Opinion
12 Civ. 6682 (PAE)
01-21-2015
STREAM SICAV and TIEN CHUNG, individually and on behalf of all others similarly situated, Plaintiffs, v. JAMES JUN WANG, SMARTHEAT, INC., and JOHN DOES 1-4, Defendants.
OPINION & ORDER
:
On March 17, 2014, the Court issued an order denying, without prejudice, plaintiffs' motion for class certification. See Dkt. 114. The Court did so for several reasons, including because plaintiffs had not yet articulated a credible theory under which injury could be proven on a classwide basis, because fact discovery remained ongoing, and because the Supreme Court's then-forthcoming decision in Halliburton Co. v. Erica P. John Fund, No. 13-317, had the potential to alter the legal analysis governing class certification in putative securities class actions premised on a fraud-on-the-market theory.
With fact discovery now complete, and with the Halliburton decision handed down, plaintiffs have moved again for class certification. See Dkt. 144. The Court has carefully reviewed that motion and the materials submitted in support of it, see Dkt. 145 ("Pl. Br."), 46 ("Pl. Decl"), 156-58, as well as defendants' submissions in opposition, see Dkt. 149 ("Def. Br."), 150 ("Def. Decl.").
For the reasons that follow, the Court denies plaintiffs' motion for class certification, on the grounds that plaintiffs' submission is insufficiently rigorous, and fails to satisfy multiple requirements for class certification set by Federal Rule of Civil Procedure 23, including those of typicality, adequacy, and predominance. However, this ruling is without prejudice. The Court will grant plaintiffs a final opportunity, on the schedule set forth below, to establish that class certification is merited.
By way of background, plaintiffs claim that SmartHeat, Inc. violated federal securities laws by secretly amending a written lock-up agreement under which certain company insiders were barred from selling shares of SmartHeat stock until the start of 2012, and which SmartHeat had publicly touted as a sign of its executives' confidence in the company. Plaintiffs seek to certify a class of "[a]ll persons who purchased or acquired shares of SmartHeat, Inc. stock between February 24, 2010 and May 30, 2012." Dkt. 144, at 1.
The proposed class excludes defendants, other SmartHeat officers and directors, and their families and affiliates.
As explained in its March 17, 2014 order, the Court understood plaintiffs, in seeking class certification, to articulate two distinct theories as to how the undisclosed amendment of the lock-up agreement that enabled insiders to sell their SmartHeat shares caused classwide injury. See Dkt. 114. The first theory was that SmartHeat's public statements that certain insiders had committed to hold their shares had boosted the value of the shares; that purchasers of shares following the announcement of the lock-up agreement had therefore paid a premium to buy the shares; and that these purchasers had suffered losses when it was revealed that the agreement had been amended to permit insiders to sell their shares. The Court referred to this as the "revision theory." As the Court noted, such an approach to proving classwide injury was in line with the manner in which classwide injury is commonly proven. See id.
But as the Court's order also noted, to be sustainable, this theory required that there have been a legally cognizable "corrective disclosure"—i.e., an announcement or other revelation of the fact that insiders had been freed to sell their shares—that caused a drop in the trading price of SmartHeat's stock. There did not appear to have been any such announcement. Rather, the share price had gradually dropped to below $1 per share, to the point at which it had been delisted and trading had ceased, with no such revelation of the alleged amendment to the lock-up agreement having ever been made. The Court therefore held that, before a class could be certified on this theory, plaintiffs were required to identify some form of corrective disclosure, for otherwise SmartHeat's failure to disclose the amendment of the lock-up agreement could not be said to have caused classwide injury. See id.
In their renewed motion for class certification, plaintiffs have abandoned the "revision theory" of liability. Presumably, this is because, as the Court anticipated in its earlier order, the historical record simply does not reveal any disclosure (by SmartHeat or otherwise) while the stock was still trading of the abandonment of the company's lock-up agreement, let alone that the company's stock price dropped following such a disclosure.
Instead, plaintiffs now pursue solely the second theory of classwide injury identified in the Court's March 17, 2014 order. See Pl. Br. 2-3. This theory is that the once-prohibited insider sales themselves had driven down the stock's trading price. In other words, plaintiffs contend, the sale of these SmartHeat shares, and the ensuing availability on the market of an unexpectedly large number of freely tradable SmartHeat shares, drove down the stock's price during the 27-month class period, so as to injure persons who had previously purchased the stock. The Court referred to this as the "insider sale" theory of liability. Dkt. 114.
Plaintiffs' "insider sale" theory is an unusual theory of classwide injury. In a typical securities fraud class action, plaintiffs allege that the market reacted negatively on a given day or several-day period to a corrective disclosure, causing a decline in stock value that simultaneously harmed all shareholder class members in the same way. See, e.g., In re Bank of Am. Corp. Sec., Derivative, & Employee Ret. Income Sec. Act (ERISA) Litig., 281 F.R.D. 134, 143 (S.D.N.Y. 2012); Wagner v. Barrick Gold Corp., 251 F.R.D. 112, 119 (S.D.N.Y. 2008). Here, however, plaintiffs do not allege that their injury resulted from a one-time stock-price drop that affected all class members. Instead, they make an argument based on the mechanics by which trading of SmartHeat shares occurred on the NASDAQ market between February 2010 and May 2012, Specifically, plaintiffs argue that, as a consequence of unlocking insider shares, all sales of SmartHeat stock during that 27-month period were made at a price below that which would have been set had those shares remained locked up. Although the precise cause of this long period of alleged price diminution is not clear from plaintiffs' papers, plaintiffs' expert, Dr. Steven P. Feinstein, appears to attribute it to factors including the increased number of SmartHeat shares available for purchase and sale, the percentage increase that the shares unlocked for sale represented over all other shares, and/or the timing at which previously locked-up shares were sold in the market. See Pl. Decl., Ex. 1 ("Feinstein Rep."), at 29-34.
In claiming classwide injury on these grounds, plaintiffs face daunting precedent. Historically, claims of injury due not to corrective disclosures but rather to the mechanics by which shares of stock were priced during a protracted period of open-market trading have almost always been held ill-suited to classwide resolution. Instead, as the courts that have considered such claims have recognized, trade-specific inquiries are almost always needed to ascertain whether the price a customer paid on a particular trade was, or was not, compromised by a defendant's allegedly unlawful conduct. As these cases have noted, a defendant may be liable as to some trades but not others, and a judgment of liability (or non-liability) on any particular trade may not bear on (let alone determine) whether there was liability on another. Accordingly, because of the need for a trade-by-trade inquiry into whether or not there was persistent price inflation (or deflation), Rule 23(b)(3)'s requirement that common issues predominate over individualized ones is not met. See, e.g., In re Initial Pub. Offerings Sec. Litig. ("IPO"), 471 F.3d 24, 44 (2d Cir. 2006) (finding class certification unwarranted based on claims that customers suffered price injury based on issuers' practices in allocating IPO shares and trading such shares in the IPO "aftermarket"); Newton v. Merrill Lynch, 259 F.3d 154, 187 (3d Cir. 2001) (upholding denial of motion for class certification based on claim that buyers and sellers of shares on NASDAQ were denied best execution by broker-dealer's policy of executing trades only at the National Best Bid and Offer price); Pearce v. UBS Paine Webber, Inc., No. 02 Civ. 2409-17 (GRA), 2004 WL 5282962, at *10-11 (D.S.C. Aug. 13, 2004) (denying motion for class certification based on claim that broker-dealer added two cents per share to best execution price charged to customers for NASDAQ stocks); Grandon v. Merrill Lynch, No. 95 Civ. 10742 (SWK), 2003 WL 22118979, at *9 (S.D.N.Y. Sept. 11, 2003) (denying motion for class certification based on claim that broker-dealer marked up municipal bond prices by 4.6%).
For a number of reasons, the Court holds, a class similarly cannot be certified here on plaintiffs' insider sale theory, based on the present record:
1. Inadequate explication of theory of consistent price-decrease during class period: Plaintiffs' expert, Dr. Feinstein, opines that it is theoretically possible for the infusion of a large number of shares to affect a stock's price over a long period of time. See Feinstein Rep. 29-34. Perhaps so: Although the Court is quite skeptical that the existence of a discrete number of even large sell orders and/or the availability for purchase and sale on a stock market of larger number of shares would tend to deflate a stock's trading price over a long period of time so as to harm all shareholders, it is not the Court's place on a class-certification motion to find against plaintiffs on this theoretical point. And the Court does not rule out that a colorable showing to this effect could be made, based on a close analysis of trading patterns of a particular security during a particular time period.
The issue here does not involve the issuance of new shares, which would predictably reduce the per-share value of outstanding shares. As alleged, the amendment of the lock-up agreement did not increase the number of outstanding SmartHeat shares, and thus did not affect the portion of SmartHeat's market capitalization represented by a particular share. It affected only the number of shares available for purchase and sale on NASDAQ.
However, Dr. Feinstein does not undertake a close analysis, or indeed any analysis, of the trading activity at issue in this case to determine if a credible claim can be made, based on actual evidence, that the trading price of SmartHeat shares on NASDAQ over the 27-month class period consistently was reduced by the fact that previously locked-up shares were available for purchase and sale. Instead, his brief discussion of the point is limited to citing research literature supporting the proposition that "sustained large-scale selling generally has a persistent negative impact on the subject stock's price." Id. at 31. But there is no showing in his report that there was "sustained large-scale selling" of the security here throughout the 27-month class period sufficient to create a persistent negative impact on stock price. There is no attempt, for example, to identify the number of "large trading orders" or "metaorders" of SmartHeat stock. Id. at 32 (citation omitted). Nor is there analysis as to the impact of the release into the market on particular dates of particular quantities of previously locked-up shares. There is also no analysis as to the factors that affected how the market came to price SmartHeat stock on the hundreds of individual days comprising the class period. And there is no explanation as to why the volume of shares available for sale (as opposed to a stream of negative news about the company) accounted for the long period of decline in SmartHeat's share price during the class period. Instead, plaintiffs' claim of a 27-month-long artificial depression in SmartHeat share price is cursory and wholly theoretical. Absent a concrete presentation of the evidence on which such a finding could be made in this case, the Court cannot find that price injury to all class members can be established by common proof. As in IPO, Newton, Pearce, and Grandon, the absence of common proof of the essential elements of economic injury and loss causation compels a finding that the common issues do not predominate over individual ones. See Fed. R. Civ. P. 23(b)(3).
2. Adequacy issues with respect to lead plaintiff: Based on plaintiffs' renewed submission for class certification, there is only one class representative on whom the class case turns: Tien C. Chung. The Court has substantial questions about Mr. Chung's adequacy as a class representative. See Fed. R. Civ. P. 23(a)(4). His sworn certification, dated June 21, 2012, represents that he purchased SmartHeat stock on one occasion: He attests that he bought 3,305 shares of that stock on April 1, 2010, at a price of $11.99 per share. See Dkt. 108-2. However, based on an attachment to the report of plaintiffs' expert, Dr. Feinstein, which tracks SmartHeat's trading history, that appears to be an impossibility: On April 1, 2010, SmartHeat's closing price was $101.60, and the stock's price did not drop to the vicinity of $12 per share until more than 14 months later. See Feinstein Rep., Ex. 4, at 57, 66. It therefore appears either that Mr. Chung purchased a different security or that his certification, sworn to under penalty of perjury, was inaccurate. The Court also has not been supplied with any evidence bearing on Mr. Chung's qualifications to lead the class. Beyond his June 21, 2012 certification, the only declaration he has submitted merely recites the procedural history of the case and his duties as purported class representative. See Pl. Decl., Ex. 3.
The share prices in the Feinstein Report have been adjusted to reflect a 10-for-1 reverse stock split effective February 6, 2012. Feinstein Rep. Ex. 4, at 73. However, even taking that into account, the share price between March 31, 2010 and April 5, 2010—the trading days before and after April 1, 2012—was between $10.74 and $9.93, well below the price that Mr. Chung claims to have paid. Id. at 57.
3. Other predominance and typicality issues: Even assuming that Mr. Chung purchased SmartHeat stock on April 1, 2010, his claims as class representative would have to be "typical of the claims . . . of the class," Fed. R. Civ. P. 23(a)(3), such that a finding of liability as to him could fairly bind defendants as to the class's claims. Cf. AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1751 (2011) ("For a class-action money judgment to bind absentees in litigation, class representatives must at all times adequately represent absent class members."). And issues common to the class would have to predominate over issues specific to individual stock purchases (e.g., Mr. Chung's or those by other class members). For several reasons, Dr. Feinstein's expert report, on which plaintiffs' class-certification motion relies, does not provide the Court with a concrete, non-speculative basis on which to reach the conclusion that a determination of SmartHeat's liability to Mr. Chung would fairly resolve defendants' liability to the broadly-defined class.
a. There is no explanation in that report as to how or on what concrete basis a factfinder could determine that the price at which Mr. Chung purchased his shares on April 1, 2010 was inflated above what it would have been had the insider shares remained locked up, as plaintiffs appear to claim. To the contrary, the gist of Dr. Feinstein's report is that the price of SmartHeat stock was lower during the class period because the previously locked-up shares were in fact for sale. And plaintiffs allege that defendant Wang covertly sold a total of 380,000 shares beginning on February 24, 2010. Dkt. 52 ("SAC") ¶ 7. If anything, then, a logical inference from Dr. Feinstein's report would appear to be that there was an unexpectedly high volume of selling activity due to insider sales between February 24 and April 1, 2010, which presumably drove the stock's trading price down by the time of Mr. Chung's purchase.
b. Nor is there any explanation in that report as to how or on what concrete basis a factfinder could determine that the decline in SmartHeat's share price after April 1, 2010 was due, in material part, to sales of previously locked-up shares or to the higher volume of tradable shares as a result of the fact that once-locked-up shares were now tradable. There is no attempt made to distinguish the portion, if any, of the stock's decline during the class period attributable to insider sales from the decline caused by other company-specific factors.
c. Most important, there is no explanation as to how, concretely, the evidence pertaining to Mr. Chung's April 1, 2010 purchase could support the conclusion that similar stock price inflation and deflation affected the holdings of other purchasers of SmartHeat stock on all other stock-purchase dates during the 27-month-long class period. The class period commenced five weeks before Mr. Chung's purchase and lasted for 26 months afterward. On February 24, 2010, the stock closed at $123.70; on April 1, 2010, the stock closed at $101.60; and on May 30, 2012, the stock closed at $4.04. See Feinstein Rep., Ex. 4, at 57, 73. Dr. Feinstein's report theorizes that it is possible that the release of previously locked-up shares into a trading market could have a long-term price impact. But there is no attempt made to show that common proof in this case could establish a price injury to Mr. Chung that would serve as a fair proxy to prove the fact of such injury to purchasers of SmartHeat stock on each day during the lengthy class period.
4. Logistical inconsistencies within plaintiffs' pleadings: Even assuming that there were a showing that common proof could establish both price inflation on the date of Mr. Chung's purchase and all dates (before and after) when other class members bought SmartHeat stock—i.e., that defendants' liability to the entire class could be extrapolated from defendants' liability to Mr. Chung—plaintiffs' definition of the class appears inconsistent with aspects of plaintiffs' pleadings. Plaintiffs allege that the lock-up agreement, as announced, was to last until the start of 2012. SAC ¶ 13. If so, there is no apparent basis for including, within the class, persons who purchased SmartHeat stock between January 2012 and May 30, 2012, during which period a purchaser would have been on notice that the previously locked-up shares were freely tradable. And there is no apparent basis for including, within the class, persons who continued to hold SmartHeat stock after January 2012, when a shareholder undisputedly would be on notice that the stock price could be affected by a higher volume of tradable shares. Notably, this latter category includes Mr. Chung.
5. Lack of explanation as to how damages would be proven: Even assuming that all elements of liability (including injury) were established as to Mr. Chung and that his claims were representative such that resolution of defendants' liability as to him would resolve defendants' liability as to the remainder of the class, plaintiffs have not shown or explained, concretely, how damages would be calculated as to any individual purchaser. Individualized damages inquiries are not necessarily inconsistent with class certification. See Wallace v. IntraLinks, 302 F.R.D. 310, 318 (S.D.N.Y. 2014). However, to evaluate the Rule 23 elements of predominance and superiority, see Fed. R. Civ. P. 23(b)(3), the Court must understand, concretely, how plaintiffs propose to reliably establish damages. See Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1433 (2013) (Absent a sound model "establishing that damages are capable of measurement on a classwide basis," plaintiffs "cannot show Rule 23(b)(3) predominance: Questions of individual damage calculations will inevitably overwhelm questions common to the class."); In re U.S. Foodserv. Inc. Pricing Litig., 729 F.3d 108, 123 n.8 (2d Cir. 2013) ("[C]ourts should examine the proposed damages methodology at the certification stage to ensure that it is consistent with the classwide theory of liability and capable of measurement on a classwide basis.").
6. Lack of citations to admissible evidence: The period for fact discovery closed on October 27, 2014, see Dkt. 140, but the authority cited for virtually all factual representations in plaintiffs' class certification papers are the allegations made in plaintiffs' class action complaint. See, e.g., Pl. Br. 4-5 (citing only the SAC for the propositions that, inter alia, SmartHeat officers entered a lock-up agreement in January 2009, secretly amended the agreement in February 2010, and sold 380,000 shares between February and August 2010). And defendants dispute many of plaintiffs' factual claims (e.g., whether the lock-up agreement was secretly amended, when the insider sales occurred, etc.). Although the Court's role at the class-certification stage is not to resolve factual disputes, the Court must ensure that there is a basis in admissible evidence for each factual representation made in support of class certification to assure that a class is not certified based on conjecture as opposed to provable facts. See, e.g., Cuevas v. Citizens Fin. Grp., Inc., 526 F. App'x 19, 21-22 (2d Cir. 2013) (summary order) ("[T]he district court must conduct a 'rigorous analysis' in determining whether [Rule 23's] requirements have been met. In making this determination, the 'district judge is to assess all of the relevant evidence admitted at the class certification stage' and resolve all material disputed facts." (quoting Comcast, 133 S. Ct. at 1432)).
For these reasons, on the present record, the Court cannot, and does not, find that the Rule 23 requirements of commonality, typicality, adequacy, predominance, or superiority have been established "'by a preponderance of the evidence.'" See Amara v. CIGNA Corp., No. 13 Civ. 447, 2014 WL 7272283, at *7 (2d Cir. Dec. 23, 2014) (quoting Malarkey v. Texaco, Inc., 983 F.2d 1204, 1214 (2d Cir. 1993)). For a class to be certified in this case, a far more rigorous, and a far more convincing, submission must be made. The Court, accordingly, denies plaintiffs' motion for class certification. The denial is without prejudice—the Court will give plaintiffs one final opportunity to seek class certification.
Should plaintiffs seek to move again for class certification, a revised motion in support of class certification, accompanied by all supporting documentation, is due February 17, 2015. The Court expects that any such submission will attempt to satisfactorily address the shortcomings and questions identified above.
If plaintiffs file a new certification motion, defendants' opposition papers will be due March 16, 2015. In the most recent round of class-certification briefing, it appears that defendants did not depose either the class representative, Mr. Chung, or plaintiffs' expert, Dr. Feinstein. The Court would benefit from such depositions and from defendants' close attention to the issues raised above. Should defendants pursue these depositions, the Court expects the parties to schedule them for sufficiently soon after the February 17, 2015 due date for plaintiffs' motion so as to enable defendants to meet their March 17, 2015 filing deadline.
Plaintiffs' reply brief will be due March 31, 2015. If defendants' opposition to class certification includes submissions from an expert or other declarant whom plaintiffs wish to depose, the Court will, on motion from plaintiffs, extend the due date of the reply brief by one week to enable plaintiffs to conduct that deposition and incorporate it into their reply.
Plaintiffs move to strike the Declaration of William McGrath (Dkt. 150), which defendants filed as part of their attempt to show that the market for SmartHeat shares was not efficient. See Dkt. 159. Defendants concede that McGrath is not an expert witness and that the Court may strike any portions of his declaration that constitute improper lay opinion. Dkt. 165, at 5; see also Fed. R. Evid. 701 ("If a witness is not testifying as an expert, testimony in the form of an opinion is limited to one that is . . . not based on scientific, technical, or other specialized knowledge."). Accordingly, the Court strikes McGrath's calculations and his subjective assessments of, for example, the nature of SmartHeat's trading volume, whether analysts provided objective coverage, and whether SmartHeat stock reacted rationally to publicly disclosed information. This leaves only: (a) the first sentence of paragraph 1, which identifies McGrath's credentials based on personal knowledge, (b) the sentences in paragraphs 2 and 15 that authenticate the exhibits attached to McGrath's declaration, and (c) the eight exhibits, which are all publicly available documents. If and when a new round of class-certification briefing occurs, the Court would benefit from an expert declaration from defendants on the points addressed in this opinion. Defendants are at liberty, but are not required, to pursue the argument they articulated here through the McGrath Declaration that the market for SmartHeat shares was not efficient.
--------
In the event plaintiffs elect not to seek class certification, they are to notify defendants promptly, and the parties are to make ajoint submission, due February 17, 2015, as to the future course of this litigation.
CONCLUSION
For the reasons set out above, plaintiffs' motion for class certification is denied. The Clerk of Court is respectfully directed to terminate the motions pending at docket numbers 144, 153, and 159. The schedule set out in the body of this opinion is hereby ordered.
SO ORDERED.
/s/_________
Paul A. Engelmayer
United States District Judge
Dated: January 21, 2015
New York, New York