In a move that will have commodities traders on high alert, the Seventh Circuit Court of Appeals has upheld the conviction of Michael Coscia, who was sentenced to three years in prison after a federal jury found the former trader guilty of spoofing and commodities fraud. In its 42-page opinion, a three-judge panel denied Coscia’s argument that the anti-spoofing statute is void for vagueness, finding the provision “provides clear notice and does not allow for arbitrary enforcement.” As a result, Coscia’s first-of-its-kind conviction was affirmed.

Coscia’s trial in October 2015 was followed closely by market participants because Coscia was the first person criminally prosecuted under the anti-spoofing law at issue. After the jury found Coscia guilty, the prosecutor argued that traders such as Coscia, “contemplating sophisticated scams will think twice if they know that there are more significant consequences than a civil lawsuit or a regulatory action.” U.S. District Judge Leinenweber subsequently imposed a three-year sentence with two years’ probation, which put the shocked trading community on notice.

As the former owner of Panther Energy Trading in New Jersey, Coscia practiced in high-frequency trading, a form of automated trading with programmed algorithms that allowed him to place a high volume of orders in a matter of milliseconds. His conviction was based on this automated trading strategy, which the prosecutors successfully characterized as spoofing, in addition to his suspect order-to-fill ratio, and testimony from other traders. Spoofing—a form of disruptive trading where a trader places bids to buy or sell futures contracts with the intent to cancel before execution—was implemented by the Obama administration as part of the 2010 Dodd-Frank financial reform. In creating false demand, a spoofer can artificially move prices for financial gain.

In finding Coscia guilty, the jury concluded that Coscia used computer algorithms to place large orders he never intended to have filled in the markets. Coscia’s trading strategy, as explained in the Seventh Circuit opinion, involved placing small orders to sell higher than current market price, then placing much larger volume orders on the buy side of the market. These large orders created the “illusion of market movement, swelling the perceived value of any given futures contract.” This allowed Coscia to execute his small volume sell orders at a higher price he created with artificial market movement. Once he sold the small volume contracts, he would buy back at a lower price to make a profit. Coscia did this by first placing small buy orders below the price he had created, then placing several large volume orders on the sell side, causing the price to drop in that market. Coscia then bought the small orders at the much lower price, and immediately cancelled the large volume orders. Executing this strategy tens of thousands of times in less than three months resulted in a $1.4 million profit for Coscia, which the prosecution successfully argued was ill-gotten gains.

The Seventh Circuit rejected Coscia’s vagueness argument in part because it found the above-described conduct fell squarely within prohibited spoofing conduct. Coscia’s computer algorithms were designed to “act like a decoy,” automatically placing orders to pump or deflate the market with large orders, which were then cancelled by design if ever at risk of getting filled. His commissioned program would cancel the orders if either a certain amount of time passed, the small orders were filled, or if any one of the large orders was filled. “Read together, these parameters clearly indicate an intent to cancel, which was further supported by his actual trading record.”

Additionally, the Seventh Circuit rejected Coscia’s contention that the evidence of record did not support his spoofing conviction. In doing so, the court pointed to a list of circumstantial evidence: Coscia’s cancellations represented 96% of all Brent futures cancellations on the Intercontinental Exchange, Coscia filled only 0.08% of his large orders on the Chicago Mercantile Exchange, Coscia’s algorithm developer testified that the algorithms were designed to prevent large orders from being filled and that the orders were designed to “pump [the] market,” only 0.57% of Coscia’s large orders were on the market for more than one second, and Coscia’s order-to-trade ratio was 1,592% while the average trader’s ratio ranged from 91% to 264%. Viewing the circumstantial evidence in its totality, the Seventh Circuit found a rational trier of fact could have found Coscia intended to cancel before execution, in violation of the anti-spoofing statute.

In today’s markets, the point-and-click traders are often outpaced by advanced computer algorithms such as those created by Coscia and Panther Energy. However, the decision to uphold Coscia’s spoofing conviction should be a warning to those employing such advanced strategies. Going forward, traders can expect that prosecutors—emboldened by the Seventh Circuit’s ruling—will look for others acting with the intent to cancel bids in order to favorably push the market as potential targets in criminal investigations.

The criminal spoofing trial in United States v. Vorley kicked off in the U.S. District Court for the Northern District of Illinois on September 14, 2020.  Less than 10 days later, on the first full day of deliberations, jurors sent a note to the court indicating they had reached an impasse, with two jurors holding out against a consensus on the verdict.  Following this development, the court denied the defendants’ request to declare a mistrial and instructed the jury to continue deliberations.

The jury’s difficulty in reaching a verdict on the complicated charges may foreshadow a similar outcome that occurred last year in the criminal trial of software developer Jitesh Thakkar.  In that case, Jitesh faced spoofing charges stemming from his company’s development of software that was later used by a London-based trader to spoof E-Mini S&P 500 futures contracts, which allegedly led to the “flash crash” of 2010.  The trial judge granted Thakkar’s mid-trial motion for a judgment of acquittal on a conspiracy charge based on the lack of evidence of any agreement between Thakkar and the London trader, but the judge allowed the spoofing counts to proceed to the jury.  The jury deadlocked 10-2 in favor of Thakkar on those charges and the government eventually dropped its case. Continue Reading Latest Criminal Spoofing Trial Hampered by Obstacles

After software developer Jitesh Thakkar‘s criminal trial on “spoofing”-related charges ended in a mistrial two weeks ago, the Government determined it will not seek to retry the case.  Thakkar was originally charged with conspiracy to commit spoofing and with two counts of spoofing arising out of his company’s development of software that enabled a London-based trader to more efficiently spoof the market for E-Mini S&P 500 futures contracts.  The trial judge granted Thakkar’s mid-trial motion for a judgment of acquittal on the conspiracy charge based on the lack of evidence of any agreement between Thakkar and the London trader, but the judge allowed the spoofing counts to proceed to the jury.  The jury deadlocked 10-2 in favor of Thakkar on those charges.     

On April 23, 2019, the U.S. Department of Justice (DOJ) filed a motion to dismiss the indictment—two days before the parties were to appear for a status conference to discuss whether DOJ would seek to try Thakkar again.  The trial judge granted the motion, effectively ending the case.

This result adds to DOJ’s mixed track record in spoofing cases.  Despite the landmark conviction of Michael Coscia for spoofing, DOJ has now lost the other two spoofing cases it has taken to trial.  In 2018, a Connecticut jury acquitted Andre Flotron of conspiracy to commit commodities fraud by spoofing after prosecutorial missteps left only that charge remaining against him.  The difficulty of proving an agreement to spoof or specific knowledge of wrongdoing in cases involving spoofing suggests that DOJ may reassess its burden of proof before trying cases against secondary actors.

Can a software programmer be held criminally responsible for designing a program that a trader uses to “spoof” the commodity futures market?  This is the question posed to the jury in U.S. v. Thakkar, 18-cr-36 (N.D. Ill.), which trial began this week in federal court.  The case grew out of the manipulative trading activities of Navinder Sarao, a London-based commodities trader who “spoofed” (i.e., placed bids or offers with the intention of canceling them before execution) futures on the Chicago Mercantile Exchange (CME).  Sarao’s activity allegedly contributed to the May 6, 2010, “Flash Crash” in which the Dow Jones Industrial Average dropped nearly 1,000 points within minutes.  Sarao pleaded guilty to fraud and spoofing charges in November 2016.

Jittesh Thakkar, the software programmer currently on trial, was indicted in February 2018 on charges that he conspired with Sarao to commit spoofing and that he aided and abetted Sarao’s spoofing by developing a customized software program that Sarao used to execute manipulative trades.  The indictment against Thakkar marks the first time the U.S. Department of Justice (DOJ) has prosecuted an individual other than a trader with a spoofing-based crime.

Continue Reading Bellwether Spoofing Case Goes to Trial in Chicago