Steven F. Reddy and John W. Sorkvist petition for review of an order of the Commodity Futures Trading Commission which found them guilty of several violations of the Commodity Exchange Act (“CEA”), 7 U.S.C. §§ 1-26, and imposed various sanctions.
BACKGROUND
a) In re Reddy
In April 1992, the Enforcement Division of the Commission (“Division”) filed an eight-count complaint against petitioners Reddy and Sorkvist, along with two other traders, accusing them of multiple violations of the CEA and the Commission’s Rules in connection with their trading activities in the sugar pit of the Coffee, Sugar & Cocoa Exchange (“CSCE”). The complaint alleged that from June 29 through October 31, 1988, and during March 1989, Reddy and Sorkvist — -both “dual traders”
A broker buckets a customer’s order by trading opposite the order for the broker’s own account or for an account in which the broker has an interest. “Indirect bucketing” occurs when a broker, aided by an accommodating trader, trades opposite his own customer while appearing to trade opposite the aecommodator.
A wash trade is a transaction made • without an intent to take a genuine, bona fide position in the market, such as a simultaneous purchase and sale designed to negate each other so that there is no change in financial position. See Sundheimer,
On November 2,1995, the Administrative Law Judge (“ALJ”) found Reddy liable for CEA violations in 35 trade sequences and Sorkvist liable in 16 trade sequences. See In re Reddy, No. 92-19,
b) In re Mayer
In April 1992, petitioners SMayer, BMayer, SHB, MCC, and Gelbstein were, along with other traders, charged in a 23-count complaint of multiple recordkeeping and trade practice violations in connection
On May 15, 1996, the ALJ found petitioners liable on 22 of the 28 counts alleged in the complaint
Petitioners appealed the ALJ’s decision to the Commission. The Division did not seek an increase in the sanctions imposed by the ALJ. On February 3, 1998, the Commission issued an opinion and order (“Original Order”), in part affirming, vacating, and modifying the ALJ’s decision. In particular, the Commission sua sponte increased the sanctions imposed by the ALJ pursuant to a policy of reviewing sanctions de novo that was adopted several months after the ALJ’s decision and petitioners’ appeals to the Commission. See In re Grossfeld, No. 89-23,
Shortly thereafter, it was brought to the Commission’s attention that it had erroneously found BMayer and Gelbstein liable for fraud under Section 4b of the CEA even though the complaint contained no such charges.' The Commission vacated its Original Order and issued an amended opinion and order (“Amended Order”) on February 25, 1998, correcting the error and altering some of the sanctions it had imposed.
DISCUSSION
a) Liability
1) Standard of Review
The Commission’s liability findings are conclusive if supported by the weight — or preponderance — of the evidence. See 7 U.S.C. § 9; Haltmier v. CFTC,
2) Reddy and Sorkvist
With regard to Reddy and Sorkvist, the evidence was as follows. The Division’s expert, Martha Kozlowski, examined thousands of trading cards and identified transactions in which a broker had simultaneously bought and sold identical sugar futures contracts in identical, or nearly so, quantities at or about the same price opposite the same trader. In the transactions pertinent to Reddy and Sorkvist, the broker’s trades were both for a customer and for the broker’s personal account. The accommodating trader bought and sold only for his personal account.
Kozlowski testified that trading with those characteristics' — principally simultaneous trades of the same contracts, at the same price, in the same quantity, and between the same broker and trader — is unlikely to occur in a competitive open outcry. In her view, such trades exhibited the characteristics of indirect bucketing.
Petitioners countered this evidence with expert and other testimony that such trading configurations can just as plausibly involve lawful dual trading, in particular a practice known as “scalping.” t Scalping involves the buying and selling of the same contracts within a very short period of time on small market fluctuations, and sometimes, if the attempt to profit fails, buying and selling at the same price. See In re Collins, No. 77-15, Comm. Fut. L. Rep. (CCH) (C.F.T.C. Nov. 26, 1986) ¶ 22,401,
Moreover, the trade sequences described all were accompanied by “audit trail irregularities.” Petitioners were required to record the terms and times of trades on trading cards. The cards reflecting the trades described above showed two kinds of irregularities in particular. First, some cards showed sequencing irregularities; e.g., the broker and accommodating trader each recorded the first transaction ás a buy and the second as a sell, a clear misdescription of the transaction. Second, other cards showed that both the broker and trader subsequently altered the quantity first recorded by identical amounts. In the Commission’s not implausible view, these alterations reflected the number of contracts traded on the broker’s own behalf. Both the sequencing irregularities and the mutual alteration of quantities are, in the context of trades with the characteristics described above, strong evidence of artificial trading.
Given the evidence, we cannot overturn the Commission’s finding of liability. It is true that “the Division must do more than present suspicious circumstances raising the possibility of knowing wrongdoing,” In re Rosenberg, No. 80-29,
To be sure, the evidence here is circumstantial, but circumstantial evidence is often all that is available in cases involving artificial trading. See, e.g., In re Buckwalter, No. 80-28,
Sorkvist, whose only proven roles were as a money-passer or accommodating trader, argues that the Division failed to establish that he had a motive to participate in noncompetitive trading and thereby failed to prove scienter. He correctly argues that proof of participation in noncompetitive trades alone is not a sufficient basis for inferring knowing participation in trades. See Gilchrist,
It is true that the Division must prove intent to establish a violation of either Section 4b or 4c of the CEA. See CFTC v. Savage,
In holding Sorkvist liable, the Commission did not improperly rely solely “upon its theory of systematic fraud as a substitute for specific evidence that [Sorkvist] knew in what, capacity an opposite broker was trading.” Rosenberg,
Reddy also claims that he was denied an impartial trial because the ALJ was biased. He relies on a gratuitous critique by the ALJ of floor trading practices, the ALJ’s crediting of Kozlowski’s analysis over that of his own expert, and the ALJ’s alleged refusal to permit meaningful cross-examination by Reddy’s counsel. The claim is meritless.
Whether a trier is impartial depends upon whether there was either an extrajudicial source of bias or a bias that demonstrates “a deep-seated favoritism or antagonism that would make [a] fair judgment impossible.” Liteky v. United States,
The ALJ’s critique of trading practices exhibited his concerns about the commodity exchanges’ methods of operation. His comments were not inappropriate, let alone indicative of deep-seated antagonism toward petitioners. An ALJ’s credibility assessments, moreover, do not constitute bias. See NLRB v. Pittsburgh S.S. Co.,
Reddy finally contends that the Division’s delay in bringing and prosecuting the case violated his right to a speedy trial. According to Reddy, the long delay between the trades in question and the bringing of the complaint hindered his ability to defend himself because it made it “totally impossible for him to recollect, or for him to find any other witnesses who would recollect, the actual circumstances of the trades so as to provide direct evidence ... with respect to [his] culpability or lack thereof.”
Section 6(b) of the Administrative Procedure Act (“APA”) requires that an agency conclude proceedings “within a reasonable time.” 5 U.S.C. § 555(b). In determining reasonableness, we look to the source of delay — e.g., the complexity of the investigation as well as the extent to which the defendant participated in delaying the proceeding. See Public Citizen Health Research Group v. Commissioner,
In any event, petitioners suffered no prejudice. Reddy claims that, if the complaint had been filed sooner, he could have presented a better defense because he and other witnesses might have recalled the trades at issue, a purely speculative assertion. However, his argument is undermined by his prior assertions that he executed “at least a hundred thousand trades a year,” and that he “would have a hard time recollecting a trade [he] did last week.” Indeed, petitioners have, if anything, benefited from the delay, which has allowed them to continue trading and deferred their obligation to pay their civil monetary penalties.
3) SMayer et al.
SMayer, BMayer, SHB, and MCC challenge the Commission’s liability determina
A. SMayer, BMayer, SHB, and MCC
SMayer, BMayer, SHB, and MCC claim that the Commission erred as a matter of law by finding that the Schedule A trades constituted illegal wash .sales. These trades were between various accounts alleged by the division to be controlled jointly by SMayer and BMayer, and, if so, all gains were offset by all losses among the same parties. Petitioners argue that there is legally insufficient evidence of either common control of the accounts or identity of control of the trading. We disagree.
Although SMayer denied any ownership interest in MCC, he reported capital gains from MCC in an amount equal to that of his mother, the avowed owner, and the Commission quite plausibly found that he had such an interest. See Mayer,
Petitioners claim that even if the trades were wash sales, there was legally insufficient evidence that their participation in such transactions was knowing. This claim is entirely meritless.
Again, an agency may find the requisite intent based on circumstantial evidence. The Commission based its finding on several compelling factors. SMayer was in the heating oil pit daily while BMayer entered the heating oil pit only rarely and sometimes only long enough to execute a single trade. For example, the NYMEX streetbook shows that on more than one occasion, BMayer executed trades in both the heating oil pit and the platinum and palladium pit during the same minute. It is also highly significant that although BMayer rarely executed heating oil trades, a substantial percentage of those trades were with SMayer on the other side. Finally, the brothers on one occasion misused a certification procedure, suggesting that they were trying to cover up unlawful trading. See Mayer,
With regard to the Schedule B trades, there was evidence of trading patterns similar to that in Reddy. As the Commission noted, “[t]he NYMEX streetbook indicates that SMayer repeatedly traded, for the Mayer family accounts, the opposite sides of the same or similar quantity of the same commodity at the same price with the same individuals within very short time intervals.” Id. at *23. It further credited the testimony of several individuals who traded opposite SMayer in these trades. They testified that during the period in question they engaged in accommodation trading with SMayer, albeit being understandably unable to identify particular trades. Finally, petitioners failed to comply with the Commission’s reporting requirements, a failure that supports an inference that had the records been kept, they would have been unfavorable to petitioners’ defense. See Note 8, supra. Given our discussion of liability in Reddy, the Commission’s findings as to Schedule B trades are plainly reasonable.
Similarly, with respect to the Schedule C and D trades, involving allegations of bucketing, the evidence was clearly sufficient. Again there were trade sequences similar to those in Reddy, albeit in greater volume. Again, scalping or other lawful practices were proffered in defense. However, the Commission rejected those hypotheses, stating that “[g]enerally, when a broker or trader legitimately publishes a flat market, it would be expected that each side of the flat trade would be executed with different traders.” Mayer,
B. Gelbstein
Gelbstein first contends that his allegedly unlawful trades did not constitute a distinctive “pattern” sufficient to support an inference of artificial trading. Specifically, with respect to the trade sequences in Schedule B, he claims that three trades are too few to constitute a “pattern,” especially where, as here, they were executed months apart — in May, July, and December, 1987. As to the trade sequences charged in Schedules C and D, Gelbstein argues that at least 78 of the alleged trades did not constitute a “pattern” of volatile trading since 61 of the trades were not executed at the same time, 56 were not executed for the same quantity, and 18 were not executed at the same price.
To be sure, three trades may, without more, be insufficient to establish a “pattern” supporting an inference of artificial trading. However, the Commission found that the three trade sequences in Schedule B were indistinguishable from the trading patterns of witnesses who testified to having unlawfully accommodated SMayer. See Mayer,
Contrary to Gelbstein’s contention that the trade sequences alleged in Schedules C and D did not constitute a distinctive pattern, the evidence of pattern was powerful. These sale and repurchase transactions were virtually simultaneous and were executed at substantially the same price. “For each bucketed trade, Gelbstein executed a buy order and a sell order for the same commodity at the same time or shortly thereafter at the same or similar price with brokers who were acting on behalf of an SHB customer for one trade and a Mayer family account for the other.” Id. at *25. We therefore agree with the Commission that such “de minim-is differences in price, time, and quantity do not change the substantive nature of the wash trades evident in the schedules.” Id.
Gelbstein maintains that, even if the trade sequences in Schedules B-D constitute a distinctive and suspicious pattern of unlawful trading, it was not shown by a preponderance of the evidence that his participation in these trades was intentional. However, in its Amended Order, the Commission identified a number of independent factors — i.e., factors significant apart from the suspicious trading pattern — that supported its inference of intent. Specifically, it found that Gelbstein was at the center of the bucketed trades, that he was a longtime friend of SMayer, that he stood next to SMayer in the trading pit, and that he could not help but be aware of the pattern of trading that he was facilitating. See id. The Commission thus reasonably concluded that Gelbstein not only had the opportunity to accommodate SMayer, but also that his close relationship with SMayer gave him a motive to do so. Cf. United States v. Knowles,
Finally, Gelbstein argues that the Commission’s decision should be set aside because of the delay in prosecuting the case. We disagree for the reasons stated in Red-dy.
b) Sanctions
We agree with the Commission that the purpose of sanctions under the CEA is twofold: “to further the [CEA]’s remedial policies and to deter others in the industry from committing similar violations.” In re Miller, No. 92-4,
Both sets of petitioners argue that the Commission abused its discretion by failing to provide sufficient explanations of why it chose the sanctions it did. See Motor Vehicle Mfrs. Ass’n,
1) Choice of Sanctions
Petitioners in both cases argue that the Commission’s choice of sanctions constituted an abuse of discretion for two reasons. First, they assert that the Commission failed to explain adequately its choice of sanctions. Second, they claim that the sanctions imposed were too severe.
Before examining these arguments, we note that the transgressions here are extremely serious. Artificial trades, such as wash transactions or bucketing, have no purpose save to avoid some legal, statutory, regulatory, or contractual obligation or to manipulate the market. They are used, for example, to avoid margin requirements, see In re Three Eight Corp., No. 88-33,
We also note that, while artificial trading can over time be profitable, it is also difficult to detect. Because the gains available from artificial trading can be great and the danger of detection may seem low, the temptation to engage in such practices may be great. If deterrence is to be achieved, substantial penalties may be necessary.
Petitioners’ criticism of the Commission’s explanation of its choice of sanctions seems largely based on its failure to state explicitly the matters discussed in the prior two paragraphs. These propositions, however, are so well known in the area of the Commission’s jurisdiction that we see no reason to require the Commission to repeat them in every artificial trading case. These propositions can be found in Commission decisions, as described above, and that is sufficient to assure us that the Commission had them in mind. In criminal cases, sentencing courts do not have to explain in every case the various differences between jaywalking, embezzlement, and torture murders. Moreover, the Commission’s selection of sanctions involves judgments that cannot be accompanied by arithmetic exactitude or extended meaningful explication, in particular because they involve a judgment with regard to deterrence of other traders who might be tempted to engage in similar conduct. Such a judgment may be difficult to express in non-conclusory terms, as was so often the case in criminal sentencing before the Guidelines imposed, for better or worse, semi-mechanical rules based on the characteristics of the offense, criminal history of the defendant, etc., that allow somewhat extended discussion, albeit largely of the mechanical aspects.
What is required in the way of Commission explanation is not a ritualistic incantation of what is well-known in the area of the agency’s jurisdiction but simply an indication sufficiently discernible to allow us to exclude arbitrariness as the explanation for a sanction. See Armstrong,
A. Cease and Desist Orders
In deciding whether to impose a cease and desist order, the Commission must determine whether there is a reasonable likelihood that the misconduct will be repeated. See In re Fetchenhier, No. 91-12,
In Reddy, the Commission imposed a cease and desist order on the grounds that petitioners’ “misconduct was repetitive and ... difficult to detect and ... therefore capable of repetition.”
To be sure, the Commission did not explicitly determine that “there is a reasonable likelihood that the misconduct will be repeated,” only that such conduct is “capable of repetition.” We are not, however, “disposed to overturn a sound deci
B. Registration Revocations
A registration revocation is appropriate where the trader’s continued registration will pose a substantial risk to the market. See In re Gordon, No. 90-19,
Both Reddy and Sorkvist were subject to statutory disqualification and a presumption of unfitness for continued registration under Section 8a(2)(E) of the CEA, 7 U.S.C. § 12a(2)(E). Reddy, however, presented no mitigation or rehabilitation evidence to rebut this presumption. Thus, the Commission acted well within its discretion in concluding that his repetitive and systematic misconduct over a five-month period supported the presumption of unfitness. See Reddy,
Sorkvist, on the other hand, did attempt to make a showing of mitigation and rehabilitation to rebut the presumption of unfitness. He noted that he did not profit from the unlawful transactions, that he fully cooperated with all phases of the administrative proceeding, that he provided records to the Division and did not violate any recordkeeping requirements, that he had no prior or subsequent violations of the Act, that he participated in trades at prevailing prices at which other brokers also would have executed trades, and that he has since given up all customer business and now trades only for his own account. See id. The Commission nevertheless rejected his showing on the grounds that his failure to make a profit from his accommodation trades was insignificant in light of the fact that he was a willing aceommodator whose ready assistance to other brokers in bucketing their customer orders undermined the integrity of the market; that his decision to trade solely for his own account did not evidence rehabilitation because his misconduct took place while he was trading for his own account; and that the passage of time, without more, does not establish rehabilitation. See id. We find that the Commission’s rejection of Sorkvist’s rebuttal evidence was not unreasonable and that its explanation adequately supports the conclusion that Sorkvist poses a substantial
SMayer, BMayer, and SHB are also subject to statutory disqualification from registration, pursuant to Section 8a(3)(A) of the CEA, 7 U.S.C. § 12a(3)(A). The Commission concluded that the gravity of petitioners’ offenses justified the presumption of unfitness and determined that the pattern of illegal trading practices established a strong likelihood that the misconduct would be repeated. Revocation under these circumstances is appropriate. First, petitioners failed to "rebut the presumption of unfitness with any mitigation or rehabilitation evidence. Second, the sheer volume and gravity of their violations support the Commission’s conclusion that petitioners are likely to participate in future misconduct at substantial risk to the market.
C. Trading Bans
Trading bans are appropriate where there is a nexus between the violation and the integrity of the futures market. See Monieson,
With respect to Reddy and Sorkvist, the Commission justified the length of the trading bans — ten and five years respectively — on the fact that their misconduct was repetitive, systematic, and difficult to detect and therefore capable of repetition. The Commission concluded that “the lengths of the trading bans are commensurate with the gravity of the violations found,” which entailed “abuse of customers and a regulated public market.” Reddy,
All petitioners argue that the Commission failed to explain with sufficient specificity why it imposed trading bans and how it arrived at the duration of the trading bans for each petitioner. We disagree.
The Commission noted that the misconduct here was serious, repetitive, and affected the integrity of the market. Those findings obviously reflect the Commission’s quite reasonable judgment that petitioners sought to profit in known and unknown ways other than by buying low and selling high. Obligations to customers were flaunted, and the amount of misconduct was substantial, as was the time period over which it occurred. Moreover, the misconduct, as discussed earlier, went to the very integrity of the market’s pricing mechanism. Finally, as the Commission noted in Mayer, the misconduct in question is difficult to detect; and such misconduct, when detected, must be heavily punished if deterrence is to be achieved.
We therefore conclude that the trading bans here were adequately explained and reasonable. The trading bans were graduated according to degree of misconduct. Sorkvist’s ban of five years and SMayer’s, SHB’s, and MCC’s lifetime bans are pro
D. Civil Monetary Penalties
In support of its imposition of civil monetary penalties on Reddy and Sorkvist, the Commission held that the fines “are commensurate with the gravity of the violations found.”
Because the civil fines imposed in both cases fall within the statutory limit of $100,000 per violation, we review the amount of the penalties for a “rational relationship” to the offenses. Monieson,
2) Sua Sponte Increase of Sanctions
SMayer, BMayer, SHB, MCC, and Gelb-stein argue that the Commission abused its discretion by sua sponte increasing the sanctions imposed by the ALJ. Although they do not dispute the Commission’s power to review the imposition of sanctions de novo, petitioners challenge the application of the de novo standard to them on the ground that they perfected their appeal before the Commission announced its adoption of a de novo standard in Grossfeld,
Petitioners contend that the Commission should have applied the old standard of review to their case because they appealed the ALJ’s decision with the understanding that the Commission’s review of sanctions would be deferential. Petitioners alternatively argue that, even if the de novo standard was appropriately applied to their case, the Commission’s sua sponte increase of sanctions violated their due process rights because the Commission did not give them an opportunity to defend their position under the new standard of review. We find no merit in either claim.
As to the first claim, a change in law that “speak[s] to the power of the court rather than to the rights or obligations of the parties” may be applied to a case pending at the time of the adoption of the new rule without raising concerns that it is impermissibly retroactive. Landgraf v. USI Film Prods.,
Petitioners’ second argument is equally unpersuasive. In support of their claim that they were unconstitutionally deprived of the right to challenge the Commission’s increase in sanctions, petitioners rely principally on Baker,
Both Baker and Fetchenhier involved the Commission’s decision to change how it considered aggravating conduct in fashioning sanctions under Section 9(b) of the CEA. Because the earlier standard had been in effect at the time of the trial before the ALJ, the Commission concluded that the respondents in Baker and Fetchenhier might reasonably have perceived certain available evidence to be legally irrelevant and might not, therefore, have proffered or developed it at trial. The Commission therefore gave the parties an opportunity to introduce additional evidence. The decisions in Baker and Fetchenhier were not based on the Commission’s disagreement with the sanctions imposed by the ALJ, but on the fact that new, mitigating evidence might be forthcoming under the new standard. See Fetchenhier,
In the instant case, petitioners had every opportunity to proffer all available evidence relevant to the merits of sanctions. The sua sponte increase of sanctions did not render previously irrelevant evidence relevant or otherwise create a need for a remand.
CONCLUSION
The petitions for review are therefore denied.
Notes
. Pursuant to 28 U.S.C. § 46(b) and an order of the Chief Judge of this Court certifying a judicial emergency, this case was heard by a panel consisting of the Chief Judge of this Court and two judges of the United States District Court sitting by designation.
. "Dual traders” are "floor brokers” who trade for customers and for their own accounts. A floor broker is "any person who, in or surrounding any pit, ring, post, or other place provided by a contract market for the meeting of persons similarly engaged, purchases or sells for any other person any commodity for future delivery on or subject to the rules of any contract market.” 7 U.S.C. § la(8). A "floor trader” is "any person who, in or surrounding any pit, ring, post, or other place provided by a contract market for the meeting of persons similarly engaged, purchases or sells solely for such person's own account, any commodity for future delivery on or subject to the rules of any contract market.” 7 U.S.C. § la(9).
. An "aecommodator” is a trader who enters into noncompetitive trading to assist others with illegal trades. See Sundheimer v. CFTC,
. Specifically, the ALJ found that Reddy had violated Section 4b(A) of the CEA, which prohibits cheating and defrauding in connection with commodity futures contracts; Section 4b(B) of the CEA, which prohibits the entering of false records; Section 4b(C) of the CEA, which prohibits deception in connection with the execution of orders for futures contracts; Section 4b(D) of the CEA, which prohibits the bucketing of customer orders and the filling of customer orders by offset; Section 138(a) of the Commission Rules, 17 C.F.R. § 1.38(a), which prohibits noncompetitive trading; Section 4c(a)(A) of the CEA, which prohibits wash sales, fictitious sales, and accommodation trades; and Section 4c(a)(B) of the CEA, which prohibits the causing of prices that are not true or bona fide to be reported, registered, or recorded. It also found that Sorkvist had violated Sections 4b(A), 4b(C), 4b(D), 4c(a)(A), and 4c(a)(B) of the CEA as well as Section 1.38(a) of the Commission Rules.
. The ALJ found that SMayer had violated CEA §§ 4b(A), 4b(D), 4c(a)(A), 4c(a)(B), and 4g(l) and Commission Rules §§ 1.31(a), 1.35(d), 1.38(a), and 166.3; that SHB had violated CEA §§ 4b(A), 4b(D), 4c(a)(A), 4c(a)(B), and 4g(l) and Commission Rules §§ 1.38(a) and 166.3; that MCC had violated CEA §§ 4b(A), 4b(D), 4c(a)(A), and 4c(a)(B) and Commission Rule § 1.38(a); that BMayer had violated CEA §§ 4c(a)(A), 4c(a)(B), and 4g(l) and Commission Rules §§ 1.31(a), 1.35(d), and 1.38(a); and that Gelbstein had violated CEA §§ 4c(a)(A) and 4c(a)(B)' and Commission Rules §§ 1.31(a), 1.35(d), and 1.38(a).
. The Amended Order deleted the fraud findings against BMayer and Gelbstein and reduced the increased penalties against them from $250,000 to $150,000 each. However, it retained the increased trading bans. The Amended Order also reduced, without explanation, the number of alleged Schedule C and D violations by SHB and MCC from 571 to 394. It did not, however, reduce the sanctions against either SHB or MCC.
. Sorkvist argues that the Commission’s order should be overturned because it applied the wrong standard in determining that the Division met its burden of proof. He contends that, instead of a "more likely than not” standard, the Commission applied a "more than coincidental” test of liability. To be sure, in finding that petitioners’ explanation for the trading irregularities was implausible, the Commission noted that the trading pattern was "more than coincidental.” This of course says only that petitioners' theory is less probable than that of the Division. Thus, the "more likely than not” standard was properly applied.
. We note that petitioners’ claims of legally insufficient evidence or erroneous findings of fact are severely undercut by their failure to keep records as required by Commission regulations. They claim that the Commission rejected the ALJ’s factual finding that their failure to retain their trading cards was designed to eliminate audit trail evidence and that no adverse inference can be drawn from that failure. However, the Commission did not overturn the ALJ's finding of fact. It held only that the failure to keep records constitutes a per se violation of its recordkeeping requirements. It otherwise deferred to the ALJ’s credibility findings. Petitioners therefore bear the heavy burden of the adverse inference drawn from that failure.