MERRITT, Circuit Judge.
In this securities fraud case, Peat, Mar-wick, Mitchell & Co., herein referred to as “Peat,” a firm of certified public accountants, appeals from a judgment of the District Court in the amount of $3.4 million, plus pre-judgment interest, plus attorneys’ fees of $1.2 million. The suit is a class action based upon causes of action implied under §§ 10(b)1 and 14(a)2 of the Securities Exchange Act of 1934 and SEC Rules [424]*42410b-53 and 14a-9.4 It is based on an allegedly false proxy solicitation issued in order to gain shareholder approval of a merger between two corporations, Chadbourn, Inc. and Standard Knitting Mills, Inc., herein referred to as “Chadbourn” and “Standard.” The primary issue is whether Peat is liable for a negligent error — the failure to point out in the proxy statement sent to stockholders of the acquired corporation that certain restrictions on the payment of dividends by the acquiring corporation applied to preferred as well as common stock. We hold that in the context of this case Peat is not liable for such conduct and reverse the District Court on the issue of liability.
I. STATEMENT OF FACTS RESPECTING RESTRICTIONS ON PAYMENT OF DIVIDENDS
A. General Terms and Purpose of the Merger
In April 1970, Chadbourn, Inc., a relatively profitable North Carolina hosiery manufacturer listed on the New York Stock Exchange, acquired all of the common stock of Standard Knitting Mills, Inc., a smaller, publicly-held, Knoxville, Tennessee, textile manufacturer, whose stock traded from time to time, although infrequently, in the over-the-counter market. On April 22, 1970, Standard’s stockholders at a special meeting agreed to exchange their stock for a package of Chadbourn securities. The meeting occurred after the stockholders received the proxy statement a month earlier from Standard transmitting information about the proposed merger and Chadbourn’s financial condition. The proxy statement contained a recommendation by Standard’s management favoring the merger, as well as financial statements of Chadbourn prepared by its accountants, Peat Marwick.
Before the merger, Standard’s stock traded at around $12.00 a share, although its book value was carried at approximately $21.00 a share, and Chadbourn’s stock fluctuated between $8.00 and $14.00 a share. Standard’s stockholders exchanged each share of Standard common for Vio of a share of Chadbourn common, plus IV2 shares of Chadbourn convertible, cumulative, preferred stock. The Chadbourn preferred was the main part of the package. According to the terms of the merger agreement, each share of Chadbourn preferred stock given in exchange was supposed to pay annual cash dividends of $.46% a share, and Chadbourn was supposed to redeem 20% of these preferred shares each year at $11.00 a share, beginning in 1975. Each preferred share carried a conversion privilege allowing the preferred stockholder to convert a share of preferred into 6/io of a share of Chadbourn common. The general purpose of the package appears to have been to give each Standard shareholder a set of Chadbourn securities with approximately the same market value as their Standard shares but with more liquidity and higher dividends.
[425]*425Approximately a year after the merger, Chadbourn’s sales of hosiery plummeted unexpectedly, and it suffered a loss of $17 million. This loss wiped out its retained earnings and left it with a capital deficit of $7 million. Chadbourn now was unable to redeem or pay dividends on the preferred stock. In October 1972, the former Standard stockholders sued Chadbourn, Standard, their management, their lawyers and the appellant, Peat, which was, as previously stated, the accounting firm that prepared and certified Chadbourn’s financial statements in the proxy materials. Plaintiffs entered into a settlement agreement with the defendants other than Peat, under which the former Standard shareholders were awarded control of Chadbourn, renamed “Stanwood Corporation.” The District Court did not take into account the value of the settlement received by the plaintiffs in determining damages against Peat. Since we reverse the findings of the District Court on liability, we need not reach questions concerning the measure of damages and the award of attorneys’ fees.
B. Restrictions on the Use of Retained Earnings Contained in the Chadbourn Loan Agreements
The first restriction on retained earnings is in a 5-year term loan agreement Chad-bourn made with three banks in September, 1969. Chadbourn borrowed $6 million from the banks repayable in installments over the 5-year period. In order to protect the banks, the loan agreement contained a provision which prohibited Chadbourn and its subsidiaries in any year during the term of the loan from redeeming or paying dividends "on its capital shares of any class” in an amount in excess of $2 million, less the amount of the repayments on the loan, plus future earnings after the 1968-69 fiscal year.5 These restrictions would apply to dividend payments on Chadbourn preferred shares issued to Standard shareholders (amounting annually to approximately $450,000) and would apply as well to any distributions to redeem these shares.
The second debt agreement was a little less restrictive. The second contractual restriction on paying out retained earnings is in another debt agreement which Chadb-ourn also entered into in 1969. Chadbourn borrowed $12.5 million in exchange for an issue of convertible, subordinated debentures. The effect of these restrictions on dividends and distributions was similar, except that under this agreement Chadbourn was free to use its 1968-69 net earnings of $3.1 million, as well as future earnings, for the payment of dividends and stock re-demptions.
The net effect of both sets of restrictions on dividends and redemptions, when taken together, was that Chadbourn would either have to continue to make money or refinance its indebtedness in order to meet fully its future dividend and redemption obligations on the preferred stock issued to Standard’s shareholders. As is explained above, when the bottom dropped out of its hosiery market and its retained earnings a year after the merger, it could do neither.
C. The Description of the Restrictions on Retained Earnings in the Proxy Statement
Standard’s proxy statement dated March 27, 1970, contained a 35-page description of the terms of the transaction and its tax-free nature, comparative earnings and stock prices of Standard and Chadbourn and a description of their history, business, managements and properties. The text was followed by 18 pages of financial statements of both companies, including the opinions of Peat with respect to the Chad-bourn financial statements and of Ernst & Ernst with respect to the Standard financial statements.
[426]*426The plan for the exchange of Standard stock for Chadbourn stock was set out at pages 4-8 of the text under the heading “SUMMARY OF PLAN.” Pages 6-7 of this portion of the proxy statement accurately describe the two sets of restrictions as follows:
Under the provisions of a term loan with three banks maturing October 1, 1974, Chadbourn cannot, without the consent of such banks, [1] declare any dividend or [2] make any distribution (other than common stock dividends), or [3] acquire any of its stock if, after such action, the aggregate of all dividends (other then stock dividends), other distributions to stockholders and all amounts paid for the acquisition of its stock plus the amounts of all payments made on the term loan, would exceed $2,000,000 plus Chadbourn’s consolidated net earnings since August 2, 1969. The Indenture dated as of March 15, 1969, hereinabove referred to contains certain restrictions on the payment of dividends on capital stock, however, such are less restrictive than those contained in the term loan agreement.
Chadbourn’s financial statements as of August 2, 1969, and Peat’s opinion, dated October 21,1969, were published in the back of the March 27, 1970, Standard proxy statement. The liabilities and stockholders’ equity side of the balance sheet was shown on page F-5 of the proxy statement. This page sets out amounts for current installments of long-term debt, the non-current portion of long-term debt, and “stockholders’ equity,” which referred in turn to certain notes, including footnote 7.
Footnote 7, paragraphs (c) and (d) erroneously described the two sets of restrictions as follows:
(c) As to the note payable to three banks, the Company has agreed to various restrictive provisions including those relating to maintenance of minimum stockholders’ equity and working capital, the purchase, sale or encumbering of fixed assets, incurrance [sic] of indebtedness, the leasing of additional assets and the payment of dividends on common stock in excess of $2,000,000 plus earnings subsequent to August 2, 1969.
(d) . Further, the indenture has certain restrictive covenants but they are less restrictive than those contained in the note agreement with the three banks. (Emphasis added.)
The word “common” in paragraph (c) referring to the loan agreement was wrong because the relevant provision of the loan agreement restricted the use of retained earnings for the payment of dividends on “capital stock of any class,” not just “common.” Thus the restriction on retained earnings would apply to all distributions to pay dividends or redeem the preferred shares issued to Standard stockholders should they approve the merger.
D. Facts Respecting Peat’s Negligence
The facts demonstrate that Peat’s omissions were the result of negligence but did not arise from an intent to deceive, or scien-ter, as found by the District Court.
Peat failed to disclose fully in the financial statement the restrictive effect of the loan agreement and indenture on Chad-bourn preferred stock. After each entry relating to long-term capitalization, the financial statement directs the attention of the reader to explanatory note 7. Note 7 alone pertains to Chadbourn’s long-term debt. Missing from the note is any reference to limitations that the debt agreements placed on Chadbourn preferred stock.
Only in notes 7(c) and 7(d) does Peat mention the restrictive provisions. Note 7(c), which discusses the loan agreement, reports only that “the Company has agreed to various restrictive provisions including . the payment of dividends on common stock in excess of $2 million plus earnings subsequent to August 2, 1969.” Note 7(d) describes the indenture. It says that “the indenture has certain restrictive covenants, but they are less restrictive than those contained in the [loan] agreement.” Thus, there is no indication in either note that the long-term debt restrictions affected the redemption and earnings of preferred stock.
[427]*427From notes 7(c) and 7(d), a reader easily could derive the following mistaken impression: The loan agreement contains certain restrictions on the payment of dividends by Chadbourn. As note 7(c) explicitly says, the loan agreement restrictions relate to the payment of dividends “on common stock.” The indenture contains limitations that are “less restrictive” than those created by the loan agreement. Since the limitations of the loan agreement apply only to common stock, the reader mistakenly could reason that the “less restrictive” indenture constraints appear to have no broader sweep. What note 7 conveys to the reader is the erroneous notion that neither the loan agreement nor the indenture restrictions apply to Chadbourn preferred stock.
The remainder of the proxy solicitation does not entirely correct the misunderstanding created by the financial statement notes. Peat argues that the textual language from the body of the proxy statement, quoted above in subsection C, adequately advised Standard shareholders that long-term debt agreements restricted certain aspects of Chadbourn’s preferred stock. We conclude, however, that contrary to Peat’s claim, the text is equivocal.
The text states that Chadbourn cannot “declare any dividends” or “make any distributions” under certain conditions specified by the loan agreement. These phrases are placed under the heading “The Chad-bourn Common Stock.” It would not be irrational to conclude from the location of these statements that the restrictions applied solely to the common stock of Chad-bourn. This conclusion would be confirmed by note 7(c), which explicitly states that the loan agreement restricts the payment of dividends on common stock. Moreover, under the section of the text labelled “Provisions Relating to the $.46% Preferred Stock,” there is no indication that any debt restrictions exist, much less that they apply to the dividends or redemption of preferred stock.
Nor does the language in the text regarding indenture restrictions correct the misleading impression of note 7(d). The text reports that the indenture contains “certain restrictions on the payment of dividends on capital stock.” Like note 7(d), the text fails to mention restrictions on the redemption of Chadbourn preferred stock. The language regarding dividend payment restrictions on capital stock is found in the “Chadbourn Common Stock” section. Its location casts doubt on the argument that “capital stock,” the restrictions on which are mentioned in part by the text, was meant in this context to include preferred stock. Adding to the doubt is the absence of any information about the indenture from the “Preferred Stock” section of the text. At best the textual discussion of indenture restrictions is equivocal regarding their reach.
The finding of the District Court that Peat acted with scienter in making the omissions is nevertheless clearly erroneous. We find in the record nothing to indicate that a desire to deceive, defraud or manipulate motivated Peat to omit from the financial statement information regarding the applicability of long-term debt restrictions. Indeed, Stanwood Corporation, the successor to Chadbourn controlled by the former Standard shareholders, hired and retained as vice president and treasurer the Peat associate, Hugh Freeze, who the shareholders’ counsel now claim sought to defraud them. If the shareholders and their representatives really believed Freeze intended to defraud them, it seems doubtful that they would have put him in charge of the financial affairs of the corporation.
At most the evidence supports a finding that Peat acted negligently in preparing the financial statements. Peat became aware that note 7 incorrectly described the debt limitation several weeks before the merger vote occurred. The Standard proxy statement was mailed to its stockholders on March 27, 1970. Between March 23, 1970, and April 1, 1970, Chadbourn’s outside counsel telephoned Freeze, the Peat manager in charge of the Chadbourn audit, and told him that a description of restrictions relating to Chadbourn’s stock had been inserted in the forepart of the Standard [428]*428proxy statement prior to mailing. The lawyer called to his attention the difference in the description in the proxy statement and the footnote, pointing out that the footnote said “common” rather than “capital stock of any class.” In the course of this conversation, Freeze took a copy of the preliminary Standard proxy statement and noted the change by hand in note 7(c). Thereafter, footnote 7(c) was not amended, and no effort was made to call the discrepancy to the attention of Standard stockholders or officials. Freeze did not foresee that the bottom would drop out of Chadbourn’s earnings and that what appeared to be a minor error at the time would become a major bone of contention.
The evidence simply suggests a mistake, an oversight, the failure to foresee a problem. We find nothing in the record indicating an intent to deceive or a motive for deception. J. B. Woolsey, Standard’s vice president for financial affairs, and presumably other Standard officers, knew of the restrictions and recommended the merger anyway. No stockholder testified that he was deceived. An erroneous statement cannot ipso facto prove fraud, and here we find no evidence of anything other than a negligent error.
II. LIABILITY FOR NEGLIGENT MISREPRESENTATION UNDER SEC RULES 10b-5 and 14a-9
In view of our conclusion that the District Court’s findings of scienter are clearly erroneous, we reverse the imposition of liability under Rule 10b-5. In Ernst & Ernst v. Hochfelder, 425 U.S. 185, 96 S.Ct. 1375, 47 L.Ed.2d 668 (1976), the Supreme Court settled the issue. It unequivocably held that liability under Rule 10b-5 requires “intentional misconduct.” Id. at 201, 96 S.Ct. at 1384. The Court said that 10b-5 requires “intentional or willful conduct designed to deceive or defraud investors.” Id. at 199, 96 S.Ct. at 1384.
We turn to the question of the standard of liability under Rule 14a-9 pertaining to statements made in proxy solicitations. There has been relatively little case law on the standard of liability following the Supreme Court decision in J. I. Case Co. v. Borak, 377 U.S. 426, 84 S.Ct. 1555, 12 L.Ed.2d 423 (1964), which established a private right of action under 14(a) and Rule 14a-9. Two circuits have examined the issue. Both have prescribed a negligence standard for the corporation issuing the proxy statement. One held that the negligence standard also applies to outside, non-management directors, Gould v. American-Hawaiian Steamship Co., 535 F.2d 761, 777-78 (3d Cir. 1976); and the other intimated in dicta, without deciding the issue, that a scienter standard probably should apply to outside directors and accountants, Gerstle v. Gamble-Skogmo, Inc., 478 F.2d 1281, 1300-1301 (2d Cir. 1973).
In view of the overall structure and collective legislative histories of the securities laws, as well as important policy considerations, we conclude that scienter should be an element of liability in private suits under the proxy provisions as they apply to outside accountants.
It is not simply a question of statutory interpretation. Federal courts created the private right of action under section 14, and they have a special responsibility to consider the consequences of their rulings and to mold liability fairly to reflect the circumstances of the parties. Although we are not called on in this case to decide the standard of liability of the corporate issuer of proxy material, we are influenced by the fact that the accountant here, unlike the corporate issuer, does not directly benefit from the proxy vote and is not in privity with the stockholder. Unlike the corporate issuer, the preparation of financial statements to be appended to proxies and other reports is the daily fare of accountants, and the accountant’s potential liability for relatively minor mistakes would be enormous under a negligence standard. In contrast to section 12(2) of the 1933 Act which imposes liability for negligent misrepresentation in a prospectus, Rule 14a-9 does not require privity. In contrast to section 11 of the 1933 Act which imposes liability for negligent misrepresentation in registration [429]*429statements, Rule 14a-9 does not require proof of actual investor reliance on the misrepresentation. Rule 14a-9, like 10b-5, substitutes the less exacting standard of materiality for reliance, TSC Ind., Inc. v. Northway, Inc., 426 U.S. 438, 96 S.Ct. 2126, 48 L.Ed.2d 757 (1976), and in the instant case there was no proof of investor reliance on the notes to the financial statements which erroneously described the restriction on payment of dividends. We can see no reason for a different standard of liability for accountants under the proxy provisions than under 10(b).6
We may not end our consideration there, however. We must turn to the legislative history of the proxy provisions. Section 14(a) and Rule 14a-9 are silent regarding the proper standard of liability. The Senate Report to the 1934 Act, commonly known as the Fletcher Report, discussed the sort of proxy abuse that Congress was trying to stop, that of corporate officers using the proxy mechanism to ratify their own frauds upon the shareholders, or outsiders soliciting shareholders’ approval to plunder a ripe company. The Report cited one example, quoted below in the footnote.7 The nature of each wrong deed depicted by the Report evidenced scienter.
An even more informative section of the Report is one describing the scope of 14(a):
It is contemplated that the rules and regulations promulgated by the Commission will protect investors from promiscuous solicitation of their proxies, on the one hand, by irresponsible outsiders seeking to wrest control of a corporation av/ay from honest and conscientious corporation officials; and, on the other hand, by unscrupulous corporate officials seeking to retain control of the management by concealing and distorting facts, (emphasis added)
Senate Committee on Banking & Currency, S.Rep. No. 1455, 73d Cong., 2d Sess. 77 (1934).
[430]*430The words “unscrupulous,” “concealing,” and “distorting” all imply knowledge or scienter; and we interpret “promiscuous” to mean reckless. In addition the characterization of irresponsible outsiders trying to “wrest control . . . from honest corporate officials,” implies dishonesty — and hence scienter — on the part of the outsiders. Consequently, the Report leads us to believe that its authors contemplated that 14(a) would be applied only against the knowing or reckless wrongdoing of outsiders.
The few times the proxy section was discussed in debate paint a similar picture of the type of misconduct against which 14(a) was directed. Roosevelt’s aide, Corcoran, who drafted the original version of the 1934 Act, spoke of “unscrupulous proxy committees” (emphasis added). See 78 Cong.Rec. 6544 (1934). Everett Dirksen, then a Representative from Illinois, stated, “[t]here is little doubt that there has been grave abuse of this authority to solicit proxies and the use of such proxies for manipulation ” (emphasis added) 78 Cong.Rec. 7961 (1934). Further debate indicates Congress’ concern that directors and officers should not be able to complete their fraud upon a company by means of a proxy solicitation that seeks ratification of their illegal acts. See 78 Cong.Rec. 7712-14 (1934). Congressman Pettengill of Indiana stated during Committee hearings exactly what he thought such undesirable activity amounted to: “larceny.” See Hearings on H.R. 7852 Before the House Comm. on Interstate & Foreign Commerce, 73d Cong., 2d Sess. 480 (1934). The common denominator of all these depictions of the problem is wrongdoing with some degree of knowledge, /. e. scienter; and nowhere, not in the committee reports nor in the House or Senate debates, does it appear that Congress desired to protect the investor against negligence of accountants as well.
Another important consideration is Congressional intent regarding subsequent amendments that are indirectly linked to 14(a). In passing the Williams Act of 1968 8 governing tender offers, Congress expressed the desire that proxy statements and tender offers be governed by the same rules and regulations. This would logically extend to standards of liability. Because 14(e)9 pertaining to tender offers requires scienter, we believe there is a strong policy reason for imposing a similar standard on 14(a).
According to the House Report for the Williams Act, “[t]he cash tender offer is similar to a proxy contest, and the committee could find no reason to continue the present gap in the Federal securities laws which leaves the cash tender offer exempt from disclosure provisions.” H.R.Rep. No. 1711, 90th Cong., 2d Sess., reprinted in [1968] U.S.Code Cong. & Admin.News, pp. 2811, 2813. Tender offers and proxy solicitations are two alternative methods of achieving the same result, corporate control; and Congress perceived that both were subject to the same type of abuse. It therefore acted to eliminate an existing loophole in the old law so that wrongful usurpation of control would not escape securities regulation whenever one combatant chooses to seize control by tender offer rather than by proxy fight. Id.
Senator Williams of New Jersey, the sponsor of the bill, stated
What this bill would do is to provide the same kind of disclosure requirements which now exist, for example, in contests [431]*431through proxies for controlling ownership in a company, (emphasis added)
113 Cong.Rec. 24665 (1967).
Senator Javits echoed this view.
The Senator [Williams] represents to the Senate, and I accept his representation fully, that this is analogous to the proxy rules. Id.
And Senator Williams repeated that “[t]his legislation is patterned on the present law and regulations which govern proxy contests.” Id. Logically the above testimony implies similar standards of liability for both proxy statements and tender offers. Otherwise some misleading solicitations which would trigger liability if shaped in the form of one transaction, would be immune if shaped as the other, or vice versa.
The language of the Williams Act clearly demonstrates that Congress envisioned scienter to be an element of 14(e). Congress used the words “fraudulent,” “deceptive,” and “manipulative.” This language indicates, in light of Ernst & Ernst, that 14(e) requires scienter. Although Ernst & Ernst was decided several years after the enactment of 14(e), we are bound by its holding that Congress intends scienter when it uses the above quoted language.
We conclude that 14(a) and 14(e) should be governed by the same standard of liability insofar as accountants’ liability is concerned, and that an action under 14(a) requires proof of scienter. Finding no evidence of scienter, we reverse the imposition of liability under 14(a) and Rule 14a-9.
III. COMPUTER DEFECTS
Chadbourn used electronic data processing to record many of its financial records from sales to inventory on hand. The District Court found various deficiencies connected with these computer accounts which, it concluded, Peat should have disclosed in the notes accompanying the proxy statement.
The record contains evidence of poorly documented computer programs, a high level of computer personnel turnover, lack of security in the computer room, erroneously coded data, and poorly designed computer programs that failed to detect improperly coded data. Peat sent several memoranda to Chadbourn’s management, documenting the computer weaknesses; and one internal Chadbourn memorandum not written by Peat, stated that the company was “pushed . to the brink of bankruptcy” by the unreliability of computer-generated information. The District Court found that Peat’s failure to disclose these weaknesses constituted fraud. We disagree.
An outside accountant examines the quality of a company’s internal accounting primarily to determine the extent to which he must test a client’s records. The more reliable the client’s accounting system proves to be, the less testing the accountant must conduct. A by-product of this testing is the discovery of weaknesses in internal accounting. The accountant may bring such weaknesses to the attention of management but he is not always obligated to inform the stockholders. This is not to say that an accountant may keep a blind eye to all wrongdoing while walking through a client’s corporate headquarters. He may be held liable to the extent that he intentionally or recklessly disregards the generally accepted, standard body of accounting knowledge. This is not the case here. Although we cannot say the District Court’s inference, that faulty computer information materially impaired management’s ability to manage, is clearly erroneous, there was no scienter in Peat’s failure to disclose. The absence of an intent to deceive is fatal to plaintiff’s claim.
According to the Statements on Auditing Procedure,10 promulgated by the Auditing Standards Executive Committee of the American Institute of Certified Public Ac[432]*432countants (AICPA), the accountant may have a duty to direct management’s attention to internal accounting weaknesses he has uncovered. But the Committee imposed no requirement that the notes to the certification of financial reports contain a similar disclosure of such weaknesses.
An auditor cannot always make an assessment of the effect of accounting weaknesses on the efficiency of a company. Often such an assessment requires a technical knowledge of a business in which accountants have no expertise. Peat’s reliance on the AICPA’s committee opinions is sufficient indication of good faith and lack of scienter.
Peat cannot be charged with knowledge of the interplay between poor or inefficient record-keeping procedures and mismanagement. Nor does the record disclose that Peat actually knew of Chadbourn’s internal memoranda evaluating the effect of poor computer information upon management. Likewise, the evidence does not support a finding of recklessness.
Finally, undisputed testimony shows that by the time of the proxy solicitations, many of the computer problems had already been solved. The only relevant disclosure would have been that, for the two quarters preceding the merger, Chadbourn’s management may not have been sufficiently informed due to temporary computer deficiencies. Accountants are not liable for failing to speculate publicly about this subject.
IV. THE AUDIT
The District Judge also held that Peat failed to conduct its audit according to generally accepted accounting principles (GAAP) and generally accepted auditing standards (GAAS), and therefore fraudulently misrepresented material facts about Chadbourn by certifying the proxy financial statements. The Court found deficiencies in two important areas of the audit, the closing inventory and accounts receivable. We conclude that there is sufficient evidence to support the Court’s finding that Peat inadequately tested Chadbourn’s financial figures in certain respects, but the evidence falls far short of proving that Peat intended to deceive the stockholders or that its negligence produced financial statements materially at odds with the real facts. The question of materiality in this context is whether, given all the financial information, there was a substantial risk that the actual value of assets or profits were significantly less than Peat stated them to be.
The District Judge found errors in five aspects of the inventory valuation: the physical count, the standard cost audit, the conversion from standard to actual costs, the inventory “work forward,” and the “lower of cost or market” test. An accurate appraisal of inventory is important in a business such as Chadbourn’s because, aside from exaggerating the primary asset listed on the balance sheet, overvaluing the closing inventory lowers the annual cost of sales and thus inflates net earnings.11 This latter figure is especially sensitive to fluctuation in inventory value when inventory value is disproportionately greater.
1. Physical Count
In conducting a physical inventory, independent accountants do not count every item on their client’s premises. The client’s employees count inventory items, and usually these counts are documented on accounting forms or “tickets” as to style, quantity, as well as other relevant characteristics. The accountant’s responsibility is to conduct statistical tests on a cross-section of these accounting tickets to determine whether the client counts were correct. The client counts need not exactly coincide with the accountant’s spot check count, as long as the percentage of discrepancies to total number of goods, is small.
[433]*433The District Court found that Peat conducted inadequate tests for counts of both finished stockings and greige goods.12 The finding regarding greige goods is based on testimony by witness Rittenberg that from a total of 1000 to 1500 bins of goods, a Peat employee tested about 100, and that the Peat count on items inside each bin conflicted with Chadbourn’s in 40 to 50 of the bins tested. In examining the exhibit from which Rittenberg drew his conclusion, we calculate for the bins tested that the average variance between Peat’s and Chad-bourn's count, weighted by the total number of goods, was only about four percent. The greige good inventory, as a whole, was valued at $2 million. This degree of variance in the greige good account is insufficient to render the inventory figure a material misstatement of fact.
The District Judge’s finding regarding the count of finished goods was based on cross examination testimony of Peat’s accountant, William Brasington. After selecting four work papers, Brasington stated that, for one ticket listing representing 199,000 dozen, one style consisting of 1655 dozen was tested and a difference of 28 dozen was found. He “admitted” that this test by itself would have been insufficient but claimed that other testing was done. He also testified that in another instance, a count of 435,000 dozen was not tested at all. It is not a reasonable conclusion, however, that the total amount of testing was insufficient.
Brasington vigorously insisted that on the whole testing was adequate. The Record does not indicate the total number of discrepancies found in either Brasington’s four work papers or all the work papers combined. Nor does it indicate for what purpose Brasington was asked to select these four work papers. No expert witness claimed that the testing of the physical count was insufficient. Plaintiff’s chief witness concerning the audit, Larry Ritten-berg, examined Peat’s work papers and testified in lengthy detail to the errors and omissions in Peat’s work. He did not testify that there were any errors in the physical count.
The District Judge inferred from the counts of the two inventory items that were not fully tested, and from Brasington’s failure to produce other work papers, that all the work papers relating to the physical count were deficient. But the record does not show that the four work papers were deficient on the whole.
The four papers were only a part of the testing of the total count. In order to determine whether the testing was deficient there must be evidence from a large enough sample of work papers from which a statistically valid conclusion about the whole may be drawn. The evidence does not support a conclusion that there was a material miscount in these inventory items.
Third, the District Judge found that Peat did not adequately test a computer listing of finished goods that was compiled from the accounting tickets. The testimony cited by the District Court was again taken from the cross examination of Brasington. Bras-ington said that 5 percent of the listings were tested and that 1 to 2 percent error was found. The District Court evidently interpreted the witness to mean that of the 5 percent tested, Vs to % of the listings were wrong. What the witness clearly meant, however, was that 1 to 2 percent of the sample tested were inaccurate, and interpolating this 5 percent sample to the whole, that 1 to 2 percent of the entire listing might have been inaccurate. This small error is immaterial and supports the validity of Peat’s testing.
The District Judge also found that Peat conducted insufficient cut-off tests. A cutoff test is one designed to tag inventory at the plant at which it was first counted so that it is not counted twice if shipped to another plant. The record does not indicate, however, that this affected a material portion of goods.
2. Standard Cost Audit
Next plaintiffs contend that Peat did not test Chadbourn’s “build-up” of standard [434]*434costs. A build-up of standard costs involves an estímate of costs per unit of goods; it starts with production records or engineering studies on the costs of labor and raw materials per good. The costs are totalled and checked style by style for consistency. The “build-up” represents what it should cost to produce goods under realistic day-today manufacturing circumstances barring unusual, unforeseen, and evanescent occurrences. Determining standard cost is a method of allocating total actual costs among different styles, and is an intermediate step in valuing the actual cost of the closing inventory.
The District Court found that Peat’s testing of the standard cost build-up was not documented in its work papers, and concluded that Peat never audited the standard cost at all. Plaintiffs have not proved, however, that there was a material risk that a proper audit of the standard cost system would have revealed materially lower costs. Indeed, the evidence tends to support Peat’s valuation. Actual costs were extremely close — and slightly higher — than the estimated standard cost, a prima facie indication of a relatively small risk that standard costs were materially lower. Nor does the record indicate unusual circumstances that unduly inflated actual production costs.
In the unlikely event that standard costs for some styles were materially lower, standard costs for other items must have been higher in order for the sum of standard costs to have approached the actual cost so closely.13 And in order for this mix of under and overvalued standard costs to have effected a material change in the valuation of the closing inventory, plaintiff must show that disproportionately more “overvalued” cost items than “undervalued” cost items were not sold and remained in the closing inventory. Consequently, even assuming Peat did not conduct further testing that the work papers directly indicate, this omission was immaterial. Moreover, there is a total absence of evidence of fraudulent intent.
3. Conversion of Standard to Actual Costs
A third complaint is that Peat knowingly failed to establish standard cost centers for each plant, from which standard costs would be converted to actual costs on a plantwide basis. Instead, Peat computed a weighted average variance between what it perceived to be the actual-to-standard variances for each plant, and then used one company-wide average variance to convert each good’s standard cost to its actual cost. The District Judge found on the basis of Rittenberg’s testimony that this procedure violated both GAAS and GAAP. Whether or not this finding was clearly erroneous, plaintiff’s chief witness specifically refused to conclude that there was a material risk that Peat’s calculations thereby inflated the value of the closing inventory.
A plantwide “standard-to-actual” variance would be computed and utilized in valuing the entire company’s inventory, only if there were an overall write-down from actual cost to market value in at least one plant. In order for Peat’s method of calculation to result in an inflated valuation, plaintiffs must prove (1) that there had been an overwhelming concentration of goods whose actual cost exceeded market value in plants which had above average actual-to-standard variances, and (2) that there was no overall concentration of such goods in the other plants. There is no evidence that such circumstances occurred, and again, the information on which Ritten-berg based his opinions could not have provided an adequate foundation for finding a material risk.
4. Inventory Work Forward
Peat observed the inventory approximately four weeks prior to the end of Chadboum’s fiscal year. Peat updated the inventory to its status as of the fiscal year’s [435]*435end, using computer-generated data on sales transacted (the “gross profit report”) subsequent to the taking of the physical inventory. The District Judge found that Peat’s reliance on Chadbourn’s computer system — with all its weaknesses — constituted a departure from GAAS, and therefore made its certification a material misrepresentation.
Of the long list of deficiencies the most relevant one was that “approximately 25% of customer service source documents [were] incorrectly coded.” Testimony by Dalton, the head of Chadbourn’s computer operations, indicates that a significant portion of the incorrect coding was of style number. This same style coding was used by the “gross profit report” program to retrieve individual costs for each item sold. This cost information was in turn used by Peat in its inventory work forward. If styles were incorrectly coded, this program which matched up styles with their respective costs would have been unable, by itself, to register the transaction correctly. But we cannot determine whether it would necessarily have provided an erroneous cost.
The District Judge found that Peat did not test the sales data for the work forward; this finding is not clearly erroneous. There is hearsay testimony that Peat conducted such tests in September or October of 1968, several months before the memo documenting the 25% error rate in coding. The District Judge evidently inferred from Peat’s failure to document such testing in its workpapers, that the testing did not occur. The record also establishes adequate foundation for this inference.
Without examining the computer program documentation and procedures used by Chadbourn’s computer operators for erroneously coded data, or conducting a test of the gross profit report data, Peat did not have any basis to evaluate the risk involved in using the reports. Although witness Rit-tenberg had no foundation to evaluate and testify on the risks either, we believe it is true that Peat conducted no testing on reports which dealt with $2 million in sales costs. The distinction here is between inadequate testing and no testing at all. The risk that 25% percent of styles were incorrectly coded presented the risk that the costs of 25% percent of the month’s cost transactions totaling $2 million were incorrectly stated. Although the failure here could lead to a material error, there is no evidence that it did and certainly no evidence of scienter.
5. “Lower of Cost or Market” Test
The final step in valuing a physical inventory is the markdown of items whose actual cost exceeds market value. The auditor must test a sampling of sales to determine whether its client has provided adequate reserves — and hence an adequate markdown in inventory value — for goods that are selling below cost. This is done in order to realize losses in inventory value as soon as they occur.
The District Judge found that Peat examined an insufficient number of sales to test Chadbourn’s reserves. This is supported by Rittenberg’s testimony. However, Brasington, one of Peat’s accountants, gave uncontradicted testimony that the valuations proved to be correct from subsequent market studies on individual styles. Subsequent valuations would not have been so close to Peat’s original predictions for so many different styles, had Peat’s testing been materially inadequate. The District Judge’s finding here is clearly erroneous.
6. Accounts Receivable
Finally, the District Judge found that Peat failed to “reconcile” confirmations taken at different dates of Chadbourn’s accounts receivable. A confirmation of an account receivable is simply a test of whether as of a certain date a customer of Chadbourn actually owes what Chadbourn’s records show them to owe. Peat claims that the confirmation tests were adequate because in all, over 80% of accounts were tested. But Peat misunderstands the thrust of the Court’s finding. When an accountant does not “reconcile” tests conducted on different dates, this leaves open the possibility that a customer who appar[436]*436ently is acknowledging an old debt, may actually be acknowledging a new one acquired between the two test periods. All confirmations must be confirmations of debts incurred as of a specific date.
The District Judge’s finding that these tests were not reconciled is not clearly erroneous. Accounts receivable was listed as approximately $10 million or 15% of total assets of $65 million, and is therefore a material asset. The record, however, is silent on the magnitude of error that could have resulted from conducting confirmation tests on different but relatively close dates. It is improbable — and there is no evidence — that customers’ debts increased so radically over a one month period to have materially altered Chadbourn’s total accounts receivable. Consequently, plaintiffs have not shown Peat’s failure to reconcile dates to have been material.
Accordingly, the judgment of the District Court is reversed.