Mathew Martoma: Harvard Cheater and Evil Computer Genius?
So the story of Mathew Martoma allegedly faking his Harvard Law School transcript is truly, genuinely nuts, but the prosecutors' motion to admit it as evidence in his insider trading trial strikes me as even nuttier. Here is the motion. It contains as an appendix the Findings of Fact and Decision of the Administrative Board at Harvard, which tells the story in some detail, and really we could talk about the story all day, it is bonkers.
First of all, Martoma was named Ajai Mathew Thomas: He changed his name after all this unpleasantness went down. Second, Harvard found that he forged transcripts that he used to apply to federal appellate clerkships in winter 1998-99, changing two Bs and a B+ to As (but leaving another B+, an A and an A- untouched). Third, when he was caught, he told Harvard administrators variously that he had faked the transcript as "a joke" or "with the intention that it be shown only to his parents," but that it accidentally got sent to judges.
Fourth, it worked like a charm:
On January 26 and 27, Mr. Thomas interviewed for a clerkship with Judge Sentelle, Judge Randolph, and Judge Ginsburg of the United States Court of Appeals for the D.C. Circuit. Mr. Thomas did not disclose to the judges that the transcript that they had received was not accurate.
Mr. Thomas has stated that it was his intention, in order to avoid any harmful effect from the altered transcript, not to be offered a clerkship and that he tried not to be a successful candidate at the interviews.
Oops, that didn't work. He got one clerkship offer and was one of two finalists for another. 1
But D.C. Circuit clerks are smart, and one of them noticed that the transcript seemed off, called the Harvard registrar, and the game was up. Or not: According to the Harvard investigation, Thomas/Martoma, having faked his transcripts, had no qualms about faking a few e-mails to support his story that the whole thing was a joke and that he didn't intend to take the clerkships. So he e-mailed a professor's secretary asking her not to send out the professor's recommendation letters because he was withdrawing his applications; that e-mail seems to have been sent on Feb. 2 -- after he'd been caught -- but backdated to Feb. 1 to make it look like he'd done it before being caught. The administrative board:
It is conceded that it is not difficult to change the date on a computer, which would then record the changed date on an e-mail message. There is no direct evidence that Mr. Thomas knew how to change the date on his computer on February 2; but he was accustomed to using a computer, and that operation is one with which many people are familiar.
Oh, man, remember 1999? Anyway. In addition to apparently changing the clock on his computer to send a backdated e-mail, which, yeah, not that hard, he also hired a fake firm called "Computer Data Forensics" -- just his buddy who, incredibly, "was facing federal fraud charges" -- to write a letter saying that he didn't backdate the e-mail. 2 And according to prosecutors, "Martoma himself created a brochure for the company replete with false information about the company's extensive experience." Harvard never figured out that the firm was fake, though now prosecutors have.
If true, all of this is terrible! I mean, really, just terrible. Faking your transcripts to get a prestigious clerkship: not good, obviously, but whatever, it happens. Getting caught and saying "nononono that was just for my parents, I never meant to send it": dishonest, but understandable. Changing the clock on your computer to backdate one e-mail to bolster your story: getting kinda weird. Creating a fake computer forensics firm and writing a fake brochure and three-page letter full of techno-gibberish to justify your backdated e-mail: no, stop, come on. Come on.
The problem is that it's so terrible that you can't just go around using it as evidence in an insider trading trial 15 years later. It's impossible -- or, at least, it's beyond me -- to read this stuff and not conclude, "this guy is a crook." 3 You show me this and then say, "also Martoma shut down the George Washington Bridge for reasons of political vengeance" and I'll convict. He just seems shady, y'know?
But our criminal justice system is generally opposed to using evidence of prior bad acts at trial, for exactly this reason: The fact that Harvard found that Martoma did an amazingly, amazingly crooked thing 15 years ago will make a jury much more likely to convict him of some new crooked thing now, but that seems unsporting. Prosecutors should have to prove beyond a reasonable doubt that he actually did the thing they're charging him with -- not just rely on evidence of his general crookedness.
Here, prosecutors have a problem. They have an 81-year-old doctor who will testify that he sent Martoma a PowerPoint full of inside information before it was released publicly. Or, rather, he will testify that "he recalls sending the PowerPoint presentation to Martoma via e-mail." But:
There is no computer forensic evidence, however, establishing that Martoma had an electronic copy of the PowerPoint presentation before the public announcement, whether obtained via e-mail or other means.
So the government can't prove that Martoma actually got the inside information they say he had. I mean: They have a witness who will testify that he told Martoma a bunch of bad stuff, and that he remembers sending that stuff to Martoma in writing, but they can't definitively show that Martoma actually got the PowerPoint.
So what do they want to do? They want to argue that they can't find the e-mail because Martoma is very good at cleverly destroying electronic evidence:
Moreover, no matter how Martoma obtained access to the document, he would have, upon recognizing the enormity of the profits and avoided losses obtained through early access to the presentation, used his knowledge of computer forensics to do everything possible to destroy or alter any electronic evidence that could incriminate him. In fact, Martoma's demonstrated capacity to create elaborate electronic forgeries -- which included expertise with respect to altering the dates of electronic records -- could have been employed by Martoma to generate phony evidence purporting to show that Martoma had received the document only once it had become public. Indeed, were such evidence to be presented, it would be sufficiently consistent with the same approach Martoma had attempted in the Ad Board proceedings by altering an e-mail date to make it appear inculpatory as to be admissible to establish Martoma's modus operandi. 4
This is nuts! The argument is that Martoma, a crafty computer villain who once changed the clock on his computer to send an e-mail, knew how to erase all traces of a different e-mail both from his own computer and also from the sender's computer and from all servers and networks everywhere so that the FBI can't find it no matter how hard they look. We know he can fool the FBI's best and brightest because he once changed the clock on his dorm-room PC. You could only make this argument if you have no idea how computers work, or if, more likely, you're relying on a judge who has no idea how computers work. Nuts.
The thing is, there are crooks in the financial industry. And, look, reading this evidence, I am persuaded that Mathew Martoma might be one of them. But it's not a good look for our justice system to rely on a general sense of financial-industry crookedness to get convictions for actual instances of actual crimes. That sort of happens, though, with prosecutors and the SEC arguing that the individual defendant in front of them did whatever he did out of"Wall Street greed."
"Wall Street greed" strikes me as code for: "Don't worry about the facts of this case. Don't you feel in your heart that everyone who works in finance is probably a crook? Isn't that enough to convict?"
The Harvard stuff is a little different, in that it gets at not a general sense of financial-industry crookedness, but a specific sense of Martoma's own (alleged, past, relatively harmless) crookedness. But it's still an effort by prosecutors to get an insider trading conviction, not by proving that Martoma got inside information -- in fact, they're admitting that they can't prove it! 6 -- but by creating a general sense that he's probably a crook so you might as well send him to jail. That's not the way it's supposed to work.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
These are ... if you haven't hung around a law school during clerkship application season, it is difficult to convey how prestigious and attractive these clerkships are. The D.C. Circuit is where you want to be if you're an ambitious Harvard Law student with a lot of As, even if most of them are fake. Sentelle, Randolph and Ginsburg are judges whose clerks go on to be Supreme Court clerks and scoff at their jealous Harvard Law School classmates. Getting a D.C. Circuit clerkship is a big deal. Getting one as a joke is amazing.
The fake letter is a pretty glorious piece of work. It's in the prosecutors' motion too. A generous sampling (emphasis added):
All evidence processing was done on a restored copy of the bit stream backup rather than on the original computer We proceeded to make a mirror image of the computer disk drive using the RSA algorithm. Although we typically use 32 bit algorithms, which are inherent In CRCHECK and CRC32, we were asked by the client to use elevated standards.
We used CRCMD5 to authenticate data at both a physical level and a logical level to demonstrate that we did not alter any of the evidence after the computer came into our possession. We proceeded with our investigation on two fronts: UNIX and PC. The UNIX Investigation is detailed in CDF Binder Number 1. The PC Investigation is detailed in CDF Binder Number 2. Both binders are in the possession of CDF. These binders merely contain work products. All pertinent findings have been delineated in this letter.
We have been asked to clearly delineate what software was utilized to perform our tests so that these tests may be reproduced. CDF utilizes the following software: GetFree, IPFilter, GetSlack, EnCase, FileList, FCG Analyzer, Partition Magic, and Anti-Clear. These software packages are utilized by law enforcement agencies and forensic experts throughout the country.
Fake binders! Love.
Is a crook: It happened 15 years ago but oh man does having it in front of you in writing compress that time period. Feels like I just read it this morning!
This sentence probably requires a [sic]; seems like some words are missing.
That link is about Fab Tourre's civil trial. But note that the prosecutors here won an argument that they should be allowed to shout about Martoma's "greed" in this criminal trial too.
I mean, they can't prove the thing about the PowerPoint. But, like, a witness who'll testify that they had lots of phone calls about illegal inside information, combined with the fact that SAC traded right after those phone calls, is ... actually pretty good evidence of insider trading? It's not like the prosecutors need either the PowerPoint or the Harvard stuff.
To contact the author on this story:
Matthew S Levine at firstname.lastname@example.org