5   Watson’s Choice

As sales of the Offshore Portfolio Investment Strategy (OPIS) were winding down in 1998, leaders of KPMG’s Personal Financial Planning (PFP) group were eager to find another product with similar revenue potential. Bond-Linked Issue Premium Structure, known under the acronym BLIPS, looked very promising. Two California lawyers, Emil Pesiri (who had been involved earlier with Son of BOSS shelters) and George Theofel, had thought up the idea for BLIPS and licensed it to Presidio, John Larson and Robert Pfaff’s investment firm. BLIPS, like the earlier shelters Pesiri had done with Daugerdas in the mid-1990s, was a type of Son of BOSS shelter. BLIPS represented an important opportunity for Presidio, and solidified its position with KPMG. Larson and Pfaff had proprietary rights in BLIPS instead of having to compete with Quadra, as they had with OPIS, to be the financial advisor implementing the strategy. Through the fall, Larson and Pfaff worked closely with a multi-firm team consisting of Jeffrey Eischeid at KPMG, R. J. Ruble at Brown & Wood, and representatives of Deutsche Bank to design the shelter.

Watson’s Rise

The technical expert put in charge of shepherding BLIPS through the approval process inside KPMG was a young tax partner and rising star named Mark Watson. Watson, an accountant by training, joined KPMG’s PFP practice in 1992, after receiving a master’s in tax from Texas A&M University. Soft-spoken, thoughtful, and quietly brilliant, Watson had contemplated following in the steps of his father, a physicist, and going into academics. After two years in KPMG’s Houston office, Watson grew bored with tax compliance work. Looking for a change, he convinced his boss to allow him to do a rotation in the firm’s Washington National Tax (WNT) office, the firm’s tax think tank. The firm offered postings at WNT to the young tax professionals who had promising futures in the organization. The fact that Watson did not have a law degree was not an impediment: by the early 1990s tax professionals with backgrounds in law and accounting played nearly interchangeable roles at major firms.

Watson flourished in D.C. The WNT environment—focused on sophisticated tax research and the intellectual give-and-take among the two dozen tax experts with training in law and accounting—came close to his ideal of an academic setting. He quickly earned a reputation as a talented, hard-working gift and estate tax specialist. Watson hoped to continue his career at WNT, but firm leaders asked him to move to KPMG’s Dallas office to revive its moribund PFP practice. In return, the firm agreed to make him partner within the following year. Typically, it took entry-level tax professionals twelve years to work their way up to partnership; with the Dallas move, Watson would be made a partner in less than six. It was a difficult offer to resist.

In the Dallas office, Watson had his first experience with KPMG’s growing tax shelter practice. When Gregg Ritchie was put in charge of developing and marketing tax shelters targeted at high-wealth individuals, he invited Watson to work on developing strategies based on gift and estate tax. Watson, who was knowledgeable about legitimate strategies in his field, was not initially averse to assisting with this project.

While working with Ritchie’s group, Watson became familiar with OPIS—the aggressive strategy aimed at eliminating capital gains that succeeded FLIP. Watson was not completely persuaded that a court would uphold the transaction if challenged by the Internal Revenue Service. At one point in 1998, Ritchie asked Watson to make a sales call to a Dallas client who might be interested in buying OPIS. Watson was not enthusiastic, but recognized the importance of being a team player and agreed to visit the prospective buyer. When Watson finished his pitch, the client asked him directly whether Watson would purchase OPIS. Relieved that the client asked, Watson bluntly answered “no.” After a second equally unsuccessful sales call, Watson was certain that he had little taste for product sales or the firm’s expanding tax shelter practice.

After a few years, Watson grew tired of running the Dallas PFP practice and had become uncomfortable with its increased focus on shelters. He decided that the time had come to leave KPMG and pursue a teaching career. The firm’s tax leaders, recognizing that KPMG was about to lose a highly talented tax expert, offered him a position back at WNT. Watson was delighted to accept. Back in D.C., he would be able to revisit the intellectual excitement he experienced during his first rotation. He would also be insulated from having to make sales calls. He assumed, moreover, that he would have a say in which products were approved for sale. When Watson returned to WNT in August 1998, he was put in charge of a growing staff charged with providing technical support to the PFP practice. In his new position, he was also responsible for overseeing the approval process for shelters to be promoted to high-wealth individuals through the group.

Soon after his arrival, Watson began to question whether his position at WNT would give him the authority to prevent conduct he considered questionable. The issue was how purchasers of the OPIS tax strategy could report their gains and losses on their tax returns. To Watson’s dismay, a member of his group circulated a memo to Ritchie and his team arguing that shelter purchasers could use a grantor trust to “net” their gains and losses, a strategy intended to avoid detection of the shelter by the IRS.¹ In essence, the trust was supposed to function as a black box so that clients could avoid separately identifying their gains for the year and the loss they claimed from the shelter. This technique was directly contrary to federal income reporting requirements. Because a grantor trust does not change the ownership status of a property for income tax purposes, income tax laws did not recognize it as an appropriate form of reporting.²

When Watson discovered that a member of his group had recommended this strategy, he made clear his view that the technique bordered on criminal conduct. In September 1998, he wrote: “When you put the OPIS transaction with this ‘stealth’ reporting approach, the whole thing stinks.”³ Realizing several months later that partners were still recommending this reporting technique to clients, Watson followed up with a detailed analysis of why it was not permitted. He concluded: “I believe that we are filing misleading, and perhaps false, returns by taking this reporting position.”⁴ As the lead expert on trusts and estates practice, Watson assumed that once he had stated his opinion, grantor trust netting would cease. He subsequently discovered that, despite his strongly worded warning, partners were still advising clients into 2000 that they could use grantor trust netting as a reporting technique to hide their participation in tax shelters. (In the summer of 2000, the IRS issued a notice stating that this reporting technique could constitute a criminal offense.)⁵

When he returned to WNT in the summer of 1998, Watson reconnected with Steven Rosenthal, a well-respected partner with expertise in the taxation of financial products and services, whose office was down the hall. Watson had gotten to know Rosenthal during his earlier rotation at WNT and both had been named partners the same year. Of similar intellectual temperament and professional outlook, Watson and Rosenthal, whose offices were near each other, would drop in on each other often to talk about substantive tax law issues.

A 1985 graduate of the University of California at Berkeley Law School, Rosenthal had substantial experience in private and government practice, particularly in the area of taxation of financial services. The field held special appeal for Rosenthal. He had majored in mathematics in college and earned a graduate degree in public policy, focusing on finance, at Harvard. After practicing at the highly respected law firm Wilmer Cutler, Rosenthal went to work for Congress’s Joint Committee on Taxation, specializing in financial institutions and products policy. When Republicans running on an antitax platform gained control of Congress in 1994, Rosenthal left the Joint Committee and joined KPMG as its National Director for Financial Services in the Financial Institutions and Products Group.

Recognizing Watson’s intelligence and enthusiasm, Rosenthal took on an informal mentoring role, acting as a sounding board and helping him think through substantive tax problems and workplace dynamics. Watson, for his part, appreciated Rosenthal’s sophisticated analysis and good judgment and looked to him for guidance on problems outside his area of expertise.

One frequent topic of discussion was the distinction under income tax law between legitimate business strategies, which might also confer substantial tax benefits, and abusive tax shelters, that is, transactions whose primary purpose was tax-motivated. Because the doctrine rarely arose in Watson’s field of gift and estate tax, he found Rosenthal’s detailed and nuanced explanations very helpful. As Rosenthal emphasized time and again, if a transaction lacked core economic characteristics, such as economic risk and reasonable opportunity for profit, he would not opine that it would “more likely than not” survive IRS challenge. This was the required standard for providing an opinion letter to a client that would avoid penalties if the tax benefits from a transaction were disallowed. As KPMG’s emphasis on product development intensified, Rosenthal and Watson would often share their worries about the direction of the firm.

In recent years, Rosenthal had blocked some of the more questionable shelters under consideration. In 1997, David Smith, who had been involved in the creation of FLIP, joined Quadra, the Seattle-based firm that had functioned as an investment advisor for that transaction. At Quadra, Smith developed a shelter known as Contingent Deferred Swaps (CDS), which was offered to KPMG. Asked to review the shelter, Rosenthal, with Watson’s support, quietly killed it. After KPMG turned it down, Smith brought the shelter first to Ernst & Young, which sold it to more than 132 clients,⁶ and later to PricewaterhouseCoopers (PwC), which sold approximately twenty-six transactions.⁷ After OPIS sales were officially discontinued at KPMG in late 1998, pressure at the firm mounted to find a strategy aimed at high-wealth individuals who were looking to shelter significant income or gains.

BLIPS

Mark Watson first saw BLIPS in early February 1999, in the form of a draft opinion letter Eischeid sent to him. Soon after Eischeid took over Ritchie’s role in running the PFP products group, he and Watson were named co-leaders of a group called the Innovative Strategies group, the successor of CaTS. Eischeid was in charge of developing and implementing products aimed at high-wealth individuals. Watson’s responsibility was to make sure that these products complied with applicable tax law.

BLIPS was developed in KPMG’s PFP practice because line partners, whose clients were high-wealth individuals, were the obvious nationwide distribution channels for the product. Throughout the fall, Pfaff and Larson at Presidio, Ruble at Brown & Wood, representatives of Deutsche Bank, and Eischeid and Bickham at KPMG had worked together to flesh out the details of BLIPS. In early 1999, the multi-firm team had a prototype to present to WNT. As technical reviewer for products in the PFP practice, Watson was charged with shepherding BLIPS through the development and approval process.

Like the Son of BOSS shelters, BLIPS was a so-called contingent liability shelter, which sought to take advantage of how the partnership taxation rules treat different liabilities. The basic idea in broad outline was to create an artificial basis, or purported investment, in a partnership.⁸ The taxpayer would borrow the money necessary to buy a financial asset, which represented a cost that could be used to reduce taxable income by increasing the basis. Later cancellation of part of that cost of indebtedness, it was claimed, would not reduce the basis.

Like the Short Options Strategy shelter described in chapter 4, BLIPS sought to take advantage of a purported asymmetry in the partnership rules, based on the Helmer case, which held that in certain situations the extinction of a contingent liability—one whose occurrence is not certain—is not deducted from the basis (or investment). In BLIPS, this was supposed to work by having a premium loan, that is, an above-market-interest loan, and a corresponding penalty for early repayment of the loan. Because the interest rate on the loan was above market, the bank would agree to advance the client an amount in excess of the face value of the loan.

The taxpayer would then transfer the loan proceeds to a partnership. When the taxpayer exited the transaction and dissolved the partnership, she would “repay” the face amount of the loan and the penalty, so she would not have made or lost money. Because the penalty only applied if the loan were repaid early, the argument was that it was “contingent” and should not be deducted from the basis in the partnership, which was equal to the amount of the premium. Under this rationale, eliminating liability for the interest on the loan would not be used to reduce the basis in the partnership. After the partnership was dissolved, its value would be equivalent to a nominal amount. (The net of the loan proceeds and premium advanced and the repayment with the penalty was zero.) But, the argument went, the taxpayer would still be able to claim the amount of the partnership’s basis, equivalent to the amount of the original premium, as a loss and deduct it against her personal income under the partnership tax “pass-through” rules.

The lynchpin of the strategy was the above-market loan, whose premium was equivalent to the amount of tax loss to be generated. To eliminate any economic risk for the client, who would otherwise be deterred from entering the transaction, the loan was made non-recourse, meaning the client was not personally liable for repayment. To make the strategy look legitimate, the promoters structured a series of complex steps that were to be executed over a seven-year period. The expectation, however, was that the client would get out after the first stage—sixty days long—and realize the loss. In this first stage, the required investment was the fee for the transaction, set at 7 percent of the anticipated tax loss. With each stage over the following seven years, the amount that a client was required to invest increased. In addition, the seven-year period made it possible to set the loan interest at a plausible rate. During the period of the transaction, some amount of the money that the client had put in was used to trade on foreign currency options to make the transaction look as if it involved actual investment activity.

The moment BLIPS landed on Watson’s desk in early February 1999, firm leaders began to exert pressure to approve it as soon as possible. To make sure Watson got the message, that same afternoon Eischeid emailed Doug Ammerman, the head of the PFP practice and Watson’s direct report, and John Lanning, the firm’s vice chair of Tax Services, to assure them that approval was expected within the month.⁹ Lanning immediately wrote back exhorting Watson to approve BLIPS in a “dramatically accelerate[d] time frame.”¹⁰ Lanning may not have known much about whether the shelter would pass muster, but he apparently trusted Pfaff’s judgment that the shelter worked—or perhaps it didn’t matter. Under pressure to generate revenue for the tax practice, Lanning had his sights on the strategy’s profitability. In the spirit of first to market, he emphasized, a month was “much too long,” given “the market potential” of the product.¹¹

Watson studied the eighty-page opinion and a PowerPoint presentation of graphs that illustrated the steps of the transaction.¹² BLIPS did not touch on gift and estate tax, Watson’s area of expertise, but drew instead on partnership tax, income tax, the taxation of financial products, and international tax, among other specialized fields. Recognizing he was out of his depth, Watson began to identify the technical issues that arose in the transaction and enlist the participation of the various WNT experts needed to be involved in the approval process.¹³ The team he assembled included Rosenthal, who would address the taxation issues around the bank loan; Richard Smith, the partnership tax expert who had recently come from the IRS and was a rising tax leader at the firm, and who would look at the applicable partnership tax rules; and Jim Sams,¹⁴ who would research the tax rules connected with the foreign currency trading that was the ostensible investment piece of the transaction. Within a few days of making his initial assignments, Watson requested an update, explaining that Ammerman and Lanning were expecting a prompt progress report. “I don’t like this pressure anymore than you do,” he apologized in a postscript to his email.¹⁵

BLIPS presented several technical tax questions, but the hot potato that no one wanted to handle was whether the transaction had economic substance. To issue an opinion that the shelter would more likely than not be upheld in court if challenged by the IRS, the firm had to rely on representations from the investor that he or she had entered the transaction for the purpose of making a profit. Under the standards that applied to tax practice, the firm was not permitted to take the investor’s representations at face value; its reliance on the investor’s representations had to be reasonable. This meant that the transaction had to have the indicia of a legitimate investment. Someone at WNT would have to take responsibility for concluding that, given all he knew about the transaction, he could still envision an investor entering into it to make a profit. Because the transaction had been designed around the client’s taking a tax loss without incurring any financial risk, reaching this conclusion in good faith was a challenge.

Because the doctrine of economic substance infrequently applied in gift and estate tax, Watson did not believe he was in a position to make this judgment. Watson first delegated the question of economic substance to Larry DeLap, head of KPMG’s Department of Professional Practice and the partner responsible for risk management throughout the tax practice. Apparently seeking to avoid responsibility for making this critical call, DeLap soon punted the question to Philip Wiesner, the head of WNT.

After some preliminary work on their assignments, the BLIPS review and development team met in mid-February at Rosenthal’s suggestion to discuss the various substantive issues and formally divide up responsibility for each one. During the brainstorming session, the group identified so many significant weaknesses and analytic flaws in the opinion letter that several participants left the meeting assuming that the product was not viable. Many of the questions went to whether tax regulations permitted the tax treatment sought and whether the opinion letter adequately discussed the applicable legal rules. The toughest issue, however—and the one about which the participants had the greatest doubts—was economic substance. Watson and Rosenthal, who had shared their many reservations privately, were relieved that BLIPS seemed to be dying a natural death, without their active intervention.

To their chagrin, BLIPS was resuscitated just two weeks later by Richard Smith. Smith and Wiesner had assumed joint responsibility for economic substance and, inexplicably, concluded that the transaction did indeed work. With the other outstanding technical issues also apparently resolved, BLIPS was back on track for quick approval.

But the problem of economic substance reared its head again. In mid-March 1999, Smith or someone else had begun to wonder what plausible business reason might lead an investor to borrow at an above-market rate. Since the loan was the heart of the shelter, an investor had to have a non-tax-based explanation for why the transaction was structured around it. Smith asked Rosenthal, an expert in taxation of financial instruments, to look into the premium loan to confirm that it was consistent with industry practices. Rosenthal had assumed his limited role in BLIPS was over, but agreed to follow up with Amir Makov, a financial markets quant recently hired at Presidio to implement the investment piece of the transaction.¹⁶ During their conversation, Makov tried to produce a plausible explanation, but Rosenthal was skeptical. Following up in an email to Smith, Rosenthal neutrally noted that he continued to question the rationale for the loan premium.¹⁷

Apparently eager to launch BLIPS, PFP leaders misconstrued Rosenthal’s email. Within a few days, Ammerman, the head of the group, announced that BLIPS was approaching final approval by DeLap, who was supposed to provide the last sign-off. According to Ammerman, Rosenthal had engaged in “extensive discussions” with Presidio and arrived at a “positive” assessment about the economic substance of the premium loan.¹⁸ When he saw Ammerman’s email, Rosenthal made clear that his assessment was not positive and that it had been based on a single conversation with Makov. In addition, Rosenthal underscored that he had not been asked to assess the business purpose or economic substance of the BLIPS investment program or the transaction generally, about which he continued to harbor doubts.¹⁹ Meanwhile, word quickly got back to Presidio that Rosenthal wasn’t predisposed to help BLIPS get approved and Larson instructed Makov to avoid any further interactions with him.²⁰ Larson also directed Makov to engage in immediate damage control and draft a memo explaining the business reason for the premium loan. Presidio did its best to prevent the memo from getting into Rosenthal’s hands, sending it instead to Randy Bickham and BLIPS’s other proponents within KPMG.²¹

A few weeks later, as part of the review process, Jeffrey Zysik, the head of the Tax Innovation Center (TIC) assisting Watson, sent DeLap a revised opinion letter, the BLIPS PowerPoint presentation, and a draft engagement letter for final approval.²² That evening, DeLap sent a lengthy email back identifying nearly thirty issues that still needed to be resolved and forwarded it to Stein to keep him updated. Behind the scenes, Stein was presumably checking with Pfaff, who would have been describing BLIPS’s progress and the obstacles it was encountering. Stein was clear that he wanted the process wrapped up quickly. Writing back to DeLap, Stein sarcastically noted his “penetrating” analysis and made clear that he expected by this point that there would only need to be “perfunctory review and approval.”²³ Addressing the remaining questions, he noted, could require “at least” two more weeks. Stein instructed DeLap to get together with “whoever was driving the process at PFP” and “resolve the issues as quickly as possible.”²⁴

It is not clear what DeLap thought. His first reaction to Stein’s email was to point out that his role was more than a “rubber stamp.”²⁵ The remaining issues were not just a matter of massaging the language of the draft opinion but raised substantive tax law questions that needed to be addressed by experts at WNT. But the very next day, in an unexplained about face, DeLap wrote Stein to reassure him that most of the issues had been worked out. Nonetheless, he still had doubts about economic substance: Could the firm reasonably rely on investors’ representations that they intended to remain in the seven-year investment program when all “indications suggested that the intent was to bail out after 60 days?”²⁶

DeLap’s thinking is difficult to fathom, but he apparently was involved in a complicated defensive maneuver. Unwilling to risk the wrath of Lanning and Stein, who could make his life unpleasant if they chose, DeLap presumably did not want to take responsibility for vetoing BLIPS on his own despite his serious reservations about the product. Having made sure earlier that he was not responsible for resolving the issue of economic substance, he could repeatedly express his doubts in the hope that someone else would decide to pull the plug. In addition, putting his concerns in writing would allow him to distance himself down the road if, after the product was approved and marketed despite his concerns, BLIPS’s lack of economic substance came back to haunt the firm.

In the following weeks, Bickham and DeLap worked on a final round of edits on the opinion letter.²⁷ Watson, meanwhile, confirmed that the technical issues were resolved, while carefully avoiding the question of economic substance.²⁸ Based on his conversations with Rosenthal, Watson continued to have serious misgivings. He did not believe, however, that he had sufficient authority to stop the transaction when Wiesner, a well-respected lawyer with significant expertise in the area and his immediate superior, had made the judgment that the transaction worked. By the end of April, BLIPS looked good to go.

Around this time, Watson began quietly to make inquiries about positions at the national tax offices of the other four accounting firms. He soon discovered that most of the firms were not about to offer him a partnership-level position given his relative inexperience. One firm expressed interest, but its hiring and recruitment process dragged on through the summer and into the following year without closure. Watson realized that he would not be able to make a gracious exit to an equivalent job at another firm. If he left, he would have to sacrifice his enviable position at the center of the firm’s sophisticated tax practice, which he had worked very hard to obtain.

“Pie in the Sky”

With BLIPS poised for final approval by DeLap in April 1999, Doug Ammerman, the head of the PFP practice, assembled a BLIPS task force to begin the roll-out. A quick launch was imperative as there needed to be a several-month period to identify prospective investors, persuade them to sign up, and close the transaction—with enough time left for them to exit the transaction and realize the loss before year’s end. During the last weekend in April, Ammerman summoned a dozen PFP client service partners, representing several regions of the United States, to a meeting at the Dallas/Fort Worth Airport to discuss implementation with the leaders of the BLIPS development team, representatives of Presidio, and Watson, who was available to discuss any technical questions that came up. Watson, for his part, secretly hoped that he would find out information about the transaction that would allow him to derail BLIPS.²⁹

At the meeting, Eischeid and Bickham described what they called the “rules of engagement.” For an investor to be eligible to participate, he had to seek to shelter a minimum of $20 million in gains or income. In addition, leaders of the Department of Professional Practice had capped BLIPS sales at fifty transactions. As Eischeid and Bickham explained, the high participation threshold and the limit on the number of transactions were intended to lower the risk that the transaction would be detected in a random IRS audit and that the agency would notice that several taxpayers had engaged in nearly identical transactions with the same preorchestrated steps. Each investor would be required to pay a 7 percent fee, 1.25 percent of which represented KPMG’s share, with the rest shared among the other promoters and facilitators. Credit for revenue generation would go exclusively to the PFP practice group, with other partners receiving cross-selling credit.³⁰ Eischeid and Bickham also explained that because the loan was non-recourse, the bank was putting strict limits on the use of the proceeds, effectively requiring the investor to keep all the funds in a money market account at the bank. The bank also required repayment by year’s end. Watson now understood why the bank would agree to a non-recourse “loan”—an issue that had been puzzling him and Rosenthal. The bank wasn’t actually going to allow the funds to leave the bank.

Eischeid and Bickham warned the audience members to be especially careful not to leave the draft opinion letter or PowerPoint presentation explaining the steps of the transaction that generated the tax loss with prospective clients. Were the IRS to get its hands on these documents during the course of an audit, they explained, investors would no longer be able to claim that they had entered the transaction for the purpose of making a profit, since the necessary steps to exit the transaction early and take a tax loss had been spelled out from the beginning.³¹

Toward the end of the session, Eischeid introduced Makov, the investment expert at Presidio, to explain the trading activity involved. The strategy involved shorting foreign currencies pegged to U.S. dollars in anticipation that the currencies would devalue during the investment period. Makov had been instructed not to be transparent about this lottery-like approach, and to describe, instead, the macroeconomic conditions in a developing economy that would lead to a sudden devaluation. As Makov spoke, a murmur of excitement rippled through his audience. Several listeners seemed to believe that investors in BLIPS stood to make a 300 percent return in a sixty-day period. Anxious to dispel this misimpression, Makov pointed out that this scenario was “pie in the sky” and that the possibility of making a profit from the transaction was “remote.”³²

As Makov described the slight prospect of profit, an arm grabbed him and pulled him back into his seat. Within minutes, Eischeid and Larson were hustling him out of the meeting, explaining that he and Larson had a plane to catch.

On hearing Makov’s words, Watson thought he finally had ammunition to derail BLIPS—the investment advisors themselves had acknowledged that the transaction had no business purpose, and the loan that drove the transaction was not a real loan. He had serious concerns, however, about the consequences of using this information. On his way home from Dallas, he visited another KPMG partner who was a friend. After Watson described his dilemma, the friend reluctantly encouraged him “to do what he had to do.” Watson recognized that if he tried to stop BLIPS at this stage, his career at KPMG, and certainly his enviable position at WNT, would be in jeopardy. Weighing his options, he decided to follow his instinct and his friend’s advice. Back in the office early Tuesday morning, Watson sent an email to DeLap informing him that he (Watson) was not comfortable issuing a more-likely-than-not opinion and acknowledging that he would “catch hell” for the email.³³ The following day, he sent a similar email, recommending that the product not be released, to his other colleagues working on BLIPS.³⁴

Meanwhile, Bickham was still struggling in the opinion letter to address yet another question that DeLap had raised back in February and that had not been satisfactorily resolved: who, for tax law purposes, was the borrower in the transaction? If the bank knew from the beginning that the investor would immediately donate the loan proceeds and the liability to the partnership, then a plausible argument could be made that it was the partnership and not the investor who was the borrower, and the investor was not entitled to the tax benefits from the deal. In late April, Rosenthal was enlisted to help Bickham address the issue. Around the same time, DeLap circled back to Rosenthal on the justification for the premium loan, sending along the memorandum that Makov had prepared after their phone conversation.

Rosenthal was not impressed with Makov’s analysis. Making his skepticism clear, he noted that the memorandum was filled with “mumbo-jumbo that [was] not particularly useful.” The argument for the premium loan relied on a concept that was “irrelevant” to the transaction and was “suspect.”³⁵ After researching the separate question of who is the borrower under tax law, Rosenthal also concluded that the tax benefits from the transaction would not be available to the investor, which he communicated to DeLap.

Having heard Watson’s renewed doubts about profit motive and the status of the bank loan and Rosenthal’s legal analysis of the identity of the borrower, DeLap and the BLIPS team passed along these concerns to KPMG’s tax leadership. Laying responsibility squarely on Watson, DeLap observed that he (DeLap) did not believe that the product should move forward “when the top PFP technical partner in WNT believes it should not be approved.”³⁶

On the afternoon of May 7, 1999, Smith and Wiesner hastily convened a meeting with Watson and Rosenthal to determine whether the problems Watson flagged could be addressed. Watson went into the meeting hoping that Wiesner and Smith would recognize the significance of the new information and reevaluate the decision to go forward with BLIPS. Over the course of the meeting, however, Smith and Wiesner concluded that appropriately drafted representations signed by Presidio, the investment firm, which stated that the transaction had a reasonable probability of making a profit, would cure the problem of the investor’s motive. They also decided that the problem of whether there was a bona fide loan could be resolved with appropriate representations from Deutsche Bank. In essence, the bank had to state that it would treat and record the loan in the same manner as other loans.³⁷

Although Watson was unsure that the representations from Presidio would suffice, he became persuaded that it was not up to the WNT to vet the investment aspects of the transaction. Amir Makov, who had supposedly been brought on because of his investment expertise, was the best judge of the transaction’s profitability. As to the proposed representations by the bank, Watson had doubts that Deutsche Bank would sign on. But if the bank did agree, Watson would have some assurance that the transaction was legitimate. Watson reasoned that Deutsche Bank, a global financial institution, and its lawyers, the renowned corporate firm Shearman & Sterling, would not risk their reputations on an abusive tax shelter. What Watson didn’t know is that Makov had been hired to give the transaction a fig leaf of legitimacy and knew from the outset that the investment arm of the transaction was a sham.³⁸ Watson also didn’t know that officials at Deutsche Bank, who had vetted the transaction internally and carefully weighed the reputational risks of being associated with a tax shelter, were in on the ruse.³⁹ At the close of the meeting, Watson was tasked with drafting the specific representations required from the investment advisory firm and the bank.

The proposed representations might well solve some problems, but the question of the borrower’s identity persisted. As Rosenthal explained to Wiesner and the other senior partners involved, “Actual indebtedness only occurs where there is an economically significant change in the taxpayer’s wealth,” which did not occur here where the investor is merely “serving as an intermediary” and the “partnership [could have] borrowed the proceeds directly.”⁴⁰ Unwilling to accept the implications of Rosenthal’s analysis, Smith wrote back: “Based on your analysis below, do you conclude that the tax results sought by the investor are ‘NOT more likely than not’ to be realized?” “Yes” Rosenthal answered tersely.⁴¹

In the meantime, Watson’s earlier email describing his concerns was making the rounds. When Ammerman and Lanning read it, they were furious. Ammerman called from the West Coast to berate Watson. Lanning sent an incensed email to the group:

I must say that I am amazed that at this late date (must now be six months into this process) our chief WNT PFP technical expert has reached this conclusion. I would have thought that Mark would have been involved in the ground floor of this process, especially on an issue as critical as profit motive. What gives? This appears to be the antithesis of “speed to market.” Is there any chance of ever getting this product off the launching pad, or should we simply give up???⁴²

Watson assumed that he would be fired. His back against the wall, he tried to explain that the problem was not due to an error or oversight on his part. His earlier analysis had been based on the only documentation that had been made available to him, the draft opinion and slide presentation sent by the BLIPS development team. It was only when he had met Presidio representatives on May 1, he explained, that he learned significant new information about the lack of profit potential and absence of a bona fide loan.

On Monday morning, May 10, 1999, Wiesner emailed his views to the tax leadership and the BLIPS development team. A 1971 graduate of Columbia Law School, Wiesner had joined KPMG in 1979 after a stint in the U.S. Department of the Treasury. He was a highly regarded tax attorney among KPMG partners and in the larger Washington tax community, with a reputation for careful and sophisticated analysis.⁴³ With the firm’s growing emphasis on tax products, however, Wiesner may have sensed that his lack of enthusiasm for shelters had rendered him vulnerable. Rumors were circulating at WNT that if he was not more gung-ho, he would soon be replaced as head of the national office.

Wiesner’s first instinct was to try to put the rabbit back in the hat or at least pretend that the rabbit was not in plain sight. In an email to the WNT review team, he noted that he “was under the impression that everyone had signed off on their respective technical issue(s)” and that he “had signed off on the overall more likely than not opinion.” He was “troubled that at this late date the issue is being revisited, and . . . a prior decision being changed on a technical issue?!” Having apparently concluded that the transaction had economic substance as long as clients signed appropriate representations about their profit motive, Wiesner’s impulse was to close his eyes to evidence that the transaction would not yield a profit—evidence that cast serious doubt on the reasonableness of the firm’s reliance on the client’s representation.⁴⁴

In a second email to the tax leadership and the WNT reviewers, Wiesner was determined to show that he could—and would—pull the trigger. Several months earlier, he noted, he had concluded that he could reach a more-likely-than-not decision based on the information and the proposed representations provided at the time. He continued to stand by that original opinion, even in the face of the new information from Presidio. What an investor had in mind in entering into the transaction was at this stage hypothetical. As Wiesner noted: “The real ‘rubber meets the road,’ will happen when the transaction is sold to investors, what the investors’ actual motive is for investing [in] the transaction, and how the transaction actually unfolds.”⁴⁵ In other words, the issue of whether a client truthfully believed that the transaction could reasonably yield a profit and the reasonableness of the firm’s reliance on that representation could be deferred to implementation.

According to Wiesner, the business considerations turned on whether the firm had sufficiently protected itself from liability in the language of the engagement letter the client would be required to sign. Wiesner also wanted to be sure that the firm was being “paid a lot of money for its opinion” because the IRS would consider BLIPS a very aggressive transaction, which clearly fell within the “tax shelter orbit.” Ultimately it came down to a “business matter.” Wiesner’s opinion was that the moment had come for the firm to make a decision. His vote was to go forward with BLIPS. “The time has come to shit and get off the pot,” he concluded.⁴⁶

Soon Stein weighed in: “It’s shit OR get off the pot. I vote shit.”⁴⁷

Ammerman, the head of the PFP practice, agreed. The team he had assembled in Dallas, he wrote to the group, was ready to start working with PFP line partners to identify prospects and begin marketing.⁴⁸ Watson’s concerns were being addressed by Presidio and Deutsche Bank and should not, he assured the group, delay the roll-out. In conclusion, he “concur[red]” with Stein.⁴⁹

Later, on May 10, Wiesner and Smith met again with Watson and Rosenthal to discuss the problem of whether the borrower was the partnership or the taxpayer for purposes of taking the loss. Wiesner and Smith argued that under applicable legal principles, the taxpayer could be treated as the borrower, and concluded that the firm was ready to issue a more-likely-than-not opinion for BLIPS. Watson, who was out of his depth, refrained from expressing a view. Rosenthal was still not persuaded. Carefully noting his disagreement, he observed, “though reasonable people could reach an opposite result,” he could not reach a more-likely-than-not opinion on the issue of who was the borrower.⁵⁰

After the meeting, Wiesner announced that the group, over Rosenthal’s dissent, had decided to move forward with the product. DeLap, apparently still hoping that someone else would stop BLIPS—so long as it was not he—asked Watson in an email where he stood. “The essential question is whether you are comfortable with a more-likely-than-not opinion on this product,” he insisted.⁵¹

Watson was astounded. In his experience, DeLap had always taken a forceful leadership role on professional practice issues, but now he was refusing to take a stance, looking instead to a junior partner to pull the plug. Watson was being thrown to the wolves, but was powerless to confront DeLap, a senior partner with much more experience, whose office was, in any case, three time zones away in California.

At this point, Watson believed that he had done all he could to stop BLIPS. The final decision was in the hands of his superiors, Wiesner and Smith. DeLap had declined to get involved. In a moment of self-doubt, Watson decided—despite deep misgivings—that it was acceptable to defer to their judgment. They had much greater expertise in the area, after all, and dissenting, when it was clear that the tax leaders were set on moving ahead, would only bring him grief. There was no point in holding out. He wrote to DeLap: “Larry, I don’t like this product and would prefer not to be associated with it. However, if the additional representations I sent to Randy on May 9 and 10 are in fact made, based on Phil Wiesner’s and Richard Smith’s input, I can reluctantly live with a more-likely-than-not opinion being issued for the product.”⁵²

DeLap passed on Watson’s lukewarm endorsement to the other tax leaders. Stein’s answer was concise: “Yikes!”⁵³ Neither he nor the other tax leaders paused to reconsider the wisdom of going forward with BLIPS.

“Get Out of Jail Card”

During a meeting of the managing partners of the PFP in Colorado Springs later in the spring of 1999, Watson tried once again to prevent BLIPS from being marketed. Ammerman, the head of the group, invited Randy Bickham to the meeting to discuss the product’s design, implementation, and revenue potential. After the presentation, Watson began to question Bickham about the reasonableness of relying on investors’ representations about their profit motive. When the argument between Watson and Bickham escalated into a shouting match, Ammerman loudly instructed Watson to sit down and be quiet. Throughout the heated discussion, the other senior partners in the group sat by silently, none willing to follow Watson’s lead and question the propriety of KPMG’s being involved in a transaction like BLIPS. Watson was ready to throw in the towel again.

But another opportunity to stop BLIPS arose in the summer. Watson and Rosenthal were put in charge of reviewing the loan documents prepared by Deutsche Bank’s lawyer, Gerald Rokoff, and the team of tax and corporate finance lawyers he had assembled at Shearman & Sterling.⁵⁴ If Watson and Rosenthal continued to harbor some lingering hope that the loan was legitimate, any doubt was dispelled in mid-July when they saw the documents, which made crystal clear that the proceeds and control over their use would remain with the bank. The loan was a loan in name only.

Meanwhile, a new player had arrived on the scene. Despite Phil Wiesner’s support for BLIPS, he had been sidelined as partner in charge of WNT to make room for David Brockway, an important senior hire. Brockway, a former chief of staff of Congress’s Joint Committee on Taxation, had left the prestigious law firm of Dewey Ballantine, after nearly two decades of tax practice, to join KPMG. A brilliant tax lawyer, Brockway was credited with designing the ground-breaking tax reform package that eliminated the first round of tax shelters in 1986 when he served as director of the Joint Committee on Taxation.⁵⁵ He had a reputation as an independent thinker, and Watson and Rosenthal were hopeful that he would put an end to the BLIPS shenanigans once he became aware of them.

His optimism rekindled, Watson sent an email in late July 1999 to the senior tax leadership, including Brockway and DeLap, describing the problems he and Rosenthal had found with the loan documents. “[O]ur primary concern,” he wrote,

is that these documents do not give the client (“the Borrower”) meaningful use or control of the loan proceeds (i.e., the bank (“DB”) will not allow the borrower access to the proceeds nor allow the Borrower effectively to determine how the proceeds will be invested)[.] As a result, we are concerned that (i) a bona fide loan may not exist, [and] (2) if a loan does exist, that the Borrower may not be considered the true borrower. . . .

Further, because DB effectively controls how the loan proceeds will be invested and because it seems to have no economic incentive to allow such proceeds to be invested in anything other than very safe investments (e.g., money market accounts), we are concerned as to whether there could be a reasonable expectation of making a profit on the loan transaction given that the cost of borrowing may far exceed the expected rate of return generated by the investments purchased with the loan proceeds. Perhaps Presidio can provide us with an economic analysis that will alleviate our fears on this matter. Such an analysis will also be important if a BLIPS transaction is ever challenged by the IRS.⁵⁶

To Watson and Rosenthal’s astonishment, the email was met with silence. Watson tried to schedule a meeting with Wiesner, to no avail. Rosenthal attempted to meet with Brockway but was unsuccessful.

Watson and Rosenthal were perplexed by Brockway’s failure to respond. In retrospect, there are plausible explanations for his inaction. Having just come on board to oversee a busy office, he was likely disinclined to revisit decisions—however questionable—that had been made prior to his tenure. BLIPS was one of several transactions in the process of implementation that had been vetted and approved in previous years. To reconsider those that appeared problematic would have required a substantial investment of time and energy that Brockway might not have wanted to divert from ongoing projects. Brockway may have also feared that reviewing BLIPS would open the door to reconsidering a number of other transactions, putting at substantial risk fees that the firm had already banked on. In addition, Brockway might not have realized that BLIPS presented a level of risk to the firm an order of magnitude greater than other transactions being implemented when he arrived. Another possible reason was that he did not want to second-guess Wiesner, since, like other members of the tax bar, he likely would have held Wiesner’s professional judgment in high regard. Wiesner’s continued presence at the firm—after having been displaced as the head of WNT—put Brockway in an awkward position. Management prerogatives and intraoffice sensitivities, in other words, may have countered any inclination he had to consider BLIPS on its merits.⁵⁷

After several days of waiting to hear from Brockway, Rosenthal persuaded Watson to give it one last try. In an email on August 4, 1999, Watson wrote: “I feel it is important to again note that I and several other WNT partners remain skeptical that the tax results purportedly generated by a BLIPS transaction would actually be sustained by a court if challenged by the IRS. We are particularly concerned about the economic substance of the BLIPS transaction, and our review of the BLIPS loan documents has increased our level of concern.”⁵⁸ To underscore that Watson and he had not been asked to opine on economic substance, Rosenthal followed up a few minutes later, emphasizing that even though he also continued to be “seriously troubled by these issues,” they were not assessing the economic substance of the transaction, but deferring to Wiesner and Smith on these issues.⁵⁹

This time the response was swift. Shortly after the email went out, Smith marched into Watson’s office and showed him a draft email by Wiesner, in which he stated his opinion that more likely than not a court would hold that the transaction had economic substance. That afternoon, Watson sent an email to the group stating that the review of the loan documents was complete and they were “ready to proceed.”⁶⁰

Bickham forwarded the email to R. J. Ruble, the lawyer at Brown & Wood who was to provide a second “independent” opinion on BLIPS. “We have received our ‘get out of jail card’ from WNT,” he wrote.⁶¹

In the meantime, Deutsche Bank, which had provided most of the funding, failed to sign on to the representations requested by the BLIPS team. Watson and the tax leaders had requested the bank to represent that the above-premium loan complied with industry standards. The bank refused, agreeing only to affirm that the loan approval and documentation process was consistent with Deutsche’s own internal procedures. By the time the issue came to a head it was already March 2000, and the firm was finalizing opinion letters for the clients who had bought BLIPS. Although Watson and Rosenthal argued that the bank’s representation was insufficient, Richard Smith overruled them, and concluded that the representation, with some vague phrases thrown in, would have to suffice at this late a stage.⁶²

Over the next six months, Watson provided technical assistance to the marketing teams and oversaw the process of issuing KPMG opinion letters. After that task was completed, he was transferred to KPMG’s Amsterdam office where he was assigned the responsibility of coordinating the firm’s various European tax practices. When he returned a year later, he discovered that the firm was still actively involved in promoting tax shelters. He resigned shortly afterward.

A few years later, Mark Watson—along with Wiesner, Stein, Smith, DeLap, Eischeid, and Bickham—was indicted for tax fraud in connection with his participation in the BLIPS approval process. Rosenthal was not.

The Push to Market BLIPS

After the roll-out meeting in the Dallas/Fort Worth Airport on April 30, tax leaders had to restrain PFP partners who were champing at the bit to line up clients for BLIPS. When the transaction was finally green-lighted in early August, marketing efforts went into high gear.⁶³ DeLap had initally only approved a total of fifty BLIPS transactions, and each region was anxious to get its share. Moreover, for BLIPS to achieve its desired tax benefits in 1999, team members had to sell and complete deals in a very brief period. To create a semblance of legitimate investment, clients were required to stay in the first stage of the transaction a minimum of sixty days. The loan with Deutsche Bank consequently had to close by mid-October so that trading on the account could begin on October 22 and clients could elect to exit the transaction by December 21. This left the tiniest of wiggle room for Presidio to unwind the transaction by year’s end. By mid-September, KPMG had more than fifty active transactions, and Larson, Presidio’s principal, recommended that sales efforts cease since a huge backlog of unfinished deals had developed.⁶⁴

The crush of deals put logistical strains on Deutsche Bank, which suddenly had an enormous amount of paperwork to handle. Principals at Presidio, KPMG, and the bank discussed placing KPMG staff on site to help. Higher-ups at the bank nixed the idea when they realized that having KPMG employees in its offices would undermine the fiction that Deutsche Bank was acting independently and not as a facilitator of the transactions.⁶⁵ Although officials at the bank, including Deutsche Bank Americas’ CEO John Ross, had reviewed the transaction carefully and understood that it was assisting in the implementation of a tax shelter transaction,⁶⁶ the bank needed to maintain a plausible degree of deniability. The deals could not look rigged. Instead, it had to appear that each investor, working with Presidio, had shown up individually to borrow money from the bank to do a transaction. Presidio also had problems handling the substantial paperwork generated by the number of transactions. KPMG devised an arrangement under which a number of its employees would assist in closing the deals, billing Presidio three-quarters of a million dollars for their efforts.⁶⁷

A bigger problem with Deutsche Bank soon arose. By late summer, the bank had made nearly $3 billion in BLIPS loans and was quickly approaching the total capacity it had established internally for the transaction.⁶⁸ Many clients, however, still remained on the waiting list. Presidio had to find a new bank that could quickly get up to speed on the transaction in time to approve and close loans by mid-October—an impossible challenge given the hundreds of documents that would need to be reviewed. Larson and Pfaff at Presidio hit on the idea of looking for other banks that already had a relationship with KPMG and were represented by Rokoff’s team of lawyers at Shearman & Sterling which was familiar with the documentation and could finalize the loans quickly.⁶⁹

Presidio approached HVB, a Germany-based bank involved in earlier KPMG tax shelters, about financing the loans. Like Deutsche Bank, HVB stood to make 1 1/4 percent based on the interest charged on the loan, which corresponded to the desired tax loss. The gain represented an infinite return on investment since the funds stayed at the bank and no moneys had to be dedicated to the loan. Officials at HVB were aware that they were facilitating tax shelter deals but the easy money proved impossible to resist.⁷⁰ The bank agreed to a $400 million credit limit, which it later extended by an additional $500 million.⁷¹ But even with so much money available, the funds were running out. And the last, biggest BLIPS deal of the year was still to come.

In the summer of 1999, Allan Abrams and his siblings had decided to sell their company Arrow Fastener, a global manufacturer of staplers and other construction fasteners. Morris Abrams, their father, had started the company in 1929, manufacturing tools in his home shop by day and selling them in his Brooklyn neighborhood at night. By the early 1990s, Arrow Fastener had grown into one of the largest and best-known manufacturer of industrial fasteners, with manufacturing and sales operations around the world. To deal with its accounting needs, the company hired KPMG, which also provided tax return and planning services. Over the years, the Abrams developed a close working relationship with several partners at the firm. With the sale of their family business on the horizon, they turned to the firm to prepare the financial data that needed to be reported to the buyer.

As a result of the sale, the Abrams family was going to realize more than $400 million in capital gains, which would be subject to a 20 percent tax.⁷² During the spring and summer, the family’s tax advisors at KPMG met with family members several times to persuade them to engage in a BLIPS transaction. During the meetings, the partners at KPMG insisted that the transaction complied with IRS rules and regulations. They also told the Abrams that they would find a way to report the transaction on their tax returns so as to not raise any red flags. KPMG’s tax professionals assured the Abrams that even if the IRS detected the transaction they would only be liable for 50 percent of the tax due.

The Abrams were initially hesitant to engage in the transaction. By the fall, however, their advisors at KPMG had convinced them that there was no downside to going forward. The problem was that insufficient time remained before the end of the year for the Abrams to stay in the first stage of the transaction the minimum sixty-day “investment” period. Although the principals at Presidio were doubtful, Jeff Stein, KPMG’s vice chair of Tax Services, made clear that the transaction was of great importance to the firm and assured the participants that KPMG would issue an opinion for the transaction. After another bank was brought in to provide a “loan” in the hundreds of millions, the Abrams entered into and unwound the transaction in a forty-seven-day period, subsequently claiming losses totaling $330 million on their tax returns.⁷³

By the time Wiesner officially released BLIPS for marketing in early August 1999, the firm’s tax leaders were no longer standing firm on their original limit of fifty transactions, presumably because significantly more than fifty client engagements were already in the pipeline. By late September, however, tax leaders had decided that BLIPS marketing had to cease.⁷⁴ In 1999, KPMG engaged in sixty-six BLIPS transactions. The following year it engaged in another handful, which were treated as “grandfathered” in under the original sales initiative because they had already been marketed to clients who wanted to use the strategy to shelter gains anticipated in 2000. All told, BLIPS generated more than $53 million in fees and turned out to be one of the firm’s most successful shelters.⁷⁵

In early 2000, KPMG’s Innovative Strategies team began to develop a replacement strategy for BLIPS that they called BLIPS 2000. After Steve Rosenthal wrote a long email to David Brockway describing significant problems with the strategy, Brockway decided to veto the transaction. Inside the group, the search was on for another strategy marketed to high-wealth individuals that would replace BLIPS.

“SELL! SELL! SELL!”

In early 2000, Wiesner sent an email to the partners in the WNT practice exhorting them to “temporarily defer non-revenue producing activities” and concentrate on meeting WNT’s revenue goals for the fiscal year.Wiesner designated the “hot products” with “significant revenue potential” that the members of the office should focus on. “Thanks for help in this critically important matter,” he added. “As Jeff [Stein] said ‘We are dealing with ruthless execution—hand to hand combat—blocking and tackling.’ Whatever the mixed metaphor, let’s just do it.”⁷⁶

While the Innovative Strategies group had not hit on a blockbuster to succeed BLIPS, another group at KPMG, known as Stratecon, was busy developing a strategy with significant revenue potential. Stratecon, devoted to designing and selling tax shelters to businesses, was headed by Walter Duer, a partner in the firm’s WNT office. In early 2000, the group was promoting a strategy under the acronym SC2 to S corporations, business entities that, like partnerships, are taxed as pass-through entities.

SC2 was created in late 1999 by Larry Manth, Robert Huber, Douglas Duncan, and Andrew Atkin, many of whom had come over from Deloitte to KPMG earlier that year. At Deloitte, they were familiar with a similar strategy that had been disallowed by the IRS.⁷⁷ In SC2, an S corporation donated a substantial portion of its nonvoting stock to a tax-exempt entity and then redeemed the stock for a small amount after two or three years. During the period that the tax-exempt entity owned the stock, the S corporation’s income was allocated to the tax-exempt entity but no distributions of earnings were made. After the stock was redeemed, the S corporation could resume making distributions. Under the applicable tax code provisions, the strategy resulted in the benefit of a charitable deduction at the time the stock was donated. It also resulted in the deferral of income during the period that the tax-exempt entity held the stock, as well as the ultimate conversion of the S corporation’s earnings from ordinary income to capital gains.

SC2’s developers had already sold two SC2s before they obtained approval of the strategy by WNT.⁷⁸ As with BLIPS, participants in the review process raised serious concerns about whether SC2 would pass muster in court. Early on, Mark Watson noted that the strategy would be difficult to distinguish from the one the IRS had earlier shut down, and that clients might not be able to obtain the tax benefits promised. He also worried that the IRS would seek to impose penalties on KPMG for aiding and abetting the understatement of taxes with SC2. William Kelliher, another partner in WNT, also initially doubted that a more-likely-than-not opinion could be issued on the version he first saw, noting his concern that the IRS “could re-characterize the overall transaction . . . as a sham.”⁷⁹

Despite early qualms, the developers of SC2, with the assistance of Mark Springer, head of the TIC, continued to work toward obtaining approval of the strategy. The firm’s economic incentives were strong. As Galbreath, a member of the TIC, emphasized during the review process, SC2 was receiving “very high level (Stein/[Richard] Rosenthal) attention.”⁸⁰ Indeed, even before the strategy had been approved by the head of the Department for Professional Practice, Richard Rosenthal, the area managing partner for the Western Region, sent around an email to the SC2 team urging them to “SELL, SELL, SELL!!”⁸¹

The risk of detection by the IRS was not supposed to be a factor in deciding whether the firm would issue an opinion for the strategy; indeed, taking it into consideration was explicitly forbidden. Some reviewers nonetheless took into account the fact that the strategy would only be marketed to a small number of S corporations. Looking back a year later, Kelliher, who was involved in approving the final version, noted: “Going back to February, when SC2 first raised its head, my recollection is that SC2 was intended to be limited to a relatively small number of S Corps. That plan made sense because there was and is a strong risk of a successful IRS attack on SC2 if the IRS gets wind of it.”⁸²

By late March, Richard Bailine, a tax partner and expert in S corporation taxation, had also reviewed and approved a version of the strategy, and Larry DeLap had signed off. Before releasing the strategy, DeLap was emphatic that the engagement letter make clear the potential risks to clients. The risks that were to be highlighted to the client included that the IRS could reallocate the income allocated to the tax-exempt organization to the other shareholders; the IRS might disallow the charitable deduction contribution sought; the corporation could lose its subchapter S status and be subject to corporate tax; and the IRS might attempt to assess penalties against the shareholders of the business for understating their taxes.⁸³

As the sales toolkit for the strategy was being finalized, the SC2 “champions” began an intensive firm-wide effort to find suitable clients. They combed through internal, commercial, and public databases and contacted professionals within the firm, including audit colleagues, clients, and other referral sources in the hope of identifying prospects. The firm also used the Indiana telemarketing firm it had hired to cold call S corporations that qualified for the strategy.

To facilitate sales, SC2 champions distributed a “Sticking Points” memo containing advice about dealing with recalcitrant clients. If a buyer thought the strategy was too good to be true, the KPMG representative was to remind the buyer that the strategy had been vetted in an extensive review by firm specialists, including some who were former IRS employees. In addition, the buyer was to be told that a number of sophisticated clients had already bought the strategy, and at least one outside law firm was available to provide an opinion. If a client needed to think about it, the memo suggested three approaches: the “get even” approach, which involved contacting the client again right around the time a large estimated payment is due, when she or he is likely to be “extremely irritated”; the “Beanie Baby” approach—named for the popular stuffed toy whose limited availability caused sales crazes in the late 1990s—which involved informing the client the firm has established a cap on the product that is quickly filling up; and the “break-up” approach, which involved informing the client that because the cap has been reached, he or she should no longer consider purchasing the product.⁸⁴ (For obvious reasons, this last one was risky and could only be used in the limited number of cases when the client could be expected to fall for the ruse.) The memo also offered strategies to do end runs around “stubborn outside counsel” who had advised against purchasing SC2.⁸⁵ As part of the sales process, the firm also offered to arrange insurance on the strategy to cover losses in the event the IRS challenged it and prevailed. Almost half a dozen SC2 purchasers took the firm up on the offer and purchased insurance.⁸⁶

The firm’s efforts to market SC2 were so extensive that in December 2001, Kelliher, who earlier had approved it, complained that KPMG seemed “intent on marketing the SC2 strategy to virtually every S corporation with a pulse.”⁸⁷ With all this activity, he worried, the IRS would find out about the product. “Call me paranoid,” he added, but the “widespread marketing is likely to bring KPMG and SC2 unwelcome attention from the IRS. . . . I realize the fees are attractive, but does the firm’s tax leadership think that this is an appropriate strategy to mass market?”⁸⁸ To reassure Kelliher, DeLap explained that Richard Rosenthal and other tax leaders had already agreed, some nine months earlier, that SC2 would no longer be mass marketed.

The firm finally discontinued marketing SC2 when it received a subpoena from the IRS requesting documents relating to the transaction in the spring of 2002.⁸⁹ During the more than two years it was on the market, SC2 was sold to more than fifty subchapter S corporations, generating $26 million in revenue.⁹⁰ In 2004, the IRS prohibited SC2 as a “listed transaction.”⁹¹

SC2 was only one of several shelters developed by Stratecon members and sold to businesses. The most lucrative strategy that has come to light was marketed under the name Contingent Liability Acceleration Strategy or CLAS. CLAS was purchased by twenty-nine companies, including Delta Airlines, Whirlpool, Clear Channel, WorldCom, Tenet Healthcare, and the U.S. units of the global biopharmaceutical AstraZeneca. The brainchild of Carol Conjura, a former IRS official and WNT partner who specialized in tax accounting, CLAS permitted a company to accelerate the timing of deductions for settlements of lawsuits and other claims. Under the strategy, a corporation created a contingent liability trust with itself as beneficiary and then donated noncash assets, for example an intracompany note, whose value was equal to the amount of the anticipated settlements. A company could then take deductions for the anticipated settlement amounts years before the claims were settled and the amounts had to be paid.

Under the fee agreement for CLAS, KPMG charged 0.4 percent of the accelerated deduction, with a minimum fee of $500,000. Target clients were required to have a minimum of $150 million in pending claims against them. KPMG focused its sales efforts on companies that had implemented aggressive strategies in the past, targeting their CFO or vice president for tax. The firm’s sales pitch emphasized “the true beauty” of the strategy: “What is not required—cash!” CLAS, which earned the firm approximately $20 million, resulted in $1.7 billion in lost tax revenues to the Treasury, more than any other single shelter that has come to light.⁹² In 2003, the IRS listed CLAS as an abusive strategy.

While Stratecon was rolling out SC2, CLAS, and other corporate tax shelters, the Innovative Strategies group in the PFP was trying to replicate its success of previous years. In 2000, the group was pursuing two approaches that together would deliver, Eischeid hoped, $38 million in sales. One was a version of the Short Options Strategy, described in chapter 4, that the team planned to market without a more-likely-than-not opinion letter from the firm under a special finder’s arrangement. The second involved more customized solutions targeted at the upper end of the market. This approach would focus on deals exceeding $100 million in which KPMG’s fees were expected to surpass $1 million per deal.

The group was well on its way to implementing these plans when the IRS issued Notice 2000-44 in August 2000, listing Son of BOSS shelters as strategies it considered abusive and whose claimed tax benefits it would challenge in court. As the Innovative Strategies leaders acknowledged, the notice “specifically described both the retired BLIPS strategy and the then current SOS strategy.”⁹³ They concluded the time had come to lie low. “Business decisions” were made “to stop the implementation of ‘sold’ SOS transactions and to stay off the ‘loss generator’ business for an appropriate period of time.’”⁹⁴

But not for very long. By May 2001, the Innovative Strategies practice was anticipating generating $27 million for the coming fiscal year. Among the strategies it was promoting was one it called Partnership Option Portfolio Securities or POPS which, if approved, was expected to produce $12 million in fees.

By 2002, the tax leaders who had spearheaded KPMG’s focus on developing and promoting abusive tax shelters had risen to the highest positions at the firm. In April, Jeff Stein, who had succeeded Lanning as vice chair of Tax Services when the latter retired in 2000, was named KPMG’s deputy chair, the second in command of the firm. Smith, the partnership tax expert who had helped the firm arrive at a more-likely-than-not opinion on BLIPS, had been named area managing partner for the Western Region at the beginning of the year. He was then promoted to vice chair of Tax Services when Stein became deputy chair that spring. Richard Rosenthal, who had helped propel Stratecon’s corporate tax strategies, was named vice chair of Tax Operations directly under Stein in the summer of 2000. Two years later, he was promoted to CFO of KPMG.⁹⁵ Eischeid, who had headed the Innovative Strategies practice, was named partner in charge of the firm’s Personal Financial Planning practice.⁹⁶

At the turn of the century, tax shelters were the driving force behind the enormous profits produced by KPMG’s tax services. As we describe in chapter 12, the firm’s tax shelter activities were eventually stopped, but not before they generated more than $100 million in fees and billions of dollars in fake tax losses.⁹⁷