United States v. Manufacturers Hanover Trust Co.

MacMAHON, District Judge. The government, claiming violation of Section 7 of the Clayton Act 1 and Section 1 of the Sherman Act, 2 seeks equitable relief undoing a merger of Manufacturers Trust Company and The Hanover Bank which resulted in the creation of defendant, Manufacturers Hanover Trust Company. The merger has now been in effect for more than three years. It was consummated on September 8, 1961, shortly before this suit was filed, 3 ****following prior *876 approval of the New York Superintendent of Banks, as required by the New York Banking Law, 4 and of the Board of Governors of the Federal Reserve System, as required by the federal Bank Merger Act. 5 Both the state and federal statutes require the respective banking agencies to consider not merely banking factors and the overall public interest before approving a merger but also the effect of the transaction on competition, including any tendency toward monopoly. Both agencies approved this merger, without adversary hearings, but after consideration of documentary evidence, statistical data, investigation, and, in the case of the Board, the reports of the Attorney General, the Federal Deposit Insurance Corporation (F.D.I.C.), and the Comptroller of the Currency, and interrogation of officers of the merging banks relative to the merger’s competitive effect. The Superintendent concluded, after thorough analysis of relevant data, that the merger “would not result in a concentration of assets beyond limits consistent with effective competition. The proposed merger would not result in such *877 a lessening of competition as to be injurious to the interest of the public, nor in such a lessening of competition as to tend toward monopoly.” The Board, despite objection on antitrust grounds interposed by the Attorney General, concluded, with the concurrence of the Comptroller of the Currency, that the merger would have no adverse competitive effect but “would tend to stimulate competition without significantly affecting the number or competitive strength of alternative sources of banking services.” The F.D.I.C. found the pro and anti competitive effects of the merger so in balance that it based its approval on banking factors and the public interest. The government makes no claim that the agencies relied on incorrect facts or misapplied the statutory standards under which they operate. It does not seek to review their decisions. Rather, it invokes original jurisdiction of this court under the antitrust laws 6 asking us to destroy a merger which the banking agencies validated. The government contends that the merger is a combination in unreasonable restraint of trade, violative of Section 1 of the Sherman Act, and that its effect “may be substantially to lessen competition or to tend to create a monopoly,” violative of Section 7 of the Clayton Act. We state at the outset that there is no evidence of predatory conduct, anticom-petitive behavior or motive, conspiracy, price-fixing, or any other intentional injury to competitors or to the public either by the constitutent banks, the resulting bank, or any of its largest competitors. The government asserts, however, that an inference of the proscribed anticompeti-tive effect in commercial banking in the City of New York, the metropolitan area, and the United States is compelled by market structure, specifically, the size of the constitutent banks, the permanent elimination of Hanover as a separate competitor, the resulting bank’s size and its share of the relevant markets, a history of mergers and a trend toward concentration, and the increase in concentration caused and threatened by the merger. Defendant contends that the facts give rise to no anticompetitive inference in the relevant markets and that the government has failed to prove its case. It argues that the government inflates its relative size by commingling the local and national markets, that mere size is not an offense, that mergers of competitors are not per se unlawful, that the merger has not eliminated significant competition, that the resulting bank does not control an undue share of the relevant markets, that concentration has not been unduly increased, and that customers are well served by numerous, strong and vigorous competitors, both locally and nationally. JURISDICTION — THE EFFECT OF AGENCY APPROVAL Defendant, caught in a cross fire, originally challenged the court’s jurisdiction over the subject matter. A similar argument was made unsuccessfully in the district court, but abandoned on appeal, in United States v. Philadelphia Nat’l Bank, 374 U.S. 321, 350 n. 26, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963). As a result, it was briefed but not argued on appeal by the government, and neither briefed nor argued by the banks. Nevertheless, it was considered but rejected by the Supreme Court. We are bound, therefore, to reject it here, whatever its merit. *878 Defendant does not press the point. 7 Instead, it urges that we should regard the opinions of the impartial banking agencies as extremely persuasive evidence. It points to the government’s failure to produce any material evidence here which was not before, or within the knowledge of, the banking agencies, to the similarity of the statutory standards governing the agencies’ consideration of the competitive factor to those of the antitrust laws, and to the agencies’ application of antitrust principles in their analyses of the competitive effect of the merger. The government, however, relying on Philadelphia, contends that we must judge the validity of the merger under a different standard from that governing the banking agencies and are bound, therefore, to assess its competitive effect on the basis of established antitrust principles, independent of the opinions of the banking agencies. In support of its position, the government asserts that the Bank Merger Act does not authorize the Federal Reserve Board to decide antitrust questions as such, United States v. Radio Corp. of America, 358 U.S. 334, 79 S.Ct. 457, 3 L.Ed.2d 354 (1959), and stresses the Board’s required consideration of public interest factors, irrelevant to the antitrust laws, and the lack of adversary hearings. In view of the Supreme Court’s decisions in United States v. Philadelphia Nat’l Bank, supra, and in United States v. First Nat’l Bank & Trust Co. of Lexington, 376 U.S. 665, 84 S.Ct. 1033, 12 L.Ed.2d 1 (1964), there is no longer any question that the Clayton and Sherman Acts apply to bank mergers and that the Bank Merger Act neither impairs the plenary jurisdiction of this court to adjudicate their validity, nor immunizes them from collateral attack here despite agency approval. It does not follow, however, that the court should ignore agency views. PRIMARY JURISDICTION In United States v. Philadelphia Nat’l Bank, supra, the Supreme Court rejected an argument that the doctrine of primary jurisdiction should be applied to the Comptroller’s approval of that merger under the Bank Merger Act. We think, however, that there are facts here, not present in Philadelphia, which, with all respect, call for a contrary holding and ad hoc application of the doctrine of primary jurisdiction to the claim made under Section 7 of the Clayton Act. This merger was effected by an acquisition of assets. There can be no question since Philadelphia that Section 7 of the Clayton Act has always applied to such bank mergers. Philadelphia and Lexington also make clear, as the government puts it, “that the Bank Merger Act of 1960 does not in any way diminish the full thrust of the Sherman and Clay-' ton Acts in bank merger cases.” An integral and vital part of the “full thrust” of the Clayton Act is Section 11, 15 U.S.C. § 21 (1958), which unequivocally vests “authority to enforce compliance” with Section 7 “in the Federal Reserve Board where applicable to banks.” That statute has never been repealed by Congress notwithstanding its amendment of Section 7 in 1950. Referring to it in a footnote in Philadelphia, 374 U.S. at 344-345 n. 22, 83 S.Ct. 1731, the Supreme Court said: “ * * * The Bank Merger Act of 1960, assigning roles in merger applications to the FDIC and the Comptroller of the Currency as well as to the FRB, plainly supplanted, we think, whatever authority the FRB may have acquired under § 11, by virtue of the amendment of § 7, to enforce § 7 against bank mergers.” *879 Whatever validity pro tanto repeal of Section 11 by implication may have had in Philadelphia, we think there is neither reason, nor basis, for such drastic surgery here. Repeal by implication on the facts in this case would be a radical departure from settled precedents. We have been taught by a long line of decisions, including Philadelphia, that of all the instruments for accommodation of regulatory statutes to the antitrust laws, pro tanto repeal of the antitrust laws by implication is the very last that ought to be employed. Indeed, we have been urged to strain the doctrine of primary jurisdiction, if necessary, to avoid such a result. Pan American World Airways, Inc. v. United States, 371 U.S. 296, 320-321, 83 S.Ct. 476, 9 L.Ed. 2d 325 (1963) (Brennan, J., dissenting) (and cases cited therein). In United States v. Borden Co., 308 U.S. 188, 197-206, 60 S.Ct. 182, 84 L.Ed. 181 (1939), the Court emphasized that where a later regulatory statute, such as the Bank Merger Act, shares common ground with the antitrust laws, we should not resort to repeal of the antitrust laws by implication unless there is “a positive repugnancy” and then only to the extent of the repugnancy. There is no need to labor the point; its roots go deep. 8 There is no repugnancy, positive or otherwise, between the Bank Merger Act and Clayton § 11 in their application to the facts of this case. That is why, with deference, the Supreme Court’s refusal to apply the doctrine of primary jurisdiction in Philadelphia is not controlling here. In Philadelphia, pro tanto repeal of Clayton § 11 by implication was grounded on a positive repugnancy resulting from the fact that the Bank Merger Act mandated sole authority to approve that particular merger to the Comptroller of the Currency. Such authority in the Comptroller is clearly repugnant to the Federal Reserve Board’s sole authority under Clayton § 11 to enforce compliance with § 7. In this case, however, because the banks involved are state banks, the Bank Merger Act vests sole authority to approve the merger in the Federal Reserve Board (12 U.S.C. § 1828(c) (3) (ii)). Advisory roles only are assigned to the F.D.I.C., the Comptroller of the Currency, and the Attorney General, and the Board is in no way bound by their reports. Clearly there is no repugnancy between sole power in the Board under the Bank Merger Act to prohibit or approve a merger in advance of its consummation, and its power under Clayton § 11 to enjoin or acquiesce in its consummation. Quite the contrary. Power under the Bank Merger Act to destroy the seed before it sprouts is perfectly consistent with power under Clayton §§11 and 7 to nip incipient restraints and monopolies “in the bud” or order divestiture. See Transamerica Corp. v. Board of Governors, 206 F.2d 163, 166, 169 (3 Cir.), cert, denied, 346 U.S. 901, 74 S.Ct. 225, 98 L.Ed. 401 (1953). Every single reason upon which Philadelphia predicates its holding that the Bank Merger Act did not repeal Clayton § 7 by implication applies here with *880 even greater force to Clayton § 11. 9 Clayton § 11 suffers from none of the infirmities of the impotent Bank Merger Act. It grants express authority to the Board to enforce compliance with Section 7 and provides for hearings, intervention by the Attorney General, findings of fact, and appeal to the Court of Appeals. Authority to enforce compliance empowers the Board to decide antitrust issues as such, to enlist the injunctive aid of the courts, Board of Governors v. Transamerica Corp., 184 F.2d 311 (9 Cir. 1950); contra, FTC v. International Paper Co., 241 F.2d 372 (2 Cir. 1956), or to fashion appropriate relief, including divestiture, by its own decree, Clayton Act § 11, 15 U.S.C. § 21(b) (Supp. V, 1964); Pan American World Airways, Inc. v. United States, supra, 371 U.S. at 312-313 nn. 17 & 18, 83 S.Ct. 476. We see no conceivable reason on the facts of this case for repealing by implication this comprehensive legislative plan for Board enforcement of Section 7. We hold, therefore, that Clayton § 11 applies to this case. The Clayton Act, however, contains a scheme of dual enforcement. 10 United States v. W. T. Grant Co., 345 U.S. 629, 631, 73 S.Ct. 894, 97 L.Ed. 1303 (1953). The Board’s jurisdiction to enforce Section 7 “where applicable to banks” is, therefore, not exclusive but concurrent with the district court’s jurisdiction under Section 15 of the Clayton Act, 15 U.S.C. § 25 (1958). It seems obvious that Congress intended that the technical and complex problems involved in the application of Clayton § 7 to the banking industry should be resolved, at least in the first instance, by a body of experts who know the competitive realities of the banking business. Indeed, the legislative histories of both Clayton § 11 11 and the Bank Merger Act 12 are replete with that very theme. *881 We think that if ever there were a field requiring administrative expertise to unravel the tangled threads of the evidence and weave them into a meaningful fabric, this is it. This case involves a multitude of technical, intricate and complex problems in the field of money and banking, a subject within the special competence of the Board and outside the conventional experience of judges. The Board is intimately familiar with this technical subject matter, as well as the competitive realities involved, from its long experience as the administrator of the nation’s banking system, periodic reports, examinations, studies, etc. It knows the relevant products; the parties to this merger and the pattern of their business; the extent, locus and significance of previous competition between them; the number, strength and pattern of business of remaining competitors and the vigor of competition, both locally and nationally; the banking habits of all customers, great and small; their practicable banking alternatives; the geographic areas of effective competition for their banking business; the history, interrelationships and trends of the geographic markets; the probable impact of the merger on depositors, borrowers and competitors, great and small, local and national; the degree of concentration, locally and nationally; the reasons for concentration and its effect on the competitive picture; the operations of the nation’s money markets; and the effect of the government’s monetary and fiscal operations on the markets and the impact of this merger upon them. The fact that the banking agencies know the banking business undoubtedly explains why the evidence before us, which was largely prepared for the agencies, contains so little to acquaint us with the industry. As we shall see, the record is inadequate to the point that there is no direct evidence as to the market shares of the banks involved or any of their competitors in either of the relevant geographic markets. We have been obliged, therefore, to resort to circumstantial evidence, with all its imperfections, at many vital points in order to decide this case. We have also been compelled to dredge through reference material and take judicial notice of essential information which the Board knows, or could gather, out of hand. Much of the evidence consists of stale statistics and studies prepared for other purposes by the Board or other banking agencies. Often, the only information available is many years old and is a precarious basis for decision in this rapidly changing world. Many statistical facts which we need are either wholly lacking or so scrambled as to be useless. The Board has the staff and facilities to obtain the missing pieces of the puzzle and to organize the swamp of statistics into a meaningful pattern. Our tedious ef *882 forts, at best, have left distressing gaps in pertinent information. These considerations, we think, cry out loud for application of the doctrine of primary jurisdiction. 13 Whitney Nat’l Bank v. Bank of New Orleans, 379 U.S. 411, 85 S.Ct. 551, 13 L.Ed.2d 386 (1965). Ordinarily, application of the doctrine of primary jurisdiction would require either dismissal of the action, thereby relegating the parties to the Board and the Court of Appeals, or suspension of the judicial process pending referral of the issue to the Board for its views. Far East Conference v. United States, 342 U.S. 570, 576, 72 S.Ct. 492, 96 L.Ed. 576 (1952). We think, however, that resort to either course at this stage of the litigation would be little short of ridiculous. There are compelling practical reasons for refusing to shuttle this case back and forth between court and agency and for meeting the problem in a different and novel way. The litigation has already been prolonged awaiting the outcome of the Philadelphia and Lexington cases, the presentation of additional evidence, revised findings, briefs, and this court’s labors. All the evidence which the parties have desired to present is now before us, and to relegate the matter to the Board for still further proceedings would greatly increase the length and cost of the litigation, exhaust the litigants, delay ultimate determination, serve no useful purpose, and defeat the ends of justice. The court must decide the Sherman Act claim in any event, and, since it poses issues requiring much the same analysis and consideration as those under the Clayton Act, the economies' of judicial administration demand decision of both claims now. Finally referral to the Board would be an idle gesture imposing needless duplication of effort for, although the Board’s opinion is short on factual analysis and not of much help to the court, we already have the benefit of its ultimate *883 conclusion which was based on facts set forth in the application to merge or otherwise within its knowledge. The Board also had the benefit of the thorough analysis of the New York Superintendent of Banks and the F.D.I.C. The doctrine of primary jurisdiction is flexible, and we should shape it and, if necessary, strain it to fit the peculiar posture of this case in order to reach a practical accommodation of court and agency. The Board approved this merger long before the Philadelphia decision. It never invoked jurisdiction under Clayton § 11 for it was undoubtedly laboring under the common mistake that Clayton § 7 did not apply to bank mergers effected by an acquisition of assets. 14 Although the Board failed to make helpful findings of fact, there is no reason to suppose that its views about the competitive effect of this merger would have been any different from what they were under the Bank Merger Act if it had predicted the outcome of Philadelphia and invoked jurisdiction over this merger under Clayton § 11. This follows, first, from the fact that the evidence is largely undisputed and there is little material evidence here which was not before, or within the knowledge of, the Board when it approved this merger under the Bank Merger Act, and, second, from the fact that the norm for assessing the competitive effect of a merger under that Act is substantively indistinguishable from that laid down in the antitrust laws. We do not understand what sort of issues bear on the effect of a merger on competition, including any tendency toward monopoly, unless they be antitrust issues. Nor does the government offer any suggestions, although pressed to do so on oral argument. A clearer or more comprehensive grant of power to consider antitrust questions is difficult to conceive. Language far less specific (unfair competition) passed muster in Pan American World Airways, Inc. v. United States, supra 371 U.S. at 306-307, 83 S.Ct. 476. Plainly, the Board is required to consider all the ramified competitive effects of a merger on competition. Necessarily, that broad authority requires consideration of the narrower question of whether, in any line of commerce in any section of the country, the effect of a particular bank merger may be a combination in restraint of trade, or a monopoly, 15 or a substantial lessening of competition, or to tend to create a monopoly. We think it too clear for argument that such rudimentary questions lie at the threshold of the most cursory consideration of “the effect of the transaction on competition, including any tendency toward monopoly.” That this is so is further apparent on the face of the statute from its mandatory requirement that, in the interest of uniform.standards, before approving a merger, the Board “shall request” a report on “the competitive factors involved” from the Attorney General, as well as the other two bank supervisory agencies. There is no point whatever in requiring such a report from the Attorney General unless it be to insure that his special competence on antitrust questions as such, particularly the impact of the Sherman and Clayton Acts on the proposed merger, is brought to bear on the Board’s consideration of its competitive effect. 16 This is evident on the face of the report of the Assistant Attor *884 ney General in charge of the Antitrust Division of the Department of Justice, submitted to the Board in this very case. 17 As we shall see, his report was superficial and permeated with erroneous assumptions of fact as well as errors of law. It cannot be gainsaid, however, that it purported to be an opinion on antitrust questions as such. Thus, the ultimate antitrust issues raised here were component ingredients of the competitive factors which the Board was required to, and did, consider and decide in approving the merger. It is absurd to suppose that Congress intended merely to burden the Board with vain contemplation. Rather, we must assume that Congress had some meaningful purpose, when it enacted a specific statute like the Bank Merger Act, of sole application to particular transactions in a single industry. That purpose, on the face of the statute, was to grant the agencies broad authority not only to weigh and consider all the antitrust questions involved, but also to make a meaningful judgment about them, along with the banking factors, as essential components of the ultimate standard of public interest. Thus, the Bank Merger Act, to the extent that it deals with competitive factors involved in a merger per ses shares common ground with the Sherman and Clayton Acts. The Acts differ, however, in at least two important respects: First, the sole standard for determining the validity of a merger under the Sherman and Clayton Acts is the actual or potential anticompetitive effect, while under the Bank Merger Act, the overall public interest, and not the anti-competitive effect, is the governing criterion. Thus, the Bank Merger Act would appear to sanction agency approval of a merger, even though it violated the antitrust laws, if, on a balance of all the designated factors, the agency decided that, nevertheless, it was in the overall public interest. 18 A court, however, would be obliged to invalidate a merger found to violate the antitrust laws even though it served the public interest. 19 The difference in controlling standards, the lack of a pervasive regulatory scheme, *885 the absence of an express exemption from, or repeal of, the antitrust laws, and the failure of the Act to provide for adversary hearings led the Supreme Court in Philadelphia to the conclusion that, since repeals by implication are not favored, neither the Bank Merger Act, nor agency approval of a merger, immunized the merger from challenge under the antitrust laws. United States v. Philadelphia Nat’l Bank, supra 374 U.S. at 851, 83 S.Ct. 1715; California v. FPC, 369 U.S. 482, 82 S.Ct. 901, 8 L.Ed.2d 54 (1962); United States v. Radio Corp. of America, supra. We think that the lack of adversary hearings in this case is of no practical significance because the material evidentiary facts were before the Board, and they were, and still are, largely undisputed. Thus, there was nothing to hear but argument, and the Attorney General provided that in his written report. 20 Nor is it of any moment that the Bank Merger Act fails to state what weight shall be given to the competitive factor, for that argument is germane only if the agency finds a merger anti-competitive but approves it anyway because of favorable banking factors, which was not the situation here. Second, the Sherman and Clayton Acts do not prohibit mergers outright but forbid only those having a specified anticompetitive effect, and neither Act contains any provision for prior approval, while the Bank Merger Act unqualifiedly prohibits all mergers between insured banks without prior written consent of the designated bank supervisory agency. This difference is not discussed in Philadelphia, but it is important to the question of what weight we should give to Board approval of this merger. The obvious purpose of such a provision was to plug a hole in Clayton § 7 20 (a) to the end that anticompetitive mergers would be blocked in their incipiency and to enable customers, businessmen, and the community to place confidence in the lawfulness and stability of approved mergers. Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 362, 83 S.Ct. 1715. The legislative history shows that Congress was mindful that, in essence, commercial banks, unlike industrial corporations, are akin to fiduciaries entrusted with depositors’ money. After all, the Bank Merger Act is part of the Federal Deposit Insurance Act and partakes of its purpose to safeguard depositors from loss and rival banks from failure which might result if a merger threatened destructive competition. Nor should we ignore the Board’s interest, as the nation’s monetary manager, in preserving vigorous competition and ample banking alternatives to insure that credit is neither too scarce, nor too hard to get, nor too dear. The competitive effect of this merger, therefore, was, and will continue to be, of vital concern to the Board. Plainly, Congress sought to minimize, if not to remove, the hazards, vagaries, *886 burdens, and penalties which might be visited upon depositors, business, and the community if an approved bank merger were later found by a court to be illegal under the antitrust laws. It recognized that the unscrambling complexities inherent in divestiture, troublesome enough in other industries, might well wreak havoc upon a bank, with incalculable harm, to depositors, borrowers, and the public weal. 21 Indeed, those considerations were the theme of the government’s tardy application for a temporary restraining order in this very case. We think all of the foregoing considerations compel us to heed the views of the Board. “After all, these anti-trust problems are largely factual and their true solution depends in the last analysis upon an intimate familiarity with the characteristic features of a particular industry in which these problems arise.” United States v. Morgan, 118 F.Supp. 621, 699 (S.D.N.Y.1953). It seems obvious, therefore, that we should treat the Board’s views as though they had been rendered under Clayton § 11. We do not stop to inquire whether our function is, therefore, limited to review. Rather, because Clayton § 11 itself provides that no order of the Board confers immunity under the antitrust laws, the decision in Philadelphia, our plenary jurisdiction, and the novel posture of this case, we will accept the agencies’ views about banking facts, including the nature of the business of the constituent banks, the existence and locus of effective competition for all types of accounts, and the practicable banking alternatives of particular classes of eus-tomers as persuasive and helpful evidence in our analysis of the competitive effect of this merger, International Shoe Co. v. FTC, 280 U.S. 291, 299, 50 S.Ct. 89, 74 L.Ed. 431 (1930), but the agencies’ conclusions of law on antitrust questions as such, including the competitive effect of this merger, are not binding upon us in our independent application of the antitrust laws. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 367, 83 S.Ct. 1715; United States v. Morgan, supra, 118 F.Supp. at 699; 7 Wigmore, Evidence § 1918, at 10, § 1952, at 81 (3d ed. 1940). Anything less would turn the Bank Merger Act into a trap so repugnant to fundamental fairness that it would make a mockery of a court of equity. The principal reasons for the doctrine of primary jurisdiction are the need for expertise and orderly and sensible coordination of agency and court. See 3 Davis, Administrative Law § 19.01 (1958). We find nothing in Philadelphia requiring us to disregard the Board’s views. On the contrary, the Supreme Court took pains to note that its holding did not deprive the Bank Merger Act of its intended force or thwart its objectives, 374 U.S. at 354, 83 S.Ct. 1715 and, indeed, in its own analysis of the antitrust questions presented, found help in the views of the banking agencies, 374 U.S. at 361, 83 S.Ct. 1715. We recognize, however, that in leaving bank mergers subject to the antitrust laws, Congress intended that their validity should be definitively judged, not by the banking agencies, however expert in the banking field, but by judges presumably skilled in antitrust adjudication. We turn to that task. *887 THE FOUNDATION FOR ANALYSIS OF THE COMPETITIVE EFFECT OF THE MERGER UNDER THE ANTITRUST LAWS. The Statutory Tests. The ultimate question under Sherman § 1 is whether the merger constitutes a combination in unreasonable restraint of trade. United States v. First Nat’l Bank & Trust Co. of Lexington, supra ; United States v. Columbia Steel Co., 334 U.S. 495, 522, 68 S.Ct. 1107, 92 L.Ed. 1533 (1948); Standard Oil Co. of New Jersey v. United States, 221 U.S. 1, 31 S.Ct. 502, 55 L.Ed. 619 (1911). The question under Clayton § 7 is whether the effect of the merger may be substantially to lessen competition or to tend to create a monopoly in the relevant market. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 362, 83 S.Ct. 1715. As the Supreme Court noted in United States v. Philadelphia Nat’l Bank, supra, these are not the kind of questions which are “susceptible of a ready and precise answer.” They require “not merely an appraisal of the immediate impact of the merger upon competition, but a prediction of its impact upon competitive conditions in the future. * * * Such a prediction is sound only if it is based upon a firm understanding of the structure of the relevant market; yet the relevant economic data are both complex and elusive.” 374 U.S. at 362, 83 S.Ct. 1741. Obviously the structure of the relevant market and the competitive effect of a given merger vary with the setting and unique facts of each case. Brown Shoe Co. v. United States, 370 U.S. 294, 322 n. 38, 82 S.Ct. 1502, 8 L. Ed.2d 510 (1962). As we shall see, the market facts here are so different from those in Philadelphia and other recent merger decisions of the Supreme Court under Clayton § 7 that those cases do not supply “a ready and precise answer” to the questions raised in this one. The primary problem in ascertaining market structure under either the Sherman or Clayton Acts is definition of the product (relevant product or services market) and “section of the country” (relevant geographic market) in which to appraise the actual or probable competitive effect of the merger. United States v. First Nat’l Bank & Trust Co. of Lexington, supra; United States v. Philadelphia Nat’l Bank, supra; United States v. E. I. Du Pont De Nemours & Co., supra; United States v. Columbia Steel Co., supra. Market Structure. A. The Economic Setting. The structure of the relevant market, like the structure of anything else, starts with a foundation. Commercial banks provide a broad range of services built upon the needs of their customers. Those needs — from the smallest deposit to the largest loan — are created by, and vary with, the forces of the underlying economy. This case centers on the merger of two New York banks. A realistic grasp of the economy of metropolitan New York and its links with the rest of the nation and the world is essential to understand the commercial banking structure built upon it, define the relevant product and geographic markets, and view the competitive picture in true perspective. Commercial banking in metropolitan New York is built upon a massive economic foundation. New York has long been the largest city in the nation and is now the third largest in the world. 22 It has a cosmopolitan population of almost 8,000,000, 23 another 7,000,000 live on its perimeter, and a fourth of the *888 nation within a 250-mile radius. 24 It is the nation’s largest center of finance, trade, manufacture, and commerce. It is the world’s largest consumer market. Its port is the most active in the world; each year 26,000 ships, one every twenty-two minutes, enter or leave carrying an estimated $9 billion worth of cargo, or two-fifths by dollar volume, of the maritime trade of the country. 25 The two major metropolitan air terminals annually serve 13,000,000 passengers, handle 800.000 aircraft arrivals and departures, and contribute $66 million in customs duties. 26 The city is headquarters for nearly 2.000 businesses, each with a net worth of $1 million or more, including 135 of the country’s 500 largest industrial corporations and 11 of the 50 largest utilities. 27 Moreover, a vast majority of the remaining industrial leaders, both foreign and domestic, maintain New York offices, 28 and it is home to more trade and business associations than any other American city. 29 A significant complement of this managerial density is that New York is the nation’s largest financial center. It houses the largest of the nation’s twelve regional Reserve banks, six of the nation’s ten largest commercial banks, and five of the ten largest insurance companies. 30 Its two stock exchanges trade an annual volume of well above $1.5 billion and account for 93% of all securities traded. Expressed in dollars, their volume is overwhelming, no less than $61 billion in 1961. 31 The city’s cotton and commodity exchanges, for the same period, acted as clearing houses for $2.2 billion in cotton, wool, rubber, hides, copper, lead, and zinc. 32 The city, when examined from the standpoint of its indigenous economic activity, is once again the national champion. The five boroughs contain North America’s most expansive and diversified manufacturing community, housing some 35,000 plants engaged in 326 different lines of manufacture. 33 These firms, although most are small by contemporary standards (two-thirds in the 1 to 19 *889 employee category), provide work for almost 900,000 persons. The number so employed is 68% greater than those similarly employed in Chicago and more than three times that in Los Angeles, the next two leading manufacturing cities. 34 The product of their labor is over $7.5 billion each year, and their payroll of over $4 billion rivals the entire foreign aid expenditure. 35 The gross personal income to residents per annum is $24 billion, 36 or 6% of the national total. 37 It is the hub of the wealthiest metropolitan area in the country in terms of personal income. In 1961, the total income of individuals in the metropolitan area was $36.79 billion. 38 This partially explains why the city’s nearly 90,000 retail establishments ring up almost $10 billion in sales annually, exceeding the combined total for Chicago, Philadelphia, and Boston. 39 Yet the sales for the city’s 29,000 wholesale establishments are even greater, amounting to over $45 billion. 40 New York is thus the chief market for the nation’s consumer and industrial goods. The city also boasts a service industry doing a volume of $5.3 billion annually, three times that of its closest competitor, and larger than that of any state other than New York. 41 Each year, the city plays host to some 14,000,000 visitors, on business or pleasure, who add $1 billion to the city’s sales and services. 42 It is not surprising that the business of managing New York’s affairs requires a governmental staff second only in size to that of the United States itself and an expense budget of over $2.7 billion. 43 The city has long since outgrown its political boundaries and expanded into contiguous counties. Adjoining Nassau and Westchester counties are fused to the city. They are virtually a continuation of its factories, stores, apartment houses, and private residences, and with the city constitute a single economic and trading area. Both counties are densely populated, heavily developed, and rapidly growing, largely at the expense of the city. Middle income families have moved into them from the city and have been replaced by families living on a lower, and often subsidized, economic scale. 44 Nassau and Westchester now have a population of 1,300,171 and 808,891, respectively, which combined exceeds that of Philadelphia, the fourth largest city in the country 45 Their combined gross personal income of $8.049 billion is greater per capita than that of any state in the Union. 46 *890 Despite the massive scale of the economy of the New York metropolitan area, for the past twenty years the city proper has lagged behind the growth of the rest of the country and the suburban counties. Postwar expansion and decentralization of industry, population growth and shifts, the rise and change in the distribution of wealth and income, and the concomitant growth of other trade and financial centers have made heavy inroads in the economic pre-eminence of the metropolitan area. 47 We think it plain from this glimpse of the size, variety, complexity, pace, and diverse geographical structure of the relevant economy that its elaborate processes would collapse overnight without financial institutions equal to its needs. B. The Financial Industry It is written that the cornerstone of New York’s economic structure was laid over three centuries ago, in 1627, by Peter Minuit when he bartered a few trinkets for all of Manhattan. After that, perhaps because the bargain was transparently suspect, men have used more subtle media of exchange, and money in one form or another has ever since been the lifeblood of the immense economy of the metropolitan area. History and the community’s expanding commerce with the rest of a developing nation and the world have made New York a national and international financial center. Many of the nation’s giant financial institutions are located in New York because they had the advantage of an early start and gained momentum with the economic growth of the metropolitan area, the country, and the world. 48 Billions of dollars must move from buyer to seller, from exporter to importer, from lender to borrower, from depositor to banker, or back again, every day. 49 Money must be found in enormous quantities to meet the daily needs of millions of individuals, countless small firms, giant nationwide corporations, governmental bodies, banks, other financial institutions, and a myriad of brokers, dealers, merchants, and industrialists who operate in or through New York’s great markets. It takes money in enormous quantities to meet the payrolls of thousands of small business firms and hundreds of giant corporations, pay dividends, finance skyscrapers, pay for cargoes arriving every day from all over the world, carry countless commercial, industrial, and consumer loans, float million dollar blocks of corporate securities, or buy a billion dollars of government obligations in a single day. The money essential to the smooth workings of this complex and widespread commerce is provided by highly specialized financial institutions, insurance companies, savings and commercial banks, finance companies, factors, investment bankers, pension funds, credit unions, and personal loan companies. All play roles often interrelated, overlapping, and competitive. The volume of business is so great and the pace of transactions so swift that money in the form of coin and currency, like Peter Minuit’s trinkets before it, is almost outmoded as a medium of exchange. It has been replaced by a more convenient, simpler, safer, and swifter medium capable of keeping pace with the dizzying velocity of transactions. Today’s medium of exchange is checkbook money. It constitutes 80% of the national money supply. 50 Providing it’ in huge quantities is the unique job of commercial banks. C. Commercial Banking, Generally Fundamentally, commercial banking is an endless cycle of borrowing and lending. Commercial banks are department stores of finance. They do their borrow- *891 mg and lending in a variety of forms. 51 Not all commercial banks offer a full range of services. Smaller banks, because of lower lending limits and lack of large and specialized staffs, tend to concentrate on the local needs of local individuals and small businesses. Conversely, some large banks, with greater lending limits, specialization, and organization, focus on the nationwide needs of large corporations. Others deal in large and small units with customers standing in a broad economic spectrum. Commercial banks, large or small, are the only financial institutions authorized by law to accept demand deposits subject to check. This unique power enables them to create money 52 and sets them apart from savings banks and other financial institutions. They, therefore, act as reservoirs gathering, expanding, supplying, and, with the help of the Federal Reserve System, clearing checkbook money in huge amounts to meet the economic needs of the community and the country. *892 When a bank accepts a demand deposit, it borrows money without interest 53 which it pumps back into the economy by honoring the depositor’s checks and making loans. The cycle is repeated over and over through the banking system and makes commercial banks the intermediaries in most financial transactions. The importance of this function is evidenced by the fact that in 1958 Americans had 52,000,000 checking accounts, with a total of $110 billion on deposit with commercial banks, and paid an estimated 90% of their bills by check. 54 The second most important function of commercial banks is supplying short-term credit to individuals, business entities, and the government. Thus, in 1961 commercial banks throughout the country carried a total of about $118 billion in loans. 55 In practice, loans are usually credited to the borrower’s account and, therefore, generate corresponding, but volatile, demand deposits. 56 Commercial banking is subject to comprehensive regulation, state and federal. Entry into the business, branching, and merging are controlled to preserve sound banking, protect depositors, and foster strong, but prevent destructive, competition. 57 Maximum interest rates, with some exceptions, are fixed by state usury laws, 58 and minimum (prime) rates indirectly by the Federal Reserve’s regulation of the money supply. 59 Interest on demand deposits is prohibited, and it is regulated on time and savings deposits. 60 Some service charges are free of regulation. Banks are required to submit periodic reports to state and federal banking agencies, 61 submit to examinations whenever the banking authorities deem it necessary, 62 and, if unsound or unsafe practices are found, their charters are subject to summary suspension and even termination. 63 Within these limits, banks are free to set or negotiate the terms and conditions of deposits and loans in competition with their rivals. It is clear, however, that absolute competition in the banking industry is not only inconsistent with, but also in many respects prohibited by, regulatory laws. D. Commercial Banking in the Metro politan Area. The commercial' banking structure of any community mirrors the size, complexity, geographic relationships, tempo, and characteristics of the underlying economy. It is not surprising, therefore, that the size, volume, complexity, variety, tempo, and geographic dispersion of the transactions flowing through commercial banks in the metropolitan area dwarfs that of other communities. It is hard to realize the immense scale of operations. It is starkly revealed, however, in the fact that 607,000,000 checks worth $624 billion were handled by the New York Federal Reserve Bank in 1958, and on a typical day $2 billion worth of checks cleared through the New York Clearing House, giving New York *893 a volume greater than that of the next 23 largest cities combined. 64 Part of the flow is generated by the local needs of millions of individuals and thousands of business firms, another part by transactions between outsiders and insiders, still other parts by outsiders and insiders buying and selling through New York’s stock and commodity exchanges, by nationwide corporations borrowing, receiving, or disbursing funds, by the Treasury’s gathering funds from its borrowings or taxes, or disbursing them from its account at the Federal Reserve banks, by outside banks shifting funds into and out of New York in response to customers’ needs or money market pressures, and, finally, by the movement of funds in and out of the area from abroad. 65 Thus, billions of dollars move in and around, and in and out of, the metropolitan area every day in a multitude of banking transactions of infinite variety in size, nature, class of customer, and geographic dispersion. The sheer volume and churning of transactions generate billions in demand deposits, make the area a traffic center for checkbook money, and create a vast reservoir of loanable funds. The area banks, therefore, handle an enormous volume and variety of loans ranging from installment loans to local consumers to multimillion dollar loans to nationwide corporations. As one might expect, the size, volume, pace, and variety of banking services in the metropolitan area has spawned and shaped a commercial banking industry commensurate with the demand for services. There are numerous competitors, great and small, geared „to general and specialized banking services in varying degrees. Thus, at the time of the merger of Manufacturers and Hanover, there were seventy-two commercial banks operating 778 offices in the metropolitan area. They held a total of $37.3 billion in deposits, $20.9 billion in loans, and $44.1 billion in assets. Eight of them held over $1 billion in assets, five others over a half billion, eleven others over $100 million, forty-seven others from $2 to $100 million, and the smallest one $166,000. 66 With this preliminary look at commercial banking in metropolitan New York, we turn to the place occupied by the parties to this merger. E. The Parties to the Merger and Their Place in the Markets. Both Manufacturers and Hanover were New York corporations with their principal place of business in New York City. Manufacturers traces its origin to 1905, and Hanover to 1873. Both were successful, well managed, 67 and enjoyed substantial growth. Nevertheless, prior to the merger, their growth rates, as well as that of their largest competitors, lagged those of representative smaller banks in the metropolitan area, 68 as well as those of the nation’s 100 largest banks. 69 Hanover had no history of mergers, but Manufacturers had one major merger fifteen years ago 70 and five minor ones *894 during the twenty-year period ending in I960. 71 Traditionally Hanover was predominantly, though not entirely, a wholesale bank. It generated most of its business from the deposits and loans of large corporate customers and wealthy individuals. It also served small business firms, but to a very much lesser extent than Manufacturers, and six months before the merger it had ventured into some retail banking services in the local market. It was one of the foremost fiduciary institutions in the city, both in the corporate and personal trust fields. Hanover had 10 branches, and all of them were confined to Manhattan, largely in the uptown business community and in the downtown industrial and financial center. Hanover’s operations in the mass retail market were so recent and so small compared to other large integrated banks that the Federal Reserve Board characterized it as a wholesale bank confining its business “almost exclusively to banks and large corporate customers.” 72 Manufacturers, on the other hand, had a tradition of retail banking. Historically its growth was due largely to emphasis on service to smaller customers. Retail business, as we shall see, depends primarily upon convenience of location. Manufacturers, therefore, by 1960 had established the largest branch system in the area, with 120 banking offices spread throughout the five boroughs of the city, where it catered to a mass local market offering a wide range of popular banking services to the general public and small businesses. These included savings ae-counts, special checking accounts, Christmas Club accounts, gift checks, night depositories, personal loans, automobile loans, education loans, repair and modernization loans, loans on accounts receivable, commodity and industrial equipment financing, and small business loans. In the last fifteen years, Manufacturers also extended its operations into wholesale banking in a nationwide market. The banking agencies, therefore, characterized Manufacturers as primarily a retail bank which was also engaged in wholesale banking. The complementary nature of the business of the constituent banks is demonstrated by the fact that Hanover had only 44,099 depositors with an average balance of $39,000 compared to Manufacturers’ 1,070,606 depositors with an average balance of $3,000. Hanover had only 4,527 borrowers with loans averaging $214,000 compared to Manufacturers’ 208,322 borrowers with loans averaging $7,000. Measured by assets, before the merger Manufacturers was the fifth largest among 72 commercial banks in the metropolitan area and the sixth among 13,484 in the nation, while Hanover ranked eighth in the area and fourteenth in the nation. After the merger, Manufacturers Hanover ranked third in the metropolitan area. Six other banks remained with assets of over $2 billion each. Two of them, Chase Manhattan Bank and First National City Bank of New York, the nation’s second and third ranking banks, respectively, are much larger than Manufacturers Hanover. 73 Between December 31, 1960 and December *895 31, 1962 one new bank opened, two industrial banks converted to commercial banks, five banks were merged and applications for four new banks were approved by regulatory authorities. At December 31, 1962 there were 70 banks in the metropolitan area. 74 Twenty-one of defendant’s local competitors, other than the top six, have assets ranging from approximately $100 million to $1 billion, and forty-three more have assets ranging up to approximately $100 million. In addition, under a recent change in New York law 75 permitting foreign banks to open full service branch offices, eleven foreign commercial banks have opened 15 branches in the city. Seven of these foreign banks have assets in excess of $1 billion. 76 Before the merger, there were 778 banking offices spread throughout the metropolitan area, and there are now 897. Most of them provide a broad range of banking services to the general public, but a few wholesale banks, which publicly spurn the minutiae of retail banking or operate only in narrowly specialized fields, remain in the picture. 77 Immediately after the merger, Manufacturers Hanover had 130 branches, and now has 136, but its relative share of offices in the whole local market has declined since the merger from 16.71% to 15.16%. 78 Since the merger, the number of commercial banks in the nation has increased to 13,569, and, while Manufacturers Hanover now ranks fourth, the Bank of America still remains the largest bank m the country by a wide margin. Moreover, the number of billion dollar banks in the nation has increased from 24 to 31, and there are 169 others with ever-growing assets gradually ranging from $179 million to approximately $1 billion. A cursory comparison of the roster of the nation’s 200 largest banks, as of December 31, 1950, December 31, 1960, and December 31, 1963 shows that entry into the top 200 has been, and continues to be, wide open and that numerous banks throughout the country have so increased in relative size during the last fifteen years that they have necessarily entered the national market, and since this merger 13 banks have climbed over their competitors and into the top 200. 79 F. The Line of Commerce. Much of the evidence presented before the Philadelphia decision related to the issue of whether commercial banking is sufficiently distinct from services offered by other types of financial institutions to constitute a distinct line of commerce. That problem has been solved by defendant’s concession, inspired, if not compelled, by United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 356, 83 S.Ct. 1737, that “the cluster of products (various kinds of credit) and services (such as checking accounts and trust administration) denoted by the term ‘commercial banking’ * * * composes a distinct line of commerce.” The government, however, contends that certain banking services 80 are sub-produet markets requiring separate anal *896 ysis. “The boundaries of * * * a submarket may be determined by examining such practical indicia as industry or public recognition of the submarket as a separate economic entity, the product’s peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors.” Brown Shoe Co. v. United States, supra, 370 U.S. at 325, 82 S.Ct. 1524. We reject the government’s contention because it distorts the true competitive picture, obfuscates analysis, and there is no evidence that any of its selected banking services meet any of the criteria laid down in Brown Shoe Co. v. United States, supra. Cf. United States v. E. I. Du Pont De Nemours & Co., supra, 353 U.S. at 594, 77 S.Ct. 872. The purpose of defining the line of commerce is to focus the impact of a merger on competition. United States v. E. I. Du Pont De Nemours & Co., supra, at 592, 77 S.Ct. 872. Competition in banking is the rivalry among banks for customers. Our quest for the relevant line of commerce is, therefore, directed to the point of the merger’s impact on customers and other banks. The evidence shows that commercial banking in the metropolitan area falls into two distinct subproducts — “wholesale” and “retail” accounts. The late Dr. Jules I. Bogen, an acknowledged expert in the banking field, officers of the merged banks, and an officer of a competitor so testified. Their testimony is undisputed. There can be no question that the banking industry, business, and the public regard wholesale and retail banking as a trade reality. 81 All of the banking agencies, and even the Attorney General, acknowledged both lines in their analysis of this merger. In the parlance of the industry, a wholesale bank is one handling a small number of large accounts, concentrating its efforts primarily on large corporations, governmental bodies, financial institutions, and wealthy individuals, while a retail bank is one handling a large number of small and intermediate accounts, catering to the mass needs of the general public and small business. The economic scale of the customer, size of account, size of the bank, specialized nature of the service, tradition, reputation, and public image determine whether the pattern of a bank’s business is wholesale, retail, or wholesale-retail. Traditionally, many of the large banks located in New York City were wholesale banks. In order to serve their large customers, staffs larger than many fair-sized towns are required. Such banks are department stores of finance, many banks within a bank. They are usually divided into departments by area of the country, by industry, or both. Specialists in a host of industries, products, and markets keep abreast of countless economic and financial details to serve the particular needs of thousands of diversified industries. Skilled research staffs evaluate companies, forecast growth potential, survey market opportunities, plan financing, and program investments. These banks are also “foreign banks” with overseas offices to assist clients in foreign ventures. They maintain working relationships with foreign correspondent banks, finance exports and imports from all over the world, and are experts in foreign exchange handling billions of foreign funds every year. They keep abreast of the rates of exchange, regulations, and monetary conditions of hundreds of countries. They are also bankers’ banks linked to a network of domestic correspondents. Frequently they participate in loans with other or correspondent banks because the credit needs are big or specialized and banks are limited by law in the amount they can loan to a single borrower, 82 and their *897 overall lending ability is governed by the amount of deposits over reserve requirements. 83 Another aspect of wholesale banking is the short-term, impersonal, “open market” — the money market proper — where money temporarily idle is loaned in enormous sums for a brief period at a small return. Such funds must be invested and liquidated quickly. The most important assets in which the open market deals, therefore, are federal funds, 84 treasury bills, and other short-term governmental obligations. 85 Other less important assets include bankers’ acceptances, commercial and finance paper, municipal bonds, and short-term government agency obligations. 86 Funds in this open market, aside from Federal Reserve funds, are loaned primarily by commercial banks, business corporations, insurance companies, and foreign central banks. Such funds are in turn borrowed by other commercial banks, the federal government, finance companies, business firms, and others. The two most important institutions in the open market, aside from the Federal Reserve itself, are the big commercial banks 87 and the government securities dealers. The large banks, with their intensive use of money, are like mass production factories compared with the custom tailored lending of the average commercial bank. The nub of the difference stems from the huge accounts and special needs *898 of their corporate customers, their size, lending limits, and organization. Yet, in other ways, the large integrated banks are like the average small commercial bank. Spurred by a postwar lag in the growth of deposits attributable to non-bank competition, dispersion of industry, population shifts, and the growth of the suburbs, as well as competing financial centers throughout the country, 88 the trend for the past twenty years has been to retail banking. Many former wholesale banks have ventured into retail operations through a network of branches offering a broad range of services to local businesses and the general public. They provide millions of individuals and countless small businesses operating in the local market with check cashing and clearing facilities. They finance the purchase of everything from household furniture to machinery. Their loans make payment possible while materials are being manufactured, stored, or transported in or around the metropolitan area, or sitting on dealers’ shelves awaiting sale. Thus, as a result of its historical development, commercial banking in the metropolitan area involves both wholesale and retail banking in varying degrees. The distinction between the categories is clear under the test laid down in Brown Shoe Co. v. United States, supra, but the demarcation line between retail and wholesale accounts is fuzzy. That the dividing line is debatable, however, is of no moment. It simply means that workable guidelines must be found in order to sketch a true competitive picture. 89 The legislative history of Clayton § 7 is permeated with congressional concern for the public and small business. 90 Manifestly, small and intermediate customers are more limited in their choice of a bank and less able to bargain for terms and conditions than large customers with nationwide standing and the economic strength to conduct their banking business in any of the nation’s financial centers. The elimination of a banking alternative by a merger is, therefore, more likely to have a direct and immediate impact on the many small and intermediate customers than on the relatively few large ones. The law, however, applies equally to the great and the small. There is no doubt that this case concerns both wholesale and retail customers and rivalry for their business. We, therefore, recognize both groups as perfectly good lines of commerce 91 and will assess the competitive impact of this merger on both large and small customers and upon all of the commercial banks competing -for their patronage in the relevant geographic markets. G. The Geographic Markets. The parties differ as to the appropriate “section of the country” or relevant geographic market for assessing the probable competitive effect of the merger. The government contends that the primary market is confined to the City of New York but that the metropolitan area and the whole nation are also appropriate markets. Defendant contends that the primary geographic market is not confined to the City of New York but embraces the metropolitan area, and that we should exclude non-area banking business in calculating local market shares. The main offices and all of the branches of both Manufacturers and Hanover were located within the City of New York. It is clear, however, that they did considerable business with customers located in the metropolitan area and elsewhere throughout the country and the world. The question to be answered, however, is not where the parties to the merger did business, or even where they competed, but where, within the area of competitive overlap, the effect of *899 the merger on competition will be direct and immediate. “This depends upon ‘the geographic structure of supplier-customer relations.’ ” United States v. Philadelphia Nat'l Bank, supra, 374 U.S. at 357, 83 S.Ct. 1738. That relationship here is complex, chaotic, and elusive. It is plain, however, that it is by no means confined to the City of New York. The only clear fact emerging from the swamp of statistics and the facts behind them is that the geographic market varies with the geographic ties and the economic scale of the customer, and the size and nature of the banking service. The short of the matter is that the facts do not permit precise definition of the geographic market. Whatever market or markets we choose, therefore, will be somewhat artificial. Confronted with a similar problem in Philadelphia, the Supreme Court taught that we must find a workable compromise which avoids the indefensible extremes of drawing the market either too expansively or too narrowly. 374 U.S. at 361, 83 S.Ct. 1715. This means, we take it, that above all else the geographic markets must not be gerrymandered to rationalize theories but defined with fidelity to competitive realities. “Congress prescribed a pragmatic, factual approach to the definition of the relevant market and not a formal, legalistic one. The geographic market selected must, therefore, both ‘correspond to the commercial realities’ of the industry and be economically significant. Thus, although the geographic market in some instances may encompass the entire Nation, under other circumstances it may be as small as a single metropolitan area.” Brown Shoe Co. v. United States, supra, 370 U.S. at 336-337, 82 S.Ct. 1530. The fact that Clayton § 7 speaks of the lessening of competition in “any section of the country” does not, of course, mean that the geographic area in which the competitive effect of the merger is to be measured under the Clayton and Sherman Acts may be defined as small as the government chooses. Cf. Brown Shoe Co. v. United States, supra, at 367, 82 S.Ct. 1542 (Harlan, J., concurring in part and dissenting in part). Rather “the boundaries of the relevant market [or markets] must be drawn with sufficient breadth to include the competing products of each of the merging companies and to recognize competition where, in fact, competition exists.” Brown Shoe Co. v. United States, supra, at 326, 82 S.Ct. 1524. Our quest is directed to the geographic area of effective competition. That area depends first upon the geographic area in which competitors market the relevant products and second upon where, within the area of competitive overlap, the customer can practicably turn for supplies. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 359, 83 S.Ct. 1715; Brown Shoe Co. v. United States, supra, 370 U.S. at 320 n. 35, 82 S.Ct. 1502; Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 327, 81 S.Ct. 623, 5 L.Ed.2d 580 (1961); Standard Oil Co. of California v. United States, 337 U.S. 293, 299 n. 5, 69 S.Ct. 1051, 93 L.Ed. 1371 (1949). H. The Local Market. Our search for the local area of effective competition is guided to a workable segment of the market which is segregated from the balance of the market and large enough to be economically significant. Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 327, 81 S.Ct. 623; Standard Oil Co. of California v. United States, supra, 337 U.S. at 299 n. 5, 69 S. Ct. 1051; United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 596; n. 42. “In determining the area of effective competition for a given product, it will be necessary to decide what comprises an appreciable segment of the market. An appreciable segment of the market may not only be a segment which covers an appreciable segment of the trade, but it may also be a segment which is largely segregated from, independent of, or not affected by the trade in that product in other parts of the country.” *900 S.Rep.No.1775, 81st Cong., 2d Sess. 6 (1950), U.S.Code Cong.Service 1950, p. 4298. We note at the outset that the Federal Reserve Board did not even discuss the problem, but the New York Superintendent of Banks, the F. D. I. C., and the Comptroller of the Currency, though in-explicit, regarded the local market as coextensive with the metropolitan area. The parties have stipulated that the metropolitan area shall be defined as the City of New York, plus Nassau and Westches-ter Counties. 92 There can be no question that banks located in New York City market their services and compete with banks throughout the metropolitan area. 93 City banks aggressively seek the patronage of suburban residents through all advertising media. Competitive overlap is inherent also in the fact that the city is the hub of a local trading and marketing area which pulls suburban businessmen and some of their banking business into the city. There is intense rivalry between city and suburban banks for the business of over 200,000 commuters who sleep in the suburbs but spend their days working in the city. 94 Commuters, particularly those with small and intermediate ac-courts, often find it convenient to bank both where they work and where they live. The fact of effective intrametro-politan competition is demonstrated by evidence that both Manufacturers and Hanover derived considerable business from customers residing outside the city, in the metropolitan area. 95 Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 359 n. 36, 83 S.Ct. 1715. The existence of effective competition between city banks and banks located in Nassau and Westchester is confirmed by a 1960 amendment to New York Banking Law, § 105, subd. 1(b) which recognizes the city, plus Nassau and West-chester, as a unified banking community and permits cross-branching within the area. Under this legislation, New York City banks had been authorized to open thirty-two branches in Nassau or West-chester by the end of 1962. At least two large Nassau banks have now branched into the city, and more applications are pending. 96 The foregoing facts show that New York City commercial banks are in direct and effective competition with the banks *901 of Nassau and Westchester counties. Moreover, the population shift to the suburbs and the growth of suburban banks create a continual threat of even greater cross-branching, thereby intensifying the effect of potential intrametro-politan competition. Cf. United States v. Penn-Olin Chemical Co., 378 U.S. 158, 173, 84 S.Ct. 1710, 12 L.Ed.2d 775 (1964). The problem of defining the local geographic market is thus reduced to whether customers can practicably turn to banks within the metropolitan area for their banking needs. The evidence shows that most individuals depend upon banks in the immediate neighborhood of their home or place of work because their resources are small, their banking needs limited, and it is neither necessary nor convenient for them to bank elsewhere. Small and intermediate business enterprises tend to do business in a broader area, but within their own region, because they are known there and because limited resources, local needs, and convenience tend to localize their banking alternatives. Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 358-359 nn. 35, 36, 83 S.Ct. 1715. These practical restrictions on retail customers also orient retail banking, which caters to the general public, to the area where the bank is located and operate to insulate such banks from actual and potential competition from outside banks for retail accounts. 97 As a result, the metropolitan area is largely segregated from, independent of, and not affected by commerce in retail banking in other parts of the country. See S.Rep.No.1775, 81st Cong., 2d Sess. 5-6 (1950). The evidence shows, however, that as customers grow from intermediate to large, from local to nationally-known concerns, from single to multiple geographic locations, the factor of inconvenience diminishes to a point where it ceases to localize the customer’s banking alternatives and no longer walls off outside competitors. The purpose of defining geographic markets is to bring the competitive picture into clear focus. Cf. United States v. Continental Can Co., 378 U.S. 441, 84 S.Ct. 1738, 12 L.Ed.2d 953 (1964). Our search for a clear picture of the local market is thus directed not to the smallest or largest possible geographic area, but to the smallest workable area of effective competition. That area, we think, is the smallest separate competitive arena big enough to accommodate most of the customers. The metropolitan area unmistakably meets the specifications for, as we shall later develop, 99.74% of Manufacturers’ depositors and 96.77% of its borrowers were retail customers located within the metropolitan area, insulated from outside competitors, actual and potential, and unable as a practical matter to bank elsewhere. The corresponding percentages for Hanover are 94.59% and 67.39%. 98 The foregoing facts clearly segregate the metropolitan area from the rest of the nation as an economically significant market and roughly delineate the local area of effective competition. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 359-361, 83 S.Ct. 1715. See S.Rep.No.1775, 81st Cong., 2d Sess. 5-6 (1950). Accordingly, we find that the metropolitan area is a more appropriate and workable segment of the market for separate analysis of the competitive effect of this merger than any smaller section of the country. Our problem is not ended, however, for there is no dispute that the merging banks also operated in a national market. I. The National Market. The government claims that the merger is violative of the antitrust laws not only in the local market but also *902 in the United States. 99 The evidence overwhelmingly establishes, and, indeed, there is no dispute that Manufacturers and Hanover, along with other leading metropolitan banks, in addition to local business, also received substantial deposits from sources outside the metropolitan area, made substantial loans to borrowers located in other states and in foreign countries, and conducted a variety of banking business with bankers and representatives located throughout the country and the world. 100 Likewise there is no question that large banks located throughout the country compete in a nationwide area for the patronage of large customers located throughout the country. Such customers regularly use the mails, telephone, and telegraph to effect deposits, apply for and obtain loans, and carry out other banking transactions. All of the banking agencies and the Attorney General, therefore, found that the merging banks and their large metropolitan competitors operated in both a local and a national market. We agree. Thus, unlike Philadelphia or Lexington, we are confronted with two geographic markets and the problem of separating one from the other in order to ascertain the true competitive structure of each. The government bears the burden of proving its claim that the merger has the proscribed anticompetitive effects in either or both of the geographic markets. Definition of the geographic markets and proof of the competitive structure of each of them lie at the very threshold of the government’s burden of proof for “proper definition of the market is a ‘necessary predicate’ to an examination of the competition that may be affected by the horizontal aspects of the merger”, Brown Shoe Co. v. United States, supra, 370 U.S. at 335, 82 S.Ct. 1529, and a sound appraisal of the immediate and future impact of this merger on competition must be based on “a firm understanding of the structure of the relevant market”, United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 362, 83 S. Ct. 1741. The government, however, makes no attempt to separate the local from the national market, nor to define the competitive structure of either. Rather it commingles the geographic markets in reliance on a contention that New York City should be considered separate and apart from the rest of the *903 country in appraising the competitive effect of this merger. We have already found that the area of effective competition for retail accounts is not confined to the City but embraces the metropolitan area. All that remains for our consideration, therefore, is whether the geographic area of effective competition for so-called “wholesale accounts” is nationwide or confined to New York City. The government’s contention is based on the thesis that New York City is the nation’s financial center and that the demand for large business loans “centers on” New York’s billion dollar banks because interest rates are lower 101 and the banking needs of wholesale customers can be served adequately only by big banks. Therefore: (1) wholesale customers located entirely outside the metropolitan area tend to place their loans in New York; (2) wholesale customers located entirely inside the metropolitan area have easy access to the big banks of the nation’s financial center and find it unnecessary, impracticable, and inconvenient to go elsewhere for their credit needs; and (3) wholesale customers with an office 102 in the metropolitan area, but with plants and facilities elsewhere throughout the country, have easy access to the big banks of New York City and find it unnecessary, impracticable, and inconvenient to go elsewhere for their banking needs, regardless of their national standing, economic strength, or nationwide connections, operations, or commerce. Defendant contends that the area of effective competition for wholesale accounts is not confined to New York City but embraces the nation because wholesale customers are the cream of the banking business, and large banks of all the nation’s many financial centers aggressively compete with each other throughout the country for the patronage of wholesale customers regardless of the location of the bank or the customer. Moreover, defendant argues, wholesale customers are not restricted to their own locality for their banking needs but can, and do, turn to many competing financial centers throughout the country for banking services. Such customers, according to defendant, have the economic strength, national standing, and nationwide geographic connections which enable them to bank in a nationwide area and find it necessary, practicable, and convenient to do so. In support of its contention, the government first points to testimony that wholesale customers can be served adequately only by large banks and to evidence that the concentration of large business, financial, and government fiscal activities in New York has made it the traditional financial center of the na *904 tion. 103 There can be no question that New York City is the nation’s leading financial center, and there is no dispute that wholesale customers require the services of large banks. Nor can there be any question that New York’s big banks, like big banks everywhere, owe their existence to the massive banking needs of a big population, big business, big financial institutions, and big government. It by no means follows, however, that New York City is the nation’s only financial center or that the area of effective competition for the accounts of wholesale customers is confined to New York City. The government claims that surveys of the Federal Reserve Board prove that the demand for large business loans “centers on” New York. One survey, of 1955 vintage, shows the distribution of business loans throughout the nation’s twelve Federal Reserve Districts. It does not segregate the share of business loans held by New York City banks but shows the aggregate share of all banks within the much larger Second Federal Reserve District. Even if we overlook the discrepancy in geographic areas, 104 the survey more refutes than confirms the government’s thesis for it would still show that New York held only 20% of the customers, and 38% of the amount, of the nation’s business loans, and that banks located elsewhere throughout the country held the rest. Although New York’s share would be larger than that of any other single financial center and would progressively increase with the size of the loans, banks elsewhere throughout the country would hold a greater share than New York of all business loans up to $5,000,000 and one-third of the loans in the highest bracket— $5,000,000 and over. 105 An economist of the Department of Justice avers that New York’s share of large business loans has not changed in the past nine years. If so, the geographic area of effective competition, actual and potential, for such loans is clearly nationwide both in terms of competitors and customer alternatives. We think, however, that the government’s affiant overlooks material facts which show that the historical imbalance between New York and the rest of the country in the supply of, and demand for, money is largely a relic of a bygone age, and that New York is confronted with more extensive and powerful nationwide competition now than it was ten years ago. We take notice that New York City has not kept pace with the population and economic growth of other areas of *905 the country. 106 We note also that growth elsewhere has caused expansion of the nation’s banking system and bred other giant financial centers. Thus, in 1958, when Congress was studying the need for legislation to regulate bank mergers, there were 13,540 commercial banks in the United States, 107 but by the end of 1963 the number of banks had increased to 13,569 108 These facts indicate that the Bank Merger Act and the nation’s growth have combined to preserve and increase the number of viable commercial banking alternatives throughout the country. Thus, the notion that there is no room for newcomers and that ample nationwide banking alternatives are about to disappear would seem to be more the product of myth than fact. 109 Nationwide competition for large business loans cannot be measured, however, merely by counting the number of banks. Our search is for the area of effective competition for wholesale accounts, the relevant product. We are concerned, therefore, less with the number of banks than with the number of effective competitors for wholesale accounts and the vigor, intensity, and power of competition. United States v. Penn-Olin Chemical Co., supra, 378 U.S. at 177, 84 S.Ct. 1710; Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 329, 81 S.Ct. 623; United States v. Columbia Steel Co., supra, 334 U.S. at 527-528, 68 S.Ct. 1107. The evidence shows that there is a direct relationship between the size of a bank and its ability to attract, serve, and keep wholesale customers. As a practical matter, small banks are not effective competitors for wholesale accounts. 110 A bank, for proper reasons of diffusion of risk, seldom wishes to loan the full amount of its legal lending limit to one borrower. Conversely, a large borrower seldom wishes to confine his loans to one bank with a small lending limit because such a bank cannot meet his expanding or emergency credit needs. Smaller banks may be able to accommodate large borrowers by participation loans with other banks, or by the borrower’s maintaining multiple lines of credit. Wholesale loans or lines of credit are, therefore, often made under complex and cumbersome participation arrangements to which a number of banks are parties. Participation arrangements, however, are inconvenient and disadvantageous to a large borrower for they involve multiple credit appraisals, notes, and notifications, and the borrower loses the benefit of negotiating with more than one bank. The negotiations are conducted by the lead bank, and there is no competitive interchange either between the participant banks and the customer, or among the participating banks. The climate is, therefore, essentially non-competitive (DX 10). Most large business loans are made by large banks simply because their legal lending limits and uncommitted funds over required reserve balances are greater than those of their smaller competitors. Large banks, with more money, wider diffusion of risk, and greater specialization are also better able, and more prone, to take the risk involved in large loans. They can also meet the borrower’s expanding needs and handle the *906 transaction with much less inconvenience to the borrower. Moreover, with greater resources and wider relationships, large banks are more able to operate beyond their own locality. Finally, the borrower has the competitive advantage of playing one large bank against another in negotiating terms. Large banks are, thus, essential to serve the massive credit needs of an expanding population, industry, commerce, and government, 111 and, although large customers borrow much more from non-bank sources than from banks, big banks are still an important source of short-term credit required for working capital. These factors explain why throughout the country banks with deposits of $100 million or more account for about 95% of the loans to large borrowers. 112 The relationship of size of loan to size of bank, as well as the mobility of larger borrowers, is also shown in a 1957 survey of the Federal Reserve Board (PX 2). The survey shows that smaller New York City banks derived only 14% of the amount and 10% of the number of their business loans from customers located outside New York City, while the city’s billion dollar banks derived more than half the amount and 20% of the number of their business loans from outside customers and accounted for 90% of the dollar amount of loans by Second District banks to borrowers with assets of $100 million or more. New York, however, is not the only city with big banks, and, although it still retains the leading position among the nation’s credit markets, the gap between it and other financial centers is rapidly shrinking. Banks throughout the country, in response to postwar changes in the geographic and social distribution of wealth 113 and the dispersion and growth of population and business, have increased in size and extended the area of their operations beyond their own localities. These normal competitive and decentralizing forces have resulted in a steady decline in the proportion of resources held by banks in New York City and an increase in the proportion held by banks located elsewhere. 114 We notice that in 1940 the three largest commercial banks in the nation were located in New York City. The Bank of America, located in California, then *907 ranked fourth, but by 1947 it had passed over all the New York banks and has ever since held the lead. It gained a phenomenal 584% increase in deposits from 1940 to 1957. 115 The Bank of America climbed to the top as a retail bank. It has 846 branches sprawled throughout California, from Oregon to Mexico, more branches, incidentally, than all of the banks of the New York metropolitan area combined. While New York’s once staid wholesale banks began welcoming retail customers to bolster lagging deposit growth, 116 the Bank of America entered the wholesale market to relieve its bulging vaults. As California companies began buying and selling nationally and internationally, and as companies elsewhere began buying and selling in California, the bank followed them and extended its operations beyond California into the national and international areas. The fact that a bank among those most remote from New York City effectively competes for wholesale accounts, regardless of the customer’s geo-, graphic location, is clear from the Bank of America’s boast that it now has f<more of the top 100 corporations as customers than any other bank.” 117 The story has its counterparts throughout the country, though some states have a unit banking system which inhibits growth, and others, such as California and New York, permit statewide or limited branching which promote growth. In 1950 there were 18 commercial banks in the United States with deposits of $1 billion or more, and 9 of them were located outside New York City. 118 By 1960, the year of this merger, the number of such banks had increased to 22, while the number outside New York City increased to 14. 119 Since this merger, the number of such banks has grown to 31, and 24 of them are located outside the City of New York. Thus, in the last thirteen years, the number of billion dollar banks outside New York has almost tripled. As a result, there are now one or more billion dollar banks located in San Francisco, Los Angeles, Seattle, Portland (Oregon), Dallas, Chicago, Detroit, Cleveland, Buffalo, Pittsburgh, Philadelphia, Boston, and Mineóla. 120 Manifestly banks are flourishing and entry into the billion dollar class is wide open in every region of the country. 121 In addition there are 43 other outside banks pushing for admission into the billion dollar class with deposits ranging from $500 million to $987 million, and 125 more with deposits from $179 million to $496 million. These facts overwhelmingly support the F.D.I.C.’s observation that: “For large business concerns it seems quite probable that there has been an increase during the past 40 years in the competition among banks for their business. That is to say, the larger banks, mostly located in large cities, complete with each *908 other for the patronage of concerns doing a nationwide business. With the greatly increased facilities of transportation and communication of recent years, there is more competition of this sort now than formerly, regardless of the changes which have occurred in the number of banks or the number and location of banking offices.” 1960 F.D.I.C. Ann.Rep. 60 (emphasis added). 122 Indeed, the number of nationwide competitors is so great, and the vigor of competition for wholesale accounts so intense, that it has forced large New York City banks to charge lower interest rates on loans and to suffer a lower ratio of net earnings to capital than banks located in other areas in order to remain competitive. 123 That outside competition has an effective influence on New York, and vice versa, is further demonstrated by evidence that all changes in the prime interest rate since 1947 and all reductions, without exception, have been initiated by New York banks, the last by Manufacturers in I960. 124 Nevertheless, New York’s big banks have not kept pace with the growth of their big outside competitors. 125 Indeed, prior to the merger, both *909 Manufacturers and Hanover lagged behind the growth of the country’s 100 largest banks. 126 The large outside banks prove the existence of effective nationwide competition and the sensitivity of the nationwide market to what happens in New York by meeting the prime rate, which tends to set the level for all rates, and remaining competitive without initiating reductions. As a result, large borrowers can obtain the same rate from large banks throughout the country. 127 It seems clear that regional credit markets are interrelated, that large sophisticated borrowers and depositors shop around, and that competitive changes in one region have a direct and immediate impact in the others. In short, money has no home but, spurred by the urge to propagate, ebbs and flows from one area to another in response to opportunities for greater yield. The interrelationship of the nation’s credit markets is explained in a recent publication of the Federal Reserve Board: “Creditmarkets * * * are * * * closely linked through the activities of borrowers, lenders, and investors. As these groups seek the most favorable opportunities for borrowing or for investment of available funds, they may find it advantageous to move from one market to another. “From a geographical standpoint, the national credit market is made up of a large number of regional and local credit markets. The rate of interest charged and other conditions in these regional or local markets may vary, but these rates and markets are nonetheless related through conditions affecting both the supply of and demand for credit.” 128 Thus, the evidence establishes that there are numerous large and growing banks spread throughout every region of the country and that they operate and compete, actually and potentially, outside their own locality, in a nation-wide area, for the accounts of large customers. Indeed, many of them have offices in New York City. Plainly the banks competing effectively for such accounts are not confined to New York, but are located throughout the country in many thriving and interrelated financial centers. We conclude, therefore, that the government’s basic premise that banks do not compete outside their own locality and that effective competition for wholesale accounts is confined to banks located in New York City is outdated and unsupported by contemporary facts. As we have seen, however, definition of the geographic boundaries of a relevant market depends not only upon the area where the parties to the merger and their competitors operate and compete, but also upon the area to which customers “ ‘can practicably turn for supplies’ ” and the place “where, within the area of competitive overlap, the effect of the merger on competition will be direct and immediate.” United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 357-359, 83 S.Ct. 1738. The banking agencies and defendant’s experts advise that large borrowers do turn to a nationwide market for bank *910 credit. According to the Federal Reserve Board: “While local markets handle most of the relatively small loans originating from local needs and based on local conditions, regionally or nationally known concerns, whose borrowings involve large sums, obtain most of their credit in a broader, even nationwide market. The changing allocation of their borrowing demand, region by region, in response to changing financial conditions helps keep interest rates in fairly close alignment. “In such ways geographically separated markets are linked in a broad national market. If lendable funds are scarce and costly in one center, the local supply will tend to be augmented by an inflow from centers where funds are more abundant and less costly. As a result, well established borrowers with a high credit rating can obtain loans from banks or others, on much the same conditions in one city as in another. There are many regional credit centers — such as Chicago, Boston, San Francisco — but the largest share of the nation’s credit and money market business is transacted in or through New York City.” 129 The banking agencies, particularly the New York Superintendent and the F.D. I.C., also inform us that there are a considerable number of commercial banks aggressively competing for deposits, both locally and nationally, that the market for large deposits is national “to a considerable degree,” and that large depositors do have nationwide deposit alternatives. Defendant’s experts testified to the same effect, and there is no evidence to the contrary. As we have seen, deposits constitute the major source of a bank’s working capital. Self-interest thus impels banks to seek all the deposits they can get. In effect, banks “buy” deposits but “sell” loans. Depositors are thus in a sellers’ market. We notice that in 1955 the New York State Banking Department investigated the types of business done by large commercial banks in New York City. It investigated their customers, the location of their plants and offices, their banking habits, and their financial practices. The department concluded that “the customers of the large banks in New York City were, for the predominant part, national corporations. Their factories and offices were located in every state of the Union. Their cash balances were distributed among banks all over the country, especially in other financial centers which competed vigorously with Manhattan for a greater share of deposits and credit lines.” 130 That conclusion is fully supported by the evidence presented here, as we shall see. The placement of large demand deposits, however, is predicated, according to the F.D.I.C. and defendant’s experts, on lending accommodations, and, according to the Federal Reserve Board, loans are the primary source of demand deposits because in practice loans are credited, at least temporarily, to the borrower’s deposit account. 131 Moreover, large loans are often conditioned upon the borrower’s maintaining compensating balances. This interrelationship between large loans and large deposits indicates that the area of effective competition is the same for both. Defendant relies heavily upon its experts and the opinions of the banking agencies in support of its contention that the relevant geographic market for wholesale accounts is nationwide. The banking agencies’ publications support defendant’s contention, but their reports on this merger conclude that the relevant geographic market for certain wholesale accounts is partly local and partly na *911 tional. That conclusion, we think, is understandable, but erroneous as a matter of law. In the applications before the banking agencies, received here either as stipulated facts or expert testimony, 132 the applicant, focusing not on the geographic area of effective competition, but upon the fact that banks located inside and outside the metropolitan area competed for wholesale customers located both inside and outside the area, mistakenly claimed, as it still does in its analysis of each category of the accounts of the constituent banks, that the relevant geographic market for certain types of wholesale accounts 133 is partly local and partly national. The Attorney General, in his report to the Federal Reserve Board, took the same equivocal position for the same erroneous reasons. The federal banking agencies, misguided, agreed. Neither the applicant, nor the Attorney General, nor the banking agencies separated what they conceived to be the local part of wholesale accounts from the national part. Rather, all commingle the product and geographic markets, and both parties repeat that root error here. We, of course, are not bound by the parties’ claims, nor the opinions of defendant’s banking experts, nor those of the Attorney General, nor of the banking agencies respecting the relevant geographic markets. Definition of the market is a typical antitrust problem which requires not the expertness of bankers, economists, or banking agencies but that of the court. Pan American World Airways, Inc. v. United States, supra, 371 U.S. at 330, 83 S.Ct. 489 (dissenting opinion); Brown Shoe Co. v. United States, supra, 370 U.S. at 368, 82 S.Ct. 1548 (Harlan, J., concurring in part and dissenting in part). Cf. United States v. Radio Corp. of America, supra. Accordingly, we reject all “expert” conclusions defining the relevant geographic market for wholesale accounts first because, except for the Federal Reserve Board, 134 it is superfluous lay opinion on a complex economic-legal problem, United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 367, 83 *912 S.Ct. 1715; 7 Wigmore, Evidence § 1918, at 10, § 1952, at 81 (3d ed. 1940), 135 and second because the opinion or contention that the geographic market for wholesale accounts is confined to the City of New York as the government urges, or partly local and partly national respecting some wholesale accounts, as defendant and the Attorney General led the banking agencies to conclude, is erroneous as a matter of law and, as we shall see later, leads to unsound conclusions. Expert opinion on facts about the banking business, however, is a different matter. Experts in the banking field are manifestly competent to express an opinion on the competitive realities of the industry. Surely the agencies, authorized to pry into confidential matters and competitive secrets, know more about who is banking with whom than we can ever hope to learn from the fragments of evidence before us. 136 They assure us that large banks of the nation’s many large cities aggressively compete with banks outside their own locality for large accounts, and that large customers do turn to banks outside their own locality, to a nation-wide area, for banking services. Defendant’s experts, all experienced bankers or economists of unquestioned qualifications, gave the same opinion. This is indeed persuasive evidence that effective competition for wholesale accounts is nationwide, for the government offers no expert or other evidence to the contrary. Nor did it in any way impair the testimony of defendant’s experts on cross-examination. No reason appears for doubting the accuracy or credibility of their testimony. “The existence of competition is a fact disclosed by observation rather than by the processes of logic; and when these officers, skilled in the business which they have carried on, assert that there * * * [is] real competition in respect of the particular product, their testimony is to be weighed like that in respect of other matters of fact. And since there is no testimony to the contrary and no reason appears for doubting the accuracy of observation or credibility of the witnesses, their statements should be accepted.” International Shoe Co. v. FTC, supra, 280 U.S. at 299, 50 S.Ct. 91. Instead of refuting the expert opinions offered by defendant, confirmed by the agencies, and supported by substantial objective evidence, the government jumps from evidence that many wholesale customers have an address or are physically located wholly or partially inside the City of New York to the conclusion that the geographic area of effective competition for wholesale accounts is confined to New York City. *913 The government then seizes on a wholesale customer’s geographic location where it favors its definition of the relevant market and ignores that factor where it works to its disadvantage. Thus, it seizes on geographic location to localize competition for large concerns with an address or physical location entirely or partially within the City of New York but ignores geographic location to sweep outside customers and those with multiple geographic locations into the local market. 137 Similarly, looking to the location of competitors and customers, defendant commingles the geographic markets by claiming that the market for seven categories of wholesale accounts is partly local and partly national. We think, however, that both parties and all the agencies have labored under a mistaken concept of a relevant geographic market under the antitrust laws. The phrase “any section of the country,” contained in Clayton § 7, refers not to a definite geographic area of the country but to the geographic area of effective competition in the relevant line of commerce. Brown Shoe Co. v. United States, supra, 370 U.S. at 320 n. 35, 82 S.Ct. 1502. The proper question to be asked, therefore, is not where is the customer located, but what is the geographic area of effective competition for his patronage. When the area of effective competition is confined to the customer’s own locality, as is the case with retail customers, so is the relevant geographic market, but when the area of effective competition embraces the nation, as is the case with wholesale customers, so does the relevant geographic market. Thus, in order to analyze the impact of this merger in true perspective, we must direct our quest to the geographic area in which competitors operate and to which wholesale customers can practicably turn for the relevant banking service. Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 327, 81 S.Ct. 623. The geographic boundaries of the area of effective competition must, therefore, be drawn broadly enough to include the location of all effective competitors, actual and potential, and the location of all customers, actual and potential, who can practicably turn to them for the relevant product. Cf. United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 599-600. The evidence shows that large banks of the nation’s large cities do reach outside their own locality to a nationwide area, for wholesale customers and that wholesale customers do choose banks outside their own locality and conduct their banking business throughout the nation. The relevant product, wholesale banking service, is not a heavy commodity where differentials in freight costs and delivery intervals make proximity of supplier and customer essential to effective competition, 138 as in Crown Zellerbach Corp. v. FTC, 296 F.2d 800 (9th Cir. 1961), cert, denied, 370 U.S. 937, 82 S.Ct. 1581, 8 L.Ed.2d 807 (1962) or American Crystal Sugar Co. v. Cuban-American Sugar Co., 259 F.2d 524 (2d Cir. 1958). Nor does the factor of inconvenience localize effective competition for wholesale accounts. The relevant product in the national market is an intangible— large sums of money in the form of credits and debits. The product is capable of convenient, cheap, and instantaneous transmission from one bank to another, or from one customer to another, in any part of the nation or the world simply by effecting debits or credits to the wholesale customer’s account by telegraph, cable, or telephone. Indeed, the government admits that the constituent banks and their customers regularly used the telephone and telegraph for that very purpose (Complaint, para. 14). Surely such facilities! are also used by all *914 large banks and wholesale customers located in large cities throughout the country. 139 The most distant bank is therefore no further away and no more inconvenient to a wholesale customer than the nearest telephone, telegraph office or mail box, as the complaint acknowledges. 140 The time dimension of money makes instantaneous transmission of large loans and deposits from one section of the country to another imperative, otherwise large sums would be tied up in float, with interest running and working capital idle, while checks are in transit through the mails. 141 Consequently, large concerns find it both convenient and necessary to channel their receipts into post office boxes, and thence to a bank acting as its agent in the locality of collections, and to transmit disbursements by bank wire to banks located in the area where money is needed to cover payrolls and conduct operations and trade. The nature of the product, mobility of demand, nationwide commerce, and sophisticated customers thus impel large banks and large customers to do business with each other at a distance, in a nationwide area. That fact favors a choice of a nationwide market for wholesale accounts. Brown Shoe Co. v. United States, supra, 370 U.S. at 328, 82 S.Ct. 1502. There is no dispute that both Manufacturers and Hanover did a large amount of business throughout the United States with large customers far removed from the metropolitan area. Nor is there dispute that large customers far removed from the area, and located in every region of the country, borrowed large sums and made large deposits at both banks. 142 That banks located close to such customers effectively compete against the New York intruders for wholesale accounts is supported by the very survey upon which the government relies, showing the distribution of the nation’s business loans (PX 2), and by evidence here that over 83% of Manufacturers’ and over 91% of Hanover’s commercial and industrial borrowers with *915 loans over $100,000, who were located entirely outside the metropolitan area, also had loans or lines of credit with outside commercial banks (DX 62(c)). Likewise, as we have seen, there are numerous banks throughout the country with resources great enough to drive and enable them to solicit outside their own locality for wholesale accounts. 143 It is not surprising, therefore, that in wholesale banking the metropolitan area is by no means the exclusive domain of New York banks. The banking agencies and defendant’s experts tell us that large banks located in the nation’s other financial centers aggressively compete for wholesale accounts within the metropolitan area and that many of the largest outside banks have offices in the City. The uncontradicted testimony is that banks located elsewhere throughout the country and the world solicit and make loans to large concerns located in the metropolitan area, either directly or through agencies or New York City money brokers. That testimony is corroborated by evidence that 41% of Manufacturers’ and nearly 60% of Hanover’s commereial and industrial borrowers of $100,000 or more, who have offices within but plants and facilities' outside the metropolitan area, also had loans or lines of credit with banks located outside the metropolitan area (DX 62(b)). That was also true of over 8% of Manufacturers’ and over 6% of Hanover’s borrowers of such an amount who were located entirely inside the metropolitan area (DX 62(a)). 144 Thus, the obstacle of customer inconvenience, claimed by the government to localize every customer’s choice of a bank 145 and bar outside competitors, has been successfully hurdled with respect to wholesale customers not only by Manufacturers and Hanover but also by banks located outside the metropolitan area. Large outside banks have penetrated New York, and large New York banks have operated throughout the country. Each has been able to enter the locality of the other and compete effectively with distant commercial banks for the patronage of distant wholesale customers. Manifestly wholesale customers, irrespective of their geographic location, have turned to distant *916 banks, outside their own locality, for part of their banking needs. Indeed, the Supreme Court recognized in Philadelphia, 374 U.S. at 360, 83 S.Ct. 1740, that “[l]arge borrowers and large depositors * * * may find it practical to do a large part of their banking business outside their home community * * We find, therefore, that large customers do not confine their patronage to banks in their local community, as the government contends, and that they do find it practical and convenient to conduct a large part of their banking business at a distance, in a nationwide area. A relevant geographic market cannot be defined, however, solely on the basis of where the constituent banks and their competitors have actually done business, or even where customers have actually turned for their banking needs. The market must be drawn also on the basis of potential competition. Where could the banks do business, and where could customers practically turn for alternative sources of supply? United States v. Penn-Olin Chem. Co., 378 U.S. 158, 174, 84 S.Ct. 1710, 12 L.Ed.2d 775 (1964); United States v. El Paso Natural Gas Co., supra, 376 U.S. at 661, 662, 84 S.Ct. 1044; United States v. Bethlehem Steel Corp., supra, 168 F. Supp. at 599. As we have seen, banks located inside and outside the metropolitan area are continually bidding in each other’s locality for wholesale customers. This actual and potential competition restrains both the metropolitan banks and their outside competitors from charging excessive interest rates and operates to keep interest rates in the nation’s financial centers in close alignment. The mere presence of numerous strong competitors, aggressively pushing into each other’s locality, exerts a powerful force which must be reckoned with in all of the nation’s financial centers, even though outside competitors never win inside wholesale customers and although inside wholesale customers never turn to outside banks. United States v. Penn-Olin Chem. Co., supra, 378 U.S. at 174, 84 S.Ct. 1710. “Unsuccessful bidders are no less competitors than the successful one. The presence of two or more suppliers gives buyers a choice.” United States v. El Paso Natural Gas Co., supra, 376 U.S. at 661, 84 S.Ct. 1049. A for-tiori, when outside competitors have successfully made inroads and are continuously bidding for more customers in the local bank’s territory, there is an insistent, constant, effective, and substantial competitive confrontation between the inside bank and the outside bank. Each is attempting to expand its business at the other’s expense. Each is thus engaged in a struggle with the other to hold old customers and win new ones. Each poses a constant threat to the other’s wholesale business, and in an integrated country, spanned in seconds by telephone and in hours by jet plane, the threat is real. Likewise, if inside wholesale customers have in fact turned to outside banks, others are readily able to do so if the inside bank is unable or unwilling to satisfy their needs on satisfactory terms. 146 Potential competition thus operates openly and subtly to prevent both inside and outside competitors from charging excessive prices or exacting onerous conditions and provides a bargaining lever for wholesale customers. Cf. United States v. Penn-Olin Chem. Co., supra, 378 U.S. at 174, 84 S.Ct. 1710. That this is no fanciful theory but a competitive reality in wholesale banking is demonstrated by the fact that the potential, if not the actual, competition of a distant bank acts as a ceiling or control on the interest rates, terms, and conditions which a bank may impose upon wholesale borrowers located in its home area. The Federal Reserve Board tells us that regional credit markets for wholesale accounts are in *917 terrelated, and none is so separate that the large customers and large competitors of one region are not affected by the price and the supply of, and demand for, credit in the others. Plainly, interest rates in one region affect interest rates in the others, and large borrowers move from one region to another in response to changing costs. Thus, competitive happenings in one region have a direct and immediate impact on wholesale banking throughout the country. That fact, we think, favors a choice of a nationwide market for wholesale banking, for “the geographic market for the purposes of determining the impact of a merger can include all areas where the trade in a product is affected by, and is not independent of, the trade in that product in other areas — for example, if a change in price in one area has an effect on price in another area both areas may be included in one geographic market.” United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 599-600; S.Rep. No. 1775, 81st Cong., 2d Sess. 5-6 (1950). The fact, moreover, that Manufacturers and Hanover were substantial alternative sources of supply for wholesale customers throughout the nation also favors the choice of a nationwide market for wholesale accounts. United States v. Bethelehem Steel Corp., supra, 168 F.Supp. at 601. We are taught that “the boundaries of the relevant market must be drawn with sufficient breadth to include the competing products of each of the merging companies and to recognize competition where in fact, competition exists.” Brown Shoe Co. v. United States, supra, 370 U.S. at 326, 82 S.Ct. 1524. Finally, the market must not be drawn “so narrowly as to place * * * [the parties to the merger] in different markets, because only the smallest customers are considered.” United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 361, 83 S.Ct. 1740. We recognize, however, that the national market for wholesale accounts is not uniform. Populous regions and those with a highly developed economy obviously have greater need for wholesale banking services than sparsely settled or underdeveloped areas. We realize also that the impact of this merger on wholesale banking may not be uniform throughout the country. It obviously has a greater significance in the nation’s financial centers than elsewhere. That fact, however, does not preclude our choice of a nationwide market. Cf. Brown Shoe Co. v. United States, supra, 370 U.S. at 337-338, 82 S.Ct. 1502. Nor do we see any reason for separating the New York metropolitan area from the balance of the country merely because the constituent banks were located in New York, or because many wholesale customers are located there, wholly or partially, or because the area is the financial capital of the nation and may now have a larger share of the nation’s wholesale banking business than any other single financial center. Those facts, we think, all favor the choice of a nationwide market. Furthermore, as we have seen, New York’s share is steadily declining, and the trend is clearly against the City with the disparate growth rate of the rest of the country. That fact bears on our prediction of the future competitive effect of this merger in judging its validity under Clayton § 7. Finally, as we have seen, the massive and elaborate underlying economy of the metropolitan area and the wholesale banking industry built upon it is so integrated with, interrelated to, interdependent upon, and enmeshed with the rest of the nation that wholesale banking in the area and in the rest-of the country are inseparably linked together. Each segment is dependent upon and influenced by what happens in the other, and both should be analyzed together, as one nationwide market. The foregoing considerations demonstrate, we think, that the entire nation is the place where, within the area of competitive overlap, this merger is likely to have a direct and immediate impact, if it has had any impact at all, on whole *918 sale banking. Cf. United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 601. Accordingly, we conclude that the relevant geographic area of effective competition for all wholesale accounts is the entire United States, regardless of the customer’s location. Thus, we must deal with two geographic markets and the problem of ascertaining the competitive effect of this merger in each. We turn, therefore, to the allocation of the accounts of the constituent banks to their relevant geographic market. J. Geographic Allocation of the Accounts of the Parties to the Merger. The government makes no attempt to allocate accounts, but relies on its one market theory and proposes that we separately analyze the quantity of competition previously existing between the constituent banks in certain banking services 147 which it arbitrarily selects as product submarkets. Defendant contends that there is neither authority nor proof upon which to base a determination that the government’s selected categories of services are separate lines of commerce. We agree. Defendant proposes that we analyze all categories of services separately and measure the degree of competition between the constituent banks and the effect of its elimination on all competitors in both markets. Both sides present a maze of statistics to prove their contentions. Those offered by the government commingle the geographic markets and dissect selected categories of deposits and loans. Those offered by defendant dissect deposits and loans of all categories of banking services and homogenize the geographic markets in some, but not in other, categories. We have examined the maze thoroughly and find neither position sound. Both obfuscate analysis and would lead us to serious error. The government’s premise that a look at a selected part reveals the whole is specious. It would have us consider only the debit side of the ledger and ignore the fact of two geographic markets. Obviously, if we ignore the many and substantial categories where competition between the constituent banks was either non-existent or comparatively minimal and focus only upon those categories where competition was substantial, we will reach the distorted conclusion that the activities of the banks were substantially competitive and not complementary. In short, the government cannot gerrymander the market any way it chooses. Likewise defendant’s premise is specious. While we agree that a valid conclusion can only be reached by looking at the whole picture, we must recognize meaningful competition where competition, in fact, exists. Cf. United States v. Continental Can Co., supra, 378 U.S. at 453, 84 S.Ct. 1738; Brown Shoe Co. v. United States, supra, 370 U.S. at 326, 82 S.Ct. 1502. Defendant cannot, on the one hand, insist that national business be subtracted from total deposits and loans before assessing the competitive impact in the local market and, on the other, ask us to fragmentize the deposits and loans of the constituent banks and homogenize the markets in some, but not in other, categories for the purpose of determining the degree of competition formerly existing between the parties to the merger. In short, defendant can neither have it both ways, nor scramble the markets any way it chooses. The competitive effect of the merger must be judged with regard to the fact that both banks competed in varying degrees in retail and wholesale banking and operated in two geographic markets. If the product and geographic markets are scrambled, as defendant urges, or selected services homogenized and analyzed while others are ignored, as the *919 government urges, probable anticompeti-tive effects in one market might be offset by pro-competitive consequences in the other, leading us to a false conclusion that on balance the merger may, or may not, restrain trade, substantially lessen competition, or tend to monopoly in either market. Such an approach to the problem is fundamentally unsound and was definitively rejected by the Supreme Court in United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 370, 83 S.Ct. 1715. That error lies at the root of the banking agencies’ conclusions and on the doorstep of both litigants. Thus, we find neither the government’s nor the defendant’s approach helpful to analysis of the competitive effect of this merger in true perspective. We must therefore allocate the accounts of the constituent banks to their respective geographic markets for the purpose of fashioning a tool for separate analysis of the competitive effect of the elimination of former competition between the constituent banks in each of the markets. Obviously we cannot allocate accounts to their respective geographic markets unless we can identify and separate the accounts subject to nationwide competition from those subject only to local competition. The banking agencies and defendant’s experts, in giving their opinions on whether competition for an account is local or nationwide, relied not only on their own observations and expertise, but supported their views with substantial evidence consisting of economic data relevant to the size and strength of the customer, the nature of his commerce, the location of his offices, plants, and facilities, his net worth and credit standing, the liquidity of collateral security, his actual banking relationships, and the presence or absence of nationwide competition. In applying the economic evidence, the experts and the agencies used norms such as local firms, nationally-known concerns, firms doing a nationwide business, firms of national stature, wholesale customers, retail customers, large customers, small customers, large loans or deposits, small loans or deposits, established borrowers, borrowers with a high credit rating or substantial net worth, prime borrowers, etc. We have examined the list of large customers, and it is fair to say that it is a roster of “Who’s Who” in American industry. The government, on cross-examination of defendant’s experts, established only the obvious fact that the norms employed are imprecise and that there is no definite criterion for differentiating large from small, local from national, retail from wholesale, etc. Vague terms, however, are not meaningless. They are useful so long as they mark some useful distinction. Those employed by the agencies and experts are all ordinary terms used to express ideas in common usage and understanding and certainly are no fuzzier than norms applied by courts and juries every day in giving concrete expression to imprecise concepts. Indeed, the Supreme Court itself speaks in terms of large and small customers, local and nationally-known concerns, etc., in both Philadelphia and Lexington. Defendant has presented voluminous statistics and economic data, like that considered by the agencies, all designed to prove that large customers have such economic strength that as a practical matter they can, and do, turn to a nationwide, and in some cases worldwide, area for their banking needs. There is merit to defendant’s basic contention that large customers do find it practical and convenient to bank where they choose. Evaluation of such evidence and application of such norms, however, to the problem of differentiating local from national accounts lies not in the domain of the court but in the province of the banking agencies, for the task requires experienced observation of the competitive realities of the banking industry, subjective economic judgments, inferences and interpretations based on expertise, nice discretionary choices, specialized knowledge, skill and competence, practical experience, intimate familiarity with the industry, and, in part, upon facts ob *920 served or known to the agencies and experts, outside this record. Courts are neither equipped to embark on the broad economic inquiry essential to the unaided application of such imprecise standards, 148 nor competent to second-guess the experts. That, of course, is the compelling reason for invoking the doctrine of primary jurisdiction in this case. 149 We can solve the problem of allocating the accounts of the constitutent banks, however, without relying on the doctrine. Our concern is not so much with whether an account is large or small, retail or wholesale, etc., but with the definite and observable fact of whether or not banks encounter effective nationwide competition for a particular kind or size of an account. Surely, at the very least, experts in the field are competent to express an opinion on such objective banking facts based on their observations and experience. International Shoe Co. v. FTC, supra, 280 U.S. at 361, 50 S.Ct. 89. Indeed, the findings of the Federal Reserve Board on the subject may well be conclusive under our ad hoc application of the doctrine of primary jurisdiction, for they were clearly based on substantial evidence. 15 U.S.C. § 21(c) (Supp. Y, 1964). Cf. Whitney Nat’l Bank v. Bank of New Orleans, supra. Although the Board’s “findings” are inex-plicit, they may be implied from the evidence before it, including the reports of the other state and federal banking agencies, and from the Board’s conclusion that Hanover was almost exclusively a wholesale bank engaged in national operations. The banking agencies and defendant’s experts have thus given us workable guidelines for differentiating local from national accounts. If their demarcation lines are debatable, the government offers no debate. It tenders no expert opinion, nor did it cross-examine defendant’s experts on the subject of whether New York banks do encounter effective nationwide competition for particular accounts. The experts thus stand uncontradicted and unimpeached. Their opinions, we find, were neither arbitrary nor capricious, but were rationally based on sound reasoning and substantial concrete evidence which we do not stop to detail. 150 Imperfections are inherent in the problem, but the antitrust laws do not demand mathematical certainty in the solution of complex and elusive socio-economic problems. Rather, workable compromises based on substantial evidence and rational judgments suffice. Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 361, 83 *921 S.Ct. 1715. The agencies’ opinions were given in the impartial discharge of their duties to the public, and there is no reason why we should not accept their opinions wherever helpful to our analysis and every reason why we should. 7 Wig-more, Evidence § 1923, at 21 (ed. 1940); International Shoe Co. v. FTC, supra, 280 U.S. at 29, 50 S.Ct. 89; cf. Administrative Procedure Act, 5 U.S.C. § 1009 (e) (1958); Whitney Nat’l Bank v. Bank of New Orleans, supra; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 361, 83 S.Ct. 1715. We think sensible accommodation of court and agency permits nothing less if this six-headed monster, 151 created to regulate bank mergers, means anything other than senseless feuding and a collossal waste of time, effort, and money. Finally, as a practical matter, we must either resort to expert help in fashioning an analytical tool or ignore the competitive reality that this case involves both a local and a national geographic market. Accordingly, we find, on the basis of the foregoing analysis of the relevant geographic markets, the uncontradicted expert opinions, and other evidence and stipulations before us that the accounts of the constituent banks are allocable to the local and national markets, as follows: Local Market Special Checking Accounts Demand Deposits under $100,000 Savings Accounts, including Christmas Club Accounts Consumer Installment Loans Commercial and Industrial Loans under $100,000 Real Estate Loans on Properties Located in the Metropolitan Area Single Payment Loans to Individuals under $25,000 Voluntary and Court Trusts Stock Transfer Agencies and Regis-trarships for Stocks Listed on the New York Stock Exchange National Market Demand Deposits over $100,000 Time Deposits other than Savings Accounts Commercial and Industrial Loans over $100,000 Loans to Financial Institutions Loans to Brokers and Dealers to Purchase or Carry Securities Loans to Others to Purchase or Carry Securities Real Estate Loans on Properties Located Outside the Metropolitan Area Single Payment Loans to Individuals over $25,000 Agency and Custody Accounts Trusteeships for Securities Issues Trusteeships for Pension and other Employee Welfare Funds Stock Transfer Agencies and Regis-trarships for Stocks Listed on the American Stock Exchange Dividend Disbursing Agencies Bond and Coupon Paying Agencies A detailed computation of the number, amount, and percentage of the deposits and loans of Manufacturers and Hanover which are allocable to the local and na *922 tional markets is set forth in Appendix A. We conclude that 50.37% of the amount of the deposits and 24.10% of the amount of the loans of Manufacturers are allocable to the local market. The comparable percentages for Hanover are 18.34% and 4.34%. The complementary-percentages allocable to the national market are 49.63% of the deposits and 75.90% of the loans of Manufacturers, and 82.68% of the deposits and 95.66% of the loans of Hanover. Having determined the portion of Manufacturers’ and Hanover’s business allo-cable to each geographic market, we turn to the parties’ contentions respecting market shares. K. Market Shares. We note at the outset that at the administrative stage neither the applicant, the Attorney General, nor any of the banking agencies made any attempt to calculate market shares in either geographic market, although all recognized that there were at least two markets. Rather, the market share statistics on which all relied respecting relative size, rank, and concentration commingled the product and geographic markets and were, therefore, spurious. The banking agencies, following the approach of the application, avoided the problem of calculating market shares by considering each category of banking service separately and analyzing the competitive effect of the merger with respect to each category on the basis of whether competition for the size and type of account was local, national, international, or partly local and partly national, as defendant claimed as to some accounts. The government here avoids the problem of separating the geographic markets and calculating market shares in each of them by reliance upon its theory that New York City should be treated separately from the rest of the country. We have rejected that contention both as to the local and national markets. Nevertheless, we set forth the government’s figures in order to deal with the parties’ contentions. Relative shares immediately before and after the merger, according to the government’s one market theory, if applied to the metropolitan area, are detailed in Appendix B and are as follows: There is nothing wrong with the government’s arithmetic, and its one market approach has the appeal of simplicity, We find, however, that we cannot float *923 downstream with the government’s easy solution for its raft is too fragile to bear its burden of proof. The government’s simplified case is specious. Its market is artificial. The uncontradicted evidence is that wholesale accounts originate in a nationwide market — the geographic area of effective competition — and not in the local market. Ignoring that evidence, the government creates its market by equating the local geographic area with the relevant market for wholesale accounts. Its equation is constructed by substituting the fallacy that the bulk of wholesale accounts originates in the local market area for the fact that most wholesale customers have some sort of physical location or address wholly or partially within the local geographic area. The switch allows the government to draw its facts from its conclusion and then circle back and use them to lay a foundation for its market. The government then erects a superstructure of statistics which appears to represent market shares in one geographic market, but which actually homogenizes the deposits and loans of two. The government’s calculation of market shares thus rests on a fiction, blinks the evidence, and ignores competitive realities. The picture which emerges definitely exaggerates the local market shares of the larger wholesale banks and does not depict the true local market shares of large integrated banks which may be greater or less than the shares shown in the government’s statistics for, as we have seen, the big integrated banks like Manufacturers originate a vast but varying portion of their business in both geographic markets, while small banks originate practically all of theirs in the local market. Thus, the larger the bank the greater the likelihood of distortion, and the smaller the bank the lesser the distortion. As a result, the government’s statistics are spurious, but we will work with them bearing the distortions in mind whenever they are helpful to analysis. Defendant, relying on Philadelphia, urges us to “shade” the “shares” of the constituent banks by subtracting the amount of their national business from their total deposits and loans in order to arrive at their true shares of the local market. Defendant suggested on oral argument that its share should be shaded by about 40% to exclude customers located outside the metropolitan area and those with multiple geographic locations. Defendant had no suggestion as to what other banks should receive similar treatment or how much their shares should be shaded. Nor has defendant attempted a mathematical calculation of market shares in either the local or the national market in accordance with its contention. Rather, it repeats its analysis category by category, as in its application to the banking agencies, and in the process commingles the geographic markets in seven categories of accounts. We reject the suggestion that defendant’s apparent local market share be shaded by 40% for it is premised on an erroneous concept of the relevant geographic market. Likewise we reject its method of analyzing each category of accounts separately because it commingles the geographic markets. Thus, neither party’s analytical tools are fashioned with fidelity to competitive realities, both obfuscate analysis, and, as we have seen, both led the banking agencies and would lead us to a fundamental error. Defendant’s failure to calculate local market shares according to its basic and sound contention that national and international business (hereafter called national business) should be shaded from the local market is understandable for, as we shall see, it is impossible for lack of evidence. It is too plain for argument that national business must be excluded in calculating local market shares, and local business excluded in calculating national market shares, otherwise our analysis of the competitive effect of this merger would be predicated upon a false picture leading us to erroneous conclusions based upon a complete misunderstanding of the competitive structure of the relevant geographic markets. United *924 States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 364 n. 40, 83 S.Ct. 1715. Determination of the local market shares of the constituent banks, however, is not a simple matter of subtracting what we have determined to be the percentages of their deposits and loans which are allocable to the national market from their total deposits and loans. Manifestly the true local market is the total deposits and loans of all the banks located in the metropolitan area, less the aggregate of their deposits and loans which originate in the national market. 152 We cannot deduct all national business from the area’s total deposits and loans, however, without knowing the percentage of business which each local bank originates in the national market. Except for the constituent banks, there is no evidence before us on that subject. 153 In *925 deed, in the main, this case was presented by the government as though Manufacturers and Hanover operated in a world by themselves. It is, therefore, impossible to ascertain the local market share of any bank located in the metropolitan area with certainty, although we can, and later will, estimate Hanover’s relative position in the local market. Likewise, indefinite evidence respecting the number and identity of national competitors, and no evidence whatever showing the portion of each competitor’s deposits and loans which originates in the national market, make it impossible to ascertain market shares in the national market. The number and identity of the national competitors were properly the subjects of expert opinion, but none was presented. Our concern, however, is not with certainties but with reasonable probabilities. 154 We must, therefore, take a pragmatic approach and fashion some tool in order to test the validity of the government’s claim that this merger offends the antitrust laws in the national market. The evidence before us on which of the nation’s 13,000-odd commercial banks compete effectively in the national market is that large businesses of necessity borrow primarily from large banks and only to a limited extent from small banks, 155 that banks with deposits of $100 million and over account for about 95% of the loans to large borrowers, 156 that deposits track loans, that the large banks of the large cities compete in a nationwide area for wholesale accounts, 157 and we take notice that something over 200 banks in 60 large cities are linked together by a nationwide bank wire. 158 We cannot iden *926 tify the effective competitors in the national market with perfect accuracy, but we think that the country’s 200 largest banks roughly fit the description. 159 The distribution of assets, deposits, and loans among the national competitors immediately before and after the merger is detailed in Appendix C. It shows the following: The aggregate figures set forth in the above table are not a true measure of the total national market, nor of any bank’s relative share, for there can be no question that most of the national competitors also originate a vast, but varying and unknown, portion of their business in their own local markets. Our market share figures, therefore, are also spurious and distorted, but inversely to those of the local market, for the smaller the bank the greater the distortion because small banks originate less business in the national market than large banks. 160 Thus, it is impossible for want of evidence to ascertain relative shares in the national market. We will use the spurious statistics in analyzing the competitive effect of this merger, but their infirmities and distortions hamper application of the market share test employed in United States v. Continental Can Co., supra; United States v. Aluminum Co. of America, 377 U.S. 271, 84 S.Ct. 1283, 12 L.Ed.2d 314 (1964); United States v. First Nat’l Bank & Trust Co. of Lexington, supra; and United States v. Philadelphia Nat’l Bank, supra. Urging us to rely completely on its spurious statistics, the government reminds us that wherever possible, without doing violence to the legislative objectives underlying the antitrust laws, we should “lighten the burden of proof,” “simplify the test of illegality,” “dispense with elaborate proof of market structure, market behavior or probable anticompeti-tive effects,” and shift the onus of justifying the transaction to defendant in the interest of sound and practicable judicial administration. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at *927 362-363, 83 S.Ct. 1715. This is not such a case. However tempting the easy course, exhausting the labor, complex, elusive and fragmentary the evidence, desirable simplified tests, or great the pressures for quick solution and mass production in this congested court, if the judicial process means anything at all, we see no substitute for reliable proof and the time-consuming hunt for figures and bits of evidence buried in the record, followed by a patient sifting, fitting, weighing, and marshalling of the evidence into a meaningful pattern in order to determine the merits of the parties’ contentions. Standard Oil Co. of California v. United States, supra, 337 U.S. at 322, 69 S.Ct. 1063 (Jackson, J., dissenting). To accede to the government’s proposal and rely completely on spurious statistics would do violence to deeper values than those embodied in the antitrust laws. Although it is the subject of much debate, 161 a trial judge, we think, must function as something more than a notary public rubber-stamping the proposed findings of either party with “never thought of thinking for himself at all.” 162 We have no choice, therefore, except to struggle through the maze of statistics, put the puzzle together, find the facts, and reach our conclusion without cutting corners. 163 Market share is only one of a gamut of criteria for analyzing the competitive effect of a merger. United States v. Penn-Olin Chem. Co., supra, 878 U.S. at 176-177, 84 S.Ct. 1710; Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 329, 81 S.Ct. 623; United States v. Columbia Steel Co., supra, 334 U.S. at 527-528, 68 S.Ct. 1107; see H.R. Rep.No.1191, 81st Cong., 1st Sess., p. 8 (1949). Congress did not “adopt a definition of the word ‘substantially,’ whether in quantitative terms of sales or assets or market shares or in designated qualitative terms, by which a merger’s effects on competition were to be measured.” Brown Shoe Co. v. United States, supra, 370 U.S. at 321, 82 S.Ct. 1521 (emphasis added). Indeed, the legislative history of Clayton § 7 “reflects a conscious avoidance of exclusively mathematical tests.” Id. at 321 n. 36, 322 n. 38, 82 S.Ct. 1522. “Obviously no magic inheres in numbers” in judging the validity of a merger under Sherman § 1, Times-Picayune Pub. Co. v. United States, 345 U.S. 594, 612, 73 S.Ct. 872, 882, 97 L.Ed. 1277 (1953), and “a judgment under [Clayton] § 7 is not to be made by any single qualitative or quantitative test. The merger must be viewed functionally in the context of the particular market involved, its structure, history and probable future.” United States v. Continental Can Co., supra, 378 U.S. at 458, 84 S.Ct. 1747. We add our view that numbers are for computers not courts. The competitive effect of this; merger must, therefore, be judged on the basis of the evidence and in the context of the market facts proved here, Brown Shoe Co. v. United States, supra, 370 U.S. at 322 n. 38, 82 S.Ct. 1502, and not. in the context of the completely different market facts proved by numerology in Alcoa, Continental Can, Lexington, and Philadelphia. Certain that we must gauge the validity of the merger in two. geographic markets, and hampered by spurious figures, we must base our ulti *928 mate conclusion on the solid facts behind the mist of misleading statistics. 164 Nevertheless, we must deal with such statistics as we have to explore many of the parties’ contentions which are focused within the framework of the spurious statistics and the number rules fashioned in Philadelphia and its progeny of recent merger cases in the Supreme Court. We turn then, at last, to an examination of the parties’ contentions respecting the competitive effect of this merger. THE COMPETITIVE EFFECT OF THE MERGER A. Size is not illegal. The government places heavy emphasis upon the size of the constituent banks and the dollar volume of their business. It contends that they were such “giants” that the merger is inherently suspect and should be condemned out of hand in the absence of proof by defendant that it is not likely to have the anticompetitive effects proscribed by Clayton § 7 or Sherman § 1. Defendant contends that the constituent banks were not “giants” when measured against the scale of the economy and size of the relevant geographic markets and that, in any event, mere size neither justifies an inference of illegality nor satisfies the government’s burden of proof. There can be no question that the banks involved are “giants” in absolute terms, but in relative terms they are dwarfs compared to their markets, many of their customers, and some of their competitors. Arguments based on size, big or little, appeal not to reason or fairness but to emotion and prejudice. The government’s cries may stir resonant chimes in some ears, but they strike a gong of alarm in ours. Daily experience in the trial court teaches that such pleas generally mask a meritless case and, if anything, compel an examination of the evidence with special scrutiny and cold objectivity lest our oath of office become a hollow mockery and equal justice for the rich and the poor alike an empty platitude. That is no doubt the premise underlying the rule that absolute size (wealth) standing alone proves nothing violative of the antitrust laws. 165 Were this a jury trial we would be bound to give such an instruction, and surely we can do no less than observe the law ourselves. Size is a relative concept. Thus, if all of the commercial banks in Philadelphia were combined into one, it would be smaller than the third largest bank in the New York metropolitan area. 166 Yet, even the largest bank in the metropolitan area is smaller relative to either of its markets than the largest and second largest banks in Philadelphia were relative to theirs. 167 Similarly, all of the commercial banks in Lexington combined would be smaller in assets than First National City Trust Co., which ranks twenty-second in the New York metropolitan area. 168 Yet, in relation to their respective markets, the largest bank in the metropolitan area is twice as small *929 as the largest bank in Lexington. 169 It would indeed be impossible for Manufacturers to attain the relative size of the defendant in Philadelphia or Lexington by merger with any other bank in the metropolitan area. These comparisons, we think, make readily apparent that the government’s emphasis on size is misleading, unwarranted, and untenable when we measure the banks involved here against the scale of the enormous markets of the metropolitan area and whole nation which they serve. Moreover, even Philadelphia and Lexington, on which the government relies, base decision not on absolute, but on relative, size. The government next complains that not all of the commercial banks in the area “enjoy equal stature.” The fact is true, but its implicit tenet is neither a precept of the antitrust laws nor, we had always supposed, of a free society. Inequalities in size are inherent in the system of free, dynamic, and competitive enterprise which the antitrust laws are designed to preserve and promote. Indeed, quite probably, the mere existence of static equality in the size of competitors, however cherished by some, would be so repugnant to a truly competitive market that it might spell an unlawful conspiracy. The isolated fact, therefore, that Manufacturers and Hanover ranked high in the lineup of competitors only proves that some firms are larger than others, and that, if disparities are not undue, proves nothing except the existence of competition. Thus, even a percentage command of a market may, or may not, create or portend forbidden anticompeti-tive effects depending upon the setting in which that factor is placed. Times-Picayune Pub. Co. v. United States, supra, 345 U.S. at 612, 73 S.Ct. 872; United States v. Columbia Steel Co., supra, 334 U.S. at 527-528, 68 S.Ct. 1107; United States v. National Lead Co., 332 U.S. 319, 352-353, 67 S.Ct. 1634, 91 L.Ed. 2077 (1947). The decisions in Philadelphia, Lexington, Alcoa, and Continental Can teach nothing to the contrary. Rather, all are prototypes of the principle that the validity of a merger must be judged in its own unique setting. Specifically, Philadelphia holds that a merger which creates a firm controlling an “undue” percentage of business in the relevant market, or threatens “undue” concentration of business in a single firm, or results in a “significant” increase in concentration of the firms in that market, is so inherently likely to lessen competition substantially that it raises a re-buttable inference of illegality under Clayton § 7. We proceed, therefore, to the question of whether this merger gives defendant control of an undue share of either market or threatens undue concentration in a single firm. B. The market share foreclosed by the merger is not presumptively illegal. It is common ground among economists, Congress, and the Supreme Court that competition is likely to be greatest when there are many sellers, none of which has a “significant” advantage in market share over smaller rivals. 170 A merger which threatens to create such an advantage in one firm tends to monopoly; one which threatens to create such an aggregate advantage in a few large firms tends first to oligopoly and ultimately to monopoly and, therefore, violates Clayton § 7. United States v. Continental Can Co., supra, 378 U.S. at 461, 84 S.Ct. 1738; United States v. Aluminum Co. of America, supra, 377 U.S. at 280, 84 S.Ct. 1283; United States *930 v. Philadelphia Nat’l Bank, supra, 374 U. S. at 363, 83 S.Ct. 1715. Whether a given merger increases the market share of the resulting firm to forbidden proportions or threatens a “significant” rise in concentration depends on the setting. Brown Shoe Co. v. United States, supra, 370 U.S. at 321-322, 82 S.Ct. 1502. The government contends that this merger, by combining the shares of the constituent banks, created a new firm with such an increased share of the market that the merger is inherently anticompetitive, presumptively illegal, and should be invalidated without further proof of anticompetitive effect unless defendant justifies the transaction. Defendant answers that its increased share is not presumptively illegal and that the government cannot so easily escape from its burden of proof. A merger between firms occupying the same markets is known as a horizontal merger. Necessarily, such a merger combines the shares of the constituent parties and eliminates one firm from the market. It thereby automatically creates a firm with an increased share and increases concentration of the number of firms in the market. Yet, Congress, in enacting Clayton § 7, did not forbid all horizontal mergers but only those which may lessen competition substantially or tend to create a monopoly. Indeed, the legislative history of Clayton § 7 shows that Congress did not intend to impede mergers of smaller companies in order to enable them to compete more effectively with larger firms. 171 The government, therefore, proves nothing unlawful by the naked fact that the acquiring firm has increased its market share or that a merger has increased concentration. We first explore whether the market share obtained by this merger is presumptively illegal or decisive without regard to other factors. In United States v. Philadelphia Nat’l Bank, supra, 172 the setting was that the *931 consummation of the merger of PNB and Girard would have increased concentration in a single firm by 64% and have created a dominant bank far larger than any of its competitors. It would have foreclosed 30% 173 of the business in the relevant market and the resulting bank would have enjoyed better than a 50% advantage in market share over its nearest rival. Moreover, if the merger were consummated, significant disparities would separate each bank except two from its closest smaller competitor thereby upsetting the competitive balance of the market structure. 174 In United States v. Aluminum Co. of America, supra, Alcoa’s merger with Rome increased concentration by only 4.6%, but that was significant in the setting for it was enough to cause foreclosure of 29.1% of the market by a single firm which already enjoyed a commanding lead in a market of too few competitors. 175 Again, significant disparities in market share separated each competitor except two from its nearest rival, and, as in Philadelphia, the increment of only 1.3% obtained by merger concentrated nearly 30% of the market in a single dominant firm. In United States v. Continental Can Co., supra, the merger of Continental with Hazel-Atlas increased concentration by 14% but resulted in control of 25% *932 of the market in a single firm which already ranked second. 176 Although the share foreclosed was 5% less than that found undue in Philadelphia and Alcoa, the court, holding that it approached the presumptively illegal line, lowered the bar five points, and Continental was too big to crawl under. The local market share foreclosed by this merger does not approach those in the trilogy of cases the government summons to its aid. Here, immediately after the merger, according to the spurious shares inherent in the hypothesis of one relevant geographic market, the merger created, not a firm with a dominant one-third or one-fourth, but a bank with only a tenuous hold on about one-eighth, with 13.6% of the metropolitan area’s total assets, 13.9% of the deposits, and 12% of the loans. Two years after the merger, the percentages had decreased to 11.9%, 12%, and 11.1%, respectively, and they have since declined to 11.5%, 11.6%, and 10.8%, respectively. There can be no question that the merger created a firm with an increased share, but the increment resulted in a bank one-third to one-half smaller than its two larger local competitors, and only slightly larger than the next three. The merger narrowed the gap previously existing in the market structure between the second and third largest banks, but widened the gap between the seventh and eighth largest banks 177 As noted earlier, we cannot ascertain defendant’s actual share of the true local market, and it may be greater or less than the shares shown by the homogenized statistics employed in the foregoing comparisons. We are confident, however, that the government has failed to prove that defendant’s share approaches the presumptively illegal lines drawn in Philadelphia, Alcoa, and Continental Can. Viewing the merger in the national market, we note at the outset that it did not increase the share held by the three largest firms. Immediately after the merger, defendant moved to fourth place, but its incremental share did not remotely approach a decisive one-third or one-fourth of the market, but amounted only to an inconclusive and vulnerable 4.22% of the assets, 4.11% of the deposits, and 3.51% of the loans, and since the merger, with the entry of newcomers into the national market and greater relative growth rates of smaller competitors, defendant’s share has declined to 3.29% of the assets, 3.41% of the deposits, and 3.27% of the loans. Although defendant’s incremental share increased deposit concentration in a single bank, the increase narrowed the gap previously existing in the competitive structure where the third largest *933 bank had nearly a 100% advantage in market share over its closest smaller rival. The merger thereby improved the balance of the competitive structure and intensified competition for the three leaders. Moreover, unlike Philadelphia, Alcoa, and Continental Can, defendant’s incremental share did not give it a decisive lead over its next smaller competitor. The gap in market share between defendant and the next ranking competitor ranged from a minimum of .35% in loans to 1.12% in deposits, and the shares of the remaining competitors were so graduated from top to bottom that no one bank had any decisive advantage in market share over its next smaller competitor. 178 We think the foregoing analysis, though based on spurious figures, demonstrates that the government has failed to prove that defendant’s true national market share approaches the 30% or 25% market shares held inherently anticompetitive in Philadelphia, Alcoa, and Continental Can. We, therefore, find nothing in the numerology of the trilogy of cases which would warrant a conclusion that this merger is inherently illegal in either market merely on the basis of the market share foreclosed by defendant or the percentage of increase in concentration in a single firm. Recognizing that even on its one market theory the share foreclosed by defendant here falls short of the line drawn in the trilogy, the government next urges that the merger is inherently anticom-petitive because defendant’s spurious share is greater than that found undue in other cases involving different industries, different markets, different histories, and different market structures. Defendant counters with cases sustaining mergers which created market shares greater than that foreclosed here. We think neither contention sound. 179 Manifestly, the same percentage share may be undue in one market setting and devoid of any anticompetitive repercussions in another. Like the Congress, we think that antitrust cases, with their infinite variables set in a dynamic economy, can no more be cast into rigid and static molds than can the size and shape of a cloud. Moreover, we have been taught, before, by, and since Philadelphia, that no particular percentage share can be deemed critical, and that the required prediction of future anticompetitive effects cannot be made without a firm understanding of the unique structure and history of the relevant markets involved in the case under consideration. Brown Shoe Co. v. United States, supra, 370 U.S. at 321, 82 S.Ct. 1502; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 362, 364 n. 41, 83 S.Ct. 1715; United States v. Continental Can Co., supra, 378 U.S. at 456, 458, 84 S.Ct. 1738. We cannot, therefore, conclude that this merger is illegal merely because the market shares foreclosed may be greater here than the shares found undue in other cases. Nor can we conclude that the merger is lawful simply because the shares here may be less than those found undue in Philadelphia, Alcoa, and Continental Can. In short, neither Philadelphia nor any other case fashions a template which mechanically supplies the answers in all other merger cases as to what percentage constitutes an “undue” market share or threatens “undue” concentration in a single firm. We must look beyond numerology, therefore, to determine whether this merger has created a firm foreclosing an “undue” share of either market or threatening “undue” concentration in one firm. Implicitly conceding as much, the government urges that the increase here offends the Clayton Act because defendant’s market share gives it an advantage over smaller competitors in greater lending limits, more advertising resources, and more branch offices. There can be no question that greater lending limits *934 give defendant an advantage over smaller competitors. Nor can there be any doubt that defendant can advertise on a scale that smaller competitors cannot afford, but, for sure, the merger has not hushed the din of rival commercials. Likewise defendant’s expanded branch system— larger than any other in the area — does give it an advantage in the local market. The fact that a merger creates a firm with an advantage over competitors is a relevant, though not a decisive, factor. “[E]xpansion [by merger] is not rendered unlawful by the mere fact that small * * * [competitors] may be adversely affected.” Brown Shoe Co. v. United States, supra, 370 U.S. at 344, 82 S.Ct. 1534. A fortiori, there is nothing unlawful in the sole fact that a merger reduces the advantage of larger competitors; quite the contrary. A struggle for advantage over competitors, large and small, is the essence of competition, 180 and “[i]t is competition, not competitors, which the Act protects.” Brown Shoe Co. v. United States, supra, at 344, 82 S.Ct. 1534. The Clayton Act, therefore, neither grants an inviolate domain to the industry leaders nor freezes smaller competitors with the status quo. 181 “Taken as a whole, the legislative history [of Clayton § 7] illuminates congressional concern with the protection of competition, not competitors, and its desire to restrain mergers only to the extent that such combinations may tend to lessen competition” substantially. Brown Shoe Co. v. United States, supra, at 320, 82 S.Ct. 1521. The government is thus required to prove something more than the naked fact that the merger creates a firm with an advantage over competitors. Philadelphia teaches that when market share advantage over competitors is the decisive factor, the “something more” required by Clayton § 7 is proof that the merger creates a firm foreclosing an “undue” share, or that it threatens “undue” concentration either in a single firm or in a handful of competitors. Philadelphia, Alcoa, and Continental Can, however, give little 'guidance, aside from the context of their own facts, on the underlying criteria for determining whether a given merger forecloses an “undue” share or threatens “undue” concentration in a single firm. Those cases thus afford no easy answer to this one where the share foreclosed by the resulting bank in both markets is uncertain, and on the basis of spurious statistics, far less than the shares found inherently anticompetitive there. It appears, however, from the market facts in the three cases, the legislative history of Clayton § 7, and the teaching of Brown Shoe and earlier cases, 182 that the key to the problem lies in whether the share of the market foreclosed by one, or a few competitors, as a result of a merger is so large relative to the shares held by remaining competitors that it threatens to arm one, or a few, with a decisive advantage in the competitive struggle, or significantly limits the opportunity for other firms to enter into, or remain in, the market. Clayton § 7 does not contain the words “undue” or “significant.” We look, therefore, to its legislative history for guidance. The Report of the House Judiciary Committee 183 observes that a significant reduction in the vigor of competition could be perceived if “the relative size of the acquiring corporation *935 had increased to such a point that its advantage over competitors threatened to be ‘decisive.’ ” Brown Shoe Co. v. United States, supra, 370 U.S. at 321 n. 36, 82 S.Ct. 1521 (emphasis added). It cannot be denied that a firm actually having a decisive advantage over its competitors is in the driver’s seat and able to control prices and curtail competition. In short, it possesses monopoly power. United States v. E. I. Du Pont De Nemours & Co., 351 U.S. 377, 391, 76 S.Ct. 994,100 L.Ed. 1264 (1956). A merger, therefore, which threatens to create such a decisive advantage tends to create a monopoly and is illegal within the meaning of Clayton § 7. We find support for that approach both in scholarly studies 184 and in the teaching of Brown Shoe Co. v. United States, supra, where, considering the same problem, but in the setting of a vertical merger, the court said, 370 U.S. at 328-329, 82 S.Ct. at 1525: “If the share of the market foreclosed is so large that it approaches monopoly proportions, the Clayton Act will, of course, have been violated ; but the arrangement will also have run afoul of the Sherman Act. And the legislative history of § 7 indicates clearly that the tests for measuring the legality of any particular economic arrangement under the Clayton Act are to be less stringent than those used in applying the Sherman Act. On the other hand, foreclosure of a de minimis share of the market will not tend ‘substantially to lessen competition.’ “Between these extremes, in cases such as the one before us, in which the foreclosure is neither of monopoly nor de minimis proportions, the percentage of the market foreclosed [5%] by the vertical arrangement cannot itself be decisive.” We think this is just such a ease for, standing alone, the share foreclosed by the defendant here is neither of incipient monopoly, nor of de minimis, proportions in either market. Plainly defendant’s share does not give it incipient monopoly power or threaten to arm it with a decisive advantage over its remaining larger competitors, and, standing alone, its gain over smaller competitors threatens no such forbidden power in either market. Nor, standing alone, does it significantly limit the opportunities for smaller competitors to enter into, or remain in, the market. Moreover, the fact that defendant is steadily losing ground, according to the spurious statistics, speaks against any rational prediction that defendant’s market share, standing alone, is undue or threatens undue concentration in a single firm in either market. Surely, in the light of the facts, we cannot speculate on the basis of spurious statistics that defendant has acquired incipient monopoly power, or an undue share of either market, as a result of this merger. Quite the contrary. The evidence indicates that defendant’s share is too small, its competitors too strong, and its rivals too numerous to warrant any such prediction. We conclude, therefore, that we cannot predicate a violation of either the Clayton or Sherman Acts solely on the factor of the market share foreclosed by defendant in either market as a result of the merger. Nor is there anything in that factor, standing alone, which relieves the government of its burden of proof. C. Concentration. The government asserts that both markets are highly concentrated and have witnessed a recent trend toward oligopoly. Therefore, the government urges, even though the percentage of increase in defendant’s share in each market may be indecisive standing alone, nonetheless, the increase in concentration caused by the merger becomes significant and decisive when viewed in the structural setting and history of the markets. The government then predicts that this merger will trigger others and urges that this is the place to call a halt in order to curb a tendency to *936 monopoly in its incipiency before customer alternatives disappear. Defendant contends that this merger does not significantly augment concentration or any trend in that direction because the business of the constituent banks was more complementary than competitive, commercial banking in both markets is fragmented rather than concentrated, newcomers have entered, and the combined market share of the larger banks has remained well within the limits necessary to effective and vigorous competition among many enterprises despite past mergers. The merger, defendant asserts, does not threaten to spark a chain reaction toward undue concentration in either market because contemporary market conditions and recent changes in state and federal law make a myth of the government’s predictions. “The dominant theme pervading congressional consideration of the 1950 amendments [to Clayton § 7] was a fear of what was considered to be a rising tide of economic concentration in the American economy.” Brown Shoe Co. v. United States, supra, 370 U.S. at 315, 82 S.Ct. 1518. See also United States v. Continental Can Co., supra, 378 U.S. at 461-462, 84 S.Ct. 1638; United States v. Aluminum Co. of America, supra, 377 U.S. at 278-281, 84 S.Ct. 1283; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 362-370, 83 S.Ct. 1715. The “keystone in the erection of a barrier to what Congress saw was the rising tide of economic concentration, was its provision of authority for arresting mergers at a time when the trend to a lessening of competition in a line of commerce was still in its incipiency. Congress saw the process of concentration in American business as a dynamic force; it sought to assure * * * the courts the power to brake this force at its outset and before it gathered momentum.” Brown Shoe Co. v. United States, supra, 370 U.S. at 317-318, 82. S.Ct. 1520 (emphasis added). Every horizontal merger causes some increase in, and tends toward, concentration by permanently eliminating one firm from the field. Congress, however, did not make all horizontal mergers illegal per se, although it was plainly aware that banking, like almost every other industry in the American economy, had a history of mergers. There is, thus, no basis either in the statute or in its legislative history for equating every increase in, or tendency toward, concentration with a tendency toward oligopoly or either of the anticompetitive effects expressly proscribed by amended Clayton § 7. 185 The grand design of Clayton § 7 was, of course, to curb tendencies toward monopoly in their incipiency, but “incipiency” is an elastic term, and it cannot be stretched to the extreme of curbing every increase in, or tendency toward, concentration. The government, therefore does not prove its case simply by reciting “Ten Little Indians,” for while Clayton § 7 bans “all mergers having demonstrable anticompetitive effects, Congress recognized the stimulation to competition that might flow from particular mergers.” Brown Shoe Co. v. United States, supra, at 319, 82 S.Ct. 1521 (emphasis added). Whether “incipiency” should be stretched to curb an increase in, or tendency toward, concentration caused, or threatened, by a given merger depends on the setting in which those factors are placed. If concentration is already high, or if there is a strong trend in that direction, even a slight or minute increase runs afoul of Clayton § 7. United States v. Continental Can Co., supra; United States v. Aluminium Co. of America, supra. The ultimate question is whether this merger, viewed in the structure, history, and probable future of the market, is of the type which Congress intended to proscribe. Brown Shoe Co. v. United States, supra, 370 U.S. at 329, 82 S.Ct. 1502. *937 Having laid out the ground rules, we proceed to the probable impact of the merger on concentration in the local market. D. The Level of Concentration Existing in the Local Market Before the Merger. The initial problem is determination of the level of concentration existing at the time of this merger. The government’s contention that the local market was oli-gopolistic before the merger rests not on predatory conduct or collusive behavior but on the structure and history of the market. 186 The government does not, and indeed could not, claim that the structure of the market is one of few sellers. Rather, it is the government’s thesis that a comparatively few banks control a disproportionately large share of the market and that a merger trend is rapidly concentrating more and more local business and formerly independent banks into fewer and fewer hands. Therefore, the government urges, the addition to concentration, or to the trend, caused or threatened by this merger, thwarts the design of Clayton § 7 to arrest a trend toward oligopoly — the tendency to monopoly — in its incipiency. The government suggests no criteria for determining the critical level of concentration in the local market. Instead, it invites us to compare its spurious statistics with the statistics in Philadelphia, Alcoa, and Continental Can. We find such a comparison neither necessary nor productive. The government’s fragile foundation and different market facts would undermine any conclusion we might draw from matching numbers. Defendant argues that a comparison of the level of concentration in New York with that of other metropolitan areas demonstrates that the local market is not unduly concentrated. It points to evidence that, even on the government’s figures, the combined “market share” of the five largest banks is less in New York than in 57 out of 65 other metropolitan areas, the share of the three largest less than in 58 out of 65 others, and the share of the largest bank less than in 60 others. The comparisons do not afford an index for determining whether concentration is undue in the market under consideration here, for the number of competitors was more in New York than in 60 out of 65 other metropolitan areas, 187 and manifestly the share of the five largest banks in a market of 72 competitors cannot be equated to the share of the five largest banks in a market where, for example, there are only five or six. Concentration ratios might be computed and compared, but defendant makes no such attempt, and we think the results would still leave us without a valid gauge. Moreover, the question is not whether concentration is greater or less than elsewhere, but whether it threatens to reach anticompetitive proportions here as a result of the merger. Thus, we find both the government’s and defendant’s tests for measuring the critical level of concentration unsound. The number and relative strength of firms necessary to effective competition cannot be compressed into a formula, and obviously what constitutes undue concentration in one industry or one market may not in another. 188 Banking is a regulated industry. Defendant, therefore, argues that the banking agencies are the best judges of the critical level of concentration and asks us to accept their conclusions that neither market was overly concentrated and that the merger did not increase concentration significantly. The government complains, however, that in assessing the competitive effect of the merger, the banking agencies relied only on aggregate figures; yet the government asks us to accept precisely the same *938 sort of figures as one basis for determining concentration. Philadelphia teaches that we are not bound by the banking agencies’ conclusions and that we must judge the validity of the merger solely in the context of the antitrust laws. But surely, in the face of the banking agencies’ opinions, we cannot speculate that either market is unduly concentrated or base our decision on a distorted picture merely to make things easy for the government. Once again, therefore, we must look behind the spurious statistics for some meaningful basis for rational judgment. The aggregate figures on which the government relies are only the distorted shadow of the true local market, and it cannot be transformed into the real image by the mere incantation of numbers, dollars, and shibboleths. The real size and shape of the local market lies invisible beneath the overlapping national market, and we cannot extract it from the government’s statistics unless we indulge in sheer guesswork. 189 We think, however, that the structure of the local market can be perceived from the evidence with fidelity to competitive realities if we focus, not upon its distorted shadow, but upon its substance — the many accounts of the general public and small business. 190 Well over 90% of all customers are al-locable to the local market. Their accounts are the countless small deposits and loans of the general public and small *939 business (see Appendix A). 191 Such customers are fettered by convenience, local needs, etc. to their own residential or working locality for their banking alternatives. The local market, unlike the national market, deals not in a few, large, custom tailored wholesale transactions of high unit profit, but in millions of small, repetitive transactions of low unit profit. Efficient, low cost, and profitable handling of retail accounts depends on great volume 192 and mass processing of millions of transactions. 193 The business dictates economies of scale which impel retail banks to acquire or establish large branch systems with offices strategically located in commercial hubs, manufacturing centers, shopping areas, transportation intersections, and in canyons walled with the vertical population of skyscraper office buildings and high-rise apartments. 194 The amount of local business generated per branch is not uniform. A bank with one branch may do more business than another with two or three. An individual bank’s relative share of the market, therefore, may be greater or less than its relative share of branches. Nonetheless, at some point advantages unique to the location, management, etc. of one branch are averaged out and overcome by other branches with like advantages and sheer numbers, and once that point is passed, there is a direct and progressive relationship between the number of branches, the number of customers, the amount of business, and market share. The greater the number of branches, the greater the volume and the greater the market share. 195 Thus, Manufacturers with 120 branches won 1,254,313 retail customers and generated $1,383,252,000 in local deposits and $360,810,000 in loans, while Hanover with 10 branches won only 36,521 retail customers and generated only $258,383,000 in local deposits and $38,686,000 in loans. The direct relationship between number of branches, number of customers, and amount of business demonstrates that significant disparities in relative share of branches breed significant disparities in market share. As the disparities in relative share of branches increase, so do the disparities in market share. Were this not so, there is no logical explanation for the proliferation of branches. Commercial realities thus warrant an inference that concentration of branches reflects collective market share of the largest integrated banks and concentration of the local market. 196 *940 The distribution of branches in the City of New York, as of December 31, 1960, was as follows: Distribution of Banking Offices of All Commercial Banks Located in New York City As of December 31, 1960. The disparity in distribution of branches manifested by the foregoing table compels the conclusion that at the time of this merger, a .handful of large banks controlled a disproportionately large share of the local market. 197 Ascertainment of the level of concentration existing before the merger, however, is only part of our problem. We must also discern the trend and determine the probable impact of this merger on concentration, and ultimately the impact of con *941 centration and this merger upon competition. E. The Trend. Clayton § 7 “prohibits a given merger only if the effect of that merger may be substantially to lessen competition” or tend to monopoly. Brown Shoe Co. v. United States, supra, 370 U.S. at 332, 82 S.Ct. 1527. The statute, however, “requires a prognosis of the probable future effect of the merger.” Ibid. Determination of the probable future impact calls for a prediction, and a sound prediction demands a valid basis for projection. We think it plain that in most cases a comparison of the number of firms and the level of concentration in a market at the instants immediately before and after a merger is not an adequate basis for discerning a trend. Rather, the evolution of the market structure, and the effect of its merger history over some reasonable period of time, 198 must be examined to discern whether there is a trend which is likely to result in undue concentration if a given merger is allowed to stand. Brown Shoe Co. v. United States, supra, at 332, 82 S.Ct. 1502. Our quest is to discover whether there is a trend toward concentration and, if so, whether the trend has resulted in such concentration, or gathered such momentum, that this merger threatens to increase concentration, or to accelerate the trend, to proscribed anticompetitive proportions. Id. at 315, 322, 332-333, 82 S.Ct. 1502. In determining whether there has been a trend toward concentration, there must be consistency in the number of firms employed for measuring changes in the level of concentration during the historical period considered, and, in computing the percentage of increase in concentration over the period, much turns on the number of firms considered. In Philadelphia, the number considered in computing changes extended down the lineup of competitors only to the point necessary to include the pre-merger position of the larger firm. 199 ** We, therefore, employ that method and will use five banks in the local market as a basis for discerning whether there is a trend toward concentration. We think that the history of the market since the end of World War II is the relevant period to be considered here in order to understand how concentration reached the level existing before this merger and predict what the impact on concentration is likely to be as a result of this merger. We will survey the market history and ascertain whether, as a matter of fact, over the relevant period the aggregate share of the five largest firms in the local market has remained fairly consistent or has grown so disproportionately to the shares of smaller competitors as a result of mergers that it is reasonably probable that this merger threatens to result in concentration of proscribed anticompetitive proportions. As we have seen, following World War II, most of the large New York banks which had previously engaged almost exclusively in wholesale banking found it necessary to enter the mass local retail market in response to adverse changes in economic conditions. 200 The pace of the *942 change was so swift that mergers with existing retail banks, rather than slow and costly entry by internal expansion, offered the only quick and economically feasible means of acquiring a ready made branch system to generate the necessary volume. The adjustment to integrated wholesale-retail operations culminated in 27 mergers during the decade ending in 1960. The immediate impact of most of the major mergers in the local market 201 was more vertical than horizontal, for the wholesale bank had comparatively little local business to bring to the retail bank’s share. The mergers, therefore, as a practical matter, brought newcomers into the local market, and their vast resources were made immediately available to the general public and small businesses on a broad scale. While it is true that one of the banks ceased to be a separate competitor, the retail partners did not vanish from the arena, for, so far as the record reveals, no bank closed its doors or ceased operations. Rather, the mergers were followed by rapid development and internal expansion of the retail side of the business, but the resulting bank integrated its local operations with extensive nationwide and worldwide wholesale banking. Manufacturers had a different history. It grew as a retail bank, and its position in the wholesale market was built upon the aggregate deposits of over one million customers. As a result, at the time of this merger, it was already an integrated bank. Hanover was predominantly a wholesale bank operating in the national market, but it also operated in the local market, and, six months before the merger, it had ventured into some retail banking. According to the evidence, only two remaining banks — Morgan Guaranty and Bank of New York— still continue to operate as purely wholesale banks. Defendant argues that the government has failed to prove that any of the earlier mergers had any anticompetitive effect. It insists that earlier mergers of wholesale and retail banks have not lessened, but increased, the vig- or of local competition, and have not impaired, but improved, available sources of credit for the general public and small marginal businesses. It stresses that competition among the survivors is strong and vigorous and points to uncontradicted evidence that large integrated banks do not discriminate against small borrowers or other retail customers, but make a special effort to attract and serve their needs, and, indeed, have blazed new trails in consumer credit, special checking accounts, savings accounts, small businesses loans, and in a broad extension of branch facilities to communities and areas which could not support their own independent bank. 202 We have no quarrel with defendant’s *943 facts, and it cannot be denied that interest rates, service charges, etc. are low. 203 Nor do we doubt that, competitive factors aside, the general public and small businesses have benefited from the mergers. Nonetheless, practices harmless in themselves, or even those conferring benefits upon the community, cannot be tolerated when they tend to create a monopoly; those which restrict competition are unlawful no matter how beneficent they may be. It is of no moment that a short-run by-product may have been greater efficiency and lower costs, for it is the theory of the antitrust laws that the long-run advantage to the community depends upon the removal of restraints upon competition. Brown Shoe Co. v. United States, supra, 370 U.S. at 344, 82 S.Ct. 15C2; Standard Oil Co. of California v. United States, supra, 337 U.S. at 309, 69 S.Ct. 1014; United States v. Aluminum Co. of America, 148 F.2d 416, 427-428 (2d Cir. 1945); United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 617-618. Although the earlier mergers were primarily vertical in their immediate aim and effect, there can be no blinking the fact that, with the passage of time, all of them have had an impact on concentration. Necessarily, the earlier mergers and the special efforts of integrated banks to attract retail customers have brought ever-greater numbers of formerly independent banks, and more and more of the local market, into fewer and fewer hands. Brown Shoe Co. v. United States, supra, 370 U.S. at 345, 82 S.Ct. 1502. In 1950 there were 70 independent commercial banks in New York City. 204 During the decade 27 banks were absorbed by mergers, while only three newcomers entered the local market, and they were established by existing banks. Thus, 38.6% of the commercial banks existing in 1950 ceased to exist as independent competitors by 1960, and Manufacturers was no stranger to the process. 205 The reach for branches gathered 222 banking offices, together with all their local business, into the hands of a diminishing group of survivors. 206 As a result, in the decade preceding this merger, the general public and small businesses had been deprived of over one-third of their independent banking alternatives. As we have seen, at the time of this merger, a handful of firms occupied 79% of the branches in the City of New York. There is no need to rely on deductive reasoning to learn how the five largest integrated banks achieved their dominant position. Hanover’s former chairman tells us: “[A]ll of the five largest banks in the complete market— by that I mean the local, national and international markets — have achieved their present position through important mergers.” 207 The relative share of branches, and therefore the aggregate market share of the five largest banks, did not remain consistent while they were expanding their retail operations. Their growth was not offset by growth of smaller banks or the entry of newcomers. Rather, as a result of “important mergers” and de novo branching, the structure of the market was altered from one of many enterprises to one of a handful of firms at the top of an inverted pyramid, with a significant, and threatening to obtain *944 a decisive, advantage in market share, The graphs below demonstrate that there is no mistaking a strong trend toward oligopoly: CHANGES IN THE DISTRIBUTION OF BANKING OFFICES IN NEW YORK CITY FROM 1954 TO SEPTEMBER 9, 1961, THE DAY FOLLOWING THIS MERGER. Having ascertained the level of concentration and discerned the trend toward oligopoly, we turn to consideration of whether this merger threatens to augment concentration or the trend, *945 F. The Increase in Concentration Caused or Threatened by the Merger. The government argues that Manufacturers and Hanover were substantial competitors in both markets, and that the merger therefore significantly increased concentration and accelerated the trend by eliminating former competition between them and removing a substantial factor from competition as a whole. Defendant admits that the constituent banks were competitors, but argues that competition between them was more complementary than competitive, and therefore the merger did not, and does not threaten to, increase concentration significantly whatever the pre-existing level or trend. Clayton § 7, of course, applies to mergers between actual competitors even though the firms have competed directly on the horizontal level in but a fraction of the relevant product and geographic markets in which either has operated. Cf. Brown Shoe Co. v. United States, supra, 370 U.S. at 337, 82 S.Ct. 1502; United States v. Bethlehem Steel Corp., supra, 168 F.Supp. at 618. The statute also bans any acquisition by one corporation of another, competitor or not, whenever the reasonable likelihood appears that the acquisition may lessen competition substantially or tend to create a monopoly. Brown Shoe Co. v. United States, supra, 370 U.S. at 317, 82 S.Ct. 1502; United States v. E. I. Du Pont De Nemours & Co., supra, 353 U.S. at 592, 77 S.Ct. 872. As we have seen, Manufacturers was a retail-wholesale bank, while Hanover was predominantly a wholesale bank. Hanover operated extensively in the national market and originated only a comparatively small percentage of its business in the local market. We, therefore, have no difficulty in concluding that the business of Manufacturers and Hanover in the local market was more complementary than competitive or to put it another way, more vertical than horizontal. It does not follow, however, that competition between Manufacturers and Hanover on the horizontal plane was minimal or insignificant, or that Hanover’s contribution to the vigor of local competition as a whole was insubstantial. Nor, as we have seen of earlier mergers, does it follow that this merger, in its vertical aspects, will have no impact on concentration in the local market. A detailed comparison of the local and national business of the constituent banks, exclusive of trust services, is set forth in Appendix A. It shows that 36,521 (91.5%) of Hanover’s customers were in the local market compared to 1,264,313 (99.6%) of Manufacturers’. Hanover originated $258,383,000 (18.-34%) of its deposits, and $38,686,000 (4.34%) of its loans in the local market compared to $1,383,252,000 (50.37%) and $368,810,000 (24.10%), respectively, for Manufacturers. Hanover had 424 loans aggregating $14,234,000 to small business firms who borrowed $100,000 or less, compared to 9,012 such loans aggregating $127,774,000 for Manufacturers. Hanover had 8.7% of the voluntary and court trusteeships held by New York City banks and trust companies compared to 4.9% for Manufacturers. (DX 1-Memo-randum, pp. 41-42.) 208 Although Hanover’s contribution to local competition may have seemed small compared to Manufacturers and other large integrated banks, it does not follow that it was insubstantial. The evidence, we think, justifies an estimate that Hanover had a greater share of the local market’s deposits than 48 smaller banks, and a greater share of its loans than 40 others. 209 In a market of 72 competitors, *946 we cannot say that the competition of a bank aggressive enough to win a greater share of deposits than two-thirds of the banks in the arena was minimal, insignificant, or insubstantial. Nor can we say that it did not contribute substantially to the vigor of competition, even though the range of its retail services was limited, and its share small, relative to larger competitors. Quite the contrary. In the setting, we think, Hanover was a substantial, vigorous, and major competitor. United States v. Aluminum Co. of America, supra, 377 U.S. at 280 et seq., 84 S.Ct. 1283; United States v. First Nat’l Bank & Trust Co. of Lexington, supra, 376 U.S. at 671-672, 84 S.Ct. 1033. Defendant argues, however, that even though the merger eliminated former competition between the constituent banks, it did not lessen competition as a whole, but increased the vigor, by giving small businesses and other retail customers not one less, but one more, practical banking alternative. Defendant points to evidence that before the merger Hanover offered no savings accounts whatever compared to 465,221 for Manufacturers. Since the merger 7,479 new savings accounts have been opened at former Hanover offices (DX 58). Hanover did not have a special checking account service until April 1960, and six months later, at the time of the merger, Hanover had developed only 4,280 special checking accounts, and 2,953 of them were accounts of its own employees (DX 5). Against this, Manufacturers had 250,021 special checking accounts (Appendix A). Since the merger, the number of special checking accounts at former Hanover banking offices has increased to 5,477, exclusive of employee accounts (DX 58). Until April, 1960 Hanover did not offer a consumer installment loan service and had only 927 such loans at the time of the merger compared to 189,460 for Manufacturers (Appendix A). By July 25, 1963 the number of such loans at former Hanover offices had increased to 3,491 (DX 58). There can be no question that one of the immediate effects of the merger was that small customers gained a practical, though not an independent, banking alternative to the extent that Hanover instituted personal banking services where previously it had none. 210 It cannot be gainsaid, however, that customers lost an actual and substantial banking alternative to the extent that services previously offered by Hanover duplicated those offered by Manufacturers and other competitors. Small business loans, single payment loans to individuals, IPC demand deposits under $100,000, and trust services loom large in that anticompeti-tive effect. 211 More significantly, customers lost a potential banking alternative, for even though Hanover had a long tradition as a wholesale bank, it already had 10 branches — more than 37 other rivals in the city — , 212 ample resources to acquire more, and the same economic motivations which had pressed other wholesale banks to integrate operations and expand into the mass retail market. 213 That it was sufficiently aggressive and in a position to respond to the urge appears clear. 214 Indeed, it had already established the nucleus of a personal banking department in 1960. True as a late arrival in the mass market it had made relatively little headway against the encircling branches of en *947 trenched competitors, 215 but during the six months before the merger, it had developed enough retail business to show some promise. Opportunities to exploit its advantages were on its doorstep. Its branches were located in areas of tremendous population density. Some 38,-700 people could have banked with Hanover without leaving the buildings in which they worked, another 1,125,000 potential customers swarmed in the areas of its 10 branches, and growth projections for the areas were impressive. 216 Hanover would, of course, as Manufacturers’ chairman put it, face “tremendous obstacles” in expanding its branch system internally to generate a volume remotely approaching that held by integrated banks like Manufacturers. “Principal among these would be the difficulties of finding new locations, the enormous capital investment required, and the lack of trained personnel.” 217 An underlying premise of the Clayton Act, however, is that internal expansion is more healthy for the economy than expansion by merger. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 370, 83 S.Ct. 1715; Brown Shoe Co. v. United States, supra, 370 U.S. at 345 n. 72,82 S.Ct. 1502. Surely we are not to suppose that a bank which at the end of six months’ experience was so enthusiastic about retail banking that it swallowed 120 branches in one gulp, was not ready, able, and eager to exploit its present advantages, as well as expand its own branch system internally, if it had not acquired the lion’s share of the market’s branches by merger. Any such prediction is precluded by defendant’s own witnesses who were emphatic that the future of commercial banking lies in the local retail market, and it is plain that Hanover had made the same forecast. 218 . Hanover was no feeble, obscure, struggling newcomer. It had been in New York for almost a century. It had vast resources, enjoyed great prestige, enormous good will, and public confidence. 219 It was doing well on its own, and there was no need to merge. 220 The independent existence of a well-equipped, well-financed, well-known, old, and thriving competitor, already established in the local market and threatening to expand into the coveted mass market, is a powerful and substantial stimulant to the vigor of competition. United States v. Penn-Olin Chem. Co., supra; United States v. El Paso Natural Gas Co., supra. Indeed, defendant insists that competition before the merger was more vigorous in New York than anywhere else in the country. We think it plain that Hanover’s contribution to the vigor was substantial, both actually and potentially. Viewed vertically, as defendant urges, the long-range impact on concentration and the vigor of competition threatened by the merger is not better, but worse. In vertical perspective, one of the market’s foremost retail banks acquired the *948 resources (production facilities) of one of the nation’s foremost wholesale banks, while the wholesale bank acquired an additional 120 branches (loan outlets and deposit inlets). Cf. Brown Shoe Co. v. United States, supra, 370 U.S. at 345, and at 367, 82 S.Ct. 1547 (Harlan, J., concurring). The combined resources can be used interchangeably in either market and shifted from one to the other in response to opportunities. That, of course, was the self-evident reason for the merger. The chairmen of both banks laid bare their plans to use the combined resources for that purpose in both markets and to convert Hanover’s branches immediately to broad scale retail operations. 221 The plans were carried out; former Hanover offices have developed retail business at a mounting rate, as we have seen, and they will continue to do so 222 Thus, the reasonable probabilities are that more and more of the local market will be gathered into still fewer hands by this merger. As a result of this merger, 80.7% of the branch banking offices in New York City were concentrated into the hands of the five largest integrated banks. While the percentage of increase in branch concentration caused by the merger was only 2.15%, the factual setting admits of no conclusion other than that this merger significantly increases concentration. United States v. Aluminum Co. of America, supra. We proceed to the question of whether it will accelerate the trend toward oligopoly. G. The Impact of this Merger on the Trend. This merger gathered into still fewer hands 10 more branches, over $38 million in local loans, and over $250 million in local deposits. Since this merger, there have been seven others, and four of the banks eliminated were acquired by four of the six largest banks in New York City. As a result of these later mergers, 18 more offices were absorbed along with over $200 million in assets. The customers’ loss of independent banking alternatives had mounted, as of September 1963, to 44% of the competitors existing in the local market at the beginning of 1950. Against this, since the merger, two new banks have opened, three other applications are pending, and 13 foreign banks have established 15 full service branches. Witnesses from some of the foreign banks testified, but there was no testimony or other evidence that any of them had gotten off the ground in the local market. Flushing National Bank of Queens, the only entirely new bank to open in the city since 1950, commenced operations December 31, 1962, and after eighteen months had wrested only 0.015% of the market’s assets from entrenched competitors. The other newcomer, County, National Bank, located in Nassau, did worse; it scratched out 0.003% (DX 56). Since the merger, the market share of the five largest banks has declined according to the spurious statistics. 223 We *949 think, however, that the decline probably reflects lost ground in the national market, but, in any event, the decline is an ebbing ripple against a flood tide toward oligopoly. Despite the post-merger facts, defendant argues that this merger does not threaten to trigger others. It points to uncontradicted testimony that there are only two purely wholesale banks left in New York and to evidence that there are no remaining retail banks in the city with extensive branch systems. Defendant argues that these commercial realities speak against future mergers for the purpose of entering into, or integrating wholesale operations with, the local market. That argument flaunts the reasons for the present merger. If Hanover can merge with Manufacturers, either of the other wholesale banks can, with equal logic, merge with any of the integrated banks. More significantly, if Hanover was pressed to merge in order to get into full scale retail operations, surely this merger puts added pressure on smaller banks to do the same thing. The argument also ignores the urge to merge in order to follow customers into the suburbs where the mere presence of branch systems tempts easy cross-branching by merger. Finally, defendant argues that the New York omnibus banking law and the federal Bank Merger Act of 1960 were enacted to check future mergers which might have an anticompetitive effect, and that we ought not speculate that the agencies entrusted with their administration will ignore their duties, or that, if they do, the Departmentof Justice will allow bank mergers to go unchallenged now that Philadephia and Lexington have subjected them to the full thrust of the antitrust laws. We are mindful that in the decade 1950-1960, the Department of Justice, like Congress, the federal banking agencies, and most of the legal profession, believed that Clayton § 7 did not apply to bank mergers accomplished by the acquisition of assets, 224 and there was doubt as to the applicability of the Sherman Act. 225 The applicability of the antitrust laws, however, had not been authoritatively determined by the Supreme Court. Now that Philadelphia and Lexington have removed all doubt as to the applicability of both Clayton § 7 and Sherman § 1 to bank mergers, we share defendant’s confidence that if the banking agencies do approve any anticompetitive mergers, the Department of Justice will promptly bring suit to enjoin them. Nevertheless, we cannot conceive how the banking agencies or the court could fairly withhold approval of any future merger of smaller banks if this one were allowed to stand. However, the facts with respect to any future mergers, the need for them, the reasons for them, and their probable competitive effects are not before us. Our concern is not with speculation about the competitive effect of the next merger, but with a rational judgment of the probable effect of this one. This merger offends Clayton § 7 in so many ways that to allow it to stand would be to ignore the statute altogether. It tends to create a monopoly by significantly increasing concentration and accelerating a trend toward oligopoly. The case more than satisfies the rule that where concentration is already great, even slight increases must be prevented. United States v. Aluminum Co. of Ameri *950 ca, supra, 377 U.S. at 280, 84 S.Ct. 1283; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 367 n. 43, 83 S.Ct. 1715; Brown Shoe Co. v. United States, supra, 370 U.S. at 334, 82 S.Ct. 1502. It also runs afoul of the principle that where there is a strong trend toward oligopoly, further tendencies in that direction are to be curbed in their incipiency, whatever the number, or vigor, of remaining competitors. United States v. Continental Can Co., supra, 378 U.S. at 461, 84 S.Ct. 1738; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 367 n. 43, 83 S.Ct. 1715; Brown Shoe Co. v. United States, supra, 370 U.S. at 333, 345-346, 82 S.Ct. 1502. The merger substantially lessens competition and restrains trade by the permanent elimination of significant competition formerly existing between major competitors, and that in “itself constitutes a violation of § 1 of the Sherman Act,” and, a fortiori, of the Clayton Act. United States v. First Nat’l Bank & Trust Co. of Lexington, supra, 376 U.S. at 671-672, 84 S.Ct. 1037. It substantially lessens the vigor of competition as a whole for it not only eliminated a substantial and vigorous competitor from actual and potential competition, but also removed one of the few remaining competitors capable of challenging the dominant position of the five largest banks and threatening to intensify the struggle. United States v. Continental Can Co., supra, 378 U.S. at 461, 84 S.Ct. 1738; United States v. Penn-Olin Chem. Co., supra, 378 U.S. at 170, 84 S.Ct. 1710; United States v. El Paso Natural Gas Co., supra, 376 U.S. at 658-662, 84 S. Ct. 1044. It permanently deprives customers of a substantial and important banking alternative, thereby impairing the ability of small businesses and other retail customers to bargain for better terms. It is no answer to such a loss that large banks make a special effort to serve small customers, for, whatever the special effort, one branch of the same bank does not compete with another. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 369,83 S.Ct. 1715. Nor is it any answer that competition among surviving competitors is vigorous, for the more that individual firms coalesce, the greater the likelihood that a coalition and parallel policies of mutual advantage will emerge from the ashes of competition. United States v. Aluminum Co. of America, supra, 377 U.S. at 280, 84 S.Ct. 1283; United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 367 n. 43, 83 S.Ct. 1715. Finally, in the setting, this merger significantly increases concentration and speeds the trend toward oligopoly. There is no need to rely on a crystal ball to predict the probable future effect of the merger upon concentration and the probable impact of increased concentration upon competition. Hanover’s plight and fate, the scarcity of newcomers since 1950, and the insignificant shares wrested from entrenched competitors by late arrivals, bear telling witness to the dimmer prospects which will confront newcomers attempting to establish a foothold, the heightened obstacles to growth of smaller banks by internal expansion, and the added pressure on them to leave the market by merger. The level of concentration and momentum of the trend resulting from earlier “important mergers” demonstrate, we think, that one more “important merger” is too much for a market structure already close-hauled. In the circumstances, this merger creates a clear and imminent threat that a handful of banks are likely to sew the local market up tight unless we call a halt. We think it clear that the merger threatens to result in five firms with a decisive advantage in the competitive struggle, with inert power to impair the opportunities for new firms to enter into, and smaller firms to remain in, the market. A merger which imminently threatens to create inert power in so few, collectively, even though not collusively, to exclude competitors, restrict growth of rivals, and squeeze smaller firms out of the market, *951 is a combination in unreasonable restraint of trade violative of § 1 of the Sherman Act. FTC v. Motion Picture Advertising Serv. Co., 344 U.S. 392, 395, 73 S.Ct. 361, 97 L.Ed. 426 (1953). Cf. Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 328, 81 S.Ct. 623; Standard Oil Co. of California v. United States, supra, 337 U.S. at 309, 69 S.Ct. 1051; United States v. E. I. Du Pont De Nemours & Co., supra, 351 U.S. 389-393, 76 S.Ct. 994; American Tobacco Co. v. United States, 328 U.S. 781, 811 et seq., 66 S.Ct. 1125, 90 L.Ed. 1575 (1946); United States v. Aluminum Co. of America, supra, 148 F.2d 427-432. A fortiori, such a merger tends to create a monopoly violative of § 7 of the Clayton Act. Brown Shoe Co. v. United States, supra, 370 U.S. at 328-329, 82 S.Ct. 1502. We conclude, therefore, that the government has sustained its burden of proof, and that the merger violates both Clayton § 7 and Sherman § 1 in the local market as defined herein. We turn to determination of the competitive effect of the merger in the national market. H. The Effect of the Merger in the National Market. Competitive power, strength, and position in the national market depend on lending limits which, in turn, depend on deposits. 226 Our statistics (Appendix C), though spurious as to relative shares of the national market, are genuine as to total deposits and, therefore, demonstrate the true relative position, power, strength, and vigor of the national competitors. There can be no question that Manufacturers and Hanover were in substantial competition with each other in the national market and that both were vigorous, substantial, and major factors in competition generally. That conclusion is overwhelmingly demonstrated by a cursory comparison of the national business of the constituent banks which is set forth in Appendix A, 227 and by a comparison of their deposits, relative to those of the 199 remaining competitors, which is set forth in Appendix C. It is reinforced by the opinions of all of the banking agencies which clearly recognized that both banks engaged in wholesale banking in a nationwide area of effective competition. The merger, therefore, permanently eliminated substantial competition existing between major competitors and permanently removed a substantial competitor and a significant lead bank from competition as a whole. These findings of fact raise the question of whether the mere elimination of former competition between Manufacturers and Hanover without regard to other factors, violates Clayton § 7 and Sherman § 1. The government argues that since the competition between Manufacturers and Hanover was substantial, the permanent elimination of that competition by merger necessarily results in a substantial lessening of competition and, therefore, violates Clayton § 7. In support of its contention, the government relied originally on United States v. Koppers Co., 202 F.Supp. 437 (W.D.Pa.1962), and United States v. Bethlehem Steel Corp., supra. We think, however, that neither case stands for the proposition that the mere elimination of substantial competition between parties to a merger is itself a violation of Clayton § 7. Clayton § 7, prior to its amendment in 1950, focused upon the elimination of competition between the merging firms by proscribing acquisitions which might result in a lessening of competition between the acquiring and the acquired company. 228 The courts, however, stead *952 fastly refused to find a violation, even under old Clayton § 7, merely on the basis of the elimination of competition formerly existing between competitors. 229 The legislative history of the 1950 amendment to Clayton § 7 itself refutes the government’s theory 230 Significantly, during debate on the 1950 amendment, Senator Kefauver, a co-sponsor of the bill, said that a merger “would not be illegal merely because of the elimination of the competition which had previously existed between the acquiring and acquired firms. That is to say, within the section [of the country], the merger would have to have the effect of lessening competition generally.” 231 Defendant’s position is supported by Tampa Elec. Co. v. Nashville Coal Co., supra, 232 and the government’s theory definitively rejected by the Supreme Court in Brown Shoe Co. v. United States, supra, 370 U.S. at 335, 82 S.Ct. 1529, where the court said: “Where the arrangement effects a horizontal merger between companies occupying the same product and geographic market, whatever competition previously may have existed in that market between the parties to the merger is eliminated. Section 7 of the Clayton Act, prior to its amendment, focused upon this aspect of horizontal combinations by proscribing acquisitions which might result in a lessening of competition between the acquiring and the acquired companies. The 1950 amendments made plain Congress’ intent that the validity of such combinations was to be gauged on a broader scale: their effect on competition generally in an economically significant market.” (emphasis added). Finally, even Philadelphia, upon which the government so heavily relies, teaches that the competitive effect of a merger must be judged, not merely on the basis of the elimination of former competition between the merging banks, but on the basis of a firm understanding of the structure of the relevant market. Thus, the text of the statute, its legislative history, and authoritative interpretations by the Supreme Court, make clear that the critical question under Clayton § 7 is whether the merger may substantially lessen competition or tend to create a monopoly when its effects are assessed within the structure and history of the whole relevant market. 233 However, all *953 of the cases on which defendant relies were decided before United States v. First Nat’l Bank & Trust Co. of Lexington, supra, and we must reckon with the Supreme Court’s latest ruling on the question. The government argues that Lexington compels a decision that the mere elimination of former substantial competition between major competitors is per se a violation of Sherman § 1 and, therefore, necessarily offends Clayton § 7. Defendant, relying on United States v. Columbia Steel Co., supra, and a line of earlier cases, 234 contends that the government’s position is unsound as a matter of law, that we must assess the competitive effect of the merger in its setting, and determine its probable effect on competition in the whole structure of the relevant market, particularly the number, vigor, and strength of remaining competitors. There can be no question that in Columbia Steel the Supreme Court held that the validity of a horizontal merger under Sherman § 1 depended on an assessment of many factors other than the mere cessation of competition between former competitors. The teaching of the case is: “In determining what constitutes unreasonable restraint, we do not think the dollar volume is in itself of compelling significance; we look rather to the percentage of business controlled, the strength of the- remaining competition, whether the action springs from business requirements or purpose to monopolize, the probable development of the industry, consumer demands, and other characteristics of the market. We do not undertake to prescribe any set of percentage figures by which to 'measure the reasonableness of a corporation’s enlargement of its activities by the purchase of the assets of a competitor. The relative effect of percentage command of a market varies with the setting in which that factor is placed.” 334 U.S. 527-528, 68 S.Ct. 1124. Looking to the factors specified in Columbia Steel, we find: the dollar volume great, but not of compelling significance ; defendant does not control an undue, dominant, or decisive share of the national market; 235 the merger springs from legitimate business requirements to enable defendant to satisfy the credit needs of very large customers and to compete more vigorously as a lead bank against larger competitors; 236 there is no purpose to monopolize, curtail competition, or control prices; the structure of the national market is one of many enterprises, none of which has any significant advantage over its next smaller competitor ; 237 newcomers abound; 238 the rela *954 tive aggregate share of the six largest banks has remained fairly consistent since 1950 and has declined since the merger; 239 the market is not oligopolistic, and this merger does not threaten monopoly or oligopoly; 240 the government has failed to prove a merger trend toward concentration in the national market; 241 small competitors flourish; and the national market and number of competitors is growing by leaps and bounds. 242 Defendant’s increased lend *955 ing limits reduced the advantage of the three larger competitors, but handicapped smaller competitors (DX 1-Mem-orandum, p. 10), though the advantage over smaller competitors did not threaten to become decisive, for the Federal Reserve Board found that the pro-competitive effect outweighed the anticompeti-tive effect in the same market. We incline to agree. See Appendix E. There is neither need nor purpose, however, to consider or determine whether the Columbia Steel factors just enumerated offset the anticompetitive effect resulting from the elimination of substantial competition between major competitors, the elimination of a substantial factor from competition as a whole, and the customers’ loss of an important alternative lead bank. That course has been foreclosed by Lexington, where the Supreme Court held that the Columbia Steel case must be confined to its special facts, and that “where merging companies are major competitive factors in a relevant market, the elimination of significant competition between them, by merger or consolidation, itself constitutes a violation of § 1 of the Sherman Act.” 376 U.S. at 672, 84 S.Ct. 1037 (emphasis added). This case falls squarely within that rule. Defendant argues that the market facts in Lexington and those in the old railroad cases on which Lexington relies, 243 have no relevance to the market facts here, and that the elimination of one out of six competitors may very well constitute an unreasonable restraint of trade, but that no such restraint is inherent in the mere elimination of one out of at least 200 competitors, particularly where the merger has not resulted in anything like a dominant 50% market share and does not threaten monopoly or oligopoly. Defendant also argues that four of the Justices in Lexington were unwilling to predicate decision solely upon the factor of the elimination of competition between the parties to the merger, and that even the majority relied also on the presence of the other factors specified in Columbia Steel. Defendant urges that if the Sherman Act were so easily satisfied, there was no need to enact even the original Clayton § 7, and refers us to the legislative history of amended Clayton § 7 which shows that Congress intended to make the test of illegality under Clayton § 7 easier than under Sherman § 1. The argument proceeds that it is anomolous to apply an easier test under the Sherman Act than would be warranted under the Clayton Act. In that connection, defendant points to United States v. Penn-Olin Chem. Co., supra, 378 U.S. at 177, 84 S. Ct. 1720 decided under Clayton § 7 and after Lexington, where the Supreme Court, in remanding the case, directed the district court to observe a number of factors in assessing the probability of a substantial lessening of competition, including “the number and power of the competitors in the relevant market”. Whatever the merits of defendant’s arguments, they are addressed to the wrong forum, for we must apply the Sherman Act as it is interpreted, and in accordance with rules enunciated, by the Supreme Court. We hold, therefore, that the elimination of substantial competition previously existing between the parties to this merger in the national market itself constitues an unreasonable restraint ox trade, violative of § 1 of the Sherman Act, and, a fortiori, establishes a reasonable probability of a substantial lessening of competition, violative of § 7 of the Clayton Act. *956 CONCLUSIONS 1. This court has jurisdiction to determine the cause. 2. The approval of the merger by the Board of Governors of the Federal Reserve System is not a final and exclusive determination of its legality. 3. The merger of Manufacturers and Hanover is a combination in restraint of trade in commercial banking in the New York metropolitan area, as defined herein, and in the United States, in violation of Section 1 of the Sherman Act. 4. There is a reasonable probability that the merger of Manufacturers and Hanover may substantially lessen competition or tend to create a monopoly in commercial banking in the New York metropolitan area, as defined herein, and in the United States, in violation of Section 7 of the Clayton Act. The foregoing opinion, footnotes, and Appendices constitute the court’s findings of fact and conclusions of law, as required by Rule 52(a), Fed.R.Civ.P., 28 U.S.C. Both sides have submitted numerous proposed findings of fact, many of which we find inconsistent with our findings, irrelevant, immaterial, bad in form, or not supported by the evidence. We, therefore, reject all of them except to the extent that they are embodied within this opinion. The parties may submit a reasonable number of further findings of fact within twenty (20) days from the filing of this opinion, in accordance with, and supplemental to, but not inconsistent with, this opinion. All pending motions to strike opinion evidence and conclusions respecting the competitive effect of the merger are granted, except as to the opinions of the banking agencies. All other motions to strike are denied, as is the motion to dismiss the complaint. If the parties are able, within ten (10) days, to agree on appropriate relief and the form of the decree to be entered, a decree may be submitted for our consideration, otherwise either party may apply to the court, by notice of motion, within ten (10) days from the filing of this opinion, and the court will set a time and conduct hearings to determine the equitable relief necessary and appropriate in the public interest to eliminate the effects of the merger offensive to Section 1 of the Sherman Act and Section 7 of the Clayton Act. *957 *958 *959 *960 *961 Footnotes [12] “This [the Bank Merger Act] puts the responsibility for acting on a proposed merger where it belongs — in the agency charged with supervising and examining the bank which will result from the merger. Out of their years of experience in supervising banks, our Federal banking agencies have developed specialized knowledge of banking and the people who engage in it. They are experts at judging the condition of the banks involved, their prospects, their management, and the needs of the community for banking services. They should have primary responsibility in deciding whether a proposed merger would be in the public interest.” 106 Cong.Ree. 7257 (1960) (emphasis supplied). See S.Rep.No. 196, 86th Cong., 1st Sess. 3-8 (1959), where the Committee approves the statutory scheme placing banking under the agencies. See also id. at 21; 106 Cong.Ree. 9711 (1960) (remarks of Senator Fulbright). [1] 15 U.S.C. § 18 (1958). Section 7 of the Clayton Act provides in pertinent part: “No corporation engaged in commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital and no corporation subject to the jurisdiction of the Federal Trade Commission shall acquire the whole or any part of the assets of another corporation engaged also in commerce, where in any line of commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.” [2] 15 U.S.C. § 1 (1958). Section 1 of the Sherman Act provides in pertinent part: “Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal * * [3] This action was one of several instituted by the Department of Justice around the country to test the applicability of the Clayton and Sherman Acts to bank mergers. The complaint was not filed and injunctive relief was not sought until a half hour after the merger had been consummated, although the government had been on notice of the proposal for many months. My Brother Cashin denied an application for a temporary restraining order on Saturday, September 9, 1961, after argument, apparently on the grounds that the government’s application was tardy and failed to show a reasonable probability of ultimate success in view of the novelty of the case and the then unsettled state of the law. Cf. United States v. Crocker-Anglo Nat’l Bank, 228 F.Supp. 849 (N.D.Cal.1963). An application for a preliminary injunction reached the Motion Part on Tuesday, September 12, 1961, after the merger had been in effect and the assets scrambled for four days. The government, therefore, withdrew the application for a preliminary injunction - and applied for an alternative order directing segregation of the assets of the merged banks pendente Ute. On oral argument, counsel represented that there was relatively little dispute about the evidentiary facts. The court, therefore, suggested immediate pretrial conference with a view to expediting procedures for the submission of additional evidence and a prompt trial on the merits. Conferences were held, procedures devised for presentation of additional evidence, and a timetable for argument of motions and trial was arranged. Accordingly, the government withdrew its application for segregation of the assets pendente lite. Most of the evidence was presented by stipulation and affidavits subject to the opposing party’s right to call any affiant for cross-examination. Trial began on December 6, 1961, and further hearings were held on December 8 and 14, 1961 and March 8, 1962. Proposed findings of fact, conclusions of law, and briefs were submitted during January, February, March and April 1962. At that juncture, the court became engaged in a protracted conspiracy trial which continued through June 1962. See United States v. Bent-vena, 319 F.2d 916 (2 Cir.), cert, denied, 375 U.S. 940, 84 S.Ct. 345, 11 L. Ed.2d 271 (1963). In the interim, on May 21, 1962, the Supreme Court noted probable jurisdiction in United States v. Philadelphia Nat’l Bank, 369 U.S. 883, 82 S.Ct. 1156, 8 L.Ed.2d 285 (1962), and, with counsels’ consent, this matter was held under advisement awaiting the Philadelphia opinion. The opinion was rendered on June 17, 1963, holding the Clayton Act applicable to bank mergers, 374 U.S. 321, 83 S.Ct. 1715, 10 L.Ed.2d 915, but leaving for later decision in United States v. First Nat’l Bank & Trust Co. of Lexington the question of whether the Sherman Act also applied. Following Philadelphia, at counsels’ request, the *876 record was reopened for presentation of supplementary evidence by both sides. Further testimony was taken on September 4, 1963, and revised findings and briefs were submitted during September and October 1963. Final argument was held on January 17, 1964. The Lexington opinion came down in April 1964, 376 U.S. 665, 84 S.Ct. 1033, 12 L.Ed.2d 1, and was followed by a series of merger decisions during April and June, all of which have an important bearing on this case. A number of informal post-trial conferences were held with counsel, the last on December 16, 1964, for the purpose of receiving voluminous stipulated supplementary evidence material to issues raised by the series of post-trial decisions. [4] N.Y. Banking Law, McKinney’s Consol. Laws, 6.2, § 601-b provides in pertinent part: “In determining whether to so approve [a bank merger], the superintendent shall take into consideration * * * whether the effect of such merger or acquisition shall * * * result in a concentration of assets beyond limits consistent with effective competition, * * * whether such merger or acquisition may result in such a lessening of competition as to be injurious to the interests of the public or tend toward monopoly and * * * primarily, the public interest and the needs and convenience thereof.” [5] The Bank Merger Act, 74 Stat. 129 (1960), 12 U.S.C. § 1828 (c) (Supp. V, 1964), amending the Federal Deposit Insurance Act § 2 [18] (c), 64 Stat. 892 (1950), provides in pertinent part: “No insured bank shall merge or consolidate with any other insured bank * * * without the prior written consent (i) of the Comptroller of the Currency if the acquiring, assuming, or resulting bank is to be a national bank or a District bank, or (ii) of the Board of Governors of the Federal Reserve System if the acquiring, assuming, or resulting bank is to be a State member bank (except a District bank), or (iii) of the Corporation if the acquiring, assuming, or resulting bank is to be a nonmember insured bank (except a District bank). ' * • * In granting or withholding consent under this subsection, the Comptroller, the Board, or the Corporation, as the case may be, shall consider the financial history and condition of each of the banks involved, the adequacy of its capital structure, its future earnings prospects, the general character of its management, the convenience and needs of the community to be served, and whether or not its corporate powers are consistent with the purposes of this chapter. In the ease of a merger * » * the appropriate agency shall also take into consideration the effect of the transaction on competition (including any tendency toward monopoly), and shall not approve the transaction unless, after considering all of such factors, it finds the transaction to be in the public interest. In the interests of uniform standards, before acting on a merger, * * * the agency * * * shall request a report on the competitive factors involved from the Attorney General and the other two banking agencies referred to in this subsection * * *. The Comptroller, the Board, and the Corporation shall each include in its annual report to the Congress a description of each merger * * * approved by it during the period covered by the report, along with the following information: the name and total resources of each bank involved; whether a report has been submitted by the Attorney General hereunder, and, if so, a summary by the Attorney General of the substance of such report; and a statement by the Comptroller, the Board, or the Corporation, as the case may be, of the basis for its approval.” [6] Section 4 of the Sherman Act, 15 U.S.C. § 4 (1958), provides in pertinent part: “The several district courts of the United States are invested with jurisdiction to prevent and restrain violations of sections 1-7 of this title * * Section 15 of the Clayton Act, 15 U.S.C. § 25 (1958), provides in pertinent part: “The several district courts of the United States are invested with jurisdiction to prevent and restrain violations of this Act, and it shall be the duty of the several United States Attorneys, in their respective districts, under the direction of the Attorney General, to institute proceedings in equity to prevent and restrain such violations.” [7] Defendant does not abandon the point and is entitled to preserve it for appeal especially in view of the fact that its abandonment in Philadelphia was obviously a tactical choice by counsel for the banks, and, as a result, the Bank Merger Act lost the day without the benefit of adversary argument. [8] United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 350, 83 S.Ct. 1715 (and cases cited therein); Pan American World Airways, Inc. v. United States, supra (dissenting opinion); California v. FPC, 369 U.S. 482, 485, 82 S.Ct. 901, 8 L.Ed.2d 54 (1962); United States v. Borden Co., supra; Terminal Warehouse Co. v. Pennsylvania R.R., 297 U.S. 500, 513-515, 56 S.Ct. 546, 80 L.Ed. 827 (1936); Central Transfer Co. v. Terminal R.R. Ass’n, 288 U.S. 469, 474-475, 53 S.Ct. 444, 77 L.Ed. 899 (1933); Keogh v. Chicago & N. W. Ry., 260 U.S. 156, 161-162, 43 S.Ct. 47, 67 L.Ed. 183 (1922); United States v. Trans-Missouri Freight Ass’n, 166 U.S. 290, 314, 17 S.Ct. 540, 41 L.Ed. 1007 (1897); Henderson’s Tobacco, 78 U.S. [11 Wall.] 652, 657, 20 L.Ed. 235 (1870); United States v. Ty-nen, 78 U.S. [11 Wall.] 88, 92, 20 L.Ed. 153 (1870). [9] Thus, the Bank Merger Act contains no specific provision repealing Clayton § 11; nor does it immunize approved mergers from § 7 or § 11. The legislative history “of the Act seems clearly to refute any suggestion that the applicability of the antitrust laws was to be affected.” The Act gives no authority to enforce the antitrust laws; nor does it state what weight is to be given to the competitive factor or provide for hearings or judicial review. [10] See United States v. El Paso Natural Gas Co., 376 U.S. 651, 664, 84 S.Ct. 1044, 12 L.Ed.2d 12 (1964) (Harlan, X, dissenting), suggesting that this “duplica-tive” anachronistic system of dual regulation should be reexamined. The du-plicative work which this very case imposes upon the parties and this congested court is proof positive of the merits of the suggestion. [11] Section 11 of the original Clayton Act was no mere accident or afterthought. Rather, its terms mirror a very basic conflict which then divided the houses of Congress. The bill, as first proposed by the House, would have made all violations, including Section 7, punishable by fine or imprisonment, H.Rep.No. 627, 63d Cong., 2d Sess. 3 (1914). This penal approach was rejected by the Senate which, by amendment, struck the criminal sanctions and substituted in their place enforcement by governmental agency, S. Rep.No. 698, 63d Cong., 2d Sess. (1914). “This was done because it was thought best, especially in view of the experimental stage of this legislation, that the harshness of the criminal law should not be applied but that enforcement * * * should be given to * * * [the appropriate agency].” Id. at 43. The Federal Reserve Board, however, was not at that time numbered among those agencies. The Board’s absence is explained by the Senate’s treatment of the House proposals. Included in the initial draft was a provision prohibiting interlocking directorates between certain categories of banks but this was deleted by the Senate. The Senate “believed that such legislation as this more properly belongs to the domain of banking rather than of commerce and such additional regulation * * * as may be wise and just should be made by amendments to the national bank acts, and the enforcement of it given to the Comptroller of the Currency and the Federal Reserve Board.” Id., at 48. The House vigorously opposed what it called the Senate’s process of “emasculation,” but the Senate was equally adamant in defense of its position, 51 Cong. Rec. 16170, 16344 (1914). The gist of the arrangement, in pertinent part, was that the Senate agreed to the inclusion [12] See note 12 on page 881. *881 of banks in the scheme of antitrust legislation, while the House acceded to the Senate’s method of enforcement. And it is here, in the Conferees’ rewritten Section 11, that the Federal Reserve Board first appears as an enforcing arm of the statute. S.Doc.Nos. 583 & 585, 63d Cong., 2d Sess. 6 (1914). It is obvious, then, from even a cursory examination of the Clayton Act, through its stages of development to the final compromise, that it was the Senate’s insistence upon keeping the control of banking, including antitrust matters, in the hands of banking experts that carried the day in Congress- and therefore represents its true intent. [13] “‘Primary jurisdiction,’ * * * applies where a claim is originally cognizable in the courts, and comes into play whenever enforcement of the claim requires the resolution of issues which, under a regulatory scheme, have been placed within the special competence of an administrative body; in such a case the judicial process is suspended pending referral of such issues to the administrative body for its views. General American Tank Oar Corp. v. El Dorado Terminal Co., 308 U.S. 422, 433 [60 S.Ct. 325, 331, 84 L.Ed. 361], “No fixed formula exists for applying the doctrine of primary jurisdiction. In every case the question is whether the reasons for the existence of the doctrine are present and whether the purposes it serves will be aided by its application in the particular litigation. These reasons and purposes have often been given expression by this Court. In the earlier cases emphasis was laid on the desirable uniformity which would obtain if initially a specialized agency passed on certain types of administrative questions. See Texas & Pacific R. Co. v. Abilene Cotton Oil Co., 204 U.S. 426 [27 S.Ct. 350, 51 L.Ed. 553]. More recently the expert and specialized knowledge of the agencies involved has been particularly stressed. See Far East Conference v. United States, 342 U.S. 570 [72 S.Ct. 492, 96 L. Ed. 576]. The two factors are part of the same principle, ‘now firmly established, that in cases raising issues of fact not within the conventional experience of judges or cases requiring the exercise of administrative discretion, agencies created by Congress for regulating the subject matter should not be passed over. This is so even though the facts after they have been appraised by specialized competence serve as a premise for legal consequences to be judicially defined. Uniformity and consistency in the regulation of business entrusted to a particular agency are secured, and the limited functions of review by the judiciary are more rationally exercised, by preliminary resort for ascertaining and interpreting the circumstances underlying legal issues to agencies that are better equipped than courts by specialization, by insight gained through experience, and by more flexible procedure.’ Id., at 574-575 [72 S.Ct. 494]. “The doctrine of primary jurisdiction thus does ‘more than prescribe the mere procedural time table of the lawsuit.’ ” United States v. Western Pac. R. Co., 352 U.S. 59, 63-65, 77 S.Ct. 161, 165, 1 L.Ed.2d 126 (1956). [14] “Section 7 of the Clayton Act covers only acquisitions of bank stocks and since bank mergers rarely, if ever, involve stock acquisitions, they are beyond the scope of that Act. Thus, there is presently a gap in the coverage of bank mergers by Federal Law.” Hearings on S. 1062 Before Senate Committee on Banking and Currency, 86th Cong., 1st Sess. 65 (1959) (statement of J. L. Robertson, Member, Board of Governors, Federal Reserve System). [15] There is no claim here under Section 2 of the Sherman Act. [16] “This report was required in order that the Attorney General’s knowledge and background of experience in the field of anticompetitive and monopolistic matters should be made available to the banking agencies in their consideration of bank mergers.” 106 Cong.Rec. 9712 (1960). See S.Rep.No. 196, 86th Cong., 1st Sess. 23, 24 (1959). [17] “The immediate effects of the proposed merger of Manufacturers Trust and Hanover will be to substantially increase the high degree of banking concentration in the financial capital of the nation, eliminate substantial existing and potential competition between the merging banks, further adversely affect the ability of smaller banks to effectively compete with New York’s large banks, and increase the tendency toward monopoly (oligopoly) in commercial banking in New York City and the New York metropolitan area. It will also have an adverse effect on commercial banking in the United States as a whole.” [18] “In any given merger, competitive factors unfavorable to the merger may be outweighed by banking factors favorable to the merger. * * *” 106 Cong.Ree. 9712 (1960) (remarks of Senator Fulbright). “In cases such as these the benefits would be demonstrable, the merger would be a ‘good merger,’ and approval should be granted in spite of the lessening of competition.” Congress wanted “to make crystal clear its intention that the various banking factors in a particular case may be held to outweigh the competitive factors, and that the competi-five factors, however favorable or unfavorable, are not, in and of themselves, controlling on the decision.” S.Rep.No. 196, 86th Cong., 1st Sess. 20, 24 (1959). [19] United States v. Jerrold Electronics Corp., 187 F.Supp. 545, 565 (E.D.Pa. 1960), aff’d, 365 U.S. 567, 81 S.Ct. 755, 5 L.Ed.2d 806 (1961); United States v. Bethlehem Steel Corp., 168 F.Supp. 576, 617-618 (S.D.N.Y.1958); see United States v. E. I. Du Pont De Nemours & Co., 353 U.S. 586, 589, 77 S.Ct. 872, 1 L. Ed.2d 1057 (1957). The government originally seized upon the difference in standards to support its argument that the Bank Merger Act did not repeal Section 7 by implication. That argument, we think, carried to its logical conclusion, euts the other way for the difference in standards is precisely the source of a clear conflict between later and earlier statutes on the same subject. That might well lead to repeal of the antitrust laws by implication or resort to the doctrine of primary jurisdiction to avoid that drastic result. Cf. Pan American World Airways, Inc. v. United States, supra. See Wemple and Cutler, The Federal Bank Merger Law and the Antitrust Laws, 16 Bus.Law. 994 (1961). [20] We think, in the circumstances, that the Attorney General’s report is akin to intervention under Olayton § 11 and the Board’s decision tantamount to a summary judgment from which the Attorney General could have appealed under Clayton § 11 or, perhaps, the Administrative Procedure Act, 5 U.S.C. § 1009 (1958). See 1 Davis, Administrative Law § 7.07 (1958) ; 3 id. § 22.15. Were it not for dual enforcement and the lack of a “pervasive regulatory scheme” governing the commercial banking industry, the government’s grievances in this court might be barred on principles of collateral estoppel or res judicata. Cf. United States v. Radio Corp. of America, supra; Pan American World Airways, Inc. v. United States, supra. See also 2 Davis, Administrative Law § 18.02, at 554-55 (1958). [20] (a). Olayton § 7 provides that nothing in it shall apply to transactions duly consummated pursuant to authority given by the CAB, the FOG, the FPO, the ICC, the SEC, the USMC, and the Secretary of Agriculture under any statutory provision vesting such power in such Commission, Secretary, or Board. The obvious reason for the omission of the FRB was that at the time of the amendment to § 7 in 1950, the Board had no authority to grant prior approval, a flaw in the Clayton Act corrected by the Bank Merger Act. [21] “The advance approval feature is important in halting bank acquisitions before they are consummated and in preserving the depositors’ confidence in an institution which might otherwise be destroyed by an attempt to unscramble assets after an acquisition has been completed.” S.Rep.No. 196, 86th Cong., 1st Sess. 22 (1959) (emphasis added). Quoting Berle, Banking Under the Anti-Trust Laws, 49 Colum.L.Rev. 589, 592 (1949), Congress noted that: “A bank failure is a community disaster, however, wherever, and whenever it occurs.” S.Rep.No. 196, 86th Cong., 1st Sess. 18, 19 (1959). Congress was fully cognizant that “the very nature of the banking business would make it well nigh impossible to restore the status quo.” Hearings on S. 1062 Before Senate Committee on Banking and Currency, 86th Cong., 1st Sess. 66 (1959) (statement of J. L. Robertson, Member, Board of Governors, Federal Reserve System). See Funk, Antitrust Legislation Affecting Bank Mergers, 75 Banking L.J. 369, 381 (1958). [22] New York’s population of 7,781,984 is exceeded by that of Tokyo, Japan with 9,311,774 (May 1960) and London, England with 8,222,340 (June 1958). New York World-Telegram and The Sun, The World Almanac and Book of Facts 1961 at 82 and 396 (hereafter referred to as The World Almanac and Book of Facts). [23] Contrast 2,002,512 for Philadelphia, Pennsylvania, and 62,810 for Lexington, Kentucky. The World Almanac and Book of Facts, at 82 and 89. [24] Those counties which besides the five counties of the city (New York, Kings, Queens, Bronx, and Richmond) comprise the metropolitan area (in New York State: Suffolk, Nassau, Westchester, and Rockland; in New Jersey: Essex, Bergen and Hudson), alone account for an additional 5.2 million people, while the population of the entire State of New York is 16.7 million. The World Almanac and Book of Facts, at 113-14; Dep’t of Commerce, State of New York, This is • New York State, pp. 16-17 (1962). [25] Over three-quarters of the Port of New York is located in New York City (578 miles of water frontage); the remainder is in New Jersey. The total value of tonnage for 1961 was $9,354,584,000; the exact number of ship movements 26,342. Dep’t of, Commerce and Industrial Development, City of New York, Statistical Guide for New York City, 1962 (hereafter referred to as Statistical Guide), at 53, 55. See New York Chamber of Commerce Bull., Vol. 55, No. 3, Oct. 1963, p. 166. [26] Kennedy (New York International) and LaGuardia Airports handled 10,147,056 and 3,292,803 passengers, respectively, in 1961. Statistical Guide, p. 54. [27] Statistical Guide, pp. 72-73. See New York Chamber of Commerce Bull., Vol. 55, No. 3, Oct. 1963, pp. 166-167; Madden, The Money Side of “The Street,” at 12-14 (Federal Reserve Bank of New York, Pub.1959) (hereafter referred to as Madden, The Money Side of “The Street”); Dep’t of Commerce, State of New York, Business Fact Book, 1962, Pt. I, at 1-7 (hereafter referred to as Business Fact Book). [28] New York Chamber of Commerce Bull., Vol. 55, No. 3 Oct. 1963, pp. 166-167; Madden, The Money Side of “The Street,” p. 13. [29] New York Chamber of Commerce Bull., Vol. 55, No. 3, Oct. 1963, p. 167. [30] Ibid. [31] Statistical Guide, p. 57; Dep’t of Commerce, State of New York, This is New York State, p. 34. [32] Statistical Guide, p. 58. [33] Statistical Guide, p. 64; Dep’t of Commerce and Public Events, City of New York, New York — The City that Belongs to the World, p. 43 (1956). [34] Business Fact Book, p. 6. [35] Total payroll for 1958 of $4,030,354,-000 compared to 1960 foreign aid expenditure of $4 billion. Information Please 'Almanac, Atlas and Yearbook, 1962, p. 589. [36] Contrast $3.75 billion for Philadelphia and $91 million for Lexington, Dep’t of Commerce Bureau of Census, County and City Data Book, 1962, pp. 507 and 557. [37] Exact figure $24,503,800. Dep’t of Commerce, State of New York, Personal Income in Counties of New York State 1960, p. 10, Research Bull. No. 3, Apr. 1962. [38] The No. 2 metropolis, the Los Angeles-Long Beach area, had individual incomes for 1961 of $16.576 billion, less than half that of New York; Chicago-Evanston, the No. 3 metropolis, $15.4 billion; Philadelphia-Camden, the No. 4 area, $8.9 billion; and Detroit, the fifth largest, $7.5 billion. N.Y. Times, Mar. 29,1964, p. 46. [39] Statistical Guide, p. 65 ; Business Fact Book, p. 1. [40] Statistical Guide, p. 69. [41] Business Fact Book, p. 3. [42] Statistical Guide, p. 3. [43] Id. at 60 and 80. [44] Nadler, The Banking Situation in New York State, 11 (1956) ; DX8 p. 3. [45] Philadelphia has a population of 2,002,-512. The World Almanac and Book of Facts, pp. 82, 113-114. The population of either county dwarfs the 131,906 population of Fayette County, Kentucky, id. at 109, and the Westchester city of Yonkers alone, with a population of 190,634, is more than three times the size of Lexington and nearly a fifth as large as Philadelphia. id. at 82, 89 and 95. [46] U.S. Dep’t of Commerce, Survey of Current Business, 9 (1963). [47] gee Nadler, The Banking gituation in New York State, 129 (1956). [48] Madden, The Money Side of “The Street,” p. 12. [49] Id. at 22. [50] Board of Governors, Federal Reserve System, The Federal Reserve System: Purposes and Functions, Feb. 1961 at 5 et seq. (hereafter referred to as The Federal Reserve System: Purposes and Functions). [51] Commercial banks provide many services. They receive savings and other time deposits, as well as demand deposits in the form of regular and special checking accounts; make consumer single payment and installment loans, real estate loans, commercial and= industrial loans, agricultural loans, loans to purchase or carry securities, and loans to other banks and other financial institutions; invest substantial amounts in governmental and other securities; provide international banking services, including commercial letters of credit, commercial remittances abroad, and travelers checks; serve as correspondent banks for other domestic commercial banks; serve as trustees under voluntary and court trusts, agency and custody accounts, pension and other employee welfare funds, and corporate securities issues; and act as stock transfer, registrar, dividend disbursement and bond and coupon paying agencies. [52] Individual member banks of the Federal Reserve System are required to hold only a fraction of their deposits as reserves and may loan the balance. If the borrower pays the proceeds of the loan to someone who deposits them in a hank, the process is repeated. Thus, as money cycles from deposit to loan through the banking system as a whole, deposits are generated and credit created on a multiple scale, as is illustrated by the following chart: [53] Member banks of the Federal Reserve System are prohibited by law from paying interest on demand deposits. 12 U. S.C. § 371a (1958). [54] Madden, The Money Side of “The Street,” p. 18. [55] Federal Reserve Bull. No. 47, p. 314 (1962). [56] The Federal Reserve System: Purposes and Functions, pp. 8 and 24. [57] 12 U.S.C. §§ 26, 27, 29, 51, 1828(c) (1958 & Supp. Y, 1964); N.Y. Banking Law, §§ 10,22,24,105, 601-b. [58] N.Y. Banking Law, § 108. [59] This is accomplished by the Federal Reserve’s buying and selling government securities daily and weekly in the open market, and by changes in the discount rates and reserve requirements. 12 U. S.C. §§ 263(c), 353-359, 462, 462b (1958). [60] N.Y. Banking Law, ■§ 14, subd. 1 (h) (1); 12 U.S.C. §§ 371a, 371b (1958). [61] N.Y. Banking Law, §§ 14, subd. l(m), 37; 12 U.S.C. §§ 161, 324, 1817(a), 1817 (c) (1958 & Supp. Y, 1964). [62] N.Y. Banking Law, §§ 36, 38; 12 U.S. C. §§ 325, 481, 483, 1820(b) (1958 & Supp. V, 1964). [63] 12 U.S.C. §§ 301, 1818(a), (b) (1958); N.Y. Banking Law, §§ 39, 40. [64] Madden, The Money Side of “The Street,” at 19, 25 and 31. [65] Id. at 22 et seq. [66] By way of comparison, forty-two banks in. the Philadelphia metropolitan area held $4.3 billion in deposits, $2.7 billion in loans, and $5 billion in assets. See United States v. Philadelphia Nat’l Bank, 201 F.Supp. 348, 355, 363 (E.D.Pa.1962). Six banks in Eayette County, Kentucky held $146,445,000 in deposits, $88,107,000 in loans, and $163,378,000 in assets. See United States v. First Nat’l Bank & Trust Co. of Lexington, 208 F.Supp. 457, 459-460 (E.D.Ky.1962). [67] During the period 1956 to 1960, Manufacturers increased its current operating income by 43.9% and averaged $42,000,-000 per year in net earnings between 1958 and 1960. During the period 1956 to 1960, Hanover had a 22.7% increase in net operating income and averaged $33,000,000 per year in net earnings between 1958 and 1960. [68] DX 40, pp. 8-9; DX 44, p. 3, para. 6; PX 28. [69] DX 1-Memorandum, p. 17. [70] On October 13, 1950 Manufacturers merged with Brooklyn Trust Co. The transaction involved banking assets of $244,069,821. [71] The five minor mergers involved banking assets aggregating $85,792,141, and the last occurred twelve years ago, in 1953. [72] DX 12, p. 2. [73] Immediately after the merger, Chase Manhattan held 20.2% of the total assets of commercial banks in the metropolitan area. First National City 17.7%, and Manufacturers Hanover 13.6%. Since the merger, Manufacturers Hanover’s share has declined to 11.5%, and the five largest banks from 71.1% to 70.8% (stipulation filed December 10, 1964). Since the merger, the asset, deposit and loan growth of Manufacturers Hanover has not kept pace with that of its smaller competitors (DX 53). Manufacturers Hanover’s percentage growth in assets from 1960 to 1961 was 1.9%, and from 1960 to 1962 was 1.6%. In contrast, the percentage growth in assets of its four largest local competitors for the same periods was 11.4% and 17.3%, respectively, for all other metropolitan area banks was 12.3% and 20.0%, respectively, and for all other New York City banks was 12.2% and 17.3%, respectively (DX 53). [74] DX 56; DX 64. [75] N.Y. Banking Law, § 200. [76] There are no statistics in evidence respecting the deposits and loans of the branches opened by the foreign banks. [77] Morgan Guaranty (N.Y. Times, May 12, 1964, p. 15) ; Bank of New York (DX 42) ; Belgium American Bank & Trust Co.; and Royal Bank of Canada & Trust Co. (Tr. 246). [78] There are no statistics showing the distribution of branches throughout the metropolitan area either at the time of the merger or since. Our later analysis of branch concentration is, therefore, based on the distribution of branches in the City of New York, and the number of branches considered later differs from the number in the above text. [79] DX 59; DX 61. [80] (1) Demand Deposits. Regular Checking Accounts. (2) Commercial and Industrial Loans. (3) Single Payment Loans to Individuals. (4) Loans to Brokers and Dealers to Purchase and Carry Securities. (5) Loans to Finance Companies. (6) Bankers’ Acceptances. (7) Personal and Corporate Trust Services. [81] Madden, The Money Side of “The Street,” p. 28. See Business Week, Nov. 17, 1956, p. 51. [82] With certain exceptions related to secured loans, no national bank may loan to any one borrower a sum more than *897 10% of the bank’s unimpaired surplus and paid-in capital stock. 12 U.S.C. § 84 (1958 & Supp V, 1964). No New York State bank may lend to any one borrower a sum greater than 10% of the aggregate of the bank’s capital, surplus, and undivided profits. N.Y. Banking Law, § 103. [83] A bank’s overall ability to lend is determined by the aggregate amount of its deposits. The law, however, requires that a percentage of deposits be kept as reserves at Federal Reserve banks. These percentages vary according to the member bank’s location and are subject to change by the Federal Reserve Board. 12 U.S.C. §§ 462, 462b (1958 & Supp V, 1964). The Board, by increasing or decreasing the percentage can effect an industrywide expansion or contraction of credit. The excess of deposits over reserve balances, therefore, measures a bank’s credit capacity. [84] Funds deposited with the Federal Reserve Bank in excess of reserve requirements. These are the most liquid of assets as they are transferred merely by immediate adjustment of balances as opposed to Clearing House checks where adjustments are not made until the following day. Banks with an excess of federal funds will lend them at interest to banks with a deficiency for periods as short as a day. The usual unit of trading is $1 million, but as the usual lending limits must be complied with trading is generally limited to the largest banks. Despite the tremendous dollar volume traded, the number of transactions is small, so there is no special market. Stock Exchange brokers and New York City banks themselves act as brokers. It has become the major way for banks to invest temporarily idle cash because it enables a bank to gain interest on money held only a day or two. Madden, The Money Side of “The Street,” pp. 42-46. [85] A treasury hill is a bill of exchange drawn on the United States. They are issued weekly by the Treasury and payable in 91 or 182 days, without interest. Profit is the difference between face value and the discount price set by competitive bidding which is centered in the New York area. The popularity of the bill stems from its ability to meet the need for a temporary investment. The interest yield is low, but the investment is safe and quickly liquidable. Madden, The Money Side of “The Street,” pp. 47-50. [86] Madden, The Money Side of “The Street,” pp. 63-67. [87] The big banks have a special problem which pulls them into the open market. The problem arises from the fact that large deposits and withdrawals of correspondent hanks, big business firms, and government securities dealers are often unpredictable. This results in fluctuating cash flows and unexpected excesses and deficiencies of funds in large amounts that may last for only a day or two. Excess funds resulting from huge deposits cannot lie idle even for a day without courting disaster. Yet the whole of such deposits may be withdrawn either on demand or on a few days’ notice. The banks must, therefore, provide for unexpected drains. The open market, with its safe, but quickly liquidable, investments provides the answer. Madden, The Money Side of “The Street,” p. 31 et seq. [88] See Nadler, The Banking Situation in New York State, at 11 (1956). [89] This will be developed infra. [90] H.Rep.No. 1191, 81st Cong., 1st Sess. 13 (1949). [91] Cf. United States v. Philadelphia Na.t’1 Bank, supra, 374 U.S. at 360 n. 37, 83 S.Ct. 1715. [92] According to the evidence, the metropolitan area is comprised of the five counties of the City, plus Westchester, Rockland, Nassau, and Suffolk Counties in New York State; Bergen, Essex, and Hudson Counties in New Jersey; and Fairfield County in Connecticut. “The general concept adopted in defining a standard metropolitan area (is] that of an integrated economic area with a large volume of daily travel and communication between a central city of 50,000 inhabitants or more and the outlying parts of the area. * * * ” II United States Bureau of the Census, U.S.Census of Business: 1954, p. 3. [93] Cf. DX 41, p. 3, excerpting from 1960 FDIC Ann.Rep. 59-61. [94] Nassau has a population of 1,300,171; and a work force of 467,233, of whom 182,669 or 39% are employed in the city. Westchester has a population of 808,891, a work force of 323,026, of whom 84,669 or 26% are employed in the city. Dep’t of Commerce, Bureau of Census, 1960 Census Tracts 150, 152. [95] Tbe parties have stipulated (PX 14) that 75% of Manufacturers and Hanover’s borrowers and depositors in five categories of banking services are located in the city and tbe other 25% in Nassau and Westchester Counties. The five categories are: IPC regular checking accounts under $100,000; commercial and industrial loans under $100,000; all special checking accounts; all savings accounts, including Christmas Club accounts ; and single payment loans to individuals $25,000 and under in size. [96] See New York State Banking Dep’t, Branch Banking, Bank Mergers and the Public Interest, at 38-39 (1964). We have no evidence showing how much of the business of Westchester and Nassau banks without branch offices in the city originates from customers located in the city, but we note that the number of reverse commuters and city businesses with suburban branches is steadily increasing with the expanding economies of those counties. [97] As we shall see, the evidence shows that most of the business done by local banks outside the metropolitan area, and by outside banks within the area, is with large borrowers and large depositors. Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 358-360, nn. 35, 36, 37, 83 S.Ct. 1715. [98] See Appendix A. [99] The complaint alleges that the merger will have proscribed anticompetitive effects “in the New York City area” (para. 19). It also alleges that: “Customers of Manufacturers and Hanover have regularly utilized interstate communications, -including the mails, telephone and telegraph, to effect deposits, apply for and obtain other services made available by ■these banks. The banks have regularly (utilized interstate communications, including the mails, telephone and telegraph, to conduct banking business with customers, and with other banks located in states other than the state in which they are located. The banks have received a substantial amount of deposits from states other than the state in which they are located, and have made a substantial amount of loans to customers in such other states.” (Para. 14.) The complaint was not amended, but the case was tried on the issue of whether the merger offends the antitrust laws “in New York City, the metropolitan area, and the United -States.” (Plaintiff’s Statement of (Contentions.) Defendant claims neither prejudice nor lack of fair notice of the full breadth of the government’s claim, and, indeed, in its answer admits that it competed with banks and others not only “in the New York City area” but also throughout the nation with respect to certain commercial banking functions (Answer, para. 9). The only purpose of a complaint under the Federal Rules is to give fair notice of a claim. Fed.R.Civ.P. 8; Conley v. Gibson, 355 U.S. 41, 47, 78 S.Ct. 99, 2 L.Ed.2d 80 (1957). The government’s broad claim is fairly embraced within the framework of the parties’ practical construction of the pleadings, and we must, therefore, deal with it on the merits. [100] The complaint does not charge violation of the antitrust laws in foreign commerce, and, in any event, the evidence is wholly insufficient to enable us separately to define the international market or to ascertain its structure in order to assess the competitive effect of the merger on foreign commerce. [101] All Sine of Loans (thousands of dollars) Banks in Loans 1-10 10-100 100-200 200 and over New York City....... 4.75% 5.65% 5.36% 5.06% 4.64% 7 Other Northern and Eastern Cities...... 5.05 5.86 5.53 5.18 4.93 11 Southern and Western Cities.......... 5.26 5.97 5.62 5.28 5.04 Source: DX 40, p. 7. [102] Some of the government’s proposed findings are pegged to “principal oflice,” others to customers “located in New York City,” and still others to a “Treasurer’s office or other executive office in New York.” There is no evidence beyond such bare description as to the activities of any such offices. We find that the only workable approach is one classification for all companies with multiple geographic locations. [103] Of. Madden, The Money Side of “The Street,” p. 30. [104] We note that the Second District covers the whole State of New York and the northern half of New Jersey and, exclusive of many other big cities, four of the nation’s largest metropolitan areas other than New York: Buffalo; Albany-Schenectady-Troy; Newark; and Paterson-Clifton-Passaie. The Federal Reserve System: Purposes and Functions, p. 67; The World Almanac and Book of Facts, pp. 82, 396. [105] Geographic Distribution of Business Loans [106] The population of the New York metropolitan area increased by 13.2% in the decade 1950-60, which equalled the growth for the Northeast, but lagged that of the West with 38.9%, the North Central with 16.1%, the South with 16.5%, and the country as a whole with 18.5%. The World Almanac and Book of Facts, p. 81. [107] Hearings on Regulation of Bank Mergers Before the Senate Committee on Banking and Currency, 86th Cong., 1st Sess. 181 (1959). [108] Fiftieth Annual Report of the Board of Governors of the Federal Reserve System Covering Operations for the Year Ended December 31, 1963, p. 234 (DX 60). [109] We notice that “this year more than 300 banks will get started, almost three times more than in 1961.” Time Magazine, Aug. 28, 1964, p. 74. [110] DX 49, pp. 403-405. Of. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 364 n. 40 (3) & at 368 n. 45, S3 S.Ct. 1715. [111] “[A]n important financial center like New York serves a function of providing wholesale banking in the form of large business credits for national and international enterprises. Such loans, which become bank assets, are attracted only by the adequacy of the institutions from a size standpoint to furnish the credit either by themselves or in association with other large financial institutions.” New York State Banking Department, Recent Bank Mergers in New York City, p. 12, (1955). [112] DX 49, p. 404; DX 40, pp. 3, 10; DX 48, p. 331. [113] The flight of business and middle income population from the city continues apace. N. Y. Herald Tribune, Jan. 25, 1965, p. 1, col. 2. [114] Federal Reserve Bank of New York, Monthly Review, Yol. 42, June 1960, p. 98; Recent Bank Mergers in New York City, supra, pp. 31-32: Relative Growth in Total Deposits of Member Banks New York City and Other Large Cities 1946-54 [115] Report of the Special Deposit Study Committee of the New York Clearing House, Sept. 25, 1959, pp. 51-65. [116] Madden, The Money Side of “The Street,” p. 28. The proportion of loan volume of banks with deposits of $1 million or more to businesses with assets of $5 million and over declined substantially between 1946 and 1955. DX 48, p. 331. [117] N. Y. Times, Dec. 9, 1963, p. 57, col. 3. Competition with New York banks is apparent from Morgan Guaranty Trust Company’s boast that “96 of America’s 100 largest corporations” also bank with it. N. Y. Times, May 12, 1964, p. 15. [118] DX 59. [119] DX 61. [120] DX 61. [121] Yet banks throughout the nation have not kept pace with the growth of their large corporate borrowers. Since 1949 the assets of the 100 largest corporations have increased 145.2% while the assets of the 100 largest banks have increased only 72.4%. DX 1-Memorandum, p. 17; Recent Bank Mergers in New York City, supra, pp. 30-31; Fortune Magazine, July 1960, p. 131; Moody’s Industrial Manual, 1960; American Banker, 100 Largest Banks in the United States, p. 53 1960; Moody’s Bank and Finance Manual, 1950. [122] Compare United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 359 n. 36, 83 S.Ct. 1739: “The evidence discloses that most of the business done outside the [Philadelphia] area is with large borrowers and large depositors;” at 360 n. 37, 83 S.Ct. 1740: “[Competition from outside the area would only be important to the larger borrowers and depositors ;” and at 368 n. 45, 83 S.Ct. at 1744: “ ‘Q. In what area does competition exist? * * * •“ ‘A. I think the stiffest, sternest competition of all is in the field to obtain demand deposits and loans. * * * “ ‘Q. What form does the competition take? “ ‘A. It takes many forms. If we are dealing with the deposits of large corporations, wealthy individuals, I would say that most, if not all, of the major hanks of the country are competing for such deposits. The same would hold true as regards loans to those corporations or toealthy individuals. “ ‘If we go into the field of smaller loans, smaller deposits, the competition is more regional — wide but nevertheless regional — and there the large banks as well as the small banks are after that business with everything they have.’ ” (Emphasis added.) Compare also United States v. First Nat’l Bank & Trust Co. of Lexington, supra, 376 U.S. at 668, 84 S.Ct. 1035: “Apart from large national companies, businesses in the area are resti-icted to the Fayette County banks for their working capital loans * * -.” [123] PX 2, p. 9; DX 40, p. 7. [124] PX 20, p. 12. We notice that recent increases in the prime rate, in the wake of the increase in the British bank rate and the Federal Reserve Board’s increase in the discount rate, were neither initiated nor followed by New York banks. N.Y. Times, Nov. 29, 1964, § 3, p. F-l, col. 5. [125] “Although New York City holds by a wide margin its position as the banking capital of the United States, it has suffered relative loss of ground in that deposit growth here, in recent years, has lagged behind that in other parts of the country.” Report of the Special Deposit Study Committee of the New York Clearing House, Sept. 25, 1959, p. 1. This conclusion was recently confirmed by the Federal Reserve Bank of New York in the following terms: “A number of New York City banks felt that, despite their unique ability to attract nationwide customers, the long-term prospects facing them were unpromising. The financial center’s share in the nation-wide bank deposits has been declining. From the outset of World War II through the end of 1959, deposits of all commercial banks tripled, increasing from approximately $71 billion to $216 billion. By contrast, deposits of the large downtown banks of New York City expanded only 67 per cent, from $18 billion to $30 billion, and most of this increase occurred during the war. As a corollary of the disparate rates of expansion, the share of the New York central reserve city banks in the deposits of all commercial banks declined from 25.4 per cent to 13.9 per cent between 1941 and 1959. The virtual absence of any growth of deposits at the City banks in the postwar period has raised doubts concerning the long-run ability of the New York City banks to keep pace with the rising credit needs of their traditional customers, the nation’s *909 largest corporations.” Federal Reserve Bank of New York, Monthly Review, Vol. 42, June 1960, p. 98. [126] The assets of the nation’s 100 largest banks have increased 72.4% since 1949, while in the case of Hanover the increase for the same period was only 14.5% and for Manufacturers it was 27.7%. DX 1-Memorandum, p. 17. [127] Bank Stock Quarterly, June 1961, p. 11 et seq. (PX 20); Bogen (Tr. 280). [128] The Federal Reserve System: Purposes and Functions, pp. 88-89 (emphasis added). “Geographically separated markets maintain contact with one another in a variety of ways: through the correspondent relations of local banks with banks in other markets; through local contacts with large savings and financial institutions, whose operations may be either regional or national; through arrangements between local dealers in investment securities and either the underwriting houses or stock exchange members of the financial centers; and through the facilities of the Federal Reserve system.” Ibid. [129] The Federal Reserve System: Purposes and Functions, pp. 88-89. [130] Recent Bank Mergers in New York City, supra, p. 12. [131] See DX 8, p. 7; DX 19, p. 9; The Federal Reserve System: Purposes and Functions, p. 24. [132] All of the facts presented to the Board (DX 1 through DX 6, inclusive, and DX 10) and to the Superintendent (DX 13 through DX 33, inclusive) are, by stipulation, part of the record in this case (DX 39). Substantial portions of the contents of those exhibits [those portions which have not been underscored] were stipulated by the parties to be true and correct (DX 39, paras. 39 and 40). The entire remaining portions of said exhibits [the underscored portions] were, by stipulation, received in evidence as part of defendant’s case, with the same force and effect as if the contents had been testified to on direct examination by expert witnesses called by defendant, subject to the right of plaintiff to cross-examine expert witnesses produced by defendant (DX 39, para. 41). Defendant produced Dr. Gabriel Hauge, then vice chairman of the board of directors of Manufacturers Hanover and now its president, both an economist and a banker, with previous experience with the New York State Banking Department and the Federal Reserve Board, for such cross-examination (Tr. 445). The government did not cross-examine Dr. Hauge with respect to the underscored portions of DX 1 through DX 6, inclusive, DX 10, and DX 13 through DX 33, inclusive, except with respect to pages 2, 9 and 16 of DX 1, and page 9 of the Memorandum to DX 1 (Tr. 452-518). [133] DX 1-Memorandum, p. 20, Table 5; DX 13-Memorandum, p. 21, Table 5, Items: 1(e) Demand Deposits over 8100,000; 2 (e) Commercial and Industrial Loans to Larger Business Borrowers (over 8100,000); 2(f) Loans to Brokers and Dealers to Purchase or Carry Securities; 3(a) International Banking Credit Accommodations and Other Services; 4(c) Trusteeships for Pension and Profit-Sharing Plans; 5(a) Trusteeships for Securities Issues; 5(d) Dividend Disbursing Agencies; and 5(e) Bond and Coupon Paying Agencies. [134] We have assumed here, under our ad hoc application of the doctrine of primary jurisdiction, that the Federal Reserve Board acted under Clayton § 11, where it is authorized to decide legal antitrust questions, and its findings of fact and conclusions may well be final if supported by substantial evidence. Its errors of law are, nonetheless, subject to judicial correction. [135] The same vice permeates four affidavits offered by the government (PX 3, 4, 5, and 6) and four counter-affidavits offered by defendant (DX 42, 43, 44, and 45) giving competitors’ opinions, pro and con, on the present and future competitive effect of the merger. This is a law suit, not an opinion poll. We neither need nor welcome a poll of competitors as to which side ought to win. PX 3, 4, 5, and 6 are also barren of facts and saturated with hearsay, assumptions, speculation, prophecies, arguments, theories, and conclusions. The four affiants echo the government’s brief like a quartet of parrots and their opinions are manifestly contaminated with bias and self-interest. The antitrust laws are neither a shield nor a sword for 'competitors motivated by fear of competition. Brown Shoe Co. v. United States, supra, 370 U.S. at 320, 344, 82 S.Ct. 1502. PX 3, 4, 5, and 6 are thus unreliable and do not rise to the dignity of evidence under the most complacent standards. We, therefore, grant defendant’s motion to strike them in their entirety. DX 42, 43, 44, 45, 46, and 47 do contain concrete facts and proper opinions on banking matters. We, therefore, accept them to that extent, but strike all opinions contained in them respecting the competitive effect of the merger. We also grant the government’s motion to strike all opinions and conclusions as to the competitive effect of the merger contained in DX 1, DX 1-Memorandum, and several other defense exhibits. [136] Of. Whitney Nat’l Bank v. Bank of New Orleans, supra, 379 U.S. at 412-413, 85 S.Ct. 551, 13 L.Ed.2d 386. [137] See PX 15 and numerous proposed findings based on the exhibit. [138] The Senate Committee Report on Clayton § 7 recognizes that “[wjhat constitutes a section -will vary -with the nature of the product. * * * A section ■which would be economically significant for a heavy, durable product * * * might well be meaningless for a light product * * S.Rep. No. 1775, 81st Cong., 2d Sess., 5-6 (1950), U.S.Code Cong.Service 1950, p. 4297. [139] The Federal Reserve leased wire system provides easy nationwide accessibility to the country’s banks. All twelve Federal Reserve Banks, their 24 branches and Treasury offices in Washington and Chicago, are joined by this network. More than 200 commercial banks located in 60 large cities are also linked together by a bank wire. This system is used by commercial banks to conduct nationwide banking transactions for themselves or their customers in a matter of minutes. Madden, The Money Side of “The Street,” pp. 25-26. [140] Id. at 30. [141] Id. at 25-26. [142] THE CONSTITUENT BANKS’ BUSINESS WITH CUSTOMERS LOCATED ENTIRELY OUTSIDE THE METROPOLITAN AREA, IN 35 DIFFERENT STATES (December 31, 1960). [143] The nation has “many hanks [which] place loans and solicit deposits outside their home area.” United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 325, 83 S.Ct. 1721. [144] There is no evidence showing the amount of business done by outside competitors within the metropolitan area. We should not assume, however, that the banking agencies could not supply the missing data or at least make a rational estimate. We infer that the amount is substantial from the size and obvious credit rating of the borrowers and the fact that, irrespective of geographic location, 40% of defendant’s commercial and industrial borrowers with loans of $100,-000 or more also had credit relationships with outside banks, and over 85% of its depositors with accounts that large had deposit relationships with outside banks. [145] The government did not produce a shred of evidence to support its claim that the factor of inconvenience restricts large customers to their own locality in their choice of a bank. Instead, it seized upon an isolated sentence from Philadelphia, at 358, 83 S.Ct. 1738, as proof here that “[individuals and corporations typically confer the bulk of their patronage on banks in their local community; they find it impractical to conduct their banking business at a distance.” The statement, of • course, is not evidence here, and surely we cannot rely on “canned” facts. In any event, its inapplicability to large customers was plainly recognized by the Supreme Court in Philadelphia, at 358-360 nn. 35, 36 & 37, 368 n. 45, 83 S.Ct. 1715. Moreover, the quoted statement was imported into the Philadelphia opinion from Transamerica Corp. v. Board of Governors, supra, 206 F.2d at 169, where the Federal Reserve Board sought to define a geographic market embracing five states but was hoisted on its own petard because of its own ill-advised administrative finding, based on an assumption, “that the local community in which a commercial bank is located is its area of competition.” In any event, the evidence here clearly compels the conclusion that wholesale customers find it practical and convenient to spread their deposits and loans around the country in response to their changing needs and the supply of, and demand for, money. See PX 2; The Federal Reserve System: Purposes and Functions, p. 88. [146] Cf. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 371, 83 S.Ct. 1745: “The only businesses located in the Philadelphia area which find * * * [the Philadelphia banks’ lending] limits inadequate are large enough readily to obtain bank credit in other cities.” [147] Commercial and industrial loans; regular checking accounts of individuals, partnerships, and corporations; single payment loans; loans to brokers and dealers; loans to finance companies; bankers’ acceptances; and personal and corporate trust services. [148] United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 363, 83 S.Ct. 1715. [149] Cf. Whitney Nat’l Bank v. Bank of New Orleans, supra, 379 U.S. at 411, 85 S.Ct. 551, 13 L.Ed.2d 386; United States v. Western Pac. R. Co., supra, 352 U.S. at 63-65, 77 S.Ct. 161; Far East Conference v. United States, supra, 342 U.S. at 574, 72 S.Ct. 492. “The dual system of enforcement provided for by the Clayton Act must have contemplated standards of proof capable of administration by the courts as well as by the Federal Trade Commission and other designated agencies. See 38 Stat. 734, 736, as amended, 15 U.S.C. §§ 21, 25 [15 U.S. C.A. §§ 21, 25]. Our interpretation of the Act, therefore, should recognize that an appraisal of economic data which might be practicable if only the latter were faced with the task may be quite otherwise for judges unequipped for it either by experience or by the availability of skilled assistance.” Standard Oil Co. of California v. United States, supra, 337 U.S. at 310 n. 13, 69 S.Ct. 1060. [150] The opinions of the experts and the evidence on which they relied are set forth in DX 1 and several other defense exhibits. The opinions of the banking agencies are in their reports, and the evidence on which their opinions were based is detailed in DX 1. We adopt all the opinions respecting objective banking facts as to the kind, type, and size of account subject to nationwide competition as part of our findings of fact. The parties may, if they wish, submit supplementary findings detailing such opinions and the evidence on which they were based, as well as other evidence on the subject which was presented to us but not to the agencies. We will assume that such additional data was within the knowledge of the experts and agencies. [151] The New York Superintendent of Banks, the Federal Reserve Board, the F.D.I.O., the Comptroller, the Attorney General, and the court. [152] The true local market (the denominator for purposes of calculating market share) is, thus, smaller than the total deposits and loans of the metropolitan area. The local market share of each local bank would be greater or less than its share of the area’s total deposits and loans, depending on whether its pattern of business relative to other local banks is predominantly national or predominantly local. We think it obvious that Hanover, for example, with 82.68% of its deposits and 95.66% of its loans originating in the national market, was not the eighth largest bank in the local market. Whether Manufacturers ranked fifth, or higher or lower, is impossible to ascertain because it had a much larger share of local business than did Hanover, and we think it quite probable that it may have outranked Morgan Guaranty in the local market. See United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 364 n. 40, 83 S.Ct. 1715. [153] Apparently there was no such evidence in the Philadelphia case, but the Supreme Court, nonetheless, shaded the share of the resulting bank by about 15%. There is nothing in the opinion or the record, however, to enable us to fathom how the court arrived at that percentage. In any event, having shaded the share, the court there had no further reason to deal with the 15% for it was no longer material to its determination of the impact of the merger in the Philadelphia area. We cannot adopt that course here, except to estimate Hanover’s relative position in the local market, because: (1) the applicable percentage is unknown and unknowable for want of evidence; (2) the government charges that this merger is anticompetitive not only in the New York area but also in the entire United States; (3) the amount of business originated in the national market is not insignificant for there are at least 21 banks in the area, other than the constituent banks, which are among the nation’s 200 largest banks and, therefore, originate a considerable, but varying portion of their deposits and loans in the national market (Appendix O), whereas in Philadelphia, apparently no banks in the area, other than the appellees, originated any significant business outside the local market, 374 U.S. at 364 n. 40, 83 S.Ct. 1715; and (4) we should not assume that the missing information is unobtainable, and we do not accept ■ the government’s excuse that the problem of separating one geographic market from the other poses an insuperable administrative burden of polling every customer and every bank. The excuse exaggerates the problem. It also ignores the role of the banking agencies. The agencies gave the Department of Justice their opinions as to the categories of accounts subject to national competition. Their reports make clear that wholesale customers and banks competing in the national market are relatively few in number. Surely in this age of computers, detailed records, periodic reports, examinations, and exjjertise, we ought not to assume that the banking agencies are unfamiliar with the pattern of business of major banks and unable to ascertain, or at least make a rational estimate of, the amount of national business and the share of each national competitor. Calculation of local market shares would then have been a simple matter of subtracting the amount of each local competitor’s national business from its total deposits and loans. The dual system of enforcement under the Clayton Act plainly assigns the task of ascertaining the banking facts to the Federal Reserve Board, and the cooperative scheme of the Bank Merger Act assumes that the Department of Justice will get its banking facts straight, in the interest of uniform standards, before it gives its advisory opinion in order to obviate, rather than spawn, divestiture litigation. S.Rep. No. 196, 86th Cong., 1st Sess. 22 (1959). Yet the Department of Justice gave its opinion opposing this merger, admittedly on deficient information, and brought this suit knowing full *925 well that there were two markets and charging violations in both without, so far as the record reveals, asking the agencies a single question, much less their help in separating one market from the other. Its report acknowledges that it was difficult to determine what portion of each bank’s business was allocable to the local and national markets, but fails to perceive that the same information gap existed as to the other large banks. Note, Federal Regulation of Bank Mergers : The Opposing Views of the Federal Banking Agencies and the Department of Justice, 75 Harv.L.Rev. 756 (1962); Waxberg and Robinson, “Chaos in Federal Regulation of Bank Mergers: A Need for Legislative Revision,” N. Y. L. J., Jan. 12, 1965, p. 1, col. 4. [154] “‘While sufficient data to support a conclusion is required, sufficient data to give the enforcement agencies, the courts and business certainty as to competitive consequences would nullify the words “Where the effect may be” in the Clayton Act and convert them into “Where the effect is.” ’ ” Brown Shoe Co. v. United States, supra, 370 U.S. at 341 n. 68, 82 S.Ct. 1532, quoting from Att’y Gen. Nat’l Comm. Antitrust Rep. 126 (1955). [155] DX 49, p. 403; DX 48, pp. 331-332. [156] RELATIVE SIZE OF BUSINESS BORROWERS AT MEMBER BANKS, BY SIZE OF BANK [157] 1960 F.D.I.C. Ann.Rep. 6; and other evidence discussed supra. [158] Madden, The Money Side of “The Street,” p. 25. [159] DX 59. [160] See note 155 supra. [161] Handler, Recent Antitrust Developments, 63 Mich.L.Rev. 59, 69 (1964) ; Merger Developments in the Supreme Court, 26 ABA Antitrust Section 233, 248 (1964) ; Handler & Robinson, The Supreme Court vs. Corporate Mergers, Fortune Magazine, Jan. 1965. [162] W. S. Gilbert, “H.M.S. Pinafore,” Act I, “When I Was a Lad.” [163] United States v. Aluminum Co. of America, supra, 377 U.S. at 282, 84 S.Ct. 1290 (Stewart, J., dissenting); United States v. El Paso Natural Gas Co., supra, 377 U.S. at 656-657, 84 S.Ct. 1044, and opinion of Harlan, J.; White Motor Co. v. United States, 372 U.S. 253, 264, 83 S.Ct. 696, 9 L.Ed.2d 738 (1963); United States v. Oregon State Medical Soc’y, 343 U.S. 326, 332, 72 S.Ct. 690, 96 L.Ed. 978 (1952); Fed.R.Civ. P. 52(a). [164] See Harfield, Legal Restraints on Expanding Banking Facilities, Competition and the Public Interest, 14 Bus.Law. 1016, 1027-1030 (1959). [165] Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 329, 334, 81 S.Ct. 623; United States v. Swift & Co., 286 U.S. 106, 116, 52 S.Ct. 460, 76 L.Ed. 999 (1932); United States v. International Harvester Co., 274 U.S. 693, 708, 47 S. Ct. 748, 71 L.Ed. 1302 (1927); United States v. United States Steel Corp., 251 U.S. 417, 451, 40 S.Ct. 293, 64 L.Ed. 343 (1920). [166] United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 370, 83 S.Ct. 1715. [167] Chase Manhattan Bank, with 20.25% of the area’s assets, is smaller than First Pennsylvania Bank, with 22.9%, and Philadelphia National Bank, with 21%. [168] First National City Trust Co. had assets of $164,026,000 (DX 64). The seven banks in Lexington combined held $163,378,000 in assets. See Record on Appeal, Yol. 2, p. 728, United States v. First Nat’l Bank & Trust Co. of Lexington, supra. [169] Chase Manhattan Bank had 20.25% of the area’s assets (DX 64), but First National Bank of Lexington held 39.83%. See United States v. First Nat’l Bank & Trust Co. of Lexington, supra, 376 U.S. at 668, 84 S.Ct. 1033. [170] See United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 363 n. 38, 83 S. Ct. 1715; Hale and Hale, Market Power; Size and Shape Under the Sherman Act, 131-38 (1958) ; Bok, Mergers and the Clayton Act, 74 Harv.L.Rev. 226, 308-21 (1961); Kronstein, Miller & Schwartz, Modern American Anti-trust Law, 43-47 (1958); Att’y Gen. Nat’l Comm. Antitrust Rep. 325-26 (1955). [171] See Brown Shoe Co. v. United States, supra, 370 U.S. at 319 n. 34, 82 S.Ct. 1502. The government would have us invert the failing company defense enunciated in International Shoe Co. v. FTC, supra, into the proposition that thriving competitors may not merge. Such a rule would invalidate most horizontal mergers as per se violations of Clayton § 7. There is nothing in the text or in the legislative history of Clayton § 7 to support any such proposition. The government, thus, does not meet its burden of proof and shift the onus of justifying the transaction to a defendant merely by showing that the parties to a horizontal merger were previously successful. The critical question under Sherman § 1 and Clayton § 7 is not how big or prosperous the parties to a merger were, but whether their union unreasonably restrains trade, foreshadows a substantial lessening of competition, or tends to create a monopoly. Those proscribed effects may, or may not, follow from the merger of big or prosperous competitors, but, as we shall see, the answer depends on the setting. [173] The government claimed that the new firm would control 37% of the assets, 36% of the deposits, and 34% of the loans, but the Supreme Court “shaded” the percentages to 30% by eliminating business which originated in a different geographic market. United States v. Philadelphia Nat’l Bank, supra, at 364 n. 40, 83 S.Ct. 1715; 201 F.Supp. at 366. [176] MARKET STRUCTURE IN CONTINENTAL CAN Before Merger After Merger Ranlc Name Share Concentration Ranh Name Share Concentration 1 American Can 26.8% 26.8% 1 American Can 26.8% 26.8% 2 Continental 21.9 48.7% 2 Continental 25.0 51.8% 3 Owens-Ill. 11.2 59.9% 3 Owens-Ill. 11.2 63.0% 4 Anchor-Hock. 3.8 63.7% 4 Anchor-Hock. 3.8 66.8% 5 National Can 3.3 67.0% 5 National Can 3.3 70.1% 6 Hazel-Atlas 3.1 70.1% 6 ? ? ? Total 6 Firms 70.1% Total 5 Firms 70.1% All Others 29.9% All Others 29.9% Source: United States v. Continental Can Co., supra, 378 U.S. at 461, 84 S.Ci 1738. [177] See Appendix D. [178] See Appendix B. [179] Congress, in amending Clayton § 7, neither adopted nor rejected such criteria. Brown Shoe Co. v. United States, supra, 370 U.S. at 320, 82 S.Ct. 1502. [180] Comment, “Substantially to Lessen Competition * * * ”: Current Problems of Horizontal Mergers, 68 Yale L.J. 1627, 1638 (1959). [181] See H.R. Rep. No. 1191, 81st Cong. 1st Sess. p. 6 (1949) ; cf. Brown Shoe Co. v. United States, supra, at 319, 82 S.Ct, 1502. [182] Earlier cases taugbt that the share foreclosed must be found to constitute a substantial share of the relevant market. Standard Oil Co. of California v. United States, supra, 337 U.S. at 314, 69 S.Ct. 1051. “That is to say, the opportunities for other traders to enter into or remain in that market must be significantly limited.” Tampa Elec. Co. v. Nashville Coal Co., supra, 365 U.S. at 328, 81 S.Ct. 628. [183] H.R. Rep. No. 1191, 81st Cong., 1st Sess. p. 8 (1949). [184] See Handler & Robinson, A Decade of Administration of the Celler-Kefauver Antimerger Act, 61 Colum.L.Rev. 629, 670 (1961). [185] Handler v. Robinson, The Supreme Court vs. Corporate Mergers, Fortune Magazine, Jan. 1965. [186] The theory requires proof of market structure. Hale and Hale, Market Power: Size and Shape Under the Sherman Act ch. 3, p. 89 (1958). There is no requirement, however, that market structure be proved with mathematical certainty or perfect accuracy. United States v. Philadelphia Nat’l Bank, supra. [187] 1960 P.D.I.O. Ann.Rep. 102, table 40. [188] Att’y Gen. Nat’l Comm. Antitrust Rep. 325 (1955). [189] If we guess that the pattern of business of each of the five largest integrated banks is exactly like Manufacturers’, that the pattern of business of wholesale banks is exactly like Hanover’s, and that smaller retail banks, originate all of their business in the local market, we could shade the shares of all the other banks in the local market (Appendix B) by applying to integrated banks the percentage of local business originated by Manufacturers, and to wholesale banks the percentage for Hanover (Appendix A). Market shares at December 31, 1960 would then be as follows: Shaded Deposits And Loans, As of December 81, I960.* Shaded Shaded Shaded Ranh Name of Banh Deposits Loans 1 Chase Manhattan Bank ......................... . 21.68% 21.34% 2 First Nat’l City Bank ........................... . 19.05 18.13 3 Chemical Bk. N. Y. Tr. ......................... . 11.01 10.58 4 Manufacturers Trust............................ . 10.01 7.45 5 Bankers Trust ................................. 8.46 7.34 Pre-Merger Concentration....................... . 70.21% 64.84% 14 Hanover Bank ................................. 1.83 .83 66 Remaining Banks ........................... . 27.96 34.33 Total All Banks................................ . 100.00% 100.00% * Formula: Sum of Deposits (Loans) of each bank allocable to Local Market: =Total Local Market. Share. Each bank’s Deposits (Loans) so alloeated-i-Total Local Market=Local Market Source: Appendix A; Appendix B; PX 17. Many controlling decisions of the Supreme Court were decided after the trial. The court, therefore, conducted a number of post-trial conferences with counsel, and the recora was opened many times. During such conferences, without disclosing our purpose, we sought a stipulation as to the pattern of business of every bank in the metropolitan area. Laboring in the dark, the parties reached a tentative understanding on the subject in June 1964, but when a stipulation was submitted on August 19, 1964, the government inserted an escape clause, and we rejected it. The above table is the product of our efforts in the interim, but without a firm stipulation, it has no solid support in the evidence. [190] DX 1, pp. 27 and 28. [191] DX 1, pp. 27, 28, and 37. DX 1-Memorandum, pp. 9, 10, and 13. Studies of the F.D.I.C. show that over 95% of all deposit accounts are under $10,000. That figure was therefore selected by Congress as the appropriate limit for Federal Deposit Insurance, but Congress is currently considering increasing the limit to $25,000 which would cover over 99% of all deposit accounts. See Hearings Before House Committee on Banking and Currency on H.R. 5130, 88th Cong., 1st Sess. 5, 9, and 74 (April 24-26, 1963); N.Y.Times, April 13, 1963, p. 59, col. 5; N.Y.Times, January 10, 1964, p. 59, col. 3. [192] DX 1, p. 28; DX 10, p. 14. [193] DX 1-Memorandum, p. 29. [194] DX 1, p. 18. [195] DX 1, p. 18; DX 10, p. 14; Appendix A. [196] Our equation is supported to some extent by the spurious statistics (Appendix B). Absent direct evidence of market share, imperfections are inherent in the problem. The only clues in the evidence are the actual local business of the constituent banks and the distribution of branches among competitors. We must work within the limitations of the evidence, draw reasonable inferences from known facts, and fashion tools for analysis. We are not concerned with mathematical certainties but with reasonable probabilities. Perfect accuracy must yield to workable compromises. United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 361, 83 S.Ct. 1715. [197] There is no evidence as to the distribution of branches in Nassau and Westchester beyond the fact that, together 10 Westchester banks have 104 offices and 16 Nassau banks have 96 (DX 64; DX 1, Supp. Map of Metropolitan Area). The parties have stipulated in effect that 75% of retail customers are located in New York City and 25% in the metropolitan area outside the city (PX 14). Since those customers located outside the city necessarily do part of their banking business in the city, branch concentration in the city has a direct impact throughout the metropolitan area. In any event, we think the picture in the city affords a fair sample of the general picture throughout the area and provides a meaningful gauge for measuring concentration. Cf. Brown Shoe Co. v. United States, supra, at 339-342 n. 69, 82 S.Ct. 1502. [198] In United States v. Philadelphia Nat’l Bank, supra, at 325-326, 83 S.Ct. 1715, the court surveyed the decade before the merger. Bok suggests a period between five to ten years prior to the merger. Bok, Mergers and the Clayton Act, 74 Harv.L.Kev. 226, 314 (1961). We think the relevant period varies with and depends on the unique history of the market under consideration. [199] See United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 365 n. 42, 83 S.Ct. 1715; Bok, Mergers and the Clayton Act, supra, at 313. [200] DX 10; Bogen, Tr. 427-437; Recent Bank Mergers in New York City, supra, p. 26 et seq.; Branch Banking, Bank Mergers and the Public Interest, supra, p. 163 et seq. The latter reference dramatically shows one adverse economic change. It notes that between 1940 and 1960 the population of Nassau and West-chester grew by 220% and 41%, respectively, while that of New York City increased only 4%. Id. at 38. [201] The government aims its heaviest guns at five out of 27 mergers which occurred in New York City during the decade beginning in 1950: (1) Chase National Bank and Bank of Manhattan Co., after the latter had absorbed Bronx County Trust Co.; (2) First National Bank and National City Bank; (3) Chemical Bank & Trust Co. and Corn Exchange Bank & Trust Co.; (4) Bankers Trust Co. and seven other banks; and (5) J. P. Morgan & Co. and Guaranty Trust Co. (PX 28; PX 9). Morgan Guaranty was a union of two banks which had been engaged almost exclusively in wholesale banking in a nationwide market, and the resulting bank still publicly adheres to that policy. 105 Cong.Rec. 7832 (1959); N.Y.Times, May 12, 1964, p. 15. The other four earlier mergers were all investigated by the New York State Banking Department which found that they did not increase concentration significantly in the local market, either because they were unions of wholesale and retail banks, like the Chase-Manhattan merger, or because the resulting bank derived a large portion of its business from the national market, as with the resulting bank in all four mergers. Recent Bank Mergers in New York City, supra. [202] DX 1, p. 28. [203] DX 1, pp. 37-39. [204] PX 28. Again, there are no merger figures in evidence for the whole metropolitan area, but we take notice that Nassau and Westchester have not escaped from the merger trend. Branch Banking, Bank Mergers and the Public Interest, supra, p. 104. [205] PX 24; DX 1, p. 3. [206] PX 28. [207] DX 10, p. 9. See also PX 28; Recent Bank Mergers in New York City, supra. [208] There are no figures in evidence showing the percentage of trust business in the metropolitan area as a whole, but those for the City of New York constitute a fair sample. Cf. Brown Shoe Co. v. United States, supra, 370 U.S. at 339-342, 82 S.Ct. 1502. [209] We have largely, though not entirely, corrected the distortion of Hanover’s market share by allocating its accounts (Appendix A). The correction, plus evidence that the smaller the bank the lesser the distortion of the spurious statistics, permit our estimate. Cf. United *946 States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 364 n. 40, 83 S.Ct. 1715. Compare Hanover’s deposits and loans al-locable to the local market with the total deposits and loans of smaller banks shown in Appendix B and in DX 64. Compare the number of Hanover’s branches with the number held by smaller competitors, supra, at 102. [210] DX 1-Memorandum, p. 12. [211] DX 1, pp. 33, 40, and 41; Appendix A. [212] See p. 940 supra. [213] DX 10, pp. 24-25. [214] DX 10, pp. 15 and 24. [215] The 10 former Hanover branches are encircled by 65 competing branches, exclusive of 14 former Manufacturers branches, and in the immediate vicinity of each pair of former Manufacturers and Hanover branches, there are at least 12 offices of competitors, all offering a broad range of retail banking services to the general public and small businesses. DX 1-Supp. Map. [216] DX 1, p. 26; DX 10, p. 16. [217] DX 10, pp. 14-15 (Flanigan). [218] DX 10, p. 24 (Gray): “The trend and the future of banking in New York City, and in other financial centers, lies in fully integrated banks offering both retail and wholesale services to all members of the banking public. The retail function complements the wholesale function in the modern bank because it provides a broad, fairly constant base for deposit growth.” [219] DX 1, pp. 3-9, 28. [220] DX 10, p. 35 (Gray): “I don’t think, Governor, we have to make this proposed merger. We can both do well on our own. But I think we will do better in the merged institution than we will separately, both of us.” [221] DX 10, pp. 15, 24, and 35. [222] DX 1, p. 27; DX 1-Memorandum, pp. 12-13. [223] DECLINE SINCE THE MERGER IN THE SPURIOUS PERCENTAGES OF ASSETS, DEPOSITS AND LOANS OF THE FIVE LARGEST BANKS LOCATED IN THE METROPOLITAN AREA (DX 66). [224] See United States v. Philadelphia Nat’l Bank, supra, 374 U.S. at 373, 83 S.Ct. 1746 (Harlan, J., dissenting) & at 396, 83 S.Ct. 1759 (Goldberg, J., dissenting). [225] Repugnancies lurking in the Bank Merger Act fed doubt as to the applicability of the Sherman Act. McCulloch v. Maryland, 17 U.S. [4 Wheat.] 316, 4 L. Ed. 579 (1819). See also Toolson v. N. Y. Yankees, 346 U.S. 356, 74 S.Ct. 78, 98 L.Ed. 64 (1953); United States v. SouthEastern Underwriters Ass’n, 322 U.S. 533, 64 S.Ct. 1162, 88 L.Ed. 1440 (1944); Transamerica Corp. v. Board of Governors, supra, at 206 F.2d 166. [226] DX 10. [227] There was also direct competition. The banks had 372 common depositors accounting for approximately $1,000,000 and 67 common borrowers accounting for $361,760. DX 1, p. 33. [228] Timken Roller Bearing Co. v. United States, 341 U.S. 593, 71 S.Ct. 971, 95 L. Ed. 1199 (1951); United States v. Paramount Pictures, Inc., 344 U.S. 131, 68 S. Ct. 915, 92 L.Ed. 1260 (1948); Sugar Institute, Inc. v. United States, 297 U.S. *952 553, 56 S.Ct. 629, 80 L.Ed. 859 (1936); United States v. Trenton Potteries Co., 273 U.S. 392, 47 S.Ct. 377, 71 L.Ed. 700 (1927). [229] American Crystal Sugar Co. v. Cuban-American Sugar Co., supra, 259 F.2d at 526, and cases there cited. [230] “While on the one hand it was desired that the test be more inclusive and stricter than that of the Sherman Act, on the other hand it was not desired that the bill go to the extreme of prohibiting all acquisitions between competing companies.” S.Rep.No. 1775, 81st Cong., 2d Sess. 4 (1950), U.S.Code Cong.Serviee 1950, p. 4296. [231] 96 Cong.Rec. 16456 (1950). See also H.R.Rep.No. 1191, 81st Cong. 1st Sess. p. 7 (1949) ; 96 Cong.Rec. 16435 (1950) ; 95 Cong.Rec. 11487 (1949). [232] “To determine substantiality in a given case, it is necessary to weigh the probable effect of the contract on the relevant area of effective competition, taking into account the relative strength of the parties, the proportionate volume of commerce involved in relation to the total volume of commerce in the relevant market area, and the probable immediate and future effects which pre-emption of that share of the market might have on effective competition therein.” 365 U.S. at 329, 81 S.Ct. 629. [233] This is not to say that forbidden anti-competitive effects may not be perceived *953 from the elimination of former substantial competition between major competitors, or from the elimination of a substantial factor from competition as a whole, when a merger’s impact is assessed in the setting in which those fac: tors are placed. Handler and Robinson, A Decade of Administration of the Celler-Kefauver Antimerger Act, 61 Colum. L.Rev. 629 (1961). [234] In addition to a series of decisions in lower federal courts, defendant cites: Times-Pieayune Pub. Co. v. United States, supra; Appalachian Coals, Inc. v. United States, 288 U.S. 344, 53 S.Ct. 471, 77 L.Ed. 825 (1933); United States v. United States Steel Corp., supra; Chicago Board of Trade v. United States, 246 U.S. 231, 38 S.Ct. 242, 62 L.Ed. 683 (1918); United States v. American Tobacco Co., 221 U.S. 106, 31 S.Ct. 632, 55 L.Ed. 663 (1911); Standard Oil Co. of New Jersey v. United States, supra. [235] Appendix C. [236] DX 10. [237] See Appendices C and E. [238] Compare low bank’s deposits and number of competitors listed in 1950, 1960, and 1963 (DX 61). [239] PERCENTAGE OF ASSETS, DEPOSITS AND LOANS OF THE SIX LARGEST, OUT OP 200 LARGEST, BANKS IN THE UNITED STATES* [240] The combined spurious share of the six largest banks in the national market is not remotely close to the figures in Philadelphia, Lexington, Alcoa, or Continental Can, nor to those considered oligopolistic by the Congress. H.R.Rep. No. 1191, p. 2. The percentages of increase in concentration on the basis of the spurious figures the instants before and after the merger were 5.12% in assets, 5.02% in deposits, and 4.51% in loans (computed from DX 61). [241] The facts detailed earlier demonstrate that the government’s reliance on the total number of mergers in the entire country during the decade beginning in 1950 is misplaced, for the evidence establishes overwhelmingly that only a relatively few of the nation’s large banks compete in the national market. [242] Compare growth of lowest bank from 1950 to 1963 (DX 61). [243] Northern Securities Co. v. United States, 193 U.S. 197, 24 S.Ct. 436, 48 L. Ed. 679 (1904); United States v. Union Pac. R. R. Co., 226 U.S. 61, 33 S.Ct. 53, 57 L.Ed. 124 (1912); United States v. Reading Co., 253 U.S. 26, 40 S.Ct. 425, 64 L.Ed. 760 (1920); United States v. Lehigh Valley R. R. Co., 254 U.S. 255, 41 S.Ct. 104, 65 L.Ed. 253 (1920).