The traditional norm of our legal system is the adversary trial by an individual plaintiff claiming redress for a particular wrong. The question before us is how far the courts should go in allowing class action, mass tort cases to deviate from that tradition. More specifically, this appeal asks us to interpret and apply the recent Supreme Court class action case of Ortiz v. Fibreboard Corp., 119 S. Ct. 2295 (1999). It holds that a "mandatory" class (a class that generally does not give individual notice to members or allow them to opt out) may not be certified, or a settlement approved, under Federal Rule of Civil Procedure 23(b)(1)(B)1 based simply on an unconventional"limited fund" created by the defendants through a settlement of their liability. (The traditional "limited fund" is a pool of money coming from an outside source, the amount of which is not subject to manipulation by the parties.) We must apply Ortiz to this class action claiming that defective pacemakers were implanted in the hearts of approximately 40,000 individuals. The appeal is from the district court's order certifying, on a "limited fund" rationale, a non-opt-out class and approving a mandatory class-action settlement of $57 million. Members of the class object to the settlement on grounds that it unfairly releases from liability the parent corporations of the manufacturers of the defective pacemakers which hold substantial assets, was not the result of arms-length negotiations among the interested parties and overcompensates the plaintiffs' lawyers as an incentive for them to settle the cases of absent class members. The district court below had relied on the Fifth Circuit decision that the Supreme Court later reversed in Ortiz. Flanagan v. Ahearn (In re Asbestos Litig.), 134 F.3d 668 (5th Cir. 1998), rev'd, Ortiz v. Fibreboard Corp., 119 S. Ct. 2295 (1999). We conclude that the Supreme Court's opinion in Ortiz, reversing the Fifth Circuit, requires that the mandatory Rule 23(b)(1)(B) class certified by the district court here must be decertified and that the settlement approved by the district court must be disapproved.

In Ortiz the Supreme Court was again faced with a large class of asbestos claimants suing a manufacturer, which had in turn sued its two insurance carriers for funds to pay the claimants. Negotiations between the lawyers for the class and the manufacturer and the two insurance companies produced a settlement fund of $1.525 billion, contingent on certification under Rule 23(b)(1)(B) as a mandatory class, and approval of the settlement on a limited fund theory. The lower courts certified the claimants as a Rule 23(b)(1)(B) mandatory, non-opt-out class and approved the settlement because they believed that on balance it was in the best interests of the claimants who otherwise stood to lose the fund should the insurance companies win their pending no-coverage cases.

The Supreme Court reversed the Fifth Circuit in Ortiz by a 7 to 2 vote. The Court concluded that applicants for certification on a limited fund theory under Rule 23(b)(1)(B) "must show that the fund is limited by more than the agreement of the parties." Ortiz, 119 S. Ct. at 2302. The Court reached this conclusion because such a mandatory class-action settlement runs head long into long-established principles of due process, the Seventh Amendment right of trial by jury and the "principle of general application in Anglo-American jurisprudence that one is not bound by a judgment in personam in a litigation in which he is not a designated party or to which he has not been made a party by service of process." 119 S. Ct. at 2314 (quoting Hansberry v. Lee, 311 U.S. 32, 40 (1940)).

One of the problems with compromising the rights of absent class members under Rule 23(b)(1)(B) through global mass tort settlements distributed on a mandatory basis arises from the perverse set of incentivesit may provide defendants and class action lawyers -- "the potential for gigantic fees." 119 S. Ct. at 2317. The defendants may be able to settle cases by providing, relatively speaking, a small amount of money for seriously injured class members while providing large attorney fees for lawyers for the class as an inducement to settlement. If the courts deviate very far from the traditional or strict limited fund theory by allowing a limited fund to be created purely by settlement, the legal system runs the risk of eliminating adversary trials conducted to redress wrongs individually by actual plaintiffs through a process by which defendants pay off a small group of plaintiffs' class action lawyers who actually represent other parties. We must apply these principles discussed in Ortiz to the case at hand.

While there are factual differences between the present case and Ortiz, the similarities with the structural problems identified in Ortiz make approval of the settlement in this case inappropriate. Like Ortiz, the settlement herein varies greatly from the traditional model. Many aspects of the settlement undermine the protections for the class that are inherently guaranteed by the traditional model. In addition, this case presents a less appropriate case on the facts than Ortiz. TPLC's corporate parents were released from liability without close scrutiny by the parties as to whether they might be liable. The facts that came out during the personal jurisdiction phase of the litigation below make it doubtful that they would have escaped all liability had they been forced to go to trial. District Judge Spiegel, who has given careful and intelligent consideration to this important case, so found at one point in his rulings on motions. There seems to be no dispute that the parent corporations have sufficient funds to undertake individual litigation and to pay any claims that might result. Their release, therefore, undermines the appropriateness of the settlement even more than the settlement in Ortiz. Like the settlement in Ortiz, the funds available are limited only by agreement of the parties, not because the funds do not exist as a factual matter, and the amount contributed by the parents is small compared to their potential liability.

This products liability class-action litigation was brought on behalf of individuals implanted with the Telectronics Accufix Atrial "J" pacemaker lead. The lead is implanted in the atrium of the heart as part of a pacemaker device used to restore normal heartbeat. It was determined in 1994 that some of the lead wires had a tendency to break, coming through the polyurethane coating and potentially causing injury to the heart and blood vessels. TPLC-manufactured leads of this type were implanted in about 40,000 persons world-wide, including about 25,000 persons in the United States.

Defendant TPLC, a Delaware corporation, manufactured and distributed these leads in the United States between 1988 and 1994. Defendant Telectronic Pacing Systems, also a Delaware corporation, is the sole owner of TPLC. Telectronic Pacing Systems' sole business is to hold certain industrial property rights, real estate and the equity interest in TPLC. Additional named defendants are Pacific Dunlop, Ltd. and Nucleus, Ltd, both Australian companies. Nucleus is a holding company that owns companies that design, manufacture and sell medical products, including pacemakers and pacemaker leads. It was the holding company for TPLC and Telectronic Pacing Systems. Pacific Dunlop purchased Nucleus in 1988 and became the beneficial owner of TPLC and Telectronic Pacing Systems. Pacific Dunlop is a publicly-held Australian company consisting of over 225 separate corporate affiliates and subsidiaries. It has annual world-wide sales of about $5.5 billion. Nucleus and Pacific Dunlop were made defendants on the ground that they are alter egos or agents of their subsidiary, TPLC2. In 1996,TPLC sold all its assets, but not its liabilities, to Pacesetter, Inc. Accufix Research Institute, Inc. became the successor company to TPLC. For clarity, we will continue to use the name TPLC to refer to the company at issue.

Numerous state and federal court actions were filed against defendants. Many of the suits became part of a multi-district litigation proceeding in the Southern District of Ohio. The amended and consolidated master class action complaint asserted claims against TPLC, Nucleus and Pacific Dunlop for negligence, strict liability, failure to warn, breach of implied and express warranty, fraud, medical monitoring, fear of future product failure, intentional and negligent infliction of emotional distress, loss of consortium, misrepresentation and compensatory and punitive damages. A 17-member Plaintiff's Steering Committee was appointed by the district court to coordinate discovery and other pretrial proceedings on behalf of all transferred plaintiffs. The district court addressed the issue of class certification several times before entering the final decision that we review today. We will not review that procedural history here in detail.

The Australian defendants, Pacific Dunlop and Nucleus, moved to dismiss for lack of personal jurisdiction. In February 1997, the district court denied the motion, finding that the companies maintained sufficient contacts with the United States for the court to exercise personal jurisdiction over them. In re Telectronics Pacing Systems, Inc., 953 F. Supp. 909 (S.D. Ohio 1997). On plaintiffs' motion, the district court later certified the class with respect to defendants Pacific Dunlop and Nucleus on plaintiffs' claims of alter ego and agency. In re Telectronics Pacing Systems, Inc., MDL-1057 (S.D. Ohio Dec. 18, 1997) (unpublished).

A one-week, nonbinding, summary jury trial was held in February 1998. The summary jury found TPLC liable under theories of strict liability, negligence and negligence per se and awarded class members between $150,000 and $3 million each, depending on the extent of their injuries. The summary jury did not award monetary damages to plaintiffs whose leads had not broken, but did find that TPLC should pay $265 million for medical monitoring of all implant recipients. The summary jury did not find Nucleus or Pacific Dunlop liable as alter egos of TPLC or under any agency theory.

Defendants and the Plaintiffs' Steering Committee entered into settlement negotiations shortly after the summary trial. The parties filed a joint motion for certification of a mandatory class and approval of the proposed settlement. On July 22, 1998, the district court preliminarily approved the class action settlement proposed by the parties. TPLC's assets, determined to be about $78 million, were divided into four funds: (1) a Patient Benefit Fund of about $47 million out of which class member would be compensated; (2) an Operating Fund of about $20 million that TPLC will use to pay operating expenses; (3) a Litigation Fund of about $7 million that TPLC would use to pay expenses related to non-lead-related litigation and (4) a Reserve Fund of $4 million to be used by defendants to pay expenses in other, unrelated litigation. Pacific Dunlop agreed to contribute $10 million to the Patient Benefit Fund, raising that fund to $57 million and the parties agreed that any unused funds from the other three funds would be added to the Patient Benefit Fund. The defendants and Plaintiffs'Steering Committee also determined categories of class members, based on the extent of injury to date and whether a lead was still implanted in the class member.

The district court approved the settlement and ultimately certified the class as a mandatory, non-opt-out class under Rule 23(b)(1)(B) as requested by the parties. The district court certified the no-opt out class because it found, based on economic information provided by TPLC, that there was a "limited fund" from which injured plaintiffs could be paid. The district court stated that it did not take into account the assets of Nucleus or Pacific Dunlop in determining the total assets available to the settlement fund because (1) it believed that the court was unlikely to obtain jurisdiction over the Australian companies, (2) the time and cost of litigating against a foreign defendants made litigation infeasible and (3) the jury in the summary trial had not found the two Australian companies liable.

As required by Rule 23(e), the district court held a fairness hearing. It reviewed the settlement for fairness, reasonableness and adequacy and found that the settlement as a whole satisfied the standards of Rule 23(e). Fifty-three class members objected to the settlement.

The district court also approved an award of 28%, or about $ 19 million, of the net Patient Benefit Fund as attorney fees. As part of the settlement agreement, defendants did not object to this fee request. The attorney fee amount was objected to by various unnamed class members.

Five different groups of class members have appealed the approval of the settlement and their appeals have been consolidated: (1) Unnamed class member-objector Harold Reed (No. 99-3476); (2) Bruce Hopkins, et al. (No. 99-3477) is comprised of 67 class members who, prior to the settlement, had filed suit against defendants in the Southern District of New York; (3) Class members-objectors Miriam Beasley, et al. (99-3478);3 (4) Kenneth Adam, et al. (No. 99-3479), are objectors, proposed intervenors and putative class members from Louisiana who received the defective leads;4 and (5) Charles Badami, et al. (No. 99-3480), also known as the "Colorado Plaintiffs," is comprised of a group of unnamed class member plaintiffs, objectors and proposed intervenors who had cases pending in District Court for the City and County of Denver that were not made part of the multi-district litigation and their cases were stayed after the class was certified in the multi-district litigation.

Among them, the appellants raise essentially five issues on appeal: (1) whether the district court abused its discretion in certifying the class as a "limited fund" class action under Federal Rule of Civil Procedure 23(b)(1)(B) where solvent and potentially liable companies were released from liability; (2) whether the settlement and certification violates due process because it does not allow plaintiffs with claims for money damages to opt-out of the settlement; (3) whether the class representatives and class counsel adequately represented the interests of all class members; (4) whether the district court erred in awarding class counsel fees of 28% of the total settlement fund; and (5) whether the district court abused its discretion in denying the motions to intervene by various class members/objectors.

Applying the characteristics of appropriate limited fund actions after Ortiz, we are compelled to reject certification of the class under subsection (b)(1)(B) and hold that approval of the settlement wasan abuse of discretion. Ortiz instructed the lower courts to look to the "traditional" or historical nature of certification under Rule 23(b)(1)(B) and stated that courts should not stray too far from these traditional models in determining if certification is suitable under Rule 23(B)(1)(B). Ortiz, 119 S. Ct. at 2311 ("[T]he greater the leniency in departing from the historical limited fund model, the greater the likelihood of abuse . . . ."). As emphasized in Ortiz, the limited fund concept in subsection (b)(1)(B) contemplates a fixed fund in the traditional sense: a fixed resource, such as a mineral deposit, or a fixed amount of money, such as a trust5. The traditional and most common use of subsection (b)(1)(B) class actions is in "limited fund" cases where claims are aggregated against a res or preexisting fund insufficient to satisfy all claims. The Supreme Court noted that classic examples of such actions include actions by shareholders to declare a dividend or otherwise to declare their rights and actions charging a breach of trust by an indenture trustee or other fiduciary that requires an accounting or similar procedure to restore the subject of the trust. Ortiz, 119 S. Ct. at 2308-09 and n.4 (quoting advisory committee notes).

While Rule 23 speaks to "the risk of impairment" of future claims, courts have long recognized that the meaning of subsection (b)(1)(B) is not as broad as the plain language implies. Ortiz, 119 S. Ct. at 2313 ("It is true, of course, that the text of Rule 23(b)(1)(B) is on its face open to a more lenient limited fund concept . . . .") Clearly any potentially large judgment creates the risk of depletion of a defendant's assets and sets up the possibility that, as a practical matter, adjudication may be "dispositive of the interest of other members not parties to the adjudications" or may "substantially impair or impede their ability to protect their interests." Fed. R. Civ. P. 23(b)(1)(B). Ortiz confirmed that a literal reading of the Rule is inappropriate and that mandatory class treatment is to be confined to a narrow category of cases. The Supreme Court directed that when looking to limited fund actions, the "object was to stay close to the historical model." Ortiz, 119 S. Ct. at 2311.

Drawing from the paradigmatic examples identified above, the Ortiz Court articulated three "common characteristics" of limited fund class actions that the drafters of Rule 23(b)(1)(B) "must have assumed would be at least a sufficient set of conditions to justify binding absent members of a class." Ortiz, 119 S. Ct. at 2311. "The first and most distinctive characteristic," the Court explained, "is that the totals of the aggregated liquidated claims and the fund available for satisfying them, set definitely at their maximums, demonstrate the inadequacy of the fund to pay all the claims." Id. The second characteristic of typical limited fund cases is that "the whole of the inadequate fund was to be devoted to the overwhelming claims." Id. "Third, the claimants identified by a common theory of recovery were treated equitably among themselves." Id. The Court reasoned that these characteristics should be treated "as presumptively necessary, and not merely sufficient, to satisfy the limited fund rationale for a mandatory class action." Id. at 2312.

Significantly, the Ortiz Court explicitly refused to decide the ultimate issue of whether, even if these three requirements were met, Rule 23(b)(1)(B) may ever be used to aggregate individual tort claims. Because we find that the settlement in the instant case fails to satisfy all three characteristics of limited fund actions, we need not answer the ultimate question either. We note, however, that the applicability of Rule 23(b)(1)(B) to a fund purporting to liquidate actual and potential tort claims is "subject to question." Id. at 2323. As theSupreme Court explained, the drafters "would have thought such an application of the Rule surprising." Id. at 2314. Moreover, as we shall explain in more detail later, there are "serious constitutional concerns that come with any attempt to aggregate individual tort claims on a limited fund rationale." Id. at 2314.

The primary problem with this settlement is that it fails to meet the first "traditional" characteristic set out by the Court in Ortiz: "the totals of the aggregated liquidated claims and the fund available for satisfying them, set definitely at their maximums, demonstrate the inadequacy of the fund to pay all the claims." 119 S. Ct. at 2311. There are no "liquidated" claims here, so the parties must first estimate the total potential liability to TPLC. TPLC maintains that it has inadequate funds to cover even its legal expenses if the claims are brought individually, let alone satisfy any judgments against it. The parties to this appeal apparently agree that TPLC, standing alone, does not have the necessary assets to cover the expected liability. However, establishing the "fund available" for satisfying the claims is the crux of the problem, because TPLC seeks to exclude the assets of its parent corporations from the calculation.

Specifically, the objectors contend that this is not a true "limited fund" case because the Australian defendants - the parent companies of TPLC - are solvent and potentially liable and their assets should not be excluded when determining the amount available to class members as redress for their injuries.6 Other courts have refused to certify a class under subsection (b)(1)(B) unless all potential sources of recovery are shown to have limited funds. Hum v. Dericks, 162 F.R.D. 628 (D. Hawaii 1995); In re Dennis Greenman Sec. Litig., 829 F.2d 1539 (11th Cir. 1987).

Although we have no factual findings from the district court on the matter, it appears undisputed that the two companies do not have "limited funds" in the traditional sense and would be able to bear the expense of litigation and pay damages if found liable. TPLC informed the district court, however, that it would not settle without the two Australian companies being dismissed. The district court, believing that the settlement was in jeopardy if it did not agree to this part of the settlement, approved it. The district court acknowledged that only TPLC was a limited fund, but held that because TPLC would not settle unless Pacific Dunlop and Nucleus were released, "the loss of settlement can constitute a 'risk' within the meaning of Rule 23(b)(1)(B)" justifying certification under that subsection. Without settlement, the district court worried that some class members might be unable to recover for their injuries because TPLC might run out of funds before all class members could be compensated for their injuries and the parent corporations might not be found liable.

The settlement reached by the Plaintiff's Steering Committee and the defendants released the Australian parent companies from any liability, now or in the future. In justifying its approval of the settlement that does not hold the parent companies accountable, the district court found a "serious question" as to whether the court could exercise personal jurisdiction over Pacific Dunlop and Nucleus. In re Telectronics Pacing Sys., Inc., 186 F.R.D. 459, 475 (S. D. Ohio 1999). This finding appears to contradict the court's finding in its earlier order denying the Australian defendants' motion to dismiss for lack of personal jurisdiction.

In a lengthy and well-reasoned order denying the motion to dismiss of the Australian defendants for lack of personal jurisdiction, the district court stated that "the Australian defendants exercised a great deal of control over [TPLC and Telectronics Pacing Systems]." 953 F. Supp. at 920. The district court noted that Nucleus orPacific Dunlop approved all large capital expenditures of TPLC and Telectronics Pacing Systems and that Nucleus management met monthly to review the monthly reports on the subsidiaries and to provide strategic planning and advice to the subsidiaries. Id. The district court found that a committee of Nucleus and Pacific Dunlop officers oversaw the operations and approved the actions of TPLC and Telectronic Pacing Systems. In addition, the court noted that William Thomas, the CEO of the subsidiaries, was an employee of Pacific Dunlop and reported to the managing director of Pacific Dunlop. The district court concluded that the association between the subsidiaries and the parent companies is "deep and wide-ranging" and evidences "undue control" by the parent companies. Id.

Furthermore, on two occasions, Pacific Dunlop signed indemnification agreements with TPLC suppliers to ensure that TPLC could continue to operate. The district court concluded that this action "demonstrates that [Pacific Dunlop] is willing, when it so desires, to accept responsibility for the obligations of its subsidiaries." Id. at 920-21. Finally, the court found Pacific Dunlop's involvement in the controversy about the leads to be "especially telling" as to the "intimate affiliation" between parent and subsidiary. It was the managing director of Pacific Dunlop that corresponded with officials from the FDA concerning problems with the lead.

The district court found these facts to "create an inference that the absent parent and the subsidiary are in a fact a single legal entity" for the purpose of exercising jurisdiction. Id. While we recognize that the court limited its analysis to examining the due process limits to personal jurisdiction and did not conclude that the corporate veil could be pierced, the facts that came out during this inquiry demonstrate the unfairness of releasing the Australian defendants through a parties-created settlement.

The district court does not explain how it came to change its findings on the court's personal jurisdiction over the Australian companies. The district court apparently based its decision to approve a settlement releasing the parent companies from all liability primarily on the summary jury's finding that the companies would not be liable on an alter ego or agency theory, thereby increasing the risk that the first-arriving plaintiffs would deplete the admittedly limited funds of TPLC, leaving the later-arriving plaintiffs without a remedy. The problem with the district court's approach is that it confuses the ability of plaintiffs to prevail on the merits with the ability to pay a judgment. The issues are separate. If not for the settlement, there would be no limited fund because the class members could pursue their claims against the Australian defendants individually if they chose, as well as against TPLC. We can only conclude that in its desire to approve a settlement and conclude the case by providing some money to class members, the court ignored its earlier findings. We cannot approve a settlement that releases these parent companies from all liability and leaves class members with no recourse against them.

Although not entirely clear, the district court also may have been suggesting that the assets of all three companies - TPLC, Nucleus and Pacific Dunlop - together constituted a limited fund due to the risk that the latter two companies would not be held liable by a jury. In Ortiz, the Supreme Court explained the ways in which the insurance assets at issue in that case could have been limited. First, "[t]he insurance assets would obviously be 'limited' in the traditional sense if the total of the demonstrable claims would render the insurers insolvent, or if the policies provided aggregate limits falling short of that total." Ortiz, 119 S. Ct. at 2317. Because this type of traditional limitation was not present in Ortiz, the Court explained that "any limit of the insurance asset here had to be a product of potentially unlimited policy coverage discounted by the risk that Fibreboard would ultimately lose the coverage dispute litigation." Id. The Court did not decide whether "[t]his sense of limit as avalue discounted by risk," which it explained to be a step removed from the historical model, would suffice for limited fund treatment. Id. Instead, assuming that such a risk analysis would suffice, the Court concluded that it had not been undertaken in that case. As in Ortiz, the district court in the instant case did not undertake an independent risk analysis, but instead accepted the $10 million settlement figure as representing the maximum amount the Australian defendants could be required to pay claimants, which is plainly improper. Moreover, we are doubtful that this "value discounted by risk" theory is sufficient to support a finding that the fund is limited, for there is always risk inherent in litigation.

In addition, the threat of bankruptcy alone, an argument put forth by TPLC as a primary reason to approve the limited fund settlement, cannot be the basis for finding a limited fund. Presumably all companies have limited funds at some point - there is always the possibility that a large mass tort action or other litigation will put a company into bankruptcy. Should that eventuality threaten, we have a comprehensive bankruptcy scheme in this country for just such an occurrence. Simply demonstrating that there is a possibility, even a likelihood, that bankruptcy might at some point occur cannot be the basis for finding that there is a "limited fund" in an ongoing corporate concern. The district court cannot discharge the debt in advance of the occurrence, thereby usurping the bankruptcy scheme through settlement, even it believes such an avenue to be in the best interests of most of the plaintiffs. See In re Asbestos Litigation, 134 F.3d at 672-74 (Smith, J., dissenting) (discussing the impropriety of using a limited fund settlement to avoid bankruptcy).

Moreover, bankruptcy would require vigorous examination of administrative expenses. including attorney fees, and would provide for a creditors' committee that would examine all claims and could object to any claim it found excessive. It is difficult to believe that administrative expenses in bankruptcy would amount to more than 25% of the total estate, or almost $20 million, the approximate amount of attorney fees and costs to be deducted from the amount of funds to be paid out to injured plaintiffs. Compare Barbara J. Houser, Chapter 11 as a Mass Tort Solution, 31 Loy. L.A.L. Rev. 451 (1998) (bankruptcy proceeding efficient way to handle mass tort claims) with Joseph F. Rice & Nancy Worth Davis, The Future of Mass Tort Claims: Comparison of Settlement Class Action to Bankruptcy Treatment of Mass Tort Claims, 50 S.C.L. Rev. 405 (1999) (transaction costs higher in bankruptcy proceeding than settlement class actions). See also Report on Mass Tort Litig., 187 F.R.D. 293, 319-22 (1999) (Report of the Advisory Comm. on Civ. Rules and the Working Group on Mass Torts to the Chief Justice of the United States and to the Judicial Conf. of the United States) (discussing reform proposals for mass tort class actions).

We also cannot approve this settlement because it appears not to be the result of arms-length negotiation among the parties. A significant aspect of this settlement appears to be to limit the liability of the parent companies. The bootstrapping of a Rule 23(b)(3) class into a Rule 23(b)(1)(B) class is impermissible and highlights the problem with defining and certifying class actions by reference to a proposed settlement. Ortiz makes clear that subsection (b)(1)(B) was not intended for the lawyers representing the parties essentially to "create" a limited fund by threatening that there would be no settlement unless the deepest pockets are totally released from liability. We therefore agree with the objectors that the Australian parent companies should not be totally released from liability based solely on agreement of the parties and for that reason alone this settlement cannot be approved.

We do not decide whether the other two traditional characteristics of a limited fundcase are met here because the first requirement is not met.

The Supreme Court in Ortiz also articulated several constitutional considerations compromised in a non-opt-out class action regarding a mass tort. Both Seventh Amendment jury trial rights and the Fifth Amendment due process principle regarding the right to a "day in court" are implicated in aggregating individual claims sounding in tort. 119 S. Ct. at 2314-15. The Supreme Court has repeatedly emphasized the importance of allowing affected persons to opt out of representative suits. See, e.g., Phillips Petro. Co. v. Shutts, 472 U.S. 797, 811-12 (1985); Mulane v. Central Hanover Bank & Trust Co., 339 U.S. 306, 314-15 (1950); Hansberry v. Lee, 311 U.S. 32, 42-45 (1940).

Class certification, whether mandatory or not, necessarily compromises various rights of absent class members. Rule 23(b)(3), with its notice and opt-out provisions, strikes a balance between the value of aggregating similar claims and the right of an individual to have his or her day in court. Certification under subsection (b)(1)(B), which does not include these protections, must be carefully scrutinized and sparingly utilized. The Supreme Court also stressed that in a proper interpretation of Rule 23, principles of sound judicial management and constitutional considerations of due process and the right to jury trial all lead to the conclusion that in an action for money damages class members are entitled to personal notice and an opportunity to opt out. 119 S. Ct. at 2314-15. This entitlement can be overcome only when individual suits would confound the interests of other plaintiffs, such as with a limited fund that must be distributed ratably or an injunction that affects all plaintiffs similarly. In one of the few appellate court opinions issued since Ortiz, Judge Easterbrook stated, "Ortiz disapproved a creative use of Rule 23(b)(1) that employed the 'limited fund' rationale to eliminate notice and opt-out rights." Jefferson v. Ingersoll Int'l Inc., 195 F.3d 894, 897 (7th Cir. 1999) (vacating class certification granted pursuant to Rule 23(b)(2), a no-notice, non-opt-out section pertaining to claims for injunctive relief). He went on to conclude that Ortiz "says in no uncertain terms that class members' rights to notice and an opportunity to opt out should be preserved whenever possible." Id. at 899.

Generally, due process requires that class members bringing particularized tort claims for money damages be provided an opportunity to opt-out of the class. Shutts, 472 U.S. at 811-12 & n.3. From a due process point of view, the opt-out choice is of less concern when there is a definite fund or res from which plaintiffs will receive damages. When there is a true limited fund, the only question is how to divide up the pie. Where defendants have sufficient funds to compensate class members through individual litigation, however, as Pacific Dunlop and Nucleus apparently do, the choice to opt out becomes much more meaningful and due process demands that class members be afforded that right where possible. If certain plaintiffs wish to opt-out and take their chances at suing a foreign corporation, due process would seem to require that they be allowed to do so absent strong considerations to the contrary not present here.

Although not necessary to our holding herein, we wish briefly to address the denial of the motions to intervene by the district court. Various unnamed class members filed motions to intervene in the district court pursuant to Federal Rule of Civil Procedure 247. The district courtdenied their motions to intervene as "untimely." In this Circuit, several factors are considered in determining timeliness: (1) the point to which the suit has progressed; (2) the purpose for which intervention is sought; (3) the length of time preceding the application during which the proposed intervenor knew or reasonably should have known of his interest in the case; (4) the prejudice to the original parties due to the proposed intervenor's failure after he knew or reasonably should have known of his interest in the case to apply promptly for intervention and (5) the existence of unusual or mitigating circumstances militating against or in favor of intervention. Michigan Ass'n for Retarded Citizens v. Smith, 657 F.2d 102, 105 (6th Cir. 1981).

The settlement notice was sent out to class members in September 1998, giving notice of the terms of the settlement and the date of the fairness hearing in November 1998. The motions to intervene were all filed before the fairness hearing. Given that motions to intervene can be filed as late as 10 days after judgment, it does not appear that the motions were untimely. The district court indicated that because the litigation was ongoing for two years at that point, the objectors should have known about their interest in the case before September 1998. However, many of them had cases pending in other districts and did not have reason to intervene until after they received notice of the settlement and found reason to object. Unnamed members of the class will rarely suspect a shortfall in the adequacy of representation before learning the terms of the settlement or problems in defining the class. In addition, class members here had every reason to believe that they would be given an opportunity to opt out right up until the settlement was finalized. We believe it was error for the district court to deny the motions on the ground of untimeliness.

In light of our disposition, all other issues are pretermitted. Our holding here does not end the matter or foreclose injured plaintiffs from recovery. If TPLC is or becomes insolvent, then, as discussed above, bankruptcy is a partial solution. This would leave open for adjudication the liability of the parent companies, which all parties recognize to be the defendants with the deepest pockets. Our decision here, therefore, should not adversely affect the members of the class who have real injuries to be redressed and compensated.

While there are differences between the settlement in Ortiz and the one at issue here, the settlement as approved strays too far from the traditional model and undermines many of the protections built into Rule 23. Moreover, the form of the settlement calls into question its fairness and raises constitutional concerns. While we do not decide if a limited fund class, or any type of class certification, can work in this case, the settlement cannot deprive class members of the protections available under Rule 23 generally and the traditional model of limited fund cases set forth by the Supreme Court in Ortiz. For that reason we reverse and remand to the district court.