In an effort to settle a civil lawsuit, the attorney for three former UBS employees accused of taking algorithmic code says his offer to prove his clients did nothing wrong has been rebuffed.
UBS sued three former employees in its algorithmic group, Jatin Suryawanshi, Partha Sarkar and Sanjay Girdhar, when they moved to Jefferies & Co earlier this year. In the firm’s complaint, UBS charged the three with misappropriation of trade secrets, breach of contract, breach of fiduciary duty, unfair competition and other wrongdoing.
But Lance Gotko, who represents Suryawanshi, Sarkar and Girdhar, said he offered to have an impartial forensics expert check algo code line-by-line to confirm to UBS that none of its proprietary algorithmic code was taken and used by Jefferies. He said UBS did not respond.
"We’ve tried several times to set up a procedure that would give them reasonable assurances that their code never left the hands of UBS employees, never made its way into Jefferies," Gotko said. "One would think that if UBS were interested in finding out the truth, they would have hopped at such a thing."
Neither UBS, nor its attorneys–Joel Miller, Adam Safer or Claire Huene–said they would comment while the lawsuit was pending. Jefferies said they would not comment, either.
The UBS lawsuit joins a recent increase in accusations of code theft by employees bolting to new firms that has caught the industry by surprise. In another prominent case, the U.S. government earlier this month charged a former Goldman Sachs computer programmer with theft of trade secrets and transportation of stolen property in foreign commerce. The programmer, Sergey Aleynikov, allegedly copied 32 megabytes of Goldman’s proprietary system code and sent it to a server in Germany.
Electronic trading executives note that cases of algo code theft are conceivable, but a relatively new development. But according to Blake Reese, an attorney in the intellectual property litigation department at Milbank, Tweed, Hadley & McCloy, in New York, they might be more common going forward.
They already are in the software industry. There, because of the portability of source code, high turnover of software engineers, difficulty and cost of protecting and enforcing code theft and low barriers to entry all are reasons why the misappropriation of trade secrets is an epidemic in the field, Reese said. It wouldn’t be a stretch to extend these factors to brokerages’ electronic trading divisions.
"It’s not a new phenomenon that a disgruntled software engineer may try to jump ship with the treasure," Reese said. "However, because algo trading has become more mainstream, the underlying code more valuable, and the competition more intense, it’s not surprising that civil and criminal suits against quants in this area seem to be on the rise."
In the complaint, UBS alleged that Sarkar copied from UBS 25,000 lines of physical source code representing trade secret algorithmic trading programs–the equivalent of one algorithm, or sections of several, sources said. Sarkar emailed the code to his personal email account to develop or reproduce for later use, according to the complaint. Suryawanshi allegedly attempted to hide Sarkar’s actions by deleting documents on UBS computers, the complaint said.
Gotko said the complaint is completely inaccurate. He said his clients were willing to have an impartial expert compare source code the three have created for Jefferies with the code UBS alleges they copied to verify or refute the complaint.
Most algorithms only work–and have value–within a firm’s proprietary electronic trading framework, several in the industry said. This is the case at Instinet, according to John Comerford, its global head of trading research, the group that develops its algos and trade-cost analysis tools.
"It’d be very difficult to really create our algorithms on someone else’s platform if you didn’t have the entire platform," Comerford said. "Of course, there are pieces that one could use. But lock, stock and barrel, you’d have to have the entire platform."
UBS also alleges Suryawanshi, Sarkar and Girdhar started working for Jefferies before completing their required notice periods, and that they had conspired to move from UBS to Jefferies together. The complaint alleges that the three intended "to disclose, and may have already disclosed," UBS trade secrets and other proprietary information to Jefferies.
Jefferies brought in Suryawanshi, Sarkar and Girdhar to beef up its sales and trading for its quantitative strategies effort. The three were part of a group Jefferies hired responsible for algorithmic and other quantitative-based trading products. Suryawanshi currently leads the group as a managing director and head of global quantitative strategies. He develops the firm’s algorithmic strategy and products.
At UBS, he led a team of quants and developers responsible for the algorithmic offering as head of algorithmic trading, Americas. Partha Sarkar and Sanjay Girdhar have also joined Jefferies as senior vice presidents within quantitative strategies. Sarkar, a 10-year veteran, and Girdhar, a five-year veteran, developed algorithmic trading products, most recently at UBS.
The three and UBS disagree over when Suryawanshi, Sarkar and Girdhar started working for Jefferies. The plaintiff’s complaint alleges they did so prior to completing their notice period. Gotko, in reply, said his clients started working for Jefferies after their notice period ended.
According to the complaint, Suryawanshi was required to give 60 days notice. Sarkar and Girdhar were each required to give 30 days notice. They each were to remain UBS employees until the respective notice periods ended.
The lawsuit is before the New York State Supreme Court. On July 30, the next court date, the state court is scheduled to hear several pending motions from both sides, Gotko said.
