WOOD, Circuit Judge.

In Spano v. Boeing Co., 633 F.3d 574 (7th Cir.2011), we confronted for the first time the question whether an action for breach of fiduciary duty under Section 502(a)(2) of the Employee Retirement Income Security Act of 1974 (ERISA), 29 U.S.C. § 1132(a)(2), may be maintained as a class action when a defined-contribution retirement savings plan is at issue. We concluded in Spano that the answer was "maybe." The proposed classes before us in that case, however, were too broad to meet the certification requirements of Federal Rule of Civil Procedure 23. Spano thus left for another day the resolution of many questions concerning the use of the class-action device for a Section 502(a)(2) claim about a defined-contribution plan.

This case requires us to take the next step. It involves a proposed class of plaintiffs who are participants in two defined-contribution plans run by Lockheed Martin. The class is more focused than those we rejected in Spano, and it reflects Spano's guidance about how to define a certifiable Section 502(a)(2) class. Notwithstanding these improvements, the district court thought that it still came up short, and so the court declined to certify the class. We granted Plaintiffs' petition under Federal Rule of Civil Procedure 23(f) to appeal that ruling. We now reverse, and we hope that our explanation for doing so will further refine the discussion we began in Spano.

I

A

Plaintiffs have brought a number of claims against Lockheed Martin Corporation and Lockheed Martin Investment Management Company (collectively, Lockheed) regarding the management of Lockheed's two retirement savings plans, the Salaried Savings Plan and the Hourly Savings Plan. (The two plans are indistinguishable for purposes of this appeal, and we refer to them collectively as the "Plan" from here on unless the distinction is relevant.) In general they allege that Lockheed breached its fiduciary duty to the Plan in a number of ways, in violation of Sections 409 and 502 of ERISA, 29 U.S.C. §§ 1109(a), 1132(a)(2)-(3). The Plan is a defined-contribution plan, often referred to as a 401(k), which allows employees to direct a portion of their earnings to a tax-deferred retirement savings account; the employee's contribution is often augmented by the employer. These plans offer a range of investment options to participants, who are permitted to allocate the funds in their accounts as they choose. Defined-contribution plans are common in this country, and they "play a vital role in the retirement planning of millions of Americans." Spano, 633 F.3d at 576.

Among the investment options Lockheed offered Plan participants was something called the "stable-value fund" (SVF). SVFs are recognized investment vehicles that are available only through employer-sponsored retirement plans and some college-savings plans. See, e.g., Adam Zoll, For Safety-First Savers, Stable-Value Funds Are Tough to Beat, http://news.morningstar.com/articlenet/article.aspx?id=592164 (last visited Aug. 5, 2013). They typically invest in a mix of short- and intermediate-term securities, such as Treasury securities, corporate bonds, and mortgage-backed securities. Because they hold longer-duration instruments, SVFs generally outperform money market funds, which invest exclusively in short-term securities. Id. To provide the stability advertised in the name, SVFs are provided through "wrap" contracts with banks or insurance companies that guarantee the fund's principal and shield it from interest-rate volatility. Id.; see also Paul J. Donahue, Plan Sponsor Fiduciary Duty for the Selection of Options in Participant-Directed Defined Contribution Plans and the Choice Between Stable Value and Money Market, 39 AKRON L.REV. 9, 20-22 (2006).

Plaintiffs allege that the SVF that Lockheed offered through its Plan failed to conform to this general description. Rather than containing a mix of short- and intermediate-term investments, Lockheed's SVF was heavily invested in short-term money market investments. This resulted in a low rate of return, such that in Lockheed's own words, the SVF did "not beat inflation by a sufficient margin to provide a meaningful retirement asset." Plaintiffs contend that structuring the SVF in this manner amounted to imprudent management and violated Lockheed's duty to manage the Plan "with [] care, skill, prudence, and diligence under the circumstances." 29 U.S.C. § 1104(a)(1)(B).

B

Plaintiffs filed this suit in 2006. Lockheed eventually moved for summary judgment, and in March 2009 the district court granted the motion with respect to some claims and denied it for others. The SVF claim is one that survived. Several days later, the district court certified two classes under Federal Rule of Civil Procedure 23(b)(1)(A) and (B), one for the Salaried Savings Plan and one for the Hourly Savings Plan. Each class was certified for all claims. The Salaried Savings Plan class was defined as:

The Hourly Savings Plan class definition was materially identical. Lockheed petitioned for permission to appeal the certification orders under Rule 23(f), which permits the courts of appeals to accept an interlocutory review of the grant or denial of class certification. We held the petition pending our decision in Spano. After Spano was issued, we vacated the district court's certification order and remanded for further proceedings.

On remand, Plaintiffs moved to modify the class definitions and to amend their complaint to add additional named plaintiffs to serve as class representatives. To conform to our statement in Spano that "a class representative in a defined-contribution case would at a minimum need to have invested in the same funds as the class members," id. at 586, Plaintiffs proposed separate classes for each of their remaining claims, with class membership in each one limited to those Plan participants who invested in the relevant funds during the class period. To conform to Spano's warning that the class must not be "defined so broadly that some members will actually be harmed" by the relief sought, id. at 587, Plaintiffs limited their definition of the SVF class to those who suffered damages as a result of Lockheed's purportedly imprudent management of the fund. To achieve this latter result, Plaintiffs proposed to use as a benchmark for class certification purposes the Hueler FirstSource Universe index (Hueler Index). That index tracks the performance of a variety of stable value funds over time — as relevant here, throughout the class period. By providing a reference point for how an average, prudently managed stable value fund would have performed throughout the class period, Plaintiffs reasoned that the Hueler Index offered a reasonable counterfactual estimate of how Lockheed's SVF would have performed if not for Lockheed's imprudence. By limiting the SVF class to only those Plan participants who suffered harm under this measure, Plaintiffs further reasoned that they had avoided including anyone in the class who may have benefited from Lockheed's conduct. The new proposed class was as follows:

The district court was still not satisfied with this narrowed class definition. It acknowledged that the class was "better-defined and more targeted" than both the previous class certified in the case and the classes in Spano, but it found that the SVF claim was "not suitable for class treatment" nevertheless. In the district court's view, including the Hueler Index in the class definition was an improper attempt to "use class certification to `back door' a resolution of this contested issue [i.e., the proper measure of loss] in [Plaintiffs'] favor." The court concluded that Plaintiffs' SVF claims were not "typical" of those of the class, as required by Rule 23(a)(3). The district court also declined to certify the class provisionally under Rule 23(c)(1)(C), which enables the district court to alter or amend any class definition at any point prior to final judgment. It took the position that certifying a class containing a reference to the Hueler Index was not an "inherently tentative" decision amenable to later modification.

Plaintiffs petitioned for permission to appeal under Rule 23(f). We granted permission with respect to the SVF claims. For the reasons discussed below, we now reverse and remand for further proceedings.

II

At the outset, we must address standing. Lockheed insists that the district court lacked subject-matter jurisdiction over the SVF claim because none of the original named plaintiffs had Article III standing to bring the action. Only one of the original named plaintiffs, Lloyd DeMartini, invested in the SVF at any point during the class period, and Lockheed asserts that he cannot show he was injured by his investment. Without injury, there can be no Article III standing, which requires a plaintiff to show an injury-in-fact that is fairly traceable to the defendant's conduct and that could likely be redressed by a favorable court decision. See, e.g., Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992); United States v. 5 S. 351 Tuthill Rd., Naperville, Ill., 233 F.3d 1017, 1022 (7th Cir.2000). Because we reject Lockheed's contention that DeMartini cannot show injury, we conclude that the district court's jurisdiction was proper. (In light of this conclusion, we need not, and do not, address Plaintiffs' argument that the later addition of David Ketterer, another SVF investor who indisputably has standing, as a named plaintiff cures any standing defect that may have existed at the outset of the case, nor do we explore the possibility that Article III standing is satisfied by Section 502(a)(2)'s express authorization of suit by any Plan member on behalf of the Plan.)

Lockheed bases its argument that DeMartini lacks standing on Plaintiffs' use of the Hueler Index to measure damages and define the SVF class. If damages are measured exclusively by the Hueler Index, DeMartini does not appear to have suffered any damages, since he invested in the SVF during a brief and apparently unusual period during which the Hueler Index did not outperform the SVF. Seizing on this, Lockheed concludes that DeMartini must be incapable of showing injury under any measure of damages. But this does not follow. As Plaintiffs emphasize throughout their briefs, the Hueler Index is intended only as a provisional estimate of damages, useful only as a mechanism to ensure that the class meets the requirements of Rule 23; by the time all is said and done, the damages measure will likely become more refined, and it is possible that DeMartini will be entitled to damages under whatever measure is used. This is just one of many instances in which we must resist the urge to make a preliminary question depend on the final resolution of the merits. See Payton v. Cnty. of Kane, 308 F.3d 673, 677 (7th Cir.2002). Injury-in-fact for standing purposes is not the same thing as the ultimate measure of recovery. The fact that a plaintiff may have difficulty proving damages does not mean that he cannot have been harmed. DeMartini's lack of damages as measured by the Hueler Index suggests that he may have a problem proving the degree of his injury, but Lockheed overreads both Article III's injury-in-fact requirement and the facts in this case when it interprets the absence of damages under the Hueler Index as dispositive proof that DeMartini was not injured. (It is possible, for instance, that if the Plan had been managed prudently, it might have outperformed the Hueler Index at all times, and thus DeMartini would have done even better. All of that remains to be shown.)

It is often the case in class litigation that by the time the remedial phase is reached, some of the original plaintiffs will not be entitled to recover, either because they lost on the merits or because they cannot show damages. Sometimes the reason a particular plaintiff cannot recover may be related to one of the three Article III standing requirements: the plaintiff may not have shown that the defendant caused her injury (in which case, we could also say that her injury was not "fairly traceable" to the defendant), or she might have failed to show that she suffered an injury at all. But in such cases, the plaintiff has lost on the merits; we do not reach back in time and enter a judgment dismissing the case for want of an Article III case or controversy. Yet that is effectively what Lockheed is asking us to do here; it wants us to use the hindsight acquired as the claims in this case have evolved to find that there was never jurisdiction over the case to begin with. We have previously rejected this unworkable view of Article III standing, and we do so again here. See, e.g., Kohen v. Pac. Inv. Mgmt. Co., 571 F.3d 672, 677 (7th Cir.2009) ("Jurisdiction established at the pleading stage by a claim of injury that is not successfully challenged at that stage is not lost when at trial the plaintiff fails to substantiate the allegation of injury; instead the suit is dismissed on the merits."); Bruggeman ex rel. Bruggeman v. Blagojevich, 324 F.3d 906, 909 (7th Cir.2003) ("[I]f [a plaintiff's] claim has no merit, then he has not been injured by any wrongful conduct of the defendant; but if the consequence were that he lacked standing, then every decision in favor of a defendant would be a decision that the court lacked jurisdiction, entitling the plaintiff to start over in another court.").

Finally, Lockheed harps on the point that it is Plaintiffs' burden to show standing. That is true but irrelevant: Plaintiffs have satisfied that burden. Their complaint alleged that they were harmed by Lockheed's mismanagement of the SVF. This was sufficient to establish injury-in-fact for pleading purposes. See Lujan, 504 U.S. at 561, 112 S.Ct. 2130 ("general factual allegations of injury resulting from the defendant's conduct may suffice" to establish standing at the pleading stage); Alliant Energy Corp. v. Bie, 277 F.3d 916, 919-20 (7th Cir.2002). Lockheed first challenged subject-matter jurisdiction in relation to the SVF claim in its motion for summary judgment, but it argued only that no plaintiff had shown that he was invested in the SVF at any point during the class period. This was incorrect, as Plaintiffs had already demonstrated through evidence that they attached to their motion for class certification; that evidence showed that DeMartini was invested in the SVF during the relevant period. This was all that was required to refute Lockheed's standing objection. See Lujan, 504 U.S. at 561, 112 S.Ct. 2130 (plaintiff can satisfy burden to show standing at summary judgment by providing "specific facts" that support standing, which are accepted as true for purposes of summary judgment). At every step in the litigation, Plaintiffs have met their burden of demonstrating standing "in the same way as any other matter on which the plaintiff bears the burden of proof ... with the manner and degree of evidence required at the successive stages of the litigation." Id.

III

A

Turning to the heart of the appeal, Plaintiffs ask us to reverse the district court's denial of class certification on the SVF claim. They argue that the proposed class, in accordance with our decisions in Spano, 633 F.3d 574 (7th Cir.2011), and Ross v. RBS Citizens, N.A., 667 F.3d 900 (7th Cir.2012), vacated on other grounds, ___ U.S. ___, 133 S.Ct. 1722, 185 L.Ed.2d 782 (2013), is precisely defined and carefully tailored to ensure that no plaintiff who may actually have benefited from Lockheed's management of the SVF will be swept into a class that seeks relief in which he has no interest (or may actively oppose). The district court did not necessarily disagree with this description. It was concerned instead that the reference in the class definition to the Hueler Index improperly prejudged the merits of the SVF claim. We review a denial of a motion for class certification for an abuse of discretion. Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir. 2012).

In concluding that the reference to the Hueler Index prejudged the merits of the SVF claim, the district court appears to have assumed that accepting the class definition also required him to accept the conclusion that the SVF was mismanaged because it underperformed relative to the Hueler Index. Any such assumption would be mistaken. It misunderstands both the nature of the SVF claim and the relation between the class definition and the merits. Plaintiffs are not arguing that the SVF was imprudently managed in violation of ERISA because it did not match or outperform the Hueler Index; rather, Plaintiffs allege that the SVF was imprudently managed because its mix of investments was not structured to allow the fund to beat inflation and therefore that it could not serve as a prudent retirement investment for Lockheed employees. If Plaintiffs prevail on this theory, they may offer the Hueler Index as one basis for calculating damages. For now, however, the reference to the Hueler Index in the class definition in no way binds the district court to the use of the Hueler Index as the damages measure should Plaintiffs prevail. If the court concludes that a different measure would be better, it is free to use one.

A decision on a class definition should not, in principle, influence the merits of the case. All class definitions allude to the merits, in that they assume either implicitly or explicitly that the defendant's conduct has adversely affected the defined group of people. Compare Ross, 667 F.3d at 903 (approving a class defined as "[a]ll current and former non-exempt employees of [defendant] who have worked at [one of defendant's] retail branch locations in Illinois at any time during the last three years, who were subject to [defendant's] unlawful compensation policies of failing to pay overtime compensation for all hours worked in excess of forty per work week"), and Messner, 669 F.3d at 810 (proposed class of "[a]ll persons or entities ... who purchased or paid for inpatient hospital services or hospital-based outpatient services directly from Northshore ... its wholly-owned hospitals, predecessors, subsidiaries, or affiliates ... from at least as early as January 1, 2000 to the present") (omissions in original). We do not worry that certifying a class in such cases somehow prevents the defendant from proving that it is not liable for unlawful conduct. The class definition is a tool of case management. It settles the question who the adversaries are, and so it enables the defendant to gauge the extent of its exposure to liability and it alerts excluded parties to consider whether they need to undertake separate actions in order to protect their rights. See Payton, 308 F.3d at 678. What it does not tell us is who will win the case. Cf. Messner, 669 F.3d at 823 (whether some class members' claims will fail on the merits is "a fact generally irrelevant to the district court's decision on class certification"); Schleicher v. Wendt, 618 F.3d 679, 687 (7th Cir.2010) ("The chance, even the certainty, that a class will lose on the merits does not prevent its certification.") There is no cause for concern that certifying a particular class will bind the court when it comes time to resolve the case.

B

On the merits, Lockheed argues that the real problem with the proposed class definition is that it attempts to sneak into the case a theory of liability that was rejected at summary judgment. Lockheed contends that Plaintiffs are precluded from raising any claim that the SVF was imprudently managed. As it sees things, the sole theory still in the case rests on misrepresentation through omission: namely, that Lockheed allegedly inadequately disclosed the nature of the SVF to Plan participants. Because many misrepresentation claims are poorly suited to class treatment, accord Spano, 633 F.3d at 589, Lockheed urges us to find that the SVF claim is unsuitable for class treatment no matter how the class is defined. This argument fails on several levels.

First, Lockheed distorts Plaintiffs' SVF claim when it characterizes their theory as one in which the SVF was imprudently managed because it deviated from the mix of investments held by other funds bearing the "stable value" label. Plaintiffs' claim is not so narrow. Plaintiffs allege that the SVF was an imprudent investment, full stop. They aim to show that the SVF was not structured to beat inflation, that it did not conform to its own Plan documents, and that Lockheed failed to alter the SVF's investment portfolio even after members of its own pension committee voiced concerns that the SVF was not structured to provide a suitable retirement asset. The fact that the SVF's investment mix apparently deviated from that of other, similarly named funds may be relevant evidence on which Plaintiffs will rely, but it does not exhaust their theory of imprudence.

From the First Amended Complaint through this appeal, Plaintiffs have made clear that they believe Lockheed's management of the SVF violated ERISA because "it was an imprudent investment for participants." This allegation appears, among other places, in the First Amended Complaint, the original motion for class certification, Plaintiffs' opposition to summary judgment, the Second Amended Complaint, Plaintiffs' amended motion for class certification, and finally Plaintiffs' appellate briefs. They allude rarely, if at all, to misrepresentation.

Most importantly, Lockheed's argument that the district court rejected Plaintiffs' imprudent management claim at summary judgment is belied by the record. The district court's order denying summary judgment on the SVF claim reads in its entirety: "Defendants' motion is DENIED as to their claim that the Stable Value Fund was properly disclosed to Plan participants and was a prudent investment option for them." All this order says is that the imprudent management claim survives. (Lest there be any doubt, the district court referred again to the imprudent management claim in its class certification decision when it stated that among Plaintiffs' surviving claims was the question "whether the Stable Value Fund [] was properly disclosed to Plan participants and was a prudent investment option for them.")

Lockheed ignores this language and instead points to isolated statements from the court's summary judgment memorandum to support its contention that the court implicitly foreclosed the imprudent management claim. It leans heavily on the district court's discussion of DeBruyne v. Equitable Life Assurance Society, 920 F.2d 457 (7th Cir.1990), reasoning that the district court's acknowledgment of DeBruyne's holding can only mean that it rejected a theory of imprudent management that relies on evidence that other stable value funds had a different mix of investments from the SVF. This interpretation stretches both the district court's order and DeBruyne beyond what either can bear.

DeBruyne arose out of the "Black Monday" stock market crash of 1987. Id. at 461. The plaintiffs were investors in an American Bar Association-sponsored retirement fund known as the "Balanced Fund," which purported to offer a balanced mix of low- and high-risk investments. Id. at 460. After losing money in the 1987 crash, the plaintiffs sued, claiming that the Balanced Fund did not contain the mixture of investments advertised in the plan documents and was not prudently managed. Id. at 462. Their sole evidence backing up these assertions was an expert report that included: (a) a comparison of the Balanced Fund's losses with those of other, similarly named funds; (b) a calculation of the Balance Fund's investment risk for several years in the 1980s (though not for 1987, the critical year in the case); and (c) an unsupported claim that the Balanced Fund was not constituted in the way a "typical" balanced fund would have been managed in 1987. Id. at 462-63. Unswayed by this submission, the district court granted summary judgment to the defendants.

This court affirmed. We noted that the plaintiffs could not show that the Balanced Fund was improperly managed based only on an expert's say-so. Id. at 464. We also observed that the defendants did not "on using the term `balanced,' become wed to a pre-established definition that could not be changed by disclosure." Id. The expert's statement about what a "typical" fund manager would have done in 1987, we concluded, "say[s] little about the wisdom of [defendant's] investments, only that [defendants] may not have followed the crowd." Id. at 465.

These are the statements from DeBruyne to which Lockheed clings. Even in isolation they do not carry the day for Lockheed, and other aspects of the case show that its holding is far narrower than Lockheed asserts. The defendants in DeBruyne submitted evidence that their fund's composition was in line with several recognized definitions of the term "balanced" used in the industry, as well as that of many other balanced funds. Id. at 464. The opinion discussed this evidence twice and relied on the fact that the plaintiffs offered nothing to rebut it; their silence indicated that the defendants' evidence was both relevant and probative. Id. at 464-65. In addition, it is not clear that the expert in DeBruyne actually offered any evidence that the Balanced Fund contained an unusual mixture of investments relative to other "balanced" funds; the only concrete comparison the expert offered was of such funds' losses, but this says nothing about the composition of the funds. Id. at 462-63. Indeed, the expert's conclusion that the management of the Balanced Fund was not "typical" does not appear to have been based on any evidence whatsoever. Id. DeBruyne does not support Lockheed's sweeping and counterintuitive proposition that the makeup and performance of similar funds is irrelevant to an imprudent management claim.

In any event, the district court did not hold that DeBruyne precludes Plaintiffs from arguing that Lockheed's SVF was imprudent by relying on evidence of the composition of other stable value funds. It said only that "[a]s in DeBruyne, using the term `stable value' does not `wed' the Fund to a specific mix of investments. That does not mean, however, that the Fund need not be managed with care and prudence." This statement does not bar Plaintiffs from pursuing their claim of imprudent management, nor does it bar them from presenting their case in any particular manner.

C

Because Plaintiffs' proposed class definition was crafted with Spano in mind, we take a moment to explain why our decision to uphold the class definition now before us is consistent with that case. In Spano, the district court had certified classes in two separate cases, Spano v. Boeing Co. (No. 09-3001), and Beesley v. International Paper Co. (No. 09-3018); both cases involved alleged breaches of fiduciary duty in violation of ERISA Sections 409 and 502(a)(2)-(3). 633 F.3d at 576-77. The class definitions in each case were extraordinarily broad and essentially identical to one another. The class in Spano was defined to include:

Id. at 577. On top of these "breathtaking[ly]" broad definitions, id. at 586, the allegations in both complaints were somewhat vague. In Spano, the plaintiffs objected to the inclusion of certain funds in the plan, but it was unclear exactly which ones or why. Id. Meanwhile, in Beesley, the plaintiffs objected to various misrepresentations and allegedly excessive administrative fees, but it was impossible to pin down how many misrepresentations the plaintiffs accused International Paper of making or whether the challenged fees applied to specific investment options or to the plan as a whole. Id. at 589-90.

The combination of exceedingly broad class definitions and murky claims made it difficult to assess the district court's certification orders. Id. at 586. Against that background, we were certain only that the particular classes before us could not stand. While we may have offered some guidance for how to approach class certification in actions under Section 502(a)(2), we emphasized that we were deciding only the cases before us. Id. at 578 ("We are not here to review any or all hypothetical orders that the court might have crafted."); id. at 588 ("Nothing we have said should be understood as ruling out the possibility of class treatment for one or more better-defined and more-targeted classes.").

It is against this backdrop that readers must understand Spano and its warnings that plaintiffs and courts must take care to avoid certifying classes in which a significant portion of the class may have interests adverse to that of the class representative. See, e.g., id. at 587 ("It is not enough to say that the named plaintiffs want relief for the plan as a whole, if the class is defined so broadly that some members will actually be harmed by that relief."); id. at 591 ("[A] fund that turns out to be an imprudent investment over a particular time for one participant may be a fine investment for another participant who invests in the same fund over a slightly different period. If both are included in the same class, a conflict will result and class treatment will become untenable."). Given the breadth of the classes at issue in Spano and the vagueness surrounding plaintiffs' claims, we were concerned that intra-class conflict of the sort that defeats both the typicality and adequacy-of-representation requirements of Rule 23(a) was all but inevitable. In such cases, a district court should not certify a class that fails to address that danger (say, through the use of subclasses or by defining the class more narrowly). But this court has never held, and Spano did not imply, that the mere possibility that a trivial level of intra-class conflict may materialize as the litigation progresses forecloses class certification entirely. See, e.g., Johnson v. Meriter Health Servs. Emp. Ret. Plan, 702 F.3d 364, 372 (7th Cir.2012) ("It is premature to declare the alleged conflicts of interest an insoluble bar to the class action."); Kohen, 571 F.3d at 680 ("At this stage in the litigation, the existence of such conflicts is hypothetical. If and when they become real, the district court can certify subclasses with separate representation of each...."). This is as true in the Section 502(a)(2) context as in any other area.

The appropriateness of class treatment in a Section 502(a)(2) case (as in other class actions) depends on the claims for which certification is sought. Here, the specifics of the SVF claim make it unlikely that the sorts of conflicts that concerned us in Spano will arise. Plaintiffs emphasize that a Section 502(a)(2) action seeks only to make the fiduciary refund to the Plan any losses caused by the breach. 29 U.S.C. § 1109(a) ("Any person who is a fiduciary with respect to a plan who breaches any of the responsibilities, obligations, or duties imposed upon fiduciaries by this subchapter shall be personally liable to make good to such plan any losses to the plan resulting from each such breach...."). There appears to be no risk that any SVF investor who benefited from Lockheed's imprudent management would have her Plan assets reduced as a result of this lawsuit. Moreover, unlike many imprudent management claims — in which the allegation is that fraud or undue risk inflated the value of a fund and then caused it to crash, see, e.g., In re Schering Plough Corp. ERISA Litig., 589 F.3d 585, 592 (3d Cir.2009) — Plaintiffs' allegation is that the SVF was so low-risk that its growth was insufficient for a retirement asset. A very low-risk fund is by nature not subject to the wide swings in value that would enable some investors to reap a windfall from a fund's mismanagement. Finally, the fact that the SVF underperformed relative to the Hueler Index for all but a very brief portion of the class period reinforces the intuition that few, if any, SVF investors profited from Lockheed's conduct. Should any of these statements turn out to be wrong, the district court can make further adjustments to the class definition later.

Finally, we repeat that this class definition is considerably narrower than those at issue in Spano. Plaintiffs have taken care to limit the class to those Plan participants who invested in the SVF during the class period. Their reference to the Hueler Index is one reasonable way to exclude from the class any persons who did not experience injury. These details make all the difference. We conclude both that Spano poses no bar to the proposed SVF class and that the district court's reservations about the class were unfounded. We leave it to the district court to decide in the first instance whether the remaining requirements for class certification have been met.

IV

We note in concluding that, to the extent the district court had concerns that the proposed class definition might not align with the ultimate outcome of the case, it may have misapprehended its authority under Rule 23(c)(1) to alter or amend its class certification order before final judgment. The district court thought itself foreclosed from this option because ruling on the class definition would not be the sort of "inherently tentative" decision amenable to later modification. But there is nothing more permanent about this proposed class definition than any other. As we explained above, adopting Plaintiffs' class definition in no way binds the district court when it comes time to rule on the merits, and we cannot detect any other feature of this class that removes it from eligibility for adaptation.

The order denying class certification for the proposed SVF class is REVERSED and the case is REMANDED for further proceedings.