Joseph C. HUBBARD, individually and on behalf of all others similarly situated, Plaintiff, State–Boston Retirement System, Plaintiff–Appellant, v. BANKATLANTIC BANCORP, INC., James A. White, Valarie C. Toalson, Jarett S. Levan, Alan B. Levan, Defendants–Appellees.
This appeal concerns a private securities fraud class action brought under § 10(b) of the Securities Exchange Act of 19341 and SEC Rule 10b–52 against a bank holding company, BankAtlantic Bancorp, Inc., and its management (collectively, “Bancorp”)3 by State–Boston Retirement System, a shareholder and the lead plaintiff. State–Boston sought to prove at trial that the holding company had misrepresented the level of risk associated with commercial real estate loans held by its subsidiary, BankAtlantic. After the trial, the District Court submitted the case to the jury on a verdict form seeking general verdicts and answers to special interrogatories under Federal Rule of Civil Procedure 49(b). When the jury returned a verdict partially in favor of State–Boston, Bancorp moved for judgment as a matter of law under Federal Rule of Civil Procedure 50. Perceiving an inconsistency between two of the jury's interrogatory answers, the District Court discarded one of them and granted the motion on the basis of the remaining findings.
This was error. When a court considers a motion for judgment as a matter of law—even after the jury has rendered a verdict—only the sufficiency of the evidence matters. Chaney v. City of Orlando, 483 F.3d 1221, 1227 (11th Cir.2007). The jury's findings are irrelevant. See id. at 1227–28. Despite the District Court's error, we may affirm for any reason supported by the record. E.g., United States v. Harris, 608 F.3d 1222, 1227 (11th Cir.2010). In this case, we conclude that the evidence was insufficient to support a finding of loss causation, an element required to make out a securities fraud claim under Rule 10b–5. See, e.g., Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 342, 125 S.Ct. 1627, 1631, 161 L.Ed.2d 577 (2005) (listing among the elements of a § 10(b) securities fraud claim “loss causation, i.e., a causal connection between the material misrepresentation and the loss” (emphasis omitted) (internal quotation marks omitted)). We therefore affirm.
BankAtlantic Bancorp, Inc., is a publicly traded bank holding company incorporated and headquartered in Florida. Its subsidiary, BankAtlantic, is a federally chartered bank that offers consumer and commercial banking and lending services throughout Florida. This case concerns allegations that from October 19, 2006, until October 25, 2007 (the “class period”),4 Bancorp fraudulently misled the public about the deteriorating credit quality of BankAtlantic's commercial real estate portfolio.5 That portfolio comprised land acquisition and development loans; land acquisition, development, and construction loans; and builder land bank loans (“BLB loans”). Each of these categories comprised loans to investors to buy land for initial development followed by sale for further development. The relevant distinction for purposes of this case is between BLB loans, which were made to investors after they had sold options to purchase lots to homebuilders, and non-BLB loans, which involved no such pre-disbursement option contracts.
BankAtlantic internally monitored the risk associated with these land loans by assigning each loan a grade on a scale from one to thirteen—the lower the grade, the safer the loan. Grades one through nine were considered passing. But once a loan was assigned a grade of ten—and therefore classified as a “special mention” asset—or a grade of eleven—and therefore classified as a “substandard” asset—it was placed on a “Loan Watch List” to allow the bank's management to keep track of loans that might pose problems.6
The Loan Watch List, which was updated monthly, was a purely internal risk-monitoring tool; it was not released to the public. Bancorp's public disclosures did regularly reveal the amount of loans designated “nonaccrual,” as opposed to “accruing.” That designation indicated the bank's judgment that a loan was unlikely to be repaid according to the terms of the loan agreement. But many commercial real estate loans that were graded ten or eleven, and therefore catalogued on the Loan Watch List, were not designated nonaccrual. Concern about these loans, therefore, was not revealed to the public.
State–Boston's case concerns these commercial real estate loans designated special-mention or substandard, or otherwise identified as potentially problematic, but not designated nonaccrual and therefore not disclosed to the public as a source of concern. According to State–Boston, Bancorp knew at least as early as the fall of 2006 that it had reason to be worried about the credit quality of the commercial real estate portfolio. State–Boston introduced evidence of what it viewed as lax underwriting, as well as internal communications within BankAtlantic that revealed concern that some borrowers might be unable to sell the land securing their loans to homebuilders, making their loans difficult to pay off.
In public statements, however, Bancorp denied concern about BankAtlantic's commercial real estate portfolio. On October 19, 2006, in a quarterly earnings conference call open to the public, James White, then Bancorp's Chief Financial Officer, said, “There's really nothing significant to note on the credit quality front[,] which in itself[,] given the current real estate environment [,] I think is a favorable commentary.” White noted that the land securing BankAtlantic's commercial real estate loans might be developed more slowly than anticipated, but suggested that the bank's underwriting gave investors reason for confidence: “[I]n our situation and because of our insistence on hard equity in projects,” he said, “we believe we're dealing with borrowers with staying power that will enable them to ride through this if indeed that trend does manifest.” On November 29, 2006, at an investor conference in New York, Jarett Levan, then President of BankAtlantic, announced that there were no troubling credit quality trends ahead.
Over time, however, Bancorp's private concerns about the commercial real estate portfolio—as reflected in the increasing number of accruing but special-mention or substandard loans on the Loan Watch List—intensified. The first iteration of the Loan Watch List generated during the class period showed no special-mention or substandard commercial real estate loans that were accruing and therefore unknown to the public. On March 31, 2007, however, the Loan Watch List showed, in addition to more than $20.5 million in nonaccrual land loans, an accruing but substandard land loan of more than $21.2 million.
By then, concern about the commercial real estate portfolio had registered in other internal communications as well. For example, in an email to several BankAtlantic officers dated March 14, 2007, Alan Levan, then Bancorp's CEO and Chairman of the Board of Directors, noted a “parade of land loans coming in for extensions recently” and warned, “It's pretty obvious the music has stopped. In most cases, the presold contract to a builder has either gone away or is in dispute or being modified.” And at a Credit Policy Committee meeting on March 21, 2007, several BankAtlantic officers discussed the recent “migrat[ion]” of about $90 million in commercial real estate loans from grade four, in the bank's risk-grading system, to the lower, but still passing, grades of five, six, and seven because of the weakening housing market.
On April 25, 2007, in an 8–K7 reporting financial results for the first quarter of 2007, and in another earnings conference call the next day, Bancorp partially disclosed its concern about the commercial real estate portfolio. In addition to disclosing the amount of BankAtlantic's nonaccrual loans, the 8–K warned,
The current environment for residential land acquisition and development loans is a concern, particularly in Florida, and represents an area where we remain very cautious in our credit management. In view of market conditions, we anticipate we may experience further deterioration in the portfolio over the next several quarters as the market attempts to absorb an oversupply of available lot inventory.
During the April 26 conference call, Alan Levan mentioned two nonaccrual loans in BankAtlantic's commercial real estate portfolio—one for about $12.5 million and one for about $7.5 million. Levan said both were BLB loans. He warned that the Florida housing market was slowing and that, as a result, the risk associated with the BLB portfolio could worsen. Because of market conditions, Levan said, homebuilders were becoming more reluctant to buy lots on the land that secured the BLB loans. That day, Bancorp's stock price dropped from $10.55 per share to $9.99.
Some of the risk associated with BankAtlantic's commercial real estate portfolio remained undisclosed. During the April 26 conference call, Levan did not disclose that a $21.2 million BLB loan had been designated “substandard” and placed on the Loan Watch List. Nor did he disclose concern about the non-BLB portion of the commercial real estate portfolio. But in fact, the smaller of the two loans Levan mentioned during the conference call was a non-BLB loan, and the “parade” of extensions mentioned in his March 14 email had included non-BLB loans.
Over the year, the risk associated with the commercial real estate portfolio worsened considerably. By the end of April 2007, the Loan Watch List showed $60.3 million in special-mention or substandard BLB loans, in addition to the $12.5 million loan Levan had mentioned, and more than $38.7 million in special-mention or substandard non-BLB loans, in addition to the $7.5 million nonaccrual loan Levan had mentioned. By the end of June 2007, BankAtlantic had $12.6 million in nonaccrual BLB loans and $3.2 million in nonaccrual non-BLB loans, which were reflected in the total amount of nonaccrual loans disclosed in an 8–K on July 24, 2007. Undisclosed, however, were the additional $90.7 million in accruing but special-mention or substandard BLB loans and $55.5 million in accruing but special-mention or substandard non-BLB loans.
As had Alan Levan during the April 26 conference call, Bancorp's public statements continued to note the risk associated with BankAtlantic's BLB loans but minimize the risk associated with the non-BLB portion of the commercial real estate portfolio. Bancorp's 10–Q8 for the first quarter of 2007, filed on May 10, told the public that “management consider[ed] these other loans”—the non-BLB loans—“to be of relatively lower risk than the ‘builder land loans.’ ”9 In another publicly accessible quarterly earnings conference call on July 25, 2007, Alan Levan said BankAtlantic's commercial real estate portfolio had “performed extremely well.” “The one category that we just are focused on,” he said, “is this land loan builder portfolio”—that is, the BLB portfolio—“because, you know, just from one day to the next, the entire homebuilding industry, you know, went into a state of flux and turmoil and is impacting that particular class.” But, he said, “there are no particular asset classes that we're concerned about other than that one class.” And in both its first-quarter and second-quarter 10–Qs, Bancorp gave a figure purporting to represent BankAtlantic's “total potential problem loans”10 —$7.8 million in the first-quarter 10–Q and $8.3 million in the second-quarter 10–Q—that could not possibly have included all of the commercial real estate loans on the Loan Watch List.11
On October 25, 2007, Bancorp released an 8–K that, according to State–Boston, brought the fraud to an end. The 8–K reported that BankAtlantic's provision for loan losses in the third quarter of 2007 was $48.9 million, up from $4.9 million in the previous quarter. It also revealed that BankAtlantic held $156.3 million in nonaccrual commercial real estate loans,12 up from $21.8 million in total nonaccrual loans at the end of June 2007. And it revealed for the first time the amount of substandard commercial real estate loans held by BankAtlantic—at that point, $90.3 million, not counting nonaccrual loans.13 By the time the 8–K was released, Bancorp's stock price had already fallen gradually over the class period, from $12.66 per share to $7.65. On October 26, 2007, it fell by another $2.93, or 38 percent, to $4.72.
Three days later, the present action was filed in the United States District Court for the Southern District of Florida.14 State–Boston asserted claims under § 10(b) of the Securities Exchange Act and Rule 10b–5, alleging that Bancorp had fraudulently misrepresented the risk associated with BankAtlantic's commercial real estate portfolio.15 State–Boston alleged that although some of that risk was revealed in the 8–K released on April 25, 2007, Bancorp had continued to mislead the public by suggesting that the risk was limited to the BLB segment of the portfolio.16 State–Boston asserted a fraud-on-the-market theory, invoking the presumption of reliance approved by the Supreme Court in Basic, Inc. v. Levinson, 485 U.S. 224, 245–47, 108 S.Ct. 978, 990–92, 99 L.Ed.2d 194 (1988). It claimed that Bancorp's misrepresentations had artificially inflated the price of its stock and that investors had suffered damages when the price dropped in response to the disclosures on April 25, 2007, and October 25, 2007. The District Court certified a plaintiff class of “[a]ll persons or entities who purchased or otherwise acquired shares of the Class A common stock of BankAtlantic Bancorp, Inc. between November 9, 2005 and October 25, 2007 and were damaged thereby.”17
State–Boston's only evidence of loss causation and damages was the expert testimony of Candace Preston, a financial analyst. Preston performed an event study, a statistical technique for measuring the effect of new information on the market price of a security, to determine how much of the decline in the price of Bancorp's stock on April 26, 2007, and October 26, 2007, was attributable to factors specific to Bancorp, rather than to general market or industry factors.18 Preston used the S & P 500 Index to eliminate any portion of the price declines attributable to market-wide factors. To remove the effect of industry-specific factors, she relied on the NASDAQ Bank Index, an index of the stock prices of hundreds of banks and bank holding companies traded on the NASDAQ. She testified that these were the indexes used by Bancorp to compare its stock price movements to market and industry trends and that she had found a “statistical fit” between Bancorp's stock price movements and those indexes.
On April 26, 2007, although Bancorp's stock price dropped more than 5 percent, the S & P 500 and the NASDAQ Bank Index each fell less than 1 percent. Preston concluded based on those indexes that, of the 56–cent April 26 price decline, 55 cents could not be explained by market or industry factors and therefore must have resulted from company-specific factors. To isolate the amount attributable to the alleged fraud, as opposed to other company-specific factors, Preston looked at several analysts' projections of Bancorp's earnings per share for 2007. Those projections, she observed, dropped by an average of 15 cents after the April 26 disclosures. Based on information in the analysts' reports, Preston concluded that two-thirds of that drop in Bancorp's projected earnings were attributable to the disclosure of previously concealed risk in the commercial real estate portfolio on April 26. She then reasoned that the same proportion of the 55–cent residual decline19 in Bancorp's stock price—two-thirds, or 37 cents—was attributable to the fraud.
On October 26, 2007, as Bancorp's stock fell 38 percent, the S & P 500 rose about 1 percent, and the NASDAQ Bank Index rose 2 percent. Preston concluded that but for company-specific factors, Bancorp's stock price would have risen on that day. She thus found a residual decline of $3.15, even more than the actual decline of $2.93. To exclude company-specific factors other than the fraud, Preston looked at analyst reports responding to the October 25 disclosures. Because analysts seemed most concerned about the deterioration of the commercial real estate portfolio, Preston concluded that all of the residual decline was attributable to the disclosure of previously concealed risk in that portfolio. She therefore opined that the entire October 26 price decline of $2.93 was attributable to the fraud.
The District Court submitted the case to the jury on a verdict form that divided the case into two periods, the first from October 19, 2006, to April 25, 2007—that is, up to the partial disclosure in the April 25 8–K—and the second from April 26, 2007, to October 25, 2007. For each period, the verdict form requested a general verdict on liability and damages, along with answers to several interrogatories asking whether each of nineteen alleged misstatements was fraudulent and made with scienter, and a number of questions related to control-person liability under § 20(a) of the Securities Exchange Act.20 See Fed.R.Civ.P. 49(b)(1) (“The court may submit to the jury forms for a general verdict, together with written questions on one or more issues of fact that the jury must decide.”).
With respect to the first period, the jury found that Bancorp had violated § 10(b) but awarded no damages. With respect to the second period, the jury found that Bancorp, in eight statements, had violated § 10(b) and awarded damages of $2.41 per share. Upon the return of the verdict form, the District Court noted an apparent inconsistency in the jury's findings on the first statement from the second period.21 The jury found that Alan Levan had “acted knowingly with respect to that statement.” It also found, however, that Levan had “acted in good faith and did not directly or indirectly induce the Section 10(b) violation.”22
Bancorp moved for judgment as a matter of law under Federal Rule of Civil Procedure 50(b) and for a new trial under Rule 59.23 In ruling on the Rule 50(b) motion, the District Court focused on the jury's findings. Because of the perceived inconsistency noted above, the District Court discarded the jury's finding that the first statement from the second period had been made knowingly and refused to consider that statement as a potential basis for liability. The court then turned to the other statements the jury had found to violate § 10(b), ignoring the others. Concluding that State–Boston had not presented evidence sufficient to support a finding that the statements found by the jury to violate § 10(b) had caused their losses, the District Court granted Bancorp's motion for judgment as a matter of law. The court also conditionally denied Bancorp's motion for a new trial. See Fed.R.Civ.P. 50(c)(1) (“If the court grants a renewed motion for judgment as a matter of law, it must also conditionally rule on any motion for a new trial by determining whether a new trial should be granted if the judgment is later vacated or reversed.”). State–Boston now appeals the District Court's grant of judgment as a matter of law, and Bancorp cross-appeals the conditional denial of its motion for a new trial.
We review a district court's ruling on a motion for judgment as a matter of law de novo. Goodman–Gable–Gould Co. v. Tiara Condo. Ass'n, 595 F.3d 1203, 1213 n. 29 (11th Cir.2010). In deciding a motion for judgment as a matter of law, we review all the evidence, drawing all reasonable inferences in favor of the nonmoving party. Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 150, 120 S.Ct. 2097, 2110, 147 L.Ed.2d 105 (2000). We do not make credibility determinations or weigh the evidence. Id. We give credence to evidence supporting the nonmoving party's case, as well as “uncontradicted and unimpeached” evidence supporting the moving party, “at least to the extent that that evidence comes from disinterested witnesses.” Id. at 151, 120 S.Ct. at 2110 (internal quotation marks omitted).
The District Court erred when it relied on the jury's findings in granting Bancorp's renewed motion for judgment as a matter of law. Federal Rule of Civil Procedure 50 allows a district court to grant a motion for judgment as a matter of law if “the court finds that a reasonable jury would not have a legally sufficient evidentiary basis to find for the [nonmoving party].” Fed.R.Civ.P. 50(a). This is the standard whether the motion is made before or after the case is submitted to the jury. Chaney v. City of Orlando, 483 F.3d 1221, 1227 (11th Cir.2007); see also 9B Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2537, at 619–24 (3d ed. 2008) (“The standard for granting a renewed motion for judgment as a matter of law under Rule 50(b) is precisely the same as the standard for granting the pre-submission motion under Rule 50(a). Thus, the post-verdict motion for judgment can be granted only if the prior motion should have been granted.” (footnote omitted)). Only the sufficiency of the evidence matters; what the jury actually found is irrelevant. See Chaney, 483 F.3d at 1227–28.
Thus, we held in Chaney that the District Court had erred when, in granting a renewed motion for judgment as a matter of law, it “effectively based its conclusions on the jury findings contained in the special verdict form.” Id. at 1227. In Chaney, the plaintiff sued the City of Orlando and a city police officer under 42 U.S.C. § 1983, alleging wrongful arrest, excessive force, and malicious prosecution. Id. at 1222. The District Court submitted the case to the jury on a special-verdict form, and the jury found, among other things, that the plaintiff's arrest had been supported by probable cause. Id. at 1225. The court relied heavily on that finding in its order granting the officer's renewed motion for judgment as a matter of law. See id. at 1225–26. The court reasoned that because the arrest was supported by probable cause, it could not be wrongful. Id. at 1225. The court also relied on the probable-cause finding to justify its rejection of the plaintiff's other claims. Id. at 1226. We held that it was error to grant the Rule 50 motion based in part on the jury's findings instead of focusing solely on the sufficiency of the evidence. Id. at 1228.
The District Court made a similar error here. Instead of considering whether the evidence was sufficient to support a verdict in favor of State–Boston, the court relied on the perceived inconsistency of two of the jury's answers to the special interrogatories24 as a ground for discarding the possibility that one of Bancorp's statements could support liability. The court also ignored statements not found by the jury to violate § 10(b), solely on the basis of the jury's findings. This the court could not do.
Nevertheless, we may affirm for any reason supported by the record. E.g., United States v. Harris, 608 F.3d 1222, 1227 (11th Cir.2010). In this case, we discern such a reason: State–Boston did not introduce evidence sufficient to support a finding in its favor on the element of loss causation. It failed to adequately separate losses caused by fraud from those caused by the 2007 collapse of the Florida real estate market. The jury therefore did not have a sufficient evidentiary basis to conclude that the fraud was a substantial contributing factor in bringing about the class's losses.
In a Rule 10b–5 securities fraud action, the plaintiff must prove (1) a material misrepresentation or omission, (2) scienter, (3) a connection between the misrepresentation or omission and the purchase or sale of a security, (4) reliance upon the misrepresentation or omission, (5) economic loss, and (6) loss causation. Ledford v. Peeples, 657 F.3d 1222, 1248 (11th Cir.2011). Proving loss causation requires more than showing that the plaintiff bought the security at a price that was artificially inflated by fraudulent misrepresentations or omissions. Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 338, 125 S.Ct. 1627, 1629, 161 L.Ed.2d 577 (2005); Robbins v. Koger Props., Inc., 116 F.3d 1441, 1448 (11th Cir.1997). The plaintiff must offer “proof of a causal connection between the misrepresentation and the investment's subsequent decline in value.” Robbins, 116 F.3d at 1448. In other words, in a fraud-on-the-market case, the plaintiff must prove not only that a fraudulent misrepresentation artificially inflated the security's value but also that “the fraud-induced inflation that was baked into the plaintiff's purchase price was subsequently removed from the stock's price, thereby causing losses to the plaintiff.” FindWhat Investor Grp. v. FindWhat.com, 658 F.3d 1282, 1311 (11th Cir.2011) (citing Robbins, 116 F.3d at 1448).
As the Supreme Court observed in Dura, when an investor buys stock at an artificially inflated price and resells at a lower price, the price decline, and the investor's consequent loss, may result in part from factors other than the dissipation of fraud-induced inflation. “[T]hat lower price,” the Court explained, “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, 125 S.Ct. at 1632. Recognizing as much in Robbins, we concluded that loss causation does not require proof that the fraud—or, more precisely, the dissipation of the fraud-induced inflation—was the sole cause of the security's decline in value. 116 F.3d at 1447. We held, however, that the plaintiff must show that it was a “substantial” or “significant contributing cause.” Id. (quoting Bruschi v. Brown, 876 F.2d 1526, 1531 (11th Cir.1989)) (internal quotation marks omitted). We also noted that the standard for proof of damages is more demanding. As we explained in Robbins,
The distinction between the loss causation requirement and proof of damages is important. To satisfy the loss causation element, a plaintiff need not show that a misrepresentation was the sole reason for the investment's decline in value. Ultimately, however, a plaintiff will be allowed to recover only damages actually caused by the misrepresentation․ [A]s long as the misrepresentation is one substantial cause of the investment's decline in value, other contributing forces will not bar recovery under the loss causation requirement. But in determining recoverable damages, these contributing forces must be isolated and removed.
Id. at 1447 n. 5.
Thus, to succeed in a fraud-on-the-market case, it is not enough to point to a decline in the security's price after the truth of the misrepresented matter was revealed to the public. The plaintiff must also offer evidence sufficient to allow the jury to separate portions of the price decline attributable to causes unrelated to the fraud, leaving only the part of the price decline attributable to the dissipation of the fraud-induced inflation. A precise apportionment, as we explained in Robbins, is needed only to prove the amount of damages owed to the plaintiff. See id. at 1447 & n. 5; see also Miller v. Asensio & Co., 364 F.3d 223, 232 (4th Cir.2004) (“[I]n a given case, a jury could properly conclude that (1) the plaintiff proved the defendant's fraud constituted a substantial cause of plaintiff's loss and so find the defendant liable but (2) the plaintiff failed to provide a method to discern by just and reasonable inference the amount of plaintiff's loss solely caused by defendant's fraud, and so refuse to award the plaintiff any damages.” (citation omitted) (internal quotation marks omitted)). But if there are confounding factors that could account for much of the decline in the price of the security, the plaintiff must offer some evidence separating the various causes of the decline in the security's price even to establish loss causation. Otherwise the jury has no basis on which to conclude that the dissipation of the fraud-induced inflation was a substantial factor in bringing about the plaintiff's loss. See In re Scientific Atlanta, Inc. Sec. Litig., 754 F.Supp.2d 1339, 1376 (N.D.Ga.2010) (“[A] determination of the substantiality of the fraudulent conduct's effect requires some measurement of the loss attributable to that conduct.”).
In this case, State–Boston claims class members purchased Bancorp stock at prices that were artificially inflated because Bancorp fraudulently concealed the poor credit quality of BankAtlantic's commercial real estate portfolio, and that those shares lost value when the portfolio's deterioration was revealed to the market. In effect, State–Boston relies on what some courts have called a “materialization of the concealed risk” theory of loss causation. Lentell v. Merrill Lynch & Co., 396 F.3d 161, 173 (2d Cir.2005); see also Ray v. Citigroup Global Mkts., Inc., 482 F.3d 991, 995 (7th Cir.2007) (“There are several ways in which a plaintiff might go about proving loss causation. The first is sometimes called the ‘materialization of risk’ standard.”). That theory allows liability on a securities fraud claim even if the decline in a security's price is not caused by the market's reaction to a corrective disclosure revealing precisely the facts concealed by the fraud, as they existed at the time of the defendant's misstatements. Under the theory, the plaintiff may prove loss causation by showing, instead, that the materialization of a fraudulently concealed risk caused the price inflation induced by the concealment of that risk to dissipate.25
In its briefs to this court, State–Boston does not appear to commit to either a straightforward corrective-disclosure or materialization-of-risk theory of loss causation. It characterizes Preston's testimony as showing that “the entirety of the October 26, 2007 stock price drop was caused by the revelation of previously undisclosed information regarding BLB and non-BLB risk, or [ ] the materialization of previously concealed BLB and non-BLB risk.” Appellants' Br. 34 (emphasis added). But State–Boston can only be relying on a materialization-of-risk theory. Suppose for the moment that, as State–Boston's expert testified, the entire decline in Bancorp's stock price on October 26, 2007, did result from the market's reaction to the 8–K's disclosure of problems in BankAtlantic's commercial real estate portfolio. Even on that assumption, the portfolio's condition as disclosed in the October 2007 8–K was worse than the portfolio's condition at the time of the misstatements.26 Some of the facts the 8–K disclosed—some of the facts that, according to State–Boston, made the stock price fall—did not exist at the time of Bancorp's misstatements. And the facts concealed or misrepresented by Bancorp's misstatements, as they existed at the time of those misstatements—for example, the number, type, and dollar value of the land loans on the Loan Watch List—were not in fact disclosed in the October 2007 8–K. Thus, even if the price decline on October 26, 2007, reflected only the market's reaction to the 8–K's disclosure of weakness in the commercial real estate portfolio, there is no reason to believe it reflected only the market's reaction to the correction of Bancorp's previous misstatements.27 State–Boston's argument, then, can only be that the October 2007 8–K reflected the materialization of a risk—the risk that the commercial real estate portfolio would deteriorate—that Bancorp had concealed from the public, and that the materialization of that risk removed the inflation created by its concealment.
State–Boston's expert, Candace Preston, testified that the entire 38 percent decrease in Bancorp's stock price on October 26, 2007, resulted from the materialization of that risk. As described in part II, supra, Preston attempted to isolate the effect of company-specific factors from the effect of general market trends by comparing the change in the Bancorp's stock price to the change in the S & P 500. She also attempted to separate the effect of company-specific factors from industry-wide trends by comparing Bancorp stock to the NASDAQ Bank Index, an index of the stock prices of hundreds of banks and bank holding companies traded on the NASDAQ.
Preston failed, however, to account for the effects of the collapse of the Florida real estate market. The NASDAQ Bank Index may be well suited to capture the effects of national trends in the banking industry, such as the broader national financial crisis that reached its nadir in 2008. But in 2007, Florida, having benefitted more than most states from the real estate boom of the previous years, was hit harder than most by the ensuing bust.28 And Florida financial institutions, as Preston admitted on cross-examination, made up only a small percentage of the NASDAQ Bank Index. That index, therefore, would be inappropriate for the task of filtering out the effects of industry-wide factors that might affect the stock price of a bank, or of the holding company of a bank, whose assets were concentrated in loans tied to Florida real estate in 2007.
BankAtlantic is just such a bank. As Bancorp acknowledged in several public SEC filings during the class period, BankAtlantic's assets were concentrated in loans tied to Florida real estate. As a result, BankAtlantic and Bancorp were particularly susceptible to any deterioration in the Florida real estate market, in addition to any national developments. To support a finding that Bancorp's misstatements were a substantial factor in bringing about its losses, therefore, State–Boston had to present evidence that would give a jury some indication, however rough, of how much of the decline in Bancorp's stock price resulted not from the fraud but from the general downturn in the Florida real estate market—the risk of which Bancorp is not alleged to have concealed.29 Cf. Lentell, 396 F.3d at 176–77 (concluding that the plaintiffs did not adequately allege loss causation because none of the alleged misstatements had concealed the price-volatility risk that materialized and caused at least some of the plaintiffs' losses); Bastian v. Petren Res. Corp., 892 F.2d 680, 686 (7th Cir.1990) (rejecting a Rule 10b–5 claim because the plaintiffs, investors in oil and gas partnerships, “knew they were assuming a risk that oil prices might drop unexpectedly” and were “unwilling to try to prove that anything beyond the materializing of that risk caused their loss”); In re Merrill Lynch & Co. Research Reports Sec. Litig., 568 F.Supp.2d 349, 364 (S.D.N.Y.2008) (“[T]he [defendants] warned investors that the fate of CMGI, a company that owned numerous subsidiary Internet-based holdings, was dependent on the fate of the Internet sector as a whole․ [T]he Complaint does not assert facts that distinguish between the alleged fraud and the market-wide collapse of Internet stocks as the cause of [the plaintiff's] losses.”). State–Boston failed to do so. None of its evidence excluded the possibility that class members' losses resulted not from anything specific about BankAtlantic's commercial real estate portfolio that Bancorp hid from the public, but from market forces that it had warned of—and that would likely have caused significant losses for an investor in any bank with a significant credit portfolio in commercial real estate in Florida in 2007. Bancorp is therefore entitled to judgment as a matter of law.30
For the foregoing reasons, the judgment of the District Court is
1. Section 10(b) provides, in relevant part:It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce or of the mails, or of any facility of any national securities exchange—․(b) To use or employ, in connection with the purchase or sale of any security registered on a national securities exchange or any security not so registered, ․ any manipulative or deceptive device or contrivance in contravention of such rules and regulations as the [Securities and Exchange] Commission may prescribe as necessary or appropriate in the public interest or for the protection of investors.15 U.S.C. § 78j(b).
2. Rule 10b–5, promulgated under § 10(b) of the Securities Exchange Act, provides:It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange,(a) To employ any device, scheme, or artifice to defraud,(b) To make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading, or(c) To engage in any act, practice, or course of business which operates or would operate as a fraud or deceit upon any person,in connection with the purchase or sale of any security.17 C.F.R. § 240.10b–5.
3. The officers named as defendants were James White, the former Executive Vice President and Chief Financial Officer (“CFO”) of the holding company and former CFO of its subsidiary, BankAtlantic; John Abdo, the Vice Chairman of the Board of Directors for the holding company and the subsidiary; Valerie Toalson, the CFO of the holding company and Executive Vice President and CFO of the subsidiary; Jarett Levan, the President of the subsidiary and, after January 16, 2007, the President of the holding company and CEO of the subsidiary; and Alan Levan, the former Chairman of the Board and CEO of the holding company and former Chairman of the Board and President and CEO of the subsidiary.
4. The class certified by the District Court originally included everyone who bought Bancorp stock between November 9, 2005, and October 25, 2007, and suffered damages as a result. But on August 18, 2010, the District Court granted Bancorp summary judgment on all claims arising from the period from November 9, 2005, through October 18, 2006. Because State–Boston does not challenge that disposition on appeal, no claims from that period are before us.
5. Because of the procedural posture of this case, in reciting the facts, we give credence to State–Boston's evidence, as well as uncontradicted and unimpeached evidence supporting Bancorp, and draw all reasonable inferences in State–Boston's favor. See Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 150–51, 120 S.Ct. 2097, 2110, 147 L.Ed.2d 105 (2000). For a fuller statement of the standard of review of a ruling on a motion for judgment as a matter of law, see infra part III.
6. According to BankAtlantic's Commercial Real Estate and Commercial Loan Policy, special-mention assets, assigned a grade of ten, “have potential weaknesses that deserve management's close attention. If left uncorrected, these potential weaknesses may result in deterioration of the repayment prospects for the asset or in the institution's credit position at some future date.” A grade of eleven—or, equivalently, a designation of “substandard”—reflects an assessment that the loan “is inadequately protected by the current sound worth and paying capacity of the obligor or the collateral pledged, if any.” Substandard assets have a “well defined weakness or set of weaknesses” and “are characterized by the distinct ‘possibility’ that the Bank will sustain some loss.” That loss potential, however, need only exist “in the aggregate amount of Substandard assets”; it “does not have to exist in individual assets classified Substandard.”BankAtlantic's policies also provided that a grade of twelve or thirteen warranted placement on the Loan Watch List. Those grades, however, were seldom assigned in practice. Typically, by the time a loan had deteriorated enough to warrant a grade of twelve or thirteen, a write-down of its value had already occurred, making a further downgrade within the one-to-thirteen risk-grading system unnecessary. Notably, the Loan Watch List also included any loans designated “nonaccrual,” a term we explain below, or past due 90 days or more.
7. According to the SEC's website, “Form 8–K is the ‘current report’ companies must file with the SEC to announce major events that shareholders should know about.” Form 8–K, U.S. Sec. & Exch. Comm'n, http://www.sec.gov/answers/form8k.htm (last modified July 22, 2009).
8. Form 10–Q, a report that “must be filed [with the SEC] for each of the first three fiscal quarters of the company's fiscal year,” “includes unaudited financial statements and provides a continuing view of the company's financial position during the year.” Form 10–Q, U.S. Sec. & Exch. Comm'n, http://www.sec.gov/answers/form10q.htm (last modified Sept. 2, 2011).
9. The context of this remark was as follows:Conditions in the residential real estate market nationally and in Florida continued to deteriorate during the first quarter of 2007. New home sales and applications for building permits fell significantly from peak levels during 2005 and inventories of unsold homes have significantly increased. The Bank's commercial real estate loan portfolio consists of several sub-categories of loans, each with differing collateral and different levels of risk. The “builder land loan” segment, at approximately $140 million, consists of land loans to borrowers who have option agreements with regional and/or national builders. These loans were originally underwritten based on projected sales of the developed lots to the builders/option holders and timely repayment of the loans is primarily dependent upon the acquisition of the property pursuant to the options. If the lots are not acquired as originally anticipated, we anticipate that the borrower may not be in a position to service the loan with the likely result being an increase in nonperforming loans and loan losses in this category.The “builder land loan” segment discussed above is part of BankAtlantic's total commercial real estate acquisition and development portfolio of approximately $562 million as of March 31, 2007. The loans other than the “builder land loans” in this category are generally secured by residential and commercial real estate which will be fully developed by the borrower or sold to third parties. These loans generally involve property with a longer investment and development horizon and are guaranteed by the borrower or individuals and/or secured by additional collateral such that it is expected that the borrower will have the ability to service the debt under current conditions for a longer period of time. Accordingly, management considers these other loans to be of relatively lower risk than the “builder land loans.”Similar language appeared in Bancorp's second-quarter 10–Q filed on August 9, 2007.
10. Total potential problem loans, as presented in the 10–Qs, represented the sum of loans past due 90 days or more, performing-impaired loans, and restructured loans.
11. At the end of the first quarter of 2007—that is, at the end of March—the Loan Watch List showed more than $21.2 million in accruing but special-mention or substandard commercial real estate loans. And as we note in part I.B, supra, at the end of the second quarter of 2007—that is, at the end of June—the Loan Watch List showed about $146.2 million in accruing but special-mention or substandard commercial real estate loans.
12. BLB loans accounted for $81.1 million of this total, and non-BLB loans for $75.2 million.
13. Of this total, BLB loans accounted for $28.7 million, and non-BLB loans for $61.6 million. The 8–K referred to these loans as “classified,” but Jay McClung, BankAtlantic's Chief Risk Officer, testified that “classified” was equivalent to “substandard.”
14. Although the initial complaint was filed on October 29, 2007, by an individual shareholder named Joseph Hubbard, our characterization of the allegations relies on the First Amended Consolidated Complaint filed by State–Boston on January 12, 2009, after it was appointed lead plaintiff.
15. The individuals named as defendants were also alleged to be liable as control persons under § 20(a) of the Securities Exchange Act. See 15 U.S.C. § 78t(a) (“Every person who, directly or indirectly, controls any person liable under any provision of this chapter or of any rule or regulation thereunder shall also be liable jointly and severally with and to the same extent as such controlled person ․ is liable ․, unless the controlling person acted in good faith and did not directly or indirectly induce the act or acts constituting the violation or cause of action.”).
16. Although the First Amended Consolidated Complaint did not allege any other fraud after April 25, 2007, State–Boston argued at trial that Bancorp's statements during that period had also fraudulently concealed the level of risk associated with the BLB portfolio by omitting to mention accruing but special-mention or substandard BLB loans.
17. As noted above, however, the District Court later granted summary judgment for Bancorp on all claims arising from the period between November 9, 2005, and October 18, 2006, effectively shortening the class period. See supra note 4.
18. The methodology of an event study is as follows:Event study analysis compares the day-to-day percentage change in the market price of a company's common stock (known as a “return”) to the return predicted by a market model that uses a market index, such as the S & P 500 Index or the NASDAQ Composite Index, and possibly an industry index. The market model describes the normal relation between the return on the company's common stock and the return on the market and industry indexes. When significant new information about the company (e.g., corrective disclosures, earnings reports, dividend changes, stock splits, regulatory rulings, acquisition bids, asset sales, or tax legislation) is disclosed to the market, the market model is used to determine the component of the stock return that would be expected based on the return of the overall market and industry. The remaining component of the stock return (that which cannot be explained by the return on the market and industry) is attributed to the new company-specific information or to chance. If the disclosure of the new information is accompanied by a stock return that is outside of the stock's normal volatility range (as measured by the market model), then the return is said to be “statistically significant.”Frank Torchio, Proper Event Study Analysis in Securities Litigation, 35 J. Corp. L. 159, 160–61 (2009) (footnotes omitted).
19. Preston used the term “residual decline” to refer to the difference between the stock price's predicted behavior, based on the indexes chosen, and its actual behavior—that is, the portion of the stock price decline not explainable by market or industry forces, as determined by the event study.
20. Section 20(a) provides, in relevant part:Every person who, directly or indirectly, controls any person liable under any provision of this chapter or of any rule or regulation thereunder shall also be liable jointly and severally with and to the same extent as such controlled person ․ is liable ․, unless the controlling person acted in good faith and did not directly or indirectly induce the act or acts constituting the violation or cause of action.15 U.S.C. § 78t(a).
21. This was Alan Levan's statement during the earnings conference call on April 26, 2007, that both of the nonaccrual commercial real estate loans were BLB loans.
22. The verdict form asked for a determination of that fact because it is a defense to control-person liability under § 20(a) that “the controlling person acted in good faith and did not directly or indirectly induce the act or acts constituting the violation or cause of action.” 15 U.S.C. § 78t(a).
23. Bancorp had timely moved for judgment as a matter of law at the close of State–Boston's evidence at trial.We note that because the jury returned a general verdict awarding no damages with respect to the first part of the class period, Bancorp's renewed motion for judgment as a matter of law—and the ruling under review here—concerned only the § 10(b) claim based on statements made during the second period, from April 26, 2007, to October 25, 2007. See In re BankAtlantic Bancorp, Inc. Sec. Litig., No. 07–61542–CIV–UNGARO, slip op. at 19–20 (S.D.Fla. Apr. 25, 2011) (“The parties agree on most of the judgment compelled by the verdict—all Defendants are entitled to judgment in their favor for the first period and Defendants [John] Abdo, Jarett Levan, and [James] White are entitled to judgment in their favor for the second. The parties dispute only the proper judgment regarding Defendants Bancorp, Alan Levan, and [Valerie] Toalson as to the second period.”).
24. We leave unanswered the question whether the jury's findings were in fact inconsistent and, if they were, what the effect of any such inconsistency might be. No answer is necessary to the disposition of this appeal.
25. For example, in In re Omnicom Group, Inc. Securities Litigation, 597 F.3d 501 (2d Cir.2010), investors brought a securities fraud class action alleging, under a fraud-on-the-market theory, that Omnicom, a marketing and advertising holding company, had failed to accurately report in its books the declining value of Internet companies it owned. Id. at 508. On appeal from a grant of summary judgment for the defendants, the Second Circuit concluded that the plaintiffs had not shown that any corrective disclosure of the inaccurate bookkeeping had caused a price decline. Id. at 513.The court nevertheless went on to consider whether the plaintiffs had shown that the stock price decline “was foreseeable and caused by the materialization of the risk concealed by the fraudulent statement.” Id. (quoting ATSI Commc'ns, Inc. v. Shaar Fund, Ltd., 493 F.3d 87, 107 (2d Cir.2007)) (internal quotation marks omitted). Even without a corrective disclosure, the court explained, the plaintiffs could show loss causation if they showed that the subsequent “resignation [of Robert Callander, a director and the chair of Omnicom's audit committee,] and the ensuing negative media attention were foreseeable risks of the fraudulent [bookkeeping] and caused the temporary share price decline in June 2002.” Id.; see also id. at 511 (noting that “[e]stablishing either [of the plaintiffs'] theor[ies]”—“that the market reacted negatively to a corrective disclosure of the fraud” or “that negative investor inferences drawn from Callander's resignation and from the news stories in June 2002 caused the loss and were a foreseeable materialization of the risk concealed by the fraudulent statement”—“would suffice to show loss causation”). The court concluded, however, that Callander's resignation and the media's reaction—which ultimately caused the stock price decline—were too tenuously related to the fraudulent bookkeeping to be considered within the “zone of risk” concealed by the fraud. Id. at 513–14.The Second Circuit entertained a similar theory in Lentell v. Merrill Lynch & Co., 396 F.3d 161 (2d Cir.2005), an earlier fraud-on-the-market case in which investors sued Merrill Lynch asserting a claim based on allegedly false reports making investment recommendations. Id. at 164. There, the court suggested that the plaintiffs would have adequately pled loss causation had they alleged facts showing that the market reacted negatively either to a disclosure of the falsity of Merrill's recommendations or to the materialization of a risk concealed by the reports. See id. at 175 (“There is no allegation that the market reacted negatively to a corrective disclosure regarding the falsity of Merrill's ‘buy’ and ‘accumulate’ recommendations and no allegation that Merrill misstated or omitted risks that did lead to the loss. This is fatal under Second Circuit precedent.” (footnote omitted)).This court has never decided whether the materialization-of-concealed-risk theory may be used to prove loss causation in a fraud-on-the-market case. We need not reach the issue here. Instead, we assume, without deciding, that this approach is valid and explain why, even on that assumption, State–Boston failed to offer evidence sufficient to support a verdict in its favor.
26. Consider the dollar amount of land loans considered “classified” (in the 8–K's terms) or “substandard” (in the terms of BankAtlantic's internal risk-grading system) at the time of the October 2007 8–K and at the times of the various misstatements. The October 2007 8–K disclosed $90.3 million in classified or substandard land loans, not including loans designated nonaccrual. But at the time of the April 2007 earnings conference call, according to what was then the most recent iteration of the Loan Watch List, BankAtlantic had one commercial real estate loan—a BLB loan—of about $21.2 million that was accruing but substandard. As late as June 30, 2007—after several of the misstatements were made—the Loan Watch List showed only about $53.8 million in accruing but substandard commercial real estate loans, comprising $34.2 million in non-BLB loans and $19.6 million in BLB loans.The dollar amount of nonaccrual commercial real estate loans disclosed in the October 2007 8–K was also much greater than the amount of such loans at the time of any of Bancorp's misstatements. The 8–K reported that $156.3 million in commercial real estate loans, both BLB and non-BLB, were designated nonaccrual. But according to the Loan Watch List, as late as August 31, 2007—after all the misstatements were made—only $25.5 million in commercial real estate loans were designated nonaccrual.
27. Another way to put the point is that even if, as Preston testified, the price decline resulted solely from the 8–K's disclosure of weakness in the commercial real estate portfolio, not all of that decline can be attributed to the dissipation of inflation caused by Bancorp's misstatements. That decline still may have reflected the market's reaction to facts that could not possibly have been disclosed at the time of Bancorp's misstatements. There is therefore no reason, even accepting Preston's conclusions, to assume that the amount of that decline is equal to the difference between the price at which class members bought Bancorp's stock and the price at which they would have bought it if not for the fraud—the usual measure of damages. See Robbins v. Koger Props., Inc., 116 F.3d 1441, 1447 n. 5 (11th Cir.1997) (“The proper measure of damages utilizes the out-of-pocket rule: the plaintiff can recover ‘the difference between the price paid and the “real” value of the security, i.e., the fair market value absent the misrepresentations, at the time of the initial purchase by the defrauded buyer.’ ” (quoting Huddleston v. Herman & MacLean, 640 F.2d 534, 556 (5th Cir. Unit A 1981), aff'd in part and rev'd in part on other grounds, 459 U.S. 375, 103 S.Ct. 683, 74 L.Ed.2d 548 (1983))). Thus, to allow a jury to award damages on that usual measure, State–Boston would still have to separate the amount of the stock price decline attributable to the ultimate disclosure of facts about the portfolio that were known at the time of Bancorp's misstatements from the portion of the decline caused by the disclosure of new facts about the portfolio. State–Boston has, however, made no attempt to do so.We need not decide whether some other measure of damages would be appropriate in a case—like this one, on State–Boston's theory—in which the materialization of a concealed risk caused the price to decline by more than the difference between the inflated price at which the plaintiff purchased the security and the price the plaintiff would have paid had the risk not been fraudulently concealed. As we explain below, State–Boston has failed for other reasons to present evidence sufficient to support a finding of loss causation. It therefore cannot prevail on the issue of liability, and it is unnecessary to reach the damages issue.
28. Data compiled by the Office of Federal Housing Enterprise Oversight, which has since become the Federal Housing Finance Agency, help give a sense of the relative severity of Florida's real estate market downturn during 2007. Of the 20 metropolitan statistical areas with the lowest rates of house price appreciation over the year ending on September 30, 2007, eight were in Florida. Office of Fed. Hous. Enter. Oversight, House Prices Weaken Further in Most Recent Quarter 30 (2007), available at http://www.fhfa.gov/webfiles/1174/3q07hpi.pdf. Depending on the index used, Florida saw either the fourth-lowest or sixth-lowest rate of house price appreciation over the same period, see id. at 10, and had the fourth-highest foreclosure rate, see id. at 14.
29. In fact, Bancorp affirmatively warned the public of that risk. For example, in its 2006 10–K, filed on March 1, 2007, Bancorp cautioned:The national real estate market is showing signs of a slow down, particularly in areas that have seen significant growth, including Florida and California. Our loan portfolio is concentrated in commercial real estate loans (virtually all of which are located in Florida), residential mortgages (nationwide), and consumer home-equity loans (Florida). We have exposure to credit losses that may arise from this concentration in what many believe is a softening real estate sector. Included in the commercial real estate loans are approximately $389 million of land development loans, which are susceptible to extended maturities or borrower default due to a slow-down in Florida construction activity, and $95.1 million of development loans for low and mid-rise condominium projects in Florida, where there is an increasing supply of new construction in the face of falling demand.․The majority of BankAtlantic's loan portfolio consists of loans secured by real estate. BankAtlantic's loan portfolio included $2.2 billion of loans secured by residential real estate and $1.4 billion of commercial real estate, construction and development loans at December 31, 2006. At December 31, 2006, BankAtlantic's commercial real estate, construction and development loans, which are concentrated mainly in Florida, represented approximately 37.6% of its loan portfolio. The real estate market in Florida is currently exhibiting signs of weakness. If real estate values in Florida were to decline, the credit quality of BankAtlantic's loan portfolio and its earnings could be adversely impacted.
30. Because we affirm the District Court's grant of judgment as a matter of law, we dismiss as moot Bancorp's cross-appeal from the District Court's conditional denial of its motion for a new trial. See Rankin v. Evans, 133 F.3d 1425, 1428, 1443 (11th Cir.1998) (affirming a post-verdict grant of judgment as a matter of law and dismissing as moot the defendants' cross-appeal from the conditional denial of their motion for a new trial).
TJOFLAT, Circuit Judge: