RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 25a0044p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

┐ UNITED STATES OF AMERICA, │ Plaintiff-Appellee, │ > No. 23-3109 │ v. │ │ RAYMOND A. ERKER, │ Defendant-Appellant. │ ┘

Appeal from the United States District Court for the Northern District of Ohio at Cleveland. No. 1:20-cr-00478-1—Dan A. Polster, District Judge.

Argued: October 29, 2024

Decided and Filed: March 3, 2025

Before: STRANCH, THAPAR, and MURPHY, Circuit Judges.

_________________

COUNSEL

ARGUED: Benton C. Martin, FEDERAL COMMUNITY DEFENDER OFFICE, Detroit, Michigan, for Appellant. Jason Manion, UNITED STATES ATTORNEY’S OFFICE, Cleveland, Ohio, for Appellee. ON BRIEF: Benton C. Martin, FEDERAL COMMUNITY DEFENDER OFFICE, Detroit, Michigan, for Appellant. Jason Manion, Stephanie Wojtasik, UNITED STATES ATTORNEY’S OFFICE, Cleveland, Ohio, for Appellee. _________________

OPINION _________________

THAPAR, Circuit Judge. Raymond Erker operated a Ponzi scheme that swindled over fifty people, mainly senior citizens, out of nine million dollars. A jury convicted Erker of mail fraud, wire fraud, money laundering, and making a false statement under oath. Erker appeals his No. 23-3109 United States v. Erker Page 2

money laundering conviction, raises an ineffective-assistance-of-counsel claim, and objects to various aspects of his sentence. Erker’s arguments fall short, so we affirm his sentence. But we remand so that the district court can consider one of Erker’s sentence-reduction arguments.

I.

Raymond Erker took money from investors, used the money for his own personal consumption, invested it in risky business endeavors, lied about it, and solicited more investors to cover his tracks.

First, Erker created two companies called GenSource and Provident Securities. Next, Erker and his co-conspirators solicited investors for GenSource and Provident (without disclosing that Erker owned both entities). They lured investors with promises of annuities and senior secured notes that in turn would generate safe, guaranteed rates of return for investors. But that wasn’t true. GenSource and Provident couldn’t give investors annuities—they didn’t have the required licenses. And Provident couldn’t offer investors senior secured notes because it had no property that could secure the notes.

Rather than making good on his promises, Erker misappropriated investor funds. He transferred money to his own personal bank account and other entities he owned. And he used the investors’ money to invest in risky start-ups. The scheme was elaborate. For instance, to disguise the fact that he owned and operated GenSource and Provident, Erker created office fronts and mailing addresses in Delaware and Nevada. And to assuage investor concerns, Erker set up call centers and fabricated account statements to make it look like all was well.

Then things went south. For the most part, Erker’s investments in the start-ups didn’t pay off. So he solicited more investors and used this new money to pay old investors. The old investors believed they were receiving their ordinary rates of return. This scheme worked for about five years. Eventually, however, Erker ran out of the money he had squandered and couldn’t repay his investors.

The government indicted Erker for conspiracy to commit mail and wire fraud, mail fraud, wire fraud, money laundering, and making a false statement under oath. After a four-day trial, No. 23-3109 United States v. Erker Page 3

the jury convicted Erker on all counts. The district court sentenced him to 262 months in prison and subsequently ordered restitution.

Erker raises four arguments on appeal: (1) the government failed to prove beyond a reasonable doubt that he withdrew more than $10,000 of criminally derived property, rather than his own personal funds, from the GenSource business account; (2) the district court imposed a procedurally and substantively unreasonable sentence by not addressing national sentencing disparities; (3) Amendment 821 to the Sentencing Guidelines entitles him to a shortened sentence; and (4) his attorney provided ineffective assistance of counsel by not moving to suppress Erker’s incriminating statements to a detective while in jail.

II.

Erker first challenges his money laundering convictions under 18 U.S.C. § 1957. This statute criminalizes “knowingly engag[ing]” “in a monetary transaction in criminally derived property.” 18 U.S.C. § 1957(a). That property must be “of a value greater than $10,000” and be “derived from specified unlawful activity.” Id. And “monetary transaction” means “the deposit, withdrawal, transfer, or exchange” of funds, meaning both deposits and withdrawals fall under the statute. Id. § 1957(f)(1). Erker’s claim is simple: Because he put both “clean” and “dirty” funds in the same commingled account, and the government showed only that he made withdrawals from that account, the government didn’t prove that he withdrew “criminally derived property” when he took money out. Instead, it just proved he took some money out of the account. And that’s not enough.

Though framed as a sufficiency-of-the-evidence case, this is really a question of statutory interpretation that we review de novo. See United States v. Wagner, 382 F.3d 598, 606–07 (6th Cir. 2004). What constitutes “criminally derived property”?

A.

Judges must “say what the law is.” Marbury v. Madison, 5 U.S. 137, 177 (1803). But sometimes Congress drafts statutes so ambiguous that it’s hard for us to do that. When those No. 23-3109 United States v. Erker Page 4

unclear laws threaten a defendant’s liberty, we must ask: Should we decide who goes to jail based on one judge’s interpretation of inartful drafting?

Fortunately, the answer is no. The government has the burden to prove that its interpretation of a statute is the best and true reading of the law. See Gary Lawson, Proving the Law, 86 Nw. U. L. Rev. 859, 888 (1992). Under this logic, the government must convince us that a law covers a defendant’s conduct. Another way courts apply this principle is through the substantive canon known as the rule of lenity. This rule is hundreds of years old—dating back to the sixteenth century. See Amy Coney Barrett, Substantive Canons and Faithful Agency, 90 B.U. L. Rev. 109, 128–30 (2010). In the American tradition, the rule of lenity demands that the accused be on notice of what the law proscribes. Id. at 130. And, if the legality of certain conduct is unclear, courts give defendants the benefit of the doubt. Wooden v. United States, 595 U.S. 360, 390 (2022) (Gorsuch, J., concurring).

All told, both ways of thinking yield a similar result. Courts can’t send a defendant to jail if they aren’t sure that the statute criminalizes his conduct. After all, if a court isn’t sure what a statute means, then a defendant can’t be on notice that his conduct is illegal. See Bouie v. City of Columbia, 378 U.S. 347, 350–51 (1964); Sorich v. United States, 555 U.S. 1204, 1207 (2009) (Scalia, J., dissenting from denial of certiorari). And punishing him for that conduct sits uncomfortably next to fundamental due process principles. Wooden, 595 U.S. at 389 (Gorsuch, J., concurring).

Thus, the critical question becomes whether 18 U.S.C. § 1957 clearly criminalizes Erker’s conduct.

B.

Section 1957 punishes offenders who “knowingly engage[] or attempt[] to engage in a monetary transaction in criminally derived property of a value greater than $10,000.” 18 U.S.C. § 1957(a). The whole ballgame comes down to what “criminally derived property” means. Section 1957(f)(2) defines it as “any property constituting, or derived from, proceeds obtained from a criminal offense.” Id. § 1957(f)(2). The term “proceeds,” in turn, means “any No. 23-3109 United States v. Erker Page 5

property derived from or obtained or retained, directly or indirectly, through some form of unlawful activity, including the gross receipts of such activity.” Id. § 1956(c)(9).

In some cases, it’s easy to see what would fall under § 1957’s sweep. If a drug dealer opened a bank account and put $15,000 that he made selling drugs into the account, and then withdrew $11,000 to buy a car, he’d have withdrawn more than $10,000 in criminally derived proceeds. That’s clear enough.

But a harder situation arises when the defendant mixes dirty money with clean money. In such a context, the defendant deposits “dirty” money into an account that also contains money he earned through lawful means. Amending the hypothetical slightly, if the drug dealer put $100,000 from his legitimate day job in his bank account, deposited $15,000 of drug money, and then withdrew $11,000, it’s hard to say whether his withdrawal consisted of clean money, dirty money, or both.

The withdrawal of commingled funds poses challenges because it’s not clear how the statute applies to such a situation. Consider the statute’s text, which is always the main source of meaning. See Kasten v. Saint-Gobain Performance Plastics Corp., 563 U.S. 1, 7 (2011). The statute says it covers “property constituting, or derived from, proceeds obtained from a criminal offense.” 18 U.S.C. § 1957(f)(2). And while we could make conclusory assertions that a withdrawal from a commingled account meets that standard, doing so just begs the question of whether the funds withdrawn from such an account “constitut[e]” or are “derived from” criminal proceeds.

But it isn’t clear that the circular reading is the best one. After all, “derived from” means “to take, receive, or obtain especially from a specified source.” Merriam-Webster Dictionary, available at: https://www.merriam-webster.com/dictionary/derive (emphasis added). That definition suggests that money must come from a particular source of funds—so pointing to any withdrawal from a commingled account might not be enough. Who can say whether those funds were originally the product of illicit activity or honest work? Thus, it’s not clear when a withdrawal from a commingled account violates the statute. No. 23-3109 United States v. Erker Page 6

But that lack of clarity doesn’t end our inquiry. After all, Congress doesn’t legislate against a blank canvas. Instead, Congress enacts laws against many background presumptions, and often incorporates terms of art into statutes. So we turn to that context to make sense of § 1957.

C.

While § 1957 only dates to 1986, courts have dealt with commingled funds for centuries. To be sure, problems of commingled funds have arisen most frequently in private-law cases, which aren’t a precise analogue to criminal prosecutions. But given the centuries of precedent dealing with commingled assets, we shouldn’t overlook courts’ long history of sorting through tricky questions about which funds stem from which source. These practices provide a starting point for how Congress may have thought about § 1957.

1.

One place where disputes about commingled funds are common is the law of trusts. In that context, a “trustee” is appointed to manage the money of a “beneficiary.” A trustee may wrongfully combine his personal funds with the beneficiary’s money in the trust fund, and then make withdrawals from the now-commingled account. Though not criminal, the trustee’s unlawful commingling and later withdrawals are analogous to Erker’s situation: A person withdraws funds from an account with both “clean” money (the beneficiary’s funds) and “dirty” money (the trustee’s money). It’s appropriate to analogize the trustee’s funds with dirty money because the trustee’s deposit of personal funds is a breach of trust.1 See Raymond Radigan, Defining Responsibilities When Multiple Parties Administer Trusts, 40 Est. Plan. 12, 13 (2013). In addition, such commingling doesn’t change the fact that the beneficiary’s funds are legitimately in the account.

Congress may well have considered how trust law resolves disputes about commingled funds. After all, we presume Congress is well aware of longstanding background presumptions.

1 Alternatively, a reader could consider the trustee’s funds to be “clean” and the beneficiary’s funds to be “dirty,” and our conclusions wouldn’t change. Nothing about this opinion hinges on whether the trustee’s funds are properly viewed as clean or dirty. No. 23-3109 United States v. Erker Page 7

Young v. United States, 535 U.S. 43, 49–50 (2002). But the key is that such principles inform our reasoning because they speak to what courts have always done: resolve difficult financial disputes like the one at hand.2 See United States v. Luis, 578 U.S. 5, 22–23 (2016) (plurality opinion) (noting that courts have experience with common law tracing rules in a criminal law case).

i.

At common law, the law of trusts provided four distinct ways to resolve difficult questions about commingled funds. The first came from Clayton’s Case. Devaynes v. Noble (Clayton’s Case) (1816) 35 Eng. Rep. 767, 793–94; 1 Mer 529, 608–12. There, an English court set out a “first-in, first-out” principle. Id. Under that logic, a court presumes that the first funds deposited in a mixed account are the first funds withdrawn. Id. at 793.

A hypothetical is helpful. Return to our drug dealer, who works a steady day job that brings him $100,000 in clean income, which he places in a savings account by making four deposits of $25,000. He then deposits his $15,000 of drug money. Then, he makes a withdrawal of $25,000. The court would presume that the initial withdrawal of $25,000 came exclusively from his first deposit of $25,000. The same would be true for his next three withdrawals. But if he deposited his drug money first, he’d have a problem—the first withdrawal would stem from his drug money.

The second rule crystallized in an 1879 dispute. See Knatchbull v. Hallett (In re Hallett’s Estate) (1879) 13 Ch D 696. There, the court established an early iteration of what would become known as the “lowest intermediate balance rule.” In brief, the court assumed that the beneficiary’s funds remained in the trust account when a trustee withdrew money from a commingled account. The trustee’s own funds would have to be completely withdrawn before a court said the beneficiary’s funds were ever touched. In other words, the court would say that the beneficiary’s funds were withdrawn only when it was mathematically impossible for them to still be in the account.

2 Of course, trust law is not a perfect match with criminal law. However, given that any robust interpretation of § 1957’s “derived from” language requires examining background accounting principles, it makes sense to turn to private law, which is where accounting principles developed. No. 23-3109 United States v. Erker Page 8

Consider the following hypothetical, which illustrates how the lowest intermediate balance rule would function in the criminal context. Again, the drug dealer has a day job that pays him $100,000 of clean income, all of which he has deposited. Next, he deposits $15,000 of drug money into that same account. Then, he withdraws $11,000. Because the dealer’s withdrawal was within the balance of legitimate funds, no court operating under this framework would say he withdrew illicit funds.

Next is Hertslet v. Oatway (In re Oatway) (1903) 2 Ch 356. There, another court adopted a third rule that courts now call the “proceeds-first” or “proceeds-out-first” approach. The Oatway court explained that when a trust contains mixed funds, and a trustee withdraws money, one of two things can happen. If the withdrawn money wasn’t recoverable, meaning that the trustee spent the funds on items that couldn’t be sold to get the funds back or that the court couldn’t figure out where the money went, the court would say the withdrawn funds belong to the trustee. Id. at 360. But if the court could identify the removed funds, then they came from the trust and the beneficiary could recover them. Id. In other words, if the trustee lost the funds, the court would say the funds were the trustee’s. If the funds weren’t lost, the court would say the money came from the beneficiary’s funds.

Returning to our hypothetical, if a drug dealer deposited $100,000 in clean funds and then deposited $15,000 in dirty money in the same account, the proceeds-first rule would proceed as follows. If the drug dealer purchased an item for $11,000, and the item’s value couldn’t be recovered, a court would consider the withdrawn funds to be dirty money. But if the item’s value could be recovered, then the court would presume that the withdrawn funds were clean.3

Fourth, some courts adopted a proportionate or “pro rata” tracing analysis. In the case In re British Red Cross Balkan Fund, a British court explained that donors to a fund were entitled to the return of a proportional share of the fund upon a request for a withdrawal. 2 Ch 419 (1914).

3 The Oatway rule functions to protect beneficiaries. Thus, a reader might question why we’d apply it in the criminal context, where there’s no beneficiary to protect. But there are several reasons to consider the Oatway principle here. First, the Oatway case is a significant development in how courts considered commingled funds. Second, even if we accept that there’s a tension between rules used in private and criminal law contexts, the Oatway court’s turn to accounting principles cuts against arguments that it’s simply unreasonable for parties to prove which funds stemmed from which source. No. 23-3109 United States v. Erker Page 9

Thus, where a trustee commingled funds, parties could seek a proportion of the remaining balance.

The proportional rule would work like this in the criminal context: Assume our drug dealer had $100,000 of clean money and $15,000 of dirty money in one account. Using some basic math, a court would conclude that 13% of his account came from dirty money. If he withdrew $11,000, a court would presume that 13% of that withdrawal was also dirty. And 13% of 11,000 is 1,430. So a court would say that $1,430 of the withdrawal was from the dirty money, while the remaining balance came from clean funds.

All told, our survey of English trust law reveals four options when dealing with commingled funds in trusts: (1) a first-in, first-out principle; (2) a “lowest intermediate balance” approach; (3) a “proceeds-first” rule; and (4) a proportional methodology.

ii.

But what matters is how the common law treated commingled funds in 1986, when Congress enacted the statute at issue in this case. Pasquantino v. United States, 544 U.S. 349, 360–61 (2005). And not all the British cases survived their journey across the Atlantic. Thus, to derive the American rules for commingled funds, consider the following developments.

The first-in, first-out rule from Clayton’s Case met with disfavor in American courts. See, e.g., In re Walter J. Schmidt & Co., 298 F. 314, 316 (S.D.N.Y. 1923) (Learned Hand, J.). Indeed, one commentator described the Clayton’s Case approach as a “blind mechanical rule of thumb with a stern disregard of the justice of any particular result.” Note, Application of the Rule in Clayton’s Case to the Distribution of Property Held under Constructive and Resulting Trusts, 28 Harv. L. Rev. 193, 194 (1914). Thus, it comes as no surprise that the Second Restatement of Trusts rejected the first-in, first-out rule. See Restatement (Second) of Trusts § 202 cmt. i (Am. Law Inst. 1959); Ruddle v. Moore, 411 F.2d 718, 718–19 (D.C. Cir. 1969) (per curiam). Instead, the Restatement asserts that if a trustee withdraws funds from a commingled fund, “the beneficiary is entitled to a proportionate share” both “in the part which remains and in the part which is withdrawn” from the account. Restatement (Second) of Trusts § 202 cmt. i. The Restatement also favored the lowest intermediate balance principle. See id. Thus, under the No. 23-3109 United States v. Erker Page 10

Second Restatement, courts have eliminated the Clayton’s Case approach and apply a proportional or lowest intermediate balance framework.4

But the story doesn’t end there. Courts declined to follow the Restatement’s proportionality approach. Instead, they either used the lowest intermediate balance test or required tracing commingled funds using a proceeds-out-first rule. See Brown & Williamson Tobacco Corp. v. First Nat. Bank of Blue Island, 504 F.2d 998, 1002–04 (7th Cir. 1974); Sean Michael Welsh, Tracing Commingled Funds in Asset Forfeiture, 88 Miss. L. J. 179, 234–35 (2019). This approach is best distilled in the Third Restatement of Restitution and Unjust Enrichment, which explains that courts typically trace to the lowest intermediate balance or to property acquired from a commingled account, which we might call a modified proceeds-first method. See Restatement (Third) of Restitution and Unjust Enrichment § 59 cmt. d (Am. Law Inst. 2011).

And trust law doesn’t stand alone. Article 9 of the Uniform Commercial Code also addresses commingled funds. Section 9-315 describes a secured party’s right to proceeds after a debtor sells secured collateral—in other words, proceeds derived from a sale. It explains that parties must trace the funds. One comment to § 9-315(b)(2) explains that parties should apply equitable principles to trace funds, noting that one such principle is the “lowest intermediate balance rule.” § 9-315 cmt. 3 (quotation omitted).

4 To be sure, one might read the various restatements to support most circuits’ views on § 1957. For example, there’s a reading of the Second Restatement that supports a victim’s right to recover under any accounting method. See Restatement (Second) of Trusts § 202 cmt. i (Am. Law Inst. 1959). That view could also draw support from the First Restatement of Restitution, which allowed a claimant to obtain an equitable lien for withdrawn funds. Restatement (First) of Restitution § 211 (1937). However, some language in the Second Restatement of Trusts § 202 also cuts against that broad view, as does the Third Restatement of Restitution and Unjust Enrichment. Start with trust law. Section 202 of the Second Restatement of Trusts explains that it is “immaterial in what order the deposits were made, whether the trust funds were first deposited or the trustee’s individual funds, since there is no inference that the money first deposited is the money first withdrawn.” Restatement (Second) of Trusts § 202 cmt. i. And the Third Restatement of Restitution and Unjust Enrichment looks to both intermediate balance and a modified proceeds-first approach. Restatement (Third) of Restitution and Unjust Enrichment § 59 cmt. d (Am. Law Inst. 2011). On net, there is considerable tension even within and among restatements about how to resolve disputes about commingled funds. No. 23-3109 United States v. Erker Page 11

In sum, modern private law doctrines have rejected a first-in-first-out approach and proportional method. What’s left standing? The lowest intermediate balance test and the proceeds-first approach.

2.

Our review of private law suggests that the government must do more than just point to a withdrawal from a commingled account to prove the existence of “criminally derived proceeds.” But which rule should apply? That depends on the background presumption against which Congress legislated. And here, we’ve had no briefing on that topic. Rather than put forward our own thoughts, we note only that there is significant debate about how the statute should function and that none of the existing caselaw correctly frames the issue. Indeed, no sister circuit has conducted an in-depth textual or historical analysis of the provision in question.

Precedent from one circuit supports an intermediate balance approach—although it doesn’t reference that rule by name. The Fifth Circuit requires prosecutors to prove that “the aggregate amount withdrawn from an account containing commingled funds exceeds the clean funds.” United States v. Davis, 226 F.3d 346, 357 (5th Cir. 2000). Only then can “individual withdrawals . . . be said to be of tainted money, even if a particular withdrawal was less than the amount of clean money in the account.” Id. In subsequent cases, the Fifth Circuit explained that without such mathematical certainty, “[n]o reasonable juror could conclude that these money laundering convictions were warranted beyond a reasonable doubt.” United States v. Loe, 248 F.3d 449, 467 (5th Cir. 2001).

But the Fifth Circuit declined to provide a robust rationale for why it adopted this approach. That court merely concluded that a rigorous tracing requirement would “defeat the purposes of the statute,” which it left unstated. Davis, 226 F.3d at 357. In other words, the court declined to base its interpretation on the text of § 1957 or applicable background principles.

Other circuits have come to different conclusions. The majority view is that § 1957 doesn’t require any sort of tracing. Under this framework, courts assume that placing any dirty money in an account renders the whole account dirty. Indeed, the Second, Third, and Eleventh Circuits have made this rule explicit. See United States v. Sokolow, 91 F.3d 396, 409 (3d Cir. No. 23-3109 United States v. Erker Page 12

1996); United States v. Silver, 864 F.3d 102, 115 (2d Cir. 2017); United States v. Ward, 197 F.3d 1076, 1083 (11th Cir. 1999).5 And the First, Fourth, Seventh, Eighth, and Tenth Circuits have at least hinted they don’t require tracing, either, although intra-circuit doctrine is hardly clear and sometimes contradicts itself without comment.6 See Audrey Spensley, Note, Untangling Laundered Funds: The Tracing Requirement under 18 U.S.C. § 1957, 75 Stan. L. Rev. 1157, 1174–82 (2023). On balance, though, the majority rule is that the government doesn’t have to trace funds at all. Instead, these circuits merely point to a withdrawal from an account that contains commingled funds.

But there are serious questions surrounding this approach. For one, it arguably departs from courts’ traditional approaches to commingled funds. We have found no doctrine in private law which assumes that any withdrawal comes from any deposit—even the long-rejected Clayton’s Case required some accounting principle. The government contends that the “dirty- money-out-first” presumption is an accounting principle. But the government doesn’t explain whether or how Congress incorporated this specific rule into the statute. Moreover, it places courts in the perverse position of making it easier to take a person’s liberty than to seize a debtor’s property.

The no-tracing approach is also an awkward fit for money laundering. Consider a hypothetical where a defendant has $20 million in clean funds and then deposits $10,000 of dirty

5 While the Eleventh Circuit hasn’t explicitly ruled on how much tracing § 1957 requires, it cited a Fourth Circuit case, United States v. Moore, to support its argument that § 1957’s sister statute, § 1956, doesn’t require tracing. United States v. Ward, 197 F.3d 1076, 1083 (11th Cir. 1999) (citing United States v. Moore, 27 F.3d 969, 976–77 (4th Cir. 1994)). In Moore, the Fourth Circuit noted that “[a] requirement that the government trace each dollar of the transaction to the criminal, as opposed to the non-criminal activity, would allow individuals effectively to defeat prosecution for money laundering by simply commingling legitimate funds with criminal proceeds.” 27 F.3d at 977. The Eleventh Circuit’s citation to Moore suggests that the Circuit believes that the same tracing analysis applies to both § 1956 and § 1957. What’s more, at least one district court in the Eleventh Circuit has adopted a no-tracing rule. See United States v. Long, No. 1:08-CR-043-CC, 2009 WL 10675289, at *2 (N.D. Ga. Sept. 10, 2009) (explaining “the law in the Eleventh Circuit does not require that the Government trace the funds in a commingled account”). 6 United States v. Rivera-Izquierdo, 850 F.3d 38, 48 (1st Cir. 2017) (explaining that the record “overwhelmingly indicates” funds were derived from fraud); United States v. Haddad, 462 F.3d 783, 786–89, 792 (7th Cir. 2006) (“[T]he vast majority of funds transferred . . . were not supported by evidence of legitimate . . . sales.”); United States v. Pennington, 168 F.3d 1060, 1066 (8th Cir. 1999) (stating that the “government need not trace funds to prove a violation of § 1957”); United States v. Mooney, 401 F.3d 940, 946 (8th Cir.) (per curiam), aff’d en banc, 425 F.3d 1093 (8th Cir. 2005) (“[T]he government need not trace each dollar to a criminal source to prove a violation.”); United States v. Battles, 745 F.3d 436, 456 (10th Cir. 2014). No. 23-3109 United States v. Erker Page 13

funds. If he withdraws $11,000, can we say that his withdrawal is “derived from” the dirty deposit? Would the “specified source” of that withdrawal be the meager deposit of drug money, which accounts for less than one percent of the balance? And, at the very least, this reading of § 1957 is no more plausible than a defendant-friendly reading. So either under a “proving law” framework or the rule of lenity, these conflicting readings present problems for the government.

What’s more, the opinions setting out a no-tracing rule didn’t perform an in-depth textual or historical analysis. See, e.g., Moore, 27 F.3d at 976–77; Silver, 864 F.3d at 115. Instead, they relied on a related statute—Section 1956. Moore, 27 F.3d at 976–77. Congress enacted § 1956 and § 1957 together along with several shared definition provisions, so these courts interpreted the two statutes in tandem. And they didn’t resolve what “derived from” meant because they assumed that Congress simply intended to capture all money laundering. Cf. United States v. Jackson, 935 F.2d 832, 840 (7th Cir. 1991). This court, for instance, “refuse[d] to read [§ 1956] in a manner that would reward the more creative money-launderer by allowing him to escape liability altogether.” United States v. Bencs, 28 F.3d 555, 562 (6th Cir. 1994). However, fears that a “creative money-launderer” could escape the reach of the money laundering statutes shouldn’t replace an independent analysis of § 1957. That analysis should be a “holistic endeavor,” which requires looking to text, context, and history. United Sav. Assn. of Tex v. Timbers of Inwood Forest Associates, Ltd., 484 U. S. 365, 371 (1988).

All told, there’s significant debate about what § 1957 means. And the existing debate among our circuits is unsatisfying, since it’s divorced from the accounting presumptions that courts use to resolve complex questions about commingled assets.

3.

Despite that significant ambiguity, however, we find it easy to reject the Ninth Circuit’s approach—or, at least, Erker’s reading of that approach, which would require a strict tracing rule.7 That court’s precedent could be read to adopt a blanket presumption that the government

7 Adding to the confusion, some district courts have read the Ninth Circuit’s caselaw as adopting a version of the lowest intermediate balance approach in other contexts. “While the Ninth Circuit Court of Appeals has never specifically identified the [lowest intermediate balance approach] as a method to trace tainted funds, the court has endorsed the accounting principles underlying the [framework] in money laundering cases,” including § 1957 cases. No. 23-3109 United States v. Erker Page 14

must trace every charged transaction to criminally derived proceeds. United States v. Rutgard, 116 F.3d 1270, 1292–93 (9th Cir. 1997). While one reading of Rutgard suggests a “clean funds out first” presumption, district courts have interpreted the case to require dollar-for-dollar tracing. For example, in United States v. Yagman, a defendant deposited $150,000 in illicit proceeds into an account containing about $17,000 in clean money. 502 F. Supp. 2d 1084, 1087–93 (C.D. Cal. 2007). The defendant wrote checks for $50,000 and $20,000. Id. at 1088– 89. The government claimed that, even assuming that the first $50,000 check included the $17,000 in clean funds, the second $20,000 check must have come from the remaining dirty funds—thereby allowing the government to charge the defendant with a § 1957 violation for writing the $20,000 check. But the district court disagreed, finding that “no rational fact-finder could determine beyond mere speculation” whether the second check “contained tainted or untainted funds.” Id. at 1089. In other words, the government had to show that particular dollars made up each check.

But the Ninth Circuit’s approach—to the extent it adopts this strict tracing rule—is a bridge too far. For one, that rule is not rooted in the statutory text. After all, the statute doesn’t say that every dollar must be “criminally derived proceeds.” Instead, the government must only show that more than $10,000 were for a given monetary transaction. So, while we might not be entirely sure what § 1957 means, we can say with certainty that it does not require strict tracing.

D.

We need not ultimately decide which background accounting presumption controls because Erker’s conduct stacks up poorly under either of our potentially acceptable approaches. Here, he challenges six money laundering convictions, which were Counts 6 through 11 at trial. And in each charge, the government showed it was mathematically impossible for any of the six withdrawals to include less than $10,000 of dirty money. In each case, Erker depleted his bank account, received infusions of client funds, and then withdrew those client funds to pay his bills. All told, regardless of whether we apply a proceeds-first or lowest-intermediate-balance rule, Erker loses.

United States v. Haleamau, 887 F. Supp. 2d 1051, 1057 (D. Haw. 2012) (citing United States v. Hanley, 190 F.3d 1017, 1024–25 (9th Cir. 1999); Rutgard, 116 F.3d at 1293). No. 23-3109 United States v. Erker Page 15

Start with Count 6. The government showed that Erker’s account had a balance of $143.70 in clean funds, to which Erker added $152,149.62 of dirty funds and then withdrew $17,000. At no point did Erker have sufficient funds to cover this withdrawal—the lowest intermediate balance of clean funds was just $143.70. So Erker couldn’t have withdrawn clean funds.

Next, look at Counts 7 and 8. Erker ran down his balance to $6,299.02 of clean funds. He then obtained an infusion of $107,636.89 from a client, meaning dirty funds. Next, he withdrew $25,145 and $40,000. Thus, regardless of what rule applies, Erker didn’t have enough clean funds to make these withdrawals.

Now consider Counts 9 and 10. Erker had $539.69 in his bank account. He then received $103,249.59 from a new client, who later sent another $14,994.85. Erker then withdrew $20,000 and then an additional $21,547.18. Because there was no more than $539.69 in clean funds, Erker must’ve taken more than $10,000 in criminally derived proceeds from the account.

Finally, consider Count 11. That Count involved Erker’s withdrawal of $14,482.09 from the same bank account as in Counts 9 and 10—and Erker hadn’t deposited any more clean funds in between. Thus, it was impossible for the charged withdrawals to have come from clean money.

All told, under any acceptable framework—including most circuits’ presumption of “dirty funds out first”—Erker’s actions fall within the sweep of 18 U.S.C. § 1957.

In response, Erker argues that the government’s trial evidence could not support his conviction here. His argument boils down to the fact that the government relied on summaries of bank statements, which he alleges omitted relevant deposits. Thus, because the summaries were incomplete, they weren’t enough to sustain his conviction.

But there are at least two problems with that argument. First, and most notably, the government introduced the full bank statements in question. So the jury had all the required information. Second, Erker failed to object to the government’s introduction of the summary. So the jury had evidence from which it could conclude that Erker’s withdrawals were from ill- No. 23-3109 United States v. Erker Page 16

gotten gains. And that’s a problem for Erker—we must take the jury’s determination at face value. Here, they determined that Erker’s withdrawals were made up of criminally derived proceeds. That’s enough to sustain his conviction.

III.

Next, Erker brings procedural and substantive challenges to his within-Guidelines sentence of 262 months. He argues his sentence was procedurally unreasonable because the district court didn’t address unwarranted sentencing disparities and didn’t justify Erker’s longer- than-average sentence. And he claims his sentence was substantively unreasonable because the district court gave insufficient weight to the disparities his sentence allegedly created.

A.

We usually review procedural reasonableness for abuse of discretion. See United States v. Potts, 947 F.3d 357, 364 (6th Cir. 2020). But here, Erker didn’t object to the procedural reasonableness of his sentence. Instead, he objected “to the Court’s sentence in total.” R. 144, Pg. ID 1274. And the district court clarified that Erker was simply preserving an objection to his sentence as unreasonably high. Id. at 1274–75. That is a substantive reasonableness challenge, not a procedural one. See United States v. Rayyan, 885 F.3d 436, 442 (6th Cir. 2018). Indeed, at sentencing Erker never brought up national sentencing disparities. Thus, we review Erker’s procedural reasonableness objection for plain error.

Erker contends that the district court failed to adequately address the “need to avoid unwarranted sentence disparities” by ignoring national sentencing statistics. 18 U.S.C. § 3553(a)(6). But we have “expressly reject[ed]” a requirement that district courts consider national sentencing statistics. United States v. Hymes, 19 F.4th 928, 936 (6th Cir. 2021). Indeed, calculating the proper Guidelines range necessarily means that the district court considered national sentencing disparities. Id. at 935. After all, “one of the purposes of the Guidelines is to maintain national uniformity in sentences.” United States v. Simmons, 501 F.3d 620, 626 (6th Cir. 2007). What’s more, the final presentence investigation report, which the court adopted without change, discussed the average sentences of similarly situated offenders. No. 23-3109 United States v. Erker Page 17

Erker’s efforts to avoid Hymes’s precedential force are unconvincing. He argues that we should decline to follow Hymes because the Sentencing Commission data is now included in presentence reports and presented at sentencing. Thus, it’s less burdensome for district courts to address national sentencing statistics. But Hymes didn’t express a view about the burdens of consulting this data. Instead, Hymes rejected a requirement to consult sentencing statistics as a “loose approach” that would “elevate the Commission’s statistical data over the text of the Guidelines themselves.” 19 F.4th at 936.

Erker also places undue weight on United States v. Perez-Rodriguez. 960 F.3d 748 (6th Cir. 2020). There, our court reversed a sentence for being substantively unreasonable, in part because the district court ignored unwarranted national disparities. In doing so, we noted that sentencing data “should serve as a starting point” to avoid disparities. Id. at 756–57 (cleaned up). But that case dealt with a substantial upward variance from the Guidelines range. Here, by contrast, Erker received a sentence at the very bottom of his uncontested Guidelines range. And Hymes has since clarified that district courts don’t need to consider sentencing data at all. See 19 F.4th at 931. Thus, the district court didn’t err in failing to raise sentencing statistics.

B.

Erker received the shortest possible sentence within his Guidelines range. He nonetheless argues that his sentence is substantively unreasonable. See Rayyan, 885 F.3d at 442. We disagree.

For this claim, Erker rehashes his arguments from his procedural challenge, contending that the district court gave insufficient weight to the disparities its sentence created. But that’s an argument that the district court should have given him a lighter sentence—and thus should have varied downward from the Guidelines. And that’s usually not an appealable claim, unless the district court believed it lacked the authority to vary downward from the Guidelines. See United States v. Prince, 214 F.3d 740, 766 (6th Cir. 2000). That’s a nonstarter here because the district court expressly acknowledged that the Guidelines were advisory.

The record also supplies ample evidence of the reasonableness of Erker’s 22-year sentence. After hearing from victims of Erker’s fraudulent scheme, the district court described No. 23-3109 United States v. Erker Page 18

his conduct as “unimaginable and incomprehensible.” R. 144, Pg. ID 1269. Erker took the money of elderly, vulnerable people. In doing so, Erker betrayed the trust of people he knew. His sentence is proportionate to the severity of his offenses. We will not disturb the district court’s substantively reasonable sentence.

C.

Erker also asks us to vacate his sentence and remand so he can receive a sentence reduction under Amendment 821 of the Sentencing Guidelines. Amendment 821 eliminated “status points” for being on probation or parole at the time of the charged offense. See U.S.S.G. § 4A1.1. And U.S.S.G. § 1B1.10 permits retroactive modifications of sentences when Amendment 821 would lower a defendant’s Guidelines range. See U.S.S.G. § 1B1.10(d) (including Amendment 821 as a covered amendment). Erker argues that without his status points, he would have had a lower criminal history category and thus a lower Guidelines range.

But there’s no need to vacate Erker’s sentence. A retroactive amendment to the Guidelines doesn’t make Erker’s sentence unlawful. See United States v. Ursery, 109 F.3d 1129, 1137 (6th Cir. 1997). We instead affirm his sentence and remand so that the district court can consider Erker’s sentence-reduction argument. See id.; United States v. Ralston, 110 F.4th 909, 924 (6th Cir. 2024). The district court need not reduce Erker’s sentence but may do so after considering the § 3553(a) factors. See 18 U.S.C. § 3582(c)(2).

IV.

Finally, Erker brings an ineffective-assistance-of-counsel claim because his lawyer failed to object to his jailhouse statements at trial. But as Erker acknowledged, we usually don’t address the merits of ineffective-assistance claims on direct appeal. See United States v. Jackson, 181 F.3d 740, 747 (6th Cir. 1999). Erker hasn’t offered any reason “to justify departing from that practice here.” United States v. Hunter, 558 F.3d 495, 508 (6th Cir. 2009).

* * *

All in all, Erker’s claims fail. His arguments about § 1957 don’t hold water because, applying any rule, he didn’t have enough clean funds to make his withdrawals. And we reject No. 23-3109 United States v. Erker Page 19

Erker’s sufficiency-of-the-evidence challenge to his § 1957 conviction, his sentencing objections, and his ineffective-assistance-of-counsel claim. We affirm Erker’s sentence and remand for consideration of his eligibility for a sentence reduction under Amendment 821.