I've occasionally entertained myself in the last few weeks by building a fantasy-universe defense for Rajat Gupta in which he was a well-meaning but somewhat bumbling director trying to raise money for Goldman and socialize with his buddy Raj Rajaratnam but never in his wildest dreams considering giving Rajaratnam illegal insider tips in exchange for "some modest benefit" to himself. That doesn't seem to be going well, does it? If you're rooting for the defense, you can't be thrilled by the sheer number of board-meeting/call-to-Raj/profitable-trade sequences, the evidence of possible financial benefits to Gupta, and most disappointingly of all the fact that Gupta is not going to get up on the stand and explain how this was all a big misunderstanding.
So let's overcorrect and declare him guilty and move right into sentencing. A while back I put up a chart that used some basic inputs and rudimentary knowledge of the sentencing guidelines to predict how insider traders would be sentenced, and it worked okay. In particular, it gave Raj Rajaratnam just over 10 years in jail; a judge gave him 11.*
Since then there have been more Galleon convictions and sentences, none of which are wildly out of line with predictions. There was also Garett Bauer and Matthew Kluger, the "thuggish" insider trader who got a record-setting sentence for their 17-year, $37mm insider trading scheme. Here's an update of the chart for Rajat Gupta's consideration:
The lines are rudimentary model predictions, based on (1) whether or not you're in the financial industry (including being a director of a public company, so Gupta qualifies), (2) whether or not you went to trial (Gupta did), and (3) how much money you made. (As discussed previously, the law about who gets how much time for insider trading is more of a set of recommendations and guesses than an actual "law.") The dots/diamonds/etc. are actual sentences for Galleon-related defendants who pled, Galleon-related defendants who went to trial, the Kluger defendants in New Jersey, and a hypothesized 6-to-8-or-so-year guess for Gupta.
This chart is a bit more scattered than the last one. There are a couple of reasons for that: one, as more Galleon insiders who either went to trial or cooperated get sentenced, there's more dispersion, since going to trial is Bad and wearing a wire is Good. Second, the New Jersey case skews the numbers, reminding you that thuggishness and simple difference among judges matters - perhaps Kluger's New Jersey judge just wanted to get into the record books and so gave him more time for a guilty plea to a $37mm scheme than Rajaratnam got for going to trial on a $63 million scheme.
But there's also the fact that the "how much did you steal" axis, which seems to be the most important thing in determining the sentence, is a slippery beast. For example, Kluger got 12 years and Bauer got 9 even though:
Alan Zegas, Kluger’s attorney, argued that his client deserved far less time because he made less than $1 million. ... Bauer made about $32 million in illicit profits in the scheme’s last four years. Zegas said Kluger believed the three men were splitting the profit equally, and his client had no idea that Bauer made 30 times more than he did.
The third man mentioned there, Kenneth Robinson, got 2.5 years because he wore a wire. (So did Gautham Shankar, the zero among the Galleon figures.) How much they each got, though, doesn't seem to be that important - the question is how much the insider trading scheme made, all in.
This was actually an issue in Raj Rajaratnam's sentencing: he argued that he should only be held accountable for his 2-and-20 share on Galleon's profits, not the full ~$63mm that Galleon made on his trading. The judge disagreed, writing about the relevant federal sentencing guidelines:
It makes far more sense to interpret the phrase “the total increase in value realized through trading in securities by the defendant and persons acting in concert with the defendant or to whom the defendant provided inside information,” USSG § 2B1.4 cmt. background (emphasis added), as referring to the increase in value realized by the defendant or others’ trading vel non rather than the value realized by the defendant himself from that trading. That interpretation would include gains from trading by “persons . . . to whom the defendant provided inside information” and thereby hold tippers responsible for gains by their tippees.
So construed, the commentary is indifferent to whether Rajaratnam, Galleon Management L.P., or Galleon investors took home that increase in value. Rather, the relevant number is the size of the increase, namely, the increase in the price of a company’s shares from the time that Rajaratnam purchased them to the time that the public learned the inside information.
Note the "and thereby hold tippers responsible for gains by their tippees." There is some dispute about what Gupta got out of his relationship with Rajaratnam that was worth breaking the law for, with some possibilities including over-valuing Gupta's investment in Galleon and the general networking benefits that come from befriending a rich hedge fund manager. Those things matter because Gupta can only be guilty of illegal insider trading if he tipped Rajaratnam to get some benefit for himself - pure charity or accident, weirdly, don't count. (Friendship ... sort of does?) But if he is guilty, what matters is not what he made for his information, but what Rajaratnam made trading on it. The government's expert thinks that's $10mm or so.
If you buy that - you don't have to, but it's the same expert who testified against Rajaratnam about the same facts, and the judge there bought it - then that makes Gupta a pretty big-time insider trader; the dumb math gets him 8 years in jail or so. Of course other facts matter, both for calculating the sentence recommended by the federal sentencing guidelines (e.g. unlike Rajaratnam, Gupta wasn't the ringleader of a conspiracy) and for guessing what the judge will actually do (where both Gupta's charitable record and the fact that, in the real world, he didn't make $10mm might actually matter). So this shouldn't be taken as too serious a prediction. At the same time, though, the chart's done pretty well so far.
* And the difference is my fault, not the chart's: I was using a $40mm-ish amount of profit from Rajaratnam's trading; the judge found $63mm, which puts his 11 year sentence right on the predicted line.