Opinion for the Court filed by Circuit Judge WALD.
Dissenting Opinion filed by Chief Judge HARRY T. EDWARDS.
WALD, Circuit Judge:
Petitioner Money Station, Inc. seeks review of ah order of the Federal Reserve Board (“Board”), allowing a joint venture known as Electronic Payment Services (“EPS”), which operates the nation’s largest Automatic Teller Machine (“ATM”) network, to acquire the assets of a smaller ATM network operated by National City Corporation. [1130]*1130In its written submissions to the Board, Money Station expressed concern that the proposed transaction would allow EPS’s network of ATMs (known as the MAC network) to further increase its dominant market position in Pennsylvania, Kentucky, and Ohio, weakening smaller ATM networks like Money Station, and precluding smaller networks from themselves purchasing National City’s assets or combining with National City in an effort to create a viable large competitor to MAC. Money Station also protested that EPS had failed to show — as required by the Bank Holding Company Act (“BHC Act” or “Act”), 12 U.S.C. § 1841, et seq. — that there would be sufficient public benefits from this transaction to outweigh the possible adverse effects. Finally, Money Station requested that the Board hold an evidentiary hearing to address the disputed issues it had raised— such as its claim that this transaction would not in fact accelerate the development of new consumer banking products.
The Board, however, rejected Money Station’s request for a hearing, and approved the transaction, finding that the adverse effects from this transaction were not significant and were outweighed by the potential benefits to the public. We hold that the Board’s finding of no significant adverse effects, even if supportable, did not permit the Board to approve the transaction without an adequate evaluation and explanation of the public benefits that would be likely to arise from this transaction. We find not only that the Board failed to provide such an explanation, but that it was required to hold an evidentiary hearing to resolve disputes raised by Money Station about the putative public benefits. Accordingly, we grant Money Station’s petition for review and vacate the Board’s order.
I. Background
The question here is whether the Federal Reserve Board properly approved a proposal by a group of bank holding companies to acquire the ATM-related assets of another bank holding company. The acquiring companies — BancOne Corp., CoreStates Financial Corp., PNC Bank Corp., and KeyCorp— are large banking organizations based in Pennsylvania and Ohio, who are the four joint venturers in Electronic Payment Services, a company which operates a network of ATMs known as MAC. Under the applicants’ proposal, EPS would acquire the ATM-related assets of National City Corporation, which operates an ATM network under the name of MoneyCenter, and National City would in turn become the fifth equity owner of EPS. As part of the proposed transaction, EPS would receive approximately $74 million in cash.
Before the proposed transaction, EPS’s MAC network was the largest ATM network in the nation by transaction volume, handling nearly 100 million transactions a month through more than 13,000 ATM machines.1 As part of its network business, MAC provided all of the services essential to ATM operations, such as operating the computer links which connect the machines, processing the transactions, and providing a clearinghouse service for reporting the transactions to financial institutions. The company whose assets EPS proposed to acquire — National City — ran a smaller ATM network known as MoneyCenter, which operated about 900 ATM machines, and provided a less comprehensive range of services.
MAC serviced not only the banks of its corporate parents but many other banks throughout the Mideast region. Although MAC’s customers were free not to join a network, and could instead operate their own ATM machines (as most banks did in the early days of ATMs), banks felt pressure to join a network like MAC,
in order to give their depositors ubiquitous access to their accounts. While a bank can deploy its own ATMs, the advantage to a shared ATM network is that a bank’s depositors will be able to use ATMs at many more locations than one bank alone could practicably support- A bank — particularly a small bank, thrift or credit union with one or only a few offices — would be at [1131]*1131a competitive disadvantage if it could not offer its depositors access to many conveniently located ATMs.
United States v. Electronic Payment Servs., Inc., Proposed Final Judgment and Competitive Impact Statement, 59 Fed.Reg. 24,711, 24,713 (1994) (“Proposed Final Judgment”).2
Two months before EPS’s application was filed, the Department of Justice (“DOJ”), in a separate action, had filed a complaint against EPS, alleging that through its MAC network it unlawfully “maintained a monopoly in access to regional networks in ... Pennsylvania and ... substantial portions of Ohio,” in violation of the Sherman Act. See Proposed Final Judgment, 59 Fed.Reg. at 24,712. The complaint alleged that EPS was maintaining an illegal tying arrangement by requiring the members of its network to buy their ATM processing services from EPS instead of third-party processors, and was in effect prohibiting its members from affiliating with competing networks.3 Ultimately, the Justice Department and EPS entered into a consent decree, whereby EPS agreed to end its practice of requiring the banks which participate in its MAC network to also purchase ATM processing services from EPS, and also agreed to allow the members of its network to become members of networks which compete with MAC. See United States v. Electronic Payment Servs., Inc., 1994-2 Trade Cases (CCH) ¶ 70,796, 1994 WL 730003 (D.Del.1994).
In the meantime, the Board published a notice of EPS’s proposed acquisition of National City’s MoneyCenter network in the Federal Register, 59 Fed.Reg. 44,149 (1994), and Money Station and other protestants filed comments in opposition. Money Station alleged that the transaction would result in significant anticompetitive effects in the Pennsylvania-Ohio-Kentucky region, and specifically claimed that:
[i]f the current applications are approved, the already high barriers to entry for effective competition will be at prohibitive levels because no other network — whether MSI or any new network — will be able to attain the critical mass of ATMs necessary to support a network that could realistically exercise a competitive restraint on EPS’ pricing and related practices. The many hundreds of ATMs controlled by NCC and Mellon will not be available to another network, and neither NCC nor Mellon— both of which have experience running successful branded networks involving other banks — will be available to establish (or assist in establishing) networks competitive with MAC. This condition is not altered by the fact that the recently announced consent decree in the EPS case opens up opportunities for third-party processors. Once MAC locks up virtually all the ATMs in the Pennsylvania-Ohio-Kentucky area as customers of its branded network through the transaction with Mellon and NCC, there will be an insufficient number of banks to create a “critical mass” necessary for a new network to be formed by a processor.4
[1132]*1132Comment of Money Station, reprinted in App. 201-02. In making these allegations, Money Station echoed many of the findings of the Department of Justice, which in its action against EPS had alleged:
The MAC network is the dominant ATM network in the affected states [of Pennsylvania, New Jersey, Delaware, West Virginia and New Hampshire].... Nearly all banks in the affected states believe they have no choice but to participate in the MAC network. Banks in the affected states affiliate with MAC because MAC is the only ATM network that provides ubiquitous ATM network access throughout all or most of the contiguous affected states.... Banks in the affected states often obtain ATM network access from MAC even though defendant’s switching and processing fees, and other costs of doing business with MAC, are higher than those charged by other networks and by independent processors.
Proposed Final Judgment, 59 Fed.Reg. at 24,713.
In addition to raising concerns about the adverse effects from the transaction, Money Station claimed that the transaction was not likely to result in benefits to the public which would outweigh the possible adverse effects, and requested that the Board hold an evidentiary hearing on the proposal. In particular, Money Station questioned whether the products identified by the applicants — the development of at-home banking services and stored value cards — would really yield net public benefits, since these products were already under development by National City and other companies, and there was no evidence that this transaction would accelerate their development. As Money Station put it: “The public benefits that the applicants claim will flow from the transaction do not meet the standards of objectivity and specificity established in prior Board decisions and could be achieved in less anti-competitive ways.” Comment of Money Station, reprinted in App. 203. The Board accepted a number of written submissions from Money Station, and members of the Board’s staff met once with Money Station and the applicants, but the Board refused to hold an evidentiary hearing.
On March 1, 1995, the Board issued an order approving the transaction, and five days later issued an explanatory statement, finding that “this proposal would not result in significant adverse effects on the competitive considerations required to be reviewed under the section 4(c)(8) standard.” Board Statement at 16, reprinted in App. 648. In addition, the Board found that the “balance of the public interest factors that ... [we must] consider under section 4(c)(8) of the BHC Act is favorable, and consistent with the approval of this proposal.” Id. at 21, reprinted in App. 653. Finally, the Board rejected Money Station’s request for a hearing, finding that “Protestant’s request disputes the weight accorded to, and the conclusions that may be drawn from, all the facts of record, and does not identify disputed issues of fact that are material to the Board’s decision.” Id. at 23-24, reprinted in App. 655-56. Board Vice Chairman Blinder dissented, finding that
it seems undeniable that allowing National City’s ATM network to be merged into the MAC network would result in some adverse effect on competition. Therefore, to approve this transaction, the Board must find that there are sufficient public benefits to outweigh the loss of competition. The application, per se, demonstrates no such benefits to the public in my view.
Dissenting Statement at 1, reprinted in App. 662 (emphasis in original).
II. Discussion
The Bank Holding Company Act of 1956 reflects a policy judgment by Congress that bank holding companies should generally not acquire nonbanking enterprises. See Bank Holding Company Act, 12 U.S.C. § 1841 et seq.; S.Rep. No. 1095, 84th Cong., 1st Sess. 1 (1956) (“bank holding companies ought not to manage or control nonbanking assets having no close relationship to banking”); Connecticut Bankers Ass’n v. Board of Governors, 627 F.2d 245, 249 (D.C.Cir.1980) (“Congress heeded the frequently voiced fear that banks could wield their economic power in such a way as to dominate other elements of the [1133]*1133private sector.”). In the Act, however, Congress made an exception to this general rule in those cases where
performance by an affiliate of a holding company can reasonably be expected to produce benefits to the public, such as greater convenience, increased competition, or gains in efficiency, that outweigh possible adverse effects, such as undue concentration of resources, decreased or unfair competition, conflicts of interests, or unsound banking practices.
BHC Act § 4(c)(8), 12 U.S.C. § 1843(c)(8). The Federal Reserve Board is the agency entrusted by Congress with the initial responsibility for conducting this balancing test and determining whether the reasonably expected public benefits from a proposed transaction outweigh the possible adverse effects.5 On review, “[t]he findings of the Board as to the facts, if supported by substantial evidence, shall be conclusive.” 12 U.S.C. § 1848.
Importantly, under this statutory framework, it is not enough for a bank holding company to show an absence of potential adverse effects from a proposed transaction. Rather, the burden is on the holding company to affirmatively show that public benefits from the transaction could reasonably be expected, and would outweigh the possible adverse effects. See Citicorp v. Board of Governors, 589 F.2d 1182, 1190 (2d Cir.) (“legislative history indicates the burden is on the applicant affirmatively to establish the net public benefit of its proposal”), cert. denied, 442 U.S. 929, 99 S.Ct. 2860, 61 L.Ed.2d 297 (1979). Put another way, it is not enough that a proposed transaction would do no harm; rather, it must be likely to do some public good.
A. Adverse Effects
Although we grant Money Station’s petition for review principally on the grounds that the Board’s public benefits investigation and explanation were inadequate, we do note concern with the Board’s treatment of the adverse effects that might arise from this transaction, such as the fact that National City’s MoneyCenter network would no longer compete with MAC, and — perhaps more importantly — that the MoneyCenter assets could no longer be combined with another smaller network in order to create a viable competitor to MAC.6
In addressing the issue of MAC’S size and dominant market position, the Board basically assumed away the issue by focusing narrowly on the marginal effects of this transaction rather than on the entire competitive situation in the ATM market. Thus, instead of meeting head-on the Act’s proscription on an “undue concentration of resources,” the Board merely said:
It has been recognized that MAC has a significant position in the ATM network access services in certain states in the Mideast region. However, the significant position of a regional ATM network is not, standing alone, contrary to the public interest. Network externalities, such as the economies of ubiquity, tend to promote the consolidation of regional ATM networks. As a result, in various geographic areas, like the Mideast region, dominant ATM networks have been emerging throughout the EFT industry. One recent study indicates that the ten largest regional networks now account for 80 percent of all regional ATM transactions in the United States. In this light, the Board believes that, as a result of economic and market structure conditions, regions are likely to have one dominant ATM network.
In other words, the Board said: “It doesn’t matter if MAC is big, or if it gets even bigger, because the economics of the situation naturally favor bigness.” While this approach conveniently allowed the Board to dismiss any concerns about monopoly concentration, it certainly cannot be deemed a [1134]*1134conclusion that no adverse effects would arise from this transaction.7
We are troubled, however, by the Board’s implication that through a slow process of accretion, a network like MAC can establish a large and potentially harmful monopoly position, provided that none of the company’s individual acquisitions standing by itself is too large. To treat a company’s size and market position before a proposed transaction as irrelevant in determining whether there are potential adverse effects from a transaction is scarcely consistent with the Act’s goals of “increasing] competition” and preventing an “undue concentration of resources,” nor does it appear consistent with the Board’s own precedents of looking at the degree of monopolization of markets in analyzing transactions. See, e.g., Citicorp, 69 Fed.Res.Bull. 648, 649 (1983) (finding that “the elimination of probable future competi-tipn is not generally significant where the market is unconcentrated”) (emphasis added); Deutsche Bank AG, 67 Fed.Res.Bull. 449, 451 n. 7 (1981) (discussing possible adverse effect that “the firm resulting from such a joint venture might be unduly strengthened relative to its competitors”).
The Board’s analysis here of the possible adverse consequences, however, is barely salvaged by the fact that despite its disclaimers, the Board did acknowledge and reflect some concern about MAC’S near-monopoly position (although that concern was reflected in carefully couched language). Thus, though it found that the adverse effects arising from this transaction would not be “significant,” the Board took pains not to rule out the possibility there would be some adverse effects. See, e.g., Board Statement at 12, reprinted in App. 644 (“[t]he facts of the record do not support the view that National City would be particularly likely to enter any relevant product market in the Mideast region independently, or through another joint venture in competition with MAC....”) (emphasis supplied); Board Statement at 19, reprinted in App. 651 (“proposal is not likely to result in any significant unfair competition”) (emphasis supplied); see also Dissenting Statement of Vice Chairman Blinder, reprinted in App. 662 (there would be a “modest reduction in competition ... it seems undeniable that allowing National City’s ATM network to be merged into the MAC network would result in some adverse effect on competition”).
As discussed above, under the BHC Act, even had the Board found no adverse effects, it would still have been required to find some reasonable expectation of public benefits. But given that the Board did find some adverse effects, it was required to find more than a speculative or scant public benefit.8
B. Public Benefits
We find, however, that the Board’s public benefits findings were in fact too speculative, and were not based on substantial evidence in the record. Those findings, in their entirety, were:
Both National City and KeyCorp propose to make significant cash investments to purchase or increase their respective equity positions in the operations of EPS, including the MAC network. The capital infusions resulting from these investments should enable EPS to continue and to ex[1135]*1135pand its research and development efforts, and thereby improve its ability to develop and offer to the public innovative electronic banking and fund transfer products, such as stored value cards and home banking services.
Protestants also dispute generally EPS’s claims that public benefits would result from this proposal and, in particular, the claim that this proposal would result in innovative electronic banking products and services. While the Board recognizes that some of these products are already offered in some form, the Board believes that the availability of these products throughout a broad-based ATM network such as MAC represents some public benefit. Similarly, while the Board believes that EPS already has made substantial progress in developing these products without the contemplated capital infusions, the enhanced research and development capabilities generated by these investments should improve EPS’s ability to introduce these products sooner, to ensure the quality of the products being offered, and to present the products to a broad customer base. The Board also believes that the broader ownership base of EPS should improve the probability of success for new products by increasing the number of financial institutions and consumers that are likely to use these products in earlier stages of development.
Board Statement at 20-21, reprinted in App. 652-63. Essentially, the Board identified two public benefits — stored value cards and home banking services — and two ways this transaction would help develop these products more quickly: (1) the capital infusion from National City and KeyCorp into EPS would potentially enhance development; and (2) the broader ownership base would potentially expand the market (and thus the likelihood of success) for the new products.
We have previously recognized that the Board’s “reasoned judgments” with respect to potential public benefits are entitled to “some deference.” Connecticut Bankers Ass’n, 627 F.2d at 254 (D.C.Cir.1980). But it is equally true that the Board, like other agencies, must “articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.” Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 2866, 77 L.Ed.2d 443 (1983). Here, we do not find such a “satisfactory explanation.”
Although the Board suggested that the new cash input would be used to increase Research & Development activities, in reality KeyCorp had represented that “Cash received from KeyCorp and National City as part of the Proposed Transaction will be used to reduce [] debts and for general working capital purposes.” App. 65 (emphasis added). Neither the Board nor our dissenting colleague cites to any firm statement by EPS, National City, or KeyCorp suggesting that any of the approximately $74 million in cash EPS would receive as part of this transaction would be used to increase Research & Development activities. Moreover, although the Board acknowledged that substantial progress had already been made in developing these products, it speculated that the capital infusion would improve EPS’s ability to introduce the products sooner — though it never identified how much sooner, nor explained how increased capital would in any way speed up the development process.
Nor was it clear why EPS could not get the capital to develop these products in the absence of this transaction. If these products were really commercially viable and would benefit the public, there is no reason to suspect that EPS (or other companies) would have any difficulty in bringing them to market as quickly as possible. The record is devoid of any evidence that EPS faced obstacles in obtaining capital in the market, and thus that it was unlikely that these products would fail to be developed in the absence of this transaction.9
[1136]*1136In the past, however, the Board has looked to whether the claimed public benefits were likely to be obtained absent the transaction. For example, in Chase Manhattan Corp., 60 Fed.Res.Bull. 142 (1974), Chase Manhattan Bank had applied to acquire all the shares of Dial Financial Corporation, and cited as a public benefit its plans to open 150 new offices in the three years following the transaction, as well as plans to expand the financial services offered by Dial, such as small business loans, farm loans, and mortgages. Rejecting these claimed public benefits as inadequate, the Board said:
The record does not show to what extent such services are presently competitively unavailable in the markets served by Dial.... Dial, which has demonstrated its ability to respond to competitive challenges, would appear likely to so diversify irrespective of its affiliation with Applicant. ... [Irrespective of the proposed affiliation, it appears that Dial would plan to open a substantial number of new offices each year.... Applicant’s proposal is not considered substantially different in effect from the policy implicit in action taken already by Dial. While the proposed acquisition would clearly lead to some public benefits, there is little indication that the above or other claimed benefits are not likely to be obtained in the absence of the acquisition.
Id. at 145. Here, however, rather than following this approach, the Board failed to analyze in any meaningful way the extent to which at-home banking or stored value cards are already available in the marketplace or the chances that EPS would have brought these products to market in the absence of the transaction. If the Board had a good reason for departing here from its past practice of looking at what would occur absent the transaction, it never chose to reveal it.10
C. Failure to Hold Evidentiary Hearing
The weakness of the Board’s public benefit findings points up what we see as the fatal flaw in this process — the Board’s failure to hold an evidentiary hearing on the disputed issue of the public benefits arising from this proposed transaction. Under section 4(c)(8), and the Board’s own regulations, the Board is required to hold a hearing where “there are disputed issues of material fact that cannot be resolved in some other manner.” 12 C.F.R. § 225.23(f)(2).
We have held that if the Board wishes to deny a hearing, it “carries a heavy burden of justification ... the agency must show that the parties could gain nothing thereby, because they disputed none of the material facts on which the agency’s decision could rest.” Independent Bankers Ass’n of Geor[1137]*1137gia v. Board of Governors, 516 F.2d 1206, 1220 (D.C.Cir.1975).11 As we said in Connecticut Bankers Ass’n:
The Board cannot lightly dismiss a protes-tante request for a hearing. Where a contest exists with respect to a material fact, the Board must conduct a full eviden-tiary hearing on that issue. Moreover, the burden of making the requisite showing to trigger the hearing requirement is not great.... The protestant must make a minimal showing that material facts are in dispute, thereby demonstrating that “an ‘inquiry in depth’ is appropriate.”
627 F.2d at 251 (quoting Independent Bankers Ass’n of Georgia, 516 F.2d at 1220 n. 57). Here, the record reflects that Money Station went well beyond the “minimal showing” it was required to make.
In its submissions to the Board, Money Station pointed out that this transaction would not necessarily increase Research & Development activities because “NCC Mellon and EPS can readily engage in ad hoc research projects” without the proposed transaction. App. 368. Money Station further noted that EPS had never responded to. the Board’s questions as to the “amount of the efficiencies that would arise from the transaction.” App. 427. In addition, Money Station pointed out that the new products that would allegedly be brought into the market were not “new or distinctive, and, with respect to which, there is no basis for concluding that any contribution by EPS — as opposed to many other entities — would be distinctive.” App. 429. To the contrary, MSI listed many instances in which the products to be developed by EPS had already been developed by other companies, App. 429-30, a statement never rebutted by EPS. Although our dissenting colleague implies that these many statements were “bald or conclusory,” see Diss. op. at 1146, he does not, in our opinion, explain why.12
An evidentiary hearing was precisely what was required to address this dispute as to the existence and extent of the potential public benefits. At a hearing, officials from EPS would have been required to testify as to what use they planned to put the increased capital. They would have had to specify how much they planned to increase Research & Development, how much more quickly product development would be accelerated, and to explain just how a broader ownership base for EPS would enhance development — all questions which went unanswered in this truncated process.
Since the use to which the capital would be put — and its consequences — was both material and in dispute, a hearing was required. This court has held that “the question of the intent of the applicant under section 4(c)(8) is essentially an adjudicative issue.” Independent Bankers Ass’n of Georgia, 516 F.2d at 1224.13 And in similar circumstances, other [1138]*1138courts have held that the Board was required to hold a hearing. See American Bancorp., Inc. v. Board of Governors, 509 F.2d 29, 87 (8th Cir.1974) (finding Board improperly denied a hearing where protestants had, inter alia, raised question about “to what uses any capital infusion by [acquiring company] into [acquired company] would be put”). Given the skeletal nature of EPS’s public benefit claims, and this court’s clear statement that the Board has a heavy burden in justifying the denial of a hearing, the Board’s summary rejection of Money Station’s hearing request was contrary to law.
III. Conclusion
Given the uncertainty surrounding the public benefits that would be likely to emerge from this transaction, the Board was required, at a minimum, to hold an evidentia-ry hearing. Accordingly, Money Station’s petition for review is granted, and the order of the Board is
Vacated.