6   Accounting for Fraud

KPMG was not the only firm among the then Big Six to play a major role in the tax shelter industry. Ernst & Young (E&Y) was eager to participate too. PricewaterhouseCoopers (PwC) was involved but withdrew after it was publicly embarrassed during Congressional hearings. Arthur Andersen was active, but its demise in the wake of the Enron debacle limited the amount of information about its shelter activities that came to light. Only Deloitte Touche remained unscathed. Although that firm was described in Janet Novack and Laura Sanders’s 1998 article in Forbes as a tax shelter promoter, no official account of wrongdoing at the firm has emerged.¹

Ernst & Young: The VIPER Strikes

Getting in the Game

As 1998 began, Arthur Andersen, KPMG, and PwC were all involved in developing and marketing generic shelters aimed at high-wealth individuals, and E&Y decided it needed to play catch-up. The firm had offices in 120 countries and 70,000 employees worldwide; like the other Big Six firms, it was feeling significant market pressure. Tax shelters were viewed as a potentially major new source of revenue.² A natural location for such work was the firm’s Personal Financial Consulting (PFC) unit, headed by Robert Garner, which offered tax advice and planning to individuals. In early 1998, Garner got approval from Richard Bobrow, then the partner in charge of the firm’s National Tax Department,³ to form a group devoted to pursuing high-fee strategies for high-wealth individuals.

Emphasizing E&Y’s new aggressiveness, the group was called VIPER: “Value Ideas Produce Extraordinary Results.” Inside the firm, the acronym sometimes stood for “Value Ideas Produce Extraordinary Revenues.”⁴ VIPER’s mission was to work with financial institutions, investment companies, law firms, and tax shelter promoters to design and market tax strategies. To lead the group, Garner tapped tax lawyer Robert Coplan, a former branch chief of the IRS Estate and Gift Tax Regulations Division, who was in the PFC group within E&Y’s National Tax Department and had been named partner four years earlier. Garner also recruited Martin Nissenbaum, a lawyer and accountant who was a specialist in income tax and executive compensation in the PFC unit. In addition, while he didn’t formally join VIPER until 2000, Richard Shapiro, a lawyer who focused on the taxation of financial instruments, worked closely with the group from its inception. Shapiro had joined E&Y as a partner in 1995 after working at both law and accounting firms.

E&Y management decided to establish a sales team dedicated to marketing the strategies developed by the VIPER group, and recruited Brian Vaughn to lead the new team. Hailing from Louisiana, a certified public accountant and financial planner with experience marketing tax shelters for Deloitte Touche, Vaughn was a brilliant salesman who exuded charm and self-confidence. E&Y first heard of him when several of the firm’s clients told E&Y that Vaughn had been offering them a tax shelter known as a flexible gain deferral strategy, which Vaughn had brought to Deloitte from KPMG, where he had earlier been employed. At Deloitte Vaughn worked with Belle Six, who was also a CPA and financial planner. Although both Vaughn and Six were married to others, they had developed a romantic relationship at Deloitte.⁵ Vaughn told E&Y that he and Six were a package deal. Vaughn also insisted that their positions be in the National Tax Department so that they would have a central role in designing and marketing tax shelters. Vaughn’s further requirements were that E&Y give them a percentage of sales on top of their salaries, relief from needing to keep detailed billing and expense records, and carte blanche with regard to travel expenses.

The firm balked at first but eventually agreed, and the two CPA/financial planners joined Ernst & Young in late January 1998 as the VIPER sales force. Vaughn and Six were excited that E&Y was committed to providing significant resources to developing a shelter practice aimed at high-wealth individuals, especially since Deloitte did not seem interested in doing so.⁶ They also knew that they were going to have to deliver serious sales results to justify the employment packages they had obtained. Six months after they had come on board, Vaughn circulated a presentation to VIPER members in which he anticipated revenues of $3.5 million for 1998, increasing to $15 million in 1999, and $25 million in 2000.⁷ The pressure to produce was on.

The VIPER team spent most of its time brainstorming on possible high-value tax strategies. Team members believed that it was unlikely they would get authorization to offer shelters that completely eliminated taxes. Ernst & Young had been more cautious than other firms in entering the personal shelter market, and a tax elimination strategy might draw unwelcome IRS attention. VIPER therefore focused its efforts on strategies designed to convert ordinary income to capital gains and to defer taxation until a later time. By late 1998, however, the group had produced little in the way of marketable products. Vaughn and Six became concerned that their jobs might be in jeopardy. The other VIPER team members had other responsibilities at the firm, but the two-person sales team was dedicated solely to marketing VIPER strategies.

At a VIPER meeting at the end of 1998, Richard Shapiro mentioned a tax shelter that the group had discussed before and had rejected as too complicated. The idea had originated with David Smith, the former KPMG tax manager who had moved to Quadra and then formed Private Capital Management Group (PCMG), a purported “investment” firm. Vaughn assured the group that if someone could explain the shelter to him, he would be able to explain it to potential clients. He and Six were told to meet with Smith and learn more about it.

Vaughn, Six, and Smith met at the Mansion on Turtle Creek hotel complex in Dallas for the better part of a day in December 1998. Smith explained something known as a “contingent deferred swap” (CDS).⁸ The objective of the transaction was first to defer taxation for one year. The shelter then converted ordinary income into long-term capital gains. This would cut the tax rate in half for a client from about 40 percent to 20 percent. This combination of deferral and conversion seemed a good fit for E&Y, given the firm’s wariness toward shelters that intended to eliminate tax liability completely.

As with other tax shelters, the transaction was organized around forming a partnership that would show losses that taxpayers could claim on their individual tax returns. For CDS to work, this partnership needed to be characterized for tax purposes as a “trade or business,” to generate “business deductions” that clients could use to offset their taxable income. To make it appear that the “trading partnership” was engaged in trading for profit, funds contributed by the taxpayer were placed in a trading account that engaged in a high volume of short-term trades. The partnership would also engage in a swap with a bank, which had the effect of creating a loss in the first year equivalent to the amount of income the client wanted sheltered. Because the partnership purported to be in the business of trading, IRS rules allowed it to deduct this loss from the ordinary income earned by the taxpayer in the same year.⁹

In the second year, the partnership took steps designed to convert the taxpayer’s ordinary income to capital gains, which were taxed at about half the rate. The most important step was terminating the swap contract before it matured. Early termination triggered a payment to the partnership from the financial institution that was the counterparty to the swap. Under the applicable tax provisions, the payment would be treated as a capital gain rather than ordinary income because it had been earned from an early termination rather than from a payment at the maturity date of the swap. Conceived as a tax strategy, CDS permitted a taxpayer with a large amount of ordinary income to defer paying taxes on it for one year and then convert it into capital gains. In economic terms, taxpayers ended up in virtually the same condition they were before the transaction—but with a significantly lower tax liability.

As with other shelters, for CDS to have any plausibility, it had to have economic substance. The taxpayer, therefore, had to have some funds at risk in the transaction. The trading activity of the partnership, however, was explicitly designed to result in little or no change in the economic position of the taxpayer, belying the claim that the partnership was involved in actual trading. Equally important, the early termination of the swap had to purport to represent a business decision by one of the partners in the transaction, not a planned step in a preorchestrated process.

After meeting with Smith, Vaughn and Six reported back to the other VIPER members. Over the next several months, the team engaged in further research to develop the strategy, consulting regularly with David Smith and Brent Clifton, a partner at Locke, Liddell & Sapp, the law firm that had agreed to provide the opinion letter for the strategy. During the same period, Smith had persuaded PwC to design and market CDS,¹⁰ a fact that may have been a factor in Ernst & Young’s decision to jump in.

Locke Liddell would receive at least $50,000 for each opinion concluding that the transaction more likely than not would be upheld if challenged by the IRS. E&Y and PCMG would receive 1.25 percent of the amount of the desired tax loss, or a quarter of a million dollars on a $20 million transaction. As the shelter neared implementation, UBS was recruited to serve as the bank that would engage in a swap with the partnership. Andrew Krieger, a principal at Deerhurst Management, was enlisted to do the trades, which were overseen by Bolton Capital Trading, one of the investment companies in a group owned by Memphis-based Charles Bolton. At one time Krieger was a highly regarded derivatives trader who pioneered currency options trading at Banker’s Trust when the bank moved aggressively into proprietary trading in the mid-1980s. At some point, Krieger had met Smith and Michael Schwartz of KPMG and then PwC. Schwartz got Krieger involved in executing the trading for a number of other tax shelters.¹¹

In late summer 1999, E&Y began to move forward with its first CDS transactions. Although one partner at the firm’s National Tax Department had expressed concerns about the shelter, the firm did not have a formal process of review. One outside lawyer who had been shown CDS by a client wrote to the firm characterizing the strategy as a “classic ‘sham’ tax shelter” that would be successfully challenged on audit by the IRS.¹² No one in the VIPER group paid any mind. The group was intent on pressing forward. As sales of CDS surged during 2000, the team became more and more ambitious. At the beginning of 2000, the team set its goal as a total of $1 billion of CDS losses. In a CDS update in late April, Brian Vaughn crowed that the team was “well on our way” to meeting this goal.

As part of its firm-wide efforts, the VIPER group had developed an internal “Action Plan” that set forth fifty-four steps to implement a transaction. One was to have the client sign an engagement letter with a fixed dollar amount as the fee, rather than to refer to tax losses as the basis. “DO NOT reference tax losses in the engagement letter,” directed the Action Plan.¹³ Step 45 in the draft plan dealt with another sensitive issue: “At the appropriate time within the swap period, G[eneral] P[artner] will terminate the swaps with the bank.”¹⁴ At least one member of the team was concerned that putting this so starkly in writing would be a red flag for the IRS. As Richard Shapiro observed,

One of the problems with tax advantaged transactions when they are reviewed is that they are perceived—correctly, I might add—as too scripted. While having a plan is important, should we have in writing “before the fact” such things as the fact that our swap will be terminated early? Clearly that is necessary for the flow of the transaction. But should there be a document in existence (such as this) that has all chapters and verses laid out? I question that seriously. The fact that no materials are to be left behind at a sales call is not enough, in my opinion. Before anything is in stone here, we should consider what our record should look like.¹⁵

In response, the reference to early termination was removed. Going forward, the VIPER team was careful to eliminate references in CDS documents suggesting that taxpayers had planned on terminating the swap early from the time they entered the transaction. Team members were also very careful to refer to the CDS as an “investment” and not a “tax” strategy.¹⁶

The team went to great lengths to create material for the files so that clients who purchased CDS would be able to document ostensibly nontax business purposes for electing to terminate the swaps early. A convenient excuse arose with the terrorist attacks in September 2001. Three days after the attacks, Coplan sent a letter to clients suggesting that they use 9/11 as the reason for exiting the transactions. Coplan proposed to clients that they write: “Based on the recent terrorist attacks and the possible economic repercussions resulting from such attacks I have decided to reevaluate my investment activities and market exposure.” The client, “after carefully considering the effects of the recent events,” should express the intention to begin closing down positions in the trading account.¹⁷

By late 2000, CDS began to attract attention from the IRS. Despite inquiries from the agency, Brian Vaughn’s enthusiasm for the transaction did not wane. Vaughn emphasized to Coplan that Vaughn would buy the transaction knowing full well that the IRS was investigating it. “As I told you,” Vaughn wrote,

I am a fighter. I don’t enjoy giving up before I get my chance to fight. Remember our opinion on CDS is a should. Let them bring their guns!!!! I believe they will turn their tales [sic] and run the other direction. CDS has economic substance and has the best promoter in the business associated with the trade. I think we owe it to Belle [who by then had gone to work for David Smith implementing CDS] and ourselves not to give up and stop the sales process at this point. Let the clients decide.¹⁸

Ernst & Young continued to promote CDS aggressively into 2001. All told, it sold seventy transactions involving 132 taxpayers, obtaining nearly $28 million in fees. The IRS listed the transaction as a potentially abusive shelter in May 2002.¹⁹

From VIPER to COBRA

Belle Six had left E&Y to work at PCMG with David Smith in August 1999. Six was struggling to extricate herself from her affair with Vaughn, which had become very tumultuous. She hoped that a different workplace would make it easier.²⁰ The financial benefits of the move were also substantial. At PCMG, Six was going to earn 40 percent of the company’s fees, which were 1.5 percent of the losses claimed by a client. In her new position, she continued to work on CDS transactions as a liaison to E&Y.

The next month, Vaughn sent an email to more than two dozen people at E&Y. “Based on meetings today with Brent Clifton at Locke, Liddell, David Smith and Belle Six of PCMG,” he wrote, “we believe we will have a permanent capital loss/ordinary loss strategy by September 15th.”²¹ Despite E&Y’s earlier reluctance at offering tax elimination strategies, Vaughn and the remainder of the VIPER team were preparing to move into this market, where KPMG, PwC, and Arthur Andersen were already active. As Vaughn explained, “The strategy . . . involves the use of foreign currency contracts, ‘contingent liabilities,’ trading partnership, and NO bank loans. . . . The new strategy will create a 1999 permanent capital or ordinary loss. We anticipate a $1 billion worth of loss for 1999. A great start to the new fiscal year. . . . Let’s have fun with this new strategy and kick some KPMG, PWC and AA. . . . No more BLIPS and BOSS.”²²

The shelter that E&Y was developing was the short option strategy that it eventually marketed as Currency Options Bring Reward Alternatives (COBRA), a version of the Son of BOSS shelter described in chapters 4 and 5. This shelter sought to increase taxpayers’ basis in a partnership by the amount they paid to buy an option, while ignoring a reduction in basis that should have been triggered by the partnership assuming the taxpayers’ liability to deliver shares under an option they sold.

Rather than working with Locke Liddell, E&Y sought to jump-start the shelter by collaborating with Paul Daugerdas at Jenkens & Gilchrist, whose experience designing and promoting contingent liability shelters dated back to his days at Arthur Andersen during the mid-1990s. The E&Y version of the option strategy that Daugerdas was offering provided for a slightly greater chance of a profit, although the likelihood was still miniscule. In addition to issuing an opinion letter, Jenkens prepared most of the transaction documents and arranged for Deutsche Bank to execute the trades. Jenkens would receive a fee of about 3 percent of the tax loss generated by the transaction, while E&Y would get 1.5 percent. R. J. Ruble of Brown & Wood would also receive a fee for issuing a second legal opinion.

After reviewing the shelter, the VIPER partners—Coplan, Nissenbaum, and Shapiro—concluded that it would more likely than not be upheld if challenged by the IRS. The sole dissenter was David Garlock, a partner in E&Y’s National Tax Department, who did not believe the transaction had merit. By October, the VIPER team was ready to roll out the strategy—“I smell fees” Vaughn chortled.²³ Because of the strategy’s riskiness, the VIPER leaders had decided to limit sales of COBRA to a small number of eligible clients. The “quick strike” team, as the sales group called itself, had to sell and implement the strategy by year’s end so that clients could take advantage of the losses generated to shelter 1999 income.

A month or so later, sales efforts came to halt, at least temporarily. In December 1999, COBRA came to the attention of Michael Kelley, the partner in charge of overseeing E&Y’s geographic area practices and the number two person in the firm’s tax practice. At about the same time, the IRS issued Notice 99-59, which targeted the BOSS shelter that PwC had been marketing. Concerned about the propriety of marketing a loss elimination strategy, Kelley asked several members of the firm’s Strategic Business Solutions group (SBS) who were experts in partnership taxation to review the transaction. Although E&Y had been selling COBRA aggressively to high-wealth individuals through the Personal Financial Counseling group, apparently no one with specific knowledge of partnership tax had actually reviewed the strategy before it had been approved.²⁴ By this time, the firm had already completed sixteen transactions, earning $13 million in the month the strategy became available.²⁵ When Coplan heard from Kelley, he reluctantly put a hold on COBRA deals in the works pending a decision.

As part of the SBS’s belated review, Jerred Blanchard, a partner in that group, was asked to analyze the transaction. Blanchard was highly critical. In his review, Blanchard noted that the argument for excluding the partnership’s assumption of the taxpayer’s short option obligation from the calculation of basis was unpersuasive, and that the mere paper loss that the taxpayer incurred was unlikely to be recognized by the government.²⁶ As Blanchard further emphasized, “The transaction does not seem to be one that a reasonable investor would entertain in the absence of the anticipated generation of a $20 million federal income tax benefit.”²⁷ The expected value of the loss exceeded the expected value of the potential gain, he noted, by $570,000. In other words, from a business perspective, the deal was almost certainly a loser. When Coplan saw Blanchard’s analysis, he dubbed it a “hatchet job.”²⁸

In early January, Kelley convened a meeting in the Washington office with several members of SBS; Coplan; Ron Friedman, the head of the firm’s National Tax Quality and Standards group; Robert Carney, a partner in Tax Controversy Matters; and Michael Frank, in-house counsel, to discuss whether E&Y should continue to sell COBRA.²⁹ After discussion, Kelley directed the group to stop marketing COBRA, a decision subsequently supported by William Lipton, the vice chair of tax.³⁰ Kelley proposed that, going forward, new strategies should require sign-off from two tax partners with subject matter expertise before a product was approved for marketing. Kelley also advised the tax strategy group to stop using clever reptilian acronyms, which, he noted, could be given an “inflammatory characterization by critics.”³¹ In response, Coplan agreed to “defang” the names the group was using.³² In February, VIPER was changed to the “less sinister sounding acronym”³³ SISG, which stood for Strategic Individual Solutions Group.

At around this time, tensions were surfacing between Coplan and Vaughn. Vaughn was miffed that he had not been included in the review meeting and was feeling left out when it came to team decisions. “I cannot afford to be excluded from this process,” Vaughn complained to Coplan. “Many people in the field view me as the leader of the VIPER team and in charge of the sales process. By excluding me from these calls and meetings I am not able to fulfill this role and help the local offices.”³⁴ For his part, Coplan was annoyed that Vaughn was a little careless with regard to communications with the team. Over time, he and Shapiro also noted a tendency of Vaughn’s to go around the team to deal with higher-ups in the National Tax Department.³⁵ Hired to market the group’s product and out of his league when it came to substantive tax expertise, Vaughn nevertheless considered himself an important participant in substantive decisions about the strategies.

Despite Kelley’s decision that E&Y cease marketing COBRA, he apparently made an exception for one last, lucrative transaction. In early 2000, Coplan asked him for permission for one more deal, emphasizing that the clients, who had signed an engagement letter in late 1999, were still very eager to go forward and understood the attendant risks. The Murphy family were highly successful hog farmers in North Carolina who were planning to sell their operations in early 2000 for about $200 million, with an expected profit of nearly $100 million. Kelley gave Coplan the green light, but insisted that the family sign a “hold harmless” agreement, under which they agreed that they would not bring any claims against the accounting firm if it turned out that the tax benefits of the strategy were disallowed. After the Murphys were informed that the strategy was very risky and the firm was seemingly shielded from potential lawsuits, Coplan got the go-ahead to sell the Murphys COBRA.

By the end of 1999, the VIPER group had decided to stop using Jenkens & Gilchrist to provide legal opinions for COBRA. Members of the group had become concerned, somewhat belatedly, that since Jenkens had been involved in developing the strategy, the law firm would not be viewed as independent for purposes of providing an opinion that protected a taxpayer against penalties. In addition, members of the VIPER group were displeased at the huge fee Jenkens was charging, a whopping 3 percent of the losses claimed in COBRA transactions. For the Murphy family, or “Murfam” COBRA, Coplan brought in Ira Akselrad, of the New York law firm Proskauer Rose, who was willing to provide the necessary legal opinions for about $400,000 each. Ernst & Young declined to pass on the full savings to the Murphys, instead increasing its fee from 1.5 percent to 2 percent. The firm ended up earning $2 million from the transaction.

Despite the substantial fees, the firm’s implementation of several COBRA transactions hit some bumps along the way. In the Murphys’ case, several important documents were finalized months after they should have been. As a result, the Proskauer law firm had the Murphys execute backdated documents. With a number of earlier COBRA transactions, E&Y missed a deadline for IRS filing extensions. When Coplan became aware of this “COBRA screw-up,” in June 2000, he got nervous. “If this leads to an audit,” he worried in an email, “we are dead meat.”³⁶

After the COBRA review in early 2000, Kelley and members of the SBS team decided to take a look at CDS.³⁷ Kelley concluded that sales of CDS shelters could continue as long as clients fit the appropriate risk profile. A celebratory Coplan titled the subject of an email reporting this news “CDS lives!”³⁸ As he underscored to the SISG team, “One point Mike Kelley stressed is that we should be very certain that the individuals we approach with this transaction are sophisticated investors who fully understand the economic and tax risks of the transaction, and would not likely seek compensation from us if the anticipated tax benefits are not ultimately realized.”³⁹

Coplan closed his email to the SISG team with a promise to find a new strategy that would produce sufficient revenues to replace those lost when COBRA was discontinued. “Finally let me congratulate you on a phenomenal first quarter,” he wrote. “The next quarter will pose challenges without COBRA, but we will strive to bring you new solutions and complete those in the pipeline to restock your quivers.”⁴⁰

In the meantime, some SISG leaders continued to get negative feedback from lawyers in the tax bar who were repelled by the transactions E&Y was promoting. In May 2000 Shapiro, who practiced in the firm’s New York office, complained to Coplan: “I got one of those ‘unpleasant’ calls I now get from time to time from dan shapiro [sic], senior tax partner at schulte roth [sic], who used to think highly of me he severely questioned COBRA; now I think he is about to question CDS.”⁴¹

Restocking the Quiver

The search for new shelters was rewarded a few months later. In early May 2000 Vaughn met with Six, who had moved with David Smith to work at Bolton, and other staff at E&Y to discuss a new strategy for clients who participated in the CDS shelter. The basic idea was to graft a COBRA-like tax eliminator onto the CDS shelter so that taxpayers would be able to defer taxation, turn it into capital gains, and then postpone any tax indefinitely. After the taxpayer died, the gains would not be subject to income tax.⁴²

Excited about this new strategy, Vaughn contacted Coplan: “If we could integrate CDS and the foreign currency trading program with the short option transaction, we could have a great transaction,” he enthused.⁴³ Coplan forwarded Vaughn’s message to Nissenbaum and Shapiro. Vaughn urged the firm to move ahead, saying, “I am only trying to seize the moment. If not us, another firm will.”⁴⁴ Coplan tried to encourage Vaughn, but emphasized: “It will not be easy to turn our folks around on a COBRA-like strategy” in light of the earlier directive to discontinue the COBRA shelter.⁴⁵ He and others in SISG knew that they had to provide a credible business purpose for the add-on feature to get it approved within E&Y. The story that Coplan and others concocted was that the idea to consolidate the partnerships and their trading activity had come from a manager handling CDS trades who believed that this consolidation would make trading “exotic” options with nonstandard features on behalf of the partnerships more efficient. Coplan and his colleagues represented that the trader already had planned to take these steps before they had learned of it and that whatever tax benefits might result were simply an incidental advantage. To get what they dubbed CDS Add-On past the firm’s SBS group, Coplan used the ruse to obtain the assistance of Dale Hortenstine, a member of the group and a partnership tax specialist. After the COBRA review, Hortenstine had explained to others in the firm that the reason this earlier strategy had been shut down by his group was that “we perceived there to be a lack of meaningful potential for economic profit or other business purpose to justify the transaction under an ACM analysis.”⁴⁶ The fabricated story about CDS Add-On persuaded Hortenstine to assist SISG, and the strategy was approved.

Brent Clifton’s opinion letter for CDS Add-On on behalf of Locke Liddell described the ostensible business purpose for the steps necessary to effectuate the transaction for a given partnership. The letter stated that the general manager of the partnership had enlisted the services of Andrew Krieger of Deerhurst Management based on his sterling track rec-ord. Krieger had expressed his wish “to consolidate his trading strategy in one partnership rather than among the several partnerships” because a single, consolidated account would enable him to “obtain accounting efficiencies, a greater pool of capital, and risk diversification.” In addition, such consolidation would enable Deerhurst to engage in exotic option trades through Bear Stearns, which had indicated that it would not execute such trades on behalf of individual partnerships because of administrative burdens. On the basis of these discussions and conversations with investors in the partnerships, the letter explained, Bolton had formed the limited liability company (LLC).⁴⁷

As investor experience indicated, these considerations were in fact irrelevant to the creation of the LLC involved in the CDS Add-On shelter. Joel Leonard, for instance, was a partner in a proprietary options trading company in Chicago known as Cornerstone Partners. In mid-2000, Leonard’s accountant suggested that he contact Jeff Brodsky at Ernst & Young about some transactions in which Leonard might be interested. Brodsky and other E&Y personnel then came to Leonard’s office and arranged for Brian Vaughn to participate in the meeting by phone. The first question they asked Leonard was whether Cornerstone would be earning at least $20 million that year. When Leonard indicated that the firm would come close to that, Brodsky told him that E&Y had a shelter that would defer taxes on that income indefinitely. During the meeting, SISG members tried to explain the steps of the transaction, but Leonard, despite his financial acumen, could not make heads or tails of the presentation. As he later recalled, his most important question was, “Can I go to jail for this?” After being reassured that it was legal, Leonard decided to do the transaction.⁴⁸

As with the earlier tax strategies, the SISG team was sensitive to the need to establish a record that the CDS Add-On transaction had a nontax business purpose. Whereas documents inside SISG were explicit about the tax purpose of the strategy, materials the group was to provide to clients had to underscore the purported business purpose of the transaction and could not refer to any facts suggesting that the purpose of the transaction was to obtain tax benefits. As with earlier techniques, the steps of the strategy had to appear to be the product of the taxpayer’s independent decisions over the period of the transaction. Coplan instructed: “There should be no materials in the clients’ hands or even in their memory that describes CDS as a single strategy that includes the Add On feature.”⁴⁹

In August 2000 the IRS issued Notice 2000-44, which was clearly directed at contingent liability shelters like COBRA. The SISG team never doubted that this strategy was encompassed within the IRS’s notice.⁵⁰ Rather than revisit the transactions and perhaps unwind them, as the team knew PwC had done with BOSS a year earlier, the E&Y team resisted the IRS’s position. The notice “only represented the Service’s views, which we knew would not be favorable as to the transaction,” Coplan remarked.⁵¹ Assuming that the notice had no retroactive implications, Coplan believed that since the transactions had already closed, the law firms engaged to provide opinions could analyze the facts as they existed before the issuance of the notice and still conclude that a court would more likely than not uphold the transaction.⁵² Soon after the notice came out, E&Y instructed offices that had sold COBRAs to set aside reserves to cover client audits, which it now believed were likely.⁵³ Around the same time, the Wall Street Journal published an article that singled out PwC’s tax shelter activities. Tax practice leaders were relieved that E&Y had been overlooked in the article.⁵⁴ PwC had been “too greedy” remarked Michael Kelley, the partner in charge of geographic tax practice areas who had approved the Murfam shelter.⁵⁵

Into the fall of 2000, Coplan felt it was necessary to continue to paper client files to substantiate business reasons for engaging in transactions. In the case of CDS Add-On, Coplan emailed Belle Six seeking a brochure describing Deerhurst’s trading record prior to the date of the consolidation of a client’s partnership into the LLC. “It would be useful to have such a statement in the client’s files,” Coplan observed, “showing the hopefully impressive performance of Deerhurst prior to the client’s decision to . . . transfer a portion of their trading account to the LLC.”⁵⁶ The problem was that by the time Coplan requested the brochure, clients had already made their decisions to engage in the strategy.⁵⁷

In 2000, E&Y had also begun to market another shelter that proved to be quite lucrative. The objective of this strategy, sold under the name of Personal Investment Corporation (PICO) was to defer taxation and convert ordinary income into capital gains. The strategy involved the creation of an S corporation jointly owned by the client and an investment advisor. An S corporation, like a partnership, does not pay taxes. Instead, gains or losses from its investment activities are reported directly by its shareholders. In this tax shelter strategy, after the S corporation engaged in some options trading that would produce nearly equivalent gains and losses, the taxpayer and investment advisor terminated the transaction through a series of steps that allowed the gains to be allocated to the investment advisor, while the losses could be claimed on the taxpayer’s return. As usual, the taxpayer would depict the transaction as profit-seeking investment activity. There would be no meaningful change in economic position, however, and the client would simply follow a prescribed series set of steps that produced an economic loss on paper.

To develop the PICO transaction, E&Y worked with Andrew Beer of an investment company called Bricolage. Beer had dreamed up the transaction in the “wee hours” of the night.⁵⁸ To find out if the transaction “worked,” Beer turned to Peter Cinquegrani, a tax specialist at Arnold & Porter and former lawyer in the IRS’s Chief Counsel Office. Beer had previously worked on tax strategies with Cinquegrani, including two shelters, COYNS and POPS, marketed by Arthur Andersen. After reviewing the PICO documentation in early 2000, Cinquegrani concluded that the transaction was a viable shelter. After Cinquegrani gave the transaction a thumbs up, Beer put him in touch with Richard Shapiro on the SISG team, the lead contact at E&Y during the development process.⁵⁹

For the PICO transaction to have any hope of passing IRS scrutiny, there had to be a plausible nontax reason why someone would create an S corporation in collaboration with an investment company and then buy that party out after only sixty or ninety days. The story that E&Y fabricated was that the client formed the S corporation for asset protection and estate planning purposes. The taxpayer included the investment advisor because she wanted to try out Bricolage’s trading strategy for a limited period to determine whether it was sound. At the end of that period, if she was satisfied with the results she would buy out the other members of the corporation and enter into a two-year asset management contract with the advisor or one of its affiliates.

Following a now familiar modus operandi, the SISG team had to make sure that no client materials suggested that PICO was anything other than a business strategy. One aspect of the PICO shelter that created a risk of unfavorable documentation was the E&Y fee. The firm’s fee was about 2 percent of the loss that the client wanted to generate, an amount that was quite large compared to the size of the initial investment. The size of the fee made it difficult to argue that the taxpayer expected any economic gain from the transaction. Coplan directed that a fee of only $50,000 be listed in the client’s engagement letter with E&Y for the PICO transaction. E&Y then arranged for the remainder of its fee to be paid by the client to a Bricolage entity, and for that company to pay this amount to E&Y. To support this payment, E&Y fabricated a contract under which E&Y supposedly provided consulting services to one of Bricolage’s affiliates. These contracts often were created after the fees had already been passed on to E&Y and so had to be backdated to avoid any discrepancies.

All told, E&Y marketed the CDS Add-On shelter to sixty-one individuals in 2000. CDS Add-On generated more than $24 million in revenues for the firm. The firm sold ninety-six PICO transactions to 150 individuals, generating more than $56 million in fees.⁶⁰

Dead Meat

Coplan’s concern that the files be papered to deflect IRS attention was not far-fetched. In December 2000 the IRS had begun to make inquiries of Bolton about a number of CDS shelters that E&Y had registered. In response to a letter from the IRS, Bolton insisted that “[a]t no time did [Bolton Capital] present to any investor or potential investor any other analysis or opinion related to the intended tax benefits of such a transaction.” Bolton claimed that the partnership’s investment objective was “capital appreciation for the investment through a variety of financial instruments.” In fact, the actual goal of the trading was “capital preservation,” which would leave the client’s financial position essentially unchanged. Bolton also maintained that “[t]he partnership incurred recourse debt in order to pursue trading activities, as well as make investments in swap transactions with large notional amounts.”⁶¹ In fact, debt was incurred only for the swap, not for trading activity. In addition, Bolton represented that “[a] portion of the assets of each partnership was invested to seek profits through short-term trading strategies.” In reality, trading activity was aimed at establishing the partnership as a trading business only for tax purposes. Bolton’s description also declared, “If the general partner based on market fluctuations terminates contracts early, capital gain would arise on such termination.”⁶² In fact, the plan from the outset in all cases was to terminate early; termination had nothing to do with market movements.

When the IRS began to audit David Smith’s Private Capital Management Group (PCMP) and sought information on the CDS transactions, Coplan sent an email in June 2001 to Belle Six about how the CDS Add-On shelters were numbered. “It may be too late,” he said, “but would you please reconsider your rejection of my suggestion that the different partnerships be named individually by the limited [partners] instead of being named numbers in sequence. We are obviously making it too easy for the IRS to simply find one transaction and then look for all other partnerships with a similar name filed in the same service center. The fact that they are going in sequence has me reasonably convinced they are mowing through all the PCMG returns.”⁶³

During the same period the IRS was looking into CDS, Coplan received worrisome news about COBRA. The IRS had started to investigate one of the partnerships that had used that strategy. In mid-March 2001, the partnership had received a notice from the IRS seeking information and documents. Over the next few weeks, Shapiro, Coplan, Thomas Dougherty, who had sold the strategy, and Dennis Conlon, a lead partner in E&Y’s Tax Controversy group, brainstormed about plausible business purposes that might have led the taxpayers to form the partnership and invest in foreign currency options. In May, Dougherty, Conlon, and Shapiro—the team assembled to deal with the IRS audit—appeared for an IRS interview on behalf of their clients. The team suggested to the IRS agent that in their preliminary discussions with the clients, E&Y had discussed the financial implications of entering into the transaction and that this was the basis for their clients’ decision. The E&Y audit team also claimed that the fee paid to the firm was for “financial planning and investment advisory services.”⁶⁴ When asked by the IRS agent whether the clients had received promotional materials, team members prevaricated.⁶⁵ At the close of the meeting, the team was handed an Information Document Request asking for additional documents. Request No. 9 sought marketing documents and all other documents relating to the financial or tax consequences of participating in the transaction.⁶⁶

Over the next few months, team members struggled to figure out how to answer Request No. 9 to avoid handing over COBRA marketing materials. Following Coplan’s instructions when COBRA was first rolled out, Dougherty had not left any materials explaining the tax benefits of the transaction when he pitched it to the clients, so they had no promotional documents in their possession. The problem was that Dougherty had documents in his files that he had shown clients. Over the next two months, the audit team went back and forth about how to answer the request truthfully but still avoid handing over COBRA promotional material. They did not want to suggest that they had anything in response to the request. At the same time, they did not want to indicate misleadingly that they had nothing. In early July, Dougherty had come up with a draft response that stated that the clients did not believe they had promotional documents in their files and that the request as worded was seeking information covered by the attorney-client, work product, and statutory accountant privileges.⁶⁷ Coplan continued to be dissatisfied with this response, which apparently raised too many red flags. On July 17, the E&Y group assembled to deal with the IRS audit and Coplan held another phone conversation during which they tried to come up with more satisfactory language.⁶⁸

At this point—perhaps believing that no matter how the team answered Request No. 9 the IRS would insist on obtaining COBRA documents in the firm’s files—Coplan must have panicked. That same afternoon, he sent out a blast email to the SISG team and other E&Y personnel involved in selling COBRA, instructing them to destroy all documents discussing the tax implications of the strategy. In an email with the subject line “Important—Purge of all key COBRA documents,” he wrote:

In an effort to eliminate unnecessary material from files—both paper and electronic—you are hereby instructed to immediately delete and dispose of any and all materials in your drawers and on your computers related to the COBRA transaction other than 1) documents related to the currency trades made by the client and research done to assist in deciding on those trades, 2) documents supporting the economic purpose and bona fide nature of the investment in the transaction, and 3) your copies of the opinion letters issued to the client.⁶⁹

Coplan’s purge order must have come to the attention of someone in E&Y’s general counsel’s office. The following day, Coplan sent out a follow-up email with the subject heading “CANCEL PURGE REQUESTS OF COBRA DOCS.” “Based on some information I just received from general counsel’s office,” he explained, “I must ask you all NOT to purge ANY COBRA documents to the extent you have not done so already.” By then, E&Y personnel had already destroyed COBRA promotional materials.⁷⁰

As audits of clients who participated in CDS, COBRA, and CDS Add-On increased over the next two years, tax professionals at E&Y continued their pattern of trying to mislead the IRS about the nature of the transactions. Clients were coached to tell the IRS they had engaged in the transactions for business reasons. Dennis Conlon, for example, emphasized to Leonard (the options trader who had not been able to understand the basic mechanics of CDS Add-On) that Leonard needed to say that he participated in the transaction “for the chance to make a lot of money.”⁷¹ Conlon used a spreadsheet to show Leonard the range of outcomes from the digital trading that went on as part of the transaction—trading of which Leonard had been wholly unaware. As he went over the figures with Leonard, he would say things like, “As you know, Joel,” and “So you remember, Joel, that the reason you did this trade was because . . .”⁷² Pointing to the “sweet spot” that represented the lottery payout, Conlon said, “See, you had a chance if you did this trade, that you could have made . . . almost $38 million, remember? That’s why you did this trade. See? That’s why you did this trade, right? It had that sweet spot where you could have made that money, as you know.”⁷³

In coaching clients, E&Y tax professionals seemed more intent on protecting the firm than looking out for the interests of their clients, who might have suffered fewer repercussions by coming clean to the IRS. In any event, it would only be a matter of time before the government trained its sights directly on E&Y.

Arthur Andersen

Arthur Andersen appears to have participated extensively in the market for tax shelters, and in particular in the promotion of Son of BOSS contingent liability shelters like those described in chapter 5. Because the firm went under in the wake of the Enron scandal, however, only piecemeal evidence of its involvement has come to light.

The public documents available suggest that Arthur Andersen was involved in promoting shelters as early as the mid-1990s when Paul Daugerdas, then a partner at Andersen, developed his first versions of contingent liability shelters using Treasury bills.⁷⁴ By 2000, the firm was marketing a slew of tax elimination shelters, under names and acronyms such as Short Sales, COINS, POPS, and Leveraged Options Strategy. Many of these transactions were designed and implemented with the help of Deutsche Bank and the financial boutique Bricolage. The transactions involved legal opinions obtained from Peter Cinquegrani, the Arnold & Porter tax lawyer who had provided opinions to E&Y clients on PICO, and William Bricker, a tax lawyer at Curtis, Mallet-Prevost.⁷⁵ What documents have emerged also suggest that tax leadership at Arthur Andersen drove its shelter activities. By the early 2000s, the firm had a team dedicated to developing and selling shelters. A formal “opinion committee” consisting of tax partners in the Office of Federal Tax Services, the firm’s Washington office, reviewed and approved shelters before they were marketed.⁷⁶

Arthur Andersen got in the game early and stayed in late. In 1999, Ken Mandel, the partner responsible for designing, selling, and implementing “leading-edge tax solutions” in the firm’s Western U.S. area, had developed, with the approval of firm leaders, a variation of Son of BOSS, known internally as a “call-option spread.” In 2000, the firm had about twenty of these transactions in the works, which ended up generating almost $15 million in fees. After IRS Notice 2000-44 was announced, which targeted so-called contingent liability or Son of BOSS shelters, Mandel and tax leaders at the firm persuaded themselves that it did not apply to the call-option spread and continued to promote it through at least the end of 2001.

That fall James Thomas and Edward Fox, two highly successful real estate investors and developers—and, among other things, owners of the Sacramento Kings basketball franchise—purchased the call-option spread in an ultimately unsuccessful attempt to shelter approximately $45 million in income. The ubiquitous R. J. Ruble, now a partner at the recently merged Sidley Austin Brown & Wood, provided opinion letters for the strategy. The multinational insurance company AIG, which routinely paid Arthur Andersen a finder’s fee to participate in the spread strategy, assisted in its execution.⁷⁷ After the IRS disallowed the claimed tax deductions, Thomas and Fox challenged the determination in court. In 2009, a district court upheld the IRS’s decision, concluding that the transaction lacked economic substance. It also found that the taxpayers were liable for penalties because they had failed to show that they acted in good faith in claiming tax benefits from the strategy.⁷⁸

Since that ruling, the government has presumably gone after individual taxpayers for tax benefits claimed in connection with Arthur Andersen’s shelters. It has not, however, pursued actions against the firm or the professionals involved. After its indictment in connection with its conduct in Enron, Arthur Andersen lost its audit license, and effectively ceased operations. The government had little to gain from bringing charges against the defunct accounting firm. Former clients with potential claims against the firm settled quickly to make sure they obtained some recovery. As a consequence, no detailed chronology of the firm’s tax shelter involvement has emerged.

PricewaterhouseCoopers

PwC sold about fifty FLIP transactions to clients in 1997 and 1998, twenty-six contingent deferred swap (CDS) transactions in 1998 and 1999, and around 120 BOSS transactions in 1999. FLIP was brought to PwC by Michael Schwartz, who in 1997 had come to Coopers & Lybrand, one of PwC’s predecessor firms, by way of KPMG. Schwartz had left KPMG because he was unhappy with the compensation he received for his involvement in designing and marketing FLIP.⁷⁹ After the merger between Coopers & Lybrand and Price Waterhouse, Schwartz worked in the firm’s Finance and Treasury Group. He introduced FLIP to various partners and presented the product to potential clients. By 1998, the Personal Financial Services group within the firm had assumed primary responsibility for marketing and implementing the shelter. In 1997 and 1998, Coopers & Lybrand participated in twelve FLIP transactions and PwC was involved in thirty-eight, for a total of fifty altogether.

At PwC, Schwartz recreated the type of arrangement for marketing and implementing FLIP that KPMG had used. Under the arrangement, First Union National Bank, which was later acquired by Wachovia, referred its customers to PwC for a FLIP presentation. PwC also used the investment firm Quadra, later renamed Quellos, to implement the transactions involved in the shelter. Schwartz worked with David Smith, with whom he had collaborated in designing FLIP when both were at KPMG. Implementing the shelter involved helping to set up offshore partnerships and working with various banks to arrange millions of dollars in financing. PwC did not register FLIP with the IRS but advised Quellos to do so, which it did in 1998 and 1999. PwC issued opinion letters to clients who purchased FLIP stating that the shelter more likely than not would be upheld in court if it was challenged by the IRS.

Schwartz was also involved in the marketing and sale of BOSS, which was brought to him by Smith. In the BOSS strategy, the client made a payment of 8.5 percent of the desired tax loss or sheltered income, about half of which went to PwC, the investment firm, and Refco Bank, which provided financing. As with FLIP, PwC relied on First Union for referrals, and the bank sent twenty-five investors to the firm for presentations on the BOSS shelter. BOSS had originally been reviewed and rejected by partners in KPMG’s Washington National Tax office, including Steve Rosenthal, the expert in taxation of financial instruments who, with Mark Watson, had unsuccessfully tried to derail BLIPS (as described in chapter 5). PwC decided to promote the shelter. It first used PCMG and then Bolton Asset Management, where Smith later moved his projects, to implement the transactions. Smith also brought PwC the CDS strategy that he had created.

One reason why Schwartz was so successful in marketing shelter ideas within PwC was the weakness of the firm’s internal review process. Each individual business unit within PwC created its own ad hoc review committee, which usually comprised senior tax members chosen by the tax partners advocating a new idea. The committee then analyzed whether the proposed product complied with the technical requirements of federal tax law, but did not take into account the economic substance doctrine or reputational or ethical considerations. Each committee was supposed to reach a consensus on whether the proposed transaction more likely than not would be upheld in court if subject to challenge. Individual committee members, however, were not personally required to reach this conclusion. Instead, the standard was whether a member could reasonably believe that others could come to such a conclusion based on the technical characteristics of the transaction. Once a committee approved a new product, the business unit was then free to market it without obtaining approval from anyone else at the firm. As the U.S. Senate committee investigating tax shelters later noted, “this process lacked independence from the business unit which stood to profit if the product was approved.”⁸⁰

PwC also developed and marketed at least one individualized contingent liability strategy in 2000 aimed at eliminating taxes. In the late 1990s, infighting among the heirs of the founder of the San Francisco Chronicle and owners of Chronicle Publishing Company, the media conglomerate that owned the paper, prompted the family to sell the company. In the summer of 2000, family members turned to PwC, which designed a complex transaction that allowed the family to shelter the gains from the sale using the section 752 technique on which the Son of BOSS shelters relied (described in chapter 4). Ruble, predictably, wrote the opinion letters for the transaction. After the IRS disallowed the claimed tax benefits, the family members challenged its decision in court and lost. Not only did the district court conclude that the transaction had no economic substance, it also upheld the imposition of penalties, concluding that the family did not have good cause for understating its income and had not relied in good faith on Ruble’s professional advice. The fact that the transaction was executed after the IRS issued Notice 2000-44, which made clear that it viewed Son of BOSS transactions as abusive, was not lost on the court.⁸¹

Purveyors of the tax shelters of the 1970s and 1980s for the most part had been institutions on the margins of professional respectability. In the tax shelter boom at the turn of the twenty-first century, in contrast, global accounting firms considered to be the epitome of professionalism played major roles in generating fraudulent tax losses of billions of dollars. As chapter 7 describes, the financial success of first-tier accounting firms from tax shelter work inspired at least one second-tier accounting firm to launch an aggressive effort to join the tax shelter game.