MARTIN

Anne Ruth Schultz, Wifredo A. Ferrer, Christopher J. Hunter, Assistant U.S. Attorney, Kathleen Mary Salyer, Harold E. Schimkat, U.S. Attorney's Office, Miami, FL, Jack B. Patrick, Washington, DC, for Plaintiff–Appellee. Sheryl Joyce Lowenthal, Attorney at Law, Miami, FL, for Defendant–Appellant.
Appeals from the United States District Court for the Southern District of Florida, D.C. Docket No. 1:08–cr–20071–JAL–4.
Before MARTIN, FAY, and SENTELLE, Circuit Judges.

MARTIN, Circuit Judge:

Laurence Isaacson was convicted after a jury trial on account of his participation in a conspiracy to commit securities fraud in violation of 18 U.S.C. § 371. He was sentenced to 36–months imprisonment and ordered to pay $8 million in restitution. In this consolidated appeal, Mr. Isaacson challenges his conviction, his sentence, and the District Court's denial of his motion for a new trial. After careful review, and with the benefit of oral argument, we affirm Mr. Isaacson's conviction and the District Court's denial of his motion for a new trial, but vacate his sentence and remand with instructions for resentencing.

I. Background

A. The Criminal Conspiracy

This case arises out of a complex scheme designed to defraud investors through a group of hedge funds we will call the Lancer Fund. At the core of the criminal enterprise were Michael Lauer, Martin Garvey, and Eric Hauser, all of whom had ownership interests in the Lancer Fund. Also central to the scheme was Bruce Cowen, who served as the Lancer Fund's managing director.

The fraud charged in this case began in earnest in late 1999, when the Lancer Fund began investing in publicly traded shell companies, or companies that typically had no real assets or business operations. Once the Lancer Fund controlled the shell companies, it drove up the share prices by buying stock in the companies at artificially inflated prices. This caused the shell companies to appear much more valuable than they really were. By the conspiracy's end, these worthless and overpriced shell companies made up the majority of the Lancer Fund's investment portfolio. Among the shell companies the Lancer Fund invested in were ServiceMax of America (SMX), Augment Systems (AUG), and Nu–D–Zine, each of which Mr. Isaacson helped the Lancer Fund acquire and manage.

Investors had no way to know what the Lancer Fund was doing because it did not disclose the companies in which it invested. This policy prevented investors from verifying for themselves the value and performance of the Lancer Fund. Instead, investors generally depended on evaluationsgiven by Lancer Fund managers and to some extent on annual audits by independent auditors.

Mr. Isaacson's criminal involvement with the conspiracy began when the Lancer Fund's auditors became suspicious about the value of the shell companies in the investment portfolio and sought support for their assigned market value. In an effort to placate the auditors, Mr. Cowen sought valuations in early 2002 from consultants corroborating the market value of the companies. One of the consultants was Milton Barbarosh, who shared an office with Mr. Isaacson in Florida.

At first, Mr. Barbarosh was not sure how he would fulfill Mr. Cowen's request, because “there was no basis to show that the value of the companies were worth anywhere near the value of the public market.” Eventually, Mr. Barbarosh decided that he could evaluate the shell companies by producing hypothetical valuations based on future business plans that were not intended to be implemented. Mr. Barbarosh shared this fraudulent plan with Mr. Isaacson, who reported that he had already suggested the idea to Mr. Cowen, and Mr. Cowen had said that it “would be fine if [Mr. Barbarosh did] the reports on that basis.” Mr. Barbarosh testified that he told Mr. Isaacson that the purpose of the valuations was to support the market price of the companies, and that he “probably” summarized the rest of the conversation with Mr. Cowen for Mr. Isaacson.

With a plan in place, Mr. Barbarosh began work on the valuation reports for SMX and Nu–D–Zine. To do that, he needed some model business plans that he could say were being implemented at the companies. Mr. Isaacson agreed to help and found two plans with figures that would allow Mr. Barbarosh to come close to the market value. Neither of these plans was ever actually going to be implemented at either SMX or Nu–D–Zine.

Mr. Barbarosh completed the valuations consistent with Mr. Cowen's request and sent him the final reports, which were dated May 23 and June 7, 2002. After Mr. Barbarosh sent the last of the reports, Mr. Isaacson spoke with Mr. Cowen and passed the message along to Mr. Barbarosh that “the reports were great, and that the auditors had accepted them.” The record does not indicate when exactly the Lancer Fund sent the reports to the auditors.

Messrs. Isaacson and Barbarosh were also involved in similar efforts to produce inflated valuations in 2003. Investors eventually became suspicious, however, in part because the auditors had not issued a report about the Lancer Fund's 2001 performance by 2003. In an effort to stave off investors' attempts to withdraw their money—which the Lancer Fund did not have—the Lancer Fund continued to misrepresent its performance to investors and encourage them to accept redemption in the form of Lancer Fund securities as opposed to cash.

Mr. Isaacson was indicted, along with Messrs. Lauer, Garvey, Hauser, and Barbarosh, on January 29, 2008. The indictment charged the co-conspirators with conspiracy to commit wire, mail, and securities fraud, in violation of 18 U.S.C. § 371 (Count One), and six counts of wire fraud, in violation of 18 U.S.C. §§ 1343 and 2 (Counts Two through Seven). Mr. Cowen was charged separately.

B. The Prosecution

Due in large part to three continuances, which were granted at various defendants' requests, Mr. Isaacson was not brought to trial until spring of 2010. In March 2010, the District Court began jury selection based on prospective jurors' responses to written questionnaires, which the parties had helped to prepare. The delay between indictment and Mr. Isaacson's trial prompted him to file a Speedy Trial Act motion to dismiss on April 19, 2010, which the District Court denied.

After a lengthy trial, Mr. Isaacson was convicted of conspiracy to commit securities fraud, as charged in Count One. The District Court dismissed Count Seven before submitting the case to the jury; the jury acquitted Mr. Isaacson of Counts Two, Three, and Four; and did not reach a verdict on Counts Five and Six. Ultimately, Messrs. Hauser, Cowen, and Barbarosh pleaded guilty based on their involvement in the scheme. Messrs. Lauer and Garvey were, however, acquitted after a jury trial.

C. Sentencing

Mr. Isaacson's presentence investigation report (PSR) noted that his offense corresponded to a base offense level of 6, and also recommended that the Court impose the following enhancements: a 20–level loss amount enhancement, pursuant to United States Sentencing Guidelines (USSG) § 2B1.1(b)(1)(K); a 4–level number of victims enhancement, pursuant to USSG § 2B1.1(b)(2)(B); a 2–level sophisticated means enhancement, pursuant to USSG § 2B1.1(b)(9)(C); and a 4–level enhancement pursuant to USSG § 2B1.1(17)(A)(ii), because Mr. Isaacson was licensed with the Securities and Exchange Commission at the time he committed the offense. These enhancements brought Mr. Isaacson's offense level to 36, which corresponded to a guideline range of 188– to 235–months imprisonment. The statutory maximum sentence for a violation of 18 U.S.C. § 371 is 60–months imprisonment.

Mr. Isaacson objected to the loss amount enhancement, the number of victims enhancement, and the sophisticated means enhancement, and also objected to the lack of a 2–level minor role reduction pursuant to USSG § 3B1.2(b). At sentencing, Mr. Isaacson's loss amount objection proved the most controversial, and ultimately drove the guideline calculation.

The PSR attributed a loss of $15 million to Mr. Isaacson based on an investment Morgan Stanley made on June 28, 2002. The District Court eventually agreed with the attribution, after much deliberation. In sentencing Mr. Isaacson, the District Court defined his agreement to participate in the conspiracy narrowly, as a conspiracy to defraud the auditors. Despite this narrow definition, the District Court concluded that by defrauding the auditors, Mr. Isaacson participated in the broader conspiracy that caused Morgan Stanley to make its investment.

This loss amount enhancement brought Mr. Isaacson's guideline range above the maximum sentence permitted by his statute of conviction, and so the Court declined to rule on the remainder of Mr. Isaacson's objections. The Court ultimately sentenced Mr. Isaacson to 36–months imprisonment, a downward variance from his 60–month guideline range. Cf. USSG § 5G1.1(a) ( “Where the statutorily authorized maximum sentence is less than the minimum of the applicable guideline range, the statutory authorized maximum sentence shall be the guideline range.”). The District Court also ordered Mr. Isaacson to pay $8 million in restitution to Morgan Stanley. This was the amount of loss the Court found the government proved at sentencing.

D. The Rule 33 Motion

About two years after Mr. Isaacson's conviction, he filed a motion pursuant to Federal Rule of Criminal Procedure 33 seeking a new trial, as well as an evidentiary hearing, based on newly discovered evidence about an alleged prosecutorial conflict of interest. Mr. Isaacson asserted his discovery that the lead prosecutor's wife was a shareholder at the firm representing Mr. Barbarosh, one of the primary witnesses against Mr. Isaacson. The attorney who represented Mr. Isaacson at trial filed an affidavit explaining that, had he known about this conflict of interest, he would have “employed different trial tactics ..., especially with regard to cross-examination of Mr. Barbarosh and [the] closing argument.”

The government filed a response, which included an affidavit from the prosecutor's wife. The prosecutor's wife explained that although she was once a shareholder with the firm, at the time of Messrs. Barbarosh and Isaacson's prosecution she was not. Rather, she was employed as a part-time contract attorney at that time, and was paid an hourly wage. She explained that she was not a member of Mr. Barbarosh's defense team and did not have any contact with the lead counsel assigned to Mr. Barbarosh's case. The government's response also noted that the prosecutor had disclosed his wife's employment to his supervisors, who concluded there was no need for recusal because there was no actual conflict or an appearance of one. Given the lack of any appearance of conflict, the government maintained that recusal was not required and that information about the conflict was not subject to disclosure.

The District Court denied Mr. Isaacson's request for an evidentiary hearing and a new trial, finding “no conflict of interest or Brady[ v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963),] violation warranting a new trial.” To the conflict of interest issue, the District Court emphasized that the wife was not actually a shareholder at the firm at the time of the representation. On the Brady issue, the Court found the wife's employment gave no additional incentive for Mr. Barbarosh to lie on the stand beyond that already explored during cross examination. The Court thus concluded that the information had no material impact on the outcome of Mr. Isaacson's prosecution.

E. This Appeal

This is a consolidated appeal, including Mr. Isaacson's direct appeal from his conviction and sentence as well as his appeal from the denial of his Rule 33 motion based on his charge of prosecutorial conflict. Mr. Isaacson makes the following arguments on appeal: (1) his conduct lies outside the scope of conduct punishable under 18 U.S.C. § 371; (2) the District Court should have granted his motion to dismiss for Speedy Trial Act violations; (3) the government failed to prove an overt act within the statute of limitations; (4) the evidence is insufficient to support the conviction; (5) his sentence is unreasonable because the loss and restitution amounts are not sufficiently attributable to his participation in the conspiracy, and because he was ordered to serve a much longer sentence and pay much more restitution than any of his more culpable co-defendants; and (6) his motion for a new trial should have been granted. We address each argument in turn.

II. The Scope of 18 U.S.C. § 371

We first address Mr. Isaacson's argument that 18 U.S.C. § 371 does not reach his conduct because the Lancer Fund made the investments through a hedge fund based in the British Virgin Islands. Mr. Isaacson relies upon the holding in a civil case, Morrison v. National Australia Bank Ltd., 561 U.S. 247, 130 S.Ct. 2869, 177 L.Ed.2d 535 (2010), to argue that his criminal conviction based on violations of the Securities Exchange Act cannot survive because the transactions were based outside the United States. The issue Mr. Isaacson raises—whether the criminal statute reaches his conduct—is subject to de novo review. United States v. Lopez–Vanegas, 493 F.3d 1305, 1311 (11th Cir.2007).

To begin, it has never been established in this Circuit that Morrison limits the scope of criminal liability for violations of the Securities Exchange Act to the same extent it limits the scope of civil liability. We need not explore that question in this appeal. That is because, even if we assume here that it does, Mr. Isaacson's conviction is still due to be affirmed.

In Morrison, the Supreme Court applied the presumption against extraterritoriality, a tool of statutory construction limiting a statute's applicability to domestic conditions unless the statute gives a “clear indication of an extraterritorial application.” 561 U.S. at 255, 130 S.Ct. at 2877–78. The Supreme Court held that the scope of civil liability based on the Securities Exchange Act is limited to cases in which the defendant committed a fraud “in connection with the purchase or sale of a security listed on an American stock exchange,” or in connection with the purchase or sale of any other security that took place in the United States. Id. at 273, 130 S.Ct. at 2888. If the conduct underlying the claim does not meet at least one of these requirements, then under Morrison, there can be no civil liability. Id.

We have no trouble concluding that Mr. Isaacson's conduct meets Morrison's requirements for a U.S. nexus. The government's expert witness analyzed the SMX, AUG, and Nu–D–Zine securities and testified that they all “traded on the Over–the–Counter Bulletin Board or the Pink Sheets.” The expert explained that these exchanges are “similar to” the NYSE and the NASDAQ, which are both American markets. Beyond that, there was evidence that the Lancer Fund was “run out of New York City” and that Mr. Isaacson's office was located in Florida, which supports the inference that the Lancer Fund purchased the securities in the United States. This evidence is enough to satisfy Morrison's conditions.

III. The Speedy Trial Act Motion

We next address Mr. Isaacson's argument that the District Court should have granted his motion to dismiss under the Speedy Trial Act. His challenge turns on the question of whether eliminating prospective jurors based only on their answers to written questionnaires before they make any personal appearance in Court constitutes part of voir dire. This is a question of law which we review de novo. United States v. Jernigan, 341 F.3d 1273, 1285 (11th Cir.2003) (“We review a claim under the Speedy Trial Act de novo and review a district court's factual determinations on excludable time for clear error.” (quotation marks omitted)).

The Speedy Trial Act generally requires a federal criminal trial to “commence within seventy days” of the defendant's indictment or initial appearance, whichever is later. 18 U.S.C. § 3161(c)(1). The seventy-day requirement is tempered to some extent by a number of exclusions which extend the time the government has to bring a defendant to trial. Id. § 3161(h).

The remedy for a violation of the seventy-day provision is mandatory dismissal upon the defendant's motion. Id. § 3162(a)(2). If, however, the defendant fails to move for enforcement of his Speedy Trial Act rights “prior to trial,” he is deemed to have waived his rights and the remedy of dismissal is no longer available to him, even if the trial technically violates the Act's requirements. Id. We have said that trial begins for purposes of the Speedy Trial Act “when the court begins the voir dire.” United States v. Gonzalez, 671 F.2d 441, 443 (11th Cir.1982). But this begs the question now before us—when (again, for purposes of the Speedy Trial Act) does voir dire begin?

Mr. Isaacson argues that the Speedy Trial clock began to run on March 6, 2008, when Mr. Garvey made his initial appearance. See18 U.S.C. § 3161(c)(1) (noting that the Speedy Trial Act clock is triggered when the indictment is filed or when the defendant makes an appearance in the court in which he is being charged, whichever is later); United States v. Mathis, 96 F.3d 1577, 1579 n. 1 (11th Cir.1996) (“In a multiple defendant case, the speedy trial clock begins to run when the last codefendant is indicted or arraigned.”). The District Court granted three continuances at different co-defendants' requests and also held a hearing on a motion for reconsideration, which Mr. Isaacson joined. Each of these events arguably extended the time the government had to bring Mr. Isaacson to trial. See18 U.S.C. § 3161(h)(7)(A) (noting that the period of delay resulting from a continuance is excludable “if the judge granted such continuance on the basis of his findings that the ends of justice served by taking such action outweigh the best interest of the public and the defendant in a speedy trial”); id. § 3161(h)(1)(D) (noting that the delay resulting from any pretrial motion is excludable “from the filing of the motion through the conclusion of the hearing on, or other prompt disposition of, such motion”). Based on his view that the time resulting from these delays was not wholly excludable, Mr. Isaacson filed a Speedy Trial Act motion on April 19, 2010.

After careful consideration, we conclude that Mr. Isaacson's motion to dismiss was properly denied because it was filed after his trial had already begun. In reaching this conclusion, we agree with the District Court's decision that jury selection based on written responses is part of the voir dire process, at least once the District Court begins ruling on opposed challenges to particular jurors. The District Court began discussing with counsel the responses to the juror questionnaires on March 18, 2010, and the parties identified several jurors they jointly agreed should be struck for cause or hardship. The District Court announced that it would eliminate several of those agreed-upon prospective jurors, which the Court characterized as “the beginning of jury selection.” This process continued on March 23, when the District Court excluded several more prospective jurors for which there was a consensus, and excused at least one juror over the government's opposition. Mr. Isaacson did not file his Speedy Trial Act motion until April 19, 2010. This being the case, Mr. Isaacson's motion was untimely, even though jurors did not appear for oral examination until April 26.

Mr. Isaacson may well be right when he posits that the common understanding of voir dire is jurors being examined under oath in the courtroom where the judge, the defendant, and the government can decide who should be placed on the jury based on the content of the jurors' answers as well as their demeanor. See Reynolds v. United States, 98 U.S. 145, 156–57, 25 L.Ed. 244 (1878) (“[T]he manner of the juror while testifying is oftentimes more indicative of the real character of his opinion than his words. That is seen below, but cannot always be spread upon the record.”). But this limited definition of voir dire cannot withstand a broader consideration of the discretion vested in District Courts for developing procedures for jury selection.

We have long recognized that District Courts have considerable discretion in deciding how best to pick a jury. This includes deciding who should ask questions of the prospective jurors; what the questions should be; and most important here, how they should be asked. See, e.g., United States v. Tegzes, 715 F.2d 505, 507 (11th Cir.1983). The only limit on this discretion this Court has articulated is that the District Court's chosen methods must create a “reasonable assurance that prejudice would be discovered if present,” id. (quotation mark omitted). Of course, the selection process must also comply with Federal Rule of Criminal Procedure 24, which addresses jury selection.

Written questionnaires are now a common complement to oral examination when selecting an effective and impartial jury. As the Second Circuit observed not too long ago, courts “routinely employ questionnaires to facilitate voir dire in a number of circumstances.” United States v. Quinones, 511 F.3d 289, 299 (2d Cir.2007) (italics omitted) (finding that written questionnaires can be used as part of the voir dire process, although outside the Speedy Trial Act context). With this in mind, the Second Circuit recognized that, “[a]lthough voir dire ordinarily contemplates seeing the jurors and hearing them speak, any court-supervised examination of prospective jurors is reasonably understood to be part of voir dire.” Id. (italics, citation, and footnote omitted).

As a practical matter, written questionnaires pose questions commonly asked, and materially indistinguishable from, what we expect to see during oral examination in the courtroom. Beyond that, the expectation that jurors will tell the truth is the same whether their answers are spoken or written. The written questionnaire in Mr. Isaacson's case informed jurors that they were “sworn to give true and complete answers,” and warned that jurors might be subject to prosecution for perjury if they did not uphold that oath. This tracks the oath the jurors took once they were called into Court in this case.

Because written questionnaires facilitate jury selection in ways not unlike traditional oral examination, we cannot agree with Mr. Isaacson's suggestion that jury selection based on written questionnaires can never constitute part of voir dire for purposes of the Speedy Trial Act. Where, as here, the District Court relies on written questionnaires to aid in jury selection, voir dire and thus the trial begins when the Court starts to rule on opposed juror challenges based on those written questionnaires. In this case, that date was March 23, several weeks before Mr. Isaacson's Speedy Trial Act motion. Mr. Isaacson's April 19 motion was therefore untimely, and he was not entitled to dismissal.

We caution that District Courts should not take our decision today “as a license to evade the Act's spirit,” Gonzalez, 671 F.2d at 444, by starting to eliminate jurors based on written questionnaires long before beginning oral examination. As we said in Gonzalez, “[w]e will not hesitate to find that a trial has not actually ‘commenced’ ... if we perceive an intent to merely pay the Act lip service.” Id. Thus, there may be cases where a deviation from the general rule about written examinations we set forth today would be warranted, where the record reveals some purposeful manipulation of the Speedy Trial Act's technical requirements.

The particular procedural background of this case, even with the rather prolonged delay of trial, gives us comfort that this voir dire was not commenced in order to avoid the requirements of the Speedy Trial Act. When the District Court first ruled on the written questionnaires, it had already granted a continuance, at two defendants' request, to April 26. Also, the Court still had a motion before it which challenged certain evidence against Mr. Isaacson and his co-defendants, for which it held an evidentiary hearing in the first week of April. Given this context, we do not see the District Court's decision to begin jury selection in March based on the written questionnaires as an intentional effort to manipulate the Speedy Trial Act.

IV. The Statute of Limitations

Next, Mr. Isaacson argues that his conviction should be vacated because the government failed to prove an overt act within the statute of limitations. In a conspiracy prosecution brought under 18 U.S.C. § 371, the government must prove at least one overt act in furtherance of the conspiracy by one of the conspirators within the five years before the return of the indictment. 18 U.S.C. § 3282(a); United States v. McNair, 605 F.3d 1152, 1213 (11th Cir.2010). Whether there is sufficient evidence of an overt act within the statute of limitations is a question of law which this Court reviews de novo. See id. at 1212 n. 87.

The grand jury returned Mr. Isaacson's indictment on January 29, 2008, meaning that the government had to prove an overt act in furtherance of the conspiracy occurring on or after January 29, 2003. See18 U.S.C. § 3282(a). When the District Court rejected Mr. Isaacson's statute of limitations argument, it found “that the government sufficiently proved a number of overt acts within the limitations period (e.g., overt acts 24, 26, and 27).” Because one overt act in furtherance of the conspiracy is enough to affirm the conviction, we focus on overt act 24, which alleged that Mr. Lauer sent letters to the Lancer Fund's investors in February 2003. Mr. Isaacson's brief does not challenge the District Court's conclusion that the sending of these letters was an overt act in furtherance of the conspiracy. Instead, Mr. Isaacson argues that this overt act does not mention him, and that Mr. Lauer's overt act cannot be used to bootstrap Mr. Isaacson's conduct within the statute of limitations.

This argument misunderstands the nature of conspiracy liability. “[A]n individual conspirator need not participate in the overt act in furtherance of the conspiracy. Once a conspiracy is established, and an individual is linked to that conspiracy, an overt act committed by any conspirator is sufficient.” United States v. Thomas, 8 F.3d 1552, 1560 n. 21 (11th Cir.1993); see also United States v. Arias, 431 F.3d 1327, 1340 (11th Cir.2005) (“An accused conspirator's participation is presumed to have continued ... until the last overt act has been committed by any of the conspirators.”). Mr. Isaacson's statute of limitations argument is therefore without merit.

V. The Sufficiency of the Evidence

Mr. Isaacson also challenges the sufficiency of the evidence supporting his conviction. He argues that several government witnesses, and one government agent who did not testify at trial, were not credible, and that two of the government's witnesses testified that they did not know Mr. Isaacson. In Mr. Isaacson's view, the case against him was weak, as evidenced by the government's failure to secure a conviction on six of the seven counts charged. Also, he points to shortcomings in the government's case he surmises could have led to his acquittal.

We review de novo the sufficiency of the evidence supporting a conviction, viewing the evidence and drawing all inferences in favor of the verdict. United States v. Benbow, 539 F.3d 1327, 1331 (11th Cir.2008). We must affirm Mr. Isaacson's conviction “unless there is no reasonable construction of the evidence from which the jury could have found [him] guilty beyond a reasonable doubt.” United States v. Joseph, 709 F.3d 1082, 1093 (11th Cir.2013). To the extent Mr. Isaacson's argument “depends upon challenges to the credibility of witnesses, the jury has exclusive province over that determination and the court of appeals may not revisit this question.” United States v. Chastain, 198 F.3d 1338, 1351 (11th Cir.1999). We will upset a jury's decision to credit a witness's testimony only in the rare circumstance that the testimony is incredible as a matter of law. United States v. Calderon, 127 F.3d 1314, 1325 (11th Cir.1997). Testimony is incredible as a matter of law only if it concerns facts that the witness “physically could not have possibly observed or events that could not have occurred under the laws of nature.” Id. (quotation marks omitted).

Applying these well-established principles, we find the evidence against Mr. Isaacson sufficient to support his conviction. He primarily argues that Messrs. Cowen and Barbarosh were not credible because the witnesses had incentives to lie. But Mr. Isaacson has not pointed to any circumstances that would render their testimony incredible as a matter of law, and instead points to circumstances which were thoroughly explored during the examinations of both witnesses. Mr. Isaacson also notes that more evidence has come out since his conviction suggesting that Mr. Barbarosh lied about his own culpability and about whether Mr. Isaacson was really the source of the business plans upon which the fraudulent valuations were based. But, at least on direct review from the conviction, we consider only the evidence that was actually presented to the jury at the time it reached its verdict. See United States v. Lopez–Ramirez, 68 F.3d 438, 441 & n. 3 (11th Cir.1995).

By contrast, the government presented significant evidence implicating Mr. Isaacson. Mr. Barbarosh testified that Mr. Isaacson gave him the business plans, knowing they would be used in valuations provided to auditors, and knowing that the plans had nothing to do with the businesses being valued. Beyond that, the shell companies Mr. Barbarosh was valuing were actually managed by Mr. Isaacson (suggesting that he knew their true value, or rather lack thereof), and had been purchased by the Lancer Fund with Mr. Isaacson's help. Finally, Mr. Isaacson testified on his own behalf, and the jury was “permitted to reject that testimony, as we must assume it did, and consider that testimony as substantive evidence of the defendant's guilt.” United States v. Jiminez, 564 F.3d 1280, 1285 (11th Cir.2009) (quotation marks and emphasis omitted). This evidence supports the inference that Mr. Isaacson was guilty, and his evidence arguably suggesting otherwise does not undermine it to the extent necessary for us to upset the jury's verdict.

VI. The Sentence

We next consider Mr. Isaacson's argument that the government did not carry its burden to attribute Morgan Stanley's losses to him for purposes of the loss amount enhancement and the restitution order. Based upon the findings of the District Court, the government's evidence did not adequately tie Morgan Stanley's investment to the scope of Mr. Isaacson's participation in the conspiracy. Thus, we vacate his sentence and remand with instructions.

In his objections to the PSR and at sentencing, Mr. Isaacson asserted that the government could not adequately attribute Morgan Stanley's $15.625 million investment to him. “When the government seeks to apply an enhancement under the Sentencing Guidelines over a defendant's factual objection, it has the burden of introducing sufficient and reliable evidence to prove the necessary facts by a preponderance of the evidence.” United States v. Washington, 714 F.3d 1358, 1361 (11th Cir.2013) (quotation marks omitted). “We review de novo the application of the sentencing guidelines and findings of fact for clear error.” United States v. Louis, 559 F.3d 1220, 1224 (11th Cir.2009).

The District Court “may hold participants in a conspiracy responsible for the losses resulting from the reasonably foreseeable acts of co-conspirators in furtherance of the conspiracy.” United States v. Hunter, 323 F.3d 1314, 1319 (11th Cir.2003). But “[t]he limits of sentencing accountability are not coextensive with the scope of criminal liability.” Id. For sentencing purposes, a defendant is responsible for the conduct of his co-conspirators only if that conduct was “in furtherance of the jointly undertaken criminal activity” and “reasonably foreseeable in connection with that criminal activity.” USSG § 1B1.3, comment. (n.2). The Guidelines have established a two-pronged test which District Courts must apply at sentencing before they can attribute losses caused by a coconspirator's conduct to a defendant. First, the District Court must “make individualized findings concerning the scope of criminal activity undertaken by a particular defendant.” Hunter, 323 F.3d at 1319 (quotation mark omitted); see alsoUSSG § 1B1.3, comment. (n.2) (“In order to determine the defendant's accountability for the conduct of others ... the court must first determine the scope of the criminal activity the particular defendant agreed to jointly undertake (i.e., the scope of the specific conduct and objectives embraced by the defendant's agreement).”). “[T]he fact that the defendant knows about the larger operation, and has agreed to perform a particular act, does not amount to acquiescence in the acts of the criminal enterprise as a whole.” Hunter, 323 F.3d at 1320.

Once the District Court determines the precise scope of the defendant's agreement, the Court must then consider whether the conduct of his co-conspirators was “in furtherance of, and reasonably foreseeable in connection with, the criminal activity jointly undertaken by the defendant.” USSG § 1B1.3, comment. (n.2). Said another way, the defendant's sentence can only be enhanced by those reasonably foreseeable losses caused by co-conspirators acting in furtherance of the part of the conspiracy in which the defendant agreed to participate. See Hunter, 323 F.3d at 1319–20;see alsoUSSG § 2B1.1, comment. (n.3(A)(i)-(ii)) (defining loss amount as the greater of the actual losses “that resulted from the offense” or the “pecuniary harm that was intended to result from the offense”).

Here, the government has not established that Morgan Stanley's losses were the result of either Mr. Isaacson's or his co-conspirators' conduct in furtherance of the part of the conspiracy in which he agreed to participate. The District Court's “finding with regards to the scope of criminal activity that was jointly undertaken by Mr. Isaacson” was this: “Mr. Isaacson knowingly participated in a scheme to produce or create inflated valuations for what I will call the shell companies which Lancer had in its portfolio.... The purpose of this aspect of the conspiracy was to help Lancer deal with questions being asked by auditors and administrators of the Lancer Funds....” Specifically, then, the District Court concluded that Mr. Isaacson agreed to jointly undertake the criminal activity of defrauding the auditors, not the investors.

As a result, to attribute Morgan Stanley's losses to Mr. Isaacson, the government was required to show that its investment was the result of acts by Mr. Isaacson or his co-conspirators taken to defraud the auditors. In assessing the link between the actions taken to further that subset of the criminal enterprise and Morgan Stanley's investment, we “must not speculate concerning the existence of a fact which would permit a more severe sentence under the guidelines.” United States v. Sepulveda, 115 F.3d 882, 890 (11th Cir.1997) (quotation marks omitted). A close look at the conspiracy to defraud the auditors and Morgan Stanley's investment reveals that far too much speculation is necessary to attribute that investment loss to the conspiracy to defraud the auditors.

Morgan Stanley never saw the valuations Mr. Isaacson helped Mr. Barbarosh to prepare. And, as the District Court noted, Morgan Stanley knew the auditors had not signed off on the Lancer Fund's financial statements for 2001, the year for which the fraudulent valuations were provided. But Morgan Stanley invested anyway. That Morgan Stanley invested knowing that the audit reports were delayed, suggesting there might be a problem, strongly supports the inference that Morgan Stanley made its investment decision entirely independent of any audit reports that would or should have been completed after Mr. Isaacson agreed to join the conspiracy. This inference is further supported by the fact that Morgan Stanley's June 28, 2002 investment was the last in a series of several investments Morgan Stanley made, dating back to October 2001, after doing its own due diligence. These facts preclude the government from showing that Morgan Stanley's investment “resulted from,” USSG § 2B1.1, comment. (n.3(A)(i)), the conspiracy to defraud the auditors. In the words of the District Court, “Morgan Stanley was going to make this investment, and had made its own internal decision that despite significant risk factors, it was going to invest, because it thought it was a good deal, and it was willing to overlook certain red flags, like the audited financial statements for 2001 being late.”

To be sure, the conspiracy to defraud the auditors delayed and possibly prevented the auditors' discovery of the fraud. In this way, the government argues, the conspiracy to defraud the auditors caused the investment because “Morgan Stanley almost certainly would have learned of Lancer's fraud if that fraud had been discovered and publicly reported by Lancer's auditors.” But this hypothetical causal chain requires inference upon inference for which there is not sufficient support in the record. See Washington, 714 F.3d at 1361. Would the auditors have actually discovered the broader conspiracy to defraud the investors if not for the conspiracy to defraud the auditors? Would the auditors have made that information public? If the auditors would have unearthed the conspiracy and released that information to the public, would they have done so in time to prevent Morgan Stanley's investment?

The record does not support an affirmative answer to this last question. The fraudulent valuation reports Mr. Barbarosh produced were dated May 3, May 23, and June 7, 2002, and Morgan Stanley made its investment on June 28, 2002. The government acknowledged at Mr. Isaacson's sentencing that there was no evidence about the precise date on which the valuation reports were provided to the auditors. Given that the final valuation report is dated June 7 and Morgan Stanley invested on June 28, we reject the government's argument that we should infer, without more information, that without those reports, the auditors would have discovered the fraud in time to save Morgan Stanley from making this bad investment decision.

To the contrary, the evidence strongly suggests that Morgan Stanley was set on making the June 2002 investment, and does not support the inference that absent the specific conspiracy Mr. Isaacson joined, the auditors would have discovered the fraud before Morgan Stanley invested. To say that Morgan Stanley's investment resulted from the conspiracy to defraud the auditors would require a number of inferential leaps and assumptions not supported by the record. See Sepulveda, 115 F.3d at 890–91. Therefore, we conclude that the link between the conspiracy to defraud the auditors and Morgan Stanley's investment is too speculative to say the latter “resulted from” the former, as the Guidelines require. USSG § 2B1.1, comment. (n.3(A)(i)-(ii)); Sepulveda, 115 F.3d at 890–91. The District Court clearly erred in concluding otherwise and Mr. Isaacson's sentence should not have been enhanced based on Morgan Stanley's losses.

The same problem plagues the District Court's restitution order. This Court reviews the legality of a restitution order de novo, and factual findings regarding the specific amount of restitution for clear error. United States v. Huff, 609 F.3d 1240, 1247 (11th Cir.2010). The Mandatory Victims Restitution Act requires a District Court to order restitution for any person who is “directly and proximately harmed as a result of” a defendant's participation in a conspiracy. See18 U.S.C. § 3663A(a)(2). Because we find that the government did not carry its burden to show that Morgan Stanley's investment was the result of the conspiracy in which Mr. Isaacson participated, he cannot be ordered to pay restitution based on that investment. Cf. United States v. Dickerson, 370 F.3d 1330, 1341 (11th Cir.2004) (“[A] criminal defendant cannot be compelled to pay restitution for conduct committed outside of the scheme, conspiracy, or pattern of criminal behavior underlying the offense of conviction.”).

Mr. Isaacson's sentence is vacated and we remand this case for his resentencing without the loss amount enhancement or the $8 million restitution order. When the District Court resentences Mr. Isaacson, it should resolve all of his remaining sentencing objections not already addressed. Becauseall of the objections Mr. Isaacson raised were before the District Court at the original sentencing, and because the Court stood ready to accept all relevant evidence from both parties going to each of those objections, we remand with instructions that the District Court is limited to the record the parties previously made as to those objections. See Washington, 714 F.3d at 1362 (describing situations in which this Court will not remand for further evidentiary hearings after a defendant's successful appeal). The same evidentiary limitation applies to the District Court's reconsideration of the appropriate restitution amount. The government has already had an opportunity to litigate the restitution question, and did not appeal the District Court's denial of the government's request to present further evidence on restitution.

VII. The Rule 33 Motion

Mr. Isaacson's second appeal challenges the District Court's denial of his Federal Rule of Criminal Procedure 33 motion for a new trial and an evidentiary hearing based on the prosecutor's failure to disclose that his wife was a lawyer at the law firm representing Mr. Barbarosh.

We review a District Court's denial of a Rule 33 motion for a new trial and an evidentiary hearing for abuse of discretion. United States v. Thompson, 422 F.3d 1285, 1294–95 (11th Cir.2005); United States v. Schlei, 122 F.3d 944, 990 (11th Cir.1997). Mr. Isaacson presents two reasons the District Court should have granted the motion: (1) because the prosecutor's recusal was required by statute, see28 U.S.C. § 528 (providing that a prosecutor is not supposed to participate in a prosecution “if such participation may result in a personal, financial, or political conflict of interest, or the appearance thereof”); or alternatively (2) because the alleged conflict should have been disclosed pursuant to Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963).

Both of these claims miss the mark. First, the prosecutor was not required to recuse himself under § 528. The prosecutor's wife had no involvement in Mr. Barbarosh's defense and did not even work in the firm's criminal defense practice group. Beyond that, she was only employed as an at-will, part-time attorney and was paid an hourly wage. See United States ex rel. Weinberger v. Equifax, Inc., 557 F.2d 456, 463–64 (5th Cir.1977) (holding that a District Court judge was not required to recuse himself where the judge's son was an attorney with the firm representing one of the parties because the son's “salary interest as an associate is too remote” to give rise to any arguable conflict); see also United States v. Reagan, 725 F.3d 471, 488 (5th Cir.2013) (noting that the “mere fact that the prosecutor's husband worked at a firm that represented one of the witnesses did not create a conflict, as [the defendant] did not show that the prosecutor or her husband had a financial stake in the outcome of [the defendant's] trial creating a conflict or requiring her disqualification,” even though the prosecutor's husband was a partner rather than an associate). Because Mr. Isaacson cannot show that the prosecutor or his wife had any financial interest in a good outcome for Mr. Barbarosh, there is no conflict of interest that would have required the prosecutor's recusal. The District Court was right to deny Mr. Isaacson's Rule 33 motion based on the statutory recusal provision.

To the extent Mr. Isaacson raises a Brady claim that disclosure of the conflict was required even if recusal was not, the challenge is still without merit. While there may be some cases where an arguable prosecutorial conflict must be disclosed even if recusal is not statutorily required, this is not one of them. To prevail on a Rule 33 motion based on an alleged Brady violation, a defendant must show that there is a reasonable probability the outcome would have been different had the evidence been disclosed. See United States v. Vallejo, 297 F.3d 1154, 1164 (11th Cir.2002) (noting that defendants must prove four things, including prejudice, before a Rule 33 motion based on a Brady violation will be granted).

Mr. Isaacson has shown no reasonable probability that the outcome would have been different if his attorney had been able to cross examine Mr. Barbarosh about the prosecutor's alleged conflict of interest. There is substantial evidence implicating Mr. Isaacson in the conspiracy, and his attorney examined Mr. Barbarosh about his plea agreement with the government and his incentive to testify against Mr. Isaacson. We agree with the District Court's finding that this purported prosecutorial conflict would not have given Mr. Barbarosh any additional incentive to lie on the stand. There is therefore no reasonable probability that the outcome would have been different had the alleged conflict been disclosed.

VIII. Conclusion

We affirm Mr. Isaacson's conviction as well as the denial of his Rule 33 motion. However, we find that the District Court erred when it enhanced Mr. Isaacson's sentence and ordered restitution based on the losses from Morgan Stanley's investment. We therefore vacate the sentence and remand for resentencing based on the record from the first sentencing.

AFFIRMED IN PART; VACATED AND REMANDED IN PART.