Streetwise Professor

August 2, 2010

If it Waddles Like a Duck, Quacks Like a Duck, and Flies Like a Duck . . .

Filed under: Commodities,Derivatives — The Professor @ 8:29 pm

I finally had a chance to do some preliminary analysis of cocoa data.  My initial findings are that price movements and price-stock relationships during July, 2010 were consistent with manipulation, and unlikely to have resulted from chance in a competitive market.

In brief, I executed the kind of event study described in my 2004 American Law and Economics article on manipulation, which examined the Ferruzzi soybean case.  For a January, 2000-June, 2010 control period, I regressed the change in the log of the July London cocoa price against the change in the logs of (a) the September London cocoa price, (b) the November London cocoa price, and (c) the July ICE New York cocoa price.  (I used the settlement prices for London, and the opening price from New York).  For each year, the data included in the sample runs from 1 March to the last trading day of the July contract.

Given the coefficients from the regression, I then forecast what the log change in the July price would have been for each day in the month of July, and take the difference between the actual log change and that forecast.  I then cumulated these residuals over the month of July, and calculated t-statistics using conventional methods.

The basic result is that the July, 2010 price rose about 6 percent more than one would have predicted, given the movements in the September, November, and July ICE prices.  The t-stat on this cumulative residual is as high as 2.95, which is statistically significant at all conventional levels.  (The p-value is about .1 percent.)  This rise in the relative price of July cocoa is exactly what you would expect to observe during a corner, and given the typical co-movements of all these prices, are highly unlikely to have occurred by chance in a competitive market.

Moreover, the July-September spread was the highest observed–by far–when the contract went off the board, then at any time dating back to 2000.  July was 235 GBP over September when the contract went off the board.  Prior to this year, the biggest back was under 100 GBP/ton.

Stocks rose rapidly in July at the same time that spreads were blowing out.  In a competitive market, you would expect stocks to plummet when spreads moved into a strong backwardation.

The spreads-stocks relation exhibits the typical “supply of storage” (or Working) curve shape, although it is somewhat noisy; backwardation is larger, the smaller are stocks.  By 12 July, 2010,  the stocks were very large for the amount of backwardation in the market.  Or put differently, given the level of stocks, backwardation was unusually large.  This is especially true if one adjusts for the strong downward time trend in stocks observed 2002-mid-2010.  After adjusting for the time trend in stocks, the spread usually associated with the level of stocks observed in mid-July 2010 was a contango of about 20 GBP–nowhere near a backwardation of 250 GBP.

Thus, the market exhibited all the indicia of market power manipulation during July, 2010.  There is no plausible competitive explanation.

The data cast the Sergeant Schultz “I see NUTHINK!” attitude of LIFFE, and the FT’s slobbering coverage of Tony Ward, in a very, very bad light.  The data support a manipulation claim quite strongly; LIFFE’s denials and the FT’s breathless Tiger Beat-esque coverage are very hard to square with this data.   One would have hoped that they would have taken at least a cursory–and skeptical look–at what transpired in the market before declaring that nothing is amiss.  One would have hoped in vain.

Print Friendly, PDF & Email


  1. […] …yes: cocoa market manipulation says Craig Pirrong.  He comes to that conclusion after his examination of price movements in cocoa markets revealed all the fingerprints of a classic squeeze. […]

    Pingback by Cocoa market manipulation? The evidence suggests… « Knowledge Problem — August 3, 2010 @ 12:23 pm

  2. Have you ever looked at manipulation in cheese prices:

    Comment by michael webster — August 3, 2010 @ 2:09 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress