Rengan Rajaratnam, on his way back to Brazil I guess. Photographer: Peter Foley/Bloomberg
Rengan Rajaratnam, on his way back to Brazil I guess. Photographer: Peter Foley/Bloomberg

It's probably not a great idea to put too much meaning on Rengan Rajaratnam's acquittal on insider trading charges. A jury trial tells you only what 12 randomly chosen laypeople think of one particular individual. As a way of deciding important legal or economic questions, or of setting precedent for the future, it is singularly unsatisfying.1

The interesting thing, I think, is not the jury's decision2 but the stuff that came before it. Roughly speaking, there are two ways to get accused of insider trading. One is the old-fashioned, stupid way. You know a guy who works at a company that's going to be acquired. He tells you, you buy short-dated out-of-the-money call options, you make a lot of money, you give him a cut. You forget to eat your Post-It note or whatever. The government catches you. You go to jail. I write something about you to the effect of, come on, insider trading is dumb, you are dumb. The end.

The other way is ... well let's not start with the insider trading. The other way is, you run a fundamental equity hedge fund. You do lots of research, which means some amount of reading financial statements but mostly talking to lots of people. You call up investor relations people at companies and say "how's the quarter coming?" You meet with executives and study their body language. You call up doctors and say "what do you think about the idea of a drug that does X to treat condition Y?" You call up people in China and say "hey does this company actually have trees there or what?" You call up helicopter rental companies and say "hey can you fly over that company's oil tanks and see if there's oil in them?"

All of this is perfectly, tiresomely legal. And all of it gives you a little bit of edge over the people in equity markets who don't spend all their time doing research or who are not good at it, so you do well and grow your assets under management and hire some analysts and tell them to go do the same thing. "Do some research, learn some stuff, find me some trades," you tell them. And off they go and call up investor relations people and doctors and people in China and helicopter companies for you, and you watch them learn the business and smile fondly and think back on your own younger, hungrier days.

And they in their turn develop their own networks, and so on recursively. That is the business; it is a business of learning stuff that not too many other people know, but the only way to learn it is if someone knows it. So you end up sitting at the center of a web of people who know (people who know (...)) stuff about companies, and you monitor the signals you get from that network and make trading decisions based on those signals.

At some point, someone in your network probably learns things about companies by calling his dumb friends at those companies and being like "hey tell me if you're gonna miss earnings and I'll trade on it and give you a cut of the profits," because some people are dumb, and a large enough sample of people will include some dumb ones.

And that signal is illegal. It's illegal to trade on it.3 The dumb guy in your network, and his dumb friend, are going to jail, in this schematic description.4 But the signal travels back to you, and it carries with it some indication of its illegality, though that indication gets weaker as it travels back through the network. So the dumb guy tells your analyst's friend "hey my dirtbag source at Company X told me they're gonna miss earnings" and the friend tells your analyst "hey there's a guy at Company X who says they're gonna miss earnings" and your analyst tells you "hey I think Company X is gonna miss earnings."

And the question is, should you go to jail?

In the generic version of the story ... that seems like a hard question to answer? Like, if you set up your network with the goal of getting tons of illegal insider information but laundering it through enough layers that you can always plausibly deny that you knew the source, then, sure, maybe you should go to jail.5 If you really thought everyone in your network was aboveboard and had no reason to think there was anything nefarious happening, then you probably shouldn't.

Obviously one solution to this problem is to give it to a jury. If the jury thinks you were basically a good guy, no jail. If the jury thinks you were basically a schemer, jail. This is a perfectly plausible approach, and really sort of the purpose of the jury system, but there seems to be a growing sense that it doesn't work that well for this specific set of cases. For me, this story is flabbergasting:

Two jurors still held doubts about whether Michael S. Steinberg, the highest-ranking SAC Capital Advisors employee to be accused of insider trading, explicitly knew that his analyst had given him illegal tips. Dozens of emails, phone records and trading charts had failed to clear up their lingering hesitations about Jon Horvath, the analyst and the government’s star witness.

Then, on Wednesday afternoon -- after a lunch of vegetable-and-turkey wraps and burgers -- the two holdouts had a breakthrough. One juror asked Demethress Gordon, the jury forewoman and one of the holdouts, to walk through a doorway. She complied, and the fellow juror said, “I told you to go through the door, but I didn’t tell you explicitly how to go through the door,” Ms. Gordon recalled in an interview.

“It’s like the elephant in the room -- it’s obvious you know what to do,” she added.

If your view of Wall Street is that it's greedy banksters doing illegal things, then, I guess, it's obvious to you that telling someone to get information about a company really means that he should get illegal inside information about the company. And, sure, 12 randomly selected laypeople might have that view.6 But for people who know something about how the financial world works, that view is ... questionable. Like I said, you could just mean that he should rent a helicopter and fly over the company. Totally legal.

Now, judges see more of these cases, so you'd expect them to have a better sense of things than the one-time-use randomly chosen jurors. And it turns out that there's just a switch that judges can flip about whether a jury should get to decide this question. It seems like -- and the appeals court is still deciding this, but it seems more likely than not to turn out this way -- the rule in future insider trading prosecutions will be that you can only be convicted if you knew, not only that the original tip came in violation of some duty to keep it confidential, but also that the tipper received some benefit for giving the tip. That second part -- that you knew there was a benefit -- is an easy thing to prove if you are some dope trading on your buddy's merger news and giving him a cut of the proceeds, but it is much harder in the case of hedge fund managers with sophisticated research networks. If your analyst comes to you and says "I know a guy who knows a guy who says that Dell will miss earnings," you might have reason to suspect that the original guy wasn't supposed to tell the guy who told your guy about the earnings miss. But I don't think there's any way to read that sentence to mean "... and my guy paid that guy for his information."

The old rule basically allowed prosecutors to point to some well-timed trades based on information from your network, tell the jury how much you got paid, say "hey isn't that suspicious," raise their eyebrows, and let the jury decide that those trades were too well-timed to be luck. The new rule would require prosecutors to really show that you were knowingly running an insider trading ring. Like, that you saw the receipts for all the bribes that your minions paid to get inside information -- not just that you got some fishily-sourced information and went ahead and traded without asking too many questions. If you take this requirement seriously, then it's hard to imagine any sophisticated insider trader ever being convicted.

I exaggerate tremendously, and this is not legal or illegal advice. But I think it explains what's going on at the Second Circuit, whose judges are pretty skeptical of the convictions of Todd Newman and Anthony Chiasson. That case is technically over some jury-instruction stuff, but it's hard for me to imagine that the real issue is the jury instructions. It's: Are you really going to require prosecutors to prove that these guys knew that their tips were paid for? Or are you going to let them present some suspicious trades and ask a jury to decide if the traders seem like bad guys?

And so in Rengan Rajaratnam's case, the judge didn't give a jury instruction telling the jurors to decide if he knew that the sources of the inside information got any benefit. Instead, she just tossed out two of the insider trading counts entirely:

Judge Buchwald, taking her cue from the appeals court, required the government to prove that the younger Mr. Rajaratnam was aware that his brother had provided some benefit to the source of the information. ...

The judge delivered an even heavier blow to the government when she dismissed the two charges accusing Mr. Rajaratnam of trading on inside information about Clearwire. She found that the government failed to show he was aware of any benefit received by the tipper.

If you're the judge, why not leave that question to the jury -- with a clear instruction -- and hope they make the right decision? The answer has to be that

  1. you're sure that Rajaratnam didn't know of any benefit received by the tipper, and
  2. you don't actually trust a jury to consider that question, instead of the more generic is-he-a-bad-guy/are-these-trades-fishy questions.

So she tossed those charges, leaving only a conspiracy charge that the jury sensibly viewed as an afterthought.

I think, then, that the lesson of Rengan Rajaratnam is that judges are getting worried that it's too easy to convict hedge fund managers of insider trading. (Despite, of course, the fact that a jury acquitted Rajaratnam.) People who do fundamental research, and rely on other people to do that research, are at risk of coming into contact with information obtained illegally and passed on in murky circumstances. Prosecutors -- and, apparently, juries -- tend to think that that should be enough reason to send them to prison. Judges are beginning to doubt that.

1 I mean, don't get me wrong, as a criminal justice mechanism -- as a democratic check on the government's power to put people in prison -- it's fantastic. But as a method of policy-making and national catharsis it strikes me as suspect. And a lot of people like it for those purposes. Here's Jesse Eisinger on financial-crisis cases:

The solution is to move away from settlements and toward trials and convictions, especially of individuals. There is the possibility of abuse and arbitrariness in such prosecutions as well. But they would be more public and prosecutors would be called on to prove their case.

And here's the Bloomberg View editors:

If prosecutors can extract big-ticket settlements without even filing charges, as they do in so-called nonprosecution agreements, there's little to protect companies from shakedowns. If U.S. attorneys spend their time cutting deals with companies, they lose the skills needed to investigate complex cases and bring them to trial.

And Judge Jed Rakoff has expressed related views, which I've (sort of) written about before.

For myself, I tend to think that journalists love jury trials because they expose lots of facts in an adversarial and narratively compelling way. But a jury trial is not supposed to be a storytelling mechanism; it's supposed to be a decision-making mechanism. At the end of it, somebody maybe goes to prison! That's not a good thing to risk if you just want to expose a lot of facts.

Of course you might like it as a decision-making mechanism too, but this is a little weird. It's possible that the right way to decide whom to blame for the financial crisis, or whatever, is by randomly choosing 12 laypeople and letting them decide without asking them to issue written findings or explain their rationale. But that would be weird, no? And even in the simple criminal-justice case -- just, should this guy go to prison for what he did? -- it's not as simple as saying "try everyone and let a jury decide." The system is actually meant to work as a set of checks in which the investigators, the prosecutors, the judge, the jury, the appeals court, etc., can all decide that a crime wasn't committed. That's part of why prosecutors win so many insider trading cases: They don't bring cases unless they're pretty sure there was insider trading. That's a good thing! The jury trial is a bit of a roll of the dice, and you shouldn't expose someone to a random risk of going to prison unless you're sure you want him to go to prison. Juries are an important check on the risk of sending innocent people to prison, but it would be a scary world if they were the only one.

2 That said, though, the jury is pretty entertaining?

“Raj kept the cards to himself and Rengan was like a puppy playing in a field, trying to get his attention,” [one juror] said. ... “The whole feeling of the trial was, ‘Hey, let’s give this a shot,’ ” she said. She cited the government’s indictment of Rengan on the last day it could do so, a point made by Gitner in his closing argument. “Basically we were all waiting for the government to come up with more of a smoking gun and we were like, really?” she said.

3 With an asterisk, the asterisk being the rest of this post. It's not just "illegal to trade on it." It's complicated!

4 In actual fact not so much: In this round of insider trading prosecutions, the tippers have done much better than the tippees, and the direct tippees have done better than the ones at greater remove. This is partly because the prosecutions are weird and maybe the behavior wasn't illegal, and partly because the prosecutorial instinct is to cut a deal with the small fry in order to get them to cooperate against the bigger fish.

5 I mean, if anyone should go to jail for insider trading, blah blah blah.

6 Or they might not. Again, Rengan Rajaratnam's jury seems to have taken the other approach. But be careful here: Being the "puppy"-ish younger brother probably makes it easier to acquit; if you're the boss, you don't have the same sort of no-one-told-me-anything defense.

To contact the writer of this article: Matt Levine at mlevine51@bloomberg.net.

To contact the editor responsible for this article: Tobin Harshaw at tharshaw@bloomberg.net.