Drexel Burnham Lambert Inc. ("Drexel") and one of its former brokers, David Ragan, petition for review of a decision by the Commodity Futures Trading Commission ("Commission") ordering them to pay Sansom Refining Company ("Sansom") approximately $1.3 million plus interest and costs for trading losses. The Commission found Drexel liable for the losses in Sansom's account because Richard Robinson, the Sansom employee who initiated the unprofitable trades, lacked actual and apparent authority to buy or sell commodity futures on Sansom's behalf, and because Sansom never ratified these trades. The Commission further ruled that Sansom was not estopped from claiming injury.

We affirm the Commission's decision in substantial part. We agree that Ragan's reckless response to Robinson's unauthorized trading orders violated section 4b of the Commodity Exchange Act, 7 U.S.C. Sec. 6b, and that Drexel was liable for his actions under section 2(a)(1)(A) of the Act, 7 U.S.C. Sec. 4. We have concluded, however, that the Commission erred in ruling that the petitioners violated section 4d(2) of the Act, 7 U.S.C. Sec. 6d(2). Accordingly, we grant the petition for review with respect to the latter finding and reverse the Commission's decision on that point. The Commission also failed to address Drexel's argument that it should not bear the entire loss because Sansom failed to mitigate damages. We therefore remand the case to the Commission for further findings on this issue.

In 1980, Sansom was engaged in the business of buying precious metal as scrap, refining it, and reselling the purified product to larger refiners. Its two officers were Jackson Loughridge, President and Treasurer, and Albert Waterman, Vice-President and Secretary. Each owned 45% of Sansom's stock. In March 1980, Loughridge discussed various means of reducing Sansom's income tax liabilities with Drexel brokers. Throughout these discussions, Loughridge considered the advice of Richard Robinson, a Sansom employee.1 Robinson apparently knew more about tax law and commodity trading than did Loughridge, and Loughridge often consulted with him in such matters.

In April 1980, Loughridge and Robinson met with Ragan at Drexel's Houston office. Ragan suggested that Sansom establish an interest rate arbitrage program involving the purchase and sale of Treasury bills and Treasury bill futures, in order to convert its profits from ordinary income to long-term capital gains. Shortly thereafter, Loughridge agreed on Sansom's behalf to accept Ragan's proposal, which entailed an initial commitment of between $300,000 and $500,000. In early May 1980, Loughridge and Waterman executed Drexel's standard account-opening documents. They granted Ragan discretionary authority to trade Sansom's account. They also signed a corporate resolution form provided by Drexel, which Loughridge modified (with Waterman's permission) to authorize Loughridge alone--not Waterman as well--to trade the account. Although the corporate resolution form permitted Loughridge "to appoint any other person or persons to do any and all things which [he] is hereby empowered to do, and generally to do and take all action necessary in connection with the account," Appendix ("App.") at 812, neither Robinson nor any other person was ever authorized by Loughridge to initiate trades on Sansom's behalf.

On May 15, 1980, Ragan began buying Treasury bills for Sansom's account. Ragan telephoned Loughridge to report these initial purchases, and Drexel sent Sansom a statement, marked "Attn. Jack Loughridge," confirming the transactions. After several telephone calls, Loughridge told Ragan not to call him to report his dealings in the account; all telephonic reports, he said, should be made directly to Robinson. After reviewing the first few written account statements, Loughridge apparently ceased reading them as well. Instead, he relied on Robinson to monitor the statements and to inform him of the account's status every week or so. Between May 15 and June 10, 1980, Sansom deposited $400,000 in its account with Drexel.

On June 11, 1980, Robinson telephoned Ragan and placed an order to sell 48 pork belly futures for Sansom's account. Ragan executed the order, without asking Loughridge whether he had authorized the sale or delegated to Robinson his exclusive authority to trade. Ragan had good reasons to be suspicious of the order, both because he had never been advised that Robinson had authority to trade for Sansom, and because the order was patently at odds with the well-understood tax goals of the client's account.

On June 19, 1980, Ragan, Robinson and Loughridge met in Philadelphia. Ragan spoke very generally about Sansom's trading program and reported that all was well. Ragan did not mention the sale of pork belly futures he had made eight days before at Robinson's behest, although at that meeting he could easily have verified Robinson's authority to initiate trades. Ragan also furnished Loughridge with a list of transactions in Sansom's account through June 11. Significantly, however, the list did not include the sale of pork belly futures on June 11 that Robinson had requested, even though three other transactions involving Treasury bills were listed for that date.

On July 9, 1980, Robinson directed Ragan to offset half of the pork belly futures at a loss of more than $77,000. Ragan did so, again without telephoning Loughridge to confirm Robinson's authority to initiate trades. During the rest of July, Robinson ordered numerous other unprofitable trades in pork belly and live cattle futures. In August and September, he speculated even more heavily, accumulating huge losses.

At no time did Ragan ask Loughridge whether Robinson was authorized to trade for Sansom. When Sansom's losses began to mount, however, he did express his worries to Robinson, who stated that Sansom had hedged the unprofitable trades through orders placed at Bache Halsey Stuart Shields Inc. ("Bache"). Ragan telephoned a broker at Bache to confirm Robinson's story. He was told that Sansom's account at Bache enjoyed a surplus roughly equal to Sansom's aggregate losses at Drexel. The Bache broker refused to tell Ragan, however, what trades had been made through that account. Hence, Ragan could not corroborate Robinson's assertion, although his fears were somewhat allayed. In fact, Sansom's account at Bache was not used to hedge Robinson's trades at Drexel.

Throughout this period, Drexel regularly sent account statements to Sansom, marked to the attention of Loughridge. The statements requested the client to report any inaccuracies immediately. Sansom never complained about the unauthorized trades, because Loughridge trusted Robinson to read the statements and apprise him of the account's status, and Robinson never mentioned the unauthorized, speculative commodity trades that he had placed with Ragan. Sansom deposited over $1.3 million in the account between July and September 1980 in order to cover its losses. Most of the checks were signed by Waterman. Neither Loughridge nor Waterman questioned Robinson when he presented the checks for their signatures.

In mid-September 1980, Ragan and Loughridge discussed a gold trade that Ragan had made for Sansom's account. It is unclear who initiated the call, or how Loughridge learned of the trade. However, it is clear that, as soon as Loughridge became aware of the unauthorized trade, he was explicit in instructing Ragan not to trade in gold. Even though Loughridge expressed concern about commodity trading in the company's account, Ragan never volunteered any information about the orders that had been placed by Robinson, nor did he use the occasion of their September discussion to inquire regarding the efficacy of the commodity trades.

On September 21, 1980, Robinson finally confessed to Loughridge that he had speculated in Sansom's account and had incurred colossal losses. Loughridge promptly closed out all the open commodity contracts the following day. Sansom maintained its account at Drexel, however, so as not to lose the tax benefits it expected to reap from its Treasury bill spreads.

On June 3, 1982, Sansom filed a reparations claim with the Commission against Drexel and Ragan. Sansom's complaint alleged violations of section 4b of the Commodity Exchange Act, 7 U.S.C. Sec. 6b.2 The Administrative Law Judge ("ALJ") who conducted a hearing on Sansom's complaint ruled in favor of Drexel and Ragan and dismissed Sansom's action on May 27, 1986.3 He found that, although Robinson lacked both actual and apparent authority to initiate trades in Sansom's account, Sansom was estopped from claiming injury from the unauthorized trades because Sansom had not complained of them to Drexel despite having received reports of the transactions and having paid $1.3 million to meet its margin requirements.

Sansom appealed and the Commission reversed.4 The Commission agreed that Robinson lacked actual and apparent authority to trade Sansom's account. Having found that Ragan accepted orders even though Robinson had no actual or apparent authority to trade for Sansom, the Commission ruled that Sansom could not be estopped from claiming injury. The Commission concluded that "Ragan was unreasonable under the circumstances to allow Robinson to begin trading the account and remained unreasonable in allowing trading to continue, even in light of Sanson's failure to protest and payment of margin...." Commission Op. at 34,108. In the Commission's view, "Ragan's conduct amount[ed] to willful disregard of whether he was acting in accordance with Sansom's instructions," and thus constituted a violation of section 4b. Id. at 34,108 n. 10.

The Commission also found Drexel and Ragan liable under section 4d(2), 7 U.S.C. Sec. 6d(2).5 According to the Commission, section 4d(2) "places the burden on the commodity professional to ascertain the authority of an individual purporting to act for a customer. If the associated person fails to take reasonable steps to learn the limits of an individual's authority, and the futures commission merchant fails to take reasonable steps to prevent, detect, and correct such employee errors, complainants will not be denied recovery merely because they could have done a better job of protecting their own interest." Commission Op. at 34,108. Without considering whether Sansom had failed to mitigate damages, the Commission ordered Drexel and Ragan to pay Sansom $1,322,074.50 with interest from September 22, 1980. Drexel and Ragan seek review of the Commission's decision.

The Commodity Exchange Act provides that "the findings of the Commission as to the facts, if supported by the weight of evidence, shall ... be conclusive." 7 U.S.C. Sec. 9. In Great Western Food Distribs. v. Brannan, 201 F.2d 476, 479-80 (7th Cir.), cert. denied, 345 U.S. 997, 73 S. Ct. 1140, 97 L. Ed. 1404 (1953), the Seventh Circuit stated that a reviewing court's function under this standard

is something other than that of mechanically reweighing the evidence to ascertain in which direction it preponderates; it is rather to review the record with the purpose of determining whether the finder of the fact was justified, i.e. acted reasonably, in concluding that the evidence, including the demeanor of the witnesses, the reasonable inferences drawn therefrom and other pertinent circumstances, supported his findings.

This circuit endorsed the preceding gloss on the "weight of evidence" standard in Schor v. CFTC, 740 F.2d 1262, 1272 (D.C.Cir.1984), vacated and remanded on other grounds, 473 U.S. 922, 105 S. Ct. 3551, 87 L. Ed. 2d 674 (1985), reinstated, 770 F.2d 211 (D.C.Cir.1985), rev'd on other grounds, 478 U.S. 833, 106 S. Ct. 3245, 92 L. Ed. 2d 675 (1986). We may only set aside the Commission's factual findings, and the legal conclusions they entail, if they are not reasonably supported by the record.

It is also clear that it is the decision of the Commission, not that of the ALJ, that is subject to judicial review. And when, as in the instant case, the Commission and an ALJ disagree on factual inferences to be drawn from the record, the Supreme Court has told us that the question to be decided is not whether the agency has "erred" in "overruling" the ALJ's findings, but whether its own findings are reasonably supported on the entire record. See, e.g., FCC v. Allentown Broadcasting Corp., 349 U.S. 358, 364, 75 S. Ct. 855, 859, 99 L. Ed. 1147 (1955); Universal Camera Corp. v. NLRB, 340 U.S. 474, 492-97, 71 S. Ct. 456, 466-69, 95 L. Ed. 456 (1951).6

In this case, the Commission and the ALJ disagreed in their views of the record with respect to several critical points, including the significance of Robinson's lack of actual or apparent authority to trade; Ragan's unreasonableness in simply assuming that Robinson was authorized to trade; and Ragan's failure to act on his suspicions and inquire regarding Robinson's unauthorized trades, especially when he knew them to be inconsistent with the tax goals of Sansom's account. In weighing the evidence in the record, the principal differences between the initial decision of the ALJ and the final judgment of the Commission are the findings (1) that Ragan's conduct constituted a willful disregard of whether he was acting in accordance with Sansom's instructions, and (2) that Sansom could not be estopped from denying Robinson's authority to trade when both the ALJ and the Commission agreed that Robinson had neither actual nor apparent authority to place orders for Sansom. These are the findings--including the underlying inferences drawn by the Commission and its weighing of the entire record of testimony--that are before this court on review. Unless otherwise restricted by statute or rule, an administrative agency "is not bound by [an ALJ's] 'secondary inferences,' or 'derivative inferences,' i.e., facts to which no witness orally testified but which the [ALJ] inferred from facts orally testified by witnesses whom the examiner believed. [An agency] may reach its own 'secondary inferences,' and we must abide by them unless they are irrational...." NLRB v. Universal Camera Corp., 190 F.2d 429, 432 (2d Cir.1951) (Frank, J., concurring) (footnote omitted).7 If the Commission's findings are reasonably supported by the record, then the petition for review must be rejected.

With respect to "a pure question of statutory construction," however, such as the Commission's reading of section 4d(2), "our first job is to try to determine congressional intent, using 'traditional tools of statutory construction.' If we can do so, then that interpretation must be given effect...." NLRB v. United Food & Commercial Workers Union, Local 23, --- U.S. ----, 108 S. Ct. 413, 421, 98 L. Ed. 2d 429 (1987) (quoting INS v. Cardoza-Fonseca, 480 U.S. 421, 107 S. Ct. 1207, 1221, 94 L. Ed. 2d 434 (1987)). If, on the other hand, "the statute is silent or ambiguous with respect to the specific issue," Chevron U.S.A. Inc. v. Natural Resources Defense Council, 467 U.S. 837, 843, 104 S. Ct. 2778, 2781, 81 L. Ed. 2d 694 (1984), then the question for us becomes whether the Commission's construction of the statute is "permissible," id., that is, one that is "rational and consistent with the statute." United Food & Commercial Workers, 108 S. Ct. at 421.

Section 4b renders it unlawful "to cheat or defraud" or "willfully to deceive" any person in regard to any commodity contract in interstate commerce. Other circuits that have construed this provision are agreed that "mere negligence, mistake, or inadvertence" fails to meet section 4b's scienter requirement; "a degree of intent beyond carelessness or negligence" is necessary to violate this provision. Hill v. Bache Halsey Stuart Shields Inc., 790 F.2d 817, 822 (10th Cir.1986); see also Greenwood v. Dittmer, 776 F.2d 785, 789 (8th Cir.1985); Haltmier v. CFTC, 554 F.2d 556, 562 (2d Cir.1977). The question here is whether section 4b encompasses reckless conduct.

We hold that recklessness is sufficient to satisfy section 4b's scienter requirement. A reckless action, as the First Circuit said in reaching the same result, "is one that departs so far from the standards of ordinary care that it is very difficult to believe the [actor] was not aware of what he was doing." First Commodity Corp. v. CFTC, 676 F.2d 1, 7 (1st Cir.1982). The language of section 4b, together with the virtually unanimous agreement among the circuits that recklessness may serve as the predicate for liability under the analogous provisions of section 10(b) of the Securities Exchange Act of 1934 and SEC Rule 10b-5, see, e.g., Rolf v. Blyth, Eastman Dillon & Co., 570 F.2d 38, 46 (2d Cir.), cert. denied, 439 U.S. 1039, 99 S. Ct. 642, 58 L. Ed. 2d 698 (1978); McLean v. Alexander, 599 F.2d 1190, 1197-98 (3d Cir.1979), convinces us that reckless inattention to obvious dangers to a client's interests in arranging a purchase or sale for the client's account triggers liability under section 4b. As Drexel's counsel acknowledged at oral argument before this court, the Commission's finding that "Ragan's conduct amounts to willful disregard of whether he was acting in accordance with Sansom's instructions," Commission Op. at 34,108 n. 10, plainly fulfills section 4b's scienter requirement if it is supported by the weight of evidence.

In our view, the Commission's determination is firmly rooted in the record before it. Loughridge was the only Sansom officer authorized to trade Sansom's account, apart from the discretionary authority vested in Ragan with respect to the purchase and sale of government securities to attain Sansom's tax objectives. Loughridge modified the corporate resolution form provided by Drexel to so provide, and Ragan was or should have been aware of this fact. At no time, moreover, did Loughridge inform Ragan or Drexel that he had delegated to Robinson his authority to trade. Nor did Ragan avail himself of numerous opportunities to ascertain Robinson's authority to initiate trades when he placed commodity futures orders, even though he had reason to be suspicious of those orders. As the Commission noted, Loughridge initiated no trades himself, and those that Robinson ordered "were unmistakably intended for speculative purposes in commodities bearing no relation to Sansom's tax goals." Commission Op. at 34,107. Ragan was also aware that Loughridge was not knowledgeable about commodity trading, and he had no reason to surmise that Loughridge was eager to speculate with the money he hired Ragan to shield from income tax. Under these circumstances, the Commission properly concluded that it was highly irresponsible for Ragan to place Robinson's orders without verifying his authority. Yet, he did not even call Loughridge--not when Robinson phoned in his first trade, and not even when Sansom's losses ran to hundreds of thousands of dollars. Ragan did not mention the first pork belly futures trade when he met with Loughridge and Robinson on June 19, 1980, and the list of trades he tendered to Loughridge that day inexplicably omitted the pork belly futures trade, despite the fact that Treasury bill transactions on the same day were included. He further failed to mention Robinson's commodity futures trades when Loughridge rebuked him for trading in gold in mid-September. Given this record, the Commission had ample support for the finding that Robinson lacked both actual and apparent authority to trade Sansom's account. See ALJ Op. at 32,198; Commission Op. at 34,107. Furthermore, we deem the evidence adduced by the Commission more than sufficient to sustain its conclusion that Ragan's conduct constituted "willful disregard of whether he was acting in accordance with Sansom's instructions," in clear violation of section 4b. Commission Op. at 34,108 n. 10.

The ALJ "concluded that Ragan, with some justification believed that Robinson had authority to trade on behalf of Sansom." ALJ Op. at 32,201. But the Commission weighed the evidence entirely differently when it found that, "[i]n light of the facts available to Ragan, he was unreasonable in simply assuming that Robinson was authorized to trade the account. At a minimum, Ragan should have sought clarification from Loughridge, an individual Ragan knew and to whom he had reasonable access." Commission Op. at 34,107. In short, if the ALJ thought that Ragan had "some justification" for his unauthorized dealings with Robinson, the Commission found it to be both "unreasonable" and reflective of a "willful disregard of whether he was acting in accordance with Sansom's instructions." Because the Commission's findings are reasonably supported by the record, we are bound to accept these determinations.8

The Commission also expressly noted that the ALJ's estoppel ruling, which derived from the same facts that the ALJ thought warranted Ragan's erroneous belief that Robinson had been empowered to trade Sansom's account, was "fundamentally inconsistent with the judge's prior determination that Ragan did not reasonably rely on statements or activity attributed to Sansom in allowing Robinson to trade." Commission Op. at 34,108. The Commission further rejected the claim that Sansom's failure to protest and its continued payment of margin rendered reasonable Ragan's persistent neglect to inquire into Robinson's authority to trade. See id. In support of this conclusion, the Commission pointed out that Ragan's awareness that Loughridge looked to Robinson to supervise the account "diminishe[d] the weight he could reasonably attribute to the failure to protest." Id. at 34,108 n. 8. It also noted that Ragan decided not to contact Loughridge when Robinson's trades produced large losses, but instead contented himself with a hasty and unsatisfactory inquiry into the status of Sansom's account at a rival investment bank--a decision which the Commission, in its expert judgment, found unjustifiable. See id. Finally, the Commission reasoned that Sansom's payment of amounts in excess of $500,000 to maintain its account "are not persuasive evidence of approval of Robinson's trades in light of the complex nature of the straddle strategy at issue. It is not uncommon that margin requirements will be underestimated in the course of soliciting participation in such programs." Id. at 34,108 n. 9.

In view of the deference we owe to an agency's factual determinations, including the reasonable inferences drawn therefrom, we conclude that the Commission's finding that Ragan acted with "willful disregard" of whether he was heeding Sansom's instructions is reasonably supported by the great weight of the evidence in the record of this case. That the Commission cast the relevant facts in a different light and drew different inferences than the ALJ is not in itself reason for us to gainsay the Commission's judgment, provided--as is here the case--that it was reasonable. See notes 6 and 7 supra and accompanying text. The Commission's conclusion fully accords with its findings of fact, and its explication of that conclusion suffices, particularly given the inconsistency at the heart of the ALJ's opinion.9C. Ratification

As a possible affirmative defense against a finding of liability, Drexel and Ragan contend that Sansom ratified Robinson's unauthorized trades when it continued to fund its account, without protest, after receipt of Drexel's confirmation statements. This contention is insupportable in view of the prevailing legal standard and the facts of this case. The Commission has held that a customer ratifies unauthorized trading for his account "only where it is clear from all the circumstances presented that the intent of the customer was to adopt as his own and for all time the trades executed for his account without authorization." Sherwood v. Madda Trading Co., [1977-1980 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 20,728 at 23,020 (CFTC Jan. 5, 1979). Circuit courts that have confronted the issue have also deemed knowledge of the relevant facts and an intent to approve the unauthorized action after its occurrence to be preconditions to ratification. See, e.g., Hill v. Bache Halsey Stuart Shields Inc., 790 F.2d 817, 827 (10th Cir.1986); Karlen v. Ray E. Friedman & Co. Commodities, 688 F.2d 1193, 1198 (8th Cir.1982); Thropp v. Bache Halsey Stuart Shields, Inc., 650 F.2d 817, 822 (6th Cir.1981); Shearson Hayden Stone, Inc. v. Leach, 583 F.2d 367, 369-70 (7th Cir.1978).

We embrace the Commission's standard for the affirmative defense of ratification, and agree with the Commission that Drexel and Ragan have failed to meet it in this case. Commission Op. at 34,107. There is no evidence that either Loughridge or Waterman was aware that the checks they signed were to fund Robinson's trading losses rather than the Treasury bill arbitrage program they were pursuing. See ALJ Op. at 32,198-99; Commission Op. at 34,107. Indeed, Loughridge's decision to close out all commodity positions as soon as Robinson confessed his unauthorized dealings, despite the heavy losses Sansom thereby incurred, strongly supports the finding that he was ignorant of Robinson's trading and that he did not intend to approve it after the fact. See id. Furthermore, Loughridge's decision to maintain Sansom's account at Drexel after Robinson's trading was discovered, in order not to lose the tax advantages he sought, cannot be deemed ratification of those trades Loughridge did not authorize. See ALJ Op. at 32,198.

Drexel and Ragan also argue that Sansom is estopped from claiming injury because it failed to protest upon receipt of Drexel's account statements listing the unauthorized trades, because it repeatedly sent checks to Drexel to cover Robinson's trading losses, and because it was aware of Robinson's history of gambling and embezzlement whereas Drexel was not.

We reject this argument. Normally, four elements must be present to establish the affirmative defense of estoppel: (1) the party to be estopped must have known the facts; (2) the party against whom estoppel is asserted must have acted in a manner that caused the other party reasonably to believe that it intended whatever action it is allegedly estopped from citing as the basis of its claim; (3) the party asserting estoppel must have been justifiably ignorant of the relevant facts; and (4) the party asserting estoppel must have relied on the other party's conduct to his injury. See Sherwood v. Madda Trading Co., [1977-1980 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 20,728 at 23,021 (CFTC Jan. 5, 1979). The Commission found, and we agree, that both the second and third elements were absent in this case.

Drexel and Ragan manifestly lacked a reasonable belief that Sansom sanctioned Robinson's trades. Both the ALJ and the Commission found that Robinson lacked actual and apparent authority to initiate trades. If Robinson had neither actual nor apparent authority to trade, then Ragan and Drexel are hard-pressed to suggest that they reasonably believed that Robinson's orders were authorized. Furthermore, the Commission pointed out that Ragan was aware of Loughridge's reliance on Robinson and that he knew or should have realized that Loughridge and Waterman might have signed the checks necessary to maintain Sansom's account because they were ignorant of the details of Ragan's trading program and did not know how much money was necessary to finance it. See Commission Op. at 34,108 & nn. 8-9. Hence, Drexel and Ragan cannot plausibly contend that their asserted belief in Sansom's acquiescence in Robinson's speculative trading was justified by Sansom's failure to protest and its payment of margin requirements. Ragan had clear instructions that only Loughridge had the authority to trade for Sansom, and these instructions were never amended during the course of Robinson's unauthorized trading.

Moreover, Ragan's failure to inquire into Robinson's authority to trade precludes the petitioners from claiming that they were justifiably ignorant of Sansom's disapproval of Robinson's trades. A party urging estoppel must show that it took reasonable steps to discover relevant facts. See Keller v. Scoular-Bishop of Missouri, Inc., [1986-1987 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 23,128 at 32,336 (CFTC June 26, 1986); Sherwood, p 20,728 at 23,021.10 The petitioners failed to do so. Ragan could easily have ascertained, at any time, whether Robinson had been authorized to trade. See Commission Op. at 34,108. Moreover, he plainly had a duty to verify Robinson's authorization, both because Robinson did not possess even apparent authority to trade and because his commands were highly suspicious, given that the speculative commodity trades Robinson ordered would not advance the objectives of Sansom's Treasury bill arbitrage program. Under the circumstances, Ragan did not display reasonable diligence in determining Robinson's authority or lack thereof, and a modicum of effort--a brief telephone call, or a single question to Loughridge when they met in person on June 19, 1980--would have revealed that Robinson did not possess authority to place orders for commodity futures. The Commission was therefore correct in ruling that Drexel and Ragan were not justifiably ignorant of Sansom's disapproval of Robinson's trades, and thus that Sansom was not estopped from claiming injury.

In addition to their claims with respect to ratification and estoppel, Drexel and Ragan also assert that Sansom should bear at least part of the losses incurred because it failed to take reasonable steps to mitigate damages.11 The Commission has recognized that "each of these defenses is slightly different," Sherwood v. Madda Trading Co., [1977-1980 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 20,728 at 23,018 (CFTC Jan. 5, 1979), yet no judgment was offered in this case on the claim of mitigation.

Although Commission precedent admits of some confusion on this point,12 it would appear that the doctrine of mitigation of damages, unlike that of estoppel, looks solely to the conduct of the party requesting damages. As was noted in Sherwood,

[c]omplaining to the responsible officer or agent of one's futures commission merchant is, in the Commission's view, the mandatory first step which a customer must take to mitigate damages upon discovery of unauthorized trading.

Id. at 23,021.13 Often, as in Sherwood, judgments on ratification and estoppel appear to subsume the inquiry on mitigation, even though the doctrinal analyses are distinct. Nonetheless, as is also clear from Sherwood, there are cases in which different results may obtain with respect to different trades, depending upon whether the complainant "was aware that unauthorized trades had been credited to his account." Id. The Commission must reach a determination on this question in order to dispose of the claim of mitigation. Accordingly, we will remand the case for the Commission's ruling on this issue.

The final issue before us concerns the Commission's purported reliance on section 4d(2) in finding Drexel and Ragan liable under the Act. Section 4d(2) requires a futures commission merchant to "treat and deal with all money ... received by [him] to margin, guarantee, or secure the trades or contracts of any customer ... as belonging to such customer." 7 U.S.C. Sec. 6d(2). The Commission ruled that Ragan's trading in accordance with Robinson's unauthorized orders violated section 4d(2). The Commission explained:

Section 4d(2) of the Act ... places the burden on the commodity professional to ascertain the authority of an individual purporting to act for a customer. If the associated person fails to take reasonable steps to learn the limits of an individual's authority, and the futures commission merchant fails to take reasonable steps to prevent, detect, and correct such employee errors, complainants will not be denied recovery merely because they could have done a better job of protecting their own interest.

Commission Op. at 34,108. In support of this claim, the Commission cited its decision in Hunter v. Madda Trading Co., [1980-1982 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 21,242 at 25,204 (CFTC Sept. 2, 1981).

The Commission's reading of section 4d(2) is unpersuasive. Neither the express terms of the statute nor the legislative history of section 4d(2) buttresses the Commission's assertions. Contrary to the Commission's ill-defended claims in Hunter, section 4d(2) was intended for one purpose: to prevent an unscrupulous broker from commingling clients' margin funds with his own and then using those funds to speculate for the broker's own account, thereby imperiling his clients' prospects of obtaining a refund of their margin deposits should the broker's gambles fail. See 80 CONG.REC. 6162, 6612-13, 7910-12 (1936); see also Craig v. Refco, Inc., 624 F. Supp. 944, 946-47 (N.D.Ill.1985) (recounting legislative history), aff'd, 816 F.2d 347 (7th Cir.1987); Marchese v. Shearson Hayden Stone, Inc., 644 F. Supp. 1381, 1384 (C.D.Cal.1986) (same), aff'd, 822 F.2d 876 (9th Cir.1987). It does not reach every failure by a broker to act in accordance with a client's orders.

Moreover, as the petitioners and amicus point out, the Commission's construction of section 4d(2) would eviscerate section 4b, rendering nugatory its scienter requirement. For if the Commission's interpretation of this section were correct, the innocent or negligent mishandling of client funds, which the Commission recognizes is immune from liability under section 4b in virtue of its scienter requirement, see Hunter, p 21,242 at 25,204 n. 8, would automatically generate liability under section 4d(2). Congress cannot have intended this irrational result. Hence, we reverse the Commission's decision that Drexel and Ragan are liable under section 4d(2).

We deny Drexel and Ragan's petition for review with regard to the Commission's finding that the petitioners violated section 4b of the Commodity Exchange Act by dint of Ragan's reckless compliance with Robinson's orders without attempting to verify Robinson's authority to trade Sansom's account. The Commission was also correct in ruling that Sansom was not estopped from claiming injury. We find, however, that the Commission's construction of section 4d(2) was erroneous. We therefore grant the petition for review insofar as it concerns the petitioners' liability under that section. Finally, we remand this case to the Commission with instructions to consider whether Sansom failed to mitigate damages.14

So ordered.

STARR, Circuit Judge, concurring in part and dissenting in part:

I am in accord with much of Judge Edwards' thorough opinion for the court. My disagreement is limited, but it is fundamental. In my view, this record will not support the Commission's conclusion that Ragan's conduct satisfied the scienter requirement of section 4b, 7 U.S.C. Sec. 6b (1982). To the contrary, this record fails to yield substantial evidence to support the Commission's inference of recklessness.

In reaching this conclusion, I am fully mindful of our limited role in analyzing the record and the deference which we rightly are to show to the agency in its factual determinations. But it is odd in the extreme to conclude, on this record, that Ragan acted recklessly.

* In large part, the Commission's recklessness determination followed from its rejection of the ALJ's conclusion that Ragan "with some justification" believed Robinson to be authorized to trade in Sansom's account with Drexel. The Commission, and now the court, thus finds a glaring inconsistency in the ALJ's opinion: that is, the ALJ concluded that there was no apparent or actual authority for Robinson to trade in the account, yet the ALJ went on to conclude that Ragan's (undisputed) good-faith belief in that authority was supported by "some justification." But the "inconsistency," upon reflection, is readily resolved. As a matter of law, for apparent authority to exist, there must be some conduct on the part of the principal to give rise to the inference of the agent's authority. See Restatement (Second) of Agency Secs. 8, 8B. On this record, the ALJ properly found that Sansom did not act in such a manner as to establish apparent authority. But as a factual matter, Ragan could readily and in good faith (although in error) have believed Robinson to have been so authorized. Here is why.

First, although Sansom may not have acted in a fashion that would require a finding of apparent authority, Sansom surely gave some indications of authority. Second, and more fundamentally, that Robinson lacked authority represents a legal conclusion; in contrast, the finding that Ragan's good-faith belief in Robinson's authority had some justification is a factual conclusion. The Commission did not reverse the ALJ's factual finding that Ragan acted in good faith; indeed, to do so would have been remarkable, inasmuch as the ALJ's conclusion was reached after hearing all the testimony and evaluating the credibility of the witnesses. Thus, while I agree with the legal conclusion that Robinson lacked actual or apparent authority, I fundamentally disagree that that conclusion is inconsistent with a finding of good faith on Ragan's part. The inescapable upshot of the ALJ's uncontradicted finding of Ragan's good faith is, it seems to me, that Robinson succeeded in bilking Ragan as well as Sansom.

After noting the "inconsistency" in the ALJ's opinion, the Commission did not more than mention in a footnote that Ragan's "unreasonable" conduct amounted to willful misconduct sufficient to satisfy the scienter requirement of section 4b. See Commission Opinion, [1986-87 Transfer Binder] Comm.Fut.L.Rep. (CCH) p 23,796, 34,108 n. 10. The Commission supplied no reasoning to support this conclusion; it did not deign to detail the record evidence supporting its conclusion that Ragan acted willfully. In short, all the Commission did was reject the ALJ's conclusion that Sansom was estopped from repudiating the trades and set forth its conclusion that Ragan violated section 4b. Unlike the court, which has undertaken the task of supplying reasoning and citing evidence in the record, the Commission acted in an utterly conclusory fashion.

Thus it is that I have no quarrel with my colleagues' statements of law. If the Commission's conclusions are supported by the weight of evidence, we are bound to uphold them. It may well be that the Commission's conclusions are supportable; but it is emphatically not the court's task to weigh the evidence for the Commission. Here, where the weight of the evidence is far from clear (and the ALJ's conclusions were rejected by the Commission) the court ought not to supply the Commission's reasoning for it. As we "review the record with the purpose of determining whether the finder of fact was justified, i.e. acted reasonably, in concluding that the evidence ... supported his findings," Great Western Food Distribs. v. Brannan, 201 F.2d 476, 479-80 (7th Cir.), cert. denied, 345 U.S. 997, 73 S. Ct. 1140, 97 L. Ed. 1404 (1953), we cannot be at all confident that the Commission's decision warrants affirmance. There is abundant evidence that Sansom and Loughridge acted negligently, even recklessly; and there is, to be sure, evidence showing that Ragan acted negligently (although not recklessly). In this uncomfortable state of affairs, with the evidence conflicting and the Commission failing to supply reasoning to support reversal of its ALJ, a remand to the CFTC for reconsideration is plainly in order.

The court ably recounts the evidence supporting the Commission's decision. The court fails, however, to note the Commission's error in viewing the ALJ's position as incoherent; indeed, the court succumbs to the same error. Moreover, the court fails to evaluate the entirety of the record, which brims with evidence casting the most elemental doubts as to whether the Commission's conclusions are adequately supported. We, of course, have no idea what the course of the Commission's weighing of the evidence was; we have only its broad statement of a conclusion. Here are the salient facts that, in my judgment, rob the Commission's ill-conceived reversal of its ALJ of the requisite evidentiary support needed to pass judicial muster.

To begin at the beginning, the ill-fated relationship between Drexel Burnham and Sansom began, after an initial contact in Philadelphia, in Drexel's offices in Houston. There, on the Texas coastal plain, two Philadelphians sat down in the office of a financial wizard of sorts, the now-bankrupt Ragan. The sophisticates from the East Coast provided a study in contrasts: Sansom's president, Loughridge, was a successful Philadelphia businessman, anxious to avoid paying the costs of civilization by entering into highly creative tax shelter schemes. His companion, Robinson, was a tax lawyer by profession, but a lawyer who had descended to the depths of unethical (indeed, felonious) conduct resulting in his involuntarily leaving the practice of law. Robinson, it seems, was a gambler. Loughridge, who frequented the same country club and was old friends with the gambling ex-lawyer, hired Robinson when the latter found himself disbarred by the Commonwealth of Pennsylvania, and, sadly, out on the street. Robinson began serving Sansom in mainly clerical ways, but later became (by virtue of his training and expertise) involved in exploring creative ways to shelter Sansom's recent upsurge in income. It was in the latter capacity that Robinson joined in the meeting deep in the heart of Texas with Ragan, a scholar turned broker, trained in philosophy but who had turned his analytical abilities to more materialistic matters than the life of the mind.

The evidence is undisputed that Robinson called the Sansomian shots at the Houston gathering. Loughridge, who was gracious enough not to inform Drexel of Robinson's distinctly shady, embarrassingly recent past, annointed Robinson as Sansom's strategic architect in the mysteries of tax shelters. Loughridge looked on silently while Robinson discussed the ins and outs of tax shelters with the obviously skilled Ragan.

The upshot was a contractual relationship between Drexel, acting through Ragan, and Sansom. In short order, transactions designed to effectuate Sansom's income-sheltering goals were undertaken. In the early stages of this relationship, Ragan tried on several occasions to reach the aloof Loughridge by phone to report on his trading activities. Loughridge would not deign to involve himself in such pedestrian matters, however, and informed Ragan that the Houstonian's contact in Philadelphia was Robinson. Robinson was, in short, Drexel's principal contact at Sansom. Ragan politely complied with Loughridge's hands-off wishes, but dutifully dispatched written confirmations and other records of the account to Sansom, addressed to Loughridge's personal attention. The record shows that Loughridge, for whatever reason, was unaware of the pork belly trades; but the record likewise shows that Ragan had no particular reason to think that Loughridge was not receiving the statements.

And so the relationship ultimately came to grief, for reasons canvassed in the court's opinion. But the course of conduct engaged in by Sansom, through Loughridge, is remarkable for its invitation to disaster. Loughridge would not be bothered after he put the fox in charge of the chickenhouse. He chose, on the contrary, not to check from time to time on the chickens. He was even too busy to open the envelopes that Drexel sent to him. Those went, like Ragan's telephone calls, to Robinson. Indeed, Loughridge's trust in Robinson, his gambler pal from the country club, was apparently too deep to alert Drexel or the philosophical Ragan to keep a wary eye on Robinson's conduct. Under these circumstances, it is hardly surprising that the ALJ upon hearing the testimony concluded that Ragan's belief in Robinson's authority was "with some justification."

The Commission found glaring fault with Ragan's failure at the June 19, 1980 meeting in Philadelphia to alert Loughridge to the first pork belly trades. This, with all respect, is poppycock. At that juncture, there were no danger signs on the horizon; my colleagues, however, fail to note this in their rendition of the facts. Robinson (surprisingly, in light of his later abysmal performance in the pork belly markets) managed to make money on the early trades. So why would Ragan try to "hide" the fact that Sansom's in-house gambler had expanded Sansom's investment strategy to include pork belly speculation when the new effort was profitable? No evidence whatever has been adduced, and no reasonable inference has been drawn, to suggest that Ragan was somehow hiding the happy fact that the pork belly trades had made money for the company. There was nothing to hide, and no reason to reduce the conversation to the level of reporting on a few bucks eked out in the gritty market of pork bellies. Remarkable conclusions are being drawn from Ragan's failure to toot his own horn; not only could Ragan shelter income, but he could pick up some extra spending money for Sansom on the side by playing the pork belly market.*

The most damning aspect of Ragan's conduct, it would appear, was his failure to communicate directly with Loughridge when the freewheeling investment approach championed by Robinson began to turn sour. This is indeed misconduct, as Drexel almost brings itself to concede. The critical issue, however, is whether this conduct rose above the level of negligence to the much more demanding level of recklessness. The answer, in a word, is no. The totality of the circumstances is overwhelmingly to the effect that Ragan's course of conduct amounted only to negligence. As noted above, Loughridge was too busy, or too disinterested, to be bothered with the details of the Drexel account. After expressly being rebuffed by Loughridge in trying to arrest his attention on Drexelian matters, Ragan obeyed (negligently) his client's command; he bowed to Loughridge's cold directive to involve only Robinson in trading matters.

Ragan's negligence did not, however, rise to recklessness. After all, Ragan sought to verify the bona fides of Robinson's representation about a surplus in the Bache account (which checked out); moreover, the calls for funds to Sansom were promptly and fully honored without a peep. It simply will not do for the Commission to upbraid Ragan for failing to place an humble telephone call of notification to Loughridge when Ragan would reasonably have assumed that the call would be transferred to Robinson. As far as Drexel was concerned, all roads at Sansom lead to Robinson. Now that the piper has to be paid, the Commission has overturned its own ALJ and in a remarkable view of the record has said, in effect, that Drexel should have assaulted Sansom with facts that Loughridge went to some lengths to avoid knowing. This result, one might well conclude, is fundamentally unfair as well as remarkably undemanding in its standard of what rises to the level of reckless conduct.

For the foregoing reasons, I respectfully dissent from that portion of today's result that vindicates the Commission's ill-conceived, ill-reasoned reversal of its Administrative Law Judge. It is the job of the Commission, and not the court, to supply reasoning to support the agency's conclusions. Although the record contains some items of evidence that might be argued to support the Commission's conclusion of recklessness, it is a significant stretch to say that the weight of the evidence supports the Commission; and that, as the majority states, is the standard of review governing this case. Ragan's undisputed good faith means that he was Robinson's victim just as much as Sansom was.