CHARLES R. BREYER United States District Judge

ORDER DENYING MOTIONS FOR NEW TRIAL AND JUDGMENT OF ACQUITTAL

Following a fiercely litigated, 29-day jury trial, Sushovan Hussain was convicted of 14 counts of wire fraud, one count of conspiracy to commit wire fraud, and one count of securities fraud for falsely representing the finances of Autonomy Corporation PLC ("Autonomy") in the run-up to Autonomy's acquisition by Hewlett-Packard ("HP"). The government presented evidence at trial establishing that Hussain and his co-conspirators represented Autonomy's revenues as being higher than they actually were for years, inflating the company's perceived value in an effort to deceive the market and, ultimately, potential acquirers. Hussain moves for a new trial, arguing that the Court erred in a number of its rulings, particularly its decisions to exclude evidence regarding Hewlett-Packard's response to the alleged fraud after it acquired Autonomy. Hussain also moves for a judgment of acquittal, contending that the evidence at trial was insufficient to sustain a conviction. Both motions are denied.

I. BACKGROUND

A. Government Allegations and Lead-Up to Trial

The government brought an indictment against Hussain in this Court on Nov. 10, 2016, see Indictment (dkt. 1), and superseded on May 4, 2017, see Superseding Indictment ("SI") (dkt. 52). According to the Superseding Indictment, Hussain, a resident of the United Kingdom, had served as Chief Financial Officer of Autonomy from approximately June 2001 to November 2011, leaving soon after HP agreed to acquire the company for approximately $11 billion in August 2011. SI ¶¶ 4, 8. Hussain was responsible for preparing Autonomy's quarterly financial reports, and for certifying that they had been independently audited and complied with UK regulations for publicly traded companies. SI ¶¶ 5, 13.

Autonomy claimed to be a "pure software" business with gross margins ranging from 85-90%. SI ¶¶ 15-16. The government alleged, however, that these representations were false—part of a calculated scheme by Hussain and other Autonomy principals to deceive the market as to Autonomy's true financial picture and growth prospects. SI ¶ 19. Autonomy's principals sought to ensure that the company exceeded revenue projections in order "to artificially increase and maintain the share price of Autonomy securities to, among other things, make Autonomy attractive to potential purchasers." SI ¶ 20. They used a variety of stratagems to improperly recognize revenue; falsified the books and records; otherwise inflated revenue by secretly selling hardware at a loss; lied to market analysts and the company's independent auditors; and intimidated those who questioned its financial practices and performance. SI ¶ 22.

Hussain moved to dismiss on Aug. 25, contending among other things that the wire-fraud charges represented an impermissible extraterritorial application of the governing statutes, and that he could not be held liable for securities fraud because he never directly communicated with HP shareholders. Mot. to Dismiss (dkt. 113). The Court rejected these arguments and denied the motion. Order Denying MTD (dkt. 129). Following extensive pre-trial litigation, trial got underway on Feb. 27, 2018.

B. Government's Case-in-Chief

At trial, the government portrayed Autonomy as a "middle-of-the-pack tech company with a decent piece of search software" that cooked its books to make it appear as though it had achieved extraordinary growth during the Great Recession, a time when most of its peers' revenues were either flat or shrinking. Sometime in late 2008, Hussain and his co-conspirators set out to improperly inflate Autonomy's revenue figures to influence market analysts' projections about its growth and estimates of its market value, in the hopes of an eventual acquisition and a big payday. (Hussain netted $16 million from the HP acquisition.) In Hussain, the conspiracy also included CEO Mike Lynch, as well as Christopher "Stouffer" Egan, Peter Menell, Andy Kanter, and Steve Chamberlain, all executives at Autonomy.

Hussain and his co-conspirators began to "cut corners" by recognizing deals with value-added resellers before a sale had been made to the end user, essentially pulling revenue from future quarters into the current quarter. They used several improper techniques to pad revenue: back-dating contracts; "selling" software to resellers when it never actually expected to be paid back; and re-selling hardware at a loss while representing to auditors, market analysts, and ultimately HP that it never sold hardware separately from software. Autonomy, in other words, was "buying its own revenue."

Autonomy set revenue targets for each quarter based on analyst expectations, and would use a combination of these techniques to fill the gap between those expectations and its actual revenue, to the tune of 10 to 25 percent each quarter. The government demonstrative below presents a picture of how this worked, separating out (1) accurately reported revenue, (2) hardware sales that were misrepresented to the market, and (3) revenue that should not have been recognized at all ("adjustments").

Image materials not available for display. In the government's telling, this made it appear as though Autonomy's core software business was growing when in fact revenues were stagnant.

Government witnesses explained that growth rate is a particularly important determinant of market valuation for a tech company, and ultimately of the price that company can command in an acquisition. Accordingly, falsely reporting increases in revenue will tend to lead to significant exaggerated valuations. Meanwhile, representing that a company sells only software when in fact it sells a significant amount of hardware will also tend to seriously exaggerate the company's value—even if the revenue figures themselves are accurate—because the profit margins on software generally outpace hardware margins by at least an order of magnitude. That is, a business that returns 50 cents per dollar to profit is much more valuable than one that returns only 5 cents per dollar.

The government's trial evidence may be grouped into three basic categories: (1) evidence regarding Autonomy's misrepresentations about its financial state, including (a) its improper accounting and (b) its inflation of revenue through hardware sales and its concealment of those sales; (2) evidence regarding the significance of these misrepresentations; and (3) evidence tying Hussain himself to these misrepresentations.

1. Evidence regarding Autonomy's misleading representations

a. Improper accounting

The government put on extensive evidence regarding Autonomy's allegedly fraudulent accounting. Because Autonomy was a publicly traded company, it was required to prepare audited financial statements, and thus had to deceive its auditor, Deloitte, into believing that any revenue it recognized comported with the applicable accounting standard. Moreover, when the company sold to HP, Autonomy's books had to match the amount of money transmitted. Accordingly, it had to do more than simply lie about the amount of money it was receiving in order to successfully pull off the scheme.

Autonomy developed a stable of reliable counterparties, many of them "value-added resellers," to whom it would turn at the end of a quarter if it anticipated falling short of its revenue target. Frequently, Autonomy would represent to its auditors that it had sold software to a reseller that was in turn planning to sell the software to an end user. In fact, however, Autonomy generally did not actually expect to be paid by the reseller, or only expected to be paid if the deal ultimately closed with the end user. The resellers had no role in negotiating with the end user; Autonomy paid the resellers a commission solely for allowing it to recognize revenue early. Many of these contracts were signed after the end of a quarter, but back-dated to make it appear as though they had actually been completed in the previous quarter—another reason that it was improper to recognize revenue.

In addition to deceiving its auditors, Autonomy also had to square its books with the actual cash on hand. Sometimes, it would collect money on these deals from the ultimate end user instead of the reseller. Sometimes, it would buy worthless software from the reseller to give it the money to pay Autonomy. And sometimes it would simply write off the amounts owed. These write-offs occurred in the "dark period" between Aug. 18 and Oct. 3, 2011—the date when HP announced it would acquire Autonomy, and the date the deal actually closed, respectively.

The government also introduced evidence that Autonomy made reciprocal transactions on which it should not have recognized revenue with its counterparties—agreeing to buy software it did not need if its counterparty in turn bought Autonomy software for a slightly lower price. And on at least one occasion, Autonomy misrepresented the sale of a license as the sale of a good to its auditors, enabling it to recognize revenue earlier than it otherwise could have.

A detailed description of the deals that the government argued involved improper accounting follows.

• Capax EDD deals :

Capax was one of several resellers with which Autonomy repeatedly did business. Capax was ostensibly an e-discovery vendor that would use Autonomy software to perform discovery work for end users. However, during the period that Capax worked with Autonomy, it lacked the capability to actually do any e-discovery work. Instead, Capax's real value to Autonomy was that it would agree to sign contracts that Autonomy could use to convince its auditors that it had made

a legitimate sale. The parties had an understanding that Capax would make a 10 percent commission on revenue it allowed Autonomy to recognize.

Autonomy's first fraudulent deal with Capax came in the first quarter of 2009, when Autonomy purported to sell Capax a five-year license to use Autonomy software in the provision of electronic-data discovery ("EDD") services, as well as a one-year right to sell Autonomy services to third parties, for $7.5 million. A senior architect of Autonomy products testified that he did not know of any excess e-Discovery work that Autonomy had from 2008 to 2011, and that, to his knowledge, Autonomy never actually outsourced such work. Stouffer Egan, who negotiated the deal for Autonomy, testified that Capax and Autonomy had an oral side agreement for Autonomy to underwrite the cost of the software, regardless of whether Capax did anything of value for Autonomy. The purpose of the deal was to enable Autonomy to recognize revenue early on e-discovery work Autonomy was performing for other companies. Autonomy eventually re-routed payments from its customers to Capax to pay it back for its software purchases. Every purchase order Capax signed as part of the deal to get paid for this work was false. Autonomy signed two similar agreements with Capax in the fourth quarter of 2009 and the first quarter of 2011, for $4 million and $1.6 million, respectively.

Autonomy deceived its auditors about the deals, concealing its oral side

agreements with Capax. It also concealed that the 2011 deal was back-dated: while the end of the quarter came on March 31, Hussain first proposed the deal to Baiocco on April 6. Autonomy's revenue recognition was improper on all three of these deals.

• Morgan Stanley hardware deal : In the second quarter of 2009, Autonomy recognized revenue on a $6 million sale of hardware to Morgan Stanley. However, it did not actually deliver the hardware. Hussain lied to Autonomy's auditors, confirming to them that the hardware had been dispatched or delivered when it had not. Autonomy should not have recognized revenue on the sale.

• Morgan Stanley software license : In the fourth quarter of 2009, Autonomy recorded an eight-year license with Morgan Stanley, on which revenue should have been recognized over a period of years, as an immediate sales of goods, allowing it to inflate revenues in earlier quarters. Autonomy represented to its auditors that it was recognizing the deal as a sale of goods because it was including a software product called SPE Basic in the package. In a call with market analysts, Lynch, the CEO, described SPE Basic as the "key winning differentiator" in closing the deal. However, the evidence at trial tended to establish that the software was

included gratuitously and did not form part of the consideration for deal.

• MicroLink/MicroTech deal : In the fourth quarter of 2009, Autonomy purchased a company called MicroLink. Four days before the acquisition closed, it agreed to sell $9.5 million in software to a company called MicroTech for resale to another company called DiscoverTech. This was not an arm's-length transaction, however, but was instead part of the MicroLink acquisition; Dave Truitt had an ownership stake in all three companies, MicroLink, MicroTech, and DiscoverTech. Truitt said he would have had no interest in purchasing the software had the MicroLink acquisition not been contingent on it, and that he did not in fact use the software. Nevertheless, Autonomy falsely represented to its auditors that it was "not aware of any . . . connection" between MicroLink and MicroTech. Had the auditors understood the relationship between these companies, they would have advised Autonomy not to recognize revenue on the deal, but would instead have treated it as part of the MicroLink acquisition.

• MicroLink/DiscoverTech deal : In the same quarter, MicroLink purported to buy $2.3 million in software from Autonomy for resale to DiscoverTech. Revenue should not have been recognized, however, because the deal was back-dated, was

not properly approved by MicroLink, and apparently was never expected to actually close; MicroLink never asked DicoverTech for payment. Alan Rizek, the CFO of MicroLink, said that he first heard of the deal when Hussain called him after the quarter had ended and proposed it to him, and that Hussain recognized the deal even though Rizek did not have the authority to bind MicroLink. Autonomy misrepresented the deal to its auditors.

• FileTek direct deals : The government also introduced evidence of transactions with end users on which Autonomy recognized revenue while simultaneously paying the end user to acquire valueless software. These were reciprocal deals that should have been accounted for as payments to end users, not receipts of revenue. Autonomy completed two of these deals with a company called FileTek—recording an $8 million sale of software in the fourth quarter of 2009 and an $8.5 million software sale in the first quarter of 2010. In the same quarters, Autonomy paid FileTek $10.3 million and $11.6 million, respectively, for its StorHouse software. However, Autonomy had no use for this software. It was merely paying FileTek to enable it to recognize revenue. Autonomy deceived its auditors by telling them the transactions were not linked. Because no value was actually exchanged, Autonomy should not have recognized revenue on these deals.

• Sales Consulting / Poste Italiane deal : In the fourth quarter of 2009, Autonomy purported to sell $2.2 million in software to Sales Consulting, with Poste Italiane as the end user. This deal was problematic for two reasons. First, the contract was back-dated. Second, it was not clear that Sales Consulting would actually be able to pay for the software. Autonomy in fact never collected on the deal, and wrote it off. Revenue should not have been recognized.

• MicroTech/Vatican deal : Autonomy sought to license software to the Vatican as part of a project to digitize the latter's libraries. Negotiations stalled, however. When the deal failed to close in the first quarter of 2010, Autonomy, after the quarter had ended, asked MicroTech to take the deal as a reseller, with the Vatican as the intended end user. Autonomy and MicroTech purported to agree to an $11 million software license in the first quarter of 2010. However, revenue should not have been recognized: the contract was back-dated, and the parties understood that MicroTech was not actually on the hook for paying. (MicroTech eventually obtained other funds from Autonomy to pay for what it owed on the Vatican deal.)

Autonomy concealed these facts from its auditors.

• Auxilium/Vatican deal : In the first quarter of 2010, Autonomy purported to sell $1.9 million in software to the Vatican through a reseller called Auxilium. However, as in other deals, it never appeared likely that Autonomy would be able to collect, particularly given that the Vatican never committed to pay for the software. The deal was ultimately written off. In addition, the contract was back-dated, signed 11 days after the quarter had ended.

• Capax/FSA deal : Autonomy recognized $4.285 million in the first quarter of 2010 on a sale of software licenses to Capax for resale to the Financial Services Authority ("FSA"). However, Autonomy was in direct negotiations with the FSA on the deal, and never delivered the software to Capax. Capax and Autonomy had an understanding that Capax would make a 10% commission for allowing Autonomy to recognize revenue through reseller deals like this one.

• FileTek/VA deal : Autonomy recognized $10 million of revenue in the third quarter of 2010 on a sale of software to FileTek, with the Veterans Administration (VA) as the supposed end user. It had become clear in September of that year that a deal Autonomy was negotiating directly with the VA would not close by the end of the quarter. Instead, Autonomy purported to sell the software to FileTek for resale to

the VA at a later date. However, FileTek was operating with the understanding that it would not have to pay if it was unable to collect from the VA, and only did the deal because it stood to make a commission for allowing Autonomy to recognize revenue. The deal indeed never closed with the VA, and Autonomy ended up making a third purchase of the useless StorHouse software, this one for $11.6 million, to enable FileTek to pay Autonomy. Autonomy should not have recognized revenue.

• Capax/Amgen deal : Autonomy purported to sell a $9 million software license to Capax for resale to Amgen in the third quarter of 2010. However, again, Capax was not actually acting as a value-added reseller. Autonomy did not get Capax involved until the very end of the quarter, when it became clear that Autonomy would not be able to close the deal directly with Amgen; Capax never intended or attempted to sell the software to Amgen; Autonomy never sought to collect from Capax; Autonomy continued to try to sell to Amgen; and Autonomy and Capax later signed and back-dated a side letter that sent the software originally intended for Amgen to Bank of America (see below). It did not disclose these facts to its

auditors. Autonomy should not have recognized revenue on the deal.

• DiscoverTech / Capax / MicroTech / Bank of America deal : Autonomy realized in the fourth quarter of 2010 that a planned sale of software to Bank of America was not in fact going to close in that quarter. After the quarter ended, still seeking to recognize the revenue in the fourth quarter, Autonomy purported to make three separate transactions with value-added resellers—DiscoverTech, MicroTech, and Capax—to sell data-hosting services to Bank of America. While the deal eventually sold through to Bank of America, Autonomy should not have recognized the revenue in the fourth quarter of 2010: the agreements were back-dated, and Autonomy's transactions with the resellers had no economic substance. Autonomy continued to negotiate with Bank of America even after signing the contracts with Capax and DiscoverTech, and the latter never assumed responsibility for paying Autonomy should its deal with Bank of America fall through. Indeed, a witness from Bank of America testified that he had never discussed the possibility of doing the deal through Capax or DiscoverTech. Autonomy misrepresented these transactions to its auditors.

• MicroTech/DOI deal : Autonomy recognized $4 million in revenue in the fourth quarter of 2010 on a software contract with MicroTech, with Department of the Interior ("DOI") as the intended end user. However, there was no economic substance to the deal: MicroTech was never obligated to pay Autonomy, and Autonomy continued to negotiate directly with DOI, though the deal never actually closed. Autonomy misrepresented the transaction to its auditors, and should not have recognized revenue.

• VMS hardware deal : In the fourth quarter of 2010, Autonomy recognized revenue on a sale of $6 million in hardware to Video Monitoring Services ("VMS"). However, Autonomy was unable to actually obtain the hardware it was supposed to be selling. Accordingly, Hussain suggested that the company seek to sell hardware that Autonomy itself was using in its offices. Though much of the hardware was never delivered, Autonomy recognized revenue on the sale and represented to its auditors that it had in fact closed the deal in that quarter. This was improper. Hussain instructed employees to lie to auditors.

• Tikit/KPMG deal : Autonomy purportedly agreed to license $6 million in software to a reseller called Tikit in the fourth quarter of 2010. Autonomy planned for Tikit to then resell the license to KPMG, an accounting firm. However, because

Tikit did not have a longstanding relationship with Autonomy, it required a side letter from Autonomy guaranteeing that Tikit would not be required to repay Autonomy if it failed to resell the software to KPMG. Autonomy falsely represented to its auditors that there was no side letter. In fact, KPMG refused to deal with Tikit, and Autonomy ultimately completed the deal directly with KPMG, never receiving payment from Tikit. Accordingly, Autonomy should not have recognized the revenue in the quarter it did.

• DiscoverTech/Prisa Deal : In the first quarter of 2011, Autonomy purportedly sold DiscoverTech the right to sub-license Autonomy software to an Italian company called Prisa for $3.6 million. However, this was linked to a $4.4 million purchase by Autonomy of worthless technology from DiscoverTech; DiscoverTech never actually sold anything to Prisa, and indeed, Prisa was not aware of DiscoverTech. In addition, the Prisa contract was back-dated. Autonomy misrepresented the transaction to its auditors. There was no economic substance to the transaction, and revenue should not have been recognized.

• MicroTech direct deal : In the first quarter of 2011, Autonomy purportedly agreed

to license software to MicroTech for $3.86 million to use for its own purposes, and to provide services to Autonomy customers. Simultaneously, however, Autonomy gave MicroTech the right to collect $4.5 million from Bank of America in exchange for no consideration (see above). A Bank of America witness testified that it had never accepted services from MicroTech. This indicates that the license Autonomy sold to MicroTech was worthless, and it accordingly should not have recognized revenue on the transaction.

• Capax/McAfee deal : Autonomy purported to license $5 million in software to Capax for resale to McAfee in the first quarter of 2011. However, Autonomy and Capax had agreed that Capax would not owe Autonomy anything unless and until Capax actually sold the software to McAfee, and Autonomy knew when it agreed to the contract with Capax that the software would not actually sell through (McAfee cancelled a contract with Autonomy the day before Autonomy signed the reseller agreement with Capax). While Capax paid Autonomy the $5 million, Autonomy later paid Capax $6 million for tools it had no use for in order to reimburse it. Autonomy should not have recognized revenue.

• DiscoverTech/Abbott Labs deal : In the second quarter of 2011, Autonomy purportedly licensed $9 million in software to DiscoverTech for sublicense to

Abbott Laboratories. Autonomy represented to its auditors that the commercial rationale for the deal was that DiscoverTech was a veteran-owned business, enabling Abbott to fulfill government requirements that it work with such a business. DiscoverTech was not actually veteran-owned, however. Moreover, there was never any prospect of the deal actually selling through to Abbott, because Abbott refused the deal on the same day Autonomy contracted with DiscoverTech. DiscoverTech never made efforts to sell software to Abbott. Autonomy later cancelled the transaction with DiscoverTech. It should not have recognized revenue on the deal.

• DiscoverTech / Dell / Hyatt deal : In the second quarter of 2011, Autonomy purportedly licensed another $9 million in software to DiscoverTech to sublicense to Dell. Dell would then itself sublicense the software to Hyatt. However, this was a sham arrangement; DiscoverTech made no efforts to resell the software to Dell, and Autonomy continued to exercise control over the terms of the negotiations with Dell. Autonomy was ultimately unable to sell the software to Dell, and cancelled the agreement with DiscoverTech later that same year. It should not have recognized revenue. Autonomy misrepresented the nature of the deal to its

auditors.

• MicroTech/HP deal : Finally, Autonomy purported to sell software to MicroTech for $7.35 million in the second quarter of 2011, with the understanding that MicroTech would resell that software to Hewlett-Packard. However, Autonomy was not even in negotiations with HP when it signed the agreement with MicroTech. MicroTech did eventually pay Autonomy, but only after Autonomy agreed to purchase $8.2 million worth of software it did not need from MicroTech. Autonomy should not have recognized revenue on the deal.

b. Misrepresentations regarding hardware deals

The government put on evidence that Autonomy sold hardware at a loss in order to inflate its revenues while misrepresenting itself as a "pure software" company and misrepresenting the nature of the sales to its auditors. Autonomy claimed that, unlike most software companies, it did not "diversify into hardware." As discussed later in this order, this is significant because the gross margins on software sales are much higher than on hardware sales—meaning that a pure software company would tend to be more valuable than a mixed software/hardware company. According to one analyst who testified at trial, cloud software businesses trade on 10 to 15 times their total subscriptions, whereas hardware businesses trade on about 0.6 times revenue.

The government introduced evidence that, contrary to its frequent representations to the contrary, Autonomy used hardware sales to pad its revenues and enable it to meet analyst projections in every quarter from the third quarter of 2009 to the second quarter of 2011—selling over $180 million in hardware over that period, with hardware sales reaching as much as $37.6 million (19.6% of total sales) in a quarter. According to the government, Hussain and Lynch hatched the plan in the third quarter of 2009, when they sold approximately $36 million worth of hardware from EMC Corporation at a loss. After that, Autonomy would sell ordinary Dell hardware—"the stuff that you or I could buy at Fry's"—at a loss, buying at $110 and selling at $100. Hussain repeatedly pushed Michael Sullivan, a salesperson at Autonomy, to complete more hardware deals.

Meanwhile, Autonomy continued to represent itself as a "pure software" company. Because recognizing the costs of the hardware on its books would have raised a red flag, making others wonder how involved in the hardware business Autonomy really was, it concocted a number of different explanations regarding what it had actually incurred the costs for—telling auditors that most of the costs were associated with software marketing, while telling market analysts that the costs were associated with the launch of a new software product known as IDOL SPE. When one analyst ask Lynch why Autonomy was maintaining a large amount of hardware inventory in early 2010, Lynch told the analyst that Autonomy was not selling hardware alone, but was instead selling what are known in the industry as "appliances"—hardware that is pre-loaded with software for the convenience of the customer.

All of Autonomy's statements, however, were lies. Sullivan said Hussain charged him with selling "only hardware, nothing else"—the hardware was not accompanied by software or services. Joel Scott, an associate general counsel at Autonomy, concurred, saying that Autonomy sold hardware solely for the purpose of padding revenue.

Fernando Lucini, who was involved in product development at Autonomy, said that very little work being done to develop and market SPE when Lynch said it was. While Autonomy attributed $7.3 million in expenses to research and development for SPE, the actual costs were no more than $100,000. Marc Geall, who worked in investor relations at Autonomy before later covering the company as an analyst, testified that he was uncomfortable at being asked to tell analysts that the revenue was related to SPE because this was not true.

Moreover, several Autonomy salespeople said they did not know that Autonomy was selling hardware at a loss, and thus did not use this fact to build goodwill with software clients, contrary to the company's representations. Several clients testified that they were unaware of Autonomy's hardware sales to their companies, as well. Meanwhile, market analysts had "no idea" that Autonomy was reselling a significant amount of hardware at a loss.

Autonomy concealed the hardware sales from HP during the due diligence process in the run-up to the acquisition. Manish Sarin, an HP employee who evaluated HP's potential acquisition of Autonomy, said that, in light of Autonomy's actual hardware sales, answers Hussain gave to questions posed by Sarin during the due diligence process regarding Autonomy's revenue were non-responsive. Andy Gersh, a managing director at KPMG LLP, an accounting firm HP hired to aid in the financial due diligence work, gave similar testimony. He said that he interpreted Autonomy's statement in its financial report that it sold appliances, and that the value of its appliance sales were similar to that of its stand-alone licenses, to mean that "there really wasn't a whole lot of hardware in the overall solution or the value of the hardware wasn't significant to the overall solution." He said that he "had no idea" of the extent of Autonomy's hardware sales. He said he "would have expected them to tell me" about the nature and extent of the hardware sales in response to his questions.

c. Extent of misrepresentations

Christopher Yelland testified regarding the proper accounting treatment of Autonomy's transactions, as well as the extent of Autonomy's misrepresentations. Yelland became the chief financial officer of the Autonomy unit within HP in April 2012, several months after the acquisition. Lynch and Hussain left the company that May, after Autonomy missed its first-quarter revenue goals by a substantial margin. After reviewing Autonomy's books, Yelland raised concerns about the company's accounting practices. Along with Antonia Anderson, a former Deloitte auditor who had moved to Autonomy shortly before the HP acquisition, Yelland began a "rebasing" exercise in June 2012—revisiting the company's accounting treatment of deals over $1 million in the prior two years. The rebasing exercise turned up significant problems.

As a director of Autonomy, Yelland had a legal responsibility to ensure that Autonomy's accounts were stated correctly, and came to believe that accounts in prior years had not been. He and his group thus prepared a restatement of the accounts of Autonomy Systems Limited—a subsidiary of Autonomy through which most of the revenue flowed—completing their work in early 2014. Yelland certified under penalty of legal sanction that the group's work was accurate.

The government walked through Yelland's conclusions regarding the correct accounting for the transactions he identified as problematic. The government also asked Yelland to identify hardware sales. Per Yelland's calculations, the total amount of revenue adjustments, including hardware, was 21.8% in 2009, 29.3% in 2010, 30.7% in the first quarter of 2011, and 20.5% in the second quarter of 2011. The government summarized Yelland's work in the following demonstrative.

Image materials not available for display.

2. Evidence regarding significance of misrepresentations

A number of witnesses testified regarding the significance of the alleged misrepresentations. Among these witnesses were (1) market analysts, (2) people who were involved in HP's decision buy Autonomy and in the due diligence process that followed, and (3) HP shareholders.

a. Market analysts

Several market analysts testified that the alleged misrepresentations had a significant effect on their valuations of the company. Paul Morland, an analyst who had covered Autonomy prior to the HP acquisition, when Autonomy was publicly traded on the United Kingdom market, said that he had found Autonomy attractive because it was a software company. He noted that software companies tend to generate high profits because marginal costs are very low, and that they "can grow very, very quickly when they've got good software." Autonomy was in the business of providing electronic discovery software and services to help clients with litigation, and had signed a number of contracts with banks during the financial crisis that began in 2008. Morland believed that the litigation that was likely to follow the crash in the financial services industry would generate significant e-discovery work, and might enable Autonomy to buck the recession.

Morland's valuation of Autonomy depended significantly on Autonomy's growth rate, rather than its current profitability. Autonomy's price-to-earnings (P/E) ratio was very high, so a decline in the growth rate would translate to a significant decline in valuation. In evaluating Autonomy's growth rate and other issues, Morland relied on all publicly available information about the firm. He read the quarterly reports "in detail," attended quarterly meetings at which Autonomy would field questions from analysts, and, "crucially," had "discussions with management around what the future looked like, what the order book was like, what the growth rate was likely to be, how sustainable the margins were, whether the margins were going to go up or down." He would then use that information and to forecast future earnings and arrive at a valuation.

Morland recalled that, when he first learned that Autonomy sold large amounts of hardware at a loss, divorced from any software sales, he said, "Blimey!": a British-ism meaning "awesomely surprising." This information would have had a "massive impact" on his valuation of the company, because the operating margin of software businesses exceeds that of hardware businesses by a factor of about 10. With this new information, "you would probably reach the conclusion that the actual software business itself was not growing at all." Other analysts said that the new information brought them to similar conclusions.

b. HP due diligence

Witnesses who had been involved in HP's acquisition and due-diligence process also spoke to the significance of Autonomy's misrepresentations. Leo Apotheker, who was hired as CEO shortly before the Autonomy acquisition, testified that he viewed HP as a low-margin hardware manufacturer, and that he wanted to shift the company to a model based more on high-margin products such as software. Part of his strategy for accomplishing this was to acquire another company, and this strategy ultimately led to the Autonomy acquisition. Apotheker said that, in reading Autonomy's financial reports, he understood that its hardware sales were minimal, and were largely associated with appliance sales. He said that the actual nature and extent of Autonomy's hardware sales was relevant because it contradicted Autonomy's statement that it was a "pure software company," and would mean that Autonomy was growing much more slowly than it claimed. Apotheker also said that he was satisfied with the diligence HP performed, though HP did not uncover Autonomy's alleged misrepresentations—suggesting HP had no choice but to rely on Autonomy's representations regarding its sales.

Sarin, the HP employee who helped oversee the due diligence process, testified that Autonomy's misrepresentations significantly affected his valuation of the company, and his recommendation as to whether HP should buy. Sarin said that Autonomy was attractive as a potential acquisition because it was a software business; because it was growing at a rate of 15-20 percent per year; because its operating margins were "the highest in the software industry"; and because Autonomy was in the business of organizing unstructured data, which Sarin and others viewed as a growing market.

Like Apotheker, Sarin said that he relied on the accuracy of both the information Autonomy provided, and information already in the public domain about the company. He assumed Autonomy had been forthright with its auditors. HP sought more information from Autonomy than what the company had disclosed publicly, but Autonomy balked at providing it. HP nevertheless persisted, asking questions and trying to "test the information they've given us."

Sarin said the fact that Autonomy was selling hardware at a loss, with no software loaded onto the hardware, would have been "very relevant" to HP's valuation model. "[W]e wanted to understand the full picture around Autonomy's revenues. . . . [that Autonomy was selling standalone hardware] would have been meaningful and absolutely new news to us." Had HP known of the hardware sales, it would likely have estimated a lower future growth rate, and hence valued the company less.

In addition, Sarin said he would have wanted to know that Autonomy's resellers were in debt to it, a fact Hussain did not disclose during their daily diligence calls. He also said that Hussain did not disclose the extent to which Autonomy relied on resellers to book revenue at the ends of quarters—a disclosure that would have been responsive to Sarin's questions. This evidence would have been relevant because it meant Autonomy "would underperform based on what the [analyst] community was expecting them to do, which would certainly have a detrimental effect on how they're perceived in the market and most certainly their stock price."

Nor did Hussain tell Sarin that Autonomy was engaging in reciprocal transactions with resellers—a fact that would have been relevant because it would have indicated that "the implied revenues and sales of the products isn't entirely correct," which in turn would call into question the company's growth rate and value.

Gersh, the KPMG witness, testified similarly. Like Apotheker and Sarin, Gersh relied on the representations in Autonomy's public documents. Gersh said that the amount of hardware Autonomy was selling was relevant "[b]ecause the profitability of software is significantly different than the profitability of hardware." He added that learning about Autonomy's hardware sales changed the way he thought about its business: "This is a completely different business from . . . what Autonomy is in. . . . [I]t's almost the same as if Autonomy was selling motor cars. It's that unusual."

c. HP shareholders

Two HP shareholders also testified regarding the significance of the alleged misrepresentations. The first was Nigel Upton, who purchased HP stock after reading a press release announcing the Autonomy acquisition. Upton thought Autonomy was a particularly good acquisition because of its high growth rate, and said that the press release had been what caused him to buy the stock. He said he "wouldn't have made the same decision" had he known that the representation of the growth rate was false.

Thomas Garner similarly testified that he first heard of HP's acquisition of Autonomy through a corporate announcement. Like Upton, he invested based on the high growth rate reported; in fact, "It was the only reason for making the decision." Like Upton, he would have made a different decision had he known that the numbers were not as reported.

3. Evidence regarding Hussain's personal involvement in the fraud

The government portrayed Hussain as a member of a larger conspiracy to defraud, and thus argued that he was legally responsible for the criminal actions of his co-conspirators. However, it also put forth a significant amount of evidence linking Hussain directly to the alleged fraud. According to the government's evidence, Hussain was heavily involved in both sales and financial reporting, overseeing the deal-making process as well as the company's books. He was also responsible for Autonomy's representations to auditors, and was involved in analyst calls as well as the HP due-diligence process, making misrepresentations in those interactions, as well. Finally, Hussain sought to cover up the fraud: instructing subordinates not to put side agreements in writing, sending intentionally misleading emails, retaliating against an internal whistleblower, and barring a market analyst who had suggested that Autonomy was misstating its revenues from participating in quarterly calls. The evidence suggested that Hussain played a central role in the conspiracy, though he was perhaps not the driving force behind it.

a. Hussain's role in sales

Autonomy witnesses testified that Hussain was intimately involved in the sales process. During one quarter, Hussain wrote to Lynch that "Stouff and I were all over the big deals." Hussain discussed and planned quid-pro-quo deals with Egan, and later misrepresented those deals to auditors. Hussain also directed Egan to back-date a deal with DiscoverTech. When Dave Truitt, the DiscoverTech CEO, told Hussain that he had been uncomfortable back-dating the deal, Hussain reassured him (falsely) that doing so was appropriate because the rules governing Autonomy's accounting differed from those with which Truitt was familiar, given that Autonomy based in Britain.

Hussain repeatedly sent e-mails at quarter's end in which he stated that he was searching for more revenue in an effort to meet analyst projections, suggesting that his role was to help sell and create revenue—not merely to recognize sales Autonomy had already made. On several occasions, Hussain himself negotiated directly with the customer. Hussain meticulously tracked the progress of sales on a spreadsheet—an indication of his heavy involvement in the sales process, as well as his knowledge that certain deals were being back-dated.

Sullivan, the Autonomy salesperson, testified that Hussain was very involved in the plan to generate more revenue through hardware sales, hatching the plan with Mike Lynch. Hussain told Sullivan that Autonomy's goal in the first quarter of 2010 was to do $20 million in hardware re-sales with Dell. "We have to get it bigger and more," Hussain told Sullivan shortly after Autonomy began working with Dell. Sullivan would periodically update Hussain and Lynch on sales of "low margin revenue," referring to these sales as the "low margin jigsaw."

b. Hussain's oversight of financial reporting

Witnesses also testified that Hussain was responsible for Autonomy's financial reporting, and in particular for deciding whether to recognize revenue on particular transactions. Hussain was tasked with ensuring the accuracy of Autonomy's financial reporting.

Matt Stephan, a senior finance manager, said that Hussain's instructions on revenue recognition made him uneasy. Regarding the deals with resellers, Stephan said he was concerned they were "only being done to get to that revenue figure," and that the deals were "a problem on many fronts from an audit perspective, [and] from [an] ongoing credit collection perspective." He perceived that Autonomy "wouldn't want to have anything to do with them" were it not trying to hit the consensus revenue estimates. Stephan complained to his supervisor, Steve Chamberlain, about the deals: "[T]hese deals were, you know, garbage. They were not worth the paper they were written on, and I wasn't happy to front them up as good deals to our auditors." Chamberlain's response was that it was Mr. Hussain's decision whether to recognize revenue on the deals. "The general [Hussain] says what to do and we follow our orders."

Discussing several back-dated hardware deals, Stephan said the deals he saw concerned him because the company seemed to be "picking and choosing" which quarter a particular deal was recognized in, rather than following accounting rules. "I recall instances where we had a whole mixture of delivery dates on hardware, and at different times they were trying to argue that they should be in or out of a given quarter, despite it being black and white about whether they were delivered or not."

Scott, the general counsel, also raised revenue-recognition issues with Chamberlain, to similar effect. Noticing that an agreement had been back-dated, Scott said he told Chamberlain he was concerned Autonomy was "taking revenue improperly." Chamberlain's response was that there was a plausible argument that the back-dating was proper, and in any event "Sushovan would share everything with the auditors and ultimately would . . . defer to the auditors on their decision." Chamberlain and Hussain did in fact not do so, however.

c. Hussain's misrepresentations

Autonomy's auditors, meanwhile, testified that Hussain repeatedly misrepresented transactions to them. Antonia Anderson, who audited Autonomy for Deloitte as an assistant manager before being hired by Autonomy in 2011, testified that Autonomy was responsible for preparing its own financial statements, with Deloitte testing particular transactions to ensure that the financial reporting was correct. Hussain confirmed in a series of "management representation letters" that it was Autonomy's responsibility, not Deloitte's, to prepare financial statements in accordance with international accounting standards. He also confirmed that he had made "all books of account and supporting documentation" available to Deloitte. However, as recounted above, Hussain withheld information from Deloitte that would have changed Deloitte's opinion of whether transactions were properly accounted for. Both Anderson and Lee Welham, a manager at Deloitte, testified extensively regarding Hussain's misrepresentations.

Most of the evidence the government introduced regarding misrepresentations to market analysts came from Lynch, not Hussain. However, Hussain participated in the quarterly calls, and the government argued that he was responsible for Lynch's misstatements as a co-conspirator.

Finally, Hussain made misrepresentations to HP representatives during the course of HP's due diligence. Sarin testified that Hussain participated in daily interactions between HP and Autonomy management in late July and early August of 2011, prior to the acquisition. Sarin described a number of specific meetings and calls in which Hussain participated. He spoke to Hussain on a daily basis regarding financial issues in the due diligence process. Hussain misrepresented the sources of Autonomy's revenue in these calls. Hussain also provided a false list of Autonomy's top customers, omitting resellers and hardware purchasers.

d. Hussain's efforts to cover up the fraud

The government also put on evidence that Hussain made concerted efforts to cover up Autonomy's fraudulent practices, suggesting that he was aware of and involved in those practices. Hussain repeatedly instructed Egan not to put side agreements that would have called revenue recognition into question in writing. He further directed Egan to craft "make the case" e-mails laying out a false rationale for a deal, creating a deceptive paper trail to fool auditors and investigators.

The government introduced evidence tending to establish that Hussain retaliated against an employee named Brent Hogenson who had made whistleblower allegations against Autonomy. Two days after Hogenson made allegations to Deloitte that Autonomy was improperly recognizing revenue, Autonomy told Scott to investigate allegations of payroll fraud in Hogenson's finance department, without disclosing to Scott that Hogenson had made whistleblower allegations. Hussain sent several e-mails to Scott during the payroll-fraud investigation raising concerns about Hogenson. When he brought these concerns to Hogenson, Hogenson accused Hussain and Autonomy of retaliation. A day later, management, including Hussain, told Scott to fire Hogenson. Scott fired two other employees in the finance department at Hussain's direction as well, saying that Hussain "did not want them at the company anymore." Hussain told Scott that one of the employees in particular "had raised her head above the parapet and that he wanted her gone." Autonomy later settled Hogenson's claims of retaliation for $750,000, with the government suggesting that the amount was an indication of the claims' validity.

The government also introduced evidence that Hussain and his co-conspirators excluded a market analyst who had questioned Autonomy's revenue recognition and growth, Daud Khan, from quarterly earnings calls.

e. Hussain's role in the conspiracy

The evidence regarding the start of the conspiracy was somewhat sketchy, and came primarily from Egan (an executive in the sales department) and Hussain's own e-mails. The government's evidence did not tend to paint Hussain as the main driver behind the conspiracy. This role instead seems to have been played by Lynch, Autonomy's CEO—an extremely ambitious, magnetic boss who exerted a great deal of pressure on other executives at the company, including Hussain, to meet financial goals. While acknowledging that it "might sound crazy," Egan described Hussain to FBI agents as "a highly ethical person" who made relatively little money in the HP acquisition. Egan said that Hussain participated only because he was "massively under the spell of one of the most powerful human characters on the planet"—that is, Lynch. When Egan once complained to Hussain about the pressure Lynch was exerting on him, Hussain said, "It's worse for me." Asked whether Egan believed Hussain, he said, "I knew it. . . . I knew he had more pressure on him than I had on me. I viewed mine as a subset of what he had." At one point Hussain expressed anxiety to Lynch about the analyst calls:

Mike, I'm burnt out. Given my anal nature, I'm spending all my time worrying about the 10 minutes on Tuesday where I have to answer the analysts questions and I'm not doing anything else. I need you to take the questions this time around unless they are really easy. I don't want to deal with the analysts anymore. All I want to do is deliver Q4, deliver the audit, deliver the acquisition and capital raising. The analysts take up too much time and I worry about it all the time rather than the business.

The government speculated that the goal of the fraud was to inflate Autonomy's growth rate to make the company attractive to potential acquirers. The government portrayed the practice of cannibalizing future revenue as unsustainable; the strategy only made sense if the members of the conspiracy could cash out before the fraud was uncovered, and/or the growth rate declined. As Egan put it, "[Y]ou start the new period with a hangover or an obligation, extra work. As opposed to starting at ground level, I'm starting in the basement, if you will." In addition, it became increasingly difficult to meet analyst projections because each inflation of revenue would in turn lead to a higher projection for the following quarter. This created a "snowball" effect, and led to more desperate measures to improperly recognize revenue. "Really don't know what to do, Mike," Hussain wrote to Lynch in an email in late 2010, lamenting the company's stagnant actual growth rate. "[R]adical action is required. Really radical. We can't wait anymore." In the government's interpretation, Hussain meant that Autonomy needed to find an acquisition partner soon.

C. Defense Case

The defense made two arguments. First, it contended that the government's witnesses were not credible, and that the government's interpretation of the evidence was incorrect. Second, it made a veiled case for jury nullification.

1. Credibility / evidence interpretation argument

For all intents and purposes, the defense did not put on an affirmative case. It was accordingly left to build its case by cross-examining the government's witnesses.

It argued that those witnesses were simply not credible—either because they were intentionally lying, or because they were mistaken about the facts. "There have been 37 witnesses, each of them or most of them with various motivations to help Hewlett-Packard or themselves," the defense told the jury during closing arguments. Stouffer Egan, Joel Scott, Alan Rizek, Dave Truitt, and John Baiocco perjured themselves, either to curry favor with HP or avoid government prosecution and shift blame on others. Meanwhile, the witnesses who worked for HP during the acquisition—Apotheker, Sarin, Gersh, Lesjak, Yelland, and Johnson—gave biased testimony because they were trying to aid HP's civil litigation against Autonomy and/or minimize their own failings. "[T]he prosecutors in this case . . . have been selling an absolute fantasy that the HP machine has created for them."

The defense also contended that the government's theory that Hussain had planned to defraud HP is belied by the fact that he went to work there after the acquisition. "Why would Mr. Hussain, knowing that all the Autonomy books and records would be in the possession of his new employer . . . lie to them before the acquisition?" Moreover, Autonomy was by no means desperate to get acquired: "Dr. Lynch was willing to walk away if HP tried to lower the price."

The defense argued that most of the transactions on which the government presented evidence were properly accounted for; that Hussain was not aware of the ones that were improper; that Autonomy properly disclosed the nature of its hardware sales; and that any misrepresentations were not material.

a. Accounting

The defense offered its own explanations of the transactions on which the government provided evidence. It argued that many of the transactions were legitimate and properly accounted for. Any misrepresentations Hussain made on other transactions were unknowing, with others, primarily Egan, scheming behind his back for their own purposes. In the defense's telling, "Most of the e-mail traffic shows scrupulous attention to the revenue recognition criteria, making sure that the risks and rewards pass with a final deal before the end of the quarter, making sure the collectability was probable." Similarly, the contracts Autonomy signed with resellers "make absolutely plain" that the resellers were still on the hook for paying Autonomy if the deals did not sell through to the intended end user. The resellers "all testified that they were on risk" and could pay. Autonomy continued to negotiate with the end user after selling to the resellers only in order to maintain good relationships, and Deloitte was aware of this face. And while Joel Scott, the general counsel, and Matt Stephan, the Autonomy finance manager, testified that the reseller deals made them uncomfortable, their continuing participation in those deals put the lie to that testimony.

The government's accounting witnesses used impermissible hindsight to conclude that Autonomy was never likely to be paid on specific deals. In fact, the Deloitte auditors "[were] all over it," and analyzed each reciprocal transaction for fair value. The auditors the government called, Anderson and Welham, were inexperienced and did not properly understand the reasons that Deloitte signed off on Autonomy's financial statements in the first place. Deloitte had access to Hogenson's whistleblower report, and "concluded that this guy doesn't know what he's talking about." HP apparently came to the same conclusion.

Meanwhile, Yelland was "fed" to the government by the "HP machine," and prepared the restatement not because of his duties to British regulators, but rather in order to bolster HP's civil case. "The restatement was designed for litigation." It is "simply absurd" to think that Yelland developed concerns about Autonomy's accounting practices on his own, given that he did not raise those issues immediately with Autonomy's auditors. Instead, he began work on the restatement "[a]t the prompting of these lawyers and accountants who were working on the civil case in England against Dr. Lynch and Mr. Hussain." That HP's auditors declined to opine on the restatement's accuracy is further evidence that it was unreliable. Moreover, Yelland lacked the appropriate expertise to evaluate Autonomy's accounts. And he "never tried to understand why Deloitte believed that the revenue recognition requirements had been met" during its initial audits.

The opinions of the auditors at trial differed from Deloitte's opinions at the time, and this disagreement introduces at least a reasonable doubt about the government's case. Moreover, far from deceiving the Deloitte auditors, Hussain relied on their opinions. The defense borrowed a line from the prosecution of O.J. Simpson to convey this point: "[T]o paraphrase a very famous lawyer, the late Johnnie Cochran: If the accountants disagree, Sushovan must go free."

The defense's specific arguments about the deals the government relied on were as follows.

• Capax EDD deals : Contrary to the testimony of Baiocco, Lucini, and Egan, Baiocco actually was doing EDD work, so these were perfectly legitimate transactions. Baiocco perjured himself by insisting that he never actually did any work pursuant to these contracts because he sought to curry favor with HP, with whom he currently does business. Meanwhile, Egan perjured himself by representing perfectly legitimate transactions as fraudulent in order to curry favor

with the government so as to avoid prosecution. The defense did not offer an explanation for Lucini's testimony in its closing, but suggested on cross that he had limited knowledge of Autonomy's operations and a faulty memory. As for the evidence that Capax's third EDD deal with Autonomy was back-dated, the defense argued that Baiocco and Autonomy had an undocumented agreement that predated the end of the quarter, and that Baiocco perjured himself in claiming that Hussain urged him to back-date the deal.

• Capax reseller deals : Contrary to the testimony of Baiocco (FSA, Amgen, Bank of America, McAfee), Egan (Amgen, Bank of America), Stephan (FSA, Amgen), Scott (Bank of America) and Smith (Bank of America), the full risks and rewards transferred to Capax on all these deals. Capax was "a major provider of support and maintenance," was a legitimate business, and confirmed the validity of these transactions to Deloitte. Baiocco and Egan perjured themselves, while Stephan and Scott used improper hindsight in evaluating the transactions. The defense did not explain why Smith, the Bank of America employee, had never heard of Capax, despite purportedly having bought the software from them. Meanwhile, the defense contended that Hussain reasonably believed back-dating the Bank of America agreements was proper because they were amendments to previous contracts, rather than new contracts.

• MicroLink/MicroTech deal : Contrary to the Dave Truitt's testimony that the sale

of MicroLink was contingent on Truitt buying software from Autonomy through MicroTech and DiscoverTech, Autonomy bought MicroLink for fair value; Truitt "wasn't going to sell the company for anything less" than the $55 million he received. The $9.5 million software purchase by MicroTech was not linked to the acquisition, but was instead a legitimate deal in which both sides received value. "The fact of the matter is that DiscoverTech paid MicroTech, MicroTech paid Autonomy, [and] Deloitte reviewed and approved the transaction." The defense did not explain Truitt's testimony that the deals were linked, other than to generally question his credibility. Meanwhile, the defense argued there was "no evidence that Mr. Hussain knew that Dave Truitt was a minority shareholder in MicroTech," though it conceded that Truitt in fact was. In any event, it said, that Truitt was a shareholder in MicroTech was not an issue from an accounting perspective because he had "no responsibilities and no control of MicroTech."

• MicroLink/DiscoverTech deal : As for the $2.3 million sale of software to MicroLink to resell to DiscoverTech, DiscoverTech needed the software it bought from Autonomy because it improved DiscoverTech's offerings, contrary to the testimony of Truitt, Egan and Rizek. Also contrary to their testimony, Hussain was not involved in back-dating the deal. Instead, Truitt, Egan, Rizek, and John Cronin conspired to back-date it without Hussain's knowledge.

• Other MicroTech reseller deals : Contrary to the testimony of Scott (DOI, Bank of America, HP), Dave Truitt (Vatican, Bank of America), Steve Truitt (DOI,

Vatican), Egan (HP, Vatican, Bank of America), Smith (Bank of America), and Lucini, Blanchflower and Goodfellow (Vatican), these were legitimate deals in which fair value was exchanged. Scott, not Hussain, negotiated the terms of the software purchases the government contended were linked to the HP and Vatican deals. Contrary to the testimony of Steve Truitt, Dave Truitt, and Egan, the Vatican deal was not back-dated: in actuality, John Cronin, a MicroTech employee who did not testify at trial, and Egan, who perjured himself on this point, had reached an agreement in principle prior to the end of the quarter. And if the deal was back-dated, Hussain was not aware of it.

• DiscoverTech reseller deals : Contrary to the testimony of Egan, Scott, Dave Truitt, and Smith, fair value was exchanged in the Bank of America deal. Contrary to the testimony of Dave Truitt, Snider, Sullivan, and Scott, fair value was also exchanged in the Abbott deal. The defense conceded that DiscoverTech was not actually a veteran-owned business, as Autonomy represented to its auditors following the Abbott Labs deal, but said there was "no indication that it was Mr. Hussain" who gave out the erroneous information. Regarding the back-dating of the Prisa deal, Egan and Dave Truitt consulted before trial and agreed to lie about being told by Hussain to back-date the contract, according to the defense. In fact,

Egan came up with the idea to back-date it because doing so would help him get a bigger bonus, and hid the back-dating from Hussain (the defense did not rebut the government's argument that the Prisa contract was not an exchange for fair value).

• FileTek direct deals : Contrary to the testimony of Sullivan, Blanchflower, Lucini, Goodfellow, Egan, and Loomis, Autonomy did in fact have a use for FileTek's StorHouse software, and exchanged fair value for it. Autonomy's "most senior engineers" determined that it was "a very useful piece of technology" that could enable Autonomy's products to handle structured data. The three licenses Autonomy bought were not cumulative because each covered different customers and/or products. And the price was fair because Autonomy negotiated a significant discount from the price FileTek originally quoted.

• FileTek/VA deal : Contrary to the testimony of Egan and Loomis, FileTek was in fact on the hook for paying Autonomy for the software, though it did not ultimately sell through to the VA.

• Morgan Stanley hardware deal : Though revenue should not have been recognized, Hussain reasonably believed the hardware actually had shipped on time, and any misrepresentation on his part was therefore unintentional.

• VMS hardware deal : Contrary to the testimony of Goodfellow, Stephan, Anderson and Welham, VMS in fact wanted staggered delivery of the hardware, so it was

proper to recognize revenue.

• Auxilium/Vatican deal , Sales Consulting / Poste Italiane deal : The defense attacked the government's inclusion of these deals on the ground that the government only relied on paper documentation of them, along with Yelland's testimony.

• Tikit/KPMG deal : Contrary to the testimony of Anderson, Welham, and Yelland, Hussain did in fact disclose the side letter to auditors. If revenue was improperly recognized, he was only relying on their advice.

• Morgan Stanley software licensing deal : The defense did not appear to address this deal in its closing argument.

• DiscoverTech/Dell/Hyatt deal : The defense did not appear to address this deal specifically in its closing argument. However, it implied that any misrepresentations were attributable to Egan.

b. Hardware sales

The defense contended that Autonomy did not misrepresent the extent or nature of its hardware sales. It urged the jury to take at face value Autonomy's representations to its auditors that its hardware sales were part of a software-marketing program. "[T]he purpose was relationship building and selling more software." In addition, the allocation of costs to SPE was appropriate because Deloitte signed off on the accounting at the time, and in any event Hussain "really wasn't much involved in making these decisions" about how to allocate costs. Apotheker perjured himself in saying he did not know Autonomy sold hardware for resale, according to the defense. Meanwhile, Sarin and Gersh did not ask the correct questions during the due diligence process to elicit information about Autonomy's hardware sales.

c. Significance of misrepresentations

Any misrepresentations or improper accounting were not significant because Autonomy disclosed all pertinent issues to HP. HP knew that Autonomy used resellers because Autonomy's annual report mentions them. Moreover, in its report to HP, KPMG noted that Autonomy recognized revenue when deals sell to resellers, rather than then they sell through to the end users. So "[t]here can't be any question that Hewlett-Packard understood how Autonomy was dealing with resellers."

Moreover, Autonomy gave HP what it believed the latter wanted in turning over lists of its top 40 customers and contracts. Autonomy believed that HP was not interested in sales to resellers or hardware sales. "The lists were the product of very limited due diligence under the English takeover rules. . . . [T]he information that was exchanged was heavily negotiated and very limited."

Any misrepresentations were insignificant, moreover, because Apotheker, the HP CEO, only cared about Autonomy's core software product. "Mr. Apotheker didn't care what he paid." "All he cared about was his vision" for integrating Autonomy's product into HP's offering "and mak[ing] a whole bunch of money."

HP, in other words, was not actually deceived. The timing of HP's announcement that Autonomy had defrauded it tends to establish this. "A year after the acquisition closed . . . Hewlett-Packard stands up, Ms. Whitman, the CEO, and screams fraud." HP fired Deloitte, in the defense's words, "because they're not going to help [HP] say that HP has been defrauded."

2. Jury nullification argument

The nullification argument picks up somewhere around where the argument about the significance of the misrepresentations leaves off. According to the defense, when HP acquired Autonomy, it did not care about Autonomy's hardware sales and any other misrepresentations; but once Apotheker had been fired, HP went to work discrediting his vision to integrate Autonomy and HP and blaming Autonomy for HP's failings. "The vision went away, and a year later, Hewlett-Packard, with its army of lawyers, consultants, and so on, began working on this case." HP's "machine" "supported the Department of Justice in bringing this case," and "brought this almost incomprehensible accounting dispute between Hewlett-Packard and Autonomy into a U.S. courtroom where they're trying to make this English man, Sushovan Hussain, into a criminal."

Hussain is a scapegoat: "Everybody in this case gets a pass, everybody that you have heard about, all the immunized witnesses, everybody that they mentioned. . . . [B]ut he's supposed to be a criminal." This is not the government carrying out justice, but rather the government doing HP's bidding: "Why is he here alone? We know why he's here. Because the HP machine wants him convicted. With countless lawyers and consultants working on the civil case for money damages in England, HP wants a criminal conviction that they can point to in that civil case." Indeed, criminal charges are not appropriate for accounting fraud cases. "This case is a dispute about accounting. That accounting dispute is being litigated in English courts for money damages, and that's where it belongs."

While the defense did not argue outright that the jury should acquit Hussain even if it found that the government had met its burden, it implied as much. Hussain "doesn't have a lot of great faith in the prosecutorial system, but you, the jury, are the only ones that can protect him from a prosecution that is trying to crush him." The jury's role is not merely to weigh the evidence, but also to serve as a "bulwark" against abuses of power. A guilty verdict would mean that the jurors had bought into an unjust bureaucratic system. "I ask you not to become part of the HP machine and to return a verdict of not guilty."

D. Verdict and Post-Trial Motions

The jury returned a verdict on April 30, finding Hussain guilty on all counts. See dkt. 394. The defense filed post-trial motions on May 23.

II. LEGAL STANDARD

A. Judgment of Acquittal

Federal Rule of Criminal Procedure 29 provides that the Court "on the defendant's motion must enter a judgment of acquittal of any offense for which the evidence is insufficient to sustain a conviction." Fed. R. Crim. P. 29(a). Evidence is insufficient to sustain a conviction if, viewing the evidence in the light most favorable to the prosecution, no rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt . See Jackson v. Virginia, 443 U.S. 307, 319-20 (1979); United States v. Nevils, 598 F.3d 1158, 1163-64 (9th Cir. 2010) (en banc).

B. New Trial

A district court may grant a new trial "if the interest of justice so requires." Fed. R. Crim. P. 33(a). A new trial is only warranted on the basis of an incorrect evidentiary ruling or an erroneous jury instruction if a party was substantially prejudiced. United States v. 99.66 Acres of Land, 970 F.2d 651, 658 (9th Cir. 1992) (evidentiary ruling); Wilkerson v. Wheeler, 772 F.3d 834, 841 (9th Cir. 2014) (jury instruction). Meanwhile, "[i]mproprieties in counsels' arguments to the jury do not require a new trial unless they are so gross as probably to prejudice the defendant, and the prejudice has not been neutralized by the trial judge." United States v. Parker, 549 F.2d 1217, 1222 (9th Cir. 1977).

III. DISCUSSION

A. Motion for Judgment of Acquittal

The defense moves for a judgment of acquittal on the basis that the evidence was insufficient to sustain convictions for conspiracy to commit wire fraud (count 1), wire fraud (counts 2-15), and securities fraud (count 16).

1. Wire fraud counts

Wire fraud has three elements: (1) the making of material misstatements or omissions as part of a scheme to defraud, (2) the use of interstate or foreign wires in furtherance of the scheme, and (3) intent to defraud. United States v. Jinian, 725 F.3d 954, 960 (9th Cir. 2013). First, the defense argues that the government failed to introduce sufficient evidence of misrepresentations or omissions: "The nature of the acts in question here was to maximize Autonomy's apparent profitability and maintain a high share price, which is not a crime." Renewed Rule 29(c) Mot. (dkt. 408) at 7. While it is of course true that maximizing profitability is not in itself a crime, the government presented copious, often duplicative evidence that Hussain and his alleged co-conspirators repeatedly misrepresented Autonomy's accounts and the nature of its business in order to deceptively inflate its perceived value. See supra Part I.B.1. The government's evidence was more than sufficient.

Next, the defense argues that the government failed to put forth sufficient evidence that Hussain possessed the requisite intent: it "lumped together a mishmash of different transactions with only occasional and tangential connections to Hussain." Renewed Rule 29(c) Mot. at 1. On the contrary, the government introduced substantial evidence that Hussain knew of the scheme to defraud and intentionally deceived auditors, market analysts, and ultimately HP. See supra Part I.B.3. So the defense's arguments that the evidence was not sufficient to convict Hussain of wire fraud fails.

2. Conspiracy count

The defense next contends that the evidence at trial was insufficient to sustain a conviction on the conspiracy count. Conspiracy requires an agreement among two or more persons to engage in conduct that would satisfy the elements of the substantive offense. See Salinas v. United States, 522 U.S. 52, 65 (1997). The intent element is satisfied when a conspirator adopts the goal of furthering or facilitating the criminal endeavor. Id.

The defense's argument that Hussain lacked the requisite intent fails for the reasons already given with respect to the substantive wire fraud charge. The defense's other contention, that the government only introduced evidence of an agreement to defraud Autonomy's shareholders, not HP, also fails. See Renewed Rule 29(c) Mot. at 7-9. While the defense is of course correct that a scheme to defraud Autonomy shareholders does not necessarily equate to a scheme to defraud HP, the government in fact introduced a significant amount of evidence that Hussain and his co-conspirators specifically intended to defraud HP. See supra Part I.B.2.b (describing misrepresentations during HP due diligence process).

3. Securities fraud count

Finally, the defense argues that the evidence was insufficient to sustain a securities fraud conviction because there was no evidence that any misrepresentation by Hussain came "in connection with" HP securities. See 18 U.S.C. § 1348. The government argued that Hussain's involvement in drafting a deceptive press release issued by HP regarding the Autonomy acquisition was a sufficient connection to HP securities to sustain a conviction on this count. The element is satisfied even where the defendant did not himself disseminate or sign a public announcement where he "can fairly be said to have caused [the speaker] to make the relevant statements," and "knew or should have known that the statements would reach investors." United States v. Hussain, No. 16-CR-00462-CRB-1, 2017 WL 4865562, at *8 (N.D. Cal. Oct. 27, 2017) (quoting S.E.C. v. Wolfson, 539 F.3d 1249, 1261 (10th Cir. 2008)). The evidence was sufficient for the trier of fact to find that Hussain caused HP to make misleading statements to its investors by deceiving HP regarding the nature of Autonomy's business, and that he should have known the statements would reach investors, given that HP is a publicly traded company. See supra Part I.B. In addition, the government introduced evidence that Hussain specifically warranted that the information to be included in HP's public announcements was correct. See Exs. 2309, 2295. The defense's argument that the government introduced "no evidence" on this issue therefore fails.

B. Motion for New Trial

The defense also argues that Hussain is entitled to a new trial on the basis of several of the Court's rulings on discovery matters, witness availability issues, evidentiary issues, and jury instructions, in addition to what it contends was an improper closing argument by the government. Those contentions are taken up below.

1. Discovery rulings

The defense litigated a number of motions pre-trial to obtain various discovery materials. It moved the Court on four separate occasions to order the government to comply with its Brady obligations (dkts. 44, 138, 211, 302); twice moved for a ruling that HP was subject to the dictates of Brady, though a private entity (dkts. 45, 140); moved to compel disclosure of grand jury colloquies (dkt. 42); moved to compel production of government files (dkt. 74); litigated HP's motion to quash a defense subpoena (dkt. 110); and moved to compel production of certain files by HP (dkt. 213) (later withdrawn, see dkt. 223). With limited exceptions regarding interviews of witnesses by HP lawyers (see dkts. 255, 260, 264, 266), the Court denied the defense's motions.

In its motion for a new trial, the defense argues that the Court's denial of its motions to compel the government to review HP's files warrants a new trial. Rule 33 Mot. (dkt. 409) at 2 n.1. The defense had argued that HP, HP's outside counsel, and the government worked together so closely in bringing this prosecution that the former should be subjected to the same disclosure requirements as the government under Brady, Giglio, and Kyles. See dkts. 45, 140. Specifically, the defense argued that the logic of Kyles v. Whitley, 514 U.S. 419 (1995), extends to any case in which the prosecution works closely with another entity. It seized on the statement in Kyles that "the individual prosecutor has a duty to learn of any favorable evidence known to others acting on the government's behalf on the case, including the police" to contend that Kyles encompasses private parties that give assistance to the government. Id. at 437.

The reason that courts have extended the Kyles duty to other branches of government is that (1) those branches' incentives are perfectly aligned with the prosecution's, and (2) the prosecution presumably has access to their files. If the prosecution chooses to work on an investigation with another branch, it makes sense to assume that the prosecution knows (or is able to find out) everything that other branch knows. See id. at 438 ("Since, then, the prosecutor has the means to discharge the government's Brady responsibility if he will, any argument for excusing a prosecutor from disclosing what he does not happen to know about boils down to a plea to substitute the police for the prosecutor, and even for the courts themselves, as the final arbiters of the government's obligation to ensure fair trials.").

The defense cited no case, however, in which a court has extended this duty to a private party, rather than a branch of government. Indeed, several courts have held that Kyles does not apply to private parties. For instance, in United States v. Josleyn, the First Circuit refused to extend a prosecutor's obligation under Kyles to a private party because it reasoned that private parties' interests may not be perfectly aligned with the government's. 206 F.3d 144, 154 (1st Cir. 2000). For instance, a private party may have an incentive not to disclose all the material in its possession to the government because it may itself become the subject of an investigation, and therefore has "an interest in seeing that the government knew some information about the kickback schemes, but not too much." Id.. Or, it may have such an incentive because it fears it would weaken its own civil case against the defendant by turning over certain information to the government, as, perhaps, here.

Accordingly, charging prosecutors with the knowledge possessed by private parties that are cooperating with them is inconsistent with the logic of Kyles. Contrary to the defense's arguments, this is true even where the private party receives some benefit for the information in turns over. Indeed, there is no reason to think that receiving a benefit for turning over some files makes it any less likely that the private party is withholding access to other files. In addition, such a rule would be inconsistent with the private party's Fourth Amendment rights.

Accordingly, the Court properly denied the defense's motions to require the government to search HP's files.

2. Witness availability issues

The defense also filed several motions related to the availability of witnesses at trial. Prior to trial, it had moved to compel the government to offer immunity to several of its witnesses (dkt. 194), and to compel the government to request the attendance at trial of overseas witnesses (dkt. 212). During trial, it raised a separate issue related to a confidential witness matter. (Agreeing that good cause exists to keep this matter confidential, the Court has redacted its discussion of it below). The defense argues in its Rule 33 brief that the Court's failure to grant its motions on the witness immunity issue and the confidential matter warrant a new trial. Rule 33 Mot. at 2 n.1.

a. Witness immunity issue

Prior to trial, Hussain urged the Court to require the government to grant immunity from prosecution to six defense witnesses. See dkt. 194. Hussain argued that fairness required this result because the government had immunized from prosecution several witness it believed would be favorable to its own case. However, the decision of whether or not to immunize witnesses rests squarely within the prosecutor's discretion, and "the prosecution does not abuse its discretion when it refuses to grant use immunity to a defense witness who has been indicted or is the subject of a criminal investigation." Williams v. Wollford, 384 F.3d 567, 602 (9th Cir. 2002). The Court had no grounds on which to issue such an order, and properly refused to do so.

b. Confidential witness issue

[Redacted]

3. Evidentiary rulings

The defense also moves for a new trial based on several of the Court's evidentiary rulings. It devotes its new trial motion mainly to the exclusion of certain evidence regarding HP's actions following the Autonomy acquisition, and incorporates by reference other arguments it raised during trial. See Rule 33 Mot. at 2 n.1. The Court takes the various issues the defense raises in turn. It also addresses other significant evidentiary rulings to which the defense preserved objections.

a. HP's payment of witnesses' legal fees

The defense twice moved during trial for the Court to admit evidence that HP was covering various witnesses' legal fees. See dkts. 262, 280. The defense argued that this evidence was relevant because it tended to show that those witnesses would be unconsciously biased to testify in a way that would aid the government's case. The government argued that this was irrelevant because HP was required under its bylaws to indemnify current and former employees in connection with government investigations and prosecutions.

The Court agreed that the evidence was marginally relevant, but determined that its probative value was substantially outweighed by the dangers of unfair prejudice, confusing the issues, and undue delay. See Fed. R. Evid. 403; Tr. at 805-808. Raising the issue would have required a lengthy detour into state law and/or HP's bylaws to explain whether HP was paying the fees by choice or by compulsion, a collateral issue that ran the risk of confusing the jury and wasting time. The Court also perceived a danger that, instead of relying on the evidence to determine whether witnesses had testified favorably to HP out of some sense of gratitude (a permissible inference), jurors would make the inference that HP was directing the government's case (an impermissible one). This danger was heightened by the defense's repeated insistence that jurors consider whether they would be doing HP's bidding in convicting Hussain. See supra Part I.C.2. The evidence was properly excluded under Rule 403.

The defense's argument that Wood v. Georgia, 450 U.S. 261 (1981), compelled a different result failed because that case involved an employer paying legal fees for a criminal defendant where the interests of the defendant and his employer, the operator of a criminal enterprise, were plainly at odds. Id. at 267-69. This potential conflict of interest implicated the Sixth Amendment right to counsel. Id. at 271-72. Similar issues were not implicated here.

b. Restatement

The defense objected to the Court's decision to allow the government to introduce the restatement of Autonomy Systems Limited's finances prepared by Chris Yelland and HP. See dkt. 158; Restatement (dkt. 143-2). (ASL was a subsidiary of Autonomy through which most of Autonomy's revenue flowed.) The restatement was relevant under Rule 401 because it spoke to whether Autonomy misrepresented its finances in its audited financial statements. See Restatement at 2.

The main grounds for the defense's objection to allowing the restatement in was that it was hearsay. The government countered that the restatement was admissible as a business record. Rule 803(6) provides that hearsay records are admissible if they are kept in the course of a regularly conducted activity of a business, made at or near the time by someone with knowledge, properly authenticated, and not shown by the opponent to lack trustworthiness. Restated financial reports generally satisfy this rule, so long as they are properly authenticated. See SEC v. Jasper, 678 F.3d 1116, 1122-23 (9th Cir. 2012).

In Jasper, the SEC alleged that Carl Jasper, a CFO of a publicly traded semiconductor company called Maxim, had concealed expenses and overstated income via a scheme to illegally back-date stock options. Id. at 1119. Maxim began an internal investigation into the backdating, and announced in September 2008 that "previously filed financial statements for our fiscal years ended in 1997 through 2005 . . . should no longer be relied upon." Id. at 1121. It issued a restatement of the financials, disclosing a reduction in pre-tax income for the period from 2000 through 2005. Id.

Jasper argued on appeal that the district court had abused its discretion in admitting the document because it represented "the culmination of an extraordinary two-year investigation costing tens of millions of dollars, and represented the technical accounting conclusions prepared many years after the facts, by teams of outside investigators and accountants with no personal knowledge." Id. Furthermore, the restatement "took place in a context of tremendous liability risk, and was explicitly created with an eye toward pending litigation." Id. The Ninth Circuit rejected this argument, holding that the restatement was admissible as a business record "because it was 'a report made at or near the time of the accounting review by those with knowledge of Maxim's books,' and '[t]he circumstances of its creation do not indicate that it lacks trustworthiness.'" Id. at 1122 (emphasis added). It continued: "That is, the [restatement] is a business record of the accounting review itself, not of the misconduct that gave rise to the need for the restatement." Id. at 1123.

Hussain sought to distinguish Jasper by arguing that the restatement was the product of a "special one-off investigation," not a record kept in the course of a regularly conducted business activity. See dkt. 158 at 3-4. He cited Paddack v. Dave Christensen, Inc., where the Ninth Circuit held that "audit reports" prepared by an accounting firm in the course of a "special audit" ordered by the Trustee of certain trust funds were not business records. 745 F.2d 1254, 1258-59 (9th Cir. 1984). The difference is that the restatements here and in Jasper were prepared as part of an ongoing duty by the company to correctly state its financials. While the trustees who ordered the audit reports in Paddack "had no regular compliance audit procedure," Yelland testified that he had a continuing obligation "to ensure that the accounts show a true and fair view of the balance sheet and profit and loss account." Tr. at 5048-49 (Yelland). Like the 10-K in Jasper, then, the restatement here "was a paradigmatic 'financial statement audit' and not a 'special audit,'" as in Paddack. See Jasper, 678 F.3d at 1123.

Hussain also contended that the restatement should not be admitted because it lacked trustworthiness. First, he argued that HP ordered the investigation—and subsequent publication of the restatement—with litigation in mind. But this does not differentiate this case from Jasper, where the restatement was prepared "in a context of tremendous liability risk." 678 F.3d at 1122-23. Because, as in Jasper, the restatement was prepared pursuant to an ongoing duty, the concern the Paddack court identified that admitting the special audit "would allow any firm to produce 'business records' that would be automatically admissible" whenever it sought an advantage in litigation, 745 F.2d at 1258, was accordingly not present here.

Hussain also argued the restatement was not trustworthy because it included a disclaimer from HP's auditor, Ernst & Young ("E&Y"), stating that E&Y "[was] not able to obtain sufficient appropriate audit evidence to provide a basis for an audit opinion." Restatement at 12. But these concerns go to the weight of the evidence rather than its admissibility. See United States v. Scholl, 166 F.3d 964, 978 (9th Cir. 1999) (quoting United States v. Keplinger, 776 F.2d 678, 694 (7th Cir. 1985)) ("Generally, objections that an exhibit may contain inaccuracies, ambiguities, or omissions go to the weight and not the admissibility of the evidence."). This is particularly true given that E&Y only said it did not express an opinion one way or another regarding the restatement's accuracy—not that it was not accurate. Accordingly, the restatement was properly admitted as a business record.

The defendant also argued that Yelland's testimony should have been excluded because he was not qualified to opine on international accounting standards. But Hussain failed to explain why the distinction among accounting rules mattered in terms of revenue recognition; the restatement indicated that the change principally affected the expensing of research and development costs. See Restatement at 3, 14. In any event, this argument similarly goes to the weight of Yelland's testimony, not its admissibility.

c. Evidence regarding HP's post-acquisition actions

In its Rule 33 motion, the defense focuses most of its fire on the Court's exclusion of certain evidence regarding what HP did or did not do following its acquisition of Autonomy. The defense sought to put on evidence suggesting that HP witnesses were not actually surprised when they learned that Autonomy had been selling a significant amount of hardware; that it was HP's failure to integrate Autonomy, not any misrepresentations by Hussain, that caused the deal to fail; and that HP's purchase-price accounting continued to "value" Autonomy at $11 billion even after the alleged fraud had been revealed. Rule 33 Mot. at 1-2. The Court excluded much of this evidence under Rules 401 and 403, concluding that the evidence was generally much less probative than the defense believed; represented impeachment on collateral issues; and would be unduly prejudicial, particularly given the defense's improper attempts to encourage the jury to decide Hussain's guilt by weighing the relative culpability of Hussain and HP. See supra Part I.C.2.

The defense suggests that the evidence it sought to introduce was of equal probative value to the restatement, and that if the Court admitted one, it had to admit the other. Not so. The restatement went directly to the question of falsity. The defense evidence, on the other hand, consisted largely of impeachment on collateral or irrelevant issues, as described below. The defense's more general arguments regarding the procedure by which the Court addressed evidentiary objections are addressed at the end of this section.

i. Evidence regarding when HP learned of hardware sales

The defense sought to introduce evidence which it claimed showed that HP learned of Autonomy's hardware sales in October or November of 2011, but nevertheless waited a year before "crying fraud." The defense claims that this evidence tends to establish that: (1) there was no fraud, because the hardware sales were "open and obvious"; (2) any misrepresentations regarding hardware were immaterial, because HP did not appear to react to the news with surprise and/or outrage; and (3) Hussain lacked the requisite intent: if he knew he was going to join HP, and that it would immediately have access to Autonomy's books and records, he would not have defrauded HP.

1. "Open and obvious" argument. To the extent the defense argues that evidence of what HP learned about Autonomy's hardware sales shortly after the acquisition gives rise to an inference that there was no fraud because any hardware sales were "open and obvious," it is wrong. Any information HP received after the acquisition closed simply is not relevant for purposes of Rule 401 to the question of whether hardware sales should have been obvious to HP prior to the acquisition. The defense had every opportunity to argue that Autonomy's hardware sales were obvious during the due diligence process, and to introduce evidence on that point.

In the alternative, even if this evidence were relevant, its relevance is quite limited, and the risk that jurors would become confused about which documents HP had access to after the acquisition and which documents it had access to before the acquisition was sufficiently great that it was proper to exclude the evidence under Rule 403. Meanwhile, to the extent the defense argues that this post-acquisition evidence would have helped jurors determine whether HP asked questions that should have elicited information regarding the true nature of Autonomy's hardware sales, it is also wrong. The actual questions HP asked were in evidence, and the defense had every opportunity to cross-examine the government's witnesses on what was actually asked. Post-acquisition evidence could not have added anything to the assessment of what questions were asked pre-acquisition.

2. Materiality argument. The defense also states that it wished to use evidence of HP's post-acquisition analysis of Autonomy's books to impeach testimony by government witnesses that they were stunned to learn of the extent and nature of HP's hardware sales, and that had they known of those sales, this would have affected HP's valuation of Autonomy. The Court agrees that evidence of words or actions that tend to contradict the defendants' in-court assertions regarding the importance of the hardware misrepresentations would be relevant impeachment material. However, the probative value of this evidence was limited by two factors. First, materiality is an objective standard, and actual reliance is not an element of any of the charges the government brought. See United States v. Lindsey, 850 F.3d 1009, 1014 (9th Cir. 2017). The misrepresentation need only have an objective propensity to influence its intended target. Id. So to the extent that HP had an anomalous approach to valuation, the evidence is not relevant at all. Second, the evidence the defense proffered did not actually tend to contradict the witnesses' testimony. Finally, this line of inquiry would have led to a confusing and time-consuming mini-trial on an attempt to impeach on a collateral issue. Accordingly, it was well within the Court's discretion to exclude the evidence under Rule 403.

Reliance is not an element of either the wire fraud or securities fraud statutes at issue here; the government was not required to prove that HP actually relied on Hussain's misrepresentations. So, to the extent the evidence tended to show that hardware sales were not important to HP, that fact itself would only have been relevant insofar as HP's view of things was indicative of the views of the reasonable, ordinary purchaser of Autonomy securities. In other words, the HP witnesses' testimony on this point was no more crucial to the government's case than the testimony of the market analysts.

The defense's task, accordingly, was to put on evidence that would tend to rebut or impeach the assertions of not only the HP witnesses but also the market analysts who testified that Autonomy's misrepresentations regarding its software sales were material—that is, increased the company's perceived value. But it offered no evidence to rebut market analysts' testimony that pure software companies are far more valuable than mixed hardware/software companies because the former achieve much higher margins, and thus trade at a higher price. Tr. at 870 (Geall [market analyst]); 4698 (Morland [market analyst]).

Instead, it sought to argue that HP's valuation of Autonomy did not depend on Autonomy's actual margins or projected growth rate, but instead only on the quality of its product offerings. The reason HP bought Autonomy "wasn't the books and records. It was the IDOL software." Tr. at 5846 (def. closing). "Mr. Apotheker didn't care what he paid." Id. at 5942. For the fact-finder to determine that the hardware misrepresentations were not material, however, she would have to make the further inference that the typical purchaser of Autonomy shares would be insensitive to the price at which it was currently trading—an inference that would have been wholly unsupported by the evidence, even if the post-closing evidence the defense sought to introduce had not been excluded. If a reasonable buyer in HP's position would still have acquired Autonomy even if it had known of the hardware sales, but would have paid less, then the misrepresentations regarding hardware sales were material. So even if the defense could have shown that HP was indifferent to Autonomy's hardware sales (and hence its actual market value), this would only tend to establish that HP overpaid—not that the misrepresentations were immaterial. There was no evidence at all suggesting that, even if HP did not negotiate on price, the ordinary, reasonable purchaser of securities would not have.

Even so, the evidence the defense sought to introduce did not actually tend to rebut government witnesses' contention that HP did not learn until months after the acquisition that Autonomy was selling hardware, separate and apart from any software sales. For instance, the defense sought to introduce an e-mail exchange between Sarin and HP employee Kathryn Harvey in which Harvey wrote HP had learned that Autonomy "had approx. $100/M year [sic.] in revenue coming from . . . Dell HW products." Ex. 2451. Sarin replied: "[Autonomy is] a predominantly software company with little to no services or hardware. I am guessing they are trying to grow their 'appliance' business i.e. Autonomy software bundled on industry-standard Dell hardware. I suspect this is sell-through revenue where they are getting a margin as they sell Dell appliances. . . . Once we learn more what kind of Dell revenue Autonomy is getting, we can have a better view." Ex. 2451. This does not suggest that Sarin "knew the truth" about the hardware sales—indeed, it suggests the contrary. And in any event, the Court allowed the defense to inquire into Sarin's response to Harvey's e-mail, further reducing the probative value of additional investigations of the topic. See Tr. at 3715.

Similarly, the defense sought to introduce evidence that Autonomy's books, received by HP following the acquisition, listed over $100 million in hardware revenue, and that E&Y accountants had e-mailed people within HP describing their knowledge that Autonomy had completed over $100 million in hardware sales. But this is consistent with Autonomy's representations during the due-diligence process that it sold hardware with software loaded on it, and in connection with marketing of software products. See supra Part I.B.1.b.

In addition to being of little probative value, this evidence would have been collateral and confusing. Allowing the defense to argue that HP did not react to the news of hardware sales with sufficient surprise, and that this tended to impeach witnesses' testimony that the alleged misrepresentations regarding hardware sales were important to HP, would have produced a mini-trial on the issue of what HP learned when, how it reacted, and why it reacted the way it did. There are any number of reasons HP might not have responded by immediately announcing to the public that it had been defrauded, including considerations of business strategy and compliance with the securities laws. As the government argued at trial, "[I]f the defendant is permitted to introduce such evidence, the government will be required to respond by calling percipient HP witnesses to explain exactly what they learned, how that information fit with their specific responsibilities within HP, and what they did or did not do with the information and why." Dkt. 308 at 3. This is far afield of the materiality inquiry.

3. Intent argument. The Court agreed that the evidence that HP received Autonomy's books and records shortly after the acquisition was relevant to Hussain's intent. However, it enabled the defense to make its intent argument by requiring the government to either drop its objections to the evidence the defense sought to admit, or stipulate that HP received all of Autonomy's books and records immediately after the acquisition. See dkt. 368; Tr. at 5700. Indeed, the defense relied on this stipulation in its closing argument. Tr. at 5940 ("When the deal closed on October 3rd, Hewlett Packard got all the books and records of Autonomy and it got the Deloitte work papers and could review them, and those work papers and books and records clearly showed—this is the stipulation that says that—clearly showed the hardware sales. You couldn't miss it."). The defense protests that the stipulation does not have the same visceral impact the books themselves would have, Rule 33 Mot. at 14; but the government never disputed that HP received the books, and indeed Yelland's testimony suggested that HP had access to everything it needed to restate Autonomy's financial reports. Introducing the books themselves would have required testimony regarding bookkeeping practices and what the books did or did not show, further complicating a case that was already plenty complex. The additional probative value of the books themselves, if any, was substantially outweighed by the dangers of waste of time and juror confusion.

ii. HP's various write-downs and integration efforts

The defense argues that it should have been allowed to introduce evidence regarding HP's $8.8 billion write-down of its Autonomy purchase, as well as of HP's purchases of other companies. The Court excluded evidence of the Autonomy write-down on the ground that the probative value of how HP accounted for Autonomy on its own books following the acquisition was substantially outweighed by the danger of undue prejudice, confusion, and waste of time. Tr. at 3503. Instead, the Court sought to focus the evidence on whether Autonomy's accounting was appropriate, and on whether the alleged misrepresentations were material from the perspective of a reasonable buyer at the time the misrepresentations were made.

The defense contends that the government introduced the evidence of HP's Autonomy write-down through a witness who was an HP shareholder. This is not quite right. The witness said he was prompted to sell his shares by "[t]he press release that announced the write-down of the value of Autonomy purchased by HP." Tr. at 371 (Upton [HP shareholder]). This is equivalent to saying he sold the shares once HP announced it had been defrauded. The government did not introduce any detail about HP's write-down, and indeed when it accidentally elicited the amount of the write-down with a later witness, the Court instructed the jury to disregard it. Tr. at 4179-80.

So the argument that the government was allowed to discuss the write-down while the defense was not is simply untrue. Indeed, the defense brought out many of the write-down-related issues on which it sought to introduce extrinsic evidence during cross-examination:

Q. You didn't know that HP had written down the Compaq acquisition in May of 2012?

A. To be honest, I don't recall. I mean, HP made many, many acquisitions and some it wrote down and some it didn't.

Q. You recall the write-down of a company called EDS, don't you?

A. Yes, I do.

...

Q. And as you testified earlier, you don't know what caused the write-down of the Autonomy acquisition?

A. No, I don't.

What the defense is really saying, then, is that it should have been allowed to extensively impeach the testimony that there was a write-down—and the implication that the write-down was justified—with a blizzard of extrinsic evidence relating to HP's integration of Autonomy and its failure to properly manage other acquisitions. Rule 33 Mot. at 16-17. The relevance of this evidence is highly attenuated. Actual reliance and loss causation are not elements of criminal fraud charges, so this evidence is only possibly relevant insofar as it bears on materiality. HP's public representations that it lost money on the Autonomy acquisition bear very tangentially, if at all, on the materiality issue; indeed, even a successful integration of Autonomy into HP would have been consistent with a finding that Hussain had committed fraud.

With this in mind, the Court excluded all but the most general evidence regarding the outcome of the Autonomy acquisition for HP. See Tr. at 3492-93. It was the defense, not the government, that elicited the most detail about the write-down. See Tr. at 3144-45 (Anderson [auditor]); 5183-84 (Yelland [HP accountant]). The defense cannot complain about not being allowed to extensively impeach details that it itself introduced. Allowing the defendant to make an extensive inquiry into how HP accounted for the acquisition—separate and apart from how it restated Autonomy's own books—would have been to allow impeachment on a collateral issue. This would have been time-consuming, requiring the government to introduce evidence the Court had excluded regarding the amount of and reasons behind the write-down; confusing, introducing yet another standard for judging the extent of Autonomy's misrepresentations; and prejudicial, feeding into the impermissible inference that the jury was required to determine whether Autonomy or HP was more at fault for the failed acquisition. See supra Part I.C.2. It was properly excluded under Rule 403.

iii. Evidence regarding HP's post-acquisition assessments of Autonomy's value

Given that the government was allowed to introduce evidence regarding HP's post-acquisition valuation of Autonomy by way of the restatement, the defense "should have been permitted to introduce evidence of other valuations of Autonomy," it insists. Rule 33 Mot. at 15. Had HP in fact completed a clean-slate assessment of Autonomy's value following the acquisition, apart from the restatement, the Court would be inclined to agree. But this was not the case.

Two of the reports the defense points to are 2012 reports (one prepared by Duff & Phelps, the other by Economic Partners) "valuing" Autonomy at around $11 billion. The problem is that these reports employed purchase-price allocation: a form of accounting that allocates the price paid for a company into various categories of assets and liabilities. These exercises took the purchase price for granted in determining how to allocate that price on HP's own books; they did not claim to reach a conclusion regarding whether the price was a fair or "correct" one.

The other assessment the defense raises is a report by E&Y that evaluated whether the impairment charge taken by HP when it announced that Autonomy had misrepresented its finances "should be reflected in the original purchase accounting." See E&Y Forecast Memo (dkt. 351-25). E&Y concluded that it would be appropriate to record the impairment in 2012, not 2011. Id. at 4. This does not tend to contradict or impeach the restatement, and makes no representations about whether Autonomy made misstatements, or whether those misstatements were material. Accordingly, it is not relevant.

Even if these reports did have some relevance to the issues in this case, they were highly confusing, needlessly introducing new accounting standards into the case. Accordingly, this evidence was also properly excluded under Rule 403. The Court gave the defense every opportunity to impeach Yelland and the circumstances under which the restatement was prepared, to put on expert testimony criticizing the methods used in the restatement, and to offer its own evidence suggesting alternative accounting treatments of the transactions at issue. It did not do so.

iv. Defense complaints about process

The defense also suggests that the manner in which the Court went about making evidentiary rulings indicates that the Court applied different rules to the respective parties. The defense contends that the Court initially "placed no limits on introducing evidence of events that occurred" after HP acquired Autonomy; decided in the middle of trial that no such evidence could come in; then changed its mind and allowed the government's post-closing evidence to come in, while barring the defense's similar evidence. As a result, Hussain was unable to put on a "complete defense." Rule 33 Mot. at 12.

This is not quite accurate. While its views regarding the admissibility of the government's evidence did indeed evolve over the course of the trial, the Court consistently represented throughout the trial that it was not inclined to admit post-acquisition evidence that did not bear on the truth or falsity of the restatement, or otherwise go directly to the issues in this case. The answer to the defense's complaint that it did not understand exactly what defense evidence the Court was inclined to allow until late in the trial is that it did not present the Court with its proposed evidence until that point.

The Court's perspective is as follows. At a pre-trial conference, in response to a motion in limine filed by the government, the Court ruled that the restatement and Yelland's testimony on it were admissible. Pre-Trial Tr. (dkt. 249) at 7. The defense did not make a similar motion to introduce "post-closing" evidence. However, the government moved to exclude defense evidence that would tend to put the focus on HP's failures to make the deal a success post-acquisition, among other things. See dkt. 147. The Court declined to rule on this motion because the defense had not yet offered specific examples of what evidence it intended to introduce regarding HP's conduct. Pre-Trial Tr. at 15-16. Indeed, the defense suggested that the Court defer. Id. at 17.

The Court did, however, offer general guidance on what types of evidence it would and would not allow in: "I think [the defense] can't say or they shouldn't say: Well, look, Hewlett-Packard was negligent. They were careless. They could have done this, they could have done that. I don't know that that's relevant." Id. at 16. The defense appeared to represent that it did not intend to put on such evidence, clarifying that it was focused primarily on the due diligence period: "[W]hat we need to introduce as evidence is questions that HP asked, questions that they chose not to ask during due diligence. . . . If somebody during due diligence expressed that HP was not interested in Autonomy's hardware sales, well, that is plainly relevant to disproving an allegation, an allegation that's in the indictment, that Mr. Hussain and others concealed hardware sales." Id. at 16-17. The Court agreed, and indeed, went on to allow the parties to introduce every piece of evidence they proffered regarding the correspondence between HP and Autonomy during the due diligence period.

During trial, the defense represented for the first time that it intended to go substantially beyond the scope of due diligence and introduce evidence regarding HP's purported mismanagement of Autonomy's business after the acquisition. Tr. at 3488-89. In the defense's view, this became fair game when the government elicited Upton's testimony that the value of his shares decreased following the acquisition (notably, the defense did not object to this testimony, and was allowed a full opportunity to call it into question on cross-examination, as described above). Id. The government elicited the testimony simply to "close the loop" on the shareholder's testimony, in the Court's view. Id. at 3493. Because, again, actual loss is not an element of either wire fraud or securities fraud, the Court explained that it viewed the contradictory evidence the defense intended to put on as irrelevant, or, at best, as impeachment on a collateral issue that was inadmissible under Rule 403. Id. at 3489-90. Seeking to clear up any confusion on its views regarding the admissibility of post-acquisition evidence, the Court reiterated that it generally did not find HP's actions to be relevant, apart from the questions it asked during due diligence: "[I'm concerned about an] open-ended inquiry into what else is going on at HP post-acquisition." Id. at 3495.

The Court did waver in its views on whether the restatement should be admitted, along with Yelland's testimony—but if this wavering made things more difficult for anyone, it was the government, not the defense. The Court's evaluation of the restatement—which went to the issue of whether there had been a misrepresentation—was not linked to its evaluation of evidence tending to show whether or not HP lost money on the acquisition, which (arguably) went to materiality.

The government planned to buttress its evidence regarding the various transactions in which Autonomy had engaged with testimony from an accountant who had reviewed Autonomy's books in light of the information Autonomy had allegedly concealed. That accountant would either be Yelland, an expert witness, or both.

As the trial progressed, the Court grew quite concerned that the evidence was coming in behind schedule, risking the possibility of a mistrial. The defense suggested that it would spend weeks impeaching Yelland if the government called him: "[H]e opens every can of worms, every door that we've been complaining about. . . . [W]e are going to challenge everything he has done and we are going to challenge his motivation and we are going to challenge the circumstances under which he" prepared the restatement. Id. at 3879, 3890. In the Court's view, this would have been perfectly proper.

Wary of how long Yelland's testimony would take, and the Court strongly encouraged the government—"forced" may be the better term—to put on its accounting expert instead of Yelland. The government said that it preferred Yelland because he had evaluated assessed every single transaction, and because Yelland had attested to the restatement's accuracy under penalty of sanctions. Id. at 3889, 3892, 3897. Nevertheless, the government reluctantly agreed to put on its expert instead, assuming the expert could testify to the restatement, and assuming that it would be able to call Yelland if the defense sought to argue that the expert's report was too cursory. Id. at 3889, 3898-89. The Court cautioned that it would revisit its decision to exclude Yelland after hearing the expert testify. Id. at 3911.

However, it soon became clear that the defense and the government had a different understanding of the Court's ruling, and that the government would need to call Yelland in any event to fill holes in its expert's report. Id. at 4769-4820. Accordingly, the Court gave the government its choice of who it would put on, and the government elected to put on Yelland. To the extent the Court believed it had a legitimate basis to exclude Yelland and perhaps the restatement in the first place, it was mistaken, and acknowledged as much at the time. See id. at 4820 ("The Government is entitled to prove their case the way they want to, assuming there is evidence and not cumulative, and I can't say, you know, this witness is better than that witness for you.").

The statement the defense quotes for the proposition that the Court waffled on its evidence—"if the Government produces this evidence, don't I have to let all this other evidence in? I think the answer to that is absolutely yes"—in fact referred specifically to the evidence the defense had promised impeaching Yelland. As it turned out, however, the defense did not, in the Court's view, offer much proper impeachment evidence. Instead, it offered evidence on a host of issues unrelated to the restatement or Yelland, as described above. The Court never intended to allow this evidence in, for the reasons already given. (While the Court prevented the defense from getting into the weeds of the circumstances of the civil litigation between HP and Hussain on relevance and/or 403 grounds, the defense received a full opportunity to question Yelland regarding any motivation he might have had to aid that litigation in preparing the restatement.)

The Court regrets the confusion and inconvenience its rulings on the restatement may have caused the lawyers in this case, but it has no doubt that its final evidentiary rulings regarding both sides' post-acquisition evidence were both justified and equitable. The views the Court expressed regarding the defense's post-acquisition evidence were consistent throughout the trial. If the defense wished to receive more specific evidentiary rulings earlier than it did, it should have asked for them.

d. Other evidentiary rulings

The defense objected to a number of other evidentiary rulings during trial. Though it does not specifically mention these arguments in its new trial motion, it has preserved its objections for appeal. Accordingly, the Court addresses them briefly.

i. "Top 40" summary charts

The government sought to admit two sets of summary charts pursuant to Federal Rule of Evidence 1006, which allows a party to summarize voluminous evidence where the summary is "accurate, authentic and properly introduced." United States v. Scales, 594 F.2d 558, 563 (6th Cir. 1979). The first was a series of charts summarizing revenue that the government claimed "needed to be adjusted as a result of the alleged fraud"—in effect an expanded version of the restatement. See dkt. 144 at 3. The second was a single chart that showed Autonomy's "top 40" customers and compared the revenue for these customers based on underlying Autonomy records with the revenue information provided to HP during the due diligence process. See id. at 4 and Ex. B. The government sought to use the latter chart to show that Autonomy's actual list of its top 40 customers by revenue was different from what it represented to HP.

The Court excluded the charts summarizing Autonomy's revenue on the ground that the charts were argumentative because the accounting in the restatement was disputed. However, it allowed the government to introduce those charts as demonstratives.

Meanwhile, the Court allowed in the summary chart of the top 40 list. The defense conceded that the chart accurately summarized Autonomy's revenue inclusive of sales to resellers and hardware sales. Tr. at 5075-76. It contended that the chart was improperly argumentative, however, because the parties disputed whether that was in fact what HP had asked for. Id. But this dispute did not make the chart itself argumentative. The parties agreed that the chart displayed precisely what the government said it did: "underlying revenue information from the billing system," irrespective of the nature of the customer. Tr. at 5386-94 (Bryant [gov't special agent]). The jury was left to decide whether that was in fact what HP had asked for during due diligence.

ii. HP's tax treatment of Autonomy acquisition

The defense sought to introduce evidence regarding HP's attempt to reduce its tax bill related to the Autonomy acquisition. See dkt. 145. The defense argued this was relevant because it tended to show that Hussain's misrepresentations were not material. See dkt. 159. In other words, Hussain argued that HP's tax windfall somehow balanced out the impact of Hussain's alleged misrepresentations, such that HP would have paid the same price for Autonomy even if it had known about his misrepresentations. But HP's reduction in taxes was much smaller than the claimed loss, and in any event, again, actual loss is not an element of a criminal fraud charge. Accordingly, this evidence was properly excluded.

iii. Evidence regarding attempts to exclude Daud Khan from analyst calls

The defense moved to exclude evidence that Hussain and his alleged co-conspirators retaliated against Daud Khan, a market analyst, for questioning Autonomy's accounting by, among other things, excluding Khan from Autonomy's earning calls. See supra Part I.B.3.d. First, the defense contended that the evidence should be excluded under Rule 401 because it was irrelevant. See dkt. 151. On the contrary, the evidence gave rise to a permissible inference that Hussain sought to cover up the misrepresentations Autonomy was making to the market, which in turn made it more likely that he was aware of the misrepresentations and involved in the alleged conspiracy. While Hussain argued that the government failed to tie the actions against Khan specifically to Hussain, in fact the government put on sufficient evidence for jurors to infer that Hussain knew about or directed the actions against Khan, given his role as a director of the company and his involvement in analyst calls. See United States v. Boise, 916 F.2d 497, 502 (9th Cir. 1990) (evidence of prior injuries relevant "only if the jury could reasonably conclude that the injuries occurred and that [the defendant] inflicted them"). Moreover, Khan testified that he had a face-to-face meeting with Hussain and other directors in which Lynch threatened consequences if he published an unfavorable note. Tr. at 2986-87 (Khan). The evidence was properly admitted.

iv. Evidence regarding Autonomy's termination of Brent Hogenson, Percy Tejada, and Reena Presad

Hussain also moved to exclude evidence tending to establish that he retaliated against Brent Hogenson and other Autonomy employees for pointing out accounting irregularities during the period of the charged conduct. See dkt. 151. While the defense argued that Hussain "had little connection to these events," the testimony at trial was to the contrary. See supra Part I.B.3.d. The evidence was relevant and was properly admitted.

v. Evidence of Autonomy's correspondence with the UK Financial Reporting Review Panel (FRRP)

The defense moved to exclude correspondence with the UK Financial Reporting Review Panel (FRRP) in which Hussain made what the government claims were misrepresentations about Autonomy's finances. See dkt. 151. The defense argued that the statements were not materially misleading; represented impermissible propensity evidence that merely revealed Hussain had a tendency to lie; and would confuse jurors by making them think Hussain was on trial for defrauding British regulators, rather than HP. However, the government laid an appropriate foundation for the evidence, making juror confusion less likely; the alleged misrepresentations were part of an ongoing course of conduct that culminated in Autonomy's misrepresentations to HP (said another way, Autonomy needed to lie to the FRRP if it was going to lie to HP); and the government put on competent evidence tending to show that Hussain did indeed make misrepresentations. See Tr. at 4920-51 (Lindsell [FRRP witness]). The evidence was properly admitted.

vi. Evidence of FileTek accounting

Finally, Hussain moved to exclude evidence that FileTek treated its transactions with Autonomy as barter deals on its books, declining to recognize revenue. See dkt. 272. The defense argued that the impact of the transactions on FileTek was different than on Autonomy, and that different accounting standards applied to each firm (US GAAP for FileTek, IFRS for Autonomy). However, the defense failed to explain how these transactions could have been barter deals from FileTek's perspective but not Autonomy's, and failed to identify specific differences between the accounting rules that would have justified different treatment. Accordingly, these were issues for cross-examination, not rationales for exclusion.

e. Prejudice

Even if the Court had sided with Hussain on these evidentiary issues, the defense cannot demonstrate prejudice. The testimony of each of the 37 witnesses at trial supported the government's case, and the Court infers from the verdict that the jury found those witnesses credible. The government presented evidence from multiple witnesses to support each element of the charged conduct. Even if the defense had been allowed to present all the evidence it sought to, that evidence would not have been particularly probative, as described above; and even if the jury had adopted all the inferences the defense hoped it would from that evidence, this still would not have negated the government's evidence on any element of any of the charged offenses, again as described above. See supra Part I.B. The government's evidence was exhaustive and exhausting, and would have been even more so had the Court not counseled it to scale back. It had multiple sources of corroboration for a great number of facts. The jury would all but certainly have reached the same verdict no matter the Court's evidentiary rulings.

4. Jury Instructions

The defense next argues that Hussain is entitled to a new trial on the ground of instructional error. See Rule 33 Mot. at 19 n.7. The Court reviews the defense's various objections to its jury instructions. See Jury Inst. (dkt. 374).

a. Elements of conspiracy charge

The defense argued at trial that conspiracy to commit wire fraud under 18 U.S.C. § 1349 requires an overt act. But the case it cited for this proposition, United States v. Green, 592 F.3d 1057, 1067 (9th Cir. 2010), is plainly inapplicable. The Green court was applying 18 U.S.C. § 371, which expressly includes an overt act requirement, whereas Hussain has been charged under § 1349, which does not.

Meanwhile, any argument that § 1349 implicitly features an overt-act requirement is unavailing. In Whitfield v. United States, the Supreme Court unanimously held that 18 U.S.C. § 1956(h), conspiracy to commit money laundering, did not require proof of an overt act because the statute's text does not require one. 543 U.S. 209 (2005). Applying the plain meaning rule, it declined to look beyond the statutory text. Id. at 213-17. The Court noted that it was especially wary of reading an overt-act requirement into conspiracy statutes because none was required at common law. Id. at 209, 213-14.

While the Ninth Circuit has not yet weighed in on the issue, a number of other courts have relied on Whitley to hold that § 1349 does not require an overt act. See, e.g., United States v. Roy, 783 F.3d 418, 420 (2d Cir. 2015); United States v. Pascacio-Rodriguez, 749 F.3d 353, 363, 364 n.49 (5th Cir. 2014); United States v. Rogers, 769 F.3d 372, 380-82 (6th Cir. 2014); United States v. Fishman, 645 F.3d 1175, 1186 (10th Cir. 2011) ; United States v. Eason, 579 F. App'x 807, 810 n.3 (11th Cir. 2014) (unpublished); United States v. Chinasa, 489 F. App'x 682, 685-86 (4th Cir. 2012) (unpublished). Moreover, the Ninth Circuit has interpreted Whitfield in the same manner as these courts do, citing it for the principle that a court need not look beyond a statute's plain meaning where the statutory text is clear. United States v. 144,774 Pounds of Blue King Crab, 410 F.3d 1131, 1136 (9th Cir. 2005); United States v. TRW Rifle 7.62x51mm Caliber, 447 F.3d 686, 691 (9th Cir. 2006). This Court accordingly finds that § 1349 contains no overt act requirement.

b. Conspiracy: specific unanimity instruction

The Court instructed the jury on the conspiracy count as follows:

First, beginning in or about Autonomy's first quarter 2009, which began in January of 2009, and continuing until in or about October 2011, there was an agreement between two or more persons to commit wire fraud; and

Second, the defendant became a member of the conspiracy knowing of at least one of its objects and intending to help accomplish it.

"Normally, a general instruction on the requirement of unanimity suffices to instruct the jury that they must be unanimous on whatever specifications form the basis of the guilty verdict." United States v. Lapier, 796 F.3d 1090, 1096 (9th Cir. 2015) (internal quotation marks omitted). "However, a specific unanimity instruction is required if there is a 'genuine possibility of jury confusion' or a possibility 'that a conviction may occur as the result of different jurors concluding that the defendant committed different acts.'" Id. In conspiracy cases, jurors must unanimously agree that a defendant entered into an agreement to "pursue a particular criminal goal." Id. Factors to consider in determining whether a genuine possibility of confusion exists include "the text of the indictment, the clarity and presentation of the government's argument, the complexity of the evidence, and the clarity or ambiguity of the jury instructions." Id., at 1097.

Relevant factors in determining whether there is a single conspiracy or multiple conspiracies include "the nature of the scheme; the identity of the participants; the quality, frequency and duration of each conspirator's transactions; and the commonality of time and goals." United States v. Zemek, 634 F.2d 1159, 1168 (9th Cir. 1980). It is important to keep in mind that "[a] single conspiracy can include subgroups or subagreements and the evidence does not have to exclude every hypothesis other than that of a single conspiracy." United States v. Bauer, 84 F.3d 1549, 1560 (9th Cir. 1996). Evidence that a conspiracy involves a shifting cast of collaborators and transactional structures is not necessarily inconsistent with a single conspiracy. United States v. Williams, 673 F. App'x 620, 622 (9th Cir. 2016) (unpublished).

In Lapier, the government acknowledged during a colloquy with the district court that it had presented evidence of two separate conspiracies in which the defendant had been involved. 796 F.3d at 1097. The appellate court found that there was "no evidence" of any overlap between the two conspiracies. Id. Under these circumstances, a "genuine possibility of jury confusion" existed because jurors might have all agreed that the defendant had committed one of the two conspiracies, while disagreeing on which of the two it was. Id. at 1097-98.

The defense argues it was "possible" that a jury could find that multiple conspiracies existed in this case, as well. Rule 33 Mot. at 20. But the standard is not whether it is possible that jurors might find multiple conspiracies, but whether there is a "genuine possibility of confusion" among jurors. Lapier, 796 F.3d at 1096 (emphasis added). There was no genuine possibility here. The government charged Hussain with conspiracy to commit wire fraud, and introduced no evidence of wires except those used to defraud HP—leaving no genuine possibility that the finder of fact would find a conspiracy existed to commit the other acts Hussain mentions in his motion. See Rule 33 Mot. at 21. The government never suggested that Hussain was involved in "suspected bank fraud by Microlink CFO Alan Rizek" or "an effort by Microlink CEO Dave Truitt to steal Autonomy software," as the defense implies. See id. And while the defense hypothesizes that jurors may have convicted Hussain of securities fraud based on a Pinkerton theory, and may have found different conspiracies to commit securities fraud, the jury instruction on the securities-fraud count was sufficiently specific to prevent this possibility because it told jurors they had to find that the scheme was "in connection with the purchase or sale of securities of Hewlett-Packard Company." Jury Inst. at 25 (emphasis added).

Throughout trial, the government consistently described the conspiracy as an attempt to deceive potential acquirers into paying more for the company than it was really worth. See, e.g., Tr. at 5823-24 (gov't closing). The defense argues that the jury might have found different conspiracies involving particular categories of fraudulent transactions carried out by Autonomy. Rule 33 Mot. at 20-21. But while the government introduced evidence of a number of different transactions, nothing at trial suggested that these were part of separate conspiracies. As the government described the evidence at trial, the transactions were in fact quite similar to each other: "The evidence in this case leaves no doubt that Autonomy was buying its own revenue. Strip away the pretext, and what you see is money flowing in a circle again and again and again." Tr. at 5721 (gov't closing). And the government introduced no evidence that the six co-conspirators it identified—Hussain, Egan, Lynch, Kanter, Menell, and Chamberlain—had anything to gain from causing Autonomy to overpay on particular types of transactions. Several activities unified in time that "promote [one] overall goal" are "consistent with one overall agreement." Zemek, 634 F.2d at 1168. And based on the evidence at trial, this strategy could have made sense only if the transactions led to higher valuations of the company through fraudulent accounting.

In sum, the government clearly and consistently argued that Hussain had been involved in a single conspiracy; the evidence, while complex in general, was quite simple on the conspiracy issue; and the jury instructions clearly specified a single conspiracy. See Lapier, 796 F.3d at 1097. There was no basis for a specific unanimity instruction. Indeed, by suggesting that jurors had heard evidence regarding multiple conspiracies when they had not, a specific unanimity instruction could only have caused confusion.

c. Pinkerton and vicarious liability

The Court gave instructions on both Pinkerton and vicarious liability, over the defense's objections. The Pinkerton instruction was appropriate because the government introduced evidence that the defendant was involved in a conspiracy. See supra Part I.B.3.e; III.A.2. Meanwhile, the vicarious liability instruction was appropriate because the elements required for "co-schemer" liability are less demanding than those required for a conspiracy. United States v. Stapleton, 293 F.3d 1111, 1117 (9th Cir. 2002) ("Vicarious liability for substantive counts of mail fraud or wire fraud does not require that a conspiracy be charged or proved."). In other words, the jury could have found that there was no conspiracy (making Pinkerton liability inapplicable), but nevertheless found Hussain vicariously liable for the actions of co-schemers.

d. Wire fraud—victim's conduct

The Court gave the following instruction regarding HP's conduct, over the defense's objection:

You have heard evidence regarding Hewlett Packard's process for deciding whether to purchase Autonomy. You are to consider this evidence to the extent that it helps you determine whether the defendant made false or fraudulent pretenses, representations, or promises as part of a scheme or plan to defraud (see p. 19). You may also consider this evidence to the extent that it helps you determine whether the statements made or facts omitted as part of the alleged scheme were material; that is, whether they had a natural tendency to influence, or were capable of influencing, a person to part with money or property (see p. 19). If you find that these elements have been met, then whether there were additional things Hewlett Packard could have done to avoid being impacted by the alleged misstatements and/or omissions is irrelevant to your verdict.

The defense contended that its only argument regarding HP was that it did not care about certain information, making it more likely that (1) HP did not seek the information the government contended Autonomy deceptively withheld, and (2) any misrepresentations were not material. Id. at 5654-56. However, it would only have been natural for jurors to seek to compare the "blameworthiness" of Hussain and HP in determining whether a conviction was warranted. And the defense encouraged this comparison in its closing argument. See supra Part I.B.2. The instruction was warranted.

5. Closing argument

Finally, the defense contends that the government made an improper argument on rebuttal. Responding to the defense's "battle of the accountants" narrative, the prosecutor said:

[T]here is no duel here, ladies and gentlemen. There is no battle. That's a ruse. No accountant came in here, with all the lawyers and folks over here—no accountant came in here, took an oath, and said Sushovan Hussain and the way he ran the accounting at Autonomy was just fine.

The defense's interpretation of the government's argument may well have been correct had the government not been responding to the defense's argument that the accountants disagreed. However, the prosecutor's statement must be interpreted in that context. The government may respond during closing if defense counsel opens the door. United States v. Williams, 990 F.2d 507, 510 (9th Cir. 1993). Here, what the prosecutor was saying was, "Contrary to what the defense suggested, the accounting witnesses in this case all agreed." The prosecutor was not saying, "The defense did not put on an accounting witness, therefore you must find for the government."

In any event, the Court immediately neutralized any chance of prejudice to the defendant by giving a curative instruction: "The Defense need to prove—needs not to prove anything. They have no burden of proving anything." Tr. at 5971 (rebuttal). So the defense is not entitled to a new trial on this ground. See Parker, 549 F.2d at 1222.

IV. CONCLUSION

For the foregoing reasons, Hussain's motions for a new trial and judgment of acquittal are DENIED.

IT IS SO ORDERED.

Dated: July 30, 2018

/s/_________

CHARLES R. BREYER

United States District Judge