holding that evidence of actual and collateral loss was admissible because the defendant argued that the victims were the fraudulent partiesSummary of this case from United States v. Gibson
Id. at 516, 128 S.Ct. 2020 (footnote omitted). Id. at 278.
ARGUED:Arza Feldman, Feldman & Feldman, Uniondale, New York, for Appellant. Kurt William Meyers, Office of the United States Attorney, Charlotte, North Carolina, for Appellee. ON BRIEF:Claire J. Rauscher, Executive Director, Matthew R. Segal, Assistant Federal Defender, Alexandra Cury, Adeel Bashir, Federal Defenders of Western North Carolina, Inc., Asheville, North Carolina; Kevin Tate, Assistant Federal Defender, Federal Defenders of Western North Carolina, Inc., Charlotte, North Carolina, for Appellant. Anne M. Tompkins, United States Attorney, Charlotte, North Carolina, for Appellee.
Before GREGORY, DAVIS, and , Circuit Judges.
Affirmed in part, reversed in part, vacated in part, and remanded by published opinion. Judge wrote the opinion, in which Judge GREGORY and Judge DAVIS joined.
, Circuit Judge:
A jury convicted William Roosevelt Cloud of various offenses stemming from an extensive mortgage fraud conspiracy. On appeal, Cloud challenges the district court's evidentiary rulings, loss calculation, and order directing him to reimburse his court-appointed attorneys' fees. We affirm the district court's judgment on the first two issues, but vacate the court's reimbursement order. Cloud also argues that his money laundering convictions must be reversed under United States v. Santos, 553 U.S. 507, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008), and thus, that the district court erred in rejecting his motion for judgment of acquittal on this ground. Applying Santos, as interpreted by United States v. Halstead, 634 F.3d 270 (4th Cir.2011), to the facts underlying Cloud's substantive money laundering convictions, we agree and therefore reverse those convictions.
Following his indictment in the Western District of North Carolina, Cloud was convicted of one count of mortgage fraud conspiracy, in violation of 18 U.S.C. § 371; three counts of mail fraud, in violation of 18 U.S.C. § 1341; thirteen counts of bank fraud, in violation of 18 U.S.C. § 1344; one count of money laundering conspiracy, in violation of 18 U.S.C. § 1956(h); and six counts of money laundering, in violation of 18 U.S.C. § 1956(a)(1). The convictions stemmed from Cloud's leadership of a mortgage-fraud conspiracy from 1999 until 2005. The scheme involved multiple coconspirators, at least fourteen of whom—including two mortgage brokers, two attorneys, an appraiser, a bank insider, and Cloud's wife—pleaded guilty to various offenses.
Based on the jury's verdict, we recite the facts in the light most favorable to the government. United States v. Kelly, 510 F.3d 433, 435 n. 1 (4th Cir.2007).
The government alleged that Cloud recruited buyers—friends, church members, and members of the Filipino–American community—with good credit to purchase various properties. He offered the buyers an opportunity to become real estate investors with no money down, earning cash upon buying a home. To lull the buyers, Cloud offered to make ownership of the properties easy, assisting with rent or mortgage payments, taking care of repairs to the properties, and ensuring that rental income would cover the mortgage payments. Cloud claimed that he would eventually sell the properties, splitting the profit with the buyers.
Unbeknownst to the buyers, Cloud bought the properties shortly before the buyers, and then “flipped” the property to the buyer, making money off each transaction.Cloud arranged for the buyers to purchase multiple properties within a short time frame, to prevent the earlier sales from appearing on their credit reports. Once a buyer purchased his or her last home, Cloud ceased contact with the buyer. Although Cloud received some rental income from tenants, the money was not used to pay the mortgages. Ultimately, many of the homes went into foreclosure.
To perpetrate the scheme, Cloud falsified the loan applications. Although the buyers supplied accurate information, the loan applications misrepresented the buyers' income, falsely indicated that the buyers were purchasing the property as a residence, and made it appear as though the buyers had sufficient funds to cover the down payment. Because the buyers trusted Cloud, they often signed the documents without reviewing them. In some instances, Cloud forged the buyers' signatures on the applications. Further, because Cloud offered the buyers an opportunity to invest in real estate with no money down, he provided cashier's checks to closing attorneys, which falsely indicated that the buyers were providing the down payment. The attorneys, in turn, falsely represented on the HUD–1 Settlement Statements that the buyers were providing the down payment. Cloud then signed the HUD–1 forms containing the false information. Cloud also paid thousands of dollars in kickbacks to buyers, at least one mortgage broker, and the recruiters responsible for finding the buyers. These payments were not disclosed on the HUD–1 forms.
The government called numerous witnesses at trial to support its theory of the fraud. For example, Joseph Goines testified that Cloud paid him a commission—between $1,500 and $3,000 per head—to obtain buyers with good credit. As a “pitch” to the buyers, Cloud instructed Goines to tell them that Cloud would pay $5,000 for the use of their credit. Robert Moore, a buyer, testified in part that he was not aware that the property he bought for $120,000 was purchased by Cloud for $64,000 less than one month prior. Moore also testified that Cloud typically paid him between $2,000 and $3,000 at closing, and that all ten properties he purchased ultimately went into foreclosure. Kim Dauria, a loan officer, testified that Cloud brought buyers to her, that she was involved in the mortgage fraud conspiracy, and that Cloud paid her. Another government witness, Daniel Greene, testified that as a mortgage broker, he conducted fraudulent real estate transactions with Cloud, that Cloud was responsible for the down payments, and that his transactions with Cloud involved false lease agreements.
At trial, Cloud insisted that he was “a neophyte real estate investor who had to rely on professionals—criminals, as it turned out—to structure his transactions,” but that he “believed his buyers would make money by renting out their properties and benefitting from rising housing prices.” Appellant's Br. 1.
Following the close of the government's case and again before the case was submitted to the jury, Cloud moved for a judgment of acquittal under Federal Rule of Criminal Procedure 29. The district court denied both motions. Although the jury acquitted Cloud on one money laundering count, it convicted him on all others.
Cloud later moved for a judgment of acquittal on his money laundering convictions based on United States v. Santos, 553 U.S. 507, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008), in which the Supreme Court affirmed the vacatur of the defendants' money laundering convictions after concluding that the offenses merged with the defendants' convictions for operating an illegal gambling business. The district court denied Cloud's motion, concluding that United States v. Howard, 309 Fed.Appx. 760 (4th Cir.2009) (unpublished), limited Santos to illegal-gambling offenses.
Prior to sentencing, the probation officer prepared a presentence report (“PSR”), calculating Cloud's total offense level as 42 and his criminal history category as I, resulting in an advisory Guidelines range of 360 months to life imprisonment. The offense level calculation was based on Cloud's estimated intended loss to lenders of $10 million as well as approximately $11 million in collateral loss to the surrounding communities. At sentencing, the court adopted the intended loss figure, but limited Cloud's collateral loss to $9 million, reducing the total loss amount to just over $19 million. This reduction decreased Cloud's offense level to 39, and his Guidelines range to 262–327 months. The court sentenced Cloud to 324 months' imprisonment. Cloud timely appealed.
On appeal, Cloud first challenges the admission of victim-impact testimony and hearsay statements at trial. Second, he argues that the district court wrongly rejected his Rule 29 motions as to his money laundering convictions, arguing that the convictions cannot stand under Santos and our subsequent decision in United States v. Halstead, 634 F.3d 270 (4th Cir.2011), and that there was insufficient evidence to support these convictions. Third, Cloud contends that the district court's loss calculation was clearly erroneous. Finally, he asserts that the district court erred in ordering him to reimburse the United States for the services of his court-appointed attorneys. We address these arguments in turn.
Cloud challenges the district court's admission of victim-impact testimony from buyers and hearsay statements from tenants. We review evidentiary rulings for an abuse of discretion and “will only overturn an evidentiary ruling that is ‘arbitrary and irrational.’ ” United States v. Cole, 631 F.3d 146, 153 (4th Cir.2011) (quoting United States v. Blake, 571 F.3d 331, 346 (4th Cir.2009)). Under Federal Rule of Criminal Procedure 52(a), evidentiary rulings are subject to harmless error review. “ ‘[I]n order to find a district court's error harmless, we need only be able to say with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error.’ ” United States v. Johnson, 617 F.3d 286, 292 (4th Cir.2010) (quoting United States v. Brooks, 111 F.3d 365, 371 (4th Cir.1997)).
Cloud first argues that victim-impact testimony from five buyers was irrelevant and highly prejudicial. The buyers testified that as a result of their dealings with Cloud, they suffered liens on their property, were unable to borrow money, were sued, suffered bad credit, or faced foreclosure. Cloud assigns particular prejudice to the testimony of Rodney Thompson, who testified that all six properties he purchased as part of Cloud's scheme went into foreclosure and that he was forced to file for bankruptcy. In response to government questioning about his first purchase, Thompson responded that he bought a home next to his parents “with the idea when they got older,” he could “take care of” them. J.A. 2136. Thereafter, Thompson began to cry.
We conclude that the district court did not abuse its discretion in admitting the victim-impact testimony. First, testimony regarding the impact of the offenses on the victims met the low bar of relevancy, given Cloud's defense that the buyers were guilty of bank fraud. In his opening statement to the jury, Cloud identified the buyers associated with the bank fraud charges and argued that “[t]hese people are not victims.... [T]hey are the ones that committed the bank fraud.” Id. 427. In United States v. Copple, 24 F.3d 535, 545 (3d Cir.1994), the Third Circuit concluded that victim-impact testimony was relevant to show intent to defraud in a mail fraud prosecution where the defendant claimed that “he had simply made a bad business decision,” explaining that “[p]roof that someone was victimized by the fraud is ... treated as some evidence of the schemer's intent.” So too here. We further find that the probative value of the victims' testimony was not “substantially outweighed by a danger of ... unfair prejudice,” Fed.R.Evid. 403.
For that reason, we also reject Cloud's separate due process challenge to the victim-impact testimony. See Payne v. Tennessee, 501 U.S. 808, 825, 111 S.Ct. 2597, 115 L.Ed.2d 720 (1991) (“In the event that evidence is introduced that is so unduly prejudicial that it renders the trial fundamentally unfair, the Due Process Clause of the Fourteenth Amendment provides a mechanism for relief.”).
Even assuming that the court erred in admitting this testimony, we find any error to be harmless. Cloud was convicted following a lengthy trial that included testimony from multiple witnesses regarding various elements of Cloud's conspiracy. Given the overwhelming evidence against Cloud, we are satisfied “ ‘with fair assurance ... that the judgment was not substantially swayed by the error,’ ” Johnson, 617 F.3d at 292 (quoting Brooks, 111 F.3d at 371).
Cloud next argues that the district court erred in admitting hearsay testimony from buyers detailing conversations between Cloud and several tenants. In one instance, Reginald Foster, a buyer, testified that a tenant told Foster that Cloud described Foster—an African–American—as a racist Caucasian who did not want to deal with his African–American tenants. In another, Rodney Thompson testified that a tenant told him that Cloud kept tenants “in the fog” about who owned the properties and that Cloud indicated that he was the landlord. J.A. 2182.
Cloud asserts that the admission of this testimony cannot be dismissed as harmless because it buttressed the government's theory that Cloud was dishonest with tenants to disguise his scheme and portrayed Cloud as an “immoral” person who would falsely accuse an African–American of racism. Appellant's Br. 41. Cloud's concerns as to Foster's testimony, however, were specifically addressed by the district court in its instructions. Following Foster's testimony, the court instructed the jury that the statement “obviously [was] not being admitted for what we call the truth of the matter asserted,” and that it should not “inject” race into the trial, but should be used, if the jury so chose, “to weigh ... [whether] Mr. Cloud [would] have made this statement to help in a cover up ... because he didn't want ... the tenants [to] talk to Mr. Foster.” J.A. 1245. Cloud challenged the last part of the instruction and in response, the court further instructed the jury that its instruction was “in no way ... an expression of the Court's view of the guilt or innocence” of Cloud. Id. 1257.
Cloud raised no objection to Thompson's testimony, and thus, the district court provided no instruction to the jury.
In any event, we conclude that any error in admitting the tenants' hearsay statements was harmless. Looking to the weight of the evidence against Cloud, and the comparative insignificance of the tenants' statements, we again find “ ‘with fair assurance ... that the judgment was not substantially swayed by the error,’ ” Johnson, 617 F.3d at 292 (quoting Brooks, 111 F.3d at 371).
Cloud also contends that the district court erred in denying his Rule 29 motions. Cloud argues that the government failed to prove Counts 27–33 of the indictment, which charged Cloud with money laundering under 18 U.S.C. § 1956(a)(1) and money laundering conspiracy under 18 U.S.C. § 1956(h) by (1) failing to show that the transactions involved the profits of unlawful activity as required by Santos, and (2) offering insufficient evidence. We review the denial of a Rule 29 motion for judgment of acquittal de novo, sustaining a guilty verdict that, when viewed in the light most favorable to the government, is supported by substantial evidence. United States v. Alerre, 430 F.3d 681, 693 (4th Cir.2005).
We find that Cloud's money laundering convictions on Counts 28–33 suffer from a merger problem, as they charge the payment of the “essential expenses” of the fraud. Halstead, 634 F.3d at 279 (quoting Santos, 553 U.S. at 528, 128 S.Ct. 2020 (Stevens, J., concurring)). Because we reverse these convictions under Santos and Halstead, we do not address Cloud's additional argument that Counts 28–33 are not supported by sufficient evidence. See United States v. Moreland, 622 F.3d 1147, 1166–68 (9th Cir.2010) (reversing defendant's convictions under 18 U.S.C. § 1956(a)(1)(A)(i) based on Santos, and declining to address defendant's additional sufficiency of the evidence challenge). As to Cloud's money laundering conspiracy conviction on Count 27, however, we affirm, finding no merger problem and sufficient evidence.
Cloud was convicted of promotional money laundering under 18 U.S.C. § 1956(a)(1) and of money laundering conspiracy under 18 U.S.C. § 1956(h). To prove promotional money laundering, the government was required to show that Cloud (1) conducted or attempted to conduct a financial transaction; (2) involving the proceeds of a specified unlawful activity; (3) knowing that the property involved proceeds of an unlawful activity; and (4) intending to promote the carrying on of the specified unlawful activity. See United States v. Singh, 518 F.3d 236, 246 (4th Cir.2008).
In Santos, the Supreme Court considered whether the term “proceeds” (which was not then defined in the federal money laundering statute) means “receipts” or “profits.” 553 U.S. at 509, 128 S.Ct. 2020 (plurality opinion). Santos was convicted of operating an illegal gambling business and of laundering the proceeds of that business. Santos's runners collected bets from gamblers, retaining a percentage as a commission and delivering the rest to Santos's collectors. The collectors delivered the money to Santos, who used it to pay the salaries of the collectors and to pay the winners. The district court vacated Santos's money laundering convictions under § 1956(a)(1)(A)(i) and § 1956(h), finding that “the proceeds admittedly used by Santos to pay winners and couriers could only have been gross proceeds,” rather than net proceeds. United States v. Santos, 342 F.Supp.2d 781, 799 (N.D.Ind.2004), aff'd, 461 F.3d 886 (7th Cir.2006).
The Supreme Court affirmed in a 4–1–4 decision. Although yielding a fractured opinion, the judgment of the Court was driven by a concern regarding a so-called “ ‘merger problem.’ ” Santos, 553 U.S. at 515, 128 S.Ct. 2020 (plurality opinion). As Justice Scalia, writing for the plurality, explained, if “proceeds” meant “gross receipts,” “nearly every violation of the illegal-lottery statute would also be a violation of the money-laundering statute, because paying a winning bettor is a transaction involving receipts that the defendant intends to promote the carrying on of the lottery.” Id. “Since few lotteries, if any, will not pay their winners, the statute criminalizing illegal lotteries, 18 U.S.C. § 1955, would ‘merge’ with the money-laundering statute.” Id. at 515–16, 128 S.Ct. 2020. Looking beyond illegal gambling, Justice Scalia added the following:
For a host of predicate crimes, merger would depend on the manner and timing of payment for the expenses associated with the commission of the crime. Few crimes are entirely free of cost, and costs are not always paid in advance. Anyone who pays for the costs of a crime with its proceeds—for example, the felon who uses the stolen money to pay for the rented getaway car—would violate the money-laundering statute. And any wealth-acquiring crime with multiple participants would become money laundering when the initial recipient of the wealth gives his confederates their shares. Generally speaking, any specified unlawful activity, an episode of which includes transactions which are not elements of the offense and in which a participant passes receipts on to someone else, would merge with money laundering.
Id. at 516, 128 S.Ct. 2020 (footnote omitted).
Justice Stevens, writing separately and concurring in the judgment only, agreed, finding that “[a]llowing the Government to treat the mere payment of the expense of operating an illegal gambling business as a separate offense is in practical effect tantamount to double jeopardy.” Id. at 527, 128 S.Ct. 2020 (Stevens, J., concurring) (emphasis added).
In Halstead, we considered the impact of Santos in the context of a healthcare fraud scheme. Before turning to the record evidence, we observed generally that “ ‘[w]hen a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.’ ” Halstead, 634 F.3d at 277 (quoting Marks v. United States, 430 U.S. 188, 193, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977)). Looking to Santos, we concluded the following:
In its narrowest sense ... the five justices could be found to have held that the money laundering term “proceeds” means “net profits” when the proceeds are from an illegal gambling operation. They also agreed that the driving force for their holding was the merger problem resulting from the circumstances in that case—a prosecution of defendants for transactions that supported convictions for both illegal gambling and money laundering. Finally, they agreed that the merger problem had to be solved by defining the term “proceeds” in the money laundering statute to mean “net profits,” because if “proceeds” were defined to mean “gross receipts,” any crime involving costs would automatically become money laundering when the money received from the crime was used to pay expenses.
Id. at 278.
In evaluating how to apply this narrowest holding of Santos, we determined that because Justice Stevens' concurrence was limited to the facts of the case, it did not control cases where “a merger problem arises in the context of money laundering and an illegal activity other than illegal gambling.” Id. at 279 (emphasis added). Contemplating such other offenses, we explained that “when the illegal activity includes money transactions to pay for the costs of the illegal activity, a merger problem can occur if the government uses those transactions also to prosecute the defendant for money laundering.” Id. And, we added that Santos made “clear” that “when a merger problem arises, a judicial solution must be found to eliminate its unfairness.” Id. When, however, “the financial transactions of the predicate offense are different from the transactions prosecuted as money laundering, the merger problem ... does not even arise.” Id. at 279–80. Turning to the facts in Halstead, we found no merger problem.
Halstead was convicted of healthcare fraud under 18 U.S.C. § 1347 and money laundering conspiracy under 18 U.S.C. § 1956(h). Trained as a chiropractor, Halstead led a scheme that fraudulently billed insurance companies for non-legitimate services for patients. As part of the scheme, Halstead “orchestrated the cash flow thus obtained from the insurance companies and healthcare providers.” Id. at 273. Priority One—a clinic opened at Halstead's direction—billed for the services and received the insurance payments. Upon receiving those payments, Priority One transferred the monies to West Virginia Healthcare Management, another company formed by a coconspirator. Finally, West Virginia Healthcare Management transferred the monies to checking accounts for Halstead and a coconspirator.
Finding no merger problem, we determined that Halstead's healthcare fraud was “complete” “the moment that the healthcare provider paid money to Priority One.” Id. at 280. The money paid was the “proceeds” of the healthcare fraud. Id. When Halstead later directed that the proceeds be transferred from Priority One to West Virginia Medical Corporation and then to himself, “[t]hese transfers constituted the ‘transactions' of money laundering.” Id. We noted that these transactions “were separate from the transactions constituting healthcare fraud,” adding that while “[t]he healthcare fraud charges were defined by the obtaining of money from the fraudulent billing of healthcare providers, ... the money laundering charge was defined by transferring the proceeds thereafter.” Id. at 280–81. Accordingly, we concluded that “the merger problem never arises in the circumstances of this case.” Id. at 281.
Because Cloud was not convicted of operating an illegal gambling business, we must determine whether a merger problem arises on the facts of this case. Halstead, 634 F.3d at 279 (noting that “Justice Stevens' opinion ... would require addressing that situation on a case-by-case approach” and “leav[ing] further development of a solution to a future case that presents the problem”).
The indictment charges Cloud in Count 27 with conspiracy to commit money laundering and in Counts 28–33 with six substantive money laundering transactions. Counts 28 and 30 stem from commissions Cloud paid to Joseph Goines for recruiting buyers William Lytle and Reggie Thompson. Goines testified that he was paid between $1,500 and $3,000 for each investorhe “brought” to Cloud. J.A. 1812. Count 29 relates to a kickback Cloud paid to Lytle, which Lytle testified was paid only after he signed the relevant documents. Count 31 charges Cloud with repaying Kenneth Strong for money Strong loaned to Cloud for a previous closing. And Counts 32 and 33 relate to payments Cloud made to Burt Bartniski, another buyer. For example, Count 32 was based in part on $5,000 that Bartniski received approximately one week after Bartniski purchased a property. Bartniski also testified that Cloud paid the first mortgage payment for a property he purchased, forming the basis of Count 33. As Bartniski testified, Cloud informed him that he would help cover the mortgage.
We have stated that “[a]n individual cannot be convicted of money laundering for paying the ‘essential expenses of operating’ the underlying crime.” Halstead, 634 F.3d at 279 (quoting Santos, 553 U.S. at 528, 128 S.Ct. 2020 (Stevens, J., concurring)). In this case, Cloud's money laundering convictions are based on payments to recruiters, buyers, and other coconspirators for the role each person played in the mortgage fraud scheme. Cloud's mortgage fraud depended on the help of others, and their help, in turn, depended on payments from Cloud. Such payments are no different than “the felon who uses the stolen money to pay for the rented getaway car” or “the initial recipient of the wealth” in “any wealth-acquiring crime with multiple participants ... [who] gives his confederates their shares.” Santos, 553 U.S. at 516, 128 S.Ct. 2020 (plurality opinion). Because Cloud's money laundering convictions on Counts 28–33 were based on paying the “essential expenses” of his underlying fraud, we find a merger problem.
In resisting this conclusion, the government (1) seizes on our language in Halstead that the underlying fraud was “complete” before the money laundering transactions occurred, and (2) argues that Cloud's money laundering transactions were not for services rendered, but were designed to perpetuate the scheme. We are not persuaded.
First, the government emphasizes that in finding no merger in Halstead, we noted that “ ‘Halstead's commission of healthcare fraud was complete before he committed money laundering.’ ” Appellee's Supp. Br. 1 (quoting Halstead, 634 F.3d at 272). The government's argument, however, does not reflect a complete reading of Halstead and fails to account for important factual differences in the two cases.
Initially, we note that Halstead was a standard money laundering case, where the defendant transferred the fraudulent proceeds through various companies and into his own coffers. Cloud, conversely, was not charged with transferring the proceeds into his personal account, but with paying persons involved in the underlying fraud for services necessary to the operation of the fraud. This factual difference is important, as Halstead makes clear that a merger problem can occur in these circumstances. See634 F.3d at 279 (“[A] merger problem can occur” “when the illegal activity includes money transactions to pay for the costs of the illegal activity ... [and] the government uses those transactions also to prosecute the defendant for money laundering.”). In transferring the proceeds from the fraud into his own account, the defendant in Halstead was not paying the “essential expenses” of the underlying fraud.
We acknowledge that the fraudulent proceeds in Halstead were eventually transferred into two coffers. See Halstead, 634 F.3d at 273 (explaining that “Halstead and Burns [a coconspirator] also orchestrated the cash flow thus obtained from the insurance companies and healthcare providers” with the money ultimately “transferred ... to Halstead and Burns' checking accounts”). The simple fact that a coconspirator was paid, however, does not result in a merger problem. Rather, we must look to the nature of the payment to determine if it charges the “money transactions to pay for the costs of the illegal activity.” Id. at 279. In Halstead, both men received proceeds from the fraud according to a plan which they orchestrated to ensure that the monies ultimately flowed to their coffers. Indeed, the indictment charged Burns in at least six instances with causing the transfer of funds from Healthcare Management to his account. Thus, in facilitating these transfers, Halstead and Burns were not paying the expenses of the fraud, but rather were reaping the fruits of their crimes.
Moreover, the distinction we made in Halstead between fraudulently procuring funds from a healthcare provider and then funneling those funds from one corporate entity to another was in response to the defendant's argument that “both crimes were the product of a single conspiracy, implying that a conspiracy can involve only the commission of one crime.” Id. at 280. As we explained in Halstead and reaffirm now, it is without doubt that a single conspiracy can encompass an agreement to commit multiple discrete offenses. In rejecting Halstead's argument to the contrary, we emphasized that the healthcare fraud was “complete” once the fraudulent funds were paid by the healthcare provider—before the transactions underlying Halstead's money laundering convictions occurred. Here, in reversing Cloud's money laundering convictions, we do not suggest that the crimes merge because they were part of one conspiracy. Rather, we conclude that the specific money laundering transactions were the payment of the “essential expenses” of Cloud's underlying fraud, thus presenting a merger problem.
Nonetheless, the government argues that Cloud's fraud—“obtaining money by the submission of false information to the victim—precisely parallels the underlying fraud in Halstead.” Appellee's Supp. Br. 1. This argument, however, focuses exclusively on the underlying fraud offense, ignoring the critical other half of our merger analysis—the money laundering transactions. In Halstead, the money laundering transactions were not based on the payment of the essential expenses of his fraud. It was for that reason that we noted there that the money laundering transactions were “ distinct from and subsequent to the transactions involved in the healthcare fraud itself.” Halstead, 634 F.3d at 271 (emphasis added). In other words, Halstead's healthcare fraud did not require the cooperation, procured via payment, of coconspirators to succeed. Cloud's mortgage fraud conspiracy, on the other hand, could not have succeeded otherwise.
Simply put, Cloud lured his coconspirators with promises of payment. Without these payments, there would have been no recruiters, no buyers, no coconspirators. And without the recruiters to provide the buyers, or the buyers to provide the good credit, or the coconspirators more generally, there would have been no mortgage fraud. Applying Santos as interpreted in Halstead, we conclude that Cloud's substantive money laundering transactions were simply the “essential expenses” of his underlying fraud, thus presenting a merger problem.
We also reject the government's second argument that Cloud's money laundering transactions were intended to perpetuate the scheme, not to pay for services rendered. On this point, the government urges that the transactions underlying Cloud's money laundering convictions were designed to “lure [buyers] into engaging in additional fraudulent transactions” and “to perpetuate the scheme.” Appellee's Supp. Br. 7–8. The evidence, however, shows that Cloud paid these recruiters, buyers, and coconspirators for services performed. And, as Justice Scalia noted in Santos, “costs are not always paid in advance.” 553 U.S. at 516, 128 S.Ct. 2020 (plurality opinion). That these same individuals were often repeat players in the conspiracy does not negate the fact that the specific transactions alleged in Counts 28–33 of the indictment were for the past and essential expenses of Cloud's mortgage fraud conspiracy.
In rejecting the government's arguments, we recognize the difficult line-drawing problems cases like this one present. See Santos, 553 U.S. at 545, 128 S.Ct. 2020 (Alito, J., dissenting) (describing the difficulty of distinguishing between payments to participants who “expect to receive a certain amount for their services whether or not the operation is profitable,” and payments to those who expect to receive “a certain percentage of the gross revenue (perhaps even in addition to a salary)”). Nonetheless, the similarity of the payments charged in Counts 28–33 to those the defendant in Santos made to the runners, collectors, and winners in the illegal gambling operation compels us to find a merger problem.
We reach a different conclusion as to Count 27, which alleges conspiracy to commit money laundering. Unlike Cloud's substantive money laundering charges, the conspiracy charge was not tied to any specific payment to a recruiter, buyer, or coconspirator. Moreover, there was evidence that Cloud used the profits from his previous flips to finance additional purchases. Specifically, attorney John Lee testified that Cloud left monies he acquired from previous closings in Lee's trust account, and later “us[ed] those monies to buy” the property underlying Count 6. J.A. 945. In utilizing monies from previous properties to finance future purchases, Cloud was not paying the “essential expenses” of the underlying crime. Thus, Count 27 does not present a merger problem.
Cloud asserts that because the jury instructions referenced “proceeds” not “profits,” the government's argument that the jury “could have” concluded that his money laundering conspiracy was tied to “profits” is “invalid.” Appellant's Br. 51–52. We are not persuaded, as jury instructions are subject to harmless error review. See United States v. Ramos–Cruz, 667 F.3d 487, 496 (4th Cir.2012) (“We find an error in instructing the jury harmless if it is clear beyond a reasonable doubt that a rational jury would have found the defendant guilty absent the error.”) (internal quotations omitted). For the reasons supporting our finding of no merger, we also find that any error in the court's instruction did not affect Cloud's conviction on Count 27.
Nor do we believe that Count 27 fails for insufficient evidence. In reviewing the denial of a Rule 29 motion for judgment of acquittal, we must affirm a guilty verdict that, when viewed in the light most favorable to the government, is supported by “ ‘substantial evidence.’ ” Alerre, 430 F.3d at 693 (quoting United States v. Burgos, 94 F.3d 849, 862 (4th Cir.1996) (en banc)). “Substantial evidence” is defined as “ ‘evidence that a reasonable finder of fact could accept as adequate and sufficient to support a conclusion of a defendant's guilt beyond a reasonable doubt.’ ” Id. (quoting Burgos, 94 F.3d at 862). Based on Lee's testimony, we are satisfied that the government met its burden of proof on Count 27.
In reversing Cloud's substantive money laundering convictions, we again emphasize that wide-ranging conspiracies such as Cloud's can support multiple discrete convictions. See Halstead, 634 F.3d at 280. And, we do not suggest that the government may not successfully prosecute promotional money laundering in connection with similar conspiracies where the transactions are designed “to promote the carrying on of specified unlawful activity,” 18 U.S.C. § 1956(a)(1)(A)(i). On the facts of this case, however, we are confronted with a merger problem that Halstead directs us to correct.
In Halstead, we concluded that Santos made “clear” that “when a merger problem arises, a judicial solution must be found to eliminate its unfairness.” 634 F.3d at 279. And we added that the solution dictated by Santos for a merger problem in the context of an illegal gambling business “is to define the proceeds ... as its net profits.” Id. In the context of another predicate crime, however, we left “further development of a solution to a future case that presents the problem.” Id. That case is now before us, and we see no better course to correct the merger problem here than that adopted in Santos, defining “proceeds” as “profits.” So defined, we reverse Cloud's substantive money laundering convictions on Counts 28–33 because they do not comport with the definition of “proceeds” announced in Santos.
Following the Court's holding in Santos, Congress amended the money-laundering statute to provide a definition of “proceeds.” Fraud Enforcement and Regulatory Act of 2009, Pub.L. No. 111–21, § 2(f)(1), 123 Stat. 1617, 1618 (2009). As amended, the statute defines “proceeds” broadly as “any property derived from or obtained or retained, directly or indirectly, through some form of unlawful activity, including the gross receipts of such activity.” 18 U.S.C. § 1956(c)(9). See Garland v. Roy, 615 F.3d 391, 394 n. 1 (5th Cir.2010). Thus, the issue we address today is not likely to arise in many more cases.
Cloud also challenges the district court's loss calculation. We review for clear error the district court's factual determination of the amount of loss attributable to Cloud, United States v. Miller, 316 F.3d 495, 503 (4th Cir.2003), mindful that the court “need only make a reasonable estimate of the loss,” U.S. Sentencing Guidelines Manual (“U.S.S.G.”) § 2B1.1 cmt. n. 3 (C).
The district court concluded that Cloud was responsible for an intended loss equal to the entire loan amount on the properties, minus a proxy for the value of the house. The proxy was set by the price Cloud paid for the properties prior to the “flip.” After subtracting the proxy from the value of the loan, the court arrived at an intended loss figure of approximately $10 million. As to the collateral loss, the district court relied on expert testimony regarding the impact that foreclosed properties have on surrounding communities. The court, however, reduced the expert's collateral figure—originally encompassing properties within 800 to 900 meters of the foreclosed property—based on the court's conclusion that an impact “over a half mile away” was not reasonably foreseeable to Cloud. J.A. 3237. Reducing the collateral loss radius to 100 to 200 meters, the court found that Cloud was responsible for over $9 million in collateral loss.
On appeal, Cloud argues that neither the intended nor collateral loss figures were tied to evidence of Cloud's intent. Specifically, Cloud asserts that the court “overlooked evidence that Cloud intended that the lenders would be repaid at least in part and that Cloud did not intend any losses to the community.” Appellant's Br. 55.
Cloud also alleges that the district court “did not apply [a] subjective measure” of intent as to the intended loss. Appellant's Br. 57. We find that “[t]his conclusory remark is insufficient to raise on appeal any merits-based challenge to the district court's ruling,” Eriline Co. S.A. v. Johnson, 440 F.3d 648, 653 n. 7 (4th Cir.2006) (citing Fed. R.App. P. 28(a)(9)(A) and its requirement that argument section of brief contain “appellant's contentions and the reasons for them, with citations to the authorities and parts of the record on which the appellant relies”). Moreover, we note that in making its loss ruling, the district court rejected Cloud's argument that there was no intended loss, finding that “there was really an intent to defraud lending institutions and to run away with that spread.” J.A. 3108. In support, the court noted that Cloud received phone calls from “investors who said they [were] receiving foreclosure notices, and [that] they thought that [Cloud] was covering the debt service,” id. 3101, and that “[i]f there truly was no intent to cause a loss to a financial institution or to an investor, then [Cloud] would have done what he did not do, and that was stay tied close to these investors when they were ... complaining to him that they were getting foreclosure notices,” id. 3108. Thus, we think the district court adequately accounted for Cloud's subjective intent.
As to intended loss, Cloud asserts that “even the most government-friendly reading of the evidence does not establish that Cloud intended the lenders to lose the entire spread between his purchase and sale prices.” Id. 58. In support, Cloud argues that he could have intended this amount of loss only if he intended that the buyers would not make a single mortgage payment. According to Cloud, evidence of his property management shows otherwise.
Although the district court noted that Cloud “did do some property management,” it clarified that this management was “very minor ... and a lot of it was to make the property look more appropriate for closing.” J.A. 3108–09. And, it concluded that the “small amount of property management [Cloud] did does not offset the fact that virtually the vast majority of the properties went into foreclosure.” Id. 3109. We find no clear error in the district court's assessment of Cloud's property management as it influenced the intended loss calculation.
As to collateral loss, Cloud faults the district court for overlooking that many of the properties were in foreclosure when Cloud purchased them, and argues that the government failed to present evidence demonstrating why Cloud should have foreseen that additional loss would occur if previously foreclosed properties were later returned to foreclosure. We disagree.
First, no evidence as to the loss caused by a subsequent foreclosure was presented because Cloud did not raise this objection at sentencing. Second, Cloud does not dispute the general rule that foreclosures have a negative impact on the surrounding communities; indeed, Cloud's own expert conceded as much. See id. 3221 (testimony by Cloud's expert that foreclosures “[g]enerally” have a negative impact on surrounding communities). Nor does Cloud suggest that all of the properties were previously foreclosed. Third, Cloud's argument assumes that none of the previously foreclosed properties would have been purchased by legitimate buyers, an assumption that is simply not supported by the evidence or by common sense. After careful review, we are satisfied that the district court did not clearly err in finding that the foreclosures that resulted from Cloud's scheme had a collateral impact on the surrounding communities.
Cloud also argues that “once the district court found that the collateral damage victims were ‘unintended,’ it could not permissibly resort to collateral-loss findings based on Cloud's expectations.” Appellant's Br. 60–61. This argument, however, seizes on a misstatement by the district court that the collateral loss victims were “unintended.” J.A. 3103. The court subsequently corrected its mistake, noting that it should have referred to the victims as collateral, not unintended victims, because although specific intent was not required, “the defendant has to have some recognition that the collateral damage will be broad.” Id. 3115. Moreover, in the case Cloud cites in support, United States v. Pendergraph, 388 F.3d 109, 113 (4th Cir.2004), the district court found that there was “ no intended loss.” Here, the district court found Cloud responsible for $10 million in intended loss.
In any event, even assuming some error in the district court's loss calculation, we are satisfied that such error was harmless. Cloud's offense level was increased by 20 levels based on a loss of more than $7 million but less than $20 million. SeeU.S.S.G. § 2B1.1(b)(1)(K). To qualify for the next lowest offense level, Cloud would have to show that the district court's loss finding was wrong to the tune of $12 million, taking the loss to more than $2.5 million but less than $7 million. See id.§ 2B1.1(b)(1)(J). Based on his arguments about the impact of his property management and the double foreclosures, this he cannot do. Accordingly, any error in the loss calculation was harmless. See United States v. Mehta, 594 F.3d 277, 283 (4th Cir.2010) (noting that in evaluating a district court's sentencing calculations, any error “is harmless if the resulting sentence was not longer than that to which the defendant would otherwise be subject” and finding that because the defendant “received the same sentence that he would have received had the district court not erred in its calculations.... the error is harmless” (internal quotations omitted)); United States v. Allen, 491 F.3d 178, 194 (4th Cir.2007) (finding loss calculation harmless because the challenged loss figure resulted in the same offense level as the calculated loss level).
The government initially argued that the district court did not plainly err in its reimbursement order. Before oral argument and citing our decision in United States v. Moore, 666 F.3d 313 (4th Cir.2012), however, the government conceded error on this point.
In United States v. Moore, 666 F.3d 313 (4th Cir.2012), we noted that under the Criminal Justice Act, 18 U.S.C. § 3006A (“CJA”), the government must provide adequate legal representation to criminal defendants charged with a federal felony who are unable to pay, but if the court subsequently finds that the defendant “ ‘is financially able to obtain counsel or to make partial payment for the representation,’ ” repayment is authorized under subsection (f). Id. at 321 (quoting 18 U.S.C. § 3006A(c)). Subsection (f) authorizes a court to order repayment of attorneys' fees “[w]henever ... the court finds that funds are available for payment from or on behalf of a person furnished representation.” 18 U.S.C. § 3006A(f).
In Moore, we held that to order reimbursement of attorneys' fees, a court must “find[ ] that there are specific funds, assets, or asset streams (or the fixed right to those funds, assets or asset streams) that are (1) identified by the court and (2) available to the defendant for the repayment of the court-appointed attorneys' fees.” 666 F.3d at 322. We noted that the district court made no findings that Moore “is financially able ... to make partial payment for the representation” or that “funds are available for payment.” Id. at 323 (internal quotations omitted). And we took “particular note” of the fact that in the absence of such findings, the district court simultaneously concluded that Moore was unable to pay a fine or interest. Id. Finding that the district court's reimbursement order conflicted with the statutory requirements, we vacated that portion of the judgment and remanded for resentencing.
Similarly, the district court here made no findings regarding Cloud's ability to reimburse the government for attorneys' fees or the availability of such funds. To the contrary, the district court concluded that Cloud was unable to pay a fine or interest. Because Cloud's reimbursement order is of the same type we rejected in Moore, we vacate that portion of the district court's judgment and remand for resentencing.
For the reasons stated, we affirm the district court's evidentiary rulings and loss calculation. We reverse Cloud's money laundering convictions on Counts 28–33, but otherwise affirm his convictions. Finally, we vacate the attorneys' fees reimbursement order, and remand for resentencing.
AFFIRMED IN PART, REVERSED IN PART, VACATED IN PART, AND REMANDED