dissenting.
*1441I respectfully dissent from the majority opinion for several reasons. First, I believe that the majority places too much emphasis on the SEC’s decision not to appeal the district court’s order on the merits of the Section 10(b) and Rule 10b-5 claims. Second, I feel that the facts of this case, which will be elaborated on below, require a reversal of the attorney’s fees award because the SEC’s position before the district court was substantially justified. Finally, the special circumstances of this case would make an award of attorney’s fees unjust.
I. BACKGROUND
In order to fully understand the SEC’s position it is important to explain the context in which Kluesner was acting when he released the BMC and EMC press releases. Wordtronix was founded by Kluesner and Frederick Zimmerman in early 1981. Kluesner was primarily responsible for raising operating capital. Shortly after Wordtronix’s initial public offering in 1982, Kluesner approached several investment banking firms to obtain additional operating capital. Kluesner’s efforts were unsuccessful because Wordtronix had only completed three prototypes of its word processors and did not record any profits or sales since its inception. It was suggested to Kluesner that if Wordtronix could demonstrate “marketing activities” it might be able to raise capital in the public markets. During this same time frame the BMC agreement was signed and the press release announcing this “sales” agreement was disseminated to the same people that Kluesner had previously approached for operating capital. The press release failed to disclose that BMC was not obligated under the agreement to buy any Wordtronix word processors other than four prototypes. In essence the BMC agreement was a distributorship agreement. Further, the agreement was conditioned on the development of mathematics software by Wordtronix. Wordtronix’s stock price increased by more than 100% in just 30 trading days after the issuance of the BMC press release. Zimmerman, Wordtronix’s co-founder, expressed serious reservations about the BMC release stating that the release “sounded like that if someone looked at the next three and a half years that Wordtro-nix was guaranteed $15 million in revenue from just one customer * * * * I felt [that] was not the case.” Joint Appendix at 371-73.
In late April Wordtronix issued the EMC press release which also characterized the EMC agreement as a sales agreement. In addition, the EMC release restated the statements in the BMC release concerning the BMC agreement. Kluesner admits that at the time of the second release BMC was in default because it failed to meet one of the express conditions in its agreement with Wordtronix. The EMC release was sent to many of the same people who received the BMC release.
Shortly thereafter Wordtronix signed an investment banking commitment for a best efforts private placement of Wordtronix securities. This placement raised the capital that Wordtronix so desperately needed.
II. DISCUSSION
The majority opinion places great emphasis on the SEC’s decision not to appeal the district court’s order on the merits of the Section 10(b) and Rule 10b-5 claim. The majority states, “[significant here is the fact that the government does not challenge any of the district court’s findings as clearly erroneous. It did not file an appeal on the merits and as such this court must accept [the district court’s] findings as not clearly erroneous.” Supra at 1440.
This simply is incorrect. To follow the majority’s reasoning, appellate review of an attorney’s fees award when no appeal has been pursued in the merits phase would be limited to a determination of the correctness of the district court’s conclusions of law. Thus, this court would merely rubberstamp the district court’s findings of fact, because if we are bound by the district court's unappealed findings of fact in the merits phase then there would be no opportunity to review these findings in the context of the attorney’s fees award. The “substantial justification” test requires a *1442determination of the reasonableness of the government’s position. In order to make such a determination this court must make an independent review of both the district court’s conclusions of law and findings of fact. Merely losing the case and deciding not to appeal is not an admission by the government that its position was unreasonable, as this court noted in Keasler v. United States, 766 F.2d 1227 (8th Cir.1985):
In determining the reasonableness of the government’s position, the legislative history of section 2412(d) is helpful. The House Judiciary Committee stated that the “reasonableness” standard “should not be read to raise a presumption that the Government position was not substantially justified, simply because it lost the case. Nor, in fact, does the standard require the Government to establish that its decision to litigate was based on a substantial probability of prevailing.”
Id. at 1231 (quoting H.R.Rep. No. 1418, 96th Cong., 2d Sess., at 11, reprinted in 1980 U.S.Code Cong. & Ad.News 4953, 4990); Federal Election Com’n v. Rose, 806 F.2d 1081, 1087 n. 13 (D.C.Cir.1986).
To require the government to appeal every case in the merits phase merely to establish its EAJA justification would unduly increase the burden both on private litigants and the government; and, thereby undermine the fundamental purposes of the EAJA.
Therefore, I believe that the government’s decision not to appeal the district court's judgment on the merits in no way detracts from its justification in initiating the case. The government is obligated to evaluate an appeal independently from both a policy and an EAJA perspective.
Further, if this court is bound by the district court’s findings of fact in every case where no appeal is taken, then an appeal of attorney’s fees under the EAJA would become meaningless. “Only through a fresh look occasioned by application of the ‘substantially justified’ standard can the court honor Congress’ intent, manifest in the inclusion of this standard, not to permit a prevailing party automatically to recover fees.” Federal Election Com’n v. Rose, 806 F.2d at 1087. We must make an independent evaluation through an EAJA perspective when determining whether the government’s position was substantially justified.
A. Substantial Justification
Although the decision of the district court on the merits of the Section 10(b) and Rule 10b-5 claim is not currently before the court, I believe that the district court erred in dismissing the action against Kluesner. It goes without saying that if the district court erred in dismissing the SEC’s action and the SEC should have prevailed then the SEC was substantially justified in initiating the action. Therefore, I believe it is necessary to consider the district court’s dismissal on the merits in determining whether the SEC was substantially justified. Further, the district court’s Findings of Fact, Conclusions of Law and Order for Judgment on Kluesner’s application for attorney’s fees is so conclusory that we must look to the court's Findings of Fact and Conclusions of Law in the merits phase in order to determine the basis of the court’s order.
In cases such as this it is especially important that the district court carefully and specifically state its findings of fact and conclusions of law. In the context of an EAJA application the court’s findings of fact and conclusions of law serve a dual purpose. First, specific findings of the district court aid and facilitate appellate review. Second, specific findings are necessary to advise the litigants of the reasonableness of the positions taken before the district court. This is important in cases where the district court finds the government’s position lacks substantial justification because it may prevent the government from taking such a position in the future, thus, advancing the goals of the EAJA.
Although the district court conducted a separate hearing concerning Kluesner’s fee application, its decision awarding fees fails to discuss the basis for the award. The order simply holds that the SEC’s position was not substantially justified and points *1443to no flaw in the legal theory advanced by the Commission or the facts supporting that theory.
Finally, it is important that the district court make specific findings of fact and conclusions of law on an EAJA application because these findings are to be made after the court has taken a “fresh look” at the case with an EAJA perspective. The danger this court risks in reviewing the findings made at the conclusion of the merits phase is that these findings do not reflect the district court’s judgment from such a perspective.
The only element of the SEC’s case that was seriously disputed was whether Klues-ner acted with scienter in issuing the press releases. The district court’s findings of fact in the merits phase state that “Klues-ner’s conduct amounted to neither an intentional nor a reckless misstatement or omission of material fact.” Reviewing this finding of the district court under the clearly erroneous standard, I believe that it was erroneous.
The evidence before the district court on the issue of Kluesner’s scienter establishes that, at a minimum, Kluesner was reckless in issuing the second press release. The press releases were factually flawed, were erroneous in fact and were purposely released to raise capital from prospective investors who would rely on the truthfulness and accuracy of the releases. The facts show that Wordtronix’s co-founder expressed grave reservations concerning the impression created by the first press release. Additionally, the price of Wordtro-nix stock skyrocketed shortly after issuing the release. These facts as well as the other evidence presented at trial suggest that market participants relied on the press releases. The evidence also establishes that the first press release did not accurately reflect the true nature of the BMC agreement. Thus, to issue a second release restating the contents of the first release, as well as announcing a second “sales” agreement with EMC — an agreement that did not obligate EMC to purchase Wordtronix products — was, at a minimum reckless, if not intentional.
Once it is established that Kluesner was at least reckless in issuing the second press release, the question then becomes whether recklessness can establish the scienter requirement of a Section 10(b) and Rule 10b-5 violation. As the district court noted, this circuit has not ruled on this issue and there currently is a split of authority among the circuits which have. See Harris v. Union Elec. Co., 787 F.2d 355, 369 n. 12 (8th Cir.), cert. denied, — U.S.—, 107 S.Ct. 94, 93 L.Ed.2d 45 (1986) (noting that the prevailing approach in the courts of appeals is that recklessness satisfies the § 10(b) scienter requirement). However, it is neither necessary nor appropriate that this issue be decided in the context of this EAJA attorney’s fees award case. The SEC need only show that its theory of scienter was a good faith extension of existing law in order to be substantially justified. See Mattson v. Bowen, 824 F.2d 655, 657 (8th Cir.1987) (“The clarity of the governing law is an important factor to consider in determining whether the government’s position was substantially justified.”). Because many circuits have adopted the theory the SEC urged before the district court, I believe the SEC’s position was substantially justified. See, e.g., Harris v. Union Elec. Co., 787 F.2d at 369 n. 12 (and authorities cited therein); Woods v. Barnett Bank, 765 F.2d 1004, 1010 (11th Cir.), rehearing denied, 772 F.2d 918 (1985).
B. Special Circumstances
The majority opinion also fails to address the crux of the SEC’s argument that special circumstances existed which would make an award of attorney’s fees unjust. The EAJA expressly provides that an application for fees should be denied where “special circumstances make an award unjust.” 28 U.S.C. § 2412(d)(1)(A).
This “safety valve” [exception] helps to insure that the Government is not deterred from advancing in good faith the novel but credible extensions and interpretations of the law that often underlie vigorous enforcement efforts. It also gives the court discretion to deny awards *1444where equitable considerations dictate an award should not be made.
H.R.Rep. No. 1418, 96th Cong., 2d Sess. at 11, reprinted in 1980 U.S.Code Cong. & Ad.News 4990. See also Oguachuba v. I.N.S., 706 F.2d 93, 98 (2d Cir.1983).
As already discussed, this case is the very case to which the “special circumstances” provision was intended to apply. The SEC was advancing in good faith a novel and credible extension and interpretation of law on the scienter element of a § 10(b) violation. Awarding fees in such cases will likely chill the vigorous enforcement of the securities laws. And, in light of recent events vigorous enforcement by the SEC ought to be encouraged, not discouraged.
In addition, equitable considerations require a reversal of the attorney’s fees award. During the SEC’s preliminary investigation of Kluesner and Wordtronix, Kluesner testified that the BMC agreement was a dealer or quota arrangement whereas at trial Kluesner testified that “3,000 units were to be delivered over a three and a half year period,” indicating that the BMC agreement was a firm sale. Joint Appendix at 892. The earlier testimony during the preliminary investigation weighed heavily in the SEC’s decision to initiate this action. The majority opinion ignores the significance of Kluesner’s testimony and the district court’s finding that “Roy Zabierek, Director of Marketing at Wordtronix, and Kluesner believed that BMC’s commitment to purchase 3,000 units and resell them was firm and attainable.”
This changed testimony critically weakened the SEC’s case on the element of scienter. If Kluesner had testified at trial, consistent with his prior testimony, that the BMC and EMC agreements were distributorship agreements and not sales agreements then this would have strengthened the SEC's case because it would establish that at the time the press releases were issued Kluesner knew that they contained misstatements of material fact.
The Commission pursued the case against Kluesner relying in part on Klues-ner’s statements in the early stages of their investigation. The change of testimony that had a vital impact on the outcome of the trial compels a denial of Kluesner’s attorney’s fees. The SEC relied to its detriment on the prior testimony of Kluesner. United States v. Fidelity & Casualty Co., 402 F.2d 893, 897 (4th Cir.1968) (“Equitable estoppel is a well-established concept invoked by courts to aid a party who, in good faith, has relied, to his detriment, upon the representations of another.”) (footnote omitted). Equity requires a reversal of Kluesner’s attorney’s fees award.
III. CONCLUSION
Because at least one permissible view of the evidence leads to the conclusion that the SEC has shown a reasonable basis in law and fact for its position, I would hold that the trial court abused its discretion in awarding attorney’s fees under the EAJA. In addition, equity requires a reversal of the attorney’s fees award. Kluesner’s changed testimony is the type of special circumstance that would make an award of attorney’s fees unjust.