This appeal, set in the context of an ERISA breach of fiduciary duty action, largely concerns the doctrine of judicial estoppel. The District Court applied the doctrine to bar Plaintiffs Montrose General Hospital, Inc. (Hospital) and Montrose Medical Group Participating Savings Plan (Plan) from asserting that the Plan is covered by ERISA on account of representations they had made in a related prior litigation. Because this suit is based on the premise that ERISA governs the Plan, the District Court's ruling rendered the Hospital and the Plan unable to state a prima facie case. The court therefore entered summary judgment in favor of Defendants Mutual Life Insurance Company of New York (MONY), whose insurance policies funded the Plan, and Richard Bulger, an outside consultant affiliated with MONY who had brought the parties together.

Judicial estoppel may be imposed only if: (1) the party to be estopped is asserting a position that is irreconcilably inconsistent with one he or she asserted in a prior proceeding; (2) the party changed his or her position in bad faith, i.e., in a culpable manner threatening to the court's authority or integrity; and (3) the use of judicial estoppel is tailored to address the affront to the court's authority or integrity. Though we agree that the inconsistency prong is satisfied in this case, the other two are not. Guided by Cleveland v. Policy Management Systems Corp., 526 U.S. 795 (1999), we hold that a party has not displayed bad faith for judicial estoppel purposes if the initial claim was never accepted or adopted by a court or agency. Because the earlier statements in this case were never accepted or adopted, judicial estoppel was inappropriate.

We hold in the alternative that application of judicial estoppel was not tailored to address any harm occasioned by the change of positions. First, the only "harm" identified by the District Court was inflicted upon third parties--fourteen plan participants who had sued the Hospital, the Plan, MONY, and Bulger in the prior litigation. Judicial estoppel's sole valid use, however, is to remedy an affront to the court's integrity. Second, judicial estoppel is an inappropriate sanction here because its effects would be borne not by any wrongdoers, but by innocent third parties.

Having determined that the District Court was wrong to invoke judicial estoppel, we turn to MONY's and Bulger's alternate grounds for affirmance. We ultimately decline to rule on most of them, concluding instead that it would be better to let the District Court pass on them in the first instance. We do, however, reach and reject MONY's and Bulger's assertion that they are entitled to summary judgment on statute of limitations grounds.

In the late 1970s, the Hospital decided to create a retirement plan. It informed its accountant, Defendant Walter Garvey, of its intentions.1 Garvey, in turn, contacted Bulger, an outside consultant who was affiliated with MONY. Bulger proposed a plan, which the Hospital ultimately adopted. The Plan was plagued by financial troubles from the beginning, and, acting on advice from Bulger, the Hospital altered its funding mechanism on several occasions. These efforts were ultimately unsuccessful, and the Hospital ceased paying premiums in connection with the Plan in either late 1991 or early 1992.

Soon thereafter, fourteen of the sixty-seven plan participants sued the Hospital, the Plan, MONY, Bulger, and Garvey. We will refer to this suit as either the "Hickok action" or the "Hickok litigation," after its first named plaintiff, June Hickok. The Hickok plaintiffs alleged that the Plan was governed by ERISA, and charged the defendants with numerous violations of their purported fiduciary duties under that statute. In their Answer, the Hospital and the Plan raised eight defenses, two of which are pertinent here. Paragraph 7 "specifically denied that the plan [was] an employee pension benefit plan within the meaning of section 3 of ERISA," and Paragraph 11 averred that "[t]he claims of the Plaintiffs [were] barred by the statute of limitations." The Hospital and the Plan repeated these claims in their Amended Answer and Pre-Trial Memorandum.

The Hickok action settled for $600,000 in May 1994. MONY and Bulger assumed responsibility for $500,000, and the Hospital and the Plan were required to pay the remaining $100,000. The settlement was distributed among the fourteen plan participants who were plaintiffs in Hickok; nothing was paid to the fifty-three who were not.

Following closely on the heels of the Hickok settlement, the Hospital and the Plan brought this action against MONY, Bulger, and Garvey, seeking to press claims on behalf of the remaining fifty-three plan participants. The claims in this case are essentially the same as those against which the Hospital and the Plan were co-defendants in Hickok.2 The Complaint avers that "[t]he plaintiff Plan is an employee benefit plan within the meaning of S 3(2)(A) of ERISA," and that the Hospital is bringing this suit in its capacity as fiduciary of the Plan. The Hospital and the Plan have not countered the charge that if the claims in Hickok were time-barred, then those in this case are as well.

Discovery ensued and both MONY and Bulger eventually moved for summary judgment. In support of their motions, MONY and Bulger averred that: (1) judicial estoppel should bar the claims against them; (2) the claims were untimely; (3) they were not ERISA fiduciaries; (4) the Hospital and the Plan were not entitled to equitable relief; and (5) the Hospital's and the Plan's "prohibited transaction" claims were without merit. Ruling on the motions, the District Court invoked judicial estoppel to bar the Hospital and the Plan from repudiating their previously expressed position that ERISA did not apply to the Plan. Because the claims pressed in this suit rest on an assertion that ERISA governs the Plan, the District Court's holding rendered the Hospital and the Plan unable to state a prima facie case, and the court entered summary judgment on behalf of MONY and Bulger. With regard to the other proffered bases for summary judgment, the court remarked that "[a]n examination of the record reveals... material issues of fact that would militate against granting summary judgment. In light of the application of judicial estoppel..., these other issues, however, need not be addressed." This appeal followed.

Federal courts possess inherent equitable authority to sanction malfeasance. One such sanction is judicial estoppel. See Klein v. Stahl GMBH & Co. Maschinefabrik, 185 F.3d 98, 109 (3d Cir. 1999). For reasons explained in the margin, judicial estoppel is distinct from both equitable and collateral estoppel.3 When properly invoked, judicial estoppel bars a litigant from asserting a position that is inconsistent with one he or she previously took before a court or agency. Summary judgment is appropriate when operation of judicial estoppel renders a litigant unable to state a prima facie case.

Three requirements must be met before a district court may properly apply judicial estoppel. First, the party to be estopped must have taken two positions that are irreconcilably inconsistent. See Ryan Operations G.P. v. Santiam-Midwest Lumber Co., 81 F.3d 355, 361 (3d Cir. 1996). Second, judicial estoppel is unwarranted unless the party changed his or her position "in bad faith--i.e., with intent to play fast and loose with the court." Id. Finally, a district court may not employ judicial estoppel unless it is "tailored to address the harm identified" and no lesser sanction would adequately remedy the damage done by the litigant's misconduct. Klein, 185 F.3d at 108 (quotation marks and citation omitted).4

Though a district court's ultimate decision to invoke the doctrine is reviewed only for abuse of discretion, see Anjelino v. New York Times Co., 200 F.3d 73, 100 (3d Cir. 2000), a court "abuses its discretion when its ruling is founded on an error of law or a misapplication of law to the facts," In re O'Brien, 186 F.3d 116, 125 (3d Cir. 1999). In this case, we agree with the District Court that the Hospital and the Plan have taken inconsistent positions. We hold, however, that the District Court's finding of bad faith was built upon an error of law, and was therefore unsound. We hold also that the District Court abused its discretion in concluding that judicial estoppel was an appropriate sanction in this case because it was not tailored to address an affront to the court's integrity and because its use would create rather than defeat a miscarriage of justice.5

The Hospital and the Plan have taken inconsistent positions. Three times during the Hickok action they specifically denied that the Plan was covered by ERISA, but this suit is based on the premise that it is. Furthermore, the Hospital and the Plan do not deny that the claims they press in this suit are materially identical to the ones brought in Hickok. The Hospital and the Plan argued that the Hickok claims were time-barred, and the claims in this case were brought after those in Hickok. If the Hickok action was time-barred, then this one is as well. We therefore agree with the District Court that the inconsistency element is satisfied.

Inconsistencies are not sanctionable unless a litigant has taken one or both positions "in bad faith--i.e., with intent to play fast and loose with the court." Ryan Operations G.P. v. Santiam-Midwest Lumber Co., 81 F.3d 355, 361 (3d Cir. 1996). A finding of bad faith "must be based on more than" the existence of an inconsistency, Klein v. Stahl GMBH & Co. Maschinefabrik, 185 F.3d 98, 111 (3d Cir. 1999) (emphasis added); indeed, a litigant has not acted in "bad faith" for judicial estoppel purposes unless two requirements are met. First, he or she must have behaved in a manner that is somehow culpable. See Ryan Operations, 81 F.3d at 362 (stating that judicial estoppel may not be employed unless "`intentional self contradiction is... used as a means of obtaining unfair advantage' " (quoting Scarano v. Central R. Co. of N.J., 203 F.2d 510, 513 (3d Cir. 1953) (emphasis added))); id. ("An inconsistent argument sufficient to invoke judicial estoppel must be attributable to intentional wrongdoing." (emphasis added)); see also In re Chambers Dev. Co. Inc., 148 F.3d 214, 229 (3d Cir. 1998) (quoting this language from Ryan Operations).

Second, a litigant may not be estopped unless he or she has engaged in culpable behavior vis-a-vis the court. As we have stressed time and time again, judicial estoppel is concerned with the relationship between litigants and the legal system, and not with the way that adversaries treat each other. See, e.g., Ryan Operations, 81 F.3d at 360 ("Judicial estoppel `is intended to protect the courts rather than the litigants.' " (quoting Fleck v. KDI Sylvan Pools, Inc., 981 F.2d 107, 121-22 (3d Cir. 1992))); Delgrosso v. Spang & Co., 903 F.2d 234, 241 (3d Cir. 1990) (same). Accordingly, judicial estoppel may not be employed unless a litigant's culpable conduct has assaulted the dignity or authority of the court.

To assess whether the Hospital and the Plan have engaged in wrongful conduct that may fairly be described as a threat to the integrity of the courts, we must review what they did. In the Hickok action, fourteen plan participants charged the Hospital and the Plan with violating ERISA-imposed fiduciary duties. In their Answer, Amended Answer, and Pre-Trial Memorandum, the Hospital and the Plan averred, among other defenses, that the Plan was not subject to ERISA and that the plaintiffs' claims were time-barred. Before the district court ruled on any dispositive motions and before the case went to trial, the parties settled, and the case was dismissed. Shortly thereafter, the Hospital and the Plan brought the present suit on behalf of the fifty-three plan participants who had not been plaintiffs in Hickok. In this litigation, the Hospital and the Plan assert--in direct contravention of their positions in Hickok--that the Plan is covered by ERISA and that the specific claims (which are, in all relevant respects, identical to those they argued were untimely while defending Hickok) are timely.

The important threshold question--the answer to which we find dispositive in this case--is whether a district court may properly find the existence of bad faith if the initial inconsistent statement was never accepted or adopted by a court or agency. MONY and Bulger apparently assume that it may. Guided by the Supreme Court's recent decision in Cleveland v. Policy Management Systems Corp., 526 U.S. 795 (1999), we disagree.

The issue in Cleveland was whether a person who sought Social Security Disability Insurance (SSDI) benefits could later be judicially estopped from claiming protected status under the Americans with Disabilities Act (ADA). In seeking SSDI benefits, the claimant certified that she was "disabled" and "unable to work," but in a later ADA suit she submitted that she could "perform the essential functions" of a job "with... a reasonable accommodation." See id. at 798-99. Observing "that, in context, these two seeming divergent statutory contentions are often consistent with each other," the Court held that "pursuit, and receipt, of SSDI benefits does not automatically estop the recipient from pursuing an ADA claim." Id. at 797.

Though Cleveland's earlier claim had been accepted by the administrative agency, see id. at 802 (stating that she had "both applied for, and received, SSDI benefits"), the Court laid down guidance highly pertinent to this case. Quoting Federal Rule of Civil Procedure 8(e)(2), it noted that "[o]ur ordinary Rules recognize that a person may not be sure in advance upon which legal theory she will succeed, and so permit parties to `set forth two or more statements of a claim or defense alternatively or hypothetically' and to `state as many separate claims and defenses as the party has regardless of consistency.' " Id. at 805. Stressing that "if an individual has merely applied for, but had not been awarded, SSDI benefits, any inconsistency in the theory of the claims is of the sort normally tolerated by our legal system," the Court opined that it did "not see why the law in respect to the assertion of SSDI and ADA claims should differ." Id.

Guided by Cleveland, we hold that it does not constitute bad faith to assert contrary positions in different proceedings when the initial claim was never accepted or adopted by a court or agency. Because the practice is specifically sanctioned by the Federal Rules, asserting inconsistent claims within a single action obviously does not constitute misconduct that threatens the court's integrity. In Cleveland, the Supreme Court drew a direct parallel between pleading inconsistently in a single case and doing so in subsequent ones, so long as the initial claim was never sustained. Moreover, the Court described the latter type of inconsistencies as "the sort normally tolerated by our legal system." Though the Court did not use the magic words--"it is not bad faith to assert inconsistent claims in separate actions so long as the initial position was never accepted by a court or agency"--Cleveland's import is clear.

The rule we adopt is consistent with judicial estoppel's purpose of protecting the integrity of the courts. "Judicial estoppel addresses the incongruity of allowing a party to assert a position in one tribunal and the opposite in another tribunal. If the second tribunal adopted the party's inconsistent position, then at least one court has probably been misled." Edwards v. Aetna Life Ins. Co., 690 F.2d 595, 599 (6th Cir. 1982). But if a party's initial position was never accepted by a court or agency, then it is difficult to see how a later change manifests an "intent to play fast and loose with the court[s]," Ryan Operations G.P. v. Santiam-Midwest Lumber Co., 81 F.3d at 361 (3d Cir. 1996) (emphasis added), any more than pleading inconsistently in a single action does. We think this insight explains why the consensus view among our sister circuits is that judicial estoppel is inappropriate unless the earlier position was accepted by a court or agency.6 This rule also has support in our cases. See Fleck v. KDI Sylvan Pools, Inc., 981 F.2d 107, 121 (3d Cir. 1992) ("[W]here a party assumes a certain position in a legal proceeding, and succeeds in maintaining that position, he may not thereafter, simply because his interests have changed, assume a contrary position...." (quotation marks and citation omitted) (emphasis added)).

We are unpersuaded by MONY's and Bulger's contentions that Cleveland is inapplicable here, or that stare decisis precludes adoption of the rule we announce today. Citing Gruber v. Hubbard Bert Karle Weber, Inc., 159 F.3d 780, 789 (3d Cir. 1998), and Deibler v. United Food & Commercial Workers' Local Union 23, 973 F.2d 206, 209 (3d Cir. 1992), they submit that the question whether a plan is covered by ERISA is one of fact rather than law. And because in Cleveland the Supreme Court expressly declined to disturb the law of judicial estoppel relating to "purely factual matters, such as `The light was red/green,' or `I can/cannot raise my arm above my head,' " 526 U.S. at 802, they suggest that Cleveland has no applicability to the issue now before us. We reject this contention for two reasons. First, it is waived because it was raised for the first time at oral argument. See Warren G. v. Cumberland County Sch. Dist., 190 F.3d 80, 84 (3d Cir. 1999). Second, we conclude that it is simply wrong on the merits. Though the question whether a particular plan is covered by ERISA may not be one of pure law, it is also not a "purely factual matter" in the sense the phrase was used in Cleveland.7

MONY and Bulger also submit that our pre-Cleveland case law precludes us from holding that there can be no bad faith for judicial estoppel purposes if the earlier statement was never accepted by a court or agency. First, to the extent this claim is true, we note simply that we owe greater fidelity to the decisions of the Supreme Court than to our own. Second, we disagree that any of our cases have actually held that judicial estoppel may be imposed in a situation such as this one.

The only case that MONY and Bulger cite in support of their claim that judicial estoppel may lie in situations where the initial claim was never accepted or adopted by a court or agency is Ryan Operations G.P. v. Santiam-Midwest Lumber Co., 81 F.3d 355 (3d Cir. 1996). Their reliance is misplaced. Ryan Operations held that a party seeking estoppel need not have been a party to the earlier proceedings, see id. at 359-60, and that the party facing estoppel need not have necessarily "benefitted" from its switch in position, see id. at 361. But Ryan Operations never stated that judicial estoppel could validly be applied in a case where the initial position was never accepted by a court or agency.8

Though our holding today may appear to be in some tension with our statement in Ryan Operations that there is no "independent requirement" that a party have "benefitted from its earlier position" to be estopped from changing it later, id. at 361, this tension is more apparent than real. First, the Ryan Operations principle remains true today: so long as the initial claim was in some way accepted or adopted, no further showing is necessary that the party "benefitted" in any particular way. See, e.g., Anjelino v. New York Times Co., 200 F.3d 73, 100 (3d Cir. 1999) (upholding a district court's use of judicial estoppel where a litigant sought to withdraw its previous representation to the court that no further discovery was needed). Second, our rule is consistent with Ryan Operations's admonition that "benefit may be relevant insofar as it evidences an intent to play fast and loose with the courts." 81 F.3d at 361. We do not hold that judicial or administrative acceptance is a prerequisite for its own sake, but rather conclude that a change of position simply cannot evidence bad faith vis-a-vis a court unless the initial statement was accepted or adopted.9

During the course of the Hickok action, the Hospital and the Plan averred that ERISA did not apply to the Plan and that the plaintiffs' claims were barred by the statute of limitations. These claims, however, were never accepted or adopted by the district court. Accordingly, their later change in position cannot, as a matter of law, constitute bad faith. We therefore hold that the District Court abused its discretion by invoking judicial estoppel.

We also hold in the alternative that the District Court abused its discretion by concluding that judicial estoppel was tailored to address any harm caused by the inconsistent statements in this case. Judicial estoppel "is an `extraordinary remedy' " that should be employed only " `when a party's inconsistent behavior would otherwise result in a miscarriage of justice.' " Ryan Operations G.P. v. Santiam-Midwest Lumber Co., 81 F.3d 355, 365 (3d Cir. 1996) (quoting Oneida Motor Freight, Inc. v. United Jersey Bank, 848 F.2d 414, 419 (3d Cir. 1988) (Stapleton, J., dissenting)). Observing that judicial estoppel "is often the harshest remedy" that a court can impose for inequitable conduct, we have held that a district court may not invoke the doctrine unless: (1) "no sanction established by the Federal Rules or a pertinent statute is up to the task of remedying the damage done by a litigant's malfeasance;" and (2) "the sanction [of judicial estoppel] is tailored to address the harm identified." Klein v. Stahl GMBH & Co. Maschinefabrik, 185 F.3d 98, 108, 110 (3d Cir. 1999) (internal quotation marks and citations omitted). In this case, the District Court failed to conduct the former inquiry, and we hold that its conclusion that judicial estoppel was tailored to address any harm caused by the inconsistent representations was not an exercise of sound discretion.

The application of judicial estoppel constitutes an exercise of a court's inherent power to sanction misconduct. See id. at 109. "Because of their very potency, inherent powers must be exercised with restraint and discretion." Chambers v. NASCO, Inc., 501 U.S. 32, 44 (1991). In Chambers, the Supreme Court held that where "bad-faith conduct in the course of litigation [can] be adequately sanctioned under" either the Federal Rules or a particular statute, then a "court ordinarily should rely on" the Rules or the statute "rather than the inherent power." Id. But, said the Court, "if in the informed discretion of the court" these other sources of authority are not "up to the task, the court may safely rely on its inherent power." Id. In Klein, we interpreted Chambers to mean "that the Rules are not `up to the task' when they would not provide a district court with the authority to sanction all of the conduct deserving of sanction." 185 F.3d at 109. But we squarely held that before utilizing its inherent powers, a district court should consider whether any Rule- or statute-based sanctions are up to the task. See id. at 110. In this case, the District Court did not consider whether any such sanctions (some of which are set forth in the margin) would have sufficed to deal with any misconduct that occurred in this case.10 That was error.

Moreover, even had the District Court concluded that use of its inherent sanctioning power was necessary, we would still hold that judicial estoppel was inappropriate here. In Klein we held that judicial estoppel, like all exercises of a court's inherent sanctioning power, may not be used unless it is "tailored to address the harm." Id. at 111. And we stated that judicial estoppel is not so tailored unless, "at a minimum," the party to be estopped took inconsistent positions in bad faith--implicitly recognizing that more would sometimes be required. Id. (emphasis added). We noted the same possibility in Ryan Operations. See 81 F.3d at 365 ("As we have already concluded that the district court erred [in employing judicial estoppel], we need not reach Ryan's argument that [its use] under the circumstances of this case would violate principles of equity and justice.... [However, i]n this case, application of judicial estoppel would be unduly harsh and inequitable. While we need not and do not decide whether we would reverse the district court's order on this ground alone, our equitable concerns lend support to our overall conclusion.").

The District Court erred in determining that judicial estoppel would be tailored to address any harm in this case for two reasons. First, judicial estoppel is not an appropriate response to the only type of harm identified by the court. In its explanation of why judicial estoppel was "appropriate relief in this case," the court faulted the Hospital and the Plan for "abandon[ing]" the fourteen plan participants who were plaintiffs in Hickok, but now seeking to assert precisely the same claims on behalf of fifty-three other participants who were not involved in Hickok. The difficulty with the District Court's reasoning is that judicial estoppel may not be used to punish litigants for how they treat other litigants or third parties;11 its only legitimate purpose is to remedy an affront to the court's integrity. See, e.g., Ryan Operations, 81 F.3d at 360 ("Judicial estoppel `is intended to protect the courts rather than the litigants.' " (quoting Fleck v. KDI Sylvan Pools, Inc., 981 F.2d 107, 121-22 (3d Cir. 1992))). Because the court's opinion contains no hint that it invoked judicial estoppel to respond to a threat to its own authority, the sanction was not tailored to address the harm in this case.

Perhaps more fundamentally, judicial estoppel was simply not tailored to address any malfeasance that may have occurred here. The only potential wrongdoers are the Hospital and the Plan, and the District Court's application of judicial estoppel did result in the dismissal of their claims against MONY and Bulger. The problem arises because the Hospital and the Plan do not seek personal gain in this case, but rather bring this action solely in their fiduciary capacities on behalf of fifty-three plan participants. It is those participants, not the Hospital and the Plan, that will be harmed by the District Court's dismissal. Even assuming that the Hospital and the Plan acted wrongly in "abandon[ing]" the Hickok plaintiffs, it is difficult to see how equity would be served by punishing fifty-three other plan participants in return.

In sum, the District Court erred in not considering whether any Rule or statute was "up to the task" before deciding to utilize its inherent sanctioning power, and abused its discretion in concluding that judicial estoppel was tailored to address any harm in this case.

MONY and Bulger advance several alternate grounds for affirming the District Court's judgment. They aver that, as a matter of law: (1) the claims against them are time-barred; (2) they cannot be held liable under ERISA because they were not fiduciaries of the Plan; (3) the Hospital and the Plan are not entitled to "equitable relief "; and (4) the Hospital and the Plan cannot prevail on their "prohibited transactions" claim. MONY and Bulger raised these arguments before the District Court, which declined to reach them in light of its judicial estoppel holding. The court did comment, however, that: "An examination of the record in relation to these other grounds asserted as bases for summary judgment reveals material issues of fact that would militate against granting summary judgment."

Though we certainly could reach and rule on each of the alternate grounds, we conclude--subject to one exception--that interests of sound judicial administration compel that we remand the case without considering them.12 This is a complicated case with a voluminous record. The able district judge plainly pondered these issues, and at one point suggested that there were genuine issues of material fact as to at least some of them. We think it better under these circumstances to let the District Court review in the first instance the arguments that neither Bulger nor MONY were ERISA fiduciaries, that the request for equitable relief should be denied, and that the prohibited transactions claim fails as a matter of law. Because the issue is so straightforward, however, we reach and reject MONY's claim that it is entitled to summary judgment on statute of limitations grounds.

ERISA's statute of limitations for fiduciary violations expires on "the earlier of ": (1)"six years after... the date of the last action which constituted a part of the breach or violation;" or (2) "three years after the earliest date on which the plaintiff had actual knowledge of the breach or violation." The statute also provides, however, that "in the case of fraud or concealment," the period is extended to "six years after the date of discovery of such breach or violation." 29 U.S.C. S 1113. We have described Section 1113 as creating "a general six year statute of limitations, shortened to three years in cases where the plaintiff has actual knowledge, and potentially extended to six years from the date of discovery in cases involving fraud or concealment." Kurz v. Philadelphia Elec. Co., 96 F.3d 1544, 1551 (3d Cir. 1996).

MONY and Bulger first contend that this suit is barred by ERISA's three year limitations period, which does not begin to run until "the plaintiff ha[s] actual knowledge of the breach or violation," 29 U.S.C. S 1113. We have interpreted the actual knowledge requirement "stringent[ly]." Gluck v. Unisys Corp., 960 F.2d 1168, 1176 (3d Cir. 1992); see also id. ("Section 1113 sets a high standard for barring claims against fiduciaries prior to the expiration of the section's six-year limitations period."). Because other sections of ERISA demonstrate that "Congress knew how to require constructive knowledge," we have opined that "[w]e do not think that Congress' failure to" provide such a standard "in section 1113 was accidental." Id. Accordingly, we have held that "actual knowledge... requires that a plaintiff have actual knowledge of all material facts necessary to understand that some claim exists," but we have emphasized "that our holding does not mean that the statute of limitations can never begin to run until a plaintiff first consults with a lawyer." Id. at 1177.

MONY and Bulger recite seven facts that they claim show that the Hospital and the Plan had "actual knowledge of the facts necessary to understand that some claim existed" more than three years prior to filing this suit in December 1994. They stress that:

* "Bulger warned [the Hospital] in writing in 1988 about not paying premiums";

* The Hospital "knew of persistent funding problems for a ten year period";

* The Plan Administrator "knew of the financial problems by, at the latest, the late 1980s ";

* The Plan Administrator "knew [the Hospital] could not make the payments by 1987";

* The Hospital "stopped paying benefits in the summer of 1991 and disclosed the problems to the participants";

* The Hospital's Administrator "reported to the [Hospital's] Board before 1991 his conclusion that the Plan could not continue"; and

* The Hospital received a letter from Plaintiff's counsel in the Hickok action "in November 1991 outlining potential ERISA violations and claims."

These facts, MONY and Bulger contend, demonstrate that "by November 1991 (at the latest) [the Hospital and the Plan] had actual knowledge sufficient to understand that (as they allege) a fiduciary duty had been breached or ERISA provision violated."

We are unpersuaded. "Gluck... requires a showing that plaintiffs actually knew not only of the events that occurred which constitute the breach or violation but also that those events supported a claim of breach of fiduciary duty or violation under ERISA." International Union of Elec., Elec., Salaried, Mach. & Furniture Workers v. Murata Erie N. Am., 980 F.2d 889, 900 (3d Cir. 1992) (emphasis added). Until the Hospital and the Plan had actual knowledge that the Plan might be covered by ERISA, they obviously had no reason to suspect that any actions by MONY or Bulger could support a claim for breach of fiduciary duty under that statute.

The only piece of evidence to which MONY and Bulger point that could have put the Hospital and the Plan on notice that the Plan was covered by ERISA was the letter the Hospital received in 1991 from the lawyer for the Hickok plaintiffs. Though the letter suggested that the Plan was subject to ERISA, two reasons counsel against reading this letter as establishing--as a matter of law--that the Hospital and the Plan thereafter possessed actual knowledge that they had ERISA claims against MONY and Bulger. First, the letter came from an attorney who was threatening to sue the Hospital and the Plan for ERISA violations. Parties are not required to believe every claim hurled by their adversaries, nor are they likely to do so. Second, the letter in no way suggested that the Hospital and the Plan might have an ERISA action against MONY and Bulger. Though MONY and Bulger argue that this information was supplied by the other pieces of evidence to which they point to establish actual knowledge, we do not believe that the evidence must, as a matter of law, be read that way. We therefore decline to affirm the District Court's judgment on this alternate ground.

Nor is this suit barred as a matter of law under the six year statute of limitations. ERISA's default limitations period expires "six years after... the date of the last action which constituted a part of the breach or violation." 29 U.S.C. S 1113. "[I]n the case of fraud or concealment," however, this period is extended to "six years after the date of discovery of such breach or violation." Id. Even assuming that this suit was not brought within the general six year limitations period, we conclude that there is at least a genuine dispute of material fact as to whether the fraud or concealment exception is applicable.

We have interpreted S 1113 "as incorporating the federal doctrine of fraudulent concealment: The statute of limitations is tolled until the plaintiff in the exercise of reasonable diligence discovered or should have discovered the alleged fraud or concealment." Kurz v. Philadelphia Elec. Co., 96 F.3d 1544 (3d Cir. 1996). Section 1113 applies "when a lawsuit has been delayed because the defendant itself has taken steps to hide its breach of fiduciary duty," and "[t]he relevant question is... not whether the complaint `sounds in concealment,' but rather whether there is evidence that the defendant took affirmative steps to hide its breach of fiduciary duty." Id. It is generally accepted that "there must be actual concealment,--i.e., some trick or contrivance intended to exclude suspicion and prevent injury." Larson v. Northrop Corp., 21 F.3d 1164, 1173 (D.C. Cir. 1994) (quotation marks and citation omitted).

In arguing against the applicability of this exception, MONY and Bulger assert that neither of them concealed anything. But the Hospital and the Plan assert, with support in the record, that "from the time of the Plan's creation and throughout its 14-year operation, Defendants consistently deceived the Hospital by misrepresenting that the Plan was not even subject to ERISA." They also submit, with record support, that although they were "generally aware that MONY, Bulger, and other MONY representatives were replacing various life insurance policies with new policies of the same or different types[,]... Bulger falsely represented to Hospital representatives that they would reduce costs while substantially increasing benefits." Finally, Eudora Bennett, the Plan Administrator, claimed in an affidavit that Bulger and Garvey thwarted her efforts "to gain access to information about the operations of the Plan."

Assuming that these allegations are true, which we must for summary judgment purposes, we cannot conclude as a matter of law that no fraud or concealment occurred in this case. MONY and Bulger's (alleged) repeated denials that ERISA applied to the Plan could reasonably have hindered the Hospital's and the Plan's ability to realize that any breach of ERISA-imposed fiduciary duties had occurred. Further, it is possible that Bulger's (alleged) misrepresentations as to the reasons for replacing the life insurance policies inhibited their capacity to discover that the Plan had been imprudently designed. Finally, the (alleged) conduct of Bulger and Garvey may have actively impeded Bennett's ability to discover facts that could have led her to conclude that fiduciary violations had taken place.

MONY and Bulger offer two responses. They aver that because " `[t]he problems surfaced soon after the establishment of the Plan,' " "the alleged design defects constituted information readily available to" the Hospital and the Plan. But MONY and Bulger provide no citations to the record, and fail to explain why the mere existence of problems means that the Hospital and the Plan were on notice that ERISA applied to the Plan or that it was designed in violation of ERISA-imposed fiduciary duties. Because conclusory allegations unsupported by explanation or facts in the record do not suffice to meet a movant's burden of persuasion, see 11 James Wm. Moore et al., Moore's Federal Practice S 56.13[1] (3d ed. 2000), we conclude that MONY and Bulger cannot prevail on this point.

Finally, MONY and Bulger submit that there was no "reasonable reliance as is required to trigger the fraud or concealment exception." They contend that the Hospital and the Plan "did not delay this lawsuit because of misrepresentations; instead, they delayed as long as possible to avoid subjecting themselves to liability and filed suit only after Hickok was resolved and they could no longer hope to avoid similar claims." MONY and Bulger point to no undisputed facts that demonstrate why the Hospital and the Plan brought this case when they did, and, accordingly, MONY and Bulger are not entitled to summary judgment on this ground. We therefore hold that MONY and Bulger are not entitled to summary judgment on statute of limitations grounds.

For the foregoing reasons, the judgment of the District Court will be reversed and this case remanded for further proceedings consistent with this opinion.