Bamberg v. Goldman, Sachs & Co.

Dragon Systems, Inc. (“Dragon”), a leading voice recognition software company in the late 1990s, needed infusions of capital to continue operations and so sought an acquisition partner. It hired an investment banker, Goldman Sachs (“Goldman”), to assist it. Dragon was acquired in June 2000 by Lernout & Hauspie Speech Products N.V. But Lernout & Hauspie had fraudulently overstated its earnings. When that was learned, bankruptcy ensued for the merged company, and the name Dragon and its technology were sold from the estate.

Naturally, considerable litigation followed out of this debacle, including these suits against Goldman by two groups of Dragon shareholders. Goldman has been found not liable both by a jury on claims of negligent performance of services, gross negligence, intentional and negligent misrepresentation, and breach of fiduciary duty, and also by a court on claims of violations of Mass. Gen. Laws ch. 93A. After a jury found in favor of Goldman on all of plaintiffs’ common-law claims on January 23, 2013, the district court found that Goldman had not engaged in unfair or deceptive conduct in violation of ch. 93A. The two groups of shareholder plaintiffs appeal from the district court’s ruling on their 93A claims, essentially arguing that there is an incongruity between the court’s *43 findings of fact and its non-liability finding, such as to justify reversal. As to the jury verdict, plaintiffs argue that they are entitled to a new trial on their common law claims because of evidentiary errors and erroneous jury instructions. Finding no error, we affirm.

i:

A. Factual Background

We tell the facts as found by the jury and the court.

Plaintiffs and now appellants James and Janet Baker, Robert Roth, and Paul Bam-berg in 1982 founded Dragon, a speech recognition technology company. Dragon, a closely-held corporation headquartered in Newton, Massachusetts, manufactured and sold software which recognized spoken commands and transcribed ordinary conversational speech. The Bakers, Bam-berg, and Roth were principal shareholders in the company and served at various times as members of Dragon’s board of directors and senior management. At the time of the events giving rise to this lawsuit, the Bakers and Seagate Technology (a principal investor in Dragon) owned 90% of the company, while Bamberg and Roth owned 8%. At oral argument, counsel for Bamberg and Roth referred to Bam-berg, Roth, and James Baker as the “brains behind [Dragon’s] technology.” Janet Baker was Dragon’s CEO from 1998 until 1999, when she was asked to resign. The district court found that Janet Baker was “considered difficult to work with” while she was CEO.

By the end of the 1990s, “Dragon had an extensive research and development pipeline for future products and opportunities ... which included speech recognition for mobile telephones and handheld devices.” These were called Dragon’s “golden eggs.” The company was considered “a leader in speech technology products” and was valued at roughly $600 million. Despite Dragon’s apparent eminence in the field, the company’s financial condition was in fact perilous. Dragon lost money in every year of its existence, save one. By the end of the 1990s, Dragon employees and executives were concerned about the company’s ability to make payroll. Dragon began to consider merging with another company in order to obtain a capital infusion so that it could develop the “golden eggs” and make them profitable.

In the fall of 1999, competing bidders Lernout & Hauspie Speech Products N.V. (“L & H”) and Visteon Automotive Systems (“Visteon”), a subsidiary of Ford, each offered to acquire Dragon for approximately $580 million. Dragon then sought out Goldman to be its investment banker. On December 8, 1999, Ellen Chamberlain, as Dragon’s Chief Financial Officer, signed an agreement (the “Engagement Letter”) with Goldman. Goldman agreed to “provide [Dragon] with financial advice and assistance in connection with this potential transaction, which may include performing valuation analyses, searching for a purchaser acceptable to [Dragon], coordinating visits of potential purchasers and assisting [Dragon] in negotiating the financial aspects of the transaction.” 1 The Engagement Letter stated *44 that Goldman would receive a $150,000 quarterly fee, as well as a payment of at least $2 million if the sale were consummated.

Significantly, the Engagement Letter also contained an exculpation clause (“Annex A”) providing that Dragon, Seagate, and Janet Baker would not hold Goldman liable for any derivative claims arising out of Goldman’s services to Dragon “except to the extent that any ... claims ... result from the gross negligence, willful misconduct or bad faith of Goldman Sachs.... ” Janet Baker and Seagate signed the Engagement Letter agreeing to the above-quoted sentence from Annex A (the only provision of the agreement that involved them in their personal capacities).

A prior draft of the Engagement Letter, which had been rejected, had, by contrast, provided that Goldman was engaged by Dragon and by Janet Baker and Seagate in their capacity as stockholders. The earlier rejected draft had also made those stockholders guarantors of Dragon’s obligation to pay Goldman for its services. The final Engagement Letter omitted the reference to “stockholders” because Janet Baker did not wish to be personally liable for Goldman’s fee. Thus, the Engagement Letter was between Dragon (the company) and Goldman (with the exception of the exculpation clause, to which, as noted above, Janet Baker and Seagate also agreed).

Goldman assembled a four-person team to work on the Dragon merger: T. Otey Smith, Alexander Berzofsky, Richard Wayner, and Christopher Fine. Wayner was the leader of the team. The record shows that Wayner was an experienced banker, having been involved in numerous transactions during his career at Goldman. Even before the Engagement Letter was signed, the Goldman team began to involve itself in the process of conducting financial due diligence on L & H. A jury could easily have concluded that Ellen Chamberlain, the CFO of Dragon, was ultimately in charge of the due diligence, and Goldman’s role was to assist her in multiple ways.

On December 16 and 17, 1999, the Goldman team met with both sets of plaintiffs and other senior Dragon management to discuss the buyout proposals. Some of the Dragon management, particularly Bam-berg, had serious reservations about merging with L & H. Bamberg expressed concern over L & H’s employment practices and “questionable financials.” At trial, he testified that “everyone around the table had got the message that I was skeptical about L & H.” The Goldman team identified several potential “issues” with the L & H proposal — for example, the volatility of L & H’s stock, “[pjast accounting practices re: R & D,” and concerns about the percentage of L & H’s growth that was due to organic growth versus growth by acquisition. Dragon’s then-President John Sha-goury testified that Goldman “dicuss[ed] these issues with the Board” and that, for the most part, the Board’s “concerns were allayed with explanation of what those [issues] were all about.”

On December 20, 1999, Dragon entered into a period of exclusive negotiation with Visteon. The negotiations did not prove fruitful, and in February 2000 Dragon and Visteon allowed the exclusivity period to lapse.

Dragon then turned its full attention to a possible merger with L & H. At this point, the Board was “focused on speed and certainty” because of Dragon’s deteri *45 orating financial situation. Dragon needed the capital from a merger.

On February 9, 2000, L & H, for its part and independently of the merger discussions, had an earnings conference call with analysts and released its Q4 1999 earnings. L & H stated during that call that its earnings in Asia had increased over 1000 percent, while growth in the United States and Europe was much slower.

On February 14, the Bakers received a news article by e-mail that both reported the L & H earnings call and raised questions regarding the reported disparity between ample growth in Asia and growth elsewhere. The record does not indicate whether the Bakers raised this news article with other Dragon executives or with Goldman. The district court found that “[pllaintiffs were aware of L & H’s growing Asian revenues, and considered the matter to be an important issue.” Goldman did not have an analyst covering L & H at the time, and no one at Goldman participated in the earnings call or communicated information about the call to anyone at Dragon.

On February 17, 2000, Chamberlain emailed the Goldman team to complain about the lack of progress on doing due diligence on L & H. She “specifically requested that Goldman ‘drive and analyze’ the due diligence” and identified several outstanding issues with respect to the process. The Goldman team continued to send due diligence requests to L & H, but it did not receive satisfactory answers. Consequently, when Goldman prepared draft evaluation books, it “simply included the future revenue projections [it] received from L & H.”

A key event occurred on February 29, 2000, when Goldman faxed Dragon a memo addressed to Dragon and copied to Hale & Dorr, Dragon’s legal counsel. The facsimile transmission sheet instructed the recipient to “forward to Paul Cohen [Dragon’s in-house counsel], Don Waite [Dragon’s CEO], and Janet Baker immediately.” It was not copied to Roth or Bamberg. The memo was entitled “Lernout & Haus-pie — Due Diligence & Accounting Issues,” and read in part:

[T]here are several areas where we feel greater insight and clarity needs to be gained with respect to both due diligence and accounting.... [W]e would like tó reiterate our point of view that additional due diligence, led by accounting professionals on both sides, is important to gain greater comfort with respect to some of the issues indicated below.

Our experience shows that companies like ... L & H[ ], which grow via acquisition, necessitate an extra level of care at this stage of the process. This only becomes more important as Dragon shareholders and employees are ready to receive, and in some instances, hold L & H common shares and/or options, as part of the merger consideration. Our recommendation is that Dragon’s certified public accountants perform comprehensive due diligence onL & H, side by side with management, Hale & Dorr and Goldman Sachs.

The memo then listed several specific areas of concern.

Bamberg and Roth did not receive a copy of this memo, and they maintain that neither the Bakers nor any other members of Dragon’s senior management ever informed them of its contents.

Chamberlain, for her part, was dissatisfied -with Goldman’s work. She testified that she found it “ironic that [the] memo was written after I wrote a memo to Goldman Sachs telling them what my expectations were from a banker and pretty much they cut and pasted it back....” It ap *46 pears that this was an internal gripe; Chamberlain did not testify that she told anyone at Goldman of her frustration at receiving the February 29 memo. Indeed, we have not been pointed to any evidence that anyone at Dragon ever told Goldman that its performance was unsatisfactory, or that it was expected to play a larger role with regard to the due diligence process.

There is conflicting evidence in the record as to whether Dragon actually followed the advice contained in the February 29 memorandum from Goldman. Janet Baker testified that Dragon took Goldman’s concerns seriously and that the Dragon team, led by Chamberlain, went “through each and every one of these points and discuss[ed] them at length.” She explained further that Goldman “had made some recommendations about things that [Dragon’s accountant] Arthur Andersen should look at, and we had Arthur Andersen look at those things.” But Catherine Moy of Arthur Andersen testified that Arthur Andersen was not aware of the February 29 memo and was never asked to implement Goldman’s recommendations. Goldman’s Wayner testified that Chamberlain was “resistant to move forward on these items” because “she and Dragon did not want to do this level of additional detail.”

The district court found that the Goldman team remained unsatisfied with L & H’s due diligence responses, but did not repeat their concerns to anyone at Dragon after sending the February 29 memo.

Also on February 29, all plaintiffs participated in a conference call with Goldman’s Wayner and Charles Elliott, a European software research analyst whom Goldman had recruited to help the Bakers assess the value of L & H’s stock. Goldman, as said, did not have an analyst covering L & H at the time, and Elliott was unaware of L & H’s February 9 earnings report. Apparently, the L & H earnings report was not mentioned on the call. Elliott commented generally on the state of the European software market and speculated (correctly, as it turned out) that the stock market would react positively to Dragon’s merger with L & H. Elliott later testified that, had he known of L & H’s skyrocketing Asian revenues, he would have been skeptical of them because he “kn[ew] something about Asian languages, and ... the phonetic structure makes it incredibly difficult to take a European dictation software system and apply it ... to Asian languages.”

On March 7, 2000, L & H announced that it had agreed to acquire a company called Dictaphone Corporation and to assume $425 million of Dictaphone’s debt. The Goldman team did not update its evaluation of the L & H-Dragon merger to include analysis of the effects of that transaction. Neither the district court’s findings nor the submissions of the parties indicate whether Goldman communicated this information to anyone at Dragon or whether Dragon knew of this acquisition from other sources.

Discussions between Dragon and L & H on a proposed merger had continued. On March 8, 2000, CEO Waite and the Bakers met with a team from L & H to discuss the terms of the proposed merger. Roth, Bamberg, and the members of the Goldman team were not at this meeting. L & H offered to buy Dragon for $500 million, half cash and half stock, but the Bakers declined the offer and made a counteroffer.

Janet Baker, on behalf of Dragon, then signed a handwritten agreement drafted by CEO Waite under which L & H would buy Dragon for $580 million, all stock. The Bakers and Waite did not consult Goldman, Chamberlain, or the other Dragon board members before Janet Baker executed the handwritten agreement.

*47 After the execution of the agreement, on March 22, 2000, CFO Chamberlain held a conference call with accountants from Dragon and L & H to discuss the due diligence issues identified by Goldman in the February 29 memo. The parties dispute whether any Goldman employees were on this call, but the district court found that Goldman’s Berzofsky participated in this call. 2 Chamberlain testified that there were still “open issues” with respect to due diligence after this call.

Importantly, the next day, March 23, the Goldman team and the Dragon board of directors had a conference call to discuss the status of due diligence. There is conflicting evidence regarding what precisely was said at the meeting. The district court found that the participants all agreed due diligence was completed. That finding is supported by the record. 3 The court also found that “the Goldman team did not state they were still dissatisfied with due diligence, even though they were.” That is also supported by the record.

On March 27, 2000, Dragon board members and executives held a meeting to consider the transaction as to which Janet Baker had executed the March 8 handwritten agreement. All of the plaintiffs attended the meeting, as did Waite, Chamberlain, Shagoury, Fine, and Smith. Goldman’s Wayner called into the meeting. The evidence shows that Goldman offered tempered but positive comments about the proposed merger. For example, Fine testified that Goldman’s assessment was “positive in the overall,” but he also stated that Goldman “did not give an opinion pro or con on the transaction” and that Waite “was displeased that [Goldman] had not ... been more positive or given more support or [a] more definitive opinion.” Still, Wayner testified that he did not disclose any of the Goldman team’s concerns about the status of due diligence on L & H because “the client did not ask.” Wayner also noted that he had already raised those issues with Chamberlain and Janet Baker. Similarly, Fine testified that the Goldman team did not bring up its due diligence concerns “because th[ey] had all been raised with substantially the same group that was in there.” Significantly, plaintiffs testified, and the district court found, that plaintiffs would not have gone forward with the deal had Wayner voiced his concerns about the due diligence issues at the March 27 meeting. Ultimately, the Dragon board voted to approve the merger, and the transaction closed on June 7, 2000. Dragon merged into L & H and ceased to be an independent entity.

From there, the situation deteriorated quickly. In August 2000, the Wall Street Journal reported that L & H had vastly overstated its Asian revenues, and indeed had identified as customers some South Korean companies who had never done business with L & H. The SEC began investigating L & H’s financial statements, and eventually L & H admitted that it had improperly recorded $373 million in revenue. L & H restated its financials for two and a half years and filed for bankruptcy in November 2000. Following the bankruptcy, plaintiffs’ shares of L & H stock became worthless. The Dragon name and technology were sold from the estate.

*48 B. Other Proceedings

Before initiating this lawsuit, plaintiffs sued numerous other parties to recover the losses they suffered as a result of L & H’s fraud. The defendants were L & H and several of its officers and directors, entities related to L & H, entities and individuals affiliated with L & H’s auditor, L & H’s investment bank, and certain banks that provided financing to L & H. Plaintiffs collectively received over $75 million in settlements from those parties.

James and Janet Baker (“the Baker plaintiffs”) and Roth and Bamberg (“the Roth plaintiffs”) then sued Goldman in separate actions in state and federal court, respectively. Goldman removed the Bakers’ suit to federal court, and the two actions were then consolidated for purposes of discovery and trial. They have been consolidated for purposes of appeal.

Each set of plaintiffs brought claims for negligent and intentional misrepresentation, negligence, gross negligence, breach of fiduciary duty, and violations of Mass. Gen. Laws ch. 93A. 4

After a 20-day trial, the jury found in favor of Goldman on all of plaintiffs’ common law claims. The verdict form instructed the jury to reach Goldman’s third-party claims against the Bakers only if they determined that Goldman was liable to the Roth plaintiffs. Nonetheless, the jury also found that Janet Baker made negligent misrepresentations and had committed a breach of her fiduciary duty to the Roth plaintiffs, and that James Baker had committed a breach of his fiduciary duty to Bamberg.

The district court found for Goldman on plaintiffs’ ch. 93A claims. 5 In a thorough opinion, the, court ruled that, although Goldman may have committed negligence by failing to (1) disclose to plaintiffs that no one at Goldman was covering L & H at the time of the transaction; (2) repeat the Goldman team’s due diligence concerns after expressing those concerns once in the February 29 memo; and (3) adequately analyze L & H’s Asian revenues and its revenue projections, Goldman’s conduct “was not so egregious as to warrant ch. 93A relief.” In reaching its conclusion, the court gave weight to the jury’s verdict exonerating Goldman of any liability, as it was entitled to do.

The court also rejected the Roth plaintiffs’ theory, presented for the first time in a post-trial brief, that Goldman had violated ch. 93A because it violated 940 Mass. Code Regs. 3.16(2). That section provides that “an act or practice is a violation of [ch. 93A] if ... [ajny person or other legal entity ... fails to disclose to a buyer or prospective buyer any fact, the disclosure of which may have influenced the buyer or prospective buyer not to enter into the transaction.” The court ruled that the Roth plaintiffs had waived any claim under § 3.16(2) by failing to present it before trial, and held that the theory was merit-less in any event because the regulation does not apply to business-to-business transactions. The court later denied both sets of plaintiffs’ motions for reconsideration of the ch. 93A ruling, as well as their motions for a new trial. This appeal followed.

*49 II.

Both the Baker plaintiffs and the Roth plaintiffs argue that the district court erred as a matter of law in holding that Goldman’s conduct was not “unfair or deceptive” within the meaning of ch. 93A. Plaintiffs’ briefs contend that ch. 93A does not require a showing of “egregious” conduct and that, even if “egregiousness” is the correct standard, Goldman’s conduct rose to that level. After de novo review, we hold that the district court correctly articulated the legal standard applicable to plaintiffs’ ch. 93A claims, and correctly applied that standard to its factual findings. We affirm the dismissal of those claims.

A. Standard, of Review

“Following a bench trial on a chapter 93A claim, we review the district court’s legal conclusions de novo and its underlying factual findings for clear error.” Fed. Ins. Co. v. HPSC, Inc., 480 F.3d 26, 34 (1st Cir.2007). “[W]hether a particular set of acts, in their factual setting, is unfair or deceptive is a question of fact,” Arthur D. Little, Inc. v. Dooyang Corp., 147 F.3d 47, 54 (1st Cir.1998) (quoting Ahem v. Scholz, 85 F.3d 774, 797 (1st Cir.1996)), “but whether that conduct rises to the level of a chapter 93A violation is a question of law.” Fed. Ins. Co., 480 F.3d at 34; see also Casavant v. Norwegian Cruise Line Ltd., 460 Mass. 500, 952 N.E.2d 908, 911-12 (2011) (whether an act is unfair or deceptive is a factual question, but “[a] ruling that conduct violates [ch. 93A] is a legal, not a factual, determination” (quoting R.W. Granger & Sons v. J & 5 Insulation, Inc., 435 Mass. 66, 754 N.E.2d 668, 675 (2001))). 6 Under the clear error standard of review, we accept the district court’s findings of fact unless, after careful consideration of the entire record, “we are ‘left with the definite and firm conviction that a mistake has been committed.’ ” Vinick v. United States, 205 F.3d 1, 6 (1st Cir.2000) (quoting Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985)). If the district court’s factual conclusions are based on an erroneous view of the controlling law, however, “the ease for deference vanishes,” and we review those conclusions de novo. Id. at 6-7.

In ruling on a ch. 93A claim, a trial court is not bound by a jury’s verdict on parallel common law claims. E.g., Klairmont v. Gainsboro Rest, Inc., 465 Mass. 165, 987 N.E.2d 1247, 1263-64 (2013); Specialized Tech. Res., Inc. v. JPS Elastomer ics Corp., 80 Mass.App.Ct. 841, 957 N.E.2d 1116, 1119-20 (2011). But the court may, if it chooses, consider the verdict in reaching its conclusions, or adopt the verdict entirely. See Serv. Publ'ns, Inc. v. Gover- *50 man, 396 Mass. 567, 487 N.E.2d 520, 527 (1986).

B. Ch. 93A Legal Standard

Chapter 93A proscribes “unfair or deceptive acts or practices in the conduct of any trade or commerce.” Mass. Gen. Laws ch. 93A, § 2 (emphasis added). The statute provides a cause of action to “[a]ny person who engages in the conduct of any trade or commerce and who suffers any loss of money or property ... as a result of the use or employment by another person ... [of] an unfair or deceptive act or practice.” Id. § 11. If successful, a plaintiff is entitled to actual damages, or double or treble damages if the defendant’s violation of § 2 was willful or knowing. Id. Here, the district court found that Goldman’s “conduct was not unfair or deceptive under ch. 93A.” 7

Here, plaintiffs essentially do not dispute the district court’s factual determinations-indeed, both sets of plaintiffs base the statement of facts in their briefs almost entirely on the district court’s findings of fact. Instead, plaintiffs argue that there is a disconnect between the district court’s factual findings and its ultimate conclusion that Goldman was not liable under ch. 93A. The latter conclusion, plaintiffs contend, was both wrong as a matter of law and clearly erroneous as a matter of fact.

Specifically, plaintiffs’ briefs argue that the district court applied the wrong legal standard to its findings of fact when it held that, in order for a defendant’s conduct to violate ch. 93A, it “must be not only wrong, but also egregiously wrong.” 8 Plaintiffs offer an extensive catalogue of Massachusetts cases in which the court found liability under ch. 93A without mentioning an “egregiousness” standard. Goldman responds that cases from both the Massachusetts state courts and this Circuit have been consistent in requiring a heightened showing of “egregiousness” or “rascality” in adjudicating ch. 93A claims.

Chapter 93A “was ‘designed to encourage more equitable behavior in the marketplace.’ ” Commercial Union Ins. Co. v. Seven Provinces Ins. Co., 217 F.3d 33, 40 (1st Cir.2000) (quoting Arthur D. Little, 147 F.3d at 55). However, “it ‘does not contemplate an overly precise standard of ethical or moral behavior. It is the standard of the commercial marketplace.’ ” Id. (quoting Ahern, 85 F.3d at 798).

The language that courts have used to describe the ch. 93A standard has varied considerably over the years. See id. (collecting cases). Early Massachusetts decisions suggested that, in order to violate ch. *51 93A § 11, conduct must “attain a level of rascality that would raise an eyebrow of someone inured to the rough and tumble of the world of commerce,” Levings v. Forbes & Wallace, Inc., 8 Mass.App.Ct. 498, 396 N.E.2d 149, 153 (1979); see also Spence v. Bos. Edison Co., 390 Mass. 604, 459 N.E.2d 80, 88 (1983), or have a “rancid flavor of unfairness,” Atkinson v. Rosenthal, 33 Mass.App.Ct. 219, 598 N.E.2d 666, 670 (1992). The First Circuit followed the Massachusetts courts’ lead in articulating the ch. 93A standard, as it is required to under Eñe. See, e.g., Quaker State Oil Ref. Corp. v. Garrity Oil Co., 884 F.2d 1510, 1513 (1st Cir.1989).

In 1995, th'e Supreme Judicial Court of Massachusetts stated that it found “uninstructive phrases such as ‘level of rascality’ ” and instead would “focus on the nature of challenged conduct and on the purpose and effect of that conduct as the crucial factors in making a [ch.] 93A fairness determination.” Mass. Emp’rs Ins. Exch. v. Propac-Mass, Inc., 420 Mass. 39, 648 N.E.2d 435, 438 (1995) (citation omitted). But even after Propac-Mass, Massachusetts courts, when addressing claims that a defendant’s negligent act constituted a violation of ch. 93A, continued using labels akin to the “level of rascality” phrase to describe the level of negligence necessary for a finding of liability. See, e.g., Ross v. Cont’l Res., Inc., 73 Mass. App.Ct. 497, 899 N.E.2d 847, 861 (2009). Moreover, the SJC has repeatedly held that “mere negligence,” standing alone, is not sufficient for a violation of ch. 93A— something more is required. E.g., Klair-mont, 987 N.E.2d at 1257; Darviris v. Petros, 442 Mass. 274, 812 N.E.2d 1188, 1192 (2004); Swanson v. Bankers Life Co., 389 Mass. 345, 450 N.E.2d 577, 580 (1983). In Marram v. Kobrick Offshore Fund, Ltd., 442 Mass. 43, 809 N.E.2d 1017 (2004), which the district court relied on for the legal standard that it applied in this case, the SJC described that “something more” as “extreme or egregious” negligence. 9 See id. at 1032; accord Lily Transp. Co. v. Royal Institutional Servs., Inc., 64 Mass. App.Ct. 179, 832 N.E.2d 666, 687 n. 15 (2005); cf. Stonehill Coll. v. Mass. Comm’n Against Discrimination, 441 Mass. 549, 808 N.E.2d 205, 229-30 (2004) (in order for breach of contract to constitute a eh. 93A violation, there must be “some egregious circumstance surrounding that breach”). And our Circuit has again followed the Massachusetts courts’ lead in using the term “egregious” to state the standard of ch. 93A liability. See, e.g., In re Pharm. Indus., 582 F.3d at 185 (quoting Mass. Sch. of Law at Andover, Inc. v. Am. Bar Ass’n, 142 F.3d 26, 41 (1st Cir.1998)); In re TJX Cos. Retail Sec. Breach Litig., 564 F.3d 489, 497 (1st Cir.2009). 10

Thus, while we share the SJC’s sentiment in Propac-Mass that phrases such as “level of rascality” are uninstructive, we find no legal error in the district court’s analysis. It drew the “egregious” stan *52 dard directly from caselaw from the SJC and this Circuit. If the standard for ch. 93A liability requires clarification, the SJC can provide it in an appropriate case. See Gill v. Gulfstream Park Racing Ass’n, Inc., 399 F.3d 391, 402 (1st Cir.2005) (“A federal court sitting in diversity cannot be expected to create new doctrines expanding state law.”).

C. The District Court’s Factual Findings

The district court identified three instances of questionable conduct on the part of Goldman: (1) its failure to disclose that Elliott was not covering L & H at the time of plaintiffs’ February 29 conference call with him; (2) its failure to reiterate at a later date (and in particular, at the final meeting on March 27) the due diligence-related concerns expressed in the February 29 memo; and (3) its work on the valuation analysis of L & H. There was ample evidence in the record upon which the judge could have concluded that this conduct was neither unfair nor deceptive. We consider each of the three instances identified above in turn.

First, with regard to the Elliott call, the trial testimony and Janet Baker’s contemporaneous notes of the call suggest that Wayner introduced Elliott as a European equities analyst-not, as L & H now contend, as an expert on L & H specifically. The district court did not clearly err in rejecting plaintiffs’ argument that Goldman misrepresented Elliott’s knowledge of L & H or otherwise acted unfairly or deceptively during the call. Indeed, it is unclear what additional information plaintiffs think they would have gained had Elliott been covering L & H, or had Wayner mentioned the Asian earnings report to Elliott. Janet Baker was aware of L & H’s surging Asian revenues, and it is doubtful that the Bakers would have changed their views on the merger based on Elliott’s telling them something that they already knew. 11 In fact, Chamberlain testified that it would not have been important to her whether or not Elliott was covering L & H.

Second, turning to Goldman’s due diligence concerns, the district court found that “the Goldman team should have disclosed their continuing due diligence concerns at the March 27 meeting or on the March 23 conference call.” Nonetheless, the court held that their failure to do so was not “egregious” because Goldman had already informed Chamberlain, who was in charge of due diligence at Dragon, and the rest of the Dragon team about the due diligence concerns. That finding was not error. Members of the Goldman team testified at trial that they saw no need to raise those issues yet again at the March 27 meeting because Goldman had made clear to Dragon personnel that it was not satisfied with L & H’s due diligence responses. Plaintiffs have pointed to no evidence that anyone from Dragon ever complained to Goldman that its work on the transaction had been unsatisfactory or that it was expected to play a larger role in the due diligence process. Further, despite the concerns raised in the February 29 *53 memo, Janet Baker had already signed a handwritten deal with L & H on March 8. Goldman could have reasonably believed, based on this development, that the deal had essentially been agreed to and hence that raising further concerns after this point would simply serve to irritate its client.

The Roth plaintiffs, noting that they were never informed of Goldman’s due diligence concerns because they never received a copy of the February 29 memo, argue that Goldman’s conduct was unfair or deceptive as to them, even if was not unfair or deceptive as to the Bakers. According to the Roth plaintiffs, Goldman deceived them because it painted a rosy picture of the transaction’s prospects in the February 29 call with Elliott and in the March 27 meeting and never told Bamberg and Roth of the outstanding due diligence issues.

The district court rejected this contention “because the Goldman bankers reasonably believed that Janet Baker and Chamberlain, their main contacts at Dragon, would inform the rest of Dragon’s board and senior management about important events and documents leading up to the merger.” This conclusion is amply supported in the record. 12 Roth testified that Janet Baker and Chamberlain were his “interface” with regard to the status of due diligence and that he “relied on Janet Baker to obtain information on [L & H] including its financial condition in order to determine whether to vote to approve the merger with L & H.” Bamberg similarly testified that he relied on Janet Baker to “keep[ ] both Bob and myself up to date on major issues while not troubling us with details.”

The Roth plaintiffs’ contention that “[t]hree Goldman witnesses admitted knowing that the Roth [p]laintiffs ... were unaware of its due diligence concerns” is not supported by the record. The Roth plaintiffs cite the following pieces of testimony for their argument:

• From Wayner’s deposition: “Q: You did not tell the people assembled [at the March 27 meeting] that Goldman Sachs had not obtained information it had requested? A: Depends on how you phrase that question. There’s a subset of persons that knew that was our point of view. So are you asking me whether no one in that room knew that or did I mention it at this meeting?”

• From Fine’s deposition: “We gave our advice ... to all the principals who were in the room at the board meeting. We had talked to — the Baker’s [sic ] had seen our books where some of the issues had been raised. The memo had gone to Mr. Waite. The majority or [sic ] the people who were capable of approving and not approv-. ing had seen our reservations.”

• From Smith’s deposition: “Q: If Rich Wayner was not satisfied with the due diligence that was obtained regarding Asian revenues, customer agreements, licensing agreements, related party transactions, would you have expected Wayner to have expressed those views at the March 27, 2000 board meeting? A: No.... I would not have — that to have been done in that form, no. It would have been an embarrassment to Ms. Baker and the deal leaders.”

The Roth plaintiffs read this testimony to suggest that Goldman knew that Roth and Bamberg were, as counsel put it at *54 oral argument, “in the blind,” and intentionally withheld information from them in order to make sure the transaction would go forward. 13 To the contrary, it is reasonable to read Wayner’s and Fine’s statements to mean that they saw no need to raise their due diligence concerns yet again because they had already communicated those concerns to most, if not all, of the individuals who had a stake in the transaction. And it is reasonable to read Smith’s testimony as simply stating that he did not want to impede the progress of the transaction because Janet Baker had already made it known that she wanted to consummate the deal as quickly as possible — indeed, she had already agreed to the transaction in writing three weeks before. The district court did not err in its conclusion that Goldman did not act unfairly or deceptively in failing to raise its due diligence concerns specifically with Roth and Bamberg.

The Roth plaintiffs have suggested that, because Judge Saris found that the L & H transaction would likely not have gone forward if Goldman had raised its ongoing concerns about due diligence at the March 27 meeting, she was required to find a violation of ch. 93A as a matter of law. We disagree. It is true that cases from Massachusetts and this Circuit have defined a “deceptive” act as one that “could reasonably be found to have caused a person to act differently from the way he or she otherwise would have acted.” In-case Inc. v. Timex Corp., 488 F.3d 46, 57 (1st Cir.2007) (quoting Aspinall v. Philip Morris Cos., 442 Mass. 381, 813 N.E.2d 476, 486 (2004)); see also Grossman v. Waltham Chem. Co., 14 Mass.App.Ct. 932, 436 N.E.2d 1243, 1245 (1982). But it cannot be that any conduct gives rise to liability under ch. 93A by virtue of the mere fact that the conduct affects a person’s actions in a way that eventually causes that person harm. Were that so, one could be liable under ch. 93A simply for giving bad advice, no matter how well-intentioned and well-founded the advice. That is not the law. Indeed, as said, the SJC has instructed that even negligent misrepresentations (which, by definition, “could reasonably be found to have caused a person to act differently from the way he or she otherwise would have acted”) give rise to ch. 93A liability only if they are “extreme” or “egregious.” Marram, 809 N.E.2d at 1032.

Finally, with regard to Goldman’s “professionally negligent” financial analysis of the Dragon-L & H transaction, the district court did not err in finding that this conduct was neither unfair nor deceptive. In reaching its conclusion, the district court properly took into account the jury’s finding, based on sufficient evidence, that Goldman was not negligent in rendering professional services. Goldman raised a bevy of due diligence issues in the February 29 memo. Plaintiffs have presented no evidence that Dragon ever responded by asking Goldman to do further due diligence work. Moreover, there is testimony that Goldman was engaged primarily to raise questions and facilitate the merger, not to lead the due diligence on Dragon’s eventual merger partner. Chamberlain— *55 not Goldman-was the “quarterback” of due diligence. The jury called it one way whether there was any negligence at all; the district court called it another, but was certainly entitled to consider the jury finding in weighing whether Goldman had been unfair or deceptive.

In short, the district court properly determined that Goldman’s conduct, even if sloppy and unforthcoming, was not unfair or deceptive. The court’s factual findings are supported by the record, and it correctly applied the ch. 93A legal standard to those findings. We find no error in the district court’s analysis of plaintiffs’ ch. 93A claims.

III.

The Roth plaintiffs contend that the district court erred in rejecting their belated theory of Goldman’s liability under 940 Mass.Code Regs. § 3.16(2). Not so. The Roth plaintiffs waived any claim under § 3.16(2) by failing to raise it before trial and waiting until the jury had ruled against them on their common law theories.

Under Mass. Gen. Laws ch. 93A § 2(c), the attorney general is empowered to make rules and regulations defining the “[u]nfair methods of competition and unfair or deceptive acts or practices” that violate § 2(a). One such regulation provides that an “act or practice is a violation of [ch. 93A] if ... [a]ny person or other legal entity subject to this act fails to disclose to a buyer or prospective buyer any fact, the disclosure of which may have influenced the buyer or prospective buyer not to enter into the transaction.” 940 Mass.Code Regs. § 3.16. In their post-trial brief to the district court, the Roth plaintiffs argued that Goldman’s conduct in this case violated § 3.16 because Goldman failed to disclose facts that would have influenced whether the Roth plaintiffs agreed to vote for the Dragon-L & H merger (and thus “buy” L & H stock). Plaintiffs did not raise this argument either before or at trial. We agree with the district court that the Roth plaintiffs waived any claim under § 3.16. See DCPB, Inc. v. City of Lebanon, 957 F.2d 913, 917 (1st Cir.1992) (plaintiff cannot, after trial, “superimpose a new (untried) theory on evidence introduced for other purposes”), superseded on other grounds, as recognized in Lamboy-Ortiz v. Ortiz-Vélez, 630 F.3d 228, 243 n. 25 (1st Cir.2010).

The Roth plaintiffs contend that their general argument that Goldman’s failure to disclose relevant facts about the L & H transaction violated ch. 93A was sufficient to put Goldman on notice of the § 3.16 claim, but we are not persuaded. Goldman knew only that it was defending against claims of “unfair or deceptive” acts under ch. 93A, § 2. It had no warning of any claim that it had per se violated § 2 via § 3.16 until after trial. See Rodriguez v. Doral Mortg. Corp., 57 F.3d 1168, 1172 (1st Cir.1995) (plaintiff may not “leave defendants to forage in forests of facts, searching at their peril for every legal theory that a court may some day find lurking in the penumbra of the record”). To allow the Roth plaintiffs to raise this claim so late would have undoubtedly prejudiced Goldman. See Grand Light & Supply Co., Inc. v. Honeywell, Inc., 771 F.2d 672, 680 (2d Cir.1985) (“Where a party seeks to apply evidence presented on a separate issue already in the case to a new claim added after the conclusion of the trial, the opponent may be unfairly prejudiced.”); see also Lebanon, 957 F.2d at 917 (citing Honeywell for this proposition).

Even if the Roth plaintiffs had properly raised an argument under § 3.16, we see no basis in present Massachusetts law to credit the claim. The SJC’s decision in Knapp Shoes Inc. v. Sylvania Shoe *56 Manufacturing Corp., 418 Mass. 737, 640 N.E.2d 1101 (1994), strongly suggests that § 3.16 applies only to transactions involving consumers and not to transactions involving sophisticated business entities.

In Knapp, the SJC held that 940 Mass. Code Regs. § 3.08(2), providing in relevant part that “[i]t shall be an unfair and deceptive act or practice to fail to perform or fulfill any promises or obligations arising under a warranty,” applies only to consumer claims under ch. 93A. Knapp, 640 N.E.2d at 1104. The court reasoned that the other subsections in § 3.08 “use the term ‘consumer’ to denote the persons protected by their provisions, and concern matters generally involved in consumer transactions.” Id. at 1105. Even though subsection (2), by its terms, did not limit its application to consumers, the court found that the context of the statute indicated that it did not apply to business-to-business transactions. Id. (“Where the bulk of the regulation applies only to consumers and their interests, and subsection (2) contains no language suggesting that it was meant to apply to a broader class of persons or interests, we conclude that the portion of subsection (2) at issue was not intended to encompass a contract dispute between businessmen based on a breach of ... warranty....”).

The same reasoning is applicable here. Two of the four subsections of § 3.16 mention “consumers” and concern consumer protection issues. Subsection (3) makes an act or practice a violation of ch. 93A if “[i]t fails to comply with existing statutes, rules, regulations or laws, meant for the protection of the public’s health, safety, or welfare promulgated by the Commonwealth or any political subdivision thereof intended to provide the consumers of this Commonwealth protection.” Subsection (4), in similar fashion, makes an act or practice a violation of ch. 93A if “[i]t violates the Federal Trade Commission Act, the Federal Consumer Credit Protection Act or other Federal consumer protection statutes within the purview of [ch.] 93A, § 2.” Thus, just as in Knapp, “[i]t is reasonably clear that, in drafting the regulation, the Attorney General had in mind protection for consumers against unfair or deceptive acts or practices.” 640 N.E.2d at 1105; see also In re First New Eng. Dental Ctrs., 291 B.R. 229, 241 (D.Mass. 2003) (applying Knapp’s reasoning to conclude that § 3.16 does not apply to “business to business transactions”); Callahan, Note, Massachusetts General Laws Chapter 93A, Section 11: The Evolution of the “Raised Eyebrow” Standard, 36 Suffolk U.L.Rev. 139, 157-58 (2002) (after Knapp, “courts may apply similar reasoning to invalidate applicability of other regulations to business-to-business transactions”); cf. Indus. Gen. Corp. v. Sequoia Pac. Sys. Corp., 44 F.3d 40, 44 (1st Cir.1995) (“A commentator has noted that section 11 ‘probably does not contain a general duty of disclosure’.... ” (quoting Gilleran, The Law of Chapter 93A § 4:10 (1989 & Supp. 1994))). But see Lechoslaw v. Bank of Am., 618 F.3d 49, 58 (1st Cir.2010) (suggesting that § 3.16 applies to businesses); Lily Transp. Corp. v. Royal Inst’l Servs., Inc., 64 Mass.App.Ct. 179, 832 N.E.2d 666, 673-74 (2005) (same).

The Roth plaintiffs argue that § 3.16 is “general” and so should not be read to exclude businesses from its scope. They note that the preamble to the section provides that it does not “limit[ ] the scope of any other rule, regulation or statute.” But § 3.16 is no more, general than § 3.08, which likewise covered multiple subjects and provided that it “in no way limits, modifies, or supersedes any other statutory or regulatory provisions dealing with warranties.” See Knapp, 640 N.E.2d at 1104. In short, we simply see no meaningful distinction between § 3.16 and § 3.08 *57 that 'would counsel against applying Knapp’s reasoning to the former. This is ultimately an issue for the SJC to resolve. 14

IV.

Finally, plaintiffs argue that they are entitled to a new trial because the district court erred in (1) admitting the drafting history 'and Annex A of the Engagement Letter between Goldman and Dragon and (2) instructing the jury regarding the relevance of the Engagement Letter and its drafting history. Both of these contentions are without merit.

A. Admission of the Drafting History and Annex A

We review the district court’s evi-dentiary rulings for abuse of discretion. Enos v. Union Stone, Inc., 732 F.3d 45, 49 (1st Cir.2013). If we find error, we reverse unless “it is highly probable that the error did not affect the outcome of the case.” McDonough v. City of Quincy, 452 F.3d 8, 19-20 (1st Cir.2006).

The district court properly applied state law in admitting the Engagement Letter and its drafting history. In Nycal Corp. v. KPMG Peat Marwick LLP, 426 Mass. 491, 688 N.E.2d 1368 (1998), the SJC set forth the requirements for a plaintiff asserting a claim of negligent misrepresentation against a defendant who supplies information for the guidance of others in business transactions. Under Nycal, if the plaintiff and defendant are not in contractual privity (as is the case here, because Goldman was engaged by Dragon, not by plaintiffs), in order to succeed on a negligent misrepresentation claim, the plaintiff must show that the defendant had “actual knowledge ... of the limited-though unnamed-group of potential [parties] that will rely on the [defendant’s advice], as well as actual knowledge of the particular financial transaction that such information is designed to influence.” 688 N.E.2d at 1371-72 (quoting First Nat’l Bank of Commerce v. Moneo Agency Inc., 911 F.2d 1053, 1062 (5th Cir.1990)); see also Restatement (Second) of Torts § 552 (1977).

Provisions of the draft Engagement Letter indicating that Goldman was to be employed by individual stockholders were explicitly removed from the final agreement. That is clearly relevant to Goldman’s knowledge as to whether individual shareholders would rely on Goldman’s financial advice, when the plaintiffs expressly chose not to sign the agreement in order to avoid the indemnification obligations which the signatory, Dragon, undertook. Accordingly, the Engagement Letter and its drafting history were relevant to both sets of plaintiffs’ negligent misrepresentation claims, see Fed.R.Evid. 401, and the plaintiffs have not shown that their relevance was substantially outweighed by a risk of unfair prejudice, see Fed.R.Evid. 403. The evidence was admissible. 15

Annex A was relevant to the case for the same reason as was the drafting history. The final version of Annex A, like the rest of the agreement, excluded any' relevant reference to “Stockholders,” which further strengthens the inference that Goldman did not intend for individual *58 stockholders to rely on its financial advice. Thus, even though Annex A concerned Goldman’s liability for derivative, rather than direct, claims, the district court did not abuse its discretion in admitting, it. This is all the more so given the district court’s explicit instruction to the jury that the exculpatory clause in Annex A “is inapplicable ... because the plaintiffs’ claims are ... direct claims for themselves as individual shareholders.” We assume that the jury followed this instruction. United States v. George, 761 F.3d 42, 57 (1st Cir.2014). 16

Even if admission of Annex A was arguably an abuse of discretion, any error was harmless, given the court’s cautionary instruction and the substantial amount of other evidence tending to suggest that Goldman did not intend individual shareholders to rely on its advice. See SEC v. Happ, 392 F.3d 12, 28-29 (1st Cir.2004) (admission of cumulative evidence was harmless error).

B. Jury Instructions

We review de novo a claim that a jury instruction was based upon an erroneous statement of the law. Hatch v. Trail King Indus., Inc., 656 F.3d 59, 64 (1st Cir.2011). “We review for abuse of discretion “whether the instructions adequately explained the law or whether they tended to confuse or mislead the jury on the controlling issues.’ ” Id. (quoting United States v. Silva, 554 F.3d 13, 21 (1st Cir.2009)); see also Johnson v. Spencer Press of Me., Inc., 364 F.3d 368, 378 (1st Cir.2004) (“So long as th[e] language' properly explains the controlling legal standards and is not unduly confusing or misleading, it will not be second-guessed on appeal.”). ‘We look at the instructions as a whole, not in isolated fragments.” Hatch, 656 F.3d at 64.

The plaintiffs objected to the last paragraph of the jury instructions concerning the Engagement Letter on the ground that it was “somewhat confusing.” The instructions on the Engagement Letter read as follows:

[L]et me just mention briefly this engagement letter you’ve heard so much about, the engagement letter.

The engagement letter is a contract between Goldman Sachs and the company, Dragon Systems. The shareholders were not parties to the contract with one exception-the so-called exculpation clause. This clause only applies to something in the law called derivative actions by shareholders on behalf of the corporation for damages suffered by the corporation. That provision is inapplicable here because the corporation no longer existed after the merger and because the plaintiffs’ claims are what are known as direct claims for themselves as individual shareholders. That’s why I keep saying, it’s always the plaintiffs as individual shareholders.

Because the shareholders were not parties to the engagement letter, they cannot sue for breach of contract. That’s why you don’t see “breach of *59 contract” in here anywhere. However, Goldman Sachs may still be held responsible for certain common law causes of action. You may have heard the term “torts.” That’s what we’ve just been talking about for the last hour. These torts include negligence, gross negligence, negligent misrepresentation[,] fraud or intentional misrepresentation, and breach of fiduciary duty. Those claims require that the plaintiffs establish that Goldman Sachs owed them a duty as I have just instructed you.

However, in determining whether Goldman Sachs is liable, you can consider all of the evidence, including the engagement agreement, the course of dealing between the parties before and after the agreement, and the history of negotiating the agreement.

There was no error in this jury instruction, nor was it confusing. 17 The jury was entitled to consider the agreement and its drafting history in considering both sets of plaintiffs’ tort claims. As said, the evidence was relevant to plaintiffs’ negligence claims because it suggested that Goldman did not foresee that individual shareholders would rely on its advice. It was relevant to the intentional misrepresentation claims because, to prove such a claim, a plaintiff must show that the defendant intended to induce the plaintiff to act upon a false statement. Masingill v. EMC Corp., 449 Mass. 532, 870 N.E.2d 81, 88 (2007). And it was relevant to the breach of fiduciary duty claims because a fiduciary relationship exists only if the plaintiff justifiably reposed trust in the defendant and the defendant knew of and accepted that trust. Broom-field v. Kosow, 349 Mass. 749, 212 N.E.2d 556, 560 (1965); see also Maffei v. Roman Catholic Archbishop of Bos., 449 Mass. 235, 867 N.E.2d 300, 313 (2007); Patsos v. First Albany Corp., 433 Mass. 323, 741 N.E.2d 841, 851 (2001). 18 The last paragraph of the quoted instruction was both clear and substantively correct.

V.

We. affirm the decision of the district court. Costs are awarded to Goldman.