We revisit a plethora of issues left open in our decision in Leigh v. Engle, 727 F.2d 113 (7th Cir.1984) ("Leigh I" ). The district court's thorough opinion makes our task less arduous. See Leigh v. Engle, 669 F. Supp. 1390 (N.D.Ill.1987) ("Leigh II" ). We conclude that the trial judge ably applied Leigh I on remand; therefore we affirm.

The facts are fully discussed in Leigh I, 727 F.2d at 115-21. We will sketch out only those necessary to this appeal.

Plaintiffs are beneficiaries of the Reliable Manufacturing Corporation Employees Profit Sharing Trust ("Reliable Trust").1 They contend that defendants violated fiduciary duties under ERISA, by investing trust assets in corporate control contests with which defendants were associated.

The defendants are a network of individuals and companies associated with defendant Clyde Engle, a financier and investor. In particular, Libco is a company controlled by Engle. The administrators of the Reliable Trust, Nathan Dardick and Ronald Zuckerman, were officers or directors of several Engle-controlled entities. We will refer to the complete Engle network, fully described in Leigh I, 727 F.2d at 116-18, as the "Engle group."2

The suit concerns the trust's investments in three common stocks of interest to the Engle group. For the details of these transactions, see id. at 118-21. The Engle Group and the Reliable Trust bought into Berkeley Bio Medical, Inc. ("Berkeley"), at the request of that company's management to help defend against a takeover attempt by Cooper Laboratories, Inc. ("Cooper"). Cooper eventually bought the Engle group's shares, giving the group a large profit. In the case of Outdoor Sports Industries, Inc. ("OSI"), the Engle group was the raider. Eventually, a "white knight" came to OSI's rescue, and the Engle group and the trust sold their shares, garnering a profit in excess of one hundred percent in approximately a one-year period. The group and Reliable Trust also bought stock in Hickory Furniture Company ("Hickory"). Engle gained control of the company, and the trust sold its shares after a year at a four percent profit.

The Reliable Trust's return on these three investments was a whopping 72 percent over a relatively brief time span. Nonetheless, plaintiffs brought this action, claiming the administrators breached their duty of loyalty to the trust under the Employee Retirement Income Security Act of 1974, 29 U.S.C. Secs. 1001-1461 ("ERISA"), and seeking the Engle group's entire profits from its investments in the three stocks, estimated at up to ten million dollars.

The trial court initially found for the defendants. We reversed in Leigh I, holding that the administrators had breached their fiduciary duties under ERISA sections 404 and 406, 29 U.S.C. Secs. 1104, 1106, and that plaintiffs might be entitled to some damages, albeit not the amount they requested. We remanded the case for a determination whether Engle and Libco were liable as fiduciaries with respect to the investments; whether the administrators and the bank unduly delayed distribution of trust assets; whether plaintiffs suffered any damages; and for a determination of fees. Id. at 140-41.

The district court held that Engle and Libco breached their fiduciary duties by failing to adequately supervise the administrators selected by them. The court found no undue delay in asset distribution. It awarded plaintiffs $6,704 in damages, the difference between the return on Hickory stock and that from a prudent alternative investment. Finally, the court allowed plaintiffs fees incurred through the time of Leigh I, and allowed some defendants partial fee reimbursement from the trust.3 Both sides appeal.

In a probably futile attempt to clarify the analysis, we will divide our discussion into three categories. Initially, we will review the district court's findings on liability to determine whether they are factually or legally erroneous. Then we will examine the court's calculation of damages. Finally, we will look at the fees questions.

The district court made two findings on liability; the losing parties appeal. We will reverse factual findings only if they are clearly erroneous. See Fed.R.Civ.P. 52(a); Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985). "Where there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous." Anderson, 470 U.S. at 574, 105 S.Ct. at 1511. We review legal determinations de novo.

In Leigh I, we ordered the district court to determine whether Engle and Libco acted reasonably and prudently in light of their knowledge of the administrators' conflicting interests and the trust's investments. 727 F.2d at 136. We held that Engle and Libco were fiduciaries of the trust with respect to selection and retention of the Plan's administrators, id. at 134, and stated that they had a duty "to take prudent and reasonable action to determine whether the administrators were fulfilling their fiduciary obligations." Id. at 135.

Judge Duff held on remand that "Engle and Libco did not take reasonable action to ensure that Dardick and Zuckerman were fulfilling their fiduciary duties." Leigh II, 669 F. Supp. at 1395. His key factual finding was that Engle and Libco "knew of, but chose to ignore" the administrators' improper investment decisions. Id. That finding, in conjunction with our opinion in Leigh I, made the ultimate finding of liability "almost inevitable." Id. at 1417.

Engle and Libco appeal. They do not attack the finding that they knew of the investments. Instead, they contend that even assuming total knowledge of the administrators' actions, they were under no duty to respond. This attack takes two approaches. The first is a thinly-veiled assault on Leigh I 's central holding. Engle and Libco argue that the investments were not speculative and that "[t]here was no actual conflict of interest." Reply Brief of Engle and Libco at 5. The latter assertion directly contradicts our holding in Leigh I that the administrators breached their duty of loyalty to the trust. 727 F.2d at 132. We will not reopen that can of worms. Also, whether the investments were speculative is irrelevant. The administrators' breach did not consist of investment in speculative assets. Rather, the administrators breached their duties when they made investment decisions out of personal motivations, without making adequate provision that the trust's best interests would be served.4 That breach, known to Engle and Libco, created a duty on their part to take action to rectify the situation. They did not do so, thereby breaching their supervisory duties.

Engle and Libco's more substantial argument is that Leigh I should not apply retroactively. They say that Leigh I created a new basis for liability from whole cloth, and that it would be "punitive" to hold them liable when they could not have recognized that the administrators were violating the statute. They point us to City of Los Angeles, Dept. of Water and Power v. Manhart, 435 U.S. 702, 98 S. Ct. 1370, 55 L. Ed. 2d 657 (1978), where the Court refused to award retroactive relief in a Title VII action brought against plan administrators.

Manhart is inapposite. First of all, it is a Title VII case, construing that statute's unique retroactive relief provisions. Id. at 718, 98 S.Ct. at 1380. Second, the Manhart Court's reasoning cuts against Engle and Libco. The Manhart Court was concerned that "major unforeseen contingencies" could wreak havoc on the nation's pension plans, destroying the entire system. Id. at 721, 98 S.Ct. at 1382. No such concerns are present here. Leigh I did not announce any fundamental change in the rules governing ERISA plan administration. At bottom, the case simply held that administrators violate their duty of loyalty when they use plan assets to dabble in their business associates' takeover attempts, unless there have been special efforts to identify and preserve the trust's best interests. Leigh I, 727 F.2d at 132. We fail to see anything novel or unpredictable in this holding. While ERISA is a fairly modern creature, the law of fiduciaries, codified in the Act, is of venerable vintage. Leigh I was an application of principles at the heart of corporate and trust law, principles that should be common knowledge to every fiduciary. Cf. Fulton Nat'l Bank v. Tate, 363 F.2d 562, 570-71 (5th Cir.1966); Restatement (Second) of Trusts Sec. 170 (1959) ("the trustee is under a duty ... to administer the trust solely in the interest of the beneficiary").

Engle and Libco briefly raise other points, all assaulting the notion that they could ever he held liable for this sort of conduct. Those arguments must fail in light of Leigh I.

The district court's second finding on liability was that "the defendants did not breach their fiduciary duties by delaying the distribution of trust assets." Leigh II, 669 F. Supp. at 1412. Plaintiffs appeal this determination.

Judge Duff's holding on this point consists of two parts. He first found that "there was neither a complete discontinuance of contributions nor a complete termination of the plan prior to February 28, 1979." Id. Appellants concede this point. Opening Brief of Intervening Plaintiffs-Appellants at 42. They thus must dispute the court's second finding that, given the 1979 termination date, initial distributions were not due until March 1, 1980. Id.

Plaintiffs' first argument is that Judge Duff ignored this court's mandate in Leigh I. In that opinion, we remanded so that the district court could reconsider the delay issue. 727 F.2d at 136-37. The district court's exhaustive factual findings and its common-sense construction of the relevant documents and regulations compel affirmance. The reasons for any delay become relevant only when distribution is delayed beyond the time mandated by the trust instruments.

Appellants act as though this approach is somehow dishonest. We are hardly inclined to discipline a trial judge for finding the crux of an issue and beginning at that point. Contrary to appellants' overzealous assertions, we did not require the district court to find liability or to follow any one theory in determining liability. In fact, the mode of analysis adopted by the district court was suggested in our opinion, when we urged the court to consider the trust documents and relevant regulations to determine at what point distribution was required. See id. at 136.

As to the factual finding that distributions were not due prior to March 1, 1980, we are persuaded by Judge Duff's reasoning. Section 7.2 of the Restated Plan says that "[u]pon termination of the Trust, the Committee shall direct the Trustee to distribute all assets remaining in the Trust...." Intervening Plaintiffs say that use of the word "upon" requires immediate distribution.

The district court properly rejected that rigid interpretation, for several reasons. Nothing in section 7.2 purports to establish the time or manner of distribution. Use of the word "upon" provides little or no guidance. Therefore, Judge Duff properly looked elsewhere for interpretive assistance. As directed by Leigh I, 727 F.2d at 136, he looked to the governing law, 26 U.S.C. Sec. 401(a)(14), which, although not applying directly to this type of termination, gives some guidance. That section requires that payments begin no later than sixty days after the close of the termination year. Likewise, in construing the word "upon" in other sections of the Restated Plan, section 5.6 adopts the same rule. We believe Judge Duff correctly looked to closely analogous termination provisions to determine when payments must begin under section 7.2.

Appellants contend that "upon" means something different in section 7.2 than elsewhere in the Restated Plan. But there is not a scintilla of evidence, either in the trust documents or elsewhere in the record, to support that assertion. We will not reject the district court's analysis in favor of mere speculation.

Likewise, the court's alternative holding independently supports its result. Judge Duff held that "[e]ven if defendants had delayed distribution of trust assets, the damages assessed in ... this court's decision would compensate plaintiffs for any loss they suffered by reason of the delay." 669 F. Supp. at 1410. That finding is clearly correct. The only possible loss from delay is the opportunity cost of not being able to invest the money elsewhere. The trust assets greatly appreciated in value during the period of the alleged delay. Appellants do not challenge this fact, which is dispositive on the damages question. We affirm the district court's finding on this point.

The parties raise two questions as to the district court's calculation of damages. The court rejected plaintiffs' contention that they were entitled to all profits on all Engle group investments in OSI, Berkeley and Hickory, holding that defendants showed that their profits were not obtained through use of the trust assets. Judge Duff did, however, award plaintiffs $6,704 in damages from the trust's investment in Hickory, since that stock did not perform as well as more appropriate investment vehicles. We are not surprised that both sides appeal. In Leigh I, we called this damage calculation a "formidable task." The district court did an admirable job with it.

Plaintiffs ask for millions of dollars in damages, claiming that all or most of the Engle group's profits from its investments in the three companies were attributable to use of trust assets. We thought we had laid this argument to rest in Leigh I, where we examined the facts and the relevant statutes and said that "[t]he plaintiffs' argument reaches too far." 727 F.2d at 137.

Our holding in Leigh I was that 29 U.S.C. section 1109 only allows recovery "where there is a causal connection between the use of the plan's assets and the profits made by fiduciaries on the investment of their own assets." 727 F.2d at 137. We also held that "the trustee has the burden of showing which property and profits are his." Id. at 138. We tried to make clear to the plaintiffs how unlikely they were to obtain substantial damages, given the investments' excellent returns and the miniscule percentage of the Engle group's total investment made up of trust assets. See id. at 137 n. 35 (trust never held even one percent of target companies' stock). Appellants characterize the Engle group's profits as a "windfall," Opening Brief of Appellants at 5, but, if that windfall resulted from defendants' own efforts, it in no way belongs to the trust. See Leigh I, 727 F.2d at 138. Plaintiffs once again accuse Judge Duff of ignoring this court's mandate (an accusation they make far too often and with no apparent thought as to whether it is appropriate, see Opening Brief of Appellants at 22 n. 5), but that "mandate" included a caveat that the judge on remand might very well find no damages appropriate. Leigh I, 727 F.2d at 138. In order to avoid disgorgement of part of its profits, the Engle group had to show that they did not result from misuse of trust assets. Judge Duff correctly held that defendants met this burden.

There are at least five theories that might explain how use of trust assets increased the Engle group's profits on its own investments. First, perhaps the Engle group effectively "parked" stock by buying it with trust assets. That is, the group might have delayed disclosure of its investment in a target, allowing it to purchase more shares at a lower, pre-disclosure, price, by not including the trust's shares in determining when the group's investments met the threshold for disclosure mandated by federal securities laws. See Securities Exchange Act Sec. 13(d), 15 U.S.C. Sec. 78m(d). That contention will not fly here, since the Engle group included the trust's holdings in its Schedule 13D filings. Likewise, any assertion that the Engle group would not have been able to purchase the stock with its own money lacks support in the record. Engle's reputation as an effective financier was based in part on his ability to quickly raise capital.

Third, the increase in the Engle group's holdings due to the trust's investments might have convinced the management of OSI to seek a "white knight," motivated Berkeley's buyer to pay a premium for the Engle group's shares or allowed the group to gain control of Hickory. Judge Duff found these contentions implausible. Those findings are not clearly erroneous. The trust's holdings were a tiny fraction of the Engle group's holdings. The Engle group's efforts succeeded because of its large stake, its ability to raise capital for a larger investment and Engle's reputation as an effective takeover artist. If anything, the trust benefited from its association with Engle, not vice versa.

Fourth, the administrators did not, in the case of Berkeley, have the potential to undermine a settlement between the Engle group and Cooper by refusing to tender the trust's shares. Cooper Laboratories would not have refused to purchase the Engle group's large stake but for the inclusion of the few shares held by the Reliable Trust. There is substantial evidence in the record to support Judge Duff's finding on this point, most notably the testimony of expert witness Daniel Fischel. Plaintiffs offered no persuasive evidence to the contrary, referring primarily to an affidavit of Joseph Dornig, a Cooper officer, never accepted into evidence. Plaintiffs rely on a single inconclusive statement taken out of context. They make no detailed argument as to why the affidavit should have been admitted. Even if it had been, the result would not have changed.

Finally, Judge Duff's finding that trust purchases did not substantially affect the market price of stock is not clearly erroneous. Plaintiffs' arguments on this point consist mostly of attempts to denigrate the testimony of Daniel Fischel, an expert called by the defendants. We have reviewed Professor Fischel's testimony and agree with Judge Duff's characterization of it as "highly credible and persuasive." Leigh II, 669 F. Supp. at 1401. Professor Fischel exhaustively analyzed the economic facts of the case. Essentially the district court believed Fischel and disbelieved plaintiff's experts. We cannot call findings based on such thorough analysis clearly erroneous.

Once the district court determined that the Engle group's profits were not tainted by the trust's investments, it had to determine whether the Trust suffered any losses attributable to the distorted investment decisions that may have resulted from the trustees' conflicts of interest. Judge Duff undertook the straightforward approach of comparing the return on the improper investments with that of a reasonably prudent alternative investment--in this instance the Harris Bank's common stock funds. In Leigh I we placed the burden of disproving damages on the defendants. 727 F.2d at 138. Therefore, the court adopted the "most generous" of the reasonable damage calculations submitted to it. Leigh II, 669 F. Supp. at 1405. Looking at each of the three trust investments in isolation, it found that the trust suffered no losses from the OSI and Berkeley investments. It did, however, find a loss of $6,704 on the Hickory investment. While that investment made a small profit, the gain was less than that from the alternative investment used as a standard by the court.

Defendants appeal this award. They contend, not without foundation, that the court erred in not looking at the value of the entire portfolio in determining whether the trust suffered any loss from the investments. Once again a party alleges that Judge Duff misconstrued or ignored this court's mandate; once again we do not agree. We did say in Leigh I that "[i]t is clear that the trust lost no money in the challenged transactions." 727 F.2d at 121-22. Judge Duff did not find differently. He merely found that the gain from the Hickory investment was less than that which would have been obtained through prudent alternative investments.

Turning to the theoretical challenge, we find that defendants' arguments contain a kernel of truth. When investment advisors make decisions, they do not view individual investments in isolation. Rather, the goal is to create a diversified portfolio that balances appropriate levels of risk and return for the investor. The risk of a given investment is neutralized somewhat when the investment is combined with others in a diversified portfolio. The risk inherent in the entire portfolio is less than that of certain assets within that portfolio. Ideally, after diversification only market risk remains. Likewise, the return from a portfolio over time should be more stable than that of isolated investments within that portfolio. (This discussion is greatly simplified; for a somewhat more technical explanation, see R. Brealey & S. Myers, Principles of Corporate Finance 119-32 (2d ed. 1984).)

Given the facts that investment advisors generally follow a portfolio strategy of investment and that beneficiaries whose assets are being managed are concerned with the end result of that strategy, not with the return on a single element in the portfolio, it makes sense for courts to look at the whole portfolio to determine the investment strategy's success. Cf. Donovan v. Bierwirth, 754 F.2d 1049, 1057 (2d Cir.1985); Langbein & Posner, The Revolution in Trust Investment Law, 63 A.B.A.J. 887, 889-90 (1976); Note, Fiduciary Standards and the Prudent Man Rule Under the Employment Retirement Income Security Act of 1974, 88 Harv.L.Rev. 960, 967 (1975).

Judge Duff acknowledged this theory's force as a general principle, but determined that it should not apply in this case. Although the question is a close one, we do not believe he abused his discretion. Portfolio theory gains its force from the fact that it reflects investment decisions in the real world. But that assumption does not hold in this case. As we held in Leigh I, defendants did not make their investment decisions with the sole goal of creating a diversified, safe portfolio for the trust beneficiaries. They argue that their decisions had that effect and that the trust made a great deal of money; but the result does not change the fact that the trustees' purposes were not those of portfolio investors. The administrators looked at the stocks in isolation; so a court is justified in taking the same view when calculating the loss from those investments. Our holding may be reduced to this: where fiduciaries breach their duty of loyalty by making individual investments with an eye toward some goal other than the creation of a proper portfolio for their clients, a court may return the favor, viewing the investments in isolation to determine damages.

One might argue that this rule results in overdeterrence, that trustees will be deterred from making potentially profitable and prudent investment decisions by the risk of future damages. That might be the case if the rule were applied to marginal conduct, e.g., to an investment strategy that borders on an unacceptable risk level but is close to the range of reasonableness. That is not this case. Here the trustees did not err slightly in their attempt to create a proper portfolio; they simply made no such attempt, at least with respect to the decisions to purchase stock in OSI, Berkeley and Hickory. There is no way to overdeter such conduct. It should never occur, no matter how profitable the end result.

In short, the appeal and cross-appeal on damages lack merit. Plaintiffs ask for the moon, a possibility we firmly rejected in Leigh I. Defendants, on the other hand, seek to avoid even a modest penalty for their misconduct. The district court struck a defensible middle approach, one much more in tune with the case's realities than the arguments of any party. Cf. Patton v. Mid-Continent Sys., Inc., 841 F.2d 742, 748 (7th Cir.1988).

Finally, we must visit the issue of fees. Judge Duff held first that the trust instruments allowed for reimbursement of fees incurred by Dardick, Zuckerman and National Boulevard Bank, for defense of claims on which plaintiffs did not prevail. For their efforts in prosecuting the sole issue on which they did prevail he awarded plaintiffs fees. Because determination of exact fees incurred in litigating specific issues is well-nigh impossible in this case, and because the parties have evidenced an uncanny proclivity for making every molehill into a mountain, he made the solomonic decision to award plaintiffs all fees incurred through Leigh I, but none thereafter. Extended fee litigation would serve no purpose except to further deplete the trust, and the issues on which plaintiffs prevailed were substantially resolved by our first decision. Moreover, "[a]warding attorneys' fee to plaintiffs for this most recent portion of the case would reward litigation that was ill-conceived, often poorly executed, and fractious." Leigh II, 669 F. Supp. at 1417. Both sides appeal.

Plaintiffs ask us to reverse the decision allowing reimbursement of fees to Dardick, Zuckerman and National Boulevard Bank, incurred in fighting claims that ultimately failed. We cannot say that the trial court abused its discretion.

The trust instruments allow for reimbursement of fees, and allow a reserve to be set up to ensure payment. The only real question is whether the trust provisions conflict with ERISA. They do not. While an award of fees to a losing defendant certainly would contravene Congress' intent, see 29 U.S.C. Sec. 1110(a), plaintiffs point us to no statutory or common-law basis for denying fees to a prevailing trustee where the trust documents specifically contemplate such reimbursement.

Plaintiffs argue that this result unjustly deprives the beneficiaries of their funds. We sympathize, to the extent that plaintiffs think it is disgraceful for pensioners to be deprived of their benefits because of this seemingly endless litigation. To the extent they try to place the blame for this result entirely on the defendants, however, plaintiffs are mistaken. Defendants prevailed as to most of the claims. The record makes painfully clear that plaintiffs (or their attorneys) needlessly prolonged this litigation through quixotic, short-sighted and hardball tactics that were apparently undertaken with little thought as to the effect on the ultimate victims--the beneficiaries. Defendants expended great sums of money defending meritless claims. Very few people would become plan administrators if subjected to such unjust, extensive potential costs. Reimbursement here conformed with the trust documents and ERISA. We therefore affirm that aspect of the district court's holding.

Judge Duff awarded plaintiffs fees incurred through Leigh I. We give great deference to this determination, committed by statute to the trial court's discretion. 29 U.S.C. Sec. 1132(g)(1). Both sides appeal. Plaintiffs want more money; defendants say plaintiffs should get nothing. In Leigh I we set out the legal standard to be applied on remand. See Leigh I, 727 F.2d at 139 n. 39. The court below followed that analysis, considering five factors:

(1) The degree of the opposing parties' culpability or bad faith; (2) the ability of the opposing parties to satisfy an award of fees; (3) whether an award of fees against the opposing parties would deter others from acting under similar circumstances; (4) whether the parties requesting fees sought to benefit all participants and beneficiaries of an ERISA plan or to resolve a significant legal question regarding ERISA; and (5) the relative merits of the parties' positions.

Id. (quoting Marquardt v. North Am. Car Corp., 652 F.2d 715, 717 (7th Cir.1981)). The district court found:

(1) Dardick, Zuckerman, Libco, and Engle were grossly negligent in investing a substantial portion of the trust's assets in three speculative stocks. (2) The four breaching fiduciaries have the resources to pay a fee award and are better able than plaintiffs to bear this cost. (3) An award of attorneys' fees to prevailing plaintiffs will tend to deter similar fiduciary misconduct in the future. (4) Plaintiffs (or at least intervening plaintiffs) brought this lawsuit on behalf of all plan participants, not merely for individual gain. (5) The liability of the four breaching fiduciaries is not a close question; even minimal reflection should have led them to realize it was unlawful to invest 30 percent of the trust's assets in three speculative stocks.

Leigh II, 669 F. Supp. at 1416.

Defendants challenge this characterization of the case, but we cannot say it is so wild as to be an abuse of discretion. In particular, the deterrent effect of a fee award looms large on these facts. As matters turned out, the trust suffered little damage. But that does not make defendants' misconduct less censurable. Defendants' briefs in this court exhibit a disturbing inclination to downplay the gravity of their offense because it made money. We agree with the court below: "Perhaps the defendants are too young, too wealthy, and too comfortable to have contemplated the enormity of the risk they took, and the human suffering so narrowly averted." Id. at 1417. A fee award is one small way to impress upon them the gravity of their conduct, and to deter such action in the future.

Plaintiffs challenge the amount of the award. Again, the abuse of discretion standard applies. Bright v. Land O'Lakes, Inc., 844 F.2d 436, 442 (7th Cir.1988). In this context, "an abuse of discretion occurs only when no reasonable person could take the view adopted by the trial court. If reasonable persons could differ, no abuse of discretion can be found." Harrington v. De Vito, 656 F.2d 264, 269 (7th Cir.1981), cert. denied, 455 U.S. 993, 102 S. Ct. 1621, 71 L. Ed. 2d 854 (1982).

The court below allowed plaintiffs' fees incurred prior to Leigh I. Although the judge found that they prevailed only as to one issue, he felt that attempting to coerce these parties into dividing fee requests by issue would be costly, time-consuming and probably not fruitful.

The district court's approach is far from unreasonable. Judge Duff correctly asserts that Leigh I "established the only significant legal principles to arise from this lawsuit, and also resolved the principal factual issues on which plaintiffs prevailed." Leigh II, 669 F. Supp. at 1416. Our first opinion made clear that while plaintiffs proved a breach of duty, they were probably not greatly injured by that breach. To continue in hope of a jackpot was mere folly, and it should not be rewarded. The district court's approach ensures that the trust will not be further depleted by years of fee litigation. We affirm.5

Judge Duff best characterized this case.

Perhaps only Charles Dickens could savor this litigation. For nearly a decade now, the parties have fought bitterly over the Reliable trust, the only noticeable effect being the steady diminution of its assets. The advocacy has been harsh and often vituperative: lawyers have accused each other of personal wrongdoing, discovery disputes continued through the last day of trial, and shouting matches have broken out. If it were possible to bottle the contempt, even hatred, which the lawyers and parties feel for one another, there would be enough to sustain a small civil war for months. The great length and extraordinary difficulty of this litigation owes much to the depth of the combatants' animosity.

Leigh II, 669 F. Supp. at 1417. This is a case that existed, through much of its ten-year life, almost solely for the benefit of the lawyers. That is indeed unfortunate, for the costs are borne most heavily by those whom ERISA was intended to protect--beneficiaries and their families. We wish the trustees had considered the consequences of their actions before breaching their duty of loyalty to the beneficiaries. We also wish, however, that plaintiffs' counsel had paused for a moment to realistically assess, at each stage of the game, the impact of extensive litigation on their clients. Had they done so, everyone involved would have been better off.

AFFIRMED.