Concurrence by Judge WALLACE.
*1298OPINION
FISHER, Circuit Judge:We must decide when mailings sent to avoid detection or responsibility for a fraudulent scheme are sent “for the purpose of executing such scheme.” 18 U.S.C. § 1341. We hold that such mailings fall within the mail fraud statute when they are sent prior to the scheme’s completion and that, to determine when a scheme is completed, we look to the scope of the scheme as devised by the perpetrator. Because a reasonable jury could have found that defendant Thomas Tanke’s September 16, 2004 letter was sent before the completion of the embezzlement scheme he devised, we affirm his mail fraud conviction on count 2 of the indictment. We also hold that the district court properly applied sentencing enhancements under United States Sentencing Guidelines Manual § 2Bl.l(b)(9)(B) and (b)(10)(C).
I. BACKGROUND1
A. Tanke’s Embezzlement
Rafael Martin and his family operated two construction businesses, Azteca Construction Company and Construction Equipment Rental and Service (CERS). Defendant Thomas Tanke worked for Az-teca between 1999 and 2004, serving as vice president of operations between 2000 and July 2004, with authority to approve or issue checks from Azteca accounts. He was not employed by CERS and had no authority to receive CERS income or pay CERS’s obligations. Tanke also maintained his own consulting business, Cedar Creek Associates.
Over a 20-month period, Tanke embezzled more than $192,000 from Azteca and CERS. From November 2002 through February 2004, Tanke caused the issuance of 21 Azteca checks, totaling $74,762.82, for his personal expenses. The Azteca checks, most signed by Tanke, were paid to his creditors or businesses he patronized, including General Motors Acceptance Corporation (for an auto lease), Audi Financial Services (financing for an Audi A4 Quattro), Capital One Services (for a credit card), Household Credit Services and HSBC Card Services, Inc. (for a credit card), Providian National Bank (for a credit card), Onyx Acceptance Corporation (financing for a BMW), Steve Larsen’s Wheel Works (a retail bicycle shop) and Kenny G. and Company (a jeweler). Tanke was never authorized to use Azteca checks to pay personal obligations or to pay these specific creditors or businesses.
Tanke falsified records to conceal these payments. He used at least six false invoices to make it appear that the checks were issued for legitimate business expenses. For example, a false invoice from “Onyx Corporation,” ostensibly for “[pegging of lines and testing of all pipe,” was actually a $10,027.92 payment for a BMW Z4. Tanke also falsified carbon copies of checks in Azteca’s check register on at least 10 occasions, also to conceal the true nature of the payments. For example, a cheek made payable to “Providian” (to pay Tanke’s personal Providian Visa card) contained no memo notation, whereas the carbon copy falsely stated that the check was paid to “Providian Products” and referenced “02420-0046-M,” an existing Azteca job number and cost code.
*1299Between February and July 2004, Tanke also diverted seven checks payable to CERS, Martin and Azteca, totaling $117,486.25, into his Cedar Creek bank account. The four checks payable to Martin and Azteca had a handwritten endorsement making the check payable to Cedar Creek, a forged signature and a printed endorsement making the check payable to Cedar Creek. The three checks payable to CERS had a printed endorsement making them payable to Cedar Creek.
B. Cedar Creek Cover Story
Tanke left Azteca in July 2004. It appears that Martin at least suspected Tanke’s embezzlement a short time later. In an August 5, 2004 email, Martin asked Tanke about a missing April 2004 check, in the amount of $39,380, from Southern Quality Trucks to CERS. Tanke responded by email that he did “not have any check from Southern Quality Trucks” and that all checks he had received had been turned over to accounting or used to pay outstanding vendors and debts. In fact, Tanke had long since deposited the check into his Cedar Creek account.
In an August 6 email, Martin told Tanke that he knew the check had been deposited into the Cedar Creek account. In an email response that day, Tanke acknowledged that he had diverted the money into his account, but told Martin that this and similar diversions were legitimate transactions to compensate him for consulting work Cedar Creek had allegedly performed for Azteca and CERS. Tanke’s email attached what he claimed were three “very old invoices” from Cedar Creek and asserted that his actions had “ensured that Cedar Creek was paid for these long overdue invoices.” In addition to casting the diversions as legitimate payments to Cedar Creek, Tanke’s email also used thinly veiled threats to discourage Martin from reporting the matter to authorities. Tanke wrote that this was “purely a ‘business issue’ ” and that he “hope[d Martin would] not try to expand it to anything else.” Tanke warned that, if Martin pursued the matter further, he would report Martin to federal and state agencies, writing: “You do not want to see this happen, as it could involve you personally in criminal or civil action that could put you in prison or forfeit all of your personal holdings.”
On August 17, Tanke emailed Martin that he should have received invoices by mail showing that over $98,000 was due to Cedar Creek for consulting fees. Tanke noted that Martin had filed documents leading to a hold on the Cedar Creek account and that, if Martin did not lift that hold, he would report Martin’s “numerous frauds and false claims” to authorities. Tanke concluded the email by writing: “I just want what has been and is due to me and we can part friends. Please do not force me to take action that will cause you harm.”
Martin replied by email on August 18 that the invoices were “not real,” that Tanke was a salaried employee who received wages and that Azteca did not owe Tanke or Cedar Creek any consulting fees. Martin added, “[i]n response to your threats, if you ha[d] knowledge of any wrongdoing by Azteca ..., you should have reported it.”
On September 16, Tanke caused a letter to be sent from Cedar Creek to Martin by U.S. mail. This mailing, charged as count 2 of the indictment, stated that Azteca and CERS had previously paid some invoices due to Cedar Creek (again, an apparent reference to Tanke’s diversion of checks *1300into his Cedar Creek account) and alleged that Azteca had provided “false information” to Cedar Creek’s bank, which resulted in reversal of these payments. The letter enclosed an invoice from Cedar Creek that requested payment for the previous invoices as well as interest, totaling $159,990.95.
In a letter dated September 23, Martin rejected the new invoice and stated that Azteca had never paid any of the previous fictitious invoices either. Martin later testified at trial that Cedar Creek did not perform any of the claimed services reflected on the September 16 Cedar Creek invoice. He also testified that neither Tanke nor Cedar Creek had ever consulted for CERS.
C. Criminal Proceedings
In 2009, a grand jury indicted Tanke on five counts of bank fraud, for violation of 18 U.S.C. § 1344, and two counts of mail fraud, for violation of 18 U.S.C. § 1341. The first mail fraud count, charged as count 1 of the indictment, was for a July 22, 2004 check from Industrial Tractor Co. to CERS for approximately $33,000, delivered via Federal Express from South Carolina to California. Tanke does not challenge his conviction on count 1. The second mail fraud count, charged as count 2 of the indictment, was for the September 16, 2004 mailing.
A jury found Tanke guilty on all counts, and the district court sentenced him to 70 months’ imprisonment, 60 months’ supervised release and $243,403.96 in restitution. The Sentencing Guidelines range of 70-87 months reflected a 2-level enhancement for use of sophisticated means and another 2-level enhancement for misrepresentation or fraud during the course of a bankruptcy proceeding.
II. DISCUSSION
Tanke raises five issues on appeal. He contends that (1) there was insufficient evidence to support his mail fraud conviction on count 2 of the indictment; (2) the district court erred by imposing a 2-level enhancement for making a misrepresentation during a bankruptcy proceeding; (3) the district court erred by imposing a 2-level enhancement for sophisticated means; (4) the district court erred by including $44,715.21 in restitution for credit card charges that were not part of the offenses of conviction; and (5) the judgment should be amended to show that interest on the restitution award was waived. The government concedes error on the fourth and fifth issues but urges us to affirm on the first three issues, which we address in turn. We have jurisdiction under 28 U.S.C. § 1291.
A. Mail Fraud
Tanke argues that the evidence is insufficient to sustain his conviction on count 2. He argues that the September 16, 2004 letter was not “for the purpose of executing” his fraudulent scheme, as 18 U.S.C. § 1341 requires, because it was mailed after he had received all of the money from the embezzlement scheme, Martin had uncovered the fraud and the scheme had been completed. We review claims of insufficient evidence under Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979). Such a challenge can succeed only if, viewing the evidence in the light most favorable to the prosecution, no rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt. See Flyer, 633 F.3d at 917.
1.
When do mailings sent to facilitate concealment of a fraudulent scheme — to *1301avoid detection, prosecution or conviction — fall within the federal mail fraud statute?2 The statute makes it unlawful to devise a scheme to defraud and then to use the mails for the purpose of executing such scheme:
Whoever, having devised or intending to devise any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises, or to sell, dispose of, loan, exchange, alter, give away, distribute, supply, or furnish or procure for unlawful use any counterfeit or spurious coin, obligation, security, or other article, or anything represented to be or intimated or held out to be such counterfeit or spurious article, for the purpose of executing such scheme or artifice or attempting so to do, places in any post office or authorized depository for mail matter, any matter or thing whatever to be sent or delivered by the Postal Service, or deposits or causes to be deposited any matter or thing whatever to be sent or delivered by any private or commercial interstate carrier, or takes or receives therefrom, any such matter or thing, or knowingly causes to be delivered by mail or such carrier according to the direction thereon, or at the place at which it is directed to be delivered by the person to whom it is addressed, any such matter or thing, shall be fined under this title or imprisoned not more than 20 years, or both.
18 U.S.C. § 1341 (emphasis added). Accordingly, we have held that mail fraud includes two elements: “first, the government must prove that a defendant devised or intended to devise a scheme to defraud a victim of his money or property; second, it must prove that in executing the scheme, the defendant made use of or caused the use of the mails.” Lo, 231 F.3d at 475.
In deciding when mailings designed to avoid detection or responsibility for a fraudulent scheme fall within the mail fraud statute, we begin with the general principle that the mailing must be sent “prior to the scheme’s completion.” United States v. Lane, 474 U.S. 438, 453, 106 S.Ct. 725, 88 L.Ed.2d 814 (1986) (emphasis in original). But when is a scheme complete?
A partial answer is that it plainly does not end before the perpetrator has obtained money or property from the victims. Thus, before proceeds of the fraud have been obtained, a mailing need only in some way further the scheme, as mailings designed to avoid detection or responsibility clearly do.
Once proceeds of the fraud have been obtained, it is harder to say when a scheme ends. The only easy case is that of an ongoing scheme, in which the perpetrator has received some of the proceeds of the fraud but expects to receive additional proceeds of the fraud in the future. It is well settled that mailings sent in the midst of such a scheme, including those designed to avoid detection or responsibility, fall within the statute. See Schmuck v. United States, 489 U.S. 705, 711-12, 714, 109 S.Ct. *13021443, 103 L.Ed.2d 734 (1989); United States v. Jinian, 725 F.3d 954, 962-63 (9th Cir.2013); Sun Sav. & Loan Ass’n v. Dierdorff, 825 F.2d 187, 196 (9th Cir.1987); United States v. Price, 623 F.2d 587, 590 & n. 2, 593 (9th Cir.1980), overruled on other grounds by United States v. De Bright, 730 F.2d 1255 (9th Cir.1984).
The more difficult cases are those in which all the intended proceeds of the scheme have been obtained, but the perpetrator uses the mails to avoid detection or responsibility for the scheme. To determine whether such mailings occurred before or after the scheme’s completion, we have to establish when the scheme ended. To do so, we have to look to the scope of the scheme as devised by the perpetrator.
In United States v. Sampson, 371 U.S. 75, 83 S.Ct. 173, 9 L.Ed.2d 136 (1962), for example, the defendants were charged with mail fraud based on letters they sent to their victims after they obtained the victims’ money. See id. at 78-79, 83 S.Ct. 173. The letters were designed to lull the victims into a false sense of security in order to postpone their ultimate complaint to authorities. See id. at 78, 83 S.Ct. 173. The district court said there could be no liability because the letters were sent after the money was received, but the Supreme Court disagreed. See id. at 79, 83 S.Ct. 173. The Court explained that the question was not whether the money had been obtained but whether the defendants’ plan had been fully executed. See id. at 80, 83 S.Ct. 173. The Court held that the defendants’ scheme had not been fully executed at the time the letters were sent because the scheme had contemplated lulling activities from the start:
[T]he indictment in this case alleged that the defendants’ scheme contemplated from the start the commission of fraudulent activities which were to be and actually were carried out both before and after the money was obtained from the victims. The indictment specifically alleged that the signed copies of the accepted applications and the covering letters were mailed by the defendants to the victims for the purpose of lulling them by assurances that the promised services would be performed. We cannot hold that such a deliberate and planned use of the United States mails by defendants engaged in a nationwide, fraudulent scheme in pursuance of a previously formulated plan could not, if established by evidence, be found by a jury under proper instructions to be “for the purpose of executing” a scheme within the meaning of the mail fraud statute.
Id. at 80-81, 83 S.Ct. 173 (emphasis added).
We applied the same principle in United States v. Lazarenko, 564 F.3d 1026 (9th Cir.2009). The defendant, a Ukranian politician, engaged in a series of fraudulent business transactions that netted him millions of dollars. See id. at 1029. By 1994, the defendant had obtained the proceeds from the scheme and concealed them in bank accounts in Switzerland and the Bahamas. See id. at 1036. Some years later, in 1997 and 1998, the defendant transferred the money to banks in California to hide his fraudulent activity as he sought political office. See id. We held that the 1997 and 1998 transfers fell outside the wire fraud statute because, in the absence of any evidence that subsequent transfers were part of the scheme as it was originally conceived, the scheme had been completed in 1994:
Under the facts of this case, we conclude no rational trier of fact could find *1303that these transfers in 1997 and 1998 were “in furtherance” of the Naukovy fraud, which centered on allegedly shady agricultural and personal purchases. The fraudulent activity was completed, and the money concealed, in 1994, when the money reached Lazarenko’s control and he deposited it into coded bank accounts where it remained for three years. Subsequent transfers were not part of the scheme as it was originally conceived. Cf. Lo, 281 F.3d at 478. Nothing in the evidence supports an inference, let alone a conviction, on the grounds that the transfers were simply a delayed link in the fraudulent chain.
Id. at 1037.3
These cases teach that mailings designed to avoid detection or responsibility for a fraudulent scheme fall within the mail fraud statute when they are sent pri- or to the scheme’s completion and that the scope of the scheme as devised by the perpetrator determines when that scheme is completed. If the scheme, as conceived by the perpetrator, has been fully executed, then the mailing, even if sent to facilitate concealment of the scheme, falls outside the statute.
We recognize that some would extend liability further, as the government urges us to do here. Some courts appear to require only that the mails are used to avoid detection or responsibility for the fraud, irrespective of whether the mailings postdate completion of the scheme. See United States v. Lopez, 71 F.3d 954, 961-62 (1st Cir.1995), abrogated on other grounds by United States v. Wells, 519 U.S. 482, 117 S.Ct. 921, 137 L.Ed.2d 107 (1997); United States v. Young, 955 F.2d 99, 108 (1st Cir.1992). Because such a rule would be inconsistent with Sampson and Lazarenko, and contrary to Lane’s requirement that the use of the mails take place “prior to the scheme’s completion,” 474 U.S. at 453, 106 S.Ct. 725 (emphasis in original), we decline to follow that approach.
Other courts, relying on “the self-evident proposition that the aim of virtually all criminal actors, including those who commit mail fraud, is not only to accomplish their criminal goals, but also to escape detection and liability for these misdeeds,” would appear to make avoiding detection and prosecution an implicit component of every scheme to defraud. United States v. Hoffman, 229 Fed.Appx. 157, 158 (3d Cir.2007) (unpublished). Hoffman’s suggestion, that every scheme to defraud inherently includes taking any necessary step to avoid detection and prosecution, has intuitive appeal. But it is in tension with Sampson, which required an actual and specific plan to conceal, not a generalized one that could be inferred from the mere existence of a fraudulent scheme. See 371 U.S. at 80-81, 83 S.Ct. 173. Nor do we see how that approach could be reconciled with the outcome in Lazarenko. Under Sampson and Lazarenko, it cannot be enough that the scheme included a general desire not to get caught; the government must be required to prove, at a minimum, that the already conceived scheme included a specific plan for evading detection.
Under Hoffman, moreover, no scheme to defraud would ever end so long as the perpetrator could take some action to avoid detection, prosecution or conviction because any such action would be treated *1304as carrying out an implicit plan to conceal that was part of the scheme from the outset. Such open-ended liability is of great concern, as we recognized in Lazar-enko:
If the government’s theory were correct, then it would be possible for an ordinary fraud to be converted into wire fraud simply by the perpetrator picking up the telephone three years later and asking a friend if he can store some fraudulently-obtained property in his garage before the police execute a search warrant or later taking the proceeds of fraud and transferring them to another bank. The government’s theory extends an already broad statute too far.
564 F.3d at 1037. As the Supreme Court cautioned in Grunewald v. United States, 353 U.S. 391, 77 S.Ct. 963, 1 L.Ed.2d 931 (1957), albeit in the conspiracy context:
[A]fter the central criminal purposes of a conspiracy have been attained, a subsidiary conspiracy to conceal may not be implied from circumstantial evidence showing merely that the conspiracy was kept a secret and that the conspirators took care to cover up their crime in order to escape detection and punishment. ... [Allowing such a conspiracy to conceal to be inferred or implied from mere overt acts of concealment would result in a great widening of the scope of conspiracy prosecutions, since it would extend the life of a conspiracy indefinitely. Acts of covering up, even though done in the context of a mutually understood need for secrecy, cannot themselves constitute proof that concealment of the crime after its commission was part of the initial agreement among the conspirators. For every conspiracy is by its very nature secret; a case can hardly be supposed where men concert together for crime and advertise their purpose to the world. And again, every conspiracy will inevitably be followed by actions taken to cover the conspirators’ traces. Sanctioning the Government’s theory would for all practical purposes wipe out the statute of limitations in conspiracy cases....
... We cannot accede to the proposition that the duration of a conspiracy can be indefinitely lengthened merely because the conspiracy is kept a secret, and merely because the conspirators take steps to bury their traces, in order to avoid detection and punishment after the central criminal purpose has been accomplished.
Id. at 401-02, 405, 77 S.Ct. 963; see also Lutwak v. United States, 344 U.S. 604, 616, 73 S.Ct. 481, 97 L.Ed. 593 (1953) (distinguishing between an actual “agreement to conceal” and “an afterthought by the conspirator for the purpose of covering up”).
Because the government’s theory here would authorize new charges, rather than merely belated charges, it is arguably even more troubling than the theory proposed by the government in Grünewald. In the conspiracy context, recognizing an implied agreement to conceal would only extend the limitations period; the same charges could be brought, albeit later. In the mail fraud context, however, recognizing an implied scheme to conceal would not only extend the limitations period but also give rise to an additional crime every time the mails were used to execute that scheme. That is, because the “scheme” would be said to continue after the embezzlement was completed, the period during which additional substantive crimes could be committed would also be indefinitely ex*1305tended. The theory broadens substantive liability and extends the limitations period.
In sum, we hold that mailings designed to avoid detection or responsibility for a fraudulent scheme fall within the mail fraud statute when they are sent before the scheme is completed. To determine when a scheme ends, we look to the scope of the scheme as devised by the perpetrator.
A scheme may be devised over time, however. Not every perpetrator “deliberately plan[s] and devise[s] a well-integrated, long-range, and effective scheme” from the outset. Sampson, 371 U.S. at 77, 83 S.Ct. 173. Allowance must be made for the reality that embezzlements and other schemes to defraud are often open-ended, opportunistic enterprises. They may evolve over time, contemplate no fixed end date or adapt to changed circumstances. Just as Lopez, Young and Hoffman suggest too expansive a reading of the mail fraud statute by covering any acts of concealment regardless of when they occur, it would be overly restrictive to look only at the scope of the plan as it was originally conceived. Lines will have to be drawn between cases in which acts of concealment can fairly be seen as part of the perpetrator’s evolving plan devised during the life of the scheme and acts of concealment that must be viewed as an after-the-fact event, as in Lazarenko. Cf. Grunewald, 353 U.S. at 405, 77 S.Ct. 963 (making “a vital distinction ... between acts of concealment done in furtherance of the main criminal objectives of the conspiracy, and acts of concealment done after the central objectives have been attained, for the purpose only of covering up after the crime”).
2.
Applying these principles here, we conclude that a reasonable jury could have found that Tanke sent the September 16 letter prior to the scheme’s completion. To be sure, one reasonable inference from the facts is that the fraudulent scheme, as Tanke conceived it, was fully executed by July 2004, when he obtained the last of the proceeds of his embezzlement and terminated his employment with Azteca. A jury could have found that the scheme ended at that time, and that the efforts Tanke later made to intimidate Martin and portray the diversions as legitimate payments for Cedar Creek’s services were simply an after-the-fact coverup Tanke conceived only after Martin discovered the embezzlement and confronted Tanke in early August.
It is also plausible, however, that a reasonable jury could have found that the Cedar Creek cover story was part of Tanke’s embezzlement scheme from the outset or as it evolved over time — just another misrepresentation to facilitate diversion of Azteca’s and CERS’s funds into Tanke’s bank accounts. Lazarenko held that no reasonable jury could find that 1997 and 1998 wire transfers were part of a scheme whose central objectives had been achieved three years earlier. See Lazarenko, 564 F.3d at 1037. Grünewald involved a similar three- to four-year time gap. See Grunewald, 353 U.S. at 395-96, 77 S.Ct. 963. Here, by contrast, there was no such gap, and the September 16 letter was just the tail end of a series of false and misleading financial transactions and statements that comprised the embezzlement scheme. Under the totality of the circumstances, a reasonable jury could have found that the Cedar Creek cover story was a link in the fraudulent chain rather than a post-completion coverup. Cf. Lazarenko, 564 F.3d at 1037.
We note that different jury instructions would have been helpful. The district *1306court could have instructed the jury that it had to find that the mailings were sent prior to the scheme’s completion and that completion of the scheme turned on the scope of the scheme as devised by Tanke. Had the jury been so instructed, it might have been more likely to find that the scheme was completed in July and that the Cedar Creek cover story was conceived and executed separately. No such instructions were given, however, and Tanke does not challenge the instructions provided to the jury. The only question he raises is whether, judged by the statutory requirements, a reasonable jury could have found that the letter of September 16 was sent in furtherance of his fraudulent scheme. We reject Tanke’s argument that his conviction on count 2 must be reversed because the scheme was completed before the September 16 letter was mailed.
3.
We also reject Tanke’s alternative argument that the September 16 letter cannot support a conviction for mail fraud because it was sent after Martin had uncovered the fraud. First, as a factual matter, it is not clear that Martin had uncovered the full extent of the fraud when the letter was sent. Second, even if the fraud had been fully exposed, we are not aware of any authority supporting the proposition that a mailing cannot further a fraudulent scheme merely because the scheme has already been uncovered. To be sure, mailings that serve only to make detection more likely do not further the scheme. See Lo, 231 F.3d at 479; United States v. Manarite, 44 F.3d 1407, 1413 (9th Cir.1995). But we have never laid down a categorical rule that post-detection mailings, if sent before the scheme’s completion, fall outside the statute. Such mailings can further the scheme by, for example, persuading the victim that the fraud did not in fact occur, confusing the issues, discouraging the victim from going to the authorities or establishing a cover story that might be helpful at trial.
For the above reasons, we hold that sufficient evidence supported Tanke’s mail fraud conviction on count 2 of the indictment, and we affirm the conviction.
B. Bankruptcy Misrepresentation Enhancement
Tanke challenges the district court’s application of a 2-level sentencing enhancement for making a misrepresentation during the course of a bankruptcy proceeding. We review the district court’s interpretation of the Sentencing Guidelines de novo and its factual findings for clear error. United States v. Swank, 676 F.3d 919, 921 (9th Cir.2012). There is an intracircuit split as to whether the standard of review for application of the Guidelines to the facts is de novo or abuse of discretion. See id. at 921-22. There is no need to resolve this split where, as here, the choice of the standard does not affect the outcome of the case. See id. at 922.
The United States Sentencing Guidelines Manual (U.S.S.G.) provides: “If the offense involved ... a misrepresentation or other fraudulent action during the course of a bankruptcy proceeding ..., increase by 2 levels.” U.S.S.G. § 2B1.1(b)(9)(B). Tanke argues that his offense did not “involve” a misrepresentation because his fraudulent scheme was completed in 2004, four years before he gave false testimony in a bankruptcy pro*1307ceeding.4 Under § 1B1.3, however, relevant conduct for purposes § 2Bl.l(b) includes all acts committed “during the commission of the offense of conviction, in preparation for that offense, or in the course of attempting to avoid detection or responsibility for that offense.” Id. § 1B1.3(a)(1) (emphasis added). Tanke’s false testimony in the bankruptcy proceeding may not have occurred in preparation for or during the commission of the offense, but it plainly occurred “in the course of attempting to avoid detection or responsibility for that offense.” Id.; accord United States v. Rivera-Gomez, 634 F.3d 507, 513 (9th Cir.2010) (concealing conduct that “occurred long after the” initial offense is still covered by the Sentencing Guidelines, because “nothing in the Guidelines establishes that conduct ceases to be relevant after a specified period of time”). The district court therefore did not err by applying the enhancement.
C. Sophisticated Means Enhancement
We also reject Tanke’s argument that the district court erred by imposing a 2-level enhancement for using sophisticated means. Under U.S.S.G. § 2B1.1, “[i]f ... the offense ... involved sophisticated means, increase by 2 levels.” Id. § 2B1.1(b)(10)(C). The commentary describes sophisticated means as:
especially complex or especially intricate offense conduct pertaining to the execution or concealment of an offense. For example, in a telemarketing scheme, locating the main office of the scheme in one jurisdiction but locating soliciting operations in another jurisdiction ordinarily indicates sophisticated means. Conduct such as hiding assets or transactions, or both, through the use of fictitious entities, corporate shells, or offshore financial accounts also ordinarily indicates sophisticated means.
Id. § 2B1.1 cmt. n.9(B). The district court found that this enhancement applied because “the trial testimony clearly showed a high level of planning and concealment of defendant’s theft, much more than simply diverting business checks into the defendant’s company’s bank accounts for his personal use as the defendant argues.” The court found that Tanke, “as vice president of Azteca, carefully engaged in dozens of various acts over a period of over 16 months to execute and conceal three separate types of fraud on the victim and his two business entities.”
Although Tanke did not use “fictitious entities, corporate shells, or offshore financial accounts,” as the Sentencing Commission’s commentary contemplates, he created at least six false invoices and falsified carbon copies of checks in Azteca’s check register on at least 10 occasions to conceal the payments. These means as a whole were sufficiently sophisticated to support the district court’s decision. See United States v. Horob, 735 F.3d 866, 872 (9th Cir.2013) (per curiam) (affirming the application of the sophisticated means enhancement because, among other things, the defendant “fabricated numerous documents” and “the complicated and fabricated paper trail made discovery of his fraud difficult”); cf. United States v. Jennings, 711 F.3d 1144, 1145 (9th Cir.2013) (apply*1308ing U.S.S.G. § 2T1.1(b)(2)) (“Conduct need not involve highly complex schemes or exhibit exceptional brilliance to justify a sophisticated means enhancement”).
III. CONCLUSION
We affirm Tanke’s conviction on count 2 of the indictment. We hold that the district court did not err by imposing sentencing enhancements under U.S.S.G. § 2Bl.l(b)(9)(B) and (b)(10)(C). We hold, in accord with the government’s concession, that the district court plainly erred by including $44,715.21 in restitution for fraudulent credit card charges and $1,851.38 in restitution for wage overpay-ments that were not part of the offenses of conviction and by failing to note the waiver of interest on restitution on the judgment. The case is remanded to the district court for appropriate proceedings.
AFFIRMED IN PART, VACATED IN PART AND REMANDED.
. Because Tanke challenges the sufficiency of the evidence, we recite the facts in the light most favorable to the government. See United States v. Flyer, 633 F.3d 911, 917 (9th Cir.2011).
. "The federal mail fraud statute does not purport to reach all frauds, but only those limited instances in which the use of the mails is a part of the execution of the fraud, leaving all other cases to be dealt with by appropriate state law.” Kann v. United States, 323 U.S. 88, 95, 65 S.Ct. 148, 89 L.Ed. 88 (1944). Consequently, "there is reason to be scrupulous in reviewing the evidence to assure that the mailing element is adequately proven.” United States v. Lo, 231 F.3d 471, 477 (9th Cir.2000).
. "It is well settled that cases construing the mail fraud and wire fraud statutes are applicable to either.” United States v. Shipsey, 363 F.3d 962, 971 n. 10 (9th Cir.2004).
. Azteca filed for bankruptcy protection in 2004. In 2006, the bankruptcy trustee sued Tanke to recover embezzled funds. Tanke testified in that proceeding in 2008, asserting that as vice president of Azteca he had full authority to settle Azteca's obligations and, because the company allegedly owed Cedar Creek money on old invoices, his diversion of checks was done with authority — testimony that the bankruptcy court found not credible.