Defendant Richard I. Berger appeals his conviction of twelve counts of conspiracy, loan fraud, falsifying corporate books, and various securities fraud violations. Berger argues that: (1) the district court improperly coerced the jury into reaching a verdict, (2) the district court violated his constitutional right to be present during trial when the district court — with counsel’s consent — made certain comments at an informal meeting with the jury outside of Berger’s presence, (3) the district court used the wrong materiality standard for securities fraud violations, (4) the indictment did not charge with sufficient particularity the materiality element for securities fraud violations, and (5) the district court erred when it ordered Berger to pay restitution. The government crossappeals the sentence imposed by the district court, arguing that the district court erred when it refused to increase Berger’s sentence based on judicially-found facts. We have jurisdiction over Berger’s appeal pursuant to 28 U.S.C. § 1291 and the government’s cross-appeal pursuant to 18 U.S.C. § 3742 (b). For the reasons given below, we affirm the conviction, affirm the restitution order, vacate the sentence and fine, and remand for resentencing under United States v. Booker, 543 U.S. 220 , 125 S.Ct. 738 , 160 L.Ed.2d 621 (2005).

Factual Background

I. Offense Conduct

A. Craig Consumer Electronics, Inc. and the Revolving Credit Agreement

Craig Consumer Electronics, Inc. (“Craig Electronics”) operated a consumer electronics business that sold products such as car stereos, compact music cen ters, and small personal stereos to retail stores. Berger was Craig Electronics’ President, Chief Executive Officer, and Chairman of the Board. Donna Richardson, a co-conspirator, pled guilty to three counts of the indictment prior to trial. Richardson was the Chief Financial Officer of Craig Electronics until May 31, 1997, when she left the company. Defendant Bonnie Metz was at various times a Vice President in Craig Electronics’ Hong Kong and Cerritos, California locations. Metz is not a party to this appeal.

On August 5, 1994, Craig Electronics entered into a $50 million revolving credit agreement (“Credit Agreement”) with a consortium of banks including BT Commercial Corporation (“Bankers Trust”), La Salle National Bank, Nationsbank of Texas, and Sanwa Business Credit Corporation. Bankers Trust acted as the agent for the consortium (collectively “lending banks”). Under the Credit Agreement, Craig Electronics could, subject to certain exclusions, borrow up to:

Craig Electronics was prohibited from borrowing against goods that had been returned to Craig Electronics but not yet inspected, or goods that were defective, sometimes referred to as “C” goods.

Craig Electronics was required to provide Bankers Trust with a Borrowing Base Certificate (“Borrowing Certificate”) every business day. Each Borrowing Certificate was supposed to report accurately the amount of Craig Electronics’ accounts receivable eligible for borrowing, updated on a daily basis, and the value of its inventory eligible for borrowing, updated on a weekly basis. The Credit Agreement required that either Berger or Richardson supervise the preparation of each Borrowing Certificate and certify in writing that the information it contained was true, correct, and complete in all material respects.

Based on the information in the Borrowing Certificates, Bankers Trust determined the amount of money Craig Electronics could borrow on each business day. Specifically, the lending banks conditioned their lending decisions on whether Craig Electronics had excess borrowing availability based on the information — particularly the accounts receivable and inventory eligible for borrowing purposes — set forth in the daily Borrowing Certificates.

Any materially false or misleading representation made in the Borrowing Certificates was identified as an event of default under the Credit Agreement. Craig Electronics was required to notify the lending banks of the nature of any default no later than two business days after it occurred.

B. Falsification of Information

Starting as early as 1995 and continuing through September 1997, Craig Electronics did not have sufficient accounts receivable and inventory to continue to borrow the money needed to fund its operations. Presumably to hide Craig Electronics’ true financial condition from the lending banks, Berger, Richardson, and Metz regularly falsified the information contained in the Borrowing Certificates. They used the following methods: First, Berger, Rich ardson, and Metz inflated the accounts receivable reported to the lending banks by: (1) pre-billing retail stores for goods that had not yet been shipped and, in some instances, had never been purchased; (2) deliberately delaying the processing of credits for returned goods that had been received and identified in an off-the-books accounting ledger called the “queue,” and not reporting this substantial volume of credits; and (3) falsely reporting that $1 million in accounts receivable from a company in Brazil remained valid through May 1997 when, in fact, the underlying sales were reversed and the goods re-routed back to Craig Electronics approximately two months earlier.

Second, Berger, Richardson, and Metz distorted the inventory figures submitted to the lending banks by: (1) improperly classifying “C” goods as “A” or “B” goods, and (2) misreporting that Craig Electronics had requisite title to certain shipments of goods originating with its overseas suppliers when, in fact, Craig Electronics either did not have proper title to the shipments for borrowing purposes, or the shipments did not exist. The financial misreporting that falsely inflated Craig Electronics’ accounts receivable and inventory caused the lending banks to lend more money to Craig Electronics than it was allowed to borrow under the Credit Agreement.

Third, to conceal the fraudulent nature of Craig Electronics’ reported accounts receivable and inventory, Berger, Richardson, and Metz deceived and attempted to deceive Craig Electronics’ outside accountants as well as auditors from the lending banks. For example, they instructed Craig Electronics employees not to reveal to the accountants and auditors the true status of Craig Electronics’ accounts receivable and inventory.

The false statements resulted in the lending banks loaning millions of dollars to Craig Electronics based on either nonexistent or substantially overstated collateral. The lending banks did not discover the full extent of the fraud until after Craig Electronics filed for bankruptcy on August 1, 1997. According to one witness, the lending banks suffered approximately $8.4 million in losses.

Finally, Berger and Richardson failed to disclose Craig Electronics’ true financial condition in several mandatory reports they filed with the Securities and Exchange Commission (“SEC”). Those SEC filings were Craig Electronics’ Amended S-1 Registration Statement, Amended 1996 10-K Report, and First Quarter 1997 10-Q Report. As a result of the activities described above, Craig Electronics was operating while in default of the Credit Agreement and was substantially overdrawn on its line of credit. None of this information, however, was disclosed in Craig Electronics’ mandatory SEC filings.

II. Trial, Jury Deliberations, and Verdict

The grand jury returned an indictment against Berger and Metz. The indictment alleged conspiracy, loan fraud, falsification of corporate books and records, making false statements to accountants of a publicly-traded company, and making false statements in reports filed with the SEC. Berger’s and Metz’s trial began on May 20, 2003 and lasted forty-one days.

A. Status Conference with the Parties

On the morning of August 29, 2003, after the jury had deliberated for three-and-a-half days, the district court held a status conference with all parties and their respective attorneys. The court hoped to discuss some of the jurors’ conflicting schedules, that appeared to limit the number of days available for deliberation. The court believed that some of these requests for days off might be related to “the issue of stress and responsibility on the part of the jury.”

The court suggested that it might be helpful to engage in an informal discussion with jurors on the record but outside the presence of the parties and their attorneys. The court explained its proposal:

Defense counsel expressed concern that the need for the judge to address the jurors was not yet necessary. After further discussion, however, counsel for Berger and Metz indicated that “doing it informally” might be acceptable. Berger’s counsel agreed with the prosecutor that it would be more productive for the court to indicate what it intended to say, and then the parties “could come to an agreement pretty quickly.” The court explained:

Counsel for defendants declined the court’s invitation to write a “script.” Metz’s counsel stated that Metz would be willing to have the court informally address the jury “without counsel being present,” because the court’s “vague outline” of what it intended to say “does not sound to us like it will in any way pressure the jury to rush to judgment,” but “will do just the opposite.” Berger’s counsel agreed, stating that he had conferred with Metz’s counsel and agreed “with everything he said.” Berger’s counsel’s agreement was qualified only by his request that, “in discussing search for justice,” the jurors “be reminded that the government bears the burden of proof and guilt beyond a reasonable doubt.”

Based on this representation from counsel, the court accepted Berger’s personal waiver of his right to be present for the court’s meeting with the jury. Berger waived his “right to be present for the particular communication that[the court] anticipate^] having with the jury.” The court found Berger’s and Metz’s waivers to be free and voluntary.

B. Informal Meeting with the Jury

In its discussion with the jury, the court first observed that several jurors had conflicting medical appointments. The judge stated:

After some laughter, two jurors indicated that they were willing to modify their plans so that deliberations could go forward on September 2nd and 5th. Juror Roux, who lived 170 miles from the courthouse, changed her child care plans; Juror Morgan delayed her medical appointment. When Morgan commented that she was not happy about the delay, the court responded, “I understand. The day that I gave instructions, I think Friday, I hadn’t slept for four nights. That’s rough, not sleeping. Then not breathing.” -

When the court turned to dates for the week starting September 9th, Juror Roux interjected that she was “ninety-nine percent sure we’ll be done by then,” to which .the court responded, ‘Wonderful., I’m glad to hear that.” Juror Roux asked whether the others agreed they would be finished by the 9th. Juror Morgan stated, “There’s no way to say.” In response, Juror Roux said, “I pretty much — I do— we all have our set minds pretty piueh.” The court then engaged in the following discussion with the jury:

The court then called on the foreperson, Juror Lynch, who had a personal conflict the week of September 9th. Lynch prefaced her answer by emphasizing that the jury still had not reviewed all the evidence and therefore could not “say [they were] stuck on things,” rather, “everythingfwas] still undecided completely.” After scheduling a portion of the week of September 9th for deliberation, the court indicated that it probably did not need to worry about dates during the week of September 15th through 18th because “[y]ou feel by that time a verdict will be reached.” At that point, an unidentified juror stated, “I sure hope so. If not, I’ll jump out this window.” Juror Roux stated, “I’m with you, buddy. Can I go first? I will even let you push me.”

In response to these concerns, the judge closed the meeting with the following comment:

C. Post-Meeting Ameliorative Instruction

Upon returning from its discussion with the jury, the judge stated that he “may”

have gone “beyond the script” by responding to jurors who had “blurted out ... that they made up their minds and [were] not going to change[them]” to “reexamine your views to see if you’re correct in your view.” At the request of Berger’s counsel, the judge agreed to have a transcript of his colloquy with the jury made available immediately, so that a corrective instruction could be given if necessary.

At a mid-afternoon hearing that same day, Berger moved for a mistrial, arguing that the court’s final comments to the jury misstated the burden of proof and took “the worst part of an Allen charge” without the balancing language. The government argued that the challenged passage essentially reiterated the instruction in the court’s initial charge concerning the duty to deliberate. The court denied Berger’s motion for a mistrial. The court, however, brought the jury back into the courtroom and read the following clarifying instruction:

The jurors resumed deliberations at 9:30 a.m. on September 2, 2003, after the Labor Day weekend. At 1:20 p.m. that day, they sent a note indicating that they: (1) had reached unanimous verdicts against Berger on twelve counts but were deadlocked as to the remaining twenty-four counts against Berger and as to all twenty-one counts against Metz, and (2) needed further instructions as to how to proceed. Upon the jury’s return for further proceedings on September 4, 2003, the court took the jury’s guilty verdicts as to twelve counts against Berger and declared a mistrial as to the rest of the counts against Berger and on all of the counts against Metz.

III. Sentencing

The court sentenced Berger to six months in prison and imposed a $1.25 million fine. The court also ordered Berger to pay the lending banks $3.14 million in restitution.

Discussion

I. Alleged Allen Instruction

A. Standard of Review

A district court’s decision to issue an Allen instruction is reviewed for abuse of discretion. See United States v. Plunk, 153 F.3d 1011, 1027 (9th Cir.1998), amended by 161 F.3d 1195 (9th Cir.1998). An Allen instruction “must be upheld unless it is clear from the record that the charge had an impermissibly coercive effect on the jury.” Id. Whether a judge has improperly coerced a jury’s verdict is a mixed question of law and fact we review de novo. See Jiminez v. Myers, 40 F.3d 976, 979 (9th Cir.1994).

B. The Court Did Not Coerce the Jury into Reaching Unanimous Verdicts

Berger contends that the district judge gave an improper Allen charge when he met informally with the jury and that the charge effectively coerced the jury to reach unanimous verdicts against him on twelve counts in violation of Berger’s right to an impartial jury. We disagree.

This court has explained:

United States v. Mason, 658 F.2d 1263 , 1265 n. 1 (9th Cir.1981) (citation omitted). The Allen instruction is most often used in cases of “apparent juror deadlock” to “admonish jurors to keep trying.” Id. at 1265; see also Weaver v. Thompson, 197 F.3d 359, 365 (9th Cir.1999) (“In the archetypal Allen charge context, the judge instructs a deadlocked jury to strive for a unanimous verdict.”).

In Weaver , this court stated that “[s]o long as the defendant has offered facts that fairly support an inference that jurors who did not agree with the majority felt pressure from the court to give up their conscientiously held beliefs in order to secure a verdict, we must proceed to the Allen charge analysis.” Weaver, 197 F.3d at 365 . In the instant case, the district judge made comments that echoed a mild Allen instruction when he mentioned: “[W]ith respect to those of you who reached a particular conclusion, and that you will not change your minds[, i]t wouldn’t be wrong for you to reconsider your position if you can be convinced that perhaps your position was not accurate, that it could be wrong.” The court’s remarks came soon after Juror Roux commented: “There’s some of us that are dead set on our verdicts and others that are dead set. You could probably leave us in here a hundred years and we’ve deliberated, we’re gone over it and my vote will not change.” These comments fairly support the inference that the jury might have been deadlocked — at least as to some counts — and that the court’s comments could have been construed as instructing the jurors to reconsider their respective positions if convinced their position was not correct. Accordingly, a full Allen analysis is appropriate.

We have stated that there is “nothing talismanic about any single element either making the charge valid or invalid; the fundamental question is whether the jury was improperly coerced, thus infringing the defendant’s due process rights.” Id. We apply a “totality of the circumstances” analysis when examining whether a judge’s statements to a jury were impermissibly coercive. Jiminez, 40 F.3d at 980 . In performing the Allen analysis, it is helpful to consider three relevant factors: “(1) the form of the instruction, (2) the time the jury deliberated after receiving the charge in relation to the total time of deliberation and (3) any other indicia of coerciveness.” United States v. Steele, 298 F.3d 906, 911 (9th Cir.2002).

1. The form of the instruction was not coercive

Berger takes issue with three aspects of the “form” of the so-called Allen instruetion the court delivered during its informal meeting with the jury. First, Berger complains that the judge’s comments omitted an instruction that jurors should not abandon their conscientiously held beliefs. Second, Berger alleges that the judge’s statements about his poor health coerced the jury into reaching a verdict. Finally, Berger claims that the judge strongly suggested to the jurors that they should work to a unanimous verdict when he said that if they did not follow his instruction it would be like “the Hatfields fighting the McCoys.” We conclude that Judge Taka-sugi’s remarks were not coercive. In addition, the ameliorative instruction cured any coerciveness that may have resulted from the judge’s informal comments to the jurors.

a. The district court instructed the jurors to hold on to their beliefs

At the end of the judge’s informal meeting with the jurors, he told them that “[i]t wouldn’t be wrong for you to reconsider your position if you can be convinced that perhaps your position was not accurate, that it could be wrong.” If the jury was truly deadlocked, these words could be interpreted as directing the jury to “reconsider potentially unreasonable positions.” Mason, 658 F.2d at 1265 n. 1. Berger contends that because the judge overlooked telling the jury to hold on to their “conscientiously held beliefs,” there was no adequate counterbalance to the coercive aspects of the so-called Allen instruction.

Generally, when a judge tells jurors to reconsider their positions, the judge must also warn the jurors to hold on to their conscientiously-held beliefs. See Mason, 658 F.2d at 1268 (reversing where instruction was more coercive than the instruction approved in Allen and failed to tell jurors in the minority not to abandon their conscientiously-held views); see also Jiminez, 40 F.3d at 981 & n. 5 (noting that failure to instruct jurors to hold on to conscientiously held beliefs “weighs heavily in favor of the conclusion that the defendant’s right to a fair trial and impartial jury has been violated”); but see United States v. Cuozzo, 962 F.2d 945, 952 (9th Cir.1992) (“While it is helpful for an Allen charge to include such ameliorative language, its lack does not itself necessarily require reversal.”). Berger, however, misconstrues the record when he contends that Judge Takasugi’s informal comments to the jury — which Berger characterizes as an Allen instruc tion — were not offset by other comments by the judge telling the jury to hold on to their beliefs.

First, the judge told the jurors: “The way I look at it, as long as your position is sincere, the position taken ... is based upon your recollection of the evidence and law, there’s not much more we can ask for.” The judge’s statement was substantially the same as an instruction telling jurors to hold on to their conscientiously held beliefs. Telling the jury to hold on to a “sincere” position based on a “recollection of the evidence and law” neutralized any potentially coercive effect and undercuts Berger’s contention.

Second, at the request of the defendants, the court gave an ameliorative instruction later that afternoon. The court instructed the jury: ‘You should not take from my remarks this morning any suggestion that you should change your views simply in order to reach an agreement or because other jurors think it is right.” That corrective instruction also neutralized any coercive effect of the court’s earlier informal comments. The instruction explained that each juror’s personal views were more important than the parties’ or the court’s interest in obtaining a unanimous verdict. Thus, we find that Berger’s first challenge to the court’s informal comments lacks merit. See United States v. Bonam, 772 F.2d 1449, 1451 (9th Cir.1985) (“When the portion of the instruction that asks the minority to re-examine its views is counterbalanced by the caution that a juror should not abandon his conscientiously held views, we have generally upheld the instruction as not coercive.”); see also United States v. Ajiboye, 961 F.2d 892, 894 (9th Cir.1992).

b. It was not coercive for the judge to mention his recent illness

Berger next argues that the judge engaged in coercion when he told the jury he had been hospitalized during trial with a serious illness. Berger implies that because Judge Takasugi is a senior judge who made reference to his illness, the jury felt compelled to reach a verdict. Such an interpretation is a misreading of the transcript. The judge’s comments regarding his health cannot reasonably be read to have had any coercive effect on the jury. First, the judge tempered his comments about his health by saying, “But the heart and lungs are very strong, so I guess I’m going to stick around for a while.” Further, Berger’s contention that references to the judge’s illness pushed the jury to reach a verdict is undermined by reading the judge’s next words, which were: “Jurors should not be forced to reach a verdict. Please understand that. Because any time you’re forced to reach a verdict you’re going to reach an improper verdict or for improper reasons. And so far as that’s concerned, there’s no time lim-it____” The judge’s discussion of his illness did not create a coercive atmosphere.

c. The judge’s statement that failure to reconsider a position would “be like the Hatfields fighting the McCoys” was not coercive

Berger next focuses on the judge’s informal comment that it “wouldn’t be wrong for you to reconsider your position,” and that jurors “have to have that state of mind throughout the deliberations. Otherwise it’s going to be like the Hatfields fighting the McCoys. It’s not going to be promotive of a final conclusion.” Berger asserts that “[t]he court’s statements told the jurors that honest and conscientious beliefs were less important than getting along — indeed, these statements told the jurors that hanging on to such beliefs would make them responsible for inappropriate feuding.”

The comparison of a deadlocked jury to the Hatfields and the McCoys might have conveyed to the jury that the judge favored a unanimous agreement over a deadlock. But we must not consider that analogy in isolation. We must instead consider it in the context of the judge’s other informal remarks and his ameliorative instruction. See Jiminez, 40 F.3d at 980 . When so viewed, the “Hatfields fighting the McCoys” comment was not coercive. As noted above, the judge told the jury that they would not be “forced to reach a verdict” and that jurors should not “changeftheir] views simply in order to reach an agreement or because other jurors think it is right.” Furthermore, the court later gave the ameliorative instruction that the jurors should not “change [their] views simply in order to reach an agreement or because other jurors think it is right.” Any coercive effect that emanated from the court’s “Hatfields fighting the McCoys” comment was neutralized by the court’s counterbalancing statements and corrective instruction.

2. The time the jury deliberated after the informal meeting in relation to the total time it deliberated does not suggest that the court’s statements were coercive

“A jury verdict reached immediately after an Allen charge can be an indication of coercion.” Bonam, 772 F.2d at 1451 . In contrast, this court has found no coercion existed in circumstances where the deliberations lasted some significant amount of time after an Allen instruction was given. See id. at 1450-51 (finding no coercion with just over one day of total deliberation, one- and-a-half hours of which came after Allen charge); see also Lorenzo, 43 F.3d at 1307 & n. 3 (finding no coercion with five-and-a-half hours of deliberation coming after Allen charge); Cuozzo, 962 F.2d at 952 & n. 6 (finding no coercion with two days and six hours of deliberation, six hours of which came after Allen charge); but see Weaver, 197 F.3d at 366 (finding coercion existed when jury returned with unanimous verdict five minutes after Allen charge).

The jury deliberated at least seven hours following the challenged comments by the judge. On the morning of August 29, 2003, the court, with counsel’s consent, had its informal meeting after the jury had deliberated about three-and-a-half days. Following the judge’s meeting with the jurors, the jury deliberated until 3:19 p.m. At that time, the court gave the ameliorative instruction that the jury members should not “change [their] views simply in order to reach an agreement or because other jurors think it is right.” The court recessed for the weekend. The jury resumed deliberations at 9:30 a.m. on September 2, 2003, and reached a verdict that afternoon at 1:20 p.m.

Seven hours was a substantial amount of time after the informal meeting. During that time, the jury could consider the charges against both defendants. In our opinion, seven hours was a “sufficient period of time to allow the jury to reach a reasoned decision in this case.” Cuozzo, 962 F.2d at 952 n. 6.

Finally, we note that the jury’s verdict on September 2nd came after the three-day Labor Day holiday. We have held:

Steele,. 298 F.3d at 911 . Here, as in Steele , any potentially coercive effect of the judge’s remarks was diluted by the long holiday weekend and the time the jury deliberated after the court’s informal remarks.

3. There Were No Other Indicia of Coercion

a. The Court’s informal statements to the jury were not directed toward a specific juror or set of jurors

Berger argues that the judge’s informal comments were particularly coercive because they were directed at a specific juror, Juror Roux. This argument misrepresents the record, because the judge addressed his remarks toward the jurors as a group. The judge only knew that “some” of the jurors were set on one position and that “others” were set on another. The judge did not know which jurors favored guilt and which favored acquittal. Moreover, because of the complicated nature of the case, Juror Roux’s comment that “some of us ... are dead set on our verdicts” might have related only to some of the dozens of charges the jurors had to consider. The record also does not reveal whether Juror Roux was in the majority or minority on any particular charge, nor how many jurors she was referring to when she said “some of us that are dead set on our verdicts.”

Here, “the judge did not ask what the division was. There is no evidence that he knew who the dissenting jurorfs were.] Further, the judge did not even know whether the majority position was to convict or acquit.” Lorenzo, 43 F.3d at 1307. Thus, Juror Roux could not have been singled out by the judge in his comments, which were made to the multiple jurors who were “dead set” on their verdicts.

b. It is not clear that the jurors were in fact deadlocked

Contrary to Berger’s contention, the jury was not deadlocked at the time the judge, with the consent of counsel, had an informal meeting with the jurors regarding scheduling issues. Berger’s case is distinguishable from other cases where an allegedly coercive Allen charge is made to an obviously deadlocked jury. In those cases, the jury typically sends out a note indicating that it is deadlocked and inquiring about the next appropriate step. See Jim-inez, 40 F.3d at 978-79 (explaining that jury sent notes to judge that stated “We are unable to reach a verdict and feel strongly that we would not be able to reach a verdict” and “We are at an impasse and request further direction.”).

There is simply no evidence in the record that any of the jurors considered further deliberations to be futile. The jury knew that deliberations were ongoing. With the consent of counsel, the judge met informally with the jurors, who had served for over three months, about their issues regarding child care, travel to the courthouse, medical appointments, and vacation plans — hoping to schedule dates for more deliberation. Juror Lynch stated that the jury had not reviewed all the evidence and could not “say [the jury was] stuck on things,” rather, “everything [was] still undecided completely.” Thus, the judge did not make his remarks in an atmosphere where the jurors would have felt that unanimity was their only escape from the jury room.

c. The jury was not unanimous on all counts

The jury reached agreement on twelve counts against Berger. The jury deadlocked on the remaining twenty-four counts against Berger. The jury also deadlocked on all twenty-one counts against Metz. This result clearly tells us that the jury exercised “their rational and independent review of the evidence” and did not succumb to the court’s alleged coercion. See Plunk, 153 F.3d at 1027 ; Cuozzo, 962 F.2d at 952 .

In light of the foregoing, we hold that Berger’s contention that the district court impermissibly coerced the jury to arrive at unanimous verdicts is without merit.

II. Berger’s Constitutional Right to Be Present

A. A Defendant’s Right to Be Present

Berger next argues that his right to be present at every stage of the proceedings was violated when the judge made the challenged informal comments to the jury outside the presence of Berger and his counsel. We conclude that Berger waived his right to be present but that the scope of the waiver did not include the full range of comments made by the judge during the informal meeting with the jury. Consequently, even though the so-called Allen instruction was not coercive, the challenged language was spoken by the court without Berger’s knowledge and consent, and thus impinged his due process right to be present at every critical stage of trial. We nevertheless hold that the error, if any, was harmless beyond a reasonable doubt and therefore not a ground for reversal.

This court has stated that the “Constitution, the fundamental principles of jury trial, and the Federal Rules of Criminal Procedure guarantee a defendant the right to be present at every stage of trial.” United States v. Frazin, 780 F.2d 1461, 1469 (9th Cir.1986) (citing Illinois v. Allen, 397 U.S. 337, 338 , 90 S.Ct. 1057 , 25 L.Ed.2d 353 (1970) (holding right of presence secured by the confrontation clause)). In United States v. Gagnon, 470 U.S. 522 , 105 S.Ct. 1482 , 84 L.Ed.2d 486 (1985), in discussing a defendant’s right to be present at critical stages of trial, the Supreme Court explained:

Id. at 526 , 105 S.Ct. 1482 (quoting Snyder v. Massachusetts, 291 U.S. 97, 105-06, 108 , 54 S.Ct. 330 , 78 L.Ed. 674 (1934)) (internal quotations and alterations omitted). Such an error, however, is not a basis for automatic reversal. If an ex parte communication “rises to the level of a constitutional violation, then the burden is on the prosecution to prove that the error was harmless beyond a reasonable doubt.” United States v. Rosales-Rodriguez, 289 F.3d 1106, 1109 (9th Cir.2002).

Our case law is clear that communication between the judge and jury-outside of counsel’s presence, without a proper waiver, violates a defendant’s right to due process of law. See Frazin, 780 F.2d at 1469 ; see also Rosales-Rodriguez, 289 F.3d at 1110 (noting that delivery of a supplemental jury instruction is a “critical” stage of a trial that requires a defendant’s or defense counsel’s presence). A defendant, however, may waive his or her constitutional right to be present at all critical stages of the proceedings “provided such waiver is voluntary, knowing, and intelligent.” Campbell v. Wood, 18 F.3d 662, 671-72 (9th Cir.1994) (citing Johnson v. Zerbst, 304 U.S. 458, 464 , 58 S.Ct. 1019 , 82 L.Ed. 1461 (1938)).

Below, we consider the scope of Berger’s waiver of his right to be present during the court’s informal meeting with the jury. We then discuss whether the constitutional error, if any, was harmless beyond a reasonable doubt.

B. Berger Did Not Waive His Right to Be Present During the So-Called Allen Instruction

The parties do not dispute that Berger knowingly, intelligently, and voluntarily waived his right to be present during the judge’s informal communication with the jury. They do, however, disagree as to the scope of his waiver. We agree with Berger that he only waived his right to be present during the court’s informal meeting with the jury, so long as the discussion involved the setting of future deliberation dates that would not conflict with the child care, vacation, and travel arrangements of some of the jurors. He did not waive his right to be present during the administration of the so-called Allen instruction.

We narrowly construe Berger’s waiver and only read it to include whatever Berger explicitly waived. See Gete v. INS, 121 F.3d 1285, 1293 (9th Cir.1997) (“[A] waiver of constitutional right is not to be implied and is not lightly to be found.” (internal quotations omitted)); see also Johnson, 304 U.S. at 464 , 58 S.Ct. 1019 . Before accepting Berger’s waiver of his right to be present, the judge explained to the parties what he planned to say to the jury at the informal meeting. The judge stated, “If there had been any suggestions about an Allen instruction I want the record to indicate that I do not believe in the Allen instruction, I will not give it, never have given it.” When asked what he was specifically planning to say, the judge narrowed his plan to two specific points: first, he would “address the issue ... regarding the jury[’s] request for days off’; second, he would “make certain that [he could] convey to them the thought that a rush to judgment is probably the worst form of verdict you could receive.” Berger expressly waived his “right to be present for the particular communication that [the court] anticipate^] having with the jury.”

The record shows that the court intended to (1) discuss the scheduling conflicts some of the jurors were having, and (2) explain to the jurors that a rush to judgment would not be appropriate. More im portantly, the court emphatically stated that it would not give an Allen instruction. Because Berger’s waiver only encompassed the “particular communication” that was “anticipated” by the district court, the waiver was restricted to the court’s two planned topics of discussion.

The jurors’ comments, however, led the judge away from his stated plan. Juror Roux stated that some jurors were “dead set” on their verdicts. In response to that comment the judge noted that “[i]t wouldn’t be wrong for you to reconsider your position if you can be convinced that perhaps your position was not accurate, that it could be wrong.” Berger argues that he did not waive his right to be present when that comment was uttered, which was, at worst, the mildest form of a mild Allen instruction. See Mason, 658 F.2d at 1265 n. 1 (“In their mildest form, [Allen ] instructions carry reminders of the importance of securing a verdict and ask jurors to reconsider potentially unreasonable positions.”).

C. The Constitutional Error Was Harmless

A violation of a defendant’s due process right to be present at critical stages of trial is subject to harmless error analysis. See Frazin, 780 F.2d at 1469 . Constitutional error is harmless if a court concludes “beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained.” Id. at 1469-70 (quoting Chapman v. California, 386 U.S. 18, 24 , 87 S.Ct. 824 , 17 L.Ed.2d 705 (1967)). As we have thoroughly discussed, the portion of the judge’s comments that arguably exceeded the scope of Berger’s waiver did not coerce the jury into reaching a verdict. The comments were, at worst, the mildest form of an Allen instruction. Further, the jury deliberated a substantial amount of time — seven hours — after the comments at issue; and the jury did not reach unanimous verdicts on all counts. All of the factors that we considered above demonstrate that the district court’s statements were not coercive and did not affect the jury’s verdict. See Rosales-Rodriguez, 289 F.3d at 1111 (holding that district court’s delivery of supplemental instruction without consulting parties was harmless error because “[t]he instruction ... was not coercive and did not cause the jury to rush to judgment”); Frazin, 780 F.2d at 1469-71 (finding that the judge’s failure to consult parties when responding to a jury note was harmless error because jury reached its own conclusion and “not as the result of actual or perceived judicial pressure”). Moreover, the defendants had an immediate opportunity to review the transcript of the meeting. The same afternoon, the court provided an ameliorative corrective instruction. Looking at the totality of the circumstances, we are convinced beyond a reasonable doubt that the jury’s verdict was not affected by any part of the court’s informal meeting with the jurors and any perceived error was harmless. Frazin, 780 F.2d at 1471 .

III. Materiality in Indictment Counts 34, 35, and 36

A. Indictment Counts 34, 35, and 36— The Materiality Element

Counts 34, 35, and 36 charged that, in three SEC filings, Berger knowingly omitted material facts about Craig Electronics’ (1) fraudulent accounting practices, (2) default of the Credit Agreement, and (3) overdrawn status on its line of credit.

At trial, the government gave notice of its intention to call a victim investor and an SEC expert witness to testify about the materiality of omissions in the SEC filings. Berger objected on the ground that Kungys v. United States, 485 U.S. 759 , 108 S.Ct. 1537, 99 L.Ed.2d 839 (1988), required the government to prove that the false statements were material to the SEC. Berger argued that the investor had no relevant testimony concerning the effect of the falsehoods on the SEC and that expert testimony on the materiality of omissions would invade the province of the jury.

The government responded that Kungys governs false statements to government agencies was not controlling because the relevant materiality standard in securities fraud cases is whether a misrepresentation or omission would influence a reasonable investor, not the SEC. The district court agreed but concluded that under the reasonable investor standard, the proposed testimony of the investor and of the SEC expert witness were unnecessary.

In a motion for judgment of acquittal at the close of the government’s case-in-chief, Berger asserted that there was insufficient evidence that the falsehoods were material to the SEC. Berger also contended that the indictment charged that the false statements were material to the lending banks, not to individual investors. In a post-verdict motion for judgment of acquittal, Berger repeated these claims and also argued that there was a fatal variance between the indictment — which alleged that the false statements were material to the lending banks — and the proof — which established that the false statements were material to reasonable investors. The court denied both motions.

B. Standard of Review

The materiality standard to be applied is a question of law reviewed de novo. See, e.g., United States v. Rosenthal, 445 F.3d 1239, 1244 (9th Cir.2006). The sufficiency of an indictment is reviewed de novo. See United States v. Pernillo-Fuentes, 252 F.3d 1030, 1032 (9th Cir.2001).

C. Counts 35 and 36 — Materiality Under Section 32(a) of the Securities Exchange Act of 1934 Assessed from the Reasonable Investor’s Perspective

Counts 35 and 36 of the indictment alleged that Berger made material omissions in mandatory filings with the SEC, in violation of section 13(a) of the Securities Exchange Act of 1934 (“1934 Act”), 15 U.S.C. § 78m(a), and section 32(a) of the 1934 Act, 15 U.S.C. § 78ff(a). Section 13(a) is a mandatory filing provision, that requires certain companies to file with the SEC “information and documents ... as the Commission shall require to keep reasonably current the information and documents required to be included in or filed with an application or registration statement” as well as “annual reports ... and ... quarterly reports.” 15 U.S.C. § 78m(a)(l)(2). Section 32(a) provides criminal penalties for “any person who willfully and knowingly makes ... any statement in any application, report, or document required to be filed under this chapter or any rule or regulation thereunder ... which statement was false or misleading with respect to any material fact .... ” 15 U.S.C. § 78ff(a) (emphasis added).

Berger contends that when applying section 32(a), courts should assess materiality from the perspective of the SEC. Berger primarily bases his argument on Kungys v. United States, 485 U.S. 759 , 108 S.Ct. 1537 , 99 L.Ed.2d 839 (1988). In Kungys, the Supreme Court interpreted a statute that penalized the “concealment of a material fact” from the Immigration and Naturalization Service. Id. at 767-72, 108 S.Ct. 1537 . The Supreme Court looked to the federal courts of appeals, which had a “uniform understanding” of “materiality” under a similar statute, 18 U.S.C. § 1001 . Kungys, 485 U.S. at 769-70 , 108 S.Ct. 1537 . Section 1001 is a general statute that criminalizes acts of material misrepresentation in “any matter within the jurisdiction of the executive, legislative or judicial branch of the Government of the United States.” 18 U.S.C. § 1001 . The courts of appeals had, in § 1001 cases, assessed materiality from the perspective of the government agency. See Kungys, 485 U.S. at 769-70 , 108 S.Ct. 1537 . The Supreme Court agreed with that interpretation and held that a false statement made to a government agency is “material” if it “has a natural tendency to influence, or was capable of influencing, the decision of the decisionmaking body to which it was addressed.” Id. at 770 , 108 S.Ct. 1537 (internal quotations omitted).

Berger identifies cases that apply the Kungys materiality standard to a number of different agencies, including the Environmental Protection Agency, see United States v. Self, 2 F.3d 1071, 1083-84 (10th Cir.1993), and the Food and Drug Administration, see United States v. Diaz, 690 F.2d 1352, 1357 (11th Cir.1982). Each of those cases dealt with the materiality of false statements made in the context of specific agency decisions. See, e.g., Self 2 F.3d at 1084 (noting false statements “had a tendency to forestall any regulatory agency investigation”); Diaz, 690 F.2d at 1358 (explaining that false records could “impair the FDA in carrying out its responsibility for protection of the public health”). Berger contends that the SEC, as a regulatory body, makes decisions based on the information contained in a company’s mandatory filings. Thus, Berger argues that the materiality of false statements made to the SEC under section 32(a) must also be assessed not from the reasonable investor’s perspective, but from the SEC’s perspective, in the context of its own regulatory decisions.

We disagree with Berger. The purpose of the 1934 Act was to benefit and protect investors, with proper agency decisionmak-ing as a secondary concern. “The 1934 Act was designed to protect investors against manipulation of stock prices.” Basic Inc. v. Levinson, 485 U.S. 224, 230 , 108 S.Ct. 978 , 99 L.Ed.2d 194 (1988) (citing S.Rep. No. 73-792, at 1-5 (1934)). Furthermore, the Court “repeatedly has described the ‘fundamental purpose’ of the Act as implementing a ‘philosophy of full disclosure.’ ” Id. (quoting Santa Fe Indus., Inc. v. Green, 430 U.S. 462, 477-78 , 97 S.Ct. 1292 , 51 L.Ed.2d 480 (1977)). Accordingly, the Supreme Court has established a specific materiality standard to be used when assessing fraudulent statements filed with the SEC. In TSC Industries, Inc. v. Northway, Inc., 426 U.S. 438 , 96 S.Ct. 2126 , 48 L.Ed.2d 757 (1976), the Supreme Court explained in the proxy-solicitation context that “[a]n omitted fact is material if there is a substantial likelihood that a reasonable shareholder would consider it important in deciding how to vote.” Id. at 449 , 96 S.Ct. 2126 . In formulating that standard, the Court noted the broad remedial purpose of Rule 14a-9 of the 1934 Act:

TSC Indus., 426 U.S. at 448 , 96 S.Ct. 2126 . Thus, the Court held that under section 14(a) and Rule 14a-9, materiality should be assessed, not from the SEC’s perspective, but from the investor’s perspective. See id. at 449 , 96 S.Ct. 2126 . The Court later expanded the reach of the TSC Industries materiality standard to section 10(b) and Rule 10b-5, which prohibit fraud in connection with the purchase or sale of securities. See Basic Inc., 485 U.S. at 231-82 , 108 S.Ct. 978 ; see also United States v. Tarallo, 380 F.3d 1174, 1182 (9th Cir.2004) (assessing materiality under Section 10(b) from reasonable investor’s perspective).

Applying the “reasonable investor” materiality standard to Section 32(a) is consistent with the goals of the SEC. As amicus curiae, the SEC noted that the agency itself commences actions on filings it considers materially misleading to investors. Cf. Touche Ross & Co. v. Redington, 442 U.S. 560, 570 , 99 S.Ct. 2479 , 61 L.Ed.2d 82 (1979) (“The information contained in the[mandatory] reports is intended to provide the [SEC],[the New York Stock Exchange], and other authorities with a sufficiently early warning to enable them to take appropriate action to protect investors.... ”). In addition to being a regulatory body, the SEC acts as a repository of information intended to be disseminated to and used by the public. See United States v. Bilzerian, 926 F.2d 1285, 1301 (2d Cir.1991) (“The securities laws are designed to make accurate information available to the investing public.... ”). The mandatory filings at issue in this case, for example, were meant for investors’ use. Craig Electronics’ Form 10-K and Form 10-Q, which were periodic reports, were filed under Section 13(a) “as necessary or appropriate for the proper protection of investors and to insure fair dealing in the security.” 15 U.S.C. § 78m(a). Furthermore, Congress has explained the purpose of the mandatory reporting provisions of the 1934 Act:

S.Rep. No. 73-1455, at 74 (1934) (emphasis added). It is clear that the reporting requirements under the 1934 Act are intended to protect investors, and that materiality should be assessed from the reasonable investor’s perspective.

The SEC concedes that the mandatory filings here satisfy a related but separate purpose specific to the agency — the SEC reviews the documents to determine whether to investigate a particular transaction. False filings that relate to specific agency decisions, however, are criminalized by a separate statute, 18 U.S.C. § 1001 . Indeed, in Bilzerian , the Second Circuit held that the government can institute two separate counts against an individual for the same false filing. In that case a false filing could be charged under both (1) section 10(b), the 1934 Act provision relating to the purchase or sale of securities, and (2) § 1001, the general statute criminalizing false statements made to government agencies. See Bilzerian, 926 F.2d at 1299-1301 (“The securities laws are designed to make accurate information available to the investing public; the SEC’s authority to regulate disclosure required under those laws brings the securities filings at issue within its jurisdiction for purposes of § 1001.”).

Here, the government could have charged Berger with making false statements that related to the SEC’s regulatory decisionmaking under § 1001 in addition to charging him with filing false statements that were material to investors under Sec tion 32(a). As to the hypothetical § 1001 charge, the Kungys case’s interpretation of materiality- — -examining the misstatement’s ability to influence the SEC’s decision-making — would be appropriate. However, under Section 32(a), which was promulgated under an act specifically intended to benefit the investing public, misstated or omitted facts should be evaluated for materiality as to those investors.

Finally, Berger argues that “materiality must be assessed in the context of a decision.” He points to Sections 10(b) and 14(a) for comparison. In the Section 10(b) context, courts assess materiality by examining a fact’s potential to influence an investor’s particular decision — the decision to buy or sell a security. See Ganino v. Citizens Utils. Co., 228 F.3d 154 , 161 (2d Cir.2000). Similarly, in Section 14(a), the false statement must have a tendency to influence a decision — how an investor will vote. See TSC Indus., 426 U.S. at 449 , 96 S.Ct. 2126 . Section 32(a), however, only criminalizes the filing of false information and does not expressly implicate any specific type of investment decision. Thus, Berger argues that the materiality standard we adopt today is “unworkable and incoherent.”

We disagree. In Sections 10(b) and 14(a), affected investment decisions — the decision to buy or sell shares and the

decision to vote a particular way — are enumerated as elements of the statutes. The language of Section 32(a) is distinct; it criminalizes the mere filing of a material false statement without requiring that the statement affect a particular- investment decision. It thus appears that Congress intended Section 32(a) to act as a catch-all provision to punish those who file a false statement, whether or not the filing can be shown to affect a specific investment decision, as long as the false statement could affect a reasonable investor. See Pope, 189 F.Supp. at 21 (“Section 32(a) is a catch all provision.... ”).

We recognize the materiality standard set forth in TSC Industries as the materiality standard for Section 32(a) of the 1934 Act. Materiality must be assessed from the perspective of the reasonable investor. Accordingly, we affirm the district court’s interpretation of Section 32(a).

D. Materiality Element in Count 34 Under the Securities Act of 1933

1. Berger’s Conviction Under Count 34 May Not Be Reversed for an Error in Citation

In Berger’s reply brief, he contends that his conviction on Count 34 must be reversed. Count 34 alleged that Berger violated Section 32(a) when he filed false statements on Craig Electronics’ Amended S-1 Registration Statement. Berger points out that Section 32(a) applies to false statements made on any application, report, or document filed under the 1934 Act. See 15 U.S.C. § 78ff. S-1 Registration Statements, however, are required under the Securities Act of 1933 (“1933 Act”), 15 U.S.C. §§ 77a-77bbbb, not the 1934 Act. See Section 6, 1933 Act, 15 U.S.C. § 77f; see also FORM S-l Registration Statement Under the Securities Act of 1933, available at http://www.sec. gov/about/forms/forms-l.pdf (“This Form shall be used for the registration under the Securities Act of 1933.... ”). Thus, Berger argues that the government did not prove an essential element of Section 32(a) — that he made a materially false statement on an application, report, or document required under the 1934 Act — and accordingly, that his conviction under Count 34 must be reversed.

We begin by noting that the government could have properly charged Berger with making a materially false statement on an S-1 Registration Statement under Section 24 of the 1933 Act, 15 U.S.C. § 77x. Section 24 of the 1933 Act criminalizes the filing of materially false statements made on a registration statement filed under the 1933 Act. See 515 U.S.C. § 77x. Thus, both Section 24 of the 1933 Act and Section 32(a) of the 1934 Act criminalize the filing of material false statements on documents required under their respective acts.

Berger does not contest that Section 24 of the 1933 Act criminalizes the conduct alleged in Count 34. Accordingly, the error alleged by Berger here is one of improper statutory citation, rather than one of substantive error in the indictment. Such an error is not a proper basis for reversal.

The Federal Rules of Criminal Procedure counsel that “[u]nless the defendant was misled and thereby prejudiced, neither an error in a citation nor a citation’s omission is a ground to dismiss the indictment or information or to reverse a conviction.” Fed.R.Crim.P. 7(c)(3). Here, the language of the indictment clearly constitutes a satisfactory enumeration of the elements of Section 24 of the 1933 Act. Accordingly, Berger was not “misled to his prejudice,” nor did “defects or irregularities in the indictment affect[his] substantial rights.” Hockenberry v. United States, 422 F.2d 171, 174 (9th Cir.1970). We hold that Count 34 here “describes the activities forming the basis of the charge [under Section 24 of the 1933 Act] with sufficient particularity to assure that the Grand Jury deliberated on the elements of the crime.” Echavarria-Olarte v. Reno, 35 F.3d 395, 398 (9th Cir.1994). Because the defect here was one of form, rather than substance, we decline to reverse his conviction because of the citation error in the indictment. See id. at 399 ; United States v. Fekri, 650 F.2d 1044, 1046 (9th Cir.1981) (“When there has been no prejudice and the error is merely an error in the citation, reversal is not required.”).

2. Materiality Under Section 24 of the Securities Act of 1933 Assessed from the Reasonable Investor’s Perspective

Because we hold that Berger was properly convicted under Section 24 of the 1933 Act, we next consider the proper materiality standard applicable to that provision. We have already determined that under Section 32(a) of the 1934 Act, materiality must be assessed from the perspective of the reasonable investor. For similar reasons, we hold that materiality under Section 24 of the 1933 Act must also be assessed from the reasonable investor’s perspective.

We acknowledge the subtle differences in the purposes of the 1933 and 1934 Acts. The Supreme Court has recognized that “[t]he essential purpose of the [1933 Act] is to protect investors by requiring publication of certain information concerning securities before offered for sale.” AC. Frost & Co. v. Coeur D’Alene Mines Corp., 312 U.S. 38, 40 , 61 S.Ct. 414 , 85 L.Ed. 500 (1941) (emphasis added). This purpose is slightly different from the purpose of the 1934 Act, which applies to post-distribution securities exchanges. See Reader v. Hirsch & Co., 197 F.Supp. 111, 114 (S.D.N.Y.1961) (“The Securities Exchange Act of 1934 does not overlap the 1933 Act, but rather supplements it, i.e., it deals with post distribution trading.”).

Nevertheless, as with the 1934 Act, Congress was primarily concerned with protecting the investing public when it passed the 1933 Act. In Ernst & Ernst v. Hochfelder, 425 U.S. 185 , 96 S.Ct. 1375 , 47 L.Ed.2d 668 (1976), the Supreme Court noted: “The Securities Act of 1933 ... was designed to provide investors with full disclosure of material information concerning public offerings of securities in commerce, to protect investors against fraud and, through the imposition of specified civil liabilities, to promote ethical standards of honesty and fair dealings.” Id. at 195 , 96 S.Ct. 1375 (citing H.R.Rep. No. 73-85, at 1-5 (1933)) (emphasis added); see also SEC v. Ralston Purina Co., 346 U.S. 119, 124 , 73 S.Ct. 981 , 97 L.Ed. 1494 (1953) (“The design of the [1933 Act] is to protect investors by promoting full disclosure of information thought necessary to informed investment decisions.”). Furthermore, this court has recognized that “the purpose of the [1933 Act] is to compel full and fair disclosure in the issuance of securities so that investors will be adequately protected.” United States v. Carman, 577 F.2d 556, 564 (9th Cir.1978) (emphasis added).

Our above discussion of the 1934 Act also applies to our interpretation of the 1933 Act. Both acts are primarily designed to protect investors: the 1934 Act protects investors with respect to post-distribution trading, and the 1933 Act protects investors at the time of issuance. It makes sense that in the penalty provisions of both acts, the materiality of a misstatement must be assessed from the reasonable investor’s perspective. As we stated above, the claim that an individual made a materially false statement that affected the SEC in its regulatory decisionmaking capacity is more appropriately charged as a violation of 18 U.S.C. § 1001 . Accordingly, we apply the materiality standard set forth in TSC Industries as the materiality standard for Section 24 of the 1933 Act.

E. The Government’s Indictment Charged the Materiality Element with Sufficient Particularity

Berger next asserts that even if we address materiality from the perspective of a reasonable investor, the indictment failed to allege that the fraudulent statements were material to investors. Berger argues that the government only charged that the fraudulent statements were material to the lending banks from which the loans were obtained. Thus, Berger maintains, when the government introduced evidence pertaining to the statements’ effect on investors, that was a fatal variance requiring reversal.

Federal Rule of Criminal Procedure 7(c)(1) requires that an indictment be a “plain, concise and definite written statement of the essential facts constituting the offense charged.” This court has said that an indictment “should be read in its entirety, construed according to common sense, and interpreted to include facts which are necessarily implied.” United States v. King, 200 F.3d 1207, 1217 (9th Cir.1999). Finally, if an indictment contains an error, there must be some evidence that the error misled the defendant to the defendant’s prejudice. See id. (citing Fed. R.Crim.P. 7(c)(3)).

The structures of Counts 34, 35, and 36 are similar. Each count first indicates the SEC filings on which the fraudulent statement was made: Craig Electronics’ Amended S-1 Registration Statement, Amended 1996 10-K Report, and First Quarter 1997 10-Q Report. The indictments then detail Berger’s misrepresentations and omissions: (1) “the fraudulent transfers of ‘C’ inventory to ‘B’ inventory ... ”, (2) that “Craig was in default of the Credit Agreement because of the false statements it had made to Bankers Trust in the Borrowing Base Certificates”, and (3) that “Craig was substantially overdrawn on its line of credit.” All three counts conclude with the final paragraph:

The question before us is whether materiality to investors could be inferred from the indictment. We believe that although the indictment did not allege to whom the misstatements were material, such lack of specificity was not fatal in this case.

It is self-evident that “the materiality [to the reasonable investor] of information relating to financial condition, solvency and profitability is not subject to serious challenge.” SEC v. Murphy, 626 F.2d 633, 653 (9th Cir.1980). In this case, we find it beyond dispute that, before making reasonable investment decisions, a reasonable investor would consider Craig Electronics’ “cash flow and ability to finance inventory purchases” to be material information.

Furthermore, that the indictment alleged materiality to investors was evident from the face of the indictment. “It is well settled, at least in this circuit, that an indictment need not allege the materiality of a false representation if the facts advanced by the pleader warrant the inference of materiality.” United States v. Oren, 893 F.2d 1057, 1063 (9th Cir.1990). In Oren , this court upheld an indictment in a § 1001 case involving the National Park Service. See id. at 1063. Though the indictment did not specifically mention materiality, the indictment indicated that: (1) Oren had made a false statement regarding an offer for land; (2) the statement was within the jurisdiction of the Park Service, which would only purchase through a specific third-party nonprofit agency; (3) the Park Service would purchase land only if it had an appraisal; and (4) Oren had given his false statement to the third party. See id. at 1064. This court held that “[s]urely these allegations warrant the inference of materiality of Oren’s false statement.” Id. (internal quotations omitted).

Likewise, here, the indictment detailed the misstatements filed by Berger, each of which could have been relied upon by a reasonable investor. The S-1, 10-K, and 10-Q filings submitted to the SEC were documents upon which reasonable investors would base investment decisions. Because the indictment alleged material misstatements and omissions contained in documents upon which investors would rely, we hold that the indictment was sufficiently detailed to warrant the inference of materiality to investors.

IV. Restitution

A. Standard of Review

A court has broad discretion in ordering restitution. See United States v. Laney, 189 F.3d 954, 966 (9th Cir.1999). A restitution order is reviewed for abuse of discretion, provided it is within the bounds of the statutory framework. See United States v. Phillips, 367 F.3d 846, 854 (9th Cir.2004). The court’s valuation methodology is reviewed de novo. See United States v. Lomow, 266 F.3d 1013, 1020 (9th Cir.2001). The factual findings supporting a restitution order are reviewed for clear error. See United States v. De La Fuente, 353 F.3d 766, 772 (9th Cir.2003).

B. The District Court Properly Calculated Restitution

Neither party disputes the district court’s conclusion that restitution in this case is governed by the Mandatory Victims Restitution Act of 1996 (“MVRA”), 18 U.S.C. § 3663A. Each party submitted various proposals to the district court regarding the restitution calculation. The court adopted one of the government’s proposals, and ordered Berger to pay $3,144,832 in restitution to the lending banks. Under this approach, the court first considered thirteen falsified Borrowing Certificates, including eleven that were not charged in the underlying indictment. Based on those falsified Borrowing Certificates, the lending banks had advanced Craig Electronics $4,976,000. The court discounted that total by the percentage of funds recovered by the lending banks when they foreclosed on Craig Electronics’ collateral — the lending banks recovered $4,944,180 of an outstanding $13,331,-403 — or 36.2%. Thus, according to the district court, the lending banks’ loss attributable to the false information was $3,144,832, or 63.8% of $4,976,000.

On appeal, Berger reasserts the position he took before the district court, that the lending banks suffered no losses traceable to the fraud. We disagree and affirm the district court’s restitution calculation.

1. Berger’s fraud directly and proximately caused the lending banks’ losses

The MVRA requires a defendant to pay restitution to a victim who is “directly and proximately harmed as a result of’ the fraud. 18 U.S.C. § 3663A(a)(2). Berger argues that the thirteen Borrowing Certificates at issue did not cause any of the losses the lending banks suffered when Craig Electronics defaulted on its loan. We conclude, however, that the district court did not commit clear error when it determined that the fraudulent Borrowing Certificates caused at least some of the losses suffered by the lending banks.

a. The district court did not err when it interpreted the Credit Agreement

Berger’s main contention is that the district court improperly attributed $3.1 million of the lending banks’ loss to the fraudulent Borrowing Certificates submitted by Craig Electronics. Berger bases this contention on the structure of the Credit Agreement, which both parties recognize as a “revolving” line of credit. Under the Credit Agreement, the amount of money Craig Electronics was able to borrow on any given day was determined by the amount of eligible inventory and accounts receivable Craig Electronics had on that day. To maintain the agreement, Craig Electronics would continuously: (1) borrow money from the lending banks based on the Borrowing Certificates, (2) purchase goods overseas with the money, (3) sell the goods in the United States, and (4) use the proceeds from the sales to pay off the loan.

The thirteen Borrowing Certificates at issue were submitted to the lending banks between April 30, 1996 and January 6, 1997. Craig Electronics defaulted on the loan on May 21, 1997. Berger argues that because of the revolving cycle, by the time the lending banks foreclosed on the loan, any effect the fraudulent Borrowing Certificates had on the amount of the loan was reduced to nothing. He notes that the equivalent of each loan advancement was paid off about sixty days after each loan advancement was made. Thus, he argues that Craig Electronics paid off the balance of each “tainted” loan advancement long before Craig Electronics defaulted. Berger contends that because Craig Electronics had stopped submitting falsified Borrowing Certificates in early January 1997— while continuing to submit truthful Borrowing Certificates and paying back money for the remainder of the loan’s life — no bank loss could be attributed to the fraudulent Borrowing Certificates.

The government maintains that the district court’s approach was proper. The government’s accountant explained that the Credit Agreement was better understood if the Borrowing Certificates were analyzed together: “[E]ach borrowing base certificate reported an aggregate amount of Craig’s outstanding loan; each day’s advance was added to this aggregate amount; and the amounts collected from Craig’s accounts receivables in the ‘lock box’ were applied to the outstanding loan.” Under this interpretation, the court could not consider any loan advancement fully paid back “until the aggregate outstanding amount equaled zero.” Thus, because Craig Electronics never paid back the full balance of the loan, the remaining outstanding debt included some loan advancements that had been “tainted” by the false statements.

On this record, we conclude that the government’s interpretation, adopted by the district court, is more reasonable. By overstating the assets and accounts receivable in thirteen Borrowing Certificates, Craig Electronics received loan advancements greater than that to which it was entitled. Thus, the amount of the outstanding debt at the time of Craig Electronics’ default was greater than it would have been had the false statements never been made. Accordingly, we hold that the district court did not err when it viewed the Credit Agreement in its broader structure, rather than as smaller component cycles.

b. Craig Electronics’ false statements significantly affected the lending banks’ loans

Berger also takes issue with the district court’s decision regarding the effect each false statement had on the final loan amount. For each of the thirteen days Craig Electronics submitted a false Borrowing Certificate, the court added the entire amount of that day’s loan advancement to the final “net loss attributed to fraud” figure. Berger asserts that such a calculation was erroneous because the court failed to correlate specifically a loss amount with each fraudulent statement made on the Borrowing Certificates at issue. Berger poses the hypothetical that if he “misreported one $50 VCR on a[Borrowing Certificate] requesting a $5 million advance and the loan goes into default the following year, [he] cannot be ordered to repay the $5 million as ‘restitution’ simply because he made a $50 false statement.”

Berger overstates his case. Nowhere in the record did such an extreme situation arise. Each overstatement on the Borrowing Certificates had a substantial impact on the overall loan amount. In fact, during some weeks, Craig Electronics would not have been eligible for any loan advancements had Berger not overstated Craig Electronics’ assets. For example, in the April 30, 1996, Borrowing Certificate, Berger improperly classified $392,191 worth of “C” goods as “B” goods, bringing the eligible “B” inventory to $1,508,725.89, a 35% overstatement of inventory. Craig Electronics’ loan advancement for that week was $585,000. Had Craig Electronics truthfully reported its assets for that week, Craig Electronics would have been

ineligible to receive any loan advancements. Moreover, Berger fails to provide any alternative method for separating the amount of the loan advancement obtained by fraud from the amount that, hypothetically, was not fraudulently obtained. We hold that the district court did not err by holding that Berger’s fraud contributed significantly to the losses suffered by the lending banks.

c. Berger’s supporting case law is inap-posite

Berger relies on a number of cases that deal with false statements made during the course of the loan as opposed to the initial loan application. Those cases hold that restitution is not appropriate if the losses “would have occurred despite the ... reliance on the false reports.” United States v. Diamond, 969 F.2d 961, 967 (10th Cir. 1992); see also United States v. Meksian, 170 F.3d 1260, 1263 (9th Cir.1999) (“[T]he main inquiry for causation in restitution cases becomes whether there was an intervening cause and, if so, whether this intervening cause was directly related to the offense conduct.”); United States v. Wilson, 980 F.2d 259, 262 (4th Cir.1992) (“[I]n the event a bank loan legitimately is obtained by one who subsequently submits a statement that is required in connection with the loan and that statement is false ..., the loss under [the Sentencing Guidelines] is the loss that can be attributed to the false statement.”). In each of these cases, courts held restitution to be inappropriate because the defendant’s fraud had been unrelated to the loss. See Meksian, 170 F.3d at 1263 (holding “contaminated nature of the loan property” to have caused loss, not defendant’s false state ments); Diamond, 969 F.2d at 966-67 (noting that where loan was initially valid, but subsequent false statements were made, false statements did not actually cause loss); see also Wilson, 980 F.2d at 262 (remanding for reconsideration of amount of loss directly caused by fraud).

These cases are factually distinct from this case. In Meksian , an unanticipated intervening environmental cause, rather than the alleged fraud, was to blame for the loss. See Meksian, 170 F.3d at 1263 . In Diamond , the fraud occurred after the loan and was, more importantly, separate from the “original debt.” Diamond, 969 F.2d at 966 . Finally, in Wilson , the losses began soon after the loan was made and those losses “were not related ... to the criminal conduct,” which occurred seven months after the loans were made. Wilson, 980 F.2d at 262 .

Most importantly, each of these cases dealt only with a single decision by a lender to lend a fixed amount of money. The instant case is more complex, and the structure of the loan calls for closer scrutiny. Craig Electronics’ loan amount was in constant flux, varying with respect to the inventories and accounts receivables reported in the Borrowing Certificates. The information in the Borrowing Certificates was vital; fraudulent overstatement would, and did, cause the lending banks to advance more money than was actually due. Berger’s citation to inapposite case law is unconvincing.

d. Intervening factors did not cause the lending banks’ losses

Finally, Berger asserts that unrelated financial difficulties- — not the falsified Borrowing Certificates — caused the loan’s failure. Berger argues that the downturn in the electronics market and Craig Electronics’ financial troubles were the reason that Craig Electronics could no longer pay off its loan obligations. This court has held that the “main inquiry for causation in restitution cases [is] whether there was an intervening cause and, if so, whether this intervening cause was directly related to the offense conduct.” United States v. Gordon, 393 F.3d 1044, 1055 (9th Cir.2004). Berger is probably correct to note that Craig Electronics’ financial troubles ultimately led to the loan default. He has not demonstrated, however, that external factors were completely to blame for the lending banks’ losses. Even if the default were inevitable, the high amount of the lending banks’ losses was not. Had Berger not fraudulently stated Craig Electronics’ assets on the Borrowing Certificates, the amount of outstanding debt would have been smaller. That fact was taken into account by the district court, which did not attribute the lending banks’ entire loss to the fraud. Instead, the district court properly focused only on the amount of loss attributable to the falsified Borrowing Certificates.

For the above reasons, we conclude that it was not clearly erroneous for the district court to determine that Berger’s fraudulent statements caused, at least in part, the lending banks to suffer a loss.

2. The district court’s approach was reasonable even if it used a repealed Sentencing Guideline

Berger also argues that the district court erred when it applied a repealed Sentencing Guideline — Sentencing Guideline § 2F1.1, comment 7(b) — in deter mining the amount of restitution. That Guideline, repealed in 2001, dealt specifically with fraudulent loan applications and the calculations of corresponding loss. See U.S.S.G. § 2F1.1 cmt.7(b) (repealed). The district court, however, did not indicate that it was bound by the Guideline. In fact, the court explicitly stated that it was bound by the MVRA with respect to the imposition of restitution. The court only referenced the Guideline, which provided a method for fashioning a loss calculation in the restitution context. The district court did not abuse its discretion when it consulted the repealed Guideline section.

The district court’s restitution calculation method was, in light of the circumstances, quite reasonable and fair. In calculating the $8.1 million restitution award, the district court devised a reasonable formula for determining the amount of the lending banks’ loss that was attributable to the fraud. The district court did not order restitution for the entire $13.3 million of the lending banks’ losses. Instead, it based its restitution calculation only on the amount of the loan advancements received due to fraudulent Borrowing Certificates. Even then, the district court discounted by the percentage the lending banks recovered when they foreclosed on Craig Electronics’ collateral. To achieve the goal of making crime victims whole, the MVRA permits “district courts to engage in an expedient and reasonable restitution process, with uncertainties resolved with a view toward achieving fairness to the victim.” Gordon, 393 F.3d at 1048 . We hold that, in light of the MVRA’s goals and the complexities of the Credit Agreement, the district court did not abuse its discretion in its restitution order.

V. Government’s Sentencing Cross-Appeal

The government cross-appeals Berger’s six month sentence and requests a remand under United States v. Booker, 543 U.S. 220 , 125 S.Ct. 738 , 160 L.Ed.2d 621 (2005). Sentencing in this case occurred on September 13, 2004, before the Supreme Court issued the Booker decision, which rendered the Sentencing Guidelines advisory. The district court erroneously assumed that, under Blakely v. Washington, 542 U.S. 296 , 124 S.Ct. 2531 , 159 L.Ed.2d 403 (2004), it could not apply Guideline sentencing enhancements based on facts not previously found by the jury. See Booker, 543 U.S. at 267-68 , 125 S.Ct. 738 . Additionally, the district court sentenced Berger under the assumption that the Sentencing Guidelines were mandatory. Accordingly, we must vacate the sentence and remand for resentencing. See United States v. Labrada-Bustamante, 428 F.3d 1252, 1266 (9th Cir.2005).

Conclusion

For the above reasons, we AFFIRM the conviction; AFFIRM the restitution order; VACATE the sentence and fine; and REMAND for resentencing.

. Some cases have separated the second factor into two distinct factors: (1) the amount of time of deliberation following the charge and (2) the total time of deliberation. See Weaver, 197 F.3d at 366 . Becáuse those two considerations must be compared with each other, we find that it is more helpful to discuss them together. . Federal Rule of Criminal Procedure 43(a) provides a similar but broader right: a defendant is entitled to be present "at every stage of the trial including the impaneling of the jury and the return of the verdict.” See Rosales-Rodriguez, 289 F.3d at 1109 . A violation of Rule 43 is harmless if "there is no reasonable possibility that prejudice resulted from the absence.” Id. As this court has recognized, the constitutional harmless error standard is "more stringent" than the harmless error standard applicable to a Rule 43 violation. Frazin, 780 F.2d at 1469 n. 8 ("The reasons supporting our determination that the constitutional error was harmless ... would necessarily satisfy the more lenient standard for nonconstitutional error....”). Because we ultimately conclude that the constitutional violation in this case was harmless beyond a reasonable doubt, we need not decide whether Berger's statutory right to be present was also violated. See id. . We refer to the relevant statutory provisions according to their Sections under the 1934 Act, rather than the United States Code section. In other words, we refer to Section 13(a) for 15 U.S.C. § 78m(a), Section 32(a) for 15 U.S.C. § 78ff(a), Section 14(a) for 15 U.S.C. § 78n(a), and Section 10(b) for 15 U.S.C. § 78j(b). . Berger contends that Congress already has provisions such as Sections 10(b) and 14(a) in place to protect investors from false statements. He implies that interpreting Section 32(a) as we do today would be duplicative of those provisions. This is inconsequential, however, because "when an act violates more than one criminal statute, the Government may prosecutef] under either so long as it does not discriminate against any class of defendants.” United States v. Batchelder, 442 U.S. 114, 123-24 , 99 S.Ct. 2198 , 60 L.Ed.2d 755 (1979); see also United States v. Pope, 189 F.Supp. 12, 21-22 (S.D.N.Y.1960) (holding permissible charges under both Section 14(a) and Section 32(a)). . Berger did not raise this issue in his opening brief. "Generally, an issue is waived when the appellant does not specifically and distinctly argue the issue in his or her opening brief.” United States v. Kama, 394 F.3d 1236, 1238 (9th Cir.2005). Nevertheless, we have the discretion to review issues not previously raised if "the issue presented is purely one of law and either does not depend on the factual record developed below, or the pertinent record has been fully developed.” Pfingston v. Ronan Eng'g Co., 284 F.3d 999 , 1004 (9th Cir.2002). Because this is a pure question of law and the relevant record has been fully developed, we exercise our discretion to consider this issue. . A person convicted under Section 24 of the 1933 Act may be fined up to $10,000 and imprisoned up to five years. See 15 U.S.C. § 77x. A person convicted under Section 32(a) of the 1934 Act faces potentially harsher penalties: up to $5,000,000 in fines and up to twenty years imprisonment. See 15 U.S.C. § 78ff. As discussed below, we remand this case for resentencing. One should be mindful of the relevant statutory maximum penalties under Section 24 of the 1933 Act when considering the appropriate sentence for Berger pursuant to his conviction under Count 34. See, e.g., Hockenberry, 422 F.2d at 174 . . The SEC itself "is charged with the duty of enforcing the [1933 Act] in the 'public interest.”' Berko v. SEC, 316 F.2d 137, 141 (2d Cir.1963). . It is only by coincidence that the amount advanced by the lending banks on the fraudulent Borrowing Certificates, $4,976,000, is similar to the amount recovered by the lending banks, $4,944,180. . The government makes an insightful point: If defendant’s analysis were correct and repayments of advances had to be computed against individual [Borrowing Certificates] rather than against the total balance of loans outstanding, then misstatements of collateral in revolving lines of credit would always result in no loss for restitution pur poses, as long as the false statements at issue were made long enough before the fraud was uncovered. It would be unwise to interpret the Credit Agreement in a way that would reward Berger for the lending banks’ failure to uncover the fraud until months after the fraud had been perpetrated. . The Guideline states, in relevant part: In fraudulent loan application cases ... the loss is the actual loss to the victim.... For example, if a defendant fraudulently obtains a loan by misrepresenting the value of his assets, the loss is the amount of the loan not repaid at the time the offense is discovered, reduced by the amount the lending institution has recovered (or can expect to recover) from any assets pledged to secure the loan. U.S.S.G. § 2F1.1 cmt.7(b) (repealed). . The remand would not apply to the restitution order, which derives from the MVRA, a separate statute. See United States v. DeGeorge, 380 F.3d 1203, 1221 (9th Cir.2004). . Both parties agree that we should remand for full reconsideration of the fine imposed. In light of our full remand as to the term of imprisonment, we find it appropriate to vacate the fine and remand for the district court to consider the statutory factors enumerated in 18 U.S.C. § 3553 (a) and the Sentencing Guideline factors listed in U.S.S.G. § 5E1.2(d).