Securities & Exchange Commission v. Bonan Huang

NOT PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _____________ No. 16-2390 _____________ SECURITIES & EXCHANGE COMMISSION v. BONAN HUANG; NAN HUANG Nan Huang, Appellant ______________ Appeal from the United States District Court for the Eastern District of Pennsylvania (E.D. Pa. No. 2-15-cv-00269) District Judge: Honorable Mark A. Kearney ______________ Submitted Under Third Circuit L.A.R. 34.1(a) March 14, 2017 ______________ Before: GREENAWAY, JR., SHWARTZ, Circuit Judges, and SIMANDLE, Chief District Judge.* (Filed: April 10, 2017) ______________ OPINION** ______________ SHWARTZ, Circuit Judge. * Honorable Jerome B. Simandle, Chief District Judge of the United States District Court for the District of New Jersey, sitting by designation. ** This disposition is not an opinion of the full Court and pursuant to I.O.P. 5.7 does not constitute binding precedent. The Securities and Exchange Commission (“SEC”) brought an insider trading enforcement action against Nan Huang. A jury found that Huang engaged in unlawful insider trading, and the District Court ordered, among other things, that Huang disgorge more than $4.4 million. Huang appeals, arguing that the SEC produced insufficient evidence of the materiality of the insider information upon which he traded, and that the District Court abused its discretion by refusing to ask potential jurors a specific question during the voir dire process, permitting prejudicial arguments about his departure from the United States, and ordering him to disgorge his profits. Because Huang fails to show that the District Court committed error, we will affirm. I A Huang was employed as a senior data analyst for Capital One Financial Corporation. In violation of the company’s confidentiality policies, Huang downloaded and analyzed confidential information regarding purchases made with Capital One credit cards at over 200 consumer retail companies and used that information to conduct more than 2000 trades in the securities of those retail companies. On January 16, 2015, Capital One terminated Huang’s employment due to his violation of company policies. The next day, Huang boarded a flight to his home country of China. Four days later, the SEC filed a complaint against Huang, alleging violations of § 10(b) of the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and Rule 10b-5, 17 C.F.R. § 240.10b-5(a). 2 Before trial, the District Court heard arguments about the admissibility of evidence related to Huang’s departure from the United States and his decision not to return for trial. The District Court rejected the SEC’s request to instruct the jury that it could draw the adverse inference that Huang’s departure and nonappearance at trial showed a consciousness of guilt. It did, however, allow the parties to present a stipulation that Huang flew to China the day after he lost his job “fearing repercussions from his violation of Capital One’s policies and procedures,” and permit each party to argue its view concerning the inference to be drawn from this evidence. App. 433. Before the District Court empanelled a jury, Huang requested that the District Court ask potential jurors whether any of them possessed “a Capital One credit card during the relevant time period.” App. 26. The District Court declined to ask that question, explaining that the case did not involve the misuse of individual Capital One customers’ information, but the Court did ask whether potential jurors or anyone in their families had been employed by a credit card company or by Capital One specifically. B The primary dispute at trial was whether the Capital One data was material. Huang admitted to trading based upon nonpublic information but disputed whether, as the District Court explained, “a reasonable investor would have considered [the information] significant in making an investment decision.” App. 669. At the start of the trial, the District Court read a number of stipulated facts to the jury, including that: Huang conducted thousands of searches of the Capital One database to collect confidential credit card revenue data of over 100 consumer retail companies; he 3 “lacked a business reason to conduct the searches”; he understood that the information he “accessed and used” was nonpublic; he “personally traded in the stock of consumer retail companies” on the basis of that information; and he understood that he was violating Capital One’s policies by engaging in this conduct. App. 433. In addition to these stipulations, the SEC presented a witness from Capital One who testified that the Capital One database captured data of all Capital One credit card transactions, that Huang conducted thousands of searches of retail companies’ aggregated transaction data, and that Huang’s searches served no job-related function. The Capital One witness also testified that Huang had received training on Capital One’s confidentiality policies and on insider trading laws, and was terminated for violating the company’s confidentiality policies. The SEC also offered the testimony of Stephen Graham, an SEC employee and expert in economic and statistical analysis. Graham showed the jury examples of the many spreadsheets that Huang used to analyze nonpublic information from the Capital One database along with historical, publicly available data. The spreadsheets used a combination of public and nonpublic information to create projections of total revenue for approximately 226 publicly traded retail companies for a given quarter approximately one month before the companies issued public announcements about this information. Graham also described a correlation analysis he performed on the data Huang compiled on the 226 retail companies. Graham found that though the nonpublic Capital One data accounted for an average of only 2.4% of the companies’ total sales, for 132 of the companies, the sales data had a statistically significant relationship to the companies’ 4 total revenues. This correlation indicated that, for those 132 companies, the Capital One data was a demonstrably valuable tool to predict the total revenues for those companies in a given time period.1 Graham also showed the jury that Huang’s files compared his revenue projections to an amalgam of leading analyst reports about the companies’ expected earnings. These reports were publicly available studies of the companies based on “as much . . . information as [the analysts] can possibly glean from anywhere to try and figure out what the companies are going to report.” App. 456. Graham concluded that by comparing his revenue projections to the analyst reports, Huang was able to predict whether a given company’s quarterly revenue announcements would outperform “Wall Street’s estimates,” in which case the stock price would likely increase, or underperform them, in which case the stock price would likely decrease. App. 456. According to Graham, “having information on the revenue from Capital One’s systems is going to help a reasonable investor make better decisions on [whether] to buy or sell securities in these individual companies.” App. 455. On cross-examination, Graham conceded that he did not analyze the total mix of information available to the public at the time Huang made each trade but stated that the total mix was incorporated into the analyst reports, which “include pretty much 1 Huang’s economic analysis expert testified that a correlation between two sets of data cannot alone establish a predictive or causal relationship, and that his statistical analyses, including a correlation analysis, showed that the Capital One data could not be used to accurately predict changes in the retail companies’ stock prices. He did not dispute, however, Graham’s analysis that the Capital One data correlated with the total revenues of the retail companies. 5 everything.” App. 491. Graham also conceded that there are data points aside from revenue that might interest an investor before purchasing a security, such as a company’s earnings per share, mergers and acquisitions, and cash flow, and that a correlation does not necessarily result in an accurate prediction. Graham and a separate witness, Matthew Cain, also testified about Huang’s profits. Graham testified that Huang’s computer contained information about his trading activity which indicated that over a three-year period he earned a net profit of approximately $1.48 million, earning a 12,929% return on his initial investment.2 Cain testified that Huang traded profitably in 105 of the 226 companies whose data he downloaded and analyzed, and that Huang’s total “positive profits” from those trades was $4,403,545. App. 518. On cross-examination, Cain stated that his calculation did not subtract from the profits for any losses Huang incurred, and thus did not reflect Huang’s net profits. After the closing arguments, which focused on whether the Capital One data was material, the District Court instructed the jury on the elements of insider trading and then stated that “all of you . . . must then agree as to whether Nan Huang engaged in insider trading as to at least one specific trade.”3 App. 670. The verdict sheet simply asked 2 At a hearing addressing a disgorgement calculation, Huang argued that Graham’s net profit figure was inaccurate since it merely compared his bank account statements at two different points in time and did not specifically account for his profits from insider trades. As such, Huang’s counsel stated that “I have no idea” what Huang’s net profits were. App. 676. 3 At other points, the jury instruction seemed to indicate that the jury must find Huang liable on every trade at issue in order to find liability. For example, in explaining the elements of insider trading, the District Court instructed that “the SEC must prove by 6 “[d]id Nan Huang engage in insider trading?” and the jury unanimously answered “yes.” App. 733. C After the jury returned the verdict, the District Court conducted a hearing and requested briefing to consider remedies. The SEC argued that the District Court should impose the equitable remedy of disgorgement in the amount of Huang’s total positive profits, namely the $4,403,545 calculated by Cain. Huang declined to offer an alternative calculation of a reasonable disgorgement figure and argued instead that the verdict form was too general to impose any amount of disgorgement since the number of transactions for which the jury actually found Huang liable was not clear. The District Court, faced with no alternative evidence from Huang, ordered disgorgement of $4,403,545, finding that Cain’s calculation amounted to a “reasonable approximation of Huang’s ill-gotten gains.”4 App. 758. Huang then moved for judgment as a matter of law, arguing that the SEC failed to present evidence regarding the total mix of information available about each individual trade Huang made or the effect the Capital One data would have on a reasonable investor, and thus did not produce sufficient evidence for a reasonable jury to find that the Capital a preponderance of the evidence the information accessed and used by Mr. Huang was both material and nonpublic at the time he traded in each of the stocks.” App. 669. As discussed herein, it is immaterial to this appeal whether the jury found Huang liable for one or all of the trades. 4 The District Court also imposed prejudgment interest of $288,965, a civil penalty of $8,807,090, and enjoined Huang from committing any further violations of § 10(b) or Rule 10b-5. 7 One data was material. The District Court denied the motion, concluding that the SEC produced ample evidence to sustain the jury’s verdict. Huang appeals. II5 Huang argues that the District Court erred by: (1) denying Huang’s motion for judgment as a matter of law since the SEC failed to prove that the nonpublic information he traded upon was material; (2) refusing to voir dire potential jurors on whether they were Capital One credit card holders; (3) admitting evidence that the SEC characterized as Huang’s “flight” from the United States; and (4) ordering Huang to disgorge more than $4.4 million of profits. Huang fails to show that the District Court committed error on any of these grounds. A Huang first argues that the District Court erred in denying his motion for judgment as a matter of law, claiming that the SEC presented insufficient evidence to show that the Capital One data was material. Our review of a District Court’s denial of such a motion is plenary and we apply the same standard as the District Court. SEC v. Teo, 746 F.3d 90, 98 (3d Cir. 2014). “[I]n entertaining a motion for judgment as a matter of law, the court should review all of the evidence in the record,” and in doing so we “must draw all reasonable inferences in favor of the nonmoving party, and [we] may not make credibility determinations or weigh the evidence.” Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 150 (2000) (citations omitted). Consequently, we “disregard all evidence 5 The District Court had jurisdiction under 28 U.S.C. § 1331, and we have jurisdiction pursuant to 28 U.S.C. § 1291. 8 favorable to the moving party that the jury is not required to believe,” id., and we “determine whether the evidence is sufficient to sustain liability,” Acumed LLC v. Advanced Surgical Servs., Inc., 561 F.3d 199, 211 (3d Cir. 2009) (internal quotation marks and citation omitted). Materiality is an essential element of civil liability under § 10(b) and Rule 10b-5. Basic Inc. v. Levinson, 485 U.S. 224, 231-32 (1988); United States v. Schiff, 602 F.3d 152, 171 n.25 (3d Cir. 2010). For information to be material, “‘there must be a substantial likelihood that the . . . fact would have been viewed by the reasonable investor as having significantly altered the ‘total mix’ of information.’” Basic, 485 U.S. at 231-32 (quoting TSC Indus., Inc. v. Northway, Inc., 426 U.S. 438, 449 (1976)). To consider the effect of nonpublic information on the “total mix” of public information, we need not consider all potentially relevant data but must engage in a fact-specific inquiry and examine enough contextual factors to determine if the insider information would have affected a “reasonable investor’s” view of a particular investment decision. See Matrixx Initiatives, Inc. v. Siracusano, 563 U.S. 27, 44 (2011) (citation omitted); Basic, 485 U.S. at 234. A company’s revenue information will often satisfy the materiality requirement since “[u]nquestionably a factfinder could draw the reasonable inference that a reasonable investor would see the obvious connection between increased revenues and the likelihood of increased profits.” Rothberg v. Rosenbloom, 771 F.2d 818, 821 (3d Cir. 1985). Graham testified that Huang used the nonpublic Capital One data in tandem with publicly available information to gain a valuable informational advantage over the public. 9 Specifically, Graham testified that Huang used the nonpublic Capital One information to create a historical analysis of all Capital One credit card transactions at a given retail company on a monthly basis, and compared that information with publicly available historical data about each company’s total reported revenue. By analyzing that combination of public and nonpublic information, Huang identified the percentage of each company’s sales that historically involved a Capital One credit card. Using this percentage and the Capital One data available to him at the end of each fiscal quarter, he created accurate projections of the company’s total revenue for that quarter.6 Because a public company will typically announce its quarterly revenues approximately 30 days after the end of a given fiscal quarter, Huang was able to make revenue projections approximately one month before the companies publicly announced their actual quarterly revenues. This insider information gave Huang early and nonpublic insight into whether 6 Huang concedes that if he had insider information about the total revenue of a given company then such information would have been material, but, he argues, his access to an average of 2.4% of the companies’ revenues was too insignificant for him to have predicted total revenue information. To be sure, information about an insignificantly small percentage of revenue might not be material. See, e.g., In re Westinghouse Sec. Lit., 90 F.3d 696, 714-15 (3d Cir. 1996) (concluding that a loan loss reserve that amounted to 0.54% of Westinghouse’s net income was quantitatively immaterial). However, Graham testified as to how Huang used the nonpublic Capital One data (in tandem with publicly available information) to predict total revenue information with greater accuracy than analysts using only publicly available information. This testimony provided “evidence upon which the jury could properly” have found that the nonpublic Capital One data allowed Huang to predict the companies’ total revenues before such information was public. Lightning Lube, Inc. v. Witco Corp., 4 F.3d 1153, 1166 (3d Cir. 1993) (citation omitted). In arguing against the possibility of predicting the total revenue data from the Capital One data, Huang essentially asks us to credit the testimony of his rebuttal witness over Graham’s, which we may not do in reviewing a motion for judgment as a matter of law. Reeves, 530 U.S. at 150. 10 the companies were likely to under- or over-perform expert predictions, and thus whether their stocks were likely to increase or decrease after the quarterly announcements. Understood in the context of the publicly available historical revenue data and the analysts’ predictions, the nonpublic Capital One data altered the total mix of information an investor would have had to make investment decisions. As reflected by Huang’s own investment decisions, which he admitted were based upon the Capital One data and resulted in a 12,929% three-year return on his investment, the nonpublic Capital One data “significantly altered the ‘total mix’ of information” in the eyes of a reasonable investor. Basic, 485 U.S. at 232; see also Rothberg, 771 F.2d at 821 (concluding that the fact that the defendants, who were “experienced investors,” acted based upon the nonpublic information evidenced its materiality). Thus, the SEC presented sufficient evidence of materiality and the District Court correctly rejected Huang’s challenge to it. B Huang also argues that the District Court committed reversible error in declining to ask potential jurors his suggested voir dire question concerning whether any of the jurors were Capital One credit card holders. Generally, district courts are afforded “broad discretion as to the questions to be asked on voir dire examination of prospective jurors,” United States v. Napoleone, 349 F.2d 350, 353 (3d Cir. 1965) (internal quotation marks and citation omitted), and so we review a court’s refusal to ask a question during voir dire for abuse of discretion, Butler v. City of Camden, 352 F.3d 811, 815 (3d Cir. 2003). 11 An abuse of discretion occurs where a district court’s examination of prospective jurors is “so general that it does not adequately probe the possibility of prejudice.” Waldorf v. Shuta, 3 F.3d 705, 710 (3d Cir. 1993). Because a district court must “make those inquiries necessary to satisfy both its duty to select an impartial jury and allow for intelligent exercise of peremptory challenges,” we have only “found error and reversed in cases where the district court barred all inquiry into a relevant subject matter designed to elicit a disqualifying prejudice.” Butler, 352 F.3d at 815-16 (emphasis omitted); see also Ristaino v. Ross, 424 U.S. 589, 594 (1976) (holding that a defendant is not always entitled “to have questions posed during [v]oir dire specifically directed to matters that conceivably might prejudice veniremen against him”). Huang requested that the District Court ask potential jurors whether any of them possessed “a Capital One credit card during the relevant time period.” App. 26. Huang argued that the question was appropriate because a juror who possessed a Capital One credit card might hear the trial evidence and think that Huang misappropriated his or her personal information. The District Court declined to ask the question, reasoning that the case did not involve the misappropriation of personal information but only the misuse of aggregated credit card data. The District Court’s decision to decline to pose the question did not amount to an abuse of discretion. Although the Capital One database “contain[ed] the confidential credit card activity of its customers,” App. 433, and Huang’s job entailed monitoring individual cardholder transactions for fraud, as the District Court accurately stated, the case involved Huang’s use of aggregated Capital One credit card data. Indeed, there was 12 no evidence that Huang used any personal information of individual Capital One customers. The exhibits entered into evidence best illustrate this point; they all contained charts or graphs about the retail companies’ sales and the trades Huang made based on this data. As the Capital One witness explained, Huang was terminated for conducting searches “aggregating by merchant” and not based on cardholders. App. 440. Because the personal information of Capital One customers was not at issue in the case nor was it compromised by Huang’s activities, Huang’s proposed question did not seek information about possible juror prejudice regarding a “relevant subject matter.” Butler, 352 F.3d at 816 (emphasis omitted). Thus, the District Court did not abuse its discretion in declining to pose Huang’s proposed question to potential jurors. C We next consider whether the District Court erred by allowing the SEC to offer evidence about Huang’s departure from the United States and argue it constituted “flight” indicative of his consciousness of guilt. We review the District Court’s decision to admit the evidence for an abuse of discretion. United States v. Green, 25 F.3d 206, 210 (3d Cir. 1994). As with all evidentiary decisions, evidence of a defendant’s flight is subject to the balancing test prescribed by Fed. R. Evid. 403. Rule 403 provides that a trial court “may exclude relevant evidence if its probative value is substantially outweighed by a danger of . . . unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.” Fed. R. Evid. 403. We afford District Courts “substantial discretion when striking a Rule 403 balance with respect to proffered evidence.” McKenna v. City of Phila., 582 F.3d 447, 461 (3d Cir. 2009). We 13 will not overturn a District Court’s decision “to admit or exclude evidence under Fed. R. Evid. 403 . . . unless it is arbitrary and irrational.” Id. (internal quotation marks omitted). Evidence of a defendant’s flight after committing wrongdoing can be probative of his consciousness of guilt. See United States v. Pungitore, 910 F.2d 1084, 1151 (3d Cir. 1990). Flight evidence is generally admissible, and is not unduly prejudicial, where, for example, there is evidence that the defendant knew at the time of flight that he might be a target of investigation, whether or not formal charges had yet been filed. See, e.g., id. (holding that the district court did not abuse its discretion in admitting flight evidence where the flight occurred after defendant found out that a member of his criminal operation had begun cooperating with a government investigation).7 Before trial, Huang argued that the SEC should not be permitted to characterize Huang’s departure from the United States as flight evidence. Huang argued that he left the United States because he was allowed in the country only under the terms of a work visa that was no longer valid once he was terminated from his job. The SEC argued that the District Court should instruct the jury that it could draw the adverse inference that Huang had a consciousness of guilt based on the facts that he fled the country the day after Capital One fired him and invoked his Fifth Amendment rights, declining to return to the United States and testify at his trial. The District Court declined to instruct the jury 7 Huang argues that the probative value of flight evidence depends on the SEC’s ability to establish four specific inferences articulated in United States v. Scarfo, 711 F. Supp. 1315, 1321 (E.D. Pa. 1989). However, in our opinion affirming the Scarfo convictions, we did not adopt the district court’s test for the admissibility of flight evidence and instead analyzed the evidence as we would any other form of evidence, considering its probative value and potential prejudicial effect. See Pungitore, 910 F.2d at 1150. 14 regarding an adverse inference, but admitted evidence regarding Huang’s departure and visa status and permitted each party to argue its respective position on the inferences to be drawn from his departure. Both parties accepted the Court’s invitation and presented their positions during their closing arguments. The District Court did not abuse its discretion in admitting evidence of Huang’s departure and permitting the SEC to characterize it as flight evidence. Huang left the United States and flew to China the day after he was terminated from his job, four days before the SEC brought this case against him. Though there is no evidence showing that Huang knew at the time of his departure that the SEC was about to bring an insider trading case against him, there is sufficient evidence to infer he left the country because he feared that he could be under government investigation: Huang stipulated to having traded on nonpublic information, and to knowing, at the time he made those trades, about the insider trading laws, and that his actions violated Capital One policies. Thus, when Huang was told that he was terminated for violating the company’s policies, there is good reason to believe that he knew he had been caught trading on nonpublic information, and that he quickly left the country out of a fear that Capital One had disclosed his activities to a government entity. While it is also plausible, but not inevitable, that Huang left the country because of his immigration status, there was sufficient evidence in the record to draw an inference that Huang left the United States because of his wrongdoing and a potential investigation into it. Thus, the District Court’s decision to allow each party to argue its respective plausible interpretation of the evidence was not “arbitrary and irrational,” McKenna, 582 15 F.3d at 461, and hence it did not abuse its discretion in admitting evidence of Huang’s potential flight from the United States and allowing the parties to argue the reasonable inferences to be drawn from such evidence. D We next review the District Court’s disgorgement order. Disgorgement is an equitable remedy. Commodity Futures Trading Comm’n v. Am. Metals Exch. Corp., 991 F.2d 71, 78-79 (3d Cir. 1993). “[T]he SEC’s use of the disgorgement remedy has been constructed around two objectives: to deprive a wrongdoer of his unjust enrichment and to deter others from violating securities laws.” Teo, 746 F.3d at 104-05 (internal quotation marks and citation omitted). “In crafting any disgorgement remedy . . . the district court should keep in mind the limitation placed on its equitable powers by th[e] requirement that there be a relationship between the amount of disgorgement and the amount of ill-gotten gain.” Am. Metal Exch. Corp., 991 F.2d at 79. We have adopted a burden-shifting approach to calculating a disgorgement figure. Teo, 746 F.3d at 105-07. First, “the SEC is required to produce evidence supporting a reasonable approximation of ‘actual profits on the tainted transactions.’” Id. at 105 (quoting SEC v. First City Fin. Corp., 890 F.2d 1215, 1231 (D.C. Cir. 1989)). “Once the SEC has made this showing, the burden shifts back to the defendant to ‘demonstrate that the disgorgement figure is not a reasonable approximation,’” with any “risk of uncertainty fall[ing] on the wrongdoer whose illegal conduct created uncertainty.” Id. (quoting First City, 890 F.2d at 1232). In evaluating these steps, a District Court “may take into account facts that were not determined by the jury, but it may not base its 16 decision on factual findings that conflict with the jury’s findings.” SEC v. Capital Sols. Monthly Income Fund, LP, 818 F.3d 346, 354 (8th Cir. 2016) (quoting Salitros v. Chrysler Corp., 306 F.3d 562, 573 (8th Cir. 2002)). As a result, a District Court has “wide discretion in deciding the amount to be disgorged,” Teo, 746 F.3d at 106, and we review its disgorgement order for abuse of discretion, id. at 101. At a hearing, the SEC proposed that Huang’s total positive profits of $4,403,545 represented the appropriate disgorgement calculation. Huang argues that this figure was not reasonable because the verdict form did not reveal the number of the insider transactions for which the jury actually found Huang liable. Huang’s argument fails to account for the fact that disgorgement is an equitable remedy that a court may impose based upon its view of the facts, even those that may not have been determined by the jury. Capital Sols., 818 F.3d at 354; see also Teo, 746 F.3d at 106. Thus, the absence of specific findings by the jury concerning whether each transaction was based on material insider information is of no consequence. Even without the jury’s transaction-by- transaction finding, there is ample support for the District Court’s conclusion that the Capital One data was material to each of Huang’s trades since it allowed him to project the retail companies’ total quarterly revenues before that information went public. 8 8 One distinction between some of Huang’s trades was the fact that Graham testified that his correlation analysis indicated that the Capital One data had statistically significant predictive power for only 132 out of the 226 companies Huang analyzed. However, the correlation analysis was not necessarily the linchpin of the entire materiality finding; Graham showed that Huang created projections about all of the retail companies’ total revenues and not just those for which the correlation was statistically significant. Cf. Matrixx Initiatives, Inc., 563 U.S. at 43 (“Given that medical professionals and regulators act on the basis of evidence of causation that is not 17 Huang also failed to produce any counter-figure, despite repeated requests for one, and even argued that Graham’s net profit calculation was inaccurate since it did not specifically account for his profits from insider trades. While net profits is normally the appropriate measure for disgorgement, Teo, 746 F.3d at 106, Huang rejected the possibility of calculating net profits and offered no alternative means to calculate a disgorgement figure. Thus, the District Court was left with only the positive profits figure Cain had presented.9 Based on this record, and especially Huang’s failure to provide any alternate measure to calculate profits and his assertion that net profits were impossible to calculate from the evidence, and because “the risk of [any] uncertainty fall[s] on [him as] the wrongdoer,” id. at 105, we cannot say the District Court erred in finding that Cain’s calculation of Huang’s total positive profits amounted to a “reasonable approximation of Huang’s ill-gotten gains.” App. 758-59. Thus, the District Court did not abuse its discretion in fashioning a disgorgement remedy based on all of Huang’s insider trades. statistically significant, it stands to reason that in certain cases reasonable investors would as well.”). Moreover, of the more than $4.4 million of total positive profits, only approximately $300,000 came from companies for which Graham’s analysis showed a lack of statistically significant correlation between the Capital One data and total revenues. Thus, even if the correlation analysis was central to the materiality finding, the total positive profits amounted to a reasonable approximation of the profits from Huang’s illegal activities, particularly given that he refused to offer an alternative calculation. 9 This figure was calculated with two important limitations. First, Cain accounted only for insider trades Huang engaged in, meaning that the profits were only from trades Huang made on companies whose nonpublic Capital One data he had downloaded and analyzed. Second, Cain calculated the profits for Huang’s trades on each retail company based only on Huang’s earnings from the time Huang misused the Capital One data until the company’s revenue data was publicly announced. 18 III For the foregoing reasons, we will affirm. 19