Named plaintiffs Ernest Ross and George Miller claim the defendants were involved in a fraudulent scheme to issue unmarketable tax-exempt bonds and that the plaintiffs (and the class they purport to represent) purchased these bonds in reliance on the integrity of the market. They claim the defendants knew the bonds were not properly tax-exempt, that the defendants knew the underlying construction project would not generate sufficient income to repay the bonds, and that insolvency and default were inevitable after completion of the retirement facility. It is undisputed that neither of the named plaintiffs read the disclosure statement which accompanied the issuance of the bonds. *987Accordingly, plaintiffs rely on the “fraud on the market” theory announced by our predecessor court in Shores v. Sklar, 647 F.2d 462 (5th Cir. May 1981) (en banc),1 cert. denied, 459 U.S. 1102, 103 S.Ct. 722, 74 L.Ed.2d 949 (1983).2
I. BACKGROUND
A. Facts
At this point, a general statement of the facts is sufficient, and we will present these facts in the light most favorable to plaintiffs because there has been no fact finding in this case. Because plaintiffs bear the burden of proof on several relevant issues, however, and because there has been adequate discovery, our resolution of the summary judgment issues (see Section III infra) will depend in part upon the strength of plaintiffs’ evidence. See Celotex Corp. v. Catrett, 477 U.S. 317, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986).
The bonds in controversy were issued to finance the construction of Mount Royal Towers, a residential/medical facility for the elderly, located near the Birmingham suburb of Vestavia Hills, Alabama. Defendant Arthur Rice was the developer of the project.
In 1978, Rice was unable to obtain conventional financing for his project, so he turned to the possibility of financing through tax-exempt bonds.3 In furtherance of this plan, Mount Royal Towers, Inc. was incorporated in 1979 as an Alabama nonprofit corporation. Rice first approached the City of Homewood, Alabama about sponsoring the bonds, but the overture was rejected after the city attorney questioned the legality of sponsoring the project. Rice then contacted the defendant City of Vestavia Hills, which agreed to sponsor the project and to permit the bonds to be issued on its behalf with title to Mount Royal reverting to the city when the bonds were paid. The city established the defendant Special Care Facilities Financing Authority to issue the bonds to build Mount Royal. Rice began negotiations with the underwriting firm of Underwood, Neuhaus regarding the bond issue, but Underwood declined to underwrite the project because conditions in the market were poor even for “safe bonds” and as a result there was “no currently existing market for more speculative types of bond issues” such as Mount Royal. Exh. 108. A group of Birmingham investment bankers also declined to underwrite the project. Rice next turned to the underwriting firm of Henderson, Few & Company. Henderson, Few decided against underwriting the $18,-000,000 bond issue in December of 1980 because of poor conditions in the bond market. Henderson, Few and the accounting firm of Peat, Marwick & Mitchell (who had been retained as a feasibility consultant) advised Rice that the proposed sales price of Mount Royal apartment units was already as high as the Birmingham market *988would bear, but was still insufficient to produce enough revenue to repay bonds carrying sufficient interest rates to be marketable.
During this period, Rice had been engaging in premarketing research. There were to be 205 apartments available in Mount Royal. During 1980, Rice had presold more than 130 units, but this number declined to 107 by September, 1980, and finally declined to zero by the end of that year. This decline allegedly was due to long delays in construction and in obtaining financing. In December, 1980, the board of trustees of Mount Royal Towers, Inc. abandoned the project, citing the cancellation of deposit agreements with potential purchasers, the failure to obtain conventional financing, and the expiration of the IRS 501(C)(3) advance ruling. On January 8, 1981, Peat, Marwick notified Rice that it could not continue to act as an independent feasibility consultant until it was paid for its past work. Rice was reminded that Peat, Marwick’s fee was in no way contingent upon the result of the feasibility study or the issuance of the bonds.
Rice stood to lose a substantial personal investment in the project if the bonds were not issued. Apparently he was undaunted by the failures and warnings of 1980. After being turned down by Henderson, Few, Rice immediately hired a new underwriter (defendant Herreth, Orr & Jones), bond counsel (defendant Jones, Bird & Howell),4 and feasibility consultant (defendant La-venthol & Horwath). A joint venture (defendant Total Concept Retirement Communities) was formed to develop the project. Defendant Peter Wright, an attorney employed by Jones, Bird, drafted the joint venture agreement. The venture was composed of defendant Wellington Corporation (owned by Rice), defendant Finerock Corporation (a subsidiary of Herreth, Orr & Jones) and defendant Robinson-Hall, Inc. (an Atlanta brokerage firm whose principals, along with Attorney Wright, arranged the initial meeting between Rice and Her-reth, Orr). Under this agreement, Finerock was to advance $12,500 per month for development costs and was to be reimbursed with interest out of Mount Royal development fee funds. Although defendants La-venthol and Jones, Bird were not parties to the joint venture, payment of their fees was contingent on the closing of the bond issue. Thus, unlike their counterparts in the earlier failed attempts to market the bonds, each of these new participants had a direct financial interest in the successful marketing of the bonds due to the contingent fee arrangements and the joint venture agreement with advances by Finerock.
The new underwriter determined that the bond issue had to be raised from $18,000,-000 to $29,000,000 to ensure repayment and proper interest rates. To meet this obligation, prices for the Mount Royal apartment units had to be raised. Under the original plan, the units would have ranged in price from $17,000 (studio apartments) to $83,000 (penthouse apartments). Henderson, Few and Peat, Marwick had concluded that these prices were as high as the Birmingham market would bear. The project, as restructured in February, 1981, included units priced from $54,900 for studio apartments to $172,400 for penthouse apartments. As a specific example, plaintiffs note that the very same apartment “presold” in 1980 for $72,900 with $976 in monthly service fees would now cost $148,500 with $1,542 in monthly fees. Plaintiffs claim the prices were increased solely to make the bonds appear marketable, and that the defendants simply disregarded the advice of Henderson, Few and Peat, Marwick that the price of the units was already as high as the market would bear. Defendants counter that the higher prices were justified because, under their “new” pricing structure, purchasers of apartments would find it easier to receive refunds of their purchase prices in the event of death or dissatisfaction.
The next step in the “venture” was to pre-sell 50% of the units before issuing the bonds. This requirement was imposed by *989the original underwriters to ensure marketability of the bonds, and was carried forward into the new project’s structure. In fact, the original feasibility consultants believed that the Mount Royal bond issue would be unmarketable without substantial pre-sale commitments. By April 1, 1981, only 33 of the 205 units, or 17%, were presold. This was a loss of 74 of the 107 pre-sale customers with whom Rice had contracted prior to doubling the sales price. By June 1,1981, 47 units, or 24%, had been presold. Rice then decided to boost sales by reducing the required deposit from $1,000 to zero. Since sales were still lacking, plaintiffs allege Rice engaged in sham transactions by pre-selling units to his salesmen, his friends, his relatives, etc., who had no intention of actually buying the units. These tactics enabled Rice to obtain the 103 commitments represented in the offering circular. Within six months of the issuance of the bonds, all 48 people who were presold without deposits had can-celled their orders.
Defendant bond counsel rendered its opinion that the proposed bonds would be tax-exempt. Plaintiffs now allege that counsel knew the bonds would not be tax-exempt, and rendered its opinion only to assure that the bonds would be issued. According to plaintiffs, the bonds were and are not tax exempt because private individuals (i.e., Rice) profited from the Mount Royal construction project, and because the Mount Royal apartment units were not going to be available to the general population of elderly citizens. Because of the high price of units, they would be available only to the affluent elderly. In addition, plaintiffs claim bond counsel as well as the indenture trustee (defendant Bank South) and the issuer defendants (the City of Ves-tavia Hills and the Special Care Financing Authority) were each aware or recklessly disregarded the above facts which undermined the feasibility of the Mount Royal project. Plaintiffs contend the defendants heedlessly proceeded to issue the apparently marketable bonds without disclosing to the public the past failures to obtain financing and Rice’s efforts to inflate the number of presold apartments.
The bonds were issued on September 30, 1981, with a face amount of $29,950,000. The disclosure statement noted that the bonds were not guaranteed by public funds and were high risk bonds because repayment depended entirely on the sales of apartments. Because of this risk, the bonds bore interest rates ranging from 15.5% to 17% tax free. Some of the original $29,950,000 was put into a separate account to pay the bondholders during the construction period. By March, 1984, the bondholders had received approximately $11,000,000.
Mount Royal Towers was constructed with $16,000,000 from the bond issue. Unfortunately, due to poor sales Mount Royal filed for Chapter 11 reorganization on April 15, 1984, and the bond issue went into default in September, 1984. Mount Royal Towers continued to operate during the reorganization and is still operating although many of the residents are renters. The complex was sold pursuant to a bankruptcy approved sale in October, 1985, and $18,000,000 was distributed to the bondholders to satisfy the balance of the outstanding indebtedness to the bondholders of $29,950,000.5
B. Proceedings In The District Court
In early 1985, plaintiffs Ross and Miller brought separate actions on behalf of all Mount Royal bondholders. These were consolidated by the district court. In its order certifying the class, the district court narrowed the class to include only those bondholders who purchased Mount Royal bonds prior to their default in 1984, and who relied on a fraud on the market theory as enunciated in Shores v. Sklar. In 86-*9907350, several defendants appeal the district court’s certification order. For the reasons discussed in Section II, infra, we affirm the district court’s order certifying the class.
On September 18, 1985, the district court indicated it would grant motions to dismiss filed by the “issuer” defendants (i.e., the city, the Special Care Financing Authority). On this same date, the court converted Bank South’s motion to dismiss into a motion for summary judgment and asked the parties to file affidavits. The court granted Bank South’s motion for summary judgment on December 12, 1985. In 86-7352, the plaintiffs appeal these two orders. For the reasons discussed in Sections 111(B)(1) and (2), we affirm the district court’s orders dismissing these defendants.
Then, on August 18, 1986, the district court granted summary judgment in favor of most of the defendants on all claims. After this order was filed, the rest of the defendants in the case filed for summary judgment. The district court entered final judgment against plaintiffs on October 16, 1986. In 86-7790, plaintiffs appeal this order. For the reasons discussed in Section III, infra, we affirm the district court’s order only to the extent it dismisses the claims against the Mount Royal trustees. We reverse that portion of the district court order dismissing the claims against the remaining defendants and remand for further proceedings.
II. CLASS CERTIFICATION
The first issue considered on appeal is whether the district court abused its discretion in certifying this case as a class action with plaintiffs Ross and Miller as class representatives. The defendants/cross-appellants assert that Ross and Miller are subject to the “unique” defense that they did not rely on the integrity of the market in purchasing the bonds. Therefore, according to defendants, the named plaintiffs’ claims are not typical of the class members’ claims, and, consequently, named plaintiffs are not adequate class representatives. In order to assess this argument we must, first, discuss generally the extent to which the applicability of a unique defense jeopardizes a named plaintiff’s ability to represent a class adequately. Second, we must analyze Shores to determine the meaning of the reliance requirement imposed by that case. Third, we must review the evidence placed before the district court to determine the extent to which plaintiffs Ross and Miller were in fact subject to a nonreliance defense. Finally, we must determine whether the named plaintiffs’ susceptibility to such a defense precludes these plaintiffs from representing the class.
A. The Unique Defense Problem
This case was certified as a 23(b)(3) class action. Accordingly, the named plaintiffs were required to satisfy the elements of Fed.R.Civ.P. 23(a) (numerosity, commonality, typicality, and adequacy of representation) and the requirement of rule 23(b)(3) that common issues predominate over individual issues so that “a class action is superior to other available methods for the fair and efficient adjudication of the controversy.”
The claim by defendants that the named plaintiffs are subject to a unique defense (i.e., not generally applicable to the class) questions the commonality and typicality of the named plaintiffs’ claims and postulates that the named plaintiffs are not adequate class representatives. The district court correctly noted that the unique defense argument could be addressed under any of the rule 23(a) tests. See Appleyard v. Wallace, 754 F.2d 955, 958 (11th Cir.1985) (considerations of rule 23(a)(2), (3), & (4) tend to overlap).
The existence of a unique defense certainly is relevant to the certification decision. See generally, 7A C. Wright, A. Miller & M. Kane, Federal Practice and Procedure § 1764, p. 259-60 & n. 27 (1986) (typicality requirement “has not been met in securities class actions in which unique defenses would be applicable to the representative parties’ claims;” citing numerous cases). The existence of even an arguable defense can vitiate the adequacy of repre*991sentation if it will distract the named plaintiffs attention from the issues common to the class. See Koos v. First National Bank of Peoria, 496 F.2d 1162, 1165 (7th Cir.1974); McNichols v. Loeb Rhoades & Co., 97 F.R.D. 331, 334 (N.D.Ill.1982) (in order to defeat class certification, possible defense need only be “unique, arguable and likely to usurp a significant portion of the litigant’s time and energy”). The certification of a class is questionable “where it is predictable that a major focus of the litigation will be on an arguable defense unique to the named plaintiff or a small subclass.” Koos, 496 F.2d at 1164; Kas v. Financial General Bankshares, Inc., 105 F.R.D. 453, 461 (D.D.C.1984); see Kline v. Wolf, 88 F.R.D. 696, 699-700 (S.D.N.Y.1981) (plaintiffs deposition testimony on critical issue conflicted with testimony of plaintiffs broker; credibility problem rendered named plaintiffs claim atypical), aff'd in pertinent part, 702 F.2d 400, 403 (2d Cir.1983) (“incredibility might support a rebuttal unavailable against many other class members, i.e., that plaintiffs did not significantly rely on the integrity of the market”).
Although the existence of an arguable defense can be taken into account by the district court, the court is not required to deny certification for such speculative reasons. Questions concerning class certification are left to the sound discretion of the trial court. See Griffin v. Carlin, 755 F.2d 1516, 1531 (11th Cir.1985); Walker v. Jim Dandy Co., 747 F.2d 1360, 1363 (11th Cir.1984); McKinnon v. Talladega County, 745 F.2d 1360, 1365 (11th Cir.1984). The district court’s decision to certify the class will be upheld absent an abuse of discretion. Griffin, 755 F.2d at 1531. Accordingly, we will not reverse the district court’s decision certifying the class merely because the defendants have raised a defense which can be termed “arguable.” Some showing must be made by defendants that the district court misunderstood the potential significance of the asserted defense.
In this case, the defendants claim plaintiff Miller’s reliance on the misrepresentations of his broker (and not his reliance on the integrity of the market) caused him to purchase the bonds. Furthermore, defendants claim both named plaintiffs have indicated in their depositions that they would not have purchased the Mount Royal bonds if they had been aware of information disclosed in the offering circular. According to defendants, this testimony will make it difficult (if not impossible) for named plaintiffs to prove that they “relied on the bonds’ availability on the market as an indication of their apparent genuineness.” Shores, 647 F.2d at 469-70. The district court concluded that these asserted defenses were “not so unique as to preclude class certification. If there has been even a partial reliance on the integrity of the market, the claim is not atypical.” Record, Vol. 4, Tab 273 at 16. In making this determination, the court recognized that the law in this circuit regarding the relationship between reliance and class certification under Shores is far from clear.6 For the reasons discussed below, we believe the district court acted within its discretion in certifying the class.
B. Reliance On The Integrity Of The Market
Reliance is a requirement in all 10b-5 cases. In the traditional 10b-5 misrepresentation case, the plaintiff must establish: (1) a misstatement or omission, (2) of a material fact, (3) made with scienter, (4) on which the plaintiff relied, (5) that proximately caused his injury. Huddleston v. Herman & MacLean, 640 F.2d 534, 543 (5th Cir. Unit A Mar. 1981), rev’d on other grounds, 459 U.S. 375, 103 S.Ct. 683, *99274 L.Ed.2d 548 (1983); Dupuy v. Dupuy, 551 F.2d 1005, 1014 (5th Cir.1977), cert. denied, 434 U.S. 911, 98 S.Ct. 312, 54 L.Ed.2d 197 (1977). In these cases, plaintiffs’ reliance established “but for” causation: had the investor known the truth he would not have purchased or sold the security. Huddleston, 640 F.2d at 549. Plaintiff must further establish that “the untruth was in some reasonably direct, or proximate, way responsible for his loss” (i.e., that the misrepresentation proximately caused the price of the security to fall). Id. The reliance and causation requirements prevent 10b-5 from becoming a private enforcement mechanism or a system of investor insurance. See Lipton v. Documation, Inc., 734 F.2d 740, 742 (11th Cir.1984), cert. denied, 469 U.S. 1132, 105 S.Ct. 814, 83 L.Ed.2d 807 (1985); Dupuy, 551 F.2d at 1016.
When the defendant has a duty to disclose, and the disclosure omits a material fact, plaintiffs reliance on the omission is presumed. Affiliated Ute Citizens v. United States, 406 U.S. 128, 153-54, 92 S.Ct. 1456, 1472, 31 L.Ed.2d 741 (1972). The presumption is employed because it is so difficult for a plaintiff to prove a negative: that he purchased or sold the security because of what the defendant did not disclose. See Shores, 647 F.2d at 468. This presumption can be rebutted if defendant proves the plaintiff’s decision would have been the same even if the omitted information had been disclosed. Lipton, 734 F.2d at 742. The presumption is also rebutted where the plaintiff admits that he never read or otherwise relied on the allegedly deficient disclosure statement. Shores, 647 F.2d at 468.
Courts have adapted the reliance requirement to allow recovery in some cases where the plaintiff admits that he did not read or otherwise rely directly on the relevant disclosure materials. In these cases, plaintiff’s failure to rely directly upon the disclosure statement is not dispos-itive where the defendant’s conduct constituted a “fraud on the market.” In order to satisfy the reliance requirement in these cases, plaintiff must establish that but for the fraud on the market he would not have purchased or sold the security. Shores, 647 F.2d at 471-72. Again, proof of reliance entails proof of causation in fact.7
It is important to distinguish between the two general types of fraud on the market cases. The first type involves a claim by a plaintiff who purchases a security traded on a developed and open market, but who does not first read or otherwise rely on a prospectus or other relevant disclosure document. Typically, the plaintiff in such a case claims that the defendant’s fraud artificially inflated the price of the security. When the fraud becomes public, the price of the security falls, and the plaintiff loses money. Whether or not the plaintiff in such a case relied on a specific misrepresentation or omission, she may be entitled to relief because she relied “generally on the supposition that the market price is validly set and that no unsuspected manipulation has artificially inflated the price, and thus indirectly on the truth of the representations underlying the stock price_” Blackie v. Barrack, 524 F.2d 891, 907 (9th Cir.1975), cert. denied, 429 U.S. 816, 97 S.Ct. 57, 50 L.Ed.2d 75 (1976). This formulation of the reliance requirement is consistent with the role of reliance as proof of causation.
Blackie, the seminal case in this area, has been followed by every court (including our court) faced with a similar situation. See, e.g., Peil v. Speiser, 806 F.2d 1154 (3d Cir.1986); Lipton v. Documation, Inc., 734 F.2d 740 (11th Cir.1984), cert. denied, 469 U.S. 1132, 105 S.Ct. 814, 83 L.Ed.2d 807 (1985); Rosenberg v. Digilog, Inc., 648 F.Supp. 40 (E.D.Pa.1985); Grossman v. Waste Management, Inc., 589 F.Supp. 395 (N.D.Ill.1984); In re LTV Securities Litigation, 88 F.R.D. 134 (N.D.Tex.1980).
*993Proof of reliance in this first type of fraud on the market case raises difficulties of proof similar to those raised in material omission cases. Thus, reliance is presumed. Lipton, 734 F.2d at 748; Blackie, 524 F.2d at 906. Plaintiff establishes a prima facie case by showing that she purchased the security and that the misrepresentation (or other deception) was material. “Materiality circumstantially establishes the reliance of some market traders and hence the inflation in the stock price— when the purchase is made the causational chain between defendant’s conduct and plaintiffs loss is sufficiently established to make out a prima facie case.” Id. at 906 (citations omitted). The presumption, derived from Ute, is particularly appropriate for fraud on the open market eases because the causal nexus is “adequately established indirectly, by proof of materiality coupled with the common sense that a stock purchaser does not ordinarily seek to purchase a loss in the form of artificially inflated stock.” Id. at 908 (footnote omitted).
As in a Ute omission case, the presumption of reliance in a Blackie fraud on the market case is rebuttable. Defendant can rebut the presumption either by disproving materiality or by proving that; (1) despite materiality, an insufficient number of traders relied on the deception so as to inflate the price; or (2) the individual plaintiff purchased despite knowledge of the deception; or (3) the plaintiff would have purchased even if he had known of the deception. Blackie, 524 F.2d at 906.
The second (and far less common) type of fraud on the market cases involve fraud in the issuance of securities traded on an undeveloped (primary) market. These cases involve plaintiffs who purchased newly-issued securities without reading the offering circular or similar disclosure documents. When these investments result in a loss, plaintiffs may in some cases recover under 10b-5 if they can show that but for defendant’s fraud the security would not have been marketed and that they relied on the integrity of the market to produce “marketable” securities. Shores v. Sklar is the lead case of this type. See also T.J. Raney & Sons, Inc. v. Fort Cobb, 717 F.2d 1330 (10th Cir.1983) (adopting Shores), cert. denied, 465 U.S. 1026, 104 S.Ct. 1285, 79 L.Ed.2d 687 (1984).
In Shores, plaintiff alleged a pervasive fraudulent scheme on the part of a developer, bond counsel, underwriter, accountant, and indenture trustee designed to induce a municipality to issue, and the public to buy, otherwise unmarketable bonds. The developer knew his company did not have the financial capability to construct the mobile homes which were to be sold to repay the bonds. He nevertheless determined, as part of a scheme to defraud the investing public, to induce the town of Frisco City, Alabama, to issue the bonds. Bond counsel drafted the offering circular but intentionally or recklessly omitted facts impugning the developer’s integrity and financial stability. The defendant accountant prepared a financial statement which he knew was materially misleading as to the developer’s financial stability. The underwriter offered the bonds for sale, even though it knew the offering circular contained material misrepresentations and omissions. The indenture trustee knowingly or recklessly disregarded its duty to ensure that the developer maintained at least $400,000 in working capital. As a result of this fraudulent scheme, the town issued bonds which, but for the fraud, could not have been marketed at any price. Plaintiff Bishop purchased the bonds based solely on his broker’s oral representations that tax-free industrial bonds generally were a good investment. He did not read or otherwise rely upon the misleading offering circular. The bonds subsequently went into default and the investors received approximately one-third of their original investment. Bishop brought a class action suit on behalf of the bond holders. 647 F.2d at 463-67.
The district court dismissed Bishop’s 10b-5 claim because Bishop admitted that he had never read or otherwise relied on the offering circular. On appeal, the en banc court affirmed the dismissal to the extent the complaint alleged the “usual” 10b-5(b) misrepresentation claim. Bishop’s *994failure to rely directly on the offering circular precluded a claim that he relied directly on the misrepresentations or omissions contained therein. Id. at 468.
Notwithstanding Bishop’s failure to read the offering circular, the majority held he did state a cause of action under the broader language of 10b-5(a) and (c).8 Bishop’s complaint alleged that “the defendants engaged in an elaborate scheme to create a bond issue that would appear genuine but was so lacking in basic requirements that the bonds could never have been approved by the Board nor presented by the underwriters had any one of the participants in the scheme not acted with intent to defraud in reckless disregard of whether the other defendants were perpetrating a fraud.” 647 F.2d at 468. When the fraud alleged is of such a broad scale as to undermine the validity of the bonds’ very existence, plaintiff’s lack of reliance on the offering circular is not determinative. Id. at 469. Instead, plaintiff must show that:
(1) The defendants knowingly conspired to bring securities onto the market which were not entitled to be marketed, intending to defraud purchasers,
(2) Plaintiff reasonably relied on the bonds’ availability on the market as an indication of their apparent genuineness, and
(3) As a result of the scheme to defraud, plaintiff suffered a loss.
Id. at 469-70. See also Lipton, 734 F.2d at 743-45 (summarizing the holding and rationale of Shores).
The instant case requires our court to examine for the first time the nature of the reliance requirement in a case brought under Shores. It is only after examining this requirement that we can assess the impact of defendant’s asserted nonreliance defense on the maintainability of this class action. As is explained above, plaintiff in a Blackie “fraud on the market” case must demonstrate causation in fact in order to satisfy the reliance requirement. This same guiding principle should govern under Shores. The court pointed out that “the role of reliance in securities actions is to establish causation,” and that “[whenever the Rule 10b-5 issue shifts from misrepresentation or omission in a document to fraud on a broader scale, the search for causation must shift also.” 647 F.2d at 471-72; see T.J. Raney, 717 F.2d at 1333. Reliance in a Shores type case arises not from reading a disclosure statement but instead from the bond buyer’s reliance on the market to produce securities that were not fraudulently created. 647 F.2d at 472.
Accordingly, a prima facie showing of reliance is established in a Shores type ease upon proof of a scheme to bring securities onto the market which were not entitled to be marketed. But for the fraud in such a case, plaintiff could not have purchased the bond because the bonds would not have been offered for sale. As one district court has said:
If plaintiffs employ a fraud on the market theory, individual issues of reliance virtually disappear because a sufficient causal link is established by the mere presence of the bonds in the marketplace. Individual questions of knowledge and reliance simply become irrelevant.
Dekro v. Stem Bros. & Co., 540 F.Supp. 406, 417 (W.D.Mo.1982). If plaintiff can prove the type of pervasive fraud necessary to establish a fraudulent marketing of *995securities,9 he has perforce satisfied the cause in fact reliance requirement of 10b-5.
Defendants point to language in Shores indicating that plaintiff must make an additional showing of “reliance on the integrity of the market” in order to recover. Indeed, the court in Shores set out as a separate element of the action that plaintiff must prove he “reasonably relied on the bonds’ availability on the market as an indication of their apparent genuineness.” 647 F.2d at 469-70. Further, the court indicated that the “requisite element of causation in fact would be established if [plaintiff] proved the scheme was intended to and did bring the bonds onto the market fraudulently and proved he relied on the integrity of the offerings of the securities market.” Id. at 469 (emphasis added). According to defendants, this language requires a Shores plaintiff to prove in some way that his decision to purchase the security was based entirely on his individual assessment of the market as a filtering mechanism and that he would have purchased the securities even if he had read the offering circular. We do not read Shores to impose any such burden of proof.
Shores recognized that the “securities laws allow an investor to rely on the integrity of the market to the extent that the securities it offers to him for purchase are entitled to be in the marketplace.” 647 F.2d at 471.10 As in the Blackie case on which Shores relied, such reliance is presumed with respect to all purchasers of securities. It is common sense that purchasers do not ordinarily seek to purchase a loss in the form of a fraudulently marketed bond. Consequently, once the plaintiff proves that the deception was material (i.e., without it, the bonds would not have been marketed), reliance on the integrity of the market is presumed.
The rebuttal of this presumption is more difficult under Shores than under Blackie. In a Blackie type case, the defendant can show that his fraud had no effect on the market price. 524 F.2d at 906. This defense severs the causal link between the fraud and the plaintiff’s loss. A Shores defendant can make no such showing, however, because his fraud creates the bond itself. Without the fraud, the plaintiff could not have purchased the bond. Shores, 647 F.2d at 464 n. 2.
A Shores defendant can rebut the presumption of reliance only by showing that the plaintiff’s reliance was not reasonable. Reliance on the market is reasonable so long as the plaintiff did not intentionally refuse to investigate known or obviously present risks associated with the bonds. Shores, 647 F.2d at 470 n. 6 (quoting Dupuy v. Dupuy, 551 F.2d at 1020). This rebuttal could include a showing that “an individual plaintiff purchased despite knowledge of the falsity of a representation, or that he would have, had he known of it.” Blackie, 524 F.2d at 906. Unless the defendant shows a compelling reason, such as plaintiff’s inside information regarding the fraud or unreasonable failure to investigate, defendant has not rebutted the general presumption that investors rely on the integrity of the market to produce bonds which are entitled to be marketed.
The defendants argue that Shores turns 10b-5 into an investor insurance policy. *996Indeed, once it can be established that the defendant fraudulently marketed a bond issue, most (if not all) of the purchasers of those bonds will have a cause of action, even if they did not read the offering circular. We do not suggest that sellers of bonds guarantee to bond purchasers that economic conditions will remain stable so that the bonds will be paid off. Investors must, of course, take the risk of default into account when making their purchases, because in the absence of fraud it is the investor who must bear the risk. But where the sellers of bonds are aware that the bonds are doomed to end in early default, and are aware that the bonds would be unmarketable but for fraudulent nondisclosure, the sellers must be held accountable to all purchasers who were unaware of the facts rendering the bonds unmarketable. If this result resembles insurer liability, it is only because the defendant’s fraud affected the entire class of purchasers. Shores stands for the simple proposition that fraud on a broader scale that permeates the very issuance of the bonds gives rise to liability on a broader scale.
C. Evidence Of Nonreliance In This Case
Defendants’ primary argument is that Shores actions universally are inappropriate for class certification because individual issues of reliance are certain to predominate over issues common to the class. Our analysis of Shores in the preceding section necessarily rejects this argument. The nonreliance “defense” is a narrow one which, we suspect, will not apply to a great many bond purchasers. Accordingly, we agree with the district court that the predominant question in a Shores action normally is the existence of a fraudulent scheme to market securities.11 Cf. Kennedy v. Tallant, 710 F.2d 711, 717 (11th Cir.1983) (“the degree of reliance placed on the prospectuses by the various class members is not a controlling element of the action”).
In the alternative, defendants argue that the deposition testimony of plaintiffs Ross and Miller render these plaintiffs particularly susceptible to a nonreliance defense. As to plaintiff Ross, defendants suggest that he was unwilling to accept even those risks disclosed in the offering circular. Ross testified that he was seeking a “safe” investment and that he would have been unwilling to invest if told that the bonds were offering a high interest rate to compensate for the substantial risk involved in the venture. Defendants point out that the offering circular stated that substantial risks were involved in Mount Royal and did explain that the high interest rate was designed to compensate purchasers for the risk of nonpayment.
Defendants assert that the presumption of reliance in a Shores case can be rebutted by showing that an investor would not have made the investment if he had known of the risks disclosed in the offering document or documents. We disagree. Although plaintiffs’ actual or constructive knowledge of the fraud would be inconsistent with blind reliance on the market, an inquiry into what plaintiff would have done if he had read the disclosure statement is not relevant to the reliance issue.
In support of their position that such inquiry would be relevant, defendants refer to language in the Shores opinion describing the plaintiff in that case. Plaintiff in Shores alleged that he was “willing to accept any marketable risk.” As long as a municipal authority was willing to authorize the issuance of a bond and the underwriter was willing to present the issue to the market, that was enough for plaintiff’s purposes. 647 F.2d at 469. Defendants contend Shores should be limited to plaintiffs who are willing to accept “any marketable risk.”12 We decline to limit Shores *997to its facts in this manner. The majority set forth a three-part test for recovery. Nothing in this test indicates that plaintiffs must show their willingness to accept “any marketable risk.”13 Shores was intended to offer protection to the general investing public from frauds which cause otherwise unmarketable securities to be marketed. Whether or not the investor has relied on the offering circular she may assume that the bonds are at least entitled to be marketed.
The investor who fails to read or otherwise rely upon disclosure statements forfeits many otherwise available legal rights. If he “proves no more than that the bonds would have been offered at a lower price or at a higher rate [but for the fraud], rather than that they would never had been marketed, he cannot recover.” Shores, 647 F.2d at 470. But even the risk averse investor who fails to read disclosure statements to ensure that his purchases are safe is entitled to a bottom line reliance on the marketability of issued bonds. This rule of law accommodates the two main purposes of rule 10b-5. On the one hand, it fosters full disclosure policy by retaining the purchaser’s incentive to read disclosure statements to ensure he gets full value for his investment. Id. On the other hand, it pays due homage to the central purposes of the securities acts — protecting investors from fraudulent schemes and promoting free and honest securities markets. Id.
Accordingly, when a fraudulent scheme is so pervasive that it satisfies the first prong of the Shores test, id. at 469, only a plaintiff whose reliance was inconsistent with his right to rely on the market may be barred from recovery. Defendants have demonstrated nothing about plaintiff Ross which indicates that he knew of the fraud or that he had information which would have triggered a duty to investigate for fraud. Defendants have offered nothing which would rebut the presumption that plaintiff Ross relied on the integrity of the market in purchasing the Mount Royal bonds.
Defendants also contend that plaintiff Miller is an inadequate class representative. Miller testified in his deposition that he relied on the oral representations of his broker in making his decision to purchase the Mount Royal bonds. Miller’s broker (unlike the broker in Shores) misrepresented several apparently material facts.14 For instance, the broker asserted that the Mount Royal bonds were municipal bonds backed by the resources of the City of Vestavia Hills. The bonds in fact were not backed by the city, and the offering circular pointed out that the bonds would be repaid (if at all) from the sales of the Mount Royal apartment units. Miller testified that if his broker had disclosed the facts as reported in the offering circular, Miller would not have purchased the bonds. Defendants claim Miller’s reliance on his broker’s misrepresentations and his unwillingness to accept the risks disclosed in the offering circular are both inconsistent with his claim that he relied on the integrity of the market.
We already have rejected the claim that a Shores ’ plaintiff’s hypothetical unwillingness to accept the disclosed risks undermines his claim of reliance on the integrity of the market. The addition of the broker’s misrepresentations does not alter our conclusion. At most, the defendants (if there was fraud) and the broker are con*998current causes of Miller’s decision to purchase the bonds. Miller would not have purchased but for his broker’s misrepresentations and could not have purchased but for the defendants’ fraud on the market. We agree with the district court that Miller’s reliance in part on factors other than the market does not defeat his claim of Shores reliance.
D. Class Certification
As is noted above, the district court might have been justified in rejecting class certification in this case because of the defendants’ assertion of an arguable nonre-liance defense. The district court chose to certify the class, and given our view of the strength of the asserted defense, we do not believe that decision constituted an abuse of discretion. Accordingly, in 86-7350, we affirm the district court.
III. MERITS
The district court handled this massive litigation by entertaining various motions to dismiss and for summary judgment during the course of discovery. The court first granted the issuer defendants’ (City of Vestavia Hills, Special Care Financing Authority) motions to dismiss on the ground that plaintiffs’ complaint did not allege specifically how these defendants participated in the fraud. The court then granted summary judgment in favor of Bank South (indenture trustee) on the ground that there was insufficient evidence of Bank South’s complicity with the alleged fraud to support a 10b-5 claim. Plaintiffs’ sole contention on appeal as to this order is that the order was issued prior to the completion of discovery which might have supported plaintiffs’ allegations. Both of these orders are appealed in 86-7350. While that appeal was pending, discovery proceeded in the court below and the district court ultimately granted summary judgment in favor of the remaining defendants. This final order is appealed in 86-7790. Because this order finds insufficient evidence of a fraudulent scheme, it affects plaintiffs’ claims as to all defendants. It is only if we reverse in 86-7790 that we need to address the orders involving Bank South and the issuer defendants. Accordingly, we will proceed directly to our analysis of 86-7790.
A. Fraud As To The Feasibility Of Mount Royal Towers
The heart of plaintiffs’ claim is that the defendants marketed the Mount Royal bonds even though they knew or recklessly disregarded the fact that the project could not generate sufficient income to repay the debt. Specifically, it is alleged defendants were aware that the long delays in construction “burned” the sales territory so that consumer interest in the Mount Royal units was irreparably damaged. Second, the defendants’ failed to heed the warnings of the original underwriter and feasibility consultant that the project could not generate sufficient income to repay the bonds at the going interest rates. Instead, defendants hired new experts who allegedly agreed to modify the project (by increasing the price of apartments and by increasing the bond interest rates) to make it appear feasible, although they knew that the changes would render the apartments too expensive to be sold in the relevant market. Third, the other defendants allegedly were on notice of Rice’s sham presale agreements which made the project appear feasible. Finally, defendants were able to market the bonds only because they failed to disclose facts which seriously undermined the conclusion warranted in the offering circular that the project was feasible. The district court reviewed the evidence presented by the parties and concluded that there was no evidence from which a jury could reasonably infer a scheme to bring securities on to the market which were not entitled to be marketed.
Before discussing the avalanche of evidence adduced by the parties, we must clarify the scope of our inquiry. First, our review of the district court order granting summary judgment is plenary. See Carlin Communications v. Southern Bell, 802 F.2d 1352, 1356 (11th Cir.1986).
Second, our review is affected by the fact that substantial discovery has tak*999en place and that the nonmoving party (plaintiffs) would bear the burden of proving the relevant issues at trial. In such a case, summary judgment should be granted if plaintiffs “fail to make a showing sufficient to establish the existence of an element essential to [their] case.” See Celotex, 106 S.Ct. at 2552-53. This showing need not be in the form of admissible evidence — it can be made by reference to affidavits, depositions, answers to interrogatories, and the like. Id. at 2553-54. Summary judgment is not, however, designed to be a trial on affidavits. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 2513, 91 L.Ed.2d 202 (1986). “The evidence of the nonmovant is to be believed, and all justifiable inferences are to be drawn in his favor.” Id. Essentially, the standard for granting summary judgment mirrors the standard for granting directed verdict — it will be granted “if, under the governing law, there can be but one reasonable conclusion as to the verdict.” Id., 106 S.Ct. at 2511.
Third, our inquiry is affected by the nature of the questions at issue. Plaintiffs must prove that (1) each defendant to be held liable acted with scienter; (2) in participating in a scheme to defraud the investing public; and (3) that but for the fraud the bonds could not have been marketed. Courts have long held that summary judgment generally is inappropriate to decide questions of scienter, knowledge, and intent. See, e.g., Poller v. Columbia Broadcasting System, 368 U.S. 464, 473, 82 S.Ct. 486, 491, 7 L.Ed.2d 458 (1962) (“summary procedures should be used sparingly in complex antitrust litigation where motive and intent play leading roles, the proof is largely in the hands of the alleged conspirators, and hostile witnesses thicken the plot”); Levinson v. Basic, Inc., 786 F.2d 741, 749 (6th Cir.1986), cert. granted, — U.S.-, 107 S.Ct. 1284, 94 L.Ed.2d 142 (1987); Amey, Inc. v. Gulf Abstract & Title, Inc., 758 F.2d 1486, 1502 (11th Cir.1985), cert. denied, 475 U.S. 1107, 106 S.Ct. 1513, 89 L.Ed.2d 912 (1986). Moreover, factual questions regarding the “marketability” of bonds absent the fraud generally are not susceptible to determination as a matter of law. See Shores, 647 F.2d at 486 (Randall, J., dissenting) (“[t]he question whether ... the security, absent the fraud, could have been marketed at some price, and the question whether the scale of the fraud was broad enough are both presumably questions of fact, not readily disposed of by summary judgment”).
1. The Existence of a Fraudulent Scheme
With these standards in mind, we turn now to the evidence presented to the district court. There is no doubt that the fraud alleged in this case is of a different character than the fraud alleged in Shores and is not as pervasive. In Shores, the entire transaction was a sham, and an apparently safe investment was merely a hollow shell. In this case, on the other hand, the defendants disclosed that the project entailed substantial risk. Nevertheless, the essence of a Shores action is that but for the fraud the bond issue would not have been marketed. Thus, when a fraud passes an unmarketable security off as a risky project, it is actionable to the same extent as a fraud which attempts to pass a worthless project off as a safe investment. In either case, the defendants have defrauded the market by knowingly or recklessly issuing bonds which are destined to end in default. Thus, while Shores stresses the requirement of “fraud on a broader scale,” we note that the breadth of the fraud is measured by its effect on the marketability of the bond, and not by the extent of deception.
Plaintiffs offered the following circumstantial evidence of a fraudulent scheme. The 1978-79 presale marketing of Mount Royal units initially was quite successful, and Rice was able to obtain a substantial number of commitments secured by deposits. Rice was unable, however, to obtain financing for the project. In a December 23, 1980 meeting, the Mount Royal Board of Trustees adopted a resolution abandoning the project and noting that one-half of the presale agreements had been cancelled due to long delays in construe*1000tion. The resolution indicated that the delays had a negative impact upon the success of the project. Despite this finding, Rice began organizing financing for another try at developing Mount Royal. According to plaintiffs, the relentless pursuit of the project was motivated primarily by Rice’s desire to recoup his investment in the project. The new deal was structured so that Rice could obtain payment upon the closing of the bond issue, thus diminishing his incentive to market the units aggressively after construction was completed.
Plaintiffs also offered some evidence that defendants ignored the advice of Henderson, Few and Peat, Marwick when they restructured the bond project in early 1981. Thomas Kelly, executive vice-president of Henderson, wrote a letter to Rice in December, 1980 indicating that the Mount Royal bond issue could not be marketed given investor reluctance to purchase risky bonds. In the letter, Kelly suggested cutting construction costs to maintain the feasibility of the project:
Undoubtedly the project will be faced with costs increases as a result of the delay. I do not think we can add significant additional costs to the bond issue budget and still have a feasible financing package. Therefore, you should discuss with the architect and contractor the possibility of reducing the scope of the project by an amount at least equal to the construction increases they anticipate. Hopefully they will be able to reduce the project’s scope without effecting [sic] the revenue-producing ability of the facility.
Exhibit 109. In his deposition, Kelly testified that during 1980 the parties anticipated issuing approximately $20,000,000 of bonds at 11V2% interest. When market interest rates went up, they were unable to raise the interest rates of the bonds because there would not have been sufficient cash flow from the project to cover the increased debt. One way of increasing cash flow would have been simply to raise the price of the apartments. Kelly testified that, as far as he knew, such increases would have rendered the project infeasible:
My recollection is that it was the opinion of the financial feasibility consultant that the rates and charges that were originally proposed were appropriate to the market conditions in Birmingham, Alabama, and that any significant increases in those rates and charges could have had an effect on the marketability or the acceptability of the project in the marketplace. That is my recollection of the opinion of the financial feasibility consultant.
Kelly Deposition at 151. While this evidence is not conclusive, a reasonable jury could infer that Rice had been warned not to raise the prices of Mount Royal apartments and that he knew raising the prices would jeopardize the project’s feasibility.
Soon after Henderson, Few withdrew its support and after the board of trustees officially abandoned the project, Rice engaged defendants Herreth, Orr & Jones, Laventhol & Horwath, and Jones, Bird & Howell to assist in restructuring the project so that financing could be obtained. The “new” project differed from the abandoned project in several significant ways. The bond issue was increased from $19,-950,000 to $29,950,000. To cover the increased debt, the price of apartments was raised dramatically. The increased prices were justified by including in the purchase agreements a provision giving purchasers a right to a refund if they were dissatisfied. This right to a refund was subject, however, to the prior rights of bondholders in the event of a default. The district court found, as a matter of law, that the new project was so dissimilar from the original project that the warnings issued in regard to the feasibility of the original project had little or no bearing on the feasibility of the new project. We disagree. A reasonable jury could find from the evidence that the prices of units were raised to an unmarketable level, that the addition of a limited refund did not render the units marketable at a higher price, that the defendants knew that the new project could not be successfully marketed in the Birmingham area at the higher price, and that the deal was restructured with the intent to make the bonds appear to be genuine while they *1001were in fact doomed to end in an early default. Since these failures followed this exact pattern, it is a jury question as to whether this default was as predictable as “the night following the day” or whether there were other causes for these failures.
Plaintiffs offered evidence that Rice obtained purchase agreements without deposits from friends and relatives in order to give investors the false impression that 50% of the Mount Royal units were presold. Defendant Laventhol & Horwath’s feasibility opinion was premised on the fact that these presale agreements were not sham agreements. Plaintiffs offered additional evidence that without some form of commitment as to 50% of the units, the bonds would not have been marketable. The district court determined that the developer’s failure to obtain valid presale agreements did not jeopardize the marketability of the bonds, because the disclosure statement “fairly disclosed” that many of the presale agreements were not secured by deposits. As we have noted, the marketability of bonds normally will be a fact question for the trier of fact, and this case is no exception. A reasonable jury could find that as to this type of project, the bonds were not marketable unless the developer obtained (without using sham sales) some commitment from potential purchasers as to 50% of the units.
Finally, plaintiffs have offered evidence that it was fraudulent to market these bonds as tax exempt bonds. There is some dispute as to whether Mount Royal was formed for charitable purposes and whether its activities are “essentially public in nature.” See Rev.Rul. 63-20, 1963-1 C.B. 24. Although the taxability of these bonds is not directly relevant to this case,15 defendants’ alleged knowledge that the bonds were not properly tax-exempt is relevant to the existence of a fraudulent scheme to issue unmarketable securities and to the issues of scienter and unmarketability.
In response to plaintiffs’ evidence, defendants offered evidence which they contend negates any inference of pervasive fraud raised by plaintiffs’ evidence. First, defendants rely heavily on the fact that each bondholder has received his investment plus some positive return. Indeed, the bondholders paid $29,950,000 in to the project and received total payments of approximately $34,000,000 by 1985. Defendants urge that because plaintiffs obtained a positive return on their investment after default, they cannot contend that but for the fraud the bonds would not have been marketed. The defendants argue that because the bonds resulted in some positive return, they had some value and could have been marketed even without the alleged fraud. Alternatively, defendants argue that any project which produces enough revenue after default to pay off the face value of the bonds could not have been the subject of a pervasive scheme to defraud the market. We reject these attempts to establish “bright-line” rules of marketability. Certainly, the jury may consider the fact that the bond proceeds of $29,500,000 were used to construct an $18,000,000 ■project and that the investors have recouped some of their investment (although without significant interest). These facts should not, however, bind the jury to a verdict for defendants. It is conceivable that a bond issue to build valuable property can be the result of a scheme to defraud investors. Plaintiffs allege that defendants knew or recklessly disregarded the fact that this bond issue would end in default. A jury could reasonably find that such a bond issue is unmarketable even though the investors would receive their initial investment plus some minimal interest after the inevitable default. The inevitability of the default, and not the consequences of the default, could lead the jury to conclude that the Mount Royal bond issue would have been unmarketable but for the defendants fraud.
Whether or not the bonds would have been “marketable” absent the *1002fraud is a complex factual inquiry. It involves an analysis of purchasers and potential sellers of bonds in the relevant market at the time these bonds were issued. Defendants do not prevail merely because they can show that the public would have purchased these bonds at some price absent the fraud. This focus solely on bond purchasers is not helpful. The lower price or higher rate which buyers would accept is only relevant if the bonds would be offered at that price or rate by nonfraudulent sellers of securities. The issue is whether bond sellers (i.e., underwriters, developers, etc.) would have been able to price these bonds without fraud at a price and rate which would attract informed buyers. Under Shores, plaintiff has the burden of proving that the bonds would have been unmarketable absent the fraud. Although this is a difficult fact to prove, plaintiffs’ case is not fatally injured by the fact that they have received some return on their investment. They should be permitted to prove (perhaps through expert testimony) that the Mount Royal bond issue could not have been marketed in September, 1981 without the benefit of a fraudulent scheme to deceive investors.
Second, defendants contend that this case does not involve a pervasive scheme to defraud because the offering circular fully disclosed the significant risk that the bond issue would end in default. Indeed, the offering circular is replete with warnings that the bonds were not backed by the assets of Vestavia Hills, and that the bonds would be paid (if at all) from the sale of the units after the closing of the bond issue. There is, however, a genuine disputed issue of fact as to whether the offering circular disclosed the full extent of the risks involved in the Mount Royal project. The circular does not explain the negative impact which the construction delay allegedly had upon purchaser interest in the units. It does not, according to plaintiffs, fully explain the failure to bring the original $20,000,000 bond issue to market and does not disclose the fact that the project was resurrected only after doubling the price of apartments. Furthermore, the circular stated that “the pre-marketing program has convinced the Company that its unique fee structure can be successfully marketed.” Although the circular gives no assurance of successful marketing, plaintiffs claim it is misleading because defendants knew or recklessly disregarded the fact that the pre-marketing program was a failure and that the sales figures were artificially inflated by Rice’s sham sales to friends and relatives. In short, plaintiffs claim the offering circular explained the existence of extensive risks but did not (and could not) disclose the full extent of the risk of default because such disclosure would have rendered the bonds unmarketable.
Disclosure of some risks is not a bar to a claim of fraud for failing to disclose the full extent of the risk. “To warn that the untoward may occur when the event is contingent is prudent; to caution that it is only possible for the unfavorable events to happen when they have already occurred is deceit.” Huddleston, 640 F.2d at 543-44. Plaintiffs have adduced enough evidence from which a reasonable jury could find that the Mount Royal project was destined to default, that some of the defendants knew this, and that these defendants fraudulently placed these bonds on the market. If the jury finds these facts, the disclosure of risks in the offering circular is immaterial. Plaintiffs could prove that the risks actually known to defendants but not disclosed in the offering circular would have made the project unmarketable on the bond market. This is a question of fact, and should be decided by the jury and not by the court. Of course, the jury can and should consider the extent of disclosure in determining the existence and extent of the fraud. The jury might find that the disclosure proves the defendants did not intend to defraud the investing public. Alternatively, the jury might find that there was some undisclosed risk of default but that this risk was not so high as to prevent the bonds from being marketed “at a lower price or at a higher rate.” Shores, 647 F.2d at 470. In either event, plaintiffs could not recover. Because a reasonable jury could find that the bonds
*1003B. Scienter
In granting summary judgment, the district court found there was “no reasonable inference of pervasive fraud or conspiracy as to any defendant.” Accordingly, the district court did not address separately the proof of scienter with regard to each defendant. Because we find there was sufficient evidence of a fraudulent scheme to preclude summary judgment, we must address the argument that plaintiff failed to show a genuine issue of fact as to the scienter of each defendant.
1.Bank South
As is noted above, issues of scien-ter are generally poor candidates for summary judgment. In this case, however, we agree with the district court that the evidence adduced against the indenture trustee (Bank South) was insufficient to justify keeping that defendant in the case. Because this complaint does not allege the type of total sham involved in Shores, there is no inference that everybody involved with the issue was involved in the fraud. As to Bank South, there was no allegation or proof of specific misconduct as was alleged in Shores, where the indenture trustee failed in its duty to ensure that the developer maintained sufficient working capital to complete the construction. This bond issue was facially valid, and there is no allegation or proof that Bank South failed in any way in its duty as indenture trustee. Plaintiffs cannot keep the indenture trustee in this case merely by making conclusory allegations of the trustee’s knowledge of the fraud.
2. Issuer Defendants
The district court dismissed the claims against the City and the Authority at the pleading stage. Plaintiffs’ complaint and a supplemental submission, Record, Yol. 4, Tab 254, permitted by the court alleged that the issuer defendants knew that Mount Royal was not the charitable corporation it claimed to be and that they were “on notice” that the project was not feasible. The district court’s acceptance of the plaintiffs’ supplemental submission (which was created using documents obtained during discovery), effectively converted the issuer defendants’ motion to dismiss into a motion for summary judgment.16
Although the supplemental submission alleges that the issuer defendants acted with scienter, the evidence presented (at most) an inference that these defendants were negligent in their reliance on Rice and several of the other defendants. The inference that there was a fraud in this bond issue is not strong enough to support the further inference that peripheral parties (such as the City and the Authority), who were not intimately involved in the planning and structuring of the bond issue, were actively involved in the fraud. Even though discovery was incomplete when these defendants were dismissed, we have seen nothing in the fully developed record to suggest the inference that these defendants acted with the requisite scienter to support a 10b-5 claim. Accordingly, in 86-7352, we affirm the district court’s dismissal of the claims against Bank South and the issuer defendants.
3. Mount Royal Trustees
Plaintiffs’ claims against the trustees 17 are similar to their claims against the issuer defendants. The trustees met in December, 1980 to abandon the Mount Royal project and did not meet again until *1004August, 1981, when Rice called a meeting to resurrect the project. In their submission of evidence in opposition to the motions for summary judgment, plaintiffs made the following claims against the trustees: (1) that, after less than one hour of consideration, the trustees “rubber stamped” the resurrection of the project as a $35,000,000 bond issue even though it had been cancelled seven months earlier because it was not feasible as a $19,000,000 project; and (2) the trustees did not keep themselves informed of subsequent developments, such as the lowering of the amount of the bond issue to $29,950,000. Plaintiff’s Abstract of Evidence at 70, 71.
These charges, even if true, do not support a reasonable inference of scienter. At best, they allege negligence on the part of the trustees, who received no financial compensation for their work as trustees. Because no reasonable juror could find from this evidence that the trustees were part of any fraudulent scheme, we affirm the district court’s grant of summary judgment as to these defendants.
4. Remaining Defendants
The remaining defendants— Rice, the underwriter, bond counsel, feasibility consultant, and the parties to the joint venture — were intimately involved in the development, planning, and issuance of these bonds to the public. If there was a fraud (and we do not imply that there was), a jury could find that each of these defendants acted with knowledge or in reckless disregard of the fraud. We in no way intimate a belief that there was a fraud or that any of these defendants acted fraudulently. The fact that bond counsel and the feasibility consultant may have been retained on a contingent fee basis does not suggest, by itself, that these defendants acted improperly. But plaintiffs have demonstrated a genuine issue of fact regarding the conduct of the agents who represented these reputable firms in this matter. Plaintiffs’ charges of fraudulent conduct cannot be dismissed merely because of the undeniably fine reputation of the defendants in their respective fields. Cases involving knowledge and intent turn on the fact finder’s assessment of the testimony and other evidence. While it is unfortunate that mere allegations can tarnish even the finest reputations and that it is prohibitively expensive to defend against such allegations, we are unable after reviewing the evidence in this case to conclude that there is not a genuine issue as to material facts with respect to the knowledge and state of mind of the remaining defendants as to the marketability of the bonds. Consequently, we are unable to affirm a judgment which pretermits the jury’s consideration of evidence. Accordingly, we hold that the district court erred in granting summary judgment to these defendants.18
IV. CONCLUSION
Based on the foregoing, the district court order certifying the plaintiffs’ class (86-7350) is affirmed. Defendants’ assertion of an arguable nonreliance defense was not sufficiently compelling to render the district court’s decision an abuse of discretion. The district court order (86-7352) dismissing the issuer defendants and the indenture trustee is affirmed because plaintiffs have failed to make an adequate showing of scienter as to these defendants. Finally, however, the district court order (86-7790) dismissing the remaining defendants is affirmed in part and reversed in part because the evidence offered by plaintiffs raises genuine issues of material fact as to the culpability of several of the defendants which should be resolved by a jury.
AFFIRMED in part, REVERSED in part and REMANDED.
. In Bonner v. City of Prichard, 661 F.2d 1206 (11th Cir.1981) (en banc), this court adopted as binding precedent all of the decisions of the former Fifth Circuit handed down prior to the close of business on September 30, 1981. Id. at 1209.
. Plaintiffs also raise pendent state law claims and a civil RICO claim alleging that the fraud involved in this case constituted a pattern of racketeering activity.
. Interest on bonds issued for quasi-public purposes can be tax-exempt if the project satisfies the requirements set forth by the Internal Revenue Service in Rev.Rul. 63-20, 1963-1 C.B. 24. Under that ruling, the obligations of a nonprofit corporation organized pursuant to the general nonprofit corporation law of a state will be considered issued "on behalf’ the state or a political subdivision thereof (and therefor exempt from tax on interest payments) provided each of the following requirements is met:
(1) the corporation must engage in activities which are essentially public in nature; (2) the corporation must be one which is not organized for profit (except to the extent of retiring indebtedness); (3) the corporate income must not inure to any private person; (4) the state or political subdivision thereof must have a beneficial interest in the corporation while the indebtedness remains outstanding and it must obtain full legal title to the property of the corporation with respect to which the indebtedness was incurred upon the retirement of such indebtedness; and (5) the corporation must have been approved by the state or a political subdivision thereof, either of which must also have approved the specific obligations issued by the corporation.
Id. at 25.
. The bond counsel defendant in this case is Alston & Bird, a successor firm of Jones, Bird & Howell.
. Defendants contended that the bondholders had been paid the face amount of their bonds and some interest. The district court correctly concluded "that the mere fact that the plaintiffs recovered more than their principal outlays is indeterminative, as a matter of law, of whether they suffered any damages. The loss of use of money is an appropriate concern if, in fact, the ‘real’ value of the bonds did not equal their purchase price at the time of issue.” Dist.Ct.Op. at 7.
. The district court lamented that it "must traipse through bog below and fog above to attempt to arrive at the present status of Eleventh Circuit law.” Record, Vol. 4, Tab 273 at 11. Indeed, the court noted that "attempts to divine the status of Eleventh Circuit law on this issue may be analogous to determining the nature of Christ after considering all the arguments made during the early years of Christianity.” Id. at 16. We will attempt to clarify the circuit’s law regarding reliance under Shores. The district court's divine inquiry must be addressed to a higher court.
. Later at note 7 and accompanying text, we further discuss Shores and the reliance and causation factors. Obviously, no investor has a guarantee that a bond, stock or other property investment is free from loss and default. An investor is entitled to make an investment that is not foredoomed to failure as a consequence of the knowing marketing of a security that cannot earn a return that would prevent a default and bankruptcy.
. Rule 10b-5 provides:
Employment of manipulative and deceptive devices.
It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange,
(a) To employ any device, scheme, or artifice to defraud,
(b) To make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading, or
(c)To engage in any act, practice, or course of business which operates or would operate as a fraud or deceit upon any person, in connection with the purchase or sale of any security.
17 C.F.R. § 240.10b-5.
. The court in Shores made it clear that the fraud must go to the very essence of the bonds’ existence. The fraud must be "so pervasive that without it, the issuer would not have issued, the dealer could not have dealt in, and the buyer could not have bought these bonds, because they would not have been offered on the market at any price.” 647 F.2d at 464 n. 2. The court further stressed this point by noting that if plaintiff “proves no more than that the bonds would have been offered at a lower price or a higher rate, rather than that they would never have been issued or marketed, he cannot recover.” Id. at 470.
. The opinion also states: "Under Bishop’s broader theory, it would have availed him nothing to have read the Offering Circular. This theory is not that he bought inferior bonds, but that the Bonds he bought were fraudulently marketed.” 647 F.2d at 470-71. We understand plaintiffs case to rest upon the theory that when the bonds involved here were marketed, the defendants had inflated the selling price by manipulating the projected receipts from the proposed sales of the apartment units upward to a figure that they knew would produce few or no sales and that an early default was inevitable.
. We do not, of course, intend to preclude a district court from exercising its discretion in a proper case to deny certification, in a Shores action. We merely reject the defendants’ suggestion that a Shores action never is appropriate for class certification.
. Taken to its extreme, this view would preclude recovery by any plaintiff who is interested in the risks associated with his or her particular investments. As so construed, Shores would offer protection only to speculators (who are willing to assume great risk) and fools (who are *997unaware of risks and depend solely on the availability of bonds in the market to choose their investments). Defendants' position in this case is not quite as extreme. They contend thát only those plaintiffs who can demonstrate their willingness to accept the risks disclosed in the offering statement may recover under Shores. Even as so modified, however, such a limitation on Shores is inconsistent with the policies advocated by that case.
. Of course, a plaintiff such as the one involved in Shores who is willing to accept any marketable risk provides the most compelling case for the fraud on the market theory. This perhaps accounts for the stress placed on the particular plaintiffs characteristics in the Shores opinion. This does not, however, mean that only such a plaintiff may recover under the Shores type of case.
. The broker in Shores merely told the plaintiff in that case that the bonds were a good investment and that others in the community had purchased similar bonds. 647 F.2d at 467.
. Plaintiffs’ loss resulted from the inability of Mount Royal to repay the bonds, not from any ruling of the IRS revoking the bonds’ tax-exempt status. In fact, there has been no such ruling and plaintiffs have not paid any tax on the interest from the bonds.
. Fed.RXiv.P. 12(b) provides that:
If, on a motion asserting the defense numbered (6) to dismiss for failure of the pleading to state a claim upon which relief can be granted, matters outside the pleading are presented to and not excluded by the court, the motion shall be treated as one for summary judgment and disposed of as provided in Rule 56, and all parties shall be given reasonable opportunity to present all material made pertinent to such a motion by Rule 56.
. These are defendants Weldon, Rogers, Ogle, Blanton, Davidson, Lee, Moffett and Schuler.
. In its final summary judgment order, the district court also dismissed plaintiffs’ RICO and state law claims. Because our opinion significantly affects the district court’s underlying assumptions on the dismissed claims, we will remand for further appropriate action on these remaining claims as well. We express no opinion on these issues.