In the
United States Court of Appeals For the Seventh Circuit ____________________ Nos. 23-2840, 23-2846 & 23-2849 UNITED STATES OF AMERICA, Plaintiff-Appellee, v.
GREGG SMITH, MICHAEL NOWAK, and CHRISTOPHER JORDAN, Defendants-Appellants. ____________________
Appeals from the United States District Court for the Northern District of Illinois, Eastern Division. Nos. 1:19-CR-00669-1, 2, 4 — Edmond E. Chang, Judge. ____________________
ARGUED SEPTEMBER 5, 2024 — DECIDED AUGUST 20, 2025 ____________________
Before EASTERBROOK, KIRSCH, and KOLAR, Circuit Judges. KIRSCH, Circuit Judge. Gregg Smith, Michael Nowak, and Christopher Jordan were precious metals futures traders who manipulated the market through an unlawful practice called ę, meaning they placed deceptive orders they in- tended to cancel to push the market price a certain direction. At trial, they were all convicted of various crimes in connec- tion with this practice. They appeal their convictions on 2 Nos. 23-2840, 23-2846 & 23-2849
multiple grounds. Finding none of their challenges persua- ǰȱ ȱĜȱȱȱǯ I A Gregg Smith, Michael Nowak, and Christopher Jordan once traded precious metals futures contracts on commodities ¡ȱ ȱ ¢ȱ ȱ ȱ ȱ ¡ȱ Group (CME). They each employed a fraudulent scheme, ȱȱęǰ to game the system and manipulate the prices of the precious metals futures they traded. ęȱȱȱȱȱȱ of commodities ȱǯȱȱȱȱȱȱ¢ȱȱȬ ȱȱ¢ȱȱȱȱ¢ȱȱȱ-upon price on an agreed-upon future date. Commodities futures traders use ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ - operated exchanges. These traders ȱȱȱ¢, ȱěȱ to sell, ȱȱȱȱȱȱȱ ¡ȱȱȱ ęȱȱȱǯȱȱȱ ¡ȱȱ¢ǰȱȱ while some traders employ computer algorithms to place or- ders, others enter their orders manually, as Smith, Nowak, and Jordan did. ¡ȱȱȱȱȱȱȱȬ ǰȱȱȱęȱ ȱȱȱǰȱ¢ȱȱȱȱȱ executed. The price of futures contracts is ȱ¢ȱ¢ȱȱ demand: the price will increase if there is more demand than supply for a product, and vice versa. But a fundamental as- sumption on CME exchanges is that every order represents a legitimate, ȱ ęȱ ȱ ȱ ȱ ȱ ȱ ȱ . Each ȱȱȱ ¡ȱȱȱȱȱ¢ȱ conveying an intent to participate in it at a particular price. Nos. 23-2840, 23-2846 & 23-2849 3
Today, ¡ȱ ȱ ¢ȱ ȱ ǰȱ ȱ ȱ Smith, Nowak, and Jordan were traders, it displayed only the ȱȱȱȱȱȱȱěȱȱny time. ȱȱ ȱ¢ȱȱ ȱȱȱȱěȱȱȱȱ smaller orders will, ¡ȱ s traders ȱȱȱȱ ȱȱȱȱȱȱȱ¢ȱȱ-called ice- ȱǯȱ ȱȱ ¢ǰȱȱȱȱȱȱȬ ceal the true size of their order. Although these features were designed to mitigate the impact that large orders would have on the market price, they paved the way for fraudulent schemes to manipulate the market—¢ȱę. ȱęȱȱ large orders that he does not in- tend to execute, driving the price in a ȱȱȬ tion ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱ¢ȱȱǯȱThis typically involves four steps. First, the trader places an order, often an ȱǰȱȱhe genuinely intends to trade. Second, the ȱȱȱȱȱȱȱȱȱȱȱȱȬ ket. The trader never intends to trade this order; rather, it is designed to push the ȱȱȱȱęȱȱȱȬ mate order. For instance, if a trader wants ȱ¢ȱȱȱȱ ȱȱǰȱȱ ȱȱȱȱȱȱȱȱ ȱȱ (or a series of smaller sell orders at descending prices) to push the market down. Third, the market reacts to the illusion of ȱ ¢ȱ ȱ ¢ȱ ȱ ȱ ǰȱ ȱ ȱ trader to execute the genuine order at his desired price. ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęǯ CME rules ȱ ȱ ȱ ę. In particular, ȱ ȱ ŚřŘȱ ȱ ȱ ȱ manipulating or at- tempting to manipulate “prices of exchange futures or 4 Nos. 23-2840, 23-2846 & 23-2849
ȱȄDzȱ¢ȱȱĴȱȱ¢ȱȃ¢ȱ ȱǰȱȱȱęȱȱȄDzȱand pur- ȱ ȱ ȱ ȱ ěȱ ȱ ȱ ȱ ȱ ȃ¡ȱ futures or options contracts or any underlying commodities ȱȱȱȱȱȱĴȱȱȱȱ the market or creating a condition in which prices do not or will not rĚȱȱȱǯȄȱ¢ǰȱȱȬ ants’ employers have ȱȱęǰȱȱex- pressly and implicitly as a form of market manipulation. B Nowak and Smith worked at JPMorgan from 2008 to 2016. Nowak was a managing director who ran the precious metals ȱǰȱĴȱȱȱ ȱ ȱȱȱȬ don. Smith was an executive director and sat next to Nowak ȱȱ ȱȱĜǯȱHe joined JPMorgan after it acquired Bear Stearns in 2008. Jordan worked as a precious metals ȱ ȱ ȱ ȱ ȱĜȱ ȱŘŖŖŜȱȱŘŖŖşǯȱAfter JPMorgan terminated him in 2009, he moved to Credit Suisse, where he worked from March to August 2010. On multiple occasions, Smith, Nowak, and Jordan each placed orders ȱȱȱĴȱȱę. That conduct eventually led the government to indict them for var- ȱ ę-related crimes. Smith and Nowak were each ȱ ȱ Ĵȱ ȱ ǰȱ ŝȱ ǯǯǯȱ § 13(a)(2); wire fraud, 18 U.S.C. § 1343; commodities fraud, 18 U.S.C. § 1348(1); and violating the anti-ęȱȱȱ the Dodd-Frank Actǰȱŝȱǯǯǯȱȗȗ 6c(a)(5)(C) & 13(a)(2). Jordan was charged with wire fraud, 18 U.S.C. § 1343. All three were charged with conspiracyǰȱȱȱ ȱ. Nos. 23-2840, 23-2846 & 23-2849 5
1 Smith and Nowak were tried together. They did not mean- ingfully contest that their trading activity ȱȱ- ȱĴȱȱęǯȱǰȱ¢ȱȱȱȱȬ ȱ ȱ¢ȱȱ ȱȱȱȱȱȱ government could not prove the requisite intent. During a three-week trial, tȱ ȱ ȱ ȱ Ȭ dence to the contrary. Below, we summarize the evidence rel- evant to this appeal. Kumar Venkataraman, ȱ ęȱ ¡ǰ explained how Smith’s and Nowak’s ȱĴȱȱȱęȱȱ ȱ ȱȱ¢ȱęȱĴǯ In a set of 100 trad- ing episodes ęǰȱȂȱȱȱȱȱęȱȱ (meaning ȱ ȱ ȱ ȱ ęǼȱ ȱ ŖǯŗŞƖǰȱ Ȭ ȱȱȱŝşǯŗŗƖȱęȱȱȱȱȱǯȱ Ȃȱęȱ ȱ ȱDZȱŖǯŘŘƖȱȱȱȱȱȱşŖǯŗŗƖȱȱ his genuine orders. ȱȂȱȱȱ ȱ ȱĴ, as well. Deȱȱ¢ȱęȱȱȱthe spoof orders, ȱ ęȱ ȱ Smith and Nowak used this strategy “again and again.” He concluded that Smith and Nowak’s trading strategy was “inconsistent with a design” to ęȱthe genuine orders and was instead “designed” to “push the price” of the market and execute the order on the other side. There was no “economically rational” reason for their Ĵȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ , Venkata- raman said. ȱȱȱȱȱȱȱȱ genuine, they achieved the intended “shock to the market” and successfully moved the price on the other side. The government also presented the testimony of CME in- vestigator Brian Wika and three of Nowak and Smith’s former coworkers who Ĵȱ ȱ ęȱ ȱ ȱ ȱ 6 Nos. 23-2840, 23-2846 & 23-2849
authorities. John Edmonds sat next to Smith and Nowak at Ȃȱ ȱ ȱ Ĝǯȱ ȱ ęȱ ȱ ȱ ȱ Smith and Nowak spoof regularly, at least several times per day, and that they in fact taught him how to spoof. According ȱǰȱȱȱȱȱȱ¢ȱȱȱ “size moves the market” and would complain “[t]hey’re fuck- ȱĴȱȄȱ ȱȱ he placed with the intent to cancel were executed. After a meeting where JPMorgan ȱĜȱ ȱȱȱȱęȱȱ regulators were looking into it, Edmonds heard Smith say to another coworkerǰȱȃȱȱȱǯȄȱEdmonds also ȱȱȱȱěȱ¢ȱȱȱȱ ęȱȱthey ȱȱ ȱęȱĴǯȱȱ on his own experience at JPMorgan, he was unaware of any legitimate non-ęȱ¡ȱȱȱǯ ȱ£ȱ ȱȃȱ¢ȱȄȱ ȱǰȱ m ȱȱȱȱǯȱ£ȱĴȱȱȱȱ ȱęȱȱdeceive other market participants. ȱęȱ that he witnessed Smith spoof “all the time” using the same strategy. Trunz sent a chat to another JPMorgan salesperson ȱȱ ȱȃȱȱȱȱȱ¢ȱȱȱȱ ěǰȄȱ ȱ ȱ ȱ ȱ ȱ ȃ¢Ȅȱ ȱ ȃ ȱ ȱ glasses against the screen” when his spoof order was exe- cuted. ȱ ȱ ęȱ that Nowak spoofed while they ȱ ȃȱ ȱ Ȅȱ ȱ ǯȱ And while coaching Trunz ahead of a compliance review, Nowak once warned, ȃǰȱ¢ȱȱ ȱǰȱ ȱȱȱǯȄ Corey Flaum sat alongȱȱȱȱȱȱȱ ȱȱ¢ȱ ȱȱŘŖŖŞǯ Flaum also confessed to ę ȱęȱthat he saw Smith spoof at least several Nos. 23-2840, 23-2846 & 23-2849 ŝ
times per week. ȱęȱȱȂȱȱȱ ȱȃȱȄȱȱȱ ȱęȱĴǯ At various points, Edmonds, Trunz, and Flaum each de- ęȱ ęȱ ȱ ȱ and ȱ ȱ ȱ and impact on the market to explain why they spoofed. They ȱ ȱ ȱ ȱ ȱ ęȱ ȱ and opined that Smith and Nowak lacked an intent to trade ȱȱȱęȱȱȱ. The district court ěȱȱȱȱ¢ȱȱȱȱ ȱ ȱȱ ȱȱȱȱǰȱȱȱěȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ . Smith and Nowak declined the instruction. The government also Ĵȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ coworkers discussing the legitimate orders they sought to ex- ecute, ȱ ¢ȱ ȱ showing that they simultane- ously placed and quickly canceled large orders on the other side of the market. ȱȱȱ¡ȱ ȱęȱǰȱ ȱ traders typically achieve it, how CME investigators identify ǰȱȱ ¢ȱȱȱȱǯȱȱȱȱȱȬ dentials, Wika stated that he had worked in CME’s investiga- tions group for 14 yearȱȱȱȱȱȱȱȱȱ ǯȱȱęȱȱȱȱȱȱȱ after it received a complaint from a market participant. When discussing the results of his investigation, Wika said that Smith’s high cancellation rates were ȃȱ ȱ ęȱ activity—indicative of lack of intent to trade those orders.” He prepared an investigative report concluding that Smith had ȱ ȱ ¢ȱ ęȱ ȱ ȱ ȱ Ȭ ȱ¡ȱȱȂȱęȱǯȱOver the de- Ȃȱ ǰȱ ȱ ȱ Ĵȱ ȱ ȱ ȱ 8 Nos. 23-2840, 23-2846 & 23-2849
ȱȱȱȱȱȱȱȱȱȱ 803(6). ȱȱȱȱ ǰȱȱ¢ȱȱȱȱ was deadlocked. Although the defendants sought a mistrial, the court granted the government’s request for the standard supplemental jury instruction under United States v. Silvern, ŚŞŚȱǯŘȱŞŝşȱǻŝȱǯȱŗşŝřǼȱǻȱǼǯȱȱ ȱȱȬ ȱȱȱȱȱĴȱǰȱȱȱ ȱȱȱȱȃȱ¢ȱȱěȱȱȱ a verdict” and discuss their “ěȱ ȱȱȱǰȄȱ ȱ ȱ ȱ ȱ ȱ ȃȱ ǽǾȱ ȱ ȱ ȱȱ ȱȱěȱof evidence.” The next day, the jury ȱȱȱ¡ȱȃȱȱȱȱ¢ȱȱ juror” that suggested that the juror had “made early decisions ȱȱȱȱȱȄȱȱȱȱȂȱ opening statements and “saw everything through that lense [sic].” The government asked the court to dismiss the juror. Instead, the court reminded the jury “that the lawyers’ state- ments and arguments are not evidence. If what a lawyer said ȱ ěȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǰȱ ȱ Ȭ dence is what counts.” The next day, the jury found Smith and ȱ¢ȱȱȱȱȱȱĴȱthem of the conspiracy charges. ȱ ȱ ȱ Ĵȱ ȱ ȱ ȱ ǰȱȱȱȱȱȱ. 2 The government tried Jordan separately and, following Smith and Nowak’s trial, dropped the conspiracy counts against him. ǰȱ ȱ ȱ ǰȱ ¢ȱ Ĵed that he spoofed. Instead, his defense was that he spoofed without the criminal intent necessary to support a conviction for wire ȱěȱȱęȱǯ Nos. 23-2840, 23-2846 & 23-2849 9
FBI Aȱ ȱ ȱ ęȱ ȱ ȱ ǯȱ Agent Luca had investigated Jordan and interviewed him in ŘŖŗŞǯȱȱȱ ǰȱ ȱĴȱ as a trader ȱȱȱȃȱȱȱȱȱȱȱȱȱ and larger non-ȱȱȱȱȱȱȱȱȱ in which he intended to cancel” and acknowledged that this Ĵȱ ȱ ȱ ȱ ęǯ He explained that he spoofed “to mislead the market, to outperform the algo- rithms, and to make—and ȱ ȱ ȱ le ęs for his ǯȄȱThe government characterized this interview as a con- fession. However, Jordan also told Agent Luca that he “did not think what he was doing was wrong.” Jordan sought to admit this statement under a hearsay exception commonly ȱȱȱȱȱǰȱǯȱǯȱǯȱŗŖŜǰȱȱȱ court denied his motion. To demonstrate Jordan’s state of mind, the government in- troduced ȱȱȱȱȱŘŖŗŖȱȱȱ the Commodity Futures Trading Commission. In stark con- trast to his interview with Agent Luca, Jordan swore during the deposition that he never placed an order that he didn’t intend to execute, and “would only cancel something if [he] changed [his] mind or if it was put in error.” He further stated ȱȱȱȱȱȱȃȱȱȱȱĚȬ ing the price” of the commodity exchange. ȱȱȱȱȱȱȂȱȱȱȬ ¢ȱ ȱ ǰȱȱǰȱ ȱę. Middle- ton said ȱȱȱȱ¢ȱȱȱȱȱę— meaning they must ȃȱȱȱȱȱ¢ȱȱȱȱ particular price or quantity”—to maintain the integrity of their markets. She acknowledged, however, that the word ęȱȱȱȱȱȱ¡ȱȱȱȱŚřŘ and that 10 Nos. 23-2840, 23-2846 & 23-2849
such language would have ę that the rule ȱ ę. Representatives from JPMorgan and Credit Suisse ¢ȱęȱȱęȱ ȱȱȱȱȬ ȱȱȱ ȱ ȱ ȱȱȱǰȱȱȱ ¢ȱȱȱ¡¢ȱȱę either. Jordan sought to challenge these statements with post-2010 compliance pol- icies and CME documents that, unlike the policies in place ȱȱ ȱȱ¢ǰȱ¡¢ȱęed ęȱȱȱ ȱ ȱ ¢ǯ The district court excluded these ¡ȱȱȱȱȱȱŚŖřǰȱǰȱęȱ that they posed a risk of confusing the jury and erroneously injecting a mistake of law defense into the case. The court also rejected Jordan’s request for a jury instruc- ȱȱȱǯȱę¢ǰȱȱȱȱȱȱ if he acted in good faith, “then he lacked the intent to defraud ȱȱȱȱěȱȱ ȱǯȄȱȱȱȱ ȱȱ¢ȱȱ ȱȱ ¢ȱȬ gue that he lacked an intent to defraud and the instruction posed a similar risk of misleading the jurors into thinking that ȱȱ ȱ ȱȱȱǯȱ The court did, however, instruct the jury on the mental state necessary to show an intent to defraud. The jury was in- structed that Jordan must have acted knowingly, which the ȱęȱȱȱȱ “aware of the nature of his con- ǯȄȱ ȱ ǰȱ ȱ ¢ȱ ȱ ęȱ ȱȱęǰȱȱȱȱȱȱȃȱęȬ tion of ‘knowingly’ … does not mean that the defendant must ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ law.” The jury returned a guilty verdict, and the court denied Jordan’s ȱȱĴȱȱȱ ȱ. Nos. 23-2840, 23-2846 & 23-2849 11
II Smith, Nowak, and Jordan all appeal the denial of their motions ȱĴȱ a new trial. They also challenge sev- eral of the district judge’s rulings at trial. We take their argu- ments in turn. A All three defendants argue that their fraud convictions cannot ȱȱęȱȱ involve a misrepre- ȱ an essential element ȱȱȱȱȱ contracts. We review this threshold legal challenge de novo. United States v. Riversǰȱ ŗŖŞȱ ǯŚȱ şŝřǰȱ şŝŞȱ ǻŝȱ ǯȱ ŘŖŘŚǼȱ ǻde novo review applies to “legal questions wrapped up in chal- ȱ ȱ ȱ Ĝ¢ȱ ȱ ȱ ȄǼǯ Their argument fails for several reasons. Both wire and commodities fraud require as an element “a scheme ȱ ęȱ to defraud.” 18 U.S.C. §§ 1343 & 1348(1). ȱȱȱȱȱȱȱȱȱ only deceptive schemes where “money or property was an ” of the fraud. Ciminelli v. United States, 598 U.S. 306, 312 ǻŘŖŘřǼȱǻȱĴǼǯȱBuilding on this, some circuits pre- viously drew ȃȱęȱȱ ȱȱȱȱȱȱ than cause their victims to enter into transactions they would otherwise avoid” and “schemes that depend for their comple- ȱȱȱȱȱȱȱȱȱȱȬ gain.” United States v. ShellefǰȱśŖŝȱǯřȱŞŘǰȱŗŖŞȱǻŘȱǯȱŘŖŖŝǼǯȱIn ȱǰȱ¢ȱȱĴȱd a scheme to defraud that violated the fraud statutes. Id.; United States v. Takhalov, ŞŘŝȱǯřȱŗřŖŝǰȱŗř13–14 (11th Cir. 2016); United States v. Guer- tinǰȱŜŝȱǯŚȱŚŚśǰȱŚśŗ–52 (D.C. Cir. 2023); United States v. Bruch- hausenǰȱ şŝŝȱ ǯŘȱ ŚŜŚǰȱ ŚŜŝ–69 (9th Cir. 1992). A scheme to 12 Nos. 23-2840, 23-2846 & 23-2849
defraud, they reasoned, requires ȱȱ the “nature of the ȱǰȄȱȱ one concerning “price” or “character- istics of the good,” TakhalovǰȱŞŘŝȱǯřȱȱŗřŗř–14, or some “ex- ȱǽǾȄȱȱȱȱǰȱUnitȱȱǯȱ ĵ, 924 F.2d 410, 420 (2d Cir. 1991). The defendants ask us to apply the essential element of the ȱ ȱ here. ȱ ęȱ ȱ market supply and demand, rather than the price or another characteristic of the futures contract itself, they contend that the other parties to the transaction “received exactly what they paid for.” TakhalovǰȱŞŘŝȱǯřȱȱŗřŗŚ ǻȱĴǼ. They liken ęȱȱȱȱȃȱ positions” in arm’s length transactions, which are not fraud. United States v. WeimertǰȱŞŗşȱǯřȱřśŗǰȱřśŝȱǻŝȱǯȱŘŖŗŜǼǯȱȬ ȱȱ¢ǰȱęȱ¢ȱ not amount to a scheme to defraud within the scope of the fraud statutes, even if the parties would not have otherwise executed the transaction. ȱȱȱȱ¢ȱ¢ȱȱȱȂȱȱȱ Kousisis v. United States, 145 S. Ct. 1382 (2025), which came down after oral argument in this case. There, the defendant had secured contracts with the Pennsylvania Department of Transportation to paint two projects in Philadelphia. Id. at 1388–89. He represented that he would acquire painting sup- ȱ ȱ ȃȱ ęȱ ȱ ,” in ȱȱȱęȱȱȱȱȱȱ pass-through entity to funnel payments. Id. at 1389. The Court Ĝȱthe defendant’s wire fraud convictions and directly ȱȱȱȱȱȱthe defendants in this case in- voke. Id. at 1390–92, 1398; see also United States v. Runner, 143 F.4th 146, 154–śśȱǻŘȱǯȱŘŖŘśǼȱǻ£ȱǼǯȱȱ wire fraud statute, the Court explainedǰȱ ȃȱ ȱ ȱ Nos. 23-2840, 23-2846 & 23-2849 13
economic loss.” Kousisis, 145 S. Ct. at 1392. Rather, a fraud conviction can stand if the defendant did no more than “use[] a material misstatement to trick a victim into a contract that requires handing over her money or property.” Id. at 1388, 1391. Because Kousisis forecloses the defendants’ essential ele- ȱȱȱȱǰ ȱ¢ȱȱȱȱ well-ȱȱȱęȱȱȱȱ to defraud within the meaning of the wire and commodities fraud statutes. United States v. Pacilioǰȱ Şśȱ ǯŚȱ ŚśŖǰȱ ŚŜŖȱ ǻŝȱ Cir. 2023); United States v. Chanu, 40 F.4th 528, 541 ǻŝȱ ǯȱ 2022); United States v. Coscia, 866 F.3d ŝŞŘǰ ŝşŝ ǻŝȱǯȱŘŖŗŝǼ. ȱȱȱȱthe defendants’ argument would fail even without Kousisis. Unlike the transactions in Kousisis or Weimert, where the defendants directly lied to the purported victimsǰȱȱȱȱęȱȱęȱȱȱ parties and the market as a whole. ęȱȱȱȬ ȱ ȱ ȱ ȱ ȱ ȱ — ȃ¢ȱȱȱȱȱȱȱȱȱȱȱ private intent to cancel.” Pacilio, 85 F.4th at 460. In this way, ęȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ d and fraud-on-the-market, ȱ ȱ ¢ȱ ȱ Ȭ sentations that undermine the integrity of the marketplace. See United States v. GilbertsonǰȱşŝŖȱǯřȱş39ǰȱşŚŝ (8th Cir. 2020) (in the context of securities fraud, the “gravamen of manipu- ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ¢ȱȱȱȱȱȱȱ¢ȱȱ ȱ¢ȱȱ¢ȱȱǰȱȱȱ¢ȱȬ nipulators”) (quȱĴǼ. The defendants also claim the fraud statutes are unconsti- tutionally vague as applied to their conduct. We have already 14 Nos. 23-2840, 23-2846 & 23-2849
rejected this argument, Pacilio, 85 F.4th at 460–61, and decline ȱȱȱǯȱȱǰȱ ȱęȱȱȬ ing violates the federal wire and commodities fraud statutes, 18 U.S.C. §§ 1343 & 1348(1). B ȱȱ ȱȱȱȱȱęȱȱȱ wire and commodities fraud (it does), the evidence cannot sustain their convictions ȱ¢ȱȱȱȱȱȱ practice. They ȱȱęȱȱȱȱ intent to cancel a ȱȱěȱȱ¡ǰȱ ȱȱ ȱ ȱ ȱ ǰȱ ȱ ǰ a conditional intent to cancel the orders ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ę. Smith and Nowak concede, as they must, that each of their convictions is predicated on the ȱ ęȱ ǰȱ meaning the convictions all stand or fall with the govern- Ȃȱ ¢ȱ ȱ ȱ ȱ ¢ȱ . ęȱ ȱ ¢ȱȱȱȱȱǰȱȱǰȱȱȱ Dodd-Frank Act provides that any person who knowingly vi- olates its anti-ęȱ ǰȱ ŝȱ ǯǯǯȱ ȗ 6c(a)(5)(C), is ¢ȱ ȱ Ĵȱ ȱ ȱ ȱ ŝȱ ǯǯǯȱ § 13(a)(2). According to Smith and Nowak, since the govern- ment cannot prove that they placed trade orders with the req- ȱȱȱȱęȱǰȱ¢ȱare entitled to a ȱȱĴȱȱȱǯȱThough formally we ȱ ȱ ȱ ¢ȱ Ĵȱ ȱ ǰȱ ȃ¢ȱ Ȅȱ ȱ ȃȱ ȱ ȱ ȱ ȱ ȱ Ĝ¢ȱ ȱ the evidence.” United States v. Peterson, 823 F.3d 1113, 1120 ǻŝȱǯȱŘŖŗŜǼ; see also RiversǰȱŗŖŞȱǯŚȱȱşŝŞȱǻȱȱ ȱ for threshold legal questions). ȱȱȱȱȱȱ¢ȱȱ the standard for intent and, frankly, we struggle to Nos. 23-2840, 23-2846 & 23-2849 15
understand why. ȱȱ ȱ ȱȱĴȱȱȱȬ ȱǯȱȱȱȱȱ¢ȱȱęȱȱ an unconditional purpose to cancel the ȱȱȱě. The jury thus considered—and rejected—Smith and Nowak’s ar- gument that the evidence does not show such unconditional intent. We measure intent at the moment the order was placed. CosciaǰȱŞŜŜȱǯřȱȱŝşśȱǻȱȱȱęȱȬ quires proving that ȱȱȃ ¢ȱȱds or ěȱ ȱȱȱȱȱȱȱȱȱěȱȱȱ execution”). If a party possesses a conditional intent to trade, he necessarily lacks an intent to cancel at the time the order is placed. The word unconditional is redundant and unneces- sary. The key question ȱęȱȱis whether an order is placed on the opposite side of the commodities market with ȱ ȱ ȱ ȱ ȱ¡ȱȱȱȱȱ the market. Chanu, 40 F.4th at 540. Even if Smith and Nowak waited to cancel the spoof orders until the legitimate trade was executed, what counts is that the trades “were designed ę¢ȱ ȱ ȱ ȱ ęǯȄȱ Cosciaǰȱ ŞŜŜȱ ǯřȱ ȱ ŝŞşǯȱ It ȱȱȱȱęȱȱ ȱ to cancel until his ȱ ȱ ȱ ę—that’s the whole point of ęǯ Cf. id. (defendant’s software program canceled ȱȱȃȱȱȱDZȱǻŗǼȱȱȱ the passage of time dzDzȱǻŘǼȱȱȱęȱȱȱȱDzȱ or (3) ȱęȱȱȱȱ”) (emphasis added). The “unconditional” semantics aside, the question is whether Smith’s and Nowak’s “purpose was not to trade on those or- ǰȱȱȱȱȱȱȱȱȱȱȱȱ ǯȄȱ Id. ȱŝşśȱǭȱǯŚśǯ 16 Nos. 23-2840, 23-2846 & 23-2849
More importantly, all that Ĵ is whether a rational trier of fact could have come to that conclusion. Peterson, 823 F.3d at 1120 (“We consider the evidence in the light most fa- ȱȱȱȱȱĜȱȱȱȱ¢ȱȬ ȱȱȱȱȱęȱȱȱ¢ȱ¢ȱȱ ȱǯȄǼǯ ȱȱȱȃ¢ȱ.” United States v. Sorensenǰȱ ŗřŚȱ ǯŚȱ Śşřǰȱ ŚşŞȱ ǻŝȱ ǯȱ ŘŖŘśǼ ǻȱĴǼ. In particular, Smith and Nowak’s focus ȱȱȱȱȃȱ¢ȱ¢ǰȄȱȱȃȱȱ¢ȱȱ ȱ ȱ ȱȱ ȱ ȱ ¢ȱ ȱ ȱ ǰȱȱȱȱȱĜ¢ȱȱȱȱȱ ȱȱ¡¢ȱĜȱȱ ǯȄȱPacilio, 85 F.4th at 463 (cleaned up). ȱ ȱȱǰȱ ȱȱęȱȱȱ¢ȱ could have found Smith and Nowak placed spoof orders with the requisite intent to cancel. ȱȱȱȬ stantial data evidence depicting the scheme, and the jury heard extensive testimony from cooperating witnesses, ęȬ cial experts, and investigators supporting an inference of guilt. This evidence ȱȱ¢ȱȱȱ messages showing that Smith and Nowak placed (and then canceled) orders on the opposite side of the market from the legitimate orders that they intended to execute. Smith and Nowak argue that this evidence is all circumstantial, as the ȱ ęȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ mind. Even so, “direct evidence of intent ȱȱĴǰȱ ȱęȱȱȱȱ¢ȱȱȱ¢ȱȬ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ¡ȱ the scheme itself.” Id. ȱŚŜŚȱǻȱĴǼǯȱǰȱ Smith and Nowak counter that the circumstantial evidence cannot support ȱ ȱ ȱ ȱ ȱ ¢ȱ Ȭ sistent with their innocence. But Smith and Nowak were free Nos. 23-2840, 23-2846 & 23-2849 ŗŝ
to argue to the jury that there were legitimate explanations for their trading Ĵȱand to cross-examine the witnesses on ȱȱȱ¢—indeed, that was the heart ȱȱȱȱȱǯȱȱ ȱȱȱǰȱȃȱ ȱȱȱ¡ȱȱȱ¡ … that were ¢ȱȱ ȱȱȱȱȱȱ¢ȱȱȬ lieve them.” United States v. Maxwell, 143 F.4th 844, 858 ǻŝȱ ǯȱ ŘŖŘśǼȱ ǻȱĴǼǯȱ ȱęȱȱȱĜȱ evidence to justify the jury’s conclusions. Ě¢ǰȱȱȱ ȱalso challenge the anti-ęȱ statuteǰȱŝȱǯǯǯȱȗ 6c(a)(5)(C), as void for vagueness. As with fraud, our precedent squarely forecloses this argument. Coscia, 866 F.3d at ŝşř–95. The ę statute is not uncon- stitutionally vague, and ample evidence supported Smith’s and Nowak’s convictions under it. ȱ ǰȱ ȱ ȱ Ȃȱ Ĝ¢ȱ ȱ ȱ ȱ challenge fails as to all convictions. There was more than suf- ęȱȱȱȱȱ ȱȱȱęȱ ȱ ȱ ȱ ŝȱ ǯǯǯȱ ȗ 6c(a)(5)(C). And ȱ ęȱ constitutes a scheme to defraud under the wire and commod- ities fraud statutes, 18 U.S.C. §§ 1343 & 1348(1), and amounts to Ĵȱȱǰȱŝȱǯǯǯȱȗ 13(a)(2), there was Ĝȱȱȱ¢ȱĴȱȱȱȱ . Because the jury’s verdict is ȱ ¢ȱ ȱ Ȭ dence on all charges, they are not entitled to a new trial on these grounds either. Peterson, 823 F.3d at 1122; United States v. ConleyǰȱŞŝśȱ.3d 391, 399–400 ǻŝȱǯȱŘŖŗŝǼȱǻwhen motions ȱĴȱȱȱȱ ȱȱȱȱȱȱȱȱ that the government failed to prove the elements of the of- ȱ¢ȱȱȱǰȱ ȱ ȱĜȱȱȱȬ ęȱȱȱȱ¢ȱǼǯ 18 Nos. 23-2840, 23-2846 & 23-2849
C Smith and Nowak next ȱȱĴȱȱȱȱ court’s decision to admit much of the evidence discussed . They claim the court’s rulings were erroneous and ask ȱȱȱȱĴȱȱȱ ȱȱȱȱǯȱWe ȱȱȱȂȱ¢ȱȱȱȱȱȱ discretion. Pacilio, 85 F.4th at 464. We will only reverse if we ȱȃȱ ȱȱęȱȱęȱȱȱȱȱ ȱ ȱ ĴǯȄȱ Id. ǻȱ ĴǼǯȱ Smith and Nowak assert that the ȱȱȱȱȱ¢ȱ ĴDZȱǻŗǼȱ¢ȱȱ¢ȱȱȱȱȬ ing witnesses, (2) Wika’s lay opinion testimony, and (3) Wika’s investigative report. ȱǰȱ ęȱȱ ȱ¢ȱȱȱȱ Ĵȱ ȱ rulings. Although the report and some of Wika’s ¢ȱȱę¢ȱȱǰȱȱȱęȱ ȱęȱȱȱȱȱ¢ǯȱ¢ȱ in light of the joint conspiracy charge, even though this evi- dence focused on Smith, the jury could consider it against Nowak as relevant and in accordance with any limiting in- structions. See United States v. LopezǰȱŜȱǯřȱŗŘŞŗǰȱŗŘŞŜȱǻŝȱǯȱ 1993) (in joint trials, juries will “sort through the evidence” and “follow instructions from the court” separate con- ȱ ȱ Ǽǯȱ ȱ ȱ ȇȱ ȱ ȱ Wika's testimony and report and, importantly, did not receive or request a limiting instruction that the jury could only con- sider them against Smith. Turning to the merits, none of these decisions was an ȱȱǯȱȱȱȱȱŝŖŗȱȱ¢ȱ witnesses to testify as to “their opinions and inferences, even ȱȱȱȱȱǯȄȱUnited States v. Locke, 643 Nos. 23-2840, 23-2846 & 23-2849 19
F.řȱŘřśǰȱŘřşȱǻŝȱǯȱŘŖŗŗǼǯȱ¢ȱ¢ȱȱ¢ȱȱȬ other person’s mental state, so long as the testimony is helpful to the jury ȱȱŝŖŗȱand appropriate under Rule 403’s ȱ ǯȱ Id. at 239–40. Though lay witnesses may not ěȱȱȱȱȱȱȱȱȱ¢ȱ elements, United States v. Noelǰȱ śŞŗȱ ǯřȱ ŚşŖǰȱ ŚşŜȱ ǻŝȱ ǯȱ 2009), they can use words “in a colloquial sense” that “em- ¢ǽǾȱ ȱ ȱ ȱ ȱ ęȱ ǰȄȱ ȱ ȱ those words mirror legal standards, Locke, 643 F.3d at 241–42 (approving of lay witnesses’ use of the words fraud and mis- representation). The lay testimony was proper under this standard. Smith and Nowak take issue with the witnesses’ use of the word ęȱȱȱȱȱ¢ǰ argue that they im- ¢ȱ£ȱęȱȱȱȱȱȬ lation, and claim that the witnesses inappropriately opined on Smith’s and Nowak’s intent ȱ ȱ ȱ . But the witnesses’ ȱ ȱ ȱ ȱ ęȱ ȱ ȱ with the colloquial vernacular of the trading industry, id. at 242ǰȱ ȱ ¢ȱ ¢ȱ ȱ ’s and Nowak’s mental states was properly framed as an opinion drawn from the trading data in light of the witnesses’ own experience, see United States v. Winbush, 580 F.řȱśŖřǰȱśŗŘȱǻŝȱǯȱŘŖŖşǼǯȱȱ for the claim that the testimony improperly equated ęȱ with fraud and price manipulation, the witnesses were simply explaining why they spoofed, what they had pled guilty to, ȱ ǻȱ Ȃȱ Ǽȱ ¢ȱ ȱ ȱ ȱ ęǯȱ That Smith and Nowak declined a jury instruction that the wit- ȱ ȱȱȱȱȱȱ¢ȱęȱȱȬ ȱȱȱȱȱȱǯ 20 Nos. 23-2840, 23-2846 & 23-2849
Smith and Nowak also argue that Wika should have testi- ęȱȱȱ¡ǰȱȱȱȱ¢ȱ ǯȱInvestigators such as Wika can ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ provide their impressions of the case without crossing into ex- pert territory. United States v. Rollins, 544 F.3d 820, 832–řřȱǻŝȱ ǯȱŘŖŖŞǼǯȱȱȱȱȱȱȱȂȱ“spe- cialized knowledge informed his mental state.” United States v. OriedoǰȱŚşŞȱǯřȱśşřǰȱŜŖŘȱǻŝȱǯȱŘŖŖŝǼǯȱȱȱ¢ȱ ȱȱexperience, Wika’s testimony was limited to his investigation and what it revealed to him. Even if it argu- ¢ “approaches the line dividing lay opinion testimony from expert opinion testimony,” this testimony remained ad- ǰȱȱȱȱȱȱȱȱȱȱȱ allowing it. Rollins, 544 F.3d at 833. At the very worst, the gov- ernment’s other evidence of Smith’s and Nowak’s guilt was so extensive that even if Wika’s testimony “had crossed the Ȅȱȱ ȱȱȱȱǯȱId. Nor was it erro- neous to admit Wika’s investigative report: investigative re- ports may come in as ȱ ȱ ȱ ȱ ȱ ȱ Evidence 803(6) if ȱȱĜȱȱȱęȱ at trial, as Wika did. United States v. KingǰȱŜŗřȱǯŘȱŜŝŖǰȱŜŝŘ– ŝřȱǻŝȱǯȱŗşŞŖǼDzȱUnited States v. BlackburnǰȱşşŘȱǯŘȱŜŜŜǰȱŜŝŘȱ ǻŝȱǯȱŗşşřǼȱǻȃȱȱȱȱȱȱȱȱȱȱ for cross-examination, dzȱ Ĵȱ ȱ ȱ ȱ does not contravene Rule 803(8).”). D ȱȱęȱǰȱȱȱ ȱȱȱthe district court’s supplemental instructions to the jury were er- ȱ ȱ ¢ȱ ǯȱ ȱ ȱ ȱ ȱ court’s decision to provide supplemental jury instructions, in- cluding its decision to read (or reread) a Silvern instruction, Nos. 23-2840, 23-2846 & 23-2849 21
ȱȱȱǯȱUnited States v. Cardena, 842 F.3d 959, şŝŚȱǻŝȱǯȱŘŖŗŜǼȱ(Silvern instruction); United States v. Sims, řŘşȱ ǯřȱ şřŝǰȱ şŚŘȱ ǻŝȱ ǯȱ ŘŖŖřǼȱ ǻȱ ȱ generally). It was appropriate for the district court to reread the Sil- vern instruction. Silvern instructions are “perfectly content- ȱ ȱ ǽ¢Ǿȱ ȱ ȱ ȱ ȱ ȱ ȱ jury.” United States v. BeverlyǰȱşŗřȱǯŘȱřřŝǰȱřśŘȱǻŝȱǯȱŗşşŖǼ ǻȱ ĴǼ. Rather, the district court followed our ǰȱȱȱȱȱȱȱinstruc- tion to a deadlocked jury. United States v. Collins, 223 F.3d 502, 508–ŖşȱǻŝȱǯȱŘŖŖŖǼǯ It was similarly within the court’s discretion to remind the jury that lawyers’ statements are arguments, not evidence. ȱȱȱȱȱȱȱĴȱȱǰȱ despite Smith and Nowak’s arguments to the contrary, crafted it carefully to avoid singling out the juror. The court also refused to ask the jury whether it remained deadlockedǰȱ¢ȱȱȱ¢ȱȱȱȱȬ ations. Under the circumstances, the court proceeded thoughtfully and exercised its discretion wisely. Nowak and Smith may not like the verdict the jury reached after receiving ȱǰȱȱȱ not mean they were issued in error. III Recall that Jordan was convicted of only wire fraud under 18 U.S.C. § 1343. Jordan challenges various aspects of his sep- ȱǯȱȱȱȱǰȱ ȱęȱȱ ȱȱ ȱ Ĝ¢ȱ ȱ ȱ ȱ claim. Jordan joined Smith ȱ Ȃȱȱȱęȱȱȱȱȱ 22 Nos. 23-2840, 23-2846 & 23-2849
ȱȱȱȱȱȱȱȱȬ ȱȱȱȱǯȱȱ ȱȱȱ he spoofed, his argument hinges entirely on that theory. Since we rejected that argument in Section II.Aǰȱ ȂȱĜ¢ȱ of the evidence challenge is a nonstarter. A Jordan’s primary remaining argument concerns the dis- trict court’s decision to exclude his statement to Agent Luca that he “did not think what he was doing was wrong.” Jordan sought to admit that statement under Federal Rule of Evi- dence 106, often referred to as the rule of completeness. Rule 106 provides that “[i]f a party introduces all or part of a state- ment, an adverse party may require the introduction, at that time, of any other part—or any other statement—that in fair- ȱȱȱȱȱȱȱȱǯȄȱ¢ȱȃ¢ȱȱ sȱȱȱ¢ȱǯȄȱǯȱǯȱǯȱŗŖŜǯ Rule 106 requires a complete statement “ȱ ȱ ȱ ȱ ȱ ȱȱȱ¢ȱȱǻŗǼȱ¡ȱȱĴȱǰȱ ǻŘǼȱȱȱĴȱȱȱ¡ǰȱǻřǼȱȱȱ the trier of fact, or (4) insure a fair and impartial understand- ing.” United States v. LewisǰȱŜŚŗȱǯřȱŝŝřǰȱŝŞśȱǻŝȱǯȱŘŖŗŗǼȱ ǻȱĴǼǯȱIn United States v. Haddad, 10 F.3d 1252 ǻŝȱ ǯȱ ŗşşřǼǰȱ ȱ ǰȱ ȱ ȱ Ĵȱ ȱ Ȭ Ȃȱǰȱȃǰȱ ȱe ȱȱȱȄȱȱ¡ȱ the remainder that said, ȃȱ ȱȱȱ ȱȱȱǯȄȱ Id. at 1259. The gun was located right next to the marijuana, ȱȱĴȱȱȱȱȱȱ ȱ ȱȱȱȱ ǯȱId. We concluded that Rule 106 required the admission of the exculpatory statement, in part to avoid any misleading inference. Id. “The completeness doctrine does not, however, require introduction of portions of a Nos. 23-2840, 23-2846 & 23-2849 23
statement that are neither explanatory of nor relevant to the Ĵȱ ǯȄȱ Lewisǰȱ ŜŚŗȱ ǯřȱ ȱ ŝŞśȱ ǻquoting United States v. MarinǰȱŜŜşȱǯŘȱŝřǰȱŞŚȱǻŘȱǯȱŗşŞŘǼ). The district court was correct that Rule 106 did not compel it to admit Jordan’s statement. Jordan’s claim that he did not think he was doing anything wrong does not change the fac- ȱȱȱȱȱȱȱȱȱ ȱȱȱ to cancel for the purpose of misleading the market. Nor does ȱȱȱ¢ȱ¢ȱȱȱȱȱȬ Ĵȱ DZȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱȱȱȱȱ ȱ ȱ ȱȱęȱ was illegal or, as he put it, wrong. United States v. Blagojevich, ŝşŚȱ ǯřȱ ŝŘşǰȱ ŝřşȱ ǻŝȱ ǯȱ ŘŖŗśǼǯȱ Rather, the court rightly noted that Jordan was free to take the stand and testify that he did not think he was doing anything wrong. United States v. FarukiǰȱŞŖřȱǯřȱŞŚŝǰȱŞśŝȱǻŝȱǯȱŘŖŗśǼȱǻȱ ȱȱȬ trict court that “the appropriate vehicle for the introduction of ȱȱ ȱȱȱȱǽȱǾȱȱȱ have taken the stand,” rather than Rule 106). Defendants often try to justify their actions when speaking with FBI agents and ȱ ȱȱĜȱ¢ȱȱȱ. Such an exculpatory statement has nothing to do with Rule 106. B Next, Jordan claims ȱȱȱȱȱȱȬ ȱ¢ȱexcluding ȱ¡ȱhe sought to use for im- peachment. The trial concerned Jordan’s trading when he worked at JPMorgan and then Credit Suisse until August 2010. Although the policies in place at the time did not explic- ¢ȱȱęǰȱȱ ȱęȱȱthe practice ȱ ȱ ȱ ¢ȱ ȱ ǯ To 24 Nos. 23-2840, 23-2846 & 23-2849
controvert this testimony, Jordan moved to admit post-2010 ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭ ȱȱ¡¢ȱȱęǯȱHowever, the district court excluded this evidence under Rule 403. The policies were re- ȱ ȱ Ěȱ ȱ -Frank Wall Street and Consumer ȱǰȱ. L. No. 111-ŘŖřǰȱŗŘŚȱǯȱŗřŝŜǰȱ ȱȬ ȱȱȱ ¢ȱŘŖŗŖǯȱȱęȱ¢ȱȱ the general wire fraud statutes, the Dodd-Frank Act amended ȱ¢ȱ¡ȱȱȱ¡¢ȱ£ȱęȱ as an unlawful disruptive practice amounting to price manip- ǯȱŝȱǯǯǯȱȗȗ 6c(a)(5)(CǼȱǻęȱęȱȱȱ ȃ¢ȱ ȱȱȱȄȱȱȃȱȱěȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱ ȱ ¡ȄǼȱ ǭȱ ŗřǻǼǻŘǼǯȱ ȱȱǰȱȱȱȱȱȱ ěȱȱȱȱȱȱg the jury and erro- neously injecting a mistake of law defense into the case and excluded it. ȱ ȱ ȱ ȱ ¡ȱ ȱ ¢ȱ ȱ and directly responsive to the witnesses’ testimony, making the district court’s decision to exclude them under Rule 403 ȱȱȱǯȱWe disagree. Jordan conducted exten- sive cross-examination of the witnesses, which elicited clear testimony that the pre-2010 documents did not mention ęȱȱȱ¢ȱȱǻȱ¢ȱǼȱǯȱ The actual documents themselves were thus of limited addi- ȱ ȱ ǯȱAnd as the district court noted, evi- dence of the after-the-fact, post-Dodd-Frank policies could have led the jury to think that the wire fraud statutes were not Ĝȱȱ£ȱęǯ ȱȱ ȱęȱ ȱ this concern—defendants in other cases have made this same ǻȱǼ argument. See Chanu, 40 F.4th at 534. Nos. 23-2840, 23-2846 & 23-2849 25
ȱȃǽǾȱȱȱȱȱȱȱ is a highly discretionary assessment,” we give “great defer- ence” to a district court’s decision to exclude evidence under Rule 403 and “¢ȱǽǾȱȱȱȱȱȱȱ agree with the ruling.” Pacilio, 85 F.4th at 465 (quotation omit- ted). ȱȱȱȱȱȱȱȱȱ valid concerns of jury confusion, Jordan does not overcome this extremely deferential standard. C Last, Jordan says the district court erred in declining to ad- minister a so-called good faith instructionǯȱę¢ǰȱ ȱ sought an instruction that if he acted in good faith “then he ȱȱȱȱȱȱȱȱȱěȱȱ wire fraud.” Though we review challenges to jury instruc- tions de novo, ȃȱȱȱȱěȱȱȬ cretion with respect to the precise wording of instructions so long ȱȱęȱǰȱȱȱȱ ǰȱ¢ȱȱȬ rectly states the law.” Chanu, 40 F.4th at 542 (cleaned up). We already addressed and rejected this same argument in Chanu. As the district court noted, a good faith instruction is ¢ȱȱ ȱȱȱȱȱȱȱȱȱ is a part of the charge. Id. ȱśŚřǯȱȱȱǰȱȱȱȱ free to provide one if they wish. But “the rule is clear” that defendants such as Jordan cannot demonstrate that “the fail- ure to include the good faith instruction denied [them] a fair trial.” Id. (cleaned up). ȱĴȱȱǰȱȱȱ¢ȱ convicted him of wire fraud after a fair, properly conducted trial. AFFIRMED