KIM v. JUMP TRADING LLC LLC (2026)

TerraUSD (Terra) and LUNA were cryptocurrencies on the Terra blockchain, which was managed by Terraform Labs, a Singaporean company. Unlike a bank, which maintains centralized records of depositors' balances, a blockchain employs a decentralized network of users to verify each transaction. LUNA was a traditional cryptocurrency, much like Bitcoin, whose value fluctuated based on supply and demand. Terra, by contrast, was supposed to be a stablecoin, whose value was linked to some other asset-for Terra, the U.S. dollar. Each Terra coin was supposed to be worth $1.

Most stablecoins rely on a reserve of another asset to maintain their value. Each coin is a claim to $1 in the reserves, so the stablecoin should be worth $1. (That is, it should be redeemable for $1 in assets, such as deposits at a bank.) Terra, however, used an arbitrage mechanism to maintain its value. An algorithm allowed holders to convert each Terra coin into a dollar's worth of LUNA and vice versa. (Rather than facilitate the exchange of Terra for LUNA, the algorithm worked by destroying a Terra coin and minting LUNA to replace it or by destroying LUNA and minting a Terra.) Then, if the market price of Terra declined below $1, arbitrageurs could buy it, convert the Terra into a dollar of LUNA, and sell the LUNA, profiting from the spread. So long as this spread was greater than the cost of trading, investors would buy Terra and convert it to LUNA or convert Terra they already owned. In principle, this would restore Terra's price to $1.

Terraform's arbitrage mechanism twice failed to keep its value at $1. Both times, Terra's price declined, and while investors holding Terra could convert it to LUNA, for various reasons this became undesirable. Terraform was able to solve the first decline, in 2021, by persuading Jump Trading to buy millions of Terra coins in exchange for options to buy LUNA cheaply. Jump's purchase restored the price of Terra to $1. But instead of proclaiming Jump's role, Terraform pretended that the arbitrage mechanism had restored Terra's price. According to the complaint, it was then that investors-unaware of Jump's role during the first crisis-bought Terra, trusting the arbitrage mechanism. When it failed again, in 2022, Terra's price collapsed. This time Jump didn't intervene. Terraform halted all trading and filed for bankruptcy. About $40 billion in nominal value vanished.

Taewoo Kim owned Terra coins when their value plummeted in 2022. Kim and a proposed class of investors in Terra sued Jump under the Commodities Exchange Act and its associated regulations, as well as under the common law of unjust enrichment. Kim alleges that Jump conspired with Terraform to manipulate the price of Terra, and, in so doing, fooled him into thinking that Terraform's arbitrage mechanism alone could maintain its equivalence with the U.S. dollar. Kim couldn't have sued Terraform because it enjoys rights under an arbitration agreement that Kim signed when he joined the “Anchor Protocol”-a program that allowed holders of Terra to trade their coins and lend them to other users. The agreement provides that

[a]ny claim or controversy arising out of or relating to the Interface, this Agreement, including any question regarding this Agreement's existence, validity or termination, or any other acts or omissions for which you may contend that we are liable, including (but not limited to) any claim or controversy as to arbitrability (“Dispute”), shall be referred to and finally resolved by arbitration in Singapore in accordance with the Arbitration Rules of the Singapore International Arbitration Centre (“SIAC Rules”).

Jump argues that Kim can't sue it either because it, too, benefits from the agreement. It moved in the district court to compel arbitration. Whether it did so under 9 U.S.C. § 4 or § 206 is unclear; the motion mentions neither. Jump says that 9 U.S.C. § 202 governs the arbitration agreement between Kim and Terraform, so we will assume that § 206 applies. When the district court denied Jump's request, see 2025 WL 1359136, 2025 U.S. Dist. LEXIS 88898 (N.D. Ill. May 9, 2025), it appealed under 9 U.S.C. § 16(a).

Jump acknowledges that it is not a party to the agreement between Kim and Terraform. Instead, it says that the doctrine of equitable estoppel requires Kim to arbitrate, since Kim is suing Jump for conduct in which Terraform was a co-conspirator. In Jump's view, a lawsuit against it would deny Terraform the benefit of the arbitration agreement because Jump may sue Terraform for contribution. And that would force Terraform to litigate, rather than arbitrate, questions about its misconduct. Yet Terraform and its successor in bankruptcy have neither intervened nor filed an amicus brief to support Jump-understandably so. The bar date for claims in the bankruptcy has passed, and Terraform's remaining assets are held in a liquidating trust from which Jump cannot benefit. That makes a suit by Jump against Terraform unlikely.

Jump nonetheless contends that an arbitrator, rather than a court, should decide the issue of equitable estoppel. It reads the portion of the agreement requiring arbitration of “any claim or controversy as to arbitrability” to mean that whenever there is a dispute whether a third party can enforce the agreement, that dispute must be arbitrated. And it points to language from the Supreme Court's decision in Howsam v. Dean Witter Reynolds, Inc., 537 U.S. 79, 123 S.Ct. 588, 154 L.Ed.2d 491 (2002), and our own in Janiga v. Questar Capital Corp., 615 F.3d 735 (7th Cir. 2010), that it thinks compels this conclusion. Together, these arguments present two questions: first, who decides-the court or an arbitrator-the merits of Jump's equitable estoppel argument? And second, if a court decides, is Jump entitled to arbitrate against Kim?

Arbitration is contractual; a court can compel it only if the parties have agreed to arbitrate. AT&T Technologies, Inc. v. Communications Workers, 475 U.S. 643, 648-49, 106 S.Ct. 1415, 89 L.Ed.2d 648 (1986). Parties can agree to arbitrate disputes about arbitrability, which ask whether a substantive dispute should be arbitrated or litigated. Kim and Terraform agreed to do that in the terms of service. For example, they would need to arbitrate a dispute about whether Kim's claims (the source of the substantive legal dispute) “aris[e] out of or relat[e] to” the terms-and, if they did, Kim would need to arbitrate them. Jump says that an arbitrator also must decide whether equitable estoppel requires Kim to arbitrate his substantive legal claims against it. That's because, in its view, Jump's equitable theory presents a “claim or controversy as to arbitrability”. See Howsam, 537 U.S. at 84, 123 S.Ct. 588; Janiga, 615 F.3d at 743.

But before deciding whether equitable estoppel is a question of arbitrability, we must decide whether Kim and Jump have agreed to arbitrate any questions. Arbitration is “strictly a matter of consent,” Coinbase, Inc. v. Suski, 602 U.S. 143, 148, 144 S.Ct. 1186, 218 L.Ed.2d 615 (2024) (citation omitted), “a way to resolve those disputes-but only those disputes-that the parties have agreed to submit to arbitration.” Ibid. (quoting First Options of Chicago, Inc. v. Kaplan, 514 U.S. 938, 943, 115 S.Ct. 1920, 131 L.Ed.2d 985 (1995)). “Consequently, the first question in any arbitration dispute must be: What have these parties agreed to?” Ibid. (emphasis added). Kim and Jump haven't agreed to arbitrate. Kim and Terraform have, so Jump's best hope of arbitrating the issue of equitable estoppel would be as a third-party beneficiary of their agreement. See Morales-Posada v. Cultural Care, Inc., 141 F.4th 301, 309-10 (1st Cir. 2025). Yet it does not contend that it satisfies the requirements of this doctrine.

Who decides whether Jump has rights under that agreement? It must be a court. As the Supreme Court explained in Rent-A-Center, West, Inc. v. Jackson, 561 U.S. 63, 71, 130 S.Ct. 2772, 177 L.Ed.2d 403 (2010), courts decide not only whether parties have agreed to delegate questions of arbitrability to an arbitrator, but also whether and by whom an agreement delegating these questions can be enforced. In Rent-A-Center one of the parties argued that an agreement to arbitrate substantive disputes was void. The Justices held that an arbitrator would need to decide whether that was true, since the parties had agreed to arbitrate questions of arbitrability. But it would have fallen to a court, not an arbitrator, to decide whether the agreement to arbitrate questions of arbitrability was enforceable if the parties had challenged it (rather than the agreement to arbitrate substantive disputes).

Just so here. Kim may have agreed to arbitrate the question whether third parties can enforce his substantive arbitration agreement with Terraform. But a court must decide which third parties can enforce the agreement to arbitrate “claim[s] or controvers[ies] as to arbitrability”. Otherwise, anyone could benefit from a stranger's agreement to arbitrate these questions without ever being a party to it-a result that the ordinary law of contracts does not allow. And the Federal Arbitration Act does not “elevate [agreements to arbitrate questions of arbitrability] over other forms of contract.” Rent-A-Center, 561 U.S. at 71, 130 S.Ct. 2772.

This means that judges, not an arbitrator, must decide whether the agreement between Kim and Terraform to arbitrate arbitrability gives rights to Jump. Jump says that language requiring Kim to arbitrate claims based on Terraform's misconduct-or based on “acts or omissions for which you [Kim] may contend that we [Terraform] are liable”-gives rights to co-conspirators in that misconduct, but the text compels the opposite conclusion: that Kim agreed to arbitrate with Terraform only. See In re Automotive Parts Antitrust Litigation, 951 F.3d 377, 385 (6th Cir. 2020) (broad scope of disputes covered by bilateral arbitration agreement did not justify applying agreement to arbitrate arbitrability to third parties); Kramer v. Toyota Motor Corp., 705 F.3d 1122, 1127-28 (9th Cir. 2013). Consider also the default rule in state law that “C can[not] claim rights under a contract that has only A and B as parties.” CCC Intelligent Solutions Inc. v. Tractable Inc., 36 F.4th 721, 724 (7th Cir. 2022); Morales-Posada, 141 F.4th at 309-10. We don't see anything in the text that alters this default. So if Jump and Kim lack an agreement to arbitrate arbitrability, and if Jump doesn't have rights to enforce Terraform's agreement with Kim, then a court must resolve its argument about equitable estoppel.

Having decided that the district judge was right to analyze Jump's equitable estoppel argument, rather than leave it to the arbitrator, we now review her decision on the merits. The district judge applied Illinois's version of equitable estoppel, finding, correctly in our view, that Illinois courts foreclose estoppel unless the third party has reasonably relied on the contract providing for arbitration. Ervin v. Nokia, Inc., 349 Ill. App. 3d 508, 514-16, 285 Ill.Dec. 714, 812 N.E.2d 534 (2004); see also Sunco International Inc. v. Jiangsu Sunco Boiler Co., Ltd., No. 25-2251, ––– F.4th ––––, 2026 WL 2349275 (7th Cir. 2026) (issued contemporaneously). No one suggests that happened here.

But was the district judge correct to apply Illinois law? Jump argues that Chapter 2 of the Federal Arbitration Act, 9 U.S.C. §§ 201-08-which implements the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (the New York Convention)-provides the applicable law of equitable estoppel. Chapter 2 is silent about whether third parties can enforce arbitration agreements. Still, Jump argues that Chapter 2 authorizes federal courts to develop their own law of equitable estoppel. This stands in contrast to Chapter 1, which uses the rules of equitable estoppel from state law. Arthur Andersen LLP v. Carlisle, 556 U.S. 624, 630-31, 129 S.Ct. 1896, 173 L.Ed.2d 832 (2009).

Like the district judge, we doubt that Chapter 2 governs this dispute. It would govern a dispute between Kim and Terraform about the enforceability of their agreement to arbitrate, for Terraform is not a citizen of the United States (although Jump is). See 9 U.S.C. § 202. But for the same reasons we hesitated to adopt Jump's expansive reading of the word “arbitrability”, we hesitate now to say that Chapter 2 determines Jump's rights as a third party. Under Jump's view, any defendant, simply by asserting rights to enforce an arbitration agreement “arising out of a legal relationship” between a citizen of the United States and a foreign one, would be entitled to have the law of Chapter 2 decide his rights to enforce the agreement as a third party, no matter how remote the litigant is from the contract providing for arbitration. Jump's reading would allow an entity to benefit from an arbitration agreement's status under Chapter 2 before deciding whether it is subject to that agreement at all. That, as the district judge aptly put it, would put the cart before the horse.

We also have serious doubts about Jump's suggestion that GE Energy Power Conversion France SAS, Corp. v. Outokumpu Stainless USA, LLC, 590 U.S. 432, 140 S.Ct. 1637, 207 L.Ed.2d 1 (2020), permits federal courts to develop a common law of equitable estoppel with respect to contracts governed by Chapter 2. There the Justices held that nothing in the statute prevents federal courts from applying “domestic” doctrines of equitable estoppel to cases involving these contracts. In the United States, the prevailing “domestic” doctrines of equitable estoppel are found in state law. Arthur Andersen tells us that Chapter 1 of the Federal Arbitration Act incorporates state doctrines of equitable estoppel, obviating the need for federal rules. 556 U.S. at 630-31, 129 S.Ct. 1896. If GE Energy teaches that Chapter 2 mandates the application of domestic doctrines of equitable estoppel, and state law provides those doctrines-including with respect to arbitration agreements-then there shouldn't be any need to create federal doctrines of equitable estoppel. See also Cassirer v. Thyssen-Bornemisza Collection Foundation, 596 U.S. 107, 116-17, 142 S.Ct. 1502, 212 L.Ed.2d 451 (2022).

Let us put this a different way. Federal courts create doctrines of federal law when necessary to solve disputes arising under federal statutes. For example, the Employee Retirement Income Security Act (ERISA) requires plans' sponsors to keep the promises they make to workers in welfare-benefit plans. This rule can be enforced only by applying doctrines of contract law-which, being absent from ERISA, federal courts must adopt. Even then, however, the federal doctrines are absorbed from state law rather than being created out of whole cloth. See, e.g., M&G Polymers USA, LLC v. Tackett, 574 U.S. 427, 135 S.Ct. 926, 190 L.Ed.2d 809 (2015). See also CNH Industrial N.V. v. Reese, 583 U.S. 133, 135-36, 138 S.Ct. 761, 200 L.Ed.2d 1 (2018) (applying “ordinary principles of contract law” to interpret collective-bargaining agreements governed by federal labor law). Chapter 2 of the Federal Arbitration Act, which does not say one word about estoppel, does not need completion through federal common law.

What Chapter 2 does say is that Chapter 1 governs with respect to all subjects not covered in Chapter 2 or the New York Convention. 9 U.S.C. § 208. Estoppel is one of those subjects omitted from Chapter 2-but, per Arthur Andersen, is covered in Chapter 1 through the incorporation of state law. It follows that state law controls disputes about estoppel under Chapter 2 as well.

We recognize that four circuits have held that federal common law provides the rule of equitable estoppel for agreements governed by Chapter 2. Setty v. Shrinivas Sugandhalaya LLP, 3 F.4th 1166 (9th Cir. 2021); InterGen N.V. v. Grina, 344 F.3d 134, 143-44 (1st Cir. 2003); Smith/Enron Cogeneration Limited Partnership, Inc. v. Smith Cogeneration International, Inc., 198 F.3d 88, 96 (2d Cir. 1999); International Paper Co. v. Schwabedissen Maschinen & Anlagen GMBH, 206 F.3d 411, 417 n.4 (4th Cir. 2000). A fifth holds, to the contrary, that state law applies. Vinton v. Indian Harbor Insurance Co., 161 F.4th 282, 287-88 (5th Cir. 2025), petition for certiorari pending, No. 25-1383.

We agree with Vinton and add that, even if we were to create a federal rule of equitable estoppel, we would get the federal rule from state law. As the Supreme Court held in United States v. Kimbell Foods, Inc., the law of the forum state supplies the federal rule unless there is a need for uniformity or applying state law would frustrate federal interests. 440 U.S. 715, 728-29, 99 S.Ct. 1448, 59 L.Ed.2d 711 (1979); see also United States v. Crown Equipment Corp., 86 F.3d 700, 706-07 (7th Cir. 1996) (defining rule of federal common law for measuring damages to federal property by reference to forum state's law); West v. Louisville Gas & Electric Co., 951 F.3d 827, 832 (7th Cir. 2020) (interpreting federal statute affecting easements by looking to forum state's law of property); 19 Wright, Miller & Cooper, Federal Practice and Procedure § 4518 (3d ed. 2026) (“The Supreme Court has put increasing emphasis on the notion that when determining the content of federal common law, forum state law should be adopted as federal law absent some good reason to displace it”).

Arthur Andersen requires us to enforce arbitration agreements under Chapter 1 of the Federal Arbitration Act like any other contract: according to rules of state law. It does not prescribe uniformity. Why should we treat agreements under Chapter 2 differently? Chapter 2 covers agreements between U.S. and foreign citizens, but that shouldn't make a difference. International arbitration agreements may contain choice-of-law clauses, which should be honored, but arguments from outside the text of an agreement (as Jump's is) presumptively rest on the law of the forum state. Schoeps v. Sompo Holdings, Inc., 160 F.4th 815, 824 (7th Cir. 2025). Statutes or treaties may require national uniformity, but Jump does not point to any. See Setty, 3 F.4th at 1169-76 (Bea, J., dissenting) (citing GE Energy).

Suppose that we were to create a uniform federal rule of equitable estoppel. We would still need to determine whether it applies over rules of other countries that have signed the New York Convention. Yet Jump does not discuss the law of Singapore, Terraform's home nation, though it would have as great a claim as does common law in the United States. Nor does Jump ask us to look to the law of the European Union, Canada, Mexico, Japan, or other trading partners in forming a body of common law.

If there were such a thing as a federal common law under Chapter 2, we would need to define its content. When we have needed to create uniform rules in the past-as in cases interpreting plans under ERISA-we have developed them from “ordinary” or “general” principles of contract law. M&G Polymers, 574 U.S. at 435, 135 S.Ct. 926; Bock v. Computer Associates International, Inc., 257 F.3d 700, 704-05 (7th Cir. 2001). Jump's proposed rule would allow third parties to arbitrate claims that allege joint misconduct by it and the signatory to an arbitration agreement. See MS Dealer Service Corp. v. Franklin, 177 F.3d 942, 947 (11th Cir. 1999). This rule does not seem “ordinary” to us. It would allow third parties to seek benefits from a contract for which neither party negotiated, depriving Kim of the benefit of his bargain with Terraform. See 1 Lord, Williston on Contracts § 1:1 (4th ed. 2026) (“The goal of contract law is to hold parties to their agreements so that they receive the benefits of their bargains.”) It may make sense to apply equitable estoppel when a party to a contract does something to make a third party think he is bound to it-for example, if Kim sued Jump for breaching the terms of service. See GE Energy, 590 U.S. at 437, 140 S.Ct. 1637 (quoting 21 Lord, Williston on Contracts § 57:19 at 200 (4th ed. 2017)). But Kim isn't asserting a contractual right against Jump and, as far as we can tell, never did anything to make Jump think it was bound to the terms. Adopting Jump's rule might also disturb existing commercial relations; parties to bilateral arbitration agreements would be left wondering which third parties, possibly unknown to them, could enforce their arbitration agreements against a possible lawsuit. See Kimbell Foods, 440 U.S. at 728-29, 99 S.Ct. 1448.

Jump relies on Underwriters at Lloyd's v. Argonaut Insurance Co., 500 F.3d 571 (7th Cir. 2007), which used a rule drawn from federal law over one from California law when deciding whether a demand to arbitrate was timely. That decision does not concern estoppel-and it precedes Arthur Andersen, which told us to get rules of estoppel from state law. The panel relied on a supposed need for national uniformity that the Supreme Court rejected two years later. Because it predated Arthur Andersen, the panel in Argonaut also did not consider the effect of 9 U.S.C. § 208, which refers back to Chapter 1 to handle unprovided-for cases. Whatever force Argonaut may have if limited to timing questions, it cannot be extended to justify the creation of federal common law about estoppel.

There isn't any federal interest that compels us to create federal common law. Some courts have justified their invention of equitable estoppel doctrines as necessary to implement a federal interest in encouraging arbitration. See, e.g., MS Dealer Service, 177 F.3d at 947. But Morgan v. Sundance, Inc., 596 U.S. 411, 417-19, 142 S.Ct. 1708, 212 L.Ed.2d 753 (2022), tells us the opposite: that there isn't a federal interest in promoting arbitration. The goal of the federal statute is enforcement of contracts, nothing more.

This leads us back to Illinois's doctrine of equitable estoppel, which requires reasonable reliance by the party invoking the arbitration agreement. Because Jump doesn't contend that it relied on representations by Kim, it cannot use equitable estoppel to enforce the arbitration agreement against him. So Jump must litigate.

AFFIRMED

Easterbrook, Circuit Judge.