Trueex, LLC et al v. Markitserv Limited et al

Filing 58

MEMORANDUM OPINION re: 22 MOTION for Preliminary Injunction filed by Truepts, LLC, Trueex, LLC. The historic purpose of a preliminary injunction is to maintain the status quo pending a full trial where all facts and arguments may be considered with adequate preparation and in their complete context. As the foregoing demonstrates, this is a case in which full and expanded consideration is essential. At a minimum, there are serious questions going to the merits. In the circumstances, the balance of hardships tips decidedly in trueEX's favor. For those and all the reasons stated above, plaintiffs' motion for a preliminary injunction [DI 22] is granted with respect to trueEX. It is denied, however, with respect to truePTS, principally because truePTS has not shown any likelihood of success on its claims. It is important to recognize that the Court is dealing here only with a preliminary injunction the purpose of which is to preserve the status quo pending trial. The Court already has raised with the parties the question whether trueEX, were it to prevail, would be entitled to a permanent injunction requiring MarkitSERV to continue to deal with it in perpetuity in the manner it did prior to the termination notice, which would be a very different matter. trueEX recognizes that it would have to deal with that troublesome problem were it to win at trial. In all the circumstances, defendants Markit SERV Limited and MarkitSERV, LLC be and they hereby are enjoined and restrained, pending hearing and determination of the action, from directly or indirectly terminating the BTA or discontinuing the provision of drop-copy service to plaint iff trueEX, LLC. No bond is required because the Court finds that there is no likelihood of harm to defendants. The foregoing constitute the Court's findings of fact and conclusions of law. (Signed by Judge Lewis A. Kaplan on 7/18/2017) (mro)

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
------------------------------------------x
TRUEEX, LLC, et ano.,

7/18/2017

Plaintiffs,

-against-

17-cv-3400 (LAK)

MARKITSERV LIMITED, et ano.,
Defendants.
------------------------------------------x

MEMORANDUM OPINION

Appearances:

Daniel L. Brockett
Adam Abensohn
Thomas Lepri
Kanika Shah
QUINN EM ANUEL URQUHART & SULLIVAN, LLP
Attorneys for Plaintiffs

Colin Kass
Bradley I. Ruskin
Gregg M. Mashberg
Scott A. Eggers
David A. Munkittrick
Seth Fiur
PROSKAUER ROSE LLP
Attorneys for Defendants

2
LEWIS A. KAPLAN, District Judge.
This matter is before the Court on trueEX’s and truePTS’s (collectively, “plaintiffs”)
motion for a preliminary injunction to prevent MarkitSERV Limited and MarkitSERV, LLC
(collectively, “MarkitSERV”) from barring plaintiffs’ access to certain of MarkitSERV’s technology
and software.

Facts
This case centers around a business relationship between two entities operating in the
world of interest rate swaps (“IRS”), a type of financial derivative.

I.

The Swaps Here at Issue
A.

The Basics
The term “derivative,” as it is used in today’s financial world, refers to a financial

instrument that derives its value from the price of an underlying instrument or index. Among the
different types of derivatives are swaps, instruments whereby two counterparties agree to exchange
cash flows on two financial instruments over a specific period of time.1 These are (1) a “reference
obligation” or “underlying asset” such as a security, a bank loan, or an index, and (2) a benchmark
loan, generally with an interest rate set relative to a commonly used reference rate such as the
London Inter-Bank Offered Rate (“LIBOR”).2

1

CSX Corp. v. Children’s Inv. Fund Mgmt. (UK) LLP, 562 F. Supp. 2d 511, 519 (S.D.N.Y.
2008), aff’d in part, vacated in part, and remanded, 654 F.3d 276 (2d Cir. 2011).

2

Id.

3
B.

Interest Rate Swaps
An IRS is a particular form of swap. 3

Typically, it “is a transaction between two

counterparties in which one stream of future interest payments [on a notional debt obligation] is
exchanged for another” such stream on the same notional amount.4 Often, the counterparties in a
swap transaction are a corporation, a bank or an investor on one side (the “buy side”5) and an
investment or commercial bank, or other financial institution, on the other side (the “dealer”6).7
Additionally, there are many “dealer to dealer” trades that do not involve buy-side customers.
In the most common type of IRS swap, Counterparty A pays a fixed interest rate to
Counterparty B which, in return, pays a floating interest rate based on a benchmark such as LIBOR.8
For example, Counterparty A and Counterparty B enter into a five-year swap with the following
terms: Counterparty A agrees to pay Counterparty B an amount equal to 6 percent per annum on a
notional principal of $20 million, and Counterparty B agrees to pay Counterparty A an amount equal
to one-year LIBOR plus 1 percent per annum on the same notional principal amount. For simplicity,

3

Id.

4

Levy Decl. [DI 30] ¶ 5.
5

Other buy-side participants include “pension funds, insurance firms, asset management
funds, and hedge funds.” Hirani Decl. [DI 25] ¶ 7.
6

“The major dealers of IRS include all the large global banks, such as Goldman Sachs, JP
Morgan, Barc lays, and Citigroup. Swap dealers are designated ‘market makers’ in IRS,
meaning they make markets by providing quotes to buy-side market participants seeking
to trade IRS.” DI 25 ¶ 5.
7

DI 30 ¶ 5.
8

DI 25 ¶ 2.

4
let us assume the counterparties exchange payments annually on December 31, beginning in 2017
and concluding in 2021. At the end of 2017, Counterparty A will pay Counterparty B $1,200,000
(i.e., $20,000,000 x 6 percent). Let us assume further that on December 31, 2016, one-year LIBOR
was 5.33 percent. At the end of 2017, then, Counterparty B will pay Counterparty A $1,266,000
(i.e., $20,000,000 x (5.33 percent + 1 percent)). Nor are “fixed for floating” IRS the only varieties.
There are many others.

II.

The Impact of Dodd-Frank
Until the enactment of the Dodd-Frank Wall Street Reform and Consumer Protection

Act of 2010 (“Dodd-Frank”), IRS traded in unregulated over-the-counter (“OTC”) markets. Most
IRS were traded on a bilateral, principal-to-principal basis with the ultimate counterparty being the
entity with which the trade was executed. Many trades were intermediated by brokers. Dodd-Frank,
however, introduced new regulatory requirements with the goal of increasing transparency in the
OTC derivatives markets.

A.

Swap Execution Facilities
In Dodd-Frank, Congress mandated that certain IRS trade only on platforms called

swap execution facilities (“SEFs”).9 In response, many trade execution platforms sought to and did
register as SEFs in order to fill the new role created by Dodd-Frank. An SEF is a trading platform
regulated by the Commodity Futures Trading Commission that provides pre-trade information
(i.e., bids and offers) and an execution mechanism for swap transactions. A buy-side participant can

9

DI 30 ¶ 6.

5
use an SEF to transmit to multiple dealers on the platform requests for quotes of offers to sell IRS
having particular terms.

If the buy-side participant finds a willing dealer, the SEF provides a

mechanism for the parties to execute their trade. There are three major SEFs that deal in IRS:
TradeWeb, Bloomberg, and trueEX.10
Dodd-Frank requires that IRS that must be traded on SEFs be “cleared,”
i.e., “submitted to a central counterparty clearinghouse that functions as an intermediary between
buyer and seller to reconcile transactions and reduce risk.”11

In a cleared transaction, a

clearinghouse steps between the counterparties—effectively becoming the buyer to the original
seller and the seller to the original buyer. It processes the transaction, guarantees completion, and
remains a part of the trade throughout its life cycle. 1 2 In the United States, the major clearinghouses
are CME Group and London Clearing House.
Dodd- Frank requires also that SEFs report trade pricing and volume information to
swap data repositories (“SDRs”), which provide central facilities for swap data reporting and
recordkeeping.13 SDRs enable market participants to see trading data for all executed trades, which
promotes transparency in the market. The largest SDR in the United States is the Depository Trust
and Clearing Corporation (“DTCC”).

10

DI 25 ¶ 8; see also Levy Decl., Ex. A [DI 30-1].
11

DI 30 ¶ 7.
“A substantial volume of IRS trades are not executed on SEFs.” DI 30 ¶ 8. For purposes
of this motion, however, we are concerned only with cleared IRS trades executed on SEFs.
12

DI 25 ¶ 9; DI 30 ¶ 7.
13

DI 30 ¶ 9.

6
B.

Trade Processing
After two counterparties agree upon an IRS trade that is required to be cleared, details

about that trade must be reported to four entities in compliance with Dodd-Frank: the clearinghouse
that clears the trade, the two counterparties to the trade, and the SDR. Completion of that reporting
is a key aspect of a process known as trade processing.14 In addition to reporting IRS trades to
clearinghouses and SDRs, trade processing encompasses such tasks as “matching . . . buyer and
seller records, confirming the terms of trades, allocating aggregated trades among a client’s different
sub-accounts, and managing other life-cycle events such as trade amendments, assignments, and
payments.”15 The large majority of IRS post-trade processing is handled by MarkitSERV.

III.

MarkitSERV
MarkitSERV is an electronic trade confirmation network for OTC derivatives.16 It

provides trade processing for “OTC derivative transactions across all major asset classes, including
credit, equity, foreign exchange and interest rates products, including [IRS].”17 MarkitSERV does
not, however, offer trade execution services.18 Rather, parties execute OTC derivative trades on
other platforms, such as SEFs, and those platforms in turn send the transaction details to

14

DI 25 ¶ 12.
15

DI 25 ¶ 13.
16

DI 30 ¶ 3.
17

DI 30 ¶ 4.
18

See DI 30 ¶ 22.

7
MarkitSERV for trade processing.19 MarkitSERV’s clients include “asset managers, hedge funds,
pension funds, fund administrators, dealers and inter-dealer brokers, prime brokers and futures
commissions merchants,” as well as clearinghouses, six reporting agencies, and thirteen SEFs.20

A.

Formation of MarkitSERV
MarkitSERV’s predecessor, SwapsWire Limited, was formed in 2000 by a

consortium of leading dealer banks as an electronic trade confirmation network for the OTC
derivatives market. 2 1

The dealers made significant investments in the network infrastructure to

enable their systems to communicate with SwapsWire. In 2008, IHS Markit (“Markit”), a global
information and services company, acquired SwapsWire and renamed it MarkitWire.22 In 2009,
Markit and DTCC “combined MarkitWire and DTCC’s Deriv/SERV business to form a joint
venture called MarkitSERV to provide OTC derivative trade processing.”23 Four years later, Markit
purchased DTCC’s share of the joint venture.

B.

MarkitSERV’s Dominant Market Position
MarkitSERV is the sole provider of IRS trade processing for (1) IRS transactions

19

See, e.g., DI 25 ¶ 21.
20

DI 30 ¶ 4.
21

DI 25 ¶ 19; DI 30 ¶ 3.
22

DI 25 ¶ 20.
23

DI 30 ¶ 3.

8
between dealers, (2) cleared, direct trades that are not required to be traded on SEFs, and (3)
uncleared swaps (a limited category of IRS that need not be cleared).24

Together, those three

categories “make up approximately 92% of the overall IRS market, based on the notional values of
trades.”25
MarkitSERV concedes for this motion only that IRS trade processing is a relevant
market for antitrust purposes.26 Its market share is depicted in Figure 1.
Figure 1

The only transactions for which MarkitSERV does not control every step of post-trade processing

24

DI 25 ¶¶ 21-22.
25

DI 25 ¶ 23.
26

Prelim. Inj. Hr’g Tr. 36:1-14, June 6, 2017.

9
are trades between dealers and buy-side customers that are executed on the SEFs operated by
Bloomberg, TradeWeb, and trueEX.27 Even for trades executed on SEFs, however, most market
participants rely on MarkitSERV’s trade processing method (discussed below) to ensure that their
books and records are accurate and up-to-date.

C.

How MarkitSERV Processes Trades
When customers choose MarkitSERV to perform trade processing, MarkitSERV

“will simultaneously submit the trade to the clearinghouse for intermediation, to the two
counterparties[,] and to the SDR using an automated workflow called ‘straight-through-processing’
[(“STP”)].”28

In order to process trades using STP, MarkitSERV has established direct

communication lines to and application program interfaces (“APIs”)29 with its customers to facilitate
“real-time updates of market participants’ books and records.”30

STP ensures that market

participants’ books and records “reflect current financial and risk exposures.”31 For that reason,

27

DI 25 ¶ 24.
28

DI 25 ¶¶ 16-17.
29

APIs “allow a programer writing an application to instruct an operating system or another
program to carry out specific tasks. In essence, [APIs] act as building blocks. When
programers have the [APIs], they can create software to complete tasks through the program
providing the interfaces by putting the interfaces together. If programers do not have the
[APIs] they must completely recreate instructions for the tasks with original code.” Lantec,
Inc. v. Novell, Inc., 306 F.3d 1003, 1008 n.3 (10th Cir. 2002) (citations omitted).
30

DI 25 ¶ 16; DI 30 ¶ 34.
31

DI 25 ¶ 16.

10
“STP facilitation is vital to dealers and buy-side clients.”32 Over the fourteen years MarkitSERV
has existed, it has set up some manner of direct connection to “hundreds of dealers and close to
2,000 buy side customers.”33
In addition to its own trade-processing services, MarkitSERV facilitates SEFs’ trade
processing through a “drop copy” workflow. For example, if a customer executes a trade on an SEF,
it in some circumstances may elect to have the SEF process that trade too. When the customer
chooses to use the SEF for trade processing, the SEF “delivers the trade details directly to the
clearinghouse and SDR, and simultaneously gives (or ‘drops’) a copy of the trade to MarkitSERV,
which then uses its network to redistribute the trade confirmation to the counterparties.”34 Utilizing
MarkitSERV’s drop-copy service allows the SEF to process trades executed on its platform using
STP without having to invest in its own direct communication lines—at least with respect to
customers that already are connected to MarkitSERV’s network.

IV.

trueEX’s Relationship with MarkitSERV
A.

The Broker Terms Agreement
As discussed above, trueEX is an SEF for IRS trades that has been active since

April 2014. In addition to trade execution, trueEX provides certain post-trade processing services
to its clients. Like the other IRS SEFs, however, trueEX does not have direct communication lines
to all of the entities that trade on its platform. Thus, in order to process trades using STP for clients
32

DI 25 ¶ 28.
33

DI 30 ¶ 34.
34

DI 30 ¶ 14.

11
with which it does not have a direct connection, trueEX requires electronic access to those clients’
books and records through other means.
trueEX approached MarkitSERV to solve this STP connectivity problem. In due
course, the parties entered into a contract, referred to as the Broker Terms Agreement (“BTA”),35
in which MarkitSERV agreed, among other things, to provide drop-copy service to trueEX. trueEX
utilizes MarkitSERV’s drop-copy workflow in a similar manner to the other SEFs, TradeWeb and
Bloomberg—when two counterparties execute a trade on trueEX, trueEX submits the trade details
directly to the clearinghouse and SDR and simultaneously drops a copy of the trade to MarkitSERV
for redistribution to the counterparties. Unlike TradeWeb and Bloomberg, however, trueEX does
not give its customers a choice in terms of trade processing. In other words, any customer that uses
trueEX to execute its IRS trades also must use trueEX to process those trades.36
In cases where trueEX has built a communication pipeline to and has an API with
a trader, trueEX does not need to rely on MarkitSERV’s network for STP.37 “Only if there is no
pipeline to the trader does [t]rueEX rely on MarkitSERV’s network to fill that gap and provide a
‘drop copy’ to the customer.”38

35

DI 25-1.
36

DI 30 ¶ 15.
37

DI 30 ¶ 24.
38

DI 30 ¶ 24.

12
B.

trueEX’s Development of truePTS
In December 2015, trueEX announced that it was working on a new IRS trade

processing business.39

trueEX’s chief executive officer, Sunil Hirani, launched this new

venture—truePTS—in June 2016.40 truePTS “seeks to process trades other than, or in addition to,
trades executed on the [t]rueEX SEF,” including trades that are not permitted by law to be executed
on SEFs.41 According to Mr. Hirani, “truePTS will offer the same IRS trade processing services as
MarkitSERV, but at a substantially lower rate and with faster and more innovative technology.”42
Between August 2016 and April 2017, truePTS met with thirty-five different participants in IRS
trading to discuss truePTS’s purpose, functionality, and competitive advantage over MarkitSERV.
truePTS now “is still in the developmental stage.”43 It is a sister company of trueEX
and is “entirely supported by” it.44 trueEX provides truePTS with all of its funding, staffing, and
resources.

According to plaintiffs, “truePTS can and will develop its own network [for STP

facilitation].”45 But truePTS would have to rely on MarkitSERV’s drop-copy workflow to process

39

DI 30 ¶ 35.
40

DI 25 ¶ 37.
41

DI 30 ¶ 37.
42

DI 25 ¶ 38.
43

DI 25 ¶ 37.
44

DI 25 ¶ 37.
45

See Pls.’ Br. [DI 23] at 3.
truePTS contends also that even once it has developed its own network, it still will require

13
its customers’ trades until it does. Accordingly, truePTS sought to enter into an agreement with
MarkitSERV for STP facilitation on similar terms as trueEX.46

In January 2017, truePTS sent

MarkitSERV a proposal outlining what such a relationship might look like.47
months, truePTS repeatedly inquired as to the status of the proposal.

For the next two

Although MarkitSERV

informed truePTS that it was reviewing the request internally, MarkitSERV never provided a
definitive response.48

C.

MarkitSERV Terminates the BTA
Meanwhile, MarkitSERV concluded that trueEX “was seeking a fundamental shift

in the nature of [their] relationship in order to facilitate the development of its new business,”
truePTS.49 In MarkitSERV’s view, trueEX “sought to create a situation in which [t]ruePTS, acting
through [t]rueEX’s relationship with MarkitSERV, could use MarkitSERV’s processing network
to offer its own, competing trade processing service”—without making the financial and technical
expenditures MarkitSERV did.50

MarkitSERV feared that truePTS thus could undercut

drop-copy functionality from MarkitSERV “so it can ‘talk’ to MarkitSERV for clients
that . . . rely on MarkitSERV’s STP for their books and records.” DI 25 ¶ 62.
46

DI 25 ¶ 40.
47

DI 25 ¶ 40.
48

Brockett Decl., Ex. 4 [DI 40-4] at 1.
49

DI 30 ¶ 42.
50

DI 30 ¶ 42.

14
MarkitSERV’s pricing and appeal to customers by offering a lower price point.51 To prevent trueEX
from “leverag[ing] its relationship with MarkitSERV for the benefit of its new competing
business”—and, in turn, to thwart the competitive threat posed by truePTS—MarkitSERV notified
trueEX that it was terminating the BTA effective May 14, 2017.52
Needless to say, plaintiffs dispute MarkitSERV’s characterization of the events
leading to the termination of the BTA. In their view, MarkitSERV’s termination was “a transparent
effort to choke off . . . competition and preserve its monopoly.”53

In any event, the parties

negotiated over the next several weeks the future of their business relationship. During one of those
sessions, MarkitSERV proposed new terms for its relationship with trueEX: MarkitSERV would
give trueEX the option either to use trueEX’s own network for processing a trade or have the trade
submitted to MarkitSERV’s “standard workflow.”54

In other words, for trades involving

counterparties with which trueEX has direct connections, trueEX would be able to process the trade.
In contrast, for trades involving counterparties with which trueEX does not have direct connections,
trueEX could execute the trade but MarkitSERV would be the one to process it.55 trueEX was not

51

See DI 30 ¶ 39.
52

Hirani Decl., Ex. 2 [DI 25-2].
53

DI 23 at 1.
54

DI 30 ¶ 49; DI 25 ¶ 47.
55

DI 30 ¶ 49.
Certain buy-side counterparties do not require STP facilitation. See DI 30 ¶ 27. For trades
involving such counterparties, trueEX would not need a direc t c onnec tion to the buy-side
c ustomer in order to process trades involving them, assuming the dealer counterparty is
directly connected to the trueEX network.

15
interested in that offer.
During another meeting, according to Mr. Hirani, MarkitSERV’s chief executive
officer, Brad Levy, assured trueEX that it would not cut off trueEX’s STP connectivity on May 14,
2017, the date the termination of the BTA was to go into effect.56 However, it became clear on
May 5, 2017, that MarkitSERV would agree to provide trueEX STP connectivity only on a day-byday basis.57 According to Mr. Hirani, “Mr. Levy made it clear that MarkitSERV would not rescind
its termination of the” BTA.58
After the parties were unable to reach a new agreement, MarkitSERV, on May 8,
2017, “advised approximately 19 customers that [it was] terminating [its] relationship with [t]rueEX
as of May 15”59—which, according to Mr. Hirani, “sow[ed] confusion and uncertainty” among
trueEX’s clients about the possible loss of STP connectivity.60

D.

Threatened Impact of Termination on Plaintiffs
1.

trueEX

trueEX asserts that it “cannot survive as a business without connectivity to

56

DI 25 ¶ 49.
57

DI 25 ¶ 51.
58

DI 25 ¶ 51.
59

DI 30 ¶ 47.
60

DI 25 ¶¶ 56-60.

16
MarkitSERV’s network.”61 It claims that any dealer or buy- side client that uses MarkitSERV for
STP will leave the trueEX platform if trueEX loses the ability to send drop-copy reports to
MarkitSERV. Moreover, it contends, “the loss of drop-copy would . . . cause the dealers that
provide liquidity to the trueEX platform to abandon trueEX . . . . An exodus of dealers would drain
all liquidity from trueEX. And without that liquidity, trueEX simply could not function as a trading
platform.”62

According to Mr. Hirani, “[e]xisting dealers have already indicated that they are

rethinking their relationship with trueEX.”63
The actual extent of trueEX’s reliance on MarkitSERV for STP facilitation is hotly
contested by the parties. According to MarkitSERV, “only 20 customers—twelve dealers and eight
buy side customers used MarkitSERV’s trade processing network for [t]rueEX-executed trades in
2017.[64] And only 21 customers—13 dealers and 8 buy side firms—did so in 2016 . . . .”65 Based
on data reported for 2016, MarkitSERV estimates that 58 percent of the notional value of all trueEX

61

DI 25 ¶ 54.
62

DI 25 ¶ 55.
63

DI 25 ¶ 56.
For example, Mr. Hirani says trueEX received the following email from a dealer:
“I just received a call from [MarkitSERV] that the relationship between TrueEX and
[MarkitSERV] has dissolved . . . and they will no longer support the drop copy
deals. . . . [I]f our connection is not setup and we do not have [MarkitSERV], this will break
our STP and I will need to discuss . . . options for future activity.” DI 25 ¶ 56.
64

MarkitSERV contends that “[t]wo of the dealers use MarkitSERV for some of their
[t]rueEX-based trades, but also have a direct connection to [t]rueEX, and so could easily
bypass MarkitSERV for each of their trades if they chose to do so.” DI 30 at 10 n.6.
65

DI 30 ¶ 26.

17
trades in U.S. dollar-denominated IRS bypassed MarkitSERV entirely.

In MarkitSERV’s view,

then, the data demonstrate that trueEX “conducted a significant portion of [its] business in 2016
without any MarkitSERV involvement, [which] contradicts [t]rueEX’s position that MarkitSERV
services ‘nearly every transaction in the IRS market.’”66
Additionally, MarkitSERV asserts that “[i]n insisting that it is dependent on
MarkitSERV’s network, [t]rueEX both understates the investment MarkitSERV has made in
building its communications network and overstates the burden [trueEX] faces in building
connectivity to the remaining dealers for whom it uses MarkitSERV’s drop copy function.”67
MarkitSERV or its predecessors allegedly “have spent hundreds of millions of dollars to install,
maintain and update” its communications network.68 As for completing connectivity, MarkitSERV
contends that if trueEX were to “build connections to the 10 dealers for whom it currently relies
upon MarkitSERV for drop copy services, its network would cover all of its dealer customers and
90% of its buy side customers.”69 In MarkitSERV’s experience, it claims, the process of onboarding
customers and setting up a direct communication line to the customer’s facility “takes approximately
eight to twelve weeks.”70

Thus, in a matter of weeks, MarkitSERV suggests, trueEX could

66

DI 30 ¶ 31 (quoting DI 25 ¶ 25).
67

DI 30 ¶ 32.
68

DI 30 ¶ 33.
69

DI 30 ¶ 28.
70

DI 30 ¶ 34.

18
“completely displace MarkitSERV from the [t]rueEX ecosystem.”71
trueEX claims that MarkitSERV “misrepresents and understates the importance of
drop-copy to trueEX’s business.”72 According to trueEX, it presently has
“105 clients that have agreements in place to trade on trueEX’s platform: 24 dealers
and 81 buy-side clients. With respect to dealers:
“a.

Of the 24 dealers that have signed up with trueEX, all but three, or
88%, require MarkitSERV connectivity for transactions in at least
one currency.

“b.

Of those 24 dealers, 22 have been ‘onboarded’ to the trueEX
platform, and 14 have traded in 2017. All but three of those 14
dealers, or 79%, have engaged in trades that required a drop-copy to
MarkitSERV’s network.

“With respect to trueEX’s buy-side clients:
“a.

Of the 81 buy- side participants who have signed up with trueEX, 41,
or 51%, require MarkitSERV connectivity for transactions in at least
one currency.

“b.

Of those 81 buy- side clients, there are 71 that have been onboarded
to the trueEX platform, and 22 that have traded on the trueEX
platform in 2017. Nine of those 22 clients, or 41%, have engaged in
trades that required a drop-copy to MarkitSERV’s network.”73

Thus, in trueEX’s view, the data strongly support the conclusion that trueEX cannot survive without
drop-copy service because the majority of its clients rely on MarkitSERV for STP. Mr. Hirani
asserts too that even if some percentage of its buy-side clients do not use MarkitSERV for STP,
“[t]rading on a SEF that does not have the ability to provide drop-copy functionality
for trades to MarkitSERV’s network is a non-starter for an enormous number of
71

DI 29 at 3.
72

Hirani Reply Decl. [DI 41] ¶ 18.
73

DI 41 ¶¶ 19-20 (footnotes omitted).

19
entities that trade IRS, including virtually the entire community of dealers (one of
which [is] on at least one side of every trade). Regardless of whether it needs
MarkitSERV’s STP to maintain its own books and records, no rational buy-side
participant would choose to trade on a SEF that is completely cut off from the vast
majority of the IRS market, and therefore lacks the liquidity and efficiency that
participants expect and demand from a SEF.”74
As for MarkitSERV’s suggestion that trueEX can simply finish building out its STP
network, trueEX maintains that “establishing STP connectivity with a client is a complex and
burdensome process.”75 In trueEX’s experience, it claims, the process “on average takes 17 months
per participant”—quite a bit longer than MarkitSERV suggests.76 Thus, it would be unrealistic to
think trueEX could “flip a switch and build an STP network connecting its entire client base . . . in
a matter of weeks.”77

2.

truePTS

The threat to truePTS is far simpler. Mr. Hirani contends that “MarkitSERV’s refusal
to provide STP connectivity to truePTS will kill truePTS before it gets off the ground.”78 If trueEX
shuts down, truePTS will fail as well because it “relies on trueEX for all of its staffing, support, and

74

DI 41 ¶ 24.
75

DI 41 ¶ 26.
76

DI 41 ¶ 31.
77

DI 41 ¶ 32.
78

DI 25 ¶ 62.

20
funding.”79

V.

The Present Litigation
Plaintiffs filed this action in response to MarkitSERV’s May 8 announcement.

trueEX and truePTS each assert claims for monopolization and attempted monopolization under
Section 2 of the Sherman Act, while trueEX alone asserts also a claim for promissory estoppel under
New York law.80 The Court has set an expedited schedule for the matter, with a trial on the merits
to take place in March 2018.
Pursuant to a standstill agreement the parties entered into on May 10, 2017, the
effective date of the BTA is tolled until July 24, 2017.81 Once the agreement is terminated, however,
MarkitSERV will be free to cut off trueEX’s access to its STP network. Plaintiffs have moved the
Court for a preliminary injunction that would preserve the status quo—i.e., the BTA would remain
in effect until trial. Plaintiffs contend that, without the Court’s intervention, trueEX will go out of
business before it has its day in court, a harm for which it could not adequately be compensated.
And without trueEX’s support, truePTS “will die on the vine before it can become a viable
business.”82

79

DI 25 ¶ 63.
80

See DI 1.
81

DI 25 ¶ 53.
The parties did not provide the Court with a copy of the standstill agreement. The Court
relies instead on the parties’ representations concerning the agreement’s terms.
82

DI 25 ¶ 63.

21
Discussion
Section 16 of the Clayton Act provides that a party may obtain injunctive relief
“against threatened loss or damage by a violation of the antitrust laws.”83

For a preliminary

injunction to issue, the movant must establish “(1) ‘irreparable harm’; (2) ‘either (a) a likelihood of
success on the merits, or (b) sufficiently serious questions going to the merits of its claims to make
them fair ground for litigation, plus a balance of the hardships tipping decidedly in favor of the
moving party’; and (3) ‘that a preliminary injunction is in the public interest.’”84

I.

Likelihood of Success on the Merits
A.

The Monopolization Claim85
Section 2 of the Sherman Act provides that it is unlawful to “monopolize, or attempt

to monopolize . . . any part of the trade or commerce among the several States, or with foreign
nations.”86 “[T]his offense requires, in addition to the possession of monopoly power in the relevant
market, ‘the willful acquisition or maintenance of that power as distinguished from growth or

83

15 U.S.C. § 26.
84

N.Y. ex rel. Schneiderman v. Actavis PLC, 787 F.3d 638, 650 (2d Cir.), cert. dismissed sub
nom. Allergan PLC v. N.Y. ex. rel. Schneiderman, 136 S. Ct. 581 (2015).
85

Plaintiffs pursue also claims for attempted monopolization. As they note in their opening
brief, the elements of attempted monopolization “essentially track those required for a
successful monopolization claim[. Hence,] a finding of . . . monopolization will also lead
to a finding of attempted monopolization.” DI 23 at 13-14 n.6 (internal quotation marks
omitted). Accordingly, the Court assesses the merits only of the monopolization claims.
86

15 U.S.C. § 2.

22
development as a consequence of a superior product, business acumen, or historic accident.’”87
Importantly, “the possession of monopoly power will not be found unlawful unless it is
accompanied by an element of anticompetitive conduct.”88 “[A]nticompetitive conduct is ‘conduct
without a legitimate business purpose that makes sense only because it eliminates competition.’”89
For purposes of this motion, MarkitSERV concedes that IRS post-trade processing
is a relevant market and that it has monopoly power in that market.90 Accordingly, the Court turns
to the question of whether plaintiffs are likely to prevail on their claim that MarkitSERV engaged
in anticompetitive conduct to maintain that monopoly power.

Specifically, plaintiffs assert that

MarkitSERV violated Section 2 in two ways: (1) it refused to deal with them and (2) it denied them
access to an essential facility—i.e., MarkitSERV’s STP network.

1.

Refusal to Deal

“[A]s a general matter, the Sherman Act ‘does not restrict the long recognized right
of [a] trader or manufacturer engaged in an entirely private business, freely to exercise his own
independent discretion as to parties with whom he will deal.’”91 However, “[t]he high value that we

87

Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004)
(quoting United States v. Grinnell Corp., 384 U.S. 563, 570-71 (1966)).
88

Id. (emphasis omitted).
89

Port Dock & Stone Corp. v. Oldcastle Ne., Inc., 507 F.3d 117, 124 (2d Cir. 2007) (quoting
Morris Commc’ns Corp. v. PGA Tour, Inc., 364 F.3d 1288, 1295 (11th Cir. 2004)).
90

Hr’g Tr. 36.
91

Trinko, 540 U.S. at 408 (quoting United States v. Colgate & Co., 250 U.S. 300, 307 (1919)).

23
have placed on the right to refuse to deal with other firms does not mean that the right is
unqualified.”92

“Under certain circumstances, a refusal to cooperate with rivals can constitute

anticompetitive conduct and violate § 2.”93 The Supreme Court, though, has been “very cautious
in recognizing such exceptions [to the right to refuse to deal], because of the uncertain virtue of
forced sharing and the difficulty of identifying and remedying anticompetitive conduct by a single
firm.”94 The Second Circuit has characterized that caution as meaning that “the sole exception to
the broad right of a firm to refuse to deal with its competitors comes into play only when a
monopolist seeks to terminate a prior (voluntary) course of dealing with a competitor.”95
That exception is rooted in two Supreme Court cases: Aspen Skiing Co. v . Aspen
Highlands Skiing Corp. and Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP.
In Trinko, the Supreme Court summarized Aspen Skiing, which it called the “leading case for § 2
liability based on refusal to cooperate with a rival,” as follows:
“The Aspen ski area consisted of four mountain areas. The defendant, who owned
three of those areas, and the plaintiff, who owned the fourth, had cooperated for
years in the issuance of a joint, multiple- day, all-area ski ticket. After repeatedly
demanding an increased share of the proceeds, the defendant canceled the joint
ticket. The plaintiff, concerned that skiers would bypass its mountain without some
joint offering, tried a variety of increasingly desperate measures to re-create the joint
ticket, even to the point of in effect offering to buy the defendant’s tickets at retail
price. The defendant refused even that. We upheld a jury verdict for the plaintiff,

92

Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 601 (1985).
93

Trinko, 540 U.S. at 408.
94

Id.
95

In re Adderall XR Antitrust Litig., 754 F.3d 128, 134 (2d Cir. 2014) (internal quotation
marks omitted) (quoting In re Elevator Antitrust Litig., 502 F.3d 47, 52, 53 (2d Cir. 2007)
(per curiam)).

24
reasoning that ‘[t]he jury may well have concluded that [the defendant] elected to
forgo these short-run benefits because it was more interested in reducing
competition . . . over the long run by harming its smaller competitor.’”96
The Trinko Court noted, however, that “Aspen Sk iing is at or near the outer boundary of § 2
liability.”97
While Trinko “does not purport to set out a ‘test,’ it usefully highlights the
distinctions that made Aspen Skiing the rare case in which a refusal to deal amounted to a prohibited
act of unilateral monopolization.”98 First, the monopolist and plaintiff-competitor in Aspen Skiing
had a preexisting “voluntary (and thus presumably profitable) course of dealing.”99

Second, the

prior course of dealing had “originated in a competitive market and had persisted for several
years.”100

Third, the monopolist’s termination of the prior course of dealing “suggested a

willingness to forsake short-term profits to achieve an anticompetitive end.”101

Finally, “the

defendant’s unwillingness to renew the ticket even if compensated at retail price revealed a
distinctly anticompetitive bent.”102

Taken together, these facts led the jury to conclude that the

96

Trinko, 540 U.S. at 408.
97

Id. at 409.
98

Adderall, 754 F.3d at 135.
99

Trinko, 540 U.S. at 409.
100

Aspen Skiing, 472 U.S. at 603.
101

Trinko, 540 U.S. at 409.
102

Id. (emphasis in original).

25
defendant’s refusal to deal with its rival was “irrational but for its anticompetitive effect.”103 The
question for this Court to decide is whether the present case fits “within the limited exception
recognized in Aspen Skiing.”104
At the outset, it is plain that truePTS on the present record is unlikely to succeed on
its refusal to deal claim because it never has had any business relationship with MarkitSERV.
trueEX, however, stands in different shoes. Several key features of Aspen Skiing are present in its
relationship with MarkitSERV.105
As in Aspen Skiing, MarkitSERV voluntarily has dealt with trueEX for several years.
MarkitSERV nevertheless contends that “it is unlikely [it] made a profit at all after factoring the
time and investment costs associated with providing drop copy service for [t]rueEX trades.”106
MarkitSERV admits, however, that it has not done a “comprehensive calculation of the cost . . . of
providing drop copy services to [t]rueEX.”107

Without more, the Court declines to credit

MarkitSERV’s speculation. Hence, despite MarkitSERV’s quibble about the profitability for it of

103

Nov ell, Inc. v. Microsoft Corp., 731 F.3d 1064, 1075 (10th Cir. 2013); see also Aspen
Skiing, 472 U.S. at 604 (“Since the jury was unambiguously instructed that Ski Co.’s refusal
to deal with Highlands ‘does not violate Section 2 if valid business reasons exist for that
refusal,’ we must assume that the jury concluded that there were no valid business reasons
for the refusal.”).
104

Trinko, 540 U.S. at 409.
105

As plaintiffs’ motion treats truePTS and trueEX as separate entities despite the fact that they
are “sister companies” that share the same staff, the Court proceeds on that assumption. It
offers no view as to whether the relationship between them is such that they should be
treated as a single entity for antitrust purposes.
106

DI 30 ¶ 20.
107

DI 30 ¶ 20.

26
that relationship, the Court finds that it has been profitable for MarkitSERV to at least some extent,
else it would not have continued the relationship.108 Moreover, trueEX offered MarkitSERV
additional money to continue to provide the same services it had been providing under the BTA, but
MarkitSERV “dismissed that proposal out of hand” and claimed that the termination “ha[d] nothing
to do with economics.’”1 0 9 This Court finds, for purposes of this motion, that MarkitSERV was
motivated at least in material part by a desire to foreclose a competitive challenge to its dominant
market position. This is not to say, however, that trueEX is in a position identical in all respects to
that of the Aspen Skiing plaintiff. Three distinctions stand out at first blush.
First, in Aspen Skiing, the joint ski ticket initially was introduced “when three
independent companies operated three different ski mountains” and “continued to provide a
desirable option for skiers when the market was enlarged to four mountains.”110 It was only after
the defendant acquired monopoly power—and no longer needed cooperation of its rival in order to

108

The Court is obliged by Rule 52(a)(2) to make findings of fact in deciding a motion for an
interlocutory injunction. Such findings are provisional in the sense that they are not binding
on a motion for summary judgment or at trial and are subject to change as the litigation
progresses.
It is well to mention also that the parties explicitly waived an evidentiary hearing on this
motion [see DI 43; DI 45] and thus consented to the making of findings entirely on the
paper record before the Court. See, e.g., Consol. Gold Fields PLC v. Minorco, S.A., 871
F.2d 252, 256 & n.3 (2d Cir. 1989); Fengler v. Numismatic Americana, Inc., 832 F.2d 745,
748 (2d Cir. 1987). Review of such findings remains subject to the clearly erroneous
standard. E.g., Anderson v. City of Bessemer City, 470 U.S. 564, 574-75 (1985); 9 James
Wm. Moore et al., Moore’s Federal Practice § 52.31[6][a] (3d ed. 2017).
109

DI 41 ¶ 12; DI 25 ¶ 51.
110

Aspen Skiing, 472 U.S. at 603.

27
compete—that it decided to terminate the popular joint ticket. 1 1 1

The Supreme Court found

significance in the fact that the parties’ cooperative relationship originated in a competitive market
because the defendant’s decision to discontinue the popular joint ticket once it reached a dominant
position—but not sooner—suggested an intention to exclude the plaintiff from the Aspen market
“on some basis other than efficiency.”112
Here, in contrast to Aspen Skiing, the relationship between MarkitSERV and trueEX
did not “originate in a competitive market.”113 Rather, it began at a time when MarkitSERV was,
and always had been, the dominant player in the market. MarkitSERV had no need to cooperate
with trueEX or any other SEF in order to compete in IRS post-trade processing.
Although that fact distinguishes this case from Aspen Skiing, it is not legally
significant, at least on the present record. The genesis of the cooperative relationship at issue in a
refusal to deal case—i.e., whether it originated in a competitive or monopolistic market—has
evidentiary significance, if at all, only where it has probative value with respect to the existence or
absence of anticompetitive motives. That is to say, in general, a new monopolist’s termination of
a cooperative relationship that originated in a competitive market is more likely to suggest
anticompetitive motives than a long-time monopolist’s termination of a relationship that arose in
a monopolistic market. But where, as here, evidence shows that the monopolist was motivated at
least in material part by anticompetitive desires, any countervailing inference the Court might have
drawn from the fact that the cooperative relationship originated in a monopolistic market would be
111

Id. at 603-04.
112

Id. at 604.
113

Id. at 603.

28
improbable.
Second, whereas the monopolist’s termination of the joint ski ticket in Aspen Skiing
suggested a willingness to sacrifice short-term profits, MarkitSERV’s termination of the BTA is
more equivocal on the present record. To be sure, MarkitSERV stands to lose the fees trueEX pays
for drop-copy service. But it perhaps stands also to gain business from trueEX customers that might
migrate away from trueEX and to MarkitSERV for trade processing because of its STP network.114
In other words, MarkitSERV’s decision to terminate the BTA may be consistent also with “a desire
to maximize the company’s immediate and overall profits.”115 But the record on this point remains
to be developed.
Third, unlike the monopolist in Aspen Skiing, MarkitSERV asserts that it had at least
one legitimate business reason for discontinuing the provision of drop-copy service to
trueEX—preventing trueEX from permitting truePTS to access MarkitSERV’s network without
MarkitSERV’s permission.116 It characterizes its action as being motivated by a desire to protect
itself from “free-riding,” which in proper circumstances may be an appropriate motive.117 But the
term “free-riding” fundamentally is a pejorative characterization that is applied to one’s helping

114

See DI 30 ¶ 51; Novell, 731 F.3d at 1076-78; MetroNet Servs. Corp. v. Qwest Corp.,
383 F.3d 1124, 1132 (9th Cir. 2004).
115

Novell, 731 F.3d at 1076.
116

MarkitSERV advances other business reasons for its decision to terminate, which the Court
on the present record finds are pretextual. See DI 23 at 18 n.8; DI 38 at 4-10.
117

Morris, 364 F.3d at 1295 (“[R]efusal to deal that is designed to protect or further the
legitimate business purposes of a defendant does not violate the antitrust laws, even if that
refusal injures competition.”); id. (“PGA met its business justification burden by showing
that it seeks to prevent Morris from ‘free-riding’ on PGA’s RTSS technology.”).

29
one’s self to a service provided by another without paying for it. In the circumstances here, it
presupposes that the BTA—pursuant to which MarkitSERV has been providing services, including
the drop-copy service, to trueEX for several years—forecloses trueEX from providing that access
to its sister company, truePTS. If the BTA does not do so, then MarkitSERV’s claim that trueEX
is facilitating truePTS’s “free-riding” on MarkitSERV’s network is empty rhetoric, as trueEX pays
for that service. So the Court turns to that issue.
During the argument of the motion, the Court asked counsel for MarkitSERV to
identify the provision or provisions of the BTA that limit what trueEX may do with its access to the
MarkitSERV network. Although counsel directed the Court’s attention to certain defined terms in
the BTA, none answered the question definitively.118 Thus, the BTA, even viewed in the light most
favorable to MarkitSERV, is not clear on this point.119
When a contract is ambiguous, the Court may consider extrinsic evidence to aid in
its interpretation.120 There is some evidence of record that ultimately may support MarkitSERV’s
interpretation.

The fact that truePTS sought repeatedly to contract with MarkitSERV for STP

facilitation on terms similar to trueEX suggests that plaintiffs, or at least some of their personnel,
understood the BTA as providing network access only to trueEX.

trueEX did not dispute

MarkitSERV’s characterization of the BTA during the hearing. Nevertheless, the interpretation of
the BTA ultimately is a question of fact for trial. Moreover, as Aspen Skiing recognized, so too is
118

See Hr’g Tr. 53-57, 67-69.
119

Of course, it is possible that the BTA is utterly silent on the question and that the problem
here arises because none of the contracting parties foresaw the possibility that trueEX might
use its access to the MarkitSERV network in the present circumstances.
120

Garza v. Marine Transp. Lines, Inc., 861 F.2d 23, 27 (2d Cir. 1988).

30
the question whether MarkitSERV’s purported business justification for terminating trueEX was
legitimate, i.e., economically rational.

In sum, the question whether MarketSERV can justify its

termination of trueEX as a legitimate response to threatened “free-riding” on its network presents
serious questions that are fair grounds for litigation.
While it is debatable on this record whether trueEX has shown that it is likely to
succeed on the merits at trial, trueEX has sustained its burden of raising serious questions going to
the merits that provide fair grounds for litigation. On the one hand, trueEX presented evidence from
which a factfinder could infer that MarkitSERV terminated the BTA for anticompetitive purposes.
On the other hand, MarkitSERV adduced evidence that its decision may have furthered a legitimate
business purpose. “Although this issue cannot be finally decided until discovery is completed and
the court has the benefit of a full factual presentation,” trueEX has satisfied its burden at this
preliminary stage.121
The foregoing is sufficient to resolve the likelihood-of-success prong of the
preliminary injunction standard.

Nevertheless, prudence suggests that the Court rule also on

plaintiffs’ likelihood of success on their other antitrust and quasi-contractual claims.

2.

Essential Facilities

Plaintiffs contend that MarkitSERV’s STP network is a facility essential to
competition in the IRS post-trade processing market and that MarkitSERV’s denial of access to it
violates Section 2. To bring a successful Section 2 essential facilities claim, a plaintiff must prove:
“(1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or

121

Jacobson & Co. v. Armstrong Cork Co., 548 F.2d 438, 444 (2d Cir. 1977).

31
reasonably to duplicate the essential facility; (3) the denial of the use of the facility to a competitor;
and (4) the feasibility of providing the facility.”122
Here, plaintiffs contend that (1) MarkitSERV “has exclusive control over the dropcopy STP function, which is essential to any SEF seeking to provide IRS trade processing services”;
(2) “it is impossible for trueEX and truePTS to ‘practically or reasonably’ duplicate the provision
of connectivity to MarkitSERV clients” because “[d]uplicating MarkitSERV’s STP infrastructure
would require a massive investment of time and money, making it impractical (and unprofitable)
from a business standpoint”; (3) “MarkitSERV is acting to deny trueEX and truePTS access to and
use of this essential facility”; and (4) “past practice demonstrates that it is feasible for MarkitSERV
to provide STP connectivity.”123
As an initial matter, the continued viability of the essential facilities doctrine has been
called into question in recent years.124 A plaintiff seeking to recover on an essential facility claim
“must show more than inconvenience, or even some economic loss; [it] must show that an
alternative to the facility is not feasible.”125

122

Twin Labs., Inc. v. Weider Health & Fitness, 900 F.2d 566, 569 (2d Cir. 1990) (quoting
MCI Commc’ns Corp. v. Am. Tel. & Tel. Co., 708 F.2d 1081, 1132-33 (7th Cir. 1983)).
123

DI 23 at 21-22.
124

Trinko, 540 U.S. at 411 (“We have never recognized such a doctrine and we find no need
either to recognize it or to repudiate it here.”); RxUSA Wholesale Inc. v. Alcon Labs., 391 F.
App’x 59, 61 (2d Cir. 2010) (“[T]o the extent that such a claim is viable, RxUSA’s essential
facilities claim fails against the Manufac turers, at the very least because RxUSA is able to
obtain pharmaceutical products from other sources, albeit at a higher price.”); Elevator
Antitrust, 502 F.3d 47, 54 (2d Cir. 2007) (describing Aspen Skiing-type termination of prior
course of dealing as “the sole exception to the right of refusal to deal”).
125

Twin Labs., Inc. v. Weider Health & Fitness, 900 F.2d 566, 570 (2d Cir. 1990).

32
The fact that trueEX has built direct connections to or otherwise managed to process
trades for three of its fourteen active dealers and thirteen of its twenty- two active buy-side clients
seriously undermines its claim that MarkitSERV’s STP network is essential. Indeed, for the first
quarter of 2017, roughly forty percent of trueEX trades bypassed MarkitSERV entirely, meaning
that trueEX successfully processed those trades without relying on MarkitSERV’s STP network.126
As for truePTS, plaintiffs admit that it “can and will develop its own network” if it has an
opportunity to get off the ground.127 To be sure, it will take time and money for truePTS to do so.
Nevertheless, in view of their admission, plaintiffs cannot credibly claim that “it is impossible
for . . . truePTS to ‘practically or reasonably’ duplicate” MarkitSERV’s network in due course.128
Moreover, trueEX has not adequately shown for purposes of this claim that it was
denied access to MarkitSERV’s STP network. “[T]he indispensable requirement for invoking the
[essential facilities] doctrine is the unavailability of access to the ‘essential facilities’; where access
exists, the doctrine serves no purpose.”129 It is undisputed that MarkitSERV offered trueEX access
to its STP network via its so-called standard workflow. Under MarkitSERV’s proposal, trueEX
126

Eggers Decl., Ex. 2 [DI 31-2].
127

DI 23 at 3.
128

DI 23 at 21.
Plaintiffs argue also that even if they were able to replicate “full STP functionality,” “the
only way to coordinate transactions” with IRS counterparties that use MarkitSERV for trade
processing “would be through the transmission of drop-copy reports, which cannot occur
without MarkitSERV’s cooperation.” Id. The world plaintiffs describe, however, is not the
one in which we live. They essentially ask the Court to predic t how MarkitSERV would
behave in future circumstances quite different from those we are faced with here. The Court
declines to do so.
129

Trinko, 540 U.S. at 411.

33
would have retained the ability to process IRS trades involving parties with which it has a direct
connection (or parties that do not require STP) and, where it does not, MarkitSERV would have
processed the trade on its network. In other words, MarkitSERV’s proposal, had it been accepted,
would have given trueEX access to MarkitSERV’s STP network and allowed trueEX to continue
to compete in the IRS trade processing market. “Because reasonable access to the essential facility
exist[ed]—even if not in a way that [wa]s conducive to [trueEX’s] existing business
model—[trueEX] cannot establish an essential facilities claim.”130

B.

Promissory Estoppel
“In New York, promissory estoppel has three elements: ‘a clear and unambiguous

promise; a reasonable and foreseeable reliance by the party to whom the promise is made, and an
injury sustained by the party asserting the estoppel by reason of the reliance.’”131 trueEX claims that
(1) MarkitSERV “unambiguously promised” not to cut off drop-copy functionality after terminating
the BTA, (2) it reasonably relied on MarkitSERV’s assurance because the parties had worked
cooperatively for over five years, and (3) it was injured as a result of being lulled into not bringing
suit for three days because MarkitSERV in that time issued a notice of termination to trueEX’s
clients that caused those clients to be concerned about potential disruption of service. The Court is
not persuaded that trueEX is likely to succeed on this claim for two reasons.
First, an essential element of a promissory estoppel claim is that the promisee—here,

130

MetroNet, 383 F.3d at 1130; see also Aerotec Int’l, Inc. v. Honeywell Int’l, Inc., 836 F.3d
1171, 1185 (9th Cir. 2016).
131

Cyberchron Corp. v. Calldata Sys. Dev., Inc., 47 F.3d 39, 44 (2d Cir. 1995).

34
trueEX—“substantial[ly] change[d]” its position “in reasonable reliance upon the promise.”132
trueEX fails entirely to show how this requirement is satisfied.
Second, trueEX’s showing with respect to injury is inadequate. It contends that but
for MarkitSERV’s promise it “would have sought judicial relief sooner, seeking to enjoin
MarkitSERV from further pursuing termination of the [BTA].”133 But despite the three-day delay,
trueEX has done just that, and there is nothing to suggest that the short delay prejudiced trueEX in
any way. In fact, the parties have entered into a standstill agreement that preserves the status quo
until July 24, 2017, thereby ensuring MarkitSERV does not “further pursu[e] termination” of its
relationship with trueEX.

While trueEX contends also that it was injured by MarkitSERV’s

termination notice, it does not claim that any of its clients actually abandoned it as a result,
especially once they learned of the parties’ standstill agreement.134

In other words, there is no

evidence that its clients’ fears translated into any identifiable trueEX loss.

II.

Balance of Hardships135
When balancing the hardships, “the issue is whether [the movants] would suffer

decidedly greater harm from an erroneous denial of an injunction than the defendants would suffer

132

Schmidt v. McKay, 555 F.2d 30, 36 (2d Cir. 1977).
133

DI 23 at 24.
134

See DI 25 ¶ 60.
135

The Court does not consider potential harm to truePTS in balancing the equities of this case
in view of its conclusion that truePTS has not shown a likelihood of success on any of its
claims.

35
from an erroneous grant.”136
In this case, the risk to MarkitSERV of an erroneous injunction is slight.

True,

issuance of a preliminary injunction would require MarkitSERV to assist a competitor with which
it no longer “wish[es] to deal”—a cognizable harm, at least in the abstract.137 Practically speaking,
however, MarkitSERV would “need do only what it has done for the past [three] years—provid[e]
[trueEX] with [drop- copy service]”138 and, unless trueEX ultimately prevails at trial, only until this
case is tried in less than a year. Moreover, it continues, as far as the Court can tell, to provide those
same services to TradeWeb and Bloomberg, which not only are trueEX’s competitors but which
compete also with MarkitSERV in the IRS post-trade processing market. In the circumstances,
MarkitSERV would not suffer significant adverse consequences in the event a preliminary
injunction erroneously were granted.
In contrast, the risks to trueEX of a wrongful denial of an injunction would be more
serious.

Without a preliminary injunction, trueEX is likely to suffer economic injury in the form

136

Foulke ex rel. Foulke v. Foulke, 896 F. Supp. 158, 162 (S.D.N.Y. 1995) (c iting Holford
USA Ltd. v. Cherokee, Inc., 864 F. Supp. 364, 374 (S.D.N.Y. 1994)); see also Buffalo Forge
Co. v. Ampco-Pittsburgh Corp., 638 F.2d 568, 569 (2d Cir. 1981); Tradescape.com v.
Shivaram, 77 F. Supp. 2d 408, 411 (S.D.N.Y. 1999).
137

United Asset Coverage, Inc. v. Avaya Inc., 409 F. Supp. 2d 1008, 1042 (N.D. Ill. 2006).
The Court notes that in both of the cases on which MarkitSERV relies for that proposition,
the court concluded that the movant had not carried its burden on irreparable harm, thereby
rendering any forced dealing that would have flowed from an injunction a severe hardship
for the defendant. See Jack Kahn Music Co. v. Baldwin Piano & Organ Co., 604 F.2d 755,
764 (2d Cir. 1979) (“Having already held that Kahn has failed to prove a probability of
irreparable damage resulting from termination of its Baldwin dealership, there is little to add
on the subject of the balance of hardships.”); Avaya, 409 F. Supp. 2d at 1039-42. Here, in
contrast, trueEX has persuaded the Court that it likely will face irreparable harm if a
preliminary injunction is not granted.
138

Reuters Ltd. v. United Press Int’l, Inc., 903 F.2d 904, 909 (2d Cir. 1990).

36
of lost clients and lost revenue, injury that has the potential to be ruinous in nature. (More on that
point later.) Furthermore, trueEX almost certainly would lose some customers between now and
a final decision on the merits which would be unlikely to “return to [trueEX] even if [MarkitSERV]
ultimately were enjoined.”139
In light of the foregoing, the Court concludes that the balance of hardships tips
decidedly in trueEX’s favor. While the extent of any hardship from an erroneous ruling on this
motion is not certain, the harm trueEX faces were MarkitSERV not enjoined when it should have
been is far more severe than any potential harm facing MarkitSERV.140 trueEX but not MarkitSERV
risks sustaining serious and likely unquantifiable economic injury to its business.

III.

Public Interest
The Court “must ensure [also] that the ‘public interest would not be disserved’ by the

issuance of a preliminary injunction.”141 The public has an interest “in enforcement of the antitrust
laws and in the preservation of competition.”142 Granting a preliminary injunction in this case would
serve those interests because it would preserve competition in the IRS post-trade processing market
and would prevent MarkitSERV from engaging in potentially anticompetitive conduct until a trial

139

Tradescape.com, 77 F. Supp. 2d at 411.
140

Id. at 412.
141

Salinger v. Colting, 607 F.3d 68, 80 (2d Cir. 2010) (quoting eBay Inc. v. MercExchange,
L.L.C., 547 U.S. 388, 391 (2006)).
142

United States v. Columbia Pictures Indus., Inc., 507 F. Supp. 412, 434 (S.D.N.Y. 1980).

37
on the merits can take place.143

IV.

Threat of Irreparable Harm
Irreparable harm is “the sine qua non for preliminary injunctive relief.”144 To prove

irreparable harm, “[t]he movant must demonstrate an injury that is neither remote nor speculative,
but actual and imminent and that cannot be remedied by an award of monetary damages.”145 It is
well settled in this circuit that “[m]ajor disruption of a business can be as harmful as termination,
and a ‘threat to the continued existence of a business can constitute irreparable injury.’”146 So too

143

MarkitSERV claims that the public interest in freedom of contract would be served by
denying the injunction. DI 29 at 29. The Sherman Act, however, embodies Congress’
judgment that freedom of contract in some circumstances must bow to the public’s interest
in promoting competition and efficiency. See 15 U.S.C. § 1 (declaring “illegal” “[e]very
contract . . . in restraint of trade”). Where the contractual right is alleged to violate the
antitrust laws, the public interest in antitrust enforcement and preservation of competition
outweighs the interest in freedom of contract.
For similar reasons, MarkitSERV’s contention that “competition is ill-served by forcing [it]
to share its network, since it reduces both parties’ incentives to invest in their respective
facilities” is unpersuasive in the present c ontext. While much ink has been spilled debating
the wisdom of the refusal to deal doctrine, see, e.g., W. Carlton, A General Analysis of
Exclusionary Conduct and Refusals To Deal—Why Aspen and Kodak Are Misguided,
68 Antitrust L.J. 659, 677 (2001), such claims remain cognizable under our antitrust laws.
Forced dealing may, as MarkitSERV contends, undermine competitive incentives, but the
law recognizes that such a result may be justified in some circumstances.
144

USA Recycling, Inc. v. Town of Babylon, 66 F.3d 1272, 1295 (2d Cir. 1995).
145

Rodriguez ex rel. Rodriguez v. DeBuono, 175 F.3d 227, 234 (2d Cir. 1999).
146

Nemer Jeep-Eagle, Inc. v . Jeep-Eagle Sales Corp., 992 F.2d 430, 435 (2d Cir. 1993)
(quoting John B. Hull, Inc. v. Waterbury Petroleum Prods., Inc., 588 F.2d 24, 28-29
(2d Cir. 1978)); see also Petereit v. S.B. Thomas, Inc., 63 F.3d 1169, 1186 (2d Cir. 1995)
(“Major disruption of a business can . . . constitute irreparable injury.”).

38
can “a threatened loss of good will and customers, both present and potential.”147
trueEX contends that it will suffer irreparable harm in the form of major disruption
to its business if a preliminary injunction does not issue.

Specifically, it asserts that it “cannot

survive as a business without connectivity to MarkitSERV’s network” because “nearly every trade
that happens on trueEX’s platform includes at least one counterparty that needs MarkitSERV
connectivity to receive real-time updates for their books and records.”148 Without the ability to
update their books and records in real time, trueEX contends, clients will “leave the trueEX platform
and quickly migrate elsewhere.”149
Although the full extent to which trueEX’s business relies on the drop-copy workflow
is not clear from the record, the Court finds that drop-copy functionality is vitally important to
trueEX’s business. STP facilitation is essential to many entities that trade IRS, especially dealers.150
If trueEX does not have a direct connection to a counterparty that requires STP, trueEX must use
MarkitSERV’s drop-copy workflow in order to provide STP to that counterparty (assuming the
counterparty is connected to MarkitSERV’s network, which the vast majority of market participants
are).

The evidence shows that trueEX has direct connections to three IRS dealers (out of the

fourteen that have traded on its platform in 2017) and the majority of its active buy-side clients
(roughly thirteen out of twenty-two). In other words, trueEX need not rely on MarkitSERV’s drop-

147

Jacobson, 548 F.2d at 445.
148

DI 23 at 25.
149

Id.
150

DI 25 ¶ 28.

39
copy workflow to process trades involving those three dealers and thirteen buy-side clients, meaning
it could continue to offer its full suite of services to those clients if the Court were to deny its
motion.

The question then is whether trueEX generates sufficient business from those clients to

remain in business until a trial on the merits can take place.
MarkitSERV, unsurprisingly, contends that it does. It points to 2016 data showing
that nearly 60 percent of the notional value of trueEX’s U.S. dollar-denominated trades bypassed
MarkitSERV’s network.

From that datum, MarkitSERV asks the Court to infer that most of

trueEX’s business would be unaffected by the discontinuation of drop-copy service.

While

MarkitSERV seems to concede that trueEX would be likely to sustain some disruption to its
business, it contends that the disruption would not rise to the level of threatening trueEX’s continued
viability.
trueEX argues strenuously to the contrary. It urges the Court to reject MarkitSERV’s
argument for two primary reasons.
First, trueEX accuses MarkitSERV of “cherry-pick[ing] [its] data” because
MarkitSERV looked only at 2016 and focused only on U.S. dollar swaps, “which ignores a
substantial portion of the platform’s volume.”151 trueEX contends that when one looks at swaps
denominated in all currencies for 2017, “more than 95% of the notional trade value on trueEX
required drop-copy to MarkitSERV.”152
Second, trueEX contends that even if MarkitSERV were correct that 60 percent of
its trades currently bypass the MarkitSERV network, trueEX still could not function as an SEF
151

DI 41 ¶ 23.
152

DI 41 ¶ 23.

40
without drop-copy service.
Since both sides are guilty of cherry-picking data—as the hearing on this motion
showed—the Court is not persuaded one way or the other by the conflicting numbers relied on by
the parties. Nevertheless, the Court finds that there would be an actual and imminent threat to the
continued viability of trueEX if MarkitSERV were to discontinue drop-copy service to it. Most
market participants consider STP facilitation a deal breaker. In the cases in which trueEX does not
have a direct connection to a customer for STP facilitation, that customer will abandon trueEX’s
platform for a trading platform that can provide STP. Currently, eleven of trueEX’s active dealer
clients lack direct connections. When those dealers leave the trueEX platform, buy-side clients who
remain on the platform will have only three dealers providing liquidity—as compared to the thirtythree dealers accessible through TradeWeb, for example.

As a result, “no rational buy-side

participant would choose to trade on a SEF that is completely cut off from the vast majority of the
IRS market, and therefore lacks the liquidity and efficiency that participants expect and demand
from a SEF.”153 Thus, even if trueEX currently processes 60 percent of its trades without any
involvement of MarkitSERV, once trueEX loses drop-copy functionality, eleven of its fourteen
dealers would leave the platform thereby causing a mass exodus of buy-side clients. Without clients
trading on its platform, trueEX cannot survive as a business. That is sufficient to establish
irreparable harm.
trueEX is likely also to suffer irreparable harm in the form of lost goodwill.
“[T]erminating the delivery of a unique product to a distributor whose customers expect and rely
on the distributor for a continuous supply of that product almost inevitably creates irreparable

153

DI 41 ¶ 24.

41
damage to the good will of the distributor.”154

“This is particularly evident when many of the

distributor’s customers have indicated not only a strong preference for the terminated product, but
also have threatened to stop dealing with the distributor if it cannot continue to supply that
product.”155
Here, trueEX’s customers expect and rely on it to perform trade processing using
STP. Not only have clients indicated a strong preference for STP—in one case referring to the lack
of STP as a “deal breaker”—some have threatened to stop doing business with trueEX if it no longer
can provide STP facilitation.156 For example, one dealer client informed trueEX that it would need
to “discuss . . . options for the future” if its connection to trueEX were not set up by the time
MarkitSERV discontinued drop-copy service.157 Another client sought to accelerate a number of
planned trades so that it could execute them before drop-copy service was cut off, suggesting that
the client did not believe it could do business with trueEX in the future.
In the circumstances, termination of the BTA quite likely would cause irreparable
harm to trueEX’s “good will and business reputation, because customers will have no choice but to
resort to using a substitute product to meet their needs—possibly resulting in a permanent loss of
business.”158

154

Reuters, 903 F.2d at 907-08.
155

Id.
156

See DI 25 ¶¶ 56-60.
157

DI 25 ¶ 56.
158

Rex Med. L.P. v. Angiotech Pharms. (US), Inc., 754 F. Supp. 2d 616, 622 (S.D.N.Y. 2010).

42
Conclusion
The historic purpose of a preliminary injunction is to maintain the status quo pending
a full trial where all facts and arguments may be considered with adequate preparation and in their
complete context.

As the foregoing demonstrates, this is a case in which full and expanded

consideration is essential. At a minimum, there are serious questions going to the merits. In the
circumstances, the balance of hardships tips decidedly in trueEX’s favor. For those and all the
reasons stated above, plaintiffs’ motion for a preliminary injunction [DI 22] is granted with respect
to trueEX. It is denied, however, with respect to truePTS, principally because truePTS has not
shown any likelihood of success on its claims.
It is important to recognize that the Court is dealing here only with a preliminary
injunction the purpose of which is to preserve the status quo pending trial. The Court already has
raised with the parties the question whether trueEX, were it to prevail, would be entitled to a
permanent injunction requiring MarkitSERV to continue to deal with it in perpetuity in the manner
it did prior to the termination notice,159 which would be a very different matter. trueEX recognizes
that it would have to deal with that troublesome problem were it to win at trial.160
In all the circumstances, defendants MarkitSERV Limited and MarkitSERV, LLC
be and they hereby are enjoined and restrained, pending hearing and determination of the action,
from directly or indirectly terminating the BTA or discontinuing the provision of drop-copy service
to plaintiff trueEX, LLC.
No bond is required because the Court finds that there is no likelihood of harm to
159

Hr’g Tr. 4:14-21.
160

Id. at 64:13-65:2.

43
defendants.161
The foregoing constitute the Court’s findings of fact and conclusions of law.
SO ORDERED.
Dated:

July 18, 2017

161

Doctor’s Assocs., Inc. v. Distajo, 107 F.3d 126, 136 (2d Cir. 1997).