On July 2, a trader for a large Canadian hedge fund manager was arrested in Austin, Texas, on securities fraud and wire fraud charges.
As detailed in a press release and a complaint from the U.S. Attorney's Office for the Southern District of New York, the trader, Sean Wygovsky, allegedly perpetrated a scheme in which he used his access to information about his employer's impending trades to make profitable trades in the accounts of three close relatives by trading in advance of these trades.
The U.S. Securities and Exchange Commission, which also filed charges, described it as a "lucrative and fraudulent front-running scheme."
In announcing the case, the SEC highlighted the use of data analytics in bringing the case.
Consistent with this description, the evidence included in the U.S. Attorney's Office and U.S. Department of Justice complaints focused on a description of suspicious trade patterns and allegations that these trade patterns were consistent with illegal activity.
Thus, in supporting their respective fraud cases, both the U.S. Attorney's Office and SEC relied heavily on the sort of trade pattern evidence that data analytics can help find.
This article explains how the SEC uses data analytics to identify suspicious trading patterns like those at issue in the case against Wygovsky, how to evaluate a case based primarily on such trading evidence, and how these principles may apply in the case against Wygovsky.
The Use of Data Analytics and the Alleged Trade Pattern in Wygovsky
The SEC has consistently touted its investment in data analytics technology and emphasized its role in helping it uncover potential fraudulent schemes. In its most recent annual report, the SEC's Division of Enforcement emphasized that "a number of ... cases involving complex, abusive trading originated from the Analysis and Detection Center housed within the Division's Market Abuse unit" and explained that the group "uses data analysis tools to detect Anthony Barkow Charles Riely suspicious trading patterns."
The press release issued by the SEC concerning the Wygovsky case similarly emphasized the role of its data analytics tools in uncovering the scheme. Joseph Sansone, the head of the Market Abuse Unit, said that "[t]hanks to the SEC's development and use of sophisticated analytics tools, Wygovsky's alleged scheme was uncovered."
What these data analytics tools allow the SEC to do is identify the trade patterns such as the one identified in the case against Wygovsky. As detailed in the U.S. Attorney's Office and SEC complaints, the government claims that Wygovsky's alleged trading fit into a neat pattern consistent with fraud.
The U.S. Attorney's Office and SEC alleged that, as a part of his work for the hedge fund, Wygovsky was responsible for executing trades in his employer's accounts and thus had material nonpublic information about these yet-to-be executed trades.
The government alleged that Wygovsky's trading in employer accounts and relative accounts showed a pattern consistent with illegal activity.
First, Wygovsky would establish a position in a relative's account just before the hedge fund manager's accounts began executing a large order, or at the same time that those accounts were executing the trades.
After the trades in the relative's account went up in value as a result of the large orders in the hedge fund manager's accounts, Wygovsky would close out the position in the relative's account "nearly always at a profit."
These trades generated modest, but consistent, returns.
The four trades included in the SEC complaint each generated less than $10,000 in profits. But the government alleged that Wygovsky made trades fitting this pattern more than 600 times and amassed $3.6 million in profits in a six-year period.
Evaluating Cases or Investigations Built on Trade Patterns
The reliance on trade pattern evidence obviously presents the question of whether, and in what circumstances, the government can support fraud charges with such evidence.
Courts have recognized that evidence of trade patterns can provide at least some evidence of illegal conduct.
In insider trading cases, for example, courts have found that a pattern of unusual trading just prior to a major public announcement, such as a merger, can evidence insider trading.
Similarly, evidence that an investment professional who obtains high one-day profits on her own trades and a less desirable return rate for her clients can be evidence of cherry-picking.
Given the government's increased reliance on data analytics and courts' willingness to credit trading evidence in some circumstances, it raises the stakes for practitioners to know how to evaluate evidence based on such work.
In responding to a case based on data analytics, whether in litigation or in the investigative stage, it is essential to have a way to test the strength of the trade pattern that may be at issue.
The two primary factors that determine whether such cases will succeed are: (1) how strong the evidence created by the trade pattern is, and (2) whether the other evidence is consistent with the government's allegations of illegal conduct.
First, the question of whether trade pattern evidence will be sufficient to support fraud charges depends on whether the trade pattern alone presents compelling evidence of illegal activity.
The trade pattern identified by the government usually consists of evidence of timely and profitable trading. But it is obviously not illegal to make profitable trades, or a series of them.
The next step in the analysis is checking whether the alleged pattern is only consistent with a form of fraud, or whether it is equally consistent with a trader just being lucky or good.
Another relevant question is whether the government's analysis focuses on all relevant data, or whether it leaves out other relevant — and less profitable — trades.
Second, the question of whether a trade pattern can support fraud is also obviously informed by what other evidence is presented.
Once the government finds a trade pattern suggestive of illegal activity, both the government and the defense are well-positioned to follow up with additional steps.
These additional steps give both sides a chance to test possible explanations for the trading data that appears consistent with illegal activity. That additional investigation will likely either neatly line up with the government's theory or point to evidence that supports a defense.
As a practical matter, many cases based on trade patterns end at the investigative stage because additional work disproves the initial theory suggested by the trading records.
The Case Against Wygovsky
In the complaint against Wygovsky, the government appears to allege that the facts fit into a textbook case of fraud charges supported by trade pattern evidence.
First, as referenced above, the U.S. Attorney's Office and SEC complaints alleged that accounts in the name of three close Wygovsky relatives placed trades perfectly timed to benefit from the trades that Wygovsky entered in his employer's accounts.
The government will likely contend that only a nefarious motive can explain why these accounts traded the same stocks as Wygovsky's employer at the exact moment they stood to profit from the market impact of Wygovsky's trades on behalf of his employer.
Second, the government also alleges that the nontrading evidence also supports the front-running cases.
Although the complaints focused primarily on trading evidence, the U.S. Attorney's Office and SEC complaints also point to other evidence, such as IP login information, suggesting that Wygovsky made the trades at issue. The U.S. Attorney's Office also points to the fact that the relatives wired Wygovsky a meaningful portion of their profits.
In the end, like any other case, the strength of this evidence will be tested by trial if Wygovsky chooses to contest the charges.
The filing of the case, though, represents an important example of how the government's use of data analytics works in practice.
This article was first published in Law360.