Manipulation on Trial. By Jeffrey Williams.

Manipulation on Trial.
By Jeffrey Williams.
New York and Australia: Cambridge University Press. 1995. Pp. xxi, 244. $39.95.

Did the Hunt brothers (Bunker, Herbert and Lamar) and associates really manipulate the silver market during the 1979-80 price spike period? The jury verdict in the 1988 trial of Minpcco v. Hunt said, "Yes." But would a jury composed of efficient market economists come to the same conclusion? Perhaps not, based upon the fascinating account by an expert witness economist for the defense, Jeffrey Williams of Stanford University.
Williams identifies four kinds of manipulation: (1) A "corner" or "squeeze" whereby a player in the futures market holds so much stock of a given commodity that other players must pay exorbitant prices to fulfill their "short" contract commitments; (2) a "rumor" of impending shortages of a commodity or related futures contract that is started by an established market player that is believed by other market participants long enough to allow the player to liquidate his position at top prices; (3) a statement by the manipulator professing broad "investor interest" to hold the commodity that allows the manipulator time to sell out at a high price; and (4) a type unknown in commodity law prior to the Hunt case:
A "price-effect" manipulation, in which a trader holding an opinion about the commodity's long-run value buys (or sells) it in sufficient quantity to influence the current pricc. knowing that such an effect on price will occur [p. 7].
Williams was dismayed to learn that, ex post, the Hunts were found guilty of type (4) manipulation which had never been identified previously in commodity law as being illegal. "Thus comes the main lesson from the Hunt silver case: Law involving such complex subjects is, in effect, created retroactively" [p. 212].
The plaintiff's lawyers' mind set was that if their client suffered damages, then somebody must have caused it. and it must have been done with malicious intent (which makes the behavior illegal). Is it possible that the Hunts had innocently (with no ill will toward any other person) played in an impersonal market, one that depends upon participants trading contracts they all hope will be profitable? Williams wonders if the Hunts' only misbehavior was that the size of their positions were too large and they appeared uncooperative to regulators [p. 212]?
Williams's frustrations as an expert witness are revealed by a theme he develops in all his essays that economic experts for both sides "based their analysis on unarticulated assumptions . . . about the nature of manipulation, the purposes of futures markets, and the formation of commodity prices" [p. 13]. He believes that the assumptions should be made explicit, as is done in more academic settings, so that they can be questioned and compared with other assumptions to detect internal inconsistencies in analysis. Other laments arc revealing:
Unfortunately, no expert put forward a clear definition of manipulation or explained the range of normal price relationships. Hence the discussion was often internally contradictory and misdirected (p. 13].
Actually, the experts did not contradict one another as much as emphasize different aspects of the determination of the price of silver, in large part bccause a comprehensive picture eluded any analyst. The members of the jury, left with the task of choosing among the experts, could do so properly only by seeing the elusive comprehensive picture: their task, however necessary, was next to impossible [p. 14].
... the contributions of the economists were largely misinterpreted or ignored, in part because the manner in which economists tend to analyze a problem (especially an allegation like manipulation) is fundamentally different from, and at times incompatible with the way the legal system assesses cvidcncc and determines cause. Those differences between economic reasoning and legal reasoning, between economic analysis and trial procedure, which went virtually unrecognized by the participants, crucially affccted the outcome of the Hunt trial (pp. 14-15].
A conceptional point made by Williams of special interest to both econometricians and forensic economists is that there is no single price of silver. Rather, there is a "web of prices" encompassing time (for immediate or future delivery), location of delivery (to exchange-approved vaults or otherwise), form (bullion, coin, and scrap), and services (refining, transportation, and storage). "Perhaps no single concept is as important for comprehending the Hunt trial and the experts' testimony as this idea of a web of prices" (p. 25]. And there is the related concept of "deliverable supply" which includes not only supplies of bullion available in exchange-approved vaults but additional amounts that could be moved in quickly at minor expense from, say, producers' stocks and coin melting [p. 16].
With regard to the econometric work related to identifying the major factors that determined the price of silver during the Hunt spike period, we note that both plaintiff and defense experts pursued three common methodologies: regression analysis, correlation coefficient and causality test. The defense experts employed an ingenious approach of replacing variables related to the Hunts' silver position with the daily temperature in Buenos Aires, an unrelated variable that appeared statistically to have (but could not have) caused the price of silver to rise, and claiming that other political and economic events must have been to blame. Experts on both sides tried to demonstrate the robustness of their conclusions with various ad-hoc models, different time intervals, and selective sample periods; however, a different result or verdict might have been found had more rigorous econometric analyses been considered, say, in the areas of the specification of the linear function form, the significance of correlation coefficients and the sensitivity of causality tests to the number of lags and to other variables. The controversy over whether to use daily prices or price changes in various regressions should be less significant if one takes a view of the recent development in stationarity analysis in time series research. An interesting point, yet not pursued by either side, is how the price of silver and its changes were directly influenced by market expectations in 1979 and 1980; this would open another field of battle for the Hunt silver case. But as the frustrated Williams remarked, the accepted style of testimony and cross-examination prevents the limitations of a modeling approach from being suitably acknowledged. This results in what he calls "internal inconsistency", which means testing the cause of the price rise in silver with one model and determining the appropriate price of silver with another.
Williams's frustrating experience as an economics expert in the Minpeco vs. Hunt case should be interesting and instructive reading for business economists, especially for those who would venture into the loosely-charted, interdisciplinary' world of forensic economics.
Samuel C. Webb
Jen-Chi Cheng
Wichita State University