Shining Light on Energy Market Manipulators
Over the past few years, several high-profile market manipulation cases have emerged in the energy and gas markets, including rulings against BP America and Louis Dreyfus Energy Services.
Like other regulatory issues, market manipulation is a tricky topic because its symptom—market movement—is perfectly legal when the market moves by itself. It becomes illegal, however, when there is intent, especially if the person or group behind the manipulation makes a profit. Unfortunately for regulators, a lot of ambiguity exists in deciphering between the two.
One of the most common examples of manipulation that regulators look for in the energy sector is uneconomic trading. In this scenario, the manipulator intentionally accrues losses to bias a market outcome and benefit the value of positions tied to that outcome. A trader’s job is to perform transactions that will profit the client or firm that he or she is trading on behalf of. While losses happen, a pattern of losing trades, particularly if coupled with gains on derivative products in the trader’s portfolio, is called uneconomic trading. A prime example of this illegal trading scheme surfaced late last year when French oil firm Total SA was charged with manipulating the gas market over a three-year period.
According to the U.S. Federal Energy Regulatory Commission (FERC), Total (or at least two of its traders) made uneconomic trades on natural gas during “bidweek,” the last week of the month, which serves as a time for suppliers to sell their core production and for distributors to buy enough to meet their demand in the following month. FERC’s investigation found that bidweek allowed the Total traders to manipulate natural gas prices for subsequent months on 38 occasions from 2009 to 2012.
Market manipulation cases like the Total case have led FERC and overseas bodies like the Office of Gas and Electricity Markets (OFGEM) to increase regulatory pressure on the energy sector, which was previously less regulated than its other financial counterparts.
Until very recently, however, energy and gas trading companies did not have the means to capture the subtleties of trader intention. With the development of big data and analytics, that has all changed.
While quantitative analytical methods like calculus and econometrics have long been incorporated in all financial decision-making and monitoring in the sector, in just the past few years, methods such as natural language processing have allowed for the layering of qualitative data sources, such as emails, chat messages and phone calls, on top of trade patterns in order for trading entities and regulators to obtain a much more complete picture of trader behavior and intention.
In addition to the methodology becoming more mainstream, the information technology infrastructure—the databases and data stores that allow for storage and complex queries of very large datasets—has become cheaper and more effective. Where a market manipulation investigation once took months or years to complete, monitoring can now be done in real-time.
When FERC investigates a case, uneconomic trading patterns will raise the first red flag, but other factors can help regulators distinguish between a problem and an unlucky trade that later turned profitable. These include whether there is evidence of explicit or malicious intent in emails and chats, multiple instances or patterns of questionable trading, as was seen in the Total case, or collusion among multiple actors.
These three elements could be modeled using analytical software that layers e-communications data and trades and compares trading and communications patterns among different employees in a firm. The analytical model would search for days during which a trader registered a loss for one product and a profit for a derivative product in the same location. It could then go back and look at communications just before this trade in search of key words that indicate collusion or mentions of the trading partner or location involved. In addition, the trading behavior of other traders on the desk who might have also registered a profit for derivative products that would have benefitted from the registered loss could also be questioned.
While market manipulation is a complex issue, using big data analytics and improved modeling techniques to monitor for fraudulent behavior can help protect companies from insider threats, shield consumers from egregious hikes in energy prices and ensure the stability of the energy sector.