DeMarco v. Lehman Bros.

222 F.R.D. 243 | S.D.N.Y. | 2004

OPINION AND ORDER

RAKOFF, District Judge.

In its recent opinion in Hevesi v. Citigroup, Inc., 366 F.3d 70 (2d Cir.2004), the Court of Appeals, citing the instant action and a similar ease before Judge Lynch, took “note that several courts in the Southern District of New York are currently grappling with the application of the fraud-on-the-market doctrine to analyst reports.” Id. at n. 9. Having so grappled, this Court concludes that the fraud-on-the-market doctrine may in certain conditions apply to analyst reports but that the plaintiffs here have failed to adduce evidence adequate to satisfy such conditions for purposes of class certification.

The allegations of the three instant consolidated actions, as further refined in the course of motion practice, describe a straightforward securities violation. Specifically, it is alleged that during the latter half of the year 2000, defendant Michael Stanek, a prominent research analyst at co-defendant Lehman Brothers, Inc., issued public reports in which he strongly recommended the purchase of the common stock of a computer software company named RealNetworks, Inc., while at the same time he privately recommended to preferred clients that they sell or “short” the stock and confessed to them that he had inflated his public recommendations because Lehman Brothers was also serving as one of RealNetworks’ investment bankers. Although at present these are no more than allegations, if true they describe a clear violation of Section 10(b) of the Securities Exchange Act, 15 U.S.C. § j(b). See DeMarco v. Lehman Bros., 309 F.Supp.2d 631 (S.D.N.Y.2004)(denying motion to dismiss).

Having survived a motion to dismiss, however, plaintiffs now seek to “up the ante” by moving to represent the class of all those who purchased RealNetworks’ stock during the period in which Stanek was making his allegedly false recommendations to the public. An obvious difficulty with this motion is that no one presently knows which of these purchasers materially relied on Stanek’s recommendations in deciding to purchase this stock. In the absence of such information, defendants argue, plaintiffs cannot satisfy such prerequisites to class action status as “numerosity,” “commonality,” and “typicality,” see Rule 23(a), Fed.R.Civ.P., let alone show that the questions of law or fact common to the members of the class predominate over any questions affecting only individual members, see Rule 23(b), Fed.R.Civ.P.

Plaintiffs respond, however, by arguing that Stanek’s allegedly false statements materially impacted the “mix of information” available to the market in pricing RealNet-works’ stock, and that such “fraud-on-the-market” is a presumptive substitute for personal reliance unless otherwise rebutted. This, they argue, is because, as the Supreme Court held in Basic v. Levinson, 485 U.S. 224, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988), “ ‘in an open and developed securities market, the price of a company’s stock is determined by the available material information regarding the company and its business. ... Misleading statements will therefore defraud purchasers of [the] stock even if the purchasers do not directly rely on the misstatements.’ ” Id. at 242, 108 S.Ct. 978, quoting Peil v. Speiser, 806 F.2d 1154, 1160-1161 (3d Cir.1986).

In Basic, the information in question was the issuer’s own statements falsely denying that it was engaged in merger talks with another company. Basic, 485 U.S. at 227, 108 S.Ct. 978. Nonetheless, plaintiffs note, the Court in Basic did not expressly limit its holding to a particular declarant or a particular kind of statement. Indeed, the Court recognized that the manner in which information about a public company gets translated into a market price is through the intervening analyses of market professionals. Id. at 247, 108 S.Ct. 978 (“For purposes of accepting the presumption of reliance in this case, we need only believe that market professionals generally consider most publicly announced material statements about companies, thereby affecting stock market prices.”)

The intervening role of research analysts and other market professionals is, indeed, critical to the pricing mechanism of the securities market, for ordinary investors frequently lack sufficient expertise to interpret the wealth of information, much of it highly technical, emanating from most public companies. Research analysts do not, however, merely digest such information and spew it out in summary form. Rather, as the very name “analyst” suggests, they interpret the data from various viewpoints and offer their more-or-less “expert” opinions as to what the data show about the prospects of the issuer and the value of its stock. Typically, as in this case, they also provide recommendations as to what action the ordinary investor should take with respect to the stock, such as “buy,” “sell,” “hold,” etc.

Although, as discussed infra, such opinions and recommendations may vary widely among different analysts, and while no reasonable investor may suppose that any given analyst can guarantee future results, see In re Merrill Lynch & Co. Research Reports Sec. Litig., 273 F.Supp.2d 351, 358 (S.D.N.Y.2003)(Pollack, J.), a reasonable investor is entitled to assume that the analyst is providing his honest opinion, rather than, as here alleged, lying in order to manipulate the market for the benefit of an issuer for which the analyst’s employer is providing lucrative services. See, e.g., In re Credit Suisse First Boston Corp. Sec. Litig., 1998 WL 734365 at *5, 1998 U.S. Dist. LEXIS 16560 at *14 (S.D.N.Y.1998). Moreover, as in every field where interpretation of technical information is involved, there are research analysts who, by their reputation for accuracy, expertise, and insight, become particularly well-known and respected. The possibility therefore exists that some research analysts may have the ability to influence market prices on the basis of their recommendations. Indeed, in Carpenter v. United States, 484 U.S. 19, 108 S.Ct. 316, 98 L.Ed.2d 275 (1987), the Supreme Court took note that even a newspaper columnist’s views of a given stock could be a material factor affecting the stock’s price where, “[b]ecause of the ... column’s perceived quality and integrity, it had the potential of affecting the price of the stocks which it examined.” Id. at 22,108 S.Ct. 316.

In Carpenter, however, the Court further noted that the district court had found that the column’s impact on the market was “difficult ... to quantify in any particular ease.” Id. at 23, 108 S.Ct. 316, quoting United States v. Winans, 612 F.Supp. 827, 830 (S.D.N.Y.1985). As it happens, this inability to quantify was irrelevant to the decision in Carpenter, both because Carpenter was a criminal case in which proof of reliance was unnecessary¹ and because, since the column was found to reflect the columnist’s honest opinions, id. at 22-23, 108 S.Ct. 316, the ease was analyzed in terms of misappropriation of information from the columnist’s employer rather than in terms of fraud on a purchaser or seller of securities, id. at 22-25, 108 S.Ct. 316. But for present purposes, the issue of measurable impact is critical.

This is because there is a qualitative difference between a statement of fact emanating from an issuer and a statement of opinion emanating from a research analyst. A well-developed efficient market can reasonably be presumed to translate the former into an effect on price, whereas no such presumption attaches to the latter. This, in turn, is because statements of fact emanating from an issuer are relatively fixed, certain, and un-contradicted. Thus, if an issuer says its profits increased 10%, an efficient market, relying on that statement, fixes a price accordingly. If later it is revealed that the previous statement was untrue and that the profits only increased 5%, the market reaction is once again reasonably predictable and ascertainable.

By comparison, a statement of opinion emanating from a research analyst is far more subjective and far less certain, and often appears in tandem with conflicting opinions from other analysts as well as new statements from the issuer. As a result, no automatic impact on the price of a security can be presumed and instead must be proven and measured before the statement can be said to have “defrauded the market” in any material way that is not simply speculative.

In some cases, sophisticated statistical techniques may enable a skilled investigator to determine that a given analyst’s opinion of a given security has indeed materially impacted market price so as to warrant application of the “fraud-on-the-market” doctrine. But in other cases the claim of measurable impact on the marketplace will be too speculative to support such an application.

Sensitive to these considerations, the opinion in Hevesi notes Professor Coffee’s view that “[o]nly in a ease where the publication of the [analyst] report clearly moved the market in a measurable fashion would the ‘fraud on the market’ doctrine seem fairly applicable.” Hevesi, 366 F.3d at 79 n. 7, quoting John C. Coffee, Jr., Security Analyst Litigation, N.Y.L.J. Sept. 20, 2001 at 5 (emphasis supplied). In any event, this Court now holds that the “fraud-on-the-market” doctrine applies in a case premised on a securities analyst’s false and fraudulent opinions or recommendations only where the plaintiff can make a showing that the analyst’s statements materially impacted the market price in a reasonably quantifiable respect.

Whatever might need to be alleged to meet this standard at the pleading stage, the Court further holds that to qualify for class certification in a case where, as here, such certification is dependent on invocation of the fraud-on-the-market doctrine, the plaintiff must adduce admissible evidence that facially meets the aforementioned standard, ie., that makes a prima facie showing that the analyst’s statements alleged to be false or fraudulent materially and measurably impacted the market price of the security to which the statements relate.

In assessing the present plaintiffs’ showing in this regard, the Court is acutely aware that under the view prevailing in this Circuit, the Court may not consider on a class certification motion either the contrary evidence offered by defendants or the merits of the underlying claims. See Caridad v. Metro-North Commuter Railroad, 191 F.3d 283 (2d Cir.1999); but cf Szabo v. Bridgeport Machines, Inc., 249 F.3d 672 (7th Cir.2001). At the same time, however, this does not relieve the district court of conducting a “rigorous analysis [] that the prerequisites of Rule 23(a) have been satisfied.” Id. at 291, quoting General Telephone Co. v. Falcon, 457 U.S. 147, 157 n. 13, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982).

Nor does anything in Caridad suggest that the Court may not require a plaintiff seeking class certification to adduce admissible evidence that, taken most favorably to the plaintiff, establishes a prima facie entitlement to such certification, which, as previously noted, is dependent in this case on the applicability of the fraud-on-the-market doctrine. Indeed, in Hevesi, the Court of Appeals, while stating that “[w]e need not decide what evidentiary showing, if any, the plaintiffs must make at the class certification stage in order to benefit from the Basic presumption in an action against research analysts and their employers,” Hevesi, 366 F.3d at 79, discussed with seeming approval the Seventh Circuit’s rejection of class certification where the plaintiffs failed to make such an evidentiary showing. Id. at 78, citing West v. Prudential Securities, Inc., 282 F.3d 935, 940 (7th Cir.2002).

Here, plaintiffs’ proffered evidence does not remotely satisfy the aforementioned burden because, even when taken most favorably to plaintiffs, it does not warrant a finding that Stanek’s allegedly false statements materially impacted the market price of RealNetworks’ stock in any reasonably quantifiable respect.

To begin with, plaintiffs now disclaim even an attempt to prove that the publication of Stanek’s “buy” recommendations and accompanying statements — which were no different than the similar recommendations of many other analysts at the time — increased the price of RealNetworks’ stock. See transcript of oral argument on this motion, 6/24/04, at 30. Plaintiffs’ contention, rather, is that if Stanek had instead disclosed his true “sell” opinion, his influence was such that, despite the others’ “buy” recommendations, his “sell” recommendation would have driven the price down. Id.² But none of the evidence plaintiffs proffer in support of this contention warrants a finding of measurable impact.

Specifically, they offer evidence of two kinds. The first kind consists of Lehman Brothers’ promotional materials, mostly from a prior time period, touting Stanek’s abilities and purported influence on the market. See, e.g., Declaration of Marc I. Gross, Esq., dated June 1, 2004, Volume II, Exhibits E-K (Lehman Brothers’ “pitchbooks” and similar literature). But such conventional puffing does not constitute meaningful evidence that any of Stanek’s statements had in fact a measurable impact on market prices, let alone that the hypothesized statement by Stanek describing his real view of RealNetworks would have had a measurable impact on the price of its stock.

Second, and primarily, plaintiffs rely on the report of their expert, Frank C. Torehio, and his accompanying materials. See Supplemental Declaration of Marc I. Gross, Esq., dated June 10, 2004, Ex. A. (“Revised Torchio Report” and exhibits thereto). Superficially, Torchio purports to meet the aforementioned standard for invocation of the fraud-on-the-market doctrine, in that his ultimate opinion is that if Stanek had publicly recommended “sell” rather than “buy,” the price of RealNetworks’ stock would have declined by 10 percent. Revised Torchio Report, ¶6. But the method he uses to reach these results has virtually nothing to do with Stanek, Stanek’s alleged influence on the market (which he largely just assumes), or Stanek’s actual statements, and appears irrelevant on its face. See Fed. R. Ev. 401. Moreover, his method is so transparently unreliable as to be inadmissable as a matter of law. See Fed.R.Evid. 702; Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579, 597, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993); Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124,135 (2d Cir.2001).

Torchio first assumes as a general matter that when market analysts downgrade their recommendations to “sell,” the stock in question declines by at least 5%. Torchio Revised Report 1190. However, Torchio bases this conclusion exclusively on a study by Kent Womack (the “Womack Study”) that measured the effect of analysts’ “sell” recommendations on 200 stocks during the period 1989-1991. Id. at ¶¶ 58, 64-69, 90.³ Needless to say, a study of analysts’ impact on that preInternet market has little relevance to Stanek’s hypothesized impact on Internet-related stocks like RealNetworks during the Internet “bubble” that lasted into the year 2000 and then collapsed during the putative class period. See In re Merrill Lynch & Co. Inc. Research Reports Securities Litig., 289 F.Supp.2d 416, 417 n. 6 (taking judicial notice of the Internet bubble and its crash). Furthermore, the Womack Study measures the impact of analysts as a group, as opposed to a single analyst like Stanek. Finally, and perhaps most critically, the Womack Study does not distinguish between the effect on the market price of simultaneous similar information (“confounding news”) emanating from the issuer. Thus, there is no way to tell if the analysts’ opinions had any independent impact at all.

Nevertheless, Torchio takes the 5% measure to be the “minimum” impact that Stanek’s hypothesized “sell” recommendation would have had on RealNetworks’ price. To calculate the “maximum” impact, he then picks three dates during the class period when the price of RealNetworks’ stock, after declining the day before, rebounded after publication of analysts’ opinions continuing to recommend its purchase. The average increase on these three occasions was 15.99%, see Torchio Revised Report ¶¶ 93-96, and Torchio thereby hypothesizes that the impact of a Stanek “sell” recommendation on Real-Networks’ stock would be somewhere between the minimum of 5% and the maximum of 15.99% — say, at least 10%. Id. at ¶ 6.

But the calculation of the “maximum” impact, though different in methodology from the Womack Study, is equally flawed on its face. Even if Torchio had picked his three dates at random, such a small number of dates would not likely have statistical significance. But, in fact, he concededly did not pick the dates at random: yet he provides no objective basis for his picking these dates but only his subjective impression that they were dates when “confounding news” from the issuer was “minimal.” Torchio Revised Report 1192. As in the case of the Womack Study, moreover, Torchio is here looking at the purported effect of analysts as a group, which provides no reasonable basis for measuring the impact of Stanek individually.⁴ Finally, as for Torchio’s choice of a midpoint between his two flawed measures, it appears to be based on no methodology whatsoever.

Without multiplying examples further, the net is that Torchio’s conclusions are both so facially unreliable as to be inadmissable under Fed.R.Evid. 702 and so plainly irrelevant as to be inadmissible under Fed.R.Evid. 401. In no way, thus, do they, or any of the other evidence proffered by plaintiffs, survive the “rigorous analysis,” supra, that is required for class certification.

In short, because plaintiffs have not met their burden of adducing prima facie evidence sufficient to warrant invocation of the fraud-on-the-market presumption, on which they rely to meet the requirements of Fed. R.Civ.P. 23,⁵ plaintiffs’ motion for class certification must be, and hereby is denied.

SO ORDERED.

1 . Similarly, the allegations of the instant lawsuit, if true, would support federal and state administrative actions without necessitating any proof of reliance (in fraud-on-the-market form or otherwise).

2 . As plaintiffs’ counsel effectively conceded at oral argument, see transcript, 6/24/04, at 39, this is not an "omission” case that implicates the doctrine of Affiliated Ute Citizens v. United States, 406 U.S. 128, 92 S.Ct. 1456, 31 L.Ed.2d 741 (1972), for Stanek was under no duty to his readers to give his tine opinion of RealNetworks and could have remained totally silent. But when he voluntarily chose to give his opinion of RealNetworks, he thereupon assumed a duty to give his true opinion. Instead, according to plaintiffs, he intentionally lied. See transcript, 6/24/04, at 39. While plaintiffs also have a "fallback” position that Stanek failed to disclose his alleged conflict-of-interest and the fact that he was giving contrary recommendations to preferred clients, they make no attempt whatever to show that the failure to reveal such information had any demonstrable impact on the market at all, let alone one sufficient to support a fraud-on-the-market presumption.

3 . While Torchio also makes reference to two other studies. An Empirical Analysis of Analysts’ Target Prices: Short-term Informativeness and. Long-Term Dynamics, by Alon Brav and Reuven Lehavy, The Journal of Finance, October 2003, and The Anatomy of the Performance of Buy and Sell Recommendations, by Scott E. Stickel, Financial Analysts Journal, September-October 1995, neither was used by Torchio to arrive at the 5% figure or any other quantitative aspect of his measurement of the impact of the hypothetical Stanek "sell” recommendation.

4 . This is in notable contrast to Torchio’s report in the "WorldCom” case, In re WorldCom Inc. Securities Litigation, 219 F.R.D. 267 (S.D.N.Y. 2003), where he isolated the statements of the particular analyst there involved (Jack Grubman) on days in which there was no confounding information.

5 . While defendants allege that plaintiffs have failed to satisfy Rule 23 in various other respects, the Court need not reach these other objections.