Partnerships Argue Tax Court Erred in Economic Substance Decision

 

ENDEAVOR PARTNERS FUND, LLC, DELTA CURRENCY
TRADING, LLC, TAX MATTERS PARTNER,
Petitioners-Appellants,
v.
COMMISSIONER OF INTERNAL REVENUE SERVICE,
Respondent-Appellee.

NOT YET SCHEDULED FOR ORAL ARGUMENT

United States Court of Appeals
for the District of Columbia Circuit

On Appeal from the Decisions of the U.S. Tax Court,
in Nos. USTC-8721-12, USTC-8846-12, USTC-8698-12, USTC-8710-12

BRIEF FOR PETITIONERS-APPELLANTS

ADRIENNE B. KOCH
DAVID L. KATSKY
KATSKY KORINS LLP
605 Third Avenue
New York, New York 10158
(212) 953-6000
dkatsky@katskykorins.com
akoch@katskykorins.com

Counsel for Petitioners-Appellants

CERTIFICATE AS TO PARTIES, RULINGS AND RELATED CASES

Appellants, pursuant to the orders of this Court, dated October 3, 2018, October 4, 2018 and October 10, 2018, and the further order of the Court, dated October 10, 2018, consolidating these four appeals, hereby submit the following preliminary certificate:

1. Parties and Amici

The following parties appeared before the US Tax Court: Appellants Endeavor Partners Fund, LLC; Delta Currency Trading, LLC; Alligator Partners Fund, LLC; Cabrini Partners Fund, LLC; and Satellite Partners Fund, LLC; and Appellee Commissioner of Internal Revenue.

2. Ruling Under Review

The Ruling under review in this Court is the Memorandum Findings Of Fact And Opinion filed on June 28, 2018 in the US Tax Court (Lauber, J.). (JA1886-1951). The appeal is from the four Orders and Decisions of that court, each dated July 11, 2018. (JA1952; JA1953; JA1954; JA1955).

3. Related Cases

The case on review has not previously been before this Court or any other court. There are no related cases.¹

4. Disclosure Statement Respecting Parent Entities

The following are the entities covered by Circuit Rule 26.1:

The represented entities are the taxpayers whose tax losses are at issue in these appeals. Each appellant was a business entity which traded foreign currency derivatives. Appellant Delta Currency Trading, LLC was managing member and tax matter partner for each other appellant.

Respectfully submitted,

Adrienne B. Koch, Esq.
Email: akoch@katskykorins.com
David L. Katsky, Esq.
Email: dkatsky@katskykorins.com
Katsky Korins LLP
605 Third Avenue
New York, New York 10158
(212) 953-6000
Counsel for Petitioners-Appellants

---

TABLE OF CONTENTS

CERTIFICATE AS TO PARTIES RULINGS AND RELATED CASES

TABLE OF AUTHORITIES

GLOSSARY

JURISDICTIONAL STATEMENT

ISSUE PRESENTED

STATUTES AND REGULATIONS

STATEMENT OF THE CASE

SUMMARY OF ARGUMENT

STANDARD OF REVIEW

ARGUMENT

CONCLUSION

TABLE OF AUTHORITIES

Abbot v. Perez, __ U.S. __, 138 S. Ct. 2305 (2018)

Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986)

Armstrong v. Geithner, 608 F.3d 854 (D.C. Cir. 2010)

ASA Investerings Partnership v. Comm'r, 201 F.3d 505 (D.C. Cir. 2000)

Campbell v. Comm'r, 868 F.2d 833 (6th Cir. 1989)

Chai v. Comm'r, 851 F.3d 190 (2d Cir. 2017)

Chhetry v. U.S. Dept. of Justice, 490 F.3d 196 (2d Cir. 2007)

Close v. Comm'r, 107 T.C.M. (CCH) 1124, 2014 WL 521039 (2014)

Dang v. Comm'r, T.C. Memo 2002-117, 2002 WL 977368 (2002)

Daniels v. Hadley Mem'l Hosp., 566 F.2d 749 (D.C. Cir. 1977)

DKD Enters. v. Comm'r, 685 F.3d 730 (8th Cir. 2012)

Friedman v. Comm'r, 216 F.3d 537 (6th Cir. 2000)

Friedman v. Comm'r, 869 F.2d 785 (4th Cir. 1989)

Gaw v. Comm'r, T.C. Memo. 1995-531, 1995 WL 664592 (1994) aff'd, 111 F.3d 962 (D.C. Cir. 1997)

Horn v. Comm'r, 968 F.2d 1229 (D.C. Cir. 1992)

Huthnance v. District of Columbia, 722 F.3d 371 (D.C. Cir. 2013)

Jombo v. Comm'r, 398 F.3d 661 (D.C. Cir. 2005)

Latif v. Obama, 677 F.3d 1175 (D.C. Cir. 2011)

Moore v. Chesapeake & O. Ry. Co., 340 U.S. 573 (1951)

Overnite Transp. Co. v. N.L.R.B., 140 F.3d 259 (D.C. Cir. 1998)

Paige v. U.S. Drug Enf't Admin., 818 F. Supp. 2d 4 (D.D.C. 2010), aff'd, 665 F.3d 1355 (D.C. Cir. 2012)

Portland Golf Club v. Comm'r, 497 U.S. 154 (1990)

Shea v. Comm'r, 112 T.C. 183 (1999)

Smith v. Reitman, 389 F.2d 303 (D.C. Cir. 1967)

Southgate Master Fund, L.L.C. v. U.S., 659 F.3d 466 (5th Cir. 2011)

Stephens v. U.S. Dept. of Labor, 571 F. Supp. 2d 186 (D.D.C. 2008), aff'd, 384 Fed. Appx. 5 (D.C. Cir. 2010)

United Parcel Service of Am., Inc. v. Comm'r, 254 F.3d 1014 (11th Cir. 2001)

United States v. Hite, 769 F.3d 1154 (D.C. Cir. 2014)

Vaquería Tres Monjitas, Inc. v. Laboy, No. 04-1840 (DRD), 2007 WL 7733665 (D.P.R. July 13, 2007), aff'd sub nom. Vaquería Tres Monjitas, Inc. v. Irizarry, 587 F.3d 464 (1st Cir. 2009)

Wichita Terminal Elev. Co. v. Comm'r, 6 T.C. 1158 (1946), aff'd, 162 F.2d 513 (10th Cir. 1947)

Statutes & Other Authorities:

Black's Law Dictionary (10^th ed. 2014)

Fed. R. Civ. P. 52(a)

Fed. R. Evid. 201(b)

T. Weithers, Foreign Exchange: A Practical Guide to the FX Markets (John Wiley & Sons, 2006)

IRC § 706(b)(1)(c)

IRC § 6112

IRC § 6226(f)

IRC § 6231(a)(3)

IRC § 7482(a)(1)

IRC § 7482(b)(1)

IRC § 7701(o)

Pub. L. No. 111-152 § 1409

Tax Court Rule 142(a)(1)

GLOSSARY

FPAA	Final Partnership Administrative Adjustment
IRC	Internal Revenue Code
IRS	Internal Revenue Service
JA	Joint Appendix

---

JURISDICTIONAL STATEMENT

The Tax Court had jurisdiction over these consolidated proceedings pursuant to IRC § 6226(f)² because they involved a timely challenge to FPAAs setting forth the IRS's proposed adjustments of “partnership items” within the meaning of IRC § 6231(a)(3) with respect to Endeavor Partners Fund, LLC (“Endeavor”), Cabrini Partners Fund, LLC (“Cabrini”), Alligator Partners Fund, LLC (“Alligator”), and Satellite Partners Fund, LLC (“Satellite”; collectively, the “Partnerships”). This Court has jurisdiction under IRC § 7482(a)(1) to consider the timely appeals (filed October 1, 2018) of the Partnerships and their tax matters partner, Delta Currency Trading, LLC (“Delta”), from the Tax Court's July 11, 2018 decisions finally disposing of their petitions following a six-day bench trial. Because at the time those petitions were filed each of the Partnerships had ceased operations and accordingly had no principal place of business, venue for this appeal lies in this Circuit under IRC § 7482(b)(1).

ISSUE PRESENTED

Whether the Tax Court erred in concluding that certain foreign currency options trades lacked economic substance (such that the losses they generated should be disallowed), where that conclusion — which under this Court's governing precedent required a determination both that the trades lacked profit potential and that they were not otherwise undertaken for a business purpose — was based on:

STATUTES AND REGULATIONS

Pertinent statutes and regulations are reproduced in the Addendum to this brief.

STATEMENT OF THE CASE

Delta and its affiliates (collectively the “Delta Group”) were involved with certain so-called “tax shelters” that we recognize this Court has viewed with disfavor. At issue here, however, is not whether any transaction was a “tax shelter” or whether any “tax shelter” was valid. Rather, the issues on this appeal are principled ones of evidence and fundamental procedural fairness. We respectfully urge the Court to focus on how the Tax Court applied the rules of evidence rather than on the subject matter to which it applied them.

Although the proceedings below addressed highly technical foreign currency trades and complex tax questions, the evidentiary issues before this Court are straightforward. The Tax Court disallowed the losses at issue here based entirely on a finding that the parties to the transactions that generated them agreed in advance that, rather than settle the currency trades pursuant to their written terms (which would have entailed a potential for profit or loss), they would settle them at rates that would guarantee that there would be no profit or loss. In other words, the Tax Court found that the trades were deliberately rigged to strip away any profit potential. We emphasize that the Tax Court's finding that the parties made this separate (unwritten) “agreement” to rig the trades was the only basis on which it found that the trades lacked economic substance. The defining issue on this appeal is whether that finding is supported by competent evidence.

As detailed below, the IRS only first advanced its theory that the trades were rigged in this manner three years after it issued the FPAAs, and two and a half months after the close of discovery. It did not reveal until the eve of trial that it had five witnesses that it claimed would substantiate that the trades were rigged. Delta moved to preclude those witnesses as untimely disclosed; the Tax Court denied the motion, but acknowledged that the IRS's documents alone would be insufficient to prove the existence of this alleged agreement. It reiterated this again and again during trial. But after the IRS heard Delta's case in chief, it withdrew every single one of those witnesses. None of them testified.

The Tax Court nevertheless found that the IRS had established the alleged agreement. It did so through a combination of factual findings that are not supported by any evidence at all, assertions that there was “no evidence” with respect to other matters on which there was in fact voluminous evidence, and an adverse inference against Delta for failing to call witnesses who were peculiarly within the IRS's power to produce and who were in fact at the IRS's disposal — by virtue of their presence on the IRS's witness list and their specific obligation (on penalty of criminal prosecution) to cooperate with the IRS in providing testimony — until the IRS chose to release them. The central question before this Court is whether that was consistent with procedural fairness and the governing evidentiary principles.

* * * * * * * * * *

The following facts (except where otherwise indicated) either were undisputed and/or uncontroverted at trial or are as found by the Tax Court (Lauber, J.) in its June 28, 2018 Memorandum Findings of Fact and Opinion (the “Memorandum,” JA1886-1951, T.C. Memo. 2018-96).

I. The Business Of Delta And The Partnerships

The Delta Group³ first became involved with tax-advantaged investments in 1999. At that time, Andrew Beer (“Beer”), the group's 60% owner, was 32 years old and a recent graduate of business school; his partner Samyak Veera (“Veera,” the group's 40% owner) was in his mid-20s. The exclusive role of the Delta Group in connection with these investments was the design and management of the foreign currency derivatives that formed a part of them. The tax planning, structuring, and marketing of the investments was done by the nation's preeminent accounting firms (including Arthur Anderson, Ernst & Young and KPMG) and highly regarded major law firms (including Arnold & Porter and Curtis Mallet-Prevost). Beer was not a tax lawyer or tax accountant and had no tax law training. He and Veera worked closely with their outside general counsel, Arnold & Porter, whom they paid by the hour and by whom they were advised at every turn to ensure that their design and management of derivatives complied with the tax law. (JA1889-1894; JA2053:184-5; JA2055:193; JA2056-7:198-20; 2058:204; JA2059-60:209-2115; JA2061:218-20; JA2064:228-9; JA2065:233-5; JA2067:242; JA2112-4:315-22; JA2115-6:325-8).

During the period between 1999 and 2002, the Delta Group earned substantial fees from investors in such investments in exchange for the design and management services it provided. Its primary purpose for the foreign currency trading in which it was engaged was always the same: to earn fees from investors for its services and expertise in designing and managing trades. (JA2114-6:320-8; JA2151:383).

II. The Partnerships And Their History

The Partnerships began as an outgrowth of a tax-advantaged investment product called POPS, the structure of which is largely described in the Memorandum (see JA1892-4). They were originally created (in 2000) to enable investors in that product to “roll” losses from one year forward into a subsequent year, in tandem with a foreign currency bet that, if successful, would create substantial profit. The substance and tax treatment of the trades the Partnerships initiated in 2000 were not at issue below and are not at issue here.⁴

III. The 2001 Trades At Issue

The POPS investors left the Partnerships in late 2001. Thereafter, the Delta Group reacquired the Partnerships and they participated in the three sets of trades (the “FX Options,” to which Deutsche Bank was again the counterparty) that produced the losses at issue here. (JA2119-20:340-5).

It was not until a decade later, in 2011, that the IRS issued FPAAs challenging those losses. (JA1926-7). These proceedings went to trial four years later in 2015 (JA1956), with the Memorandum issued in 2018. By the time of trial some 14 years after the returns at issue were filed, the evidence had become distant and stale.

The IRS's position shifted substantially over time. Although it initially contended that the losses should be disallowed because the FX Options had no possibility of profit on their face (see JA4887; JA4905-6; JA4933; JA4955; JA4973-4; JA5011-12; JA5030; JA5048-9), in October 2014 (years after it issued the FPAAs and months after the close of discovery (JA737; JA749-50)) it revealed an entirely new theory: the FX Options could not make a profit because they were not actually carried out according to the terms set forth in the documents governing the trades (see, e.g., JA2678-2766; JA2952-3040; JA3125-3213; JA3312-3400). Instead, the IRS asserted, the Delta Group and Deutsche Bank (a) made a separate agreement to fix in advance the exchange rates that would apply when the trades were settled; and (b) set those agreed-upon rates in a way that ensured that the trades would never make any profit or loss, regardless of actual currency fluctuations. (JA737-8; JA1005-6). At trial, the IRS proceeded only based on this theory that the trades were rigged to remove any actual profit potential — a theory Delta sharply contested. The parties' dispute about whether the trades were rigged was and is the primary issue in this case.

As the Tax Court recognized, under this Court's controlling precedent the FX Options had economic substance if they either (a) “'ha[d] a reasonable prospect ex ante, for economic gain'”; or (b) were “'undertaken for a business purpose other than the tax benefits.'” (JA1936-7, quoting Horn v. Comm'r, 968 F.2d 1229, 1237 (D.C. Cir. 1992)). “If a transaction satisfies either prong of this test, it is deemed to have economic substance.” (JA1937).⁵ The Tax Court concluded that the trades lacked economic substance based on a finding that they were in fact rigged as the IRS claimed and therefore had neither a potential for profit nor a business purpose.⁶ We address the evidence and rulings on each of these two prongs separately.

A. Profit Potential

1. The Evidence

a. The Basic Structure Of The FX Options

The FX Options were placed on November 26, December 7 and December 18, 2001. (JA1907-22). A reasonably detailed description of them is essential to an understanding of the parties' dispute.⁷ Each trade was executed as a large block, of which the Partnerships (and other entities not at issue here) were allocated percentages. For each trade there was a pair of foreign currency options, with the price (“premium”) for one option payable in euros and the premium for the other payable in Danish kroner. Each Partnership paid Deutsche Bank a small portion of each premium in cash and borrowed the balance of the premium from Deutsche Bank. The loans were made (and repayable) in the same currency as the premium; that is, euros for the option with a euro premium, and kroner for the option with a krone premium.

The pair was structured so that one of the trades would result in a large payment in one currency (either euros or kroner) approximately one week after the transaction was executed, and the other would result in a smaller payment in the other currency (either kroner or euros) about two years thereafter. The payment on each trade would be in the same currency as its premium. Which of the two options would result in the large payment and which in the small one was not known at the outset of the transaction; this would be determined based on currency fluctuations between yet another pair of currencies. But it would always be the case that one option would result in a large payment after a week and the other would result in a small one in two years.

The large payment would always be substantially more than the premium for the option that resulted in that payment. Conversely, the small payment would always be substantially less than the premium for the other option. As a result, the Partnership would realize a gain on receipt of the large payment and a loss on the later receipt of the small one.

Whether the Partnership would ultimately make money on any given option pair would depend upon whether the large payment plus the small payment was equal to, greater than, or less than the sum of (a) the principal amount of the two loans, plus (b) the interest due on the loans, plus (c) the amount of cash invested. Half of the total amount due on the loans was denominated in euros and the other half in kroner. But because the lion's share of the total value of the large and small payments would be in only one of those currencies (that is, whichever one represented the large payment), a substantial portion of that total value would need to be converted to the other currency in order to repay the loan in that currency.

The Tax Court's analysis of the November 26, December 7 and December 18 trades makes clear if there was no agreement in advance to fix the exchange rates that would be used when the payments were made and the loans repaid, how far the payments would go toward covering the loans would depend upon the actual movements of the euro and krone against each other during the initial week between the date of the investments and the large payment on the first option. Further, because the Partnerships themselves kept their books in dollars, in the absence of an advance agreement to fix the exchange rates how far those payments would then go toward covering the Partnerships' initial investments (and the value of any excess, if any) would depend upon the dollar value of what was left after the loans were satisfied — which, in turn, would depend upon movements of the euro and krone against the dollar both during that initial week and during the remaining life of the investment until the smaller payment on the second option was made. (JA1907-22).

Currency fluctuations occur continuously. Using a sophisticated “Monte Carlo” simulation,⁸ Delta's expert explained that such fluctuations made it probable that the overall investments would result in a profit or loss (and highly unlikely that they would result in a net gain or loss of zero). (See JA361-9). In other words, in the absence of an agreement to rig the rates the trades had the potential to make a profit or loss. The Tax Court essentially agreed, emphasizing repeatedly that the basis for its conclusion that they had no such potential was its finding (discussed further below) that there was such an agreement. (JA1909; JA1914; JA1919; JA1940; JA1941; JA1944; JA1947-8).

The option pairs included an additional feature that could have made even more money for the Partnerships: they represented a bet on the euro/krone “peg.”⁹ More specifically, in 2001 the krone was pegged to the euro (and hence ordinarily traded within a specified band relative to the euro and fluctuated only within that band). However, if the peg were broken, the krone would likely have risen precipitously in value relative to the euro. Thus, in a case in which a Partnership received the large payment in kroner, the Partnership would have realized a significant gain because the cost of repaying the loan (in euros) on the other option would have effectively been reduced. (JA369-371; JA2110-12:305-12; JA2145:358-9).

In addition, if after the first week the large payment was made in euros rather than kroner (such that the later, smaller payment would be made in kroner), then the Partnership would have a further bet on the euro/krone peg: it would be entitled to a payment in kroner whose value could skyrocket if a de-peg occurred at any point in the next two years. Delta presented documentary evidence that the possibility of a de-peg and the substantial profits that could result from one were among its considerations in placing these bets on behalf of the Partnerships. (See JA5397).

Such a de-peg was an unlikely event. But it was by no means impossible, as proven by events shortly before the trial of these cases. Until January 2015, the Swiss franc was, like the krone, pegged to the euro. In January 2015, it de-pegged — despite assurances from the Swiss central banking authority only one month earlier that it would not do so. Investors who were holding Swiss francs at the time made substantial amounts of money. (See JA4923-41). The same would have been true of the Partnerships if Denmark had made such a move at any point when they were holding kroner.¹⁰ Although the Tax Court dismissed this as a source of profit potential, it made clear that its reason for doing so was its finding that the parties had rigged the rates. (See JA1944 (finding that because of such rigging the Partnerships could not have made a profit “even if a depeg occurred”)).

b. The Spreadsheets And The Deutsche Bank Witnesses

As Delta pointed out in the Tax Court, because the IRS's theory that the trades were rigged was nowhere revealed in the FPAAs (and, as the court acknowledged, was not in fact revealed until in the Fall of 2014 — see JA1005-6), as a matter of law the IRS had the burden of proof on rigging. (See infra at 38-40). In its witness list in its Pretrial Memorandum, the IRS made clear how it planned to meet that burden: it would introduce through witnesses spreadsheets that, the IRS claimed, were “exchanged” between Deutsche Bank and the Delta Group and evidenced their “agreement” to rig the exchange rates that would apply on future dates. (See JA716-8).

What the IRS offered, however, was merely a series of spreadsheets (with a certification that they came from Deutsche Bank's records), together with a handful of emails. The IRS offered no email (and no other evidence) that even suggested, much less demonstrated, that any spreadsheets were exchanged between Deutsche Bank and Delta with respect to the first set of trades on November 26, 2001. The emails the IRS did offer showed that (a) on December 7, 2001 at 4:42 p.m. — a time that may or may not have been before the second set of trades was placed on that date — Thomas Chin of Deutsche Bank sent Clayton McDonald of Delta¹¹ an email referring to “today's trades” and attaching a set of Deutsche Bank spreadsheets (JA6898-6916); (b) on December 18, 2001 at 4:34 p.m. — a time that, again, may or may not have been before the third set of trades was placed that date — Chin (Deutsche Bank) sent McDonald (Delta) another email referring to a “third tranche” and attaching another set of Deutsche Bank spreadsheets (JA6941; JA6956-63); and (c) on December 21, 2001, McDonald (Delta) sent Bryan Moran of Deutsche Bank an email referring again to “Tranche 3” and attaching a set of Delta spreadsheets (JA7038-7048).¹² The spreadsheets Deutsche Bank sent to Delta on December 18 for the “third tranche” show that Deutsche Bank calculated the anticipated payout on the seventh day using rates of 7.441476021 (EUR/DKK), 0.902191373 (EUR/USD), and 8.248223429 (USD/DKK), while the spreadsheets Delta sent to Deutsche Bank three days later for that same tranche show that Delta calculated the anticipated payout using rates of 7.44111250 (EUR/DKK), 0.9025 (EUR/USD), and 8.245 (USD/DKK) (JA6957, lines 10-26; JA7041, lines 9-37). In other words, (a) there is no email relating to the first trade; (b) it is not clear that any email between Deutsche Bank and Delta was transmitted before either of the other two trades; and (c) in all events, there was no “exchange” in which any rates were agreed.

Moreover, the Deutsche Bank spreadsheets attached to its emails are themselves opaque and, without a live witness to illuminate them, provide no explanation of why the rates therein were chosen. In sharp contrast, the Delta Group's own internal spreadsheets — which, as noted above, employed exchange rates different from those shown on Deutsche Bank's (see supra at 15-16; accord JA6456-6536; JA6537-6619; JA6620-6703; JA6898-6916; JA6941-63; JA7038-48) — were admitted into evidence with the express caveat that Delta would have to show their meaning through witness testimony. (See JA2034-5:171-5). Delta did so through the testimony of Andrew Beer, who took the court through its spreadsheets line by line, explaining all of the numbers and tying them into the transactions. (JA2170-6:401-23). He also explained that, in accordance with common practice in the industry, the purpose of the spreadsheets was not to reflect an agreement on the exchange rates that would ultimately be applied in calculating any profit or loss, but rather to make sure that the trades were priced properly — that is, to make sure no profit or loss was inadvertently built into the trades in advance, by confirming that the trades were priced so that if the market did not change at all the parties would essentially break even. (JA2173-4:411-14). This did not mean that the parties expected the market to remain static; to the contrary, they expected it to change. (See, e.g., JA2148:370-3). The purpose of the calculation was to ensure that the trades were priced in such a way that any profit or loss would result from that change and not from an error in pricing in the first instance. (JA2173-4:411-14).

Deutsche Bank's spreadsheets were likely prepared for the same reason, which would explain why the rates it used were close to (but still different from) the rates used by Delta. That is, Delta and Deutsche Bank shared the same purpose but employed slightly different approaches and thus came to slightly different numbers. (See supra at 15-16).¹³ But in all events, without some explanation from Deutsche Bank about what its spreadsheets meant, their mere existence could not serve as evidence that the Delta Group agreed in advance that instead of settling the trades at the rates in effect at the time of settlement the parties would settle them using the rates Deutsche Bank applied in its ex ante analysis. Indeed, the small differences in the rates used by Delta and Deutsche Bank are inconsistent with any agreement to fix the rates.

By the time the IRS offered the Deutsche Bank spreadsheets and emails in evidence to show the alleged rate rigging scheme, the Tax Court had already ruled (in connection with Delta's motion in limine) that “[t]hese emails do not, by themselves, prove that there was an agreement or understanding to fix FX rates.” (JA1006). Consistent with this, the Tax Court agreed with Delta that an explanation of the Deutsche Bank spreadsheets would be necessary. Again and again, it specified that the IRS had the “burden to tie the numbers in” and that it was the IRS's “burden to clarify” the spreadsheets, just as Delta had by that point already done (and was required to do) with its own spreadsheets.¹⁴ It also reiterated that the emails themselves did not prove an agreement. (See JA2220-1:521-4). The Deutsche Bank spreadsheets were admitted into evidence with that express “caveat,” which was repeated for each spreadsheet. (See supra, n.14).

The IRS made clear that in addition to demonstrating the “exchange” of spreadsheets, it would also prove the rate rigging agreement through the testimony of (among others) several employees or former employees of Deutsche Bank, including Chin, Moran, and Andrew Baxter — the Deutsche Bank representatives who sent, received, and/or were copied on the December 7, 18 and 21 emails. (JA716-8, JA720; see supra at 14-15; JA6898; JA6941; JA7038).¹⁵ These witnesses, the IRS promised:

would testify as to the process by which Deutsche Bank and Delta agreed to the parameters of the FX binary option trades, including agreements on forward FX rates, spot FX rates, and interest rates . . . [and] may also testify as to the content of specific documents exchanged between Deutsche Bank and [Delta].

(JA716-18, emphasis added). That is, the IRS squarely told the Tax Court that it would put Deutsche Bank witnesses on the stand who would testify that Deutsche Bank and the Delta Group had rigged the rates and would also testify about the exchange of documents evidencing that alleged agreement.

c. The Deutsche Bank Non-Prosecution Agreement 

We pause here for a word of background about Deutsche Bank that is essential to an understanding of these events. Deutsche Bank was heavily involved in tax-advantaged investments — not only those in which the Delta Group participated, but also numerous others. In or about December 2010 (about eleven months before the FPAAs at issue here), Deutsche Bank entered into an agreement (the “Non-Prosecution Agreement”) with the United States Department of Justice (“DOJ”) in which (a) the DOJ agreed not to “criminally prosecute” Deutsche Bank for any crimes related to its roles in those investments — which (according to the DOJ) “result[ed] in the evasion of approximately $5.9 billion in U.S. individual income taxes” (JA4757); and (b) in exchange, Deutsche Bank agreed to pay $533,633,153 in fines and penalties and to “cooperate fully” with the IRS by (among other things) “us[ing] its best efforts to secure the attendance and truthful statements or testimony of any past or current officers, agents, or employees at any meeting or interview or before the grand jury or at any trial or other court proceeding” (JA4758-9). The Non-Prosecution Agreement also provides that any violation shall subject Deutsche Bank to criminal prosecution, and further provides that any such prosecution that is not time-barred as of the date of the agreement shall not be subject to any statute of limitations defense. (JA4759-60). As a result, Deutsche Bank had a legal obligation — under penalty of criminal prosecution — to assist the IRS in this trial by using “its best efforts to secure” the attendance and truthful testimony of Chin, Moran and Baxter as present or former Deutsche Bank employees. And those witnesses, all on the IRS's witness list, apparently had been “secured” in this manner and stood ready to testify truthfully at the IRS's behest.

d. Delta's Evidence

At trial, Delta presented its own evidence first. With respect to the question of whether or not there was an agreement with Deutsche Bank fixing the exchange rates in advance, Delta offered:

• The formal documents that governed the transactions

Both the FX Options themselves and the loan transactions through which the Partnerships financed them were subject to an ISDA Master Agreement between Deutsche Bank and each of the Partnerships. (JA2725; JA2727; JA2729; JA2734; JA2739; JA2741; JA2743; JA2748; JA2753; JA2755; JA2757; JA2762; JA2999; JA3001; JA3003; JA3008; JA3013; JA3015; JA3017; JA3022; JA3027; JA3029; JA3031; JA3036; JA3172; JA3174; JA3176; JA3181; JA3186; JA3188; JA3190; JA3195; JA3200; JA3202; JA3204; JA3209; JA3359; JA3361; JA3363; JA3368; JA3373; JA3375; JA3377; JA3382; JA3387; JA3389; JA3391; JA3396). Each ISDA Master Agreement required any payment between the parties to be made in the currency specified in the contract under which the payment was made (referred to in the agreement as the “Contractual Currency”). (JA2688; JA2962; JA3135; JA3322, each at ¶8). Thus, for example, if the result of one of the transactions was that a Partnership was entitled to receive a large payment in kroner, the Partnership could insist on receiving that payment in kroner. Payment in any other currency would not satisfy the obligation unless that currency was tendered in an amount the Partnership could convert into the amount of kroner it was entitled to receive. (Id.). Further, if Deutsche Bank tried to pay the Partnership an amount in dollars, euros, or any other currency that could not be converted to the required amount of kroner, the Partnership could require Deutsche Bank to make up the shortfall. (Id.).

Each ISDA Master Agreement goes on to specify that no purported amendment, modification or waiver is effective unless it is embodied in a writing signed by both parties or confirmed by an exchange of electronic messages between the parties. (JA2689; JA2963; JA3136; JA3323, each at ¶9(b)). In other words, the written agreements between the Partnerships and Deutsche Bank prohibited the rate-fixing arrangement the IRS hypothesized unless the parties expressly agreed to it in writing (or, at the very least, in an email exchange — that is, one that includes a confirmation from the Partnerships or Delta acting on their behalf).

In keeping with these underlying documents, each block trade was made pursuant to Confirmations and OTC Deposit Annexes that set forth the specific terms of the trades and the amounts to be paid in Euros or Krone without specifying the exchange rate to be used. (See JA2725-2766; JA2999-3040; JA3172-3213; JA3359-3400).

• The testimony of Andrew Beer

The only knowledgeable witness to testify about whether or not there was an agreement to rig the exchange rates was Beer. In addition to explaining the existence of the Deutsche Bank spreadsheets by reference to Delta's own spreadsheets and walking the Tax Court through the numbers in Delta's spreadsheets in order to tie them into the transactions (see supra at 16-17), Beer testified that it would have been distinctly adverse to his business interests and those of investors to rig the rates. He explained that future investors needed to be able to use any losses generated by the transactions. He had been advised repeatedly by Arnold & Porter that those losses could not be used by future investors unless the transactions that generated them included a potential for profit and a risk of loss. (JA2581-3). The transactions were carefully designed to include such a potential and such a risk — while at the same time ensuring that the risk would be limited to the amount invested by each Partnership and that Deutsche Bank would be hedged overall. Taking the risk out altogether (and with it the profit potential) would have destroyed the value of the transactions and made them thoroughly unattractive to future investors. This would have run directly contrary both to Delta's interests and to those of Deutsche Bank — whose primary profit expectation with respect to these transactions arose from the fees they would receive in the event the contemplated investment products were ultimately sold to an investor. (JA2584). But in addition, in the event of a de-peg Beer expected the Partnerships (and, by virtue of its share in the Partnerships, the Delta Group) to make a substantial amount of money. (JA2145:358-9).¹⁶

e. The IRS's Evidenc e

The IRS's witness list included every person from Deutsche Bank and Delta who was party to the December 7, 18 and 21 emails. (See supra at 19-20). Prior to trial, Delta moved to exclude the testimony of those five witnesses on the theory that they had not been timely disclosed (JA737); the court denied that motion on the ground that the emails had been produced to Delta in or about October 2014 five months before trial and included the names of those witnesses. Said the court:

These emails do not, by themselves, prove that there was an agreement or understanding to fix the FX rates. But they surely informed petitioners of the people whom the IRS thought might be relevant to such an inquiry.

(JA1006, emphasis added). Nevertheless, once it heard Delta's evidence the IRS withdrew every single one of those five witnesses.

As a result, there was no testimony from anyone about what the emails meant or whether any of them was sent in advance of the set of trades to which it ostensibly related. Instead, the only testimony the IRS offered on this subject was that of its expert, Timothy Weithers, who had no personal knowledge concerning the spreadsheets or the emails and acknowledged in his testimony that he was not aware of any writing in which Delta and Deutsche Bank ever agreed upon the exchange rates in advance. (See JA2349-60:749-95).

* * * * * * * * * *

We note one more thing. After the small payment, the Partnerships realized in the aggregate a net profit of $128,000 on an initial cash investment of $873,000. (JA4097; JA73 at ¶104; JA73-4 at ¶107; JA74 at ¶110; JA79 at ¶134; JA80 at ¶¶137, 140; JA406 at ¶172, JA407 at ¶175, JA407-8 at ¶178, JA412 at ¶201, JA413 at ¶204, JA413-4 at ¶207). This represented a gain of roughly 15% over the period from November 26, 2001 (when the first set of FX trades was executed) to October 9, 2002 (when the outstanding options were closed) — a period when the Dow Jones declined by more than 27% and the S&P 500 declined by more than 32%. (JA144 at ¶76; JA145 at ¶¶82, 83; JA146 at ¶89). In other words, they made profits in a time period when, had they invested in the stock market, they likely would have lost money.

2. The Tax Court's Ruling

The Tax Court found that the Delta Group and Deutsche Bank agreed in advance on the exchange rates that would apply when the trades were settled, and on that basis found that they could not have produced a profit. (JA1937-41). It made this determination based on a finding that the Deutsche Bank spreadsheets were “exchanged between Delta and Deutsche Bank when executing” each of the trades. (JA1939). It found that these spreadsheets were not merely pricing calculations, but rather signified Deutsche Bank's intent to actually settle the trades at the rates set forth in the spreadsheets regardless of what the market actually did. Because the Tax Court found that the spreadsheets were “exchanged” before each trade, it concluded that the Partnerships “knew in advance how many dollars they would receive when the trades closed, [and] bore no foreign exchange risk and had no possibility of making any foreign exchange profit.” (JA1939). As a result, it concluded that the Monte Carlo simulation Delta submitted (see supra at 12) had “no probative value” because it assumed the trades were not rigged (JA1929), and further determined that the possibility of a depeg was irrelevant because (in light of the rigging) the Partnerships “could not have profited” even if one occurred (JA1944).

The court below heavily rested its conclusion that the trades were rigged on its decision to construe the Deutsche Bank documents against Delta based on an adverse inference it drew from the absence of any testimony from Deutsche Bank about what they meant. Said the court:

In any event, the most logical witness to testify about Deutsche Bank's trading practices would have been a witness from Deutsche Bank. [Delta] did not call anyone from Deutsche Bank to testify, and from this we infer that such testimony would not have been helpful to [Delta].

(JA1940). The court said nothing of the requirement it had expressly imposed that the IRS call witnesses to “tie in” or otherwise explain the spreadsheets with competent testimony: there was no finding that the IRS had met that requirement (nor could there be, as the IRS made no attempt to do so), and no explanation of why it was not enforced. Nor did the court say anything of its prior indication — which it initially made in denying Delta's motion in limine and later reiterated during trial — that the emails the IRS ultimately offered as the only evidence of any agreement to fix the rates “do not, by themselves, prove” such an agreement. (See supra at 19 and 25). Nor did the court address the fact that — far from being peculiarly available to Delta (as the law would have required them to be in order to support an adverse inference based on Delta's failure to call them — see infra at 43), the Deutsche Bank witnesses were, by virtue of the Non-Prosecution Agreement, peculiarly within the IRS's control and not Delta's. The court simply used the adverse inference to close the gaping hole in the IRS's proof.

Regarding the profit that the Partnerships actually made (that is, a 15% return in less than one year — see supra at 26), the Tax Court dismissed it as simply a return of the Partnerships' initial investments “with interest at LIBOR” (JA1941), although no party presented any version of LIBOR that would have yielded such a return.

B. Business Purpose

1. The Evidence

Delta also presented evidence at trial concerning the purpose for which it took control of the Partnerships and placed the FX Options: it wanted to continue the Partnerships' existence and trading so that they could be used in new investment products that the Delta Group hoped to market to investors at a substantial profit. More specifically, in late 2001 the Delta Group was making efforts to expand its business. The business plan was to (a) continue developing new forms of derivatives for use in new tax-advantaged investments that would have more robust investment (and profit-generating) components and would be designed in part to be more in line with positions being taken publicly by the IRS; (b) identify entities in which those derivatives could be utilized; (c) help identify new investors and also approach prior investors whose trades it had placed and managed before; and (d) separately market its other services, including asset management and currency trading, to previous investors with whom it had relationships and to new investors in new transactions. In this regard, in addition to Beer's testimony Delta submitted substantial documentary evidence concerning at least four new products on which the Delta Group was working in 2001 and 2002, for which the Partnerships with their trading history and embedded losses could have been used.¹⁷ It also submitted extensive testimony from the Delta Group's General Counsel, David Diamond (see JA2428:901), who testified — supported, again, by documents — about the group's efforts with respect to particular transactions and clients in that time frame. (See JA2429-33:907-22; JA5413-15; JA5418-98; JA5501-15; JA5518-19).

In the end, nothing came of the Delta Group's extensive efforts and no new tax-advantaged transactions were closed, because in early 2002 the IRS issued “soft letters” (pursuant to IRC § 6112) to a number of accounting and law firms. The letters, in substance, directed the firms to disclose the names of their clients who were investors in certain tax-advantaged transactions that the IRS did not believe were lawful. The Delta Group received such a letter on May 29, 2002. (JA1925).

In June and September 2002, the IRS published formal notices announcing that the tax-advantaged investments it had targeted in the soft letters would be challenged as unlawful. (JA1925). The Delta Group was advised by major accounting and law firms that the soft letters and the notices only represented the IRS's views and were not legally binding. These firms also reiterated their respective views that POPS and similar tax-advantaged transactions were lawful. (JA2062:220; JA2067:242; JA2114:320-1; JA2115-6:32708; JA5392). In the real world, however, whether the firms were right or wrong did not matter. The combination of the soft letters and the notices ended the market for tax-advantaged investments in 2002. The Delta Group was unable thereafter to successfully market a single such investment. (JA2183-4:454-51; JA5414-5; JA1925).

With the market dead and no sales possible, on October 9, 2002, the Delta Group terminated the FX Options, generating $144 million in losses. (JA2185:458-9; JA1926). These losses were only partially utilized by the Delta Group: they flowed up to a Delta Group affiliate that had earned $40 million in fees in 2002 from tax-advantaged investments. (JA2151-2:385-6; JA2185:458-9). The $40 million in gains was netted against the $144 million in losses and the balance was carried forward. (JA2185:458-9; JA2267:631-2; JA2268:635; JA2298:654). The group earned no further fees after that (and consequently could “use” no further losses); ultimately, its primary operating company went into bankruptcy and its business failed. (JA2598; JA2296:647; JA2299:656; JA5066 at ¶1; JA5276-5382).

Of equal importance, the record contains none of the evidence one would expect to see if the FX Options had been placed for the purpose of generating tax losses for Beer. For example, no effort was made to match the losses they generated to Beer's anticipated income; in fact, contrary to what occurs in a “prearranged tax shelter,” the losses at issue ($144 million) here did not match — or even come close to approximating — the $40 million in gains they were ultimately used to defray. (JA2267:631-2; JA2268:635; JA2298-9:655-6). As well, unlike the gains and losses targeted by the IRS notices, that $40 million in gains did not come from some source “unrelated” to the losses; rather, it represented profits the Delta Group had earned for services similar to the services for which it had hoped to earn profits in connection with the new investment products it was developing. Moreover, had the transactions been placed for the benefit of any particular investor (including Beer), Deutsche Bank would have received a significant fee at the time. (See, e.g., JA2255:584; JA2584). But no fees were paid to Deutsche Bank for the transactions, precisely because they were designed to be part of a product that an investor yet to be identified would pay for in the future — whereupon Deutsche Bank would be compensated pursuant to the parties' usual arrangement, with a share of the fees paid by the investor. (See JA2256:586-7; JA2301:664-5).

2. The Tax Court's Ruling

As noted above, the Tax Court recognized that the question of whether the FX Options had a business purpose is separate from their profit potential, and that under this Court's precedent the two must be considered disjunctively. (See supra at 8). Nevertheless, the court's finding that the exchange rates for the FX Options were fixed in advance in a way that ensured they could make no profit drove its conclusion that they also lacked a business purpose. (See JA1939-41). The court stated that there was “no evidence” other than Beer's own testimony to support Delta's argument that the Delta Group intended to profit by using the Partnerships as a component of new investment products. (JA1946). It therefore found that the Delta Group's purpose in reacquiring the Partnerships and placing the FX Option trades was simply to “reduce its and Mr. Beer's own tax liabilities,” rather than to further the Delta Group's profit-making business. (JA1946).

* * * * * * * * * *

Having concluded that the FX Options satisfied neither prong of Horn, the Tax Court held that they lacked economic substance. On that basis, it sustained the IRS's disallowance of the losses they generated. (JA1947-8). This appeal followed.¹⁸

SUMMARY OF ARGUMENT

The Tax Court's ruling that the FX Options lacked economic substance cannot stand unless it was correct both in its determination that the trades lacked profit potential and in its determination that they lacked a business purpose. There is ample basis to conclude that the Tax Court erred in either or both of those determinations.

The Tax Court's determination that the FX Options lacked profit potential was based entirely on its finding that the trades were rigged to produce no profit despite their written terms. As detailed below, this finding is unsupported by the evidence. The insufficiency of the evidence should have compelled the Tax Court to conclude that the IRS had not met its burden of proving that the trades were rigged — a burden that rested squarely with the IRS as a matter of law because it did not reveal its rate-rigging theory until years after it issued the FPAAs. Instead, the Tax Court filled in the substantial gaps in the IRS's proof and found in its favor by drawing an adverse inference against Delta based on the fact that Delta did not call any Deutsche Bank witnesses to testify. But no such inference should have been made against Delta because — far from being peculiarly available to it and not to the IRS (as required to support an adverse inference) — by virtue of the Non-Prosecution Agreement those witnesses were at the IRS's beck and call and not Delta's. Moreover, on its witness list the IRS had undertaken to call them, and the court had previously indicated that it would be the IRS's burden to do so in order to “tie in” the numbers on the documents the IRS submitted. When the IRS withdrew those witnesses after hearing Delta's evidence, Delta — believing that the IRS had failed to meet the burden that the court had repeatedly emphasized would be the IRS's to meet — had no reason to call them itself. To draw an adverse inference against Delta under these circumstances was error as a matter of law. (Point I.A). Upon a determination that the Tax Court erred in finding that the trades were rigged, its conclusion that they lacked profit potential should be reversed. (Point I.B).

This would be enough to require reversal of the Tax Court's ruling that the FX Options lacked economic substance. But that ruling can and should be reversed on a second ground: the Tax Court also erred in determining that Delta lacked a business purpose for the trades. The court, in essence, collapsed the business purpose analysis into the profit potential analysis: it disposed of “business purpose” based on its finding that there was no profit potential. This was error as a matter of law: this Court's precedent required the Tax Court to analyze business purpose disjunctively from profit potential. Delta submitted ample evidence that it was the Delta Group's (and Beer's) intent to obtain profits through use of the Partnerships, and that the reason they got losses instead was because of the complete failure of their business. Contrary to the Tax Court's bald statement (JA1946), that evidence consisted of more than just Beer's own testimony, it also included numerous documents and the testimony of the Delta Group's former General Counsel. This evidence was sufficient to support a finding that the FX Options had a business purpose regardless of any potential to yield a direct profit. (Point II).

The Tax Court's ruling that the FX Options lacked economic substance should be reversed on either or both of the foregoing grounds.

STANDARD OF REVIEW

As the Tax Court's findings of fact are reviewed for clear error,¹⁹ much of this Court's review here will be governed by that standard. Under it, this Court must “look to all the evidence of record to determine whether [the Tax Court's findings] can pass muster”; those findings “[should] not be permitted to stand where [they are] based on a serious mistake as to the effect of evidence or [are] clearly contrary to the weight of evidence.” Daniels v. Hadley Mem'l Hosp., 566 F.2d 749, 757 (D.C. Cir. 1977). If the record contains conflicting evidence, “the clear error standard requires [this Court], as the reviewing court, to assess the comparative weight of the evidence both for and against the [lower] court's finding. It may be that the evidence relied upon by the [lower] court is insufficiently probative to sustain its finding.” Latif v. Obama, 677 F.3d 1175, 1201 (D.C. Cir. 2011).

Certain of the matters at issue here, however, are subject to different standards of review. The Tax Court's decision that it could draw an adverse inference against Delta based on its failure to call witnesses from Deutsche Bank (see supra at 27-28 and infra at 41-46) is reviewed de novo as a question of law. See Huthnance v. District of Columbia, 722 F.3d 371, 378 (D. C. Cir. 2013) (adverse inference is “proscribed . . . where its premises do not obtain”). If such an inference was available as a matter of law, its application is reviewed for abuse of discretion. See id. at 379; accord Overnite Transp. Co. v. N.L.R.B., 140 F.3d 259, 266 n.1 (D.C. Cir. 1998).

The Tax Court's decision to admit or exclude evidence — including expert evidence — is reviewed for abuse of discretion. United States v. Hite, 769 F.3d 1154, 1168, 1169-70 (D.C. Cir. 2014). Its determination concerning the burden of proof, however, is a question of law subject to de novo review. See Abbot v. Perez, __ U.S. __, 138 S. Ct. 2305, 2326 (2018). The same is true of the manner in which the Tax Court applied the analysis mandated by Horn.

Under any of these standards, if this Court determines that the Tax Court erred it should reverse if it finds that the error was prejudicial — that is, the error impacted the outcome. See Huthnance, 722 F.3d at 381 (“A court confronting a trial error must ask whether the error substantially affected the outcome of the case. If the court cannot say with fair assurance that the error was harmless, it must conclude that the error was not.”).

ARGUMENT

I. THE TAX COURT'S FINDING THAT THE FX OPTIONS LACKED PROFIT POTENTIAL WAS CLEARLY ERRONEOUS

The Tax Court's determination that the FX Option trades lacked a potential for profit was premised entirely on its finding that Delta and Deutsche Bank agreed to “fix[ ] the applicable exchange rates in advance.” (JA1939; accord JA1929): “Because the exchange rates at which the options would be settled were agreed in advance, neither party bore any meaningful foreign-exchange risk.” (emphasis added)). Because that finding was clearly erroneous, it should be reversed — and, with it, the determination that the FX Option trades lacked profit potential.

A. The Tax Court Committed Clear Error In Finding That The IRS Had Met Its Burden Of Proving An Agreement To Rig The Trades

1. The IRS Had The Burden Of Proving Such An Agreement

The Tax Court did not rule on the burden of proof because it concluded that its findings were supported by a preponderance of the evidence. (JA1934). Although we submit that the Tax Court's factual findings were clearly erroneous regardless of who had the burden of proof, we also submit that burden of proof should have been allocated to the IRS. To begin with, the Tax Court expressly ruled that the IRS would have the “burden” to show that the spreadsheets meant what the IRS said they meant. (See supra at 19). The court never rescinded or modified that ruling, nor did the IRS ask it to do so. That burden remained with the IRS, and Delta tried its case that way — among other things, consciously electing not to call any additional witnesses concerning the spreadsheets after the IRS dropped all of its Deutsche Bank witnesses. The burden of proof should rest on the IRS at least with respect to what the spreadsheets tell us about the alleged rigging, simply because that is what the court pointedly and repeatedly told the parties (as it had also told them with respect to Delta's spreadsheets).

Even without that ruling, under Tax Court Rule 142(a)(1) — which allocates the burden of proof to the IRS “in respect of any new matter” — the IRS should have been allocated the burden of proving its assertion that the terms of the trades were not as reflected in the written Confirmations and the ISDA documents, because the IRS did not reveal that theory when it issued the FPAAs in 2011. The earlier reports of its revenue agents likewise argued only that the absence of profit potential arose from the written terms of the transactions; they made no mention of any separate agreement to rig the rates. (JA4885-4992; JA4953-5065). The IRS did not disclose its theory about rate rigging until October of 2014 (three years after the FPAAs and close to the commencement of trial), when it indicated that it had developed that theory based on documents it had received from Deutsche Bank. (JA737 at ¶1(c)). The fact that the IRS itself apparently did not even think of that theory until that late date makes doubly clear that it is “new matter” within the meaning of Rule 142(a)(1).²⁰

In all events, proof of the existence or non-existence of an off-term-sheet agreement to rig the rates requires the presentation of evidence that is different from evidence of the profit potential and business purpose of the transactions based on the written documents that govern them. In fact, any such agreement is contrary to those documents. (See supra at 22-23). Because these two theories require different proof, the rate-rigging theory is “new matter” as to which the burden of proof should have been allocated to the IRS.²¹

2. The Tax Court Erred In Finding That The Parties Agreed In Advance On The Rates At Which The FX Options Would Be Settled

The Tax Court's determination that the parties agreed in advance on the rates at which the FX Options would settle was based on (a) a finding that the Deutsche Bank spreadsheets were “exchanged” in advance of each trade and demonstrate such an agreement; and (b) Deutsche Bank's apparent use of the rates set forth in its spreadsheets when it settled the trades. But as detailed below, the finding that the spreadsheets were “exchanged” finds no support in the evidence, and Deutsche Bank's use of particular rates cannot by itself support the conclusion that the Partnerships agreed to those rates. As a result, the court erred in finding that the rates were fixed.

a. There Is No Evidence Of Any “Exchange” Of Spreadsheets

The Tax Court found that “[b]efore executing” the November 26 trade, Delta and Deutsche Bank “exchanged trade summary spreadsheets” that “included the seven-day forward rate prevailing on the date the trades were opened” and thus agreed on and fixed in advance the rates at which they would settle (JA1908). The court made the same finding with respect to the December 7 trade (JA1914) and the December 18 trade (JA1919). As detailed above, the only evidence of any “exchange” of spreadsheets was that Deutsche Bank emailed spreadsheets to Delta on December 7 and 18 in the late afternoon, and Delta separately emailed a spreadsheet (showing different rates) to Deutsche Bank on December 21. There is no evidence that either party sent the other any spreadsheet with respect to the November 26 trade. Nor is there any evidence that that the spreadsheets transmitted by email on December 7 and 18 were sent “before executing” the trades placed on those dates. The IRS's witness list included the Deutsche Bank author of those emails (who could have provided such evidence), but after hearing Delta's case in chief the IRS pointedly decided not to call him. (See supra at 14-15).

There is no reasonable basis to construe this evidence as showing an “exchange” of spreadsheets “before” each of the trades, let alone one that could signify an agreement on the rates that would vary the parties' obligations under the ISDA agreements. (See supra at 15-16). To find that it shows such an agreement was clear error — particularly given the court's prior caution that the emails “do not, by themselves, prove that there was an agreement or understanding to fix the FX rates.” (JA1006).

Although this would be enough to require reversal of that finding, there is more. The only fact witness who testified about any spreadsheets (Beer) explained that the Delta Group also prepared internal spreadsheets respecting the FX trades, that those spreadsheets were prepared only to ensure that the trades were properly priced, and that Deutsche Bank's spreadsheets were likely prepared for the same purpose. (See supra at 16-18). This might not have been the only possible explanation for the spreadsheets, but at trial it was the only explanation offered by any fact witness. Nevertheless, the Tax Court (a) pretended that Beer had not offered any such explanation (see JA1940); and then (b) drew an adverse inference against Delta based on Delta's failure to call any witness from Deutsche Bank to corroborate what Beer had said (id.).

This was error as a matter of law. A trial judge may not draw an adverse inference based on a party's failure to call a particular witness unless the witness is “peculiarly within the power of” that party. Huthnance, 722 F.3d at 378 (citation and internal quotations omitted). There was no basis for any finding that Deutsche Bank witnesses were “peculiarly within the power of” Delta; indeed, the IRS did not argue as much, and the Tax Court did not find as much. As a result, the Tax Court could not draw any inference against Delta by virtue of their absence.²²

This error was plainly prejudicial. Without the adverse inference, the only evidence about what the spreadsheets meant was Beer's testimony. As a matter of law, the Tax Court could not adopt the IRS's proffered meaning of the spreadsheets based solely on a finding that Beer's testimony was not credible.²³ This makes a reversal of the Tax Court's finding that the spreadsheets show an advance agreement on the exchange rates doubly warranted.

We submit that this is more than enough to require a reversal of that finding. But we further submit that, under the circumstances, in fact the party against whom the Tax Court should have drawn an adverse inference was the IRS. As detailed above, Deutsche Bank was entirely within the IRS's control: it had an obligation to cooperate with the IRS and would have faced criminal consequences if it failed to do so. (See supra at 20-21). This alone would support an adverse inference against the IRS based on its failure to call the Deutsche Bank witnesses on its witness list, particularly in light of the IRS's failure even to attempt to explain why, having so listed them and then seen and heard Delta's evidence, it chose not to call them.²⁴ But as detailed above, the Tax Court expressly told the parties that it would be the IRS's burden to call such witnesses to “tie the numbers in,” and the Deutsche Bank spreadsheets were admitted subject to that “caveat” — which the Tax Court repeated over and over with respect to each such spreadsheet. (See supra at 19 and n.14).

The IRS heard that warning. It then saw and heard the evidence Delta offered on its case in chief. At that point, the IRS had every reason to call the Deutsche Bank witnesses on its witness list if they would do anything other than support or corroborate Delta's position. They were, in effect, sitting in the IRS's bullpen readily available for the IRS to call in order to meet the burden that the court had specifically told the IRS it would have to meet. But rather than putting them before the court, the IRS made the opposite and calculated decision not to call any of the witnesses it said it would call to explain the spreadsheets and the alleged rigging. The IRS, in other words, used the control it had over those witnesses by virtue of the Non-Prosecution Agreement to designate them on its witness list, and then further exercised that control by releasing them after it heard Delta's evidence. At that point, Delta had absolutely no reason to call those witnesses itself: the IRS had failed to make the case that the court had specifically told the parties the IRS would have the burden to make.

The only fair inference to be drawn is that the IRS knew those witnesses would not support its case. This is not merely a technical point; it is inescapable as a matter of simple logic and common sense. In light of all of the circumstances, it was not only error as a matter of law to apply an adverse inference against Delta; it was also an abuse of discretion not to apply one against the IRS.

With or without an adverse inference against the IRS, however, the record contains no evidence of the “exchange” of spreadsheets on which the Tax Court based its finding that the parties agreed in advance to fix the rates at which the trades would settle. It was therefore error to find one.

b. The Outcome Of The Trades Cannot, By Itself, Prove Rigging

Apart from the “exchange” of spreadsheets that the evidence clearly shows never took place and the adverse inference against Delta that was error as a matter of law, the only other basis the Tax Court articulated for its determination that the trades were rigged was the fact that Deutsche Bank apparently applied the rates set forth on its spreadsheets when it recorded the payout from the winning trades on its books. (See JA1939-40). But that fact standing alone is insufficient to show an agreement to fix the trades at the outset — a species of res ipsa loquitur — unless (at a minimum) it could not be explained by anything else.²⁵ That is not the case here.

The documents that set forth the terms of the trades did not specify an exchange rate for dollar/euro, dollar/krone, or euro/krone (see JA2678-2766; JA2952-3040; JA3125-3213; JA3312-3400), and in hindsight it appears that Deutsche Bank may not have used any of the actual spot rates in effect on the settlement dates.²⁶ But that fact does not necessarily mean that the rates Deutsche Bank applied were agreed in advance with Delta.

Beer explained that, as provided in the governing documents, Delta relied on Deutsche Bank to choose among actual rates, and believed Deutsche Bank had done so. (JA2250-1:565-6; JA2299-2301:657-65). He also explained that he did not realize that Deutsche Bank apparently did not do so until these proceedings. (JA2299:658-9; JA2579-80). The reason for this was three-fold. First, the Delta Group's risk management procedures were not designed to raise red flags based on an absence of any profit or loss; rather, what would have generated scrutiny was a large profit or loss, or a situation in which the transactions uniformly resulted in a profit or uniformly resulted in a loss. (JA2300:660-1). For things like making sure the exchange rates were correct, Delta relied on Deutsche Bank's risk management system because Deutsche Bank was “spending hundreds of millions to track trades and other kinds of instruments.” (JA2300:661-2). Second, there was no de-peg. Had there been one, Beer would have looked for a substantial profit on the numerous bets the Delta Group and its affiliates had on that event — and would himself have insisted on strict adherence to the ISDA Master Agreement in order to ensure such a profit. (JA2300:660). Without such a de-peg, Beer had no such global expectation, and therefore no reason to look closely at the results of these specific trades. (JA2149:337; JA2300:660). Third, the Partnerships were not ultimately used in any transaction with an outside investor. If they had been, the transactions would have been scrutinized more carefully after the fact, and the anomaly may have come to light. Without an outside investor, there was no such subsequent scrutiny. (JA2300:661).

The Tax Court rejected this testimony as “self-serving” and not credible, and imposed an adverse inference against Delta for its failure to call any witness from Deutsche Bank to corroborate it. (JA1940). But as detailed above, it was error for the Tax Court to find in the IRS's favor simply by virtue of disbelieving Beer. (See supra at 43-44 and n.23). It was separate error for the court to impose an adverse inference against Delta, given that the Deutsche Bank witnesses were “peculiarly available” to the IRS, not to Delta; to the contrary, the IRS's deliberate choice not to call any of the Deutsche Bank witnesses who were waiting at its (but not Delta's) beck and call supports the inference that they would have corroborated Beer's testimony rather than contradicting it. (See supra at 42-43). And in any event, as a matter of law Deutsche Bank's use of the rates the IRS says the parties agreed upon in advance is not (and cannot be) sufficient evidence in and of itself to meet the IRS's burden of proving the existence of such an agreement. (See supra at 19, n.14).²⁷

Because the evidence in fact shows no advance “exchange” of spreadsheets showing matching rates and includes no explanation from Deutsche Bank as to why its spreadsheets used the rates they did, Deutsche Bank's one-sided use of those rates is the only “evidence” of any alleged agreement. On that record, the Tax Court's committed clear error in finding such an agreement.

B. Absent An Agreement To Fix The Rates, The FX Options Had Profit Potential

Although the IRS initially took the position that the FX Options lacked profit potential on their face, by the time of trial the IRS conceded that, without an agreement in advance to fix the rates at which they would settle, the trades “would have involved tremendous risk” (that is, “tremendous” potential for profit or loss). (JA1504).²⁸ Accordingly, the IRS did not even attempt to show that the trades would have lacked profit potential if the rates were not fixed; to the contrary, its expert admitted that in performing his analysis he “assume[d]” that the rates were fixed and that, in fact, the IRS did not tell him that “one of the issues at (sic) dispute is whether or not there was such an agreement.” (JA2378:803). The Tax Court, too, acknowledged that without an agreement to fix the rates there would have been a “potential for profit”; its only basis for finding no such potential — and for finding the analysis of Delta's expert showing the extent of such potential “irrelevant” (JA1930) — was its conclusion that the rates were in fact fixed. (See JA1939).

As a result, upon a reversal of the Tax Court's finding that the rates were fixed, this Court must conclude that the FX Options had a potential for profit. As a matter of law, such a potential is enough to satisfy the “for profit” prong of Horn (and thereby give them economic substance — see supra at 8) regardless of whether they actually made a profit.²⁹ The Tax Court's conclusion that they lacked such substance should be reversed on this basis alone.

II. THE TAX COURT'S SEPARATE FINDING THAT THE FX OPTIONS LACKED A BUSINESS PURPOSE WAS ALSO CLEARLY ERRONEOUS

If the Court finds that the FX Options lacked profit potential, it should still reverse based on the second disjunctive prong of Horn: Delta had a subjective business purpose for the trades — that is, an overall intent to profit in ways other than though direct gains on the trades themselves. In particular, it hoped to maintain the Partnerships as trading entities with built-in tax losses for the purpose of using them in new investment products on which it would continue to earn the kinds of profits it had earned in the past. (See supra at 29-32). As the IRS has elsewhere conceded (and indeed urged), wholly apart from the tax treatment of any outside investors in such products, these activities by Delta constitute a “business” when they are engaged in for profit. See Chai v. Comm'r, 851 F.3d 190, 210-15 (2d Cir. 2017) (affirming Tax Court's holding that Jason Chai's activities as an “accommodation party” for the Delta Group's tax advantaged investments constituted a “trade or business” because he engaged in them with a profit motive).

Although its analysis of the “business purpose” prong of Horn continues for more than six pages (see JA1941-48), the Tax Court did not actually examine Delta's business purpose for any of the trades at issue here disjunctively from their potential to generate direct gains, as Horn commands. To the contrary, it (a) first excluded the testimony of Delta's expert on whether the transactions made economic sense for reasons other than their potential to generate direct gains, finding that testimony “irrelevant” because it did not go to the question of whether those transactions could generate direct gains (see JA1930), and (b) then found that the transactions lacked a subjective business purpose because they could not generate such gains. (JA1947).

This ignores the disjunctive nature of the Horn analysis. Under Horn, a finding that a transaction lacks profit potential cannot be dispositive as to whether or not it has a business purpose — if it were, there would be no reason for the second prong of the analysis. See Horn, 968 F.2d at 1237 (“a transaction undertaken for a nontax business purpose will not be considered an economic sham even if there was no objectively reasonable possibility that the transaction could produce profits”) (emphasis in original). The Tax Court here essentially concluded that its finding that the transactions could not have produced a direct profit precluded any finding of business purpose. This was error as a matter of law.

On the second prong of Horn, the question before the Tax Court was not whether the trades could themselves have made a profit. Rather, the question was whether they were undertaken for any purpose other than to generate tax losses that would “reduce [Delta's] and Mr. Beer's own tax liabilities.” (See JA1946).³⁰ An intent to use the Partnerships and their built-in losses in order to profit in the overall business of creating derivatives for use in structured investment products is plainly such a purpose.³¹ The Tax Court's statement (JA1946) that the record contained “no evidence” of any such intent other than Beer's own testimony was plainly incorrect: the record contains numerous documents showing the products the Delta Group was developing (on which it intended to profit), as well as the corroborating testimony of its former General Counsel. (See supra at 30).

Taking the profit potential of the trades themselves out of the calculus (as we submit the Court must do in order to faithfully apply the Horn disjunctive analysis), the Tax Court's conclusion that the transactions were planned at the outset for the sole purpose of generating losses to offset Beer's income is based entirely on the fact that Beer ultimately used about one-third of the Partnerships' losses (that is, $40 million out of $144 million) for that purpose. As detailed above, the record contains none of the other evidence one would expect to see if the losses were planned for Beer's benefit. (See supra at 31-32). But the record is also clear that Beer used the losses only because of a drastic change in circumstances between the time the trades were placed (in late 2001) and the time the losses were used (in late 2002). In stark terms, what happened in the interim that caused Beer to receive the “benefit” of some of the losses in the Partnerships was the colossal failure of the most profitable arm of his business. After successfully expanding that business (and its payroll) for three years (see JA2116-9:329-42), in 2002 he suddenly lost everything he had invested. When Bricolage went into bankruptcy on October 15, 2005 (JA5067), he lost an additional $7 million that he personally put into Bricolage between 2002 and 2005 in an effort to keep it afloat. (JA2598). Use of some of the Partnerships' losses to offset income was the dramatic opposite of what he sought — he sought an ongoing income stream of tens of millions of dollars a year in fees, and instead got one year's limited tax savings.³² To suggest that Beer foresaw the failure of his business and was looking for losses (let alone planned any of this) in late 2001, when everything seemed to be on an upswing, is absurd.

We submit that the factual evidence at trial by itself — without consideration of the expert testimony — should have compelled the conclusion that the transactions were intended to produce fees and profits for the Delta Group and Beer rather than to generate losses for the sole purpose of reducing Beer's personal taxes. But we further submit that the Tax Court erred in excluding Delta's expert on this subject. That expert, Dr. Jack Yeager, is in the business of evaluating real-world business decisions for clients who pay him for his expertise. (JA2484-7:1041-53; JA2521:1093). He explained that this kind of business valuation is different from the valuation of a particular business opportunity standing alone such as the purchase or sale of a specific business or of securities: in the latter kind of valuation, the question is how much the asset or opportunity is worth, and the answer is a specific number; in contrast, in evaluating a decision the question is which alternative is preferable, and the answer depends upon how the alternatives compare with one another. The better opportunity will always be the one that has the greatest economic value. (See JA2519-20:1088-9).

Dr. Yeager performed a “real options analysis” to determine whether the actions of the Delta Group and Beer in 2001 with respect to the Partnerships were economically rational. (JA314; JA318; JA2486-7:1048-53; JA2521:1093). That analysis showed that at that time the best and most profitable potential use of the Partnerships was as a vehicle for new investments. As a result, from an analytical perspective the actions of the Delta Group and Beer in 2001 are best explained by an intent to use the Partnerships in this manner, not by an intent of Beer to take the losses for himself. (JA318-30).

The Tax Court excluded Dr. Yeager's testimony as “irrelevant” because he did not address the question of whether the trades themselves could have made a profit. (JA1930). That, however, was not the point of his testimony. Rather, the point was to show that — regardless of whether the trades themselves could have made a profit — it was economically rational for the Partnerships to engage in them for the purpose of otherwise benefitting Delta's overall business. This is not irrelevant; it goes directly to the second prong of Horn.

The Tax Court also expressed a view that Dr. Yeager “relied on many assumptions that had no support in the evidentiary record and are implausible on their face.” (JA1930). It is difficult to address this statement because the Tax Court gave no indication of which assumptions it viewed this way. We note, however, that virtually all of Dr. Yeager's assumptions were about Beer's expectations at the relevant time. In his testimony, Beer explained in detail both his expectations (which were the bases for Dr. Yeager's assumptions) and the reasons for them.³³

We emphasize that, because the facts so clearly show an overall profit motive, it should not even be necessary to consider the expert reports and testimony in evaluating whether the actions of the Delta Group and Beer in 2001 are better explained by such a motive or by a desire to confer tax benefits on Beer personally. But those reports and that testimony further confirm that a profit motive is the only explanation that makes sense. The IRS's expert (Dr. Finnerty) offered no basis for choosing between the two, while Dr. Yeager explained through professional analysis that profit motive is by far the better explanation. This goes to the heart of the business purpose analysis. To exclude it as “irrelevant” was an abuse of discretion. See Hite, supra, 769 F.3d at 1168, 1169-70 (where expert testimony went to an issue that was “relevant to the question of whether [the defendant] had the requisite intent,” District Court's decision to exclude it as having “little probative value” was an abuse of discretion).

With or without Dr. Yeager's analysis, the record supports only one conclusion: the Delta Group and Beer caused the Partnerships to enter into the FX transactions in November and December of 2001 for the purpose of positioning them for use in investment vehicles from which the Delta Group would substantially profit. The Tax Court's contrary finding amounts in essence to a another application of res ipsa loquitur: the mere fact that Beer ultimately used some of the losses in 2002 ($40 million out of $144 million) proves that the transactions were undertaken in 2001 for the purpose of providing tax losses for his benefit. Because that is not the only possible explanation for the conduct of Delta and the Partnerships, the law does not permit that kind of inference here. (See supra at 46-47 and n.25).

We note one more point. The fact that the new products on which Delta was working in 2001 and 2002 may have had tax-advantaged components does not and should not change the business purpose analysis. As detailed above, those products were designed in consultation with counsel and the accounting firms as a new generation of structured investments, and were specifically crafted to be more in line with the positions being taken publicly by the IRS. Whether or not they would have achieved that goal is something the Court cannot determine at this point — both because the designs of the investment products were never finalized due to the collapse of the market for tax-advantaged investment products in 2002 and because the question of whether deductions claimed by an unidentified future investor in a future tax year would be valid is beyond the jurisdiction of the Tax Court and this Court. Rather, the question before this Court on the issue of business purpose is whether or not the transactions were undertaken solely for the purpose of tax avoidance by the Partnerships or any of their then-current members. The answer is clear: neither the Partnerships nor any of their members, including Delta, sought to achieve anything through the transactions other than the creation of a valuable investment vehicle with respect to which the Delta Group could earn fees as part of its profit-making business. Delta was not looking for, and the Partnerships were not looking to create, tax losses for itself (or, for that matter, for anyone who had a current interest in the Partnerships). Rather, Delta was looking for, and the Partnerships sought to create, something that could generate a profit in connection with the Delta Group's usual business of creating derivatives for use in structured investment products. Inasmuch as this activity was not designed to create tax benefits for any person or entity in the Delta Group but rather designed to create income and profit, it constitutes a business purpose.³⁴

CONCLUSION

The fully-developed record supports only the conclusion that the FX Options had profit potential and/or were undertaken for a business purpose. As a result, under Horn they had economic substance. Accordingly, the Tax Court's ruling that they did not should be reversed.

Respectfully submitted,

Adrienne B. Koch
(akoch@katskykorins.com)
David L. Katsky
(dkatsky@katskykorins.com)
KATSKY KORINS LLP
605 Third Avenue
New York, New York 10158
(212) 953-6000

Attorneys for Petitioners-Appellants Delta Currency Trading, LLC, Cabrini Partners Fund, LLC, Alligator Partners Fund, LLC, Endeavor Partners Fund, LLC and Satellite Partners Fund, LLC

February 26, 2019