Citations
- 362 F. Supp. 240
Full opinion text
MEMORANDUM OF DECISION
ZAMPANO, District Judge.
In this civil antitrust action, the government is challenging the proposed consolidation of The Connecticut National Bank (CNB), the fourth largest commercial banking association in the state, and The First New Haven National Bank (FNH), which ranks eighth among Connecticut’s commercial banks. The complaint seeks an injunction, alleging that the merger will violate Section 7 of the Clayton Act, 15 U.S.C. § 18. The defendants have denied that their agreement to combine would be harmful to competition and have asked this Court to lift the stay on the Comptroller of the Currency’s (Comptroller) approval of their plans, which was automatically imposed pursuant to the provisions of the Bank Merger Act of 1966 (BMA), 12 U.S.C. § 1828(c)(7)(A), when this action was initiated.
The Comptroller has intervened and participated throughout the proceedings as provided by 12 U.S.C. § 1828(c)(7) (D). The Court’s jurisdiction under Section 15 of the Clayton Act, 15 U.S.C. § 25, is undisputed. After a lengthy trial and post-trial arguments and review of the voluminous briefs and statistical information submitted by counsel, this Court is of the opinion that the defendant banks should be allowed to merge, with the condition that certain divestitures in the so-called “Four Town Area” are accomplished as more fully set forth, infra.
In the pretrial proceedings, the Court appointed a Special Master who was of great assistance to the parties and to the Court in narrowing the issues to be presented. In addition, counsel for the parties demonstrated extraordinary competence, skill and cooperation. Their presentations were of incalculable aid to the Court.
I. FINDINGS OF FACT
A. THE DEFENDANTS
Both defendants are banking associations organized under the laws of the United States and both transact business and may be found within the District of Connecticut. CNB maintains its principal place of business in Bridgeport and FNH in New Haven, Connecticut. They are each engaged in interstate commerce.
(1) Connecticut National Bank
1. CNB was chartered in 1806 as the Bridgeport Bank, with general banking and trust powers and was converted to a national bank charter in 1865. The bank’s only acquisition since 1958 involved the Atlantic National Bank in Stamford. (D-35; P-56). The rate of which CNB has expanded by de novo■ branching has abated in recent years. Between 1961 and 1965, CNB opened 15 offices. From 1965 to 1970, the bank added only five branches. (P-73; Tr. 1544).
2. As of December 31, 1972, CNB had total assets of $463.3 million, total deposits of $412.2 million, and total loans of $253.6 million, ranking seventh in deposit size among banks headquartered in Connecticut and fourth among commercial banks headquartered in the state. (P-122; P-70; D-10). Savings deposits constitute approximately 50% of CNB’s total deposits. (Tr. 698). CNB has a legal lending limit of $2.8 million.
3. As of December 31, 1972, CNB operated 51 offices in 25 towns. Forty-one of these offices are located in Fair-field County, nine in New Haven County and one in Litchfield County. (D-89).
(2) First New Haven National Bank
4. FNH is the result of a consolidation in 1957 of the New Haven Bank, chartered under the laws of the State of Connecticut in 1792, and The First National Bank of New Haven, chartered on June 20, 1863, under the National Currency Act. Since its acquisition of Branford Trust Company in 1964, FNH has not engaged in any merger. (D-35, p. 55).
5. As of December 31, 1972, FNH had total assets of $333.4 million, total deposits of $272.4 million and total loans of $224.7 million, ranking tenth among banks headquartered in Connecticut and eighth among commercial banks in the state. (P-120; P-70; D-10). Savings deposits constitute approximately 50% of FNH’s total deposits. (Tr. 1452). FNH has a legal lending limit of $2.3 million.
6. As of December 31, 1972, FNH operated 22 offices, all in New Haven County. (D-38).
(3) Merger Agreement
7. CNB and FNH have entered into a Consolidation Agreement, dated November 19, 1970, which will result, if carried out, in the consolidation of CNB and FNH under the charter of FNH and with the title, “The First Connecticut National Bank.”
8. CNB and FNH prepared and filed an application dated February 2, 1971, for approval of the proposed consolidation and of the Consolidation Agreement with the United States Comptroller of the Currency who supervises all national banks. (D-35).
9. The consolidation was approved by the Comptroller of the Currency on July 26, 1971. (Int-1).
(4) Divestiture
10. As part of the merger agreement, CNB has contracted to sell its Milford office to the Hartford National Bank (Hartford National), and its Orange office to the organizers of a new bank to be chartered. FNH has agreed to sell its Derby office to the Connecticut Bank and Trust Co. (CBT) upon consummation of the consolidation. (Tr. 838).
11. In addition, the defendants have agreed to divest themselves of three additional offices in Ansonia, Milford and Orange within one year after the pro-' posed merger is approved. (Tr. 687-8).
12. To implement the divestitures, the acquiring banks will obtain by operation of law all of the banking business, including deposits, of the divested branches. The customers will not have to perform any act in order for their deposit business to be transferred to the acquiring banks. (Tr. 2215-6, 2218).
B. THE BANKING BUSINESS
(1) General Background and Regulation of Banking
13. Banking is distinguishable from most commercial enterprises in that both state and federal governments insist on the availability of stable banks and banking for the citizenry. If a bank fails, it is detrimental to the entire community. Banking is therefore regulated to insure there is competition; but measures are taken to assure that competition is not unbridled and yet doesn’t become over-competitive so that bank failures result.
14. Mr. John L. Donovan, Regional Administrator of National Banks of the First National Bank Region described bank regulators as people who “carry water on both shoulders,” in order to assure both competition and solvency in the banking industry. (Tr. 1829-1830). A primary concern in this business is that the competition not become so intense that bank failures result. (Tr. 1831-1832).
15. Congress has required that banks be examined by the Comptroller’s Office to insure the health of the banking industry. 12 U.S.C. § 481. The function of the Regional Administrator of National Banks and his examiners is to examine the banks in his jurisdiction at reasonably frequent intervals to carry out the Congressional mandate. (Tr. 1940-1941).
16. The Regional Administrator is required by the Bank Merger Act to consider the financial, managerial resources, and the future prospects of the existing and proposed banking institutions. (Tr. 1944).
17. FNH has been a member of the National Banking System since its inception in 1863 and is the oldest chartered national bank with Charter No. 1; the Comptroller’s Office has been examining it regularly since 1863. CNB has been a national bank since 1865. (Tr. 1928).
18. Federal statutes (see 12 U.S.C. § 481), require that national banks be examined three times in every two years. Generally, they are each examined between eight and twelve months apart. (Tr. 1796). The Comptroller’s Office spends between 250 and 280 man-days examining a bank the size of FNH. CNB requires between 420 and 450 man-days. (Tr. 1928-1929).
19. A bank examination involves complete and thorough procedures to determine the bank’s assets and liabilities. Between 80 to 85% of the dollar volume of the loan portfolio is reviewed. Examiners express opinions on the general competence of management, the bank’s condition, earnings, capital adequacy, liquidity, internal controls, audit procedures, and future prospects for growth. (Tr. 1929-1931).
20. Copies of the examination report are filed with the bank and the Comptroller’s Office in Washington along with any corrective action that may be required. Copies also are sent to the Federal Reserve Bank for the appropriate Federal Reserve District and to the FDIC in Washington. (Tr. 1929-1932). In bank mergers, the Comptroller requires a detailed application which is typified by D-35, the application of the defendants in this case. (Tr. 1935; D-35).
21. The Regional Director makes recommendations on the feasibility of branch expansions and mergers, but the final decisions rest with the Comptroller. With respect to an application for a de novo branch or consolidation with another bank, multiple factors are carefully scrutinized including the public need, profitability, economic character of the service area, population and business growth, any anticompetitive effects, management strength, capital adequacy, and legal considerations. In addition, there is verification and a critical evaluation of the application data, and competitor banks are consulted.
22. Pursuant to' the BMA, the Department of Justice is notified of the Comptroller’s approval of any merger and, after consideration of the application, may move in the courts to block the consolidation, as was done in the instant case.
(2) Connecticut Banking History and Regulation
23. The history of banking in Connecticut has been somewhat unique. For example, branch banking was not authorized prior to 1933, at which time the legislature also passed .the “Home Office Protection” law that prohibits de novo branch banking into a town where the head office of another bank is located. As a result, the biggest cities, including Hartford, New Haven and Bridgeport, are “closed” to de novo branching by out-of-town banks.
24. All state banks and loan companies are regulated by the State Banking Commission. During the past 10 years, 32 applications for new bank charters were filed and 22 were approved. However, it seems clear that, unless there is a change of policy, the Commission will not approve a new bank charter for the purpose of sale. (Tr. 2153). Moreover, the Commission requires that persons applying for a de novo branch represent in their application that they have no intention to sell the new bank or merge it with another bank. (Tr. 1627; D-7, D-84). The Commission has never received an application for a new bank charter for the purpose of sale upon formation to a holding company. (Tr. 2154).
25. According to Dr. Charles Stokes, Chairman of the Economics Department at University of Bridgeport, a specialist in regional economics and ad-visor to several Connecticut banks as well as the Connecticut Development Commission, by the early 1950’s there had emerged in Connecticut five de facto banking districts which were essentially distinct, non-overlapping areas: (1) Fairfield County and a part of Litchfield County, (2) Southern New Haven County, (3) Hartford County; excluding the towns of Bristol,. Meriden, Middletown and New Britain, but including all other towns east of the Connecticut River, (4) the town of Waterbury, Northern New Haven County and the remainder of Litchfield County, and (5) a midstate region consisting of the towns of Bristol, Meriden, Middletown and New Britain. Banks in these de facto districts branched throughout the district in which their headquarters were located and did not compete outside of their districts. (Tr. 886-7; D-38 through D-49).
26. In 1959, the pattern was torn when the Hartford National was permitted to acquire a bank in Stamford, thereby crossing over a boundary which did not exist by law but did prevail in practice. Shortly CBT followed suit and entered into the New Haven and Fair-field markets. Thus, two banks with significant state-wide influence and markets emerged on the Connecticut banking scene. (Tr. 971).
27. In the period since 1960, CBT added 42 offices to its system and entered 23 new towns, Hartford National founded 36 offices in 15 new towns, while CNB expanded by 27 in eight towns and FNH grew by 12 offices in five towns. In a theoretical time frame of the next ten to 15 years during which other banks might arguably expand by de novo branching, it is highly probable that CBT and Hartford National will be spreading across the state at a rate four or five times faster than any other bank. (Tr.1478-9; 1538-9).
28. In the one year period September 30, 1971 to September 30, 1972, CBT and Hartford National increased their total assets by $183 million and $196 million respectively, increases almost equivalent to the total assets of the tenth largest commercial bank, Second New Haven Bank. The combined increases in assets in one year for the two Hartford banks are almost equal to the total assets of CNB and exceed the total assets of FNH. (Tr. 1685-7; D-74; D-75; D-76; P-120; P-122). An increase in capital accounts is important from a competitive standpoint, since it increases lending limits and provides the means to acquire new offices and institute new services. (Tr. 932).
29. From 1965 to 1970, CNB and FNH had among the lowest rates of increase in total assets, deposits, and equity capital among the nine largest Connecticut commercial banks, and CNB had the lowest rate of increase in number of branches. (Tr. 939-40; D-16). CNB ranks eighth among the ten largest commercial banks in Connecticut in terms of its ratio of loans to capital and ninth in terms of its ratio of capital to assets. (Tr. 1551-3).
30. The additional deposits acquired by the two Hartford banks, CBT and Hartford National, from December 30, 1969 to September 30, 1972 ($745.3 million) exceeded the combined deposits of defendant banks as of September 30, 1972 ($672.8 million). (Tr. 1526; D-69).
31. The two so-called Hartford banks slowly but surely are penetrating more and more local markets throughout the state. (Tr. 670). They advertise extensively, actively solicit clients particularly in New Haven and Fairfield Counties, and have succeeded in securing as customers 17 of the 25 corporations with offices in those two counties. (Tr. 1441; D-60). As a result, it seems clear that as they are growing more and more dominant, other banks are finding themselves correspondingly in a weakened position. (Tr. 974). For example, FNH has experienced virtually no growth in demand deposits and a very modest increase in time deposits. (Tr. 1457-8).
32. In addition, the third banking giant, the Union Trust Company (“Union Trust”), was created by merger in 1970.
33. The plaintiff has permitted the state-wide growth by merger of these three banks. (Tr. 473-4).
34. The unsettling state of the banking structure in Connecticut has caused the Comptroller to insist he will not permit new national bank or branch expansion into the major banking markets of this state. (Tr. 1677; 1812; Int-2).
C. LINE OF COMMERCE
(1) Comparison of Commercial Banks and Other Financial '• Institutions
35. Economists approach analysis of markets by first defining a product or service. This is a complicated matter since it is necessary to include in the market products or services that consumers regard as reasonably close substitutes. Economists then determine whether sellers of the defined product or service are close substitutes geographically. (Tr. 1333,879).
36. Even in the opinion of Dr. Murphy who was called by the plaintiff as an expert on banking markets, increased competition among commercial banks and thrift institutions for savings-type liabilities in the 1960’s suggests that some reconsideration of the commercial bank product line specification may be in order, especially if mutual savings banks are successful in obtaining the demand deposit privilege in several of the states in which they operate and savings and loan associations make use of the capability to make third-party transfers. (Tr. 209-10; 771; 1026-7; 1231; D-58). Dr. Murphy further testified that commercial banks currently compete with savings banks for at least time and savings deposits, personal loans and real estate loans. (Tr. 150, 221).
37. According to the President of the Connecticut Savings Bank in New Haven (who is a former commercial bank officer) savings banks compete directly with commercial banks for all forms of time and savings accounts, installment loans, mortgage loans as well as for branching sites and personnel. (Tr. 1748-9). All five of the commercial bank presidents who testified cited savings banks as their chief competitors. (Tr. 672, 697; 1451-2; 627-8; 1869-70; '1764).
38. The bank regulatory officials who testified were in agreement that savings banks are strong competitors of commercial banks. (Tr. 2163; 1922-3). The Comptroller requires that an applicant for a de novo branch or for a new national bank charter furnish a list of competing savings banks in the applicant’s trade areas along with a map of their locations so that the Comptroller can determine the proximity of substantial competitors to the applicant’s proposed branch or new bank. The national bank examiner investigating the applicant’s proposal must assess, in his written report, the degree of competition from savings banks and other thrift institutions. (Tr. 1922).
39. In addition, the Connecticut State Banking Commission concurs that savings banks are “formidable competitors” of commercial banks. The presence of savings banks is taken into consideration by the State Banking Department when an application for a new branch of a commercial bank is being evaluated. (Tr. 2163).
40. The plaintiff apparently also recognizes that analysis of the competitive factors involved in a merger of two savings banks must include evidence of the amount of time and savings deposits held by commercial banks. (Tr. 527-8). Since 1961, when the Supreme Court found commercial banking to be the appropriate line of commerce in one case, the plaintiff’s expert witness acknowledges that competition between commercial banks and savings and loan associations and mutual savings banks has increased. (Tr. 210-212).
41. In short, competition between savings banks and commercial banks is at a level simply not envisioned ten years ago. A national trend toward more nearly equal powers between commercial banks and thrift institutions, such as savings banks, is. discernible from the recommendations of various studies. The Presidential Commission on Financial Structure (“Hunt Commission”), the Friends Study and the report of the New York Federal Reserve Bank recommended broadened powers for all thrift institutions, including checking accounts and third party transfer arrangements, to enable these other financial institutions to survive because commercial banks are increasingly penetrating the retail banking field. (Tr. 741; 245-247). The Hunt Commission has submitted its recommendátions to President Nixon and a task force in the Treasury Department is preparing legislation to implement it. (Tr. 575; 749-750).
42. Commercial banks in Connecticut have actively opposed the efforts of savings banks to branch into communities where they will become a competitive threat. (Tr. 1760-1763). In one case, the prospect of further competition from savings banks led a commercial bank to file separate suits against the State Banking Commissioner and the Federal Deposit Insurance Corporation to overturn the approval of a new branch in Simsbury for the Burritt Mutual Savings Bank of New Britain. (Tr. 1763-1974; 732-735; D-77; D-78).
43. Statistics gathered by three federal bank regulatory agencies demonstrate a high degree of comparability between Connecticut savings and commercial banks in the areas o,f operating income, level of earnings, loans, investment portfolios, income from securities and service charges. (Tr. 1006-1008; D-19-20).
44. The similarity between savings banks and commercial banks is such that one may merge into the other without difficulty. There are several cases in Connecticut and in other states where mutual savings banks were converted or merged into commercial banks. (Tr. 735-737).
45. In the instant action, the plaintiff has admitted that CNB competes directly with savings banks for 79% of its loans. (Tr. 231-232). It is also commonly known that the presence of savings banks has a direct influence on the branching decisions of CNB and FNH. (Tr. 707-8; 1453; 272-3). Both banks must take into account rates charged by savings banks when setting their own rates. (Tr. 697). FNH recently lost its corporate secretary and vice president to its principal savings bank rival, Connecticut Savings Bank, where he now serves as president. (Tr. 1452). Clearly this competition is keenly experienced by all parties.
46. CNB and FNH compete with other financial institutions for loans and deposits as well. Credit unions and personal finance companies are a partial substitute for personal loans. (Tr. 214-215). CNB and FNH compete directly with savings and loan associations for savings deposits and mortgage loans, with insurance companies for term loans, with credit unions for savings deposits and small consumer loans, and with sales finance and personal loan companies for instalment loans. (D-35, pp. 50-53).
(2) Savings Deposits in Connecticut Commercial Banks
47. Prior to the late 1950’s, commercial banks in Connecticut were not involved in the consumer savings market to any great extent. Since that time, however, commercial banks have accomplished significant penetration into this area of retail banking. (Tr. 725-6; 729; D-61 through D-66). Between 1960 and 1971, savings banks increased their total savings deposits from $2.5 billion to $6.1 billion. Moving at a faster clip, commercial banks increased their savings from $596 million to $2.2 billion during the same time span. The share of the total savings market held by savings banks decreased from 65% to 59% between 1960 and 1971, while the share of commercial banks increased from 15% to 22%. (Tr. 722; D-62; D-63).
48. The number of savings accounts held by Connecticut commercial banks increased by 299,704 between June, 1966 and June, 1970. During the same period, savings accounts held by savings banks increased by 210,249. Whereas commercial banks increased their savings accounts by 31.4%, savings banks experienced an increase of only 11.2%. (Tr. 719-720).
49. Insofar as the personal savings market is concerned, Connecticut commercial banks increased their share from 15.61% in 1961 to 29.09% in 1969, while the share of savings banks decreased from 67.32% to 60.01%. Another type of thrift institution, savings and .loan associations, also lost ground to commercial banks. Whereas savings and loan associations had 17.07% of the personal savings market in 1961, they had 15.90% in 1969. (Tr. 723-724; D-65; D-66).
50. The increasing penetration by Connecticut commercial banks into the personal savings market is typical of the trend in all the 16 states where savings banks are located. (Tr. 726-727; D-52). There is no question that there is direct competition between savings banks and commercial banks for all forms of savings accounts. (Tr. 150; 164; 1747; 1869; 1923).
51. Commercial banks and savings banks pay approximately the same rate of interest on savings certificates of deposit. (Tr. 221-222). The slight difference in the interest rates which may legally be paid on savings accounts by savings banks and commercial banks does not distinguish the two institutions. The difference in rate does not appear to be decisive in the mind of the public. (Tr. 1747-1748; 1661).
(3) Demand, Deposits in Connecticut Savings Banks
52. To offset the growing penetration by commercial banks into the personal savings market in Connecticut, savings banks have fought to acquire the right to furnish personal checking accounts to their depositors. Legislation to accomplish this goal was first introduced in 1967 and has made repeated advances in the legislature. In 1971, this legislation was defeated in the state Senate by a very narrow margin. Savings banks have continued to seek these powers and were able to reach a compromise in the recent legislative session, which was signed into law on May 15, 1973. (Tr. 727-729, 742; 1743; 1870).
53. The compromise bill, entitled “An Act Concerning Personal Checking Accounts in Savings Banks and Building or Savings and Loan Associations”, allows savings banks to offer personal, non-interest-bearing checking accounts, but prohibits “NOW” accounts (negotiable orders of withdrawal from interest-bearing savings accounts). This legislation had the support of both the Savings Banks Association of Connecticut and the Connecticut Bankers Association. The provisions which apparently paved the way for compromise give the State Banking Commission the power to order the savings banks to stop accepting such accounts if that agency determines that a savings bank is seriously jeopardizing the vitality of a neighboring commercial bank.
54. Proposed federal legislation would either preclude interest paid on “NOW” accounts or allow all nationally regulated financial institutions to offer them (both contingent on an extension of the federal power to set interest rate ceilings). The Connecticut bill is scheduled to go into effect at the earliest of three dates: (1) when the national legislation is passed; (2) when the difference in interest rates on all savings and time deposits under one hundred thousand dollars is eliminated by savings banks and by state banks and trust companies which are not members of the federal reserve system; or (3) on December 31, 1975. Sub. Sen. Bill No. 1646, Section 7.
55. Connecticut residents have been exposed to the “NOW” account through their use in the greater Springfield, Massachusetts area, which includes a number of Connecticut towns. The Springfield savings banks have publicized their “NOW” powers in various ways including advertisements in the Springfield newspapers, which are read regularly by northern Connecticut residents. (Tr. 1001).
56. One savings bank in Connecticut, The Savings Bank of Tolland, already offers demand deposits through a power it has historically held in its charter. (Tr. 1005). Once other Connecticut savings banks exercise this right to offer checking accounts or some variation thereof, no problems are anticipated in making this service immediately available to the public, because of the extensive research that has already been conducted, based on the experience of other states. (Tr. 1653, 1658; 1742-1746).
57. Connecticut savings banks are also seeking broader consumer lending powers in order to offer a full range of consumer borrowing services. They are expected to continue their quest for the right to hold pension funds for their own employees, the right to offer credit cards, and the right to make personal loans in excess of $5,000. (Tr. 731-732; 729-730).
58. With the increasingly likely prospect of closer competition between savings and commercial banks, these institutions have amended their bylaws to prohibit interlocking relationships in their governing boards. FNH's Board of Directors voted such a resolution in early 1972 providing that due to “changing relationships between commercial banks and savings institutions, Directors of this Bank who were also trustees of mutual savings banks or savings and loan associations, would not stand for renomination at the end of the year” CNB made a similar reform in January, 1973.
D. SECTION OF THE COUNTRY
(1) The Standard Metropolitan Statistical Areas (SMSA’sJ
59. SMSA’s are formulated by the Office of Management and Budget by aggregating individual towns on the basis of the commuting patterns of the residents. The criteria used in determining these geographic areas do not include consideration of any banking data. (Tr. 129). The concept was not developed as a tool for analyzing banking markets. (Tr. 387).
60. None of the bankers who testified consider the area within a SMSA to be an area coextensive with any meaningful banking market. The term SMSA meant nothing to the Chief Executive Officer of CNB until he saw it in the Complaint in this case. (Tr. 691). Most bankers in discussing their geographic area talk in terms of service area, not SMSA’s. (Tr. 865).
61. Moreover, the report of the Department of Justice to the Comptroller on the competitive aspects of the proposed consolidation did not employ the concept of SMSA in its analysis of the. relevant geographic area in which to measure potential competition. (Tr. 533; P-95). In fact, the analysis of the proposed consolidation’s effect on potential competition as contained in the Justice Department’s report employs the town as the unit of analysis and the State of Connecticut as the relevant geographic market. (P-95, pages 4-5).
62. Even the report of the Federal Deposit Insurance Corporation to the Comptroller of the Currency on the competitive aspects of the proposed consolidation did not employ the concept of SMSA in its analysis of the relevant geographic area in this case. (Tr. 534-536; P-96).
63. A major flaw in the approach often used by the Federal Deposit Insurance Corporation (“FDIC”) to compile market share data within a SMSA is the fact that the FDIC reports only deposits resident in bank offices in such SMSA’s and does not attempt to determine the total deposits generated from that geographic area. (Tr. 1542). For example, in responding to the FDIC request as of June 30, 1970 to report its "IPC” demand deposits in its offices in the Bridgeport SMSA, CNB reported $87 million, but two months later in preparing a true geographic source analysis of the bank’s IPC demand deposits on the basis of origination by town, it was determined that a total of only $77 million originated from towns in the Bridgeport SMSA. This discrepancy resulted in the FDIC figures overstating CNB’s position in the Bridgeport SMSA by 13%. (Tr. 1542).
64. All of the larger banks in Connecticut and many out-of-state banks derive business from areas beyond their immediate office areas. (D-60; Tr. 1543).
65. The Bridgeport SMA does not comprise CNB’s relevant banking market, since only 57% of CNB’s deposits originated from towns in the SMSA, while about 88% of CNB’s total deposit business derived from the towns in which CNB has offices. (D-l; Tr. 1539-40). Common sense, then, would indicate that the relevant market areas of CNB and FNH generally coincide with where each has established branch offices.
66. The SMSA concept is particularly useless in New England where banking organizations have offices and derive substantial deposits beyond the boundaries of the SMSA where their head offices are located. (D-l; D-3; D-5; D-6; D-39 through D-49 ; D-60; Tr. 907).
67. The three federal banking regulatory agencies define a bank’s service area as the geographic area from which the bank derives 75% of its deposits. (Tr. 349, 543; 1983). The concept of market area and the concept of service area may or may not be coterminous. One of the criteria for determining whether a bank’s service area is a market is whether service area-wide pricing is affected by the behavior of participants in the market. (Tr. 353).
68. Economists would likely determine all of the competing alternatives for banking services available to Connecticut customers, including New York City banks, and take them into account in delineating a relevant geographic banking market in which to analyze the effects of the proposed consolidation. (Tr. 991-2; D-60). Perhaps the most appropriate means ■ by which the Court may measure banking markets geographically would include determining the area in which banking products or services are sold by a group of firms which act as though they are competing and evaluating evidence from representative consumers about where they aetually conduct their banking business. (Tr. 879-80, 884).
(2) The State of Connecticut as a Banking Market
69. The service and rates provided by one bank are compared by businessmen; if one bank finds it is losing customers to another because of rates or services, the latter bank will consider meeting that competition. (Tr. 362). When a bank prices its mortgage loans, for example, on a system-wide basis, the bank must take into account the pricing practices of all other commercial banks, mutual savings banks and savings and loan associations in the area. (Tr. 377-378).
70. Connecticut banks generally price their services on a system-wide basis (Tr. 903; 624-5). According to their chief executive officers, CNB and FNH both follow a system-wide pricing practice throughout their respective service areas. (Tr. 671; 1437-8).
71. Witnesses called by both sides testified that, when a bank’s prices are set on a system-wide basis throughout its service area, its competitive behavior is influenced by every bank operating in its service area. (Tr. 359,378; 671).
72. As the plaintiff readily admits, both Hartford National and CBT, the two largest Connecticut banks, are state-wide banks and operate in the service areas of both CNB and FNH. (Tr. 887-9; D-10; D-41, D-42). Consequently, CNB and FNH each must take CBT’s prices and Hartford National’s prices into account in determining their own prices, as do all banks competing with these two Hartford banks. (Tr. 381-3; 903).
73. As of August 15, 1972, the Hartford area banks had 333,273 demand deposit accounts which originated in the State of Connecticut. These accounts total $1,078,464,945. They had 301,687 savings accounts which originated in Connecticut, with deposit totals of $661,290,837.
At this same time, the Hartford area banks had 14,917 commercial and industrial loans which originated in Connecticut, with total principal balances of $462,444,000. Mortgage loans originating in Connecticut numbered 15,168, with total principal balances of $290,887,807. (D-60).
74. As of August 15, 1972, a total of 6,382 demand deposit accounts of the Hartford area banks originated in Fair-field County. Total deposits in these accounts were $21,395,000. Savings accounts originating in Fairfield County numbered 12,344, with total deposits of $19,388,000.
In addition, the Hartford area banks had on that date 1,345 commercial and industrial loans which originated in Fairfield County, with principal balances of $20,495,000. Fairfield County mortgage loans numbered 591, with principal balances of $22,771,000. (D-60).
75. As of August 15, 1972, the Hartford area banks had 22,459 demand deposits in New Haven County. Total demand deposits for that area were $52,892,000. New Haven County savings accounts held by the Hartford banks numbered 24,048, with deposit totals of $34,805,000.
The Hartford area banks had 7,555 commercial and industrial loans which originated in New Haven County, with principal balances of $40,357,000. Mortgage loans numbered 1169, with principal balances of $32,595,000. (D-60).
76. As of August 15, 1972, the Hartford banks had 121 demand deposit accounts in the City of Bridgeport, with a total deposits of $813,000. They also held 241 savings accounts, with $341,000 in total deposits in Bridgeport, as of that date.
585 commercial and industrial loans from the Hartford banks originated in Bridgeport with total principal balances of $2,748,000. (D-60).
77. In the City of New Haven, 26,962 demand deposit accounts were held by the Hartford area banks as of August 15, 1972, with total deposits of $49,045,000. 12,559 savings accounts were held by them in New Haven, with total deposits of $16,209,000.
The Hartford area banks also had 3,230 commercial and industrial loans in the City of New Haven, with total principal balances of $21,377,000. Mortgage loans in the City of New Haven numbered 427, with total principal balances-of $6,398,000. (D-60).
78. The service area of Union Trust, Connecticut’s third largest commercial bank, is equivalent to the service areas of CNB and FNH combined, excluding Bridgeport. (P-70; D-38; D-39; D-49; Tr. 891-2). If Union Trust changed its rates on consumer services, this too would have a competitive impact on CNB and FNH.
79. A comparison of the legal lending limits available to the four largest commercial banks with offices in the Bridgeport and New Haven areas shows that Hartford National and CBT, which have offices in both areas, have legal lending limits about $8 million; Union Trust, which also has offices in both areas, has a legal lending limit of $2.8 million; CNB, which has offices in the Bridgeport area, has a legal lending limit of $2.8 million; and FNH, which has offices in the New Haven area, has a legal lending limit of $2.3 million. (D-70; Tr. 678-679).
80. Obviously, CNB or FNH alone could not effectively compete with the Hartford “giants”, or even keep pace with the fast expansion of Union Trust in their own respective areas, much less in other parts of the state-wide market. The hard facts of this struggle for survival are compounded by the very real presence of mammoth New York banking firms who by their incomparable size and capabilities threaten extinction to smaller competitors in the banking region.
In its stipulations to proposed findings, the plaintiff agreed that “Statewide banking in Connecticut should be encouraged.”
(3) New York City As An “Extension” Of The Connecticut Banking Market
81. The opinion of the Comptroller approving this proposed consolidation stated that the resultant bank could compete more effectively with the larger state-wide Connecticut banks and with the out-of-state banks that canvass the area for profitable banking business. (Tr. 2013-2014; Int-1).
82. Dr. Murphy, one of the plaintiff’s expert witnesses, agreed that Stamford and Norwalk, Connecticut and New York City, might all be part of the same banking market, based on commuting patterns and the substantial competitive interaction between these areas, and that substantial competition exists for the banking business of Norwalk SMSA residents from banks in Bridgeport, Stamford and New York City. (Tr. 442-4, 448). The exhibits fully support his opinion. (P-2; P-6; P-8;' P-9).
83. All in all, approximately 50,000 heads of households commute into Connecticut from New York or vice versa. (Tr. 434; P-2; P-9). New York banks’ prices have an inevitable and substantial impact on pricing decisions of Connecticutbanks. (Tr. 1875; 902).
84. The daily circulation of New York City morning papers in Fairfield County of the New York Daily News (44,058) and The New York Times (29,360), exceeds the circulation of Fairfield County’s only morning daily, The Bridgeport Post-Telegram (11,521). Among Sunday papers, 118,405 copies of New York City newspapers circulate in Fairfield County, as compared to 79,000 copies of the Bridgeport paper. (Tr. 901).
85. The giveaway advertisements sponsored by New York City mutual savings banks in the newspapers and on television present competition to commercial banks in Connecticut for savings. (Tr. 633). In Fairfield County, New York banks advertise heavily on television, in the newspapers, and in the railroad stations. (Tr. 1873-4; D-91).
86. The television market concept, “ADI” — area of dominant influence — for New York television advertisers includes Fairfield County. The Bridgeport, Stamford, Norwalk and Danbury area accounts for 4.1% of TV households in the New York television market. (Tr. 900-1).
87. To supplement these public relations efforts, the large New York banks canvass Connecticut, calling on potential customers throughout the state. They have acquired $500 million in Connecticut business through using road men to solicit accounts. (Tr. 1792; 693-694; 631-632; D-60).
88. New York banks competing in FNH’s service area are using international services as one means of attracting customers of FNH. (Tr. 1469). According to its president, Union Trust encounters severe competition from New York City banks, particularly in the commercial lending and trust areas. (Tr. 1871). For commercial and industrial loans, New York banks actively solicit customers in Connecticut, including customers of their Connecticut correspondent banks. (Tr. 1872-3).
89. It is not possible for CNB to raise its prices on commercial and industrial loans above the going rate for such loans in national money markets such as New York and Chicago. If the bank were to do so, a large percentage of its customers having approximately $45,000,000 in commercial and industrial lines of credit would be likely to leave the bank, being of such financial stature that they have access to other money markets. (Tr. 1579-80).
90. The data compiled in the Special Master’s Report in August and September, 1972 show that six large New York City banks obtained over $483 million in deposits and over $381 million in commercial and industrial and mortgage loans from Connecticut customers, without taking into account time and savings deposits of First National City Bank or the business at certain offices of these banks. Over 23,000 deposit accounts and 1,600 loan accounts of Connecticut customers were held by such banks. (D-60, pp. 16, 23, 24).
91. Of the 23,000 demand deposit and time and savings accounts held by New York banks with statement addresses in Connecticut, about 1,000 are large accounts and the remaining 22,000 are smaller accounts opened by commuters. The accounts generally cover the entire range in terms of size. (Tr. 995-996). A substantial number of checks clearing through CNB are drawn on New York banks. (D-35, p. 44; Tr. 694).
92. Such large department stores in Connecticut as Gimbels and Wallach’s in Bridgeport and Bloomingdale’s in Stamford clear checks through New York. (Tr. 996). M. Schiavone & Sons, with its main office in North Haven, Connecticut, maintains an office in New York because of its banking requirements in New York. (Tr. 2086).
93. Warnaco, Inc., headquartered in Bridgeport, with $300 million in 1972 sales, and 1,500 employees in Bridgeport, borrows $2 million from CNB out of total borrowing of $29 million. The balance of $27 million is obtained from banks in New York, Chicago, and elsewhere. (Tr. 2129-30).
94. In sum, as a factual matter it would be ostrich-like for a Connecticut court not to recognize the closely connected portion of the megalopolis that bridges the Long Island Sound.
E. POTENTIAL COMPETITION
95. The most important questions in measuring the significance of potential competition are: 1) Who are the available entrants in terms of capability, size or geographic proximity; 2) How many potential competitors are eliminated by the merger; 3) What, if any, anticompetitive effects result from the elimination of a potential entrant?
96. Dr. Merton Peck, Chairman of the Economics Department at Yale University and a former member of the President’s Council of Economic Advisors, prepared elaborate statistical materials to support his opinions expressed on the issue of potential competition in the instant case. The Court concurs in and adopts the opinions of this highly competent and credible witness.
97. Several of his general observations are pertinent. First, the geographic market in analyzing the competitive effects of a merger might be large when considering influences on rates and very small when considering the threat of entry. (Tr. 1334). Second, the SMS A is not an appropriate area to assess potential competition. A banker in Woodbridge, for example, would not be concerned with a new entrant into Guilford, even though the two towns are in the same SMSA. The reason is simple: the driving time between the two towns is over 45 minutes and there are seven banks with 39 offices in the intervening towns between Woodbridge and Guilford. (Tr. 1239, 1347). Third, objective factors and subjective criteria must be applied. Fourth, the most significant effects of the instant merger would occur if either CNB or FNH was already operating one or more offices in a town and the other defendant was the most likely entrant, or if the defendants were ranked the number one and two banks “standing in the wings” as potential entrants into a town. In the former situation, the merger would eliminate the highest ranked potential competitor, in the latter case the consolidation would eliminate the second ranked likely entrant. (Tr. 1228, 1244-5). Fifth, legal, regulatory, and economic barriers to entry must weigh in the balance. (Tr. 1248, 1285).
98. More specifically, Dr. Peck made a detailed study of 165 of the 169 towns in Connecticut to determine the merger’s impact on potential competition. It is important to note that, in deference to the plaintiff’s contention on the “Line of Commerce” issue, he limited his survey solely to commercial banks. He did not take into account the presence of savings banks; thus, it is evident that his conclusions would be strengthened if the competitive effects of savings banks had been considered.
99. The following objective factors were applied in Dr. Peck’s study: the size of deposits of each commercial bank, the geographic proximity of the bank, its financial resources, and finally, its incentive to enter.
100. In addition, Dr. Peck considered separately the closed towns, the open towns, the 30 smallest towns, the 15 largest towns, population per banking office and concentration of the markets.
(1) Analysis By Towns
(a) b0 Closed Towns
101. By size, no first-ranked potential entrant is eliminated by the merger; in three towns, the potential entrant excluded by the consolidation ranks second. In 21 towns, the eliminated potential entrants rank seventh out of ten. (Tr. 1315; D-24).
102. By geographic proximity, only one of the 40 towns will lose the closest potential entrant; in no town is the second potential competitor cancelled. In 33 towns, the consolidation does not oust a bank ranking closer than ten other potential entrants. (Tr. 1316; D-25).
103. Dr. Peck did not attempt to combine the ranking of potential entrants by geographic proximity and by size, since he was unaware of any objective way to weigh size as against geographic proximity. He accordingly prepared tables which cross-classify the two criteria of size and geographic proximity to show the combined effect of both criteria. (Tr. 1245; D-23; D-26).
104. The cross-classification analysis indicates that in no case does the consolidation eliminate a potential entrant ranking first in geographic proximity and first in size. The potential entrant affected by the merger ranks fourth in New Haven in geographic proximity and second in size; in Stratford — third in geographic proximity and fourth in size, and in Bridgeport — fifth in geographic proximity and third in size. (Tr. 1316-17; D-26).
105. Thus, the proposed consolidation would not substantially lessen potential competition in any of the closed towns in Connecticut, based both on the ranking by size and by geographic proximity of the potential entrant. (Tr. 1321).
(b) 125 Open Towns
106. D-21 examines the effect of the consolidation in 125 open towns in Connecticut ranked by size of the potential entrant bank. It reveals that in none of the 125 towns is the potential entrant ranked first in size eliminated by the proposed consolidation. In two towns the second ranked potential entrant is excluded by the merger. The majority of the towns, namely 63, fall into the category where there is elimination of the seventh ranked potential entrant. (D-21, Tr. 1308).
107. By geographic proximity, in only one town is the leading potential entrant eliminated by the consolidation and in no town is the second ranked potential entrant eliminated. In most cases (112 out of 125) the proposed consolidation does not affect the potential entrant ranked among the top ten banks. (Tr. 1309; D-22).
108. D-23 cross-classifies the effect of the proposed consolidation on size and geographic proximity in the 125 Connecticut towns and illustrates that in no town is the largest and nearest potential entrant eliminated. (Tr. 1309-1310; D-23).
109. With respect to potential entrants identified by the plaintiff as having state-wide capability, the analysis demonstrates that in no town is there less than four potential entrants remaining after the proposed consolidation. Actually, in 52 towns, nine such potential entrants would remain after consolidation. (D-34; Tr. 1311).
110. Therefore, the proposed merger would not have the effect of significantly reducing potential competition in the 125 towns open to de novo branching. (Tr. 1311-2; D-21; D-22; D-31; D-32; D-34).
(c) 30 Small Towns
111. Prof. Peck also conducted a survey of the 30 small towns in Connecticut which have populations projected for 1980 by the Connecticut Office of State Planning of 5,000 or less and in which there is presently no banking office. (Tr. 1300; D-32).
112. With respect to each of these towns there are presently ten banks available as potential entrants identified by the plaintiff as capable of state-wide operation, while after the proposed consolidation there would remain nine. Since it is not likely that there would be more than one bank in each of these towns in 1980, and there would remain nine potential entrants available among the larger banks in the state, the proposed consolidation would not have any tendency to substantially lessen potential competition in any of such towns. (D-32, Tr. 1301).
113. Moreover, the divestiture would still leave no fewer than six potential entrants in the towns surrounding the four-town area. (Tr. 1307; D-27).
(d) 15 Large Towns
114. Of the 15 towns in Connecticut with over 50,000 in population, there is no town where fewer than four of the banks identified by the plaintiff as capable of expanding into state-wide systems would remain as potential entrants after the proposed consolidation. The most typical case is that in six of such towns six such banks would remain after the consolidation. (Tr. 1281; D-30). A sampling of the major towns is illustrative.
HARTFORD
115. With respect to Hartford, the proposed merger would have the effect of eliminating FNH as a potential. entrant ranking fifth in size among the ten largest potential commercial bank entrants. Among potential entrants ranked by geographic proximity the proposed consolidation would not affect any of the ten closest commercial bank potential entrants. (Tr. 1252; D-27 [Hartford Data Sheet]).
116. After the proposed merger is accomplished, six of the potential entrants selected by the plaintiff as capable of developing into state-wide banking operations would still remain as potential entrants into Hartford. Hartford, being the center of the insurance industry and the state capítol, makes it an attractive market for banks to enter. However, it is a closed town which makes entry more difficult. (Tr. 1255).
117. Assuming regulatory approval, it would be possible for one of the potential entrants to enter Hartford via merger with one of three small banks headquartered in Hartford. The fact that there are only three banks available for acquisition in Hartford would increase the premium that a selling bank could extract from a purchasing organization and tend to discourage merger. (Tr. 1257).
118. Generally, the size of a bank within a market area can raise barriers to entry as to outside potential competitors if there is a significant size difference. (Tr. 1365). Hartford, being the head office town of Connecticut’s two largest banks, with the market heavily concentrated between them, does not present an encouraging prospective site to potential entrants. (Tr. 1363).
BRIDGEPORT
119. In analyzing Brigeport, FNH is by size the third most likely entrant, behind Hartford National and Union Trust. Before the merger there are six banks ranked by size which are among the ten identified by plaintiff as being capable of state-wide banking; after the merger, there are five remaining likely potential entrants. Ranked by geographic proximity, FNH is fifth so that there are four other banks operating branches nearer to Bridgeport than FNH. (Tr. 1259-60).
120. The fact that Bridgeport is closed to de novo branching means that the easiest mode of entry historically is not available to potential entrants. The other mode of entry into Bridgeport would be the possibility of merger with one of the two banks operating in Bridgeport not among Connecticut’s ten largest commercial banks, Lafayette Bank and Trust Company and James Staples Company. (Tr. 1260-1). The same problems mentioned with respect to the acquisition of the small banks in Hartford exist to a heightened degree in Bridgeport.
121. The fact that Bridgeport is a relatively slow growth area in terms of various economic indicators, compared to the State of Connecticut, is an additional factor which affects the likelihood of entry by commercial banks into Bridgeport. (Tr. 1262; D-28).
122. In order to serve certain kinds of the medium-sized commercial businesses centered in downtown Bridgeport effectively, entry by a commercial bank into Bridgeport itself would be necessary. (Tr. 1264-5). In Bridgeport the fact that four of the state’s ten largest commercial banks are operating there already as existing competitors might discourage some potential entrants from entering.
NEW HAVEN
123. New Haven is the third largest town in Connecticut and the proposed consolidation would have the effect of eliminating CNB, the second largest potential entrant ranked by size. Before the merger there are five potential entrants into New Haven as ranked by the plaintiff capable of becoming statewide banks and four into New Haven after the consolidation. CNB is ranked fourth among the potential entrants into New Haven based on geographic proximity which may be a less important factor in larger towns. (Tr. 1268-1269).
124. According to D-29, New Haven, compared with the State of Connecticut, has grown more slowly. Retail and wholesale sales are lagging, total employment is down. It appears that New Haven is a less attractive market for entry than Bridgeport. (Tr. 1276; D-29). Since New Haven is a closed town and the state’s five largest banks are already in New Haven, entry, apart from merger, must be by the highly unlikely one of a new charter. (Tr. 1278; 1449-51).
125. With respect to entry into New Haven by merger, the only bank smaller than FNH would be Second New Haven Bank, one of the ten largest banks in the state. This fact would raise the cost of acquiring it considerably. (Tr. 1271).
126. Five of Connecticut’s largest commercial banks are already in New Haven and have more impact on competition than if such banks were potential competitors. (Tr. 1279). Thus, the reduction from five to four in the number of potential entrants would clearly not be of major significance in evaluating the competitive structure of banking institutions in New Haven)
STAMFORD
127. According to population projections prepared by the Connecticut State Planning Office for 1980, six banks could establish one office each in Stamford without reducing the population per banking branch below present levels. (Tr. 1297; D-31). Currently, in Stamford, only the First Stamford Bank & Trust has just one office. All the out-of-town banks have at least four offices.
128. Without reducing the population per banking office in Stamford, one or possibly two banks could enter Stamford with a system of branches up to 1980. After the consolidation, there would be five banks in Stamford identified by the plaintiff as being capable of state-wide operations remaining as potential entrants. (Tr. 1297-1298; D-31).
129. The population per bank office in Stamford is lower than the other 14 cities on D-31 except Danbury. This indicates that some of the entries into Stamford may have occurred in anticipation of growth and therefore prospects of entry into Stamford are not as good as they might otherwise be-. (D-31). The significance of the present low population per banking office figure in Stamford and elsewhere in Connecticut is that the area is saturated or approaching saturation thereby discouraging new entrants. (Tr. 1298-1299).
DANBURY
130. Danbury, although currently the smallest of these 15 towns, presents a situation similar to Stamford. (D-31). Only five bank offices could be opened in Danbury by 1980 without reducing the present population per bank office. (Tr. 1299).
131. After the proposed consolidation, four banks would remain as potential entrants into Danbury which are capable of developing into state-wide banking systems. If all four of the potential entrants entered Danbury following the proposed consolidation, the nine largest commercial banks in the state would all be operating in Danbury.
(e) Population Per Banking Office
132. Population per banking office is a useful measure in assessing a community’s need for additional banking services in that it reflects factors of local growth and personal income. This concept is also significant because it is taken into account by bank regulatory agencies in determining whether to approve de novo branch applications. (Tr. 342;. 1288; 653-654). An indication of the amount of demand in open towns in Connecticut would be represented by the present population per banking office in these areas, which was 5,633 in 1972. (Tr. 1300).
133. In a de novo branch application the examiner reports both past and estimated census figures for the head office city, the community in which the branch would be located and the service area of the proposed branch. (P-128). The summary of information accompanying the de novo branch application furnishes the population of the service area. (Tr. 1984).
134. The Comptroller’s regional administrator for New England testified that it is his practice, in making recommendations on de novo branch applications, to consider the population per banking office figure, and if it was below the state’s average for comparable towns, then other factors would have to outweigh it before he would favorably recommend its approval. (Tr. 1986). One such factor which might influence the Comptroller’s decision is the availability of enough banking business to support a new branch in the area.
135. Available banking business must exist in an area in order for the branch to be successful, and the Comptroller requires the applicant bank to demonstrate that it has a good chance of obtaining some of it. (Tr. 1821). The deposit structure and available banking business in an area are measured by sales figures and other criteria besides population. (Tr. 1797).
136. There are 136 towns in Connecticut where neither FNH nor CNB have banking offices, 103 of which are open towns and 33 closed. Of the 103 open towns, approximately 40 have no bank because they are extremely small in population, mostly in the 1,000 to 2,000 range. A branch or bank in towns with such populations would not be very profitable. (Tr. 1823-1824). Most of the non-bank towns are in rural areas where large sections are in state or national forests in Windham, New London and Litchfeld counties. (Tr. 1824).
137. In a town of 12,000 people with no bank which might justify one bank, it would take a protracted period of time for it to become profitable. The Regional Administrator, based on the factor of population per banking office only, testified that he would be reluctant to recommend that two banks be authorized to enter. (Tr. 1986-1987).
138. CNB has offices in 17 towns in Fairfield County, six of which are closed. In the remaining 11 towns, the addition of one office by FNH would reduce the population per banking office figure below the state’s level and would make it more difficult for the new entrant to become profitable and viable within a reasonable time period.
139. In the six remaining towns in Fairfield County where neither CNB nor FNH have a branch, Greenwich is closed, Easton has a population of 4,900 and Sherman has a population of 1,500. Neither has a bank in town. Easton could use one office but it would not likely become quickly profitable. (Tr. 1910). The fact that no bank has chosen to enter these towns is also given some weight by the Comptroller and his agents.