Opinion
No. 89 C 8467
September 27, 2001
MEMORANDUM, OPINION AND ORDER
This case is before the Court on the defendants' Motion to Strike the Expert Testimony of Professor Jeffery C. Williams. Defendants argue that Professor Williams' testimony does not satisfy the criteria established by Rule 702 of the Federal Rules of Evidence and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). For the following reasons, defendants' motion is granted in part and denied in part.
BACKGROUND
This case grew out of the activities of the Board of Trade of the City of Chicago ("CBOT") and the Business Conduct Committee of the Board of Trade of the City of Chicago ("BCC") in the summer of 1989. Plaintiffs allege that defendants intentionally conspired, combined, and/or acted in concert with each other and with their associated Trading Houses, Trading Houses' Clients, and the Major Soybean Merchants to restrain trade and fix prices in the soybean cash and futures markets. Defendants contend they acted in good faith out of a duty to ensure an orderly liquidation of futures.
Factually, this case stems from events at the Chicago Board of Trade in the summer of 1989. Ferruzzi, a large commercial firm that was active in the soybeans market, held a large number of "long" soybean positions in July of 1989. (For those uninitiated in the intricacies of the futures market, "long" describes the market position of someone who has bought something, whether a physical commodity or a futures contract. In contrast, "short" describes the market position of someone who has sold something, usually a futures contract. Logically, therefore, for every long there must be a corresponding short.) The BCC of the Chicago Board of Trade had been monitoring Ferruzzi's acquisition of May soybean futures, and the defendants allege that Ferruzzi behaved like it was attempting to corner the May soybean market. Therefore, the BCC became increasingly concerned with Ferruzzi's long soybean positions in June and July of 1989.
On July 11, 1989, the Directors of CBOT issued an Emergency Order which directed all entities possessing 3, 000, 000 bushels of soybeans, or more, to liquidate 20% of their long positions for a period of five trading days. The Emergency Order directed Ferruzzi, and other similarly situated commercial entities, to reduce its holdings in July soybean futures contracts by the required amount. The parties dispute the motivation for this action. Defendants contend that it was a rational response to the artificiality in the soybean market caused by Ferruzzi's uneconomic conduct. Plaintiffs assert that the action was taken to protect the financial interests of certain CBOT Board members (and their associated Trading Houses) who held substantial short positions in the soybean futures market. The plaintiffs' primary argument is that since the CBOT Board members were exposed to considerable financial risk, if in fact Ferruzzi was attempting to corner the July soybean market, then the Emergency Order was a self-serving act intended primarily to protect those market participants with significant short positions. As additional support for this position, the plaintiffs contend that in July 1989, the soybean market was not artificial.
After the Board's action, the price of soybeans futures at the Board of Trade fell, but then shortly thereafter rebounded. The plaintiffs in this case are farmers who allege they were injured because they sold soybeans at an artificially reduced price. The source ofthe purported injury is the connection between the commodity prices negotiated on the floor of the CBOT and the actual cash prices offered to farmers at crop elevators around the region. As both parties agree, the cash price offered to farmers when they bring their crops to market is directly linked to how that commodity is being traded on the floor of the CBOT. Further, both parties agree that a significant drop in the commodity futures price at the CBOT would have a corresponding negative effect on prices paid by individual farmers at market.
The plaintiffs intend to introduce the opinions of Professor Jeffrey C. Williams to establish that the July soybean futures market was not artificial prior to the Board's issuance of the Emergency Order and that the actions of the Board actually introduced artificiality into the July soybeans market. The plaintiffs plan to proffer the opinions of Professor Williams on the effect of the Emergency Order not only on the July soybean futures market, but also on subsequent monthly soybean futures markets. Additionally, Williams would testify regarding the extent of the farmers' financial damages.
DISCUSSION
Defendants contend that the testimony of Professor Williams should be barred from this suit because the tests he ran on the relevant data do not have the statistical "power" to detect market manipulation and because his methodology fails to account for several critical variables. Additionally, defendants challenge Professor Williams opinion that the CBOT had market power in the cash market for soybeans. These issues have been fully briefed by the parties and the Court conducted a Daubert hearing from May 14 through May 16, 2001.
Expert testimony is available when "scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue." Fed.R.Evid. 702. The requirement that an expert's testimony pertain to "scientific knowledge" establishes a standard of evidentiary reliability. Daubert, 509 U.S. at 590. The proposed testimony must be derived by scientific method and not based on "subjective belief or unsupported speculation." Id; (Gruca v. Alpha Therapeutic Corp., 51 F.3d 638, 643 (7th Cir. 1995). A district court has wide discretion in determining the competency of a witness as an expert and the relevancy of his testimony with respect to a particular subject. Roback v. V.I.P Transport. Inc., 90 F.3d 1207, 1215 (7th Cir. 1996); United States v. Stevenson, F.3d 1262, 1267 (7th Cir. 1993).
In Daubert the Supreme Court directed district courts to perform a "gatekeeping" function in determining the reliability of expert testimony offered under Federal Rule of Evidence 702. In making that determination, district courts have been directed to consider four non-exclusive factors:
1) whether the proffered conclusion lends itself to verification by the scientific method through testing; 2) whether it has been subjected to peer review; 3) whether it has been evaluated in light of the potential rate of error of the scientific technique; and 4) whether it is consistent with the generally accepted method for gathering the relevant scientific evidence. 509 U.S. at 593-94. Additionally, in Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999), the Supreme Court noted that in determining the reliability of an expert opinion, district courts should apply the four factors set forth in Daubert flexibly. As the Seventh Circuit has stated more colorfully, "[t]he principle of Daubert is merely that if an expert witness is to offer an opinion based on science, it must be real science, not junk science." TufRacin Products, Inc. v. American Suzuki Motor Co., 223 F.3d 585, 591 (7th Cir. 2000) (Posner, C.J.).
I. Professor Williams' Opinion Regarding the July Soybean Futures Market
Defendants do not challenge the very impressive credentials of Professor Williams, but rather his opinions and methodology in this particular case. Briefly, however, the Court would like to note his general qualifications in this field. Professor Williams received his masters and PhD from Yale University. His doctoral dissertation was entitled "The Economic Function of Futures Markets." He has published three books, all concentrating on commodity and futures markets, of which all have been peer reviewed. One of his books, Manipulation on Trial: Economic Analysis and the Hunt Silver Case, concentrated on the manipulation of the silver market and his experience testifying as an expert witness. Professor Williams is currently the Daniel Barton DeLoach Professor in the Department of Agriculture and Resource Economics at the University of California at Davis.
Professor Williams is the general expert who plaintiffs have chosen to support all of their claims. Plaintiffs intend to use his testimony to establish all of the factors which require expert testimony, including his opinions that: the July market for soybeans was not artificial prior to the activity of the BCC; Ferruzzi was not behaving uneconomically by remaining in the futures market as opposed to increasing its participation in the cash market; the action of the BCC and CBOT to issue the Emergency Order on July 11, 1989, which forced Ferruzzzi to reduce its holding of soybeans, caused the market to behave in an uneconomic and artificial manner and drove down the price of July market soybeans; his opinion, offered for the purposes of calculating damages, on the effect of the Emergency Order on soybean prices in the time period immediately following its issuance; and his opinion on the price effect of the Emergency Order on the soybean market following the month of July in 1989. Professor Williams testified during the Daubert hearing that he used the same tests in this case as he had in the Hunt silver case, which was the subject of his book Manipulation On Trial, and that his methodology had been extensively peer reviewed.
Professor Williams believes that the market conditions in July 1989 were not artificial because many deliverable soybeans existed within range of either Chicago or Toledo, Ohio to satisfy the demands of all market participants who held long positions. Williams espouses the position that, if price conditions were in fact artificial in the July soybean futures market, then individuals or commercial entities would have started to ship soybeans for receipt in Chicago or Toledo. In large part because he could not discern any change in delivery patterns, Professor Williams theorizes that the market was not artificial in the period leading up to the Board's Emergency Order. Furthermore, he believes that the price conditions in the July soybean futures market were not caused by Ferruzzi' s alleged uneconomic conduct, but rather by benign market factors such as local weather conditions and export demands.
In order to substantiate his conclusions, Professor Williams engaged in numerous statistical tests to determine whether there was artificiality in the July 1989 soybean futures market. As he testified during theDaubert hearing, he performed a series of statistical regression analyses that focused on different variables that he determined were relevant to the consideration of whether artificiality existed in the July 1989 soybean futures market. For example, he performed one regression analysis in which he examined the correlation between the July 1989 price of soybeans and the November 1989 price of soybeans (or, as Professor Williams described it, the old crop — new crop spread) using different explanatory variables. These variables included the daily soybean positions of Ferruzzi over a period of numerous months as well as the price spread for corn (which Professor Williams determined was a reasonable control variable) over the same period of time. Additionally, Professor Williams performed a series of statistical regressions involving other market participants to determine whether price relationships in corn could be explained by the trading of corn by Ferruzzi and other major players in the futures market such as Cargill, Continental, and Bunge. In Professor Williams' own words, perhaps the most important regression analysis he performed was the one that considered the "change in the spread day to day explained by the change in the spread in corn day to day and the various ways of representing Ferruzzi's positions, old crop, new crop, warehouse receipts, futures positions, and so forth." (Daubert Hearing at 181.)
Professor Williams also testified during the Daubert hearing regarding the statistical significance, or R-squared, of his various regression analyses. R-squared is a statistics term that represents how well a particular dependent variable explains what is being measured. See Richard A. Wehmhoefer, Statistics in Litigation 83 (McGraw Hill 1985). For example, if a dependent variable perfectly explained what was being analyzed, the R-squared would be 1. On the other hand, if a dependent variable only explained what was being measured 25% of the time, the R-squared would be .25. Professor Williams testified that the R-squared values for the regression analyses that he conducted which tested the effect of Ferruzzi's trading and futures positions on the price of soybeans were very low. Specifically, in a regression analysis intended to explain the change in the spread of soybeans, old crop versus new crop, with Ferruzzi's trading in corn as a dependent variable, the R-squared was only .32. Therefore, according to Professor Williams, Ferruzzi was not acting as a corner because Ferruzzi's market activities were not related statistically to the movement in the July price of soybeans relative to the new crop price of soybeans. These statistical conclusions provide the support for Professor Williams' opinion that Ferruzzi was not acting to corner the July soybean market, but rather that the Board's Emergency Order was motivated by other, more sublime factors (i.e. the self-interestedness of the members of the Board).
Having discussed the methodological and statistical underpinnings of Professor Williams' opinions concerning the July 1989 soybean futures market it is now necessary for the Court to determine if those conclusions meet the requirements of reliability expressed in Daubert and its progeny. As mentioned earlier, the Supreme Court in Daubert provided a number of distinct factors that district courts could rely upon to determine whether testimony qualifies, for the purposes of Federal Rule of Evidence 702, as a scientific opinion. We will use those guidelines as a roadmap in resolving whether Professor Williams' opinions about the July 1989 soybean market are reliable.
The first Daubert factor is whether the proffered conclusion lends itself to verification by the scientific method through testing. There is little doubt that Professor Williams' opinion that the July 1989 soybean market was not artificial was reached through a scientific process that included the use of numerous regression analyses. In each of these regressions, Professor Williams used different explanatory variables, all of which were designed to test the artificiality of the July 1989 soybean market and particularly whether Ferruzzi was attempting to assert a corner on that market. While the defendants may object to the "power" of these tests or attack the soundness of the methodology, we feel these issues are better left for a jury to decide. Therefore, we conclude that Professor Williams' conclusions satisfy the first prong of the Daubert test.
The second Daubert factor is whether the proffered conclusion has been subjected to peer review. Again, it is the opinion of this Court that Professor Williams' conclusion regarding the July 1989 soybeans futures market satisfies this requirement. Professor Williams' theories and methodologies have been the subject of at least three books published by the Cambridge University Press. Professor Williams' peers in the field of economics have extensively reviewed all three of these books. Additionally, Professor Williams' third book, Manipulation on Trial, discussed the significant role he played as an expert witness, including the methodologies he used to support his conclusions, in the Hunt silver commodities manipulation trial in the United States District Court for the Southern District of New York. Professor Williams testified during the Daubert hearing in this case that, though the commodities markets being analyzed in the two cases were very different, the regression analyses performed were largely the same. Therefore, we find that Professor Williams' methodologies have been subjected to peer review.
The third Daubert factor is whether the expert testimony has been evaluated in light of the potential rate of error of the scientific technique. This has been a major source of contention in these proceedings. The defendants claim that Professor Williams has never tested his regressions to determine if they have the "power" to detect market manipulation. The plaintiffs and Professor Williams obviously dispute this contention and instead argue that his regression methodology is very capable of detecting any artificiality in the soybean futures market. In fact, Professor Williams conducted a "power" test on his methodologies with the result that in "97 cases out of 100, the daily deliverables (i.e. changes in warehouse receipts) is positive and statistically significant." (Plaintiffs' Supplemental Memo at 3.) In the Court's estimation, this level of confidence falls within "the known or potential rate of error" for regression analysis. Daubert, 509 U.S. at 594.
Briefly, the Court would like to tackle the difficult issue of what is and what is not an event study and whether Professor Williams performed one. According to Professor Craig Pirrong, defendants' expert witness, an event study is "a formal statistical economic analysis that attempts to determine whether particular events had effects on prices." (Daubert Transcript at 400.) Plaintiffs define what an event study by citing in their briefs to a definition provided by the United States District Court for the Southern District of New York. Specifically, according to plaintiffs, an event study "refers to a regression analysis that examines the effect of an event on some dependent variable, such as a corporation's stock price." Rmed Int'l Inc. v. Sloan's Supermarkets, Inc., 2000 WL 310352 (S.D.N.Y March 24, 2000). In our opinion, regardless of which definition of event study one prefers, we find that Professor Williams performed one. Even within the parameters offered by Professor Pirrong, we feel that Professor Williams' regression analyses "attempt to determine whether particular events [namely uneconomic activity on the part of Ferruzzi] had effects on prices." As a result, while it is certainly defendants prerogative to attack Professor Williams' event study during trial, we feel the proper audience for that attack is the jury.
The fourth and final Daubert factor is whether the expert testimony is consistent with the generally accepted method for gathering the relevant scientific evidence. Both sides to this dispute agree that regression analysis is universally accepted as one of the principal methods to interpret the type of financial and economic data at issue in this case. Therefore, the fourth Daubert factor has been satisfied.
As both the Federal Rules of Evidence and the Supreme Court of the United States instruct, it is the job of this Court to act as gatekeeper to ensure that only reliable expert testimony reaches the ears of the jury. With respect to the testimony of Professor Williams on the subjects of the artificiality of the July 1989 soybean futures market and the price effect of CBOT's Emergency Order on soybeans delivered in the month of July 1989, we open the gate.
II. Professor Williams' Opinions Regarding the Post-July Soybean Futures Market
The other component of Professor Williams' conclusions in this case involve the purported price effect of the Board's Emergency Order on the soybean market following the month of July in 1989. Professor Williams has stated there was a 40 cent drop in the price of a bushel of soybeans immediately following the issuance of the Emergency Order in July 1989. According to Professor Williams' statistical tests, 25 cents of the 40 cent drop in the July soybean price could be attributed to the Emergency Order. However, Professor Williams also opines that the Emergency Order continued to have a gradually decreasing effect on soybean prices in both August and September of 1989. Specifically, he concludes that in August the effect of the Emergency Order was to reduce the price of soybeans by approximately 10 cents a bushel, and in September the price effect was 5 cents a bushel. (Daubert Transcript at 150.)
While this Court is convinced Professor Williams' scientific conclusions concerning the artificiality of the July soybean futures market satisfies the dictates of Daubert and its progeny, that conviction disappears when the issue instead is the reliability of his opinions regarding the Emergency Order's price effect in the months following July 1989. His conclusion does not lend itself to verification by the scientific method through testing because there is no evidence that Professor Williams actually conducted any tests to substantiate these purported price effects other than visually inspecting certain charts and graphs generated for his analysis of the July soybean market. While plaintiffs argue that Professor Williams' post-July price effect conclusions have been peer reviewed, there is no viable argument that these conclusions have been evaluated in light of the potential rate of error of the scientific technique. Professor Williams simply offers no testimony or evidence of the statistical "power" of these conclusions. Finally, we do not believe that Professor Williams' testimony is consistent with the generally accepted method for gathering the relevant scientific evidence. In fact, it appears that his opinions about the post-July soybean price effect are the result of guesswork and eyeballing — neither of which satisfy the dictates of Daubert. Therefore, the defendants' motion to strike the testimony of Professor Williams with respect to the post-July soybean price effect is granted.
CONCLUSION
After a thorough review of the briefs and proceedings from the Daubert hearing, we grant in part and deny in part the defendants' motion to strike the expert testimony of Professor Jeffrey C. Williams. Because Professor Williams' testimony regarding the artificiality of the July soybean futures market and the potential price effect of the Emergency Order on July soybeans satisfies the dictates of Daubert that portion of the defendants' motion is denied. On the other hand, because Professor Williams' opinion concerning the effect of the Emergency Order on the price of soybeans in both August and September is unreliable, that portion of the defendants' motion is granted.