Michael Lewis: Did Goldman Sachs Overstep in Criminally Charging Its Ex-Programmer?
To be honest, I’m really surprised that no one's really talked about this post yet on WSO given its content. It is a very long piece, so I imagine some of you were like me at first and didn’t have the attention span to crush all eight pages of this thing. But if you have the time, I strongly encourage you to read it; I mean, it is Michael Lewis. For the rest of you lazy monkeys out there, the following is my still-not-that-short sparknotes version of the piece (so you can discuss in comments below even if you haven’t read the whole thing, just like in school), followed by some of my thoughts on it.
Sergey Aleynikov's Life Story
Sergey Aleynikov was a Russian immigrant who left his homeland because they did not let him study what he was truly interested in: computer science. Upon moving to New York, he first found a low-paying job operating computers, and quickly moved through his career: research and a master's degree at Rutgers, internet start-ups, then finally IDT, a telecom company. After a decade there, he was the star technologist. Of course, then headhunters began calling.
Working at and Leaving Goldman
Goldman Sachs was very involved in the high-frequency trading market at this time and was interested in star programmers like Sergey. He had reservations at first, but he enjoyed the challenge and went through dozens of brainteasers, programming questions, and math problems to get his offer: which was $270,000 including bonus at the time.
Serge was very good at what he was hired to do at Goldman, which was to make their algorithms faster. However, Goldman was maybe just clearing the top 10 in this competitive marketplace, and for good reason. According to Aleynikov, their systems and practices were not as organized and conducive to good programming as those of his previous firm, IDT.
Goldman's systems were from a previous firm that they had bought out and therefore most of the coding consisted of patching up problems as they came along. Even when Serge wanted to build a new system for them, management denied his request because there is not much incentive for long term profits when everyone is looking at their year-end bonus. This incentive also impacted the firm in that competition often outranked teamwork. As Sergey put it, "When two people wanted to talk they wouldn't just do it out on the floor, they would go to one of the offices around the floor and close the door. I never had that experience in telecom or academia." Goldman also had a one-way relationship with open source; they only took from it and even if they only slightly changed the code, Goldman put their proprietary stamp on it and refused to release it back out.
After a few years at Goldman, Aleynikov was known as one of the best programmers on the street. Naturally, headhunters became calling again. After turning down other banks, he decided to take an offer to work at a new hedge fund, Teza Technologies, founded by another Russian named Misha Malyshev, who was a "legendary" high-frequency trader at the time. They offered him a salary of $1 million and that wasn't even the main reason he wanted to go. Teza wanted to build its own trading system and that fact excited him, because he was sick of patching up the old system at Goldman.
When he left, he wanted to take some of the open source code that he had been using at Goldman, but since it would have been too much trouble to splice out the open source code from the proprietary code (or to find all of it again since it was scattered all over the Internet), he just took it all: 8 megabytes out of 1 gigabyte of code. He knew it was wrong, but when asked about how he felt when he did it, he said, "It felt like speeding. Speeding in the car."
Arrest, Trial, and Incarceration
Soon after, Sergey was arrested without warning and initially had no idea what he did wrong. The F.B.I. agent who was in charge of the investigation had little to no idea about high-frequency trading or programming, so Aleynikov's attempts to explain himself were futile. In fact, 48 hours after Goldman called the F.B.I., Aleynikov was arrested, and somehow during that time, these F.B.I. agents must have learned all they needed to know about the complex business. Sergey signed a confession because he did what he allegedly did, but he did not think it was anything wrong.
After that, Aleynikov was denied bail because "he had in his possession computer code which in the wrong hands could be used 'to manipulate markets in unfair ways.' (Goldman Sachs presumably used it to manipulate markets in fair ways.)" (I love this remark by Lewis)
The jury in his trial consisted of high-school graduates, none of which had computer programming experience. On top of that, Goldman's employees were more or less trained to be salespeople (they didn't lie, but they talked about things they didn't really understand in order accomplish their goal) during the trial. GS claimed that the code was all proprietary and was worth a lot of money (even though he didn't take any of the trading strategies, see below), and the jury couldn't argue against that without understanding the business. Aleynikov's lawyer knew he didn't stand a chance, and they did not really put up much of a fight. In the end, he was sentenced, without parole, to eight years in prison.
Michael Lewis's Second Trial
Lewis wanted the truth, and believing that the justice system did a poor job of that, held his own trial in a restaurant in New York, with people familiar with Goldman Sachs, high-frequency trading, and computer programming serving as jurors.
The jurors quickly realized that Sergey Aleynikov was something special. He was gifted in many areas related to computer programming but also had raw intelligence. They also realized that his methods of "stealing" the code were quite harmless and normal in the computer programming industry. Using a subversion repository (the F.B.I. thought the word "subversion" implied malicious intent) and deleting his bash commands (in order to hide your password) were common practices. He was not trying to cover his tracks because he didn't believe what he did was malevolent.
Serge also did not take Goldman's trading strategies, which shocked all of these jurors, especially when he said "it wasn't that interesting" and that "it's all one big gamble." To knowledgeable people in this industry, it was like stealing the jewelry box without the jewels. It was clear that Aleynikov had a passion for programming and only that, not necessarily in using it for trading.
The jurors also did not find it surprising that Serge had taken anything at all, even though it was not going to be very useful at Teza (apparently the new system he was going to build there was likely to be in a different programming language too). The explanation? When most people leave their job for another one, they take their notebook full of notes that are probably not directly useful to the new company, but instead may serve as templates or thoughts they can draw on. For programmers, that notebook is their code, and Aleynikov's "notebook" had very little to do with Goldman Sachs.
Lastly, given these facts, why had Goldman gone after Sergey Aleynikov in the first place? No one really knows for sure but one theory that I liked was this: "Every manager of a Wall Street tech group likes to have people believe that his guys are geniuses. Their whole persona among their peers is that what they and their team do can't be replicated. When people find out that 95 percent of their code is open-source, it kills that perception....So when the security people come to them and tell them about the downloads, they can't say, 'No big deal.' And they can't say, 'I don’t know what he took.'"
Epilogue
Sergey's appeal was finally heard after he spent a year in jail, and he was promptly released. However, the nightmare was not over. His lawyer noticed that his passport had not been returned, and when questioned about it, the government arrested Serge again for "accessing and duplicating a complex proprietary and highly confidential computer source code owned by Goldman Sachs." But because of sentencing guidelines, he could not return to jail since he had already served time. Instead, he had to plead guilty and be let go on time served. Sergey Aleynikov, however, was not broken. He learned from the experience about the world and about the importance of his family. It was always about learning for this Wall Street programmer.
Goldman's Response
Goldman Sachs has spent millions of dollars and tens of thousands of hours developing the proprietary source code and technology used in our market-making business. The firm has put in place extensive safeguards to protect this valuable technology. In addition to contractual limits on disseminating confidential information, the firm restricts access to proprietary technology to those employees whose duties designing and maintaining the technology require such access....In this case, the United States Court of Appeals for the Second Circuit determined that Sergey Aleynikov, minutes before his going-away party, "encrypted and uploaded to a server in Germany more than 500,000 lines of source code for Goldman’s HFT [high-frequency-trading] system." While some of those files included open source software, the Court determined that "a substantially greater number of the uploaded files contained proprietary code." The Court went on to note that the code "could be integrated into a competitor’s system" and that Mr. Aleynikov then "deleted the encryption program as well as the history of his computer commands."
My Thoughts
I mean what more can I say? After reading this, I was left wondering, "Why isn't this a bigger deal? Why hasn't Goldman been sued yet?" But I am biased. I love Michael Lewis's writing in general, not to mention I have a programming background and also left an investment bank early in my career because I was tired of many of the same things that Serge had mentioned. That's why I want to hear your thoughts.
Great post.
8 MB of code.. is A LOT OF FUCKING CODE
Eddie... and more specifically Michael Lewis, you lost me after the first sentence of the synopsis.
Are you completely kidding me? How contrived and ridiculous is this story when one man's quest to escape persecution for being a ---- computer programmer? WTF? Seriously. What a joke.
He likely left Russia because he saw a bunch of American movies and wanted to come to America to chase all the golden haired, rosey cheeked easy American girls and then he he would walk over to the nearest fire hydrant, give it two quarter counter-clock-wise turns and catch as many hundred dollar bills that spewed out as he could fit in his bearskin coat. Then he'd responsibly turn off the money spigot that is stationed on most major intersections in America until he needed to re-up again.
Michael Lewis is a joke. And his articles are painfully long. I have yet to finish a single book or article of his... other than the War and Peace piece he did in Vanity Fair a few years back about Michael Burry.
Honestly, I lose faith in humanity when I hear about this stuff, I think I'm going to take a break from reading about it.
Very interesting read. I love how a lot of people look up to Goldman Sachs as gods but there's always another side.
I can't believe they put high school graduates on the jury of a trial of HFT.
Yet, you're willing to believe Lewis' "jury" composed entirely of people who do exactly the same thing that Aleynikov does are nice and impartial?
There's no way that Aleynikov was not aware he was violating company policy and likely intellectual property law when he copied thousands of lines of proprietary code, deleted embeded encryption files and transferred those files between his personal computers. Even the appellate court that overturned his conviction felt that Aleynikov clearly committed a crime, but Congress' poor statutory language let Aleynikov's crime fall through a loophole.
In a move that only further illustrates the spirit of the law, Congress has since closed the loophole through which Aleynikov received an acquittal.
Do you honestly think a guy who built the entire system would be caught so easily if he tries to steal intellectual property for the new company? I'm no computer science experts, but that just doesn't make any sense to me.
Having done quite a bit of reading on this trial, I can assure you that Michael Lewis' piece in Vanity Fair is unbelievably apologetic to Aleynikov. Lewis omits details, writes with a stalwart slant against Goldman, and concocts a "second trial" with absolutely no prosecutor, where the "jurors" are expected to reach a verdict after talking to the accused over dinner! Unsurprisingly, the tale that the "second jurors" receive is just a bit favorable to Aleynikov, who doubtlessly believes he has done nothing wrong. Equally unsurprisingly, the "jury" of HF computer programmers side with their peer and are not familiar with the nuances of intellectual property laws.
Furthermore, Lewis' suggestion that Aleynikov's colleagues failed to speak up in his defense because they were "too worried about their year-end bonuses" is as unsatisfying as it is revealing of Lewis' unrelenting mission to paint everyone in finance (even the computer programmers!) as a money-hugging, two-faced asshole who would stop at nothing to savor in the satisfaction of denying a cup of porridge to Jurgis Rudkus. So to is his ludicrous insinuation that Goldman would prosecute Aleynikov out of meta-insecurity, just to make a superficial point about the value of their HFT code. Lewis has lost his sophistication. Do these arguments really satisfy anyone other than Occupy Wall Street?
Which readings have you done on this trial? I'm curious, can you direct me? I'd like to be on the same page before I draft up a counterargument.
Thanks for sharing!
Not that good of an article. Michael Lewis should try to be less biased. I still loved liar's poker and the big short.
Look, as with just about everything, the truth is a whole lot simpler and a great deal less sinister than Lewis' conspiracy theory would have you believe.
Aleynikov wasn't born a criminal, nor do I think he thought he was committing espionage when he uploaded Goldman's trading platform to his personal server space. However, neither do I believe that Aleynikov sat by none-the-wiser as he deleted encryption, uploading thousands of lines of proprietary code onto off-shore servers, deleted his bash history, and moved those files onto various personal flash drives. Any mildly reasonable person (as Aleynikov no doubt was) would realize that they - at the very least - were violating all sorts of company policies in so doing, and it's almost certainly not in the spirit of protecting intellectual property rights.
Aleynikov was convicted not because a high school-educated jury couldn't understand his complex HFT coding, but because they weren't as legally sophisticated as the appellate court judges, who were stricter in their application of stare decisis and more keenly aware of legal precedent. The judges themselves more or less agreed with the jury: Aleynikov knowingly committed a crime and got caught. People like that get in trouble, even if they are the only "I-don't-care-about-money-or-HFT-strats-because-I'm-such-a-pure-computer-scientist-from-Russia-and-I-just-took-that-$1.2-million-per-year-job-because-I-am-so-passionate-about-HFT-programs-that-run-efficiently" guys at Goldman. Give me a break, Michael Lewis. Aleynikov is a normal guy, who was great at computer programming at bad at stealing intellectual property. What a waste; the whole situation is a real shame.
Lewis is way out of his element with this material. It's embarrassing to read.
HFT is an intellectual property business and copying code is infringement. It's not any more complicated than that.
There is a definitive grey area when it comes to open source and proprietary code. Look at the many cases involve software companies.
Like software companies, GS requires their employees sign non disclosure agreements to mitigate code theft and protect intellectual property. this guy broke that "rule".
Bottom line, if the code was truly open source then it could have been downloaded elsewhere. But it wasn't so some of the code was prop to GS.
Not sure why you got MS. Sound logic to me.
who wants to read something pro goldman??? this is about book sales/publicity.
Aleynikov conveniently contends that he copied Goldman's code simply to gain access to the open source pieces of the HFT infrastructure. His download pattern, however, would suggest otherwise, given that the overwhelming majority of the content he uploaded was GS proprietary code and not open source.
My best guess is that Aleynikov downloaded GS' framework because he wanted to look over some of the modifications he made to open source files during his time at GS. By replicating those changes, he could likely have made some valuable additions to his HF's HFT platform. Unfortunately, Aleynikov voluntarily sold the intellectual property related to those modifications to GS for an income of $400k per year. He was not free to copy these files for his personal or professional use, something of which he was no doubt aware.
I think Lewis is an engaging writer but I don't trust his financial competence or his impartiality
Just Google around a bit and you'll find a fair number of law blogs, particularly those that post on intellectual property cases have commented on the proceedings. It was a fairly groundbreaking trial as far as IP goes. It has been months since I last read anything though.
Great piece by Michael Lewis. All of his writing is good IMO.
The lawsuit did seem very skewed against the programmer, although his thumb drive conduct was clearly a red flag. I believe litigation has continued into 2013.
Clearly you didn't read the article, since Aleynikov did not use a thumb drive and "litigation" has not "continued" into 2013 - he is being tried by entirely different body for somewhat similar offenses.
Oh Poler Bare
I agree the article is biased, and i agree the dinner trial idea is an absurdity. However, having worked in the industry for a decade, i DO think many people would let someone go to jail if would help them financially and I also do believe the explanation that goldman pursued the case aggressively in part because management wanted to keep the appearance that the firm has proprietary technology that is highly valuable when in reality it is worth significantly less then generally believed....wall st is very Wizard of Oz-like and no expense is spared to hide the man behind the curtain. I have seen many examples of this altho obviously not as dramatic as prosecuting someone and getting a criminal conviction. As the old daying goes, "the greatest defenses lie at the gates to nothing"...the real secret is that they didnt have anything that was worth that much.
If "many people think that's why the banks settled in equity research cases", don't those "many people" realize, by definition, that equity research is useless. And if it's useless, then does it matter whether or not they admit it?
If you think that the people you are working with would sit by idly while you are wrongfully convicted of a crime that leads to 8 years in prison, I think you should get a new job.
Aleynikov was clearly guilty of a crime, which is the much more reasonable answer as to why GS colleagues didn't "stand up for him". Moreover, what, exactly, were his GS colleagues suppose to say in his defense in Lewis' version of the world? What defensive purpose could calling another GS colleague have that is germane to the case?
Eligendi aperiam sint nesciunt molestiae. Et est libero qui deserunt atque doloremque. Reiciendis aperiam quia magni.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Vero temporibus odio vel aut nostrum. Consequatur voluptate ut voluptatum aliquam. Ut harum doloremque architecto fugiat aliquam. Nihil architecto quaerat rem animi. Et officia soluta cumque veritatis.
Rerum fugiat quaerat ea quam pariatur est. Quia et et sed modi.
Reprehenderit similique ipsam provident quasi et vero dolorem. Dignissimos saepe dolorum ipsa consectetur enim suscipit recusandae iste. Sequi sed non asperiores suscipit et ullam. Dolores magni corrupti maiores nihil asperiores.