[1081]*1081OPINION
IKUTA, Circuit Judge:
Kyle Grasso appeals his convictions for money laundering, bank fraud, loan fraud, and conspiracy to commit loan and bank fraud, stemming from a Los Angeles-based scheme to defraud mortgage lenders. We conclude that the evidence adduced at trial, taken in the light most favorable to the government, was adequate to enable a rational trier of fact to find the essential elements of each conviction. See United States v. Nevils, 598 F.3d 1158, 1163-64 (9th Cir.2010) (en banc). Accordingly, we affirm on all counts.
I
We have already explained how this scheme operated in United States v. Rizk, 660 F.3d 1125 (9th Cir.2011) (affirming conviction of Lila Rizk, a real estate appraiser in the scheme and one of Grasso’s co-defendants). Therefore, we provide only a brief overview of the scheme before detailing the particular aspects relevant to Grasso’s appeal.
A
The scheme, as crafted by Mark Abrams, a mortgage broker, and Charles Elliott Fitzgerald, a real estate developer, took advantage of the real estate frenzy of the early 2000s. The conspirators would enter into a purchase agreement for a home in an exclusive Westside Los Angeles community, and then obtain a loan for significantly more than the sale price, pocketing the extra money. For this scheme to work, the conspirators had to exert control over multiple aspects of each real estate transaction. Among other things, the sellers and agents had to keep the true purchase price of the home confidential; the appraisal reports had to show the falsely inflated values of the homes; the title and escrow companies had to prepare two sets of documents, one showing the actual purchase price (for the seller) and one showing the inflated purchase price (for the lender); .and finally the Multiple Listing Service (MLS) had to show the falsely inflated sales price of the homes. Abrams and Fitzgerald coordinated the efforts of a range of colleagues to make this conspiracy work. Abrams relied on Jamieson Matykowski, one of his employees, and Richard Maize, who controlled Americorp Funding, among others. Americorp, a mortgage broker, would send loan packages to lending agents for review and approval. The conspirators generally targeted banks that did not require their lending agents to use rigorous documentation and approval standards. Over the course of the scheme, the conspirators conducted roughly 80 fraudulent transactions. The banks that ultimately financed the loans that their lending agents approved, including Lehman Brothers (through its lending agent, Aurora Loan Services), GreenPoint Bank (through its lending agent, GreenPoint Mortgage), and RBC Mortgage Company, together lost at least $46 million in the Abrams-Fitzgerald conspiracy.
Grasso was one of Abrams and Fitzgerald’s recruits for this scheme. Grasso and his partner Joseph Babajian were successful real estate agents in the same affluent area that Abrams and Fitzgerald targeted, Westside Los Angeles. From 2000 to January 2001, Grasso and Babajian were the top-producing real estate agents at Fred Sands Realtors in Beverly Hills. Grasso and Babajian subsequently left Fred Sands and became affiliated with Prudential California Realty. Prudential compensated Grasso and Babajian for their move by conveying an ownership interest in several of Prudential’s subsidiaries to Grasso and Babajian’s wholly owned company, FSC Ventures. One of these subsidiaries was Cal Title, a title insurance company whose lax approval process would become [1082]*1082instrumental to the co-conspirators’ operation.
At trial, the government presented evidence to prove that Grasso became involved in the Abrams-Fitzgerald scheme some time in 2000 and worked with them through at least late 2002. According to the government, Grasso. participated primarily by identifying houses to be included in the scheme, ensuring that the seller would keep the sales price confidential, managing the information reported in the MLS listings, and obtaining the title insurance and escrow documents needed for the conspiracy through his access to Cal Title. The government’s case against Grasso rested heavily on evidence relating to six real estate transactions, which we describe in the order they occurred.
We begin with Grasso’s purchase of a property on Claridge Drive, Beverly Hills for his personal use. In July 2000, after Grasso separated from his wife, he was in the market for his own home, and focused on the Claridge Drive property. According to the evidence adduced at trial, Abrams thought that Grasso’s skills and contacts were useful in furthering the scheme and wanted to “ingratiate” himself with Grasso to induce him to continue helping with the fraud. Therefore, Abrams offered to help Grasso purchase the Claridge Drive property by following the same basic blueprint used for other transactions in the scheme.' Under the plán, Abrams would front Grasso the money for a down payment, Grasso would obtain an inflated loan, and then use the extra funds to pay Abrams back. According to Abrams, Grasso knew how the scheme worked and was a willing participant. Matykowski confirmed that Grasso acknowledged he was “going to do one of our deals” by inflating the purchase price of the Claridge Drive property “to forego putting a large down payment down.”
The plan moved forward over the summer. Grasso entered into a purchase agreement with Jose Menendez, the owner of the Claridge Drive property, to purchase the house for $890,000.1 In. early September 2000, Grasso submitted two fraudulent loan applications to Americorp Funding, one seeking a first mortgage of $746,250, and the second seeking a home equity credit line of $149,200. Both applications stated that the purchase price of the home was $995,000, and the first application stated that Grasso was making a down payment of some $120,000. Americorp submitted Grasso’s application to GreenPoint Mortgage. In reviewing the applications, GreenPoint Mortgage identified a red flag: Grasso already owned a home that had a higher value than the Claridge Drive property, which raised the inference that Grasso planned to use the Claridge Drive property as a rental. In response to GreenPoint Mortgage’s inquiry on this point, Grasso falsely stated that he and his wife intended to move into the home. Satisfied with that explanation, GreenPoint Mortgage approved Grasso’s loans. As was the standard procedure for all loans that GreenPoint Mortgage originated at that time, GreenPoint Bank funded the loan.
At closing, Abrams’s in-house escrow company prepared fraudulent documents reflecting a purchase price of $995,000, but sent a settlement statement to Menendez stating the correct $890,000 sales price.
At the same time he was purchasing the Claridge Drive property, from August to [1083]*1083September 2000, Grasso sought Abrams and Fitzgerald’s help with a client’s property on Alta Drive in Beverly Hills. The property had been on the market for nearly a year, and the seller, Vigen Shaghzo, was putting pressure on Grasso to sell it. Abrams and Fitzgerald agreed to buy the Alta Drive property if they could work one of their fraudulent deals.
The Alta Drive transaction unfolded as follows. In early August 2000, Abrams offered $2,000,000 to Shaghzo in the name of Abrams’s father, a straw purchaser. According to Abrams, Grasso knew that Abrams’s father was not actually purchasing the home. After Shaghzo accepted the $2,000,000 offer, Grasso withdrew the $2,050,000 MLS listing for the house and relisted it for $4,495,000. This false relisting caused a problem, however: when Shaghzo found out about it, he ordered Grasso to correct the misinformation, and Grasso agreed to do so. Abrams and Fitzgerald then submitted a fraudulent loan application for an inflated purchase price of $4,395,000 to Aurora Loan Services for approval. A loan for the full amount was funded by Lehman Brothers. After the sale closed in September 2000, and the escrow company disbursed the loan proceeds from Lehman Brothers, Abrams paid Grasso’s firm $46,436 in commissions. Several months later, MLS reported the property sold at its true price of $2,000,000. To prevent Aurora Loan Services and Lehman Brothers from seeing this accurate listing, Grasso again arranged to change the MLS listing to show a sales price of $4,495,000.
In January 2001, after Grasso became affiliated with Prudential and obtained an indirect interest in Cal Title, Abrams and Fitzgerald began using Cal Title to provide title insurance for their purchases. Cal Title was willing to prepare two title insurance policies, one with the inflated purchase price (which would be provided to- the lender) and the second with the actual purchase price (that the seller could review). Although Abrams and Fitzgerald attempted to use other title companies for their transactions, these companies soon declined to work with them, and Abrams and Fitzgerald began using Cal Title exclusively, whether or not.Grasso or Babajian was acting as the agent for the transaction.
Sometime in 2001, Grasso called Matykowski to complain that Abrams and Fitzgerald had been using Cal Title for transactions where he was not the agent. Grasso was agitated about this development, and forbade the Abrams-Fitzgerald team from using Cal Title unless Grasso and Babajian were in the deal. Abrams later told Matykowski that he agreed to pay Grasso and Babajian — via Prudential — a commission on transactions where Cal Title was involved, even when they were not serving as agents.
From approximately February 2001 through summer 2002, Grasso represented Abrams and Fitzgerald in their acquisition of four additional properties relevant to this appeal. For a property on Mandeville Canyon Road, Beverly Hills, Grasso submitted several unsuccessful purchase offers using the names of different straw buyers, before submitting a third, successful offer on behalf of “Jamieson Matykowski and/or assignee.” While the transaction was in escrow, Matykowski purportedly assigned the purchase agreement to yet another straw buyer. As in the Alta Drive transaction, the straw buyer was not the true purchaser and in fact did not know the conspirators had used his name. In addition to multiple straw purchasers, the conspirators also used two different escrow companies: the original escrow company resigned after determining it [1084]*1084was uncomfortable with the transaction, and was replaced by Cal Title. FSC Ventures (Grasso and Babajian’s wholly owned company) earned $19,231 in commissions for the deal.
Grasso also assisted the conspirators in purchasing a property on Claircrest Drive in Beverly Hills. According to Matykowski, Grasso was aware that the conspirators were using a straw purchaser for this property, as well. At one point, Matykowski testified, he was in Grasso’s office, and jokingly told Grasso, “Turn your back for a second I’m going to sign this contract.” Matykowski then forged the name “Hugo Wendt” on the purchase offer. Grasso treated this forgery as “just normal.” Once the seller accepted the offer for the Claircrest Drive property, Grasso asked the escrow officer to assign the purchase agreement to a different straw purchaser. Abrams and Fitzgerald paid FSC Ventures $21,490 in commissions for this transaction when it closed.
In summer 2002, Abrams and Fitzgerald purchased two more Beverly Hills properties, one on Yoakum Drive and the other on Benedict Canyon Drive. Although they did not use either Grasso or Babajian as their real estate agent, they did use Cal Title. Consistent with Abrams’s prior agreement, Abrams paid Prudential (which in turn paid Grasso and Babajian through FSC Ventures) a three percent commission for each transaction, which together amounted to $50,833. Abrams made these payments separately after the deals closed and the bank had already wire transferred the loan proceeds; the payments did not go through escrow or show up on the settlement statements for the transactions. Abrams and Grasso characterized these payments as “referral fees.”
B
In August 2007, a grand jury indicted Grasso for one count of conspiracy to commit bank fraud and loan fraud, in violation of 18 U.S.C. § 371.2 This charge named Grasso and three co-defendants, including Rizk and Babajian, along with six other co-conspirators, including Abrams and Fitzgerald. Listing nine transactions as overt acts, the government alleged that the co-conspirators devised and executed a scheme to commit bank fraud against Lehman Brothers, GreenPoint Bank, and other federally-insured institutions between 2000 and 2003, and to commit loan fraud by submitting false statements, reports, and valuations in connection with the listed transactions.3
The indictment also charged Grasso with one count of bank fraud and aiding and abetting, in violation of 18 U.S.C. §§ 2,4 1344(1);5 seventeen counts of loan fraud [1085]*1085and aiding and abetting, in violation of-18 U.S.C. §§ 2, 1014;6 and three counts of money laundering and aiding and abetting, in violation of 18 U.S.C. §§ 2, 1956(a)(1).7 The bank fraud charge named all four defendants and rested on the same allegations as the conspiracy charge. The money laundering and loan fraud charges related to specific real estate transactions.8 The three money laundering charges related to individual “referral” payments that Abrams made to Prudential after the Yoakum Drive and Benedict Canyon Drive transactions.
After a jury trial, Grasso was convicted on all conspiracy, bank fraud, and money laundering charges, and on the loan fraud charges relating to the Claridge Drive, Alta Drive, Mandeville Canyon Road, and Clairerest Drive transactions. Grasso moved for acquittal on all charges or in thé alternative for a new trial. The district court denied these motions, rejecting his sufficiency of the evidence claims. The court sentenced Grasso to twelve months and one day in prison and $13 million in restitution.
II
On appeal, Grasso argues that the district court erred in denying his motion for acquittal, see Fed.R.Crim.P. 29, because the evidence was insufficient to support a guilty verdict on the conspiracy, bank fraud, loan fraud, and money laundering counts for which he was convicted.9
We have jurisdiction under 28 U.S.C. § 1291. We review denials of Rule 29 motions for acquittal de novo. United States v. Gonzalez-Diaz, 630 F.3d 1239, 1242 (9th Cir.2011). In considering a claim that the evidence ‘ adduced at trial was insufficient to sustain a conviction, “[a] reviewing court must consider the evidence presented at trial in the light most favorable to the prosecution.” United States v. Nevils, 598 F.3d 1158, 1164 (9th Cir.2010) (en banc) (citing Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979)). This means that “a reviewing court ‘must presume — even if it [1086]*1086does not affirmatively appear in the record — that the trier of fact resolved any such conflicts in favor of the prosecution, and must defer to that resolution.’ ” Id. (quoting Jackson, 443 U.S. at 326, 99 S.Ct. 2781). Second, the court must determine whether the evidence, viewed in that manner, “is adequate to allow any rational trier of fact to find the essential elements of a crime beyond a reasonable doubt.” Id. (internal quotation marks and alterations omitted).
We begin with Grasso’s appeal of his conviction for conspiracy to commit loan fraud and bank fraud. To convict Grasso of conspiracy, the government must first prove that a conspiracy existed. “To prove a conspiracy under 18 U.S.C. § 371, the government must first establish; (1) an agreement to engage in criminal activity, (2) one or more overt acts taken to implement the agreement, and (3) the requisite intent to commit the substantive crime.” Rizk, 660 F.3d at 1134 (quoting United States v. Sullivan, 522 F.3d 967, 974 (9th Cir.2008)). “Once the existence of the conspiracy is shown, evidence establishing beyond a reasonable doubt a knowing connection of the defendant with the conspiracy, even though the connection is slight, is sufficient to convict him of knowing participation in the conspiracy.” United States v. Meyers, 847 F.2d 1408, 1413 (9th Cir.1988). We have explained that a defendant may have a “slight connection” to a conspiracy even if the defendant did not know all the conspirators, did not participate in the conspiracy from its beginning or participate in all its enterprises, or otherwise know all its details. See United States v. Reed, 575 F.3d 900, 924 (9th Cir.2009). However, “[i]t is not a crime to be acquainted with criminals or to be physically present when they are committing crimes.” United States v. Herrera-Gonzalez, 263 F.3d 1092, 1095 (9th Cir.2001).
Where the defendant has a connection (even if slight) to the conspiracy, the government must also show that the defendant’s connection to the conspiracy is knowledgeable; “that is, the government must prove beyond a reasonable doubt that the defendant knew of his connection to the charged conspiracy.” Meyers, 847 F.2d at 1413. To establish a knowing connection, “[t]here need not, of course, be proof that the conspirators were aware of the criminality of their objective,” Ingram v. United States, 360 U.S. 672, 678, 79 S.Ct. 1314, 3 L.Ed.2d 1503 (1959). Instead, the government must show that the defendant was aware of “the unlawful object toward which the agreement [was] directed,” United States v. Krasovich, 819 F.2d 253, 255 (9th Cir.1987); see also id. (“Knowledge of the objective of the conspiracy is an essential element of any conspiracy conviction.”) (citing Ingram, 360 U.S. at 678, 79 S.Ct. 1314). The government may rely on circumstantial evidence and inferences drawn from that evidence in order to prove the defendant’s knowing connection to the conspiracy. See United States v. Johnson, 297 F.3d 845, 868-69 (9th Cir.2002).
Here, it is undisputed that the government sufficiently proved the existence of a conspiracy to commit bank fraud and loan fraud. As we explained in Rizk, “Abrams, Fitzgerald, and others associated with them initiated and carried out a scheme to defraud mortgage lenders.” Rizk, 660 F.3d at 1128. Nor does Grasso contest that he participated in the conspiracy; rather, the disputed issue is whether he did so knowingly.
To that end, Grasso argues that the government failed to adduce sufficient evidence to prove he had “knowledge of the objective of the conspiracy,” id. at 1134. According to Grasso, Abrams and Fitzgerald targeted and manipulated him because of his “highly respected and ex[1087]*1087tremely successful” real estate practice in Los Angeles. Grasso claims that the evidence showed that his involvement in the conspiracy was limited to the legitimate aspects of the real estate transactions, and he was unaware of the fraudulent steps in Abrams and Fitzgerald’s scheme.
Viewing the facts in the light most favorable to the prosecution, Nevils, 598 F.3d at 1164, we conclude that the government presented more than sufficient evidence that Grasso knew the objectives of the fraudulent scheme. A rational jury could determine beyond a reasonable doubt that as early as July 2000, Grasso knew that the real estate transactions were structured to deceive Lehman Brothers, GreenPoint Bank, and other federally insured banks, and that he assisted in this deception. The evidence showed that Grasso assisted his co-conspirators in convincing lenders that various properties were sold for more than their actual sales prices. Grasso personally benefitted from falsely inflating the purchase price for his Claridge Drive property, and he personally inflated the MLS listing in the Alta Drive transaction. He was aware of the role Cal Title played in creating two different title insurance policies, one showing the inflated purchase price and one showing the true purchase price. Matykowski’s testimony that he joked with Grasso about forging the signature of a straw purchaser, along with other evidence, indicated that Grasso knew that the conspirators furthered the scheme by using the names of straw purchasers who might not even be aware of the transaction. Finally, the evidence showed that Grasso was an expert in his field, so the jury could reasonably conclude that he was not duped by Abrams and Matykowski into innocently carrying out the acts described above.
We reached a similar conclusion in Rizk. Like Grasso, Rizk argued that the evidence was insufficient to establish that she knew of the objective of the conspiracy. Rizk, 660 F.3d at 1134. We disagreed, holding that a rational jury could have found beyond a reasonable doubt that Rizk intended to defraud the victim lenders in light of evidence that Rizk significantly overvalued the properties in preparing her appraisals, requested that the MLS database be manipulated to show higher sales prices, and knew her appraisals were being used to get loans on the properties. Id. at 1134-35. Like Grasso, Rizk also raised the defense that she was “duped and used” by Abrams and Fitzgerald, but we held that the government had introduced sufficient evidence to overcome this defense by showing that Rizk was an “experienced appraiser,” “knew that her appraisals were being used to finance the purchase of properties,” and “was unduly influenced by the values put forth by Abrams and Fitzgerald.” Rizk, 660 F.3d at 1135. Here, too, the government’s evidence, taken in the light most favorable to the prosecution, see Nevils, 598 F.3d at 1164, was sufficient for the jury to reject Grasso’s exculpatory theory and to find beyond a reasonable doubt that he had “knowledge of the objective of the conspiracy.” Rizk, 660 F.3d at 1135.
Next, we consider Grasso’s appeal of his convictions for loan fraud associated with the Claridge Drive, Alta Drive, Mandeville Canyon Road, and Claircrest Drive transactions. To obtain a loan fraud conviction, the government must prove that the defendant “[1] knowingly [made] any false statement or report ... for the purpose of influencing in any way [2] the action of ... a bank insured by the Federal Deposit Insurance Corporation.” 18 U.S.C. § 1014.
Grasso begins by challenging his loan fraud conviction for the Claridge [1088]*1088Drive transaction. Grasso claims that the government did not prove that he had sufficient knowledge that the false statements he made to obtain loans to buy the Claridge Drive property would influence a federally insured bank, and so failed to prove that he made his false statements for that purpose.
We disagree with this argument. Contrary to Grasso’s claims, the government can prove the knowledge element of § 1014 by showing that Grasso made false statements for the purpose of obtaining a loan, knowing that those statements would be submitted to federally insured banks.10 See United States v. Bellucci, 995 F.2d 157, 159 (9th Cir.1993) (per curiam) (“Section 1014’s proscription of knowing misrepresentation reach[es] a defendant’s knowledge of the statement’s presentation to banks generally[,] as distinguished from a particular bank.”) (internal quotations omitted) (quoting United States v. Lentz, 524 F.2d 69, 71 (5th Cir.1975)). In Bellucci, the defendant submitted a loan application with false statements through his mortgage broker, who eventually conveyed them to a bank. 995 F.2d at 159. Because Bellucci had a long career as a developer and builder and thus “was familiar with the manner in which the lending process operates,” and even testified that he understood his broker “would use the loan applications to go shop for loans,” the court concluded that “a rational jury could easily infer beyond a reasonable doubt that Bellucci knew [that his broker] would present his loan application to a variety of financial institutions, including banks.” Id. Here, evidence at trial showed that Grasso was an experienced real estate agent who, like Bellucci, was well-versed in the mortgage lending process. For instance, Grasso and his partner Babajian, as a team, were the top-grossing real estate agents at Prudential in California in 2002. The evidence also showed Grasso’s understanding that the Claridge Drive deal would operate on the scheme’s basic blueprint, which was generally to obtain funding from banks that did not impose rigorous standards on their lending agents. Taking this evidence together, a jury could reasonably conclude that Grasso knew that his false statements made to obtain a loan would ultimately influence an insured bank. See Bellucci, 995 F.2d at 159.11
Grasso next argues that there was insufficient evidence to convict him of loan fraud because he was involved only in the legitimate first step of the real estate transactions, had no knowledge of the scheme to defraud banks, and did not make any false statements to a federally insured financial institution in the Alta Drive, Claircrest Drive, and Mandeville Drive transactions.12
[1089]*1089This argument also fails. Under Pinkerton v. United States, a defendant charged with participating in a conspiracy may be subject to liability for offenses committed as part of that conspiracy, even if the defendant did not directly participate in each offense. 328 U.S. 640, 647, 66 S.Ct. 1180, 90 L.Ed. 1489 (1946); see also United States v. Hernandez-Orellana, 539 F.3d 994, 1007 (9th Cir.2008). Pinkerton “renders all co-conspirators criminally liable for reasonably foreseeable overt acts committed by others in furtherance of the conspiracy they have joined, whether they were aware of them or not.” Hemandez-Orellana, 539 F.3d at 1007. Although we have noted there may be due process limitations to imposing Pinkerton liability on defendants “with extremely minor roles in the conspiracy,” United States v. Bingham, 653 F.3d 983, 997 (9th Cir.2011), or where “the relationship between the defendant and the substantive .offense is slight,” United States v. Castaneda, 9 F.3d 761, 766 (9th Cir.1993), overruled on other grounds by United States v. Nordby, 225 F.3d 1053 (9th Cir.2000), such concerns are not present in this case.
Taking the evidence adduced at trial in the light most favorable to the government, a reasonable juror could find all the necessary elements to impose Pinkerton liability, contrary to Grasso’s contention that he had nothing to do with the fraud. Abrams and Matykowski testified that Grasso was well aware of the conspiracy’s fraudulent purpose and had a substantial role in it by the time the Alta Drive, Mandeville Canyon Road, and Claircrest Drive transactions occurred. Moreover, Abrams admitted that he committed loan fraud in each of these transactions, which took place according to the scheme’s basic blueprint. Because Grasso was aware of this blueprint, he could reasonably have foreseen Abrams’s loan fraud in those transactions. Thus, there is no due process problem in holding him liable for Abrams’s acts under Pinkerton. Because there is sufficient evidence to uphold Grasso’s convictions for loan fraud under Pinkerton, it does not matter whether Grasso was aware of when or whether Abrams committed the acts constituting loan fraud in each transaction. Hernandez-Orellana, 539 F.3d at 1007.
C
We now turn to Grasso’s appeal of his conviction for bank fraud, based on his role in the scheme to defraud Lehman Brothers and GreenPoint Bank. The bank fraud statute, 18 U.S.C. § 1344(1), provides criminal penalties for anyone who “knowingly executes, or attempts to execute, a scheme or artifice — ... (1) to defraud a financial institution.”13 Grasso argues that the government failed to adduce sufficient evidence that he had knowledge of the object of the scheme to defraud banks or that he had the requisite intent to defraud. See Rizk, 660 F.3d at 1135 (stating that one of the “essential elements” of bank fraud under § 1344(1) is that the defendant executed or attempted to execute the fraudulent scheme “with the intent to defraud”).
In making its case that Grasso was guilty of bank fraud, the government re[1090]*1090lied on the same evidence of Grasso’s knowledge that supported his conviction for conspiracy to commit bank fraud. Accordingly, Grasso’s argument that the evidence was insufficient fails for the same reason that his challenge to his conspiracy conviction fails: viewing the facts in the light most favorable to the prosecution, there was ample evidence that Grasso was well aware of the scheme’s fraudulent objective and that he intentionally furthered the fraud. See supra at 1086-87.14
D
Finally, we address Grasso’s argument that the government presented insufficient evidence to convict him of money laundering when he received two “referral fees” for the Yoakum Drive and Benedict Canyon Drive transactions. Grasso argues that Abrams gave him referral fees for access to Cal Title, that the loan fraud and bank fraud scheme could not have occurred without such access, and, therefore, that the money laundering offense merged into the underlying loan fraud and bank fraud offenses and cannot be separately punished.
To address this argument, we first consider the relevant legal framework. Under the money laundering statute, 18 U.S.C. § 1956(a)(1), the government must prove that a defendant: [1] knew that money being used in a financial transaction was the “proceeds of some form of unlawful activity,” and [2] then conducted or attempted to conduct, a financial transaction with such proceeds; [3] for the purpose of either promoting an unlawful activity or for concealment. See 18 U.S.C. § 1956(a)(1) (emphasis added).
' In United States v. Santos, the Supreme Court attempted to define the term “proceeds” but could not reach a five-justice majority to do so. 553 U.S. 507, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008). In Santos, the defendant was convicted of running an illegal lottery, in which “runners” took bets from gamblers and delivered the money to “collectors.” 553 U.S. at 509, 128 S.Ct. 2020. Those collectors, in turn, passed the money along to defendant Santos, who then paid the runners and collectors for their services and disbursed awards to the lottery winners. Id. The government charged Santos with money laundering under § 1956(a) on the theory that the money used to pay the runners, collectors, and lottery winners was the proceeds of his illegal lottery and that Santos was using that money to promote the carrying on of the lottery activity. Id. at 509-10,128 S.Ct. 2020.
On appeal, the Court considered whether the word “proceeds” in § 1956 meant the gross receipts of an illegal activity, or only the net receipts, i.e., the profits of that activity. Writing for a four-justice plurality, Justice Scalia concluded that both definitions of “proceeds” were plausible, and that “because the ‘profits’ definition of proceeds is always more defendant-friendly than the ‘receipts’ definition, the rule of lenity dictates that it should be adopted.” Id. at 513, 128 S.Ct. 2020. In supporting this conclusion, the plurality noted that defining “proceeds” to mean “receipts” would frequently lead to a “merger problem”: “nearly every violation of the illegal-lottery statute would also be [1091]*1091a 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. at 515, 128 S.Ct. 2020. The plurality noted that Congress would not have “wanted a transaction that is a normal part of a crime it had duly considered and appropriately punished elsewhere in the Criminal Code to radically increase the sentence for that crime” by application of the money-laundering statute. Id. at 517, 128 S.Ct. 2020. Because Santos’s payments to the runners, collectors and lottery winners were a normal part of the cost of doing business (and not pure profit), the plurality concluded the payments did not constitute “proceeds” for purposes of the statute. Id.
Four justices rejected this analysis and dissented. The dissent contended that “proceeds” should be defined as “gross receipts,” and would leave any “merger problems” to district judges’ discretion at sentencing. Id. at 547, 128 S.Ct. 2020 (Alito, J., dissenting). Under this analysis, Santos’s payments to the runners, collectors and lottery winners were made out of the gross receipts of the lottery business and thus constituted “proceeds” for purposes of the statute.
Justice Stevens, in a concurrence, rejected both definitions of “proceeds,” and adopted a case-by-case approach. He was particularly troubled by the fact that the dissenting opinion allowed “the Government to treat the mere payment of the expense of operating an illegal gambling business as a separate' offense.” Id. at 527, 128 S.Ct. 2020 (Stevens, J., concurring). Considering both “common'sense” and the “rule of lenity,” Justice Stevens agreed with Justice Scalia that there was no reason to think Congress intended that “revenue generated by a gambling business that is used to pay the essential expenses of operating that business,” which is “a normal part of a crime it had duly considered and appropriately punished elsewhere in the Criminal Code,” should also be punished as money laundering. Id. at 528, 128 S.Ct. 2020. Justice Stevens explained that this result was “particularly unfair in [Santos ] because the penalties for money laundering are substantially more severe than those for the underlying offense of operating a gambling business.” Id. at 527, 128 S.Ct. 2020. Congress had capped the maximum sentence for engaging in an illegal lottery to five years, an “important limitation” that would be “eviscerated” if a court could impose a sentence of up to twenty years, the statutory maximum for money laundering. Id. Justice Stevens concluded’ that Congress could not have intended such a “perverse result.” Id. at 528,128 S.Ct. 2020.
Although Justice Stevens agreed that proceeds meant “profits” in the lottery scheme, he agreed with the four justices in dissent that proceeds did not mean profits in every case. Among other things, he explained, “the legislative history of § 1956 makes it clear that Congress intended the term ‘proceeds’ to include gross revenues from the sale of contraband and the operation of organized crime syndicates involving such sales.” Id. at 526-27, 128 S.Ct. 2020.
Santos quickly caught the attention of Congress, which amended § 1956 in 2009 to define “proceeds” 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.” FERA § 2(f). The amendments effectively endorsed Justice Alito’s dissent and overruled any interpretation of “proceeds” based on Justice Scalia’s and Justice Stevens’ opinions. See S. Rep. Ill— 10, at 432 (2009) (explaining that the plurality decision in Santos was “erroneous” and “contrary to Congressional intent.”). [1092]*1092Under the current version of § 1956, there could be no debate that Grasso’s “referral fees” for the Yoakum Drive and Benedict Canyon Road transactions would be separately punishable as money laundering. The 2009 amendments are not retroactive, however, see FERA § 4(f), and so we must apply Santos in assessing Grasso’s merger argument.
In analyzing the three opinions produced by the Santos court, we have derived the controlling rule that “ ‘proceeds’ means ‘profits’ where viewing ‘proceeds’ as ‘receipts’ would present a ‘merger’ problem of the kind that troubled the plurality and concurrence in Santos.” See United States v. Van Alstyne, 584 F.3d 803, 814 (2009). Our subsequent case law has identified three factors we consider in determining when such a “merger problem” arises.
First, we look to whether a given transaction was a “central component” of the underlying scheme. Van Alstyne, 584 F.3d at 815. In Van Alstyne, a defendant was charged with 19 counts of mail fraud for running a Ponzi scheme and three money laundering convictions. Id. at 809. In considering the defendant’s challenge to the money laundering convictions, we struck down the two based on regular distributions to individual investors in the Ponzi scheme. Because “issuing distribution checks that supposedly represented generous returns on his victims’ investment was a central component of the ‘scheme to defraud,’ ” we held that convicting the defendant of money laundering “for the bank transfers inherent in the ‘scheme’ central to the mail fraud charges” would raise a merger problem. Id. at 815. On the other hand, we held that the money laundering conviction based on the refund of an investor’s entire investment after the scheme began to unravel did not raise such a merger problem. Id. We reasoned that the refund check was not part of the “core scheme” because the payment “undermined rather than advanced” the Ponzi scheme. Id. at 815-16; see also United States v. Bush, 626 F.3d 527, 538 (9th Cir.2010) (reasoning that transfers of proceeds from a four-year, $36 million Ponzi scheme were not central to “carrying out the scheme’s objective[s]” in part because the defendant had “operated his scheme for several years” before transferring funds to his account, and concluding that the transfers did not raise a merger problem); United States v. Phillips, 704 F.3d 754, 766 n. 11 (9th Cir.2012) (looking to whether the money laundering transaction was a “central component” of the underlying scheme); United States v. Moreland, 622 F.3d 1147, 1166 (9th Cir.2010) (same).
Second, we consider whether the inclusion of the money laundering charges leads to “ ‘a radical increase in the statutory maximum sentence’ for the underlying offense,” because five justices in Santos also expressed this concern. Bush, 626 F.3d at 538 (quoting United States v. Kratt, 579 F.3d 558, 562 (6th Cir.2009)).15 In other words, when Congress caps a defendant’s maximum sentence for the underlying offense at something radically less than the maximum sentence for money laundering, we may infer that Congress did not intend the “important limitation” on the penalty for the underlying offense to be “eviscerated” by the penalty for a money laundering conviction. Santos, 553 U.S. at 527, 128 S.Ct. 2020. By contrast, where a defendant’s underlying mail and wire-fraud convictions carried a statutory [1093]*1093maximum sentence of thirty years, Congress’s choice of penalty would not be “eviscerated” by the ten-year statutory maximum sentence for money laundering. See Bush, 626 F.3d at 538; see also Phillips, 704 F.3d at 766 n. 11 (9th Cir.2012) (holding that Santos’s concern regarding the evisceration of Congress’s choice of a low statutory maximum penalty for an underlying offense did not apply where the underlying offenses carried maximum sentences of twenty years, and the money laundering offense carried a lower maximum sentence of ten years).
Third, we generally consider the money used in co-conspirator transfers in crimes such as the sale of contraband, fraud, or bribery as constituting the “proceeds” of such crimes for purposes of the money laundering statute. In United States v. Webster, we concluded that when “a money laundering count is based on transfers among co-conspirators of money from the sale of drugs, ‘proceeds’ includes all ‘receipts’ from such sales.” 623 F.3d 901, 906 (9th Cir.2010). We acknowledged that Justice Scalia’s plurality opinion stated that the merger problem might exist for certain payments among conspirators, but concluded that this was a minority view that was not controlling. Id. Rather, we relied on the four justices in dissent, plus Justice Stevens, who agreed that “the legislative history of § 1956 makes it clear that Congress intended the term ‘proceeds’ to include gross revenues from the sale of contraband and the operation of organized crime syndicates involving such sales.” Id.
• In United States v. Wilkes, we extended Webster’s rationale beyond drug cases and into the context of fraud and bribery. 662 F.3d 524 (9th Cir.2011). In Wilkes, the ringleader of a scheme involving kickbacks to a Congressman in exchange for lucrative government contracts challenged his conviction of concealment money laundering.16 The defendant challenged his conviction on two grounds. He first argued that his use of multiple transfers to indirectly pay off the Congressman did not constitute “concealment,” id. at 545-47. Second, relying on Santos, Wilkes argued that his transfer of funds to pay off the Congressman “was merely an expense associated with the bribery, and, thus, not ‘proceeds’” under the money laundering statute. Id. at 548-49. We rejected both arguments. We first held that the government had proven concealment because “the effort to disguise the source of the money was an additional act that is separately punishable.” Id. at 547. With respect to defendant’s Santos argument that there was insufficient evidence that the bribes constituted “proceeds,” we concluded that “[ujnder Webster, [the defendant’s] money laundering count was based on a transfer to a co-conspirator of money from honest services fraud and bribery such that ‘proceeds’ would include all ‘receipts’ from the fraud.” Id. at 549. Accordingly, we rejected defendant’s argument that because the kickback payments were an inherent part of the scheme to defraud, they could not constitute proceeds for purposes of the money laundering statute, and concluded that defendant’s transfers to the [1094]*1094Congressman “constituted ‘proceeds’ of the scheme to defraud,” id.
We disagree with the dissent’s argument that Wilkes’s determination that kickback payments to the Congressman were “proceeds” is inconsistent with Van Alstyne. Dis. op. at 1100. Van Alstyne did not have occasion to address co-conspirator transfers, while Webster and Wilkes appropriately relied on five justices’ views in Santos in concluding that such transfers did not raise a merger problem. Van Alstyne ’s statement that a merger problem may be triggered when confederates in a scheme shared profits was based on a summary of Justice Scalia’s plurality opinion, 584 F.3d at 815, and does not necessarily extend beyond the situation identified by Justice Stevens, that of illegal gambling operations. Accordingly, we are bound by our precedent in Webster and Wilkes.17
We now apply these principles to Grasso’s argument that the government’s money laundering charges for the Yoakum Drive and Benedict Canyon Road transactions raise a merger problem. Under the applicable case law, Grasso’s argument turns on whether the government has provided sufficient evidence that the referral fees are “proceeds” of loan fraud and bank fraud. We first consider whether viewing “proceeds” as “receipts” in this context would raise a merger problem “of the kind that troubled the plurality and concurrence in Santos.” Van Alstyne, 584 F.3d at 814.
On their face, the loan fraud counts do' not raise such a merger problem. The loan fraud statute, § 1014, criminalizes knowingly “makfing] any false statement or report” to defraud an insured institution, while § 1956 criminalizes knowingly “conducting] a .financial transaction” involving proceeds of an unlawful activity for specified purposes. Because these statutes criminalize different types of behavior, the money laundering counts did not increase Grasso’s sentence for the same behavior underlying the loan fraud counts.
Unlike the loan fraud counts, the bank fraud count charged Grasso with the entire Abrams-Fitzgerald “scheme to defraud,” and so encompasses the Yoakum Drive and Benedict Canyon Drive transactions. But we reject Grasso’s merger argument here as well. First, Abrams’s two kickbacks to Grasso were not a “central component” of the underlying scheme. Van Alstyne, 584 F.3d at 815. Given that the Abrams and Fitzgerald scheme lasted three years and involved roughly 80 fraudulent transactions, and that Abrams and [1095]*1095Fitzgerald generally secured access to Cal Title through Grasso’s participation in their real estate transactions, the two isolated referral payments to Grasso at the end of the scheme were not “central” to the scheme’s success. See Bush, 626 F.3d at 537. The fact that the scheme operated successfully for several years before Abrams and Fitzgerald ever gave Grasso a kickback is powerful evidence of the kickbacks’ “irrelevance to the overarching fraud scheme.” Id. Moreover, Abrams’s kickbacks to Grasso drained off funds that Abrams could have re-invested in purchasing another property to extract fraudulent loan proceeds, thus hindering the scheme in the same way the returned investment in Van Alstyne hindered the defendant’s Ponzi scheme. Van Alstyne, 584 F.3d at 815-16.
Second, the inclusion of the money laundering charges did not threaten ‘“a radical increase in the statutory maximum sentence’ for the underlying offense.” Bush, 626 F.3d at 538 (quoting Kratt, 579 F.3d at 562). Here, the underlying bank fraud count has a 30-year statutory maximum, while the money laundering count has a 20-year statutory maximum. As in Bush and Phillips, the statutory cap on the penalty for bank fraud is not “eviscerated” by the maximum penalty for money laundering, and thus raises no inference regarding Congressional intent.18 Finally, we have declined to define “proceeds” to mean “profits” when a money laundering conviction is based on kickbacks and transfers to co-conspirators in schemes to defraud such as Abrams’s. See Wilkes, 662 F.3d at 547.
Because the money laundering convictions in this case do not raise a merger problem with respect to either the loan fraud or bank fraud convictions, we define “proceeds” to mean “gross receipts,” and conclude that these referral fees may be viewed as “proceeds” of the loan and bank fraud. On that basis, we reject Grasso’s argument that the government presented insufficient evidence that his money laundering charges were based on “separate and distinct” activity from the Yoakum Drive and Benedict Canyon Drive transactions.
It is undisputed that the remaining elements of promotion money laundering, under § 1956(a)(l)(A)(i), were fulfilled here. The government adduced evidence that Grasso knew that the referral fees were derived from the Benedict Canyon and Yoakum Drive transactions, and that Grasso knew these activities involved bank and loan fraud. Further, the testimony of Abrams and Matykowski sufficiently demonstrates that Grasso had the requisite intent to promote the carrying on of loan fraud, as his demands for money were intended to ensure the conspirators’ ongoing access to Cal Title. We therefore affirm Grasso’s money laundering conviction under § 1956(a)(l)(A)(i), and need not consider whether it could be alternatively upheld under § 1966(a)(l)(B)(i).
Ill
We conclude that there was sufficient evidence to support Grasso’s Convictions for conspiracy, loan fraud, bank fraud, and money laundering, and that the district [1096]*1096court therefore did not err in denying Grasso’s motion for acquittal.
AFFIRMED.