FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

UNITED STATES OF AMERICA, No. 22-50142

Plaintiff-Appellee, D.C. No. 3:18-cr-03072- v. TWR-1

ANDREW HACKETT,

OPINION

Defendant-Appellant.

Appeal from the United States District Court for the Southern District of California Todd W. Robinson, District Judge, Presiding

Argued and Submitted August 21, 2023 Pasadena, California

Filed December 18, 2024

Before: Marsha S. Berzon, Johnnie B. Rawlinson, and Daniel A. Bress, Circuit Judges.

Opinion by Judge Rawlinson; Dissent by Judge Berzon SUMMARY *

Criminal Law

The panel affirmed the district court’s judgment in a case in which Andrew Hackett, a stock promoter, was convicted and sentenced for conspiracy to commit securities fraud and securities fraud in connection with the manipulative trading of a public company’s stock. The district court imposed a 16-level sentencing enhancement under the pre-November 1, 2024, version of U.S.S.G. § 2B1.1(b)(1)(I), which applies if the loss exceeds more than $1.5 million. (The 2024 versions of the guideline and commentary do not apply to this case.) Hackett argued on appeal that the district court erred by following the commentary to § 2B1.1, which defines “loss” as the “greater of actual loss or intended loss.” U.S.S.G. § 2B1.1 cmt. n.3(A). According to Hackett, this court should follow the framework articulated in Kisor v. Wilkie, 588 U.S. 558 (2019), to determine whether § 2B1.1 is genuinely ambiguous as it pertains to the definition. In Hackett’s view, because “loss” does not include intended loss in its ordinary meaning, applying intended loss to enhance his sentence impermissibly expanded the guideline. The panel reviewed for plain error because Hackett’s objection to the district court’s loss calculation was not sufficiently specific to preserve de novo review.

* This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

USA V. HACKETT 3

The panel held that the district court’s reliance upon the definition of “loss” set forth in the commentary withstands plain error review because any error was not clear or obvious given this court’s precedent recognizing both actual and intended loss, and because there is a lack of consensus among the circuit courts on this issue. In a concurrently filed memorandum disposition, the panel addressed Hackett’s additional challenges to his conviction and sentence. Judge Berzon dissented. She wrote (1) Hackett’s challenge on appeal includes a narrower argument than a generic challenge to the “intended loss” commentary, in that he also argues that the term “intended loss” does not include a loss that was discussed or hoped for but was never attempted to be implemented; (2) regardless of whether Hackett preserved a wholesale challenge to any inclusion of intended loss in § 2B1.1 calculations, he certainly preserved a narrower objection urging a substantial-action threshold to determine intended loss; (3) as to that argument, if not to the broader one, de novo review is appropriate; and (4) that argument is potentially meritorious, although its application to this case cannot be determined without further district court consideration. Carlton F. Gunn (argued), Law Office of Carlton F. Gunn, Los Angeles, California, for Defendant-Appellant. Zachary Howe (argued) and Mark R. Rehe, Assistant United States Attorneys; Daniel E. Zipp, Assistant United States Attorney Chief, Appellate Section, Criminal Division; Randy S. Grossman, United States Attorney; United States Department of Justice, Office of the United States Attorney, San Diego, California; Aaron P. Arnzen, Bottini & Bottini Inc., La Jolla, California; for Plaintiff-Appellee.

OPINION

RAWLINSON, Circuit Judge:

Andrew Hackett (Hackett) appeals his conviction for one count of conspiracy to commit securities fraud in violation of 18 U.S.C. §§ 371, 981(a)(1)(C), and 28 U.S.C. § 2461(c); and one count of securities fraud in violation of 15 U.S.C. §§ 78j(b), 78f(f), 18 U.S.C. § 981(a)(1)(C), 28 U.S.C. § 2461(c), and 17 C.F.R. § 240.10b-5. Hackett also appeals the forty-six months of imprisonment imposed following his conviction. Hackett specifically challenges the district court’s reliance on the commentary to United States Sentencing Guidelines (U.S.S.G.) § 2B1.1, which defines loss as “the greater of actual loss or intended loss.” U.S.S.G. § 2B1.1 cmt. n. 3(A). 1

1 As we discuss later, the Sentencing Commission recently amended the relevant Guidelines provision and commentary. Those revisions do not

USA V. HACKETT 5

We have jurisdiction under 28 U.S.C. § 1291, and we affirm the judgment of the district court. 2

I. BACKGROUND

Kevin Gillespie (Gillespie) was the founder and CEO of First Harvest, an investment bank. First Harvest primarily consulted with cannabis companies that were preparing to go public or to raise capital. 3 In 2016, First Harvest became a publicly traded company. Gillespie consulted with Annetta Budhu (Budhu), the owner of Baywall, Inc., in taking First Harvest public. Baywall, Inc. was compensated 200,000 restricted shares 4 in First Harvest in exchange for Budhu’s assistance. Gillespie testified that the company began to lose approximately $125,000 a month after going public. According to Gillespie, the company was “[s]teadily raising capital month in and month out.” But on several occasions, Gillespie acquired toxic debt. 5 Around the time Gillespie was taking on toxic debt, Budhu introduced Gillespie to Hackett, a Canadian stock

apply to this case. Unless otherwise noted, all citations of the Guidelines and commentary in this opinion are of the versions in effect prior to the recent 2024 amendments. 2 In a memorandum disposition filed concurrently with this opinion, we address Hackett’s additional challenges to his conviction and sentence. 3 First Harvest was subsequently renamed Arias Intel (Arias). 4 The restricted shares could not be sold for 180 days after First Harvest went public. 5 Gillespie described toxic debt as “like taking a very, very bad loan.” According to Gillespie, if the debt is not repaid, the debt holder will continuously sell shares of the company on the open market “until it’s basically worthless.” promoter. Gillespie and Hackett engaged in several telephone discussions regarding successful stocks that “[Hackett’s] group had participated in.” Gillespie researched one of the stocks and described it as a “pump- and-dump scheme.” 6 Through his company Free Life Investments, Hackett agreed to loan First Harvest $300,000 in exchange for a convertible promissory note. According to the promissory note, Hackett would receive 750,000 shares at $0.40 per share plus interest if after one year Harvest failed to repay Hackett $300,000 plus five percent interest. But the agreement was not executed. According to Gillespie, Hackett’s “money was [not] available and [Hackett’s partner] had capital available.” So First Harvest entered into an agreement with Hackett’s partner, Robert Farrill (Farrill) that was “substantially similar” to the agreement with Hackett. Farrill wired $300,000 to First Harvest and converted the promissory note into 750,000 shares. 7 Hackett also contracted with Budhu to receive Baywall’s restricted shares. Budhu then contacted Clear Trust, LLC (Clear Trust), a stock transfer agent, and requested that Clear

6 A pump-and-dump scheme “involve[s] the touting of a company’s stock . . . through false and misleading statements to the marketplace.” United States v. Zolp, 479 F.3d 715, 717 n.1 (9th Cir. 2007) (internal quotation marks omitted). “After pumping the stock, [f]raudsters make huge profits by selling their cheap stock into the market.” Id. (citations omitted). The critical steps in the “pump-and-dump scheme” are control, pump, and dump. 7 Although Hackett was no longer a party to the agreement, he remained involved through telephone conversations with Gillespie. Hackett encouraged Gillespie to use the $300,000 to invest in a marketing plan, and also suggested that Gillespie file an “S-1 [form] so that the shares would become free trading quickly.”

USA V. HACKETT 7

Trust transfer 200,000 shares of First Harvest stock to Free Life Investments, Hackett’s company, and lift the transfer restrictions under Rule 144 of the Securities Exchange Act of 1934. 8 Prior to lifting the restrictions on sale of the Baywall shares, Clear Trust received a share purchase agreement for 200,000 shares of stock between Baywall, Inc., as seller and Free Life Investments as purchaser; a legal opinion; and a seller’s representation letter from Hackett on behalf of Free Life Investment. In the representation letter, Hackett agreed to comply with Rule 144. After the restrictions were lifted by Clear Trust, Hackett promoted First Harvest stock, and recruited others to promote First Harvest stock. Hackett and his co-defendants paid Stellar Media Group, LLC to develop and distribute newsletters promoting First Harvest and Arias stock. Subject lines included text such as: “HVST stock has expert analysts drooling. Profit now,” “Experts love this stock. It could help you profit by 291%,” “ASNT stock could net you the biggest gains of 2018,” and “ASNT stock is in a position to gain up to 1,880%.”

8 Under Rule 144, a shareholder must present the shares, a seller’s representation letter that the shareholder will comply with Rule 144 when selling shares, and a legal opinion that the shareholder is qualified to resell under Rule 144. In the representation letter, the seller represents that: the resell of the shares would comply with Rule 144 and would be sold “within a reasonable period of time.” The seller also represents that (1) no payment was made “in connection with the offer or sale of the Shares to any person or entity except any customary broker’s commission or dealer’s charges,” (2) there was no solicitation of or arrangement “for the solicitation of orders to buy in anticipation of or in connection with the proposed sale,” (3) the seller did not act “in concert with any person in selling the Shares,” and (4) the seller did not “engage[] in a plan with anyone else to dispose of the Shares.” See Note to 17 C.F.R. § 230.144(f)(1). Hackett also used call rooms to promote stock purchases. David Wolfson (Wolfson) was the owner of several call rooms in California, and was associated with the owner of a call room located in the Philippines. Wolfson described a call room as “an office with agents, salespeople, openers, and closers who market and sell various things.” According to Wolfson, the sales agents used “predictive dialer” software that called “10, 12 lines at a time per person until it reached a contact” from a list of credited investors that Wolfson bought “from lead brokers.” Hackett introduced Wolfson to Liana Millhouse (Millhouse) to work with Wolfson. Hackett informed Wolfson that “most of the stock [Hackett] was successful in selling came as a result of Ms. Millhouse’s work.” Wolfson acknowledged that sales agents would tout the pitched stock with statements that were not always “completely honest.” For example, agents represented that the stocks had favorable growth, price, and appreciation potential. When an investor expressed interest in buying shares, the sales agents would notify Wolfson, and he would alert Millhouse of a prospective purchaser. Millhouse would “give a specific price to execute the trade.” The price provided by Millhouse was often higher than the market price. Wolfson and other sales agents were paid commissions for stock sales. The activities of Hackett and his co-defendants were exposed by FBI informant, Michael Forster (Forster), who was involved with many pump-and-dump schemes. In 2017, Forster arranged a meeting at the airport with an unknown individual. When Forster arrived, he was met with an arrest warrant from the FBI. Forster was offered a cooperation agreement, requiring him “to begin recording and capturing all correspondence with anyone and everyone involved in pump and dumps, with stock fraud.” Forster provided all

USA V. HACKETT 9

recordings, text messages, and emails with the co-defendants to the FBI. Hackett and his co-defendants were subsequently charged with one count of conspiracy to commit securities fraud, and one count of securities fraud. The indictment alleged that Hackett and his co-defendants promoted and recruited others to promote Arias and its stock “in order to artificially avoid the deflation of, maintain the price of, and inflate the share price of Arias stock.” The allegations include Hackett’s participation in the manipulative trading of Arias stock, and “engag[ing] call room operators to contact potential investors and convinc[ing] them to purchase Arias stock, in exchange for a portion of the investments made by those investors.” Once the stock was artificially inflated through these tactics, Hackett sold the stock in the open market. Hackett was convicted on both counts. The district court imposed a sixteen-level sentencing enhancement under § 2B1.1(b)(1). The court found that Hackett owned 550,000 shares and intended to sell each share at four to five dollars a share. The district court used the lower amount to calculate an intended loss amount of $2.2 million. Hackett’s counsel objected to the court’s calculation of the amount of loss, but did not object that intended loss was a legally invalid way to calculate the amount of loss. The commentary to § 2B1.1(b)(1) defines loss as “the greater of actual loss or intended loss.” See U.S.S.G. § 2B1.1 cmt. n. 3(A). The district court obviously relied upon the commentary to determine that the loss caused by Hackett was “the intended loss amount had the venture been successful.” 9 After applying a sixteen-level enhancement, the district court sentenced Hackett to forty- six months of imprisonment. Hackett filed a timely appeal. Our colleague in dissent posits that “Hackett’s challenge on appeal to his sentence includes a narrower argument than a generic challenge to the intended loss commentary. Hackett also argues that the definition of loss [in the Guideline itself] . . . does not include intended loss that never takes place, at least in contexts such as that here.” Dissenting Opinion, p. 25 (internal quotation marks omitted) (alterations in the original). However, a review of the entire paragraph from Hackett’s brief confirms that Hackett’s challenge is indeed to the guideline commentary definition of loss. The entire paragraph reads:

This [Kisor v. Wilkie, 588 U.S. 558 (2019)] makes guideline commentary authoritative only when the actual guideline is ‘genuinely ambiguous.’ The definition of ‘loss’ may vary in some respects, in some contexts, but it does not include ‘intended loss’ that never takes place, at least in contexts such as that here [United States v. Banks, 55 F.4th 246 (3d Cir. 2022)] was correct in holding the ‘intended loss’ application note

9 Our colleague in dissent takes the position that “the district court did not refer to the Guidelines’ commentary to § 2B1.1.” See Dissenting Opinion, p. 24. However, the district court’s use of the phrase “intended loss” mirrors the language used in Comment Note 3(A).

USA V. HACKETT 11

impermissibly expands the guideline for ‘loss,’ and this Court should follow Banks.

Read in its entirety the paragraph, selectively quoted from Hackett’s brief in the Dissenting Opinion, unmistakenly presents a challenge to the Guideline Commentary. The balance of the Dissenting Opinion relies on an analysis that was not included in Hackett’s briefing or mentioned during his oral argument. Accordingly, we briefly make three additional points in response:

1. Hackett’s fraudulent scheme was not “contemplated but unimplemented.” Dissenting Opinion, p. 32. As previously detailed, Hackett orchestrated a classic “pump- and-dump” scheme. 2. Tellingly, the Dissenting Opinion does not cite one case that has adopted our dissenting colleague’s proferred interpretation of § 2B1.1 to include only “losses tethered to a defendant’s substantial actions.” Dissenting Opinion, p. 29. 3. The Sentencing Commission recently amended § 2B1.1 to move the “intended loss” language from the commentary into § 2B1.1 itself. United States Sentencing Guidelines Manual, November 1, 2024, § 2B1.1(b)(1), Notes to Table (A) (“Loss is the greater of actual loss or intended loss”). We acknowledge that the 2024 amendment does not apply to Hackett’s sentencing. However, we include discussion of the amendment because the amendment supports our position that the intended loss calculation challenged by Hackett and our dissenting colleague is not inconsistent with the Guidelines’ “overall structure and purpose.” Dissenting Opinion, p. 31. Importantly, the Sentencing Commission explained that the amendment to § 2B1.1 was made due to the “conflicting court decisions” following the Supreme Court’s decision in Kisor, and to disavow the Third Circuit’s approach as articulated in Banks. See United States Sentencing Commission Guidelines Manual 2024, Supplement to Appendix C, November 1, 2024. This explanation undermines Hackett’s argument that the intended loss calculation is incompatible with other provisions of the Guidelines.

II. DISCUSSION

A. Standard of Review Section 2B1.1(b)(1) of the Sentencing Guidelines directs district courts to increase a defendant’s offense level “[i]f the loss exceeded $6,500.” It then provides a graduated schedule for increasing the offense level if the losses exceed certain amounts. Relevant here, if the loss exceeds more than $1.5 million, a 16-level enhancement applies. See U.S.S.G. § 2B1.1(b)(1)(I). The Guidelines do not define “loss.” But commentary in the Guidelines’ Application Notes states that “loss is the greater of actual loss or intended loss.” U.S.S.G. §2 B1.1 cmt. n. 3(A). On appeal, Hackett argues that under Kisor “intended loss” is not a permissible interpretation of “loss” as used in the Guidelines. The parties disagree on the standard of review that we should apply to this question. The Government contends that we should review for plain error. Hackett argues that we should review de novo. Hackett maintains that he raised an objection “to the presentence report’s loss calculation and methodology.” He also maintains that “even [if] there was not a sufficient objection,” this issue is a question of law and

USA V. HACKETT 13

the government would not be prejudiced if we decide the issue. Under the “contemporaneous-objection rule . . . a party must inform the court – when the court ruling or order is made or sought – of the action the party wishes the court to take, or the party’s objection to the court’s action and the grounds for that objection. . . .” United States v. Klensch, 87 F.4th 1159, 1162 (9th Cir. 2023) (citations, alteration, and internal quotation marks omitted). “Sentencing objections must have a specific substantive basis that provides the district court with an opportunity to address the error in the first instance and allows this court to engage in more meaningful review.” Id. (citation and internal quotation marks omitted). Hackett’s objection to the district court’s loss calculation was not sufficiently specific to preserve de novo review. See United States v. Grissom, 525 F.3d 691, 694 (9th Cir. 2008). In his objections to the calculations in the Presentencing Report (PSR), Hackett did not take issue with using intended loss as a measure of loss. He instead argued that the calculation in the PSR of intended loss in the amount of $4,750,000 was flawed. He maintained that he owned 550,000 shares rather than 950,000 shares of stock, and that the price should have been less than five dollars per share. There was thus no objection to use of intended loss, as set forth in the Application Note, but to the amount arrived at after application of that loss metric. At the sentencing hearing, Hackett made the same objection to the Government’s application of market-share analysis to calculate intended loss. At the sentencing hearing, Hackett likewise did not argue that intended loss was a legally improper measure of loss, whether under Kisor or otherwise. Nor did Hackett clearly advocate for a loss calculation based on the actual losses of the victims, as opposed to a loss amount based on “the actual market price at the time of sale,” which was effectively another way to calculate loss based on intended loss. The government had argued in its sentencing memorandum that actual loss would be difficult to determine because Hackett “hid his transactions in nominee and/or offshore accounts.” Hackett did nothing to rebut this statement at the sentencing hearing. And far from maintaining that intended loss was an improper metric of loss, Hackett’s counsel, if anything, accepted that intended loss could be an appropriate measure of loss. Among other things, counsel stated that “there seems to be a number of different manners in which one can try to calculate loss,” and that “there are different ways to calculate the intent to view the intent in these types of deals.” Counsel’s argument that “[t]he actuality of what happened here was I think even more important,” referred to counsel’s central argument at the sentencing hearing: that the district court should use the actual market price at the time of the sale to calculate loss, as the government had endorsed in the cases of Hackett’s co-defendants. But Hackett’s argument in favor of avoiding disparities among co-defendants is distinct from the claim he now advances on appeal: that “intended loss” is not a permissible interpretation of Guidelines “loss” under Kisor. Hackett points out that we have said that “[i]t is claims that are deemed waived, or forfeited, not arguments.” United States v. Kirilyuk, 29 F.4th 1128, 1135 (9th Cir. 2022). But we do not think it can be fairly said on this record that Hackett is advancing the same claim that he did in the district court. And unlike in Kirilyuk, Hackett at sentencing

USA V. HACKETT 15

accepted the premise that the interpretation of “loss” in the Guidelines’ Application Notes, which included intended loss, could be a permissible reading of “loss” in U.S.S.G. § 2B1.1(b)(1). Kirilyuk is therefore distinguishable. In short, because Hackett did not sufficiently object to the district court’s reliance on the commentary to determine the loss amount for the enhancement, we conclude that our review is for plain error. See id. Finally, we reject Hackett’s argument that we should exercise our discretion to review the Kisor argument de novo. See United States v. McAdory, 935 F.3d 838, 841-42 (9th Cir. 2019) (explaining that we are “not limited” to plain error review when the question is purely legal and the failure to raise it below would not prejudice the opposing party). As we explained above, Hackett did essentially nothing to develop in the district court the factual basis for an actual loss calculation— a factual, not legal issue. And had Hackett more properly objected under Kisor, the government could have offered, and the district court could have considered, alternative ways of calculating loss. Undertaking that analysis now, so many years after the original sentencing, would prejudice the government. For all these reasons, plain error review remains the most appropriate on this record. B. Plain Error Review “A trial court commits plain error when (1) there is error, (2) that is plain . . ., and (3) the error affects substantial rights.” United States v. Ramirez-Ramirez, 45 F.4th 1103, 1109 (9th Cir. 2022) (citation and alteration omitted). Error constitutes plain error when it is so obvious that a district court judge should be able to avoid the error without the benefit of an objection. See United States v. Klinger, 128 F.3d 705, 712 (9th Cir. 1997), as amended (citation omitted). “If those conditions are met, we have discretion to notice such error, but only if the error seriously affects the fairness, integrity, or public reputation of judicial proceedings.” Ramirez-Ramirez, 45 F.4th at 1109 (citation and internal quotation marks omitted). Hackett argues that the district court erred by following the guideline commentary, which defines “loss” as the “greater of actual loss or intended loss.” § 2B1.1 cmt. n.3(A).10 According to Hackett, the commentary’s definition expands “beyond the ordinary meaning of loss.” Hackett insists that we follow the framework articulated in Kisor to determine whether § 2B1.1 is genuinely ambiguous as it pertains to the definition of “loss.” In Hackett’s view, because “loss” does not include intended loss in its ordinary meaning, applying intended loss to enhance his sentence impermissibly expanded the guideline. In Kisor, the Supreme Court instructed courts to “exhaust all the traditional tools of construction” to determine whether a regulation is “genuinely ambiguous” after analyzing its “text, structure, history, and purpose.” 588 U.S. at 575 (citation and internal quotation marks omitted). If after exhausting these “tools of construction,” id., the regulation is not “genuinely ambiguous,” no deference should be given to the agency’s interpretation of the regulation. Id.

10 Under U.S.S.G. § 2B1.1 cmt. n.3(A)(i), “actual loss” is the “reasonably foreseeable pecuniary harm that resulted from the offense.” “Intended loss” is “the pecuniary harm that the defendant purposely sought to inflict,” including “intended pecuniary harm that would have been impossible or unlikely to occur (e.g., as in a government sting operation, or an insurance fraud in which the claim exceeded the insurance value).” Id. at cmt. n.3(A)(ii).

USA V. HACKETT 17

We have not previously held that the term “loss” under § 2B1.1 is genuinely ambiguous. Rather, we have often recognized “intended loss” as part and parcel of the plain meaning of the term “loss.” See United States v. Tulaner, 512 F.3d 576, 578 (9th Cir. 2008) (“In determining the amount of the loss, the greater of the actual or intended loss applies. . . .”) (citing U.S.S.G. § 2B1.1 cmt. n. 3(A)). See also United States v Popov, 742 F.3d 911, 915 (9th Cir. 2014) (“Section 2B1.1 of the Guidelines provides that the applicable loss is the greater of the actual loss or the intended loss. . . .”) (citation omitted); United States v. Jenkins, 633 F.3d 788, 808 (9th Cir. 2011) (“Typically, loss is the greater of actual loss or intended loss . . .”) (citation omitted). We concluded in United States v. Castillo, 69 F.4th 648, 657 (9th Cir. 2023), that the analysis in Kisor must be conducted when applying Stinson v. United States, 508 U.S. 36 (1993). In Stinson, the Supreme Court held that Sentencing Guidelines commentary “that interprets or explains a guideline” is “authoritative unless it violates the Constitution or a federal statute, or is inconsistent with, or a plainly erroneous reading of, that guideline.” 508 U.S. at 38. But Castillo held that “[the] more demanding deference standard articulated in Kisor applies to the Guidelines commentary.” Castillo, 69 F.4th at 655. In a recent opinion, decided before Castillo, we applied Stinson to determine that a $500-per-credit card loss multiplier set forth in the commentary to § 2B1.1 is “inconsistent with, or a plainly erroneous reading of” “loss” under that Guideline. Kiriliyuk, 29 F.4th at 1136-37 (citation omitted). We clarified that, while “dictionary definitions for ‘loss’ may vary, . . . [n]o reasonable person would define the loss from a stolen credit card as an automatic $500 rather than a fact-specific amount.” Id. at 1138 (citation, alteration, and internal quotation marks omitted). We specifically held that “loss” cannot mean a pre-determined, contrived amount with no connection to the crime committed. Id. (citation and internal quotation marks omitted). Rather, “§ 2B1.1 is driven by the amount of loss caused by the crime.” Id. (citation and internal quotation marks omitted) (emphasis in the original). We determined that application of the “$500- per-card multiplier” as provided in the guideline commentary “operate[d] as an enhanced punishment, rather than an assessment of ‘loss’ tied to the facts of the case.” Id. (emphasis in the original).11 In Kiriliyuk, we did not specifically address whether the commentary’s definition of “loss” as including both actual and intended loss impermissibly expanded that term as a matter of law. Id. at 1138. But, the opinion did reference both actual and intended loss. See id. (“As determined by the Probation Office, Kiriliyuk’s conspiracy involved $1.4 million in actual losses or $3.4 million in intended losses. . . .”). This inclusion is consistent with our established practice of referencing both actual loss and intended loss when interpreting the term “loss.” Tulaner, 512 F.3d at 578; Popov, 742 F.3d at 915; Jenkins, 633 F.3d at 808. In sum, we have not grappled with the effect of the Kisor decision on the deference we have afforded the definition of “loss” in the guideline commentary. We decline to do so now because any error was not clear or obvious given our precedent recognizing both actual and intended loss, and because there is a lack of consensus among the circuit courts on this issue. See United States v. Ghanem, 993 F.3d 1113,

11 Hackett has never argued that the assessment of loss in his case was not “tied to the facts of the case.” Id.

USA V. HACKETT 19

1131 (9th Cir. 2021) (“[An error] cannot be plain if there is no controlling authority on point and where the most closely analogous precedent leads to conflicting results.”) (citations and internal quotation marks omitted). Our sister circuits have varied in their determinations of whether Kisor permits application of the definition of “loss” set forth in the commentary. In Banks, 55 F.4th at 257, the Third Circuit reasoned that “[t]he ordinary meaning of ‘loss’ in the context of § 2B1.1 is ‘actual loss.’” The Third Circuit focused on the absence of the term “intended loss” in the Guideline as an indication “that the Guideline does not include intended loss.” Id. (footnote reference omitted). The Third Circuit also relied on dictionary definitions of the term “loss.” See id. at 257-58. The Third Circuit did, however, acknowledge that “loss” could mean “intended loss” when taken “in context.” Id. at 258. However, the Third Circuit observed that “in the context of a sentence enhancement for basic economic offenses, the ordinary meaning of the word loss is the loss the victim actually suffered.” Id. (footnote reference and internal quotation marks omitted). The Third Circuit “accord[ed] the commentary no weight” because in the Third Circuit’s view, “the commentary expands the definition of loss” by including intended loss. Id. (internal quotation marks omitted). In contrast, the Sixth Circuit deferred to the Commission’s interpretation of “loss” because the term “[fell] within [the] zone of ambiguity” of § 2B1.1.” United States v. You, 74 F.4th 378, 397-98 (6th Cir. 2023) (internal quotation marks omitted). The Sixth Circuit concluded that “the definition of loss has no single right answer,” and relying solely on the definition of the word “loss” does not sufficiently engage in the Kisor analysis. Id. (citation omitted). The Sixth Circuit disagreed with the Third Circuit’s analysis in Banks. The Sixth Circuit criticized Banks as “attempt[ing] to impose a one-size-fits all definition” “without consulting the traditional tools of the commentary’s structure, history, and purpose.” Id. at 397 (citations and internal quotation marks omitted). The Sixth Circuit explained that the inclusion of intended loss is consistent with the Guidelines’ purpose of “assess[ing] the seriousness of the offense and the defendant’s relative culpability.” Id. (citations and internal quotation marks omitted). The Sixth Circuit also reasoned that excluding intended loss would result in “vastly different sentences for similarly culpable defendants.” Id. at 398. The Sixth Circuit expressly linked its analysis to the defendant’s situation, observing that “[f]or someone like [the defendant], who was arrested before causing actual loss, including losses that she intended is a reasonable way to gauge her culpability.” Id. One month later, in United States v. Smith, 79 F.4th 790, 798 (6th Cir. 2023), the Sixth Circuit reiterated that the term “loss” is “ambiguous.” In Smith, the Sixth Circuit undertook to further explain why the term “loss” is ambiguous. Id. at 797. The Sixth Circuit referenced “Kisor’s demand that [a court] look at the whole structure of the Guidelines” in assessing ambiguity. Id. at 798. Included within that structure is “§ 1B.1.3, which is the relevant-conduct guideline.” Id. The Sixth Circuit emphasized that “[t]he relevant-conduct guideline instructs the court to consider all harm that resulted from the acts and omissions of the . . . undertaken criminal activity and all harm that was the object of such acts and omissions.” Id. (citation, alteration, and internal quotation marks omitted) (emphasis in the original). The Sixth Circuit observed that “[the] use of the term ‘harm’ in the relevant-conduct guideline clearly contemplates harm that actually occurred and harm that the person intended to

USA V. HACKETT 21

cause.” Id. (internal quotation marks omitted) (emphasis in the original). The court explained that “[t]he context of the Guidelines therefore renders the term ‘loss’ in [§ 2B1.1] ambiguous.” Id. The Tenth Circuit resolved a similar issue in United States v. Maloid, 71 F.4th 795, 808 (10th Cir. 2023), by focusing on whether Kisor overruled Stinson. The Tenth Circuit explicitly held that “[t]he Supreme Court has not abrogated Stinson or the deference we have routinely given the Guidelines’ commentary.” Id. at 813.12 According to the Tenth Circuit, Kisor is a “middle-ground approach to govern the relationship between the Judiciary and executive agencies.” Id. at 806. The court observed “that Kisor had everything to say about executive agencies and precious little about the Sentencing Commission,” which is “a critical distinction.” Id. The court emphasized that the Commission is different from executive agencies because “it speaks as an agent of the Judiciary to help judges properly sentence defendants.” Id. at 807. In contrast, agencies address “policy concerns as agents of the President.” Id. (citation and internal quotation marks omitted). The First, Fourth, and Eleventh Circuits have addressed under plain error review whether “loss” in § 2B1.1 encompasses intended loss. In United States v. Gadson, 77 F.4th 16, 20 (1st Cir. 2023), the First Circuit determined that even if Kisor abrogated Stinson, no binding precedent indicated that the district court’s reliance on the Guideline commentary’s definition of the term “loss” was plainly

12 In an unpublished opinion, the Tenth Circuit concluded that Maloid foreclosed the argument that the term “loss” in § 2B1.1 encompasses only “actual loss.” See United States v. Foreman, No. 22-1255, 2024 WL 548644 at *1-*2 (10th Cir. Feb. 12, 2024). erroneous. The First Circuit observed that the Banks decision did not establish plain error because the Third Circuit had “expressed no opinion as to whether its interpretation was ‘clear or obvious’” and had indicated that its holding applied only “in certain contexts.” Id. at 21. The First Circuit also reasoned that, like the Ninth Circuit, it “regularly” used both actual loss and intended loss to calculate “the harm (both actual and intended) inflicted by the fraudster’s nefarious activities, and that intended loss is frequently a better measure of culpability than actual loss.” Id. (citations and internal quotation marks omitted). In an unpublished disposition, the Fourth Circuit took a similar approach, determining that the district court did not plainly err by affording deference to the guideline’s commentary, because there is no “settled law of the Supreme Court or [the] circuit.” United States v. Limbaugh, No. 21- 4449, 2023 WL 119577 at *4 (4th Cir. Jan. 6, 2023) (citation omitted). The Fourth Circuit declined to “say that the district court committed a clear or obvious error in treating as valid longstanding Guidelines commentary to which the defendant did not object.” Id. (internal quotation marks omitted). The Fourth Circuit concluded that circuit authority provides no “robust consensus” that would “allow [the court] to label as plain any error committed here.” Id. (citation and internal quotation marks omitted). Similarly, the Eleventh Circuit held that its precedent “did not specifically and directly resolve the question of whether § 2B1.1’s definition of loss is ambiguous.” United

USA V. HACKETT 23

States v. Verdeza, 69 F.4th 780, 794 (11th Cir. 2023) (alteration and internal quotation marks omitted).13 After reviewing these cases, we feel comfortable in our conclusion that there is no consensus among the circuits on this issue.

III. CONCLUSION

The district court did not plainly err by relying upon the definition of “loss” set forth in the commentary to § 2B1.1. Because of the unsettled nature of the law on this issue, any error was not clear or obvious. See United States v. Thompson, 82 F.3d 849, 855 (9th Cir. 1996) (“[W]e do not see how an error can be plain error when the Supreme Court and this court have not spoken on the subject, and the authority in other circuits is split.”) (citation omitted).

AFFIRMED.

13 The Eleventh Circuit had previously ruled that Kisor applies to the Commission’s commentary. United States v. Dupree, 57 F.4th 1269, 1276 (11th Cir. 2023) (en banc) (“Kisor’s clarification of Auer [v. Robbins, 519 U.S. 452 (1997)] deference applies to the Guidelines and its commentary.”). BERZON, Circuit Judge, dissenting:

The majority opinion misconstrues both the procedural history of this case and the meaning of the relevant Sentencing Guideline. 1 I would reverse, vacate Hackett’s sentence, and remand to the district court for resentencing. I. On appeal, Hackett challenges a loss-based sentence enhancement under Guideline § 2B1.1. The district court calculated a loss of $2,200,000 by multiplying 550,000 stock shares by a price of $4 per share, which triggered “an adjustment of 16 levels upward for the intended loss amount in this case.” The court found that Hackett “at one time held 550,000 shares” of First Harvest, considering “only those shares that [Hackett] had, not the ones that [he] talked about acquiring.” And it determined $4 as the “target price of the pump aspect” by “credit[ing] the conversation” in which Hackett “discussed the 4 to 5 dollar per share target with respect to the pump aspect of this scheme.” In pronouncing the sentence, the district court did not refer to the Guidelines’ commentary to § 2B1.1. See Maj. Op. at 23 (“The district court . . . rel[ied] upon the definition of ‘loss’ set forth in the commentary to § 2B1.1.”).

1 As the majority notes, the relevant Guideline has now been changed, by moving the commentary regarding “intended loss” into the Notes to the loss amount table included in the Guideline. See Maj. Op. at 4–5 n.1; U.S.S.G. App. C, Amdt. 827 (Nov. 1, 2024). All references in this dissent to the Guideline and the commentary are to the version at the time of sentencing (“the Guideline”). There has been no change in the Guideline as to the issue addressed in this dissent, except that the commentary I rely on later, see pp. 29–30, infra, is now in the Notes.

USA V. HACKETT 25

The majority opinion considers Hackett’s challenge as one solely to the validity of the Guidelines’ commentary stating that “loss is the greater of actual loss or intended loss.” U.S.S.G. § 2B1.1 cmt. 3(A). And the majority reviews that challenge for plain error, reasoning that he objected to the amount of loss calculation rather than to the district court’s deference to the guideline commentary on the definition of intended loss. Maj. Op. at 13. But Hackett’s challenge on appeal to his sentence includes a narrower argument than a generic challenge to the “intended loss” commentary. Hackett also argues that the “definition of ‘loss’ [in the Guideline itself] . . . does not include ‘intended loss’ that never takes place, at least in contexts such as that here.” The majority’s suggestion that, on appeal, Hackett only “specifically challenges the district court’s reliance on the commentary to United States Sentencing Guidelines (U.S.S.G.) § 2B1.1” is thus incorrect. Maj. Op. at 4. At sentencing, Hackett preserved a context-specific objection to the district court’s broad construction of “loss” in § 2B1.1. In his objections to the Presentence Report’s (PSR) Guidelines calculations, Hackett challenged a proposed offense-level increase based on an intended loss figure of $4,750,000. Hackett’s written objection to the PSR argued that “the conclusion in the PSR of an intended loss of $4,750,000 is flawed,” objecting to the use of “some amorphous hopeful price in the future” to calculate the loss amount rather than the “market price at the time of sale.” At his June 2022 sentencing hearing, Hackett renewed his challenge to the PSR’s method of calculating and imputing a loss amount to him. He emphasized that “there are different ways to calculate the intent or to view the intent in these types of deals, because there may have been discussions, but those discussions never necessarily were anything more than mere discussions. The actuality of what happened here was[,] I think[,] even more important.” Hackett maintained that the $4 price discussed was too speculative to be used in loss calculations because no substantial action was taken toward selling at that price:

I will address your $4 calculation specifically. The problem is that these conversations all occur before anything is really ever kind of formalized, or for that matter really arrived at by these defendants. Because there is all sorts of different discussions. Budhu is talking about $2, and then they talk about $4, and then this discussion about $5.

As a result of this tentativeness and vacillation, Hackett further argued,

it could be fairly stated that those numbers that were thrown out [in discussions] weren’t even numbers that people were really planning on or intending. They were just discussing [those prices] as mere opportunities or possibilities. And I think there is a difference between mere opportunities and possibilities and something that was, in fact, intended by the parties.

(Emphasis added.) Hackett’s objection in district court thus fairly previewed the narrower argument he urges—if briefly and somewhat opaquely—on appeal: Even if “intended loss” is properly

USA V. HACKETT 27

included in the loss calculation, the sentence enhancement based on loss does not apply in the current context. Hackett’s challenge to the district court’s loss calculation emphasized that even intended loss must be predicated on the “actuality of what happened”—that is, actions performed, not actions contemplated but abandoned. Regardless of whether Hackett preserved a wholesale challenge to any inclusion of intended loss in § 2B1.1 calculations, I would hold that he has certainly preserved a narrower objection urging a substantial-action threshold to determine intended loss. As to that argument, if not to the broader one, de novo review of his legal position is appropriate. 2 Addressing it, I would hold that the argument is potentially meritorious, although its application to this case cannot be determined without further district court consideration.

II.

Section 2B1.1(b)(1) provides that “[i]f the loss exceeded $6,500,” a defendant’s offense level should be increased as the Guideline indicates. U.S.S.G. § 2B1.1(b)(1). For a loss of “[m]ore than $1,500,000,” a 16-level increase applies. Id. As the language of the Guideline indicates and as explicated by the commentary, the attributed loss does not include loss never realized because of a failure to carry out substantial action toward executing an inchoate plan. As that

2 Like the majority, I do not address de novo whether the entire concept of “intended loss” is inconsistent with § 2B1.1, applying Kisor v. Wilkie, 588 U.S. 558 (2019), and United States v. Castillo, 69 F.4th 648 (9th Cir. 2023). I agree that that issue was not raised in the district court and so is reviewable only for plain error. On my approach to the case, there is no reason to conduct that plain error review, as Hackett is entitled to resentencing in any event. interpretation was not applied by the district court, Hackett should be resentenced applying the proper understanding of the Guideline and commentary, read together. First, the text of the Guideline itself is straightforward. Section 2B1.1’s reference to situations where “the loss exceeded . . . $1,500,000” does not sweep in loss attributable to plans an individual never substantially implemented. Section 2B1.1 “does not define ‘loss.’ In interpreting the Guidelines, we apply the ordinary tools of statutory interpretation and look to the plain meaning of its terms. See Kisor v. Wilkie, 588 U.S. 558, 573-75 (2019). Such tools include ‘consult[ing] dictionary definitions, which we trust to capture the common contemporary understandings of the word.’” United States v. Kirilyuk, 29 F.4th 1128, 1137 (9th Cir. 2022) (citations omitted). The Guideline’s use of the past tense—“loss exceeded”—underscores that § 2B1.1’s plain language refers to losses that are at least readily quantifiable based on a defendant’s concrete actions. See Loss, Black’s Law Dictionary (12th ed. 2024) (defining “loss” as “the disappearance or diminution of value”); id. (“When the loss is a decrease in value, the usual method of calculating the loss is to ascertain the amount by which a thing’s original cost exceeds its later selling price.”). Although “‘loss’ can have a range of meanings,” we have recognized that it “cannot mean a . . . contrived amount with no connection to the crime committed.” Kirilyuk, 29 F.4th at 1137-38. In Kirilyuk, we rejected the imposition of an automatic assessed loss of $500 per stolen credit card number rather than an amount tethered to the facts of a particular offense. See id. at 1133-39. Here, the same insistence that sentence enhancements must bear a “connection to the crime committed,” id. at 1138, should

USA V. HACKETT 29

make us wary of reading “loss” so broadly as to reach intended loss predicated on actions a defendant discussed but never took steps to perform. Second, the history of § 2B1.1 also supports linking “loss” to concrete action by a defendant in furtherance of an offense. The 1987 Guidelines provided for a sentence enhancement pegged to “the value of the property taken,” with a tiered table of values specified under the heading of “[l]oss.” U.S.S.G. § 2B1.1(b)(1) (1987). That is, the original framing of § 2B1.1 understood “loss” as an amount attributable to a defendant’s actions, viz. taking property. See id.; see also U.S.S.G. § 2B1.1(b)(1) (Jan. 1988). The June 1988 Guidelines replaced “the value of the property taken” with, simply, “the loss,” which language (still also reflected in the Guideline’s table heading) endured at the time of sentencing. U.S.S.G. § 2B1.1(b)(1) (June 1988). Third, the commentary, read as a whole, is in this respect consistent with the Guideline’s text, confirming that “loss” in § 2B1.1 refers at least to losses tethered to a defendant’s substantial actions. The commentary’s discussion of “[i]ntended loss” includes two examples of “intended pecuniary harm that would have been impossible or unlikely to occur.” U.S.S.G. § 2B1.1 cmt. 3(A)(ii). In both “a government sting operation” and “an insurance fraud in which the claim exceeded the insured value,” a defendant takes substantial action to effectuate a loss, but the loss does not occur due to third-party actions or policies beyond the defendant’s control. Id. Hackett, in contrast, maintains that he took no similar action to effectuate the loss for which the district court held him responsible. His one-time discussion of selling at $4 per share—the price the district court used to calculate total loss—could turn out to be far afield from a nearly complete transaction that would have occurred but for police deception or an insurance maximum. Whether it is or not would depend on factors such as whether Hackett tried but failed to sell his shares at $4 per share, or whether, in contrast, he never tried to implement the inchoate plan and decided instead—as his district court challenge suggested— to sell at the market price on a particular day, whatever that price was. More broadly, the commentary to § 2B1.1 indicates that “loss” in the Guideline only sweeps in “[i]ntended loss” that a “defendant purposely sought to inflict.” Id. (emphasis added). “Sought” suggests doing something toward an end, not just talking about it. Read alongside the examples above, the commentary’s requirement of intentional pursuit lends further support to a reading of § 2B1.1 that enhances a defendant’s sentence based only on losses he proactively attempted to generate. Fourth, this reading of “loss” coheres with criminal law’s broader approach to inchoate crimes. The law of attempt generally emphasizes that legal liability for a criminal attempt requires that concrete acts be taken toward committing the underlying offense. That is, “[t]o constitute an attempt, the mere intent to commit a crime is not enough; the performance of an act is also necessary,” to “distinguish situations of ‘preparation,’ not deserving of criminal punishment, from situations of genuine attempt.” 1 Wharton's Criminal Law § 7:5 (16th ed. 2023); see id. (listing actions that support attempt liability). In the same vein, our court has long recognized that criminal liability attaches only when some substantial step has been taken toward committing a crime, with an inchoate plan insufficient to trigger liability. See United States v. Saavedra-Velazquez, 578 F.3d 1103, 1104 (9th Cir. 2009) (noting that “the definition [of attempt] at common law . . .

USA V. HACKETT 31

requires a ‘substantial step towards committing the crime’” (citation omitted)); United States v. Buffington, 815 F.2d 1292, 1301 (9th Cir. 1987) (requiring “conduct constituting a substantial step toward commission of the crime”); see also United States v. Gracidas-Ulibarry, 231 F.3d 1188, 1192 (9th Cir. 2000) (collecting sources). Fifth, looking to the Guidelines’ overall structure and purpose, other provisions make clear that sentencing must rest on injuries attributable to actions actually undertaken— not merely contemplated but abandoned—by a defendant. Section 1B1.3’s discussion of relevant conduct provides that a defendant’s offense level “shall be determined on the basis of . . . all harm that resulted from the acts and omissions specified in subsections (a)(1) and (a)(2) above, and all harm that was the object of such acts and omissions.” U.S.S.G. § 1B1.3(a)(3) (emphasis added). Subsections (a)(1) and (a)(2) sweep in “all acts and omissions committed . . . or willfully caused by the defendant” and, for joint offenses, “all acts and omissions of others . . . that occurred” within certain parameters. U.S.S.G. § 1B1.3(a)(1), (a)(2) (emphases added). This language, relied on by the government as support for the proposition that “loss” includes “intended loss,” makes clear that for Guidelines purposes an “object” counts as harm only if a concrete act or omission with that object occurred, not if it was only contemplated. These provisions support a consistent reading of the Guidelines in which sentencing is based on harms, including losses, attributable to a defendant’s actual actions. The commentary to § 1B1.3 confirms this understanding of “harm.” One note refutes the idea that the Guidelines’ conception of “harm” includes those that have not been realized but were only risked, stating that “[u]nless clearly indicated by the guidelines, harm that is merely risked is not to be treated as the equivalent of harm that occurred.” U.S.S.G. § 1B1.3 cmt. 6(B). Sixth, the fact patterns in other cases on intended loss are instructive in parsing § 2B1.1. The majority opinion invokes case law from other circuits to buttress its analysis. But most of the cases on which it relies do not match Hackett’s situation as he portrays it. Those cases permit attributing intended loss to a defendant in situations where that defendant took substantial steps toward carrying out an offense directed at causing a loss of a certain amount, but that amount of loss did not materialize due to circumstances beyond the defendant’s control. They therefore underscore that § 2B1.1’s use of “loss” does not include amounts attributable to contemplated but unimplemented plans. For example, in Banks, the Fourth Circuit determined that “Banks’s plot was to open Gain Capital [Group] accounts and make electronic deposits into those accounts, but his deposits were drawn on bank accounts with insufficient funds. He then tried to withdraw funds from these accounts, with the goal being to complete the withdrawals/transfers before the lack of supporting funds could be detected.” United States v. Banks, 55 F.4th 246, 251 (3d Cir. 2022) (internal quotation marks omitted). The defendant “made fraudulent deposits of $324,000 and unsuccessfully executed 70 withdrawals/transfers totaling $264,000” with Gain, although Gain “suffered no actual loss” and “did not transfer a single dollar to Banks.” Id. Similarly, the Sixth Circuit in Smith held that the “intended loss” amount in a scheme “defraud[ing] banks and their customers” by misusing account holders’ personal information included the actual losses “plus the funds the conspirators tried to steal but were unsuccessful at

USA V. HACKETT 33

obtaining.” United States v. Smith, 79 F.4th 790, 792-93 (6th Cir. 2023). And in Gadson, the First Circuit explained that the “coconspirators obtained the names and personal information . . . of real individuals, and then used that information to apply for [specific] loans for themselves in those persons’ names, with no intention of repaying the loans.” United States v. Gadson, 77 F.4th 16, 18 (1st Cir. 2023). 3 Finally, the Eleventh Circuit in Verdeza described the intended loss of $3.4 million as representing the total amount two clinics billed to an insurer as part of a fraudulent healthcare operation, although the insurer “grew suspicious . . . [and] denied the claims.” United States v. Verdeza, 69 F.4th 780, 785-86, 794 (11th Cir. 2023). In each of these instances, it appears, the defendant took substantial actions that were meant to achieve the amount of intended loss attributed, but was thwarted by external circumstances or their own errors. *** I am convinced that, considered along with the commentary, the Guideline’s reference to “loss” is not ambiguous as to the contextual issue Hackett preserved on appeal. See Kisor, 588 U.S. at 575. Section 2B1.1’s use of “loss” clearly excludes intended losses attributable only to a defendant’s contemplated but unperformed actions. We do not punish thought crimes and so cannot rely for sentencing

3 “To support the loan applications, Gadson and his coconspirators also created and used fraudulent supporting documents, such as counterfeit driver’s licenses, pay stubs, and lease agreements.” United States v. Gadson, 77 F.4th 16, 18 (1st Cir. 2023). The opinion does not contain more information about how the intended loss was calculated based on the factual circumstances or how “the district court here determined that intended loss was greater than actual loss.” Id. at 20. purposes on evidence of an intent never manifested by substantial action.

III.

As to how my interpretation of Section 2B1.1 applies here: The factual record in this case is, unfortunately, murky. In calculating the loss amount under § 2B1.1, the district court attributed 550,000 total shares to Hackett, although it is not immediately clear whether Hackett took steps to effectuate the sale of all of those various shares. My reading of the record suggests that Hackett did pump the value of First Harvest stock and then sell thousands of shares of that stock at the resulting elevated market price (which was higher than it would have been absent fraudulent pumping). Later, there was a near-complete loss of value in the remaining First Harvest stock because the market for that stock was flooded with shares for sale. But the underlying record is not sufficiently developed to provide a full factual account of Hackett’s relevant conduct or of the connection between such conduct and the prices and total shares the district court attributed to him in determining a $2,200,000 loss and thus a 16-level enhancement under § 2B1.1. Various questions remain: How many of the 550,000 shares attributed to Hackett were sold on the open market versus through call rooms versus via other means versus never sold at all? At what price was each of those shares ultimately sold? Who had control over which blocks of stock at which points in time? What substantial actions did Hackett take—not just consider—with respect to selling the shares under his control as part of his offense conduct? And what led Hackett not to sell his shares at the price he discussed on the phone call: his own decision to sell sooner at a different price, market forces beyond his control that foreclosed the

USA V. HACKETT 35

contemplated price, or some other factor? The answers to these important questions, which bear directly on how much loss the Guidelines should attribute to Hackett, are not evident in the record and briefing before us. “[A]s a general matter, if a district court errs in sentencing, we will remand for resentencing on an open record—that is, without limitation on the evidence that the district court may consider.” United States v. Matthews, 278 F.3d 880, 885 (9th Cir. 2002) (en banc). Because the answers to the above questions are not available at this juncture, I would remand the case on an open record so that the district court can apply the correct interpretation of “loss” in § 2B1.1 when resentencing Hackett, after conducting any necessary additional fact-finding.