UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK ------------------------------------- x UNITED STATES OF AMERICA, : : -v: : RAJAT K. GUPTA, : : Defendant. : ------------------------------------- x JED S. RAKOFF, U.S.D.J.

11 Cr. 907 (JSR) SENTENCING MEMORANDUM AND ORDER

The Court is called upon to impose sentence on Rajat K. Gupta, who on June 15, 2012, was found guilty by a jury of one count of conspiracy and three counts of substantive securities fraud, in connection with providing material non-public information to Raj Rajaratnam. Federal law requires a court to state, not only orally

but in writing, its reasons for imposing a sentence “different from” a Guidelines sentence. 18 U.S.C. § 3553(c)(2). See also United

States v. Rattoballi, 452 F.3d 127, 128-29 (2d Cir. 2006). This will be a non-guidelines sentence, and, accordingly, the Court will both read this Sentencing Memorandum in open court and docket it promptly thereafter. Imposing a sentence on a fellow human being is a formidable responsibility. It requires a court to consider, with great care and The notion that

sensitivity, a large complex of facts and factors.

this complicated analysis, and moral responsibility, can be reduced to the mechanical adding-up of a small set of numbers artificially assigned to a few arbitrarily-selected variables wars with common

sense.

Whereas apples and oranges may have but a few salient

qualities, human beings in their interactions with society are too complicated to be treated like commodities, and the attempt to do so can only lead to bizarre results. Nowhere is this more obvious than in this very case, where the Sentencing Guidelines assign just 2 points to Mr. Gupta for his abuse of a position of trust -- the very heart of his offense -- yet assign him no fewer than 18 points for the resultant but unpredictable monetary gains made by others, from which Mr. Gupta did not in any direct sense receive one penny. It may be worth remembering that the Sentencing Guidelines were originally designed to moderate unwarranted disparities in federal sentencing by enacting a set of complicated rules that, it was hypothesized, would cause federal judges to impose for any given crime a sentence approximately equal to what empirical data showed was the average sentence previously imposed by federal judges for that crime. 96 (2007). See generally Kimbrough v. United States, 552 U.S. 85, From almost the outset, however, the Guidelines deviated For example, even though a perceived racial

from this goal.

disparity in sentencing was one of the evils the Guidelines were designed to combat, in actuality the Guidelines imposed in narcotics sentencing a huge racial disparity that dwarfed any prior such problem. Specifically, the Sentencing Commission, based on limited

2

and faulty data, originally determined that an ounce of crack cocaine should be treated as the equivalent of 100 ounces of powder cocaine for sentencing purposes, even though the two substances were chemically almost identical and, as later studies showed, very similar in their effects. Since, however, 85 percent of crack

cocaine offenders were black, while most of those who dealt in powder cocaine were Caucasian or Hispanic, the result of the 100-to-1 ratio was to force upon the courts a gross racial disparity in narcotics sentencing. See id. at 97-98. It was only in 2010 that the ratio was

changed from 100-to-1 to 18-to-1; and even then as much on the basis of conjecture as evidence. See generally Dorsey v. United States, 132 S.Ct. 2321, 2326 (2012). For the Sentencing Commission had no more

empirical basis for imposing the ratio of 18-to-1 than for earlier imposing the ratio of 100-to-1. In both cases, the numbers were plucked from thin air. While this example is drawn from the area of narcotics, the fundamental point is equally applicable to the instant case. Here, as there, the numbers assigned by the Sentencing Commission to various sentencing factors appear to be more the product of speculation, whim, or abstract number-crunching than of any rigorous methodology –- thus maximizing the risk of injustice. Another example of the deviation of the Guidelines from the original goals of the Sentencing Commission –- and one more directly relevant to the instant case –- is the huge increase in the 3

recommended Guidelines sentences for securities fraud cases.

The

Guidelines’ calculations for this offense are no longer tied to the mean of what federal judges had previously imposed for such crimes, but instead reflect an ever more draconian approach to white collar crime, unsupported by any empirical data. Take the hypothetical but

typical case described by Professor Kate Stith of Yale Law School, involving a typical securities fraud defendant who pled guilty to inflating the financial figures of a public company, thereby causing at least 250 shareholders to collectively suffer a reduction of more than $12.5 million in the value of their shares. In 1987, such a defendant would have faced a Guidelines sentence of 30-37 months; but by 2003, the same defendant would have faced a Guidelines sentence of 151-188 months, a more than 500% increase. See Kate Stith, Federal Sentencing: The One-Way Ratchet, New York City Bar Association First Annual Conference on White Collar Crime (May 2012). Was such a crime really 500% worse in 2003 than it was in 1987? Had any of the

factors that underlie rational sentencing so radically changed as to warrant such a huge increase? In fairness, this vast increase in white collar sentencing was partly mandated by Congress, reacting in turn to public outcry over such massive frauds as Enron and WorldCom. But in implementing

the Congressional mandate, the Sentencing Commission chose to focus largely on a single factor as the basis for enhanced punishment: the amount of monetary loss or gain occasioned by the offense. By making

4

a Guidelines sentence turn, for all practical purposes, on this single factor, the Sentencing Commission effectively ignored the statutory requirement that federal sentencing take many factors into account, see 18 U.S.C. § 3553(a), and, by contrast, effectively guaranteed that many such sentences would be irrational on their face. This Court has already had occasion to comment on the unreasonableness of this approach in United States v. Adelson, 441 F. Supp. 2d 506 (S.D.N.Y. 2006), and hereby adopts by reference the observations made there. But there is no better illustration of the

irrationality of this approach than the instant case: for of the total of 30 Guidelines points calculated by the Probation Department and endorsed by the Government as reflecting the proper measure of Mr. Gupta’s crime and punishment, no fewer than 20 –- or two-thirds of the total -– are exclusively the product of Rajaratnam’s and his companies’ monetary gain, in which Mr. Gupta did not share in any direct sense. It might be argued that the Guidelines still work to minimize disparities. But if the sentences so calculated are the product of

placing an overwhelming emphasis on a factor that may be central to some frauds but largely incidental to others, the effect is to create, in the name of promoting uniformity, a sentencing disparity of the most unreasonable kind.

5

The heart of Mr. Gupta’s offenses here, it bears repeating, is his egregious breach of trust. Mr. Rajaratnam’s gain, though a

product of that breach, is not even part of the legal theory under which the Government here proceeded, which would have held Gupta guilty even if Rajaratnam had not made a cent. While insider trading

may work a huge unfairness on innocent investors, Congress has never treated it as a fraud on investors, the Securities Exchange Commission has explicitly opposed any such legislation, and the Supreme Court has rejected any attempt to extend coverage of the securities fraud laws on such a theory. See, e.g., Chiarella v.

United States, 445 U.S. 222, 232-235 (1980). Prosecution of insider trading therefore proceeds, as in this case, on one or more theories of defrauding the institution (or its shareholders) that owned the information. See, e.g., Dirks v. SEC, 463 U.S. 646, 660-64 (1983); In the eye of the law,

Carpenter v. U.S., 484 U.S. 19, 25-27 (1987).

Gupta’s crime was to breach his fiduciary duty of confidentiality to Goldman Sachs; or to put it another way, Goldman Sachs, not the marketplace, was the victim of Gupta’s crimes as charged. Yet the

Guidelines assess his punishment almost exclusively on the basis of how much money his accomplice gained by trading on the information. At best, this is a very rough surrogate for the harm to Goldman Sachs. The Court is nonetheless mandated to calculate the defendant’s Guidelines range, see 18 U.S.C. § 3553(a)(4)(A), even if,

6

as the Court now holds, the non-guideline sentence that it intends to impose would not vary one whit if the Guidelines calculation was that proposed by the Government, that proposed by the defendant, or anywhere in between. The parties agree that the base offense level for the offense of which Mr. Gupta stands convicted is 8 points, and that 2 points must be added for abuse of trust. To these 10 points must be added

the number of points corresponding to the amount of monetary gain resulting from the offense. Such gain is defined in the official

comment to the pertinent section of the Guidelines as “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.” U.S.S.G. § 2B1.4 cmt. Judge Holwell pointed out in connection with Mr. Rajaratnam’s sentencing, this “phrase is not a model of clarity.” United States v. Rajaratnam, No. 09 Cr. 1184 (RJH), 2012 WL 362031, at *14 (S.D.N.Y. Jan. 31, 2012). Nonetheless, it seems reasonably clear to this Court that the comment limits the calculation to gains made or losses avoided in trades that were based, in whole or in part, on the inside information. In the instant case, however, it is also clear to the Court, both from the jury’s split verdict and from the Court’s own assessment of the evidence, that the trades in question were those made by Rajaratnam and his Galleon funds on September 23, 2008 and As

7

October 24, 2008, directly and immediately as the result of tips from Gupta.1 In the former case, Gupta, late on the afternoon of September 23, tipped Rajaratnam about Warren Buffett’s soon-to-be-announced infusion of $5 billion into Goldman Sachs, whereupon Rajaratnam caused various Galleon funds to purchase large quantities of Goldman stock just before the market closed. When the Buffett investment was

announced the following morning, the stock surged, causing Galleon to realize an immediate gain of $1,231,630. In the latter case, Gupta,

on October 23, tipped Rajaratnam that Goldman Sachs would soon report third quarter losses, whereas many analysts were predicting a profit. On the next day, Rajaratnam sold 150,000 shares of Goldman. Thereafter, as word began to seep out about Goldman’s reduced prospects, the stock began to fall, and when the poor third quarter results were finally made public on December 16, 2008, it fell still further. Based on all the evidence, the Court concludes that it is more likely than not that Rajaratnam, in the absence of Gupta’s tip, would not have caused Galleon to sell its valuable Goldman stock until the morning of December 17, 2008. The tip thus enabled Galleon to avoid losses of $3,800,565. Taken together, therefore, the

If the Court were assessing the evidence without the benefit of the jury’s verdict, it might find that the Government had proved, at least by a preponderance of evidence, that some of additional Galleon trades were also, as the Government alleged, the product of Gupta’s tips. But a decent respect for the jury’s assessment of the evidence, albeit under a higher standard, properly informs this Court’s assessment. 8

1

September and October tip-based trades resulted in an illegal “gain” of $5,032,195. This figure, while large, is less than one-third of the $15,355,409 gain calculated by the Government and endorsed by the Pre-Sentence Report of the Probation Department. But in the

arbitrary world of the Guidelines, this big difference makes little difference. Instead of adding 20 points to Gupta’s Guidelines score, it adds 18 points, still overwhelming all other factors. Although the defendant propounds a number of other theories for still further reducing the gain figure, see Sentencing Memorandum of Rajat K. Gupta at 55-67, the Court rejects these arguments, essentially for the reasons given by Judge Holwell in rejecting similar arguments at the time of the Rajaratnam sentencing, 2012 WL 362031, at *13-15, as well as the additional reasons set forth in the Government’s two sentencing memoranda submitted in this case. Thus,

the Court concludes that the total offense level is 28, the criminal history category is I, and the Guidelines range is 78 to 97 months’ imprisonment. But this Guidelines range does not rationally square with the facts of this case, not only for the reasons already stated but also because it does not take adequate account of the factors this Court is required by law to consider in imposing sentence. The Court

therefore turns to the bedrock of all federal sentencing, section 3553(a) of Title 18, entitled “Factors to be considered in imposing a

9

sentence.”

The very first factor is “the nature and circumstances of

the offense and the history and characteristics of the defendant” (emphasis supplied). Thus, at the very outset, there is presented

the fundamental problem of this sentence, for Mr. Gupta’s personal history and characteristics starkly contrast with the nature and circumstances of his crimes. All the evidence before the Court -– not just the letters written on Mr. Gupta’s behalf but also the objective facts of record –- establish beyond cavil that Mr. Gupta has selflessly devoted a huge amount of time and effort to a very wide variety of socially beneficial activities, such as the Global Fund to Fight AIDS, Tuberculosis and Malaria, the Public Health Foundation of India, the Indian School of Business, the Pratham Foundation (which provides quality education to underprivileged children in India), the Cornell Medical School, the Rockefeller Foundation, and many many more. well summarized in his counsel’s sentencing memorandum, such activities are but illustrations of Mr. Gupta’s big heart and helping hand, which he extended without fanfare or self-promotion, to all with whom he came in contact. While some have suggested that the large volume of poignant letters submitted on Mr. Gupta’s behalf are simply the strategem of a rich, well-connected defendant endeavoring to derail the Court from focusing on his crimes, this is simply not the case, for the facts recited in most of the letters are well documented and, indeed, As

10

undisputed by the Government.

The Court can say without exaggeration

that it has never encountered a defendant whose prior history suggests such an extraordinary devotion, not only to humanity writ large, but also to individual human beings in their times of need. The Guidelines virtually ignore this measure of the man, but here as elsewhere the Guidelines must take second place to section 3553(a), which requires a court to take account of a defendant’s character in imposing sentence. And how could it be otherwise, for on this day of

judgment, must not one judge the man as a whole? But when one looks at the nature and circumstances of the offense, the picture darkens considerably. In the Court’s view, the evidence at trial established, to a virtual certainty, that Mr. Gupta, well knowing his fiduciary responsibilities to Goldman Sachs, brazenly disclosed material non-public information to Mr. Rajaratnam at the very time, September and October 2008, when our financial institutions were in immense distress and most in need of stability, repose, and trust. Consider, for example, his tip to Rajaratnam on September 23, 2008. With Goldman Sachs in turmoil but on the verge

of being rescued from possible ruin by an infusion of $5 billion, Gupta, within minutes of hearing of the transaction, tipped Rajaratnam, so that the latter could trade on this information in the last few minutes before the market closed. equivalent of stabbing Goldman in the back. This was the functional

11

So why did Mr. Gupta do it?

Since motive is not an element

of the offenses here in issue, it did not need to be proved at trial, and so one can only speculate. Having finished his spectacular

career at McKinsey in 2007, Gupta, for all his charitable endeavors, may have felt frustrated in not finding new business worlds to conquer; and Rajaratnam, a clever cultivator of persons with information, repeatedly held out prospects of exciting new international business opportunities that Rajaratnam would help fund but that Gupta would lead. There is also in some of the information

presented to the Court under seal an implicit suggestion that, after so many years of assuming the role of father to all, Gupta may have longed to escape the straightjacket of overwhelming responsibility, and had begun to loosen his self-restraint in ways that clouded his judgment. But whatever was operating in the recesses of his brain,

there is no doubt that Gupta, though not immediately profiting from tipping Rajaratnam, viewed it as an avenue to future benefits, opportunities, and even excitement. Thus, by any measure, Gupta’s

criminal acts represented the very antithesis of the values he had previously embodied. So how does a court balance these polar extremes? In arguing

for a non-guideline sentence in the Pre-Sentence Report, the experienced Senior U.S. Probation Officer Emily Frankelis had this to say: “We believe the defendant’s commission of the instant offenses was aberrant behavior – not aberrant as defined by the U.S.

12

Sentencing Guidelines, but rather as defined by Merriam-Webster: ‘ . . . atypical.’” The Court agrees, and finds that the aberrant nature of Mr. Gupta’s conduct by itself would warrant a non-guideline sentence, even aside from the other factors favoring leniency. in order to find just the right sentence, the Court must also consider two further mandates of section 3553(a): first, “the need for the sentence imposed” to afford specific deterrence, general deterrence, “just punishment,” and the like; and, second, the requirement that any sentence imposed be “sufficient, but not greater than necessary, to comply with [these] purposes.” As to specific deterrence, it seems obvious that, having suffered such a blow to his reputation, Mr. Gupta is unlikely to repeat his transgressions, and no further punishment is needed to achieve this result. different conclusion. General deterrence, however, suggests a As this Court has repeatedly noted in other But

cases, insider trading is an easy crime to commit but a difficult crime to catch. Others similarly situated to the defendant must

therefore be made to understand that when you get caught, you will go to jail. Defendant’s proposals to have Mr. Gupta undertake various

innovative forms of community service would, in the Court’s view, totally fail to send this message. Moreover, if the reports of Mr.

Gupta’s charitable endeavors are at all accurate, he can be counted on to devote himself to community service when he finishes any prison term, regardless of any order of the Court.

13

At the same time, no one really knows how much jail time is necessary to materially deter insider trading; but common sense suggests that most business executives fear even a modest prison term to a degree that more hardened types might not. Thus, a relatively

modest prison term should be “sufficient, but not more than necessary,” for this purpose. There are, however, still other factors set forth in § 3553(a) that the Court must, and has, considered, of which perhaps the most difficult, but most important one, is the concept of “just punishment.” While all the other factors under section 3553 partake

to a lesser or greater degree of policy considerations, “just punishment” taps a deeper vein. Human beings, as social animals, are programmed to respect moral values. This is why people without shame

or guilt are considered psychopaths, and also why violations of the moral order raise such deep passions in the human breast. have come to understand that insider trading is not only a sophisticated form of cheating but also a fundamental breach of trust and confidence, they have increasingly internalized their revulsion for its commission. While no defendant should be made a martyr to As people

public passion, meaningful punishment is still necessary to reaffirm society’s deep-seated need to see justice triumphant. No sentence of

probation, or anything close to it, could serve this purpose. After carefully weighing all these, and other, relevant factors, the Court concludes that the sentence that most fulfills all

14

requirements of section 3553(a) is two years in prison.

Rajat K.

Gupta is therefore sentenced to 24 months’ imprisonment, concurrent on all counts, to be followed by one year of supervised release, on the terms stated from the bench and here incorporated by reference. The otherwise mandatory forfeiture has been waived by the Government, but Court imposes a fine in the sum of $5,000,000. The Court will

defer the determination of restitution for up to 90 days, as permitted by federal law. A formal Judgment embodying these terms

and incorporating this Memorandum by reference will issue shortly. Meanwhile, Mr. Gupta is ordered to surrender to the designated prison by 2 p.m. on January 8, 2013.

SO ORDERED.

JED S. RAKOFF, U.S.D.J. Dated: New York, New York October 24, 2012

15