Petitioners petition for review of a final order of the Commodity Futures Trading Commission affirming violations of the Commodities and Exchange Act and assessing a civil monetary penalty of $2.375 million. We AFFIRM the finding of liability, but find that the civil monetary penalty imposed was not reasonable in light of the violations at issue and that in assessing the penalty, mitigating evidence was improperly excluded from consideration. We REVERSE the order imposing a civil penalty and REMAND for a new assessment consistent with this opinion.

From April 1993 to March 1996, the petitioners, R&W Technical Services, Ltd., sold computer software to individuals interested in trading commodity futures contracts. This software required users to provide a source of real-time financial data which the software analyzed each day. Based on preset formulas and the real-time data, the software made buy and sell recommendations that the user was advised to act upon at the open of trading the next day. In order to help sell the software, the petitioners' advertisements included claims of enormous profits made during their seven years of trading with this system.

The advertisements characterized these results as "certified." The advertisements claimed that one reason R&W was selling this proprietary software was to generate more capital which they could invest with their system. Some advertisements stated that the software was only available in "limited quantities" and encouraged buyers to act quickly. The petitioners sold upwards of 1,000 copies of their software for prices of approximately $2,500 per copy. The advertisements offered to refund the purchase price plus 10 percent if the user did not show a profit after a year of using the system; apparently, only 11 customers ever requested a refund.

What the petitioners neglected to mention in their advertisements is that they never tested their system by making any trades in actual markets with real money. Instead, all of their performance data came from "paper" trades. In other words, the petitioners ran their system on real-time data but only pretended to perform the trades which their program recommended. The petitioners kept track of these virtual gains and losses and then presented the results as having been obtained with real cash.

On March 19, 1996, the Commodity Futures Trading Commission filed a four-count administrative complaint against the petitioners and its constituents, Gregory M. Reagan and Marshall L. Worsham individually, alleging violations of the Commodity and Exchange Act (CEA).1 Count I alleged fraud in the solicitation of customers in violation of CEA 4b(a)(i) and (iii). Count II alleged that the sellers were acting as Commodity Trading Advisors (CTAs) without being registered, in violation of CEA 4m(1). Count III alleged fraudulent sales practices and fraudulent advertising as CTAs in violation of CEA 4o(1) and Commission Rule 4.41(a). Count IV charged the petitioners with failure to produce required records and books. Reagan and Worsham also were charged as aiders and abettors and controlling persons for R&W's violations. On December 1, 1997, an Administrative Law Judge found R&W and Reagan2 liable on all counts and imposed a civil penalty of $7.125 million.

On May 16, 1998, the petitioners timely moved to reopen the hearing for evidence of customer satisfaction which might mitigate the penalty. On March 16, 1999, after de novo review, the Commission affirmed in part and reversed in part the ALJ decision. The Commission affirmedviolations under Count I (fraudulent solicitation) and Count III (fraudulent advertising). The Commission declined to reach Count II (failure to register as a CTA) or Count IV (record keeping). The Commission denied the request to reopen the hearing for mitigating evidence, ordered the sellers to cease and desist from their violations, and imposed a penalty, jointly and severally, of $2.375 million. The sellers then petitioned for review of the Commission's final order.

The petitioners contend that there is insufficient evidence to support a finding of materiality, an element of fraudulent solicitation under CEA 4b(a)(i) and (iii).3 Whether a misrepresentation is material is "a mixed question of law and fact," involving the "application of a legal standard to a particular set of facts."4 A deferential standard applies to questions of law encompassed by the agency's expertise so long as the agency's conclusion is reasonable.5 However, "'[w]hen the question is of the sort that courts commonly encounter, de novo review is proper.'"6 Because materiality in allegedly fraudulent transactions is a question that courts often encounter, de novo review is proper at least insofar as the application of the law to the Commission's findings of facts, which are conclusive "if supported by the weight of evidence."7

In this case, the petitioners misrepresented hypothetical trading results as real trading results. They sold software proclaiming that high profits had been obtained through actual trading over a period of years. These results supposedly had been "certified." In fact, however, no actual trades were ever made with their system. Instead, all results were simulated, and the petitioners risked no money in testing their system.

A statement or omitted fact is "material" if there is a substantial likelihood that a reasonable investor would consider the information important in making a decision to invest.8 We have little hesitation in saying that a reasonable investor would regard as material the fact that the petitioners' trading system "had never been tested through actual trading."9 Specifically, according to the Commission's expert, "hypothetical trading results have many inherent limitations." For example, such results assume a customer can execute his trade at the opening price. They also ignore the ability of a customer to steadfastly adhere to a particular trading scheme even when confronted with an initial series of losses.

The petitioners attempt to characterize their method as not hypothetical or unreliable because their results were based on real-time data and not based on the benefit of hindsight. In the end, however, the problem with reporting hypothetical trading results as real is that it allows an unscrupulous seller to test an arbitrarily large number of potential investment systems at little cost and then merely market the ones that happen to do best. Even if each system tested used real-time rather than historical data, the choice of which system to market introduces a hindsight bias.

If real money is used in testing, a seller cannot afford to indulge in such cherry-picking, which means investors can have more confidence in the seller's claims. There is no allegation that R&W tested multiple systems and only marketed the best, but this rationale nevertheless explains why any reasonable investor would be skeptical when hypothetical results are portrayed as real.10

The petitioners' advertisements proclaimed that the money made through their software sales was plowed back into real trades using their system. This also was a misrepresentation. The Commission found that the use of a trading system by its developers is a sign of authenticity, which reasonably increases consumer confidence in buying and using the system. The petitioners contend that these representations were of no additional benefit to consumer confidence, given that the petitioners already offered a money-back guarantee and had placed their corporate logo on the product.

Actual futures trading, however, creates exposure to substantial risk. A claim that one trades pursuant to the system one sells clearly expresses a higher level of confidence than merely putting a corporate logo on the product and offering a full refund. The petitioners also misrepresented the risks of futures trading by making bold predictions of high profits. Such claims "amount[] to the type of guarantee of profit prohibited under Section 4b of the Act."11 The petitioners respond that their guarantee of profits was backed only by a refund policy, which clearly established the only risks they insured against. However, the existence of a limited refund policy coupled with extravagant claims of false profits only confirms that the petitioners misrepresented the existence of the substantial risks inherent in futures trading.12

The petitioners argue that the Commission presented little evidence that actual trading would have been very different from real trading. However, even accepting the petitioners' characterization of the evidence as true does not change the fact that their statements were fraudulent and would have made a difference to a reasonable investor, even if the difference in practice did not produce catastrophic losses. As the Commission reasoned, a claim of "actual trading can convey to a customer that 'these results have been achieved,' [whereas] the petitioners' method can only convey that 'these results might have been possible.'"13

Because simulated results inherently overstate the reliability and validity of an investment system, and because extravagant claims understate the inherent risks in commodities trading, a reasonable investor would find the petitioners' fraudulent misrepresentations to be material.14

CEA 4b(a) prohibits any person from defrauding another person "in or in connection with" a commodity futures contract.The Commission interprets "in connection with" to reach the petitioners' conduct, while the petitioners dispute that any alleged misrepresentations were made "in connection with" any commodity futures contracts. To resolve this dispute, we first must determine the proper standard of review.

To the extent that a legal question involves the interpretation of the CEA, the Commission should normally be accorded a high level of deference,15 since the Commission is entrusted with administering the CEA through rules and regulations.16 Moreover, the Commission administers the CEA not only through its rulemaking authority, but also through its adjudicative power.17 As this court has stated, "[e]ven the adjudicative interpretations of policy-making agencies are entitled to Chevron deference."18

In the present case, however, Congress has given adjudicative authority not only to the Commission, but also to the federal district and appellate courts, both in terms of reviewing Commission orders and decisions and in terms of original jurisdiction over certain types of actions.19 However, the Commission remains the only governmental body entrusted with rulemaking authority in this area of law, and thus still remains the primary authority for interpretive policy decisions.20

Although the federal appellate courts have the power to review Commission orders and decisions, and although the district courts have been granted original jurisdiction in certain cases under the CEA, nothing in the CEA purports to lower the standard of review of Commission orders and decisions with respect to questions of statutory interpretation. Thus, because the phrase "in connection with" is a term of ambiguous scope, and because the Commission is the primary policy maker in this area, we find no reason to depart from regular Chevron deference, despite the fact that federal courts at some times may be called upon in an original action to interpret the CEA regarding a provision whose meaning has not yet been interpreted by the Commission in a rule or adjudication.

However, if there is no prior interpretation by the Commission, any question of Chevron deference becomes moot. Instead, interpretations advanced by the Commission during the litigation may be construed as offered for the purpose of "provid[ing] a convenient litigating position"; if so, the Commission's interpretation would not be entitled to deference.21

If a federal court answers a question of statutory interpretation before the Commission, then that court's interpretation may later conflict with an interpretation adopted by the Commission in a later rule or case. The resolution of such a conflict, however, is not before us, since the federal cases which have interpreted "in connection with" have not constricted the phrase any narrower than the Commission has interpreted it in its decisions. Thus, we must defer to the Commission's interpretation unless it is unreasonable.22

In order for a fraudulent statement to be "in connection with" a commodities future contract as required by 4b(a), the statement must first misrepresent the fundamental risk associated with such investments.23 Misrepresentations regarding the tax status of a commodities account are not actionable,24 nor are misrepresentations regarding commissions.25 The plain language of 4b(a) requires the misrepresentation to have some connection with the trading of commodity futures contracts. At issue is how tenuous that connection can be.

While it is clear that the petitioners' advertising claims misrepresented the fundamental risk associated with commodity futures investments and trading systems, the unusual aspect of this case is that the petitioners executed no trades for customers. They only sold software. In essence, the petitioners provided investment recommendations, but did not have any additional discretion to make trades on behalf of their customers. The petitioners generally profited only from the sale of software and not from the trading of their customers.26

In Saxe v. E.F. Hutton & Co.,27 another fraud case under the CEA, a broker solicited the opening of a discretionary account by misrepresenting the experience and skills of another broker who was to trade for the account.28 In the instant case, the petitioners' misrepresentations regarding the reliability of their system are analogous to the broker in Saxe misrepresenting another broker's track record. Unlike the broker in Saxe, however, the petitioners' customers only purchased recommendations from the petitioners. The petitioners had no authority to execute any trades on their customer's behalf, whether at the customer's request or at the petitioners' discretion. In one sense, then, the "connection" between the fraud and trades in Saxe is two levels closer than in the current case. The broker in Saxe had both the customer's money and the customer's permission to execute trades.

In Clayton Brokerage v. CFTC,29 however, the Eleventh Circuit found misrepresentations to be "in connection with" commodity trading regardless of whether the account was discretionary or traded only at the request of the customer. Clayton Brokerage is thus only one level removed from our case, in which the petitioners made recommendations but did not have any direct stake in the trading.

Moreover, although the petitioners did not profit from customer trading by receiving commissions, the petitioners did not give their advice away and necessarily expected their customers to make trades. This expensive software had no purpose except as a device for choosing which trades to make. In the end, then, the question is whether the phrase "in connection with" can reasonably be read to reach the petitioners' fraud.

Under the Commission's interpretation, fraud in the sale of investment advice willbe "in connection with" the sale of a commodities future contract if the fraud relates to the risk of the trading and the primary purpose of purchasing the advice is to execute trades. Under this interpretation, the fraud here can be understood as more similar to the fraud in Clayton Brokerage and Saxe than first appears. The only purpose for depositing money in a commodity account is to subsequently make trades, either on one's own initiative or at the recommendation or discretion of the broker. Similarly, no one spends several thousand dollars on a sophisticated software package without seriously intending to execute trades.

These scenarios contrast markedly with other sales of investment advice. For example, magazine and newspaper articles often dispense investment advice on a variety of topics. A person may buy a newspaper or magazine to read such articles with no intention to follow through on any of the recommendations. Such sources are invariably used not only for educational and research purposes, but also entertainment or leisure purposes. The same, however, cannot be said about petitioners' software. As the Commission noted below:

unlike the cases that have found the connection between the fraud and trade lacking, the misrepresentations in this case are neither incidental nor secondary to the futures trading but are directly related to that trading. In fact, the gravamen of the claim is that the respondents misled potential purchases of their system concerning trading profits and trading risks in order to induce customers to trade, and there is ample evidence to show that they did trade.30

There are additional reasons to construe 4b(a) broadly rather than narrowly. Originally, 4b(a) only applied to members of a contract market. In 1968 it was extended to reach "any person."31 In fact, "[t]he legislative history [of 4b(a)] indicates a progressive trend toward broader application of the CEA."32 In 1974, Congress gave the Commission even greater enforcement powers, in part because of the fear that unscrupulous individuals were encouraging amateurs to trade in the commodities markets through fraudulent advertising.33 Remedial statutes are to be construed liberally,34 and in an era of increasing individual participation in commodities markets, the need for such protection has not lessened.

The Commission's position in its earlier adjudication of this case is consistent with CFTC v. AVCO Financial Corp.35 and does not conflict with any earlier interpretations by the federal courts. The petitioners defrauded customers regarding the reliability of a system whose only intended use was as a means of selecting commodity futures contracts. To say that such fraud is "in connection with" commodity futures contracts is not unreasonable. Given the standard of review, we must affirm the Commission's findings of liability under 4b(a).

The petitioners were also charged with violating CEA 4o and Commission Rule 4.41(a), which prohibit Commodity Trading Advisors ("CTAs") from defrauding clientsand prospective clients through various schemes including false advertising. At issue is whether the petitioners were functioning as CTAs.

A CTA is defined as any person who "for compensation or profit, engages in the business of advising others, either directly or through publications, writings, or electronic media, as to the value of or the advisability of trading in [futures contracts]."36 Excluded from this definition is any "publisher or producer of any print or electronic data of general and regular dissemination, including its employees."37 However, the publisher exception only applies if the CTA's "furnishing of such services . . . is solely incidental to the conduct of their business."38 The Commission argues that "such services" refers to "any advisory services," while the petitioners argue that it refers only to "personalized advisory services."

The plain language of the statute, however, shows that the phrase refers to "any advisory services." Section 1a(5)(A)(i) defines a CTA as any person who "engages in the business of advising others, either directly or through publications, writings, or electronic media." Absent any other distinctions, the later reference to "such services" can only refer to both the direct and indirect provision of advisory services. This might appear to vitiate the entire publisher exception, since the publishing of such advice may often be the primary business of a publisher. However, the exclusion still protects incidental publishers of such advice, such as general magazines and newspapers, even if it does not exclude publishers who specifically concentrate on commodities and futures advice. To accept the petitioners' interpretation, however, would allow any large publishing company to offer highly personalized trading advice on the side - such as a 1-800 number trading hotline - without having to conform to the rules regulating commodity trading advisors, so long as the service was "incidental" to their regular impersonalized publishing activities. For these reasons, the Commission's interpretation is the correct one.39

Alternatively, we note that the petitioners' software publishing was neither "generally" nor "regularly" disseminated, as required in order to meet the publishing exception.40 In Lowe v. SEC,41 the Supreme Court dealt with an even broader publisher exception within the Investment Advisers Act of 1940. The Court defined regular dissemination to require that "there is no indication that [dissemination] ha[s] been timed to specific market activity."42 In this case, the petitioners' recommendations were provided by software that was programmed to "speak" only when certain market conditions were met. Thus, the petitioners' recommendationswere timed to particular market activity and not "regularly" disseminated. Moreover, a publication is only of "general" dissemination when it is circulated for sale to the general public at large in an open market.43 The record here indicates that the petitioners advertised that the software would only be sold in limited numbers. While that may have been a selling tactic, such claims cut against their argument that their software was "generally" disseminated.

The petitioners contend, however, that under Lowe, "impersonal" publishers of investment advice cannot be CTAs. In Lowe, the Court found that publishers of impersonal advice were not investment advisors under the Investment Advisers Act of 1940 (IAA),44 which defined investment advisor to include persons who advised others indirectly through publications.45 The Court found that the legislative intent behind the IAA was to regulate only the business of dispensing personalized advice and not to regulate impersonalized publishing activities; therefore, impersonal publishers were not included in the definition of "investment adviser."46

The petitioners in this case were impersonal publishers. The software they offered provided impersonal recommendations as to the buying and selling of commodities. Such recommendations were not based on any knowledge regarding the user's personal financial situation. Admittedly, this is a more sophisticated form of publishing than a weekly newsletter, but in substance it is the same as if the petitioners operated their proprietary software at home and faxed reports to their subscribers on a daily basis.

Just because Lowe found that the IAA excluded such publishers, however, does not entail that the CEA must. The statutes are different, and Lowe read the statute to avoid constitutional concerns. Had impersonal advisors been included in the definition of an investment advisor, there would have been an unconstitutional prior restraint in regulating them, since the publication of impersonal advice about specific investments is fully protected speech under the First Amendment.47

The Commission does not contest that such regulations would be a prior restraint, but instead argues that we need not consider that constitutional question because the validity of the CEA's registration requirements is not before us. The petitioners are no longer charged with registration violations, and even if the registration requirements are unconstitutional, the rest of the CEA would remain intact under a severability clause.48

Instead, the Commission argues that the CEA can define impersonal publishers as CTAs, even if the CEA cannot impose registration requirements upon them. Then, once the CEA has defined such advisors as CTAs, the CEA can impose liability on them for violations of the CEA's antifraud provisions, since liability for fraud would not run afoul of the First Amendment.49

In general, statutes should be construed so as to avoid constitutional questions.50 However, the constitutional question in this case has already been avoided because the registration requirements are not before us. We find, then, that the petitioners fall within the plain meaning of the definition of CTA and thus are subject to the CEA's antifraud provisions which apply to CTAs. However, we do note that lower courts have split on this issue.51

In order to apply 4o and Commission Rule 4.41 to the petitioners, their advertising scheme must have been used to defraud potential "clients."52 The petitioners argue that their customers were not clients because no personal relationship existed between the petitioners and their customers. While the CEA does not define "client," the petitioners urge that the common meaning of the term requires a personalized relationship.

In terms of Commission Rule 4.1, the Commission has previously interpreted it to apply to advertisements directed to mere subscribers of investment advice.53 The Commission's interpretation of its own rule would, of course, be "controlling . . . unless it is plainly erroneous or inconsistent with the regulation."54

The petitioners' argument regarding 4o, however, finds superficial support in cases such as Lowe, which distinguished between the impersonal publisher-subscriber relationship and "the investment adviser-client" relationship.55 Lowe's distinction was made based on the IAA Act and not the CEA. However, even the CEA makes a distinction between clients and subscribers. For example, 4o and Rule 4.41(c) apply only to clients, while Rule 4.33(a) specifically applies to clients and subscribers.

The petitioners, however, failed to raise their argument regarding the definition of "client" before the Commission and instead simply argued that 4o did not apply to them because they were not CTAs. We have held that "[a]s a general rule, in considering a petition for review from a final agency order, the courts will not consider questions of law which were neither presented to nor passed on by the agency."56 Because we find no compelling reason to address this question of interpretation for the first time on appeal, we do not decide the issue.

An agency's refusal to reopen the record is a procedural matter which is reviewable for an abuse of discretion.57 In the present case, the Commission refused to hear testimony that demonstrated widespread customer satisfaction with the petitioners' product.

The reason the petitioners did not give such evidence at the ALJ hearing was because both the ALJ and enforcement counsel had indicated that whether the software worked as advertised was not at issue. The ALJ even excluded the Division's customer witness testimony as irrelevant.Yet after the hearing, the ALJ made findings of fact based on an assumption that no trading scheme could work as the petitioners advertised.58 When the Commission then refused to hear evidence of customer satisfaction, the Commission stated that exculpatory evidence, including evidence of customer satisfaction, was not material because the Commission agreed that the efficacy of the system was not at issue.

However, evidence of the efficacy of the petitioners' system was relevant in assessing sanctions.59 The Commission now portrays the petitioners' failure to adduce evidence at the ALJ hearing as a tactical decision.60 Such a characterization is hard to square with the fact that the ALJ and Division stated that the efficacy of the system was not at issue. Further, we see no tactical advantage the petitioners might have gained by holding back evidence that their customers were satisfied and had made rather than lost money.

When the Commission conducted its review, its review was de novo. Thus, there was no need to exclude the defendant's mitigating evidence based on deference to prior factual findings. Because of the relevance of this mitigating evidence, and because the ALJ appears to have misled the petitioners as to the admissibility of this evidence, the Commission abused its discretion in refusing to reopen the record to hear this evidence, especially given that the ALJ originally imposed over a $ 7 million penalty based on a finding that no such trading system could provide a "trader with any significant market advantage."61

Sanctions are reviewed under the abuse of discretion standard.62 The standard is that the sanction must be rationally related to the offense.63 Given that the Commission should have considered evidence of the efficacy of the system and customer satisfaction, the penalty must be reassessed in light of such mitigating evidence. The improper exclusion of evidence aside, the original penalty was unreasonably excessive.

The penalty may be determined by focusing on the "relative gravity of . . . misconduct" in light of factors such as:

(1) the relationship of the violation at issue to the regulatory purposes of the Act; (2) respondent's state of mind; (3) the consequences flowing from the violative conduct; and (4) respondent's post-violation conduct.64

Furthermore, "[t]he level of sanctions should reflect 'the particular mitigating or aggravating circumstances presented by the unique facts of the individual conduct at issue.'"65 On occasion, penalties in similar cases have been "a guide to the appropriate level of a civil monetary penalty."66 However, an undue focus on past penalties fails to account for changes in policy orinflation, and thus may undermine deterrence.67 Federal courts have also rejected the notion that uniform sanctions must be imposed by an administrative agency for similar violations.68

Even if sanctions need not be precisely uniform, they must be rational, and neither inflation nor expressed policy changes can explain the magnitude of the penalty in this case. At oral argument, counsel for the Commission was unable to say that there had ever been a fine greater than $100,000 in a case in which there had been no demonstration of harm to others. One case allowed a civil penalty of $1.8 million, but only after a defendant committed extensive solicitation fraud, his customers suffered losses of more than $2 million, and the defendant violated a cease and desist order.69

In calculating a civil penalty, "the financial benefit that accrued to the respondent and/or the loss suffered by customers as a result of the wrongdoing are especially pertinent factors."70 In this case, the Commission had no evidence of customer losses and thus focused exclusively on the gain to the petitioners when approving a $2.735 million sanction based on estimated gross revenues. When a penalty is designed for deterrence and not restitution, however, the proper measure of gain to the defendant is net profits, not gross revenues.71

Thus, the lack of demonstrated harm in this case suggests that the petitioners' violations, while actionable, were not so egregious as to warrant a $2.375 million penalty. Presumably, the petitioners' system worked better than the one at issue in CFTC v. AVCO Financial Corp.72 In both AVCO and the present case, sellers portrayed hypothetical results as real results.73 The evidence of actual losses suffered by AVCO's customers shows why such misrepresentations would be material to reasonable investors.74 The lack of any demonstrated losses in this case makes plain that a $2.375 million penalty is capricious and indefensible. Even the Commission's witness, whose testimony demonstrated that the system did not perform as well in practice as in theory, earned a $60,000 profit in one year using the system and stated that no other system worked better.

The Commission has never imposed a penalty this large on any individual and has never imposed a penalty of even the same order of magnitude absent demonstrated harm to others. Thus, on remand a new assessment of the penalty should begin with the petitioner's net profits, which then should be adjusted lower based upon any mitigating evidence the petitioners present with regard to customer satisfaction.

AFFIRMED in part, REVERSED in part, and REMANDED.