Opinion
Civil Action No. 3:04-CV-1836-D (Consolidated with Civil Action No. 3:04-CV-1869-D and Civil Action No. 3:04-CV-2156-D).
April 2, 2008.
Robert E. Jenkins, Law Office Of Robert E. Jenkins PC, Roger F. Claxton, Claxton Hill PLLC, Dallas, TX, Thomas E. Bilek, The Bilek Law Firm LLP, Houston, TX, for Plaintiffs.
Martin B. Mcnamara, M. Byron Wilder, Gibson Dunn Crutcher-Dallas, William D. Sims, Jr., Vinson Elkins, Randal G. Mathis, AnissaF. Curliss, Mark M. Donheiser, Roger Evans, Mathis Donheiser, Dallas, TX, for Defendants.
MEMORANDUM OPINION AND ORDER
The dispositive question presented by lead plaintiff's ("plaintiff's") motion for class certification is whether it has established loss causation by a preponderance of the evidence. Concluding that it has not, the court denies the motion.
The court is deciding this motion without conducting an evidentiary hearing or receiving oral testimony, as permitted under Fed.R.Civ.P. 43(c) (formerly Rule 43(e) before it was renumbered effective December 1, 2007 under the restyled civil rules). "Normally, Rule 43[(c)] provides the procedure for a motion to be heard. As provided by Rule 43[(c)], the court may decide a motion on affidavits or may, apparently in its discretion, direct that the matter be heard on oral testimony." Sanders v. Monsanto Co., 574 F.2d 198, 199-200 (5th Cir. 1978) (addressing former Rule 43(e) and holding that civil contempt hearing was trial-type proceeding that required oral hearing rather than use of Rule 43(e) procedure). So far as the court is aware, no circuit court has addressed whether a class certification motion can be decided under Rule 43(e) (now Rule 43(c)) without an evidentiary hearing or receiving oral testimony. But in the Fifth Circuit, even a preliminary injunction application can be heard on the papers using the Rule 43(c) procedure. E.g., Fed. Sav. Loan Ins. Corp. v. Dixon, 835 F.2d 554, 558 (5th Cir. 1987) (addressing former Rule 43(e)). The court has located no district court decision that holds that Rule 43(c) (or former Rule 43(e)) is inapplicable to a class certification motion. One district court class certification decision appears to have been made under the Rule 43(e) procedure. See Lucky v. Bd. of Regents, 1981 WL 234, at *1 (S.D. Fla. June 29, 1981) (involving class certification hearing at which plaintiffs relied on allegations in complaint, affidavits submitted in support of separate preliminary injunction request, and supporting memorandum; defendants offered argument, but not evidence, referred to facts outlined in memorandum and supporting affidavits in response to preliminary injunction motion, and supplemented information by referring to discovery directed to class certification issues and by additional affidavit; and court stated that although "not unmindful of the potential usefulness and frequent necessity of an evidentiary hearing in determining the propriety of class certification, . . . an evidentiary hearing is not an absolute prerequisite for initial determination particularly when, as here, other sources of relevant, adequate, and largely undisputed information have been relied upon by both parties." (citing, inter alia, Rule 43(e)). And in a securities fraud action, Westlake v. Abrams, 98 F.R.D. 1, 3 (N.D. Ga. 1981), the district court decided the class certification motion on the briefs. Although at one point in its opinion it referred to Rule 43(e), it is not clear whether the court explicitly relied on Rule 43(e) in opting to decide the motion without an evidentiary hearing.
I
The relevant background facts of this securities fraud action are set out in two published opinions and need not be repeated. In the fraud-on-the-market context, the Fifth Circuit requires a plaintiff to establish loss causation "at the class certification stage by a preponderance of all admissible evidence." See Oscar Private Equity Invs. v. Allegiance Telecom, Inc., 487 F.3d 261, 269 (5th Cir. 2007). In Oscar the Fifth Circuit also specifically required that class certification "be supported by a showing of loss causation that targets the corrective disclosure appearing among other negative disclosures made at the same time." Id. at 262 (emphasis added). "[W]hen unrelated negative statements are announced contemporaneous of a corrective disclosure, the plaintiff must prove that it is more probable than not that it was this negative statement, and not other unrelated negative statements, that caused a significant amount of the decline." Id. at 270 (citations and internal quotation marks omitted).II
Plaintiff maintains that loss causation in this case is established by the evidence of an August 6, 2004 decline in defendant Belo Corporation's ("Belo's") stock price following the company's previous-day announcement of circulation declines stemming from circulation overstatements. The announcement stated, in relevant part:
Belo . . . announced today that The Dallas Morning News . . . will report a greater than expected decline in its September 2004 circulation. An internal investigation, which is ongoing, has disclosed practices and procedures that led to an overstatement in circulation, primarily in single copy sales. Belo estimates the decline in circulation related to this matter to be approximately 1.5 percent daily and five percent Sunday. This decline, coupled with a reduction in state circulation that was first communicated publicly on March 9, 2004, of approximately 2.5 percent daily and 3.5 percent Sunday, and an anticipated lower circulation volumes for the six month period ending September 30, 2004, will result in a total decline in circulation of approximately five percent daily and 11.5 percent Sunday when compared with reported September 2003 figures.
Although the briefing is not clear on this point, plaintiff also appears to argue that it can demonstrate loss causation by showing that defendants' false statements regarding circulation figures inflated Belo's stock price. P. Reply 3. But as the Supreme Court made clear in Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336 (2005), loss causation cannot be demonstrated merely by "establishing that the price of the security on the date of purchase was inflated because of the misrepresentation." Id. at 338 (quoting Broudo v. Dura Pharmaceuticals, Inc., 339 F.3d 933, 938 (9th Cir. 2003) (internal quotation marks and emphasis omitted). Loss causation instead concerns the moment when a defendant's misrepresentation "proximately caused the plaintiff's economic loss." Id. at 346. Because the court holds that plaintiff has failed to show loss causation under the applicable standard, it need not determine whether the alleged misrepresentations actually inflated Belo's stock price.
Ds. App. 87.
Defendants contend that plaintiff has failed to prove that the August 6, 2004 stock price decline was not caused by other negative, non-fraud related information contained in the August 5 announcement. They posit that the August 5 announcement was actually composed of multiple news disclosures, and that plaintiff has failed to prove that the stock-price decline was caused by the fraud-related disclosure, as Oscar requires. Defendants offer the expert testimony of Paul A. Gompers, Ph.D. ("Dr. Gompers"), who testifies that the August 5 announcement attributed the overall circulation decline to three separate sources, only one of which is related to the alleged fraud. Specifically, defendants argue that the announcement contained three separate pieces of information related to circulation. The announcement reported (1) an expected circulation decline of 1.5% daily and 5% Sunday related to Belo's previous circulation overstatement, (2) an expected 2.5% daily and 3.5% Sunday decline resulting from a change in the computation methodology of circulation, and (3) an expected 1.0% daily and 3% Sunday decline in circulation attributable to an industry-wide decline in circulation. Combined, these figures result in a 5% daily and 11.5% Sunday total decline in circulation.
In addition to defendant Belo, defendants Robert W. Decherd, James M. Moroney III, and Barry Peckham oppose the class certification motion.
Plaintiff offers in reply the expert testimony of Scott D. Hakala, Ph.D ("Dr. Hakala"), who avers that Belo's August 5 announcement cannot be parsed as defendants contend, and that all the reported decline in circulation is related to the alleged circulation inflation scheme. Plaintiff therefore argues that the multiple items of news that defendants cite "are all directly related to the announcement of the circulation overstatement scheme." P. Reply 8.
In reply, plaintiff also argues that its "well-pleaded allegations of loss causation based on events throughout the class period . . . have been further confirmed by the expert report of [Scott D. Hakala, Ph.D ("Dr. Hakala")]." P. Reply 9. This appears to suggest that plaintiff intends to rely on dates other than August 5 to establish loss causation. Plaintiff only specifically argues, however, that "loss causation is demonstrated via the negative disclosure on August 5," id. at 3, and it does not provide in its briefing any analysis of loss causation for any other day or time period. The expert report on which plaintiff intends to rely to establish pre-August 5 loss causation is incomplete. Dr. Hakala admits that "an exact analysis of each partially corrective date prior to August 6, 2004 . . . requires further discovery and detailed analysis not possible at this point in time. . . ." As to these dates, Dr. Hakala only speculates about what a "detailed analysis would find." Hakala Decl. 12. The court therefore only considers the August 5 announcement as the basis for plaintiff's allegation of loss causation.
III
The court finds that, contrary to plaintiff's assertions, the August 5 announcement is composed of various news pieces. It concludes, in this context, that plaintiff is required to "prove that it is more probable than not that it was [the overstatement-related] negative statement, and not other unrelated negative statements, that caused a significant amount of the decline." Oscar 487 F.3d at 266.
The August 5 announcement clearly delineates a total circulation decline attributable to three separate causes. The fact that, in the announcement, Belo specifically attributed the first cause of the decline to the overstatement implies that the other two causes of circulation declines are not related to the overstatement. Although from the announcement itself it is unclear what are the other two sources of the declines, defendants offer evidence that the second — "a reduction in state circulation" — refers to Belo's March 9, 2004 announcement that circulation of The Dallas Morning News ("DMN") was expected to decrease slightly, "attributable to a change in . . . methodology for calculating State circulation from one acceptable method to another." Ds. App. 20. Defendants have also presented evidence that the third cause of the decline — "anticipated lower circulation volumes for the six month period ending September 30, 2004" — was interpreted by analysts as resulting from "natural readership losses arising from market conditions." Id. at 24. Defendants thus maintain that of the total 5% daily and 11.5% Sunday declines mentioned in the August 5, 2004 announcement, 3.5% of the daily declines and 6.5% of the Sunday declines are attributable to a combination of a change in circulation calculation methodology and industry-wide circulation declines.
Plaintiff does not refute defendants' characterization of these declines. It appears to argue that these reported declines are nevertheless related to the overstatement.
Because not all items of circulation-related news contained in the August 5 announcement are, at least facially, related to the overstatement, the court concludes that this case requires a similar multi-layered loss-causation inquiry as that conducted in Oscar. In Oscar the Fifth Circuit required that the plaintiffs prove "(1) that the negative `truthful' information causing the decrease in price is related to an allegedly false, non-confirmatory positive statement made earlier and (2) that it is more probable than not that it was this negative statement, and not other unrelated negative statements, that caused a significant amount of the decline." Oscar, 487 F.3d at 266 (emphasis added).
In the present case, plaintiff does not even attempt to address each of the multiple announcements individually, much less demonstrate how they are related to the circulation overstatement. Plaintiff simply posits that "Belo's August statement was a single piece of negative news." P. Reply 8. Although conceding at certain points that "on August 5, 2004 Belo made multiple announcements," plaintiff merely argues that "each related to the circulation overstatement problem at DMN." P. Br. 23. To the extent that plaintiff seeks to rely on a theory that each of the multiple announcements is related to the circulation overstatement, it has the burden of demonstrating this fact. But plaintiff fails to argue, much less prove, that the second and third causes — the decline resulting from the March 9, 2004 announcement of a "change in . . . methodology for calculating State circulation," and the "anticipated lower circulation volumes for the six month period ending September 30, 2004" — are in fact related to the previously announced misrepresentations concerning circulation figures that the August 5, 2004 statement disclosed. Together, the second and third sources of decline account for most of the total loss in circulation. Without a showing that these two cited causes for the total circulation decline are "related to an allegedly false, non-confirmatory positive statement made earlier," the court cannot find that plaintiff has satisfied its burden under Oscar.
In fact, analysts commenting on the August 5, 2004 announcement clearly distinguish the portion of the reported circulation decline attributed to the questionable practices leading to the overstatements from the portion of the announcement that attributed the decline to the other sources.
Beyond the state-wide circulation issue, the other portions of Belo's circulation reductions are comprised of: 1) "Natural readership losses arising from market conditions of (1.0%) daily and (3.0%) Sunday; and 2) the more disturbing overstatement of circulation, costing the DMN (1.5%) daily and (5.0%) Sunday . . . (Citigroup Smith Barney, August 9, 2004)
The company will report a total decline . . . [t]hat includes 1.5% daily and 5% Sunday reduction relating to the [overstated circulation], 2.5% daily and 3.5% Sunday relating to state circulation (previously announced), and the several hundred basis point remainder relating to the decline in upcoming September numbers . . . (Credit Suisse First Boston, August 6, 2004.)
Gompers Decl. ¶ 45.
These statements undermine plaintiff's contention that all of the August 5 news pieces are related to the overstatement. Plaintiff's expert report offers similar analyst commentary. "The decline related to [the] overstatement is expected to be 1.5% for daily and 5% for Sundays. . . . The paper was already reporting daily and Sunday circulation declines of 2.5% and 3.5% respectively." Davenport Equity Research August 6, 2004. Hakala Decl. 105. Tellingly, although some analysts emphasized the overstatement-related portion of the announcement, no analyst appears to have conflated all the sources of decline as overstatement-related, as plaintiff now suggests. Plaintiff has therefore failed to prove "that it is more probable than not that it was "[an overstatement-related] negative statement, and not other unrelated negative statements, that caused a significant amount of the decline." Oscar 487 F.3d at 266.
Although plaintiff appears to rely primarily on the theory that all the announcements are related to the circulation overstatement, Dr. Hakala makes the additional argument in his report that the August 6, 2004 stock price drop cannot be attributed to the news of circulation decline resulting from the change in methodology because this information had been announced previously (on March 9, 2004, as explained in the August 5, 2004 announcement). He opines that the "change in methodology" information had already been incorporated by the market, and thus "factored into Belo's share price prior to August 6, 2004." Hakala Decl. 13. Dr. Hakala fails, however, to substantiate this claim.
First, Dr. Hakala admits that his analysis regarding corrective disclosures occurring before August 5 is incomplete. See supra note 5. To the extent that Dr. Hakala relies on pre-August 5 corrective disclosures as the basis for asserting that the news of the methodology change had already been "factored into Belo's share price prior to August 6, 2004," he fails to provide the analytical support for the assertion.
Dr. Hakala's event study, for example, does not analyze the March 9, 2004 news release or provide an empirical basis from which the court can find that the market in fact reacted to that announcement. See Hakala Decl. 68
Second, there is a material distinction between the March 9 and August 5 announcements: only the latter quantifies the amount of the decline. The March 9 announcement states, in relevant part: "circulation at [DMN] is expected to be down slightly, with Sunday circulation down a bit more than daily. The decrease is attributable to a change in our methodology for calculating State circulation from one acceptable method to another." Ds. App. 20. It is far from clear, based on this simple comparison reading, that the market had already fully incorporated the quantified change announced on August 5 through a March 9 change-in-methodology disclosure that did not quantify the expected decline. Plaintiff has therefore failed to prove that the stock decline on August 6 was not at least partly attributable to the information contained in the August 5 announcement that for the first time publicly quantified the magnitude of the decline resulting from the methodology change.
But even assuming that the market had in fact previously taken account of some of the decline in circulation due to the change in methodology, this would not be sufficient to persuade the court to find that plaintiff has satisfied its burden of proving loss causation. The dictates of Oscar are clear: in a context, like this one, in which there is more than one negative disclosure, the plaintiff must show "loss causation that targets the corrective disclosure[.]" Oscar, 487 F.3d at 262 (emphasis added). In this context, the plaintiff "must, in order to establish loss causation . . . offer some empirically-based showing that the corrective disclosure was more than just present at the scene." Id. at 271. Dr. Hakala's event study tends to establish that the market reacted to the bundle of August 5 news pieces with an August 6 stock price drop of 5.47%. See Hakala Decl. 74. Crucially, however, the study fails to target the corrective disclosure at issue. This showing is insufficient under Oscar. Oscar, 487 F.3d at 271 ("The plaintiffs' expert does detail event studies supporting a finding that [defendant's] stock reacted to the entire bundle of negative information contained in the 4Q01 announcement, but this reaction suggests only market efficiency, not loss causation, for there is no evidence linking the culpable disclosure to the stock-price movement.") (emphasis in original).
Although the Oscar panel appeared to agree that the plaintiff had overall presented more persuasive analyst commentary tending to establish that the stock price drop was significantly attributable to the fraud-related corrective disclosure, it nevertheless vacated the district court's decision to grant class certification as an abuse of discretion, concluding that the plaintiff had failed to provide an empirical showing that targeted the corrective disclosure. Oscar, 487 F.3d at 270-71.
Because plaintiff lacks the empirically-based showing that targets the corrective disclosure, as required by the Fifth Circuit in Oscar, the court holds that it has failed to establish loss causation by a preponderance of the evidence. And because plaintiff has failed to make this showing, its motion for class certification must be denied. See id. (holding that failure to prove loss causation results in failure to trigger the presumption of reliance provided by the fraud-on-the-market theory).
* * *
Lead plaintiff's September 17, 2007 motion for class certification is denied. In view of this disposition, defendants' December 21, 2007 motion to strike is denied as moot.
SO ORDERED.