Featured Articles

Your Government at Work on Financial Products

Posted on June 9, 2008 by Lee A. Sheppard
Document originally published in Tax Notes
on May 7, 2007.

What do women want?

Women want husbands who are willing to share the housework and talk about the relationship. Men's disinclination to do either of these tasks has resulted in a nation of dirty houses and heavy discussions in the marriage counselor's office. So let's take a narrower and more practical approach to this issue: suitable Mother's Day gifts.

A while back we told readers not to buy those tacky little machine-made 14-karat-gold chains for their wives. This touched a nerve among readers, many of whom were genuinely shocked to be told that women really don't like the wretched things. So we're trying to be constructive here.

Let's get the unsuitable gifts out of the way. Most modern men know never to buy an appliance as a gift. But also on the unsuitable list are cheap multicolored bouquets of flowers from the grocery store and pictures of the children. If it's the thought that counts, these are afterthoughts. So here are some suitable thoughts:

Spa day. Pedicures are mostly attractive to foot fetishists, and no one really needs facials except actresses, whose pores get clogged with Dermablend. The point is pampering — and relieving the wife from taking care of the children for the day — which is only possible if she is out of the house. Many are the mothers who had to clean up the dishes after the children attempted to make Mother's Day brunch.

Longchamp Le Pliage tote bag. This foldable nylon tote bag is much loved in Paris, Palm Beach, and Manhattan. It is good for travel. It comes in a zillion colors, but it is best to stick to conservative colors when buying a gift.

Crème de La Mer. It's Crisco with seaweed in it, it's very expensive, and women all love it. Don't ask why. Miraculous powers are ascribed to it.

Perfume. If the wife doesn't wear perfume, don't buy perfume. If the wife has a dresser full of perfumes, take the most empty bottle and replace it with the same thing. If you want to buy perfume and are unsure what to buy, buy Joy, an expensive, heavy floral that everyone likes.

Flowers. The Wall Street Journal wants you to buy a pot of hydrangeas for the wife. Hydrangeas are lovely, but if one were to buy them, the best thing would be to buy the whole hydrangea bush and have it planted in the yard. Hydrangeas bloom much of the year. If one wanted to buy a potted flower, it should be an orchid.

Here's how to do cut flowers right. Go to the cabinet full of unused bridal china and look for a large vase. Whether or not you find a vase, look for an unused table — that would be one free of car keys, magazines, and television remotes. Take the vase to a florist, or select a large vase there. Tell the florist the height of the table and the color of the room, and let him build an arrangement to be delivered on Mother's Day.

A DVD player and a DVD of The Queen. The best movie of 2006, and deserving of every award. Your correspondent saw it twice. You don't have to be an Anglophile to appreciate the political and personal drama surrounding the death of Princess Diana. Not only is Helen Mirren's performance amazing, but Michael Sheen nails the smarmy, ambitious Tony Blair.

Gift certificates. Is it tacky to buy gift certificates? Not necessarily. There are some ornery people who return all gifts. It's their passive-aggressive way of rejecting the giver as well as the gift. Short of a trip to the psychiatrist, the best way around the rejection problem is a gift certificate, which leaves the recipient unable to respond. It has to be in a denomination that might be more than what one would ordinarily spend.

Mother's Day is about celebrating the indispensability and competence of mothers. IRS Chief Counsel Donald Korb has been celebrating for a while, having recently put a lawyer who took time out for motherhood in charge of the IRS response to new financial products. Korb has put Phoebe Mix in charge of responding to new financial products as part of an IRS chief counsel's office reorganization that will devote more resources to financial questions.

Now, around here we question the long-term consequences of reacting to each new financial product, but for the IRS, it would be an accomplishment to be able to react quickly. As the following discussion will show, the IRS's problem with financial questions is not confined to the failure to respond to new developments quickly. This article looks at two problems — one new, one festering — for which decisions to fight appear to have preceded review of the basic questions they raise.

First, the larger failure is the failure of the policy crowd at Treasury to develop a coherent approach to basic questions like whether a financial instrument can be classified as something other than debt or equity, and if a security is debt, how to handle contingent return. The big questions cannot be put off forever, because they will and are being decided by the private sector while the government dithers.

Second, while his predecessor tended to acquiesce in questionable taxpayer interpretations, Korb's office has gotten into the habit of using litigation to determine what the law is. As important as it is to get rid of the Wall Street rule, laying novel financial questions before judges ill-equipped to deal with them is a risky strategy. Finance is not an area in which judges should be making the law, as amply demonstrated by Bank One Corp. v. Commissioner, 120 T.C. No. 11 (May 2, 2003).

In the areas of debt/equity distinction and original issue discount, the tax administrator has the power to issue legislative regulations. Writing a new regulation to change an undesirable result is a lot more productive than marching off to court.

Why Nothing Gets Done

On April 27 New York University's Stern School of Business held a conference on the intersection of taxes and investing that featured the unbecoming spectacle of former government officials explaining why they and the administrations they worked for got nothing done in the financial area.

That the government has essentially abdicated a lot of decisionmaking in the financial area to the private sector was acknowledged, even if the private sector is not always comfortable with that situation. No one on either side will acknowledge the existence of the Wall Street rule, which says that the government has to acquiesce when the volume of transactions reaches a sufficiently high level.

"The private sector shapes the terms of the debate, but we are working in a vacuum, trying to predict the result or rushing to do deals," commented Edward Kleinbard of Cleary, Gottlieb, Steen & Hamilton. "We have a tax system which is unraveling when it comes to the taxation of capital because the government doesn't act, it reacts."

Viva Hammer of Crowell & Moring, a former Treasury associate tax legislative counsel, explained that little is accomplished in the financial area because there are so many sacred cows, and because no one in government is penalized for failing to make a decision. As the pair discussed the government's two decades of acquiescence in debt treatment for DECS, she observed that it may be that the government deliberately defers decisions. Thus the government leaves itself with an option to make a decision later, because there is no payoff for making a decision.

"There's absolutely nothing that we have to do today that cannot be put off until tomorrow," seconded Jeffrey Maddrey of PricewaterhouseCoopers, a former Treasury attorney-adviser, summarizing the attitude in government. That failure to make decisions, Kleinbard commented, makes the government "an enabler of the Wall Street rule."

The government cannot create a uniform system, Hammer argued, because some questions are controlled by ancient law, so that someone loses out if the treatment is changed. The academic term for the phenomenon of small groups of potential losers defending their turf against the common weal is "pareto-optimality." Hammer compared the defenders of the financial status quo to people with rent-controlled apartments in Chelsea (now a desirable neighborhood in Manhattan).

"No one advocates for the whole economy in our system," Hammer said. One would think that advocating for the whole system would be the responsibility of the executive branch.

In Rev. Rul. 2002-31, 2002-1 C.B. 1023, the IRS stated that it would permit the accrual and deduction of interest on contingent convertible debt securities with no adjustment for convertibility, essentially acquiescing in the position that issuers of those securities took to enlarge their interest deduction beyond their actual cost. In so doing, the IRS followed its contingent payment debt regulations to the letter and ignored its statutory power to say what the taxpayer's accounting for the securities should be.

The IRS could have said, in the accompanying Notice 2002-36, 2002-1 C.B. 1029, that it would rewrite the regulations, but it did not. The notice only asked for comments on the anomalous treatment of convertibles, adding some concern that the addition of a relatively insignificant contingency could change the tax result. No action was promised or taken. Five years later, no attempt has been made to review the obvious flaws in the OID rules. The government took the existing contingent payment regulations as graven in stone, even though they are just regulations, which can be changed.

Kleinbard and Hammer defended the position taken in the revenue ruling as "technically correct," which is the most that can be said for an interpretation that is at best plausible. "There are tremendous differences in tax treatment for very small, extremely arbitrary legal distinctions, defined by ancient legal norms with no current financial significance," said Hammer. She was referring to the historic treatment of convertibles as wholly debt, but could also have been alluding to the administrative decision to respect meaningless contingencies.

The long-delayed congressional response to that ruling has been to introduce legislation to change the result by restricting the issuer's deduction to its actual costs. Readers, this will be enacted eventually; it's a revenue raiser, and it's in play. Introduction of a corrective provision like this can be thought of as shelf registration; eventually it will be used.

The latest iteration of the correction is in section 210 of S. 349, the Senate Finance Committee-approved version of minimum wage legislation, which would amend section 1275 to require that the noncontingent bond method of the regulations (or any other prescribed method) be applied by reference to a convertible fixed-rate debt.

In its report, the Finance Committee noted that the contingency is likely to be meaningless and inserted only for the purpose of "significantly and artificially" increasing the OID deduction. And the committee had nothing good to say about the revenue ruling, calling the result "inappropriate from an economic standpoint and based upon a reasonable interpretation of the plain meaning of the regulations." What about the anomalous treatment of convertibles? Better to have legislation, said the committee, rather than administrative acquiescence in taxpayer self- help designed to achieve a favorable tax result.

Call Spread Convertibles

Does the IRS really need to repeat the mistake of the contingent convertibles ruling? Does the IRS really need to give license to the country's practitioners to take advantage of every boo-boo and anomaly in the OID regulations, which it retains the power to change? As the following discussion will show, a similar, but not as complex, convertible debt question has arisen on audit.

The IRS's ability to respond to financial, uh, innovation will be tested by a discrete question of integration of the elements of a public borrowing transaction that does not raise big scary issues of policy. The question is call spread convertibles, and it ought to be easy for the IRS to get to the right answer. The American Bar Association Section of Taxation Financial Transactions Committee will discuss call spread convertibles at its May meeting this week in Washington.

IRS agents are examining call spread convertibles, but the IRS National Office is noncommittal about where the inquiry is going. "Our technical people are aware of these transactions and are looking at them. At the appropriate time and manner, we will let people know our conclusions," an IRS spokesperson told Tax Analysts.

Symantec, the writer of quite excellent computer security software, was one of the many companies that put on a call spread convertible. The stated business reason was to counteract the dilution of equity that would occur when $2 billion of six-year convertible notes that Symantec had issued were converted. Most convertibles convert before maturity, and for Symantec this debt issuance was so large that it was effectively a recapitalization of the entire company.

Symantec purchased a call option from the underwriters of the convertible notes, for the same number of shares, that the convertible debt could be converted to (104,590,200), at the same conversion price stated in the notes ($19.12, which is 22.5 percent more than the then-current share price). This call option, which cost Symantec $592 million, allows the company to obtain the shares it needs on conversion at a price no higher than the conversion price.

So far so good — Symantec would be completely hedged against its obligation to deliver shares on conversion. That is, the purchased call option effectively converts the convertible notes into synthetic straight debt. That is important from a business standpoint because the small 22.5 percent conversion premium makes conversion, hence dilution, more likely. Dilution makes shareholders unhappy.

But then Symantec simultaneously sold a call option for $326 million to the same underwriters for a similar number of shares for $27.32, which is 75 percent more than the then-current share price. Here the business idea is that the dilution would take place at a much higher price, when the shares would have appreciated so much that everyone would be happy.

What was the point of that? securities analysts asked, noting that underwriters don't do anything for free. Analysts think that the $266 million difference between the put and call option prices should be considered part of the cost of issuing the convertible notes, that is, an extra 2.2 percent on top of the 0.75 percent interest Symantec was paying on the convertibles. They argued that selling the call option was "superfluous" and asked why Symantec didn't just issue a note paying 3 percent and have a conversion price of $27.32 per share. The answer to that question is that the market seems to want convertibles with small conversion premiums. (For discussion, see /taxbasehttp://software.seekingalpha.com/article/12351.)

The addition of the written call makes the package look like a straight debt plus a warrant. The question for securities analysts is whether Symantec's package is cheaper than convertible debt, which, as the stated coupon here shows, is mighty cheap. The analysts came up with an effective interest rate for the synthetic straight debt that exceeds 7 percent, which they thought might be more expensive than the rate Symantec could borrow at, by nearly half a percentage point.

Well, is Symantec enriching bankers or enriching tax practitioners? Both, apparently. Taxpayers like Symantec have effectively lowered their borrowing costs by writing a call, while not taking that reduction into account in calculating the interest cost they deduct on their tax returns.

Taxpayers are taking the position that only the purchased call option — the one on the same number of shares at the conversion price — need be integrated with the convertible debt under reg. section 1.1275-6. This allows them to amortize the cost of the purchased call over the period of the debt as a deduction for original issue discount.

What about the written call option, which reduced the total cost of the options? Oh, uh, that's a "separate warrant transaction" that need not be taken into account in determining the amount of interest the taxpayer will deduct. Tax lawyers tweak the terms of the written call option so that it cannot be integrated under the terms of reg. section 1.1275-6. The usual way to keep the written call separate is to give it a term that is 90 days longer than the purchased call. The mismatch in maturities prevents integration under the regulation.

Taxpayer election here is not confined to the permissive integration rules. Arguably, the taxpayer could have achieved the same economic result by issuing contingent convertibles, or by issuing straight debt with separate warrants. The tax rules and regulations would not treat either of these the same way call spread convertibles are treated. Indeed, the overstatement of the interest deduction would have been larger had a contingent convertible been issued.

Reg. section 1.1275-6 provides for permissive integration of qualifying debt instruments and their hedges. The permissiveness point is important. The regulations were seen as a favor to taxpayers when they were issued more than a decade ago.

Reg. section 1.1275-6(b)(2) defines a hedge as "any financial instrument (as defined in paragraph (b)(3) of this section) if the combined cash flows of the financial instrument and the qualifying debt instrument permit the calculation of a yield to maturity (under the principles of section 1272), or the right to the combined cash flows would qualify under section 1.1275-5 as a variable rate debt instrument."

Reg. section 1.1275-6(b)(3) defines a financial instrument as "a spot, forward, or futures contract, an option, a notional principal contract, a debt instrument, or a similar instrument, or combination or series of financial instruments." The only caveat on elective integration is that the synthetic debt security should have the same remaining term as the original debt. An equity option can be integrated with debt that has an equity-referenced payment. (See reg. section 1.1275-6(h), Example 2.) A few years back, the New York State Bar Association Tax Section bemoaned the failure of the regulations to allow integration of debt securities with credit default swaps.

Reg. section 1.1275-6(c)(1) lays down the conditions for elective integration. The taxpayer must identify the hedge, and the other side of the hedge cannot be a related party. The same taxpayer that issued the debt must enter into the hedge, which must be entered substantially contemporaneously. Neither the debt nor its hedge should have been part of another integrated transaction within the previous 30 days, nor should either have been part of a straddle.

Reg. section 1.1275-6(c)(2) allows the tax administrator to compel integration "if the combined cash flows on the qualifying debt instrument and financial instrument are substantially the same as the combined cash flows required for the financial instrument to be a section 1.1275-6 hedge." But there are conditions. The taxpayer must have failed to identify the hedge, have used a related party on one or the other side of the hedge, or have legged out of an integrated transaction within the past month.

The circumstances under which the tax administrator can compel integration do not include the failure to identify and integrate every financial contract that could hedge the debt security in question. If the taxpayer merely wants to account for a hedge separately, the IRS cannot compel integration. (For discussion, see Doc 96-20560 or 96 TNT 142-78 96 TNT 142-78: Special Reports (SPR).) The IRS lost an argument to compel integration of a debt and its hedge in a tax shelter context in Klamath Strategic Investment Fund v. United States, No. 5:04CV-00278 (E.D. Tex. July 20, 2006).

Reg. section 1.1275-2(g) is a general antiabuse rule that allows the IRS to apply or depart from the OID regulations when a result is unreasonable in light of the purposes of the statute based on all the facts and circumstances. One factor in determining reasonableness is whether the treatment of a debt security has a substantial impact on the taxpayer's tax liability. The antiabuse rule was not invoked in the case of contingent convertibles.

In Example 3 of reg. section 1.1275-2(g)(3), the taxpayer issues a convertible debt rather than straight debt and a warrant as an investment unit. The latter would have OID, while the convertible would not. The example concludes that taking advantage of long- standing differences in tax treatment is reasonable. But in Example 1, the inclusion of a meaningless early call option in a debt is designed to reduce the amount of interest included by holders before the call date. This example concludes that the result is unreasonable and the tax administrator can ignore the call option in determining interest inclusion by holders.

In Rev. Rul. 2000-12, 2000-1 C.B. 744, the IRS invoked the antiabuse rule to compel integration of offsetting debt instruments that could not be integrated under the integration rule of reg. section 1.1275-6 because they cannot be readily combined and their maturities are different. The IRS concluded that the taxpayer would be able to claim an artificial loss if the two notes were not integrated. Arguably the artificial loss is more egregious than an overstated interest deduction.

Can a taxpayer integrate a debt security with more than one hedge? Yes. It is difficult, but it can be done and it has been done, mostly by real estate investment trust issuers. The problem for taxpayers desiring to integrate is that the hedge has to be perfectly matched, which is why the hedge is deliberately imperfect in the call spread convertible case being discussed here.

Some prominent practitioners argued to the government that the two parts of what they called "bond hedge and warrant transactions" should be kept separate, and agents should not be setting them up. They noted that the business purpose of call spread convertibles is to reduce the dilution from conversion by ensuring that no dilution will occur until the written call at the much higher strike price is exercised. They admitted that the arrangement produces additional OID deductions, but argued that the decision whether the written calls should be integrated should be made at a higher level than IRS examiners.

The IRS response to the call spread convertibles problem has been to examine the transactions on audit. Trouble with contesting these positions is that the regulation calling for integration of the call options is elective. There is not much point in going to court to contest a taxpayer election that is adverse to the fisc. All taxpayer elections are adverse to the fisc. (What do we always say around here about giving taxpayers elections?) A more appropriate response would be a notice stating that the regulation will be changed to require integration of all contracts that affect the yield on a particular debt security.

Of course, proper income measurement requires that only the real cost of the yield on the convertible debt be deductible, and the tax administrator has the power under section 446(b) to compel proper accounting. But section 446(b) is a general power, and a specifically prescribed accounting method trumps the general even if the specific method does not measure income properly. (Williams v. Commissioner, 94 T.C. 464 (1990), aff'd, 1 F.3d 502 (7th Cir. 1993). The OID rules are a specific method of accounting for discount debt. The IRS cannot remove a taxpayer from a permissible method and put it on a method that the IRS thinks more clearly reflects income. (Prabel v. Commissioner, 91 T.C. 1101 (1988), aff'd, 882 F.2d 820 (3d Cir. 1989).)

So while the tax administrator could fight in court for proper income measurement in the case of call spread convertibles, it would be unbecoming to argue against its own regulation. The underlying statute empowers the IRS to write legislative regulations. (Woods Investment Co. v. Commissioner, 85 T.C. 274 (1985).)

Moreover, the overstatement of the OID deduction is probably not so egregious as to bring the case within the rule of Ford Motor Co. v. Commissioner, 102 T.C. 87 (1994), aff'd, 71 F.3d 209 (6th Cir. 1995). A judge who could be made to understand the offsetting call options would be unlikely to have his conscience troubled by a deduction that could be excused as the equivalent of the interest deduction for straight debt.

Prepaid Forward Contracts

On to the festering problem. The IRS is setting up prepaid forward contract cases all over the country, using Securities and Exchange Commission filings by insiders to find out when an individual shareholder has entered a contract that the SEC considers to be a disposition of shares. Mix plans to discuss prepaid forward contracts at the ABA Tax Section Financial Transactions Committee meeting this week.

There are three pieces of guidance that together form the IRS position on prepaid forwards. Whatever one believes to be the appropriate tax result in these cases, the question has not been well handled procedurally. But throwing nonbinding bits of guidance up against the wall to see whether they stick has been a bad habit for several administrations, particularly in the financial area.

In the first, Rev. Rul. 2003-7, 2003-1 C.B. 363, the IRS ruled that the receipt of cash on a pledge of shares did not constitute a constructive sale of those shares if the taxpayer who pledged them retained the right to substitute collateral and is not economically compelled to deliver the pledged shares.

The ruling concerned a three-year variable delivery forward contract for publicly traded shares held by an individual. The counterparty to the contract was an investment bank. The taxpayer pledged the maximum number of shares that would be deliverable under the contract as collateral for the obligation to perform on the contract. The pledged shares were held by a third-party trustee. The taxpayer retained the right to vote the shares and collect dividends. The taxpayer was protected from downside risk of holding the shares during the term of the contract by the variable delivery formula, which collared a 25 percent range of prices.

The taxpayer received cash in an unstated amount. (In the usual case, the amount is the discounted present value of the maximum number of shares deliverable under the contract.) The taxpayer had the unrestricted legal right to satisfy the contractual delivery obligation by delivering the pledged shares, substitute shares of the same issuer, or cash.

The ruling posits that the taxpayer "is not otherwise economically compelled to deliver the pledged shares." The ruling contains a caveat that a sale may have occurred "if a shareholder is under any legal restraint or requirement or under any economic compulsion to deliver pledged shares rather than to exercise a right to deliver cash or other shares." Among the indicators of economic compulsion is "an expectation that a shareholder will lack sufficient resources to exercise the right to deliver cash or shares other than pledged shares."

What if the customer's only available asset to satisfy the forward contract is the shares — that is, the customer doesn't have the cash? Does that constitute economic compulsion? Robert Gordon of Twenty-First Securities, who was chairing the NYU meeting, wanted to know. It's a valid question, since prepaid forwards were sold to some cash-short customers. Matthew Stevens of Skadden, Arps, Slate, Meagher & Flom noted that many deals were done with these "core affluent" customers. Stevens was dismissive, arguing that the old cases ignored economic compulsion. Customers often use the cash for other investments, which could be liquidated to settle the forward contract in cash.

At least Rev. Rul. 2003-7 was a public piece of guidance on which all taxpayers were entitled to rely. Agree or disagree with the result, the government made a decision and put it out there. The ruling has since been undercut by less formal guidance on which taxpayers are not legally entitled to rely, although they can take return positions.

But the revenue ruling was made on incomplete facts, the IRS later learned. So it was followed by the second piece of guidance, TAM 200604033, which raised the question whether an agreement by the taxpayer to lend its shares to the counterparty would affect constructive sale treatment.

The taxpayer in TAM 200604033 had several agreements with the counterparty, who we assume for purposes of discussion was an investment bank. There was a forward purchase contract. There was a share lending agreement. There was a pledge agreement. All of the agreements were interrelated, were executed simultaneously, and covered the same identified shares.

The forward purchase contract allowed the counterparty to purchase a variable number of the taxpayer's identified shares on an agreed date in the future, while prepaying the taxpayer the purchase price in cash at a discount to reflect the time value of money. In this case, the taxpayer's cash settlement option had been deleted from the purchase contract. The counterparty had the right to accelerate the forward purchase contract if it could not hedge the shares.

The forward purchase contract required the pledge agreement, under which the taxpayer pledged the identified shares to secure his obligation to deliver the identified shares or identical shares at the date the forward purchase contract was to be settled. The forward purchase contract and the pledge agreement required the trustee holding the identified shares to enter into the share lending agreement, under which the counterparty could borrow the shares for hedging or use them in its business as a dealer. The pledge agreement contained the taxpayer's preapproval of the share lending agreement.

The IRS concluded that the taxpayer must recognize gain on a prepaid forward purchase contract when the taxpayer will never have the identified shares returned because of a compulsory share lending agreement and behavior inconsistent with open transaction treatment. The ruling hinges on the idea that the forward contract would not have been made had the taxpayer not agreed to lend the shares and the forward contract would not survive a termination of the share lending agreement.

Bankers believe that section 1058 and a 34-year-old revenue ruling separately protect any share lending that goes on in the context of a prepaid forward contract. To a banker, the collateral account that the identified shares go into is just like a margin account. Shares are routinely borrowed out of margin accounts. Bankers point out that a prepaid forward is essentially a bet that the share appreciation will exceed the interest cost on the loan element and that it offers little downside protection.

Rev. Rul. 72-478, 1972-2 C.B. 487, said that share lending was not inconsistent with a short-against-the-box transaction. The ruling says that the short sale will not be considered a sale of the taxpayer's shares as long as those particular shares were not delivered to satisfy the short sale.

Maddrey noted that the taxpayer in the ruling was completely hedged against the risks of his long position, had received cash, and had parted with the shares by lending them to the broker. One would not make this arrangement for any reason other than tax reasons, he explained. There were, according to Maddrey, "no economics" - - that is, the taxpayer would have a pretax economic loss unless he lent the shares.

Hence the facts of Rev. Rul. 72-478 are arguably more objectionable from the perspective of disposition of shares than prepaid forward contracts are. Maddrey argued that the fact that a customer could agree to lend the shares means that the pledge of the shares for the forward contract does not wrest control of the shares from him.

The ruling predates the enactment of section 1058, a narrow exception to section 1001 that is intended to encourage share lending. Section 1058 requires that the identified securities be returned to the lender, that dividends be passed through to the lender, and that the loan "not reduce the risk of loss or opportunity for gain of the transferor of the securities in the securities transferred."

The taxpayer in TAM 200604033 made a section 1058 argument. The IRS countered that the taxpayer would not get the identified shares back. Section 1058 requires that the lender of shares retain a significant amount of risk of loss and opportunity for gain, and, the IRS argued, should be narrowly construed. The IRS concluded:

      Taxpayer has given up nearly all indicia of ownership in the "lent" shares in a related transaction occurring nearly simultaneously with the lending transaction, including a transfer of its risk of loss and most of its opportunity for gain. In addition, it is doubtful that Taxpayer could regain possession of the pledged shares and terminate a Share Lending Agreement without accelerating the Transaction, leading to an offset of the two agreements.

But still the guidance was incomplete. The third piece, Advice Memorandum 2007-004, was released in February. This memorandum is supposed to restate in general terms the IRS's section 1058 argument and other arguments to be used in litigation, but it does not add a lot to the previous two pieces. The memorandum is also an acknowledgement that TAM 200604033 had some bad facts that may not be present in other cases.

The taxpayer in AM 2007-004 had a variable rate forward purchase contract on publicly traded shares with a counterparty that is some sort of financial intermediary. The forward purchase contract required the taxpayer to pledge the maximum number of shares deliverable under the contract as security for the taxpayer's obligations. A third-party trustee held the pledged shares, and the taxpayer retained dividend and voting rights. (In real life, there are no third-party trustees in these deals.)

The trustee lent the shares to the counterparty, which borrowed the shares and sold them. The share lending agreement gave the taxpayer the right to demand that the pledged shares be replaced with identical shares. The memorandum states, however, that it is not necessary for the counterparty to borrow or otherwise take possession of the shares, as long as it has the power to do so and thereby to exercise control over them.

The rambling memorandum addresses the argument that the transaction is not closed under Burnet v. Logan, 283 U.S. 404 (1931), because the number of shares ultimately deliverable is variable. The IRS argues that the open transaction doctrine should only apply when it is not possible to determine the value of either of the assets exchanged. The IRS concludes that the taxpayer has parted with the maximum number of shares deliverable under the contract in exchange for immediate cash and the right to get a variable number of identical shares in the future. "Nothing is indeterminate," the memorandum intones. Readers will recall that in TAM 200604033, the IRS had trouble articulating what was sold and for how much.

The memorandum distinguishes Rev. Rul. 2003-7 on the basis of the share lending agreement. When considered together, the memorandum opines, the forward contract and the share lending agreement "transfer dominion and control" of the shares to the financial intermediary. The memorandum argues that the trustee's independence is irrelevant because the financial intermediary has "the unconstrained right to do as it wishes with the shares."

The memorandum argues that section 1058 is not satisfied because the two agreements, taken together, reduce the taxpayer's risk of loss in the transferred shares. The memorandum cites Helvering v. Le Gierse, 312 U.S. 531 (1941), for the theory that the two contracts should be combined. In that case, the 80-year-old taxpayer took out offsetting life insurance contracts in an unsuccessful effort to keep the proceeds out of her estate.

"It is unreasonable to think that Congress would have intended for taxpayers to avoid this requirement by using a simple device of separating their agreement into two or more documents," the memorandum argues. Moreover, the two contracts have to be stepped together because not combining them "would be ignoring the economic realities of the transaction."

"If they're going to base the argument on 'economic reality,' it's hard to know where to stop," Maddrey commented. "The economic reality of a short against the box is a sale." Maddrey thinks that the 1972 ruling has to trump the statute — that is, the share loan and the short sale cannot be stepped together for purposes of analyzing whether risk has been offset under section 1058(b)(3).

Why can't the IRS just revoke Rev. Rul. 72-478? It has bad facts. It is not as though securities lending would be put at risk as a business practice because section 1058 protects it. In the ordinary case, customers agree to lend whatever shares their brokers have custody of, and this is thought to grease the wheels of commerce in an economy based on reshuffling the existing pot of capital. Section 1058 is supposed to encourage securities lending by saying that the disposition of the shares should not be taxable. Bankers view it as a safe harbor — that is, that an argument could be made that there was no disposition even if section 1058(b)(3) wasn't satisfied.

The IRS is trying to say that securities lending is inconsistent with nonrecognition of a monetization transaction. Otherwise, the memorandum is clumsily saying that share lending in the context of a prepaid forward contract is a bad fact, even if it is not a bad fact in the run-of-the-mill margin account or short-against-the-box transaction. Bankers fault the memorandum for not at least trying to distinguish Rev. Rul. 72-478.

Trouble is, in Maddrey's view, there is just too much water under the bridge respecting short sales against ownership of fungible publicly traded securities. (For this he cites Rev. Rul. 2004-15, 2004-1 C.B. 515, which respected the form of a grandfathered short sale against the box.) Before TAM 200604033, securities lending was so routine whenever publicly traded shares were involved that bankers just assumed there was no problem with it. For a securities dealer, it is normal to ask for the right to rehypothecate any shares that are in its custody. Some customers called back their shares in the wake of TAM 200604033.

Bankers are trying to fit their deals within the contours of the guidance. Prepaid forward contracts are being concluded without the counterparty borrowing the customers' shares. In AM 2007-004, the IRS is saying that it doesn't care whether the counterparty actually borrows the securities, as long as it is able to get control of them. Some customers are putting on collar contracts with separate loans. Financial intermediaries favor forward contracts, however, because they do not have to be booked as loans and are not subject to margin lending rules.

The securities lending question is curious because it only applies to investment banks and securities dealers as counterparties. If a securities dealer has custody of a customer's shares, it has to do something with them, from both a regulatory and a business standpoint. Commercial bank counterparties don't need to borrow securities. Because so many deals were being done, whether a financial intermediary could afford not to borrow the shares became a pricing question. Securities dealers had to borrow the shares to get their prices as low as commercial banks' prices. In AM 2007-004, the IRS is saying that it does not care how a particular financial intermediary runs its business.

Why are we fighting about securities lending? Why is the IRS picking a fight about prepaid forward contracts when it could have written regulations to cover them under section 1259(c)(1)(E)? Although the House Ways and Means Committee report on section 1259 states that a forward contract that provides for "significant" variation in the amount of property to be delivered does not result in a constructive sale, variable delivery forward contracts without significant variations would fall under the residual clause of section 1259(c)(1)(E) if the IRS were to bother writing implementing regulations. In Rev. Rul. 2003-7, the IRS assumed away the section 1259 question by assuming a significant variation in the number of deliverable shares.

Again, we are talking about whether judges, the private bar, or the tax administrator will write the rules. Readers will recall that the private bar effectively hijacked the collar rule of section 1259, declaring that a collar that covered most normal movements in share prices should not cause a recognition event. That was bad enough. But the tax administrator retained the power to say which nonenumerated monetization transactions should result in recognition, and that power has never been invoked.