LYNCH, Circuit Judge.

Bond purchasers brought suit alleging violations of federal securities laws in the May 1998 offering of bonds of Bradford College in Massachusetts. In January 2000, the college defaulted on its bond obligations. This suit was brought ten months later. The district court dismissed the amended complaint for failure to meet the pleading standards in the Private Securities Litigation Reform Act of 1995 ("PSLRA"), Pub.L. No. 104-67, 109 Stat. 737. McKeown v. Advest, Inc. (McKeown I), 2006 WL 2974154 (D.Mass. Sept.30, 2006). The district court also denied the plaintiffs' motion to vacate dismissal and plaintiffs' post-dismissal motion for leave to amend the complaint to address pleading deficiencies identified by the court. McKeown v. Advest, Inc. (McKeown II), 2006 WL 3842132 (D.Mass. Dec.29, 2006).

The plaintiffs, purchasers and an insurer of Bradford bonds sold in May 1998, claim they were misled by the Official Statement accompanying the offering, which allegedly concealed Bradford's dire financial straits and inability to pay the bond debt. The allegations are against three sets of defendants: Joseph Short and Donald Kiszka, the former President and Vice President of Administration and Finance of the College, respectively (the "Officers"); sixteen members of the College's Board of Trustees (the "Trustees"; together with the Officers, collectively the "Bradford defendants"); and Advest, Inc., the underwriter investment banking firm.

This is our first occasion to apply the Supreme Court's recent guidance regarding the standards for pleadings under the PSLRA in Tellabs, Inc. v. Makor Issues Rights, Ltd., ___ U.S. ___, 127 S.Ct. 2499, 168 L.Ed.2d 179 (2007). Under Tellabs, certain principles are clear. Tellabs has altered this circuit's prior standard, as set forth in In re Credit Suisse First Boston Corp., 431 F.3d 36 (1st Cir. 2005), for determining the sufficiency of pleadings of scienter in securities fraud cases under Rule 12(b)(6). Tellabs affirms our case law that plaintiffs' inferences of scienter should be weighed against competing inferences of non-culpable behavior. See, e.g., Greebel v. FTP Software, Inc., 194 F.3d 185, 203 (1st Cir. 1999). Tellabs also affirms our rule that the complaint is considered as a whole rather than piecemeal. See, e.g., In re Cabletron Sys., Inc., 311 F.3d 11, 40 (1st Cir. 2002). Finally, we hold that under the reasoning of Tellabs, the PSLRA does not alter the liberal amendment policy of Federal Rule of Civil Procedure 15. And we stress again our disinclination to require allowance of amendment of complaints when there has been undue delay.

We affirm the district court's denial of the plaintiffs' post-dismissal motion to allow a belated second motion to amend the complaint.

Confining our analysis of the motion to dismiss to the amended complaint, we affirm dismissal. Although one of the allegations presents an arguable claim of misrepresentation as to the college's budget for financial aid spending in the 1998-1999 academic year, the pleadings are insufficient to establish the requisite scienter.

I.

In reciting the facts as alleged, we draw all reasonable inferences in the plaintiffs' favor. Bradford was established as a coeducational academy in 1803. After going through incarnations as a female academy and junior college, Bradford became a coeducational college in 1971. By the 1990s, Bradford became mired in persistent cash flow problems. In spite of increasing enrollment and a growing operational budget, Bradford had operating deficits every year from 1989 to 1997. Boosting enrollment became a critical goal for the college because tuition, room, board, and other student fees constituted the largest source by far of its operating revenues. Bradford's administration invested in educational and physical improvements designed to attract more students, raise revenues, and attain financial stability. Bradford financed these improvements in part with a $1.5 million loan from the United States Department of Education and a $5.4 million bond offering in 1995.

There were indications in 1997 that Bradford's fortunes might be on the up-swing. In spite of admitting fewer students than in years past, the matriculation rate rose from 25% in 1996 to 34% in 1997, and the number of full-time enrolled students rose sharply in the. fall of 1997. Meanwhile, the combined value of Bradford's endowment and investment portfolio increased 66% between June 1995 and June 1997.

On February 6, 1998, the Trustees voted to approve issuing another series of bonds both to help settle its debt and to finance a project designed to increase Bradford's residential capacity and, in turn, accommodate even higher levels of enrollment. The Trustees approved the sale of $17.9 million worth of bonds, which the college's underwriter Advest determined to be the college's maximum bonding capacity. The bonds would be secured solely by a lien on tuition receipts. By entering into the bond transaction, Bradford committed to paving over $1.2 million a year in debt service through 2028.

The Massachusetts Industrial Finance Agency ("MIFA") issued the bonds rather than the college. The offering was on May 1, 1998; the offering closed when the transaction documents were executed twelve days later. An Official Statement, prepared by Advest, Short, and Kiszka, and dated May 1, accompanied the offering. The Official Statement outlined the mechanics of the transaction and contained various qualifications and disclaimers. For instance, under the heading "Bon-downers' Risks," the Statement provided that the college would be the sole source of repayment for the bonds, and disclaimed any assurance "that revenues will be realized by [Bradford] ill the amount necessary to make payments . . . sufficient to pay the debt service on the [bonds]." The Statement noted specific risks to bond-holders, including that the college's failure to meet self-proclaimed future enrollment targets could jeopardize its ability to support the debt. The Statement also noted as risk factors the college's ability to control "expenses, competition, costs, [and] the amount of financial aid awarded to students." Both the cover sheet and the main text of the Statement disclosed that Standard Poor's had assigned a "BBB-" rating to the bonds — indicating the highest level of risk short of junk bond status.

The Official Statement also incorporated a series of appendices, the first of which was a document, signed by Short and Kiszka, containing information about Bradford's operations. The document specified that the college would use a portion of the proceeds from the 1998 bonds to refund the 1995 bonds. Most of the remainder of the proceeds would be used to pay for renovations to existing residence halls and construction of new residence facilities. The renovations and construction were projected to require two years and cost almost $15 million.

In addition to providing details about the proposed project, this appendix also presented selected statistics illustrating recent trends in enrollment, financial aid, the growth of Bradford's endowment and investments, and fund-raising. A section entitled "Accounting Matters" set out the college's analysis of some of those numbers as well as targets for future enrollment and financial aid levels. Significantly, the Official Statement expressly advised that the college was not certifying the accuracy of any "projections and opinions" in the Statement.

The next appendix contained Bradford's audited financial statements for fiscal years 1994 to 1997. The remaining appendices comprised various transaction documents including opinion letters and a continuing disclosure agreement. The Official Statement (including its appendices) is discussed in further detail below.

Bradford College failed to reach its benchmarks for increased student enrollment and reduced financial aid awards for the 1998-1999 school year, despite the promising numbers from 1997. A revenue shortfall forced the college to rely on donations and over $1.5 million from the endowment. In spite of the deficit, the college granted financial aid to 90% of students in 1998-1999. Because Bradford itself funded about half of all financial aid, such high aid levels represented a significant drain on the college's cash flow.

Bradford College never recovered. President Short resigned in the summer of 1998, to be followed by Kiszka a year later. In November 1999, the college announced that it would cease operations after the 1999-2000 school year. The 1998 bonds were declared in default in January 2000.

II.

The plaintiffs brought suit in November 2000, alleging federal securities law violations and various state-law claims. The plaintiffs initially voluntarily dismissed the case without prejudice under tolling agreements. In the interim, the college went through bankruptcy proceedings.

The plaintiffs then filed the current action in July 2004 and amended their complaint in January 2005. The motion to dismiss was addressed to this amended complaint.

The first count of the amended complaint alleges violations of section 10(b) of the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and the concomitant Rule 10b-5, 17 C.F.R. § 240.10b-5, for material omissions and misrepresentations in the Official Statement. Four of the allegations in the amended complaint are pressed on appeal: that the Official Statement was deceptive (1) by avoiding reference to Bradford's high rate of student attrition; (2) by misrepresenting projected enrollment levels; (3) by falsely stating that the college intended to contribute $1 million of its own funds toward completion of the construction and renovation project; and (4) by including false statements regarding projected financial aid levels for the 1997-1998 and 1998-1999 academic years.

The sequence of proceedings proves important to the legal argument about further amendment of the complaint. Advest filed a motion to dismiss the amended complaint in February 2005; the Bradford defendants followed suit in March. The plaintiffs filed their opposition on May 2, 2005, and the district court heard oral argument on October 11, 2005. Almost a year after the hearing, on September 30, 2006, the district court granted the motions to dismiss the federal claims and declined to exercise supplemental jurisdiction over the remaining state-law claims. McKeown I, 2006 WL 2974154, at *1, *7.

The plaintiffs did not immediately appeal, instead filing simultaneous motions with the district court asking it to vacate dismissal under Federal Rule of Civil Procedure 59(e) and to grant plaintiffs leave to further amend the complaint under Rule 15(a). A Proposed Second Amended Complaint was attached to the Rule 15(a) motion. The plaintiffs argued that the district court erred in dismissing without granting leave to amend the complaint to cure pleading deficiencies pointed out by the district court. In addition, they claimed that the dismissal should be vacated on the basis of new evidence uncovered during a document review conducted by plaintiffs' counsel in June 2005.

The district court denied the post-dismissal motions on December 29, 2006. McKeown II, 2006 WL 3842132, at *1. The court disagreed that evidence discovered in June 2005 could qualify as new evidence warranting vacation of the judgment. Id. The court pointed out that the plaintiffs made no effort to amend their complaint after uncovering the "new" evidence and during the long pendency of the motions to dismiss. Id. at *1 n. 3. Lastly, the district court held that entry of judgment precluded jurisdiction to consider the Proposed Second Amended Complaint. Id. at *1.

III.

The plaintiffs argue that the district court erred by failing to apply the liberal amendment policy of Federal Rule of Civil Procedure 15(a).

The plaintiffs tie the Rule 59(e) issue to the denial of the motion to amend. Review of the denial of the Rule 59(e) motion is for "manifest abuse of discretion." Council of Ins. Agents Brokers v. Juarbe-Jiménez, 443 F.3d 103, 111 (1st Cir. 2006) (quoting Binkley Co. v. E. Tank, Inc., 831 F.2d 333, 337 (1st Cir. 1987)) (internal quotation marks omitted); see also Venegas-Hernandez v. Sonolux Records, 370 F.3d 183, 190 (1st Cir. 2004). Rule 59(e), while it authorizes post-judgment motions to alter a judgment, does not state what grounds would justify such an alteration. See Fed.R.Civ.P. 59(e); 11 Wright, Miller Kane, Federal Practice and Procedure § 2810.1 (2d ed. 1995). District courts enjoy considerable discretion in deciding Rule 59(e) motions, subject to circumstances developed in the case law. Venegas-Hernandez, 370 F.3d at 190. Correction of a clearly established "manifest error of law" is among the grounds for a valid Rule 59(e) motion. FDIC v. World Univ. Inc., 978 F.2d 10, 16 (1st Cir. 1992). A Rule 59(e) motion should not, however, "raise arguments which could, and should, have been made before judgment issued." Id. (quoting Harley-Davidson Motor Co. v. Bank of New England, 897 F.2d 611, 616 (1st Cir. 1990)) (internal quotation marks omitted). The "manifest error" that the plaintiffs allege consists of the district court's dismissal of the amended complaint without leave to amend. We turn to the motion to amend.

Rule 15(a) provides that a party may amend its pleading with "the court's leave," and that "[t]he court should freely give leave when justice so requires." Fed.R.Civ.P. 15(a). The rule reflects a liberal amendment policy, O'Connell v. Hyatt Hotels of PR., 357 F.3d 152, 154 (1st Cir. 2004), but even so, the district court enjoys significant latitude in deciding whether to grant leave to amend. We defer to the district court's decision "if any adequate reason for the denial is apparent on the record." LaRocca v. Borden, Inc., 276 F.3d 22, 32 n. 9 (1st Cir. 2002) (quoting Grant v. News Group Boston, Inc., 55 F.3d 1, 5 (1st Cir. 1995)) (internal quotation marks omitted). Grounds for denial include "undue delay, bad faith or dilatory motive . . . repeated failure to cure deficiencies by amendments previously allowed, undue prejudice to the opposing party . . . [and] futility of amendment." Foman v. Davis, 371 U.S. 178, 182, 83 S.Ct. 227, 9 L.Ed.2d 222 (1962).

Plaintiffs argue that these liberal standards remain unaffected by the PSLRA. Defendant Advest posits that allowing belated amendments would be "particularly inappropriate" in cases governed by the PSLRA, citing In re Bristol-Myers Squibb Securities Litigation, 228 F.R.D. 221, 229 (D.N.J. 2005), and In re Champion Enterprises, Inc. Securities Litigation, 145 F.Supp.2d 871, 872 (E.D.Mich. 2001). The circuit courts have expressed different views, and we have not addressed the issue before. Compare Miller v. Champion Enters., Inc., 346 F.3d 660, 692 (6th Cir. 2003) ("[T]he purpose of the PSLRA would be frustrated if district courts were required to allow repeated amendments to complaints filed under the PSLRA."), with Belizan v. Hershon, 434 F.3d 579, 583-84 (D.C. Cir. 2006) (holding PSLRA does not alter operation of Rule 15(a)), and Eminence Capital, LLC v. Aspeon, Inc., 316 F.3d 1048, 1052 (9th Cir. 2003) (per curiam) (same).

We hold that the PSLRA does not itself modify the liberal amendment policy of Rule 15(a). We agree with the D.C. Circuit that "had the Congress wished to make dismissal with prejudice the norm, and to that extent supercede the ordinary application of Rule 15(a), we would expect the text of the PSLRA so to provide." Belizan, 434 F.3d at 584. The text of the Act neither purports to affect Rule 15(a), nor does it require that all dismissals be with prejudice. See, e.g., 15 U.S.C. § 78u-4(b)(3)(A) (requiring dismissal for failure to meet the Act's pleading requirements, but not specifying dismissal with prejudice). In the absence of a legislative directive to the contrary, Rule 15 applies as in the normal course. That is one of the lessons of Tellabs. Cf. Tellabs, 127 S.Ct. at 2509 (applying Rule 12(b)(6) standard that all factual allegations in a complaint must be accepted as true in the securities fraud context "as with any motion to dismiss"); Aldridge v. AT Cross Corp., 284 F.3d 72, 78 (1st Cir. 2002) (applying usual Rule 12(b)(6) standard in PSLRA context).

Interpreting the PSLRA as constricting the operation of Rule 15(a) would be contrary to the purposes of the Act. The PSLRA serves "twin goals: to curb frivolous, lawyer-driven litigation, while preserving investors' ability to recover on meritorious claims." Tellabs, 127 S.Ct. at 2509. The heightened pleading standard furthers the goal of deterring frivolous litigation by erecting a significant hurdle for a plaintiff to clear before her complaint can survive a motion to dismiss. A blanket rule that the PSLRA modifies Rule 15(a) would tip the scales too far, compromising plaintiffs' ability to have meritorious claims presented in court. Cf. Eminence Capital, 316 F.3d at 1052 (holding that liberal amendment policy of Rule 15(a) should be adhered to precisely because the PSLRA requires "an unprecedented degree of specificity and detail"). To read the PSLRA to constrict Rule 15(a) would disturb the legislative balance struck by the Act. The number and nature of prior amendments to a complaint is relevant as to any motion for leave to amend. To the extent that Miller may embody a rule that the PSLRA modifies the operation of Rule 15(a), however, we disagree.

The parties dispute whether the more restrictive standard for post-judgment motions to amend under James v. Watt, 716 F.2d 71, 77-78 (1st Cir. 1983) (Breyer, J.), applies, or the usual Rule 15(a) standard. Here, the formal motion was not made until after dismissal. But the plaintiffs' opposition to the motions to dismiss did at least conditionally request an opportunity to amend "[i]n the event . . . that the [c]ourt does agree that the Plaintiffs have stumbled over some of the pleading intricacies in a federal securities action."

That does not win the day for the plaintiffs. Each case will turn on its own circumstances. The district court had abundant reason to deny the motion, however the request to amend is characterized. This suit was originally filed six years before the court's dismissal order and dismissed by the plaintiffs without prejudice. The complaint was re-filed in July 2004 and amended in January 2005. The defendants moved to dismiss in early 2005, identifying deficiencies in the amended complaint. The plaintiffs could have moved to amend then, but did not. Nor did they move to amend in June 2005, after plaintiffs' counsel claimed they had uncovered "new" evidence in a document review related to the college's bankruptcy proceedings. To continue the litany, the plaintiffs did not move to amend at oral argument on October 11, 2005, or before the court's ruling on September 30, 2006. Plaintiffs took no action to add new allegations even though they knew what they would add if they amended.

The plaintiffs argue that in the end, they were entitled to wait and see if their amended complaint was rejected by the district court before being put to the costs of filing a second amended complaint. They claim this would promote efficiency in the judicial system. Plaintiffs have it exactly backwards — their methodology would lead to delays, inefficiencies, and wasted work. The plaintiffs do not get leisurely repeated bites at the apple, forcing a district judge to decide whether each successive complaint was adequate under the PSLRA. Plaintiffs may not, having the needed information, deliberately wait in the wings for a year and a half with another amendment to a complaint should the court hold the first amended complaint was insufficient. Such an approach would impose unnecessary costs and inefficiencies on both the courts and party opponents. This court expressly disapproved a similar tactic in James, and we do so again. See id. at 78 ("Such a practice would dramatically undermine the ordinary rules governing the finality of judicial decisions, and should not be sanctioned in the absence of compelling circumstances." (citing 6 Wright Miller, Federal Practice and Procedure § 1489(1971))).

It is black-letter law that "[Regardless of the context, the longer a plaintiff delays, the more likely [a] motion to amend will be denied, as protracted delay, with its attendant burdens on the opponent and the court, is itself a sufficient reason for the court to withhold permission to amend." Steir v. Girl Scouts of the USA, 383 F.3d 7, 12 (1st Cir. 2004) (citing Acosta-Mestre v. Hilton Int'l of P.R., Inc., 156 F.3d 49, 52-53 (1st Cir. 1998)); see also Palmer v. Champion Mortgage, 465 F.3d 24, 30-31 (1st Cir. 2006) (approving rejection of request for leave to amend made fifteen months after commencement of action on the basis of previously available information). There was no error.

IV

A. PSLRA Pleading Requirements

We evaluate de novo whether the first amended complaint meets the pleading requirements imposed by the PSLRA. Aldridge, 284 F.3d at 78. We may affirm "on any independently sufficient ground." Ezra Charitable Trust v. Tyco Int'l Inc., 466 F.3d 1, 6 (1st Cir. 2006) (quoting Badillo-Santiago v. Naveira-Merly, 378 F.3d 1, 5 (1st Cir. 2004)) (internal quotation marks omitted).

Under the PSLRA, as with any motion to dismiss under Rule 12(b)(6), we accept well-pleaded factual allegations in the complaint as true and view all reasonable inferences in the plaintiffs' favor. Aldridge, 284 F.3d at 78; Greebel, 194 F.3d at 195-96; see also Tellabs, 127 S.Ct. at 2509. The Supreme Court has recently altered the Rule 12(b)(6) standard in a manner which gives it more heft. In order to survive a motion to dismiss, a complaint must allege "a plausible entitlement to relief." Bell Atl. Corp. v. Twombly, ___ U.S. ___, 127 S.Ct. 1955, 1967-69, 167 L.Ed.2d 929 (2007); Rodríguez-Ortiz v, Margo Caribe, Inc., 490 F.3d 92, 95 (1st Cir. 2007). The Court's formulation revised language from Conley v. Gibson, 355 U.S. 41, 78 S.Ct. 99, 2 L.Ed.2d 80 (1957), that a complaint should not be dismissed unless "it appears beyond doubt that the plaintiff can prove no set of facts" entitling him to relief. Id. at 45-46, 78 S.Ct. 99. Of course, plaintiffs alleging securities fraud must also meet the Rule 9(b) standard for pleading fraud with particularity.

For a complaint to state a claim for securities fraud under section 10(b) and Rule 10b-5, it must plead six elements: (1) a material misrepresentation or omission; (2) scienter, or a wrongful state of mind; (3) a connection with the purchase or sale of a security; (4) reliance; (5) economic loss; and (6) loss causation. Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 341-42, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005).

The PSLRA requires plaintiffs' complaint to "specify each statement alleged to have been misleading [and] the reason or reasons why the statement is misleading." 15 U.S.C. § 78u-4(b)(1). Further, "if an allegation regarding the statement or omission is made on information and belief, the complaint shall state with particularity all facts on which that belief is formed." Id.

The PSLRA also separately imposes a rigorous pleading standard on allegations of scienter. Scienter is "a mental state embracing intent to deceive, manipulate, or defraud." Ernst Ernst v. Hoch-felder, 425 U.S. 185, 193 n. 12, 96 S.Ct. 1375, 47 L.Ed.2d 668 (1976).

In this circuit, a plaintiff may satisfy the scienter requirement with a showing of either conscious intent to defraud or "a high degree of recklessness." Aldridge, 284 F.3d at 82 (citing Greebel, 194 F.3d at 198-201). A complaint alleging securities fraud "shall, with respect to each [alleged] act or omission . . . state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind." 15 U.S.C. § 78u-4(b)(2) (emphasis added). "While under Rule 12(b)(6) all inferences must be drawn in plaintiffs' favor, inferences of scienter do not survive if they are merely reasonable, as is true when pleadings for other causes of action are tested by motion to dismiss under Rule 12(b)(6)." Greebel, 194 F.3d at 195.

Further, Tellabs confirms this court's existing rule from Cabletron, 311 F.3d at 40, that scienter should be evaluated with reference to the complaint as a whole rather than to piecemeal allegations. Tellabs, 127 S.Ct. at 2509 ("The inquiry . . . is whether all of the facts alleged, taken collectively, give rise to a strong inference of scienter. . . ."). Tellabs also affirms our view, see, e.g., Ezra Charitable Trust, 466 F.3d at 11, that competing inferences should be weighed against plaintiffs' preferred interpretation of the facts. Tellabs, 127 S.Ct. at 2504 (holding that a court "must engage in a comparative evaluation; it must consider, not only inferences urged by the plaintiff . . . but also competing inferences rationally drawn from the facts alleged"). While it may be unusual for courts to weigh competing inferences from facts, Congress mandated this review in the PSLRA.

However, Tellabs has overruled one aspect of the rule this court stated in Credit Suisse. Credit Suisse held that where there were equally strong inferences for and against scienter, this resulted in a win for the defendant. 431 F.3d at 49 ("Scienter allegations do not pass the 'strong inference' test when, viewed in light of the complaint as a whole, there are legitimate explanations for the behavior that are equally convincing."). This is no longer the law.

Tellabs held that a "strong inference" of scienter "must be more than merely plausible or reasonable — it must be cogent and at least as compelling as any opposing inference of nonfraudulent intent." 127 S.Ct. at 2504-05 (emphasis added). In other words, where there are equally strong inferences for and against scienter, Tellabs now awards the draw to the plaintiff. 127 S.Ct. at 2510 ("A complaint will survive . . . if . . . the inference of scienter [is] cogent and at least as compelling as any opposing inference one could draw from the facts alleged.").

B. Application to the Factual Allegations

We apply these principles to the amended complaint's allegations about four categories of misleading statements and omissions in the Official Statement.

The amended complaint alleges that the Official Statement was issued within the context of a serious financial crisis at Bradford, and the defendants had motivation to conceal the crisis so that they could continue to finance their operations. For instance, the amended complaint alleges defendants were well aware of the college's financial instability. The amended complaint describes a meeting of the Board of Trustees on February 6, 1997, a year before the Board approved the bond offering. At that meeting, the Chair of the Board stated that the then-current cash flow model "indicates that the College may be able to survive for five more years. To stem cash bleeding, however, which would occur in that period would be devastating to the faculty and staff in terms of no salary increases and other cost cutting measures." Kiszka, the college's chief financial officer, apparently rejoined that this was too optimistic: "[T]here is a possibility of the College surviving five years assuming more layoffs and no salary increases, but [Kiszka] felt that it would be more like two or three years. Looking at the bigger picture, layoffs and cutbacks would be very disruptive and send a bad message." The amended complaint alleged that the college took no steps to lay off personnel or freeze salaries, which were necessary "[t]o stem cash bleeding," because that would be "disruptive and send a bad message." The amended complaint alleges that this reluctance to "send a bad message" extended to the representations in the Official Statement.

The amended complaint alleges that one trustee resigned when the Board of Trustees initially voted in 1996 to approve construction of new dormitories because he concluded that student enrollment could not support the expansion.

Despite having implemented no widespread cost-cutting measures and after a decade of budget deficits, the Bradford defendants approved the issuance of the bonds. This in turn obligated the college to incur an additional cash flow drain of over $1.2 million annually through the year 2028. At the time, the college's annual operating budget, excluding financial aid, was $11.7 million. The amended complaint alleges that the ability to meet this additional obligation was "dependent primarily upon two factors: the College's level of student enrollment, and the degree to which the College discounted its tuition revenues by funding financial aid awards to students."

The plaintiffs' lead theory of misrepresentation was that the college's ability to pay the bonds depended upon the school's meeting its increased enrollment targets, which would substantially increase the size of the student body. The Official Statement, they allege, had several flaws: It was too optimistic when it said the college believed it could meet its goals. It failed to disclose available information suggesting that the college was unlikely to meet its goals. And it fudged on the level of the college's equity contribution to the project.

The plaintiffs' secondary theory was that even if the college could increase its enrollment, this would not improve the college's financial condition if it gave high financial aid to reach this result. The plaintiffs concede the Official Statement informed investors of this connection between enrollment and high financial aid. The Statement contained the college's projections that it would reduce financial aid, and said that its financial plan called for reducing aid in the upcoming 1998-1999 academic year to 28.8% of student income. The plaintiffs argue that the district court erred in rejecting as immaterial the difference between the projections in the Official Statement, what was in the school's budget, and what actually happened.

We start with the allegations relevant to increased enrollment targets.

1. Attrition Rate

The amended complaint alleges that the Official Statement misled investors by failing to refer to Bradford's "severe, long-standing problem with student attrition." The short answer is that the Statement warned that the college's ability to pay debt service was "highly dependent upon tuition and fee revenues from students," and that a failure "to attract and retain students in sufficient numbers . . . could adversely affect the ability of [the college] to make required payments on the Series 1998 Bonds." The Official Statement also presented enough information to notify investors that student attrition factored into the risk associated with the bonds.

Tables in the Statement containing enrollment and admissions statistics covering academic years 1993 to 1997 revealed that the college enrolled new students each fall in numbers far exceeding one-quarter of total enrollment. While some portion of the additional number of incoming students might be attributable to fluctuations in the number of transfer students and others not enrolled in a traditional four-year program at the college, these tables gave notice that some portion of the discrepancy would be due to attrition.

The attrition rate at Bradford had, according to the amended complaint, been "an extraordinary 60%" since 1989. The amended complaint alleges that omitting that figure from the Official Statement concealed the college's inability to repay its bond debt. The defense argues those figures may be derived from the information provided. The plaintiffs argue that it is impossible to calculate from the tables the alleged 60% attrition rate. This may be true, but it is insufficient to make the other statements misleading. The amended complaint does not explain why omitting information about the precise attrition rate at Bradford would mislead investors when more pertinent measures of the college's financial health are presented forth-rightly in the Statement.

The Statement alerts investors that repayment of the bond debt depended entirely on Bradford's revenues, and that like most colleges, Bradford was "highly dependent upon tuition and fee revenues from students." Student-generated revenue is, at any given time, a function of total enrollment levels, not of attrition. The Statement provided information about past and current enrollment levels, and the amended complaint never disputes the accuracy of those numbers. Furthermore, the Statement contained information on admissions trends that would allow a potential investor to evaluate whether the college's enrollment level was sustainable.

The college did not have a duty to disclose in the Statement every possible material fact about its operations and finances, so long as the disclosures that were made satisfied the statute. Cf. Gross v. Summa Four, Inc., 93 F.3d 987, 992 (1st Cir. 1996) ("[A] corporation does not commit securities fraud merely by failing to disclose all nonpublic material information in its possession." (citing Roeder v. Alpha Indus,, Inc., 814 F.2d 22, 26 (1st Cir. 1987))). The Official Statement needed only disclose enough accurate information and not omit pertinent information to allow investors to make an informed decision about whether to invest. It did so.

2. Enrollment Projections

Again, on the theory that accurate enrollment projections were key, the plaintiffs find fault with this passage from the Official Statement:

As of April 3, 1998, applications received by the College to date total 879, an increase of more than 18% from April 3, 1997. The total of 879 exceeds total applications received for the fall 1997. The majority of increases have been the traditional freshman application pool.

Based on this' increase in applications, historic rates for conversion of applications into enrollments, the number of applications from freshmen and deposits received to date, the College believes that it can reach its goal of enrolling 225 new students for the fall of 1998. . . .

The plaintiffs do not challenge the facial accuracy of the college's contemporaneous figures, but allege that the use of the number of applications for the fall of 1998 was misleading for several reasons.

The plaintiffs argue that because Brad-ford had begun accepting standardized applications submitted over the internet, this "artificially inflated the applications numbers." The goal of 225 new students in fall 1998 was allegedly unrealistic because ultimately, the acceptance rate dropped and Bradford "had fewer actual acceptances" for fall 1998 than the previous year. Finally, the plaintiffs allege that by the date of the Official Statement, the defendants knew that the number of students who had placed deposits for fall 1998 "had declined by almost 20%."

These allegations fall short of establishing that the enrollment projections were materially misleading at the time of the Statement. A plaintiff may not plead "fraud by hindsight"; i.e., a complaint "may not simply contrast a defendant's past optimism with less favorable actual results" in support of a claim of securities fraud. Shaw v. Digital Equip. Corp., 82 F.3d 1194, 1223 (1st Cir. 1996). There is nothing in the amended complaint to establish that the defendants were aware of facts, at the time they made their predictions, that would have made those predictions unreasonable, if they were unreasonable.

Even now there is no basis to conclude, from the plaintiffs' premises, that the predictions were unreasonable. It has not been demonstrated that accepting applications by internet for the fall of 1998 would inflate the college's application numbers or somehow lead to lower matriculation rates, or that the defendants had any reason to think this. Similarly, while the amended complaint, without giving specific numbers, quotes the actual acceptance rate for fall 1998 and claims that Bradford "had fewer actual acceptances [in fall 1998] than it had for Fall 1997," it does not allege in any detail how the defendants could have fore-cast that outcome in May 1998.

The allegation regarding the drop in actual deposits, as opposed to projected deposits, comes closer, but it does not allege enough detail to satisfy the pleading standards of the PSLRA. The amended complaint alleges there was a 20% year-over-year drop in actual deposits as of May 1998, but does not "state with particularity all facts on which that belief is formed." 15 U.S.C. § 78u-4(b)(1). The plaintiffs have not included details about how they were able to identify the 20% figure, much less whether this information was known to the defendants at the relevant time. Standing alone, this is insufficient for an allegation of securities fraud. It is true, as the plaintiffs argue, that the PSLRA does not require plaintiffs to plead evidence. See Cabletron, 311 F.3d at 33 (citing Cooperman v. Individual Inc., 171 F.3d 43, 48-49 (1st Cir. 1999)); Shaw, 82 F.3d at 1225. But more meat was needed on these bones.

3. Equity Contribution

The amended complaint alleges that the Official Statement also skewed the school's future financial health by misrepresenting Bradford's intentions regarding a planned equity contribution to the construction and renovation project. The Statement, under the heading of "Estimated Sources and Uses of Funds," said that a $1 million contribution of the college's own funds would be used for the construction and renovation project.

The amended complaint alleges that at a February 5, 1998 meeting of two committees of the Board of Trustees, those committees determined that the bond issue would not cover the entire cost of the project and resolved to recommend to the full Board that

[t]he College will have to review the project and attempt to reduce the final two phases to correspond with the bond financing or the College will have to include an equity contribution to fund a portion of the construction. It was agreed that the bond financing would proceed as recommended with Advest and that the College would make every effort to reduce its construction costs or make an equity contribution at the end in the final phase of the project.

(Emphases added in amended complaint.) The plaintiffs interpret this as revealing the Bradford defendants' intention not to contribute any funds to the project, but rather "to cut corners on the project in an effort to bring project costs down to the level of the bond financing." The Official Statement was allegedly misleading because it failed to disclose the "contingent and delayed nature of the College's putative 'equity' contribution."

Even accepting the allegations as true, and adopting the plaintiffs' preferred inference that the committees' recommendation was implemented by the full Board prior to the bond offering, the allegations simply do not establish any misrepresentation. The Official Statement does not speak in any way to the timing of Bradford's expected contribution. By like token, the committees' statement from the February 5, 1998 meeting does not at all foreclose the college from making an equity contribution; in fact, it clearly states an intention that the college would contribute should the cost of completing the project so require.

4. Financial Aid Projections

We turn to the plaintiffs' second tier of arguments — that the financial condition of the college would be adversely affected even if enrollment increased unless financial aid did not increase, and that the statements on this point were misleading.

The amended complaint alleges that projections in the Official Statement for financial aid levels in the 1997-1998 and 1998-1999 academic years were false and misleading when made. The Official Statement acknowledges that "there has been a substantial increase in financial aid funded by the College" between 1989 and 1997. However, the plaintiffs point to the Statement's projections:

[D]uring the 1997-98 academic year, the College estimates that financial aid will be reduced to 29.9% of student income versus 30.3% the previous year. This expected reduction is a result of a change in methodology of aiding students with college-funded support versus additional loans funded by students and/or parents. The College's financial plan currently calls for a further reduction of financial aid spending for [the] 1998-1999 academic year to 28.8% of student income.

The amended complaint alleges that these statements were untrue — that the actual percentage of financial aid awarded in 1997-1998 was over 35% of student income. The plaintiffs also allege that the 1998-1999 budget, as of the date of the Official Statement, contemplated that financial aid would be 31.3% of student income, not 28.8%.

a. 1997-1998

Again, the allegations concerning the discrepancies between the actual percentage for the 1997-1998 academic year and the estimated ones in the Official Statement come from hindsight. The amended complaint obtained the 35% figure from Bradford's audited financials for the year, which were not produced until after the close of the fiscal year in June 1998, after the date of the Official Statement.

No facts are pled to support the general allegation that "[a]t the time the Official Statement was distributed, the College possessed, and the Defendants had access to, the data that proved the 1997-98 'estimate' was substantially incorrect." These allegations do not "state with particularity facts giving rise to a strong inference that the defendant acted with [scienter]." 15 U.S.C. § 78u-4(b)(2).

The plaintiffs do point to information which they say was available at the time of the Official Statement about the 1997-1998 academic year. They allege the college knew that enrollment for the spring term in 1998 was lower than in the budget and knew what its financial aid commitments were. They also say the defendants knew the financial aid commitments were almost $250,000 more than budgeted. Those two pieces of data, however, do not, without more, tell one of the actual percentage of financial aid as against student income, either as of May 1, 1998, or as of the end of that academic year.

b. 1998-1999

The most troublesome issue in the case is presented by the allegations regarding the Official Statement's description of the financial plan for the 1998-1999 school year's financial aid levels. These are not mere optimistic projections because the statement is that "[t]he College's financial plan currently calls for a further reduction of financial aid spending for [the] 1998-1999 academic year to 28.8% of student income." (Emphasis added.) This sentence would misrepresent the facts if the defendants, as of May 1, 1998, had actually decided to budget a significantly greater amount of the college's funds for student aid.

The amended complaint alleges that Bradford's 1998-1999 budget was originally submitted to the Trustees on April 29, 1998, and reviewed by the Finance Committee of the Board on May 8, 1998. The allegation is that the budget was revised at a meeting of college administrators later in May. The finalized budget pegged financial aid spending for 1998-1999 at 31.3% of student income. According to the amended complaint, that amount was $280,000 greater than the sum referred to in the Official Statement issued on May 1. There is no allegation as to the budgeted figures in the April 29 draft budget. The plaintiffs draw the inference that the final figure of 31.3% was in the initial draft budget of April 29. The claim is that the makers of the statement about the 1998-1999 budget had actual knowledge that the statement was false or misleading, thus removing it from the safe harbor provisions of the PSLRA. See 15 U.S.C. §§ 77z- 2, 78u-5; Greebel 194 F.3d at 201.

We cannot say that the discrepancy between the 28.8% figure in the Statement and the 31.3% figure alleged in the amended complaint is immaterial as a matter of law, as defendants argue. Materiality is usually a matter for the trier of fact. Shaw, 82 F.3d at 1217 (citing Basic Inc. v. Levinson, 485 U.S. 224, 236, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988)). Although $280,000 represents only a small portion of the college's operating revenues, which exceeded $13 million in fiscal year 1997, it amounts to about 7% of the amount Bradford spent on student aid that year. Put another way, $280,000 would cover about a third of the debt service due on the bonds in 1999. The 31.3% figure could indicate a trend of rising financial aid contributions, not the anticipated downward trend portrayed in the Statement. The Statement named the amount of financial aid awards as a risk factor in the college's ability to repay the bonds.

Even assuming materiality, the amended complaint provides no information on what proposed budget figures were known to the Bradford defendants before the Official Statement was completed. One might infer that at least President Short, as an ex officio member of the Board of Trustees, knew that financial aid was budgeted to be 31.3% at the time he helped draft the Statement. But few specifics are given to support the inference that the defendants knew the representation about the budget was wrong when made.

We must assess whether the plaintiffs have pleaded sufficient facts to give rise to a strong inference that the defendants wrote that passage in the Official Statement with "intent to deceive, manipulate, or defraud." Ernst Ernst, 425 U.S. at 193 n. 12, 96 S.Ct. 1375. More than mere proof that the defendants made a particular false or misleading statement is required to show scienter. Aldridge, 284 F.3d at 83; see also Geffon v. Micrion Corp., 249 F.3d 29, 36 (1st Cir. 2001). But the fact that a defendant knowingly made a false statement is "classic evidence" of scienter. Aldridge, 284 F.3d at 83.

One inference, urged by plaintiffs, is that the defendants — or at least Short and the Trustees — were in possession of a budget that clearly contradicted the numbers they planned to quote in the Official Statement. The argument is that they feared that portraying an upward trend in financial aid expenditures would warn off investors, so they opted instead to spin the numbers to suggest that Bradford was heading for a turnaround.

According to the amended complaint, the Bradford defendants determined "to operate the College at all costs rather than preserve its assets for the benefit of creditors." The defendants, on this theory, saw the 1998 bond offering as a means of continuing operations, even though the defendants knew it could not save Bradford from insolvency. With that in mind, the plaintiffs allege, the defendants were willing to misrepresent the amount of the funds they had earmarked for student aid in order to shade the college's operational health and induce hapless investors to purchase bonds to finance an already doomed project.

But there are other inferences, which in our view are stronger. After years of budget deficits, the Bradford defendants realized by early 1997 that the college would have to cease operations within five years unless they could implement a plan to stabilize its budget. Still, within a year, the college had experienced a spike in its matriculation rate and the number of incoming students, leading the defendants to believe that demand would support an increase in the college's enrollment capacity. The Bradford defendants thus settled on an expansion plan to be funded by the 1998 bonds, and they expected that expansion would save the school.

The defendants considered a draft budget for the 1998-1999 school year just a matter of days before the bond offering. That budget may or may not, at that time, have matched the defendants' sanguine financial aid projections in the Official Statement. However, the defendants would know that the draft budget would yet be revised, as the amended complaint confirms.

In addition, the defendants may have been operating under a different set of assumptions. For instance, the Statement indicates that phase one of the project, to be financed by the bonds that had not yet been issued as of the time of the draft budget, would be completed by fall 1998. The resulting addition of "new, townhouse-style dormitories" could increase enrollment capacity and Bradford's attractiveness to current and prospective students. Given the defendants' enrollment targets for the 1998-1999 school year and their assumption that the expansion project would be financed and underway by that time, it would be reasonable to infer that they believed the 28.8% figure to be achievable.

There is no set pattern of facts that will establish scienter; it is a case-by-case inquiry. Greebel, 194 F.3d at 196. There are several reasons why the plaintiffs' inference of scienter is not at least equally as strong. First, the Official Statement as a whole candidly laid out the sorry financial history of the college and, for most of its estimates and projections as to a happier future, it provided accurate and non-misleading information, as we discussed. The Official Statement fully disclosed that despite the college's enrollment growth, the college had (a) incurred operating deficits every year since 1989 and (b) done so with a substantial increase in financial aid funded by the college. It was careful to say it could only estimate that for the 1997-1998 academic year, there would be a reduction in financial aid as a proportion of student income, and that it would be a modest four-tenths of a percentage point. It explained the basis for the estimate. As to the 1998-1999 academic year, the Statement was careful to say that the college's financial plan "currently" called for a reduction of one and a tenth percentage points in financial aid over its estimates for the prior year. The Official Statement fully disclosed that some 80% of full-time students received Bradford-funded aid. And the Statement said that based on four factors, the college believed it could reach its enrollment goals and "reduc[e] slightly the average amount of financial aid awards . . . from College funds." Conversely, the Statement said, failure to meet these goals could "adversely affect the College's ability to reach Financial Equilibrium."

In addition, the Bradford College defendants have different characteristics than are typical in securities fraud cases, characteristics which make it more difficult to infer a high degree of recklessness or an intent to defraud. They are unlike the paradigmatic securities fraud defendant, who is likely to be a corporate insider standing to profit from the sale of artificially inflated securities. Here, the defendants are Officers and Trustees of a non-profit educational institution. There are no allegations that the proceeds from the Bradford bonds would be spent on anything that would personally enrich any of the Bradford defendants. There is no allegation that they are particularly sophisticated in securities transactions.

Of course "the self-interested motivation of defendants in the form of saving their salaries or jobs" is relevant — though not necessarily sufficient — to a showing of scienter. Greebel, 194 F.3d at 196. There is no reason to credit the inference that the Official Statement was made in order to save Short's job as President, as he resigned a few months later, by July 1998. Short's resignation had long been anticipated, as the Official Statement itself referred to his expected resignation on June 30, 1998. In this case there is no allegation of any additional motive other than the defendants' desire to keep the college operating and to deprive creditors of their due in an inevitable bankruptcy proceeding. These are shaky grounds for leaping to the conclusion that there is a strong inference that the defendants intentionally or recklessly disregarded the facts available to them when quoting financial aid figures in an offering statement accompanying a multi-million dollar bond offering.

We hold that the plaintiffs' allegations regarding planned financial aid expenditures for the 1998-1999 school year fail to establish an inference of scienter that is cogent and at least as compelling as available competing inferences of non-fraudulent conduct. The dismissal of the section 10(b) and Rule 10b-5 claims against all of the Bradford defendants was appropriate.

C. Section 20(a) Claim Against the Bradford Defendants

In addition to their claims under section 10(b) and Rule 10b-5, the plaintiffs pleaded a claim under section 20(a) of the Securities Exchange Act of 1934, 15 U.S.C. § 78t, against the Officer and Trustee defendants. Section 20(a) imposes joint and several liability on "[e]very person who, directly or indirectly, controls any person liable" for a securities fraud violation. Id. § 78t(a). The plaintiffs correctly point out that they need not plead scienter according to the rigorous standards of the PSLRA in order to state a claim under section 20(a). See In re Stone Webster, Inc. Sec. Litig. (Stone Webster I), 414 F.3d 187, 194 (1st Cir. 2005).

However, it does not follow that the control-person claim against the Brad-ford defendants should be allowed to proceed if, as the plaintiffs phrase it, we find that the plaintiffs would have stated a claim for securities fraud "but for their failure to plead scienter sufficiently on the part of individual defendants." The plain terms of section 20(a) indicate that it only creates liability derivative of an underlying securities violation. 15 U.S.C. § 78t(a); see also In re Stone Webster Sec. Litig. ( Stone Webster II), 424 F.3d 24, 27 (1st Cir. 2005) ("[I]t is an essential element of the § 20(a) controlling person claims in question that plaintiffs show a Rule 10b-5 violation by the controlled entity."). Here, there was no underlying 10b-5 violation. The section 20(a) claim must fail.

D. Section 12(a)(2) Claim Against Advest

The amended complaint also includes a claim against Advest under section 12(a)(2) of the Securities Act of 1933, 15 U.S.C. § 77 l. That section penalizes the sale of a security by means of a misleading prospectus. Id. As an underwriter, Advest was under a duty to exercise "reasonable care" as to the offering. Id. § 78 l(a)(2).

Advest argues that it may not be reached under section 12 because the actual issuer of the bond was a state agency, MIFA, and section 12 liability does not attach to securities issued by a state government entity. See 15 U.S.C. § 77c(a)(2). We do not need to address that argument, nor the plaintiffs' argument that they state a section 12(a)(2) claim because the bond is solely payable by a covered private entity.

We take a different approach. We have assumed arguendo that the allegations about financial aid for the 1998-1999 academic year could be found to be material against the Bradford defendants, but that there is no strong inference of scienter on the part of those defendants. There are no allegations that Advest knew of these proposed budgetary materials. There is no basis to assume that the budget materials made available to the Bradford defendants two days before the Official Statement issued were known to Advest, so that Advest knew at the time the budget materials were at variance (if they were) with the Statement. Indeed, all of the allegations in the complaint go to facts allegedly known to the Bradford defendants. The essence of the claim against Advest is fraud, and the pleading fails to meet the Rule 9(b) standard for pleading allegations of fraud with specificity. Where section 12(a)(2) claims are grounded in fraud, Rule 9(b) applies. See Shaw, 82 F.3d at 1223 ("[I]f a plaintiff were to attempt to establish violations of Sections 11 and [12(a)(2)] as well as the anti-fraud provisions of the Exchange Act through allegations in a single complaint of a unified course of fraudulent conduct . . . Rule 9(b) would probably apply to the Section 11, [12(a)(2)], and Rule 10b-5 claims alike."); see also Wagner v. First Horizon Pharm. Corp., 464 F.3d 1273, 1275 (11th Cir. 2006) ("[E]ven securities claims without a fraud element must be pled with particularity pursuant to [Rule 9(b)] when that nonfraud securities claim is alleged to be part of a defendant's fraudulent conduct."); Melder v. Morris, 27 F.3d 1097, 1100 n. 6 (5th Cir. 1994) (applying Rule 9(b) to section 12 claim "grounded in fraud"); Shapiro v. UJB Fin. Corp., 964 F.2d 272, 287 (3d Cir. 1992) (same); Sears v. Likens, 912 F.2d 889, 892-93 (7th Cir. 1990) (same). But see In re NationsMart Corp. Sec. Litig., 130 F.3d 309, 315-16 (8th Cir. 1997) (declining to apply Rule 9(b) to non-fraud securities claims).

Although the amended complaint fails to plead any federal claims upon which relief can be granted, we express no views on the plaintiffs' state-law claims, which are currently pending in a stayed state court proceeding.

The judgment of the district court is affirmed.