UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK
March 27, 2012
IN RE FEDERAL HOME LOAN MORTGAGE CORP. (FREDDIE MAC) SECURITIES LITIGATION
The opinion of the court was delivered by: Cedarbaum, J.
Jerry Jones*fn1 sues defendants Richard Syron, Freddie Mac's former CEO, and Anthony Piszel, Freddie Mac's former CFO, for violations of Sections 10(b) and 20(a) of the Securities Exchange Act of 1934, 15 U.S.C. §§ 78j(b) & 78t(a), and Rule 10b-5 thereunder, 17 C.F.R. § 240.10b-5. This is a motion to certify a class, to appoint Jones as lead plaintiff and class representative, and to name Cotchett, Pitre & McCarthy, LLP as class counsel. Defendants oppose the motion on several grounds. Primarily, they contend that Jones has not established that the market for Freddie Mac's Series Z preferred shares ("Series Z") was efficient. Thus, they argue that Jones cannot establish the "fraud on the market" presumption of collective reliance. Nor can he satisfy the predominance requirement of Fed. R. Civ. P. 23(b)(3). Defendants also argue that Jones is not an adequate class representative and that Cotchett, Pitre & McCarthy, LLP is not qualified to serve as class counsel.
Because the parties dispute the efficiency of the market for Series Z, I held an evidentiary hearing to determine whether plaintiffs can rely on the fraud on the market theory.*fn2 Both Jones and defendants moved to exclude the opposing party's expert testimony and related evidence on efficiency.
For the reasons that follow, the motion for class certification is denied. Plaintiff has not shown by a preponderance of the credible evidence that the market for Freddie Mac's Series Z preferred shares was efficient during the class period.
On November 29, 2007, Freddie Mac commenced its Series Z offering. Jones purchased 3,265 Series Z shares on February 26, 2008, at a price of $25.99 per share, and purchased 5,500 additional shares on July 11, 2008, at a price of $17.10 per share.
Jones alleges that from November 2007 through September 2008, defendants misrepresented and omitted material information about Freddie Mac's underwriting and risk management practices, as well as the adequacy of its capitalization. On July 10, 2008, the Wall Street Journal reported that Treasury officials had been meeting for months to discuss plans to bail out Freddie Mac, and former St. Louis Federal Reserve President William Poole told Bloomberg that Freddie Mac was "insolvent" and needed a government bailout. Jones alleges that the stock price fell from $21.98 per share to $17.60 per share on July 10, 2008 as a result of these disclosures. On July 11, 2008, Jones purchased 5500 shares. On Sunday, September 7, 2008, Treasury Secretary Henry Paulson and the Federal Housing Finance Agency announced that the federal government had placed Freddie Mac into conservatorship. On Monday, September 8, 2008, the stock fell from $13.56 per share to $2.87.
Jones has moved to certify a class comprised of:
All persons or entities, other than Defendants, who, between November 29, 2007 and September 6, 2008 ("Class Period"), purchased or acquired Freddie Mac's Series Z preferred stock, either on the secondary market or through an original offering pursuant to an offering circular, and suffered damages thereby (the "Class").
Pls.' Class Cert. Mot. at 3.
I. Legal Standard
The elements of a claim for securities fraud under Section 10(b) and Rule 10b-5 are: (1) a material misrepresentation or omission; (2) scienter; (3) a connection with the purchase or sale of a security; (4) reliance; (5) economic loss; and (6) loss causation. Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 341-42, 125 S. Ct. 1627, 1631, 161 L. Ed. 2d 577 (2005).
A. Class Certification
Plaintiff seeks to certify a class pursuant to Fed. R. Civ. P. 23(a) and Fed. R. Civ. P. 23(b)(3). He bears the burden of establishing by a preponderance of the credible evidence that the putative class satisfies each of the requirements of these rules. Teamsters Local 445 Freight Div. Pension Fund v. Bombardier Inc., 546 F.3d 196, 202-03 (2d Cir. 2008). Rule 23(a) provides that one or more members of a class may sue on behalf of all members "only if: (1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class." Rule 23(b)(3) states that certification is appropriate if the court finds that "the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy."
In In re Initial Public Offering Securities Litigation (In re IPO), the Second Circuit clarified that a district court must make a "definitive assessment of Rule 23 requirements, notwithstanding their overlap with merits issues," and that it must resolve the material factual disputes relevant to each Rule 23 requirement. 471 F.3d 24, 41 (2d Cir. 2006); see also Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2551, 180 L. Ed. 2d 374 (2011) ("A party seeking class certification must affirmatively demonstrate his compliance with the Rule -- that is, he must be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact, etc."). Both Teamsters and In re IPO contemplated the possibility of a district court's holding of an evidentiary hearing, as the extent of discovery granted on class certification issues is "generally left to the trial court's considerable discretion." Teamsters, 546 F.3d at 204 (quoting Heerwagen v. Clear Channel Commc'ns., 435 F.3d 219, 233 (2d Cir. 2006)); see also In re IPO, 471 F.3d at 41. A trial court must receive enough evidence to be satisfied that each Rule 23 requirement has been met. Teamsters, 546 F.3d at 204. Although defendants contest several of plaintiff's arguments in favor of class certification, the focus of the parties' briefs and expert testimony is on Rule 23(b)(3)'s predominance requirement and the efficiency of the market for Series Z.
B. Rule 23(b)(3): Predominance
Class certification is available only if plaintiff can establish a class-wide presumption of reliance through the "fraud on the market" theory.*fn3 See Basic Inc. v. Levinson, 485 U.S. 224, 241-42, 108 S. Ct. 978, 989, 99 L. Ed. 2d 194 (1988) (recognizing that private securities fraud plaintiffs may be entitled to a presumption of collective reliance). Without this presumption, individual questions about investor reliance on misrepresentations would predominate over common questions, and a class action would not be a superior mechanism for resolving the dispute. See Fed. R. Civ. P. 23(b)(3).
In Basic, the Supreme Court explained that "[t]he fraud on the market theory is based on the hypothesis that, 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." 485 U.S. at 241-42 (internal citation omitted). Because in an efficient market, the stock price necessarily reflects any material misrepresentation, and because investors who purchase stock do so in reliance on the integrity of the market, misleading statements will defraud purchasers even if they do not directly rely on the misstatement. Id. at 241-42. In order for the fraud on the market theory to apply, plaintiff must show that defendants have Ed. 2d 741 (1972). For claims involving primarily a failure to disclose, "positive proof of reliance is not a prerequisite to recovery." Id. The claims here are not primarily omission claims, and the presumption does not apply. See, e.g., Starr v. Georgeson S'holder, Inc., 412 F.3d 103, 109 n.5 (2d Cir. 2005) (holding that Affiliated Ute's presumption did not apply because plaintiff focused most heavily on allegedly false and misleading statements).
(1) publicly made (2) a material misrepresentation (3) about stock traded on an efficient market, and (4) that the plaintiff purchased the shares between the time the misrepresentations were made and the time the truth was revealed. In re Salomon Analyst Metromedia Litig., 544 F.3d 474, 481 & n.4 (2d Cir. 2008) (citing Basic, 485 U.S. at 248 n.27). Defendants assert that the market for Series Z shares was not efficient throughout the proposed class period.*fn4
Economists identify three general forms of the efficient capital markets hypothesis:
First is the weak form, which asserts simply that the current share price in an efficient market reflects all information about past share prices. If the weak form of the hypothesis accurately describes a market, it is impossible to predict future prices using only past prices. Second, the semi-strong form, which asserts that a share price in an efficient market reflects all public information concerning the security (including but not limited to past share prices). Third, the strong form, which asserts that all relevant information, public and private, is reflected in the price of securities in an efficient market. The strong form has been widely discredited.
In re Initial Pub. Offering Sec. Litig., 260 F.R.D. 81, 98 n.148 (S.D.N.Y. 2009). The fraud on the market theory is based on the semi-strong form of market efficiency. See, e.g., ATSI Commc'ns Inc. v. Shaar Fund, Ltd., 493 F.3d 87, 101 n.4 (2d Cir. 2007).
C. Factors Relevant to the Determination of Market Efficiency
The Second Circuit has not explicitly adopted a test for the market efficiency of securities, but it has noted that the factors set forth in Cammer v. Bloom, 711 F. Supp. 1264, 1279-87 (D.N.J. 1989) are routinely applied by district courts. Teamsters, 546 F.3d at 204 n.11. The Cammer factors include:
(1) the average weekly trading volume of the stock; (2) the number of securities analysts following and reporting on the stock; (3) the extent to which market makers and arbitrageurs trade in the stock; (4) the issuer's eligibility to file an SEC registration Form S-3; and (5) the demonstration of a cause-and-effect relationship between unexpected, material disclosures and changes in stock prices. Teamsters, 546 F.3d at 200. In addition, courts sometimes apply three additional Krogman factors: (1) the company's market capitalization; (2) size of the bid-ask spread; and (3) the percentage of shares available to the public (i.e., the public float). See, e.g., In re Initial Pub. Offering, 260 F.R.D. at 94 (citing Krogman v. Sterritt, 202 F.R.D. 467, 478 (N.D. Tex. 2001)).
Courts have also considered a stock's listing on a national exchange to be probative of efficiency, but this factor is not dispositive when efficiency is disputed. See, e.g., In re Initial Pub. Offering Sec. Litig., 227 F.R.D. 65, 107 n.324 (S.D.N.Y. 2004) (explaining that listing on a national exchange is a "good indicator" of efficiency); Bell v. Ascendant Solutions, Inc., 422 F.3d 307, 314 (5th Cir. 2005) (holding that plaintiffs did not demonstrate that a stock listed on NASDAQ was efficient); Cammer, 711 F. Supp. at 1281 ("It would be illogical to apply a presumption of reliance merely because a security is traded within a certain 'whole market' without considering the trading characteristics of the individual stock itself.").
Under Teamsters, the most important Cammer factor is the "cause and effect" factor -- evidence that unexpected corporate events or releases caused an immediate response in the price of a security. This is "the essence of an efficient market and the foundation for the fraud on the market theory." Teamsters, 546
F.3d at 207 (quoting Cammer at 1287). The court stated: Without the demonstration of such a causal relationship, it is difficult to presume that the market will integrate the release of material information about a security into its price. An event study that correlates the disclosures of unanticipated, material information about a security with corresponding fluctuations in price has been considered prima facie evidence of the existence of such a causal relationship.
Id. at 207-08.
The other Cammer factors do not directly address whether the market price reflects public information, which is the issue at the core of the semi-strong form of the efficient market hypothesis.
II. Market Efficiency of Series Z
A. Background on Event Studies
An event study examines the extent to which stock prices react to the release of new, material information (an "event"). For the purpose of showing market efficiency, days on which unexpected, important news entered the market are identified. The researcher also defines the parameters for the "event window," i.e., the period in which the news may affect the price.
The actual price of the security during the event is compared against the expected price, which is calculated based on the security's historical relationship to a market index. This historical relationship is measured over a "control period." The difference between the stock's actual price and the expected price is defined as an "abnormal return." A. Craig MacKinlay, Event Studies in Economics and Finance, 35 J. Econ. Lit. 13, 14-16 (1997). In an efficient market, stock prices should show statistically significant abnormal returns on days in which unexpected, material information is released into the market.
B. Expert Testimony
Each party presented an expert at the hearing, and I relied on the experts' testimony about the content of their reports. Plaintiff's expert witness, Dr. Craig McCann, received a Ph.D in economics from UCLA and is employed by Securities Litigation and Consulting Group, Inc. He has conducted market event studies in connection with consulting work for the SEC, and he has taught economics and finance at the University of South Carolina, Virginia Tech, the University of Maryland, and Georgetown University.
Defendants' expert witness, Dr. Mukesh Bajaj, has a Ph.D in finance from the University of California at Berkeley, as well as an MBA with an emphasis on finance from the University of Texas. He founded his own consulting firm, AFE Consulting, where he currently consults for corporations, regulatory agencies, and law firms. He also teaches advanced corporate finance and financial engineering at the Haas School of Business at the University of California at Berkeley.
C. Dr. McCann's Event Studies
1. McCann's First Study
McCann supplied two different event studies in two different reports during the course of discovery. In the first event study, he identified 28 news dates, 17 of which had statistically significant abnormal returns. He used a rolling control period of 126 days, ending shortly before the date being evaluated. The event window included both the day of the relevant news and the subsequent day. He identified 28 news days between June 6, 2008 and September 8, 2008 -- roughly the last third of the class period.
McCann's first study contained several significant errors. In an event study regression, the stock and index returns should be calculated in either logarithmic terms or arithmetic terms. As he conceded at the hearing, McCann initially calculated Series Z's returns in arithmetic terms but calculated the market index return in logarithmic terms.
There were other serious errors in McCann's first study. McCann failed to recognize the relationship between the direction of the price movement and the relevant news. This caused McCann improperly to classify three of his chosen dates as abnormal return dates. On two of those dates, June 27, 2008 and August 22, 2008, although the news was negative, the price of the stock went up more than could be attributed to general market conditions. On the third date, July 16, 2008, he identified both good and bad news, yet he did not explain the relationship between news and price. In addition, McCann listed July 9, 2008, as one of the 17 days with statistically significant abnormal returns, despite the fact that neither July 9 nor the day prior was one of his chosen 28 event dates. McCann also erroneously listed July 17, 2008, as a news date, and improperly included September 8, 2008 as a news date, even though it is not within the class period.
Defendants' expert witness, Bajaj, offered a convincing critique of McCann's choice of a 126-day rolling control period. Bajaj demonstrated that McCann's control period encompassed a time of relative placidity in the relationship between Series Z and the S&P Preferred Stock Index. McCann used this period to calculate his abnormal returns in the "stormy ocean" of the last third of the class period -- a time in which the relationship between Series Z and the index was much more volatile. According to Bajaj, because McCann's control period was "placid," he found an excessive number of abnormal returns in the last third of the class period.
At the hearing, McCann testified that after correcting for his logarithmic/arithmetic error by running both returns in arithmetic terms, and limiting his event window to only one day, he found statistically significant abnormal returns on only 13 of 28 news days, as opposed to the 17 he had originally reported. In his testimony about the 13 days, McCann did not exclude the three days in which he could not show that the price was affected by the news.
2. McCann's Second Study
In the second event study McCann conducted, he chose news days from a different period, December 3, 2007 through June 5, 2008 -- the first two-thirds of the class period that had been part of his control period in the first study. Unlike his first study, in which he used a 126-day rolling control period, in his second study, he used the six months under review as a control. In his first study, McCann reviewed the response to news on the day the news was released as well as the subsequent day. In his second study, he limited the event window to a single day. For news released after the close of the market, he studied only one day following.
In the second study, McCann changed his methodology. He testified that the proper way to conduct an event study is to divide the class period into days in which potentially material news entered the market and days in which no news entered the market. Then, according to McCann, the relevant test is whether there were substantially more statistically significant abnormal return days during the subset of days in which potentially material news was released. To execute this comparison, he performed Fisher's Exact and Chi-Square tests on the data. These tests calculate the probability of observing abnormal returns across two sets of dates if the market price of Series Z did not respond to material news. Using this method, McCann found that of 128 trading days in his second study, there were 31 potential news days, and he found a statistically significant abnormal return on 4 of those days. In his 97 non-news days, McCann found only 1 statistically significant abnormal return.
He concluded from this that there was a less than 1% chance that the news had no effect on the price of Series Z.
At the evidentiary hearing, McCann admitted that March 17, 2008 -- one of the four news days with an abnormal return in his second event study -- should have been classified as a non-news day because the relevant news emerged after the close of trading. With this shift in the ratios of abnormal returns on news days and non-news days, McCann's conclusions based on the Fisher's Exact test were reversed for the period from December 3, 2007 through June 5, 2008. In other words, the market for Series Z failed McCann's test for efficiency for the first two-thirds of the class period, because the percentage of days with abnormal returns on news and non-news days was not different at a statistically significant level.
McCann combined the news and non-news days from both his first and second event studies, and concluded that there was a less than 1% chance that he would observe significant abnormal returns on 16 of 57 news days and 9 of 136 non-news days, if news had no effect on the price of Series Z. Given the two different methodologies McCann used to define his control period, it is not appropriate simply to add the results of his two event studies. Even if it were appropriate, combining the two studies results in statistically significant abnormal returns on only 16 of 57 news days -- or 28%. These are the numbers that McCann testified to after conceding that March 17, 2008 and September 8, 2008 should be removed. He did not correct the numbers for the days improperly included because price movements were in the wrong direction given the news, nor for the day with both positive and negative news. Furthermore, McCann's numbers conflict with plaintiff's exhibit 267, which lists McCann's news days and abnormal returns. If September 8, 2008 is excluded, the exhibit shows statistically significant abnormal returns on 15 of 56 news days.
Even if McCann demonstrated that news probably had some effect on price, plaintiff's burden is higher. A plaintiff must show that the market price responds to most new, material news. At the hearing, McCann cited to one law review article to support his approach. See Paul Ferrillo, Frederick Dunbar & David Tabak, The "Less Than" Efficient Capital Markets Hypothesis: Requiring More Proof From Plaintiffs in Fraud-onthe-Market Cases, 78 St. John's L. Rev. 81, 120-21 (2004). After outlining this comparative methodology, the article states that "[w]hile this test addresses the question of whether the stock responds to news, it does not answer the question about whether the response is of the correct magnitude. Therefore, this test is a threshold step, not a sufficient condition, to show that a stock traded in an efficient market." Id. at 122.
The Fisher's Exact and Chi-Square tests do not show that the price of Series Z consistently responded to unexpected, material news. Furthermore, I question McCann's use of these tests for the first time in a second and inconsistent event study. At his first deposition, McCann testified that if new information had been released into the market on each of his initial 28 event days, one could expect to see statistically significant abnormal returns on each of the 28 days in an efficient market. In contrast, Bajaj testified that an economist may conclude that a market is efficient if it reacts to news 80 to 90% of the time, depending on the number of news dates at issue. Even accounting for market "noise" -- extraneous circumstances and confounding effects -- McCann's showing that the market reacted to news 28% of the time is insufficient to satisfy the most important Cammer factor,*fn5 the cause-and-effect relationship between unexpected news and price.
D. Bajaj's Study
Bajaj tested for abnormal returns on Series Z throughout the class period and evaluated McCann's results. For several reasons, Bajaj found far fewer statistically significant abnormal returns on McCann's selected news dates. Bajaj's regression included a volatility adjustment; he focused on logarithmic returns, as opposed to arithmetic returns; and he reclassified several of McCann's news dates and abnormal returns because of the staleness of the news and the direction of the return relative to price. Because Bajaj did not independently select news dates, he did not conduct a traditional event study for the purposes of this litigation. Defendants are not required to demonstrate that the market is inefficient, and it was not necessary for Bajaj to conduct his own event study.
E. Daubert Motions
Both plaintiff and defendants move to exclude the expert testimony and related evidence presented by the other party. Expert opinion testimony is admissible if:
(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert has reliably applied the principles and methods to the facts of the case.
Fed. R. Evid. 702. The expert's testimony must be relevant to the task at hand and rest on a reliable foundation. Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 584-87, 113 S. Ct. 2786, 2792-94, 125 L. Ed. 2d 469 (1993). A court should consider whether the technique has been tested, whether it has been subjected to peer review and publication, the known or potential rate of error, and whether it is generally accepted within the relevant scientific community. See id. at 594. In addition, a court should consider whether the expert fails to employ the level of intellectual rigor that characterizes the practice of an expert in the relevant field. Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152, 119 S. Ct. 1167, 1176, 143 L. Ed. 2d 238 (1999). Even if a methodology is sound, a court may conclude that there is simply too great an analytical gap between the data and the opinion offered. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S. Ct. 512, 519, 139 L. Ed. 2d 508 (1997).
After listening to the testimony of both experts and examining their study methods, I found Bajaj's testimony credible, reliable, and persuasive. In contrast, McCann's analysis changed so many times in important ways and was so internally inconsistent that I found it unreliable and unpersuasive. After listening carefully to all of his testimony about his two inconsistent event studies, I could not credit his testimony, and I have serious reservations about the admissibility of such poorly supported opinions.
F. Cammer and Krogman Factors
The Cammer factors are intended to be an analytical tool, not a checklist. See Teamsters, 546 F.3d at 210 (quoting Unger v. Amedisys, Inc., 401 F.3d 316, 325 (5th Cir. 2005)). In this case, the less important Cammer factors support a finding of efficiency. The average weekly trading turnover for Series Z was 4.6%, which weighs in favor of a finding of efficiency. See Cammer, 711 F. Supp. at 1286. The existence of a New York Stock Exchange specialist for Series Z also supports an inference of efficiency. Because Freddie Mac was a government-sponsored enterprise, it was not required to register with the SEC. Regardless, I find the SEC Form S-3 factor redundant in light of the market capitalization factor. See Cammer, 711 F. Supp. at 1287 (discussing the Form S-3 factor and noting that it is the number of shares traded and value of the shares outstanding that involve the facts which imply efficiency). Even though securities analysts covered Freddie Mac's common stock, the analysts did not cover the Series Z preferred stock. While there were rating agency reports on Series Z, these are less thorough than analyst reports and should be given less weight. Cf. Teamsters, 546 F.3d at 206 n.12 (noting that district courts may conclude that rating agencies covering bonds less directly affect price than equity analysts who provide a conduit from the company to investors). This factor provides only some support for a finding of efficiency. In addition, the three Krogman factors -- bid-ask spread, market capitalization, and public float -- support an inference of efficiency to varying degrees. See Krogman, 202 F.R.D. at 478.
Although the less important Cammer and Krogman factors support an inference of efficiency, these factors cannot substitute for evidence of a cause-and-effect relationship between unexpected news and market price. This is the critical factor -- the sine qua non of efficiency. It speaks to the "essence" of the efficient market hypothesis, and it is the foundation of the fraud on the market theory. Teamsters, 546 F.3d at 207. The other Cammer and Krogman factors do not directly address the question of efficiency. Without evidence of the prompt effect of unexpected news on market price, the market cannot be called efficient.
McCann's testimony about the efficiency of the market was unreliable and unpersuasive. His first event study was deeply flawed. He changed the relevant news dates from the first study to the second, and then changed them again in his testimony. Without excluding dates in which the abnormal return was in the wrong direction given the news, McCann found that 16 of 57 news days had a statistically significant abnormal return. McCann's showing that Series Z responded to material news 28% of the time is insufficient to satisfy plaintiff's burden of proving Cammer's cause-and-effect factor. Furthermore, after McCann corrected one of his erroneous abnormal return dates, his Fisher's Exact test for December 3, 2007 through June 5, 2008 did not support his opinion that the price of Series Z responded to unexpected news.
Because Jones has not shown by a preponderance of the credible evidence that the market for Series Z was efficient, the fraud on the market presumption of collective reliance does not apply. Accordingly, on the element of reliance, individual issues predominate over collective ones. I do not reach whether the other requirements of Rule 23 have been satisfied. For the foregoing reasons, Jones's motion for class certification is denied. The Clerk is directed to close all motions in 09 Civ. 832.
MIRIAM GOLDMAN CEDARBAUM United States District Judge