MEMORANDUM AND ORDER

KEITH P. ELLISON, District Judge.

The Court returns to the critical issue of class certification. In its decision of December 6, 2013, the Court denied class certification because of Plaintiffs' failure to demonstrate that their damages can be measured on a classwide basis, consistent with their theories of liability. (Doc. No. 709.)1 In accordance with the Court's prior order, Plaintiffs have reformulated their proposed subclass periods around a fulcrum of April 20, 2010, the date of the Deepwater Horizon explosion. Plaintiffs have also provided thorough descriptions of damages methodologies they propose to apply to the pre-explosion and post-explosion subclasses. Plaintiffs seek class certification a second time.

The issue before the Court is two-pronged but narrow. The Court must determine whether Plaintiffs' proposed damages methodologies (1) quantify the injury caused by Defendants' alleged wrongful conduct, and (2) can be deployed on a classwide basis such that common issues will predominate over individualized ones. The first prong is necessitated by the reasoning articulated by the Supreme Court in Comcast Corporation v. Behrend, 133 S.Ct. 1426 (2013). The second prong is necessitated by Rule 23(b)(3) of the Federal Rules of Civil Procedure.

The Court has been substantially aided in this inquiry by the efforts of parties' counsel and experts. Having reviewed Plaintiffs' Renewed Motion for Class Certification (Doc. Nos. 739 & 739-1); Defendants' opposition (Doc. No. 783), Plaintiffs' reply brief (Doc. No. 802), Plaintiffs' supplemental brief (Doc. No. 836), Defendants' opposition to the supplemental brief (Doc. No. 842), all papers in support thereof, and having heard oral argument, the Court finds that Plaintiffs' Renewed Motion for Class Certification (Doc. No. 739) must be DENIED as to the Pre-Explosion "Process Safety" Subclass and GRANTED as to the Post-Explosion "Spill Severity" Subclass. The Court also GRANTS Plaintiffs' Motion for Leave to Amend the Complaint (Doc. No. 777). The Court's reasoning is explained in full below.

I. THE COURT'S DECEMBER 6th ORDER

On Plaintiffs' previous motion to certify the class, the Court agreed that all prerequisites of Rule 23(a) and Rule 23(b)(3) were met except for the predominance of classwide damages.2 (Doc. No. 709 (the "December 6th Order"), at 8-9, 14, 17, 19-20, 22, 32.) Plaintiffs had indicated that damages would be calculated on a classwide basis using an event study, but provided no specifics as to how the calculations would be performed. Defendants anticipated that Plaintiffs' expert would propose a "constant dollar" model of inflation and pointed out the ways in which such a model would be inconsistent with Plaintiffs' theories of liability.

The Court largely agreed with Defendants' criticisms and noted that a classwide damages model which did not hew to Plaintiffs' theory of liability would not satisfy Rule 23(b)(3)'s predominance requirement pursuant to Comcast Corporation v. Behrend, 133 S.Ct. 1426 (2013). Specifically, the Court highlighted the following problems with a "constant dollar" damages model:

(December 6th Order at 29 (citations and footnotes omitted).) The Court permitted Plaintiffs to re-urge their motion if they could comply with Comcast and Rule 23(b)(3). (Id. at 32.)

II. PLAINTIFFS' MODIFIED SUBCLASSES AND DAMAGES MODELS

In response to the December 6th Order, Plaintiffs modified their proposed subclasses and articulated differing damages methodologies for each. The subclasses are now defined as: (1) the "Pre-Explosion" or "Process Safety" Subclass consisting of persons or entities who purchased or otherwise acquired BP American Depositary Shares ("ADSs") between November 8, 2007 and April 20, 2010 and were injured thereby; and (2) the "Post-Explosion" or "Spill Severity" Subclass consisting of persons or entities who purchased or otherwise acquired BP ADSs between either April 26, 2010 or April 29, 20103 and May 28, 2010 and were injured thereby. (Doc. No. 739-1 ("Mot."), at 1.)

A. The Pre-Explosion "Process Safety" Subclass

Plaintiffs allege that Defendants repeatedly and falsely assured the market that process safety improvements recommended by the Baker Panel—most notably, a comprehensive Operating Management System ("OMS")—were being rolled out throughout the organization, across the globe. These statements began in 2007 and 2008 with expressions of intent, design, and early implementation. (Doc. No. 339 ("SAC"), at ¶¶ 321, 323, 325, 327, 329.) They evolved into representations regarding the scope and geography of implementation in 2009 and 2010. (Id. at ¶¶ 332, 334-35, 347, 351, 353, 357, 359.) Plaintiffs allege that these misrepresentations lulled the market into believing that BP was a safer company than it actually was. In other words, according to Plaintiffs, "BP's repeated statements concerning the implementation of OMS and the Baker Panel recommendations misled investors about BP's ability to prevent catastrophic process safety events and contain them if and when they occurred." (Mot. at 10.) Plaintiffs claim that Defendants' fraud was revealed when the Deepwater Horizon exploded and BP was subsequently unable to contain the oil spill. (Id. at 11.)

To model damages for this subclass of investors, Plaintiffs propose first to calculate the total investment losses caused by the fraudulent statements by measuring the decline in the stock price on days in which "corrective" information entered the marketplace—so-called "corrective disclosure days" or "corrective events." For the Baker Panel and OMS-related statements, which allegedly understated BP's exposure to catastrophic risk, Plaintiffs characterize days on which the risk materialized as corrective events. (Doc. No. 741-1 ("Coffman Report"), at ¶ 28.) Plaintiffs have alleged eight corrective events related to the process safety subclass misrepresentations. (Id. at ¶¶ 84-132.) The stock price declined $20.38 in response to these corrective events, after netting the eight days' decline against $6.87 of stock price increase which would not have occurred absent the fraud. (Id. at ¶ 17.) Plaintiffs then reduce this amount to account for the portion of the materialized risk which was known to the market—i.e., the risk voluntarily assumed by investors, which should not be compensated. The resulting, maximum damages per share are $18.53. (Id.)

Plaintiffs then propose to apportion the $18.53 of allegedly compensable losses across the fourteen alleged misrepresentations made in the Pre-Explosion Subclass period. Plaintiffs' expert, Chad Coffman, claims that an economic modeling tool known as "Bayesian Updating" can depict how investor expectations of risk change over time based on evolving information. (Coffman Report at ¶¶ 55-62.) Mr. Coffman states that it would be nearly impossible to employ Bayesian Updating to model the influence of each alleged misstatement in the Pre-Explosion Subclass period. (Id. ¶ 53.) Nonetheless, based on the theory of Bayesian Updating, Mr. Coffman claims that a straight-line linear progression of "inflation"—beginning with $0 of "inflation" as of the date of the first misleading statement relating to the Baker Panel recommendations or OMS (whether or not the statement is actionable) and ending with the full $18.53 of "inflation" on the last day of the subclass period—would be a conservative estimate of the incremental loss attributable to each alleged misrepresentation in this period. (Id. ¶¶ 63-67.) Each alleged misrepresentation would capture the "inflation" that had been generated in the time period since the immediately preceding misrepresentation. (Id. ¶¶ 70-73.) The parties refer to this as the "step up" or "step wise" function.

The following graph depicts Plaintiffs' proposed methodology for apportioning their losses across the Pre-Explosion Subclass period:

Exhibit 1

Mr. Coffman states that the above methodology will allow damages to be calculated on a classwide basis. The only individualized facts which will need to be taken into account are when a particular subclass member purchased and sold his, her, or its shares. (Coffman Report ¶ 159.)

It is important to clarify one confusing aspect of Plaintiffs' pre-explosion damages methodology. Plaintiffs and Mr. Coffman use the term "inflation" to refer to the apportionment of alleged losses among the various process safety misstatements. Inflation typically refers to the disparity between the price paid for the security and its value, at the time of the transaction—a disparity created by the alleged fraud. See, e.g., Hubbard v. BankAtlantic Bancorp, Inc., 688 F.3d 713, 725 (11th Cir. 2012) ("[I]n 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 fraudinduced inflation that was baked into the plaintiff's purchase price was subsequently removed from the stock's price, thereby causing losses to the plaintiff.") (quotation marks and citation omitted)). But the pre-explosion damages methodology is based on total losses following the explosion, not on the distortion allegedly created by Defendants' misstatements before anyone knew that BP would suffer a deepwater well blow-out. As described in Section IV(A)(2), everyone acknowledges that the pre-explosion distortion—if Mr. Coffman had attempted to model it—would return lower damages figures than those depicted above. (Doc. No. 831, at 41.) Because the pre-explosion damages methodology does not model inflation in the traditional sense, the Court will avoid using that term in connection with the pre-explosion damages methodology, or will employ quotation marks to clarify that the use of the term is non-traditional in the pre-explosion damages methodology.

B. The Post-Explosion "Spill Severity" Subclass

Following the explosion, two BP executives—Defendants Douglas Suttles and Anthony Hayward—provided public estimates of the ensuing oil spill. These estimates ranged from an initial estimate of 1,000 barrels of oil per day ("bopd") to later estimates of 5,000 bopd. (SAC ¶¶ 371, 373, 381.) According to Plaintiffs, these estimates were overly optimistic and contradicted both contemporaneous internal estimates (which reached as high as 14,266-92,000 bopd) and the real flow rate (determined after the fact to be approximately 60,000 bopd). (Id. ¶¶ 6, 375.) Because Defendants publicly lowballed the flow rate, the market price of BP ADSs did not reflect the magnitude of the disaster facing the company. In other words, Defendants' spill severity misrepresentations prevented the ADS stock price from falling to a level which accurately reflected the magnitude of the spill. (Mot. at 16-17.)

Plaintiffs propose to model damages for the Post-Explosion Subclass by calculating the inflation caused by the fraudulent statements. Plaintiffs equate inflation with the total decline in the stock price on corrective disclosure days. Plaintiffs have identified six corrective events related to the spill severity misrepresentations. (Coffman Report at ¶¶ 140-156.) The stock price declined $17.10 in response to these corrective events, after netting the six days' decline against $6.87 of stock price increase which would not have occurred absent the fraud. (Id. at ¶ 19.)

Unlike the Pre-Explosion Subclass, however, Plaintiffs do not propose to apportion the $17.10 of allegedly compensable losses across the three or four alleged misrepresentations made in the Post-Explosion Subclass period. Instead—due to the short duration of the subclass period and the consistency of the alleged misstatements—Plaintiffs propose to utilize a "constant dollar" methodology, in which the total inflation is carried back to the very first alleged misstatement. (Mot. at 22-23.) The inflation slowly dissipates over the time period of the alleged corrective disclosures, until inflation is $0 at the time of the last corrective disclosure. The following graph depicts the proposed methodology for carrying inflation back through the Post-Explosion Subclass period:

Exhibit 8a

Mr. Coffman states that the above methodology will allow damages to be calculated on a classwide basis. The only individualized facts which will need to be taken into account are when a particular subclass member purchased and sold his, her, or its shares. (Coffman Report ¶ 159.)

III. DEFENDANTS' OPPOSITION TO CLASS CERTIFICATION

Defendants lodge a number of criticisms of Plaintiffs' proposed damages methodologies. These include, briefly:

Defendants couch the above criticisms in the language of Comcast, accusing Plaintiffs of either continuing to seek recovery for theories of liability that are no longer part of the case, or of proposing damages methodologies which are inconsistent with their liability case.

As before, Defendants have raised compelling arguments regarding the overbreadth of Plaintiffs' claimed damages. But the Court need not be satisfied that Plaintiffs have identified the correct amount of damages before it can find that this action should proceed collectively. The Court need be satisfied only that Plaintiffs have proposed a viable, internally consistent, and truly classwide approach to calculating damages. With this in mind, the Court will consider each criticism lodged by Defendants in light of the context in which the case currently sits.

IV. CLASS CERTIFICATION ANALYSIS

A. Plaintiffs have not carried their burden as to the Pre-Explosion Subclass.

The heft of Defendants' arguments in opposition to class certification is directed at the weaknesses of Plaintiffs' pre-explosion damages methodology. While the Court agrees that many features of the pre-explosion damages methodology are problematic, they are not necessarily impediments to class certification. Defendants' final criticism, however, is fatal.

1. Plaintiffs' characterization of certain days as corrective events, even if erroneous, does not prevent class certification.

Defendants fault Plaintiffs' decision to treat certain stock price declines—specifically, those associated with the market's realization that BP could not contain the oil spill—as corrective events for the alleged process safety fraud. (Opp. at 25-32.) They note that no alleged misrepresentations regarding BP's oil spill response capabilities remain in the case. The alleged misstatements from the Initial Exploration Plan ("IEP") and the Regional Oil Spill Response Plan ("OSRP") were deemed not capable of classwide proof, given their lack of publicity prior to the oil spill. (Id. at 25.) Defendants note that Plaintiffs' overall damages calculations have not materially differed from their last attempt at class certification, indicating that the Court's decision not to certify the IEP and OSRP statements for class action treatment produced no corresponding change in the damages calculations. (Id. at 26-27.) Defendants argue that this is the same problem encountered in Comcast. (Id. at 32.)

Plaintiffs—correctly—reframe this argument as an argument regarding loss causation. (Doc. No. 802 ("Reply"), at 20.) As the Supreme Court explained in Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336 (2005), buying a stock at an inflated cost does not harm the purchaser.4 A compensable injury occurs only when the fraud is revealed and causes the stock price to drop, directly devaluing the stock which remains in the purchaser's possession. See Dura, 544 U.S. at 344 (noting that a securities fraud defendant's liability attaches when the facts which he misrepresented or concealed "`become generally known' and `as a result' share value `depreciates'") (quoting Restatement (Second) of Torts § 548A, comment b, at 107 (1976)). This element of a securities fraud claim is known as "loss causation."5

"Loss causation requires proof of a causal connection between a misstatement and a subsequent decline in a stock's price." Alaska Elec. Pension Fund v. Flowserve Corp., 572 F.3d 221, 229 (5th Cir. 2009); see also Dura, 544 U.S. at 342 (describing loss causation as a "causal connection between the material misrepresentation and the loss"). There are two components to this "causal connection." First, the decline in stock price must follow "a disclosure of negative truthful information that was related to the allegedly false, non-confirmatory positive statement made earlier." Flowserve, 572 F.3d at 229 (internal quotation marks, brackets, and citation omitted). For negative information to be "related to" a prior misstatement, it need not "specifically reveal[] the fraud" by correcting "`fact-for-fact'" the prior misstatement. See id. at 229-30. Instead, the "disclosed information must reflect part of the `relevant truth'—the truth obscured by the fraudulent statements." Id. at 230. Second, it must be "`more probable than not that . . . this negative statement, and not other unrelated negative statements . . . caused a significant amount of the decline.'" Id. at 228 (quoting Greenberg v. Crossroad Sys., 364 F.3d 657, 666 (5th Cir. 2004)).

Defendants dispute that corrective events which informed the market that BP could not contain the oil spill were "related to" the process safety misstatements. In reply, Plaintiffs explain that they will argue that stock price declines associated with BP's inability to contain the oil spill revealed the falsity of the OMS statements. According to Plaintiffs, Defendants' statements implicitly reassured the market that OMS would both reduce the likelihood that BP would suffer another process safety incident and give BP the tools to respond to and mitigate the effects of an incident, should one occur. (Reply at 20-22.) Defendants have lambasted this argument, characterizing it as a sudden departure from Plaintiffs' previous litigation position, which did not characterize the OMS statements as statements regarding BP's spill containment capabilities. (Opp. at 27-30.)

The Court agrees with Defendants that this is a "new" theory, and one not altogether convincing as a matter of logic.6 Despite the initial appeal of Defendants' position, however, the Court believes it to be premature. Loss causation need not be proven at the class certification stage. See Erica P. John Fund, Inc. v. Halliburton Co., 131 S.Ct. 2179, 2185-86 (2011). And Plaintiffs' alleged failure of proof—if indeed there is one—is a classwide failure amenable to a classwide solution: the declines from the unrelated corrective events are simply removed from the damages calculations. As Plaintiffs note, their proposed methodology tolerates such alterations in inputs. (Mot. at 19; Reply at 22-23.)

"[T]he focus of the 23(b)(3) class certification inquiry—predominance—is not whether the plaintiffs will fail or succeed, but whether they will fail or succeed together."7 Erica P. John Fund, Inc. v. Halliburton Co., 718 F.3d 423, 431 (5th Cir. 2013), cert. granted, Halliburton Co. v. Erica P. John Fund Inc., 134 S.Ct. 636 (2013). Consequently, the Court is not persuaded that Plaintiffs' decision to treat containment-related disclosures as corrective events for the process safety fraud prevents class certification.

2. Plaintiffs' attempt to apportion their losses throughout the subclass period is not so arbitrary as to prevent class certification.

Defendants also criticize the "linear progression" aspect of Plaintiffs' pre-explosion damages methodology. This aspect of the damages model is explicitly intended to address the Court's concern that the Baker Panel and OMS statements were alleged to have a "cumulative" and "compounding" misleading effect on the market, rendering a "constant dollar" methodology inappropriate.

First, Defendants claim that Mr. Coffman's linear progression methodology—i.e., his "step up" or "step wise" function—is completely arbitrary because it does not take into account the content or context of the various pre-explosion statements. (Opp. at 13.) This disconnect produces irrational results. For example, the principal determining factor of how much incremental loss is attributed to any given statement is how much time has passed since the immediately preceding statement. (Id. at 13-14.) Additionally, the amount of loss attributed to a particular misstatement is constant and independent of whether any preceding misstatement is proven. For example, if Plaintiffs establish that Mr. Hayward misspoke on March 23, 2010 when he stated that the Texas City accident "has changed in a profound and fundamental way [BP's] approach to safety and operations integrity," they are awarded $18.05 of "inflation" as of that date, even if it is the only misstatement proven. (Id. at 14-15.) Defendants argue that these illogical results illustrate how divorced the damages methodology is from Plaintiffs' liability case. (Id. at 15-16.) Notably, this belies Plaintiffs' contention that the model measures the "cumulative" effects of the alleged misstatements, because the effect of any given misrepresentation is independent of whether prior misrepresentations have been shown. (Id. at 18-21.)

Plaintiffs claim that the proposed methodology utilizes a proven and acceptable means of measuring damages: working backwards from the effect of a "corrective" event on the stock price. (Reply at 8-9; Doc. No. 831, at 41.) They also defend the consistency of the damages methodology with the liability case. The proposed methodology reflects that the "inflation" built up over the course of the subclass period, beginning with a small amount of "inflation" as the statements began, and ending with the full amount of "inflation" as of the time of the explosion. (Reply at 9.) They acknowledge that Mr. Coffman's methodology provides at most an approximation of the influence of Defendants' process safety misstatements, but argue that case law permits them to approximate damages when precise measurements are unavailable. (Id. at 10-13.)

Plaintiffs also note that their proposed methodology will be responsive to the jury's liability findings, because failure to prove later misrepresentations actionable renders the "inflation" from those statements unrecoverable. (Reply at 9 & n.10.) They acknowledge the accuracy of Defendants' hypothetical—where failure to prove any misrepresentation but a latein-time misrepresentation does not affect the "inflation" produced by that misrepresentation. However, Plaintiffs caution that the consistency of their damages methodology with the liability case should not be tested on highly unlikely scenarios, when (according to them) there are 87 million possible permutations of liability findings for the 14 alleged process safety misstatements. (Id. at 16-17 & n.25.) They note that, in the event that Defendants' unlikely hypothetical scenarios transpire, Defendants are free to argue that the damages methodology cannot be accepted in those circumstances. (Id. at 17 n.25.)

Finally, and most importantly, Plaintiffs contend that Defendants' criticisms are directed to whether the proposed damages methodology is correct, not whether it is consistent with the liability case. (Reply at 13-15.) They state that the accuracy of the damages calculations is an inquiry reserved for the jury. (Id. at 15.) They argue that Comcast does not authorize denial of class certification based on alleged weaknesses in a plaintiff's damages methodology, unless those weaknesses conceal difficult or complex individualized damages determinations. (Id. at 15-16.) Because Defendants' arguments—if well-taken—indicate a classwide failure of proof, and not that individual issues predominate, class certification is proper. (Id.)

As the Court perceives it, the fundamental issue undergirding the parties' respective arguments is whether Plaintiffs may "work backwards" from the stock price declines at the time of alleged corrective events and attempt to apportion their losses over the course of the Pre-Explosion Subclass period. Plaintiffs refer to the technique as "back-casting" and claim that it is commonly employed in securities fraud cases. (Doc. No. 831, at 41.)

If Plaintiffs are correct, and "back-casting" is an accepted means of approximating the losses caused by misleading statements, then the proposed methodology—while perhaps not entirely convincing—provides at least some logical and consistent basis for apportioning Plaintiffs' losses among the 14 alleged process safety misstatements. If the approach is not acceptable as a matter of law, however, then Defendants have correctly identified a problem similar to the one found in Comcast. As Justice Scalia explained:

133 S. Ct. at 1433 (citations omitted) (emphasis original).

As to the question of whether a Section 10(b) claimant may "back-cast" damages, there is ample support for Plaintiffs' position. Typical Section 10(b) damages are governed by the "outof-pocket" rule—i.e., a Section 10(b) claimant may recover "`the difference between the price paid and the "value" of the stock when bought.'" Acticon AG v. China North East Petro. Holdings Ltd., 692 F.3d 34, 38 (2d Cir. 2012) (quoting Elkind v. Liggett & Myers, Inc., 635 F.2d 156, 168 (2d Cir. 1980)). And multiple cases indicate that these "out of pocket" damages can be measured by reference to the decline in the stock price on the day of disclosure. See, e.g., In re Enron Corp. Sec. Derivative & "ERISA" Litig., 529 F.Supp.2d 644, 716 (S.D. Tex. 2006) (stating that the "out-of-pocket measure . . . allows a purchaser to recover the difference between the purchase price and the true value of the securities absent the alleged fraud as measured by the correction in the market price following curative disclosure") (quotation marks and citation omitted) (emphasis added); In re Royal Dutch/Shell Transp. Sec. Litig., 404 F.Supp.2d 605, 610 (D.N.J. 2005) (same). Thus, case law reflects a long-standing and widespread practice of measuring the stock price impact of a given misstatement by implication from the stock price decline caused by the misstatement's disclosure. See, e.g., Goldberg v. Household Bank, F.S.B., 890 F.2d 965, 966-67 (7th Cir. 1989) ("When markets are liquid and respond quickly to news, the [stock price] drop when the truth appears is a good measure of the value of the information, making it the appropriate measure of damages."); Harris v. Union Elec. Co., 787 F.2d 355, 368 (8th Cir. 1986) ("[W]e believe that the true value of the bonds on the date they were issued is reflected by the drop in the market price once the fraud was discovered.") (emphasis original); Blackie v. Barrack, 524 F.2d 891, 909 n.25 (9th Cir. 1975) (recognizing that "the change in [stock] price after a corrective release" is "circumstantial evidence of the inflation when purchased," although "it is not the exclusive method of measuring inflation").8

Admittedly, the authorities above are not directly on point, as Plaintiffs and Mr. Coffman concede that the pre-explosion damages methodology does not calculate the amount of preexplosion inflation in BP's stock price. (Reply at 4-6; Coffman Report ¶¶ 8, 72.) Specifically, when the corrective event is the materialization of an understated risk, the stock price movement on the date of correction (i.e., on the date that the risk materialized) will not equate to inflation on the date of purchase unless the probability of the risk materializing was 100 percent.9 If the probability is less than 100 percent, the stock price correction after the risk materializes will be larger than the pre-materialization inflation.10 Here, not even Plaintiffs argue that the risk of a deepwater well blow-out and oil spill was 100 percent. The Court will address this aspect of the proposed damages methodology below, in Section IV(A)(3), as it prompts a separate discussion of what losses are redressable by the securities laws. But with regards to the arbitrariness of Mr. Coffman's model, the Court accepts the general notion that Plaintiffs may "back-cast" their losses as derived from corrective disclosure days.

With this understanding in mind, the core of Defendants' argument appears to be that Mr. Coffman's "linear progression" and "step-wise function" are imperfect means of allocating Plaintiffs' losses over the relevant time period. But Plaintiffs' damages model need not be perfect. It need not be "correct." See Vaccarino v. Midland Nat'l Life Ins. Co., No. 2:11-CV-05858-CAS(MANx), 2014 WL 572365, at *10-13 (C.D. Cal. Feb. 3, 2014); In re Cox Enters. Inc. Set-Top Cable Television Box Antitrust Litig., No. 12-ML-2048-C, 2014 WL 104964, at *12-13 (W.D. Okl. Jan. 9, 2014). Indeed, it is impossible to imagine any damages methodology that would meet Defendants' exacting standards except for proof of stock price impact on each day that a process safety misstatement was made. But the fact that Plaintiffs' task is unquestionably difficult does not render it impossible. Approximating damages in any case is an imperfect science—particularly when an alleged fraud is perpetrated over a multi-year period, with a near-continuous series of alleged misstatements.11 The pre-explosion damages methodology contains its flaws, but it is not wholly arbitrary.

3. The fact that the pre-explosion damages methodology does not measure the classwide injury caused by Defendants' alleged fraud is fatal to class certification.

As previously alluded to, Plaintiffs concede that the pre-explosion damages methodology does not calculate the amount of pre-explosion inflation in BP's stock price. As a result, Defendants characterize the methodology as "investor insurance" rather than legally compensable securities fraud damages. Defendants argue that the proper type of damages to be recovered by investors in securities fraud cases is "out-of-pocket" damages. (Opp. at 36-37.) In the context of this case, Defendants claim that pre-explosion purchasers' damages should be determined by how the market would have priced the additional, undisclosed risk of a process safety incident before one occurred—not how the market in reality priced the consequences of the Deepwater Horizon explosion. (Id.; Doc. No. 783-7 ("Stulz Report"), at ¶¶ 30-31.) Because the pre-explosion damages methodology is not based on the "but for" price that the stock would have traded at absent Defendants' alleged fraud, it represents an impermissible windfall for investors, protecting them from the downside consequences of an assumed (if allegedly understated) risk. (Opp. at 38-39.)

Plaintiffs disagree that the proper measure of pre-explosion purchasers' damages is the difference between the actual price paid for the stock and a hypothetical, ex ante "but for" price at which the stock would have traded had the market been able to appropriately price BP's risk profile. They argue that the observed post-explosion stock price declines may be recovered as consequential damages from the alleged process safety fraud. (Doc. No. 836 ("Supp. Brief"), at 3.) They expressly eschew that their recovery should be limited to the market price distortion, because they intend to prove that their post-explosion losses were proximately caused by BP's failure to institute OMS as publicly represented. (Id. at 3-8.) Plaintiffs note that the Fifth Circuit allows them to pursue consequential damages. (Id. at 10.)

Plaintiffs are correct that no single measure of compensatory damages controls all situations redressed by Section 10(b). See Abell v. Potomac Ins. Co., 858 F.2d 1104, 1137 (5th Cir. 1988), vacated on other grounds sub. nom. Fryar v. Abell, 492 U.S. 914 (1989). Plaintiffs are also correct that the Fifth Circuit has approved consequential damages in the Section 10(b) context.12 See James v. Meinke, 778 F.2d 200, 205-06 (5th Cir. 1985); Meyers v. Moody, 693 F.2d 1196, 1212 (5th Cir. 1982). As the Fifth Circuit has explained, "special [i.e., consequential] damages are defined as outlays attributable to the defendant's wrongful conduct." Meyers, 693 F.2d at 1212 (emphasis added). This definition of consequential damages dovetails with Section 28 of the Securities Exchange Act of 1934 (the "Exchange Act"), which states that no plaintiff may recover more than the "actual damages to that person on account of the act complained of." 15 U.S.C. § 78bb(a)(1). These authorities counsel that a distinction must be drawn between the alleged misrepresentation and the subject matter allegedly misrepresented—in this case, between Defendants' public misstatements of their process safety reforms, and the underlying failure to institute process safety reforms. The Exchange Act provides compensation for losses caused by the former; losses caused by the latter are beyond the scope of the Act. See Santa Fe Indus., Inc. v. Green, 430 U.S. 462, 477 (1977) ("No doubt Congress meant to prohibit the full range of ingenious devices that might be used to manipulate securities prices. But we do not think it would have chosen this `term of art' [i.e., manipulative] if it had meant to bring within the scope of [Section] 10(b) instances of corporate mismanagement such as this[.]"); City of Sterling Heights Police & Fire Ret. Sys. v. Abbey Nat'l, PLC, 423 F.Supp.2d 348, 354-55 (S.D.N.Y. 2006) ("Allegations of corporate mismanagement generally fall outside the scope of [Section] 10(b) because they fail to allege conduct that is considered `manipulative or deceptive.'").

Plaintiffs argue that Defendants' process safety misstatements were a proximate cause of all their post-explosion investment losses because Plaintiffs were deprived of the opportunity to avoid the increased risk by divesting prior to the explosion. (Reply at 6 n.6; Coffman Report at ¶¶ 35-36.) But this articulation of the causal link between the alleged misstatements and the claimed losses injects individualized inquiries into what is supposed to be a classwide model of recovery. In other words, the causal link withstands scrutiny only if Defendants' misrepresentations induced a transaction—i.e., if a particular investor would not have purchased the security had he known the true state of BP's process safety programs.13 See Garnatz v. Stifel, Nicolaus & Co., Inc., 559 F.2d 1357, 1360-61 (8th Cir. 1977) ("[T]he gravamen of the present action was not whether [plaintiff] bought the bonds for a fair price, but that he bought at all . . . Under these circumstances, we believe that a rescissory damage measure . . . is appropriate."); Chasins v. Smith, Barney & Co., Inc., 438 F.2d 1167, 1173 (2d Cir. 1970) (approving the use of an alternative damages measure when "the evil is not the price at which [plaintiff] bought but the fact of being induced to buy and invest for some future growth in these stocks without disclosure of [defendant's] interest"). But if Defendants' misrepresentations merely manipulated a price— i.e., if an investor would still have purchased the security, but for less—then the causal chain between the misrepresentations and the alleged consequential losses is severed.

For this reason, Plaintiffs' articulation of consequential damages is antithetical to the "fraud-on-the-market" theory which enables the classwide resolution of their claims. The "fraudon-the-market" theory presumes that, in an "impersonal well-developed market for securities," investors rely upon the "integrity of the market price." Basic Inc. v. Levinson, 485 U.S. 224, 247 (1988). This presumption of reliance across a putative class enables claims to be resolved on a collective basis. See Regents of the Univ. of Cal. v. Credit Suisse First Boston (USA), Inc., 482 F.3d 372, 383, 394 (5th Cir. 2007). If investors are not relying upon the integrity of the market price—if they are, as Plaintiffs suggest, determining their own risk thresholds specific to the company at issue—then Plaintiffs' proposed measurement of damages cannot be deployed without an individualized inquiry into each investor's subjective motivations. Classwide treatment would be patently inappropriate in such a case.14 This is exactly the kind of elision of classwide and individualized questions that can and should be avoided by closely examining Plaintiffs' proposed damages methodology. See Comcast, 133 S. Ct. at 1434 (noting that "a methodology that identifies damages that are not the result of the wrong" can provide no assurance that "damages [resulting from the wrong] are capable of measurement and will not require labyrinthine individual calculations").

Finally, Plaintiffs argue that the alleged flaw in their methodology does not inhibit class certification, because the "out-of-pocket" measurement suggested by Defendants can also be accomplished on a classwide basis. (Reply at 5 n.5; Supp. Brief at 2.) Plaintiffs may be correct, but they have not given the Court any information about how the alternative calculation would be performed. The Court is left, as it was in December 2013, with a conclusory assertion that damages will be calculated on a classwide basis. Plaintiffs bear the burden of proving all relevant elements of Rule 23. That burden is not met by asking the Court simply to trust them. Because Plaintiffs have not shown that the damages of pre-explosion purchasers can be calculated on a classwide basis consistent with their theory of liability, the predominance requirement of Rule 23(b)(3) has not been met as to the Pre-Explosion "Process Safety" Subclass. The Court will not certify the Pre-Explosion Subclass for class action treatment.

B. The Court will certify the Post-Explosion "Spill Severity" Subclass.

Defendants offer fewer criticisms of the proposed damages methodology for the Post-Explosion Subclass. The Court finds none of the criticisms sufficient to deny class certification.

First, Defendants argue that the Court previously rejected a "constant dollar" approach to damages. (Opp. at 4, 41.) This overstates the December 6th Order. The Court acknowledged the incongruity of using "constant dollar" to measure damages for the proposed class as a whole, but the specific concerns identified by the Court were relevant to the pre-explosion time frame only. (December 6th Order at 13 n.6, 29.) Having now had occasion to review the "constant dollar" approach in the specific and isolated context of the spill severity misstatements, the Court perceives no legal or logical impediment to its use. As the Seventh Circuit explained, a Section 10(b) claim may be predicated on an understatement of bad news just as easily as an overstatement of good news. See Schleicher v. Wendt, 618 F.3d 679, 683 (7th Cir. 2010). The harm caused by that misrepresentation, then, is the artificial delay in the stock price falling, rather than the creation of artificial inflation:

Id. at 684. Plaintiffs' theory in the post-explosion time frame is that Defendants misrepresented their internal estimates of the oil spill; that the stock market price failed to fall to the level reflecting the magnitude of the crisis facing BP; that the market learned the truth; and that the stock market price corrected. (Mot. at 16-17.) There is nothing problematic with the "constant dollar" approach in these circumstances.

Second, Defendants argue that it is improper for the Post-Explosion Subclass period to begin on April 26, 2010 when the first alleged post-explosion misstatement in Plaintiffs' live pleading occurred after the markets closed on April 28, 2010. (Opp. at 42-43.) This dispute will be resolved in accordance with the Court's ruling on Plaintiffs' Motion for Leave to Amend, addressed in Section V, below.

Third, Defendants criticize the decision to carry back the full measure of the stock price decline in the months after the explosion to the first alleged spill severity misstatement because there is no evidence that Defendants had the same information about flow rate in the days after the explosion that the market acquired by mid-June 2010.15 (Opp. at 46-48.) They also suggest that Plaintiffs are using their damages methodology—based on what was allegedly "foreseeable" to Defendants as a result of their internal estimates of the flow rates—in order to expand their liability theories beyond what has been pled. (Id.)

Fourth, Defendants note that all of the alleged corrective disclosures for the postexplosion spill severity misrepresentations are also corrective disclosures for the pre-explosion process safety misrepresentations. They fault Plaintiffs for "recycling" corrective disclosures without disaggregating inflation by type of misrepresentation "corrected" or risk materialized. (Opp. at 48-49.) Mr. Coffman notes that the overlap will not lead to "double counting of damages" because "a single physical share can never collect for the same disclosure more than once." (Coffman Report at ¶ 22.) Plaintiffs argue that it is consistent with both case law and logic for a single corrective disclosure to be linked to two separate but related fraudulent schemes. (Reply at 31-32.)

Defendants' third and fourth arguments concern loss causation, in that they challenge the "fit" between an alleged corrective event and an alleged fraudulent statement.16 The Court shares Defendants' concerns regarding the apparent disconnect between some corrective events and the fraud which they are alleged to have corrected. Most notably, it is difficult to imagine how BP's cancellation of its June 2010 dividend in response to intense political pressure "corrected" Defendants' flow rate statements.17 Nonetheless, as explained above in Section IV(A)(1), failure to prove loss causation is not, at present, an impediment to class certification. The Court reiterates its understanding that Plaintiffs' task at the class certification stage is to present a legally viable, internally consistent, and truly classwide approach to calculating damages. Whether Plaintiffs have properly executed under the approach is a question for a different day.

The damages methodology proposed for the Post-Explosion Subclass meets the requirements for Rule 23(b)(3) predominance: it attempts to quantify the injury caused by Defendants' alleged wrongful conduct, and it can be deployed on a classwide basis. The Court will certify the Post-Explosion Subclass for class action treatment.

V. PLAINTIFFS' MOTION FOR LEAVE TO AMEND

A. Description of Proposed Amendments

1. New spill severity misrepresentation

The most important and consequential amendment in Plaintiffs' proposed Third Consolidated Amended Class Action Complaint ("TAC") is the addition of a new spill severity misrepresentation on April 24, 2010. Plaintiffs acknowledge that they seek to add this misrepresentation to the case because otherwise investors who purchased BP ADSs between April 26-28, 2010 would have no remaining claims and would not be eligible to participate in the class action. (Doc. No. 777-1 ("MLA"), at 2.)

a. The SAC allegations

The Second Consolidated Amended Class Action Complaint ("SAC"), filed on April 2, 2012, alleged that an internal BP document dated April 27, 2010 set forth three estimates for the spill rate: a low estimate of 1,063 barrels of oil per day ("bopd"), a best estimate of 5,758 bopd, and a high estimate of 14,266 bopd. (SAC ¶ 375.) This document was the basis of Plaintiffs' scienter allegations for the post-explosion spill severity statements included in the SAC. There were three such statements:

b. The TAC allegations

Plaintiffs seek leave to file the TAC, which would add a fourth spill severity misstatement made by Mr. Suttles on April 24th. The statement was made during a joint press conference with the U.S. Coast Guard. Mr. Suttles stated that BP had detected ongoing releases of oil from the Macondo well at a rate of approximately 1,000 bopd. (Doc. No. 777-2 ("TAC"), at ¶ 370.) Mr. Suttles also did not correct the U.S. Coast Guard leader, Rear Admiral Landry, when she said that the U.S. government also estimated the flow rate at 1,000 bopd. (Id.) Rear Admiral Landry had asked Mr. Suttles prior to the press conference if he could support the 1,000 bopd flow rate estimate. (Id.)

Plaintiffs support Mr. Suttles's scienter for the April 24th statement with the following new allegations:

2. Other amendments

The TAC contains other amendments that are intended to reconcile the complaint with the Court's prior orders on motion to dismiss and class certification. (E.g., TAC at 1-2.) These amendments do not appear to be subject to any controversy.

B. Legal Standards

1. Rule 16(b)(4): whether to allow modification of the scheduling order

When the deadline for seeking leave to amend pleadings has expired, a court considering a motion to amend must first determine whether to modify the scheduling order under the Rule 16(b)(4) good cause standard. See S & W Enters., L.L.C. v. SouthTrust Bank of Ala., N.A., 315 F.3d 533, 536 (5th Cir. 2003). A court assesses four factors in deciding whether to grant an untimely motion for leave to amend under Rule 16(b)(4): "(1) the explanation for the failure to timely move for leave to amend; (2) the importance of the amendment; (3) potential prejudice in allowing the amendment; and (4) the availability of a continuance to cure such prejudice." Id. at 536 (internal quotation marks, brackets, and citation omitted). If the movant satisfies the requirements of Rule 16(b)(4), the court must next determine whether to grant leave to amend under the more liberal standard of Rule 15(a)(2). Id.

2. Rule 15(a)(2): whether to allow amendment of the complaint

A party may amend its pleadings once as a matter of course within certain deadlines. FED. R. CIV. P. 15(a)(1). Thereafter, a party may amend its pleadings "only with the opposing party's written consent or the court's leave." FED. R. CIV. P. 15(a)(2). "Rule 15(a) declares that leave to amend `shall be freely given when justice so requires'; this mandate is to be heeded." Foman v. Davis, 371 U.S. 178, 182 (1962). Accordingly, district courts in the Fifth Circuit "must entertain a presumption in favor of granting parties leave to amend." Mayeaux v. Louisiana Health Service and Indem. Co., 376 F.3d 420, 425 (5th Cir. 2004).

"Leave to amend, however, is by no means automatic." Little v. Liquid Air Corp., 952 F.2d 841, 845-46 (5th Cir. 1992), rev'd on other grounds, 37 F.3d 1069 (5th Cir. 1994). Factors for the court to consider in determining whether there is a substantial reason to deny a motion for leave to amend include "undue delay, bad faith or dilatory motive on the part of the movant, repeated failure to cure deficiencies by amendments previously allowed, undue prejudice to the opposing party, and futility of amendment." Wimm v. Jack Eckerd Corp., 3 F.3d 137, 139 (5th Cir. 1993).

C. Motion for Leave to Amend Analysis

1. The Court will apply Rule 16(b)(4).

Defendants suggest that Rule 16(b)(4) controls because the deadline to amend the pleadings has passed. Unfortunately, no docket control order entered in this case has ever included a deadline to amend pleadings. (Doc. Nos. 582 & 730.) Defendants argue, in the absence of an amendment deadline in the docket control order, the deadline to amend pleadings was the deadline for Plaintiffs to file the SAC, or April 2, 2012.19 (MLA Opp. at 8-9.) The Court disagrees that this date was intended to function as a deadline for all amendment. The Private Securities Litigation Reform Act ("PSLRA") did not allow Plaintiffs any case-specific discovery prior to the Court's rulings on motion to dismiss, the last of which issued in February 2013. (Doc. No. 536.) To accept Defendants' characterization of the April 2, 2012 deadline would be to lock Plaintiffs into a set of allegations based on pre-discovery knowledge.

Nonetheless, this case has progressed far enough that the Court will consider Plaintiffs' request untimely under the schedule. See, e.g., Smith v. Honeywell Int'l, Inc., No. 10-CV-03345-ES-JAD, 2014 WL 301031, at *5 (D.N.J. Jan. 27, 2014) ("Where a scheduling order sets no amendment deadline, courts have looked to when discovery closes to determine whether the motion to amend is untimely under Rule 16."). Fact discovery officially closed on January 31, 2014, and expert discovery is due to close within the next two months. (Doc. No. 730, at 1.) By all reasonable metrics, the time for amendment has passed.

2. Modification of the scheduling order is justified under Rule 16(b)(4).

Under the Rule 16(b)(4) good cause standard, the Court will consider: "(1) the explanation for the failure to timely move for leave to amend; (2) the importance of the amendment; (3) potential prejudice in allowing the amendment; and (4) the availability of a continuance to cure such prejudice." S & W Enters., 315 F.3d at 536 (internal quotation marks, brackets, and citation omitted). These factors, in the aggregate, weigh in favor of modification.

The importance of the amendment cannot be overstated. Without it, investors who purchased on three days in the post-explosion time frame will be excluded from the class action. (MLA at 2.) Additionally, $5.26 of stock price decline which Plaintiffs allege was related to the spill severity fraud would be unrecoverable. (MLA Opp. at 6.)

The amendment also threatens little prejudice to the Defendants—prejudice which can be cured, if necessary, by further modifications to the scheduling order. The Court recently extended expert discovery, which is under way and now scheduled for completion on or around June 19, 2014. To the extent that Defendants require any further extension to the expert discovery deadline on account of the new spill severity misstatement, the Court is amenable to such requests. The remainder of the schedule—which was vacated on May 7, 2014 to accommodate the extension in expert discovery—will be reset at the earliest possible date, with the input of the parties.

Defendants also suggest that they will be prejudiced because amendment is equivalent to providing Plaintiffs a "third bite" at the class certification apple. (MLA Opp. at 15-16.) If the requested amendment were the difference between certifying and not certifying the Post-Explosion Subclass, this argument would carry much more force. But permitting Plaintiffs to file the TAC affects only the dates of the Post-Explosion Subclass period. It does not change whether certification is appropriate.20

Thus, the only factor that could counsel against modification of the scheduling order is undue delay in requesting the amendment. The Court shares Defendants' consternation at the exceptionally late date of Plaintiffs' request. But Plaintiffs have given adequate explanation of why the new spill severity misstatement was not included in their prior complaints.21 And while Plaintiffs appear to have been in possession of the information which underlies their new allegations since at least early 2012, the Court cannot condemn Plaintiffs' discretion not to amend their pleadings as discovery made available new allegations that would not—at the time—have appreciably changed the landscape of the litigation. The present request appears to have been prompted by the Court's guidance in the December 6th Order that the subclass periods should be organized around the April 20, 2010 explosion. (Doc. No. 807 ("MLA Reply"), at 6.) In that light, and against the context of the case as a whole, the request is not unduly delayed.

3. The Court grants leave to amend pursuant to Rule 15(a)(2).

The Court has no trouble determining that amendment should be permitted under the liberal standard of Rule 15(a)(2). As described above, neither Plaintiffs' delay in asking for amendment nor the prejudice which will be visited upon Defendants are substantial enough to refuse leave. Nor is the proposed amendment clearly futile due to insufficient scienter allegations.22 See 6 C. Wright & A. Miller, Federal Practice & Procedure § 1487, pp. 637, 642 (2d ed. 1990) ("If a proposed amendment is not clearly futile, then denial of leave to amend is improper."). Leave to amend will be granted.

VI. CONCLUSION

Plaintiffs' Motion for Leave to Amend (Doc. No. 777) is GRANTED. Plaintiffs may file the Third Consolidated Amended Class Action Complaint (Doc. No. 777-2).

Plaintiffs' Renewed Motion for Class Certification (Doc. No. 739) is DENIED as to the proposed Pre-Explosion "Process Safety" Subclass. Plaintiffs have failed to establish that damages from Defendants' alleged misstatements in the Pre-Explosion Subclass period can be calculated on a classwide basis.

Plaintiffs' Renewed Motion for Class Certification (Doc. No. 739) is GRANTED as to the proposed Post-Explosion "Spill Severity" Subclass. Pursuant to Plaintiffs' Motion, and the Court's December 6, 2013 Order, all requirements of Rule 23(a) and Rule 23(b)(3) have been met as to the proposed Post-Explosion Subclass. The Court CERTIFIES the following plaintiff class pursuant to Federal Rule of Civil Procedure 23:

Thomas P. DiNapoli, Comptroller of the State of New York, as Administrative Head of the New York State and Local Retirement Systems and sole Trustee of the New York State Common Retirement Fund ("New York") and the Ohio Public Employees' Retirement System ("Ohio") are APPOINTED as Class Representatives for the Class. Cohen Milstein Sellers & Toll PLLC and Berman DeValerio are APPOINTED as Class Counsel.

IT IS SO ORDERED.