MEMORANDUM AND ORDER
NEWCOMER, District Judge.
This case is presently before the Court for approval of a proposed settlement pursuant to the requirements of Rule 23(e), Fed.R.Civ.P. In this class action, plaintiffs, who are mortgage holders located in the Eastern District, are suing their mortgagees for payment of interest which they allege is due to them because of the lending institutions’ policy of requiring “mortgage escrow payments.” A hearing was held on the propriety of the proposed settlement on June 16, 1978, and all parties were given opportunity to file written objections with the Court prior to that date. At the hearing, both plaintiffs’ and defendants’ counsel spoke in favor of the proposed settlement, and a number of objectors also addressed the Court. After consideration of these statements, both oral and written, and a review of other settlements and the applicable case law, the Court has decided to accept the settlement and to issue the proposed Final Judgment.
[573]*573Before discussing the merits of the objections and the settlement, a brief history of this case might be in order. In 1972, the first group of named plaintiffs filed this suit in federal court, alleging that the then 177 named defendants were liable to them and others so situated on a variety of theories arising out of the mortgage escrow system. This system involves the alleged practice of the defendants, all of whom are lending institutions in the Eastern District, requiring mortgagors to prepay monthly into escrow accounts with defendants a sum equal to one-twelfth their annual expected property taxes, mortgage insurance premiums, assessments, and water and sewer rentals. Under this practice, the mortgage institutions do not pay the mortgagors any interest on the prepaid amounts, which the mortgagees remain free to commingle and to invest for their own profit. The original complaint set forth five causes of action. In its first opinion, Kinee v. Abraham Lincoln Federal Savings and Loan Association, 365 F.Supp. 975 (E.D.Pa.1973),1 the Court dismissed four of these claims on various grounds. First, the Court dismissed two of the claims asserted under federal statutes, the “Truth-in-Lending” Act, 15 U.S.C. § 1601 et seq., and the Homeowners Loan Act, 12 U.S.C. § 1461 et seq. Second, the Court dismissed the two claims raised under state law, declining to exercise its pendent jurisdiction. The only claim remaining for any trial in this Court, therefore, would be the antitrust claim. Under the Sherman Act, 15 U.S.C. § 1, the plaintiffs allege that the defendants conspired to switch to the mortgage escrow system from the “capitalization” system. The capitalization system allows the mortgagor to reduce the amount of principle owed by the amount of each month’s prepayment. Plaintiffs also claim that the escrow system represented an illegal “tie-in,” whereby the two products of the escrow service and the loan itself were illegally “tied.”
In 1975, in Sommers v. Abraham Lincoln Federal Savings and Loan Association, 66 F.R.D. 581 (E.D.Pa.1975), this Court certified a class under Rule 23(b)(3), with subclasses consisting of the holders of the government-insured mortgages and the non-government-insured mortgages. Discovery was taken in regard to class certification, but presently is by no means complete. In 1976, a few defendants began to negotiate for settlement of this ponderous litigation, and it now appears that a substantial number of the defendants, perhaps all but one, are or will be parties to this proposed settlement.
After this Court dismissed the state law claims, the plaintiffs filed suit in Pennsylvania court. Since settlement of this action will also, by agreement, dispose of the state court action, those claims are also considered here as being still before this Court, even though ruled out earlier. In their complaint, plaintiffs basically allege that the defendants stand in a fiduciary relationship to their mortgagors. They request the Court therefore to find that an express trust has been created, and that the defendants have breached their fiduciary duty in commingling the funds and profiting from them. Also, plaintiffs contend that in absence of an express trust, the Court should impose a constructive trust to avoid defendants’ unjust enrichment.
In evaluating a proposed settlement of a class action, the Court is to determine if the settlement is fair, adequate and reasonable. Dunn v. H. K. Porter, 78 F.R.D. 41 (E.D.Pa.1978). The burden is upon the parties to convince the Court that the settlement fulfills these requirements. In a recent decision, the Court of Appeals for the Third Circuit enunciated some factors which the district court might use in reaching this decision. The most important of these is the plaintiffs’ probability of success on the merits, Kusner v. First Pennsylvania Corp., 74 F.R.D. 606 [574]*574(E.D.Pa.1977). Other factors mentioned by the appellate court include:
“ . . . the complexity, expense and likely duration of the litigation . . ; the reaction of the class to the settlement . ; the stage of the proceedings and the amount of discovery completed . ; the risks of establishing liability . . . ; the risks of establishing damages . ; the risks of maintaining the class action through the trial . ; the ability of the defendants to withstand a greater judgment . . . ; the range of reasonableness of the settlement fund in light of the best possible recovery . . . ; the range of reasonableness of the settlement fund to a possible recovery in light of all the attendant risks of litigation . . . Girsh v. Jepson, 521 F.2d 153, 157 (3d Cir. 1975).
The Court will analyze this settlement with regard to these factors below. In this process, it will deal with many of the objections filed and expressed at the hearing.
Probability of success
Plaintiffs attempted a lawsuit in this Court that has been tried and has failed in a number of other courts. Here, the Court dismissed four-fifths of the case at the pleadings stage. This, in and of itself, does not bode well regarding its ultimate success, were this to go to trial. Since the Court’s 12(b)(6) opinion in 1974, a number of cases have since ruled on many of the same issues. Since these opinions consistently support the Court’s view of the law, a brief review is all that is necessary.
The Court first dismissed the “Truth-in-Lending” claim. This Court’s opinion has been accepted as authority by other courts in reaching and dismissing this issue in mortgage escrow cases. In Stavrides v. Mellon National Bank, 353 F.Supp. 1072 (W.D.Pa.1973), that court reached the same conclusion and was affirmed by this Circuit’s appellate court, 487 F.2d 953 (3d Cir. 1973). In Umdenstock v. American Mortgage and Investment Co., 363 F.Supp. 1375 (W.D.Okl.1973), that court also concluded that the mortgage escrow cases could not be pursued under the Truth-in-Lending Act. It was affirmed on this point by the Tenth Circuit court, 495 F.2d 589 (10th Cir. 1974). Even the Pennsylvania Supreme Court agrees with the Kinee opinion on this issue. Buchanan v. Brentwood Federal Savings and Loan Association, 457 Pa. 135, 320 A.2d 117, 130 (1974). Therefore, the Court must conclude that for the reasons set forth in its earlier dismissal of the Truth-in-Lending claim, the probability of success on that issue is virtually nil.
The same holds true for the claim under the Homeowners Loan Act. One other court has dismissed this claim and has been affirmed by its Court of Appeals. Gibson v. First Federal Savings and Loan Association, 364 F.Supp. 614 (E.D.Mich.), aff’d 504 F.2d 826 (6th Cir. 1974). The Court has not been presented with any reason, nor can it see any, that gives this claim any more merit than when the Court originally considered it. Therefore, the Court finds that the plaintiffs’ probability of success on this issue is equally low.
The Court must next consider the probability of the plaintiffs’ success on the antitrust claims. Although these were not dismissed by the Court’s first opinion, that memorandum noted that the claims were somewhat weak. The Court noted that the facts to prove the conspiracy would have to be circumstantial, and could be read as probative of innocent behavior as well. Illicit intent is a difficult thing to prove, and plaintiffs’ likelihood of success on the conspiracy claim is limited by this factor. Also, the few facts that have emerged through the discovery pertinent to class certification, recited by the Court at 66 F.R.D. 585-586, show a non-uniformity of action that would militate against such a finding of conspiratorial parallel action. Although, as the Court observed in Kinee, a complete uniformity of action is not necessary to make out plaintiffs’ claim, the facts here are farther from those allowing an inference of conspiracy than might have been anticipated. The escrow practices of the. many alleged conspirators vary in material detail: some include assessments in the escrow, others do not; some insist on escrow[575]*575ing for all loans, some for only government-insured loans.
The Court cannot find any record of a conspiracy case in the mortgage escrow field having been decided. In Umdenstock, the district court granted summary judgment in favor of defendants, but the circuit court reversed, on the grounds that such an issue was not appropriately resolved by a Rule 56 motion. In Stavrides, the district court did not dismiss the antitrust conspiracy claim on the pleadings, but this Court can find no evidence of this issue having been tried or otherwise resolved. Had they been successfully tried, the Court might be more optimistic about the claim’s possible success. As it is, on the limited record before the Court, the Court cannot be anything but pessimistic about its future at trial.
Plaintiffs, in their complaint, also claim that the escrowing and the mortgage were illegally “tied” together. The problem presented by this argument, as the Stavrides court acknowledged, is that the plaintiffs must show that “escrowing” was a separate product. In its first opinion, this Court termed this contention “strained.” The Court is still of that opinion, and concludes that proof of this claim would be difficult, if not impossible.
Finally, on the state law issues, some of the objections were based on the viewpoint that at least the state law issues have a rosy future in the Pennsylvania courts. The objectors base their opinion on Buchanan v. Brentwood Federal Savings and Loan Association, supra. In that 1974 decision, the Pennsylvania Supreme Court allowed that mortgage escrow case to proceed on the pleadings. However, in the opinion of at least one of the justices, the case just squeaked by the liberal standards for a demurrer in this state’s courts. Justice Pomeroy, in a separate opinion, reflected on the rule that a complaint should be dismissed only in cases which are clear and free from doubt. The justice agreed that any doubt should be resolved at trial. Although he agreed with the decision to sustain the demurrer, he said: “In its present status, the case at bar presents, as I see it, serious doubt as to whether [it] should have been sustained.” 320 A.2d at 133-134. He criticized the majority for writing what he considered to be an overly broad opinion solely on the basis of the pleadings.
Only one other state court has sustained a demurrer on a mortgage escrow case brought under these theories. See, Carpenter v. Suffolk Franklin Savings Bank, 392 Mass. 770, 291 N.E.2d 609 (1973). At trial, the lower court decided that the elements of a trust, either constructive or express, had not been made out. This conclusion was affirmed by the Supreme Judicial Court of Massachusetts, 346 N.E.2d 892 (1976). The objectors claim that the language of Brentwood reveals that Pennsylvania is the most favorable climate for a mortgage escrow case to proceed. Although this may be true on demurrer, in light of the many cases thrown out in other states on the pleadings,2 it still does not guarantee that when the facts are presented, the claims will be sustained.
One problem facing the state law issues at trial is that of proving that the multiplicity of agreements all establish a trust.3 In addition, the trust argument can only succeed if the parties’ intent to establish a [576]*576trust is shown. A number of cases in other states have been defeated on the pleadings by the defendants’ insistence that they did not intend to create a trust relationship, even if such was sought by the plaintiffs. See, e. g. Brooks v. Valley National Bank, 113 Ariz. 169, 548 P.2d 1166 (1976); Durkee v. Franklin Savings Association, 17 Ill. App.3d 978, 309 N.E.2d 118 (1974). The Court concludes that, even though objectors may be correct in their assertion that Pennsylvania has been amenable to this suit in the pretrial proceedings, there is no indication that these circumstances will continue at trial. In fact, in view of the difficulty of proof of the issues of express trust, constructive trust and unjust enrichment, as evidenced by the results in the Carpenter case, the Court is of the opinion that even here, the likelihood of success at trial is minimal.
Objectors also complain that the damages possible at trial are significantly greater than the relief offered by the settlement proposal. However, in considering the possible recovery at trial, the Court may discount such recovery in light of the possibility of no recovery at all because of an adverse liability verdict. In Bryan v. Pittsburgh Plate Glass Co., 494 F.2d 799 (3d Cir.) cert. den. 419 U.S. 900, 95 S.Ct. 184, 42 L.Ed.2d 146 (1974), the appellate court sanctioned a settlement for a low percentage of total possible damages where it appeared that it was not probable that plaintiffs would win all damage categories. Here, plaintiffs are receiving future relief only, of two percent interest on their mortgage escrow accounts, capitalization or the possibility of not having to make escrow payments at all. Other mortgage escrow cases which have settled have not been nearly so generous. In the L. E. Lay Antitrust Litigation, MDL Docket No. 187 (E.D.Texas), Judge Justice approved a settlement which offered, as in the case at bar, no retroactive relief and prospective interest of only IV2 percent. The Graybeal v. American Savings and Loan settlements, 59 F.R.D. 7 (D.D.C.1974-75), provided for no retroactive damages and no prospective interest relief. The only thing that plaintiffs there received was the option to pay their own taxes or to keep making escrow payments without receiving interest. Finally, in the settlement of McNary v. American Savings and Loans, 76 F.R.D. 644 (N.D.Texas), the judge there approved a lump sum settlement for the plaintiffs, which was virtually split with counsel, and no prospective relief. The Court believes, after viewing these other settlements, that the prospective relief offered by this settlement is fair, adequate and reasonable, under all the circumstances. It will not require payment of any retroactive damages because of the problems with the liability case.
In assessing the probabilities of success on the merits, either as the liability or damages, the Court is not attempting to decide the merits of the case. The district court “need only evaluate the probable outcome of the litigation and is not required to weigh and decide each contention . .” Bryan v. Pittsburgh Plate Glass, supra at 801. This is a very difficult job, as the court observed in Brucker v. Thyssen-Bornemisza Europe N.V., 424 F.Supp. 679 (S.D. N.Y.1976), and the district court’s discretion is respected in absence of a clear abuse. Girsh v. Jepson, supra; Kohn v. American Metal Climax, Inc., 489 F.2d 262 (3d Cir. 1973). In evaluating the success of the case at trial, the Court is not to substitute its own business experience with that of experienced counsel, especially in a case of this length and complexity. Kusner v. First Pennsylvania Corp., supra; Cannon v. Texas Gulf Sulphur Co., 55 F.R.D. 308 (S.D.N.Y.1973). Here, where counsel for plaintiffs has been involved with this type of suit in other forums, and has remained in contact with the progress of the mortgage escrow cases around the country, his decision to accept this settlement rather than to take his case to trial is certainly an educated evaluation of the case’s strengths. The Court has decided that while a victory is still possible, the case is not such that the Court should reject a carefully arrived-upon and relatively favorable end to this litigation.
[577]*577
Other factors considered on the settlement
The Court is of the opinion that the other Girsh factors also favor acceptance of the settlement. The first factor is the complexity, expense and duration of the proceedings. The complexity of the suit is easily shown in pure numbers: over 188,000 plaintiff class members, over 120 named defendants, almost 80 named plaintiffs, and over 880 documents filed with the Court at this time. This litigation, already in progress for six years, would surely consume many more in preparation for trial, trial and the resultant appeals. Since it does involve unsettled issues, it is very appropriate for settlement. Brucker, supra. Plaintiffs’ counsel has submitted a request for expenses to date totalling over $22,000 and attorneys’ fees of almost $680,000. Without commenting on the reasonableness of these figures, they certainly indicate that a complete trial and appeals would result in extremely high costs.
The next Girsh factor, the reaction of the class, also points toward acceptance of the settlement. In a class of 188,000 plaintiffs, only some 8,000 chose to “opt-out” of the settlement. This five percent exclusion rate compares very favorably with the 20-percent rate that this class action encountered upon its initial notice. Thus, the Court can infer, of those who originally did not want to be part of the class at all, some 15 percent now approve of the settlement and wish to benefit from it.4
A class action settlement in which as many as 20 percent of the class objected has been approved and affirmed by this Circuit’s appellate court.
“While the proportion of the class opposed to a settlement is one factor to be considered in assessing its fairness, [citations omitted] a settlement is not unfair or unreasonable simply because a large number of class members oppose it. The drafters of Rule 23 chose as a means of protecting the class the requirement that the district court approve the settlement. They did not require rejection of a settlement on objection of a given part of the class.” Bryan v. Pittsburgh Glass Co., 494 F.2d at 803.
In light of the relatively large percentage of the class members choosing to stay with the settlement and the small numbers of objections filed,5 the Court finds that the factor of class reaction weighs in favor of accepting the settlement.
The next factor listed in Girsh is that of the stage of the proceedings and the amount of discovery completed. Here, some discovery was completed as part of the class certification proceedings. However much is left to be completed, and the Court stayed all discovery when it appeared that the settlement negotiations might bear fruit. Therefore, it would seem to serve the interests of judicial and financial economy to allow settlement at this stage, when the Court or the parties have not invested the time and effort of a full round of discovery or a full trial.
[578]*578The next issue not specifically reached is that of the risks of maintaining a class action throughout the trial. Here, there are two risks presented. The first problem was recognized by the Pennsylvania Supreme Court in Buchanan. In order to decide the state law issues, the court involved, be it. this Court or the state court, must examine terms of the mortgage escrow agreement. Here, where there are over 120 defendants, with an unknown number of different agreements, a new class or subclasses might have to be certified in order to try the state law issues. If a mortgagee had a unique agreement, and its mortgagors were not so numerous as to merit class treatment, they would have to be excluded on the state' claims. In Buchanan, the Supreme Court ordered that on remand, the trial court was to limit each class of plaintiffs to those holding mortgages in which the escrow clauses do not differ “materially.” 320 A.2d at 131. This might result in a number of the plaintiff class members being decertified. The second problem is presented by the question of damages. If the plaintiffs were to succeed and to reach a stage at which damages must be shown, the class action device might prove unuseable at that point. This problem was discussed by this Court in its class certification opinion. While it was not sufficient to defeat class certification, this problem certainly speaks in favor of this uniform class-wide settlement, in terms of the Court’s ability to manage this massive class.
The Court finds that the settlement amount is reasonable both in the light of the best possible recovery and the most probable recovery. Objectors have complained about the proposed two percent rate of interest, and assert that it should be at least equal to the passbook savings rate of interest. The Court finds that this would not be reasonable. A number of states have passed laws requiring payment of mortgage escrow accounts. (Pennsylvania is not one, although numerous bills have been introduced on this subject.) None of those states require lending institutions as escrowee to pay the full passbook savings rates. A few peg the escrow interest rate to the passbook savings rate, but discount it by a certain percent in order to allow the banks to recover their costs. Two of the largest states, California and New York, have selected two percent as a proper figure for escrow account interest. The Court considers these facts as evidence that requires that the objections be overruled and the settlement be approved.
The objectors also complain that the method of computation of the interest rate, on month-end balances after deductions for payments, will result in the mortgagors receiving nothing. This is simply not true. Many of the month-end balances will be significant, because the tax and rental payments are not made to the government authorities each month, but only at a lesser interval. Therefore, for some months, the mortgage escrow account will have a large balance, while other months, it will be small. Therefore, an average of these balances seems fair and reasonable to the Court. The L. E. Lay settlement, supra, also allows for interest to be paid only on the month-end balances. The Court finds this as further evidence of the propriety of this accounting method.
While the defendants have not attempted to show the Court that they are unable to withstand any greater judgment, such as was shown in Kusner v. First Pennsylvania Corporation, supra, the reasonableness of the relief sought is demonstrated in part by the studies submitted to the Court in support of the settlement.6 These stu[579]*579dies, conducted on mortgage escrow situations other than the one in this district, indicate that lending institutions make a profit, but a minimal one, on the maintenance of mortgage accounts. At least one other court has found similarly, In re L. E. Lay Antitrust Litigation, MDL Docket No. 187 (E.D.Texas, November 8,1976). In that case, Judge Justice retained an independent court expert who opined that the one and V2 percent interest allowed by that settlement was “more than fair and reasonable” in light of defendants’ profits on the accounts.7 Here, where the plaintiffs may receive even more than the L. E. Lay settlement, the Court finds this to be fair.
Other objections
Plaintiffs first complain about the release of defendants from further suit, in § 3.08. The Court does not find this release unreasonable. It was used in both the McNary and L. E. Lay settlement agreements. The objectors claim that it extends to any claims that a mortgagor might have regarding negligent handling of his escrow account. At the hearing, the liaison counsel for all defendants stated that it applied only to such claims as might grow out of the escrow interest issue. The Court considers that this statement is part and parcel of the settlement agreement, and is thus a limitation on any overly-broad reading of § 3.08.
Objectors also complain of the indemnity clause of § 3.09, whereby plaintiffs agree to indemnify the settling defendants from any action-over by a non-settling defendant in these tax escrow cases. The Court does not find this to be an unreasonable protection allowed defendants, who remain subject to possible liability because of the incomplete nature of this settlement. While, as plaintiffs’ counsel admitted at the hearing, this might ultimately result in the class members having to pay some monies, or have it deducted from their interest, the Court finds that these sums would be nominal when spread out among over 188,000 mortgagors. Since both the release and the indemnification are clearly prerequisites for any settlement by the defendants, the Court will not reject an otherwise generous settlement on these bases.
Objectors also claim that the provision for release of the defendants because of law or court decision requiring lower or no interest, and the exception for an institution that becomes non-competitive are unfair, because they favor the defendants. Both the Graybeai and L. E. Lay settlements, approved by those courts, contain similar provisions. Also, the Court does not find that the provisions of § 4.04, which do not come into play until the Court makes the appropriate finding, are such that could easily be abused by a defendant, as the objectors fear.
Objectors also have objected that the option of choosing to pay interest, capitalize, or discontinue escrowing should not remain with defendants, but should be given to mortgagors. (§ 3.03) The Court finds this provision to be reasonable. By requiring the lending institutions to choose, the plaintiff class has in fact achieved its goal of ending the lack of diversity allegedly caused by the conspiracy. This forces the defendants to be diverse. On this objection and those objections generally that the agreement is too ■ “pro-defendant,” the Court’s position is that the plaintiffs’ case [580]*580was legally insufficient to force a “pro-plaintiff” settlement. In view of the Court’s analysis of the weakness of plaintiffs’ case, the Court considers the portions of the agreement that favor defendants, such as their release if a future settlement is more favorable, to be fair and reasonable. This is simply a logical extension of the concept of reducing plaintiffs’ potential damage recovery because of the weakness of the liability case.
In conclusion, the Court hereby overrules all the objections presented by the objectors, both in writing and at the hearing. Some objections are so totally without merit, or are so closely related to others, that the court has not discussed them here. The Court has carefully considered both sides on this settlement, as it is required to do under Rule 23(e) as the protector of the absent class members, and has decided that the settlement is fair, adequate and reasonable. This is not to imply that this document is what the Court would have prepared, had it been a representative of the plaintiffs or of the defendants. However, as was stated before, it is not the Court’s job to replace the judgment of counsel with that of itself. Its job is to insure that an arms-length settlement, without collusion or fraud, adequately and fairly meets the interests of the class members. In light of the factors set forth in Girsh, and as an exercise of that discretion vested in it under the Rules, the Court hereby approves the settlement as proposed by counsel for plaintiffs and defendants.