In re BankAtlantic Bancorp, Inc.: Emerging Securities Damages Issues
The recent decision by Judge Ungaro in In re BankAtlantic Bancorp, Inc. Securities Litigation, No. 07-61542 (S.D. Fla. Apr. 25, 2011), offers a useful look at some emerging issues that arise when analyzing securities damages under the Securities Exchange Act of 1934. At its surface, the ruling is fairly simple: plaintiffs failed to meet their burden to establish loss causation or damages because plaintiffs' expert did not analyze how much of the stock drop at issue was attributable to fraudulent statements, as opposed to the other, non-actionable statements. Without such expert evidence, any attempt to attribute the stock price decline to the fraudulent statements would be speculation.
The district court's path to this result reveals at least two interesting, and unresolved, issues concerning the calculation of securities damages under Section 10(b) of the Securities Exchange Act of 1934. First, the court had to confront the question of what constitutes sufficient "materialization of the risk" absent a corrective disclosure. Second, the opinion contains an interesting discussion concerning the question of how much proof is required to establish loss causation under Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336 (2005). This article discusses these two issues in the context of the Bancorp decision.
Background
Plaintiffs filed a class action securities suit against BankAtlantic Bancorp., Inc. (Bancorp), the publicly traded parent company of BankAtlantic, a federally chartered bank providing consumer and commercial banking and lending in Florida, and some of its officers. The relevant consolidated amended complaint alleged that defendants made misrepresentations about the quality and value of assets in BankAtlantic's loan portfolio in violation of Sections 10(b), 20(a) and 20A of the Exchange Act. In a nutshell, plaintiffs alleged that defendants made misrepresentations and omissions concerning the deteriorating quality of BankAtlantic's loan portfolio, the bank's underwriting practices and the adequacy of loan reserves.
The case was tried before a jury in October and November 2010. Although the jury found for the defendants on almost all of the allegations, it concluded that a statement on an April 26, 2007 investor conference call by Alan Levan, the former Chairman and CEO of Bancorp and of BankAtlantic, violated Section 10(b) and proximately caused damages of $2.41 per share.
The district court's opinion explained that in the weeks leading up to the April 26 call, and dissemination of first quarter results, defendants had begun to distinguish two types of loans in its "land loan" portfolio, which was a significant portion of its total loan portfolio. The first category was "builder land bank" or "BLB" loans, which were made to developers who had option contracts with large regional and national homebuilders. The remaining, non-BLB loans were made to developers who developed land without entering into option contracts with national and regional homebuilders.
The court concluded that the evidence also demonstrated that, in the weeks leading up to the April 26 investor call, Bancorp had determined that there were problems throughout the entire portfolio. On the April 26 call, Levan discussed concerns about risks associated with the BLB portion of the land loan portfolio. However, in response to a question about "construction loans generally speaking," Levan stated: "But lots of our portfolio is a construction portfolio that we're not in any way concerned about." After summarizing the evidence presented to the jury, the court concluded that a jury could have found Levan's statement "to be an actionable concealment of the risk of substantial losses to the non-BLB land loans."
The court then turned to the question of loss causation and damages resulting from the misrepresentation. Plaintiffs argued that the loss materialized on October 25 and 26, 2007, when Bancorp disclosed for the first time loan losses across its entire land loan portfolio. At the same time, the bank also disclosed information about net interest margin compression, costs associated with opening new stores, and issues with other parts of its loan portfolio. The next day, Bancorp's stock price declined by $2.93. Relying on the testimony of its damages expert, Candace Preston, plaintiffs claimed the entire $2.93 as damages attributable to the fraud.
Loss Causation and the Materialization of the Risk
The district court explained that to establish loss causation under Dura, plaintiffs must establish that (i) the fraud was revealed to the market, and (ii) the revelation caused, at least in substantial part, the decline in the stock price. Opinion at 44.
As to the first prong, defendants argued that plaintiffs must demonstrate that the market acquired information revealing that the April 26 disclosure was false and that plaintiffs could not simply argue that the risk materialized. See In re Dell Inc. Sec. Litig., 591 F. Supp. 2d 877, 911 (W.D. Tex. 2008) ("A dopting the Plaintiff's theory of materialization of risk would eviscerate Dura's loss causation requirement, as any negative earnings news following a misrepresentation could, arguably, 'materialize' the risk of a prior misrepresentation."); In re eSpeed, Inc. Sec. Litig., 457 F. Supp. 2d 266, 296 (S.D.N.Y. 2006) ("Because there was no corrective disclosure regarding the spray device before the economic loss occurred, the alleged deception regarding the spray device could not possibly have caused the economic loss."). That is, according to defendants, while the October 2007 announcement may have revealed problems with non-BLB loans, nothing in that announcement established that the negative information was a result of the risk that should have been disclosed in April 2007.
The court rejected this argument, explaining that numerous courts have held that loss causation can be established "even where the defendant does not publicly correct his fraud, but instead the fraud is revealed through some other event." Opinion at 45 (citing cases); see also In re Motorola Sec. Litig., 505 F. Supp. 2d 501, 542-43 (N.D. Ill. 2007) (finding a plaintiff may still establish loss causation where a corrective disclosure does not, on its face, specifically identify or explicitly correct a previous representation); Lentell v. Merrill Lynch & Co., 396F.3d 16 1, 172-77 (2d Cir. 2005) (requiring that the disclosure reveal that the defendant's "misstatement or omission concealed something from the market that, when disclosed, negatively affected the value of the security"). The district court in Bancorp adopted this approach and reasoned that "the evidence supports a finding that the disclosures on October 25 and 26, 2007 revealed that the risk of substantial losses was not limited to the BLB loans but existed throughout the entire land loan portfolio." Opinion at 46.
Although this result seems to be consistent with the case law accepting the materialization of the risk theory, it raises the question of what level of proof is required to demonstrate a sufficient nexus between the fraudulent statement and the market learning that the information in the statement was not reliable? Where an express corrective disclosure is made, there is no doubt that an efficient market has now learned that the prior statement was not accurate. Absent such a disclosure, it can be harder to determine whether the market actually acquired that knowledge. In Bancorp, for example, while both the April 2007 and October 2007 statements concerned non-BLB loans, it is not apparent how the court could be confident that the negative information in October 2007 was a materialization of the prior risk, rather than as a result of other developments, particularly given the continuing problems in the housing sector.
Indeed, in a recent decision, Judge Easterbrook of the Seventh Circuit opined that the terminology "materialization of the risk" has no value. Schleicher v. Wendt, 6 18 F.3d 6 79 (7th Cir. 2010). He wrote: "The phrase appears in a few decisions . . . to describe particular claims, but it is not a legal doctrine or anything special as a matter of fact. . . . The phrase adds nothing to the analysis. . . . [T]he fraud lies in an intentionally false or misleading statement, and the loss is realized when the truth turns out to be worse than the statement implied." Id. at 683-84.
Loss Causation and the Inadequate Event Study
In prior years, plaintiffs would sometimes seek to offer damages experts who did not even do an event study, which is a detailed analysis of all events during and immediately following the class period, in order to "distinguish between the fraud-related and non-fraud related influences on the stock's price behavior." In re Oracle Sec. Litig., 829 F. Supp. 1176 , 1181 (N .D. Cal. 1993). Courts generally rejected such efforts. See, e.g., Id. (As a result of failure to perform an event study, "the results reached . . . cannot be evaluated by standard measures of statistical significance."); In re Northern Telecom Ltd. Sec. Litig., 116F. Supp. 2d 446, 460 (S.D.N.Y. 2000) ("Torkelsen's testimony is fatally deficient in that he did not perform an event study or similar analysis to remove the effects on stock price of market and industry information and he did not challenge the event study performed by defendants' expert"); In re Imperial Credit Indus., Inc. Sec. Litig., 252 F. Supp. 2d at 1016 (excluding expert report because "absent an event study or similar analysis, Plaintiffs cannot eliminate that portion of the price decline of ICII's and/or SPFC's stock which is unrelated to the alleged wrong").
Dura makes clear why an event study is needed, explaining that where a stock price drops after an adverse disclosure, "that lower price may reflect, not the earlier misrepresentation, but changed economic circumstances, changed investor expectations, new industry-specific or firm specific facts, conditions, or other events, which taken separately or together account for some or all of that lower price." 544 U.S. at 343. That is what happened here. Bancorp's October 2007 disclosure provided negative news about net interest margin compression, curtailed branch expansion, and changes in the performance of home equity loans. The announcement also revealed negative information about both BLB and non-BLB land portfolios, even though the jury found fraud only with respect to the non-BLB portfolio.
Defendants did not offer expert testimony to the jury on this issue. Plaintiffs presented testimony from Candace Preston, who purportedly performed a detailed event study, in which she reviewed over a hundred analyst reports to support her conclusion that the news unrelated to the loan portfolio did not affect the share price.
That was not enough, however, because Preston did not disaggregate the BLB and non-BLB loan portfolios. Preston's assumption that the jury would agree with plaintiffs that April 25, 2007 statements about both portfolios were fraudulent. But the jury did not do that. Instead, the jury found fraud only with respect to the disclosure as to non-BLB loans.
The court concluded that Preston's failure was a fatal blow to plaintiffs' effort to establish whether in the face of much negative information, the news about the non-BLB portfolio alone impacted the stock price. According to the district court, "[a]ny attempt to attribute some price decline to one particular piece without expert testimony would also be impermissible speculation. While it may be true that the negative land loan news was spread equally between the BLB and non-BLB portions, any inference that each had an equal effect on the stock price is only speculation." Opinion at 60. See also Fener v. Operating Engineers Const. Indus., 579 F.3d 401, 410 (5th Cir. 2009) ("Hakala's testimony was fatally flawed; he wedded himself to the idea that the press release was only one piece of news and conducted his event study based on that belief. We reject any event study that shows only how a 'stock reacted to the entire bundle of negative information,' rather than examining the 'evidence linking the culpable disclosure to the stock-price movement.'") (internal citations omitted). Thus, the court held that the jury had no basis to decide that the stock price decline, or that some portion of the decline, was attributable to the materialization of the risk with respect to non-BLB loans.
The court also rejected plaintiffs' argument that Preston should not have this burden because neither Bancorp nor the analysts precisely quantified this information either. It reasoned that plaintiffs are not relieved of their burden to disaggregate simply because the analysis will not be "mathematically precise." Id. at 57. The court also noted that plaintiffs' argument was undermined by the fact that Preston disaggregated other price changes, absent specific quantification, on issues where the jury ultimately found no liability.
The Failure to Prove Damages
Judge Ungaro also noted that even if the court somehow concluded that sufficient proof of loss causation had been proven, plaintiffs could not meet their burden of proof on damages based on Preston's failure to disaggregate the various statements. According to the court: "[T]o prove damages, a more rigorous showing is required, because by the express terms of the Exchange Act, a plaintiff's recovery is limited to 'actual damages on account of the act complained of.'" Opinion at 63 (quoting 15 U.S.C. §§78bb(a)). Absent a disaggregation analysis, the court reasoned, plaintiff "did not produce sufficient evidence to support an award of damages in any amount." Id. at 64.
What Is Plaintiffs' Burden Under Dura?
Although Dura is inevitably a starting point for any discussion of damages under the Exchange Act, it is important to remember that, on the facts of Dura, the court was considering the sufficiency of loss causation allegations at the pleading stage. Thus, while Justice Breyer indicated that "it should not prove burdensome for a plaintiff" to meet its burden in the complaint, 544 U.S. at 343, the decision left for another day the level of proof required at summary judgment or at trial. Id. at 346 ("We need not, and do not, consider other proximate cause or loss-related questions.").
The district court in Bancorp noted that the greater weight of authority suggests that "a securities-fraud plaintiff can satisfy his burden of proving loss causation only by producing the testimony of an expert who has completed a reliable multiple-regression analysis, event study, and financial analysis in order to quantify the extent to which the claimed losses are the result of the alleged fraud." Opinion at 61-62. The court speculates that it may be possible to establish loss causation using a less rigorous standard and indicates that it attempted to apply a less rigorous standard here, which plaintiff still could not meet.
This brief discussion by the district court highlights a critical unresolved question under Dura: what does plaintiff actually have to show at trial? Here, plaintiffs' expert made a readily apparent error in her analysis when she failed to disaggregate the disclosures based on the two types of loans, erroneously assuming that the jury would find fraud as to both types. But what if she had attempted to do this analysis and opined, for example, that one type of loan contributed 50% to the stock decline? How would a court discern whether the analysis was sufficient even to go to the jury? In Fener v. Operating Engineers Construction Industries, 579 F.3d 401, 410 (5th Cir. 2009), for example, the court noted that in the case of a securities fraud suit concerning circulation decline at a newspaper, "[c]onceivably, DMN's fraudulent practices could have resulted in 90% of the circulation decline, but if the stock price fell because the market was concerned only with the reason for the other 10%, loss causation could not be proven." But in that case, as well, the expert failed to disaggregate the fraudulent and non-fraudulent statements. Indeed, as plaintiffs begin to do more such analysis, issues of the quality of expert opinions under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), and its progeny, may come into play.
At one time, plaintiffs did not even bother with an event study. Now that the absence of one is not conceivably an option, plaintiffs are doing them more routinely. From an insurance and defense perspective, this will place greater importance on a deeper understanding of damages issues and the underlying factual assumptions that are driving the damages analysis to ensure that all avenues of attack on plaintiffs' damages theories are thoroughly evaluated.
