January 5, 2010
In television, movies and in news stories for that matter, we like to discover the guilty party. A “whodunnit”, while suspenseful throughout, is easier to digest when the culprit is identified. It provides clarity and crisp resolution of conflicts presented in the story. A guilty party makes things nice and tidy for us. We rarely are satisfied at story’s end when left with a collection of culprits or worse yet, a set of circumstances, which may or may not have led to the state of affairs. That is, unless a sequel is in the works that promises to be far better than anything that otherwise could have transpired. Our investigation into the significant price spikes and extreme volatility in the cotton markets in March of 2008 does not identify a single cause or problem. There is no smoking gun, nor a single manipulator who will go to jail. In addition, it appears that the only sequel will be dictated by the market response to the latest reports of continued volatility in the cotton markets, combined with rules and regulations that govern those markets.
Unlike television or film, however, not nailing an individual cause or culprit is particularly unsatisfying to those, through no or little fault of their own, who suffered real and significant economic repercussions from those market conditions—some losing their businesses and their livelihoods. There are a few morals to this story.
The first from the investigation and public report is that there was a collection of potential culprits. Multiple factors converged in an atypical manner.
The second is that as we go forward, we need to examine what took place and look around the corner, and be nimble and quick in a continuing effort to avoid potentially problematic market conditions. We need to be thinking about potential sequels.
Since its inception, the cotton market has been a comparatively stable market, comprised for the most part by commercial traders. Speculators certainly existed, as they do in all markets. Without speculators, it would be impossible to have orderly functioning markets. That said, there are new speculators who in recent years have shifted their focus to commodity markets, including the cotton market, seeking to diversify their investment portfolios. These new speculators have the possibility to financially overwhelm the market compared to typical hedgers and have a distinctive passive long-only trading strategy. These massive passives have come to represent a significant portion of some markets. In fact, they accounted for approximately 38% to 41% of the open interest in the May 2008 cotton futures contract.
The massive passives certainly did not single-handedly drive a price spike and contemporaneous volatility, nor is there any indication that they should have been prohibited from trading. Their trading positions as a percentage of the market were not materially different from the positions held by this investor class in comparable periods.
However, it remains possible that they, among other variables, contributed to the volatility because of their trading strategy. Specifically, while they contributed a significant amount of volume on the long side, they do not add working liquidity to the market because they characteristically do not alter their trading strategy on a daily or weekly basis as new market information becomes available. During that week of March 3rd, 2008, the massive passives did not, as is their modus operandi, materially alter their futures or option positions. Therefore, they did not act as a liquidity source of potential counter parties—a function that traditional speculators in these markets have performed over the decades. By not responding to the price fluctuations, and that was certainly their prerogative, the massive passives maintained a concentrated pool of “dead liquidity”. While not per se disruptive, the massive passives were a factor that may have contributed to the volatility and drove the natural shorts, the cotton merchants, looking to reduce their exposure in the futures market into the option market.
In my view, when dealing with a market like cotton with historically low volatility and few market participants, the Commission should be especially vigilant as to how we address the massive passives. Some have argued that we need to limit appropriately the participation in the market by these traders as a class so that they do not become excessively concentrated. That is certainly a legitimate question.
These issues, issues of new market participants generally and how they may affect markets that have by and large worked exceedingly well for decades is something that the Commission needs to address. That is why I am pleased that we are currently working on a proposed rule that would, I hope, establish a system that implements mandatory hard cap position limits on traders. Such a system could ensure that no single trader has too much concentration in a given market. As I have said in the past, we need to redefine any exemptions to those position limits. Exemptions should be approved by the Commission, not the exchanges, and be targeted, verifiable and transparent. By law, the agricultural commodities, including cotton, have certain position limits in place already, but there may be a need to revise those as we move forward. Additionally, it certainly makes sense to revisit the issue of exemptions with regard to those trading in the agricultural complex.
Another take away is the fact that the trading of cotton futures contracts on the InterContinental Exchange (ICE) was subject to price limits. These price limits are the maximum amount the price of a futures contact can increase or decrease in a single day. Once they reach a price limit (either up or down) the market is “locked limit”. At the same time, however, there were no price limits on trading of cotton futures options, and therefore, option prices were not constrained, as were futures prices.
ICE Clear U.S. is the Commission-registered entity that clears and settles all transactions executed on or subject to the rules of the ICE. When the May 2008 cotton futures price locked limit up, consistent with its rules, ICE used prices from the option market as the basis for the mark-to-market calculation for open futures positions in cotton. ICE then issued calls for the settlement variation (that is more cash for margin calls from traders) based on these prices. Because the option prices were not subject to price limits, when the option price exceeded the futures limit up price, the use of option prices in the mark-to-market calculation process resulted in larger margin calls to many market participants with short positions than if the calculation process had used the limit up futures price. This, in turn, exacerbated the urgency for the shorts to reduce their exposure. Many traders could not obtain credit to make those margin calls and lost significant sums of money as a result. Meanwhile, the consequential trading patterns caused prices to continue their rise in both the futures and option market.
In the weeks following these tumultuous market events, ICE stopped calculating mark-to-market using synthetic prices, and in June 2008, ICE self-certified to the Commission a rule amendment that removes the requirement (but not the option) that synthetic prices be used to calculate mark-to-market when futures prices are locked limit up or limit down.
Additionally, on June 11, 2008, ICE submitted an amendment to their rules for Commission approval. The change would expand the daily price limits applicable to cotton futures and apply them to cotton options. While approved by the Commission, unfortunately, the amendment with regard to setting daily price limits in cotton options is not yet in effect because it only becomes practicable once ICE has the applicable technology systems in place. To date, ICE has yet to put such technology in place. That is a significant and important concern and I am hopeful that the exchange will, in the very near future, move forward on the matter.
In retrospect, perhaps we should have stepped in earlier to say that synthetic option pricing for mark-to-market in the futures market is problematic. Maybe we should have suggested a need for a better system for margining. However, our investigation simply could not isolate with evidence that the existence of price limits, coupled with the use of implied futures prices to calculate mark-to-market, resulted in an improper market price. Certainly from the perspective of traders holding short positions, the use of a settlement price above the limit price worked against them. From the perspective of the clearinghouse and financial integrity of the market, there would have been additional risk of default taken on if the limit price, which was obviously not a true futures market price, was substituted for an identifiable market price that was above the limit price.
On a separate issue, while the report does not conclude that there is any correlation between the closing of the actual cotton trading pit at the InterContinental Exchange (ICE) and the price spikes or volatility, I tend to think that this was also an anomalous situation that certainly did not assist in decreasing market volatility. March 3rd was the first day on which no cotton pit was open. The market went limit up that day in electronic trading in the very early hours of the morning. There would be no floor traders showing up at ICE that day to trade in the cotton futures pit. There would be no actual person-to-person futures floor trading. Again, the investigation did not find any relationship in this regard, yet the lack of a cotton futures trading pit as an additional venue where traders could look each other in the eye and trade certainly did not facilitate orderly markets.
In summary, when, as in this story, there are multiple factors converging in a manner that is not typical to the market, a collection of culprits if you will, it is necessary that the exchange and the regulator vigilantly monitor those factors. Furthermore, we need to be prepared to take appropriate action in the event of a reasonable possibility of potentially calamitous market conditions.
Finally, thorough investigations in government take time, more than many people would like. This investigation is no different. Our staff have done an extremely commendable job and worked exceptionally hard on this effort. I applaud them for this work product and for their ability to juggle multiple tasks at one time. The 27-page investigative report is a more in-depth public document than the agency typically releases related to investigative matters. I hope that trend continues. This report will continue to be useful as we do the job for which we have been entrusted.
Last Updated: January 24, 2011