxes; (7) which party bears the risk or loss or damage to the property; and (8) which party receives the profits from the operation and sale of the property.[104]

The readers will recognize that the first, fourth and fifth factors made their way into this list from Segall and other nonfungible when authorities, which the Grodt & McKay court carefully cited. In fact, the fourth factor is the ultimate test used by these authorities to determine the timing of the transfer. It is not at all clear how a determination that an obligation to sell and to purchase is binding helps to decide whether a purported sale (whenever it takes place) should be treated as something else or disregarded entirely. Not surprisingly, the Grodt & McKay court did not use this factor in its analysis. On the other hand, the third, sixth, seventh and eighth factors all describe different aspects of economic exposure to the asset. The Tax Court relied on these factors in reaching its ultimate conclusion. As for title and possession, the purported owner had one (title) but not the other (possession)—a separation typical of a nonfungible whether case and more relevant in understanding why the ownership question arose than in resolving it.

Another setting in which the tax ownership issue appears front and center is a sale of a business coupled with an agreement retaining the seller to manage the business and making the purchase price payable only from the business’s profits. The famous (or infamous?) example of this transaction was addressed by the Supreme Court in Brown v. Commissioner.[105] The government argued that the buyer, a tax exempt organization, did not become an owner of the seller’s business because it made no investment and assumed no risk, all of which was retained by the seller. The Court disagreed and pointed out several factors demonstrating the buyer’s economic exposure to the business. First, the Court noted that upon payment of the stipulated price, the buyer would own the business outright. Second, if the business failed to produce a certain amount of profits and, therefore, the minimum installment payments were not made, the seller would have a recourse to the assets of the business. Finally, the Court pointed out that the purchase price was reasonable and resulted from arm’s length negotiations. Regardless of which side had the better argument, it is clear that both sides focused almost exclusively on allocation of economic exposure, and even more narrowly, on allocation of the risk from the business.

Another subset of nonfungible whether authorities deals with a situation where a seller is certain (or virtually certain) to repurchase an asset from a buyer either as a result of a contractual obligation or based on the facts and circumstances. Because the repurchase price is fixed, the buyer is not exposed to the upside or downside from the asset. Not surprisingly, the authorities find that tax ownership was never transferred from the seller to the buyer.[106]

The nonfungible whether authorities address many different factual settings. Creative taxpayers continue to design increasingly more complicated structures in order to transfer tax benefits of ownership.[107] However, the conceptual approach used by these authorities is remarkably consistent: they focus on allocation of economic risks and rewards from the asset in each specific transaction, balance each party’s exposure to the property and decide whether a purported owner has enough exposure to be respected as such for tax purposes.

B. “Nonfungible” Securities

As the discussion of the Bradford, Fordyce and Barde Steel decisions has demonstrated, it is possible in a when context that units of an otherwise fungible asset may become identified in a particular set of circumstances, making authorities generally applicable to nonfungible assets relevant in the analysis. It shouldn’t be surprising that similar identification may occur in a whether setting as well. For example, the taxpayer in Patton v. Jonas[108]entered into an option agreement to purchase a large portion of the preferred and common stock of O’Neil-Duro Company, but lacked the funds to exercise it. When a family corporation owned by the taxpayer, Mr. Patton, and his relatives, refused to lend money to the taxpayer, an alternative agreement was reached. The corporation purchased a large portion of the shares subject to an option and granted the taxpayer a right to purchase these shares at cost (i.e., granted the taxpayer a call). Mr. Patton, in turn, guaranteed to the family corporation that the dividends paid on the shares will be no less than the average rate of income earned by it from other investments. He also agreed to indemnify the corporation for any loss of principal from the transaction. The corporation retained a right to sell the shares of O’Neil-Duro Company to Mr. Patton at any time (i.e., it obtained a put). Eleven years later the taxpayer purchased the shares from the corporation for their cost plus the difference between the dividends paid during the period when the corporation held the shares and the agreed upon fixed return. Mr. Patton treated the transaction as a loan and deducted the difference as interest. The Service denied a deduction arguing that the corporation owned the shares for tax purposes.

The court agreed with the taxpayer. Focusing solely on the economics of the transaction, the court reasoned that the corporation had no risk from the shares because it could sell them to Mr. Patton at any time and because Mr. Patton guaranteed its principal investment. Similarly, the corporation had no opportunity to profit from the investment because of the taxpayer’s call. Finally, the corporation’s return was fixed and entirely independent from the dividends paid on the shares. The court concluded that the corporation was merely a conduit in plaintiff’s chain of title and, in essence, did nothing more than advance the necessary funds for the stock purchase in return for a defeasible title thereto. [The corporation’s] real economic interest in the stock contained all the essential ingredients of the ordinary security interest incident to the normal debtor-creditor relationship, as distinguished from those involved in a stock purchase.[109]

While the court did not address the issue of fungibility, it was quite clear that the specific shares purchased by the family corporation would be ultimately delivered to Mr. Patton. In addition to economic considerations making alternative dispositions unattractive, Mr. Patton was a stockholder, director, and vice-president of that corporation who, one would think, had sufficient control to prevent undesirable sales of these shares to third parties.

A very similar fact pattern was addressed in Green v. Commissioner.[110] Mr. Green was in the shoes of the corporation in Patton and his fellow attorney and friend was in Mr. Patton’s shoes. Substantively, the case was easier because instead of a combination of a put and a call, the parties in Green entered into a forward obligating Mr. Green, the initial buyer of the stock, to sell it to Mr. Smith within 12 months for a fixed sum. The shares involved represented practically all of the issuer’s stock, so there was no meaningful uncertainty regarding which specific shares will be delivered under the forward.[111 ]Neither the government nor the taxpayer argued for a loan characterization. The government asserted that the arrangement was a partnership, the taxpayer argued that the form should be respected and its 37 percent profit should be taxed at capital gains rates. Both the Tax Court and the circuit court concluded that Mr. Green was never a partner or the owner of the stock, but rather a lender, albeit at an exorbitant rate that would have violated the state usury law.

A final piece in the trilogy is Comtel Corporation v. Commissioner.[112 ]The case involved a public tender offer by Zeckendorf Hotels Corporation for the stock of Commodore Hotel, Inc. The buyer might have overpaid: over ninety one percent of the shares were tendered, but the buyer was unable to amass enough cash to purchase the shares. However, it found investors that agreed to create a new corporation—Comtel. Two investors and a Zeckendorf nominee contributed equal amounts of cash to Comtel which in turn secured a loan sufficient, together with the contributed cash, to purchase the Commodore shares. Zeckendorf borrowed these funds from Comtel, used them to acquire the Commodore shares, and sold them to Comtel for exactly the same price (repaying the transitory loan with the sale proceeds). As part of the transaction, Comtel gave Zeckendorf an option to repurchase these shares for a two-week period slightly more than six months after the date of the first transaction. In addition, Comtel shareholders entered into an indemnity and a subordination agreements that assured the two investors return of their capital plus a fixed profit of about 12 percent. Finally, Comtel was not permitted to sell or dispose of or pledge the shares, except to the bank that provided the original loan. In substance, Comtel was bound to delivery these specific shares to Zeckendorf if it exercised its call.

The court concluded that Comtel was a lender, not an owner of the stock. In addition to discussing the parties’ intent, the court observed that the investors made a risk-proof investment, that Comtel’s rights were similar to those of a mortgagee of a nonrecourse debt, that Comtel’s profit from the option had no relation to the value of the Commodore shares, but was in the nature of interest. In other words, consistently with other nonfungible whether authorities, the court analyzed the economics of the transaction. Discussing limitations on Comtel’s disposition of the Commodore stock, the court referred to it as “unique property,”[113]which it certainly was in the context of the case.[114]

While intermediaries in all three cases discussed so far were treated as lenders, a transitory stockholder may also be denied an ownership status if it is treated as an agent.[115] Not surprisingly, courts rely on lack of intermediary’s economic exposure to the stock, as well as control exercised over the intermediary’s actions by the principal, in concluding that the agent never became the owner for tax purposes.

Overall, cases discussed in this section demonstrate two points. First, a fungible asset may lose its fungibility and, if it does, the analysis applicable to nonfungible assets would govern the tax ownership inquiry. Second, the primacy of the economic analysis in the whether context does not depend on whether a particular asset is nonfungible, such as equipment or real estate, or fungible but identified, such as publicly traded stock.


A. Stock Loans and Subordination Agreements

The discussion of fungible whether authorities begins by focusing on a transaction not closely considered until now: a transaction that transfers tax ownership, but is not a sale. This transaction is a securities loan—a transfer of stock or other securities in exchange for a promise to return identical securities upon request. As discussed earlier, a stock loan is a necessary component of any short sale. Because short sales were wide- spread before the arrival of income tax, the question about the effect of a stock loan on tax ownership arose as early as 1915, and by 1925 it reached the Supreme Court.

The specific question in Provost v. United States[116]was whether a stock loan (and a return of a borrowed stock) constitutes a transfer of legal title to the shares, making it subject to stamp tax. Brokers of the New York Stock Exchange argued that a stock lender is in the same position as a pledgor of the stock. Because there is no transfer of title (meaning ownership) of the stock when it is pledged, there should be no stamp tax when it is lent. In addition, the brokers reasoned that because a borrowing broker always deposits with a stock lender cash collateral equal to the full market value of the stock, the lender should be viewed as borrowing money from the broker—a transaction specifically excluded from the scope of the stamp tax statute.[117]

The Court’s response to these arguments is quoted at length below for two reasons. First, it laid the foundation for the entire framework of analysis developed by the nonfungible whether authorities. Second, some of the critical points in the Court’s reasoning were glossed over by the later decisions, to their great detriment. Thus, it is important to set the tone with a relatively full expression of the Court’s view. The Court began by alluding to the economic features of a stock loan, recognizing that the stock lender retained the entire economics of the lent stock:

During the continuance of the loan the borrowing broker is [b]ound by the loan contract to give the lender all the benefits and the lender is bound to assume all the burdens incident to ownership of the stock which is the subject of the transaction, as though the lender had retained the stock.[118]

If this were a nonfungible whether case, the inquiry would have ended right then and there and the brokers would have won. Fungibility of the stock made a critical difference for the analysis. The Court went on to address the arguments of the brokers:

[The broker’s] arguments ignore the essential legal characteristics of the loan transaction. It may be agreed for the purpose of this discussion, . . . that the relation of the customer and the broker with whom the customer deposits stock as security for advances . . . is technically that of pledgor and pledgee, with authority and power on the part of the broker to repledge to the extent of his advance. . . . Although the broker has an implied authority to substitute other securities of the same kind and amount for the securities which he holds for his customer, and to repledge them to the extent of his advances, courts have not dispensed with the requirement that he should at least have, either in his own possession or lodged with his bank on the repledge, specific securities of the kind and amount purchased for this customer, available for delivery to the customer on payment of the balance due.

But the borrower of stock holds nothing for account of the lender. The procedure adopted and the obligations incurred in effecting a loan of stock and its delivery upon a short sale neither contemplate nor admit of the retention by either the borrower or the lender of any of the incidents of ownership in the stock loaned . . . . Unlike the pledgee of stock who must have specific stock available for the pledgor on payment of his loan, the borrower of stock has no interest in the stock nor the right to demand it from any other. For that reason he can be neither a pledgee, trustee nor bailee for the lender, and he is not one with whom stock has been deposited as collateral security for money loaned. For the incidents of ownership, the lender has substituted the personal obligation, wholly contractual, of the borrower to restore him, on demand, to the economic position in which he would have been, as owner of the stock had the loan transaction not been entered into.[119]

The analysis could not have been clearer: a pledgor does not become a tax owner of a pledged stock while a stock borrower does become a tax owner of a borrowed stock because the pledgor has a limited control over the pledged securities while the stock borrower’s control is complete. This result obtains even though a stock borrower gains no economic exposure to the borrowed stock, all of which is retained by a lender. In other words, control overrides economic exposure in determining tax ownership of a borrowed stock.

The Service followed this reasoning even before the Supreme Court had a chance to affirm the Court of Claims decision in Provost. The dilemma that the government faced was a classic example calling for an application of the single owner rule. While I introduced this rule during the discussion of fungible when authorities such as forwards, it was probably a stock loan that raised the issue of multiple potential owners of a single fungible asset for the first time. In the Solicitor’s Memorandum issued in 1925, the Service considered who should be treated as owning dividends paid on a borrowed stock.[120 ]There were two potential candidates: the stock lender (A) who lent the stock to a borrower (B), and a third party purchaser (D) who bought the stock from B (with B selling the stock short).[121] The Service observed that if after these transactions take place, the stock issuer pays a dividend, the rules of the New York Stock Exchange (NYSE) would require A’s broker to credit A’s account with an amount equal to the dividend and D’s broker would be required to do the same. Although there would be two dividend-equivalent payments, the memorandum reasoned, [t]here is but one dividend . . . . It follows that for the purposes of computing normal income tax only one purchaser is entitled to deduct it [i.e., is the owner of the dividend] . . . . The question is, as between A and D, Which of the two is entitled to take the deduction? It is the opinion of this office that D rather than A is alone entitled to take it. Although the “Street” designates the delivery of the certificate of stock from [A] to [B] as a “loan,” it was held in Provost Bros. & Co. v. United States, decided in the Court of Claims on December 1, 1924, that title to the stock passed to the borrower . . . . The credit entered in favor of A by [A’s broker] is, then, not a dividend but a sum of money measured by a dividend, and is credited to him not as an owner of stock but because of the terms of the contract . . . .[122]

Was A’s stock loan, therefore, a taxable disposition? Were payments received by A ineligible for tax-preferred treatment accorded to dividends? Not quite. The memorandum observed that brokers did not tell their customers when they borrowed the customers’ margin shares, and that the brokers held these shares in large pools, making it impossible to determine whose client’s shares were actually lent. Recognizing that “the administrative difficulties involved in the above ruling are considerable”, the memorandum concluded that “[i]nasmuch as ‘loans’ of stock incident to short sales are comparatively rare, it would seem expedient”[123]to permit the traders and brokers to ignore the reality and treat both A and D as owners as long as A’s broker does not know that she lent the shares specifically belonging to A.

Clearly, this was a ruling of convenience. What looked like a good compromise at the time lead to decades of uncertainty. The government used several different justifications to defend the result reached in 1925, growing increasingly frustrated with the issue. [124] In the end, Congress had to step in noting (in quite an understatement) that “uncertainty has developed as to the correct income tax treatment of certain securities lending transactions.”[125] To resolve the uncertainty, Congress enacted Section 1058. As long as a securities lending complies with its requirements, the lender recognizes no gain or loss from the transaction and her basis and holding period in the securities are unaffected by the lending.[126]

Notably, Section 1058 says nothing about tax ownership. If anything, it indirectly supports the conclusion that a securities loan is an ownership change because it provides for a nonrecognition treatment—something that would be needed only if the underlying transaction would (or at least could) be a realization event otherwise. In the end, despite the tortured history of securities lending, its tax consequences are clear under current law: stock loans transfer tax ownership, but are subject to a nonrecognition provision, at least as long as they meet the requirements of Section 1058.[127] Provost is as good a statement of the law today as it was in 1925.

Another set of disputes regarding tax ownership of fungible securities resulted from two acts of British government: the War Obligations Act of 1914 and the Financial Powers (U.S.A. Securities) Act of 1941. Because the Court of Claims decisions addressing each of the Acts are similar in most respects, I will focus on the more recent one. To finance its expenditures during the Second World War, the British government borrowed $425 million from the United States. The loan was collateralized by stocks of American corporations obtained by the British government from its citizens under the U.S.A. Act. Pursuant to this law, any British subject who owned American stocks was required to deliver them to the British Treasury, together with dividend orders and voting proxies. A death of a Briton whose shares were among those collateralizing the $425 million loan sparked a controversy with the IRS.

In Bickford-Smith v. United States,[128]the Service asserted that the plaintiffs’ decedent “owned and held” the stock of The Ensign-Bickford Company of Connecticut at the time of his death, and that an estate tax was payable on its value.[129] While the decedent continued to be the registered owner of the stock on the books of The Ensign- Bickford Company, he certainly did not hold the stock at the time of his death. Being a law-abiding citizen, the decedent delivered his shares to the British government and they were deposited, together with other stock, as collateral with the Federal Reserve Bank of New York where they remained at the time of the his death. In return for the shares, the decedent, as any other depositor, obtained a receipt that stated that his stock was placed “at the disposal” of the British Treasury, and that when the dividends are paid on the stock, the Treasury would pay an equivalent amount in pound sterling to the holder of the receipt. Furthermore, the Treasury was obligated to return the stock to the depositor, or to return “any security of that description” in lieu of the shares actually deposited, or, should the Treasury exercise its right to dispose of the shares, to pay the depositor fair market value of these shares.

The court recognized that the decedent retained full economic exposure to the stock, and that this constituted a “strong indicia of ownership:” “The right to income from property, and the chance of gain and risk of loss from later increases or decreases in its market value usually accompany ownership.”[130 ]However, the court concluded that it was the British government, not the decedent, who owned the stock:

[The decedent] had no right to the return of the stock, either soon or late. He had no right even that the stock be kept available, subject to the pledge to the [U.S. government], until the [British] Government either released it to him or extinguished any possibility of a release and substituted its obligation to pay him the then value of it . . . . [A]ll the documentary indicia of ownership [was] in the Government, plus a complete immunity from any claim for a return of the stock, and a complete power to do as it pleased with the stock at any time, being accountable only for paying the equivalent of dividends in the meantime, and for, at its option, paying the market value or returning the same or substituted shares at some time in the future. We think that the plaintiffs’ decedent had a chose in action, which the [ ] British Government could satisfy by alternative performances, at its option, and that he did not continue to own the stock.[131]

It is hardly surprising that this analysis is so similar to the discussion in Provost. In substance, the British government borrowed the shares of American companies from its citizens. Unlike a New York Stock Exchange broker-borrower, however, the government had no obligation to return identical stock—only its value in cash. The Service must have thought that it could limit Provost to stamp tax analysis to even litigate Bickford-Smith. It was mistaken. The Court of Claims expressly found that Provost applied and the ownership was transferred even though the stock of The Ensign-Bickford Company was neither actually sold by the British Treasury, nor transferred to the British Government on the issuer’s books. It was the power to dispose, not the actual

disposition, the court explained, that determined the outcome.

Cases and rulings dealing with subordination agreements have not made a profound impact on legal thought regarding tax ownership, although the Service may have given them a second life in recent revenue rulings.[132] In any case, these authorities provide another reminder about the importance of control over a fungible asset in a whether context. Disputes about tax consequences of subordination agreements were an indirect result of the minimum capital requirements promulgated by the several stock exchanges and the Securities and Exchange Commission. These rules required stock brokers that were members of the exchanges to have a certain amount of capital available for claims of their customers and general creditors. The subordination agreements were devised to meet these requirements without increasing the brokers’ debt. The agreements came in two main flavors—each discussed in a revenue ruling and a few cases.

Under the first type of a subordination agreement (I will call it a “note subordination agreement”), a client delivered to a broker a non-recourse “secured demand note” together with cash and marketable securities securing the note.[133] The value of the securities exceeded the face amount of the note. The client was free to withdraw any of its securities as long as it replaced them with other marketable securities or cash of equal value. The client retained the right to vote the shares and collect dividends and interest on the shares and bonds placed in the subordination account. It also was required to maintain the aggregate value of securities supporting the demand note at or above its face amount to avoid liquidation of collateral by the broker. If the broker encountered “an event of financial restriction” (such as insolvency, bankruptcy, or determination that the securities were needed to meet the capital requirements of the exchange), the broker could demand payment under the note and, if refused, liquidate the collateral and retain the proceeds, presumably not in excess of the face amount of the note. If the broker drew down the note, the broker was required to issue to the client either its shares or its junior debentures that would be subordinated to all claims of all present and future creditors of the broker.

Another type of a subordination agreement (I will call it a “pledge subordination agreement”) did not involve any debt obligations. Instead, a client deposited securities and/or cash into an account with a broker that was subordinated to all claims of all present and future creditors of the broker.[134] Unless an event of financial restriction occurred, the client retained full beneficial ownership of the securities, including the right to vote, the right to current income, and the right to withdraw the securities or dispose of them, provided that other securities or cash of equal value were deposited in the account. If an event of financial restriction occurred, the broker had a right to dispose of any securities in the account and to use the cash to satisfy claims of its other creditors. The client in this case would have a claim against the broker in an amount equal to the sale proceeds, but this claim would be the most junior claim against the broker.

Based on the authorities discussed in this section, one would expect that the first arrangement would be analyzed as a non-recourse loan secured by a pledge of fungible assets. The client would remain an owner of these assets until, if ever, it defaults under the note and the broker sells the collateral on client’s behalf to satisfy the client’s obligations. The second type of subordination agreement is a pledge that turns into something similar to a securities loan upon occurrence of a contingency. Thus, the client should remain the owner of the securities until, upon contingency, the broker becomes free to dispose of them regardless of whether the broker actually sells the securities.

The Service reached exactly these conclusions in two revenue rulings.[135] Unfortunately, neither ruling contained a legal discussion or cited any authorities. While the ensuing litigation resulted in some confusion,[136] the Second Circuit confirmed the government’s analysis of a note subordination agreement.[137] A case involving a pledge subordination agreement deserves a closer look.

In Miami National Bank v. Commissioner,[138]the Service argued that the client lost ownership of shares placed in a subordinated account with his broker, and, therefore, the client could not have transferred these shares to an acquirer. As a result, the acquirer did not own enough shares of the target to file a consolidated return with it. The broker holding the stock went bankrupt, but never sold the stock. Eventually, the client paid cash to the broker, withdrew the shares, and delivered them to the acquirer. The Service made two arguments. First, it contended that the client transferred the ownership of the stock to the broker when he delivered the shares pursuant to the agreement. Second, the government argued that, at the very least, the client lost ownership of the stock when the broker went bankrupt. The Tax Court rejected both arguments. As to the first one, the court pointed out that the government’s position had an unfortunate effect of contradicting its own revenue ruling.[139] The court reasoned that the client remained the owner of securities because he retained full economics of the stock and a right to substitute cash or other securities for it. It was particularly important for the court that, in addition to having a legal right, the client had sufficient means to substitute cash or other securities for the shares held in the subordinated account. In other words, his right of substitution was real.[140 ]The court responded to the second argument by noting that the stock was never sold—hardly a convincing answer, at least without additional explanation.

While not without their share of confusion, subordination agreement authorities are generally consistent with other fungible whether decisions discussed so far. Economic exposure to the underlying asset is relevant, but it does not dominate the analysis. The critical issue is control over the property in question, revealed by the power to dispose of it or substitute it for other property.

B. Sale Repurchase Agreements and Related Authorities

Controversies related to tax ownership of securities subject to sale-repurchase agreements have a long history. The first round took place in the 1930s, with both taxpayers and the government having some success. The parties sparred again in the 1970s, with the IRS gaining a decisive advantage. Finally, the government achieved, it appeared, a complete victory when it won a repo case in the Supreme Court in 1994.[141]

The first of several decisions issued by appellate courts in the 1970s is also the most important one because the later opinions relied heavily on its reasoning. The plaintiff in American National Bank of Austin v. United States[142]was a dominant player in underwriting local municipal bonds. Under the existing law, Texas municipalities had to refinance their bonds all at once, so a single concern or syndicate had to acquire all of the bonds in an issue.[143] Bank of Austin was such a concern. However, it could not legally sell the bonds to the public, making bond dealers another necessary link in the chain. On some occasions, bond dealers paid the purchase price for the bonds to the bank, the bank forwarded it to the municipality, received the bonds and delivered them to the dealers’ customers. In other words, the bank simply acted as an intermediary. In other cases, however, the dealers asked the bank to take the bonds “up for our account.”[144] Upon such request, the bank used its own funds to purchase the bonds on behalf of the dealers. The dealers offered the bonds for sale to their customers both before and after the bank paid for them, and a substantial portion of the bonds were usually sold before they were issued. Ordinarily, once the dealer sold a bond, it forwarded the proceeds to the bank adding, if necessary, a sufficient sum to purchase the bonds at their book value. All interest accrued on the bonds while they were held by the bank was retained by the bank. There were no written agreements between the bank and the dealers. However, during the years in question, the bank never refused to sell a bond to a dealer, and no dealer failed to pay the bank an amount equal to the value of the bond on the bank’s books regardless of the success or failure of the particular issue’s flotation.

While the court made no attempt to align its decision with other fungible whether authorities, such as Provost, the opinion leaves little doubt about the reasons for the ultimate conclusion:

[The bank] looked solely to the interest accruing on the bonds for its profit. The dealers profited if they could sell the bonds for more than their adjusted bids, but bore the risk that the bonds could not be sold for at least that much. In short, [the bank] was in effect a lender secured by collateral in its possession. . . . The dealer exercised complete dominion over the bonds after they came into the bank’s possession. He sold them at his pleasure, at prices he determined, and without reference to the bank, except that the proceeds were collected from the customer by the bank and applied to the dealer’s account. Obviously, the inventory of bonds was being held by the bank for the dealers and subject to their disposition. To us the bank was simply a lender of its funds to the dealers.[145]

The conclusion relies on two familiar arguments: the bank lacked economic exposure to the bonds and had no control over them. However, because both arguments point in the same direction, we are left with little guidance regarding their relative importance. This ambiguity is not the only problem with this opinion. First, basing its conclusions regarding lack of the bank’s economic exposure on the parties’ actions, the court neither stated explicitly that the absence of written agreements did not affect its analysis in light of the conduct of the parties, nor put any limitations on the court’s ability to treat the conduct of the parties as an equivalent of a legally binding agreement. Lack of analysis is even more problematic with respect to the court’s control argument. Why did it conclude that the dealers “exercised complete dominion over the bonds”? Was it because they “sold them at [their] pleasure” to their customers? But the dealers clearly did not deliver the bonds to the customers at the time of sale. At best, then, the dealers sold the bonds forward. Perhaps the court’s point was that the dealers must have been absolutely certain that they would obtain the bonds from the bank if they felt comfortable selling the bonds forward to their customers. However, a call on the bonds would be sufficient for that purpose. Was the court saying that a call right is equivalent to “complete control”? And if so, how would the court reconcile this with the conclusion that a stock lender ceases to own the stock despite being able to demand its return on short notice?

In light of such cases as Patton and Comtel, it seems very important that the bank could not itself sell the bonds to the public. That is, regardless of whether the dealers would have preferred that the bank held the bonds until they were delivered to the dealers’ customers, there just wasn’t much the bank could do with the bonds otherwise. Thus, as a factual matter, the uncertainty regarding which specific bonds would ultimately delivered by the bank was relatively small. However, the bank’s inability to sell the bonds was mentioned only once in a long factual description and the court did not refer to it in the analysis. It also appears that, as a matter of practice, the bank never oversold. That is, it never sold to the dealers more bonds than were issued by the municipality, covering its net short position by purchasing the bonds in the market.[146] If it did, there would have been uncertainty regarding which specific bonds (with potentially varying tax attributes) were actually delivered by the bank. Again, the opinion never addressed the question.[147]

The court in First American National Bank of Nashville v. United States[148]had to address two new arguments. First, the bank argued that the outcome in this case should be different because the bank was a licensed bond dealer, i.e., it could have sold the bonds to customers directly and did not have to rely solely on the bond dealers. The court saw no merit in this contention—perhaps a conclusion that was reached too fast in light of the authorities such as Patton, Green and Comtel. The bank’s second argument was based on North Texas Lumber. It argued that “ownership of the bonds in question did not pass to the [repo buyers], until those parties actually paid for the bonds and took delivery of them.”[149] The court dismissed this argument, explaining, in essence, that North Texas Lumber was a when case and, therefore, was entirely inapplicable to the whether case before it.[150] Although this distinction should be lauded as a great support for the four-category classification suggested in this article, the later discussion will show that perhaps the bank should have made its argument slightly differently, and perhaps the court should not have dismissed it so cavalierly.

The only decision that found a repo buyer to be the tax owner of municipal obligations (i.e., respected the repo’s form) also relied both on the economic exposure and control over the asset in reaching its conclusion.[151] The court reasoned that the “Bank had the right to sell the bonds to third parties and retain any profit made on such sale as its own property”[152]and that this was “a very important fact in this case.”[153]

Needless to say, it would have been helpful if the court explained whether it considered as more important that the bank retained the opportunity to profit from the bonds, had a right to sell them, or both.

A continuing uncertainty regarding taxation of repos eventually drew attention of the Supreme Court. Unfortunately, as much as the Court contributed to the analysis of fungible whether authorities in Provost, it failed to do the same in Loewenstein.[154] The specific question was whether interest earned by several mutual funds from repo transactions involving U.S. Treasury obligations was exempt from taxation by the State of Nebraska. The plaintiff in the case was a shareholder of two mutual funds that were repo buyers (and, arguably, money lenders). The mutual funds acquired the Treasuries from their counterparties and sold them back for the same price increased by accrued interest set at market rates unrelated to the coupon interest paid on the Treasuries. The repurchases took place either on a fixed date or upon demand of either party. The counterparties could substitute Treasury securities held by the mutual funds and the mutual funds forwarded coupon interest paid on the Treasuries to the counterparties.

The Court concluded that the mutual funds earned interest income from the loans made to their counterparties, not from the Treasuries they held, based on four features present in each transaction. First, the mutual funds paid a fixed sum to the counterparty at the commencement of a repo and the counterparty repaid that sum with interest at the repo’s termination. Second, if the counterparty defaulted, the mutual funds could liquidate the Treasuries, retain the proceeds up to the amount due from the counterparty plus expenses, and hold the counterparty liable for any shortfall. Third, the amount of the Treasuries held by the mutual funds was adjusted such that their value remained 102 percent of the original purchase price. Fourth, the counterparty could substitute other Treasuries for those held by the mutual funds as long as the incoming and outgoing securities had the same value.

The Court’s reasoning reveals that despite the criticism of its decision in Frank Lyon, it continued to use a descriptive, rather than an analytical, approach to resolving tax ownership issues. Other than repeating that the third and the fourth factors reflect an arrangement that would have existed in a debtor-creditor relationship, the opinion did not explain what meaning and relative weight it ascribed to each of the four factors. What could this meaning be?

Assuming that the Court implied, when it recited the first factor, that its description of what actually happened revealed a binding agreement existing between the parties from the inception of the transactions, the first factor stands for more than one proposition. It demonstrates, first, that that there was an agreement to repurchase, not just a put right held by the mutual funds or a call right held by the counterparty. It follows that the mutual funds could not freely sell the Treasuries to a third party without taking a short position in the Treasuries and exposing themselves to a potentially unlimited risk. It also follows that the mutual funds had neither risk of loss nor opportunity for gain from the Treasuries. Finally, it means that mutual funds’ return was fixed and guaranteed. The Court’s second and third factors only confirm the conclusion about the fixed and guaranteed return. The fourth factor, however, is highly relevant.

The counterparty in Loewenstein had the same right of substitution that the Provost court considered important and that leaves no doubt as to control over the Treasuries. Thus, all the relevant arguments flow from the four Loewenstein factors, and they are entirely consistent with the analysis of other fungible whether authorities. It would have been helpful, however, if these arguments, rather than a description of the arrangement, were stated expressly in the opinion.

The holding of Loewenstein is limited to determining the owner of interest derived by the mutual funds. “[T]he dispositive question is whether the [mutual funds] earned interest on ‘obligations of the United States Government,’ not whether the [mutual funds] ‘owned’ such obligations,” the Court explained.[155] Immediately thereafter, it addressed the tax ownership question:

Even if it did matter how repos were characterized . . . , Frank Lyon Co. does not support [taxpayer’s] position . . . . [O]ur decision in that case to honor the taxpayer’s characterization of its transaction as a “sale-and-leaseback” rather than a “financing transaction” was founded on an examination of “the substance and economic realities of the transaction.” This examination included identification of 27 specific facts. The substance and economic realities of the Trust’s repo transactions, as manifested in the specific facts discussed above, are that the Trusts do not receive either coupon interest or discount interest from federal securities by participating in repos. Rather, in economic reality, the Trusts receive interest on cash they have lent to the Seller- Borrower.[156]

Remarkably enough, the Court considered an analogy to Frank Lyon, a nonfungible whether authority, but failed to discuss Provost, a fungible whether case. After all, one of the arguments by the brokers in Provost was that a stock loan by a customer to a broker who deposited full cash collateral with the customer was the same as a money loan by the broker to the customer collateralized by the customer’s stock the exact transaction that, as the Court concluded, took place in Loewenstein. The Provost Court rejected the brokers’ argument because, unlike a money lender holding stock collateral as a pledgee, the brokers (stock borrowers) had no restrictions on disposition of the stock. One of the briefs filed in Loewenstein reveals that the mutual funds were contractually prohibited from selling the bonds. If the Loewenstein Court appreciated the overall similarity between the case before it and the Provost decision, and if it focused on the distinction between the brokers’ freedom to dispose of the stock and lack of the mutual funds’ freedom to dispose of the Treasuries, perhaps we would have a much clearer understanding of the importance of control in determining tax ownership of fungible assets. More importantly, the Court’s failure to distinguish Provost, combined with lack of attention to the control factor in the lower courts’ repo decisions, resulted in a considerable uncertainty regarding tax treatment of repos.

In addition to litigating, the Service issued several revenue rulings addressing repos and similar transactions.[157] While for the large part they were fairly uncontroversial in recasting particular repos as secured loans, the ruling that generated much excitement was the one holding that something that might have been treated as a loan was a valid sale. In Revenue Ruling 82-144,[158]the government concluded that a regulated investment company (a RIC) that purchased a portfolio of municipal obligations from a dealer and simultaneously acquired a small number of puts that gave the RIC a right to sell the obligations back to the dealer became the owner of the obligations for tax purposes. The puts were non-assignable, they were purchased for an arm’s-length price, had maturities substantially less that the remaining term of the obligations, and terminated if the RIC disposed of the obligations to which they related. The purpose for acquiring the puts was to provide liquidity.[159 ]The RIC was free to dispose of the obligations at will.

The Service stressed that the dealer could not demand a repurchase of the obligations (did not have a call to go with the RIC’s put) and did not solicit buyers for them. Analyzing the authorities, the Service concluded that “two significant factors of ownership are: (1) which party to the transaction has the right to dispose of the property; and (2) which party bears the risk of profit or loss with respect to the property.”[160]

Because the RIC was free to sell the obligations and retained “the full benefit of any appreciation in the value of the obligations,”[161]the Service concluded that the RIC became their owner. Thus, the ruling is yet another example of a repo-related authority enunciating a two-part ownership test based on economics and control without explaining the relative importance of each factor.[162]

C. Reconciling Fungible Whether Decisions

Cases dealing with repos and related authorities are an important part of fungible whether cases, yet they hardly establish clear principles. They refer to both the relative economic exposure and the parties’ control over the securities, but they do not provide a more nuanced analysis and in some cases the economic exposure seems to have an upper hand.[163] Despite this lack of clarity, the prevailing view among commentators is that the repo cases and rulings assume that the buyer would hold the specific securities

transferred to it by the seller until the repurchase.[164] The modern repo markets, however, function quite differently. The buyer is permitted to dispose of the securities received from the seller and return identical but different securities upon repurchase.[165] The government has acknowledged this reality, yet has made no attempt to state its current position. Should these modern repos be treated as loans too? If so, is there any way to reconcile this outcome with the treatment of securities loans that have been viewed as transferring tax ownership since Provost?[166]

There appears to be at least two ways to interpret repo authorities in light of Provost, Bickford-Smith and the subordination agreement decisions. One interpretation, that supported by the commentators, is that buyer’s power to dispose is inconsistent with the loan treatment. That is, a repo buyer with a power to dispose the securities is their tax owner, not a secured lender. There is certainly strong support for this conclusion,[167]but it is difficult to accept in light of the fact that some authorities failed to address this issue completely, including in circumstances when the buyer could have had such a power.[168]

Another approach is by no means revolutionary. In fact, one needs to go further back to find a basis for it. Back to Provost, that is. As the Supreme Court stressed in this opinion, the stock loan is different from a pledge not because a borrower has a right to dispose of the borrowed stock, but because it “holds nothing for account of the lender.”[169]

On the other hand, in case of a broker-pledgee, “[a]lthough the broker has an implied authority to substitute other securities of the same kind and amount for the securities which he holds for his customer, . . . he should at least have . . . specific securities of the kind and amount purchased for his customer, available for delivery . . . .”[170] Under Provost , therefore, a repo buyer need not hold the same securities that it purchased from the seller in order for a repo to be treated as a loan. It only needs to hold at all times identical securities in the sufficient amount.

If that is the rule, all repo authorities make more sense. The issue in all of them was taxation of the coupon interest paid on repo’d securities and collected, invariably, by repo buyers. To collect this interest, the buyers must have held the securities—maybe not the ones sold to them, but at least identical to those sold. This was such a basic fact, and it was so central to the entire controversy, that it is not particularly surprising that the courts did not focus on it. To be sure, there is a difference between being obligated to hold identical securities and holding them as a factual matter. But it can hardly be contested that it is easier to explain the repo authorities in a consistent manner that also comports with other fungible whether authorities if the Provost-based approach is adopted.[171]

Finally, the Provost test makes sense if we apply the basic principles discussed earlier in this article: the certainty rule and the single owner rule. Consider several different ways in which buyer or seller in a repo may control the securities in question. First, a seller may have a right to substitute the repo’d securities while a buyer has no power to dispose.[172]Here, the certainty rule supports the conclusion that no disposition takes place and the repo should be treated as a loan. The opposite result may create endless realization events upon every substitution which, as the experience has demonstrated, could be quite frequent.[173] On the other hand, there is no countervailing need to treat a repo as a sale to accommodate the single owner rule.

A second alternative is the complete opposite of the first one: a seller has no right to substitute while a buyer has an unlimited power to dispose.[174] Not finding a sale by the seller in this case would violate the single owner rule. If a repo is treated as a loan in this scenario and the buyer sells the asset to a third party, both that party and the repo seller would be treated as owners of the same asset—an unacceptable result. Not surprisingly, some authorities have expressly concluded that a party with a power to dispose is the asset’s owner, and the commentators have agreed that authorities that have not addressed the issue have assumed the same outcome.

The third scenario demonstrates why the Provost-based interpretation of the repo authorities comport with both the certainty rule and the single owner rule. In this scenario, a seller has no right to substitute, a buyer has a power to dispose, but must hold identical securities at all times. While the buyer is free to sell the asset received from the seller (“A”), the single owner rule is not violated even if the buyer does so as long as it holds an identical asset (“B”). The two assets simply switch tax attributes. Because other than tax attributes, the two assets are indistinguishable, nobody is affected by the switch. The seller that receives asset B back is indifferent as long as her holding period continues and her basis in the asset remains the same. The buyer is also indifferent as long as he is viewed as selling the asset that he owns, i.e., asset B, even if in fact asset A is sold. And the third party purchaser of the asset definitely does not care about the difference between A and B. Thus, the tax system may happily assume that no matter what actually happens, for tax purposes, it is always asset B that is sold by the buyer. If so, the buyer always retains asset A, and the seller may be treated as its continuing owner without violating the single owner rule.

There are two more combinations, both reflecting a sort of a stalemate. In the fourth scenario, a seller has no right to substitute and a buyer has no power to dispose.[175]

Treating the seller as a continuing owner in this case creates no problems with the single owner rule because the buyer cannot dispose of it. Of course, not finding an ownership transfer does not raise any issues with the certainty rule either. Hence, the authorities conclude that the ownership remains with the seller because that’s the party with the economic exposure.

Things are more complicated in the last, fifth scenario—quite unique but based in reality. Here, the seller has a right to substitute and the buyer has a power to dispose. This appears to have been the case in Miami National Bank[176]after the broker went bankrupt. The Tax Court rejected the government’s argument that buyer’s (broker’s) power to dispose resulted in a transfer of tax ownership and concluded that no sale took place because the shares were never actually sold. Perhaps, the court could have strengthened this conclusion with the following reasoning. Finding a current ownership transfer where the seller retains a right to substitute creates a problem with the certainty rule, unless the transfer is subject to a nonrecognition rule—something that cannot be assumed.[177 ]Not finding a sale while the buyer has a power to dispose violates the single owner rule. This Catch 22 might be resolved by observing that because neither solution is satisfactory ex ante, reasonable approach would be to base a decision on what occurred ex post. Because in the end the stock returned to the seller (subordination agreement customer), this approach would support the conclusion reached by both court. In essence, it would apply the open transaction doctrine to these unique circumstances.[178]

Overall, more than any other category, the fungible whether authorities support the caveat made in the beginning of this article—their outcomes are mostly consistent with the proposed contextual analysis, but their reasoning sometimes provides little support for it. Provost, Bickford-Smith, and some subordination agreement authorities strongly indicate that control determines the ownership analysis, rather than being one of the factors affecting it, the other one being economic exposure. Other subordination agreement decisions raise questions about the meaning of control. Repo and related authorities place substantial emphasis on the economics, occasionally refer to control, but ultimately do not provide a well-reasoned analysis. In the end, it seems that control should be viewed as the dispositive factor. This conclusion helps to reconcile all authorities within the fungible whether category. In addition, it is consistent with the certainty rule and the single owner rule.

D. Differences Between Fungible and Nonfungible Whether Authorities

It is hard not to notice that the analysis in stock loan and repo cases differs from the approach taken in sale-leaseback and related nonfungible whether authorities. Cases dealing with fungible assets are usually focused on identifying which of the two potential tax owners controls the asset. This issue is entirely absent if the asset is nonfungible. Instead, the tax owner is identified based on comparing the economic substance of the arrangement to the form chosen by the parties. Like the distinction between fungible and nonfungible when authorities, this disparity is caused in part by the fungibility itself. However, there is another difference between the two types of authorities that is unique to the whether context.

The analysis differs depending on the type of the asset because the separation between the big three tax ownership features—title, possession and economic exposure— usually depends on whether or not the asset is fungible. In a typical case dealing with a nonfungible asset, title is separated from possession and many, if not all, economic benefits and burdens of ownership. The question here is whether there is any reason to respect a title-holder as an owner (other than the fact of holding the title). This reason exists if the title-holder has some economic exposure to the asset, even if this exposure is less than that of the party possessing the property. Absent such exposure, title is a mere security device.

In a fungible asset case, uncertainty about tax ownership usually arises because possession and title are severed from economic exposure. The inquiry in this context is whether a party with both title and possession of the asset should not be respected as its owner. It should not if another party controls the asset in addition to being exposed to its economics. This outcome should not be surprising. Title and possession are generally important in tax ownership analysis precisely because a title-holder with possession ordinarily controls the asset. When it does not, these features tend to confuse, rather than clarify, the ownership inquiry.

Separating control from title and possession is only possible if the asset is fungible. The controlling party may substitute an otherwise identical asset, or even a different asset of the same value, without raising any objections from the counterparty possessing the asset because the counterparty is indifferent as to which specific asset it possesses as long as its has sufficient value. Thus, a client pledging stock to her broker, and a repo seller that retains a right to substitute securities held by the repo buyer, control the securities held by the broker and the buyer, respectively, and are treated as their owners. The outer boundary of control is a restriction (rather than a prohibition) on the power to dispose of the asset imposed on the party with title and possession that generally has the power to dispose. The restriction obligates this party to hold identical securities at all times.

Control manifests itself in the seller’s right to substitute and the buyer’s power to dispose of the asset. When either of these rights, but not both, is present, identifying the party with control, and, therefore, the owner of a fungible asset, does not present difficulties. When neither right is present, or when both rights coexist, neither party controls the asset and other ownership features should be brought into the analysis. This happens, for example, if a fungible asset becomes identified. The holder of the asset may not dispose of it because it cannot be certain that it would be able to reacquire the specific asset later. The substitution is impossible because there are no substitutes.

Of course, the stalemate in which neither of the two potential owners of a fungible asset controls it is precisely the conundrum encountered by the nonfungible whether authorities. They consider situations in which neither party controls the property because title is separated from possession. Not surprisingly, the solution is the same in both cases—the analysis turns to economic risks and rewards. Thus, identification of a fungible asset brings the tax ownership inquiry within the analytical framework of the nonfungible whether authorities. This is exactly what we have observed earlier while considering the when cases such as Bradford and Barde Steel: rules for nonfungible assets applied to a security or commodity in the when context once they lost their fungible nature.

This analysis reinforces a conclusion that the two-prong ownership test (i.e., control and economics) exemplified by Revenue Ruling 82-144 and present in various forms in many fungible whether authorities is a shortcut, not the real gauge. The control prong is not merely given more weight in the analysis, it answers the tax ownership question all by itself as long as it is possible to identify one party that controls a fungible asset. If neither party has control, or in a confusing scenario when two parties arguably control the same fungible asset, the inquiry turns to the benefits and burdens of ownership which, at that point, is the only factor in the analysis. This conceptualization of the test for fungible assets provides additional support for, and at the same time is supported by, the earlier conclusion that the nonfungible whether authorities ascertain tax ownership based exclusively on the parties’ relative economic exposure.

Another difference between fungible and nonfungible whether authorities arises from the distinction between simple and complex assets discussed earlier in the when context. The famous Grodt & McKay Realty test applying to nonfungible assets boasts eight factors and as many as four additional factors have been added to the mix. The Supreme Court’s record of 27 factors in Frank Lyon stands unsurpassed. Fungible assets simply do not have nearly as many features to consider, so it is not surprising that to Frank Lyon’s 27 factors Loewenstein mastered only four and Revenue Ruling 82-144 limited itself to two.

Finally, mostly for historic reasons, nonfungible whether authorities have a much stronger tax avoidance flavor than their counterparts dealing with fungible assets. The Supreme Court opinion in Brown supported a fairly aggressive and far-reaching tax avoidance transaction.[179] Sale-leasebacks played a prominent role in the tax shelter industry of the 1970s. The so-called lease-in, lease-out deals (or LILOs) and their progeny are examples of the latest wave of tax avoidance schemes. On the other hand, typical cases addressing tax ownership of fungible assets in the whether setting deal with transactions that are either primarily business-motivated or produce a widely-accepted result. To be sure, repos of municipal securities generated measurable tax savings and this fact was not lost on the courts. However, repos came into existence not as a tax play, and they continue to exist after the courts denied the contemplated tax savings. Likewise, although one might argue that avoiding gain recognition on disposition of a security when lending it to a broker produces a tax benefit, this benefit is now expressly conferred by Congress. For whatever reason, when the Service attacked clear tax avoidance transactions involving fungible assets, it simply chose not to raise the tax ownership issue and argued that the transaction lacked economic substance and was a sham instead. Although this strategy appeared to have worked in the early nineties, it backfired more recently in context of foreign tax credit arbitrage transactions discussed later.


A. Nonfungible Assets

Once the distinction is made between an inquiry into when a sale of a nonfungible asset results in transfer of tax ownership and whether a transaction transfers ownership at any point, it is difficult not to see the difference in the ownership analysis in these two contexts. Yet any suggestion that this demarcation line is clearly recognized by existing authorities is doomed to fall under an endless string of examples to the contrary. Thus, the following argument highlighting the difference is aspirational, at least in part.

To begin with the obvious, it hardly advances the analysis in a when case to observe that the parties called the transaction a sale and referred to themselves as a “buyer” and a “seller.” Unfortunately, nonfungible when authorities occasionally refer to these factors. On the other hand, it is difficult to see how a prototypical timing inquiry into “whether the contract creates a present obligation on the seller to execute and deliver a deed and a present obligation on the purchaser to make payments” helps to reach a conclusion in a whether setting. Nevertheless, this became one of the Grodt & McKay factors and every whether authority reciting these factors, and there are more than a few, refers to this inquiry as part of its analysis.

On a more fundamental level, the two types of cases are different because when cases are dynamic and whether cases are static. Resorting, not without hesitation, to the “bundle of sticks” metaphor,[180]one could say that in whether cases the sticks are clearly separated between the parties and the question is to determine whose sticks outweigh. In when cases, both the buyer and the seller are holding on to many of the sticks at the same time, with the seller’s grip weakening as the transaction progresses. Because of this difference, the whether authorities analyze the result while the when authorities consider a process. Thus, it is common in the timing cases to encounter a situation where both the buyer and the seller control the property in question to some extent and both parties have economic exposure to the property at the same time. This virtually never happens in the whether context. To be sure, a title-holder and a lessee are both exposed to the economics of the property subject to the lease, but their exposure is not contemporaneous.

Another difference is the role played by title and possession. In a vast majority of whether cases dealing with nonfungible assets, location of title and possession is predetermined and is always the same. Authorities conclude that basing the ownership analysis on this inevitable allocation would not be productive and focus on the economics instead. Thus, although both title and possession are among the Grodt & McKay factors, they usually have little bearing on the analysis in the whether setting. The situation is entirely different in the timing cases. In each such case, both title and possession are transferred from a buyer to a seller at some point. Identifying the exact moment of this transfer helps to determine when the rights and obligations of the parties have become fixed. Therefore, while title and possession play a central role in when cases, they are much less important in the whether context.

The difference in emphasis on the economic exposure follows necessarily from the above discussion. In circumstances where title and possession are inevitably split between the two potential owners, economic benefits and burdens is the critical remaining consideration by which to measure their relative claims. Not surprisingly, the nonfungible whether authorities focus painstakingly on the economics. The nonfungible when authorities deal with circumstance where all of the “sticks” composing ownership will move sooner or later, and there is no reason to base the decision on transfer of any one particular stick, including the economic exposure to the asset. Thus, the timing cases place less emphasis on the economic risks and rewards.

B. Fungible Assets

As we have seen, pinpointing the specific time when ownership of a fungible asset changes hands turns on the uncertainty regarding the asset in question. No transfer takes place until the specific asset is identified even if a seller transfers all of the economic exposure to the buyer. On the other hand, fungible whether authorities focus on identifying the party controlling the asset. Control usually manifests itself by either the seller’s right to substitute the asset or the buyer’s power to dispose of it.

It only takes a short step to observe that these two approaches are intimately related. Indeed, once a specific asset is identified as being covered by the contract, the owner of the asset no longer has a right to dispose of it or to substitute another asset for the one identified, i.e., the owner loses control at that point. Conversely, as long as a party retains a right to dispose of the asset or to substitute another one, it retains control over it. This means that the specific asset is not identified and uncertainty remains regarding which particular asset will be eventually delivered.[181]

However, ending the analysis here would produce an incomplete picture. The most obvious issue that calls for an explanation is that a large number of the fungible whether authorities repeatedly refer to benefits and burdens, while fungible when authorities mostly ignore this factor. Furthermore, there is a disconnect between the when and whether authorities dealing with identified fungible assets. In sum, despite sharing the same fundamental issue, fungible when and whether authorities have important differences. We turn to these differences next.

There is nothing surprising in the references to the benefits and burdens found, for example, in the repo cases and Revenue Ruling 82-144. One thing these references certainly do not mean is that fungible whether cases are decided based on economic exposure. As this article argues, existence of these references is only a shorthand for saying that if the control factor does not determine the owner, the answer would depend on the economic exposure because this is the test in nonfungible whether cases. Fungible when authorities could not possibly place substantial emphasis on economics not only because fungibility causes an entirely different issue to dominate the analysis. Another reason is that even if a specific asset is identified and is no longer fungible, the relevant frame of analysis is that of nonfungible when authorities. As we have seen, these authorities relegate economic exposure to a subsidiary role.

Identifying another difference between when and whether authorities is easier if we focus first on situations in which a specific fungible asset has been identified. Remarkably enough, a very similar fact pattern gave rise to at least two different conclusions regarding tax ownership. The simple transaction in question is a purchase of a fungible asset and a contemporaneous (or even preceding) agreement to sell the asset in the future. The transaction involves three parties whom I will call a Seller (the original owner of the asset), an Intermediary (a buyer / re-seller), and a Buyer (the forward purchaser). The court in Bradford held that the Intermediary—Mr. Bradford— was respected as a tax owner of the shares, but failed to acquire a long-term holding period in them because he only owned them for a brief moment. Intermediaries in Patton, Green and Comtel also acquired shares and entered into forward contracts (in substance, if not in form) to sell the shares in the future. Each of these cases held, however, that the Intermediary never became the owner of the shares for tax purposes, but was a lender to the Buyer. Can these decisions be reconciled?

The reasoning of the Patton-Comtel line of cases seems persuasive. Tax law readily disregards circular cash flows, transitory ownership of assets, and even transitory entities. Application of the most stringent variety of the step transaction doctrine—the binding commitment test—also suggests that an Intermediary that is unconditionally committed to sell the acquired asset should not be treated as owning the asset at any point. On the other hand, treating the Bradford transaction as a loan does not seem like a clearly correct result. After all, the Buyer in Bradford did not attempt to borrow from the Intermediary, and, most likely, it could have borrowed cheaper given a choice. The simple pattern addressed in Bradford and in the Patton-Comtel trilogy gives rise to a considerable tension between when and whether authorities primarily caused by a difficulty of converting a Buyer into an unwitting and unwilling lender. Perhaps, this difficulty also explains why Congress was so careful to avoid treating an Intermediary as a lender and a Buyer as a borrower in transactions that, in Congress’s own words, gave the Intermediary a return economically indistinguishable from that of a lender.[182] Ironically, where the lender is known, the tension persists if the fungible assets are not identified. Repos provide a good example.

Of course, the buy-and-sell-forward transaction also describes a repo. In a repo, a party selling the underlying asset (such as a municipality) is a Seller, a repo buyer (such as a bank) is an Intermediary, and a repo re-purchaser (such as a bond dealer, that may or may not be the same as the original seller depend) is the Buyer.[183] A sale contract that allows a seller to substitute freely another asset for the one originally delivered would be treated as executory by the when authorities because the specific asset has not been identified. Not surprisingly, the whether authorities conclude that repos in which the Seller has a right of substitution are loans—the Intermediary could not possibly be the owner of the asset regardless of any repurchase obligations because it has not acquired ownership in the first place. So far, the when and whether approaches are in harmony. If the Seller has no right of substitution, the first leg of a repo viewed independently would be treated as a completed sale according to the when reasoning. However, if the Intermediary must retain the specific securities received from the Seller (which, as many commentators believe, is a necessary prerequisite for treating it as a loan), these securities lose their fungible nature. Under nonfungible when principles,

Intermediary’s obligation to deliver these specific securities to the Buyer upon repo’s maturity is treated as an immediate transfer of the securities to the Buyer. Thus, according to the when reasoning, a repo is a sale to the Intermediary followed by an immediate re-sale to the Buyer. This could be treated exactly this way under Bradford. Fortunately, there is always a known lender (the Intermediary) and a known borrower (the Buyer) in any repo, so a loan characterization appears to fit better. Indeed, this is what the repo authorities conclude. The when and whether approaches are still not in conflict.

Things get complicated, however, if the Provost-based analysis leads one to conclude that a repo is a loan as long as the Intermediary must hold identical (but not the specific) securities throughout the repo term. In this case, the securities remain fungible in the Intermediary’s hands and the resale leg of a repo viewed by itself remains executory (no current sale) under the when reasoning. The implicit conclusion made in Patton, Green, and Comtel that a sale followed by an immediate resale should be viewed as a loan is no longer applicable because the resale is not immediate. In other words, treating repos as loans where a repo buyer is not obligated to hold specific securities received from a repo seller is inconsistent with when analysis of fungible assets. While possibly viewed as an argument against the Provost-based interpretation, this inconsistency may merely mean that tax ownership analysis is context-sensitive.

The readers may recall that a taxpayer came close to pointing out this inconsistency in Bank of Nashville.[184] Because in that case the bank (the Intermediary) was itself a licensed bond dealer, it was by no means certain that it would hold the specific bonds received from the issuers throughout the term of the repo. Instead of citing North Texas Lumber,[185]a nonfungible when case, the bank should have argued that under the short sale authorities, commodities futures rulings, and the rest of the fungible when decisions, the uncertainty about the specific bonds to be delivered in the second leg of the repo meant that this leg must have remained open until the actual delivery. Unfortunately, the bank failed to identify the relevant authorities, and the court missed an opportunity to consider the subtle distinction between a when and a whether inquiries.

As if one needed another confirmation that failing to address the issues doesn’t resolve them, the tension between when and whether approaches exemplified by the results in Bradford and the Patton-Comtel trilogy resurfaced again in a recent controversy discussed next.


A contemporary controversy raising the tax ownership issue involves a fairly simple transaction sold to several large U.S. corporations by Twenty-First Securities Corporation, an investment firm specializing in tax-advantaged strategies. For example, Compaq Computer Corporation executed it as follows:

On September 16, 1992, Twenty First, acting on Compaq’s behalf, bought ten million Royal Dutch ADRs from the designated seller, which was another client of Twenty-First. Twenty-First immediately sold the ADRs back to the seller. The trades were made in 46 separate New York Stock Exchange (NYSE) floor transactions—23 purchase transactions and 23 corresponding resale transactions—of about 450,000 ADRs each and were all completed in a little over an hour.[186]

All purchases were done pursuant to the special NYSE settlement terms and settled on September 17; all sales were executed pursuant to the regular terms of the exchange and settled on September 21. Compaq was a shareholder of record between September 17 and September 21. The brief period between September 17 and 21 was of no small significance because of the manner in which large corporations pay dividends to their shareholders.

Usually, before the dividend is paid, the issuer would declare the amount of the dividend, the record date, and the payment date (which is frequently several weeks after the record date). Owners of the issuer’s shares (or ADRs representing beneficial interest in the shares) at the close of business on the record date would receive the dividend on the payment date. The shares or ADRs would trade “cum dividend,” i.e., with price not reduced on account of the dividend, if the purchaser of a share would be entitled to the dividend. If the purchaser would become the holder of record after the record date for the dividend, the shares or ADRs would be purchased “ex dividend” and would be worth less than the “cum dividend” shares roughly by the amount of the announced dividend, all other things being equal. In Compaq, the dividend declaration date for Royal Dutch ADRs preceded September 16 and the dividend record date fell between September 17 and September 21. By executing the purchase pursuant to the special terms and the sales pursuant to the regular terms, Compaq assured itself of record ownership on the dividend record date.

Compaq did on purpose what most investors try to avoid at all costs—it bought high and sold low, realizing a short-term capital loss of about $20.5 million. Not coincidentally, it had an unrelated capital gain of over $200 million to shelter the loss.

On the other hand, Compaq received $22.5 million in gross dividend and paid about $1.5 million in fees. Furthermore, Compaq paid about $3.4 million in foreign withholding taxes and claimed the same amount as a foreign tax credit under Section 901. Because this tax credit far exceeded $640,000 of U.S. tax on the transaction,[187]the excess was available to offset tax on other income. The Service denied the tax credit arguing that the entire transaction lacked economic substance. The Tax Court held for the Service and assessed a negligence penalty against Compaq; the Court of Appeals for the Fifth Circuit reversed. A virtually identical transaction carried out by Twenty First on behalf of IES Industries followed a similar path, with the taxpayer winning in the Eighth Circuit.

It is hardly surprising that the Service chose to attack these transactions on the ground that they lacked economic substance. After all, this strategy was successful before, and the Service was determined to prove that foreign taxes should generally be treated as an expense for the purposes of an economic substance inquiry. One wonders, however, whether it was wise to stipulate, as the government did, that other than the economic substance issue, it had “no objection to how Compaq chose to report its tax benefits and liabilities concerning the transaction.”[188]

There is no shortage of opinions about the economic substance doctrine in general, and its application to the Compaq transaction in particular. Addressing this and related doctrines is beyond the scope of this article. Setting economic substance aside, it is hard not to notice that the tax ownership issue is front and center in the Compaq transaction. Although the government was not making the ownership argument, the Tax Court came close to addressing it when it asserted that Compaq “was acquiring a foreign tax credit, not substantive ownership of Royal Dutch ADRs.”[189 ]Moreover, the court’s holding—disallowance of the loss from sale of ADRs and of the dividend inclusion—is the exact outcome that would have resulted had the Tax Court concluded that Compaq never acquired the ADRs. However, the Tax Court rested its conclusion solely on the economic substance argument, and the circuit court never addressed the ownership issue. Perhaps, the court would have changed its conclusion if it asked itself whether Compaq owned the ADRs.

Following the approach suggested in this article, we should first observe that the asset arguably owned by Compaq—the ADRs—is a fungible publicly-traded security. The next question is whether, in the context of the particular transaction, the specific ADRs acquired by Compaq lost their fungibility. It seems clear that little, if anything, remained uncertain regarding the specific ADRs that were going to be delivered by Compaq on September 21. The uncertainty could have existed for one of two reasons: either Compaq had other identical ADRs that it could have delivered, or Compaq could buy such other ADRs and deliver them to settle the trades. It is not clear from the opinions whether Compaq owned any other Royal Dutch ADRs prior to September 16, but it appears fairly likely that it did not, or, at least, not in the amount comparable to 10 million ADRs acquired in the transaction. Any ADRs purchased on or after September 16 under the regular settlement terms would come into Compaq’s possession only after September 21. To be sure, Compaq could have conceivably purchased more ADRs on the special, next-day settlement terms used by it in the first place. However, Compaq would have to finance the purchase—something that it did not have to do in the transaction arranged by Twenty First because it bought and sold simultaneously. In that trade, Compaq managed to acquire 10 million ADRs worth almost $900 million by depositing about $17 million in a margin account and withdrawing it three hours later. Compaq would have been required to part with a much more substantial sum for a longer period to acquire additional securities in an amount that would create a meaningful uncertainty regarding which specific ADRs it would eventually deliver on September 21.

In sum, although not entirely clear, it appears very likely that although ADRs are fungible securities in general, the specific ADRs acquired by Compaq on September 16 were bound to be delivered to the purchaser on September 21. That is, as a practical matter, there was no possibility that Compaq would substitute any other ADRs for those it purchased at the inception of the transaction. Thus, in the specific context of the transaction in question, the Royal Dutch ADRs purchased by Compaq on September 16 were not fungible. Hence, the relevant tax ownership authorities are those dealing with nonfungible assets.

The essence of the Compaq transaction is very similar to the familiar pattern where a taxpayer purchases an asset and simultaneously contracts to sell it forward. Putting aside the settlement procedures (i.e., assuming that trades are entered into and settle simultaneously on their settlement dates), Compaq may be viewed as purchasing the ADRs on September 17 (the settlement date of the buy-side of the transaction) and simultaneously entering into a forward contract to sell the same amount of ADRs on September 21 (the settlement date of the sell-side of the transaction). Viewed this way, the transaction looks very similar to those considered in Bradford and Patton-Comtel, except that the forward in Compaq is much shorter than in any of those cases. Which line of authorities is more relevant?

Conceptualizing the Compaq transaction as a whether case immediately runs into a problem of offering an alternative characterization of the transaction. If Compq did not acquire ownership of ADRs for tax purposes, what did it do? The difficulty arises for two related reasons both arising from the fact that the transaction was executed in a public market. In order to argue by analogy to Patton that Compaq was a lender rather than the owner of the ADRs, one would have to identify the borrower. This would either be a party that both received the purchase price from Compaq and returned it to Compaq (presumably increased by the amount of the dividends paid during the term of the loan), or the party on whose behalf Compaq acquired ADRs and who acquired them from Compaq pursuant to a plan. Because the transactions were executed on a stock exchange, there could have been no assurance that the seller of the ADRs and their ultimate buyer would be one and the same person. To be sure, this was exactly what happened. Indeed, the circuit court described the transaction as a purchase “from the designated seller” followed by an “immediate[] sale . . . back to the seller.”[190] Yet, the court refused to disregard the interposition of the public market and treat the transactions as repos—a decision that was not surprising in light of a limited precedent for doing so. As long as there was no certainty that the seller and the re-purchaser were the same party, the first loan characterization appears difficult.

The second loan characterization is also problematic. No particular buyer requested Compaq to advance funds on its behalf. Even if the Compaq court looked “through” the market and concluded that the actual purchaser was predetermined, it would be difficult to conclude that Compaq acted as a lender because it never parted with $900 million—the presumed principal of the deemed loan. Because of the manner in which the trades were executed and the operation of the margin requirements, it only had to deposit a small faction of the ADRs’ price for a brief period. How could Compaq be viewed as lending $900 million to anyone when the money never left the Compaq’s coffers? Thus, there are considerable difficulties with applying the reasoning of Patton- Comtel line to the Compaq transaction.

Would Bradford be more helpful? It appears that an argument based on the nonfungible when authorities would be much more damaging to Compaq’s position as an owner. At the time when Compaq acquired the ADRs, it became unconditionally obligated to deliver them to a buyer. The ultimate question of the nonfungible when authorities is whether the respective rights and obligations of the parties have become fixed. On the facts of Compaq, there seems to be little doubt that this was the case. If so, like Mr. Bradford, Compaq lost tax ownership of the ADRs a moment after it acquired them and certainly before the record date for the dividend. Who was then the owner of the ADRs and the dividends? Their ultimate purchaser. The problem with the fact that this purchaser may be different from the original seller that needs to be addressed if the transaction is approached as a whether case, does not exist in the when context. In fact, this ultimate purchaser may not be a single taxpayer. Whoever this purchaser is, it, and not Compaq, owns the ADRs on the dividend record date. As for the cash flows, the ultimate buyer would be viewed as paying a portion of the purchase price in cash and the rest by assigning its dividends to Compaq. The Bradford conceptualization fits the Compaq transaction almost seamlessly, and produces the exact result argued for by the Service. After all, maybe it is more productive to let Bradford and Patton approaches coexist peacefully and apply each type in the appropriate setting.


Having the benefit of the analytical framework proposed in this article, it is worth taking another look at the existing tax ownership scholarship. It should come as no surprise at this point that the views of the commentators thinking about ownership are greatly affected by the type of transactions they consider. Those trying to discern the unifying principles while focusing on real estate and equipment transfers effected primarily through leases and sale-leasebacks, argue that the inquiry should be limited to economic benefits and burdens and offer detailed tests.[191] Others looking at transactions in stock and securities conclude that economics are all but irrelevant in determining an owner of a fungible asset.[192] While these differences underscore the unsettled state of the ownership doctrine, they are not inconsistent with the approach proposed here. All of these commentators are right and there is nothing inconsistent in their views, as long as the views of each group are limited to the ownership inquiry in a particular context.

A third group of commentators, however, suggests that a particular principle or set of principles applies to ownership transfers in general. The conceptual approach put forth by David Miller is not contrary to the approach proposed in this article.[193]

However, Miller ends up offering six rules of thumb that are fairly narrow and are tailored to specific transactions. Charles Kingson and Lee Sheppard make an argument that is far less consistent with my analysis.[194] Having considered the Compaq transaction in light of other ownership authorities, they argue that Compaq did not own ADRs at any time because it could neither gain nor lose from their price movements. In other words, “[w]hether the legal owner of a capital asset should be considered the owner for tax purposes is a more elaborate way of asking whether the legal owner is economically at risk.”[195] Neither Sheppard nor Kingson acknowledge that there is at least a question whether the Frank Lyon principles, or those enunciated in Provost, should govern the ownership inquiry in a particular context.

Professor Shaviro recognizes this uncertainty, but sees no way to resolve it. Considering whether a long counterparty in an equity swap (on a fungible underlying equity) should be viewed as the owner of the underlying equity, he asks whether the analogies based on short against the box authorities would dispose of the issue and concludes that the answer is essentially a matter of prediction. A court might reject this approach, suggested Shaviro, 809 on the ground that the tax treatment of fungible securities has no overall unifying logic, and thus [ ] the tax treatment of an equity swap need not be intellectually consistent with the tax treatment of a nonowner’s long position or a short against the box. The court might decide that precedents involving similarly structured transactions (for example, tangible property leases accompanied by put and call options between the lessor and lessee) were more germane than precedents involving different fact patterns but similarly fungible assets (such as a short against the box).[196]

This article is an attempt to resolve the uncertainty identified by Professor Shaviro. I argue that there is a unifying principle in the law of tax ownership, there are reasons why the ownership analysis of an equity swap should not be analogized to leases of tangible property, and these reasons rest on much more than the any specific line of authorities dealing with any particular transaction.

In sum, the law of tax ownership is vast, remarkably fragmented, and thoroughly confused. I argue that there is a substantial support for subdividing tax ownership authorities into four categories. Furthermore, I identify the tests relevant for each of the categories—the tests that are not expressly articulated by many of the authorities, but are largely followed nonetheless. This approach, I believe, clarifies the law of tax ownership by reconciling seemingly contradictory precedents, identifying critical issues, and focusing on relevant inquiries. On a more practical level, it helps to locate the appropriate precedents and to distinguish the irrelevant ones. Finally, and perhaps most importantly, the suggested framework should provide a useful guidance for resolving future difficult tax ownership questions which, if my experience is any indication, are certain to arise.

At the same time, one should fully recognize the limitations of this article. Recognizing the four categories and placing a particular transaction in the correct one by no means completes the tax ownership inquiry. Even if one accepts the tests proposed here for each of the four groups, it is clear that transactions within each of them are, and are likely to continue to be, full of subtle distinctions that will call for a nuanced analysis. Thus, using the contextual approach developed here would only help to make the first rough cut. However, as the discussion has demonstrated, making this first step can substantially advance the analysis.

No doubt, the proposed division is not the only “right” way to analyze the ownership issues. Tax ownership authorities present scholars with such rich and diverse material that it should be expected, not just possible, that other conceptual approaches will be suggested in the future. They will arrive not a minute too soon. For the moment, the goal of this article would be more than fulfilled if contextual analysis of tax ownership is used as a helpful tool in answering difficult ownership questions.

Alex Raskolnikov[1]

[1 ]Associate Professor, Columbia University School of Law. B.S., M.S., Mendeleev University of Chemical Technology, 1988; J.D., Yale Law School, 1998. I appreciate insightful comments from Marvin Chirelstein, Sam Dimon, Deborah Schenk, David Schizer, and Dana Trier. All mistakes are solely my own.

[2 ]Edward Kleinbard developed the fundamental difference between tax ownership of fungible and nonfungible assets in Risky and Riskless Positions in Securities—perhaps the most influential intellectual statement of a decade in this area. See Edward D. Kleinbard, Risky and Riskless Positions in Securities, 71 Taxes 783 (1993). However, as its title implies, Kleinbard’s article is limited to considering only one type of fungible assets—securities. Furthermore, Kleinbard devoted most of his attention to the analysis of the timing authorities. Leasing transactions, including sale-leasebacks and leveraged leases, have received a large share of attention. See, e.g., Richard E. Marsh, Jr., Tax Ownership of Real Estate, 39 Tax Law. 563, 574 (1985); Michael H. Simonson, Determining Tax Ownership of Leased Property, 38 Tax Law. 1, 2 (1984); Peter L. Faber, Determining the Owner of an Asset for Tax Purposes, 12 Taxes 795 (1983); Louis A. Del Cotto, Sale and Leaseback: A Hollow Sound When Tapped?, 37 Tax L. Rev. 1, 3 n.2 (1981)(citing

numerous earlier articles going back to 1940s). Several articles focus on certain celebrated cases. See, e.g., Bernard Wolfman, The Supreme Court in Lyon’s Den: A Failure of Judicial Process, 66 Corn. L. Rev. 1075 (1981); William H. Kinsey, Bootstraps and Capital Gain—A Participant’s View of Commissioner v. Clay Brown, 64 Mich. L. Rev. 581 (1966).

[3]See, e.g., Daniel Shaviro, Risk-Based Rules and the Taxation of Capital Income, 50 Tax L. Rev.

643 (1995); Noël B. Cunningham & Deborah H. Schenk, Taxation Without Realization: A “Revolutionary” Approach to Ownership, 47 Tax L. Rev. 725 (1992).

[4]There are at least two comprehensive discussions of tax ownership issues. See James M. Peaslee, David Z. Nirenberg, Federal Income Taxation of Securitization Transactions, at 68-85 (3d ed. 2001) (dividing numerous authorities into ten different groupings, including categories for short sales and for the timing cases); David S. Miller, Taxpayer’s Ability to Avoid Tax Ownership: Current Law and Future Prospects, 51 Tax Law. 279 (1988) (compiling a comprehensive source of tax ownership authorities and suggesting an approach to analyzing tax ownership cases). However, neither Peaslee nor Miller advocate the approach proposed in this article.

[5]See, e.g., Merrill v. Comm’r, 40 T.C. 66, 74 (1963).

[6 ]Practically every commentator who thought about tax ownership argued that the courts and the government should abandon consideration of the parties’ intent. See, e.g.,. Simonson, supra note 2, at 2 (calling inquiries into subjective intent a “red herring” and explaining that “[t]he search for intent, and the attention and discussion devoted to both the idea and the search, usually accomplish little and detract from the factors, identified or not, that in fact control the decision.”). Occasionally, courts themselves ridicule the parties or a lower court for basing a legal conclusion on the taxpayer’s intent. See, e.g., Major Realty Corp. v. Comm’r, 749 F.2d 1483, 1487 (11th Cir. 1985) (“the Tax Court’s questionable finding that the parties did not ‘intend’ to close is immaterial to the characterization of the transaction for tax purposes) (citations omitted); American National Bank of Austin v. United States, 421 F.2d 442, 451 (5th Cir. 1970) (“We note first that more than the legal opinions and otherwise self-serving testimony of the trial court, witnesses about their past intention to transfer ‘ownership’ is required under the circumstances here to rebut the presumption that the commissioner’s determination was correct.”).

[7]209 U.S. 365 (1908).

[8]Id. at 378-39 (citations omitted).

[9]See S.M. 4281, IV-2 C.B. 187 (1925).

[10]See S. Op. 1179, 1 C.B. 60 (1919).

[11]See Provost v. United States, 269 U.S. 443, 451-54 (1926).

[12]S. Op. 1179, 1 C.B. 60. A mutuum is “a loan for consumption,” see Black’s Law Dictionary (6th ed. 1990). Because it is expected that the asset will be consumed, it is necessarily acknowledged that the borrower will return an asset identical to the one borrowed, but not the very same one. In essence, a mutuum is a loan of a fungible asset. At common law, such a loan is treated as a sale, not a bailment, see id. If the specific asset that was lent must be returned to the lender, the loan is called commodatum—this is

a loan of a nonfungible asset. In a commodatum, a lender retains ownership of the asset, see id., at 937-38 (defining “loan for use”); see also, Kleinbard, supra note 2, at 787.

[13]For exactly the same reasons, one may enter into a forward contract to sell a share of stock in one year without identifying any specific share as being subject to the contract. Even if the forward buyer pays the purchase price, or a portion of it, when she enters into this forward (i.e., if the forward contract is prepaid), the buyer is not concerned whether the seller would be able to fulfill his obligations, as long as the buyer receives adequate collateral. Generally, this would be true even if at the inception of the forward the seller owns no stock at all. If an asset is nonfungible, expectations would be quite different. A request sometime in the fall to borrow a friend’s cottage on Cape Cod in order to sell it is likely to be met with skepticism. The cottage is unique and, once sold, cannot be replaced in the spring by returning a villa in the Hamptons.

[14]See Burnet v. Logan, 283 U.S. 404 (1936).

[15]See Treas. Reg. § 1.1233-1(a)(1) (“For income tax purposes, a short sale is not deemed to be consummated until delivery of property to close the short sale.”).

[16]See Rev. Rul. 2004-15, 2004-8 I.R.B. 515; see also Rev. Rul. 2002-44, 2002-2 C.B. 84.

[17]See I.R.C. §1233. As one commentator put it, “Section 1233 can be viewed as a legislative reenactment of the identification rule and the old cases holding that short sales against the box are not dispositions.” Lee A. Sheppard, Fixes to Ensure That Tax Is Paid on Capital Gain, 69 Tax Notes 1165, 1166 (1995).

[18]See Treas. Reg. § 1.1012-1(c)(2)-(7). I will refer to these rules as tax identification rules to distinguish them from the general tax ownership principle providing that tax ownership of a fungible asset is not transferred until the specific asset is identified. The scope of this principle is much broader than that of tax identification rules which, for example, do not apply to commodities and did not apply to debt instruments until well after the general identification principle had been established.

[19]This opportunity to choose is particularly helpful if the taxpayer wants to re-establish her long position. Instead of returning her existing long shares to cover the short (and, presumably, realizing taxable gain), and purchasing new shares to re-establish the long (presumably, resulting in a higher basis), the taxpayer could cover the short with the newly purchased identical shares, possibly realizing a smaller gain or a loss, and continue to hold the long shares that used to be “in the box” (albeit with a lower basis). The gain built into these shares will remain untaxed until the taxpayer eventually sells the shares, and it will escape tax entirely if the taxpayer holds the shares until death.

[20]See, e.g., Griffin v. Comm’r, 45 B.T.A. 588 (1941); Bingham v. Comm’r, 27 B.T.A. 186, 189 (1932); Rev. Rul. 72-478, 1972-2 C.B. 487.

[21]A short against the box executed by Estee Lauder on the stock of Estee Lauder Cos. deferred

(and could have avoided permanently, considering that Ms. Lauder was 87 years old at the time) more than $100 million of tax. See, e.g., Lee A. Sheppard, Fixes to Ensure That Tax Is Paid on Capital Gain, 69 Tax Notes 1165, 1166 (1995).

[22]The section clearly creates a deemed, rather than actual, realization event, providing that upon entering into a constructive sale, “the taxpayer shall recognize gain as if such position were sold,” Sec. 1259(a)(1) (emphasis supplied), and the holding period of property treated as constructively sold “shall be determined as if such position were originally acquired on the date of such constructive sale,” Sec. 1259(a)(2) (emphasis supplied). Legislative history is consistent with this analysis, see H.R. Rep. 2014, 105th Cong., 1st Sess., at 438 (1997).

[23]See Rev. Rul. 2003-31, 2003-13 I.R.B. 643 (confirming that there is no tax disposition upon entering into a short against the box).

[24]See Stevenson v. Comm’r, Docket No. 13449-02 (2002). The case quickly received publicity, including an article in Forbes magazine. See Janet Novack, Kinky Tax Tricks, Forbes, Sept. 30, 2002, at 50 (describing the type of transaction entered into by Mr. Stevenson as “a strategy apparently used by many wealthy individuals to raise cash and hedge risk while deferring taxes.”); Lee A. Sheppard, IRS Pursues Individual Constructive Sales Using Equity-Linked Securities, 96 Tax Notes 1797 (2002).

[25]The litigation was not entirely a surprise, in light of an earlier Field Service Advice addressing a similar (but not identical) transaction. See Field Serv. Adv. 200111011 (Dec. 6, 2000). For a detailed analysis of the FSA and an argument against the IRS’s position see Robert A. Rudnick & Michelle L. Petock, Forward Sale Contracts: The IRS’s Recent Attempt to View Code Sec. 1259 As a Trap for the Wary, Tax’n Fin. Prods 19 (Summer 2002). For a brief summary of the transaction and a view sympathetic to the government, see Lee A. Sheppard, IRS Pursues Individual Constructive Sales Using Equity-Linked Securities, 96 Tax Notes 1797 (2002).

[26]In another typical payout pattern, if the stock price exceeds $25, the seller will deliver a number of shares with a total value that is $500 less than the current market value of the stock.

[27]The primary purpose of introducing the uncertainty regarding the number of shares to be delivered upon maturity (i.e., the kinks) is to preclude application of the constructive sale rules of Section 1259 to the forward.

[28]2003-5 I.R.B. 363.


[30]In addition to having a right to substitute other shares, the seller’s ability to settle the forward in cash, assuming it was real, is a material conditions precedent that precludes completion of the sale under the nonfungible when authorities discussed in the next part. Although the ruling did not explain that the right to substitute the pledged shares and the cash settlement option were so important quite for the reasons just described, identifying these factors as crucial to the tax ownership inquiry was entirely consistent with the approach suggested in this article.

[31]See Rev. Rul. 2002-44, 2002-2 C.B. 84.

[32]See Rev. Rul. 66-97, 1966-1 C.B. 190.

[33]Currently, trades on the New York Stock Exchange executed “regular way” settle in 3 business days. See 17 C.F.R. § 240.15c6-1 (1993). Prior to 1995, this period has been as long as five business days and as short as one business day. See, e.g., Compaq Computer Corp. v. Comm’r, 277 F.3d 778 (5th Cir. 2001) (5 business days); Provost v. United States, 269 U.S. 443, 451 (1926); G.C.M. 12570, XII-5 C.B. 1934 (one business day).

[34]See G.C.M. 12570, XII-5 C.B. 1934.

[35]83 F.2d 257 (2d Cir. 1936).

[36]Id. at 257-58 (citations omitted).

[37]For a case where the uncertainty about specific shares to be delivered was more apparent (but also ignored by the court), see Dashiell v. Comm’r, 100 F.2d 625 (7th Cir. 1938).

[38]83 F.2d at 257 (emphasis supplied).


[40]Ironically, a decade after deciding Ruml, the same court expressly refused to rely on taxpayer’s intent in determining which shares should be treated as used to close a short position: “[The specific shares] remained under control of the taxpayer and up to the time of actual delivery could have been sold and replaced by other purchases . . . . [A] shifting intent to cover a short sale ought not to be the critical event which would determine gain or loss under a tax statute. It would leave the whole matter of fixing the event of the taxpayer’s own will. We hold that the time of delivery was the time at which the covering transactions must be regarded as closed.” Richardson v. Comm’r, 121 F.2d 1 (2d Cir. 1941) (emphasis supplied). Recently, the Service reaffirmed that the seller’s intent to deliver specific fungible asset under an executory contract does not result in a current sale of the asset. See Rev. Rul. 2003-7, 2003-5 I.R.B. 363.

[41]After losing several more cases, the Service adopted the Ruml approach, see G.C.M. 21503,

1939-2 C.B. 205, and later extended the rule to bonds, see I.T. 3442, 1941-1 C.B. 212. These decisions were later confirmed, see Rev. Rul. 66-97, 1966-1 C.B. 190; Rev. Rul. 70-344, 1970-2 C.B. 50. After some complications caused by the interaction with the installment sale rules were eliminated with the repeal of these rules, things returned to the status quo established in 1939, see Rev. Rul. 2002-44 (confirming the trade date rule, citing Rev. Rul. 66-97).

[42]A.R.M. 100, 3 C.B. 66, 70-71 (1920) (emphasis supplied).

[43 ]A.R.M. 135, 5 C.B. 67, 77 (1921).

[44 ]A.R.M. 100, 3 C.B. at 69 (citing Art. 1581, Regulations 45 which stated that “goods merely ordered for future delivery and for which no transfer of title has been effected should be excluded” for inventories).

[45]See Treas. Reg. § 1.471-1 (“A purchaser . . . should not include [in inventory] goods ordered for future delivery, transfer of title to which has not yet been effected.”). The industry eventually convinced the government to allow symmetrical treatment of physical commodities and commodities futures through marking them to market (but not on the grounds that the futures were part of the physical inventories). See A.R.M. 135, 5 C.B. 67 (1921). The Service reiterated all of these conclusion in a trio of revenue rulings issued in 1974. See Rev. Rul. 74-223, 1974-1 C.B. 23, Rev. Rul. 72-226, 1974-1 C.B. 119, Rev. Rul. 74- 227, 1974-1 C.B. 120.

[46]1974-1 C.B. 120.

[47]See G.C.M. 18658, 1937-2 C.B. 77.

[48]See Rev. Rul. 74-226, 1974-1 C.B. 119.

[49]72 F.2d 755 (10th Cir. 1934).

[50]Amalgamated Sugar, 72 F.2d at 758. Like all of its contemporaries that were addressing tax ownership issue in similar situations, the court phrased the ownership inquiry as a title transfer test.

[51]39 F.2d 381 (9th Cir. 1930).

[52]35 F.2d 880 (D.C. Cir. 1929).

[53]Haas, 39 F.2d at 382.

[54]See White Oak Transp. v. Comm’r, 24 B.T.A. 871 (1931) (deciding to follow Haas and distinguishing Amalgamated Sugar); see also Modesto Dry Yard Inc. v. Comm’r, 14 T.C. 374 (1950).

[55]Treas. Reg. § 1.471-1.

[56]See, e.g., Applegate v. Comm’r, 980 F.2d 1125, 1126 (7th Cir. 1992), Patterson v. Hightower,

245 F.2d 765 (5th Cir. 1957); Molsen v. Comm’r, 85 T.C. 485 (1985); Priv. Let. Rul. 8726007 (1987).

[57]For example, courts have characterized foreign currency forwards and futures as short sales and implicitly treated them as open transactions until they were assigned, terminated or settled. See, e.g., Carborundum Co. v. Comm’r, 74 T.C. 730 (1980) (agreeing with the IRS that a futures contract to sell foreign currency is a short sale subject to Section 1233, but rejecting an argument that a sale of such contract shortly before its expiration is subject to the rules of this section). The same is true for “when, as and if issued” contracts, i.e., forward contracts to buy or sell securities that have not yet begun trading on an exchange. See Rev. Rul. 57-29, 1957-1 C.B. 519 (“In computing the cost basis of assets for any purpose, the Internal Revenue Service does not recognize an obligation of a taxpayer reflected in an executory contract prior to the performance of the contract . . . .”). Similarly, the IRS excluded futures contracts from the coverage of a revenue ruling that allowed farmers to deduct in the year of payment the amounts paid for feed to be consumed in the following year, explaining that “[t]he purchase of commodity

future contract [ ] is considered to be the purchase of a right to acquire the specific commodity rather than a purchase of the commodity itself . . . .” Rev. Rul. 75-152, 1975-1 C.B. 144 (superseded by Rev. Rul 79-229, 1979-2 C.B. 210 without affecting the conclusion regarding commodity futures contracts). Finally, when the Service considered the appropriate method of calculating gross receipts of a subchapter S corporation engaged in commodity futures trading, it reiterated that “a commodity futures contract is merely an executory contract” and “the taxpayer [that offset one contract with another] has not purchased or sold the underlying commodity, but has merely liquidated its rights and obligations in the executory contract . . . .” Rev. Rul. 79-294, 1979-2 C.B. 305.

[58]Many fungible assets that are not capital assets for tax purposes (specifically, commodities) are subject to inventory accounting rules.

[59 ]Not all pricing conventions may be used with any ordering convention, however. For example, the lower of cost or market pricing is not available to taxpayers using last-in-first-out ordering rule. See I.R.C. § 472(b)(2).

[60]This statement is entirely true when the entire inventory is re-valued at market annually, and not quite true in other cases. It turns out, however, that majority of fungible assets that are subject to inventory accounting are likely to be covered by the mark-to-market convention. Securities and commodities represent a vast majority of fungible assets. Securities dealers have always been allowed to value their inventories at market, see Treas. Reg. § 1.471-5, and are now required to do so, see I.R.C. § 475(a). Securities traders are also permitted to mark to market their securities, see I.R.C. § 475(f). As will be discussed later in this section, commodities dealers and traders have been marking their inventories to market for over 100 years, see infra, text accompanying notes 42-45. Today, they are permitted to elect mark-to-market accounting by Section 475(e), (f). In all these cases, economically fungible assets are also fungible for tax purposes.

[61]See Kleinbard, supra note 2, at 787.

[62]This is not a mere speculation. In a particularly egregious example, when the Chrysler Corporation paid an unusually large dividend, purported “owners” of the stock claimed a dividends received deduction in respect of 160% of dividends actually paid. See Kleinbard, supra note 2, at 787, n.18.

[63]See, e.g., Miller, supra note 3, at 285.

[64]1972-2 C.B. 487.

[65]281 U.S. 11 (1930).

[66]Id. at 13.

[68]While stock is a archetypical fungible asset, cases such as Fordyce should be analyzed under the nonfungible rubric, see infra, Part III.C.

[69]Id. at 434.

[70]See, e.g., Lucas v. North Texas Lumber, 281 U.S. 11, 13 (1930) (sale didn’t take place in the earlier year because “unconditional liability of vendee for the purchase price was not created in that year”); Major Realty Corp. v. Comm’r, 749 F.2d 1483, 1486-87 (11th Cir. 1985) (concluding that buyer became unconditionally obligated to pay the full purchase price on the early date, and citing North Texas Lumber for the proposition that “an unconditional liability [is] a benchmark when a transaction is closed for tax purposes”); Claiborne v. United States, 648 F.2d 448, 451 (6th Cir. 1981) (concluding that ownership was transferred because the seller could have forced the buyer’s payment rather than rescission of the contract); Bradford v. United States, 444 F.2d 1133 (Ct. Cl. 1971) (sale completed at an early date because the buyer “had an absolute right to title” upon payment of the purchase price); Wiseman v. Scruggs, 281 F.2d 900, 902 (10th Cir. 1960) (holding that the contract was not executory because it “created a present obligation on the part of the sellers to execute and deliver deeds of conveyance as installment payments were made [and] a present obligation on the part of the purchaser to make the installment payments . . . .”); Rich Lumber Co. v. United States, 237 F.2d 424, 427 (1st Cir. 1956) (holding that no sale took place on an early date because the buyer “during that year was not unconditionally and irrevocably bound to take the property and pay the agreed price . . . .”); North Jersey Title Ins. Co. v. Comm’r, 79 F.2d 492 (3d Cir. 1935) (finding a sale at an early date because “the liability of vendee was unconditional in the earlier year”); Helvering v. Mibley-Mimnaugh Lumber Co., 70 F.2d 843, 845 (D.C. Cir. 1934) (holding that if the seller was an accrual basis taxpayer, it must have accrued the purchase price on an early date because “in addition to delivery of the property, the seller had otherwise complied with his contract, as the result of which there existed an unconditional obligation on the buyer to comply”); First American Bank of Nashville v. Oman, 209 F. Supp. 902 (M.D. Tenn. 1962) (holding that no sale took place at an early date because the buyer at that time had no unconditional duty to pay); Perry v. Comm’r, 35 T.C.M. (CCH) 1718 (1976) (holding that the sale took place on the date when the seller became entitled to buyer’s stock); Merrill v. Comm’r, 40 T.C. 66, 76 (1963) (concluding that the sale took place prior the passage of title because “subject to [the seller’s] tender of the remainder of the purchase price, they could have forced conveyance of the legal title”); Morco Corp. v. Comm’r, 20 T.C.M. (CCH) 305 (1961) (holding that a sale does not occur and a loss may not be deducted until the buyer performs to the extent that would “impose[] upon the purchaser an unconditional duty to pay”); Rev. Rul. 73-369, 1973-2 C.B. 155 (sale is not completed at an early date because “buyer was not obligated to complete the sale and burdens and benefits of ownership remained with the seller”).

[71]114 F.2d 706 (6th Cir. 1940).

[72]Id. at 709-10.

[73]Id. at 710 (emphasis supplied).

[74]See, e.g., Wagner v. Comm’r, 518 F.2d 655 (10th Cir. 1975) (sale on the early date despite executed warranty deed remaining in escrow until the late date); Maher v. Comm’r, 55 T.C. 441, 452 (1970) (retention of title by seller for security purposes should be viewed as transfer of title and taking back a mortgage); Clodfelter v. Comm’r, 48 T.C. 694 (1967) (seller’s retention of title until all payments have been received does not defer the sale).

[75]See, e.g., Int’l Paper Co. v. United States, 33 Fed. Cl. 384, 394-95 (1995) (a list of restrictions imposed on the way in which seller could run its business prior to the closing date; sale found on the late date); Rich Lumber Co. v. United States, 237 F.2d 424 (1st Cir. 1956) (seller retained possession of timber lands, but could not diminish their value, such as by cutting and removing timber)

[76]See, e.g., J.B.N. Telephone Co. v. United States, 638 F.2d 227 (10th Cir. 1981) (after signing the sale contract, seller continued to operate the manual telephone equipment while buyer installed the automatic dialing equipment); Rich Lumber Co. v. United States, 237 F.2d 424 (1st Cir. 1956) (buyer could survey the property prior to completion of the sale); Harmston v. Comm’r, 61 T.C. 216 (1973) (seller retained the property, buyer was given right to inspect at any reasonable time); Griffin Paper v. Great Northern Nekoosa Corp., 74 T.C.M. (CCH) 559 (1997) (seller’s representatives occupied two seats on the board of directors, an office position, and one seat on the executive committee, no sale until the late date).

[77]See, e.g., Wagner v. Comm’r, 518 F.2d 655 (10th Cir. 1975) (sale on the early date despite the fact that buyer did not obtain title or possession until the late date, a third party lessee possessed the property until the late date).

[78]In Major Realty Corp. v. Comm’r, 42 T.C.M. (CCH) 373 (1981), rev’d on other grounds, 749

F.2d 1483 (11th Cir. 1985), Tax Court gave the following explanation: “[T]he doctrine of equitable conversion, [provides that] equitable title passes to the purchaser at the time a contract is signed (so that any loss or damage to the property befalls the purchaser) . . . . It is clear that a contract to sell real estate which operates to invoke the doctrine . . . is insufficient of itself to effectuate a completed transaction for tax purposes since the transfer of title and full payment were conditions to the completion of the transaction.” Id. at 381, n.15.

[79]See, e.g., Harmston v. Comm’r, 61 T.C. 216 (1973) (seller of orange groves retains many benefits and burdens, buyer assumes risk of damage to the trees by acts of God, other than frost).

[80 ]One instance when this frequently occurs is when a business is bought based on a balance sheet, i.e., the purchase price is set by reference to the target company’s balance sheet as of a certain early date and later adjusted only for extraordinary changes. Thus, the acquirer has full economic exposure to the target company provided the deal goes through, and no exposure if it does not. See, e.g., Segall v. Comm’r, 114 F.2d 706 (6th Cir. 1940) (executory contract made in October, sale took place in January of the following year, purchaser assumed liabilities of the target based in August balance sheet).

[81]See, e.g., Penn-Dixie Steel Corp. v. Comm’r, 69 T.C. 837 (1978) (benefits and burdens transferred in large part by a combination of seller’s put and buyer’s call on the early date, no ownership transfer until late date); Griffin Paper v. Great Northern Nekoosa Corp., 74 T.C.M. (CCH) 559 (1997) (same);

[82]See, e.g., Int’l Paper Co. v. United States, 33 Fed. Cl. 384 (1995) (no sale until transaction approved by the Interstate Commerce Commission); Merrill v. Comm’r, 40 T.C. 66 (1963) (no sale until seller completes repairs, but issuance of a title insurance policy is not a material condition where the same title insurance company issued the policy on the same property two months earlier); Perry v. Comm’r, 35

T.C.M. (CCH) 1718 (1976) (no sale until approval by buyer’s directors and stockholders).

[83]See, e.g., Helvering v. Mibley-Mimnaugh Lumber Co., 70 F.2d 843 (D.C. Cir. 1934) (uncertainty about the amount of standing timber affected the purchase price, timing of sale depends on whether the seller is a cash or accrual basis taxpayer).

[84]See, e.g., Int’l Paper Co. v. United States, 33 Fed. Cl. 384 (1995) (no current sale where buyer, among other things, retained the right to rescind the contract if there were any material loss, casualty, or adverse change with respect to the target company).

[85]See, e.g., Morco Corp. v. Comm’r, 300 F.2d 245 (2d Cir. 1962) (about 3% prepaid in 1952 and

1954, no sale until 1955); Segall v. Comm’r, 114 F.2d 706 (6th Cir. 1940) (prepayment of about one third of consideration in 1931, no sale until 1932); Doyle v. Comm’r, 110 F.2d 157 (2d Cir. 1940) (prepayment of about one third does not trigger ownership transfer).

[86]See, e.g., Clodfelter v. Comm’r, 426 F.2d 1391 (9th Cir. 1970); Baertschi v. Comm’r, 412 F.2d

494 (6th Cir. 1969) (about 70% of the price remains unpaid at the time of sale). Recognizing this reality, Congress enacted Section 453 that specifies a method of including purchase price in income where part of the price is paid after the sale.

[87]This belief appears entirely rational for several reasons. First, because the taxpayer delivered all of his shares, it could substitute any other shares only from new purchases in the market that was likely to be fairly thin. Second, a market purchase would expose the taxpayer to a risk that the tender offer would not succeed, the market price would drop to what it had been before the tender was announced, and the taxpayer would end up holding extra shares for which he overpaid. Third, it appears unlikely that a taxpayer who decided to participate in the tender offer would acquire additional shares of the target and not tender those shares as well. Because there was no meaningful uncertainty regarding which specific shares would be delivered to the acquirer, there was no reason to treat the shares any differently than a parcel of land or a building, which is exactly what the court did.

[88]444 F.2d 1133 (Ct. Cl. 1971)

[89]The agreement provided further that all stock dividends would be delivered to the buyer together with the shares, cash dividends paid on the shares prior to the closing would be retained by the sellers, and the excess of cash dividends retained by the sellers over the cash dividends paid during the same period of the preceding year would reduce the purchase price.

[90]444 F.2d at 1143.

[91]Id. at 1144.

[92]14 B.T.A. 209 (1928).

[93]See supra, text accompanying notes 28-30.

[94]See, e.g., Northwest Acceptance Corp. v. Comm’r, 58 T.C. 836 (1972), aff’d per curiam, 500

F.2d 1222 (9th Cir. 1974); Lockhart Leasing Co. v. Comm’r, 54 T.C. 301 (1970); Judson Mills v. Comm’r, 11 T.C. 25 (1948); Holeproof Hosiery Co. v. Comm’r, 11 B.T.A. 547 (1928).

[95]1955-2 C.B. 39.

[96]See Rev. Proc. 2001-28, 2001-1 C.B. 1156.

[97]See, e.g., Swift Dodge v. Comm’r, 692 F.2d 651 (9th Cir. 1982) (lease treated as sale, lessee exposed to both the upside and the downside of the leased asset, lessor’s exposure similar to that of a secured seller in a conditional sale); Estate of Starr v. Comm’r, 274 F.2d 294 (9th Cir. 1959) (lease treated as a sale, leased property has negligible salvage value at the end of the lease to anyone but the lessee).

[98]See Helvering v. Lazarus & Co., 308 U.S. 252 (1939).

[99]435 U.S. 561 (1978).

[100]See, e.g., Charles I. Kingson, The Confusion Over Tax Ownership, 93 Tax Notes 409, 418 (2001) (arguing that “Lyon . . . polluted (and continues to pollute) tax law.”); Shaviro, supra note 3, at 677 (referring to the case as “notorious”); Bernard Wolfman, The Supreme Court in Lyon’s Den: A Failure of Judicial Process, 66 Corn. L. Rev. 1075, 1099-1100 (1981) (“Frank Lyon did not have to happen. The elucidation of principle in tax cases should not depend on irrelevant or legalistic distinctions. A Supreme Court opinion ought not become the basis for tax lawyers to make a laughingstock of the Court as they now do when quite routinely they add unnecessary third parties to financing transactions in order to qualify for the shelter of Frank Lyon.”). The most likely explanation, if not justification, for the Court’s attention to the parties’ intent was that the Court endorsed a transaction of a type widely used by the tax shelter industry, and the Court may have wanted to distinguish it from the shady deals done by the shelter promoters.

[101]435 U.S. at 572-73 (citations omitted).

[102]See id. at 582-83.

[103]77 T.C. 1221 (1981).

[104]Id. at 1237-38 (citations omitted).

[105]380 U.S. 563 (1965).

[106]Examples of circumstances in which the seller was found to be bound to repurchase include an express obligation to repurchase, see e.g., Rev. Rul. 83-47, 1983-1 C.B. 63; a combination of a put and call options exercisable at the same time and for the same price, see, e.g., Rev. Rul. 72-543, 1972-2 C.B. 87, or an option that, in the court’s opinion, is certain to be exercised, see , e.g., Vickers v. Comm’r, 36 T.C.M. (CCH) 391 (1977).

[107]See, e.g., Rev. Rul. 2002-69, 2002-44 I.R.B. 760 (describing a lease-in, lease-out transaction).

[108]249 F.2d 375 (7th Cir. 1957).

[109]Id. at 378.

[110 ] 367 F.2d 823 (7th Cir. 1966).

[111]See id. (the title-holder possessed 12,096 out of 14,519 shares).

[112]376 F.2d 791 (2d Cir. 1967).

[113]Id. at 795.

[114]Comparing Comtel with Patton and Green is a convenient opportunity to introduce the distinction between a two-party case and a three-party case. The point of describing this distinction, however, is to show that it is mostly without a difference. Comtel, a transitory owner or an intermediary, purchased the stock from, and sold the stock to, the same party—Zenkendorf. So Comtel is a two-party case. Patton and Green are three-party cases. The ultimate owner induced an intermediary to purchase the stock from a third party and immediately contracted to acquire the same stock from the intermediary. Of course, instead of lending the amount equal to the purchase price to Zenkendorf, Comtel could have purchased the Commodore shares directly from their historic shareholders, making it a three-party case. Similarly, the ultimate buyers in Patton and Green could have borrowed (from the intermediaries or from other lenders), purchased the stock, immediately resold it to the intermediaries and use the purchase price

to repay the loan (i.e., do exactly what Zenkendorf did in Comtel). Other than mechanics, these transactions have no meaningful differences, especially if, as it happened in Comtel, the party lending the purchase price is the intermediary itself. Both types of cases may be found among the traditional nonfungible whether authorities, see, e.g., Rev. Rul. 68-590, 1968-2 C.B. 66, as well as when considering fungible whether cases, as discussed later in the article. Compare Nebraska Dep’t of Revenue v. Loewenstein, 513 U.S. 123 (1994) (two-party repo) and Union Planters Nat’l Bank of Memphis v. United States, 426 F.2d 115 (6th Cir. 1970) (two-party repo), with First Am. Nat’l Bank of Nashville v. United States, 467 F.2d 1098 (6th Cir. 1972) (three-party repo) and American Nat’l Bank of Austin v. United

States, 421 F.2d 442 (5th Cir. 1970) (three party repo).

[115]See, e.g., Rupe Invest Corp. v. Comm’r, 266 F.2d 624 (5th Cir. 1959).

[116]269 U.S. 443 (1926).

[117]“[D]eposit[s] of stock certificates as collateral security for money loaned” were specifically excluded from tax. Id. at 450, n.1.

[118]269 U.S. at 452.

[119]Id. at 455-56 (citations omitted). Although the holding of the case addressed only the application of the stamp tax statute, it is impossible not to recognize that the Court concluded that a stock loan transferred ownership of the stock for tax purposes.

[120]S.M. 4281, IV-2 C.B. 187 (1925).

[121]The fact pattern is simplified by eliminating the discussion of rights and obligations of A’s, B’s and D’s brokers.

[122]S.M. 4281, IV-2 C.B. 187, 188 (1925). This conclusion was confirmed later, see Rev. Rul. 80-

135, 1980-1 C.B. 18; Rev. Rul. 60-177, 1960-1 C.B. 9.

[123]IV-2 C.B. at 188-89.

[124 ]In 1948, the Service issued a private ruling to NYSE where it stated that “the borrower does not become the owner of the stock he borrows and that he is required to return the stock any time the lender notifies him to do so” and, therefore, “the loan of stock and the return thereof to the lender under the circumstances set forth above, is not a disposition of property which results in recognized gain or loss for Federal income tax purposes . . . .” The ruling was published by two major tax services—including Commerce Clearing House, Inc. (known as CCH). See 5 CCH 1948 Stand. Fed. Tax Rep. 10158; see also, 6 P-H 1948 Fed. Taxes P76,270. Clearly, the ruling’s authors didn’t check the Supreme Court opinion in Provost. Several decades later, the government concluded that if the loan involved common or preferred stock and if the borrower returned identical stock, the overall transaction was subject to the nonrecognition rules of Section 1036, see Rev. Rul. 57-451, 1957-2 C.B. 295. Decades went by, and the IRS modified its approach again. It dismissed “some public misunderstanding” of the 1948 ruling and confessed that the ruling’s conclusion was “legally unsupportable,” see Gen. Couns. Mem. 36948 (Dec. 10, 1976). Instead the government concluded that an ownership transfer resulting from a stock loan was not taxable because “the transaction remains open and the income tax consequences cannot be determined until the borrower satisfies his obligation to the lender.” Gen. Couns. Mem. 36948 (Dec. 10, 1976). Section 1036 was mentioned as a fall-back argument in the limited circumstances when it applied. Finally, the Service refused “to issue rulings as to whether a securities lending transaction constitutes a sale or exchange or whether the transaction interrupts the lender’s holding period.” S.R. No. 95-762, 359 (1978).

[125] Id.

[126]Complying with the additional requirements of the regulations proposed under Section 1058 is also advisable.

[127]But see, Prop. Treas. Reg. § 1.1058-1(e) (providing that failure to comply with the requirements of Section 1058 triggers recognition of gain or loss under Section 1001). Another provision that could provide for nonrecognition treatment of a stock loan is Section 1036.

[128]80 F. Supp. 660 (Ct. Cl. 1948).

[129]At the time of the decedent’s death, Section 862 provided that “Stock in a domestic corporation owned and held by a nonresident not a citizen of the United States shall be deemed property within the United States” and other provisions made such stock subject to the federal estate tax. See Bickford-Smith, 80 F. Supp. at 671. Court of Claims revisited the issue addressed in Bickford-Smith several years later and reaffirmed its views without a substantial analysis. See City Bank Farmers Trust Co. v. United States, 149 F. Supp. 186 (Ct. Cl. 1957).

[130]80 F. Supp. at 672.

[131]Id. at 672-73.

[132]See Rev. Rul. 2003-97, 2003-34 I.R.B. 380; Rev. Rul. 2003-7, 2003-5 I.R.B. 363.

[133 ]The description of a note subordination agreement is a “composite sketch” that comes from several incomplete descriptions of this type of agreement. See Meisels v. United States, 732 F.2d 132, 133- 34 (Fed. Cir. 1984) (subordination agreement with Hayden Stone, a member of the NYSE and Amex); Lorch v. Comm’r, 605 F.2d 657, 658 (2d Cir. 1979) (same); Rev. Rul. 73-122, 1973-1 C.B. 66 (subordination agreement entered into by the members of the NYSE).

[134]This description is also a “composite sketch” based on descriptions in Stahl v. United States,

441 F.2d 999, 1000 (D.C. Cir. 1970), (subordination agreement with Balough & Company); Michtom v. United States, 626 F.2d 815 (Ct. Cl. 1980) (subordination agreement with Hayden Stone); Miami Nat’l Bank v. Comm’r, 67 T.C. 793, 794-96 (1977), and Rev. Rul. 69-455, 1969-2 C.B. 9.

[135]See Rev. Rul. 73-122, 1973-1 C.B. 66 (note subordination agreement); Rev. Rul. 69-455, 1969-

2 C.B. 9 (pledge subordination agreement).

[136]In one case, the Service argued that a broker in a pledge subordination agreement could return cash rather than the securities received from the customer. See Stahl v. United States, 441 F.2d 999, 1000 (D.C. Cir. 1970). The court disagreed and held that this type of agreement resulted in a bailment of the securities. For another example of confusion caused by these agreements, see Michtom v. United States, 626 F.2d 815 (Ct. Cl. 1980), en banc (reversing an earlier decision of its panel).

[137]See Lorch v. Comm’r, 605 F.2d 657 (2d Cir. 1979).

[138]67 T.C. 793 (1977).

[139]See Rev. Rul. 69-445, 1969-2 C.B. 9.

[140]See 67 T.C. at 800.

[141]As any student of tax law knows, no IRS victory is ever complete. Once the Service prevailed in treating repos as secured loans, other tax planning opportunities presented themselves in spades. For a detailed discussion of tax issued raised by repos, see, e.g., William W. Chip, Are Repos Really Loans?, 95

Tax Notes 1057 (2002); Kleinbard, supra note 2, at 797-98; see generally, M. Stigum, The Repo and

Reverse Markets (1989).

[142]421 F.2d 442 (5th Cir. 1970).

[143]Id. at 445.

[144]Id. at 446.

[145]421 F.2d at 452.

[146]It is not clear whether the bank could legally engage in such activity.

[147]An opinion in Union Planters National Bank of Memphis v. United States, 426 F.2d 115 (6th Cir. 1970), addressing a somewhat different fact pattern. The entire discussion was limited to a few sentences and a conclusion that the government’s argument in the case was stronger than in Bank of Austin. The few sentences focused solely on economics: the bank’s protection from risk of loss and its retention of coupons while holding the bonds. The only reference to control over the bonds was made in describing the government’s argument, not in the court’s analysis. See id. at 117.

[148]467 F.2d 1098 (6th Cir. 1972).

[149]Union Planters, 467 F.2d at 1101.

[150]“[A]ppellant’s reliance on the [North Texas Lumber] line of authority begs the question—the issue in this case is whether appellant ever owned the bonds for tax purposes, not whether appellant surrendered ownership at the time the contracts for sale were executed.” Id. (emphasis in original).

[151]See Citizens Nat’l Bank of Waco v. United States, 551 F.2d 832 (Ct. Cl. 1977). Unlike all the repo cases discussed to this point, involved a single transaction in which a bank purchased municipal bonds from one of its customers that needed cash and simultaneously secured a right to resell the bonds to the customer at the same price (a put). When all was said and done, the bank sold all of the bonds back to the customer at the customer’s request. The Service argued that there was an implied agreement that all of the bonds will be repurchased by the customer, but the court disagreed.

[152]Id. at 838.

[153]Id. at 842.

[154]Nebraska Dep’t of Revenue v. Loewenstein, 513 U.S. 123 (1994).

[155]Id. at 134.

[156] Id.

[160]1982-2 C.B. at 35.

[161] Id.

[157]See, e.g., Rev. Rul. 79-108, 1979-1 C.B. 75; Rev. Rul. 77-59, 1977-1 C.B. 196; Rev. Rul. 72-

47, 1974-1 C.B. 24.

[158]1982-2 C.B. 34.

[159]The ruling explained that because investors in the RIC had a right to redeem their shares at any time, the RIC needed an increased liquidity in order to provide cash for the redeeming shareholders.

[162]There are, however, several additional problems with the ruling. First, it failed to state clearly whether the puts had the same strike price as the purchase price of the obligations, unless the reference to “full benefit of any appreciation” was meant to say that this was in fact the case. Second, the government’s reliance on the taxpayer’s reason for acquiring the puts seems unwise. Obviously, if the RIC cared about liquidity only, an agreement by the dealer to repurchase the munies at their then market value would have completely satisfied the RIC’s needs. On the other hand, even if the RIC had bought the puts to protect from risk of loss (as opposed to just of obtaining liquidity), and if the puts had protected all of the obligations (not just a small part), the conclusion should not have changed because the bonds were fungible and the RIC was free to dispose of them at will. Recognizing at least some of the problems with the ruling, the Service announced in the following year that it would no longer issue advance rulings addressing ownership of securities where the purchaser has a right to put the security to the seller or a third party. See Rev. Proc. 83-55, 1983-2 C.B. 572. Although at the time many transactions in the market fit this description, see, e.g., Willard B. Taylor, Debt/Equity and Other Tax Distinctions: How Far Can We Go?, 62 Taxes 848, 853 (1984), the announcement may have produced the intended “chilling effect” and reduced the volume of the new deals, see Steven D. Conlon, Vincent M. Aquilino, Dale S. Collinson, Tax Law Fundamentals of Tax-Exempt Derivatives, 55 Tax Notes 381, 391 (1992). The chilling effect might not have been as serious as it appeared to some commentators, however. See Kleinbard, supra note 2, at 799 (remarking that the “[m]unicipal bond sale/put programs of the type contemplated by Revenue Ruling 82-144 have proliferated since then, leaving tax advisors with the near-impossible task of deciding what business terms are acceptable under that ruling.”). The conceptual questions raised by the ruling remain unanswered.

[163]Although power to dispose was important in Bank of Waco, Citizens Nat’l Bank of Waco v. United States, 551 F.2d 832 (Ct. Cl. 1977), and in Revenue Ruling 72-47, 1974-1 C.B. 24, other authorities did not discuss it even where the question was pressing as it was, for example, where the buyer-bank had a securities license and could freely dispose of the repo’d bonds. See, e.g., Union Planters Nat’l Bank of Memphis v. United States, 426 F.2d 115 (6th Cir. 1970); First Am. Nat’l Bank of Nashville v. United States, 467 F.2d 1098 (6th Cir. 1972).

[164]See, e.g., Kleinbard, supra note 2, at 798; Chip, supra note 141, at 1059.

[165]See, e.g., Kleinbard, supra note 2, at 798 (“by the time anyone noticed, untold trillions of dollars of repos had been consummated—and been reported for tax purposes as money loans” despite the buyer’s freedom to dispose of repo’d securities); Report of ABA Committee on Securities Lending, 91 T.N.T. 107-33 (1991) (noting that “[i]f the [repo] buyer-creditor is in the securities industry it often will re- hypothecate the securities.”)

[166]As one commentator cautioned, “[s]hould a taxpayer or the Service mount a court challenge to the current consensus on the tax treatment of repos [as loans], the challenge might be upheld.” Chip, supra note 141, at 1057.

[167]Compare Nebraska Dep’t of Revenue v. Loewenstein, 513 U.S. 123 (1994) (right to substitute in seller, no power to dispose in buyer, repo treated as loan); Rev. Rul. 72-47, 1974-1 C.B. 24 (no power to dispose in buyer, buyer required to hold repo’d securities, repo treated as loan); with Citizens Nat’l Bank of Waco v. United States, 551 F.2d 832 (Ct. Cl. 1977) (buyer free to dispose, treated as tax owner); Rev. Rul. 82-144, 1982-2 C.B. 34 (same).

[168]See, e.g., First Am. Nat’l Bank of Nashville v. United States, 467 F.2d 1098 (6th Cir. 1972) (buyer was a licensed bond dealer, clear opportunity to sell); American Nat’l Bank of Austin v. United States, 421 F.2d 442 (5th Cir. 1970) (no opportunity to dispose to customers, but possibly an opportunity to sell to other bond dealers).

[169]Provost v. United States, 269 U.S. 443, 455 (1926) (emphasis supplied).

[170]Id. (emphasis supplied).

[171]As a practical matter, the Provost-based approach has a benefit of presenting a much smaller problem for the modern securities industry compared to the inflexible requirement to hold specific securities in order to obtain a loan treatment. It eliminates any need for tracking fungible assets— something that could hardly be done anyway. The vast bulk of the repos are based on U. S. Treasury securities. See Kleinbard, supra note 2, at 798. These securities are traded only in book-entry form. See Andrea S. Kramer, Financial Products: Taxation, Regulation, and Design § 2.02 (3d ed. 2000). This approach also protects a securities dealer that intends to treat a repo as a secured loan as long as the dealer has somewhere among its holdings, or even pledged to its bank, securities identical to those “sold.” While not a guarantee, the likelihood that a dealer may meet this requirement has to be higher than the probability that it holds the specific securities received from the repo seller.

[172]See, e.g., Nebraska Dep’t of Revenue v. Loewenstein, 513 U.S. 123 (1994); see also Rev. Rul.

69-455, 1969-2 C.B. 9 (prior to the event of financial restriction).

[173]See, e.g., First Nat’l Bank in Wichita v. Comm’r, 57 F.2d 7, 8 (10th Cir. 1932) (noting that the dealers substituted bonds on a frequent basis, sometimes several times a day). In many instances, Sections 1058 and 1036 would provide for a nonrecognition treatment upon the substitution, but not in all cases. For example, if the securities involved are debt obligations, Section 1036 is unavailable. If the securities borrower does not have an unconditional obligation to return identical securities to the lender, or if she does not forward all interest or dividend payments to the lender, Section 1058 does not apply. See, e.g., Miami Nat’l Bank v. Comm’r, 67 T.C. 793 (1977) (obligation to return not unconditional).

[174]See, e.g., Citizens Nat’l Bank of Waco v. United States, 551 F.2d 832 (Ct. Cl. 1977); Bickford- Smith v. United States, 80 F. Supp. 660 (Ct. Cl. 1948); Rev. Rul. 82-144, 1982-2 C.B. 34.

[175]See, e.g., American Nat’l Bank of Austin v. United States, 421 F.2d 442 (5th Cir. 1970); Bank of California, N.A. v. Comm’r, 80 F.2d 389 (9th Cir. 1935).

[176]Miami Nat’l Bank v. Comm’r, 67 T.C. 793 (1977).

[177]See supra, note 173.

[178]This analysis puts Revenue Ruling 82-144 in a somewhat different light. Arguably, the Service focused, at least in substantial part, on entirely irrelevant factors and missed a critical one—the strike price of the puts. Whether there were fewer puts than bonds, whether they were acquired to provide liquidity or insulate from loss, and whether the puts had the same term as the bonds had nothing to do with the taxpayer’s freedom to sell the bonds to anyone but the seller. Although the points are related, the strike price of the puts is critical not because it demonstrates how much of the economic exposure the taxpayer retained per se, but because it would affect the likelihood that the taxpayer would sell the bonds back to the seller. If the strike is much higher than the sale price of the bonds, it is virtually certain that the taxpayer would exercise the puts because no other buyer would be prepared to pay so much for the bonds. This would make the buyer’s power to dispose theoretical and should probably lead to a conclusion that the seller retains the ownership of the bonds. See Kleinbard, supra note 2, at 799 (making similar observations). Another statement in the ruling gains a new meaning based on the suggested analysis. The ruling provided that the puts would disappear if the bonds were sold. Thus, assuming that puts were valuable, the taxpayer was required (or, more precisely, economically compelled) to hold the bonds at all times in order to preserve its rights under the puts. That is, it could not sell the bonds into the market (hoping to take advantage of an increase in price) so that it could later buy new ones (hopefully, after the price has dropped) and deliver them under the puts. Was this what the IRS had in mind when it added the provision?

[179]The potential damage to the fisc was evidenced by a swift and strong Congressional reaction that followed. See, e.g., William H. Weigel, Unrelated Debt-Financed Income Under Section 514: A Retrospective (and a Modest Proposal), 50 Tax Law. 625 (1997).

[180]See Frank Lyon Co. v. United States, 435 U.S. 561, 570 (1978).

[181]This close relation between identification and control and their importance in all cases dealing with fungible assets is what made Kleinbard’s argument so powerful. While he used authorities dealing with short sale and securities loans to illustrate his points, the discussion in this article shows that support for Kleinbard’s insights has a much broader base. In fact, most fungible authorities, both when and whether, support his approach.

[182]The Congressional solution is embodied in Section 1258 that converts a portion of capital gain from a “conversion transaction” into ordinary income. A conversion transaction includes, inter alia, “any transaction (1) substantially all of the taxpayer’s expected return from which is attributable to the time value of the taxpayer’s net investment in such transaction, and, (2) which is (A) the holding of any property (whether or not actively traded), and the entering into a contract to sell such property (or substantially identical property) at a price determined in accordance with such contract, but only if such property was acquired and such contract was entered into on a substantially contemporaneous basis . . . .” I.R.C. § 1258(c). Congress explained the need to enact this section by noting that “[t]he committee is aware that taxpayers are able to enter into transactions the economic substance of which is indistinguishable from loans in terms of the return anticipated and the risks borne by the taxpayer.” H.R. Rep. No. 103-111, at

636-37 (1993). Nonetheless, Congress specified that “gain realized by a taxpayer from disposition or other termination of a position that was part of a conversion transaction that would otherwise be treated as capital gain will be treated as ordinary income (but not as interest) for all purposes of the Internal Revenue Code.” Id. at 637 (emphasis supplied).

[183]For an explanation of a distinction between a two-party repo and a three-party repo, see supra note 114.

[184]First Am. Nat’l Bank of Nashville v. United States, 467 F.2d 1098 (6th Cir. 1972).

[185]Lucas v. North Texas Lumber Co., 281 U.S. 11 (1930).

[186]Compaq Computer Corp. v. Comm’r, 277 F.3d 778, 780 (5th Cir. 2001). ADRs, or American Depository Receipts, are trading units allowing foreign corporations to have their shares traded on the major U.S. stock exchanges. An ADR is a certificate of ownership of a trust that holds the foreign issuer’s shares. For all practical purposes, trading in ADRs is like trading in the underlying shares, such as shares of Royal Dutch Petroleum Company in the Compaq case.

[187 ]The $640,000, according to the Compaq’s calculations, was the tax on its net profit from the transaction: the excess of gross dividends received over the loss from sale of the ADRs and the fees incurred in the transaction.

[188]Compaq Computer Corp. v. Comm’r, 277 F.3d 778, 781 (5th Cir. 2001).

[189]Compaq Computer Corp. v. Comm’r, 113 T.C. 214, 223 (1999).

[190]277 F.3d at 780.

[191]See, e.g., Marsh, supra note 2, at 567; Simonson, supra note 2, at 3; Faber, supra note 2, at

[192]See, e.g., StevenM. Surdell, AnalyzingEquity-Linked Debt InstrumentsAfter Reform Proposals, 71 Tax Notes 1525 (1996)F; Debora L. Paul, Another Uneasy Compromise: The Treatment of Hedging in a Realization Income Tax, 3 Fla. Tax Rev. 1 (1996); Kevin Dolan, Carolyn Dupuy, Equity Derivatives: Principles and Practice, 15 Va. Tax Rev. 161 (1995).

[193]See Miller, supra note 3, at 285.

[194]See Lee A. Sheppard, Should Riskless Profit Equal Economic Substance?, 94 Tax Notes 153 (2002); Kingson, supra note 100.

[195]Id. at 160.

[196]Shaviro, supra note 3, at 679.


II.The Fungible When Authorities

A.Short Sales, Prepaid Forwards and the Trade Date Rule

B.Commodity Futures, Forwards and Contracts for Deferred Delivery

C.Some Preliminary Observations

III.The Nonfungible When Authorities

A.The Basic Standard

B.Title, Possession, Benefits and Burdens, and Other Factors

C.“Nonfungible” Securities

D.Differences Between Fungible and Nonfungible When Authorities

IV.The Non-Fungible Whether Authorities

A.The Basic Standard


V.The Fungible Whether Authorities

A.Stock Loans and Subordination Agreements

B.Sale Repurchase Agreements and Related Authorities

C.Reconciling Fungible Whether Decisions

D.Differences Between Fungible and Nonfungible Whether Authorities

VI.Comparing the When and the Whether Authorities

A.Nonfungible Assets

B.Fungible Assets

VII.Contextual Analysis of Tax Ownership—A Recent Example