Stobie Creek Investments LLC v. United States

608 F.3d 1366 (2010)

STOBIE CREEK INVESTMENTS LLC, JFW Enterprises, Inc., Tax Matters and Notice Partners; and
Stobie Creek Investments LLC, by and through JFW Investments, LLC, Tax Matters and Notice Partner, Plaintiffs-Appellants,
v.
UNITED STATES, Defendant-Appellee.

No. 2008-5190.

United States Court of Appeals, Federal Circuit.

June 11, 2010.

Robert E. Kolek, Schiff Hardin LLP, of Chicago, IL, argued for plaintiffs-appellants. With him on the brief were Neil Lloyd and Thomas R. Wechter. Of counsel was Colleen Feeney Romero.

Judith A. Hagley, Attorney, Appellate Section, Tax Division, United States Department of Justice, of Washington, DC, argued for defendant-appellee. With her on the brief were John A. Dicicco, Acting Assistant Attorney General, Gilbert S. Rothenberg, Acting Deputy Assistant Attorney General, and Richard Farber, Attorney.

Before BRYSON, PROST, and MOORE, Circuit Judges.

PROST, Circuit Judge.

This tax refund suit concerns a series of transactions exemplifying the Son of BOSS^[1] tax shelter, marketed here as the Jenkens & Gilchrist ("J & G") strategy. The shelter took advantage of the fact that assets and contingent liabilities were treated differently for tax purposes when contributed to a partnership, thus enabling the taxpayer to generate an artificial loss. See 26 U.S.C. §§ 722, 733, 752, 754; see also IRS Notice No.2000-44, 2000-2 C.B. 255, 2000 WL 1138430. This artificial loss is then used to offset income from other transactions.

In this case, the taxpayers used the J & G strategy to inflate the basis of their stock in the Therma-Tru family business, thereby eliminating more than $200 million in capital gains (and avoiding $40 million in taxes) resulting from the sale of that stock. The Internal Revenue Service ("IRS") subsequently determined that the partnership used to implement the J & G strategy, Stobie Creek Investments LLC ("Stobie Creek"), was a sham. Based on this determination, the IRS disallowed the partnership's stated basis in the stock, increased the partnership's capital gain from the sale of that stock, and assessed additional taxes.

Stobie Creek, JFW Enterprises, Inc., and JFW Investments, LLC (collectively, "plaintiffs") then filed this refund suit in the United States Court of Federal Claims, contesting the Notices of Final Partnership Administrative Adjustment ("FPAAs") in which these determinations were made. Following a bench trial, the Court of Federal Claims upheld the FPAAs and associated penalties, concluding that the basis-inflating transactions were properly disregarded under the economic substance doctrine. Stobie Creek Invs., LLC v. United States, 82 Fed. Cl. 636, 701-02, 721 (2008). Plaintiffs now appeal the application of the economic substance doctrine, accuracy-related penalties, and an evidentiary ruling made during trial. We affirm.

This case turns on whether a series of transactions was properly disregarded under the economic substance doctrine, despite complying with the literal terms of the tax code. We conclude that the answer is yes. The trial court properly disregarded the transactions as lacking an objective economic reality; the taxpayers failed to show that the transactions were undertaken for any business purpose beyond obtaining a tax benefit. Accordingly, Stobie Creek, acting through Jeffrey Welles, was properly subject to accuracy-related penalties pursuant to 26 U.S.C. § 6662. Because it was not reasonable for Stobie Creek to rely on advice of professionals involved in promoting and implementing the tax shelter, the narrow reasonable-cause defense in 26 U.S.C. § 6664(c)(1) does not apply.

BACKGROUND

The taxpayers in this case are six members of the Welles family and the Welles trust (collectively the "Welleses"). This case arises out of a series of events beginning in 1999, when the Welles family agreed to sell a controlling interest in the family business, Therma-Tru Corporation ("Therma-Tru") to Kenner and Company ("Kenner"). Stobie Creek, 82 Fed.Cl. at 642. Patriarch and taxpayer David Welles Sr. ("David Welles") started what became Therma-Tru in 1962, when attorneys from the law firm of Shumaker, Loop & Kendrick, LLP ("SLK") helped him purchase a lumberyard. Therma-Tru subsequently grew into one of the leading manufacturers and sellers of residential entry doors. Id. at 641.

Over the years, the Welles family retained SLK for a variety of legal matters, including the Therma-Tru deal with Kenner. David Waterman ("Waterman") was the principal SLK attorney representing the Welleses in the transaction. The deal called for Kenner to infuse Therma-Tru with cash equal to a 50% equity position. At the same time, Therma-Tru shareholders would redeem 50% of their stock for cash. Because the deal involved Kenner paying cash for stock, the Therma-Tru shareholders would be taxed on their redemption of stock for cash. The Welles family planned to redeem 50% of their stock for approximately $215 million in cash. Because of their low basis in the stock, the redemption was expected to produce more than $200 million in capital gains.

Before the sale to Kenner was finalized, Jeffrey Welles asked Waterman whether there were any strategies for reducing the taxes the Welles family would otherwise owe on the planned sale. Id. at 643. Jeffery Welles is the son of Therma-Tru founder David Welles and the primary investment adviser to Therma-Tru and the Welles family. Prior to assuming this role, Jeffrey Welles worked in investment banking with Goldman Sachs and Lazard Freres. Id. at 641. In response to Jeffrey Welles's request, SLK contacted the law firm Jenkens & Gilchrist, P.C. ("J & G"). Waterman had previously referred other SLK clients with similar requests to J & G. In those cases, as here, Waterman and SLK then helped the clients implement a strategy^[2] developed and marketed by J & G.

To learn the details of the J & G strategy, the Welles family signed confidentiality agreements prepared by Donna Guerin, a partner at J & G. Id. at 643. In January 2000, the Welles family met in Vero Beach, Florida, to discuss various matters related to the pending Therma-Tru deal with Kenner (the "Vero Beach meeting"). During the two-day meeting, Waterman gave a presentation on the J & G strategy. Id. at 645. Among the materials Waterman distributed and discussed was an executive summary prepared by J & G, which gave a detailed overview of the J & G strategy. When asked whether he would engage in the J & G strategy if he were in the Welleses' situation, Waterman said he would. Id. at 646.

The goal of the J & G strategy was to reduce the capital gain resulting from the sale of assets. The strategy reduced a taxpayer's capital gain by increasing, or "stepping up," the basis in the asset the taxpayer wanted to sell. Because a partnership does not pay taxes, the resulting stepped-up basis passes through to the partners, thereby reducing the partner's capital gain and attendant capital gains tax when the asset is sold. Id. at 645.

To create a stepped-up basis in the asset, the J & G strategy called for contributions to a partnership, followed by distribution of the partnership's assets to the taxpayers. This goal was accomplished through a sequence of six steps, carried out in a particular order to ensure the taxpayers received the desired tax benefit: (1) investment in foreign currency options through a single-member LLC; (2) formation of a partnership with a third party or wholly-owned S corporation; (3) contribution of the foreign currency options to the partnership; (4) recognition of an economic gain or loss by the partnership when the options expired or were exercised; (5) termination and liquidation of the partnership through contribution of the taxpayer's partnership interest to an S corporation; (6) sale of the partnership's assets by the S corporation or taxpayer. Id.

Because of their importance to this appeal, steps 1, 3, and 5 warrant additional discussion here. Step 1 of the J & G strategy called for a particular type of investment in foreign currency: option spreads. To create an option spread, or "collar," the taxpayer sells a short option and purchases a long option on the same currency.

When the options are contributed to the partnership (here, Stobie Creek) during step 3, the taxpayer's basis in his partnership interest is increased by the cost of the long option, but not decreased by the short option obligation. Under the J & G strategy, the short option's contribution has no effect on the taxpayer's basis because it is not treated as a "liability" under 26 U.S.C. § 752 when calculating the taxpayer's basis in his partnership interest. When the partnership is liquidated during step 5, the tax basis in the partnership's assets is "stepped up" to match the partner's outside basis. This stepped-up basis allows the taxpayer to recognize less capital gain when the asset is sold during step 6.

The Welles family decided to pursue the J & G strategy. To obtain help implementing the strategy, the Welleses agreed to pay a fixed fee to J & G and SLK. J & G received a fee equal to 2% of the total gain to be sheltered, or $4,091,500. SLK's fee was 1% of the total gain to be sheltered, or $2,045,750. Id. at 651.

On March 3, 2000, the Stobie Creek partnership was formed. Single-member limited liability companies ("LLCs") were also formed for each family member, as well as the David Welles Qualified Annuity Trust ("Welles Trust"). To allow the LLCs to join Stobie Creek as partners, SLK attorneys prepared the corresponding paperwork, originally dated March 3, 2000. SLK asked Jeffrey Welles to review drafts of various documents, including the Stobie Creek company agreement and authorization for Stobie Creek to receive all cash proceeds from the Therma-Tru sale.

Three days later, Waterman sent a letter to each of the Welleses. The letter "confirm[ed] and correct[ed] certain information [SLK] provided to" the Welleses at the Vero Beach meeting. Waterman stated that J & G would be issuing a tax opinion for the Welleses similar to the one attached to the letter, which would opine that it was "more likely than not" that the transactions would be respected for federal income tax purposes. Id. at 647. Waterman also opined that recently issued federal regulations "did not appear to apply" to the Welles family, since the J & G strategy reduced only individual income tax liability, not corporate tax liability.

Contrary to his prior recommendation, Waterman's letter made a point of not recommending the J & G strategy: "I believe we [SLK] have been clear that we are not recommending that you pursue [J & G's] proposal. In fact, we have advised you that our knowledge of J & G's proposal was obtained in confidence under a confidentiality agreement ... [and] we are therefore unable to issue the opinion being offered by J & G." While noting that the letter tried to distance Waterman and SLK from their prior promotion of the J & G strategy, the trial court nonetheless found that the legal advice of Waterman or SLK was tainted by self-interest. Id. at 648. The trial court concluded that SLK was a broker for J & G's strategy. Id.

Two weeks later, on March 20, 2000, SLK sent an email to J & G. Per the email's request, J & G instructed Deutsche Bank to open accounts for each of the Welleses' single-member LLCs. SLK provided J & G and Deutsche Bank with a list showing the amount of capital gain each of the Welleses expected to realize upon redemption of their Therma-Tru stock. Id. at 648. Deutsche Bank used this list to determine the stated premiums for the options the Welleses would be purchasing as part of the J & G strategy. Deutsche Bank subsequently sent Jeffrey Welles sample confirmations for the type of digital options the Welleses were planning to acquire.

On March 28, 2000, $2,045,750 was wired from the Welles trust to Deutsche Bank. Jeffrey Welles authorized the transfer, which paid for the options the LLCs were to acquire through step 1 of the J & G strategy. The amount corresponded to the difference between the stated premiums on the long and short options. Three days later, on March 31, each of the LLCs entered into two pairs of option contracts (collectively, the Foreign Exchange Digital Options Transactions or "FXDOTS"). The first pair was an option collar on the value of the Swiss franc ("CHF") versus the United States dollar. The second pair was an option collar on the value of the United States dollar versus the euro. Both pairs of options were to close on April 17, 2000. Each option was digital, meaning the payoff was a fixed amount if the option expired "in the money" or nothing at all if the option expired "out of the money." A long (call) option expires "out of the money" if the asset's price is lower at the option's expiration than the price of exercising the option to buy the asset. A short (put) option expires "out of the money" if the asset's price is higher at the option's expiration than the price of exercising the option to sell the asset. Id. at 649 & n. 9. On April 3, 2000, the LLCs transferred their option contracts to the Stobie Creek partnership.

The option collar on the euro consisted of a purchased long option with a strike price of $0.9912 per euro and a sold short option with a strike price of $0.9914 per euro. The two-pip (two-thousands of a unit) spread between the two options, $0.9912-$0.9914 per euro, was referred to as that option collar's "sweet spot." Using the LLC belonging to Jeffrey Welles as an example, if the euro traded at less than $0.9912 per euro on the option's close date, the LLC (Jeffrey Welles) would lose $96,625. If the euro traded at more than $0.9914 per euro, he would gain $96,625. If the euro traded in the collar's sweet spot (above $0.9912 but below $0.9914), Jeffrey Welles would gain $19,228,357.

Analogously, the option collar on the Swiss franc consisted of a purchased long option with a strike price of CHF 1.7027 per dollar and a sold short option with a strike price of CHF 1.7029 per dollar. The sweet spot for the Swiss franc option collar was again a two-pip spread, CHF 1.7027— CHF 1.7029 per dollar.^[3]

As shown in the table below, there were nine possible outcomes for Stobie Creek's investments in the euro and Swiss franc digital options. The returns from these possible outcomes varied from a gain of $407 million (hitting both collars' sweet spots) to a loss of $2 million (all options finishing out of the money).

Value of

Options        Profit Loss Outcome for FXDOTS
$1  CHF      Gain                     Gain                          Gain
1.7027     $202,529,250             $407,104,250                $204,555,000
and
$1  CHF    No profit                  Gain                         Gain
1.7029          $0                   $204,555,000                 $2,045,750
€ 1  $0.9914
€ 1 > $0.9912
and
€ 1
As the table shows, six of the nine possible outcomes yield a positive return. However, as the experts explained at trial, the outcomes were not all equally likely to occur.
On April 17, 2000, all of the options expired out of the money. Stobie Creek consequently lost its entire investment of $2,045,750. On May 9, 2000, the Therma-Tru deal with Keener closed. Id. at 650.
After the close of the Therma-Tru deal (and long after the options expired), a series of emails and faxes passed among the Welleses, SLK, and J & G regarding the proper dates for documents related to the formation and transfer of partnership interests among the different corporate entities. SLK was responsible for preparing undated versions of some forms, which it then sent to J & G for review. At trial, the government introduced several copies of the assignment and joinder agreements transferring the Stobie Creek partnership interests from the LLCs to the S corporations. Some were signed, others were not; some were undated, others bore various (conflicting) dates. Upon learning that the Therma-Tru stock certificates had to be dated April 14, 2000, an attorney at SLK asked Guerin at J & G whether this posed a "timing issue," though he believed the April 14th date "pose[d] no threat." Id. The trial court found that the "threat" referred to the proper ordering of the transactions, which was necessary to achieve the strategy's beneficial tax treatment. The "threat" was ultimately addressed by replacing the March 24, 2000 documents assigning each LLC's interest in the Therma-Tru stock with documents dated April 14, 2000. Another series of emails between SLK and J & G concerned the date on which to "document" the shift of the partnership interests from the LLCs to the S corporations. SLK and J & G ultimately settled on April 30, 2000 because it enabled them to file a tax return for a short year. Other emails and faxes between SLK and J & G show that confusion (and re-dating) of documents continued through December 2000. Id. at 651.
In February 2001, Stobie Creek filed its federal income tax return for the tax period beginning on Stobie Creek's formation date, March 3, 2000, and ending on April 30, 2000, the date when the partnership interests were documented as passing from the LLCs to the S corporations (the "2000 tax year"). Id. at 657. In February 2002, Stobie Creek filed its return for the tax period beginning May 1, 2000 and ending December 31, 2000 (the "2000 stub year").
The IRS issued a FPAA for Stobie Creek's 2000 tax year in March 2005. The IRS issued a FPAA for Stobie Creek's 2000 stub year in February 2007. The FPAAs disregarded Stobie Creek for tax purposes as a sham and disallowed the partnership's stated basis in the Therma-Tru stock, finding it attributable to transactions entered into for the purpose of tax avoidance. As a result, the FPAAs increased Stobie Creek's capital gain income from the sale of the Therma-Tru stock. The FPAAs also imposed accuracy-related  penalties pursuant to 26 U.S.C. § 6662. Id. at 702.
Stobie Creek and the other plaintiffs filed this action in the Court of Federal Claims in July 2005. Id. at 657. The complaint sought readjustment of partnership items for the 2000 tax year and 2000 stub year, as well as a tax refund of the $4.2 million assessed in the FPAAs.
During a two week bench trial, the Court of Federal Claims heard testimony from several fact witnesses, including Waterman and Jeffrey Welles. Plaintiffs also presented three expert witnesses, Dr. Robert Kolb, Dr. Richard Levich, and Dr. Jeffrey Frankel. The government offered the testimony of one expert witness, Dr. David DeRosa. Id. at 639-40 n. 3. The trial court found that Stobie Creek's basis calculations complied with the literal requirements of the tax code. Id. at 670-71. In so finding, the trial court declined to apply Treasury Regulation § 1.752-6 retroactively.[4] The trial court nonetheless disregarded the transactions implementing the J & G strategy under the economic substance, step transaction, and end result doctrines. Id. at 671-702. In doing so, the court found that the plaintiffs failed to show the FXDOTs had a business purpose beyond creating a tax advantage. Id. at 696. This finding was largely based on the nature of the investments: the trial court found that for the FXDOTs to make any profit, two historically correlated currencies—the Swiss franc and the euro—had to decouple and move in opposite directions. If the currencies moved in the same direction relative to the dollar, the most favorable outcome the Welleses could hope for was breaking even, or zero profit. Id. at 690.
The trial court further found that Stobie Creek, acting through Jeffrey Welles, the manager of the tax matters partner for Stobie Creek, was liable for accuracy-related penalties under 26 U.S.C. § 6662. In so holding, the trial court rejected plaintiffs' argument that the reasonable cause defense in 26 U.S.C. § 6664(c)(1) applied. The trial court found that Jeffrey Welles's reliance on the professional advice of J & G and SLK was not reasonable or in good faith, given that both firms had a clear conflict of interest and Jeffrey Welles's own investment experience meant he would have recognized that the J & G strategy was "too good to be true." Id. at 707-21.
The plaintiffs now appeal. We have jurisdiction under 28 U.S.C. § 1295(a)(3).
ANALYSIS
This case is governed by certain provisions of the Tax Equity and Fiscal Responsibility Act of 1982 ("TEFRA"), 26 U.S.C. §§ 6221-6234. TEFRA created a unified partnership-level procedure for auditing and litigating "partnership items," thus addressing concerns about inconsistent treatment of the same partnership items across partners. See Schell v. United States, 589 F.3d 1378, 1381 (Fed.Cir.  2009). Penalties related to adjustments of partnership items are also determined during the partnership-level proceeding. 26 U.S.C. §§ 6221, 6226(f).
I. Economic Substance Doctrine
The primary question before this court is whether the transactions implementing the J & G strategy were properly disregarded under the economic substance doctrine. We conclude that they were.
How a transaction is characterized is a question of law we review de novo. Accordingly, we review the trial court's application of the economic substance doctrine without deference. Coltec Industries, Inc. v. U.S., 454 F.3d 1340, 1357 (Fed.Cir.2006). The trial court's underlying factual findings are reviewed for clear error. Jade Trading, LLC ex rel. Ervin v. United States, 598 F.3d 1372, 1376 (Fed.Cir.2010). Because deductions are a matter of legislative grace, the taxpayer has the burden of proving that a transaction had economic substance by a preponderance of evidence. Id.
The economic substance doctrine seeks to distinguish between structuring a real transaction in a particular way to obtain a tax benefit, which is legitimate, and creating a transaction to generate a tax benefit, which is illegitimate. Coltec, 454 F.3d at 1357; see also Klamath Strategic Invest. Fund ex. rel. St. Croix v. United States, 568 F.3d 537, 543-44 (5th Cir.2009). Under this doctrine, we disregard the tax consequences of transactions that comply with the literal terms of the tax code, but nonetheless lack "economic reality." Coltec, 454 F.3d at 1355-56; see also Frank Lyon Co. v. United States, 435 U.S. 561, 583-84, 98 S. Ct. 1291, 55 L. Ed. 2d 550 (1978); Klamath, 568 F.3d at 544; United Parcel Serv. of Am., Inc. v. Comm'r, 254 F.3d 1014, 1018 (11th Cir.2001); ACM P'ship v. Comm'r, 157 F.3d 231, 247 (3d Cir.1998); James v. Comm'r, 899 F.2d 905, 908-09 (10th Cir.1990). Such transactions include those that have no business purpose beyond reducing or avoiding taxes, regardless of whether the taxpayer's subjective motivation was tax avoidance. Coltec, 454 F.3d at 1355 (citing Higgins v. Smith, 308 U.S. 473, 476, 60 S. Ct. 355, 84 L. Ed. 406 (1940)); Ballagh v. United States, 331 F.2d 874, 877-78 (Ct.Cl.1964); see also Frank Lyon, 435 U.S. at 583-84, 98 S. Ct. 1291; Klamath, 568 F.3d at 544. We also disregard transactions shaped solely by tax-avoidance features. Frank Lyon, 435 U.S. at 583-84, 98 S. Ct. 1291; see also Coltec, 454 F.3d at 1355; Gregory v. Helvering, 293 U.S. 465, 469-70, 55 S. Ct. 266, 79 L. Ed. 596 (1935); Klamath, 568 F.3d at 544. Whether a transaction lacks "economic reality," has no bona fide "business purpose" or was shaped solely by tax-avoidance features is an objective inquiry, evaluated prospectively. Coltec, 454 F.3d at 1356. In other words, the transaction is evaluated based on the information available to a prudent investor at the time the taxpayer entered into the transaction, not what may (or may not) have happened later.
As they did at trial, the parties on appeal primarily focus on whether the FXDOTs should be disregarded under the economic substance doctrine. The plaintiffs argue that the FXDOTs should not be disregarded because they were entered into with a bona fide business purpose: namely, profit from investing in foreign currencies.
In a careful, well-reasoned opinion, the Court of Federal Claims rejected plaintiffs' argument. It then disregarded the  FXDOTs as lacking economic substance. In doing so, the trial court properly gave greater weight to the testimony of government expert Dr. DeRosa. The trial court did not clearly err in finding that Dr. DeRosa's testimony provided the more convincing and complete methodology for determining how a "reasonable investor" would judge the profit potential of the FXDOTs. This analysis examined the probability of each outcome, the expected rate of return, and the price of the options, which are all factors a prudent investor might consider when deciding whether to invest. Stobie Creek, 82 Fed.Cl. at 685-89. The factors Dr. DeRosa considered were thus highly relevant to evaluating the central question about the FXDOTs: whether a prudent investor would have had a reasonable expectation of earning a profit from the transaction. See Coltec, 454 F.3d at 1357.
Similarly, the trial court did not clearly err in finding that the selective and incomplete analysis of the plaintiffs' experts undermined their opinions that the FXDOTs had a "very substantial profit potential" (Dr. Kolb) or at least a modest profit potential (Dr. Levich). Stobie Creek, 82 Fed.Cl. at 677-80. For example, although plaintiffs' expert Dr. Kolb testified that the nine possible outcomes were not equally probable, he did not calculate the probabilities of the different outcomes—even though these probabilities were essential to evaluating whether a profit potential existed. As the trial court correctly observed, no reasonable person would make an investment, no matter what the stated return, if the probability of achieving that return were zero. Id. at 691. The analysis of plaintiffs' second expert, Dr. Levich, was similarly incomplete. Dr. Levich estimated that the probability of obtaining a 2-to-1 payoff on the FXDOTS was 9-21% on the dollar/euro options and 15-27% on the Swiss franc/dollar options; as for either or both options hitting the sweet spot (5 of the 9 outcomes), Dr. Levich simply stated it would be a "relatively rare occurrence." Id. at 680. The trial court properly accorded Dr. Levich's testimony little weight for two reasons. First, the trial court found Dr. Levich's opinion was based on estimates of high volatility in the exchange rate, which the market data did not support. Second, the trial court found Dr. Levich's opinion was undermined by the nature of the trades themselves: to return any profitable outcome, the FXDOTs required two historically correlated currencies to decouple and move in opposite directions. So long as the currencies moved in the same direction (as they had historically), the most the Welleses could hope for was breaking even, or zero profit. 82 Fed.Cl. at 690. In light of the record, neither of these findings is clearly erroneous.
Based on its evaluation of the expert testimony and supporting documentation, the trial court disregarded the FXDOTs under the economic substance doctrine. It concluded that the transactions did not reflect economic reality, nor were they motivated by a business purpose. Id. at 672-98, 701-02.
We reach the same conclusion. Measured either by their economic reality or their purported business purpose, the FXDOTs were properly disregarded under the economic substance doctrine.
A. Economic Reality
A transaction lacks "economic reality" when the tax result (gain or loss) is "purely fictional." See, e.g., Jade Trading, 598 F.3d at 1377. This inquiry often  focuses on whether there was a reasonable possibility of making a profit from the transaction. See, e.g., Coltec, 454 F.3d at 1356 (quoting Black & Decker Corp. v. United States, 436 F.3d 431, 441 (4th Cir. 2006)); see also Gilman v. Comm'r, 933 F.2d 143, 146-47 (2d Cir.1991) (asking whether a prudent investor would have found that a "realistic potential for economic profit" existed). Thus, in Jade Trading, we held that the taxpayers were not entitled to a basis of over $15 million in their Jade partnership interests (and an attendant tax loss of $14.9 million) because they had not contributed $15 million to the partnership, nor had they lost $14.9 million on exiting the partnership. 598 F.3d at 1377. In so holding, we explained that the "[option] transaction's fictional loss, inability to realize a profit, lack of investment character, meaningless inclusion in a partnership, and disproportionate tax advantage as compared to the amount invested and potential return, compel a conclusion that the spread transaction objectively lacked economic substance." Id.
In this case, the FXDOTs lacked economic reality for at least two reasons: (1) the tax result flowing from the FXDOTs was purely fictional; and (2) there was no reasonable possibility that the FXDOTs would return a profit. We discuss each of these reasons in turn.
First, as in Jade Trading, the $204,575,000 stepped-up basis in the Therma-Tru stock[5] was purely fictional: although the taxpayers only paid (and lost) about $2 million for the FXDOTs, they claimed a basis of over $200 million in their partnership interests, based on the contribution of those FXDOTs to Stobie Creek. See 598 F.3d at 1377. It is true that the taxpayers did purchase and contribute long options with a stated premium of $204,575,000 to Stobie Creek. However, they also sold and contributed short options with a stated premium of $202,529,250. Even though a literal application of the tax code at that time[6] may have permitted the taxpayers to treat these transactions separately, what matters under the economic substance doctrine is whether the tax treatment accords with economic reality.
In our analysis, the FXDOTs are properly treated as a single, unified transaction. Such treatment is more consistent with what was actually paid for the FXDOTs and how Deutsche Bank, the broker for the options, treated the transaction. Because Deutsche Bank is a third party and the one that stood to lose if the investments were not properly hedged, Deutsche Bank's treatment is particularly probative. The evidence shows that Deutsche Bank netted the premiums of the long and short options against each other, rather than require full payment of the option premiums and deposits against the margin, as is typically required when such options are entered into separately. The netting of the premiums and absence of a margin requirement are strong evidence Deutsche Bank did not view the long and short options as separate (or separable) transactions. Indeed, because Deutsche Bank treated the FXDOTs as one transaction, the taxpayers only paid the difference between the premiums, $2,045,750, rather  than the long option's stated premium of $204,575,000. Thus, when the FXDOTs expired out of the money, the taxpayers lost only $2,045,750, not $204,575,000.
Because the economic reality is that the long and short options were not separate, under the economic substance doctrine they similarly should not be separate for the purpose of calculating the taxpayers' basis in Stobie Creek. Accordingly, the taxpayers' claimed basis of $204,575,000 is properly disregarded as lacking economic reality; it does not reflect what the taxpayers paid Deutsche Bank for the FXDOTs ($2,045,757), or what they lost when the FXDOTs expired out of the money.
The FXDOTs also lack economic reality because there was no reasonable possibility the FXDOTs would return a profit, due to a combination of factors. These factors included the nature of the market (i.e., the high positive correlation between the movement of the euro and Swiss franc), the overpricing of the option premiums, and the structure of the investment (i.e., the necessity of the currencies decoupling, the effectively nonexistent "sweet spot," and the narrow range of the strike price). Cf. Klamath, 568 F.3d at 545 (noting that the taxpayers designed the transactions and investment strategy "so there was no reasonable possibility of a profit").
The evidence presented at trial shows that for the FXDOTs to have made any profit, the historically correlated euro and Swiss franc would have had to decouple and move in opposite directions. Stobie Creek, 82 Fed.Cl. at 690. Taking probabilities into account reveals that five of the nine possible outcomes would never occur, because the sweet spots would never be hit. Government expert Dr. DeRosa explained that the sweet spots could never be hit because of the strike price's narrow range (two pips) and Deutsche Bank's wide latitude in deciding whether the FXDOTs were "in the money." Id. at 686. In deciding whether a sweet spot had hit, Deutsche Bank could choose a quote's bid price, ask price, or something in between. Nor was Deutsche Bank limited to a specific bank's quote; it could solicit quotes from as many banks as it wanted and could choose among them. Further, the quotes Deutsche Bank received were three pips wide, and thus always greater than the two-pip spreads for the FXDOTs' sweet spots. The unattainable nature of the sweet spots is supported by the way Deutsche Bank internally hedged the FXDOTs. The evidence shows Deutsche Bank manually changed the short component's strike price for each option pair, eliminating the need for Deutsche Bank to internally hedge against the risk of hitting the sweet spot. Because there was effectively no sweet spot, the probability of any positive return was only 11.43% for the dollar/euro options and 19.95% for the Swiss franc/dollar options. Id. at 688.
The expected rates-of-return similarly show there was no reasonable possibility the FXDOTs would earn a profit. Expected rates of return are revealing, particularly if they account for costs and fees associated with implementing the transaction; a reasonable investor would consider such expenses when evaluating an investment's likely profitability. Dr. DeRosa testified that the expected rate of return was -77.14% for the dollar/euro options and-60.10% for the Swiss franc/dollar options. Id. at 688. With J & G's and SLK's fees for implementing the transaction included in the analysis, these rates were even more negative.
Finally, the price the Welleses paid for FXDOTs strongly suggests the transaction  lacked economic reality. The trial court sensibly reasoned that a prudent investor would not want to overpay for an investment, and would thus avoid a transaction in which the premiums were greater than the investment's expected value. By this rubric, the FXDOTs were precisely the type of transaction a reasonable investor would seek to avoid: Although the theoretical value of the euro/dollar long options was only about $23.3 million, the premiums valued the options at $102.3 million, more than four times that amount. The Swiss franc/dollar options were similarly overpriced; the stated premiums valued the options at more than three times the options' theoretical value. This disparity is far greater than the marginal variation Dr. DeRosa testified could occur, and there is no evidence of a market-related reason for the significant pricing difference.
B. Business Purpose
Asking whether a transaction has a bona fide business purpose is another way to differentiate between real transactions, structured in a particular way to obtain a tax benefit (legitimate), and transactions created to generate a tax benefit (illegitimate). Coltec, 454 F.3d at 1357.
We conclude that the FXDOTs fall in the category of "illegitimate" transactions identified in Coltec. See id. The evidence shows that the FXDOTs were part of a prepackaged strategy marketed to shelter taxable gain. Stobie Creek, 82 Fed.Cl. at 693-94. The Welleses sought out the J & G strategy not because they wanted to profit from investments in foreign currency (a legitimate purpose), but only because they wanted to lower their tax liability on the Therma-Tru deal, an unrelated transaction (an illegitimate purpose). Here the tax-avoidance motive preceded the "investment" strategy and any evaluation of profit potential; the FXDOTs (and the J & G strategy more generally) were simply a means to the desired end of creating a tax benefit.
It is true that the Welleses implemented the J & G strategy for the purpose of minimizing the tax consequences of the Therma-Tru deal, a real transaction with economic substance. That connection in itself, however, does not legitimize the FXDOTs or the J & G strategy. Although the Welleses were unquestionably free to structure the Therma-Tru deal to minimize their tax liability, the J & G strategy was not a way of structuring the Therma-Tru deal. Cf. Coltec, 454 F.3d at 1357. Rather, the J & G strategy was a separate, independent set of transactions that had no purpose besides creating a tax benefit. Id.; cf. Ballagh, 331 F.2d at 878.
The Welleses' claim of a profit motive behind the FXDOTs is belied by ample evidence that the tax advantages could not have been achieved had the transaction taken another form and that, absent the tax advantages, the transaction never would have occurred. See Frank Lyon, 435 U.S. at 583 n. 18, 98 S. Ct. 1291; Gregory, 293 U.S. at 469-70, 55 S. Ct. 266. This court's decision in Jade Trading held that the "meaningless inclusion in the partnership" of options was evidence the transaction lacked economic substance. Cf. 598 F.3d at 1377. In this case, the Welleses similarly used unnecessary corporate entities to invest in the FXDOTs. Although the LLCs, S corporations, and partnership (Stobie Creek) were not necessary to the transaction and did not enhance its potential profitability, the taxpayers nevertheless went to great lengths to create these entities and transfer the FXDOTs among them. See id.; cf. Gregory, 293 U.S. at 469-70, 55 S. Ct. 266.
 Taxpayers' focus on generating tax benefits, rather than pursuing a legitimate business purpose, is also evidenced by the backdating of different transactions, including the FXDOTs, to conform to the J & G strategy. Stobie Creek, 82 Fed.Cl. at 695. The trial court found this backdating did nothing to enhance the transactions' investment potential, but was absolutely critical to achieving the desired basis enhancement and associated tax benefits. Id. at 695-96. Indeed, the FXDOTs could have been, and in fact were, carried out in a different order than J & G's prescribed strategy. We agree with the trial court that the redating of different transactions reveals an emphasis on generating tax benefits; conforming to the J & G strategy mattered only if the purpose was tax avoidance, not economic profit from the FXDOTs. Cf. Neonatology Assocs. v. Comm'r, 299 F.3d 221, 230 n. 12 (3d Cir. 2002); Rogers v. United States, 281 F.3d 1108, 1114-15 (10th Cir.2002).
Finally, the fee structure undermines plaintiffs' contention that the J & G strategy had a business purpose (besides generating tax benefits, which does not count). See Coltec, 454 F.3d at 1358-59. The fees paid to SLK, J & G, and Deutsche Bank were all computed based on the amount of gain to be sheltered by the J & G strategy, without reference to typical economic considerations, such as the amount of risk on the investment. Stobie Creek, 82 Fed.Cl. at 693-94. Plaintiffs have offered no explanation for why the taxpayers paid such high fees, particularly to Deutsche Bank. Deutsche Bank's fee exceeded not only the normal fees for foreign currency options, but the value of the FXDOTs themselves.
Thus, because the FXDOTs lacked economic reality and had no business purpose, they were properly disregarded under the economic substance doctrine.
II. Penalties
A. Jurisdiction
A threshold question is whether we have jurisdiction to review the accuracy-related penalties imposed under 26 U.S.C. § 6662. See Special Devices, Inc. v. OEA, Inc., 269 F.3d 1340, 1342-43 (Fed. Cir.2001). Because plaintiffs challenge the "applicability of a[] penalty ... which relates to an adjustment to a partnership item" and do not raise a partner-level defense, we conclude that the answer is yes. 26 U.S.C. §§ 6226(f), 6230.
The penalties challenged on appeal relate to Stobie Creek's misstatement of its inside basis in Therma-Tru stock, as well as to adjustments of its basis in that stock pursuant to 26 U.S.C. § 754. The partnership's basis in contributed property is a partnership item. Treas. Reg. § 301.6231(a)(3)-1(a)(vi), (c)(2); see also Am. Boat Co. v. United States, 583 F.3d 471 (7th Cir.2009). Adjustments made pursuant to a § 754 election are also partnership items. Treas. Reg. § 301.6231(a)(3)-1(a)(3). Accordingly, the penalties "relate[] to an adjustment to a partnership item," i.e., Stobie Creek's basis in the Therma-Tru stock. 26 U.S.C. § 6226(f).
Further, the defense at issue on appeal is a partnership-level defense, not a partner-level defense. In a partnership-level proceeding such as this, we lack jurisdiction to consider partner-level defenses. See 26 U.S.C. §§ 7422(h), 6230(c); Schell, 589 F.3d at 1382; Am. Boat, 583 F.3d at 478-79. Stobie Creek argues it had reasonable cause for stepping up the basis in the Therma-Tru stock. A reasonable-cause defense under 26 U.S.C. § 6664(c) may be a partner- or partnership-level defense, depending on who is asserting it.  See Temp. Treas. Reg. § 301.6221-1T(d); Klamath, 568 F.3d at 548; Whitehouse Hotel Ltd. v. Comm'r, 131 T.C. 112, 173 (2008). We have jurisdiction because here the partnership (Stobie Creek) is claiming it had reasonable cause based on the actions of its managing partner, Jeffrey Welles.[7]See Am. Boat, 583 F.3d at 479-80.
B. Reasonable-Cause Defense
On the merits, the only question is whether Stobie Creek, acting through Jeffrey Welles, had reasonable cause for its tax position. Stobie Creek argues reasonable cause and good faith are demonstrated by Jeffrey Welles's reliance on advice from SLK and J & G. Stobie Creek, 82 Fed.Cl. at 718. The answer turns on whether such reliance was reasonable under the circumstances. We conclude that it was not.
Mandatory, accuracy-related penalties apply to certain underpayments of tax that meet the statutory requirements. 26 U.S.C. § 6662(a), (h). Section 6664(c)(1) provides a narrow defense to § 6662 penalties if the taxpayer proves it had (1) reasonable cause for the underpayment and (2) acted in good faith. See also Treas. Reg. § 1.6664-4(c)(1). The taxpayer bears the burden of showing this exception applies. See Conway v. United States, 326 F.3d 1268, 1278 (Fed.Cir.2003). Whether a taxpayer had reasonable cause is a question of fact decided on a case-by-case basis. Id.; Treas. Reg. § 1.6664-4(b)(1). We review this determination and the findings underlying it for clear error. See Am. Boat, 583 F.3d at 483; Barrett v. United States, 561 F.3d 1140, 1148 (10th Cir.2009). In doing so, we take into account all the pertinent facts and circumstances. Treas. Reg. § 1.6664-4(b)(1). The most important of these factors is "the extent of the taxpayer's effort to assess the taxpayer's proper tax liability," judged in light of the taxpayer's "experience, knowledge, and education." Treas. Reg. § 1.6664-4(b)(1).
One way to show reasonable cause is to show reasonable reliance on the advice of a competent and independent professional adviser. Treas. Reg. § 1.6664-4(b)(1); United States v. Boyle, 469 U.S. 241, 251, 105 S. Ct. 687, 83 L. Ed. 2d 622 (1985). This advice must meet several requirements. First, the taxpayer must show that the advice was based on "all pertinent facts and circumstances and the law as it relates to those facts and circumstances." Treas. Reg. § 1.6664-4(c)(1)(i). Second, the advice relied upon must not be based on any "unreasonable factual or legal assumptions," and must not "unreasonably rely on the representations, statements, findings, or agreements of the taxpayer or any other person." Id. § 1.6664-4(c)(1)(ii). Third, the taxpayer's reliance on the advice must itself be objectively reasonable. The reasonableness of any reliance turns on the quality and objectivity of the advice. See Klamath, 568 F.3d at 548; Chamberlain v. Comm'r, 66 F.3d 729 (5th Cir.1995); Swayze v. United States, 785 F.2d 715, 719 (9th Cir.1986). Reliance is not reasonable, for example, if the adviser has an inherent conflict of interest about which the taxpayer knew or should have known. Treas. Reg. § 1.6664-4(c); Am. Boat, 583 F.3d at  481-82; Hansen v. Comm'r, 471 F.3d 1021, 1031-32 (9th Cir.2006); Neonatology, 299 F.3d at 234; Pasternak v. Comm'r, 990 F.2d 893, 903 (6th Cir.1993). Nor is it reasonable if the taxpayer knew or should have known that the transaction was "too good to be true," based on all the circumstances, including the taxpayer's education, sophistication, business experience, and purposes for entering into the transaction. Treas. Reg. § 1.6664-4(c); Hansen, 471 F.3d at 1032.
The trial court concluded that Stobie Creek was not entitled to the reasonable-cause defense because it was not reasonable for Jeffrey Welles to rely on the advice of SLK or J & G. Stobie Creek, 82 Fed.Cl. at 720-21. The court found that both firms had an inherent conflict of interest, which Jeffrey Welles knew or should have known about. This conflict of interest arose from the role of SLK and J & G in promoting, implementing, and receiving fees from the J & G strategy. Accordingly, the trial court found that the firms could hardly qualify as independent professionals, since neither was disinterested in the outcome of the strategy they were evaluating.
We agree that the reasonable-cause defense does not apply to Stobie Creek. The trial court did not clearly err in finding it objectively unreasonable for Jeffrey Welles to rely on the advice of J & G and SLK because J & G was a promoter of the shelter and SLK was an agent of the promoter, making them anything but independent. Cf. Pasternak, 990 F.2d at 903. Advice hardly qualifies as disinterested or objective if it comes from parties who actively promote or implement the transactions in question. See, e.g., id.; Mortensen v. Comm'r, 440 F.3d 375 (6th Cir.2006); Van Scoten v. Comm'r, 439 F.3d 1243, 1253 (10th Cir.2006); Goldman v. Comm'r, 39 F.3d 402, 408 (2d Cir.1994).
Further, the trial court did not clearly err in finding that Jeffrey Welles knew or should have known about this conflict of interest. At trial, the government presented extensive circumstantial evidence that Jeffrey Welles authorized, reviewed, or at minimum received updates on the strategy's progress, and thus knew about both firms' roles.
For example, J & G's role as a promoter of the strategy was evident from the initial confidentiality agreement, which stated that the "proprietary" strategy had been "developed by J & G." This role was again apparent in J & G's fee agreement, which tied the firm's compensation to the gain sheltered by the strategy. Jeffrey Welles received, reviewed, and signed these documents. J & G's role was similarly evidenced by its efforts to implement the shelter. For example, J & G helped set up the FXDOTs, draft the formation and transfer agreements for Stobie Creek, and assure that the transactions adhered to the strategy's chronology.
Similarly, the evidence supports the trial court's conclusion that Jeffrey Welles knew or should have known that SLK was an agent of J & G, and thus could not reasonably rely on SLK's advice. SLK's agency relationship was apparent from the beginning. Waterman referred the Welleses to J & G, presented the strategy at the Vero Beach meeting, and recommended the strategy. As was true for J & G, SLK's fee agreement made clear that SLK had a financial stake in the outcome, again tying compensation to the sheltered gain. SLK also helped implement the strategy by drafting and backdating documents for the different corporate entities. Indeed, SLK openly acknowledged its role  in a letter to the Welleses. The letter stated that the lower taxable gain that would be reported on Stobie Creek's return was "produced by the tax strategy that was developed by [J & G] and implemented with our [SLK's] help earlier this year." The trial court found that Jeffrey Welles received this letter. Based on that and other evidence presented at trial, it was reasonable for the trial court to infer that Jeffrey Welles (and thus Stobie Creek) knew or should have known about the conflicts of interest for J & G and SLK. It was not objectively reasonable for Jeffrey Welles to ignore evidence of these conflicts and continue to rely on the advice, regardless of the Welleses' longstanding relationship with SLK or the reputations of both firms.
Even if Jeffrey Welles had not known about the conflicts of interest, his reliance on the advice of SLK and J & G was still unreasonable. Based on Jeffrey Welles's education and experience, as well as the reason the Welleses pursued the J & G strategy, the trial court found that Jeffrey Welles should have known that the J & G strategy was "too good to be true." Cf. Neonatology, 299 F.3d at 234. This determination is not clearly erroneous. Jeffrey Welles was a highly educated professional with extensive experience in finance, having worked as an investment banker and as the manager of his family's complex finances. Stobie Creek, 82 Fed.Cl. at 715. In that managerial role, he had helped implement a number of sophisticated tax-planning strategies, giving him sufficient knowledge and experience to know when a tax-planning strategy was likely "too good to be true." Jeffrey Welles knew that the J & G strategy was marketed as a "Basis Enhancing Derivatives Structure" and that the purpose of the strategy was to boost the basis in capital assets, "generating a reduced gain for tax purposes." Moreover, Jeffrey Welles sought out and selected the J & G strategy because of a desire to avoid taxes that would otherwise be owed on the Therma-Tru deal, not because he wanted to structure the deal itself to minimize taxes.
Accordingly, Stobie Creek had no reasonable-cause defense for its tax position.
III. Evidentiary Ruling
Plaintiffs argue that they are entitled to a new trial because the trial court erroneously excluded the testimony of their expert, Stuart Smith. We disagree.
A trial court's evidentiary rulings are reviewed for abuse of discretion. Pac. Gas & Elec. Co. v. United States, 536 F.3d 1282, 1285 (Fed.Cir.2008). In this case, Smith sought to testify about the tax laws "as they existed in 2000" and whether the J & G tax opinion letter complied with the standards set out in Treasury Circular 230. The trial court excluded Smith's expert report and testimony under Federal Rule of Evidence 702, concluding that Smith's opinion would not "assist" the court because the opinion concerned a question of law, not fact. Stobie Creek Invs., LLC v. United States, 81 Fed. Cl. 358, 359-61 (2008).
Plaintiffs are not entitled to a new trial because the trial court properly excluded Smith's expert testimony. Under Rule 702, expert testimony must "assist the trier of fact to understand the evidence or to determine a fact in issue." Fed.R.Evid. 702 (emphases added). Because proper interpretation of the tax laws and Treasury Circular 230 are issues of law, it was not an abuse of discretion to exclude expert testimony related to those questions. See Mola Dev. Corp. v. United States, 516  F.3d 1370, 1379 n. 6 (Fed.Cir.2008). To the extent Smith sought to testify about whether the J & G tax opinion letter met the standards of Treasury Circular 230, that opinion similarly would not have "assist[ed]" the trial court because Smith's proposed testimony consisted of a lengthy legal analysis of past precedent and assumed key factual representations underlying the J & G opinion were accurate, when in actuality they were false (and known to be so by the Welleses). Stobie Creek, 81 Fed.Cl. at 362; see Stobie Creek, 82 Fed.Cl. at 706-07, 720-21. Excluding Smith's report and testimony was thus within the trial court's discretion.
CONCLUSION
We affirm the application of the economic substance doctrine to the J & G strategy and accuracy-related penalties imposed on Stobie Creek. Stobie Creek was not entitled to a reasonable-cause defense under § 6664(c)(1) and the testimony of plaintiff's expert Smith was properly excluded under Federal Rule of Evidence 702.
AFFIRMED
NOTES
[1]  "BOSS" is an acronym for "Bond and Option Sales Strategy." Son of BOSS is a variation on the BOSS tax shelter. Kornman & Assocs. v. United States, 527 F.3d 443, 446 n. 2 (5th Cir.2008).
[2]  J & G called its strategy the "Basis Enhancing Derivatives Structure," or BEDS. 82 Fed.Cl. at 643 n. 6. We refer to it simply as the "J & G strategy."
[3]  Again taking the LLC belonging to Jeffrey Welles as an example, if the Swiss franc traded at less than CHF 1.7027 per dollar, the LLC (Jeffrey Welles) would lose $96,625. If the Swiss franc traded at more than CHF 1.7029 per dollar, he would gain $96,625. If the euro traded in the collar's sweet spot (above CHF 1.7027 but below CHF 1.7029), Jeffrey Welles would gain $19,228,357.
[4]  Had the trial court applied Treasury Regulation § 1.752-6 retroactively, plaintiff's refund action would fail under the literal application of the tax code and treasury regulations because the short options would constitute liabilities for the purpose of 26 U.S.C. § 752, reducing the LLCs' basis in their partnership interests.
The government has not appealed the trial court's holding on the retroactivity of Treasury Regulation § 1.752-6. Accordingly, we do not decide whether Treasury Regulation § 1.752-6 applies retroactively because, even if it does not, the J & G strategy was properly disregarded under the economic substance doctrine.
[5]  When the partnership interests were transferred from the LLCs to the S corporations, the basis in the Therma-Tru stock was stepped up to match the taxpayers' outside basis in the partnership, $204,575,000.
[6]  The applicable regulations have since been amended. See Treas. Reg. §§ 1.752-6, 1.752-7 (2009).
[7]  Jeffrey Welles was the manager of North Channel, the tax-matters partner of Stobie Creek.