We remanded this multidistrict price-fixing class action for more detailed findings of fact and conclusions of law concerning various aspects of settlements approved by the district court. 643 F.2d 195, 225-26 (5th Cir. 1981). At the same time, we noted that "(n)one of the (objectors') arguments persuades us that the settlements must be set aside or modified." Id. at 202. The district court made additional findings and conclusions and once again approved the proposed settlements.1 This reapproval is now before us for further evaluation in light of our mandates in the prior remand and the district court's findings and conclusions. Because our mandates have been followed and proper findings support the settlement agreements and the formula for allocating certain of the settlement proceeds among class members as fair, reasonable, and adequate, we find the reapproval to be well within the discretion of the district court and affirm. Since Texas Industries, Inc. v. Radcliff Materials, Inc., --- U.S. ----, 101 S. Ct. 2061, 68 L. Ed. 2d 500 (1981), has established that there is no right of contribution under the antitrust laws, the distribution of settlement funds may proceed.

In remanding this action to the district court, we sought an enlarged factual and legal base upon which to evaluate the settlement agreements and allocation formula. The particular requests we made of the district court for information on these distinct issues are summarized below.

We enunciated a three-step process for passing on the fairness, reasonableness, and adequacy of the settlement agreements. That process required the district court to evaluate the plaintiffs' likelihood of prevailing at trial and the range of possible recovery at such a trial, to consider other relevant factors, and to determine the point on or below the range of possible recovery at which the settlement was fair, reasonable, and adequate. In re Corrugated Container Antitrust Litigation, 643 F.2d 195, 212. In calling for additional information on the range of recovery and likelihood of prevailing at trial, we eschewed any requirement of a high degree of specificity. Id. Nonetheless, we found the district court's initial findings inadequate in the following respects.

In crediting Dr. Richard C. Hoyt's expert testimony and establishing $200 to $800 million as the range of possible recovery, the district court failed to compile adequate findings to rebut two of the objectors' criticisms. The objectors argued that Dr. Hoyt's computation wrongly failed to take into account the period preceding the four-year statute of limitations and was the product of defective methodology. The district court was asked to explain why it only considered damages within the statute of limitations period in computing the range of possible recovery. Id. at 214. In explaining the district court was requested to evaluate objectors' argument that defendants' fraudulent concealment tolled the statute of limitations and enlarged the temporal, and thus, the monetary range of possible recovery. Id. In addition, the district court was instructed to consider the difficulties of proving a price-fixing conspiracy and assessing damages for the longer period. Id. We also called upon the district court to justify its reliance on Dr. Hoyt's estimates, rebut the objectors' criticisms of Dr. Hoyt's conclusions, and explain why Dr. Hoyt's conclusions are more accurate than those of the objectors' experts. Id. at 215. We then turned to the district court's discussion of the likelihood that the class would prevail on the merits and suggested that the district court more completely explain its conclusions. Id. at 217.

Based on the established range of possible recovery and the plaintiffs' likelihood of success, the court was to exercise its discretion and determine the fairness, reasonableness, and adequacy of the settlements. We also instructed the district court to explain and evaluate why it attached importance to factors such as the length, complexity, and expense of an actual trial and the number and importance of parties opting out of the class or objecting to the settlements. Id. at 217-18. Furthermore, the district court was directed to make particularized findings as to why each individual settlement was reasonable, or at least to explain why it was unnecessary to do so. Id. at 218.

In order to allocate the proceeds of settlements reached prior to January 5, 1979, the agreements provide that class members file claims based on their purchases between 1960 and 1978 and that the proceeds be distributed to class members in proportion to those claims. The objectors point out that this formula treats pre-1973 purchases on a par with post-1972 purchases, even though the district court expressed the opinion that claims on pre-1973 purchases might have been barred by the statute of limitations. We called upon the district court to articulate persuasive justifications for the allocation formula. Id. at 220.

Our discussion of the district court's additional findings on the adequacy of the settlements and the reasonableness of the allocation formula is best prefaced by the caveat that the essence of a settlement is compromise. A just result is often no more than an arbitrary point between competing notions of reasonableness. As we stated in our earlier remand, "our appellate function ... is a limited one, especially in light of the strong judicial policy favoring settlement of disputes." Id. at 207. We are limited in reviewing the district court's approval of this reconciliation and cannot upset that approval unless the district court clearly abused its discretion. Id. See also Cotton v. Hinton, 559 F.2d 1326, 1331 (5th Cir. 1977); Young v. Katz, 447 F.2d 431, 432 (5th Cir. 1971).

The district court, in reaching its determination that the settlement agreements, collectively and individually, were fair, reasonable, and adequate, made added findings on or evaluations of each issue and factor we requested. We will discuss, in order, the district court's additional findings with respect to the range of possible recovery, the likelihood of prevailing on the merits, and other factors which, taken together with the first two, affect the judge's exercise of discretion in passing on the fairness of a settlement agreement.

The district court's additional findings comply with our request that it confront the two central criticisms of its estimated range of possible recovery: that it (i) excludes prelimitations period damages and (ii) relies on the study of Dr. Hoyt which used an unsound method of estimating damages from excess profits. 643 F.2d at 212-15.

The district court set forth detailed findings to support its conclusion that the adequacy of the settlements should be evaluated on the basis of the range of damages potentially recoverable during the statute of limitations period since the chances of recovery for prelimitations period damages were comparatively slight. The court pointed out that in urging that defendants' fraudulent concealment tolled the statute of limitations, objectors were required to produce evidence of acts by the defendants designed to conceal the alleged conspiracy and to show that the plaintiffs lacked knowledge of the conspiracy despite the exercise of due diligence.2 United States v. Container Corporation of America, 393 U.S. 333, 89 S. Ct. 510, 21 L. Ed. 2d 526 (1969), which involved twelve of the defendants in this action, and a subsequent decree prohibiting price communications by sellers of corrugated containers, rendered proof of fraudulent concealment all the more difficult. Likewise, proving due diligence class-wide on a claimant-by-claimant basis would be particularly difficult. The district court noted that in addition to the difficulty with proof of fraudulent concealment, the plaintiffs would have trouble proving a price-fixing conspiracy and damages during the pre-1973 period due to fading memories, missing or dead witnesses, lost documents and sales records, and the unavailability of reliable economic data. The district court's findings and conclusions support its decision to discount the prelimitations period damages in its computation of the range of possible recovery.

The district court also explained its reliance on Dr. Hoyt's economic estimate of the range of possible recovery. Dr. Hoyt was eminently qualified to conduct a damages estimate because of his training, experience, knowledge of the industry, and involvement in the case. The district court concluded that his analysis was logical, persuasive, and consistent with the evidence then known to the court. This conclusion is supported by the proof. In addition, Dr. Hoyt testified in person and was subjected to cross-examination which best permitted the district court to pass on his credibility. The district court detailed its reasons for rejecting the objectors' criticisms of Dr. Hoyt's "profits" model of damages and for supporting Dr. Hoyt's methodology and assumptions. For example, it found the Hoyt study produced results consistent with the studies of other experts when those other studies were limited to the period most closely aligned with proof of the conspiracy in the related criminal cases. It found that Hoyt's profit analysis was based on data that was more reliable, or as reliable as that used by others. It found that external factors did not have any different effects on prices, costs, or profits. And it found that Hoyt's study was more thorough than others.

The district court also explained its refusal to credit the findings of the objectors' experts. First, the court noted that these experts did not testify, thus, denying the court an opportunity to observe them and judge their credibility. Second, the objectors' experts lacked Dr. Hoyt's familiarity with the industry. Third, the model of one of the experts, upon whom the others relied, assumed that the conspiracy continued into 1978 and utilized an "after" comparison period that was characterized by atypical prices and abbreviated length. Having found the methodology and assumptions of the objectors' experts flawed, the district court corrected those flaws and found that the damage estimate of those experts was not inconsistent with Dr. Hoyt's damage estimate.

Our role in reviewing the district court's reliance on Dr. Hoyt's expert testimony is limited. "The competency and qualifications required of expert witnesses is a matter committed to the broad discretion of the trial judge .... When it sits without a jury, the trial court's discretion is also the sole determinant of the weight to be given expert testimony." Ludlow Corp. v. Textile Rubber & Chemical Co., 636 F.2d 1057, 1060 (5th Cir. 1981). "(A) ppellate courts must rely heavily upon the findings and credibility choices made by trial courts. Otherwise, we would be thrust into the position of reconsidering credibility and re-resolving conflicts in expert testimony, a position that mistakes this Court's function." Ziegler v. Phillips Petroleum Co., 483 F.2d 858, 861 (5th Cir.), cert. denied, 414 U.S. 1079, 94 S. Ct. 597, 38 L. Ed. 2d 485 (1973). The district court has satisfactorily justified its reliance on Dr. Hoyt's expert study in formulating a range of possible recovery. Our review ends there.

The district court entered detailed findings and conclusions on the ability of the plaintiffs to prevail on the merits at trial. The additional findings were gleaned from materials presented to the court, the discovery proceedings, the presentence proceedings of the criminal defendants who pled nolo contendere, and the trial of the other criminal defendants. The presentence reports on the defendants who pled nolo contendere contained the government's assessment of the relative culpability of each defendant. They were placed under seal by the district court. The district court's particularized assessment of the chances that the class would prevail against each defendant was also sealed. Since the findings and conclusions publicly articulated by the district court satisfy our limited remand instructions and provide the basis for affirmance independent of the sealed material and assessment, the sealed matters do not affect the substantial rights of the parties. Fed.R.Civ.P. 61. The objectors' assertion that this procedure compromises their due process rights fails to present a reversible error.

a. Proof of Violation. While there is strong evidence that some employees and some companies violated the Sherman Act, the defendants dispute this and the jury acquitted all of those tried for violating the criminal provisions of that Act. There is little or no evidence regarding a conspiracy in some parts of the country. Furthermore, there are several characteristics of the industry that would render a nation-wide price-fixing conspiracy unlikely or even impossible: the industry is highly fragmented, the highest market share of any one company being 6.3%; a great deal of competition exists among sellers; and new companies enter the business with frequency and ease. In connection with the plaintiffs' ability to prove their case, the district court noted two additional factors. First, the plaintiff class would have significant problems obtaining evidence, particularly in light of the fact that many industry employees would invoke their fifth amendment privileges. Second, the defense counsel in the case are experienced and outstanding antitrust litigators, while plaintiffs' counsel, though competent, are relatively inexperienced. The district court concluded that the defendants raised issues of fact that "would allow a jury to find against plaintiffs."

b. Proof of Impact. The district court found that it would be a "close question" as to whether the alleged violations had a class-wide impact because of the evidence of substantial local control over pricing.

c. Proof of Damages. The district court noted that proof of damages would present "serious problems" since there was some evidence that the real price of corrugated materials declined during the alleged conspiracy and that the corrugated industry is characterized by low profits.

d. Summary. The district court judge, in addition to his discussion of the general difficulties the plaintiffs would face in proving their case, assessed the plaintiffs' likelihood of prevailing against each individual defendant. We reject the objectors' argument that those assessments are too amorphous. A district court need not establish the plaintiffs' likelihood of prevailing to a certainty. 643 F.2d at 212. The findings of the district court are clear enough to support its exercise of discretion to approve the settlements.

In our remand, we instructed the district court to take into account the range of possible recovery and the plaintiffs' likelihood of prevailing at trial before exercising its discretion in passing on the fairness, reasonableness, and adequacy of the settlements. We also suggested that the court articulate why it attached significance to a number of factors. Finally, we asked the district court to pass on the reasonableness of each individual settlement or, in the alternative, to explain why it was proper to judge them collectively.

The district court satisfactorily explained its reliance on various other factors. It found that, in general, class members support the settlements. Of the more than 200,000 class members, only 210 opted out. The parties objecting to the settlements are both qualitatively and quantitatively insignificant: they do not include a major corporation and several appear to be commonly owned. The district court also ascribed importance to the fact that the settlements would permit the avoidance of an expensive, lengthy, and complex trial. The court noted that the trial against the settling defendants would last for many months, perhaps years. The complexity of an antitrust jury trial of thirty-seven defendants gives rise to the very real possibility that the plaintiffs' case could be prejudiced. As for the expenses, they are to be measured not only in the court costs and legal fees, but also in the loss of time and services of executives and employees involved in preparing the case for trial or testifying at trial. The settlements also avoided additional costly discovery.

The district court cited four other factors that indicated the settlements should be approved. First, the court noted that the immediate payment of the settlements by the defendants vitiated any advantage to them of avoiding litigation and significantly benefitted the class since those payments are bearing interest. Second, the settlements reduced the complexity of and enhanced the class' case against the non-settling defendants. Third, the settlement agreements contained cooperation clauses that proved to be helpful to the class in estimating damages, discovering evidence, and proceeding against the non-settling defendants. Fourth, the settlements specifically reserved the plaintiffs' right to recover from the non-settling defendants the entire amount of damages caused by the conspiracy, less the amounts recovered in the settlements.3

Having weighed the plaintiffs' range of possible recovery and likelihood of prevailing at trial, and taken the additional factors noted above into consideration, the district court concluded that the settlements evaluated collectively are fair, reasonable, and adequate. The district court also explained why it was proper for it to evaluate the settlements collectively. Nonetheless, the district court undertook individual evaluations of the settlements and likewise concluded that each was fair, reasonable, and adequate.

In its detailed analysis of the individual settlements, the district court, in addition to the factors discussed above, considered: the amount of each settlement in total dollars and dollars per percentage point of the defendant's market share; when the settlement was negotiated; whether the defendant or its employees had been indicted or named a co-conspirator; whether indictments were for misdemeanors or felonies and how those indicted pleaded; whether the case against the defendant was strong; how the settlement fit into the plaintiffs' settlement strategy; the defendant's ability to pay; whether the settlement was reached before class certification or at the point when a mandamus action sought decertification; how the settlement dealt with the so-called Illinois Brick legislation4 that would have reduced plaintiffs' recoverable damages; whether the expense of litigating against a minor defendant was justified by the potential recovery; and what non-monetary benefits the settlement provided.

We affirm the district court's findings and conclusions with respect to the fairness, reasonableness, and adequacy of the settlement agreements.5

In light of the district court's conclusion that post-1972 claims have greater merit than pre-1973 claims, our limited remand called upon the district court to justify its approval of the allocation formula of the pre-January 5, 1979 settlement agreements which treated pre-1973 and post-1972 claims on a par. 643 F.2d at 218-20. The district court responded by offering several reasons why the distribution provisions are fair and reasonable.

First, the district court noted that some compensation should be provided for the pre-1973 purchases because all of the settlements contain a provision releasing claims for that period. Second, the claims submitted confirm the fact that most class members made purchases during both the pre-1973 and post-1972 periods. Third, the relative value of pre-1973 claims will be reduced because claims based upon undocumented estimates will be discounted and the pre-1973 period will contain a higher percentage of such claims; there will be a smaller volume of claims for the pre-1973 period; and because of inflation, the pre-1973 claims will be of relatively less value. Fourth, devising and implementing a distribution formula weighting the value of pre-1973 and post-1972 claims would be costly, complex, and burdensome.

The district court did not abuse its discretion in approving the plan for allocating the pre-January 5, 1979 settlement proceeds. Indeed, in our remand we intimated that if the district court found that most plaintiffs had claims for the pre-1973 and post-1972 periods and that the costs of executing an alternative distribution formula would be substantial then the share-and-share-alike formula would be appropriate. 643 F.2d at 220. The district court so found and, moreover, supplemented those findings with additional justifications. The plan approved by the district court does not cause an inequity to any one plaintiff. It is approved.

When we first considered the district court's approval of the settlement agreement and allocation formula, the defendants argued that approval of the settlements and distribution of the proceeds should be deferred pending resolution of the rights to contribution of any non-settling defendants. Id. at 224-25. Such a right to contribution would have given rise to claims of a right to rescind by the settling defendants.

These issues have been mooted by the Supreme Court's intervening decision in Texas Industries, Inc. v. Radcliff Materials, Inc., --- U.S. ----, 101 S. Ct. 2061, 68 L. Ed. 2d 500 (1981), which holds that a federal antitrust defendant has no right to contribution from his co-conspirators. Thus, this obstacle to the implementation of the settlement agreements has been effectively removed.

The district court has determined that the settlement agreements are collectively and individually fair, reasonable, and adequate. We have now been able to review that court's exercise of its discretion in light of its additional findings of fact and conclusions of law. The approval of both the settlements and the distribution plan was well within that discretion's bounds. Accordingly, the judgment of the district court is

AFFIRMED.