Thomas Herman, representative for a class of investors in Legent Corporation (Legent), appeals the district court's decision granting judgment as a matter of law to Legent and several of its corporate officers in this "fraud on the market" securities fraud class action. See Bentley v. Legent Corp., 849 F. Supp. 429 (E.D. Va.1994) (lower court opinion). Appellants allege that Legent made a series of fraudulent public statements about its past and future performance that inflated the value of Legent's stock over a six-month period. For the reasons discussed below, we agree with the district court: (1) that the statements projecting future performance were not fraudulent; (2) that Legent's statements of past performance were not false and were not made with scienter or with an intent to defraud; and, (3) that Legent was not liable for the statements of an independent market analyst because those statements could not legally be attributable to Legent. Because we agree that Appellants presented insufficient evidence of securities fraud for a reasonable jury to rule in their favor, we affirm the judgment of the district court.

Legent is an international systems software company based in Virginia that sells computer software and services for the management of information systems. In the past, Legent has performed so well that it has established a reputation as a "growth" company. To help the market stay informed of its growth, Legent disseminates information about the company's performance, called "guidance," through periodic press releases and conference calls with market analysts.1 Based in part on this guidance information, analysts attempt to predict Legent's anticipated market performance. These predictions are known as "street expectations."

Legent did not publicly reveal its budget figures, including its annual and quarterly revenue estimates, or its other internal reports to the market. Only senior management received these internal monthly reports so that they could track and forecast its financial performance on a monthly and quarterly basis. For example, one type of these confidential reports used for planning and for operating the company was the Corporate Forecast Report (CFR), which showed estimated revenues and earnings forecasts for the current month and quarter. The internal reports, however, did not account for certain large deals that sometimes closed during a late surge in sales at the end of each quarter, as is common in the software systems business. The resulting sharp increase in revenues is known as the "hockey stick effect" because a graph of quarterly sales revenues resembles a slightly inclined hockey stick. The district court made an unchallenged finding that "[t]he market was well aware that companies like Legent are subject to this effect." Bentley, 849 F. Supp. at 432.

At its annual meeting with market analysts in August 1992, Legent revealed its objective of 20% revenue growth "for 1993 and beyond."2 (J.A. 604.) During January, April, and June of 1993, Legent made a series of optimistic--and allegedly fraudulent--guidance statements to analysts. Its internal reports during the second and third quarters, however, showed a decline from its budgeted performance in earnings per share and revenues. After the close of the market on July 12, 1993, Legent publicly revealed its poor performance and lower-than-expected revenues for the third quarter. Its stock fell ten points the next day.

On July 13, 1993, several independent stockholders filed suit, alleging a series of fraudulent misstatements of projected future earnings and of past performance, in violation of Sec. 10b of the Securities & Exchange Act of 1934, 15 U.S.C.A. Sec. 78j(b) (West 1981), and SEC Rule 10b-5, 17 CFR Sec. 240.10b-5 (1993). In August, these suits were consolidated into a class action representing those who purchased Legent stock between January 27, 1993, when Legent first made public statements allegedly inflating the price of its stock, and July 12, 1993, when Legent revealed the decline in its third quarter performance. While granting partial summary judgment to Legent, the district court denied a major portion of Legent's motion for summary judgment because it found, based on documentary evidence and deposition testimony, that Appellants had raised a genuine issue of material fact as to whether Legent's internal documents actually showed that Legent was not performing as well as management implied in its public guidance statements.

At trial, after the close of Appellants' case, the district court granted Legent's motion for judgment as a matter of law.3 The district court found that Appellants' evidence of securities fraud was insufficient to put the issue before the jury because the statements made by Legent's officers were true, were couched in cautionary terms, were made in good faith, and had a reasonable basis in fact when they were made. Bentley, 849 F. Supp. at 432-33. The court also found that Legent's projections of future performance were not "guarantees," and therefore not actionable material statements under the securities laws. Id. The district court further found that Legent made no statements about its past performance because the statements concerned the growth of the company and therefore were projections of future performance which were not actionable. Id. at 432. Finally, the district court found that Legent was not liable for the market analyst's statements because they were not attributable to Legent. Id. at 432-33. This appeal followed.4

Appellants argue that the district court erred in granting Legent's motion for judgment as a matter of law. Rule 50(a) provides that the district court may grant a motion for judgment as a matter of law in actions tried by a jury if "a party has been fully heard ... and there is no legally sufficient evidentiary basis for a reasonable jury to find for that party." Fed.R.Civ.P. 50(a)(1). We review the grant of a Rule 50(a) motion de novo. Malone v. Microdyne Corp., 26 F.3d 471, 475 (4th Cir.1994). Furthermore, on review of a judgment as a matter of law, we must view the evidence and the inferences therefrom in the light most favorable to Appellants as non-movants. Id. at 472 n. 1. A mere scintilla of evidence, however, is insufficient to overturn the judgment, and the inferences drawn in favor of Appellants as the opposing party must be reasonably probable. See Lust v. Clark Equipment Co., 792 F.2d 436, 437 (4th Cir.1986).

The primary issue on appeal is whether Appellants presented sufficient evidence of actionable material statements for a reasonable jury to rule in their favor. Liability under Sec. 10(b)5 and Rule 10b-56 requires the plaintiff to prove that: "(1) the defendant made a false or misleading statement of material fact (2) with scienter (3) upon which the plaintiff justifiably relied (4) that proximately caused the plaintiff's damages." Malone, 26 F.3d at 476. Under a "fraud on the market" theory, however, the element of direct justifiable reliance need not be proved.7 Id. at 476 n. 5.

Rule 10b-5 prohibits misstatements of material facts or omissions of those material facts necessary to make the statements made not misleading. In Walker, we indicated that a company's duty to disclose financial projections depends on the surrounding circumstances, and noted that "if a company undertakes projection disclosures, it must make the full disclosure necessary to avoid making the statements misleading." Walker v. Action Indus., 802 F.2d 703, 710 (4th Cir.1986), cert. denied, 479 U.S. 1065 (1987). For an omission to violate Rule 10b-5, there must be a "duty to speak," and "if such a duty exists, only omissions of material facts are actionable." Id. at 706 (footnote omitted) (emphasis added). " 'An omitted fact is material if there is a substantial likelihood that a reasonable shareholder would consider it important in [making his decision].' " Id. at n. 6 (quoting TSC Industries v. Northway, Inc., 426 U.S. 438, 449 (1976)).8 Otherwise stated, "there must be a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the 'total mix' of information made available." TSC Industries, 426 U.S. at 449. Finally, in determining whether omission of a material fact makes a statement fraudulent under Rule 10b-5, we examine the statement in light of the circumstances under which it was made. See Rule 10b-5(b), 17 C.F.R. Sec. 240.10b-5(b).

To support the argument that Legent's statements were fraudulent, Appellants contend that evidence from Legent's internal documents was sufficient to allow a reasonable jury to conclude that, contrary to the trial testimony, Legent's management did not believe its public statements, either when they were made or when subsequent information developed that shed doubt upon them. Whether the jury should be allowed to make such an inference depends on the reasonableness of the inference:

Permissible inferences must still be within the range of reasonable probability, however, and it is the duty of the court to withdraw the case from the jury when the necessary inference is so tenuous that it rests merely upon speculation and conjecture.

Ford Motor Co. v. McDavid, 259 F.2d 261, 266 (4th Cir.), cert. denied, 358 U.S. 908 (1958).

Appellants call our attention to three different types of statements that allegedly were materially false or misleading under Sec. 10(b) and Rule 10b-5: (1) statements predicting future performance; (2) statements revealing past or historical performance; and, (3) statements made to an independent market analyst. We address each in turn.

We have recently addressed whether public predictions of future performance are actionable as securities fraud:

Misstatements or omissions regarding actual past or present facts are far more likely to be actionable than statements regarding projections of future performance. Generally, the latter will be deemed actionable under Sec. 10(b) and Rule 10b-5 only if they are supported by specific statements of fact or are worded as guarantees.

Malone, 26 F.3d at 479 (emphasis in original) (citing Raab v. General Physics Corp., 4 F.3d 286, 289-90 (4th Cir.1993)); see also Hillson Partners Ltd. Partnership v. Adage, Inc., 42 F.3d 204, 212 (4th Cir.1994).

Statements which involve optimistic opinions or predictions of future performance, rather than representations of fact, are a type of sales talk or "puffing." " 'Soft,' 'puffing' statements such as these generally lack materiality because the market price of a share is not inflated by vague statements predicting growth." Raab, 4 F.3d at 289.9 A reasonable investor would not rely on a discussion of growth which was nothing more than a "loose prediction." Id. at 290. Once a statement is determined to be commonplace commercial puffery, our inquiry ends. We do not need to consider the remaining Rule 10b-5 elements because such a statement is not material, and thus is not actionable as a matter of law. Id. at 289 n. 1.

We now review each of these statements of future performance to determine whether a reasonable jury could find that they constitute federal securities fraud.

On January 26th, John F. Burton, Legent's Chief Executive Officer, distributed a "confidential and urgent" memo to Legent's management indicating that first quarter "earnings per share was achieved [primarily] through expense savings," as opposed to revenue growth. (J.A. 2537.) The memo also noted that "unless a major change occurs, we will undershoot the [second quarter] revenue budget in all areas." (Id.) The memo was not predominantly concerned with overall first quarter revenues, but rather with revenues in certain business lines that "was far short of budget" for Legent's North American and international markets. (Id.) The memo instructed top level management to reassess their revenue forecasts for the second quarter and to incorporate a hiring freeze and other expense-saving measures. It further warned that "if we are not extraordinarily careful and expeditious in our action, we will quickly find ourselves in a position which we cannot rectify." (Id.) Yet, the next day in its January 27th press release, Legent quoted CEO Burton as stating that Legent's first quarter performance indicated it was "firmly on its 1993 plan for revenue and earnings growth."10 (J.A. 2613.) The press release announced a 20% increase in total first quarter revenues and indicated that the net income from the three months of the first quarter showed earnings per share of $0.54. During a conference call with analysts the same day, Burton again stated that Legent experienced a 20% revenue growth during the first quarter and noted that $0.05 of the $0.54 earnings per share was due to a rollover of the recently acquired Goal System International's "thirteenth month" (September 1992) earnings into Legent's first quarter earnings.11 Near the close of the conference call, in response to a request for guidance on earnings, Burton stated that

with the earnings that we are announcing, it gives us even greater confidence in what the average is of the street expectations. We feel even more confident that we will comfortably make those estimates.

(J.A. 1586.)

We discussed in Cooke v. Manufactured Homes, Inc., 998 F.2d 1256 (4th Cir.1993), the type of factual assertion necessary to make a future prediction material, and thus actionable if false or misleading. In Cooke, the projections of future prosperity were supported by specific statements of fact representing specific business projects. For instance, the Cooke defendant asserted that the company would repurchase 400,000 shares of its own stock, and that it was negotiating with an insurance company to act as a guarantor on its loans. Id. at 1259.

Burton's statements--that Legent was firmly on its plan for earnings growth and that Legent would comfortably make the street expectations for annual earnings per share--simply lack the necessary level of specificity. All we discern in the January 27th predictions are the type of "soft," "puffing" statements that we found not actionable as a matter of law in Malone. Malone, 26 F.3d at 479-80; see also Raab, 4 F.3d at 289 (such statements "lack materiality because the market price of a share is not inflated by vague statements predicting growth"). No reasonable jury could find that these statements, either individually or taken as a whole, constitute a guarantee; nor could a reasonable jury find that such statements are factually specific enough to perpetuate a fraud on the market. Therefore, as a matter of law, these statements are not actionable as fraud. Malone, 26 F.3d at 479-80.12

A few months later, during an April 13th conference call, Burton noted that the company was "on a slower track than we might have expected or wanted to be at this time," and announced second quarter revenues of $102 million and earnings per share of $0.52, which was "at the very low end." (J.A. 1595.) He stated that "we remain comfortable with street expectations" for an annual earnings per share of $2.45 and that "the pipelines [of unclosed deals] give us comfort that the expected annual earnings should be achieved." (J.A. 1600.) Burton noted that revenues were lighter than expected in the second quarter, but indicated that this was due to the failure to close approximately ten large deals worth $1 million to $9 million each that were "in the pipeline." (J.A. 1635.) He also indicated that the unclosed deals were "highly-qualified contracts" but were "not necessarily forecasted." (J.A. 1601.)

First, Appellants argue that the "comfort" statement was fraudulent because Legent was having difficulty meeting its internally forecasted estimates of earnings and revenues. However, the "comfort" statement here is much like the one we held not actionable as a matter of law in Malone. In Malone, we held that a February 1992 statement by Microdyne's president "that he was 'comfortable with [an independent analyst's] earnings estimates for fiscal 1992 and 1993' " was not actionable as a matter of law because it "obviously did not constitute a guarantee and was certainly not specific enough to perpetrate a fraud on the market." 26 F.3d at 473, 479-80. We noted "a clear distinction between a company's strict 'duty to accurately report ... its past results,' on the one hand, and the company's relative freedom to 'prognosticat[e]' or make 'predictions of its future business prospects,' on the other hand." Id. at 479 (quoting Raab, 4 F.3d at 287, 289) (emphasis in Malone ). By the same token, Burton's statement that Legent was "comfortable" with expectations that annual earnings would be met is also an unactionable future prediction.

Second, Appellants contend that characterizing the large deals as "highly-qualified" implied that the deals would close during the third quarter, and therefore constituted fraud. We disagree. There was no guarantee or assurance here that the "highly-qualified" deals would close during the third quarter. Indeed, Burton indicated at the same time that the deals were not "forecasted."13 We do not believe that a reasonable investor would rely on such soft predictions that simply are not supported by specific statements of fact or expressed as guarantees. They, therefore, do not rise to the level of material statements of fact necessary to support a claim of securities fraud. Id.; Raab, 4 F.3d at 290.

Because of increasing discrepancies between the October budget forecast and internal monthly forecasts during the second quarter, on April 26, 1993, Legent's Chief Financial Officer, Franchon Smithson, with input by Legent's Treasurer, James Buchanan, prepared an internal document entitled "Legent Forecast Range Estimates" that revised the October budget forecast numbers. This revised internal forecast gave a target range for third quarter, fourth quarter, and fiscal year 1993 revenues, as well as earnings per share, that were slightly lower than the October budget forecast figures. Legent's management testified at trial that thereafter it measured the company's performance against the April revised forecast estimates.14 Furthermore, during an April 27th conference call, Legent publicly revealed its revised estimates for third quarter and annual earnings per share.

During the April 27th conference call, Burton estimated a guidance figure for third quarter earnings per share "in the 45-to 55-cent area, giving a 10-cent swing one way or another," and stated that Legent believed its predicted annual earnings per share for 1993 would "be in the 2.25 to 2.50 range--two-and-a-quarter to two-dollars-and-fifty-cents." (J.A. 1693-94.) He also promised "to keep [analysts] apprised of significant changes throughout the quarter." (J.A. 1701.) Burton further noted that during the second quarter, Legent's performance and its difficulty in forecasting the market had surprised him, and, as a result, he noted that Legent's accuracy and precision in forecasting was no longer as good as it had been in the past. He also reminded analysts that the fourth quarter had always been Legent's strongest quarter for new license revenues, which accounted for roughly half of Legent's annual revenues.

Viewing the April 27th predictions of quarterly and annual earnings per share in the context in which they were made, and particularly in light of the other cautionary statements made during the conference call, no reasonable jury would find that Legent's projections were expressed as guarantees or supported by specific statements of fact. Thus, the statements complained of were not of the type upon which a reasonable investor would rely, and, therefore, were not actionable under the securities laws. Malone, 26 F.3d at 479; see also Luce v. Edelstein, 802 F.2d 49, 56 (2d Cir.1986) (no securities fraud liability when statements "clearly 'bespeak caution.' ").15

In sum, we find that all of Legent's public predictions of future performance lacked the factual specificity or guarantee required to make them material. The district court committed no error in granting judgment as a matter of law on this issue.

Appellants argue that the district court should have allowed the jury to determine whether several statements concerning Legent's past performance were fraudulent under Rule 10b-5 and Sec. 10(b). To establish liability for these statements under the fraud on the market theory, Appellants must prove that they were false or misleading statements of material fact made with scienter by Legent, and that they proximately caused Appellants' damages. Malone, 26 F.3d at 476. We address each of the four allegedly fraudulent statements in turn.

As discussed above in the section on statements predicting future performance, Legent issued a press release on January 27th stating that its first quarter performance indicated it was "firmly on its 1993 plan for revenue and earnings growth." (J.A. 2613.) Appellants argue that the January 27th statement was an actionable fraudulent statement of past performance because it referred to Legent's performance to date, and evidence showed that Legent's first quarter performance did not coincide with its budget estimates.16 Appellants further contend that Legent management used the terms "plan" and "budget" interchangeably, and thus a jury could infer that when Burton said Legent was "on plan" he meant that it was "on budget."17 Therefore, if Legent's actual first quarter revenues and earnings growth figures did not meet the budget, Appellants argue that Legent must have misled its investors by stating it was "on plan." Even if we assume, however, that the "plan" referred to on January 27th was actually Legent's budget, Appellants presented no evidence that Legent's overall performance for the first quarter did not meet budget predictions.

Appellants support their argument by stating that, because the earnings and revenues from Goal's "thirteenth month" were not included in the 1993 budget, the roll-over of these funds was not part of "the plan" referred to in the January 27th press release. They contend that Legent was not "on plan" without the Goal additions. For example, Appellants point out that without Goal's revenues, first quarter new license revenues of $57.7 million was 11% short of the budgeted estimate of $64.6 million.18

However, Legent informed analysts, and therefore the entire market, that earnings and revenues from the Goal merger were incorporated into Legent's actual first quarter performance figures.19 During the January 27th conference call, Burton told analysts that the actual first quarter figures included $9 million in revenues and $0.05 in earnings per share from Goal's "thirteenth month." The budget predicted first quarter earnings per share to be $0.49; actual first quarter earnings per share without Goal's "thirteenth month" earnings were $0.49 and with Goal's earnings were $0.54. The budget predicted total first quarter revenues to be $115 million; actual first quarter revenues without the addition of Goal's "thirteenth month" was $107 million, but with the addition was $116.7 million.20 Thus, Legent's January 27th statement that it was "on plan" was true regarding its past performance for first quarter revenue growth and earnings, and, therefore, could not be fraudulent.

Thus, regardless of whether the "plan" referred to in the January 27th press release was Legent's 1993 budget or, as the district court determined, an "overall business plan," Bentley, 849 F. Supp. at 432, there simply is no evidence that the statement was false as it pertained to Legent's past performance. Because there was no evidence from which a reasonable jury could conclude that the "firmly on plan" statement was fraudulent, we affirm on this issue.

An April 23rd press release noted that, although Legent fell short of expected new license revenues and although several "particularly large orders" did not close during the quarter, "tight expense controls" helped Legent achieve its $.52 earnings per share for the second quarter. (J.A. 2614-15.) Legent also made other statements in April stating that it achieved its second quarter earnings by using "tight expense controls." Appellants argue that these statements were fraudulent statements of historical fact because Legent made representations that use of these "tight expense controls" could be relied upon again in later quarters to help meet earnings expectations. Appellants, however, refer us to no such representation, and we are unable to find one in the record.21 Moreover, no analyst testified about any market expectations that Legent would repeat its use of any particular expense controls, and there is no evidence that Legent told the market that it would use these same "tight expense controls" in order to make its third quarter earnings. As with the Goal "thirteenth month" adjustment in January, use of the April expense controls was an accepted accounting practice that was not, in itself, evidence of securities fraud under Sec. 10(b) or Rule 10b-5. See Malone, 26 F.3d at 477-78.

In response to a question from an analyst near the end of the April 23rd conference call about deals that had closed in Europe during April, Burton incorrectly responded that Legent had closed three deals worth a total of $2 million in the United Kingdom. Appellants contend that Burton's misrepresentation was materially false and misleading and, therefore, fraudulent. They maintained during oral argument that this statement was material because it impressed upon analysts that Legent was closing big deals, when in fact the value of the three deals closed in the United Kingdom was only $193,000 and represented a portion of an undisclosed number of deals in Europe worth a total of $2 million. Burton admitted at trial that, in response to a question at the end of the April 23rd conference call, he mistakenly attributed the $2 million worth of sales that had closed in all of Europe to three sales in the United Kingdom. Burton further testified that he was unaware he had made this misstatement until this litigation began.

We reject Appellants' argument that the $2 million statement was fraudulent, because we find no evidence of scienter, the second essential element for a private cause of action for damages under Sec. 10(b) and Rule 10b-5. Malone, 26 F.3d at 476.

In a Rule 10b-5 action, issues of intent typically go to the jury and defendants are not entitled to judgment as a matter of law "on the ground of lack of scienter unless the plaintiff has failed to present facts that can support an inference of bad faith or an inference that defendants acted with an intent to deceive," manipulate, or defraud.

Id. at 479 (quoting Wechsler v. Steinberg, 733 F.2d 1054, 1058-59 (2d Cir.1984)).

Appellants presented no evidence from which a jury could infer that Burton spoke in bad faith or intended the $2 million dollar statement to deceive the market. Thus, even assuming the statement concerned a material fact on which the analysts relied, there is no evidence of scienter. Without evidence supporting an inference that Burton knowingly spoke in bad faith or with the intention to deceive, manipulate, or defraud the market, the misstatement cannot satisfy the elements of securities fraud. Malone, 26 F.3d at 479.

Finally, Appellants contend that statements made in an independent market analyst's report concerning Legent's past performance were materially false and misleading. However, before Legent can be held liable for the analyst's statements, Appellants must first show that the analyst's report should be attributed to Legent. Raab, 4 F.3d at 288 (company cannot be held liable for statements contained in independent analyst's report absent proof that company exercised sufficient control over report to attribute report's statements to the company).

In Raab, we found that General Physics was not liable for statements in a report by Goldman Sachs because the allegations were insufficient to show that "General Physics exercised the kind of control over the Goldman Sachs report that would render it liable for statements made therein." Id. at 289. We noted that:

The securities laws require General Physics to speak truthfully to investors; they do not require the company to police statements made by third parties for inaccuracies, even if the third party attributes the statement to General Physics. Without control over Goldman Sachs' report, any statement made by General Physics personnel could be taken out of context, incorrectly quoted, or stripped of important qualifiers.

Id. at 288. We indicated that, in order to be liable for statements in an analyst's report, a company must " 'sufficiently entangle[ ] itself with the analyst['s] forecasts to render those predictions "attributable to it." ' " Id. (citing Elkind v. Liggett & Myers, Inc., 635 F.2d 156, 163 (2d Cir.1980)). As further explained in Elkind,

We have no doubt that a company may so involve itself in the preparation of reports and projections by outsiders as to assume a duty to correct material errors in these projections. This may occur when officials of the company have, by their activity, made an implied representation that ... they have reviewed [the outsider's information and determined that it] is true or at least in accordance with the company's views.

Elkind, 635 F.2d at 163.

On June 14th, Scott Smith, an analyst with Donaldson, Lufkin, & Jenrette who had participated in the April conference calls, visited Legent's corporate headquarters and spoke with Franchon Smithson (Senior Vice President and Chief Financial Officer), David Wetmore (Executive Vice President and Chief Operating Officer), and Bill Drummey (head of North American Sales). Smith released a report on June 16th revealing Legent's "increased confidence in its ability to achieve earnings expectations of $0.45-0.55" for the third quarter and stating that "[m]anagement noted that business has been on target for April and May." (J.A. 2638.) Apparently this report was circulated among other market analysts, although the report was marked "DLJ--CONFIDENTIAL INFORMATION FOR INTERNAL USE." (J.A. 2638.) Smith testified that the report was an amalgamation of information given to him by Wetmore, Smithson, and Drummey. He further testified that he could not specifically remember who provided him with what information, but that his contemporaneous notes indicated that Smithson had told him "pipeline strong, sales so far on target."22 (J.A. 1441.) Smithson, however, testified he did not tell Smith that Legent was "on target" or that Legent had "increased confidence" in its ability to earn 45 to 55 cents a share in the third quarter. (J.A. 861-864.)

Appellants contend that a jury should decide whether the statements in Smith's report were material, and also whether they are attributable to Legent because (1) Legent provided Smith with the information in his report and (2) Smithson reviewed the report several days after its release but made no effort to correct the statements therein. They contend that Legent knew the report was fraudulent because by the time of Smith's visit, Legent's internal documents revealed that it could not make its third quarter earnings per share estimates,23 and therefore it had a duty to correct the report.24

We find no evidence that Legent sufficiently entangled itself with Smith's report to make his statements attributable to the company, or that it "place[d] its imprimatur, expressly or impliedly," on Smith's report. Elkind, 635 F.2d at 163. Smith's visit to Legent was unlike the conference calls, in which Legent released information to analysts in order to promote "much more concise and precise[street] estimates and guidance." (J.A. 1695.) Indeed, the fact that Legent used conference calls to ensure that all of the market analysts heard the same information at the same time works against any inference that we should attribute Smith's report to Legent.

The Smith report did not directly quote Legent's management or reveal who supplied the information to Smith. There is no evidence that Legent had any control over the contents of the report, or, even if Legent had such control, that the report was so inconsistent with Legent's internal estimates that it would require Legent to correct the statements in the report. Unlike the conference calls, which are transcribed in the record, there is nothing in the record from which to determine whether "any statement by [Legent] personnel [was] taken out of context, incorrectly quoted, or stripped of important qualifiers" in Smith's report. Raab, 4 F.3d at 288. Although Smith's report undoubtedly expressed his impressions of his visit with Legent's management, we find no evidence from which a reasonable jury could infer that Legent had sufficient control over Smith's report to hold it liable for the statements made therein. Thus, we affirm the judgment as a matter of law against the claims arising out of Smith's June 16th report.

In sum, none of Appellants' assertions of fraudulent conduct are borne out in the evidence. Legent's future projections were not supported by specific statements of fact or worded as guarantees, as is generally required to be actionable under the securities laws. The statements taken as a whole were cautionary and did not suggest the high degree of reliability required for predictions of future growth to be fraudulent. As we most recently noted, "imposing liability on companies for predictions of future growth, which are often and inevitably wrong, would lead to the further proliferation of lawsuits and would be contrary to the 'goal of full disclosure underlying the securities laws.' " Hillson, 42 F.3d at 220 (quoting Raab, 4 F.3d at 290) (footnote omitted). Furthermore, no evidence indicated that Legent's past projections were false or were made with scienter, or that a jury could attribute the analyst's report to Legent. Based on the foregoing analysis, we affirm the judgment as a matter of law.25