UNITED STATES OF AMERICA

Before the

COMMODITY FUTURES TRADING COMMISSION

______________________________________

In the Matter of :

STEPHEN F. REDDY, NICHOLAS

DeSALVO, and JOHN W. SORKVIST

______________________________________

Respondents Stephen F. Reddy ("Reddy"), Nicholas DeSalvo ("DeSalvo"), and John W. Sorkvist ("Sorkvist") appeal from the initial decision finding them liable for numerous noncompetitive futures transactions in the Coffee, Sugar, and Cocoa Exchange's ("CSCE") sugar pit.⁽¹⁾ Respondents argue that the Administrative Law Judge ("ALJ") denied them a fair hearing, entered liability findings unsupported by the weight of the evidence, and imposed overly severe sanctions. In response, the Division of Enforcement ("Division") urges the Commission to uphold the initial decision in all respects. For the reasons that follow, the initial decision is affirmed.

BACKGROUND

On April 24, 1992, the Commission issued an eight-count complaint charging Reddy, DeSalvo, Sorkvist, and a fourth respondent, Gary R. Bergamo, with multiple violations of the Commodity Exchange Act, 7 U.S.C. �� 1-26 (1994) ("CEA" or "Act"), and the Commission's regulations. The allegations arose from the respondents' trading activities in the CSCE sugar pit from June through October 1988 and in March 1989.⁽²⁾ Respondents were active floor brokers at the CSCE, who filled customer orders and traded for their own accounts.⁽³⁾

The complaint alleged that the respondents, together with other named and unnamed individuals, indirectly "bucketed" customer orders and committed other related trade practice violations. The complaint alleged that these violations occurred in the course of 67 separate trade sequences. It identified 35 trade sequences in which Reddy participated, 32 trade sequences for DeSalvo, 19 for Sorkvist, and 14 for Bergamo.⁽⁴⁾

The complaint alleged that by virtue of their conduct, respondents, directly and/or by aiding and abetting others, cheated and defrauded customers in violation of Section 4b(A) of the Act; made false reports in connection with the execution of futures contracts, in violation of Section 4b(B); deceived customers in connection with the execution of their orders in violation of Section 4b(C); bucketed and filled customer orders by offset in violation of Section 4b(D); entered into wash, fictitious, and accommodation trades in violation of Section 4c(a)(A); caused non-bona fide prices to be reported in violation of Section 4c(a)(B) of the Act; and violated Commission Rule 1.38, 17 C.F.R. � 1.38 (1998), by engaging in noncompetitive trading.⁽⁵⁾

Further, the complaint charged that each respondent, pursuant to Section 13(a) of the Act, 7 U.S.C. � 13(c)(a) (1994), aided and abetted the violations of others. Additionally, the complaint alleged that Reddy and DeSalvo willfully entered or caused to be entered false records in violation of Section 4b(B) of the Act. Bergamo was charged with aiding and abetting Reddy and DeSalvo's Section 4b(B) violations, and with failing to keep and maintain trading records.⁽⁶⁾

The parties engaged in discovery and filed prehearing memoranda. Shortly before the hearing began, the ALJ issued an order declaring that the Division's proposed expert witness, Martha Kozlowski, was in fact an expert entitled to offer opinion testimony and foreclosed inquiry into her credentials or expertise at the hearing. See ALJ Order of Oct. 6, 1994. The ALJ issued this order in response to a motion by Reddy and DeSalvo for subpoenas directing DOE to turn over copies of any testimony Kozlowski had offered in other cases, as well as her notes in this case. Respondents also sought access to her CFTC personnel file and reviews of her job performance by her supervisors.

In earlier prehearing action, the respondents had moved to quash Kozlowski's verified statement on the ground that it contained opinions, conclusions, and conjecture about violations that had not been charged in the complaint.⁽⁷⁾ Respondents argued that such matters were inflammatory and exceeded the authority given to the Division by the Commission when it issued the complaint.⁽⁸⁾ In response, the Division argued that it was not seeking findings of violations for any misconduct not charged in the complaint. The ALJ denied respondents' motion to quash and ruled that, "[t]o the extent the expert testifies in terms of legal conclusions, the court will disregard her opinion." See ALJ Order of September 16, 1993.

Four days of hearings were held in New York on October 17-20, 1994. The Division's case was based principally on trading cards and other documents, as analyzed and explained by Kozlowski. Kozlowski focused on "pattern" evidence, i.e., the configurations of the allegedly violative trading sequences as they appeared on the respondents' trading cards.

Kozlowski, through her verified statement, testified that indirect bucket trades, e.g., where the broker illegally trades against his customer's order, usually display a characteristic manner of execution. Most simply, the broker who wishes to trade opposite a customer without appearing to do so fills the customer's order opposite an accommodating trader. According to Kozlowski:

[A] broker buys and sells the same or nearly the same quantity of the same contract month, at or about the same price, opposite the same trader. As to the broker's trades, one side is for a customer and the other side is for his personal account. As to the opposite trader, or the accommodator, both trades are for his personal account. The trades are usually, although not always, written within a few lines of each other on the respective trader's trading cards.

Kozlowski Verified Statement at 9.

For instance, the broker may buy ten contracts for a customer from a trader who is trading for his own account and then sell ten contracts from his own personal account to the same trader at or about the same price. The round-turn transaction effectively is a wash as to the accommodating trader, who buys and sells the same amount of a commodity at the same price. The transaction achieves its purpose by allowing the broker indirectly to trade against his customer, i.e., to "bucket" his customer's order.

According to Kozlowski, this trade configuration is unlikely to occur competitively by open outcry. Therefore, when the pattern appears repeatedly, involving the same people, "it is evidence that the trades were noncompetitive." Id. at 10.

Kozlowski identified various motives that might prompt a bucketing broker to trade noncompetitively opposite his customer: to obtain a better price for himself than he could get by open outcry; to trade for his own account ahead of a customer order and then use the customer's order to offset the broker's position at a predetermined profit; or to correct an error in filling an order. Id. at 12-13. According to Kozlowski, while the accommodating trader usually neither makes nor loses money on the non-competitive trades, local traders may accommodate such trades to encourage the broker to execute future customer orders opposite them. Id. at 15.

Kozlowski also addressed audit trail irregularities, i.e., anomalous entries or markings on the trading cards of the participants in allegedly violative trading sequences. Id. at 10-12. She testified that each trade sequence alleged in the complaint was accompanied by at least one audit trail irregularity, a circumstance that strengthened her conclusion that the trades were executed noncompetitively. She identified the two irregularities that appeared most frequently on trading cards as quantity changes and sequencing aberrations. Id. at 11.

Kozlowski stated that the quantity changes most frequently involved an increase in the number of contracts bought or sold for the broker's customer, with the amount of the increase equalling the number of contracts that the broker traded for his personal account. Id. For example, the trading cards of a broker and trader might show that the broker originally bought or sold six contracts for the broker's customer. Then both cards might show the six written over and changed to, perhaps, ten. Finally, in a separate entry, the cards would show a purchase or sale of four contracts for the broker's personal account opposite the trader. Kozlowski testified that "I do not believe it is a coincidence" that both parties to a trade made the identical quantity change, with the difference being "the same as the amount the broker traded for his own account," and that "the broker executed both the buy and sell opposite the same trader, usually at the same price." Id. Kozlowski testified that these irregularities indicated that the broker and the accommodator probably entered into a competitive trade, after which the broker requested the accommodator to increase the quantity by a certain number of contracts and "give him back" the same quantity. Id.

As to the sequencing anomalies, Kozlowski stated that, if trades are being executed competitively, one would expect to see both traders record the trades in the same sequence, as required by CSCE rules. If, for instance, a broker bought a quantity of contracts on the first side of a round-turn transaction and sold on the offsetting side, one would expect to see the purchase listed on the broker's trading card before the sale. Conversely, the opposite trader's trading card should show a sale followed by a purchase. Some of the trading cards examined by the Division's expert departed from this natural sequence, with both participants to a trade listing the buy first or the sell first. Id. at 12.

According to Kozlowski, this kind of anomaly can occur in a noncompetitive trade when the bucketing broker simply directs the accommodator to record a buy and a sell on his card without telling him the sequence. Alternatively, she suggested, since the trades may have been inserted after the fact, the traders needed to find an open line on their cards on which to record the trades. The available lines may not have allowed the trader to record the trades in a certain sequence. Id.

Kozlowski's overview of the respondents' trading was followed by a detailed analysis of each of the trading sequences alleged in the complaint. Id. at 16-176. Kozlowski was cross-examined by respondents' attorneys, who challenged her assumptions and conclusions.

In addition to Kozlowski, the Division called 11 CSCE floor brokers and traders and four CSCE employees. The nonparty floor brokers and local traders who participated in the trade sequences at issue were called to authenticate their trading cards. On questioning by respondents' attorneys, they denied that any violations had occurred and testified that they always traded opposite respondents (and others) competitively, by open outcry, in the ring. Several offered alternative explanations for the trading card irregularities identified by Kozlowski. The CSCE employees gave testimony describing the exchange's trade recording and clearing functions.

After the Division presented its case, the respondents testified and presented their own expert, Hugh Cadden, to counter Kozlowski's testimony. The respondents described the general trading conditions in the sugar pit at the relevant times and described their personal trading styles and manner of recording trades. All said they could not recall the specifics of the trade sequences at issue. (Tr. at 428 (Reddy), 542 (DeSalvo), 631 (Sorkvist).) They said that they executed several hundred trades a day and could not remember the details of trades from week to week, while these trades were six years old. All denied engaging in any noncompetitive trading at any time. They testified that, as busy and successful broker-traders, they had no incentive to cheat and that the need to maintain customer goodwill provided a disincentive to abuse of customer orders.

Respondents' expert Cadden examined the documentary evidence offered in support of the complaint's allegations of indirect bucket trading and concluded that "the subject trade sequences, individually and collectively, are just as likely to have occurred competitively as not." Cadden Verified Statement at 7. According to Cadden, the trading pattern that Kozlowski identified as characteristic of indirect bucket trades was equally consistent with "lawful, competitive dual trading." Id. at 8-9. Cadden stated:

It is perfectly normal, natural and legal for a dual trader to be selling for his own account when he is buying for a customer or vice-versa. . . . When a broker, who is in the process of selling contracts for his own account through numerous transactions, is handed a customer order to buy a small number of contracts, that broker would quite likely end up trading with the same traders with whom he has just traded . . . .

In other instances where a broker is working a large order, it is very useful for the broker to trade for his own account to facilitate execution.

Cadden Verified Statement at 9.

Cadden also stated that the wash trading results which Kozlowski found indicative of accommodation trading were equally consistent with scalping and some aspects of spread trading. Id. at 10-11. He also testified that the audit trail irregularities relied upon by Kozlowski were, in his opinion, not out of the ordinary, especially in light of the heavy trading and intense volatility in sugar futures during the relevant time periods. Id. at 11-16. Cadden reviewed each of the 66 trade sequences at issue and posited a nonviolative theory of what might have transpired. Id. at 24-262.

Respondents also presented John Tubman, a former director of compliance at the CSCE, who testified that the CSCE Compliance Department conducted its own investigation into the activities underlying the allegations in this complaint and found no violations.

Respondents, through their own testimony and that of the non-party exchange members, also challenged the Division's interpretation of the trading card irregularities. They contended that 1988-1989 was a time of high volume and intense price volatility in the CSCE sugar pit and that in consequence trading card errors (cross-outs and write-overs) were the norm and not the exception. They also noted that the quantity changes were consistent with the then-common practice of "blocking" trades. In an example of blocking, a trader who bought ten contracts from another at a certain price and then bought five more from the same opposite trader at the same price might change the ten to 15 instead of entering two separate transactions. Respondents said that in a busy market it was more efficient to block trades than to enter duplicative trade data. Respondents also contended that the sequencing errors generally stemmed from the volatile conditions in the pit.

Initial Decision. The ALJ issued his initial decision in the Division's favor on November 2, 1995, finding respondents liable in 60 of the 66 trading sequences charged. Initial Decision at 7-80. With the facts of the trading sequences undisputed, the ALJ confronted competing inferences drawn by the parties from the documentary record. He found the Division's analysis persuasive and rejected respondents' view of the trading sequences. "Respondents provided nothing more than loose and sundry explanations of what 'could have happened'," the ALJ ruled. Id. at 82. "Essentially, Respondents ask the Court to believe that the most uncanny combinations of customer orders, market conditions, trading card discrepancies, trading card changes, and personal profits to a Respondent or his opposite broker were merely flukes which arose from lawful market activity. The Court has rejected these explanations as they simply do not provide a credible explanation of the facts." Id.

The ALJ further found that, "in each of the transactions" there existed "a reasonable and likely motive for the alleged wrongdoing." Id. He held that as to each proven violation, "respondents either made a profit, enhanced their trading position, or assisted another in accomplishing the same." Id.

He concluded that Reddy was liable for violations in 35 trade sequences, DeSalvo in 27, Sorkvist in 16, and Bergamo in 12. Initial Decision at 83-86. The ALJ found the evidence insufficient to sustain the Division's burden of proof with respect to six trade sequences.⁽⁹⁾

As sanctions, the ALJ issued a cease and desist order against all respondents and revoked their registrations. In addition, he ordered personal trading bans of ten years for Reddy and DeSalvo and five years for Sorkvist and Bergamo. Finally, he imposed civil penalties of $300,000 each on Reddy and DeSalvo and $150,000 each on Sorkvist and Bergamo. Id. at 89-90.

The ALJ concluded his decision with a critique of the purported practices in the pit, stating in part that:

[T]his case epitomizes the extreme difficulty the Commission encounters in investigating and prosecuting floor trading practice cases. With regard to order execution and record keeping, the exchanges persist in operating within the constraints of 19th century technology. . . .

Dual trading, order execution by hand signals and shouts, scribbled floor tickets and trading cards, trade allocation, and the oft-exercised right to make purported corrections in the trade register and time and sales record, all provide a powerful temptation to those with a proclivity for making an unlawful profit at the expense of an unknown customer. . . .

Id. at 88-89.

DISCUSSION

Respondents raise substantive and procedural arguments on appeal. On the merits, they contend that the ALJ's liability findings are not supported by the weight of the evidence. They argue that there was no "pattern" of illegal trading and that the Division's purported audit trail "irregularities" were not proven. In addition, Sorkvist argues that the Division failed to meet its burden of proof with respect to his scienter. Respondents assert that the ALJ also erred in failing to address in his initial decision the testimony of the CSCE officials and nonparty floor brokers and traders.

Procedurally, respondents claim that the ALJ prejudicially restricted their right to cross-examine the Division's expert about her qualifications and methodology. They also renew the argument raised below that Kozlowski's verified statement should not have been admitted because it addressed misconduct not charged in the complaint. They further contend that the ALJ violated their due process rights by criticizing the open outcry system.

Finally, respondents contend that the ALJ's sanctions are excessive, overstating the gravity of the offenses found, when compared to the sanctions imposed in similar cases. They ask that the complaint be dismissed or that the sanctions imposed be substantially reduced.

The Division urges affirmance of the initial decision.

Procedural Issues. We turn first to respondents' contention that the ALJ abused his discretion, violated the Commission's Rules of Practice, and denied them due process by unduly restricting their cross-examination of Kozlowski, a Commission investigator and the Division's key witness. It is a frequent practice of the Division to use the investigator who develops an enforcement case as an expert witness. The Division maintains that the ALJ's decisions regarding Kozlowski's testimony were sound and well within his authority to assure the orderly conduct of the hearing.

The ALJ did not err in his management of respondents' cross-examination of Kozlowski at the hearing. "Cross-examination should be limited to the subject matter of the direct examination and matters affecting the credibility of the witness." Fed. R. Evid. 611(b). Thus, the right to cross-examine a witness does not mean that a party can do so in "whatever way, and to whatever extent" it desires. Douglas v. Owens, 50 F.3d 1226, 1230 (3d Cir. 1995). Rather, a party is guaranteed only "an opportunity for effective cross examination," and the trier of fact may properly exercise discretion to impose reasonable limits on the scope of cross-examination. Id.; see also Maatschappij v. A.O. Smith Corp., 590 F.2d 415, 421 (2d Cir. 1976); accord, In re Air Disaster at Lockerbie, Scotland on December 21, 1988, 37 F.3d 804, 825 (2d Cir. 1994). Our review of the hearing transcript convinces us that the ALJ did no more than attempt to hold counsel for both parties to that standard.

The ALJ told the attorney for Reddy and DeSalvo at the outset of his cross-examination of Kozlowski "to indicate what portion of her direct [testimony] you're examining her over." (Tr. at 302.) Despite this reminder, the attorney proceeded to ask six pages of questions about how Kozlowski had prepared her report and how many times she had visited the CSCE, see Tr. at 302-07, before the ALJ again admonished, "I'm going to direct you to go to her direct testimony . . . ." (Tr. at 307.) The attorney still persisted in asking whether Kozlowski had conducted certain studies. (Tr. at 308-12.) The ALJ did not, however, limit the cross-examination when the attorney finally began asking Kozlowski about her analysis of specific trade sequences. (Tr. at 312-22.)

In a non-jury trial in particular, there is little to be gained by belaboring the obvious or attempting to score debating points against the opposing party's witness. Also, and as the ALJ noted, many of the points the attorney raised with Kozlowski could have been developed more efficiently through the testimony of respondents' own expert.⁽¹⁰⁾ In contrast, the attorney for Sorkvist focused his cross-examination on Kozlowski's direct testimony and was allowed to proceed largely without interruption. See generally Tr. at 359-401.

Thus, we find that the ALJ did not abuse his discretion during the course of the hearing. The ALJ has broad discretion to determine the scope of expert testimony and to limit cross-examination to the issues raised in the direct testimony. See Maatschappij, 590 F.2d at 421 (protracted questioning of an expert witness on a fact that could be determined by the judge was grounds for limiting cross-examination to avoid "a waste of time"). Furthermore, even if the ALJ were deemed to have erred in seeking to restrict the cross-examination of Kozlowski, any error was harmless, since the attorney for Reddy and DeSalvo simply ignored most of the ALJ's repeated directives to limit his questioning. For instance, after issuing several warnings not to stray from Kozlowski's direct testimony, which went largely unheeded by counsel, the ALJ ultimately ordered the attorney to cite the page of Kozlowski's direct testimony before asking a question. (Tr. at 342.)⁽¹¹⁾ Still, when the attorney failed to comply, the ALJ did not strictly enforce his order. (See generally Tr. at 342-58.) Respondents' cross-examination of Kozlowski consumed three hours, filled 100 pages of transcript, and was, in our view, comprehensive. Respondents suffered no prejudice in defending this case as a consequence of the ALJ's manner of proceeding.⁽¹²⁾

We also find that Kozlowski's verified statement was properly admitted into evidence, despite its conclusions that Reddy and DeSalvo traded ahead of customers in several sequences, a form of violative trading not charged in the complaint. Under the nonbinding guidance of the Federal Rules of Evidence, proof of misconduct other than the charged offense is not considered extrinsic, and therefore prohibited, if it arose from the same transaction or series of transactions as the charged offense, if it was inextricably intertwined with the evidence regarding the charged offense, or if it is necessary to complete the story of the offense on trial. See Fed. R. Evid. 404(b).⁽¹³⁾ The challenged portions of Kozlowski's testimony refer to acts that were inextricably intertwined with the violations charged in the complaint and without which Kozlowski's testimony might have been incomplete and confusing. Kozlowski's verified statement was probative of the offenses charged, did not contain prohibited evidence of "other" offenses, and was properly admitted by the ALJ. Likewise, the ALJ's findings of fact, while in certain instances observing that trading ahead had occurred, did so only in the context of analysis of the charged transactions. There is no indication in the initial decision that any findings of liability were premised upon instances of trading ahead. We therefore find no error in either the admission of Kozlowski's verified statement into evidence or the ALJ's findings with respect thereto.

The Trading Sequences. Respondents argue that the Division's analysis of the trade sequences did not establish a "pattern" of violative trading and that, even if it did, the pattern was not shown to have occurred "repeatedly." They contend that the Division's assertion that an indirect bucket trade has a characteristic, readily identifiable pattern is not based on any objective quantitative or measurable analysis, but only upon opinion testimony without independent support. Emphasizing this point, they note that the trades identified by the Division as characteristic of indirect bucketing are equally characteristic of lawful dual trading. The Division contends that its evidence supports the pattern identified by its expert.

An analysis of trading patterns and audit trail evidence may be the basis for establishing noncompetitive trading. See In re Rousso, [Current Transfer Binder] Comm. Fut. L. Rep. (CCH) � 27,166 at 45,308 (CFTC July 29, 1997), and cases cited therein. To prevail in a trade practice case where, typically, undisputed facts are subject to competing inferences, the Division, as the party with the burden of proof, must establish that its version of the facts is "more probable" than respondents' version. In re Rosenberg, [1990-1992 Transfer Binder] Comm. Fut. L. Rep. (CCH)

� 24,992 at 37,643 (CFTC Jan. 25, 1991). The Division has done so in this case. Respondents' version of events, in addition to being highly speculative, is inherently implausible.

A comparison of the two experts' analyses reveals why, on this record, the Division's theory must prevail. The Division's view of the case examines the roles played by the bucketing broker and the accommodating trader and the resulting trading patterns and explains them plausibly and convincingly as a single, unlawful, trading sequence. The accommodator's role enables the bucketing broker to trade opposite the broker's customer without appearing to do so.

Respondents, on the other hand, contend that each party to a trading sequence acted independently and traded competitively and that each, coincidentally, achieved trading results that, taken together, create a trading pattern consistent with the Division's theory. This confluence of results on both sides of the trade occurs repeatedly and is difficult to explain by pure coincidence. In respondents' view, even though the trading sequences when examined as a whole exhibit equivocal, possibly unlawful results for both parties, it nevertheless should be inferred that both parties to the trading were on each occasion acting lawfully and independently of each other. Respondents infer that the broker in each sequence was engaged in legitimate dual trading, not in bucketing a customer order, and that the trader was intentionally flattening a position to avoid a loss, not accommodating the broker. Respondents ask too much, building inference on inference to the point of implausibility. This, no doubt, is what the ALJ had in mind in observing that "[r]espondents ask the Court to believe that the most uncanny combinations of customer orders, market conditions, trading card discrepancies, trading card changes, and personal profits to a Respondent or his opposite broker were merely flukes which arose from lawful market activity." Initial Decision at 82.

In an example of how respondents' theory stretches plausibility to the breaking point, we note the six sequences in which Reddy, found to have been acting as a bucketing broker, traded opposite nonparty Paul Damante, described by the Division as acting as a trader/accommodator. Damante broke even or earned a profit on each sequence. In Sequence 7, for instance, Damante bought 8 October contracts at 14.33 opposite Reddy, who was trading for his own account and then sold 8 October contracts at 14.35 opposite Reddy, who this time was trading for a customer.

According to Cadden, the trades "have no relation to each other." Cadden Verified Statement at 50. Cadden opines that Damante's wash results reflect the fact that he was scalping the October sugar contract and that most of Damante's trades on the day in question involved one to 16 lots and were executed for "mostly a difference of two to five points, with a few exceptions." Id. at 52.

Cadden's interpretation is undercut by Kozlowski's observation that, of the six trade sequences involving Reddy and Damante, all have the same trading card irregularity--the above-described sequencing error. Both traders always recorded the buy side first. In his analysis of Sequence 8, Cadden asserts that Damante "records his buy side on his card first," a practice he describes as "generally true in a scalp." Id. at 57. He points to the other trading sequences involving Damante (Sequences 7, 9, 18, 23 and 55) to support his point. However, Cadden's contention runs counter to Damante's oral testimony. (Tr. at 27.)⁽¹⁴⁾

We also wish to emphasize the extent to which the respondents' theory depends upon an assumption that participants in the trading sequences who trade for their own account are risk-averse to a degree one would not expect in the inherently volatile futures markets. Under the Division's theory, the traders' quick round-turn wash results provide further evidence for inferring accommodation. Under respondents' alternative theory:

there is nothing abnormal if a local trader buys and sells for his own account at the same price. There are a number of reasons for this to occur. The most likely reason . . . is that the broker was attempting to scalp, was unsuccessful, and was quite pleased to take a pass-out [a break-even trade].

Cadden Verified Statement at 10. This general statement ignores the context in which the alleged violations in this case occurred. The round-turn wash trades were executed in unusually volatile markets, many during declared "fast" markets. Prices frequently moved up and down over broad ranges.⁽¹⁵⁾ The quick offsets at wash prices foreclosed the opportunity for gain as well as averting loss. Some of the trades at issue were executed early in the trading day, when no pressure existed to end the day flat and avoid margin fees. Furthermore, not all of the traders in the alleged sequences were scalpers unable or unwilling to establish and hold a position. In short, respondents' explanation for the frequent wash results lacks broad applicability and is an ineffective explanation for the pattern of trading and audit trail irregularities observed in this case.

For the foregoing reasons, we find that the Division's version of events regarding indirect bucket trades is "more probable" than that offered by respondents.

The foregoing discussion applies particularly to the great majority of violative trading sequences in which indirect bucketing was found.⁽¹⁶⁾ In three sequences exhibiting a related but not identical pattern, the accommodating trader helped a broker trade opposite the customer order of an affiliated broker. See Sequences 25, 40, and 60.

In three other trades, the ALJ found that Sorkvist intentionally lost money to others in the pit through illegal "money pass" transactions. In Sequence 16, he lost $672 to a non-party broker; in Sequence 21 he lost $4,480 to DeSalvo; and in Sequence 29 he lost $3,528 opposite DeSalvo, trading for himself and Reddy.

As for the remaining trades, in Sequence 2, Reddy bought 17 contracts for customers from Bergamo and then immediately sold Bergamo 17 contracts for Reddy's own account at a price one tick lower. The ALJ found that Reddy already had established a noncompetitive long position in Sequence 1, which both he and Bergamo exploited to their advantage in Sequence 2.⁽¹⁷⁾ In Sequence 30, Reddy lowered the execution price of 85 contracts sold for his customers. The price change benefited Bergamo, who was buying for his own account. In Sequence 39, DeSalvo offset separate customer orders to buy and sell opposite Bergamo, who was trading for his own account, at a price that allowed Bergamo to earn a $29,400 profit.

In these sequences, as in the indirect bucketing trades, the Division ties the various elements of each sequence into a coordinated scheme, while respondents again posit that the individual traders, acting independently, coincidentally produced trading results that appear suspicious, but are innocent. Thus, in these sequences as well, the Division's theory is more probable.

Respondents also argue that the audit trail irregularities used to bolster the pattern evidence were given undue weight. As noted above, in a number of trading sequences the trading cards of both a bucketing broker and an accommodating trader showed the same quantity change in a trading sequence. The second audit trail irregularity identified by the Division as occurring with some frequency involved round-turn trades being recorded in the same sequence by a bucketing broker and his accommodator, with both recording the buy side first or the sell side first in a round-turn transaction.

"[T]he probative value of an audit irregularity will vary with the Division's ability to suggest a plausible connection between the irregularity at issue and the type of arrangement underlying the Division's theory of the case." In re Gilchrist, [1990-1992 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 24,993 at 37,652 n.22 (CFTC Jan. 25, 1991). As described above, the Division, through Kozlowski, advanced plausible explanations for both forms of trading card irregularity. The explanation for the quantity increases in particular is tied integrally to the Division's theory of the case: the quantity changes in the transaction executed for the bucketing broker's customer match the amounts traded between the broker for his own account and the accommodating trader. The broker, having tested the market at his customer's risk in a competitive trade, then exploits it at no risk for his own benefit. The audit trail irregularities, occurring in conjunction with the basic violative trading pattern of indirect bucketing, provide further support for the Division's case.

* * *

Respondents also raise the scienter issue. Reddy and DeSalvo argue that "the testimony of Respondents and nonparty witnesses alike failed to prove either motivation or how noncompetitive trades might fit into an alleged arrangement." Reddy and DeSalvo App. Br. at 31.

Sorkvist argues that the ALJ made no specific findings as to his intent and that none of the ALJ's generic motive findings apply to him. The initial decision held Sorkvist liable as a money passer in three sequences and as an accommodator in 13 others. The two trading sequences in which he was charged as a bucketing broker were dismissed by the ALJ as unproved. Sorkvist asserts that no evidence exists that he intentionally accommodated bucketing brokers and that no motive was established for the three money pass trades.

The ALJ provided this analysis of all the respondents' states of mind:

It is the motive which strengthens the link between the trading card irregularities, trade pattern and market conditions, and weaves the element of scienter into the actions of each of the Respondents. The motive was clear in each of the proven violations set out in the findings for this case: respondents either made a profit, enhanced their trading position, or assisted another in accomplishing the same.

Initial Decision at 82.

In a trade practice case, beyond proof of the suspect conduct, the Division must establish that the respondent's participation in it was knowing. In re Buckwalter, [1990-1992 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 24,995 at 37,684-85 (CFTC Jan. 25, 1991); In re Bear Stearns & Company, [1990-1992 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 24,994 at 37,665-66 (CFTC Jan. 25, 1991); see also CFTC v. Savage, 611 F.2d 270, 284 (9th Cir. 1979). Such evidence must provide a "non-speculative basis" for inferring the respondent's knowledge--at the time of his participation--of the wrongful conduct, or it must negate the plausibility that the primary wrongdoers (here, bucketing brokers) could have achieved the results without the facilitating respondent's knowing cooperation. Bear Stearns,

� 24,994 at 37,365-66.

Like other elements of a trade practice violation, a respondent's state of mind may be inferred from circumstantial evidence. As the ALJ found, the circumstances surrounding the suspicious patterns rebutted any suggestion that the transactions could have occurred in the normal course of trading without collusion. Thus, in addition to inferring scienter from an available motive, the ALJ inferred willfulness as the only plausible explanation for the trading sequences and audit trail irregularities, frequently characterizing them in such terms as "uncanny" or "more than happenstance." See Initial Decision, passim.

Respondents, misapprehending this aspect of the initial decision, argue that the ALJ applied the wrong standard of proof. They maintain that, rather than the properly applicable "weight of the evidence" standard, the ALJ actually applied a much lower "more than mere coincidence" standard of proof.

As a threshold matter, we find that the ALJ articulated the proper standard of proof, see Initial Decision at 82, and as demonstrated by his detailed findings, correctly applied that standard. As we read the initial decision--and the record--the characterizations in question are addressed to the plausibility of the inferences to be drawn from the record. Those characterizations flow from the ALJ's analysis of the circumstances of individual trading sequences and reflect his conclusion that the totality of such circumstances did not arise by chance.

For instance, in Sequence 16, one of the three money pass trades, Sorkvist did not simply lose $672 to a nonparty trader; the $672 was exactly half the amount he had gained through an earlier transaction opposite the same trader. See Kozlowski Verified Statement at 60-62. Where the element of scienter is satisfied by proof of knowing conduct, the precise motive for each intended action becomes unnecessary (though we by no means minimize the evidence of motive in the record or findings of motive in the initial decision). We find that the scienter of all respondents is established on the record by the weight of the evidence.

Respondents also argue that the initial decision "ignor[ed]" the "uncontroverted" testimony of 11 floor brokers and local traders and four CSCE compliance officials. They argue that this testimony was exculpatory and should have been credited by the ALJ. Our review of the hearing transcript shows that the Division called the nonparty brokers and traders to authenticate their trading cards. Respondents then elicited testimony from them that Reddy and DeSalvo had never asked them to accommodate, to engage in bucketing, or any other non-competitive trading practices. The witnesses could not recall the specifics of the trade sequences at issue in the complaint.

Three of the CSCE officials limited their testimony to describing exchange procedures, such as how trading cards are processed, when cross-trading is permissible, and how time-and-sales reports are generated. A fourth CSCE official, who was called by respondents to establish that the CSCE had scrutinized the challenged trading sequences and found them competitive, was shown on cross-examination to lack a foundation for his factual testimony. (Tr. at 654-71.) In these circumstances, the ALJ was not required to make explicit findings as to the testimony of any of these witnesses. Moreover, since their testimony was limited or lacked foundation as described above, any failure of the ALJ to consider such testimony would have been harmless. In this de novo review we have considered their testimony in concluding that the weight of the evidence establishes the violations found by the ALJ.

The initial decision contains the ALJ's explicit and implicit credibility determinations of the various witnesses and their testimony. The foregoing analysis reflects our adoption of those determinations.

We turn finally to the ALJ's comments concerning the adequacy of the CSCE trading and recordkeeping system. His views do not establish extrajudicial or pervasive bias. A presiding official is permitted to take background knowledge gained on the bench into account in evaluating the evidence before him or her in a particular case.⁽¹⁸⁾ Moreover, our independent review of both the sanctions imposed and the record establishing liability should operate to allay any perception of prejudice to respondents arising from the ALJ's views.

Sanctions. The ALJ sanctioned the appealing respondents by issuing cease and desist orders, revoking their floor broker registrations, imposing trading bans (ten years for Reddy and DeSalvo; five years for Sorkvist), and assessing civil penalties ($300,000 each for Reddy and DeSalvo; $150,000 for Sorkvist). See Initial Decision at 89-90. Respondents challenge these sanctions as too severe, while the Division urges the Commission to affirm. The Division argues that the ALJ's sanctions are well within statutory limits and are fully justified by the seriousness of the violations.

The Commission reviews sanctions de novo. In re Grossfeld, [Current Transfer Binder] Comm. Fut. L. Rep. (CCH) � 26,921 at 44,467 (CFTC Dec. 10, 1996) ("[T]he focus of our analysis will be the appropriate sanction in light of the given facts and circumstances rather than the quality of the assessment made by the ALJ."). Having undertaken the review required by Grossfeld, we have determined that the sanctions imposed in the initial decision are appropriate and should be affirmed.

The fraud findings made against respondents in the initial decision and affirmed herein operate as a statutory disqualification from registration and as a presumption of unfitness for continued registration under Section 8a(2)(E) of the Act, 7 U.S.C. � 12a(2)(E) (1994). Respondents' misconduct supports the presumption of unfitness and, thus, the revocation of their registrations. See In re Sundheimer, [1980-1982 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 21,245 at 25,220 (CFTC Sept. 16, 1981), aff'd Sundheimer v. CFTC, 688 F.2d 150 (2d Cir. 1982). The misconduct was repetitive and occurred over a period of many months. It was systematic and a regular part of their trading activity. Also, because respondents' lawful and unlawful trading was so intertwined, their misconduct was difficult to detect and is therefore capable of repetition. See Rousso, � 27,166 at 45,310 and cases cited therein.

Any registrant found presumptively disqualified may produce mitigation and rehabilitation evidence to rebut the presumption. Reddy and DeSalvo have made no showing under either category.

Sorkvist asserts several factors to support his claims of mitigation and rehabilitation and argues that his evidence on this point was properly raised below but ignored by the ALJ. Sorkvist notes that he did not profit on the violative trades; that he cooperated with all phases of this proceeding; that he provided records to the Division and did not violate any recordkeeping requirements; that he has 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 given up all customer business and now trades only for his own account. Finally, he notes the age of the violations. Sorkvist App. Br. at 57-58.

We have considered Sorkvist's showing and find it inadequate to rebut the presumption of unfitness. That no profit accrued to him on the trade sequences at issue in this case is of little moment considering the fact that the record reveals him as a willing accommodator who readily assisted other brokers in abusing their customer orders and generally undermining the integrity of the market. In these circumstances, Sorkvist's decision to trade solely for his own account does not evidence rehabilitation or a changed direction, since his misconduct of record took place while he was trading for his own account. While we acknowledge the age of the misconduct, the passage of time, without more, does not establish rehabilitation. Sorkvist has failed to offer clear and convincing evidence of mitigation or rehabilitation as required by law. See generally In re Antonacci, [1990-1992 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 24,835 (CFTC April 18, 1990).

We turn now to the other sanctions. The same factors that we identified in support of respondents' unfitness for registration also warrant the cease and desist order, the civil monetary penalties, and the trading bans imposed by the ALJ. The amounts of the civil money penalties and the lengths of the trading bans are commensurate with the gravity of the violations found. That gravity is not measured in this case by the financial gain to respondents. The penalties reflect and seek to deter the betrayal of the public interest caused by respondents' abuse of customers and a regulated public market.⁽¹⁹⁾

The civil penalties are appropriate when considered in terms of gravity. Respondents have not requested a net worth hearing or proffered evidence regarding their net worth and therefore have waived the issue of whether the penalties are inappropriate in light of their resources.⁽²⁰⁾

CONCLUSION

For the foregoing reasons, and as modified herein, the initial decision is affirmed.⁽²¹⁾

IT IS SO ORDERED.⁽²²⁾

By the Commission (Chairperson BORN and Commissioners TULL, HOLUM, and SPEARS).

________________________________

Jean A. Webb

Secretary of the Commission

Commodity Futures Trading Commission

Dated: February 4, 1998

1. See In re Reddy, [1994-1996 Transfer Binder] Comm. Fut. L. Rep. (CCH) � 26,544 (ALJ Nov. 2, 1995).

2. All of the transactions at issue in this case involved the world sugar futures contract, Sugar #11, which is traded on the CSCE. A Sugar #11 contract consists of 112,000 pounds (50 long tons), and prices are quoted in cents and hundredths of a cent. The minimum price fluctuation is one one-hundredth of a cent, which is also one point. In dollar terms, one point equals $11.20 per contract. A one-cent move equals $1,120 per contract. The delivery months traded are January, March, May, July and October. Trading opens at 10:00 a.m. and continues until the completion of the closing call at 2:00 p.m. Trading is divided into eight half-hour periods, designated as brackets "B" through "J." The periods at issue in the complaint were times of record volume and intense price volatility in the Sugar #11 contract.

3. Reddy and DeSalvo executed customer orders for B & J

4 Of the sequences charged in the complaint, the first occurred on June 29, 1988; sequences 2 through 31 took place in July 1988; sequences 32 through 42 in August 1988; sequences 43 through 51 in September 1988; sequences 52 through 60 in October 1988; and sequences 61 through 67 in March 1989.

5. All references here are to Sections 4b and 4c of the Act as in force when the complaint was filed. Section 4b was renumbered and recodified by the Futures Trading Practices Act of 1992, Pub. L. No. 102-546, 106 Stat. 3590.

6. Bergamo filed a notice of appeal on November 21, 1995, but failed to perfect his appeal by filing a brief as required by Commission Rule 10.102(b), 17 C.F.R. ��10.102(b) (1997). Accordingly, the Commission dismissed Bergamo's appeal as unperfected. See Order of May 2, 1997.

7. Kozlowski's direct testimony was presented in written form (a verified statement), and she was made available at the hearing for cross-examination. See Commission Rule 10.66(d), 17 C.F.R. � 10.66(d) (1998). The ALJ followed the same procedure with respondents' expert, Hugh Cadden.

8. Respondents moved to quash Kozlowski's verified statement because it contained allegations that Reddy and DeSalvo traded ahead of their customers' orders in at least 12 trade sequences and executed non-competitive trades not charged in the complaint in another seven trade sequences. Respondents argued that they should not be required to defend against allegations not made in the complaint and urged the ALJ, at a minimum, to require the Division to amend the verified statement by deleting all testimony pertaining to these allegations.

9. The ALJ dismissed the complaint as to trade sequences 24, 32, 50, 64, 66, and 67.

10. For instance, at one point the attorney for Reddy and DeSalvo asked Kozlowski whether she had performed "a study of all the trading in the ring on any given time period of, let's say, a trading day, review every trading card of every broker who traded in the sugar 11 pit on that day." (Tr. at 309-10.) Had respondents deemed such a study important to forming an expert witness opinion, they could have had their own witness testify to performing such a study and describe its results. At one point the ALJ reminded the attorney that "you could attempt to prove a lot of what you're trying to do here through your own expert witness." (Tr. at 322.)

11. Kozlowski's verified statement was served on respondents more than a year before the hearing.

12. In addition, we find that respondents' prehearing requests to subpoena Kozlowski's employee evaluations and other personnel documents were unreasonable and vexatious and, as such, were properly rejected by the ALJ. See 17 C.F.R. � 10.68(a)(3) (1997) (requiring the ALJ to satisfy himself that a subpoena application is "not unreasonable, oppressive, excessive in scope or unduly burdensome.").

13. Cf. United States v. Aleman, 592 F.2d 881, 885 (5th Cir. 1979) ("the policies underlying the rule are simply inapplicable when some offenses committed in a single criminal episode become 'other acts' because the defendant is indicted for less than all of his actions"); accord, United States v. Weeks, 716 F.2d 830, 832 (11th Cir. 1983); United States v. DeLuna, 763 F.2d 897, 912-13 (8th Cir. 1985).

14. The following exchange took place at the hearing:

15. For instance, in Sequence 15, the price ranged 10 points over a two-and-a-half minute period, with a reported high of 13.75 and a reported low of 13.65. The nonparty accommodating trader sold at 13.75 and bought at 13.73. See Kozlowski Verified Statement at 56-59.

16. See sequences 1, 3-15, 17-20, 22-23, 26-28, 31, 33-34, 36-38, 41-49, 51-59, 61-63 and 65, in which the participating broker traded opposite his own customer. In several of these sequences, two affiliated brokers both traded opposite a customer order of one of the brokers.

17. Bergamo gained $190 on the trade. Reddy, whose sale to Bergamo offset part of an existing long position established at a lower price, gained $617.

18. See Coprue v. Nissan Motor Corp., No. Civ. A. 94-3557, reported at 1997 WL 68569 (E.D. La. Feb. 19, 1997), and cases cited therein (no recusal based on magistrate's statement that "'she had dealt with the Japanese before, that they were good at hiding the ball, and it (hiding the ball) was not going to happen in her court'").

19. Sorkvist invokes to no avail the damages rule of In re Gordon, [1992-1994 Transfer Binder] Comm. Fut. L. Rep. (CCH) ��25,667 at 40,182 (CFTC Mar. 16, 1993). Gordon held that, in determining the gravity of the violations and in order to deter fraud, the Commission will consider in establishing a civil monetary penalty the financial benefit to the respondent or the losses suffered by customers when that information is available. The Gordon case, however, involved fraudulent solicitation of retail customers, and the calculation of the civil penalty reflected those facts. These factors in gauging the gravity of the offenses are not readily transferable to trade practice cases, where violative conduct has the potential for threatening the integrity of the markets and the confidence of all those who rely on them for risk shifting and price discovery.

20. The conduct in this proceeding occurred well before the effective date of the Futures Trading Practices Act of 1992, pursuant to which net worth ceased to be a factor in penalty determinations. Therefore, in assessing civil penalty sanctions, former Section 6(d) of the Act governs, rather than current Section 6(e)(1). Gordon, ��25,667 at 40,182 n.5.

21. All issues and responses raised by the parties have been considered. Any issue not specifically addressed herein has been found lacking in merit and not worthy of extended discussion.

22. A motion to stay the effect of the decision pending reconsideration by the Commission or notice of appeal seeking review by the relevant United States Court of Appeals must be filed within 15 days of the date this order is served. See Section 6(c) of the Act, 7 U.S.C. � 9 (1994) and Commission Rule 10.106 17 C.F.R. � 10.106 (1998).