EASTERBROOK, Circuit Judge.

Cash Equivalent Fund is a money market mutual fund with a sweeps feature. Brokers offer the Fund to their customers as an adjunct to their principal accounts. When an account has a cash balance, a computer "sweeps" the money into the Fund, where it earns interest until the customer reinvests in stocks or other financial instruments; the Fund redeems shares automatically to supply the cash for these transactions. The Fund and its investment adviser, Kemper Financial Services, Inc., say that the extra costs of implementing a sweeps feature and the additional transactions it generates, plus the check-writing and wire transfer features of the Fund, justify a fee exceeding the norm for money market funds. Kemper receives an annual administration fee of 0.38% of the Fund's assets. It also receives an investment management fee starting at 0.22% of the first $500 million of the Fund's assets and dropping in increments to 0.15% of the assets exceeding $3 billion. As a result of these fees, the Fund pays interest at a rate approximately 0.2% per annum lower than money market funds that operate passively, including one that Kemper itself manages, the Kemper Money Market Fund.

Despite the difference in fees and payouts, the Fund has grown steadily and now manages more than $5 billion of assets. One might think this judgment of investors dispositive: offered extra services at lower interest, the Fund's investors chose the extra services; others have sent their money elsewhere to get a higher return. Whether the extra services are "worth" the price is the sort of judgment people make every day when deciding whether to buy a stripped down computer or pay extra for one with bells and whistles; our government does not try to determine whether extra features are worth a higher price.

Things are not so simple when the services are rendered by an investment adviser rather than a manufacturer or retailer of computers. Section 36(b) of the Investment Company Act of 1940, 15 U.S.C. § 80a-35(b), provides that the adviser of a registered investment company "shall be deemed to have a fiduciary duty with respect to the receipt of compensation for services". Fiduciary duties require honest dealing. Managers of all corporations owe fiduciary duties to their firms. These duties have never been thought to justify judicial review of levels of compensation paid, short of extreme cases amounting to waste. Nonetheless, § 36(b) has been understood in light of its legislative history to put investment advisers on leashes shorter than those of corporate managers generally, and to require the federal courts to decide whether the fees charged by investment advisers are "excessive". Daily Income Fund, Inc. v. Fox, 464 U.S. 523, 534-41, 104 S.Ct. 831, 837-41, 78 L.Ed.2d 645 (1984).

Jill S. Kamen, who owns shares of the Fund, filed this suit under § 36(b), contending that Kemper's fees are excessive and should be reduced, with excess fees for prior years returned to the Fund. Kamen added a claim that in soliciting the investors' approval of the fee structure in 1984, the Fund had misleadingly compared the fees it pays to Kemper with the fees the Kemper Money Market Fund pays. Cash Equivalent Fund pays approximately 0.2% of its assets per annum more than the Money Market Fund does; Kamen believes that the proxy statement implied that the Fund's fees are equivalent to or lower than those paid to the Money Market Fund. Section 20(a), 15 U.S.C. § 80a-20(a), forbids using the mails to send a proxy statement that violates rules established by the Securities and Exchange Commission. The SEC has by rule under the Investment Company Act adopted its rules under the Securities Exchange Act of 1934, which forbid materially misleading statements. See 17 C.F.R. §§ 270.20a-1(a), 240.14a-9(a).

Judge Nordberg first held that § 20(a) creates a private right of action, 659 F. Supp. 1153 (N.D.Ill. 1987), and then dismissed the portion of the complaint based on the proxy solicitation in 1984, holding that Kamen needed but had neglected to make a demand on the Fund's board of directors. See Fed.R.Civ.P. 23.1. He also struck Kamen's demand for a jury trial on the demand for restitution of excessive fees. We denied Kamen's petition for mandamus in an unpublished order because any error could be reviewed on appeal from a final decision, and the Supreme Court denied certiorari over Justice White's dissent, which observed that other courts of appeals disagree with First National Bank of Waukesha v. Warren, 796 F.2d 999 (7th Cir. 1986), on which the decision rested. Kamen v. Nordberg, 485 U.S. 939, 108 S.Ct. 1119, 99 L.Ed.2d 279 (1988).

Although Fox holds that an investor need not made a demand on the directors when proceeding under § 36(b), that claim did not last much longer. Judge Nordberg asked a magistrate to analyze Kemper's argument that Kamen is not an adequate representative of the other investors in the Fund. The magistrate recommended that the court grant summary judgment for the defendants on the § 36(b) claim because Kamen is not an adequate representative of the class under Fed.R.Civ.P. 23. Magistrate Balog wrote:

[N]o other shareholder has joined in this suit, instituted a claim, or inquired into plaintiff's action; the other shareholders have approved the fees charged by Kemper; after notice of plaintiff's allegations, the shareholders approved an increase in fees. Based on these facts, it can only be said that plaintiff's interests are antagonistic to those of the other shareholders. In such a case, plaintiff cannot adequately protect those interests. . . . It is apparent from the record as it stands that plaintiff's concerns are not those of a class, but are a private matter. As such, plaintiff cannot maintain this suit as a class action.

Judge Aspen, to whom the case was transferred for decision, adopted the magistrate's report and granted the "motion for summary judgment on the issue of plaintiff's adequacy as a class representative. This cause may proceed, if plaintiff so chooses, as a non-class action." After the parties pointed out that the suit was not filed as a class action, and that adequacy of representation is material (if at all) only under Rule 23.1, which governs derivative actions, the court entered judgment for the defendants, stating that because "plaintiff was adjudicated as a non-adequate representative plaintiff cannot proceed individually".

Kamen's appeal presents three questions: whether the § 20 claim should have been dismissed for failure to make a demand on the directors; whether the § 36(b) claim should have been dismissed because she stands alone among the Fund's shareholders; and whether, if the § 36(b) claim should be reinstated, she would be entitled to a jury trial. Defendants maintain that only questions about the adequacy of representation are properly before us, because only that question was resolved in the final decision, and the notice of appeal identified only that decision as the subject of appeal. Defendants misunderstand the rules governing issues that may be litigated on appeal. An appeal from the final judgment brings up for review all decisions that shaped the contours of that judgment. E.g., Chaka v. Lane, 894 F.2d 923 (7th Cir. 1990); Kaszuk v. Bakery Confectionery Union, 791 F.2d 548 (7th Cir. 1986). It is unnecessary to identify earlier interlocutory orders in the notice of appeal. These orders may not themselves be appealed; they are not "final" and so are outside the scope of 28 U.S.C. § 1291. Kamen's notice of appeal specified the final decision, and we have jurisdiction to review all of the legal questions preserved in the district court and in this court affecting the validity of that decision.

I

Kamen's complaint as finally amended alleges that she did not make a demand on the board of directors because the seven independent directors (of the ten-member board) "receive aggregate remuneration of approximately $300,000 a year for serving as directors of the Fund and all of the other funds in the Kemper group" and therefore "are dependent upon and subservient to" Kemper. It alleges in addition that demand would be futile because the Fund solicited the proxies, so a demand would request that the directors sue themselves, and that because the Fund has asked for the dismissal of the suit on the merits the directors obviously are not interested in pursuing the claims. Judge Nordberg thought these allegations insufficient to excuse a demand under Rule 23.1, as do we.

It is far from clear that Rule 23.1 applies to a suit under § 20 of the Investment Advisers Act. The Rule applies to a "derivative action brought . . . to enforce a right of a corporation . . ., the corporation . . . having failed to enforce a right which may properly be asserted by it". Violations of § 20 do not yield rights "of the corporation" in the customary sense. Kamen does not sue in the right of the Fund; she sues the Fund for injury done her by the Fund. The theory of a suit under the proxy rules is that the corporation violated a right of the investor to truthful information. If the investor recovers against the corporation, it may in turn seek to recover from its directors, but the principal wrong is by the corporation against the investors. Kamen conceded in the district court, and again at oral argument in this court, that Rule 23.1 applies to her claim under § 20. Perhaps she did this because she seeks as relief a payment to the Fund, and not a remedy for the investors personally. Whatever the reason for the concession, the question is not presented and we express no opinion on it. Similarly, we express no view on the question whether § 20 creates a private right of action, and if so what the appropriate remedy may be. The district court held that the statute creates a right of action, 659 F. Supp. at 1156-60; our conclusion concerning the demand requirement makes it unnecessary to decide whether to imply such a right. See Burks v. Lasker, 441 U.S. 471, 475-76 n. 5, 99 S.Ct. 1831, 1835-36 n. 5, 60 L.Ed.2d 404 (1979).

The district court asked whether Kamen had satisfied the demand requirement of Rule 23.1, and the briefs on appeal debate the issue in these terms. Yet as we held in Starrels v. First National Bank of Chicago, 870 F.2d 1168, 1170 (7th Cir. 1989), Rule 23.1 governs pleading but does not create a demand requirement. Rule 23.1 requires the complaint to say with particularity what has been done about demand and why. Plaintiff may comply by saying that she made a demand or that a rule of law excuses demand.

What is the source of the rules requiring or excusing demand? When the claim for relief is based on state law, we held in Starrels, the law of the state in which the defendant is incorporated governs. See also Burks, 441 U.S. at 477-78, 99 S.Ct. at 1836-37. When the claim for relief is based on federal substantive law, then federal law also governs the requirement of demand. See Burks, 441 U.S. at 475-77, 99 S.Ct. at 1835-37. But cf. Fox, 464 U.S. at 542-47, 104 S.Ct. at 841-44 (Stevens, J., concurring) (arguing that in cases under federal law there is no demand requirement unless the statute creates one). Federal common law contains a demand rule. Hawes v. Oakland, 104 U.S. 450, 26 L.Ed. 827 (1881).

Even when federal common law supplies the rule of decision, it may obtain that rule not from first principles but from state law. United States v. Kimbell Foods Inc., 440 U.S. 715, 99 S.Ct. 1448, 59 L.Ed.2d 711 (1979). This is especially appropriate under Rule 23.1, because the demand requirement is an aspect of the division of authority between corporate managers and investors, a division usually governed by state law. Burks holds that federal law with respect to directors' power to dismiss derivative suits should be derived from state law, unless the state law is hostile to federal interests. Perhaps the demand requirement, too, should be absorbed from state law. Few courts have considered this possibility. Cases both before and after Burks hold or assume that the demand requirement is a creature of federal law. See, in addition to Hawes and the cases cited in Starrels, 870 F.2d at 1170 n. 4, our own opinions in Nussbacher v. Continental Illinois National Bank, 518 F.2d 873 (7th Cir. 1975), and Thornton v. Evans, 692 F.2d 1064, 1079-81 (7th Cir. 1982). Not until Kamen filed her reply brief in this court had anyone doubted that federal law defines the demand requirement in this case. Kamen's reply brief suggested that we import the rules from Maryland law. (The Fund is a Maryland corporation.) The suggestion comes too late, Wilson v. O'Leary, 895 F.2d 378, 384 (7th Cir. 1990); we shall follow the tradition of Hawes and use federal common law.

The scope of the demand requirement depends on why demand ever is required. The demand rule could reflect a hope that the dispute will go away without litigation, that the board of directors will "do something" (or persuade the putative plaintiff that suit is pointless). Demand then initiates a form of alternative dispute resolution, much like mediation. Steps to control the volume of litigation are welcome, yet the demand rule creates more litigation than it prevents. It is difficult to identify cases in which the board's response to a demand satisfied the shareholder and thus prevented litigation; even if the board acts the shareholder may believe the board did too little. It is easy to point to hundreds of cases, including this one, in which the demand requirement was itself the centerpiece of the litigation. An approach uncertain in scope and discretionary in operation — that is, any rule except one invariably requiring or excusing demand — promotes litigation. When the stakes are high (as they frequently are in cases of this character), even a small disagreement between the parties about the application of a legal rule makes it difficult to resolve disagreements amicably.

A stronger rationale for the demand requirement is the one Hawes gives — that it allows directors to make a business decision about a business question: whether to invest the time and resources of the corporation in litigation. 104 U.S. at 457, 461-62, 26 L.Ed. 827. See also Fox, 464 U.S. at 532-33, 104 S.Ct. at 836-37; Deborah A. DeMott, Demand in Derivative Actions: Problems of Interpretation and Function, 19 U.C.Davis L.Rev. 461, 484-88 (1986); Daniel R. Fischel, The Demand and Standing Requirements in Stockholder Derivative Actions, 44 U.Chi.L.Rev. 168, 171-72 (1976). Firms must make operational decisions; if these misfire, they must decide what to do next. Each decision must be made with the interests of the corporation at heart. Whether to buy a particular combination of services at a particular price is a business decision. So too the decision to file a lawsuit about the price or pursue a different course, such as renegotiating the contract, changing the level of services, even finding a new adviser. Even doing nothing is justified when the resources of top managers required to act exceed the injury to the firm; when "something must be done", acts short of litigation could have net benefits exceeding those of litigation. If the directors run the show, then they must control litigation (versus other remedies) to the same extent as they make the initial business decision.

Choosing between litigation and some other response may be difficult, depending on information unavailable to courts and a sense of the situation in which business executives are trained. Managers who make such judgment calls poorly ultimately give way to superior executives; no such mechanism "selects out" judges who try to make business decisions. In the long run firms are better off when business decisions are made by business specialists, even granting the inevitable errors. If principles such as the "business judgment rule" preserve room for managers to err in making an operational decision, so too they preserve room to err in deciding what remedies to pursue. United Copper Securities Co. v. Amalgamated Copper Co., 244 U.S. 261, 263-64, 37 S.Ct. 509, 510-11, 61 L.Ed. 1119 (1917).

Consider now why plaintiffs may resist making demand. (a) Delay in starting the litigation while the board ponders may injure the firm, perhaps because the statute of limitations is about to expire, perhaps because a questionable transaction is about to occur and it will be hard to unscramble the eggs if it happens before the court can act. (b) Demand may be futile, in the sense that the members of the board are interested in the transaction and unwilling to sue themselves, or because they are so set against litigation that their minds are closed. (c) Demand may be pointless, in the sense that a substantive rule prevents the corporation from controlling the litigation. Fox held that only an investor or the SEC may initiate litigation under § 36(b), and it followed from the firm's inability to file its own case or prevent the investor from litigating that demand was unnecessary. Other statutes likewise may eliminate the point of making demand. (d) Demand may sometimes be imprudent from the plaintiff's perspective. Counsel who fear that the board will sue may hesitate before making a demand, because if the firm sues counsel will not reap the legal fees of victory. Or counsel may think that the board will pursue a strategy in litigation that his client disapproves, or settle for too little.

We will return to these four. Perhaps the most serious difficulty with demand from the perspective of plaintiffs is the link between the making of demand and the standard courts apply to the directors' decision not to sue. In Delaware, the Mother Court of corporate law, any shareholder who makes a demand is deemed to concede that demand was required. Spiegel v. Buntrock, 571 A.2d 767, 775 (Del. 1990); Stotland v. GAF Corp., 469 A.2d 421 (Del. 1983). If demand is required, then the disinterested members of the board are deemed to possess the ability to refuse to sue or control the litigation, provided their decisions are sufficiently reasoned to come within the capacious bounds of the business judgment doctrine. Zapata Corp. v. Maldonado, 430 A.2d 779 (Del. 1981); see also, e.g., Aronson v. Lewis, 473 A.2d 805, 813 (Del. 1984). Except in extraordinary cases, then, tendering a demand to the board puts the plaintiff out of court under Delaware law. No wonder plaintiffs stoutly resist making demands.

Federal courts have never embraced Delaware's link between the making of a demand and special deference to the board's decision not to sue. See Bach v. National Western Life Insurance Co., 810 F.2d 509, 513 (5th Cir. 1987); Joy v. North, 692 F.2d 880, 888 n. 7 (2d Cir. 1982) (Winter, J.). See also Alford v. Shaw, 320 N.C. 465, 358 S.E.2d 323, 327 (1987) (rejecting a tie between demand and the standard of review). We think it would be unwise to do so. When the standard of review depends on the existence of a demand, plaintiffs have extraordinarily strong reasons not to make a demand, and corporations extraordinarily strong reasons to insist on one. Demand then becomes a threshold issue in every derivative suit, one that must be resolved in advance of discovery and on the basis of a good deal of speculation about what the board might do. As in Spiegel, the plaintiff will assert that the board is unreasonable. Why ask persons with closed minds? The board will proclaim that Solomonic wisdom would be applied if only plaintiff would ask, while simultaneously asserting that the suit has no conceivable merit. It is not a pretty picture, but it is an extended and expensive one, made more so by some peculiarities in the way Delaware phrases its standards. See Starrels, 870 F.2d at 1174-76 (concurring opinion). As the American Law Institute has observed, "the need for demand and the standard of judicial review are logically very distinct." Principles of Corporate Governance: Analysis and Recommendations § 7.03 comment a at 65 (Tent. Draft No. 8, 1988). We conclude that when the demand requirement comes from federal common law, the making of a demand does not affect the standard with which the court will assess the board's decision not to sue.

Four reasons remain why demand may be inappropriate: (a) exigencies of time; (b) futility; (c) irrelevance given a substantive rule; (d) the risk that demand will lead to suit and so deprive counsel of fees that might have been obtained were it necessary to file a derivative suit. We may at once discard (d) as a legal excuse. Cases in category (c) obviously never require demand. Cases in category (a) justify filing the complaint before receiving the board's answer to the demand but do not justify failure to make a demand. When time is tight, the investor should make demand at the same time he files the complaint. See Delaware Hudson Co. v. Albany Susquehanna R.R., 213 U.S. 435, 447, 29 S.Ct. 540, 543, 53 L.Ed. 862 (1909). Category (b), futility, is the usual sticking point. The plaintiff asserts that the board is interested or intransigent; the board asserts that it is reasonable and wise. Courts predictably have great difficulty deciding who is right when, as is usual, it must decide such questions on the pleadings.

At least in principle the rationale of the demand requirement implies a futility exception. If courts would not respect the directors' decision not to file suit, then demand would be an empty formality. When all directors have a financial stake in the transaction, their decision not to sue themselves would carry little weight with a court. Or perhaps all of the directors are so ensnarled in the transaction that even when only the duty of care is at stake, their judgment could not be respected. Again demand seems an empty gesture. Courts dispense with futile gestures.

"In principle" is an important qualifier. In practice the futility exception to the demand rule has produced gobs of litigation. It is this exception that has sapped the potential role of the demand requirement as an alternative dispute resolution mechanism. Hundreds of cases opine on whether demand is or is not futile. Difficulties in sorting cases into demand-required and demand-excused bins are not worth incurring, once we sever the link between demand and the standard of review (as we have done). The American Law Institute recommends that courts abolish the "futility" exception to the demand rule, turning demand into an exhaustion requirement with much the same scope and function as the exhaustion requirement in the law of collateral review of criminal convictions. Principles of Corporate Governance §§ 7.03, 7.08, and commentary at 64-71 (Tent. Draft No. 8, 1988). "The futility exception . . . [is] ambiguous in scope and has proven a prodigious generator of litigation." Id. at 64. The time has come to do away with it. If demand is useful, then let the investor make one; if indeed futile, the board's response will establish that soon enough. In either case, the litigation may proceed free of arguments about whether a demand should have been made in the first place. The virtue of simplification may be seen by considering three of the common battles about the meaning of "futility".

1. The plaintiff may say that some or all of the members of the board approved or are interested in the transaction and that demand is futile because they will not sue themselves or contest their own acts. Although directors are unlikely to sue themselves, they may well take some action to palliate the consequences of poorly conceived acts, including their own. Directors want the venture to succeed, and if shown how they can improve its prospects, are likely to act. One mistake at the time of the initial decision does not imply that the member of the board opposes remedial action. Even when the "action" involves suit against some of their number, this does not disable the board. The ALI properly observes, id. at 70-71, that the board may appoint a minority of disinterested members to evaluate the demand and act for the corporation. In the extreme case in which all members are implicated, the board may expand its size and authorize the new members to act for the firm. Of course it may choose to do none of these things, but if so it will just decline the demand. Making a demand is cheap, especially so when the board is disabled from acting. Why prefer extended, costly litigation to the cheap and quick expedient of a demand?

2. The board may be determined not to sue. Perhaps by the time the judge comes to consider whether plaintiff should have made a demand, the defendant will have moved to dismiss the case on the merits. Any demand in such a case would be doomed to failure, and even at an earlier stage it may be transparent that the directors want nothing to do with litigation. This application of the "futility" exception has both a practical and a conceptual difficulty. The practical one is that it is difficult to tell in advance just what position the firm would take if asked; disputes about the demand requirement usually are resolved before the defendants plead to the merits. It is easy for the plaintiffs to say (and for the defendants to deny) that the board has a closed mind; it is much harder to tell who is right.

The conceptual difficulty is that even an adamant unwillingness to sue may reflect the merits. Boards ought not pursue silly or frivolous claims. So certain knowledge that the board is unwilling to authorize litigation may reflect only confidence that the case is feeble or injurious to the firm and other investors. Why should the plaintiffs be authorized to sue, and without so much as a request to the board, just because the complaint is all heat and no light? Once more the ALI hit the nail on the head when observing that a formulation of the futility rule that inquires whether a demand would prompt the board to correct a wrong "assumes that there is a wrong to be corrected. The director's antagonism to an action may well be justified and flow from a sound judgment that the action is either not meritorious or would otherwise subject the corporation to serious injury." Id. at 69. A decision not to file a weak lawsuit would be protected by the business judgment rule, so it makes perfect sense to ask for the board's perspective.

3. A plaintiff may insist that even the independent directors are toadies, so that their judgment could not be respected. Perhaps they are friends of the putative defendants; perhaps they draw hefty directors' fees and fear loss of their offices if they authorize suit; perhaps they believe that the courts have no business supervising corporate affairs and would not authorize litigation no matter how meritorious (and no matter how little their regard for holding onto their offices). If demand is futile in fact for any of these reasons, then the board will say no with dispatch and the case may proceed. As we have broken the link between the demand and the standard of review, the plaintiff may employ this arsenal of arguments to argue that the decision not to sue ought not be respected; the board will stand on the business judgment rule. The court will resolve the question on the merits rather than trying to treat it as a procedural hurdle. Framing questions about the independence of the directors as exceptions to the demand requirement diverts attention from the real issues.

Kamen's complaint pursues all three of these arguments for futility and has all the weaknesses we have identified. Courts frequently say that considerations of this sort do not demonstrate futility. The district court's able opinion collects many of the holdings, 659 F. Supp. at 1161-63. We are in accord, as Starrels demonstrates. Yet other cases, such as our opinions in Nussbacher and Thornton, accept the board's actual or anticipated unwillingness to sue as futility adequate to excuse demand, greatly complicating the resolution of litigation. The line between "potential futility" and "real futility" is increasingly refined.

Recent cases in several circuits display impatience with the futility exception and have been creative in denying that a demand would be futile even when it is pellucid that the board is not about to authorize a suit. E.g., Lewis v. Graves, 701 F.2d 245 (2d Cir. 1983); Greenspun v. Del E. Webb Corp., 634 F.2d 1204 (9th Cir. 1980); In re Kauffman Mutual Fund Actions, 479 F.2d 257 (1st Cir. 1973). Although none of these opinions dispenses with the exception altogether, the day is at hand — unless decisions of the Supreme Court bar the way.

Two decisions, Doctor v. Harrington, 196 U.S. 579, 25 S.Ct. 355, 49 L.Ed. 606 (1905), and Susquehanna, arguably adopt a futility exception to the demand requirement for purposes of pre- Erie general federal law. Although both of these cases speak favorably of a futility exception to the demand requirement, neither prevents the evolution of the federal common law.

Doctor raised the question whether the corporation should be aligned as a plaintiff or a defendant for purposes of diversity jurisdiction. The firm is a defendant to the extent the investor complains that it failed to bring suit; it is a plaintiff to the extent that the derivative suit may end in an order compelling the wrongdoers to pay money to the corporation. Doctor holds that the firm should be aligned as a defendant when the board is so hostile to the investors that demand would be futile; otherwise it should be aligned as a plaintiff. Many people (including the Supreme Court in Susquehanna, 213 U.S. at 449-50, 29 S.Ct. at 544) understood this as implying a futility exception to the demand requirement. Although this may have been a plausible inference, Doctor is no longer good law. It was overruled in all but name by Smith v. Sperling, 354 U.S. 91, 77 S.Ct. 1112, 1 L.Ed.2d 1205 (1957), which holds that in a derivative suit the corporation always should be aligned as a defendant for purposes of determining complete diversity. After Sperling "futility" is unimportant. When explaining why it would no longer use the board's refusal to sue as the basis of the alignment decision, the Court gave an explanation equally applicable to the futility exception to the demand rule in general: "To stop and try the charge of wrongdoing [in refusing to sue] is to delve into the merits. That does not seem to us the proper course. It is a time-consuming, wasteful exertion of energy on a preliminary issue in the case. The instant case is a good illustration, for it has been over eight years in the courts on the question of jurisdiction." 354 U.S. at 95, 77 S.Ct. at 1115.

Susquehanna interprets old Equity Rule 94, which codified the holding of Hawes. Rule 94 required demand, leaving no (apparent) room for exceptions. Susquehanna holds that the rule was not so inflexible. This holding is of no consequence; Rule 94 is no longer with us, having been amended many times in the course of the transformation to Rule 23.1. See Fox, 464 U.S. at 530-31 n. 5, 104 S.Ct. at 835-36 n. 5. But the discussion of futility, 213 U.S. at 447-52, 29 S.Ct. at 543-45, may be thought to establish a doctrine independent of the defunct equity rule. If this was the Court's meaning, the decision is nonetheless linked to its time. Susquehanna assumed that if the members of the board were implicated in the transaction, or held financial interests in the party to be sued, they would be incompetent to act on a demand. Although that is so as a matter of corporate law, the court did not consider the possibility that a committee of independent directors could act on the demand. Not until the 1970s did courts hold that an independent committee could act on demands — and control litigation in the name of the corporation — even though a majority of the board was interested in the transaction. Development of the independent committee washed away the assumption on which the discussion in Susquehanna depends. The Court has never endorsed a "futility" exception to the demand requirement under current assumptions about the ability of committees to act for boards of directors, and given the statement in Sperling about the wastefulness of inquiries into futility we are confident it would not do so.

We conclude that precedent does not prevent us from holding that claims of futility should be tested by making the demand rather than by arguing about hypotheticals. If the firm declines to sue, the court can decide whether the board's decision is entitled to respect under state corporate law, which applies in light of the holding of Burks. See also Pepper v. Litton, 308 U.S. 295, 306, 60 S.Ct. 238, 245, 84 L.Ed. 281 (1939); Hill v. Wallace, 259 U.S. 44, 61, 42 S.Ct. 453, 455, 66 L.Ed. 822 (1922); United Copper, 244 U.S. at 264, 37 S.Ct. at 510-11. As we have rejected Zapata, the making or failure to make a demand will not alter the business judgment standard that ordinarily applies to corporate decisions. Courts now may focus on the question whether the board's actual decision should be given force, rather than on hypothetical inquiries. "Futility" is the only reason Kamen gives for not making a demand on her claim under § 20(a). As this is an unsatisfactory reason, we agree with the district court's decision that the claim must be dismissed for failure to make a required demand.

Abolition of the futility exception calls into question our holdings in Thornton and Nussbacher. Thornton was a suit under ERISA, and the court extensively discussed the policies behind ERISA before deciding that demand would be futile. It may be that Thornton, like Fox, illustrates our category (c): Demand is not required when under substantive law the board may neither control nor prevent litigation. That question must be left for another day. Nussbacher, on the other hand, was founded squarely on the futility exception. The panel held that demand was excused because it was clear from the defendants' motion to dismiss the suit on the merits that demand would have been futile, 518 F.2d at 878-79. Nussbacher reasoned that because Rule 23.1 allows the plaintiff to plead reasons why demand is excused, it must follow that the futility of demand is an adequate excuse. In other words, Nussbacher treated Rule 23.1 itself as the source of the substantive requirements, a position we repudiated in Starrels, applying the rationale of Burks. Other decisions regularly hold that the board's defense of the suit on the merits does not justify failure to make demand. E.g., Grossman v. Johnson, 674 F.2d 115 (1st Cir. 1982); Cramer v. GTE Corp., 582 F.2d 259, 276 (3d Cir. 1978). Lest Nussbacher be thought to represent an independent assessment of the futility exception with continuing vitality, we now formally overrule that case.

II

After Fox the demand requirement of Rule 23.1 does not apply to a claim under § 36(b). The Court held that § 36(b) creates a right of action that only the investor and the SEC may pursue. Because the mutual fund may not assert a claim against the investment adviser under § 36(b), the Court reasoned, Rule 23.1 — which applies only to suits "brought . . . to enforce a right of the corporation . . ., the corporation . . . having failed to enforce a right which may properly be asserted by it" — does not call for a demand on the directors to file suit. What is the point of dunning the directors, if under the statute they may not sue? Our holding with respect to the need for a demand under § 20(a) therefore does not affect the claim under § 36(b).

A

Judge Aspen dismissed the claim under § 36(b) on the basis of Magistrate Balog's conclusion that Kamen is not an adequate representative of other investors in the Fund. Rule 23.1 requires adequacy: "The derivative action may not be maintained if it appears that the plaintiff does not fairly and adequately represent the interests of the shareholders or members similarly situated in enforcing the right of the corporation or association." Yet the very statement of the adequate-representation requirement repeats the theme that Rule 23.1 is limited to suits in which an investor seeks to enforce a corporate right. Rule 23.1 imposes hurdles, including both the pleading requirement and the adequate-representation requirement, before a court will strip the directors of their entitlement to manage the affairs of the corporation, including their right to control the pursuit or compromise of its legal claims. Fox holds that a claim under § 36(b) is not a claim "of the corporation", and it follows that Rule 23.1 is inapplicable. Demand requirements and adequate-representation requirements go hand in glove. If the claim under § 36(b) is really the investor's personal claim, it is unimportant whether Kamen "adequately" represents other investors. Under the statute, she need represent no one.

Kemper relies on footnote 11 of Fox, 464 U.S. at 535 n. 11, 104 S.Ct. at 838 n. 11, which allows that a suit under § 36(b) is "undeniably `derivative' in the broad sense of that word", because the fiduciary obligation runs to the mutual fund, which will receive any remedy. True enough, but it is of no assistance to defendants. Rule 23.1 does not ask whether a suit is derivative "in the broad sense of that word". It asks whether the suit seeks to enforce a right of the corporation, "the corporation . . . having failed to enforce a right which may properly be asserted by it". Fox holds that the claim under § 36(b) is not one the corporation may assert. It therefore establishes that Rule 23.1 does not apply, period. The opinion could hardly be clearer. At pages 528, 535 n. 11, 104 S.Ct. at 834, 838 n. 11, and again in stating the holding at 542, 104 S.Ct. at 841, the Court quotes this language of Rule 23.1 and observes that § 36(b) litigation does not fall within the domain of the rule. Lest there be any doubt, the last sentence reads:

It follows that Rule 23.1 does not apply to an action brought by a shareholder under § 36(b) of the Investment Company Act and that the plaintiff in such a case need not first make a demand upon the fund's directors before bringing suit.

464 U.S. at 542, 104 S.Ct. at 841. Kamen filed a suit under § 36(b), and Rule 23.1 therefore "does not apply". In Fox that meant no demand; here it means no need for adequate representation. See also Pellegrino v. Nesbit, 203 F.2d 463 (9th Cir. 1953) (similar conclusion with respect to the short-swing-profit-recovery provision of § 16(b) of the Securities Exchange Act of 1934).

If Rule 23.1 does not require adequate representation, defendants maintain, then the due process clause of the fifth amendment must. Due process requires adequate representation, though, only when the plaintiff is representing someone else. A judgment in a class action will bind persons who are not before the court. Before resolving the legal entitlements of missing persons, the court must ensure that they have an effective voice. Hansberry v. Lee, 311 U.S. 32, 41-42, 61 S.Ct. 115, 117-18, 85 L.Ed. 22 (1940). Kamen represents no one other than herself. Fox establishes that the right at stake is personal; she does not need the approval of the court or other investors to pursue it. Doubtless other investors are interested in the result, yet many suits affect the status of others as a practical matter without triggering a constitutional requirement of "adequate representation".

Rules 19 and 24 are designed for such cases. Rule 19 allows and sometimes requires the joinder of persons who will be affected by a judgment; Rule 24 allows intervention. Defendants do not maintain that Rule 19 requires the joinder of other investors. Rule 24 may allow them to intervene on the other side to argue for or against Kemper's management fees. Bethune Plaza, Inc. v. Lumpkin, 863 F.2d 525, 531-34 (7th Cir. 1988). None has done so, undoubtedly because investors who support the current fee structure believe that they already have adequate champions: Kemper and the Fund. What is more, a judgment against the defendants requiring a reduction in the level of fees would not "bind" other investors any more than a judgment in a contract or tort suit "binds" the investors and employees of the firm required to pay damages. If the court should deem the fees excessive, and the Fund then cut back on services, shareholders may take their money elsewhere. That would injure the Fund and Kemper, which as parties represent themselves. It might also reduce the consumers' surplus of the investors (the value they place on the Fund's services, less the price they pay), but such an indirect consequence does not require either actual or vicarious representation in the litigation. See O'Bannon v. Town Court Nursing Center, 447 U.S. 773, 788-89, 100 S.Ct. 2467, 2476-77, 65 L.Ed.2d 506 (1980).

B

At all events, Kamen is no less adequate a representative than are most plaintiffs in class actions. Securities actions, like many suits under Rule 23, are lawyers' vehicles. Investors diversify their holdings, so it is no surprise that Kamen, like most plaintiffs in securities cases, does not hold very much stock in the defendants and has delegated the investigation and prosecution of the suit to counsel. Class actions are valuable precisely because they allow the vindication of claims too small to prosecute individually but worth litigating in the aggregate. See In re American Reserve Corp., 840 F.2d 487 (7th Cir. 1988). When defendants' counsel took Kamen's deposition and learned that she knew little about either the Fund or the case and had given counsel free reign, they learned only that this case fits the norm. Securities markets function efficiently because of the division of labor; legal markets also rely on specialization. Kamen did not need to immerse herself in the mutual fund business to qualify as a plaintiff. Surowitz v. Hilton Hotels Corp., 383 U.S. 363, 86 S.Ct. 845, 15 L.Ed.2d 807 (1966); Lewis v. Curtis, 671 F.2d 779, 788-89 (3d Cir. 1982). Counsel to whom Kamen entrusted the litigation — perhaps more accurately, who found Kamen to wage the litigation — is a specialist in the field, having argued and won Fox in the Supreme Court. There can be no doubt that Kamen's champion will advocate the claim vigorously and skilfully.

Magistrate Balog's conclusion that Kamen has only a private grievance misses the point. Kamen is not trying to get even because she bears a grudge — say, because a member of the Fund's board trampled her petunias. She seeks a higher rate of return on her investment. So do all other shareholders in the Fund. It may well be that most other shareholders believe that the Fund's special services are worth the 0.2% cost, but shareholders have a common interest in ensuring that the Fund pays Kemper no more than the extra services are worth. Commonality of interest is the essence of adequate representation. Sosna v. Iowa, 419 U.S. 393, 403, 95 S.Ct. 553, 559, 42 L.Ed.2d 532 (1975). If the services make a 0.2% premium fee "excessive", the defendants will prevail on the merits. If the services do not justify the incremental fee, then all shareholders will gain from a decision.

Given that Kamen and other investors win or lose together, the adequacy of her representation (more realistically, of her lawyer's) matters if the suit is strong. Suppose the Fund has a well-grounded claim against Kemper for $40 million in excessive fees. A feeble litigant, or one willing to sell out for satisfaction of her personal claim, would injure other investors in the Fund by extinguishing the claim against Kemper without producing its full value for the Fund. Other investors therefore could be worried that the first one to step forward and sue will be insufficiently vigorous, or have divided loyalties. No such investor has appeared to complain that Kamen will do too little for them, and neither the Fund nor Kemper seems worried that Kamen will under-prosecute this suit. Why is it that the defendants insist that the plaintiff is a poor representative of the other stockholders? Perhaps defendants fear that Kamen will be too vigorous?

Defendants' principal fear is not of inadequate representation but of legal error — that a court playing rate regulator may think the fees excessive even when they are not. Such a mistake would injure all investors, and the injury would fall more heavily on investors other than Kamen who use the sweeps and redemption services the Fund provides. Kamen would find other money market funds suited to her passive investment strategy; active investors would be especially aggrieved by a cutback in services. In this sense, even though Kamen's interests are the same as those of other investors — all want the Fund to pay no more than the market price for the services Kemper renders — Kamen has less to fear from an error and therefore is not the optimal champion. Still, such differences in incentives pervade class actions. Even if the members of the class were perfectly homogeneous (they never are), the representative's small stake might lead her to settle for too little or to press arguments that favored her position (or her attorney's) at the potential expense of those she represents. See Kenneth W. Dam, Class Actions: Efficiency, Compensation, Deterrence, and Conflict of Interest, 4 J. Legal Studies 47 (1975); Andrew Rosenfield, An Empirical Test of Class-Action Settlement, 5 J. Legal Studies 113 (1976). Agency costs of this kind, even when coupled with differential sensitivity to error costs, are not the same as concrete conflict of interest between the "representative" and other members of the class. They inhere in representative actions. The real bogies, the costs of defense and the risk of error, haunt all litigation. False positives and the potential for self-serving conduct are endemic to the system under § 36(b); the costs of legal error in regulating prices are attributable to the existence of § 36(b) and not to the selection of Kamen as a plaintiff.

Although the investors by majority vote approved Kemper's fees in 1986, the vote was not unanimous. Of the 5.3 billion shares of the Fund's money market portfolio in 1986, 2.97 billion were voted at the meeting (almost all by proxy). Approximately 2.44 billion were voted for the management agreement, 364 million against, and 169 million abstaining. The contract was ratified, then, by 82% of the shares present at the meeting, although only 46% of the outstanding shares. These figures do not suggest that Kamen is in a teensy minority; she had as of 1986 the company of the holders of 364 million shares of the Fund. Section 36(b)(2) provides that investors' approval of a management fee should inform the court's judgment on the merits; it does not imply that approval forecloses suit by one of the dissenters — if it did, § 36(b) would be dead, for all fees challenged under the statute have been approved by the investors at one or another time. So, too, it is not dispositive that none of the other investors has intervened to support Kamen's suit. Many a class action proceeds with a single representative; conservation on the number of litigants is a virtue of the device. Even if this were a case in which the plaintiff must represent others "adequately", then, Kamen would be a proper plaintiff.

III

Because the action may proceed under § 36(b), we need to decide whether Kamen is entitled to a jury trial in the event there are disputed issues of material fact. Judge Nordberg held not, following In re Evangelist, 760 F.2d 27 (1st Cir. 1985), Schuyt v. Rowe Price Prime Reserve Fund, Inc., 835 F.2d 45 (2d Cir. 1987), and Krinsk v. Fund Asset Management, Inc., 875 F.2d 404, 414 (2d Cir. 1989). These opinions hold that the action authorized by § 36(b) is not a "suit at common law" within the meaning of the seventh amendment because the statute creates a fiduciary duty and recovery "shall in no event exceed the amount of compensation or payments received from [the] investment company". § 36(b)(3). A combination of fiduciary duty with a remedy of cancellation and restitution is traditionally equitable.

Novel statutes such as § 36(b), establishing requirements and procedures that Fox repeatedly called "unique", do not fit well into a constitutional framework requiring the rights to jury trial as of 1791 to be "preserved". Federal courts abolished the distinction between law and equity with the adoption of the Rules of Civil Procedure in 1938, and changes in both the nature of legal rights and the preferred remedies make it difficult to reconstruct what our forbears would have seen as "common law". See Douglas G. Baird, The Seventh Amendment and Jury Trials in Bankruptcy, 1989 Sup.Ct.Rev. 261. No rights comparable to those of § 36(b) existed two centuries ago; the closest equitable action dealt with corporate "waste" and not with determining whether prices are reasonable. Although enforcing fiduciary duties was equitable in English practice, awarding damages was a job for a common law court. Squeezing a hybrid action into one category or the other is bound to cause friction. Recognizing that no answer can be wholly satisfactory, and not wanting to add unnecessarily to the clutter of opinions on the subject, we adopt both the holding and rationale of Evangelist.

One case postdating the First and Second Circuits' decisions calls for comment. Teamsters Local No. 391 v. Terry, ___ U.S. ___, 110 S.Ct. 1339, 108 L.Ed.2d 519 (1990), holds that a suit alleging that a union breached its duty of fair representation, and requesting damages measured by the amount of back pay, is "at common law" for constitutional purposes and so allows either side to demand a jury. The Court rejected an argument that the foundation of the suit on a breach of the union's fiduciary duty to its members was sufficient to render it "equitable". Even the fillip that the remedy would be measured by the amount of pay lost, a yardstick from the courts of equity, did not suffice. Terry appears to call into question the foundation for Evangelist and similar holdings about § 36(b).

Although any prediction is hazardous, we conclude that the Court would think an action under § 36(b) equitable under the analysis it used in Terry. Seven Justices accepted the proposition, central to cases such as Evangelist, that "an action by a trust beneficiary against a trustee for a breach of fiduciary duty" is equitable because it was "within the exclusive jurisdiction of the courts of equity" in 1791. 110 S.Ct. at 1346 (plurality opinion); see also id. at 1355 (Kennedy, J., dissenting). The investment adviser is not a "trustee", and the relation between fund and adviser is contractual rather than one in which the beneficiary of the trust lacks authority to choose the trustee; still, the statute creates a fiduciary duty and enforces it with a remedy (disgorgement) common in trust cases.

The analogy between the union's duty and a trustee's broke down in Terry because the union was supposed to be enforcing a contract, and the claim against the union depended on proof that the employer broke its contract. The remedy, although measured by pay lost, came from the union rather than the employer, which made it look more like damages than restitution. An action under § 36(b), quite unlike the action for a breach of the duty of fair representation, is one to annul a contract rather than to enforce it. Reformation or cancellation of a contract was equitable in 1791 and until the distinction between law and equity broke down in this century. Restitution under § 36(b) comes from the party that received the benefits, which further separates this case from Terry, Four different opinions in Terry, advocating four different approaches to the constitutional question, render parlous any predictions. Nonetheless, the combination of a fiduciary duty with a restitutionary remedy in § 36(b) continues to put this statute on the equitable side of the constitutional line.

The judgment of the district court is affirmed in part, reversed in part, and remanded for further proceedings on the claim under § 36(b).