RALPH K. WINTER, Circuit Judge:

This is an appeal from a grant of summary judgment for defendants by the District Court for the District of Connecticut, Eginton, Judge, dismissing a derivative action against certain directors and officers of Citytrust upon a recommendation of a special litigation committee and placing that report under seal. 519 F.Supp. 1312 (D.Conn. 1981).

We reverse.

BACKGROUND

In October of 1977, Dr. Athalie Doris Joy brought this shareholder's derivative suit on behalf of Connecticut Financial Services Corporation (now Citytrust Bancorp, Inc.) against its wholly-owned banking subsidiary, Citytrust, and the officers and directors of Citytrust. Both corporations are incorporated in Connecticut. The complaint alleged diversity of citizenship, common law breach of trust and of fiduciary duty as well as violations of the National Bank Act, 12 U.S.C. § 84 (1976), which limits aggregate loans to a single person or entity to 10% of a bank's combined stockholder equity and capital. The allegations concern loans made by Citytrust to the Katz Corporation ("Katz") for construction of an office building in a redevelopment area of Norwalk, Connecticut. Plaintiff seeks a $6 million recovery plus interest and attorney's fees.

The underlying transactions need only be briefly summarized at this point. In 1967, Citytrust entered into a 20-year term lease agreement for approximately 9% of an office building which Katz was planning to build in Norwalk. Katz, then a respected developer, signed a $4 million construction mortgage for a one-and-a-half year term on January 12, 1971. Although the mortgage was written through and recorded in the name of Citytrust, Chase Manhattan Bank provided the bulk of the financing, $3.5 million, with Citytrust participating to the extent of $500,000. At this time, Katz had already borrowed, largely in unsecured form, an additional $250,000 from Citytrust to finance construction of the office building. As the building neared completion in early 1972, Katz had drawn down the full value of the $4 million mortgage. At its expiration in June, 1972, the Chase mortgage was replaced by a $4.5 million mortgage by First National City Bank, with Citytrust both issuing the mortgage and participating to the extent of $90,000. Meanwhile, Katz continued to receive unsecured loans from Citytrust. By December, 1972, that unsecured debt reached $900,000, for a total of $990,000 in Citytrust loans related to the building.

In June, 1973, with the building only half rented, the First National City mortgage was extended for a year. Katz's unsecured debt to Citytrust had by now climbed to $1,840,000. In November, in conjunction with the issuance of yet another loan to Katz, Citytrust obtained a blanket second mortgage on the building and on other Katz properties to secure what was now a total loan balance of $2,140,000. Shortly thereafter, the First National City mortgage was extended to August, 1975, and Citytrust lent Katz another $300,000. Just prior to this extension of credit, the National Bank Examiners classified the Katz loans.

In April, 1975, a refinancing plan was completed with Lincoln National Life Insurance Company providing a $6 million loan to a Katz-related partnership which had taken title to the building. The loan was secured by a first mortgage on the building and was used to consolidate Katz's debt. As a condition of the new financing, Citytrust was required to take a 30-year master lease on the still largely unrented building at a rental equaling the mortgage payments to Lincoln National, in effect guaranteeing Katz's $6 million obligation to Lincoln. In addition to undertaking the master lease, Citytrust had by now extended $2,665,000 in loans to Katz.

In May, 1975, the National Bank Examiners classified $2 million of the Katz loans as doubtful and required a charge off of $665,000. On August 18, 1976, in an apparent effort to salvage what was left of its position, Citytrust's Board of Directors authorized loans which exceeded the 10% federal statutory limit. After these loans were consummated, Katz's total indebtedness to Citytrust reached $3,545,000. On October 20, 1976, Citytrust charged off the $2 million remaining on the second mortgage.

On June 13, 1977, the Katz-related partnership relinquished title to the building to Citytrust in exchange for a release from its obligation to Lincoln National and a release of personal guarantees previously assumed by members of the Katz family. Citytrust thus directly assumed the $6 million Lincoln National mortgage. In October, 1977, Second Nutmeg Financial purchased the building for $9,600,000 which consisted of its assumption of the $6 million Lincoln National mortgage and a $3,600,000 note to Citytrust secured by a second mortgage. There is an indication in the District Court record that an affiliate of Second Nutmeg which later acquired the building has defaulted and Citytrust once again owns it, along with the concurrent obligations. There is no indication that rental income is now adequate to meet those obligations, and we appear free to assume that the other Katz properties covered by the second mortgage are not of any significant value.

In October, 1977, Joy commenced this action after making an unsuccessful demand on the Directors of Citytrust. During the pendency of this case, the Supreme Court decided Burks v. Lasker, 441 U.S. 471, 99 S.Ct. 1831, 60 L.Ed.2d 404 (1979), holding that federal courts must apply state law in determining the authority of a committee of independent directors to discontinue derivative suits even in many cases which arise under federal law. Immediately following the Burks decision, the Board of Directors of Citytrust and Connecticut Financial Services Corporation authorized the establishment of a Special Litigation Committee to determine whether continued prosecution of this derivative action would be in the best interests of the corporation. The Committee consisted of two Board members, Marion S. Kellogg and Ernest C. Trefz. Kellogg was elected to the Board of Directors on July 21, 1976 and commenced service on September 15, 1976. Trefz was elected to the Board on December 15, 1976 and commenced service on January 3, 1977. Neither is a defendant in this action.

By resolution dated August 15, 1979, the full Board of Directors, a majority of whom were defendants, voted to delegate to the Committee the power to review, investigate and analyze the circumstances surrounding the pending derivative action. The Committee retained independent counsel, John Murtha, Esquire, to assist its investigation.

Nine months later, the Committee issued a Report recommending that the suit be discontinued as to 23 defendants, 20 of whom were outside directors of either Citytrust or Connecticut Financial Services and three of whom were either officers or directors or both. (The 23 will hereafter be referred to as the "outside defendants"). The Committee concluded there was "no reasonable possibility" that the outside defendants would be found liable. Its Report also recommended that settlement be considered with regard to seven defendants who were the senior officers most directly involved in the Katz loans. (These seven will hereafter be referred to as the "inside defendants"). As to them, the Committee found there was a "possibility" that one or more might be found to have been negligent. Counsel for the Committee made it clear to the District Court, however, that the decision to pursue settlement was not necessarily a decision to press the litigation against the inside defendants. If settlement is not reached, the Committee will reconsider whether to recommend termination of that portion of the action also.

When plaintiff declined to withdraw the action as to the outside defendants, the corporation filed a motion to dismiss the case as to them. The District Court permitted discovery on the limited issue of the Committee's " bona fides, motivation and thoroughness." 519 F.Supp. at 1315. Portions of the Committee Report, consisting of a summary and a detailed presentation of the Committee's factual findings, supplemented by expert opinion letters and counsel's memorandum of law, were produced. These documents were put under seal pursuant to a protective order. Plaintiff was also allowed to depose a variety of persons involved in the underlying transactions and in preparation of the Report, to pose interrogatories to others, and to see various documents relating to the Report.

After discovery, Judge Eginton granted the defendants' motion for summary judgment, the protective order remaining in force. Concluding that no dispositive Connecticut case or statute exists, Judge Eginton referred to the weight of authority in cases reported elsewhere. He held that Connecticut law permits the use of a Burks committee and that the business judgment rule limits judicial scrutiny of its recommendations to the good faith, independence and thoroughness of the Committee. 519 F.Supp. at 1325. He resolved these issues favorably to the Committee and, therefore, entered summary judgment in favor of the 23 outside defendants. Plaintiff appeals from the ruling. We reverse as to both the grant of summary judgment and the sealing of the Committee report.

DISCUSSION

The grounds of liability asserted are common law claims of negligence and breach of fiduciary duty, as well as violation of the National Bank Act. We agree with Judge Eginton for the reasons stated in his opinion that the Special Litigation Committee may seek dismissal of both the state common law and federal claims if Connecticut law authorizes it to do so. 519 F.Supp. at 1318-22; Burks, supra. We also agree with him that the Connecticut statutory and case law cited by the parties is not dispositive. Our task, therefore, is to predict what the Connecticut Supreme Court would do in a case such as the one before us.

Appellees assert that, since a board of directors can delegate all its powers to a committee, Conn.Gen.Stat.Ann. § 33-318(a) (West 1982), a special litigation committee of independent directors can decide whether a derivative action should be dismissed or continued. They further argue that, when an appropriate motion is made, courts must defer to the committee's recommendation under the so-called business judgment rule, even though the delegation of power is made by directors who are defendants in the action. Judge Eginton adopted that position and limited his inquiry to the Committee's good faith, independence and thoroughness. Appellees also assert that the Committee Report in question may be kept under seal, any public use being in violation of the District Court's order. Appellant claims an absolute right to maintain a derivative action once begun and challenges the protective order on constitutional and non-constitutional grounds.

An examination of these claims requires a discussion of some underlying principles of corporate law. Our opinion first addresses the nature and function of the business judgment rule, which played a large role in persuading the District Court to dismiss this action. It turns then to the legal oddity known as the derivative action, thought by many to be an endangered species as a consequence of the evolution of special litigation committees. See Comment, Special Litigation Committees — An Expanding and Potent Threat to Shareholder Derivative Suits, 2 Cardozo L.Rev. 169 (1980); Note, The Business Judgment Rule in Derivative Suits Against Directors, 65 Cornell L.Rev. 600 (1980); Dent, The Power of Directors to Terminate Shareholder Litigation: The Death of the Derivative Suit, 75 Northwestern L.Rev. 96 (1980). Finally, it discusses the general principles applicable to attempts by special litigation committees to terminate particular derivative actions, and their relevance to the present case.

A. The Liability of Corporate Directors and Officers and the Business Judgment Rule

While it is often stated that corporate directors and officers will be liable for negligence in carrying out their corporate duties, all seem agreed that such a statement is misleading. See generally, Lattin, Corporations, 272-75 (1971). Whereas an automobile driver who makes a mistake in judgment as to speed or distance injuring a pedestrian will likely be called upon to respond in damages, a corporate officer who makes a mistake in judgment as to economic conditions, consumer tastes or production line efficiency will rarely, if ever, be found liable for damages suffered by the corporation. See generally, Symposium, Officers' and Directors' Responsibilities and Liabilities, 27 Bus.Lawyer 1 (1971); Fever, Personal Liabilities of Corporate Officers and Directors, 28-42 (2d ed. 1974). Whatever the terminology, the fact is that liability is rarely imposed upon corporate directors or officers simply for bad judgment and this reluctance to impose liability for unsuccessful business decisions has been doctrinally labelled the business judgment rule. Although the rule has suffered under academic criticism, see, e.g., Cary, Standards of Conduct Under Common Law, Present Day Statutes and the Model Act, 27 Bus.Lawyer 61 (1972), it is not without rational basis.

First, shareholders to a very real degree voluntarily undertake the risk of bad business judgment. Investors need not buy stock, for investment markets offer an array of opportunities less vulnerable to mistakes in judgment by corporate officers. Nor need investors buy stock in particular corporations. In the exercise of what is genuinely a free choice, the quality of a firm's management is often decisive and information is available from professional advisors. Since shareholders can and do select among investments partly on the basis of management, the business judgment rule merely recognizes a certain voluntariness in undertaking the risk of bad business decisions.

Second, courts recognize that after-the-fact litigation is a most imperfect device to evaluate corporate business decisions. The circumstances surrounding a corporate decision are not easily reconstructed in a courtroom years later, since business imperatives often call for quick decisions, inevitably based on less than perfect information. The entrepreneur's function is to encounter risks and to confront uncertainty, and a reasoned decision at the time made may seem a wild hunch viewed years later against a background of perfect knowledge.

Third, because potential profit often corresponds to the potential risk, it is very much in the interest of shareholders that the law not create incentives for overly cautious corporate decisions. Some opportunities offer great profits at the risk of very substantial losses, while the alternatives offer less risk of loss but also less potential profit. Shareholders can reduce the volatility of risk by diversifying their holdings. In the case of the diversified shareholder, the seemingly more risky alternatives may well be the best choice since great losses in some stocks will over time be offset by even greater gains in others. Given mutual funds and similar forms of diversified investment, courts need not bend over backwards to give special protection to shareholders who refuse to reduce the volatility of risk by not diversifying. A rule which penalizes the choice of seemingly riskier alternatives thus may not be in the interest of shareholders generally.

Whatever its merit, however, the business judgment rule extends only as far as the reasons which justify its existence. Thus, it does not apply in cases, e.g., in which the corporate decision lacks a business purpose, see Singer v. Magnavox, 380 A.2d 969 (Del.Supr. 1977), is tainted by a conflict of interest, Globe Woolen v. Utica Gas Electric Co., 224 N.Y. 483, 121 N.E. 378 (1918), is so egregious as to amount to a no-win decision, Litwin v. Allen, 25 N.Y.S.2d 667 (N.Y.Co.Sup.Ct. 1940), or results from an obvious and prolonged failure to exercise oversight or supervision, McDonnell v. American Leduc Petroleums, Ltd., 491 F.2d 380 (2d Cir. 1974); Atherton v. Anderson, 99 F.2d 883 (6th Cir. 1938). Other examples may occur.

B. Shareholder Derivative Actions

Whereas ordinary lenders may and will sue directly to enforce their rights and debentureholders look to indenture trustees to enforce obligations to them, direct actions by individual shareholders for injuries to the value of their investment would be an inefficient and wasteful method of enforcing management obligations. The stake of each shareholder in the likely return is usually too small to justify bringing a lawsuit and a multiplicity of such actions would result in corporate and judicial waste. Moreover, the costs of organizing a large number of geographically diverse shareholders to bring an action are usually prohibitively high. If an alternative remedy were not available, therefore, the fiduciary obligations of corporate management, however limited, might well be unenforceable. Moreover, state or federal law may impose other duties upon directors and officers which are designed to protect shareholders through procedural or other requirements, e.g. proxy rules. Enforcement of such obligations by shareholders, even where monetary recovery by the corporation is doubtful, may be desirable. J.I. Case Co. v. Borak, 377 U.S. 426, 84 S.Ct. 1555, 12 L.Ed.2d 423 (1964); Mills v. Electric Auto-Lite, 396 U.S. 375, 90 S.Ct. 616, 24 L.Ed.2d 593 (1970).

The derivative action is the common law's inventive solution to the problem of actions to protect shareholder interests. In its classic form, a derivative suit involves two actions brought by an individual shareholder: (i) an action against the corporation for failing to bring a specified suit and (ii) an action on behalf of the corporation for harm to it identical to the one which the corporation failed to bring. See Ross v. Bernhard, 396 U.S. 531, 90 S.Ct. 733, 24 L.Ed.2d 729 (1970). The technical structure of the derivative suit is thus quite unusual. Moreover, the shareholder plaintiffs are quite often little more than a formality for purposes of the caption rather than parties with a real interest in the outcome. Since any judgment runs to the corporation, shareholder plaintiffs at best realize an appreciation in the value of their shares. The real incentive to bring derivative actions is usually not the hope of return to the corporation but the hope of handsome fees to be recovered by plaintiffs' counsel. As two leading commentators state:

[T]he derivative action constitutes a major bulwark against managerial self-dealing. As a practical matter this means that the rules governing plaintiffs' legal fees are critical to the operation of the corporate system: Since very few shareholders would pay an attorney's fee out of their own pocket to finance a suit that is brought on the corporation's behalf and normally holds only a slight and indirect benefit for the plaintiff, very few derivative actions would be brought if the law did not allow the plaintiff's attorney to be compensated by a contingent fee payable out of the corporate recovery.

Cary and Eisenberg, Corporations 938 (5th ed. 1980).

However, there is a danger in authorizing lawyers to bring actions on behalf of unconsulted groups. Derivative suits may be brought for their nuisance value, the threat of protracted discovery and litigation forcing settlement and payment of fees even where the underlying suit has modest merit. Such suits may be harmful to shareholders because the costs offset the recovery. Thus, a continuing debate surrounding derivative actions has been over restricting their use to situations where the corporation has a reasonable chance for benefit.

C. Termination of Derivative Suits by Special Litigation Committees

In the normal course of events a decision whether to bring a lawsuit is a corporate economic decision subject to the business judgment rule. United Copper Securities Co. v. Amalgamated Copper Co., 244 U.S. 261, 37 S.Ct. 509, 61 L.Ed. 1119 (1917). Thus, shareholders upset at a corporate failure to bring actions for, say, nonpayment of a debt for goods sold and delivered, may not initiate a derivative suit without first making a demand upon the directors to bring the action. Where the directors refuse, and the derivative action challenges that refusal, courts apply the business judgment rule to the action of the directors. In a demand-required case, therefore, the directors' decision will be conclusive unless bad faith is proven.

Different rules apply, however, in the cases which primarily concern us here. When there is a conflict of interest in the directors' decision not to sue because the directors themselves have profited from the transaction underlying the litigation or are named defendants, no demand need be made and shareholders can proceed directly with a derivative suit. Note, Demand on Directors and Shareholders as a Prerequisite to a Derivative Suit, 73 Harv.L.Rev. 746, 753-56 (1960). It is in demand-not-required cases that the special litigation committee plays its role.

Appellees argue that, because special litigation committees are composed of "independent" directors — usually newly-elected directors who are not defendants — courts should treat their recommendations as the equivalent of a board refusal to bring an action in demand-required cases. If that proposition were accepted, the business judgment rule would apply in full force to the recommendation, and judicial scrutiny would be limited to the good faith, independence and thoroughness of the committee, as it is in the case of everyday business decisions. Appellant argues, on the other hand, that such committees are transparent devices enabling implicated directors to avoid liability and that derivative actions in demand-not-required cases are immune from termination whatever the recommendation of special litigation committees. We disagree with both parties.

We believe Connecticut would not adopt appellees' contention that the business judgment rule should play a major role where a special litigation committee recommends termination of an action in a demand-not-required case, such as the one before us. As a practical matter, new board members are selected by incumbents. The reality is, therefore, that special litigation committees created to evaluate the merits of certain litigation are appointed by the defendants to that litigation. It is not cynical to expect that such committees will tend to view derivative actions against the other directors with skepticism. Indeed, if the involved directors expected any result other than a recommendation of termination at least as to them, they would probably never establish the committee. The conflict of interest which renders the business judgment rule inapplicable in the case of directors who are defendants is hardly eliminated by the creation of a special litigation committee.

It is here that we part company with Judge Cardamone. While he recognizes that the business judgment rule has never applied to corporate decisions tainted by a conflict of interest, he argues that the conflict in the defendants' creation of a committee to determine whether this action should be terminated is wholly cured by a judicial finding that the committee acted independently and in good faith. This view is a major departure from the traditional scrutiny courts have given to the underlying fairness of corporate decisions which benefit directors. Lattin, Corporations at 293; see also Ferris v. Polycast Technology Corp., 180 Conn. 199, 208-09, 429 A.2d 850, 854 (1980) and cases cited therein. To be sure, Judge Cardamone is correct in anticipating difficulties in judicial review of the recommendations of special litigation committees. These difficulties are not new, however, but have confronted every court which has scrutinized the fairness of corporate transactions involving a conflict of interest.

Moreover, the difficulties courts face in evaluation of business decisions are considerably less in the case of recommendations of special litigation committees. The relevant decision — whether to continue litigation — is at hand and the danger of deceptive hindsight simply does not exist. Moreover, it can hardly be argued that terminating a lawsuit is an area in which courts have no special aptitude. Citytrust's Special Litigation Committee concluded that there was "no reasonable possibility" that 23 outside defendants would be held liable. A court is not ill-equipped to review the merits of that conclusion. Even when the Committee recommendation arises from the fear of further damage to the corporation, for example, the distraction of key personnel, the cost of complying with discovery, and the possible indemnification of defendants out of the corporate treasury, courts are not on unfamiliar terrain. The rule we predict Connecticut would establish emphasizes matters such as probable liability and extent of recovery. For these reasons we hold that the wide discretion afforded directors under the business judgment rule does not apply when a special litigation committee recommends dismissal of a suit.

We think Connecticut would reject appellees' argument for a second reason. Limiting judicial scrutiny in cases such as the one before us to the good faith, thoroughness and independence of the special litigation committee would effectively eliminate the fiduciary obligation of directors and officers. As adumbrated further below, the present action involves classic allegations and substantial evidence of mismanagement and perhaps deliberate wrongdoing resulting in a loss exceeding 10% of the shareholders' capital and equity. The traditional fiduciary obligations of directors and officers under Connecticut law can hardly be said to exist if the sole enforcement method can be eliminated on a recommendation of the defendants' appointees. Other well-understood principles of corporate law would also be altered. For example, the requirement of a demand upon directors before initiating a derivative suit does not apply to a case such as the present one, where the corporate directors and officers are defendants. One reason — the other being futility — underlying the rule is that directors and officers cannot render a fair judgment on allegations of their own misconduct. Appellees would have us substantially modify this by allowing the defendants to appoint a committee to evaluate these allegations and impose their conclusions on the plaintiffs. In derivative actions, the plaintiffs have always controlled the action subject to judicial findings of adequate representation of shareholders and approval of settlements. See Conn. Gen.Stat.Ann. § 52-572j (West 1982). Appellees' view essentially vests power in defendants' appointees to bring about their dismissal.

We are aware that Auerbach v. Bennett, 47 N.Y.2d 619, 419 N.Y.S.2d 920, 393 N.E.2d 994 (1979), held that the business judgment rule limits judicial scrutiny of the recommendations of special litigation committees to their good faith, thoroughness and independence. Because we believe that test would work a major transformation of Connecticut corporate law, we predict that Connecticut would not adopt it. While the ongoing debate over the legal obligations of corporate management has spawned literature from Connecticut commentators calling for less judicial scrutiny of corporate transactions, Wolfson, A Critique of Corporate Law, 34 Miami L.Rev. 959 (1980), the special litigation committee, as envisioned in Auerbach, seems a rather blunt instrument to accomplish that end, since it appears to allow dismissals in actions for deliberate looting as well as in nuisance suits.

We detect no signals that Connecticut law is moving away from the enforcement of directors' and officers' traditional fiduciary obligations. Connecticut legislation governing indemnification of directors and officers for expenses incurred in litigation supports the view we take here. Conn.Gen. Stat.Ann. § 33-320a(b) (West 1982). Although that provision has no direct application to this case, attitudes toward indemnification are relevant to the issue before us. By its terms, the Connecticut statute is exclusive and cannot be varied by corporate by-laws. It calls for indemnification without court approval only in circumstances in which the defendant directors and officers secure a judgment in their favor. This legislation adopts the middle ground between no indemnification and permissible indemnification without regard to outcome and thus does not bespeak a negative attitude toward enforcement of fiduciary obligations through meritorious derivative litigation.

We also disagree with appellant that recommendations of special litigation committees should be ignored by courts in which derivative actions are pending. The incentives underlying derivative litigation are such that actions may well be brought which cannot be dismissed on motion but which also are unlikely to lead to net benefits to the corporation. At least where the effectuation of an overriding federal policy is not at stake, cf. Galef v. Alexander, 615 F.2d 51, 62 (2d Cir. 1980) ("serious questions" whether a special litigation committee can seek dismissal of an action alleging federal proxy violations), a corporation can play a legitimate role in aiding a court to determine whether the maintenance of an action in its name is in fact in its interest. Surviving a motion to dismiss and for summary judgment establishes only that a claim for relief has been stated and that there is a scintilla of evidence to support it. It does not establish that continued prosecution of the action is actually in the corporation's interest.

Appellant argues that Connecticut recognizes an absolute right to bring and continue a derivative action. The statute relied upon, however, Conn.Gen.Stat.Ann. § 52-572j, is clearly of a procedural nature directing how and where such actions may be brought in a state court. It has no apparent substantive effect. Appellant also relies upon Conn.Gen.Stat.Ann. § 33-323 (West 1982) which renders transactions between the corporation and directors voidable if unfair to the corporation. She argues that creation of the Committee is thus a voidable corporate act since a majority of those voting were defendants. Since our decision permits dismissal of this action only after an independent judicial finding that it will not benefit the corporation, the statutory fairness requirement is satisfied.

The Connecticut statute permitting indemnification with court approval of a director's expenses incurred in defending litigation even when the settlement calls for payment to the corporation, Conn.Gen. Stat.Ann. § 33-320a(b), suggests a recognition by the General Assembly of the fact that some derivative actions may in the end not benefit the corporation. We believe Connecticut law allows a court before which a derivative action is pending to entertain a motion for judgment by a defendant corporation based on a recommendation of a special litigation committee. Before granting such a motion, however, it should apply far more vigorous scrutiny to that recommendation than occurs under the good faith, independence and thoroughness test. While those qualities are necessary even to consideration of the merits of such a motion, they are by no means sufficient to determine whether it should be granted. The reason for entertaining such a motion stems from the nature of the incentives underlying derivative litigation. Because they encourage actions which are not subject to pretrial dismissal but which also may not benefit the corporation, the device of the special litigation committee affords an opportunity for a judicial second look at the underlying litigation. There is no reason, however, to treat the recommendation of such a committee as having presumptive weight or to accord it deference beyond its inherent persuasiveness. The function of judicial scrutiny of a committee's recommendation is to determine independently whether the action is likely to harm the corporation rather than help it.

We strongly disagree with the implications of the dissenting opinion that by not according the recommendations of special litigation committees conclusive weight we are somehow embarking on a new course of "stricter corporate accountability" which will lead to "more derivative lawsuits." Such committees are very recent creations. Indeed, Auerbach, the first decision of a state high court to address these issues, was decided only three years ago. Our failure to accord the recommendations of such committees conclusive weight can hardly be regarded as overregulation. And our holding that the recommendations of such committees may be considered by a court and provide the basis for dismissal of a derivative action will hardly lead to an increase in the number of such actions brought.

The Auerbach decision gives excessive weight to the recommendations of special litigation committees. In rejecting Auerbach and concluding that Connecticut would adopt a rule limiting the role of such committees, we are not without eminent judicial support. In Zapata Corp. v. Maldonado, 430 A.2d 779 (Del.Supr. 1979), the Delaware Supreme Court, which has unique experience in the area, laid down the following rule: Where a derivative suit cannot be brought without prior demand upon the directors followed by refusal, the directors' decision will stand absent a demonstration of self-interest or bad faith; but where such a demand is excused (for reasons of futility, etc.) and a derivative action is properly brought, an independent committee of directors may obtain a dismissal only if the trial court finds both (a) that the committee was independent, acted in good faith and made a reasonable investigation; and (b) that in the court's independent business judgment as to the corporation's best interest, the action should be dismissed. We believe Connecticut would adopt a similar rule.

Independent judicial scrutiny of a special litigation committee's recommendation is apt to be difficult and in an effort to ease that task we establish some guidelines. We emphasize that what we say here applies to cases involving allegations of direct economic injury to the corporation diminishing the value of the shareholders' investment as a consequence of fraud, mismanagement or self-dealing. This is not a case involving allegations of ultra vires acts or acts illegal under domestic or foreign laws which do not have as a major purpose protection of the corporate shareholder's investment interest. Compare Gall v. Exxon Corp., 418 F.Supp. 508 (S.D.N.Y. 1976). Nor is it a case in which the relief sought may benefit the corporation but in ways not entailing a direct financial return. Mills v. Electric Auto-Lite, supra (action for misleading proxy statement where no monetary recovery by the corporation may result); Bosch v. Meeker Cooperative Light Power Ass'n, 257 Minn. 362, 101 N.W.2d 423 (1960) (invalidation of directors' elections). The guidelines we establish here are limited to cases involving allegations of acts resulting in direct financial harm to the corporation and a consequent diminishing of the value of shareholders' investment. We express no views on the appropriate calculus to be applied to recommendations of special litigation committees where the derivative action alleges violations of law not designed principally to protect shareholders or, which, if successful, may benefit the corporation in ways other than the recovery of compensatory damages.

In cases such as the present one, the burden is on the moving party, as in motions for summary judgment generally, to demonstrate that the action is more likely than not to be against the interests of the corporation. This showing is to be based on the underlying data developed in the course of discovery and of the committee's investigation and the committee's reasoning, not simply its naked conclusions. The weight to be given certain evidence is to be determined by conventional analysis, such as whether testimony is under oath and subject to cross-examination. Finally, the function of the court's review is to determine the balance of probabilities as to likely future benefit to the corporation, not to render a decision on the merits, fashion the appropriate legal principles or resolve issues of credibility. Where the legal rule is unclear and the likely evidence in conflict, the court need only weigh the uncertainties, not resolve them. The court's function is thus not unlike a lawyer's determining what a case is "worth" for purposes of settlement.

Where the court determines that the likely recoverable damages discounted by the probability of a finding of liability are less than the costs to the corporation in continuing the action, it should dismiss the case. The costs which may properly be taken into account are attorney's fees and other out-of-pocket expenses related to the litigation and time spent by corporate personnel preparing for and participating in the trial. The court should also weigh indemnification which is mandatory under corporate by-laws, private contract or Connecticut law, discounted of course by the probability of liability for such sums. We believe indemnification the corporation may later pay as a matter of discretion should not be taken into account since it is an avoidable cost. The existence or non-existence of insurance should not be considered in the calculation of costs, since premiums have previously been paid. The existence of insurance is relevant to the calculation of potential benefits.

Where, having completed the above analysis, the court finds a likely net return to the corporation which is not substantial in relation to shareholder equity, it may take into account two other items as costs. First, it may consider the impact of distraction of key personnel by continued litigation. Second, it may take into account potential lost profits which may result from the publicity of a trial.

Judicial scrutiny of special litigation committee recommendations should thus be limited to a comparison of the direct costs imposed upon the corporation by the litigation with the potential benefits. We are mindful that other less direct costs may be incurred, such as a negative impact on morale and upon the corporate image. Nevertheless, we believe that such factors, with the two exceptions noted, should not be taken into account. Quite apart from the elusiveness of attempting to predict such effects, they are quite likely to be directly related to the degree of wrongdoing, a spectacular fraud being generally more newsworthy and damaging to morale than a mistake in judgment as to the strength of consumer demand.

We do recognize two exceptions, however. First, where the likely net return is not substantial in relation to shareholder equity, the court can consider the degree to which key personnel may be distracted from corporate business by continuance of the litigation. We appreciate that litigation can disrupt the decision-making process and thereby impose unforeseen and undetected costs. These are not measurable and we limit consideration of them to cases where the likely return to the corporation is not great. Where that is the case and many of the key directors and officers will be heavily involved in the litigation, a court may take such potential costs into account.

Second, where the corporation deals with the general public and its level of business is dependent upon public identification and acceptance of the corporate product or service, we believe the court ought to take potential business lost as a consequence of a trial into account when the likely net return to the corporation is not substantial in relation to total shareholder equity. In such a case, there is less likelihood of a direct relationship between impact on business and degree of misconduct. Where the likely return to the corporation from the litigation is higher, however, we believe the uncertainty as to the kind of publicity which will attend a trial precludes consideration of that impact. Moreover, when potential lost profits are taken into account, the basis for calculating them must be something more solid than the conclusory opinions of alleged experts, e.g., verifiable examples in similar firms.

D. Procedural Issues

Review of special litigation committee reports also raises a number of procedural issues. First, documents used by parties moving for, or opposing, summary judgment should not remain under seal absent the most compelling reasons. Fed.R.Civ.P. 26(c) authorizes protective orders in the course of discovery, "[a]n important purpose of [which] is to preserve the confidentiality of materials which are revealed in discovery but not made public at trial." National Polymer Products v. Borg-Warner Corp., 641 F.2d 418, 424 (6th Cir. 1981). Protective orders are useful to prevent discovery from being used as a club by threatening disclosure of matters which will never be used at trial. Discovery involves the use of compulsory process to facilitate orderly preparation for trial, not to educate or titillate the public. Private matters which are discoverable may, upon a showing of cause, be put under seal under Rule 26(c), in the first instance. Martindell v. International Tel. Tel. Corp., 594 F.2d 291 (2d Cir. 1979), says no more than that.

At the adjudication stage, however, very different considerations apply. An adjudication is a formal act of government, the basis of which should, absent exceptional circumstances, be subject to public scrutiny. We simply do not understand the argument that derivative actions may be routinely dismissed on the basis of secret documents. We cannot say what the effect on investor confidence would be if special litigation committees were routinely allowed to do their work in the dark of night. We believe, however, that confidence in the administration of justice would be severely weakened. Indeed, any other rule might well create serious constitutional issues. See Globe Newspaper Co. v. Superior Court, ___ U.S. ___, 102 S.Ct. 2613, 73 L.Ed.2d 248 (1982).

We do not say that every piece of evidence, no matter how tangentially related to the issue or how damaging to a party disclosure might be, must invariably be subject to public scrutiny. An exercise of judgment is in order. The importance of the material to the adjudication, the damage disclosure might cause, and the public interest in such materials should be taken into account before a seal is imposed.

Second, if the special litigation committee recommends termination and a motion for judgment follows, the committee must disclose to the court and the parties not only its report but all underlying data. To the extent that communications arguably protected by the attorney-client privilege may be involved in that data, a motion for judgment based on the report waives the privilege. See In re John Doe Corporation, 675 F.2d 482 (2d Cir. 1982). The work-product immunity will apply to the documents usually included within its terms to the extent that they are working papers of the committee's counsel and are not communicated to the committee. Once communicated, the immunity may not be claimed, since the papers may be part of the basis for the committee's recommendations.

E. The Present Case

Given the principles outlined above, we reverse the order putting the Citytrust Committee's Report under seal and restricting its use. The showing of good cause was merely that it and the accompanying documents were protected by the attorney-client privilege and work-product immunity and involved a candid assessment of the bank's internal operations which, if made public, would adversely affect the two corporations in the banking industry and the Bridgeport community. This showing is patently inadequate. First, the submission of materials to a court in order to obtain a summary judgment utterly precludes the assertion of the attorney-client privilege or work-product immunity. Second, a naked conclusory statement that publication of the Report will injure the bank in the industry and local community falls woefully short of the kind of showing which raises even an arguable issue as to whether it may be kept under seal. The Report is no longer a private document. It is part of a court record. Since it is the basis for the adjudication, only the most compelling reasons can justify the total foreclosure of public and professional scrutiny. The potential harm asserted by the corporate defendants is in disclosure of poor management in the past. That is hardly a trade secret. The argument that disclosure of poor management is so harmful as to justify keeping the Report under seal proves too much since it is a claim which grows stronger with the degree of misconduct. Were outright looting of Citytrust disclosed by the Report, for example, the harm of publication would be even greater. Moreover, Citytrust is a publicly owned company and this litigation directly concerns management's obligations to shareholders. We believe that foreclosing public scrutiny of the grounds for this adjudication is wholly unjustifiable.

We turn now to the contents of the Special Litigation Committee's Report. We emphasize that this recitation is the Committee's version of the facts. The record suggests that a trial might reveal sharply differing versions of the same events from various witnesses as well as sharply differing inferences drawn from that testimony.

According to the Report, Nelson L. North was Citytrust's Chief Executive Officer and Norman Schaff, Jr. was its Chief Lending Officer during the period in question. The management of Citytrust was completely dominated by North. Bank officers who did not temper themselves to his regime had a short tenure at the bank. North also exercised strong control over the activities of the Board of Directors. Board members were given neither materials nor agendas prior to meetings and requests for long range planning documents were left unanswered. North's control is illustrated by the fact that contrary to the recommendation of Citytrust's outside auditors, the bank's Audit Committee was not composed solely of outside directors but instead counted among its members Mr. North and one other officer. Minutes of the Audit Committee between 1971 and 1974 are largely incomplete.

Mr. North apparently brought the initial proposal for the Katz loan to Citytrust. From 1971 to 1976, North's son was employed by Katz, and he apparently deemed this a sufficient conflict of interest to preclude his voting on the Katz transactions in Executive Committee meetings. This fastidiousness appears to have been limited to the formality of voting, for the Report strongly suggests that North was deeply involved in the Katz transactions, although the full degree of his involvement is left uncertain. The Report also adds that Mr. North has destroyed his records.

Katz appears to have been experiencing financial difficulties as early as 1971. Chase Manhattan in fact opposed financing the Katz building in part because of a $1.6 million shortfall between the building cost and available lending; it was North who persuaded Chase to make the initial mortgage loan. By 1972, Katz was falling behind in its loans and by December of that year owed Citytrust $990,000 with respect to the building. The Report concludes that by then Citytrust was effectively a joint venturer with Katz in the building, sharing the risk of loss but entitled at best only to interest and principal if things went well. There is also some indication that a portion of the unsecured advance made by Citytrust was being applied to the Chase mortgage.

Notwithstanding the increasingly evident peril in Citytrust's transactions with Katz, no appraisals of the buildings were undertaken until 1976, and no rentability study until 1974. Although Katz had suggested that a public offering would alleviate the situation, no professional review of the preliminary prospectus was undertaken. From 1972 through 1973, only one meeting of the Board of Directors or Executive Committee considered the Katz loans. The Senior Loan Committee did meet on the Katz matter late in 1972 and may have adopted a very cautious attitude toward further credit extension. Despite this, and despite the absence of Executive Committee and Board support, senior management extended almost a million dollars in loans to Katz between 1972 and 1973.

From 1973 through 1974, the number of Board meetings at which the Katz loans were considered increased to five. This is roughly contemporaneous with the recommendation of the outside auditors that a 50% special fund be set up for the Katz loans and the National Bank Examiners' classification of the total outstanding Katz indebtedness as substandard. It is, as the Report notes, "unsettling" that neither Schaff nor Citytrust's Comptroller recall being advised of the recommendation as to the special reserve. Moreover, it is not established that the Directors were advised of this recommendation.

By late 1974, the Katz loans were so clearly a problem that they were extensively considered by the Board and the Executive Committee. In fact, the Report notes that these loans were discussed at a minimum of 25 Board and Executive Committee meetings. Nevertheless, when, on August 18, 1976, the Board was presented with the request to go over the 10% limit, there was no prior mention of the issue on the agenda nor was opinion of counsel presented to the Board or even sought. Indeed, copies of the Comptroller's letter suggesting that Directors might wish to consult their personal counsel were not distributed to the Board.

The Report estimates a loss of $5.1 million to Citytrust. As stated earlier, there is an indication in the record that since the Report was issued, the new owner has defaulted and Citytrust again owns the building. If so, $5.1 million may be considerably less than the actual loss. In any event, a loss exceeding 10% of shareholder equity seems quite likely.

The Report contains the opinion of two experts. One concluded that the impact on morale of bank personnel, on the image of Citytrust among the banking public, upon persons who might be asked to become directors and upon potential new customers would offset even a recovery of $5.5 million. The other reached a different conclusion, stating that a recovery of even $2 million would be worth pursuing notwithstanding speculation about the public impact. It stated that this opinion would stand whether or not the outside directors continue as defendants, "as long as the insurance carrier is obligated through the `D and O policy'." This last phrase might have given the Committee some pause since the letter requesting the opinion indicated that the insurance carrier had raised a question as to its liability.

As to exceeding the federal statutory limit, the Committee concluded that under the compelling circumstances surrounding the vote, it is possible that no net damage to Citytrust resulted. It also concluded that even if damage did occur, the maximum recovery would be $376,000 plus interest, a sum too small in the Committee's view to justify continuance of the action.

As to the claims of breach of fiduciary duty, the Committee recommended that the suit be discontinued as to the outside defendants because there is "no reasonable possibility" that they might be found liable. As to the others, it concluded merely that

there is a possibility that a finding of negligence could be rendered against any one or more of the senior loan officers who participated in the Katz Corporation loans. Although it is emphasized that there is no evidence whatsoever of any self-dealing or of any deliberate impropriety, there is some indication that the most prudent lending principles were not adhered to during the evolution of those loans.

Applying the standard of review set out above to the Committee's recommendation, we look first to potential liability generally without regard to which defendants are responsible. As to that liability, we find that plaintiff's chances of success are rather high. The loss to Citytrust resulted from decisions which put the bank in a classic "no win" situation. The Katz venture was risky and increasingly so. By continuing extensions of substantial amounts of credit the bank subjected the principal to those risks although its potential gain was no more than the interest it could have earned in less risky, more diversified loans. In a real sense, there was a low ceiling on profits but only a distant floor for losses. It is so similar to the classic case of Litwin v. Allen, supra (bank purchase of bonds with an option in the seller to repurchase at the original price, the bank thus bearing the entire risk of a drop in price with no hope of gain beyond the stipulated interest) that we cannot agree with the Committee's conclusion that only a "possibility of a finding of negligence" exists.

The issue as to which defendants are responsible is less clear. The Committee concluded that there is "no reasonable possibility" of the outside defendants being found liable because they had neither information nor reasonable notice of the problems raised by the Katz transactions. We note first that members of the inside defendants may contradict that version and, if so, a possibility of liability in the outside group exists. Moreover, lack of knowledge is not necessarily a defense, if it is the result of an abdication of directional responsibility. McDonnell, supra; Atherton, supra. Directors who willingly allow others to make major decisions affecting the future of the corporation wholly without supervision or oversight may not defend on their lack of knowledge, for that ignorance itself is a breach of fiduciary duty. The issue turns in large part upon how and why these defendants were left in the dark. See Graham v. Allis Chalmers Mfg. Co., 41 Del.Ch. 78, 188 A.2d 125 (1963). An individual analysis of each outside defendant's role may show that some are blameless or even that they all were justified in not acting before they did, but neither is an inexorable conclusion on the basis of the present record.

The Report concluded as to the inside defendants that there was a "possibility" of liability. This conclusion is a considerable understatement and not entirely consistent with the Report's finding as to the outside defendants. The outsiders' best defense may well be that the inside group actively concealed the Katz problem. Given the fact that exoneration of the outside defendants may show culpability of the insiders and our conclusion that the probability of liability somewhere is high, we think the exposure of the inside group is considerably more than a "possibility." Nor do we agree that "there is no evidence whatsoever" of deliberate impropriety. Not only is there the problem of North's apparently inconsistent behavior with respect to the appropriateness of his participation in the considerations of Katz transactions, but his failure to keep the Board of Directors informed may well entail more than a negligent omission.

A precise estimate of potential damages is not possible since the trier must determine at what point liability begins. We think, however, that on the present record, a trier might easily find liability extending back to early 1972 or before (assuming no statute of limitations problem), resulting in a return of several million dollars to Citytrust, or perhaps 10% or more of the shareholder equity. This far exceeds the potential cost of the litigation to the corporation.

CONCLUSION

Applying the analysis described above, we conclude that the probability of a substantial net return to the corporation is high. We reject, therefore, the recommendation of the Special Litigation Committee. The grant of summary judgment is reversed, the protective order is vacated, and the case is remanded. Since we have been unable to explain our reasoning in this opinion without extensive reference to materials under seal, the mandate shall issue forthwith as to the lifting of the protective order.