STAHL, Circuit Judge.

In the late 1980’s, plaintiffs-appellants purchased shares of two separate “junk bond” funds. After the value of the purchased shares plummeted, plaintiffs alleged various federal securities law violations. In a series of related rulings, the district court dismissed some of plaintiffs’ allegations for failure to state a claim, and granted summary judgment in favor of defendants on all remaining claims. We affirm in part and reverse in part.

FACTUAL BACKGROUND AND PRIOR PROCEEDINGS

Prior to this appeal, the proceedings in these two cases were not formally consolidated. As the district court noted, the two cases raise many identical issues. Thus, our discussion, unless we specifically state otherwise, applies equally to both cases.

*172 In 1988, both New America High Income Fund, Inc. and Prospect Street High Income Portfolio, Inc. (“the New America Fund,” and “the Prospect Street Fund,” or collectively “the funds”) were first publicly offered on the New York Stock Exchange. Each fund’s purpose, as stated in their nearly identical prospectuses, was to invest in a diversified portfolio of high yield fixed-income securities, commonly known as “junk bonds.”

In April 1989, well after the initial public offerings, a study headed by Professor Paul Asquith (“the Asquith study”) disclosed that the default rate of junk bonds was much higher than had been previously believed. 1 This conclusion was reached by calculating the adverse effects of “aging” on junk bonds. 2

Within months of the study, though not necessarily as a direct result of the study, the market for junk bonds began to collapse. By November 1989, both funds had reduced their dividends, and the share value of each fund had declined considerably.

Plaintiffs, who consist of putative classes of purchasers of each fund, commenced parallel actions against the two funds. The First Amended Complaints (hereinafter “the original complaints”) were lengthy, alleging violations of a variety of federal securities laws, including section 10(b) of the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and sections 11 and 12(2) of the Securities Act of 1933, 15 U.S.C. §§ 77k, 77(2). 3 The gist of the original complaints was that the funds’ directors, advisors and underwriters (“defendants”) knew of, but failed to disclose, adverse information about the junk bond market. In particular, the complaints alleged that defendants had agreed to act, and had in fact acted, as purchasers of last resort for undesirable junk bonds; that they knew of infirmities in the junk bond market at the time they publicly offered shares of the funds and thereafter; and that misleading statistics were used in the prospectuses to portray the historical performance of junk bonds. 4

The district court dismissed many of plaintiffs’ claims on the pleadings, see Miller v. New Am. High Income Fund, 755 F.Supp. 1099 (D.Mass.1991) (“Miller 7”); Lucia v. *173 Prospect St. High Income Portfolio, Inc., 769 F.Supp. 410 (D.Mass.1991) (“Lucia I ”), but nonetheless allowed both sets of plaintiffs to replead.

Plaintiffs’ Second Amended Complaints (hereinafter “the revised complaints”) alleged causes of action only under sections 11 and 12(2). All section 10(b) claims presented in the original complaints were dropped. Among other things, the revised complaints focused on a ten-year comparison between junk bonds and United States Treasury securities (“Treasury securities”) that was included in the prospectuses. 5 Though the ten-year figure showed that junk bonds had outperformed Treasury securities, the revised complaints alleged that during the six years leading up to each fund’s public offering, Treasury securities had actually outperformed junk bonds. 6

Shortly after the revised complaints were filed, defendants moved for summary judgment. The district court began by ruling as a matter of law that the comparison to Treasury securities in the prospectuses was not misleading. See In re New Am. High Income Fund Sec. Litig., 834 F.Supp. 501, 506-07 (D.Mass.1993) (“Miller 77”). It went on to grant summary judgment in favor of defendants on all other claims. Id.; Lucia v. Prospect St. High Income Portfolio, Inc., No. 90-10781-MA (D.Mass. Aug. 26 1993) (“Lucia II ”). Plaintiffs appeal these various rulings. We address plaintiffs’ claims in the order in which they were decided by the district court.

II.

DISCUSSION

A. Section 10(b) Claims

The district court dismissed plaintiffs’ section 10(b) claims at the first of these cases’ *174 two pleading stages. We affirm that dismissal, though on somewhat narrower grounds than those relied upon by the district court.

1. Standard of Review

Rule 12(b)(6) dismissals are subject to de novo review. Northeast Doran, Inc. v. Key Bank of Maine, 15 F.3d 1, 2 (1st Cir.1994). While we generally credit all allegations in the complaint and draw all reasonable inferences favorable to the plaintiff, id., Rule 9(b) imposes heightened pleading requirements for allegations of fraud. “In all averments of fraud or mistake, the circumstances constituting fraud or mistake shall be stated with particularity.” Fed.R.Civ.P. 9(b).

As we have stated in a recent discussion of Rule 9(b) in the securities context:

[GJeneral averments of the defendants’ knowledge of material falsity will not suffice. Consistent with Fed.R.Civ.P. 9(b), the complaint must set forth specific facts that make it reasonable to believe that defendants] knew that a statement was materially false or misleading. The rule requires that the particular times, dates, places or other details of the alleged fraudulent involvement of the actors be alleged.

Serabian v. Amoskeag Bank Shares, Inc., 24 F.3d 357, 361 (1st Cir.1994) (citations and internal quotation marks omitted). ‘We have been especially rigorous in demanding such factual support in the securities context.” Romani v. Shearson Lehman Hutton, 929 F.2d 875, 878 (1st Cir.1991). Moreover, this heightened pleading is required “even when the fraud relates to matters peculiarly within the knowledge of the opposing party.” Id.

2. The Original Complaints

Plaintiffs’ original complaints alleged various wrongdoing by defendants. The common thread running throughout the original complaints, however, was that defendants knew of infirmities in the junk bond market, and that they nonetheless entered a vast web of illicit agreements with Drexel Burnham Lambert, and with former junk bond dealer Michael Milken, in order to become purchasers of last resort for undesirable junk bonds.

The district court properly concluded that these general allegations in the original complaints were wholly conclusory. No factual basis is put forward to support plaintiffs’ theory that defendants consorted with Drex-el, that they dealt with Michael Milken, that they agreed to act as purchasers of last resort for undesirable bonds, or that they knew enough to anticipate the impending fall-out of the junk bond market. Because all of plaintiffs’ 10(b) claims rely fundamentally on such unsupported allegations, the district court properly dismissed these claims for failure to meet Rule 9(b). 7 Cf. Romani, 929 F.2d at 878 (finding that complaint failed to satisfy Rule 9(b) where it contained “no factual allegations that would support a reasonable inference that adverse circumstances existed at the time of the offering, and were known and deliberately or recklessly disregarded by defendants”).

B. Section 11 and 12(2) Claims

As noted above, plaintiffs were allowed to replead. Defendants’ motions for summary judgment soon followed, and summary judgment was granted in favor of defendants.

1. Standard of Review

“A district court’s grant of summary judgment is subject to plenary review.” Calenti v. Boto, 24 F.3d 335, 338 (1st Cir.1994). We read the record indulging all inferences in favor of the non-moving party. Id. Summary judgment is appropriate only “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). In seeking to forestall the entry of summary judgment, a nonmovant may not rely upon allegations in its pleadings. Rath *175 er, the nonmovant must “set forth specific facts showing that there is a genuine issue for trial.” Fed.R.Civ.P. 56(e).

2. Parallel Paths Diverge

Both complaints alleged that the six-year comparison favored Treasury securities. And the Miller plaintiffs, unlike the Lucia plaintiffs, in their response to defendants’ motion for summary judgment, set forth facts showing that the six-year figure, as well as a shorter three-year figure, actually favored Treasury securities. Moreover, the district court squarely addressed this argument in ruling on the Miller defendants’ motion for summary judgment. See In re New Am. High Income Fund Sec. Litig., 834 F.Supp. 501, 506-07 (D.Mass.1993). Accordingly, we see no merit to defendants’ argument that the Miller plaintiffs waived this issue.

The Lucia plaintiffs, however, failed to preserve this issue. In fact, the Lucia plaintiffs’ opposition to defendants’ summary judgment motion fails to even mention the six-year comparison. Despite the striking similarities in these two cases, the Lucia plaintiffs pursued a significantly different tack in opposing defendants’ motion for summary judgment, and failed to argue that the Prospect Street prospectus was misleading due to its failure to include a shorter-term comparison to Treasury securities. As noted above, a nonmovant faced with a motion for summary judgment may not rest on its pleadings. Moreover, we see no reason in this case to relax our general rule that “theories not raised squarely before the district court cannot be surfaced for the first time on appeal.” McCoy v. Massachusetts Inst. of Technology, 950 F.2d 13, 22 (1st Cir.1991), cert. denied, — U.S. -, 112 S.Ct. 1939, 118 L.Ed.2d 545 (1992). Accordingly, our discussion of the six-year comparison applies only to the Miller case.

3. Materiality under Sections 11 and 12(2) and the Omission of the Six-Year Comparison

Sections 11 and 12(2) both prohibit, inter alia, the use of any “untrue statement of a material fact,” 15 U.S.C. 77i(2), as well the use of any information which “omits to state a material fact necessary in order to make the statements, in the light of the circumstances under which they are made, not misleading.” Id.; see also 15 U.S.C. 77k(a).

The boundaries of materiality in the securities context are clearly enunciated in our case law.

The mere fact that an investor might find information interesting or desirable is not sufficient to satisfy the materiality requirement. Rather, information is “material” only if its disclosure would alter the “total mix ” of facts available to the investor and “if there is a substantial likelihood that a reasonable shareholder would consider it important” to the investment decision.

Milton v. Van Dorn Co., 961 F.2d 965, 969 (1st Cir.1992) (quoting Basic, Inc. v. Levinson, 485 U.S. 224, 231-32, 108 S.Ct. 978, 983-84, 99 L.Ed.2d 194 (1988)). It is equally well established that “[w]hen a corporation does make a disclosure—whether it be voluntary or required—there is a duty to make it complete and accurate.” Roeder v. Alpha Indus., Inc., 814 F.2d 22, 26 (1st Cir.1987). Moreover, disclosed facts may “not be ‘so incomplete as to mislead.’ ” Backman v. Polaroid Corp., 910 F.2d 10, 16 (1st Cir.1990) (en banc) (quoting SEC v. Texas Gulf Sulphur Co., 401 F.2d 833, 862 (2d Cir.1968), cert. denied, 394 U.S. 976, 89 S.Ct. 1454, 22 L.Ed.2d 756 (1969)).

In addition, the fact that a statement is literally accurate does not preclude liability under federal securities laws. “Some statements, although literally accurate, can become, through their context and manner of presentation, devices which mislead investors. For that reason, the disclosure required by the securities laws is measured not by literal truth, but by the ability of the material to accurately inform rather than mislead prospective buyers.” McMahan v. Wherehouse Entertainment, Inc., 900 F.2d 576, 579 (2d Cir.1990), cert. denied, 501 U.S. 1249, 111 S.Ct. 2887, 115 L.Ed.2d 1052 (1991). Under the foregoing standards, “emphasis and gloss can, in the right circumstances, create liability.” Isquith v. Middle S. Utils., Inc., 847 F.2d 186, 203 (5th Cir.), *176 cert. denied, 488 U.S. 926, 109 S.Ct. 310, 102 L.Ed.2d 329 (1988).

Finally, we note that the question of whether an omission or misleading statement is material “is normally a jury question and should not be taken from it unless the court has engaged in meticulous and well articulated analysis of each item of withheld or misrepresented information.” SEC v. Seabord Corp., 677 F.2d 1301, 1306 (9th Cir.1982). See also Milton, 961 F.2d at 970 (“‘[T]he [objective] determination [of materiality] requires delicate assessments of the inferences a ‘reasonable shareholder’ would draw from a given set of [undisputed] facts and the significance of those inferences to him and those assessments are peculiarly ones for the trier of fact.’ ”) (quoting TSC Indus. Inc. v. Northway, Inc., 426 U.S. 438, 450, 96 S.Ct. 2126, 2133, 48 L.Ed.2d 757 (1976)); Isquith, 847 F.2d at 208 (stating that the adequacy of disclosures in securities cases is generally a question for a jury).

As we have said, plaintiffs argue that the ten-year comparison between Treasury securities and junk bonds, though accurate, was misleading because a shorter, six-year comparison favored Treasury securities. We begin by noting that the six years at issue are the six years leading up to the fund’s public offering. Moreover, while any one or two years might favor Treasury securities without amounting to an unfavorable trend, we think that a six-year comparison favoring Treasury securities is substantial enough to cast some doubt on the reliability of the reported ten-year figure. In other words, we cannot say as a matter of law that the undisclosed information about the six-year period would not alter the total mix of facts available to the investor. Rather, a jury could find that there is a substantial likelihood that a reasonable shareholder would consider the six-year comparison important to the investment decision. See Milton, 961 F.2d at 969.

We expressly decline to make hard and fast rules about the time length of reported investment results, i.e., we do not hold that ten-year comparisons must always be accompanied by shorter-term comparisons. Nor do we hold that a plaintiff always creates a triable issue of fact by merely unearthing unfavorable news regarding shorter time intervals than those reported.

Moreover, the unfavorable six-year figure in this case does not necessarily render the ten-year comparison misleading. Rather, a jury, knowing the individual annual returns over the ten-year period at issue (which are not now ascertainable on the record before us) and perhaps having other guideposts for determining the relative reliability of shorter- and longer-term bond comparisons, may conclude that the ten-year comparison standing alone is not misleading at all. Because the district court felt it irrelevant that defendants had not reported the claimed six-year negative trend, it gave no attention to whether the Miller plaintiffs had adequately established a factual base — viz., that defendants knew, or reasonably should have known, of that change of circumstances. While we have some doubt about the adequacy of the Miller plaintiffs’ proof of defendants’ knowledge, we. nonetheless recognize that discovery on this issue was limited. We reverse and remand to permit further discovery in this area. Following such discovery, the court may then reconsider defendants’ motion for summary judgment, if defendants choose to renew it.

Thus, on the current state of the record in the Miller case, summary judgment on this issue was improper. We agree with the district court that the ten-year comparison “paints a much rosier picture,” New America, 834 F.Supp. at 507, than the six-year comparison. Having established this fact, the district court erred in concluding in the Miller case that the comparison nonetheless was not misleading as a matter of law.

4. Other Summary Judgment Issues

While fact issues remain with regard to the Treasury security comparison in the Miller case, the district court properly granted summary judgment on all other issues in both cases. For example, plaintiffs alleged that (1) defendants knew or should have known of the effect that “aging” calculations have on determining junk bond returns, and (2) defendants should not have used the DBL composite as an indicator of past performance of *177 junk bonds because that composite failed to account for “forced bond exchanges.” 8

It is doubtless true, as plaintiffs allege, that several significant studies with regard to “aging” discovered statistical infirmities in the traditional methods of calculating junk bond returns. However, these studies were completed only after the prospectuses were issued. Moreover, according to affidavits in the record, the Asquith study was the first study of its kind to display the infirmities of previous calculation methods. Plaintiffs failed to adduce any facts which, contrary to defendants’ affidavits, would tend to show that defendants were aware of these infirmities, or that they could or should have been aware of the effects of “aging” analysis at the time the funds were initially offered to the public. Given plaintiffs’ failure to raise a triable issue of fact, we affirm the district court’s grant of summary judgment on this issue.

A similar analysis disposes of plaintiffs’ allegation that the DBL composite, relied on extensively by defendants in the prospectuses, failed to account for forced bond exchanges. Defendants offered affidavits to the effect that forced bond exchanges in fact were accounted for in the DBL composite. Plaintiffs offer no evidence to the contrary. Accordingly, we find no error in the district court’s grant of summary judgment on this issue. Because further discovery will occur in the Miller case, we leave to the district court the formulation of the extent of that discovery, consistent with the ruling made herein.

Lastly, defendant Prudential Bache, an underwriter of the New America Fund, argues that claims against it were untimely filed. The district court did not rule on when the statute of limitations in this case began to run, nor can we make such a determination on the record before us. Accordingly, we leave this important procedural issue to be determined in the first instance by the district court.

III.

CONCLUSION

We have carefully considered all other arguments and find them to be either waived or without merit. For the foregoing reasons, the various orders of the district court are

Affirmed in full as to the Lucia case, and, as to the Miller case, affirmed in part, reversed in part, and remanded for further proceedings consistent with this opinion.

Notes

. The results of the Asquith study were first made public through various financial and general periodicals in April of 1989. See, e.g., Kenneth N. Gilpin, Further Rise in Rates is Expected, N.Y. Times, Apr. 10, 1989, at D9; Linda Sandler & Michael Siconolfi, Junk Bonds are Taking Their Lumps, Wall St. J„ Apr. 14, 1989, at Cl. The study itself was not published until September 1989. See Paul Asquith, et al., Original Issue High Yield Bonds: Aging Analyses of Defaults, Exchanges and Calls, 44 J.Fin., No. 4 (September 1989).

. The record reveals that, prior to the Asquith study, the traditionally accepted method of determining annual bond default rates was to divide the total number of defaults per year by the total size of the relevant market sector for that year. As the affidavit of Professor Asquith points out, however, this method loses its accuracy in a rapidly expanding market, such as the junk bond market of the 1970’s and '80's, where new issues greatly enlarged the existing market. In other words, the traditional method does not reveal whether a preponderance of older or newer issues are defaulting in a given year.

Breaking from the traditional method of calculation, Asquith’s study tracked the default rate of bonds based on their dates of issuance. The study revealed that junk bonds become more likely to default as they grow older, hence the term "aging.”

. Sections 10(b), 11 and 12(2) all prohibit the use of materially misleading information in the sale of securities, and the same conduct may be actionable under all three sections. See, e.g., Herman & MacLean v. Huddleston, 459 U.S. 375, 382-83, 103 S.Ct. 683, 687-88, 74 L.Ed.2d 548 (1983) (stating that the same conduct may be actionable under sections 10(b) and 11); Shapiro v. UJB Fin. Corp., 964 F.2d 272, 279, 286-89 (3d Cir.) (explaining that single set of factual allegations may state claim under sections 11 and 12(2)), cert. denied,-U.S.-, 113 S.Ct. 365, 121 L.Ed.2d 278 (1992).

While sections 10(b), 11 and 12(2) differ significantly from one another, see, e.g., Herman & MacLean, 459 U.S. at 382, 103 S.Ct. at 687; Ernst & Ernst v. Hochfelder, 425 U.S. 185, 210-11, 96 S.Ct. 1375, 1389, 47 L.Ed.2d 668 (1976), the parties focus solely on the materiality requirement, which is common to all three sections. Cf. In re Donald J. Trump Casino Sec. Litig., 7 F.3d 357, 368 n. 10 (3d Cir.1993) ("Because our analysis here is predicated on the materiality requirement, which is common to [plaintiffs' section 10(b), 11 and 12(2) claims], we do not here distinguish between [those provisions.]”), cer t. denied, - U.S. -, 114 S.Ct. 1219, 127 L.Ed.2d 565 (1994).

.The original complaints also alleged RICO claims and common law fraud claims, which were dismissed by the district court. Plaintiffs do not appeal these dismissals.

. The relevant portion of the New America Fund's prospectus states:

The Fund's portfolio will consist primarily of "high yield” corporate bonds.

"High yield" bonds offer a higher yield to maturity than bonds with higher ratings as compensation for holding an obligation of an issuer perceived to be less credit worthy. The DBL composite measures the performance of the most representative bonds in the "high yield” market and is compiled monthly by Drexel Burnham Lambert Incorporated. As of December 31, 1987, the DBL Composite offered a yield spread of 484 basis points (i.e., 4.84%; 1% equals 100 basis points) over the comparable Treasury security, 7% U.S. Treasury due 1994. U.S. Treasury securities are considered to have minimal risk. The average spread between the DBL Composite and the comparable U.S. Treasury issue was 358 basis points for 1980, 397 basis points for 1981, 503 basis points for 1982, 337 basis points for 1983, 311 basis points for 1984, 362 basis points for 1985, 496 basis points for 1986 and 451 basis points for 1987.

For the years 1977 through 1986, the spread in yields between "high yield” securities and representative U.S. Treasury securities has averaged approximately 393 basis points. For this period, the loss in principal and interest due to defaults on "high yield” securities has averaged approximately 97 basis points. Thus, for the period 1977 to 1986, the net average spread between "high yield” securities and representative U.S. Treasury securities (i.e., the average spread between "high yield" securities and U.S. Treasury securities, minus the average default loss on "high yield” securities) was 296 basis points. For 1987, the loss of principal and interest due to defaults is estimated to have been 125 basis points.* However, past performance is not necessarily indicative of future performance....

The capital structure of the Fund has been designed to take advantage of the historical spread in yields between "high yield” securities and representative U.S. Treasury securities, compared with the average default loss on “high yield” securities.

* Statistical data appearing above are based on information provided by Drexel Burnham Lambert Incorporated.

The Prospect Street prospectus is similarly structured and worded.

We note in passing that the Prospect Street prospectus reports significantly different annual spreads for the years 1980 through 1987. Because neither the Miller nor the Lucia plaintiffs have argued, either below or on appeal, that these inconsistencies are actionable, we deem the issue waived.

. Both revised complaints at ¶ 29 state:

29. The [Asquith] Study also disclosed that high yield debt had not in fact produced higher realized returns and lower standard deviations of returns than either investment grade or treasury bonds for the period 1982 through 1987....

The Asquith study, in turn, relies on statistics from Marshall E. Blume & Donald B. Keim, Volatility Patterns of Fixed Income Securities, Rodney L. White Center For Financial Research, Wharton School, University of Pennsylvania (March 1989) (“the Blume and Keim study”).

. Given adequate grounds to support dismissal of plaintiffs' section 10(b) claims, we expressly decline to address the district court’s "loss causation” analysis, and its use of Bastían v. Petren Resources Corp., 892 F.2d 680 (7th Cir.), cert. denied, 496 U.S. 906, 110 S.Ct. 2590, 110 L.Ed.2d 270 (1990), in rejecting these same claims.

. Forced bond exchanges, also known as "distressed” bond exchanges, occur when a bond issuer, rather than default on its existing obligations, exchanges them for a new set of obligations.