Summary
finding a thirty-day event window too long where it included days where the price movement was not statistically significant
Summary of this case from United States v. HatfieldOpinion
Criminal No. 3:06CR137 (CFD).
October 31, 2008
Nora R. Dannehy, Acting United States Attorney, District of Connecticut, Eric J. Glover, Assistant United States Attorney, District of Connecticut, New Haven CT, Raymond E. Patricco, Jr., Assistant United States Attorney, Eastern District of Virginia, for United States.
Alan M. Vinegrad, Douglas B. Bloom, Olivia A. Radin, Pamela A. Carter, Covington Burling LLP, Frederick P. Hafetz, Michael S. Chernis, Susan R. Necheles, Tracy E. Sivitz, Hafetz Necheles, Richard L. Spinogatti, Robert J. Cleary, William C. Komaroff, Proskauer Rose, New York, NY, Hope Ivy Hamilton, Covington Burling, LLP, Peter H. White, Mayer Brown LLP, Washington, DC, Jonathan E. Rich, Anthony Pacheco, Keith L. Butler, Proskauer Rose LLP, Los Angeles, CA, Richard A. Reeve, George Gust Kouros, Sheehan Reeve, William F. Dow, III, Jacobs, Grudberg, Belt, Dow Katz, P.C., New Haven, CT, Richard R. Brown, Brown, Paindiris Scott, Hartford, CT, for Robert D. Graham, Christian M. Milton and Christopher P. Garand.
RULING ON LOSS CALCULATION, VICTIM ENHANCEMENT, AND RESTITUTION
The Superseding Indictment charged defendants Ferguson, Graham, Milton, and Monrad with one count of conspiracy, seven counts of securities fraud, five counts of making false statements to the U.S. Securities and Exchange Commission ("SEC"), and three counts of mail fraud, and charged defendant Garand with one count of conspiracy, three counts of securities fraud, three counts of making false statements to the SEC, and three counts of mail fraud. The charges resulted from a fraudulent reinsurance contract between American International Group, Inc. ("AIG") and General Reinsurance Corp. ("Gen Re"). The jury returned a verdict convicting each defendant of every offense with which he or she was charged.
The Court now addresses several sentencing issues common to all five of the defendants. First, a significant factor in reaching the defendants' sentencing guideline range is the amount of loss the fraud caused; the guidelines base offense level of 7 may increase by up to 30 levels as the amount of the loss increases. United States Sentencing Commission, Guidelines Manual, § 2B1.1 (Nov. 2007). Second, up to 6 levels may be added to the base offense level depending on the number of victims that the fraud involved. Id. at § 2B1.1(b)(2). Third, the Court must determine whether an order of restitution is appropriate. On September 25, 2008, the Court held a hearing on the loss calculation, victim impact, and restitution issues.
I. Background
A. Convictions
The defendants' convictions arose from a loss portfolio transfer ("LPT") reinsurance transaction negotiated between Gen Re and AIG. Ferguson, Garand, Graham, and Monrad were Gen Re executives and Milton was an AIG executive. Count One charged all five of the defendants with participating in a conspiracy to commit securities fraud, to make and cause to be made false and misleading statements in reports filed with the SEC, to falsify and cause to be falsified the books and records of a public company, and to commit mail fraud, in violation of 18 U.S.C. § 371.
Counts Two through Five and Eight through Ten charged defendants Ferguson, Graham, Milton, and Monrad with securities fraud, in violation of 15 U.S.C. §§ 78j(b) 78ff. Counts Eight through Ten also charged defendant Garand with securities fraud, in violation of 15 U.S.C. §§ 78j(b) 78ff.
Counts Six and Seven and Eleven through Thirteen charged defendants Ferguson, Graham, Milton, and Monrad with making and causing to be made false and misleading statements with the SEC, in violation of 15 U.S.C. §§ 78m(a) 78ff. Counts Eleven through Thirteen also charged defendant Garand with making and causing to be made false and misleading statements with the SEC, in violation of 15 U.S.C. §§ 78m(a) 78ff.
Counts Fourteen through Sixteen charged all five defendants with mail fraud, in violation of 18 U.S.C. § 1341.
B. Presentence Reports
The presentence reports for all five of the defendants concluded that loss could not reasonably be determined, and recommended that the Court use an alternative measure of gain. See U.S.S.G. § 2B1.1 cmt. 3(B). Finding the gain to be the $5 million Gen Re profited for taking part in the LPT, the presentence reports added 18 levels to each defendant's guidelines calculation. See U.S.S.G. § 2B1.1(b)(1)(J). Concluding that there were more than 50,000 AIG shareholders at the time of the offense, and therefore more than 250 victims, the presentence reports added an additional 6 levels to each defendant's guidelines calculation. See U.S.S.G. § 2B1.1(b)(2)(C). Finally, the presentence reports found that mandatory restitution is applicable under 18 U.S.C. § 3663A.
The presentence reports applied the November 1, 2007 guidelines manual. The defendants have not identified any ex post facto issues with its application.
C. Government Expert Jeffrey Davis
The Government and the defendants dispute the presentence reports' findings on loss and have provided separate expert opinions to support their alternative calculations. The Government's expert, Jeffrey Davis, opined that his best estimate of the LPT-caused loss was between $1.2 billion and $1.4 billion, though he also provided several other loss calculations ranging from $344 million to $598 million. To reach these estimates, Davis performed several statistical event studies. The event studies examined the relationship between news about the LPT fraud and AIG's stock price. Davis' modified, or "leakage," event study — which produced the $1.2 billion to $1.4 billion loss estimate — assumed that from February 14, 2005 to March 15, 2005, the stock market reacted to the continual leak of information related to the LPT-fraud. The $1.2 billion to $1.4 billion, then, is an estimate of the total decline in AIG's stock price for that thirty-day period, after controlling for market and industry factors.
Jeffrey Davis is a senior vice president of Economists Incorporated in Washington, D.C. He holds an M.A. degree in economics from the University of California at Los Angeles, a B.A. degree in economics from the University of California at Riverside, and a J.D. degree from George Washington University in Washington, D.C.
"An event study is an examination of the association between news about a company (good, bad, neutral) and stock price movements. The researcher is examining whether the association between news and share price movements is strong enough to support an inference of, among other things, causation. . . . Typically, event studies work backward from what is ultimately determined to be a fair price, after dissipation of inflation, to determine how much inflation was contained in the price due to fraud during the relevant time frame all the way back to the beginning." Madge S. Thorsen, et al., Rediscovering the Economics of Loss Causation, 6 J. Bus. Sec. L. 93, 109 (2006).
As an alternative to the leakage event study, Davis conducted a series of "standard" event studies. These studies estimated loss by reference to the decline in AIG's stock price on three specific dates, which Davis identified as days on which the market would have reacted to news about the LPT. First, on February 14, 2005, AIG issued a press release revealing that it received subpoenas from the New York Attorney General and the SEC relating to various reinsurance transactions and AIG's accounting for them. Next, on March 14, 2005 news sources indicated that Maurice Greenberg would be forced out of his position as CEO of AIG. Finally, on March 15, 2005, the market reacted to AIG's official announcement that it would replace Greenberg as CEO, that CFO Howard Smith would be taking a leave, and that AIG would delay the filing of its 2004 Form 10-K. The standard event study of all three of these days produced a $544 million to $597 million estimate of loss. The event study incorporating only February 14, 2005 and March 14, 2005 estimated the loss between $344 million and $392 million. The event study of February 14, 2005 and March 15, 2005 estimated the loss between $354 million and $419 million. Finally, the event study of March 14, 2005 and March 15, 2005 estimated the loss between $369 million and $420 million.
The Securities Exchange Act of 1934 requires all publicly-held U.S. companies to file Form 10-Ks, annual comprehensive reports of the company's performance. Thomas Lee Hazen, Federal Securities Law, IV.A (2d ed. 2003).
D. Defense Expert Rene Stulz
The defendants jointly counter with the report of Rene Stulz, who opined that the amount of loss due to the LPT fraud that can actually be calculated is zero. Although the defendants agree that an event study is the appropriate methodology to calculate loss in this type of case in general, they take issue with Davis' modified and standard event studies as applied here. Among the criticisms are: first, the defendants argue that Davis' leakage event study is flawed because it assumes all of the decline in AIG's stock price over the thirty-day period was due to the LPT fraud, as opposed to any number of AIGspecific "confounding factors"; and second, the defendants dispute that each of the three dates on which Davis based his various standard event studies does not meet the legal standard of a "corrective disclosure" from which causation may properly be deduced.
Dr. René Stulz holds the Everett D. Reese Chair in Money and Banking at the Ohio State University and is currently the Director of the Dice Center for Research in Financial Economics. He holds a Ph.D. in economics from the Massachusetts Institute of Technology.
II. Reasonable Estimate of the Loss
The United States Sentencing Guidelines § 2B1.1 calls for graduated increases to a defendant's offense level depending on the amount of loss caused by the offense. Loss is defined as "the greater of actual loss or intended loss," where "actual loss" is the "reasonably foreseeable pecuniary harm that resulted from the offense." U.S.S.G. § 2B1.1 cmt. 3(A). "Pecuniary harm" is the "harm that is monetary or that otherwise is readily measurable in money," and to be reasonably foreseeable, it must be "harm that the defendant knew or, under the circumstances, reasonably should have known, was a potential result of the offense." Id. at § 2B1.1 cmt. 3(A)(i)-(iv).
Importantly, the Guidelines explain that the sentencing court "need only make a reasonable estimate of the loss." Id. at § 2B1.1 cmt. 3(C). Further, because the "sentencing judge is in a unique position to assess the evidence and estimate the loss based upon that evidence[,] . . . the court's loss determination is entitled to appropriate deference." Id. The Guidelines include a nonexhaustive list of factors a court might consider in estimating the loss, including "the approximate number of victims multiplied by the average loss to each victim," and the "reduction that resulted from the offense in the value of equity securities or other corporate assets." Id. at § 2B1.1 cmt. 3(C)(iii)-(iv).
The Second Circuit addressed guidelines loss calculation in criminal securities fraud cases in United States v. Ebbers, 458 F.3d 110 (2d Cir. 2006) and United States v. Rutkoske, 506 F.3d 170 (2d Cir. 2007). While "loss to investors who hold during the period of an ongoing fraud is not easily quantifiable[,] . . . some estimate must be made for Guidelines' purposes, or perpetrators of fraud would get a windfall." Ebbers, 458 F.3d at 127. Nonetheless, the district court's estimate must be grounded in sound economic principles. The Second Circuit in Rutkoske vacated a sentence and remanded for resentencing where the district court's method of calculating loss relied on the difference in stock price between certain dates which had "no particular relevance to the offense conduct, and in fact was three months after the end of the charged conspiracy. This method implicitly attributed the total amount of the decline in value of [the internet company's] shares to [the defendant's] offense conduct." Id. at 178. The court concluded that "[t]he District Court's basic failure at least to approximate the amount of the loss caused by the fraud without even considering other factors relevant to decline in [the company's] share price requires" resentencing. Id. at 180. Although "[m]any factors causing a decline in a company's performance may become publicly known around the time of the fraud and be one cause in the difference in price between X-day and Y-day[,] . . . [l]osses from causes other than the fraud must be excluded form the loss calculation." Ebbers, 458 F.3d at 128; see also United States v. Olis, 429 F.3d 540, 548 (5th Cir. 2005) (vacating and remanding for resentencing where the district court based its finding of loss solely on one witnesses' trial testimony and refused to consider the defendant's expert report that pointed to causes of the stock's decline unrelated to the defendant's personal criminal culpability).
In both Ebbers and Rutkoske, the Second Circuit referenced the economic principles that guide loss calculation in civil cases. In particular, Rutkoske endorsed the "teaching of the Supreme Court in Dura Pharmaceuticals," stating, "`[T]here is no loss attributable to a misrepresentation unless and until the truth is subsequently revealed and the price of the stock accordingly declines' and that portion of a price decline caused by other factors must be excluded from the loss calculation." Rutkoske, 506 F.3d at 179 (quoting Olis, 429 F.3d at 546). The court continued that it saw "no reason why considerations relevant to loss causation in a civil fraud case should not apply, at least as strongly, to a sentencing regime in which the amount of loss caused by a fraud is a critical determinate of the length of a defendant's sentence." Id.
The defendants emphasize this language in Rutkoske, arguing that the Second Circuit adopted in the sentencing context the whole of the civil fraud standard of loss causation. They argue that after Rutkoske it is not sufficient "to show that a news piece about the fraud is followed by a stock price decline," but that according to the civil standard of "corrective disclosure," a disclosure is not "legally cognizable" unless it "reveal[s] to the market the falsity of the prior [misrepresentation]." See Defs.' Jt. Mem. at 16 (citing Lentell v. Merrill Lynch Co., 396 F.3d 161, 175 n. 5 (2d Cir. 2005), cert. denied, 546 U.S. 935, 126 S.Ct. 421, 163 L.Ed.2d 321 (2005)). Based on this reading of "corrective disclosure," the defendants' expert argues that the dates Davis identified for his standard event studies are "irrelevant." See Stulz Surrebuttal at 1. The February 14, 2005 AIG press release, the defendants argue, is not a legally cognizable corrective disclosure because "no reasonable investor reading the [press release] could have understood it to be `correcting' anything, let alone" revealing to the market the falsity of anything specifically related to the LPT. Defs.' Jt. Mem. at 20. The defendants further argue that the March 14, 2005 speculation about Greenberg's resignation was not a corrective disclosure for two reasons: first, because by this date the market already knew of key facts about the LPT scheme; and second, because the market's reaction to the news of Greenberg's possible resignation may have had only little to do with the LPT, and more to do with that fact that AIG would no longer benefit from Greenberg's leadership. Finally, the defendants argue that there was no corrective disclosure in the March 15, 2005 news because the market had already reacted to this news the previous day.
Although the principles that guide civil loss causation should also guide sentencing courts in determining loss, the strict standard of "corrective disclosure" that the defendants support can be incompatible with the guidelines instruction of a "reasonable estimate of the loss." The Second Circuit in Rutkoske acknowledged as much when, immediately after stating that the court saw no reason why "considerations relevant to loss causation in a civil fraud case" should not apply in sentencing, it qualified that "[d]etermining the extent to which a defendant's fraud, as distinguished from market or other forces, caused shareholders' losses inevitably cannot be an exact science," and the "Guidelines' allowance of a `reasonable estimate' of loss remains pertinent." 506 F.3d at 179. "Normally," the court explained, "expert opinion and some consideration of the market in general and relevant segments in particular will enable a sentencing judge to approximate the extent of loss caused by a defendant's fraud." Id. at 180. The Fifth Circuit in Olis acknowledged the same, finding that while courts should be guided by the proposition that "there is no loss attributable to a misrepresentation unless and until the truth is subsequently revealed and the price of the stock accordingly declines[,] . . . given the time and evidentiary constraints on the sentencing process, the methods adopted in these cases are necessarily less exact than the measure of damage applicable in civil securities litigation." 429 F.3d at 546, 547. The Second and Fifth Circuits instructed sentencing courts to adopt from the civil context the "considerations relevant" to civil loss causation, not necessarily and always the civil fraud standard. Namely, sentencing courts must calculate guidelines loss against the backdrop of an efficient market, and must account for confounding factors that may also have affected a company's stock price.
III. Loss Calculation
A. Leakage Event Study
The Court rejects Davis' leakage event study estimate of loss of $1.2 billion to $1.4 billion. Although some academics and courts have endorsed the general concept of leakage event studies that "may consider `event windows' or several days over time," when a "day where movement is not statistically significant may be part of a series of days which, taken together, are highly statistically significant," Thorsen, 6 J. Bus. Sec. L. at 111-12, in this case the Government has not justified sufficiently Davis' 30-day event window or accounted for non-LPT confounding factors that may have affected AIG's stock price during that time. See Defs.' Jt. Mem. at 35. Sentencing courts cannot fail to "at least to approximate the amount of the loss caused by the fraud without even considering other factors relevant to decline in [the company's] share price." Rutkoske, 506 F.3d at 180. Because Davis' leakage study attributes all non-market and non-industry related decline in AIG's stock price to the LPT fraud without accounting for other factors that may have contributed to that decline, it is not a reasonable estimate of the loss.
The Government cites several cases in which courts acknowledged the concept of evaluating decline in stock price over a period of more than one day. See Swack v. Credit Suisse First Boston, 383 F.Supp.2d 223, 243-44 (D.Mass. 2004); Fogarazzo v. Lehman Bros., Inc., 341 F.Supp.2d 274, 292 (S.D.N.Y. 2004); Danis v. USN Communications, Inc., 73 F.Supp.2d 923, 943 (N.D.Ill. 1999). The courts' discussion of leakage in these cases was confined to the appropriate standard for pleading securities fraud claims.
The defendants cited several confounding factors during the 30-day event window: February 15, 2005, news that two AIG executives pleaded guilty for their alleged roles in bid-rigging; February 24, 2005, news that AIG was the insurer for the largest product recall in UK retailing history; February 25, 2005, the New York Times published a negative article about AIG's relationship with C.V. Starr; February 28, 2005, the Financial Times reported that then-Attorney General Eliot Spitzer had commenced an investigation into AIG's relationship with C.V. Starr, as well as a separate new investigation into AIG's relationship with Coral Re; March 2, 2005, the Financial Times provided new details concerning the Coral Re investigation, revealing that it related to allegations that AIG had used finite reinsurance from Coral Re to smooth its earnings.
B. Standard Event Study
While Davis' leakage event study is not a reasonable estimate of the loss caused by the LPT fraud, Davis' standard event study of February 14, 2005, March 14, 2005, and March 15, 2005 does result in a reasonable estimate of the loss between $544 million and $597 million. On each of these days, new and significant information related to the LPT fraud was disclosed, and there was a statistically significant price reaction in AIG's stock price.
The Court adopts Mr. Davis' factual findings as to the event study of February 14, 2005, March 14, 2005, and March 15, 2005.
February 14, 2005
On February 14, 2005, AIG announced that it received subpoenas from the New York Attorney General and the SEC relating to "non-traditional insurance products and certain assumed reinsurance transactions and AIG's accounting for such transactions." Gov't Attach. 25; see also Gov't Trial Ex. 243. AIG's stock price dropped by a statistically significant amount on this day. The defendants argue that this disclosure is not the proper basis for an event study because it did not "mention [] the LPT and says absolutely nothing about AIG's financial statements or the charged scheme." Defs.' Jt. Mem. at 20.
Examining news and analyst reaction to the February 14 disclosure reveals that even though the LPT transaction was not specifically named or explained in the announcement, the market associated the disclosure with the LPT fraud. For example, Charlene Hamrah, AIG's director of investor relations, testified at trial that on and after the February 14 disclosure, she received calls from investors and analysts who were concerned about the Gen Re transaction. Trial Tr. Vol. II, 302-03; see also Gov't Attach. 24 (analyst Jay Cohen's testimony that the February 14 news was important to his clients). An article published in the Wall Street Journal on February 18 also demonstrates that investors associated the subpoenas with the LPT fraud; specifically, it reported in response to the subpoenas that
Although this article and others discussed in this section of the opinion appeared on dates later than the relevant date of the event study, they reflect the understanding of and reaction to the events on those three dates.
[r]egulators are scrutinizing a deal struck between American International Group Inc. and Berkshire Hathaway Corp.'s General Reinsurance unit in late 2000 or early 2001 to determine if it was aimed at making the giant insurer's reserves look healthier than they were. . . . Specifically, regulators are looking at a transaction — in late 2000 or early 2001 — in which AIG added hundreds of millions of dollars of claims reserves to its balance sheet, in return for cash premium payments.
Gov't Attach. 14 (Ian McDonald, "Regulators Probe an AIG Pact with General Re," Wall Street Journal, February 18, 2005, at B. 2). A February 21 article in Barron's explained that
[c]ompany insiders were reportedly stunned at the specificity of the information requests, which centered on two reinsurance deals AIG reached with Berkshire Hathaway's General Re unit in late 2000 and early 2001. . . . Specifically, investigators are said to be looking at two reinsurance deals with General Re that allowed AIG to beef up its fourth-quarter 2000 and first quarter 2001 reserves with a total of some $200 million in no-risk finite reinsurance transactions with Gen Re. In effect, investigators are looking at whether AIG borrowed the claimed loss reserves from Gen Re, boosting the reserves carried on its balance sheet.
Gov't Attach. 39 (Jonathan R. Laing, "Turning Up the Heat on AIG," Barron's, February 21, 2005, at 18). These accounts indicate that market analysts and investors regarded the February 14 subpoenas as LPT-related news, and the statistically significant drop in AIG's stock price on that day can be attributed to the fraud.
March 14, 2005
The March 14, 2005 event was the press reports that AIG would replace Maurice Greenberg as CEO. The market reacted with a statistically significant drop in AIG's stock price. The defendants again argue that this news "did not correct anything about the LPT, or otherwise reveal the truth about AIG's loss reserves." Defs.' Jt. Mem. at 21. They argue that by this date, the market already knew of the LPT and Greenberg's role in it, and further, that any decline in stock price that resulted from speculation over his departure had more to do with the fact that "investors were unhappy that Mr. Greenberg was forced to resign in light of his lengthy and successful reign as the head of AIG." Id. at 22.
At 9:42 a.m. on March 14, MarketWatch reported that Maurice Greenberg told CNBC he would step down the following week. Gov't Attach. 34 (Steve Gelsi, "Greenberg Tells CNBC He'll Step Down," MarketWatch, March 14, 2005). The media continued to show interest in his departure from AIG throughout the day. For example, at 3:24 p.m., the Associate Press reported that Greenberg "could be stepping down." Gov't Attach. 32 (Michael J. Martinez, "CEO Changes, Merger News Boosts Stocks," Associated Press, March 14, 2005). Even later, the Financial Times reported that there "are reports tonight that . . . Greenberg is stepping down as CEO. Greenberg is expected to resign as chief executive tonight." The reason the Financial Times cited for his departure is significant; it reported that "[a]nalysts say regulatory scrutiny is at the center of the change." See Gov't Attach. 33 (Paul Kangas Jeff Yastine, "Nightly Business Report," Financial Times, March 14, 2005). Thus, Greenberg's being forced out of AIG was viewed by analysts as confirmation of his role in the LPT fraud and the seriousness of the LPT fraud, and the statistically significant drop in AIG's stock price on March 14 can be attributed to this news.
March 15, 2005
At 10:01 p.m. on March 14, 2005, AIG officially announced first, that Greenberg would be replaced as CEO, second, that CFO Howard Smith would be taking leave, and third, that AIG would delay issuing its 10-K. Gov't Attach. 27. March 15, 2005, on which there was a statistically significant drop in AIG's stock price, was the day the market would have responded to the previous night's announcement. The defendants argue that this date cannot be considered in the event study because the market had already learned about and reacted to Greenberg's departure on March 14, and that the event study failed to account for confounding factors in the news about Howard Smith and the delayed 10-K.
The reporting of AIG's March 14 announcement clearly drew the connection between Greenberg's departure and the LPT fraud. For example, a report in the San Jose Mercury News discussed how Greenberg would be leaving "amid concern over a rising number of regulatory inquiries at the financial-services titan he built over nearly four decades," and that
AIG currently is under investigation by New York Attorney General Eliot Spitzer, federal prosecutors, and the Securities and Exchange Commission into the use of so-called finite insurance, or financial reinsurance, which critics say could be used to manipulate earnings. The transaction being probed took place between AIG and Berkshire Hathaway's General Reinsurance unit four years ago and apparently was intended to shore up AIG's reserves. Reports have said Maurice Greenberg was personally involved.
See Defs.' Ex. 11. (Eileen Alt Powell, "Insurance Giant AIG Ousts CEO Greenberg; Shake-Up Occurs As Firm Faces Probes," San Jose Mercury News, March 15, 2005, at 1C). Although the market did have information about the LPT fraud, Greenberg's role in it, and Greenberg's resignation prior to March 15, on that day the market continued to react to further confirmation of this news and AIG's actions of the previous evening.
Furthermore, Howard Smith's taking leave and the delayed 10-K were not confounding factors separate from the LPT. Rather, they were related to the LPT. Howard Smith was involved in the LPT fraud, see Gov't Ex. 31, and in its March 14 announcement, AIG explained the reasoning behind the delayed 10-K as follows: "This action is the result of the management changes described above as well as the Company's ongoing internal review of the accounting for certain transactions, which review was commenced in connection with previously announced regulatory inquiries." Gov't Attach. 27; see also Defs.' Ex. 12 (David Pilla, "AIG's Greenberg Steps Aside; 10-K Filing Delayed as Regulatory Probes Intensify," Best Insurance News, March 15, 2005). Greenberg's resignation, Howard Smith's taking leave, and the delayed 10-K were announced as and were perceived by the market to be LPT-related, and therefore the March 15 decline in AIG's stock price can be attributed to the fraud.
In conclusion, the Court finds that Davis' standard event study of February 14, 2005, March 14, 2005, and March 15, 2005, provides a reasonable estimate of the loss between $544 million and $597 million. Because even the lowest end of the estimate, $544 million, is more than the highest level of loss specified in U.S.S.G. § 2B1.1, it is unnecessary to identify a specific number within the range. U.S.S.G. § 2B1.1(b)(1)(P). Thirty levels will be added to each defendant's guideline calculation for a loss of more than $400 million. Id.
Davis' estimate is based on "damaged shares" held only by banks, insurance companies and mutual funds. His calculation, then, likely underestimates the loss by only including losses to institutional investors.
IV. Number of Victims
In addition to loss, the sentencing guidelines call for an increase in offense level based on the number of victims. U.S.S.G. § 2B1.1(b)(2). The guidelines define victim as "any person who sustained any part of the actual loss." Id. at § 2B1.1 cmt. 1; see also United States v. Abiodun, 536 F.3d 162, 169 (2d Cir. 2008) (concluding that when a court makes a finding as to actual loss, a victim enhancement must be based on the guidelines loss calculation). At oral argument the defendants agreed that should the Court find actual loss could be calculated, some victim enhancement would apply.
At issue is whether that enhancement is 4 levels, for 50 or more victims, or 6 levels, for 250 or more victims. Davis' event study incorporated 139,611,930 "damaged shares" that he identified as held by 154 institutional investors, including 103 mutual funds. The Government argues that because these mutual funds were comprised of thousands of investors, the number of victims of the LPT fraud clearly surpasses 250. The Court agrees with Davis that the individual shareholders of mutual funds that were invested in AIG during the fraud period and that were identified by Davis are victims as defined by the sentencing guidelines, as each suffered pecuniary harm in relation to their share of the mutual fund in which they were invested. Therefore, 6 levels will be added to each defendant's guideline calculation for 250 or more victims of the fraud.
The Government provided an example of one mutual fund holding "damaged shares" of AIG that by itself had 250 shareholders. Gov't Resp. to Inquiries by the Court at 6.
V. Restitution
The defendants were convicted of offenses defined under both Title 15 and Title 18 of the United States Code. The Mandatory Victim Restitution Act of 1996 ("MVRA") provides that sentencing courts shall award restitution for any offense against property under Title 18 "in which an identifiable victim or victims has suffered a physical injury or pecuniary loss." 18 U.S.C. § 3663A(a)-(c). The Victim and Witness Protection Act of 1982 ("VWPA") gives sentencing courts the discretion to award restitution for other offenses. 18 U.S.C. § 3663.
Counts two through twelve of the indictment charged the defendants with securities fraud under Title 15. The first count of the indictment, conspiracy to commit securities fraud, and counts fourteen through sixteen, mail fraud, charged the defendants under Title 18.
As the Court finds that an order of restitution is inappropriate, it does not address the defendants' arguments that neither the Title 15 offenses (securities fraud) nor the Title 18 offenses (conspiracy and mail fraud) are offenses against property encompassed by the MVRA.
The MVRA, however, instructs sentencing courts that an order of restitution is not mandatory if fashioning such an award is unduly complicated. More specifically, the MVRA states:
This section shall not apply in the case of an offense [against property under Title 18] if the court finds, from facts on the record, that — (A) the number of identifiable victims is so large as to make restitution impracticable; or (B) determining complex issues of fact related to the cause or amount of the victim's losses would complicate or prolong the sentencing process to a degree that the need to provide restitution to any victim is outweighed by the burden on the sentencing process.18 U.S.C. § 3663A(c)(3). The Senate Report that accompanied the MVRA's enactment sheds light on Congress' intent "that courts order full restitution to all identifiable victims of covered offenses, while guaranteeing that the sentencing phase of criminal trials do not become fora for the determination of facts and issues better suited to civil proceedings." S. Rep. No. 104-179, at 18 (1995), as reprinted in 1996 U.S.C.C.A.N. 924, 931. "Moreover," the committee wrote, the Act "requires that there be an identifiable victim who suffers a physical injury or pecuniary loss before mandatory restitution provisions would apply. In all cases, it is the committee's intent that highly complex issues related to the cause or amount of a victim's loss not be resolved under the provisions of mandatory restitution." Id. at 19.
The Second Circuit has echoed Congress' intent in several cases in which victims were not actually identified or complex factual issues made a mandatory restitution order impracticable. In United States v. Reifler, the court explained that "Congress plainly intended that sentencing courts not become embroiled in intricate issues of proof, as it provided that the MVRA is to be inapplicable if the court finds that the determination of complex factual issues related to the cause or amount of the victims' losses would unduly burden the sentencing process." 446 F.3d 65, 136 (2d Cir. 2006) (citing 18 U.S.C. § 3663A(c)(3)(B)); see also United States v. Rigas, 409 F.3d 555, 563 (2d Cir. 2005) (upholding the district court's determination that the exceptions to the MVRA applied where the victims were too numerous and that the complexity of factual issues would extend the sentencing process inordinately); United States v. Catoggio, 326 F.3d 323, 326-27 (2d Cir. 2003). "This provision reflects Congress's intention that the process of determining an appropriate order of restitution be `streamlined,' and that the restitution `determination be made quickly.'" Reifler, 446 F.3d at 136 (citing S. Rep. No. 104-179, at 20). The MVRA "is clear that restitution can only be imposed to the extent that the victims of a crime are actually identified," and it "requires courts to order restitution in the full amount of the victims' losses." Catoggio, 326 F.3d at 327-29 (citing 18 U.S.C. §§ 3663A(c)(1)(B), 3664(f)(1)(A)). In Catoggio, the Second Circuit found the district court's order of restitution violated the MVRA because the court did not actually identify the victims to whom restitution should be paid. Id. at 328.
Although not all securities fraud cases will fall within these exceptions to the MVRA, this case is sufficiently complicated so as to make the MVRA inapplicable. Actually identifying all of the victims of the LPT fraud, if even possible, would severely complicate and prolong the sentencing process. See 18 U.S.C. § 3663A(c)(1)(A), (B). While Mr. Davis' event study identified 154 institutional holders of LPT-damaged shares, in order to award restitution, this court would have to identify the individuals who had an interest in those institutional holdings, as well as all of the victims of the LPT fraud not accounted for in the Davis event study. Fashioning such an order of restitution would require a level of detail and precision that was not necessary to calculate a reasonable estimate of loss for the loss calculation. The Government has admitted that it "does not presently know the identity, let alone the location, of all or even most of the persons who were shareholders of AIG in the relevant period, and ascertainment of that information would be highly impracticable." Gov't Resp. at 22. These complicated issues are better resolved in the several pending civil proceedings against AIG and these defendants. See In re American International Group, Inc. Sec. Litig., No. 04-CV-8141 (JES) (S.D.N.Y.) (filed Oct. 15, 2004) (class action); In re American International Group, Inc. Derivative Litig., No. 04-CV-8406 (JES) (S.D.N.Y.) (filed Oct. 25, 2004) (shareholder derivative suit); American International Group, Inc. Consolidated Derivative Litig., C.A. No. 769 (VCS) (Del. Ch.) (filed Oct. 21, 2004) (shareholder derivative suit); S.E.C. v. Ferguson et al., No. 06-CV-0778 (LAS) (S.D.N.Y.) (filed Feb. 2, 2006).
As the Government explained, "because shares of AIG stock largely are bought and sold by stock brokers, precisely identifying the total number of shareholders that traded on these days is a prohibitive task. Typically, stock brokers hold shares on behalf of their individual customers in a `street name' — either their own name or the name of a convenient nominee — in order to facilitate the efficient transfer of the shares. While the brokers know the names of their individual customers, and can identify their total number, AIG typically cannot. Obtaining such information from multiple, individual stock brokers would be an exceedingly time-consuming and costly process." Gov't Resp. to Inquiries by the Court at 3.
The Crime Victims' Rights Act ("CVRA") enumerates rights that crime victims or the Government may raise "in the district court in which a defendant is being prosecuted for the crime." See 18 U.S.C. § 3771(a), (c)(1), (d)(1), (d)(3). The district court "shall take up and decide any motion asserting a victim's right forthwith." Id. at § 3771(d)(3); see also Rigas, 409 F.3d at 561-62. Among the crime victims' rights is the right "to full and timely restitution as provided in law." Id. at § 3771(a)(6). The CVRA specifies that when "the court finds that the number of crime victims makes it impracticable to accord all the crime victims [restitution], the court shall fashion a reasonable procedure to give effect to this chapter that does not unduly complicate or prolong the proceedings." Id. at § 3771(d)(2). The Court finds restitution impracticable as provided in law under the MVRA and the VWPA, id. at § 3771(a)(6), and it further finds that the several pending civil proceedings afford a reasonable procedure to give effect to the CVRA, id. at § 3771(d)(2).
Finally, for the same reasons, the Court finds discretionary restitution is also inappropriate under the VWPA. Similar to the MVRA, the VWPA provides that "[t]o the extent that the court determines that the complication and prolongation of the sentencing process resulting from the fashioning of an order of restitution under this section outweighs the need to provide restitution to any victims, the court may decline to make such an order." 18 U.S.C. § 3663(a)(1)(B)(ii).
VI. Conclusion
For the above reasons, 30 levels will be added to the defendants' guidelines for loss of more than $400 million, 6 levels will be added for more than 250 victims, and no restitution will be ordered. The parties may submit additional materials concerning sentencing of the individual defendants within 21 days.