concurring in part and dissenting in part:
My differences with Chief Judge Bauer’s incisive opinion for the majority are narrow and I write separately only to make a few essential points. The majority opinion deals admirably with the principles that *497should define fraud in arranged transactions on- the trading floor. As to the broad concepts, therefore, I have no quarrel with the majority, but in application a more precise analysis seems to be required.
The majority has correctly rejected the argument that trading in a mode other than open outcry is fraudulent per se (although, of course, it is a violation of the rules of the Chicago Board of Trade and of the Commodities Exchange Act). It is reasonably clear, however, that, when an arranged trade has generated a “surplus” to be passed on to another party, to be credited against past losses or to be retained for use against future losses, there is sufficient evidence to support a jury verdict of guilt. Nonetheless, the government bears the burden of proving beyond a reasonable doubt that each individual defendant specifically intended to perpetrate a fraud. The majority, I think, errs by failing to consider the specific evidence against each defendant.
As the majority opinion indicates, the trades involved here were not made at “scratch” and hence presumably involved some profit to the trader. In ordinary course, one may assume that a profit as such is not necessarily an indicator of fraud since in general some profit is required to cover transaction costs. The existence of transaction costs is the reason all trading, whether by open outcry, by matching or otherwise must, in the long run, generate a profit. But beyond the profit ordinarily generated by trading, there is in the ease of fraudulent transactions, what I would call an identifiable “surplus” available to be kicked back to compensate for losses or otherwise to be applied for the benefit of someone other than the customer. It is to the generation and application of such a surplus that I would look for evidence of fraud.
As to the mail fraud counts, I find the majority’s theory of culpability to be most clearly stated as follows: “By picking customer prices and opposing traders, the defendants removed their customers from the pit’s competitive marketplace and forced the customers to accept the results they selected, guaranteeing profits to the local and denying the customer the opportunity to obtain a better price.” Ante at 477-78. Although this general statement may accurately describe the operations of most of the defendants, the majority’s formulation fails to differentiate those trades, in which an identifiable surplus was applied for . a trader’s benefit, from trades where there is no evidence of such a surplus. Of course, in both circumstances there was a violation of CBOT rules and of the CEA; however, fraudulent intent may be inferred from the first-described circumstances but not from the second.
In this connection, the Government’s own witnesses testified, that customers placed MOO and MOC orders for purposes other than obtaining the most favorable price. For example, Charlotte Ohlmiller testified that an MOC order is typically made by customers “looking for a fill,” e.g., by customers desiring to have a sell order executed at the closing price, usually to close out their market position, but not by those seeking a particular (high) price. Tr. vol. 15, p. 65 (Ohlmiller direct); Tr. vol. .13, p. 294 (Ohlmiller, cross by Shine). See also Tr. vol. 21, p. 3243 (Eggum direct) (“Q: Mr. Eggum, when a customer puts in a request for an order to fill a market on close, is that customer specifying a price that he wishes? A: No. Q: What is the customer’s specified request on an MOC order? A: He’s specifying the time that he wants the order filled.”). Consequently, when executing MOC orders, traders are trying to carry out their customers’ instructions, usually to reduce the customers’ market exposure. It was this wish by customers to reduce risk that led some of the appellants to “trade on the curb,” i.e., after authorized trading hours. Some traders, however, including many of those convicted here, may have engaged in such trading to generate excess profits — as well as to serve their customers’ timing requirements. Therefore, curb trading on MOC and MOO orders can reasonably be explained on the basis of fraudulent as well as honest intent. Hence, the issue is what kind of evidence is sufficient to entitle a *498jury to reject honesty as an explanation and to accept fraud.
Many of the trades cited by the majority as examples of presumptively fraudulent transactions involve evidence of identifiable surpluses applied for the benefit of a party other than the customer. I quite agree that, as to these trades, the evidence is sufficient. But there are some trades where there is no evidence of (1) a price deviating from the auction market in a direction adverse to the customer or (2) an identifiable surplus applied for the benefit of someone other than the customer. For example, appellant Ashman was convicted of “matching” MOO and MOC orders on the curb.1 Ashman presumably made a profit on these transactions, but there is no evidence that he realized more than he could have had the trades been made by open outcry. There is also no evidence that any of the other participants in the transaction owed Ashman anything and that the trade was intended to reduce their indebtedness. This evidence is in contrast to that against, for example, appellant Schneider. In Schneider’s case, there is clear evidence that a portion of the profits was kicked back to other transaction participants. The existence of this identifiable “surplus,” which was kicked back, supports an inference of fraud. Many more traders here resemble Schneider’s situation than resemble Ashman’s.
It is not enough that failure to trade by open outcry deprives a customer of an opportunity to receive the best price, for such a failure may just as well relieve the customer of the risk of getting the worst price. Many of the examples given by the majority involve money applied for the benefit of someone other than the customer. But from these examples, we should not extrapolate to the conclusion that “the jury had ample basis to conclude that [all] the defendants manipulated orders with the intent to deprive customers of better prices so that they could generate profits for themselves.” Ante at 481. Therefore, I would require a showing of an identifiable surplus, as to each appellant, applied for a purpose other than the customer’s benefit. The vast majority of these convictions meet this criterion, but a few do not and I think that the difference must be accounted for.
Accordingly, I respectfully dissent on this point to the extent indicated.
With respect to the adequacy of the mailings here to support mail fraud convictions, Judge Wood’s opinion in United States v. Biesiadecki, 933 F.2d 539, 545 (7th Cir.1991) seems to be an adequate precedent. But I do not believe that United States v. Schmuck, 489 U.S. 705, 109 S.Ct. 1443, 103 L.Ed.2d 734 (1989), upon which Biesiadecki relied, is entirely on point. There the mailings were essential to the fraud on an ongoing basis, i.e., the retail dealer had to use the mails to transfer title to his retail customers if he was to continue buying cars from Schmuck. Here the mailing of confirmations was not integral to the fraud in any way. The confirmations did not purport to convey any information on how the trades were carried out — merely that the trades had been executed and what the resulting prices were.
I do not think one can say that the confirmations were deficient in failing to disclose that the trades were not carried on by open outcry. A confirmation says nothing about the method of trading, and no one looks to it to disclose information on this subject. If anything, by reporting the prices at which the trades were actually made— prices somewhat adverse to the customers’ interests — the confirmations might have had some tendency to alert customers to the trading improprieties. This would be precisely the reverse of the lulling effect that the majority thinks is dispositive.
As I have indicated, I think Biesiadecki controls and supports the majority opinion. But there seems to be no limitation at this point on a “lulling effect” theory. Any routine mailing in any way connected with the fraudulent activity seems to be “lull*499ing.” Not to mail confirmations here would suggest to the customers that the trades had not been made, not that they were made but at phony prices. The “lulling” rationale thus knows no real limits, but after United States v. Biesiadecki, this may be inevitable.
. In the context of this litigation, “matching” occurred when two brokers, one holding buy ' orders and the other holding sell orders (or a single broker holding both), fill both the buy and sell orders with a single local in equal quantities at prices the broker had dictated, rather than by open outcry. Matching typically occurred on the curb.